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": 6.646942576773088,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.0796,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.932664226837745,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0951,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.931724198598444,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0773,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.650092780835505,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0743,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.548548316884645,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.0853,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.052184756640213,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.017,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 4.890549883875497,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.9775,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.685017099814745,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.984,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 2.5367014377197523,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9162,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.5134222224400338,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9613,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.196620710098812,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9687,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.324800675092874,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9306,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.475295438021865,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9265,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.489345299593995,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.923,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.179513752373875,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9163,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.1114611877511313,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8699,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.112082805068068,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8706,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.7902861013555222,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.825,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.3807778709117264,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8134,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.9446500084844462,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.8135,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.6663111114580007,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.8134,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.8223864859864802,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.8172,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 2.0023263607774497,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.8026,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.870219291860517,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.8086,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.38312244928424,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7975,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1890073360764588,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7603,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.3366754558912861,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7782,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.3167868105637417,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.763,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.9740740364892309,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7427,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8497562918378271,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7502,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.0124874185486055,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7381,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.9457674827029477,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7253,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7371366254046878,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.762,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.8574002738973333,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7454,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7412097755715424,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7181,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6970368981807881,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.7127,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.6885953369022149,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.7281,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7733317704177356,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.7525,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.702734027055239,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.7215,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6233629897306007,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.7343,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7331844113770369,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.704,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.722393930807107,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6807,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.778342201781904,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6858,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6353061375351429,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.7155,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.6804355758278895,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.7064,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6295214005570192,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.7099,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6689269461513594,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.6819,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6817926717302693,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.7274,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6296947844150348,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.7057,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7322447506934461,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6947,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7368030825745415,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7082,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.6572351358692388,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.7103,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.7826066854576241,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6971,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7039688966433039,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6761,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6865315431534426,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6711,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7254647196542139,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6856,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.763146367159171,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6935,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.789375947369188,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.699,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.7753881414928597,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.6821,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.6528640126499125,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6948,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7336728077600647,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6838,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.6452359031892335,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.7075,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8313614269686481,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6969,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7904475693812617,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6777,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6821285091185317,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6557,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.6982250727022783,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.6727,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.7408211950436421,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6732,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.7140742586321746,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6589,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 186.15847923345132,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.7323,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.030819650176432,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6735,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.7919347159559709,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6621,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8969994388790768,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6519,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.0148572909089104,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6335,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.1823434485443218,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6795,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.8154791699494301,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6685,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.36827087625624,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6747,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8048920119070015,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6672,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8671881792323733,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6652,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8151578192967545,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.6463,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.9696076884585224,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.6449,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.9598689612985014,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6464,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.803350790120748,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6493,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.099433367766758,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.65,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.6921499413711566,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6731,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.1504882818926787,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6669,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8436032388133392,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.665,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 23.109509240401202,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6577,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 2.226328994873025,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6763,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.9616997800745878,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6702,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.132486189901255,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6646,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.0102632538948733,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6867,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.2521861556631295,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6466,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.9644085294498898,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6662,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.0887317930448175,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6601,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.9887182498187101,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6402,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0817060984346638,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6587,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.9422108211930462,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6372,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.0457712145998463,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6628,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.9788943641171699,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6407,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.0228115858928075,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6327,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.0054861101791226,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6457,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.8338746105788781,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6479,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.0642793530919699,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6503,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 0.8283355542877116,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6583,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.7822715048212066,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6632,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.8189049279041792,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6708,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.7234706603469003,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.648,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8832175710807038,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6176,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.7247346107007175,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6405,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.3962816736265418,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6469,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.7731910522734201,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6362,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9187973779561689,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6457,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.907498788620162,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6718,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.7174251390016275,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6669,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8287841232431589,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6352,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.7838626741569021,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6538,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.7144589411576534,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6444,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.7203734832957325,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6475,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.9390403467744568,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6297,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.7547749214565589,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.6686,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.8015283976287434,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6511,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.8858455689681912,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6371,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.6817346134417919,
+      "learning_rate": 4e-05,
+      "loss": 0.6618,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.8726059202044234,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6116,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.7936875232744518,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.64,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.6950287226909196,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6473,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.8674908124265281,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6549,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.7201653976518536,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.673,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.7827143278868864,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.6394,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.7904900330925325,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6498,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.7153784385479285,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.673,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.6684272830838692,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6662,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.8445990095867368,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6686,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.6892124114604102,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6576,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.7124783316717701,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6334,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.6401644957067286,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6593,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.668680421790019,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.647,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.7089695438793532,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6596,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.6552674803319217,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.65,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.709187957969286,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6641,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.67968692551828,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.643,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.6965199643688316,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6707,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.7037723017018684,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6358,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.7600240794639587,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6304,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7002480728431044,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6173,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.6442112441910778,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6189,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.5944789421341041,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.6381,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7878159661172829,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6259,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.700441303179936,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6338,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.7064251133170409,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6252,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.7893202668390761,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.633,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.9193388752459063,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6303,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7883502575298452,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.6123,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.9762725471209608,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6038,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.8227998008796873,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.6357,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.815969150967755,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6399,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.7075964347629348,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6504,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6511151762582738,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.621,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.684641416488046,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6372,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.7457417790194425,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.6678,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.8458954092492169,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.6365,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.7944099827747089,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.5904,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.7994303124646509,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6159,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.8407769346496955,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6105,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.7157389145137325,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6517,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.8734435551668811,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.5839,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.577645035172301,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6342,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.8557733267758636,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6221,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.5386147737469349,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6349,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.649565484288151,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.614,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.5965186330342457,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6185,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6267345097344975,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6416,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.6394678479749457,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6333,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.5103825096503766,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6059,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.6125172626338301,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.628,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6265839740952408,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6466,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.6198731141505774,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.661,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6452238404952214,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6252,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.5514052158496064,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6053,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.6203332599430572,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6165,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.5512955567760478,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.5916,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.5431857839949047,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.6167,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.6136559954838027,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6181,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.5906228967835399,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6097,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.5671716316319763,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6432,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.5167209799508392,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.6437,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.5615898859932537,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.6001,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.5138801615022194,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.6257,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.7026050296551878,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6366,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.6145229017423588,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.603,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6347802970804958,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6074,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7386794276202723,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6117,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6813801968745098,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6044,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6075117849503596,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.6468,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5775020229713701,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.6139,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5631958826337875,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6284,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6100886474063608,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.5947,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.5177718260370308,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.5842,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6631502662698153,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.5996,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.547898940965757,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6026,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5620175393964304,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.5755,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5995747500033483,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.5883,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.6019478984166559,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6466,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.6333927603109604,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6332,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.596402392529501,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6224,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6779968693767165,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6097,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5346309069563712,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6177,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.5741747080573707,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6329,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.631772377002875,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6052,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.6156845952237328,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6146,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5536374247113552,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.5962,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.6287123112746678,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6082,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.4590415118169856,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5801,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.5938713315515485,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5919,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.550778720586392,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6187,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.5396699373335692,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.569,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6560096615558779,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.631,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.5181656721386954,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.5839,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.5297668885333561,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.604,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.556920300262551,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.619,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.5596817270525541,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.6315,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.5406570397881639,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.5933,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.5293362984524463,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6188,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6287975808122311,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.5984,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6063889950888683,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6025,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.5556575956773563,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.5979,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.6553376918901264,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6213,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.4943836563053644,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.613,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6966494070804184,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.5892,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.5632196131125439,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.5707,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.603625532103983,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6053,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.5216369531061912,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5979,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.5233014314212568,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6126,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.48857735713998623,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.6123,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.542343854187665,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6337,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.4941373465870162,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.5883,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5427945449355172,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.5979,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.6344329424164002,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.6038,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6210965906390857,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.5847,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.6865014922635322,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.6166,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.562663578673065,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6003,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6594921150980523,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.6029,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5858586902366858,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5933,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6325798182553738,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6173,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5811497233668735,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5704,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.612245502782062,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.5879,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.5805930955033376,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.5924,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.6961547825520877,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4954,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.5230581757775616,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.4723,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 0.7215489170645845,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5057,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.6315245517748482,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5036,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.6615896794705022,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5119,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.6384667295756155,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4711,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.5829345867971377,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5181,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 0.6469198347464266,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5247,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.627959773576762,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.4843,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.6299034419711094,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.4791,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.5797917819334699,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.4797,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.7397620939156244,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5148,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.5286621408161396,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.4867,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 0.6926912314603775,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.4832,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.5518883109487901,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.4732,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.7450499828759439,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.49,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.5718893203151443,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.4725,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 0.6471288392395668,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.4689,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.6298479281587019,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.4671,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 0.6623638460657126,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.478,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.6504580896176616,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.4939,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.6333947453683182,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.4728,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.7029277309081066,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.4661,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.5954404249702794,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.4736,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6138328285802581,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.4849,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.5537364837727495,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5018,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.6414224393007362,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.4881,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.5136188299735318,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.476,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.635380624962008,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.4781,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.5619801206160031,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.4901,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.5566860151330484,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5101,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.550916040764026,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.4771,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6238104190402265,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.4761,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.557175127974939,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.4959,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.6509080415387997,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.4789,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.6209132381678705,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.4876,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.5530539162235477,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.4761,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.5912420414220682,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.4795,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.5151154631914837,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.4933,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.5891535096725709,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4836,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.6094564255326465,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5039,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.6194534710542892,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5051,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.533070976818571,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.4743,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.5860423624325038,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.4637,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.5317992467779296,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.4796,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.4729196136972659,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.4883,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.5389939845979693,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.4798,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.4836333631641106,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.4612,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.5995657685183258,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.4729,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.5766691595644032,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.4978,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.603847539587792,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.4813,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.6707886696095446,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.4718,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.5116719020027788,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.4893,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.6200997556298926,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.4927,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.5420846566452099,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.4739,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7100375629191196,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.5056,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.543333354594474,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.5027,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.6769809866349208,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4783,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.5044751887999919,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.4956,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.6560322968121699,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5019,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.45198323844356225,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.4849,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6097511318695122,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.4891,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.4615086550900446,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.4884,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.5378162706875855,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.475,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.5121966077771887,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.489,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.5549143761014504,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.4754,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.5681029439588637,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4741,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.49689307956649836,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.4887,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6142217789494789,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4854,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.5351691706168669,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4739,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.5890286168522497,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.465,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.5523627486684327,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.4744,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.5143944633126096,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.4776,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.5121414103052002,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.483,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.5518395355986643,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4837,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.4414839169900946,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.4782,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.5514944702455197,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4758,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.49696440434724304,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.4927,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.48455631704571056,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.4611,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.45892723254867485,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.4748,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.47582171940899987,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.4767,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.5214390584546771,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.471,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.47727703877079763,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.479,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.5852890607640341,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.4978,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.5360613958240543,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.494,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.4951977493865475,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5065,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.4588946842198721,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.475,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.4548956657542452,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.4536,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.45673916164017303,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.482,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.498181379738716,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5023,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.45945655663472396,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.4902,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.5041148708755065,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.4669,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.5504928834994939,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.4788,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.5102242001216803,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4729,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.5512793486278161,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.4946,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.5044492265498448,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.4927,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.4681360591486859,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.4831,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.5200671054696839,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5039,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.5689140680959092,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.4899,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.4653471261524222,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.4962,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.6304887961579501,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5037,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.5043563342063659,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.481,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.5052738576786306,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4816,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.4836122141889605,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.4681,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.5039249554456271,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.4724,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.5684146658747748,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.4941,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.47353117164977115,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.4906,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.6483490261475336,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.4922,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.5017419349998725,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.4948,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.5554671329320032,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.4781,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.5247055938966232,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4716,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.5080402355295142,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.48,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.48692336975060524,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.4889,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.49485436764984747,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.4812,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.4919671252873039,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.4794,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.47197653327067596,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4763,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.46981701720609165,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.4977,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.46402584086534365,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.4828,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.45430023531477903,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4792,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.4908971397761353,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.4766,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.504070919574521,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.4729,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.4880025985715224,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.4944,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.4878308334561446,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.4698,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.5183806315253033,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5048,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.47193727039080696,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.485,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.501347541766422,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.4973,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.46654334411375503,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.4801,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.47009790718641775,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.4819,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.5722903453610664,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.4696,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.54655916188772,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.509,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.5941064983869054,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.4757,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.5069011117469749,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5041,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.5353354628519301,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.4727,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.4932211942812617,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.4918,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.5000602278393358,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.4899,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5223563579981931,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5009,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.4807777762959649,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.4801,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.4563705137604807,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.4718,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.5217252701634065,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.4872,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.440854239430146,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4591,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.5975520338545791,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.4758,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.47923299722811086,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4898,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.5136659592051201,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.4966,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.4714438973876213,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.4684,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.45967017875727867,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.4896,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.4783574254101278,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4967,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.473720804362813,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.4635,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4872886448950557,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.4811,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.4877098887729168,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.4738,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.5001471711135093,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.4902,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.5505788289762429,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.4989,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.47589376528314675,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.4669,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5456411543881455,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.4612,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.46617949211620396,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.4993,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.535637780240596,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.4908,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.51885746940118,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.4905,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5506890870274848,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.4817,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.4964055076140856,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5063,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.577107471585149,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.4953,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.5050542382546536,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.4773,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.45692926267875966,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.4843,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.5156771743036387,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.4642,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.46257627905358434,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.499,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.4898026001398337,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.4868,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.4637106964509717,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5044,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.5044000765571798,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.4825,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.47586178488021763,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.4794,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.5099283959033369,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.4831,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5001404840907577,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.4746,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.4721653790981787,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4832,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5259027608068769,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.4855,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.49992946470167343,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.4878,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.4751946929674115,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.4809,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.44956799882462534,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.4966,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.4369570251845496,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.4708,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.45352028661788946,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.4913,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.48072757798802773,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.4759,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.44814498306832895,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.4781,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.44074580475791664,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4848,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.4597199069810897,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4836,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.46934683997767734,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5003,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.4465880801088606,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.4896,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.4179577350877362,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.4701,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.4870597533084163,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.4735,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.4910978043648207,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5072,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.4724261116663464,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.4755,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.48934293707194804,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.4846,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.44398514842256576,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4876,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.47721460506972774,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.4996,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.4305167949829257,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.4688,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.48866730380689616,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.4766,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.4299455916781191,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.4905,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.4209722718895507,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.4725,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.4147435203892593,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5042,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5064560103183122,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.4947,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.4344546189991301,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.465,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5584466298995078,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.4845,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.43551344779938866,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5035,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.5869684956961416,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.4902,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.44332170220503636,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.5097,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.5512394176922947,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.4956,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.42812345599201257,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.4766,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.4931398034173543,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.485,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.4409521961791151,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.4658,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.5044471991118915,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.4969,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.4401232704274646,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.4819,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.46767916404920884,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.4823,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.4974974080182284,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.4969,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.438783788301862,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.4851,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.4871803707020326,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.4639,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.4672401245324371,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.4961,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.5278136096027319,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.4706,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.46407166546968165,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.4729,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.4428189918333573,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.4894,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.42870210376435125,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.4656,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.4378938379707292,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.4679,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.5739043849426072,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.4574,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.44014139620627685,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.4786,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.5517444256458195,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.4678,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5100365291094308,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.4859,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.532771497200485,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.4601,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.45794939692326453,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.4673,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.5008265157700492,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.4716,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.44650541848537845,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5072,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.4449545607917448,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4576,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.44013409766297057,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.4936,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.45708695290661194,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.4605,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.49584905217617764,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.4792,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.4459338956254753,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4761,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.474927675440464,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.4686,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.4897080101491172,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.4743,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.46557766545613793,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.4651,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.47850383546492214,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.4859,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.4473631707083552,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.4748,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.475008896780214,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.4737,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.4819232536814007,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.4753,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.4649026267465635,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4777,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.41084342937940244,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.4551,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.4723939491464516,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.4848,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.457048403372962,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.4458,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.41281725188888085,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.4779,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.47265992095887327,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4783,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.42537407477216177,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.4691,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.4421926408022787,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.4793,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.46533828224518403,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4873,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.3854251058252337,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.4689,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.44871302351246795,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.4463,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.6089361610895275,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3561,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.4841947666355822,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3366,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.8471945392952447,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.33,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.48920185869808264,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3361,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.5255962885301699,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3363,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.49009941015479913,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3261,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.5223444570433811,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3244,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.4275015360757823,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3268,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.5078564903241047,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.335,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.5327912685754411,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3197,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.4400986102622844,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3358,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.48360266384955575,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.339,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.46955441524303654,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3354,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.4230361639500384,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3289,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.46014274463197935,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3185,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.45607689664483103,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3519,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.4502709045862769,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3187,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.45256612334480256,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3307,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.461317809622396,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3146,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.4243109938411887,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3277,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.45815139153395495,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3408,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.43383230210914253,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3496,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.43046508461029864,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3097,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.4647874094039326,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3252,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.43012865605107686,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3212,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.45746801207135634,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3306,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.42380081876437703,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3363,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.4680553345703328,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3241,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.4356710426079786,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3267,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.45505739439055676,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3278,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.456392855685041,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3487,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.45532079162088585,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3522,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.47153077759358625,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.3274,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.404534010722502,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3157,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.4656168614628304,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3451,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.42580701829758794,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3215,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.41198513856049657,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3345,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.4775104898191653,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.3434,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.4149219365588406,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3443,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.4686009766468875,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3258,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.4160615080949939,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3641,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.43888285269816435,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3435,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.4669223238177745,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3538,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.44714725369186825,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3386,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.44946690699814656,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3148,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.39270939253207443,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3425,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.4873887346233523,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3503,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.41509357818361664,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3358,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.4513460735698012,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.3417,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.44798192115574187,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3353,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.4466087772955498,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3205,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.4739307882616213,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3418,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.43664471479534384,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.319,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.43264974785617333,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3308,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.40581684035019994,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3299,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.4472459320746381,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3366,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.4243078079274474,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3564,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.4673007169514019,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3489,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.4094116981361358,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.2992,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.4570723627750309,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3533,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.46028201039509614,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3513,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.4217644389847635,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3503,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.4034773914222882,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3424,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.4685434786275368,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3246,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.3853132761353432,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3419,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.41273639284216,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.339,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.4458965366088044,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3373,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.4246454684767966,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3591,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.3883644444393653,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3627,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.41110608543260185,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3699,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.43155376129423095,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3341,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.4126019786417537,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3337,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.4151608131420393,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3368,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.4410076119607071,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3478,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.37763641277912186,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3336,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.43502942243780995,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.33,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.4062868873221939,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3478,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.40380341539766423,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3504,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.390094569092444,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3263,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.41197263328925754,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3335,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.394584888468792,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3291,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.3931286030188888,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3497,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.4151063983608213,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3281,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.3883073802332517,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3454,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.40649822788352724,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3253,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.3850771776988074,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3443,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.4026561733445534,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3323,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.3936741702179022,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3216,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.3894823226756613,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.328,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.42275520277388356,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3394,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.38596162754414204,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3137,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.3974306174712455,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3518,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.36960328214170685,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3286,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4022265105333133,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3393,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.38420498950714854,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3475,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.41791736291517,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3613,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.3877866121413008,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3556,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.40480501003043134,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3473,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.397790781017423,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3566,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.4186554169738257,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3467,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.38633070810741355,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3232,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.433759492369869,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3497,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.3813063717531377,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3177,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.4264479868184723,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3452,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.4157221433493148,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3588,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.4014759373344182,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3399,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.45196871937128696,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3188,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.4265775459159041,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.35,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.3841139035637085,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.3447,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.4797175327097477,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.332,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.39444308741544076,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3394,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.41907757682172647,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.3563,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5009357563294359,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3594,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.4322676370106741,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3512,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.43177778259555755,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3392,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.41618412195608356,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3251,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.4027382529956285,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3613,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.4211932196123468,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3429,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.37508512805312433,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3333,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.43001469836715717,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3566,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.4446907907873438,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3106,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.41058634335748506,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3453,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.4541001598586059,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3371,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.39991962441544876,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3316,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.4035232649814314,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3459,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.4451790076451386,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3582,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.429956492710246,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3448,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.43685383503148706,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3371,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.42556679001544845,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3251,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.41232347034042893,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3419,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.41578271491274793,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3297,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.42521337633162926,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3494,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.38505788594322904,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3472,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.4429459697381193,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3419,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.4054594698294223,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3036,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.3647082781961668,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3387,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.4531797350545955,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.335,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.40552046016695387,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3354,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.45007772347198005,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3398,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.39021131311441976,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3445,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.4534178948830891,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3492,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.41406613248185314,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.3568,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4902796803392371,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3377,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.40091923515792094,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3348,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.42758573798446875,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3629,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4193843781699617,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3538,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.4082387232110178,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3165,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.4461274657088827,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3244,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.43387551898981375,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3345,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.4103633422714567,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3446,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.42007074283860096,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3444,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.3854030219025788,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3365,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.41964944911654706,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3312,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.37366001219211736,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3484,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.3855255345815174,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3328,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.3867570370536385,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.358,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.40909228350856314,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3253,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.37532079368167953,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3643,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4077968615877349,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3285,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.3851862516925784,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3493,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.45148159660991305,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3517,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.37627619813866997,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.344,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.41068939850056674,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3394,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.38041517934208524,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3572,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.3923418139617557,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3233,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.380134892092584,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3398,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.40384388767081975,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3509,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.3916186589970315,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3554,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.3930496170496318,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3431,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.39023854316600315,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3367,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.4077003079071397,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.352,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.3762914417572129,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3489,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.37607858984988596,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.331,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.43252354837334545,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3436,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.38513265087804244,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3301,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.3987477237562886,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3377,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.4000538898578094,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3574,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.44477005987496365,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.3362,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.39810978492048155,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3139,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.4051116859157475,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3337,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.3993037325348538,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3593,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.4009631452292928,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3182,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.42798528032303634,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3503,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.3928178263501947,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3393,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.4200580654499311,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3518,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.38592853043913306,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3434,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4180535384249045,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3336,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.39417118919816774,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3547,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.3787916570644741,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3348,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.4217043845658041,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.345,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.39697303546774415,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3222,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.36905510540194403,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3423,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.43198375105498504,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3352,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.3826040989374758,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.344,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.39311455251717814,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3353,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.3860994908197377,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3516,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.40864715998441653,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3418,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.42192716654244083,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3267,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.41075573102214763,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3461,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.4073993888935479,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3255,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.4301968572461143,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3395,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.40605421000711234,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3359,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.4125878273102405,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3394,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.38470538973114626,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3333,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.40256817397357897,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.335,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.37549726323293586,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3335,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.3833430691506391,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.3177,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.39991496548449934,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3291,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.37754601439889385,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3242,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.36832740660178726,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3485,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.3955064291109623,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3197,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.3937058023164765,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3348,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.3904364130900761,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3297,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.4046991654914996,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3357,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.39726791760340957,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3251,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.375361659290437,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3301,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4122315600409959,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3582,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.4006988906822774,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.3427,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.36374985779466606,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3357,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.388722356091092,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3459,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.42601192477537375,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3463,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.39365306757229157,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3265,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.3841944454020071,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3221,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.3862284148611294,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3394,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.4031090852642816,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3214,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.39536633709706054,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3222,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.3979694766195916,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3443,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.4003543748327734,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3235,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.38956772046541155,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3586,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.3546245485667708,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3303,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.3739491468703554,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3357,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.37748159464394004,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.3379,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.3578127576755293,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3359,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.36449422585210056,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3358,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.3868892587204477,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3397,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4016311138767254,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3369,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.39742940778250857,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3271,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.3861726174439287,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3387,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.3767800927226333,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3339,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.4281705320286349,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3285,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.38494183214403416,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3445,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.40521890241806563,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3432,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.38096660370997987,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3317,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.46455855015746467,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3444,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.35252998305000993,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.3215,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.3773738996575007,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.335,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.4241559366854113,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.282,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.5354690704324638,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.253,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 0.3357337728532011,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2323,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.6222270356416003,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2178,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.6337506622150998,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2243,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.4027558230748313,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2444,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5248975278789321,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.2336,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.4571543440524602,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.2339,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.42036342639929175,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2494,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.3921423075538725,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2203,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.4460208736659324,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2327,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.4143060142314718,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2166,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.4119227142701445,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2241,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.39646105473112075,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.2343,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.36148748895218163,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2495,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.4015001231516159,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.217,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.3574017000770027,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2192,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.3365816799300959,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2312,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.36877722558569714,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2208,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.35631572232162545,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2187,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.35069549427783897,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2325,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.3553062451744101,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2072,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.3420403170781126,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1988,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.3486676213670596,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2291,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.3426625655168989,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.2287,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.33077342830185247,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2209,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.359148322719356,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2288,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.3289521504251788,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2604,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.3833967199578771,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2124,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.32986743959931336,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2123,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.3288378066560614,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2266,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.33963276401296877,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2334,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.3457793590230139,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2066,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.3350916627858506,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2138,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.33640719633637345,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2351,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.318341392768127,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2277,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.34092327214585055,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2158,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.3402448093998678,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2256,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.33715144458273677,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.222,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.32792073554044465,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2364,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.3460187572640008,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2087,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.35077947757336964,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.2196,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.3302958718713349,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2303,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.3396654969387613,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2076,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.3613400887344227,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.209,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.32295349739442075,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2156,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.33180248037956683,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2081,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.3259609792481737,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2466,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.3380513741717029,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2209,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.31692532349939767,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2328,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.3345387650145794,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2005,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.3265958718239215,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2423,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.3219716635600748,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2293,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.31061055886693384,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2215,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.3251488949912862,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2191,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.3374358557218252,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2181,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.44234165688982746,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.2105,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.3220334386558045,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.237,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.31800467976115254,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2177,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.34115795754948636,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2145,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.3260274558135038,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2264,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.3487592594054172,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2473,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.3285424424874941,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2214,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.34675316959256475,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.2246,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.33773880658270056,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2189,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.357611737745086,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2047,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.33525950831368084,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2349,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.3467029419246968,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2177,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.3327133678210005,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.232,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.3332459630751777,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2309,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.336463032531346,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2225,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.3311414160227271,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2178,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.3337528943121259,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2303,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.3369976536382718,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2159,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.33603493312252286,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2102,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.33394727504473504,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.2236,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.3345664717494832,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2516,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.3216790674788729,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.1893,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.33128699964981523,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.2351,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.3309338973850223,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.215,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.3364585325191218,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2133,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.34329052850302405,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2213,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.3258885096403343,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.2392,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.3227042705836817,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2274,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.3426383610251034,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2193,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.3350015219269101,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2109,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.34035153165829946,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2183,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.32032421399262073,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.22,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.3641867398972823,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2239,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.3335331531703671,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2257,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.33786900858704133,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.207,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.33457886395835595,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2409,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.33511957217114485,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.2293,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.316625786307599,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2088,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.3315206555330983,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2296,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.32652199843153795,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2441,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.33929993330306224,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.227,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.31580624131764506,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2383,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.324600198339504,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.2168,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.3258249506589105,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2219,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.33906599803135845,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2196,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.32535102836899044,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2145,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.31849876853915765,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2397,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.3128469276859356,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2606,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.3140117406481586,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.222,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.31092358599676634,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2116,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.3597447521649351,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.246,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.3419748610068261,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.21,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.3165690296814663,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2494,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.32293284249208803,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2184,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.3225868010027006,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.221,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.32127857008380845,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.226,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.31064940299208327,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2208,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.3245101633496895,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2182,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.3261495209345163,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.2015,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.30049540881198616,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2285,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.3164713496084414,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2282,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.32553432894761014,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2636,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.3302899878132656,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1943,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.31808148454879526,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2198,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.32544871697141836,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2358,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.3180779874073218,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2163,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.322594772302802,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.2311,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.32166914292353493,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2137,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.3343454513480529,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2216,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.3032769445043106,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2271,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.3193781153263876,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2337,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.3337315915827545,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2298,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.34056935068875543,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2198,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.32272656164575164,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2173,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.31936388029362744,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.2192,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.3177018526471647,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.214,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.31609485310912905,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2185,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.3119007570851822,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2144,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.30520802208748,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2313,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.31631939553423594,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2032,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.3338856362336441,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2264,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.33233935190161484,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2431,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.33091421550649636,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2297,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.3118447625215883,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2112,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.33080042888436817,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2391,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.3194335444296353,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2204,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.313229543404351,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.22,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.31940191252905903,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2144,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.3205892892112837,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2271,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.3277229069285714,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2156,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.32319870819723184,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2285,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.30678059294961824,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2355,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.3121667564789698,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2089,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.31825775380675003,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2106,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.32641320866070256,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2191,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.31093970165312473,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2024,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.3251496766618291,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2326,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.3257621178735087,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2127,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.31809494747570627,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.2307,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.308759433989013,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.239,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.31203699121089534,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.2235,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.32903957170577125,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2409,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.3254048790820728,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2248,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.34491972390164977,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2264,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.3148808240979574,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2321,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.30973119276848127,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2258,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.3172512708906623,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2043,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.3195054529632741,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2335,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.3326085588702874,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2227,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.32555935435050865,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2181,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.3185139258260917,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2289,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.33226460973104943,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.24,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.3244298868612208,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.2509,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.31681172649376427,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2379,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.31944515461544754,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.207,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.3195222297978593,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.2202,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.3181276263022896,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.2058,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.3116943177598926,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2192,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.3083134994326932,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2322,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.3003345995381344,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2247,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.3172219303649753,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2295,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.3189007913314121,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2152,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.3063099741978546,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2292,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.31219693669511284,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2247,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.3175093363618314,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2186,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.315074552802494,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2249,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.31563878152306674,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2142,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.31586932039401655,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2312,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.3032768621114548,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2439,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.3026370686782438,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2249,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.32477882315540135,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2334,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.32038525832902837,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2418,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.3126268404942415,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2258,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.29509176708711293,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2273,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.31913106136912256,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2173,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.3248707650400626,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2199,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.302224957212772,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2255,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.3054092337288155,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2273,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.32908783086443755,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2086,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.3114307864380307,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2193,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.299492092105162,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2268,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.31305441798601563,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2138,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.3314047038902724,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2217,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.3096549675292402,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2229,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.301204935316647,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2314,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.3071853030311517,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2214,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.3111530182755258,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2183,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.31850949710917464,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2052,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.3048996564219933,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2262,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.3072003463255179,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2209,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.3229105748500999,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2099,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.30784375944925696,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2192,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.3053049131703333,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2217,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.30530110177850783,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2408,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.3181733173592342,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1969,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.303858729319632,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2296,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.3099235902700981,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2324,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.32467673021780596,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2347,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.32242391447238794,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2164,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.30360620277289874,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2424,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.310013989342044,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2376,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.31852722036792563,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2314,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.31093425309377404,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2116,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.3142928079837063,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.216,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.3074510034390114,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2106,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.308881822524384,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2159,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.31371456559186034,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2182,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.32577666829930896,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2254,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.3099154686860936,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.225,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.30524545120870333,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2214,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.3050509287166143,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.219,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.31465579766183105,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2288,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.30128158480613604,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2294,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.30952387987166147,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2254,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.30552738499865795,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2274,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.3108005327104246,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1983,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.31369450044178754,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.2215,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.31685440370278317,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2304,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.30426526483547406,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2247,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.3158988911502579,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.221,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.3021008353093867,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2197,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.2975280292647164,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2273,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.3013376068913311,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1983,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.30440103949256053,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.2595,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.3124406235071509,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2134,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.30460655363951966,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2318,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.3019282095117422,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2372,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.3078659704760482,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1934,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.31178499640212565,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.2187,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.309738193754049,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2252,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.31289191224394786,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2233,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.3337602754620193,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2008,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.42998109720803834,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1726,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.3343534500219152,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1554,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.2542582943232771,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.175,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 0.28018127743192783,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.1605,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 0.40155905083740107,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.1815,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.3606263435933922,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.1463,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.4179207906548961,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.1675,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.32154553364993077,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.1818,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.28551423620404265,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.1508,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.3298741955894742,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.1402,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.3381959964594119,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.156,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.3009997582829704,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1472,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.26796643293714884,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1544,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.2692757806659747,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1691,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.2909696559407753,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.1616,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.2778027461515208,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.1619,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.2734625791267229,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.151,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.2735305464210604,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1557,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.25108675606141895,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.1511,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.25412947905038086,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1807,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.2639712077885802,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.1598,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.2787419688270002,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.1578,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.2737702888012554,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1714,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.2742665875863831,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.1652,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.2635359706215975,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.1906,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.26654052356494295,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.158,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.2633764342761335,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1613,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.265901874861318,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.1565,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.2733966951587255,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.1738,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.26592747379677883,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.1599,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.2739261369805341,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1929,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.2678019294344197,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.1765,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.2577058283590938,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.1965,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.2700969603964516,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1709,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.26741963154898785,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.1618,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.2623915096483959,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1659,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.254674153789631,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1741,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.2453695259692025,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1628,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.25880402122157226,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.1529,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.2816423431112007,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1598,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.2724364256159644,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.1353,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.26403636371565414,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.1515,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.2625585730927396,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.1645,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.26169904482853434,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.1482,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.2585165559100004,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.1647,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.2629525168986688,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1701,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.26043918622502815,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1488,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.2563765243392665,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1576,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.26399792323290966,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1701,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.2545643529871264,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1465,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.25130650580859165,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.1662,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.2561830157144275,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.1852,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.26022786704972467,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.16,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.2728077675945061,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1584,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.2659616238574817,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.1656,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.26062908369653315,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.1517,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.2592375463848676,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.1587,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.25707005076506906,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.1532,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.2627119364694698,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.1628,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.24972324103233667,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.1703,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.2541095046633754,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.1403,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.25216352449355256,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1458,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.2573591142660321,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.1577,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.253449153174683,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.1718,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.2521000056919566,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.1716,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.2573544512248772,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1548,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.2556141954939716,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1642,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.2466870615915138,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.152,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.2597955389928222,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.168,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.24812016681737786,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1693,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.253738266828648,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1704,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.24610387312600196,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.1515,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.2609880469399408,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.166,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.24996118058016958,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.174,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.258140953350627,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.1681,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.24827389929029212,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1753,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.2571827327624212,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.1529,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.2552060186880473,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.1538,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.2292231389635334,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.1667,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.2626607455940731,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.1504,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.2666621487805961,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1627,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.264290352340411,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.1555,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.24660256686484505,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1624,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.2445502907448152,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.1525,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.2565466780407326,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.1568,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.26051601432415356,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.167,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.2585198489560051,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.1678,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.2553589988952847,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1587,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.2550695565262399,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.1748,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.2675685973602138,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.1583,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.2413210648995397,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1707,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.25140200495109527,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.1572,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.25458712763484437,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1634,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.2556194900294045,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.174,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.2517692170971332,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1511,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.2391707466063267,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1684,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.24213501173195628,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.179,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.26406622366331645,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.1542,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.26149933398122627,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.1597,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.2406040950139673,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.1597,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.25246011867721246,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.1682,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.25544390204687245,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.1724,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.2588955663426414,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1629,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.25189367402790724,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.1682,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.2465532768874959,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.1573,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.25796533900439,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.1806,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.2650722277145831,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1816,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.25794572798353066,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1696,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.23999336072935465,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.1558,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.24816719596306294,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.1645,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.25280889235607373,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1706,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.24632722670962875,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1608,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.24572266051613656,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.1696,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.2524121937481412,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.1761,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.25236690234574477,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.1941,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.2456513387390916,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1608,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.25851602072263513,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.165,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.2504928698714834,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1635,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.26497442934899845,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.1471,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.2641641017333323,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.1778,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.2511733575590109,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.1723,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.2509742075727257,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1703,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.25124847098775777,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1906,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.2449095508330636,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1578,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.25883861077584663,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.1552,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.23925318105579213,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1636,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.24972459064516547,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1807,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.26079841931223846,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.1595,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.25456534931931146,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.1692,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.25585588453395464,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.1699,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.26160959243467563,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.1474,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.2516885953845625,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.1604,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.2518362543650144,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1857,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.25502566019690937,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1681,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.2545613679571293,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1624,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.255102771567199,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1635,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.2479165996992288,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1606,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.24314577909337257,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.1658,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.2660578279936598,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1621,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.243221209093378,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.1416,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.24641788673854792,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1733,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.26034638337135796,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.1719,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.2471901892249118,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1635,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.25494174659452024,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.139,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.25171729401134346,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1582,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.2512187526402761,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.1567,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.2619964338195267,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1787,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.2442880724723905,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1768,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.24810494811887443,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.1529,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.2520887557694156,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1633,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.24503007771265295,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.1629,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.2570198806086685,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1657,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.2532267774171613,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1529,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.2571962842440809,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.153,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.25016057319806717,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1483,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.24689727104863585,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.1507,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.25502657251797534,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1739,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.24593296445835724,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1578,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.24832045657757038,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.1815,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.25670954964816717,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1734,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.2507726118594846,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.1463,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.24867258301635892,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.1728,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.25458677922001843,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1472,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.2455045693612529,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1485,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.2566158190704263,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.1545,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.24960172994837632,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.1748,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.2532464242533188,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1606,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.2432370694587451,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1604,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.25505612817111956,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1451,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.23543455573143385,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.1525,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.24414326166222863,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1568,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.2489139095866761,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.1663,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.25587061360772395,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.1617,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.2567213800836501,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1583,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.2564020834287697,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.155,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.25224176883652794,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.1597,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.26122775038585316,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.186,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.23499412659911662,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.1459,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.255051107555129,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.1597,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.25576834102215995,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1843,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.254423050441407,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1849,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.24296175600543055,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.1714,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.2446289051936704,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.1533,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.26177606663745184,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.1605,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.2436705710494597,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.1728,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.24390011615727988,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.156,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.2451020075008754,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1578,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.24525665619857515,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.164,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.250357564521719,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1767,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.2545384798908961,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.1758,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.257629775097365,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1652,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.2535759120934949,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.1608,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.24869615554527558,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1665,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.24731127831533306,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.1779,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.251433989297688,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.182,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.25508374393777605,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.1874,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.24569062764332167,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1571,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.2591200246623949,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.1728,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.24764643903257175,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.1656,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.236868631584074,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.1528,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.2541909833059057,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.1504,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.25464113574510616,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.1629,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.24058883225401842,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1721,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.2566725942465741,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1698,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.26899872247709256,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.1825,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.24917748364724632,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.1761,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.24795024521067535,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1773,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.252526383543419,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.1571,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.2647631808283695,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.1808,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.2404409712654962,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1437,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.25557624003246454,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1585,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.2541913809017845,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1623,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.2600229213042745,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.1619,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.24255672536545816,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1507,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.24762079481452415,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.1402,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.2399994241196984,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1509,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.24091737042170563,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.1521,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.25715800249959747,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.1724,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.2474516139182865,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.1576,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.25250919476231787,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.1711,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.24399056569280894,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1649,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.2519193867170711,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.187,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.25852198905449875,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.1575,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.25166366123280753,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1597,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.2408879322847605,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.1571,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.24476940643004644,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.1637,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.24670085123108432,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1613,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.2628237594408418,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1661,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.23754546258042714,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.1389,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.2588961709168392,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.1711,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.25328474388794525,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.1608,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.24975981200287972,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1544,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.24645020264641082,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1654,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.24531304897779438,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.1636,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.25583308235807917,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1448,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.2509602250273933,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.1658,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.24747151138212903,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.1332,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.24673474678016463,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.1632,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.2524511150766962,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.1428,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.2425887858272819,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1579,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.25366454801510685,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1518,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.24417397474161678,
+      "learning_rate": 0.0,
+      "loss": 0.1608,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.2884355673913754e+18,
+      "train_loss": 0.3773756278360762,
+      "train_runtime": 37669.6474,
+      "train_samples_per_second": 4.194,
+      "train_steps_per_second": 0.033
+    }
+  ],
+  "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": 2.2884355673913754e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}