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": 4.983291139240507,
+  "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": 6.954410400383741,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.0837,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 6.9400501865680555,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0654,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 6.7987504048943945,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0768,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 6.653575167921407,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0483,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 6.559019603143511,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.0553,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 6.150383345282052,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.0362,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 5.046946677704143,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.0315,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 4.548099510649266,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.997,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.0248076855242343,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.9453,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.764462686471018,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.9597,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.6330409014681746,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.9552,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.8043868831683882,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.9285,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 4.119280633506145,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.9286,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.806498484710525,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.914,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.349101031015638,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.9049,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.457118661708991,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.8648,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.395131480833892,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.8447,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.027011547926397,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.8467,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.5016441178113864,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.8003,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.1862156325870468,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.7855,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.169651971202369,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.79,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.4733903642449016,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.7689,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.3007351461160763,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.768,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.0135913590505714,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.752,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.9661750296419254,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.7604,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.9929522983734735,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.7456,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.9956895280529141,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.7252,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.9310742583106661,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.7318,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.8568515449253671,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.7202,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7882826888909908,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.7111,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.7919361910508409,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.7167,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.7880585842507666,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.7192,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7751376967578392,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.7028,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.7543564519537581,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.7106,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.8329757642433349,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.7001,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.7419723430227314,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.6805,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8731532526048719,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.6994,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.7210480525465929,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.6962,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.7976769488728342,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.6924,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6495276113429324,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.6877,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7452636694535398,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6948,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8246595290829414,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.6855,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7894603713964571,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.6833,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.9393220260027626,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.6866,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7011909671509465,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.6859,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.8425658746356457,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.6613,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7112503449024806,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.669,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.8707079693405161,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.6547,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7922885285886438,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.6857,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7671348882310398,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.6805,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.6440900127496116,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.6589,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7917500664178065,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.6594,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6813330384265905,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.6485,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.6904098970937212,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.6455,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.7515573252311591,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.6488,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7250902560938726,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.6736,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.7592467420123675,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.6677,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7319093626699653,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.6615,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.941881720722119,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.6472,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.2047165847910062,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.6479,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.7321403448119913,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.6461,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.9518601971353224,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.6502,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 1.0163651023692077,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.6602,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9169486416046632,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.6413,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.6974057537080287,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.6451,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.1670836784716068,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.658,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.9236023691301253,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.6633,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.9347693368298489,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.6339,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.4303176457391067,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.628,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.9600686276432909,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.6424,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.4317751086843131,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.6525,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.9064738079875657,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.6434,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.486491805755955,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.6378,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.982784036935644,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.6485,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.4870922981152221,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.6498,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.1435257808275623,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.6425,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.4789920752305985,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.6619,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.1260615625721069,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.6461,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4740105074613266,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.619,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.3869940427845926,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.626,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.3034212101455251,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.6252,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.68289411461455,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.644,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.2626834984377235,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.6414,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.314289626288198,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.6482,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.2491504421527542,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.6405,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.8518429377389781,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.6417,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.9874957212846749,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.6309,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.1199611650893155,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.6381,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.007659922740823,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.6193,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.9381623893482547,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.6308,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.4740013444360531,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.6449,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.9208216256779698,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.6179,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.1077264548000336,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.6094,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.1613227955681877,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.6278,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.3362050082614625,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.6256,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.9076979750116068,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.6301,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.361369247370263,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.6333,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.9358892842812995,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.6539,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.2380687982341076,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.6374,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.5932098618922832,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.6381,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.0245121892036138,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.6392,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.5607390147240106,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.6328,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.6175403214791868,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.6378,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.0931575136863096,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.6309,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 2.04553597292404,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.6295,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.3641814475366492,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.6208,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 2.175539120577516,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.6454,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.7624187916450553,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.6326,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 2.1079482503983638,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.6357,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.914026520079883,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.6439,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.860470142894017,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.6276,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 2.0408716526069677,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.6291,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.7820366220310464,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.6185,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.7523377683295287,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.6338,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.7615120198101528,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.6336,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.2723732298454176,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.6174,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 2.4191326943774607,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.6313,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.8328826567416288,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.6292,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 2.1116309682757213,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.6333,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.9680226888243344,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.625,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 2.0494773164220157,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.6229,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.5500534822035332,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.6175,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.9139244213929787,
+      "learning_rate": 4e-05,
+      "loss": 0.6168,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.5010358406351174,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.6193,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 2.0321580775349424,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.5987,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.951754324964686,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.6344,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.2377681100408082,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.6133,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.1851428210674686,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.6401,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.2842171163259886,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.621,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.925367595604473,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.6202,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.6400196345433515,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.5935,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.9476606871594433,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.6134,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.9840007533590123,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.6175,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.3088630473639644,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.6253,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.9717992678008205,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.6057,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.5907043418300266,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.6493,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 2.046187139082419,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.6203,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 2.0441984178882873,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.6335,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.3247673726272031,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.6201,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.6008636317180505,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.6442,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.3132785590896519,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.6228,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.0799649469045405,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.6092,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.3084117951246697,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.6187,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.3040702214012478,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.6332,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.3246849971775365,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.6223,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.1133453711361432,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.6203,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.5212733020399984,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.64,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.1918500074346965,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.6175,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.3304436223181861,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.6205,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.389173548371737,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.6347,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.2711928984462,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.6027,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.168692041661893,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.6198,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.3976572786522525,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.601,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.9693732962843065,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.6033,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.2517932069381732,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.618,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.9431714844722686,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.6088,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.1337488036130554,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.6268,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2372875141603026,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.6285,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.011131049183032,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.6127,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.3097453136267005,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.611,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.0564549665367275,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.5992,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.9769830856598268,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.6081,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.1099181496381898,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.6224,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.2987698425103593,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.6248,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.8777740399672798,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.6043,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.5400498813956358,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.6114,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.9996587995756404,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.6217,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.9694227759225438,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.6134,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.3955738162465008,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.6191,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.821718015690657,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.6228,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.5717503184251056,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.6222,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.6115042694426303,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.6348,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.4286086413149055,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.6101,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.5349889681249824,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.6291,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.4745467264198815,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.625,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.2103984621444766,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.6064,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.1122496454477153,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.612,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.400871717877839,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.6364,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.0480487558870482,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.6073,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.6399720785424263,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.6253,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.1487687074885622,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.6078,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.7256647191494647,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.606,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.4815467405829454,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.6342,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.5456781459884399,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.6149,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.42055328929477,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.6252,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.164121811609642,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.608,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.2381164899030535,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.5928,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.9967579966414553,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.5917,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.1090706469905602,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.6033,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.1198692890397965,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.6181,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.3766137535756369,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.6123,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.2633971487281341,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.6192,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.1388142275182511,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.6125,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 1.2922040088889029,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.613,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.9544800492686206,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.5829,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 1.309587387971099,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.6062,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.927450669490755,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.6065,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 1.4028833209782143,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.6117,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.1388047255444957,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.6081,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.1534986291069729,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.6116,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.1732865302072821,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.606,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.8330794708119093,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.6098,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 1.2780310020827839,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.6106,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.9031986914771571,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.6131,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 1.1094640162127052,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.6079,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.0629140797411722,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.6214,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.8947263141622616,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.6129,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.070950132163622,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.6156,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.7313141296035833,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.6035,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.0621208232698145,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.6125,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.8297601109673262,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.603,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 1.0842415166426773,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.6245,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.9406477756620928,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.5952,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.957964930398037,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.5975,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.2555652779638378,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.6084,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.86878070978041,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.6243,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 1.0244995141168223,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.6062,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.443860249427533,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.5874,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.8663628967787877,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.5945,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 1.221101240064337,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.5959,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.8605157920856649,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.5846,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.8378182215296612,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.6093,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6404433781517037,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.6063,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 1.0888078181364826,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.6158,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.8422213424676048,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.6101,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.7957737135782372,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.6209,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 1.0194480164148838,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.6061,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 1.3466777738047577,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.6025,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.9660390241969244,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.6102,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 1.2649901891365827,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.6044,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 1.1416037966870898,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.6104,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.8753199532251423,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.5985,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.977295084131597,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.6156,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.751475625724539,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.596,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.8425923067611041,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.6148,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.9775990078269458,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.6082,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8953733267050219,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.6011,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 1.0266294493342953,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.5998,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.9750240426544625,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.5913,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.8308003291204458,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.5951,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.986076181883239,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.6039,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.2635656103741617,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.5956,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.9455852149981113,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.5861,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.0584159324395885,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.6132,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 1.0811629911718001,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.5917,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.9477403399860244,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.6071,
+      "step": 246
+    },
+    {
+      "epoch": 1.0007594936708861,
+      "grad_norm": 1.2436834332179856,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.7038,
+      "step": 247
+    },
+    {
+      "epoch": 1.0048101265822784,
+      "grad_norm": 0.9638437861566009,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.5592,
+      "step": 248
+    },
+    {
+      "epoch": 1.008860759493671,
+      "grad_norm": 1.0195205462687125,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.5238,
+      "step": 249
+    },
+    {
+      "epoch": 1.0129113924050632,
+      "grad_norm": 0.8763884395184933,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.5256,
+      "step": 250
+    },
+    {
+      "epoch": 1.0169620253164557,
+      "grad_norm": 1.2454463268794869,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.5769,
+      "step": 251
+    },
+    {
+      "epoch": 1.021012658227848,
+      "grad_norm": 0.8986589153450555,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.5288,
+      "step": 252
+    },
+    {
+      "epoch": 1.0250632911392406,
+      "grad_norm": 0.8518925757174307,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.4986,
+      "step": 253
+    },
+    {
+      "epoch": 1.0291139240506328,
+      "grad_norm": 1.348489504928881,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.5607,
+      "step": 254
+    },
+    {
+      "epoch": 1.0331645569620254,
+      "grad_norm": 0.7634655371568729,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.5261,
+      "step": 255
+    },
+    {
+      "epoch": 1.0372151898734177,
+      "grad_norm": 1.1094197341691527,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.5128,
+      "step": 256
+    },
+    {
+      "epoch": 1.0412658227848102,
+      "grad_norm": 0.8771192758331756,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.5359,
+      "step": 257
+    },
+    {
+      "epoch": 1.0453164556962025,
+      "grad_norm": 0.9501647060341649,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5477,
+      "step": 258
+    },
+    {
+      "epoch": 1.049367088607595,
+      "grad_norm": 1.1602003663008746,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.5152,
+      "step": 259
+    },
+    {
+      "epoch": 1.0534177215189873,
+      "grad_norm": 0.9029510647636,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.5615,
+      "step": 260
+    },
+    {
+      "epoch": 1.0574683544303798,
+      "grad_norm": 1.0403602720169456,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.528,
+      "step": 261
+    },
+    {
+      "epoch": 1.061518987341772,
+      "grad_norm": 1.131859261218781,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.523,
+      "step": 262
+    },
+    {
+      "epoch": 1.0655696202531646,
+      "grad_norm": 0.8401834424919907,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.5322,
+      "step": 263
+    },
+    {
+      "epoch": 1.0696202531645569,
+      "grad_norm": 0.7309117877766685,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.5351,
+      "step": 264
+    },
+    {
+      "epoch": 1.0736708860759494,
+      "grad_norm": 0.8503214419644856,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.5494,
+      "step": 265
+    },
+    {
+      "epoch": 1.0777215189873417,
+      "grad_norm": 0.6846960317707459,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.5344,
+      "step": 266
+    },
+    {
+      "epoch": 1.0817721518987342,
+      "grad_norm": 0.8696235595377376,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.5648,
+      "step": 267
+    },
+    {
+      "epoch": 1.0858227848101265,
+      "grad_norm": 0.6852078471349666,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.4944,
+      "step": 268
+    },
+    {
+      "epoch": 1.089873417721519,
+      "grad_norm": 0.9165859207315454,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.5257,
+      "step": 269
+    },
+    {
+      "epoch": 1.0939240506329113,
+      "grad_norm": 1.1448029066857075,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.5308,
+      "step": 270
+    },
+    {
+      "epoch": 1.0979746835443038,
+      "grad_norm": 0.7093320877956338,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.528,
+      "step": 271
+    },
+    {
+      "epoch": 1.1020253164556961,
+      "grad_norm": 1.3304757380361536,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.546,
+      "step": 272
+    },
+    {
+      "epoch": 1.1060759493670886,
+      "grad_norm": 0.8743743449729098,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.5049,
+      "step": 273
+    },
+    {
+      "epoch": 1.110126582278481,
+      "grad_norm": 0.92894133297133,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.5172,
+      "step": 274
+    },
+    {
+      "epoch": 1.1141772151898734,
+      "grad_norm": 0.9893456300412816,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.5424,
+      "step": 275
+    },
+    {
+      "epoch": 1.1182278481012657,
+      "grad_norm": 0.9801139239828293,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.5403,
+      "step": 276
+    },
+    {
+      "epoch": 1.1222784810126583,
+      "grad_norm": 0.6996329860296641,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.5308,
+      "step": 277
+    },
+    {
+      "epoch": 1.1263291139240506,
+      "grad_norm": 1.1345533307193476,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.5238,
+      "step": 278
+    },
+    {
+      "epoch": 1.130379746835443,
+      "grad_norm": 0.7520919989057505,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.5474,
+      "step": 279
+    },
+    {
+      "epoch": 1.1344303797468354,
+      "grad_norm": 0.794550011323583,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.5533,
+      "step": 280
+    },
+    {
+      "epoch": 1.1384810126582279,
+      "grad_norm": 0.6673607268797548,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.4961,
+      "step": 281
+    },
+    {
+      "epoch": 1.1425316455696202,
+      "grad_norm": 0.8341911995819137,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.5828,
+      "step": 282
+    },
+    {
+      "epoch": 1.1465822784810127,
+      "grad_norm": 0.6132443136973689,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.5132,
+      "step": 283
+    },
+    {
+      "epoch": 1.150632911392405,
+      "grad_norm": 0.6356379033728499,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.5196,
+      "step": 284
+    },
+    {
+      "epoch": 1.1546835443037975,
+      "grad_norm": 0.604809444480571,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5183,
+      "step": 285
+    },
+    {
+      "epoch": 1.1587341772151898,
+      "grad_norm": 0.7573407320210432,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.5175,
+      "step": 286
+    },
+    {
+      "epoch": 1.1627848101265823,
+      "grad_norm": 0.7527626341375323,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.572,
+      "step": 287
+    },
+    {
+      "epoch": 1.1668354430379746,
+      "grad_norm": 0.607057118521876,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.5068,
+      "step": 288
+    },
+    {
+      "epoch": 1.1708860759493671,
+      "grad_norm": 0.7853849014102718,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.5424,
+      "step": 289
+    },
+    {
+      "epoch": 1.1749367088607594,
+      "grad_norm": 0.7181432734011717,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.5583,
+      "step": 290
+    },
+    {
+      "epoch": 1.178987341772152,
+      "grad_norm": 0.6347936149329371,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.5218,
+      "step": 291
+    },
+    {
+      "epoch": 1.1830379746835442,
+      "grad_norm": 0.7940849882469808,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.5431,
+      "step": 292
+    },
+    {
+      "epoch": 1.1870886075949367,
+      "grad_norm": 0.6977318096600906,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.5461,
+      "step": 293
+    },
+    {
+      "epoch": 1.191139240506329,
+      "grad_norm": 0.6026651185754219,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.5162,
+      "step": 294
+    },
+    {
+      "epoch": 1.1951898734177215,
+      "grad_norm": 0.711696482947943,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.5564,
+      "step": 295
+    },
+    {
+      "epoch": 1.1992405063291138,
+      "grad_norm": 0.6026645501695609,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.5395,
+      "step": 296
+    },
+    {
+      "epoch": 1.2032911392405063,
+      "grad_norm": 0.7676258704784457,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.5349,
+      "step": 297
+    },
+    {
+      "epoch": 1.2073417721518986,
+      "grad_norm": 0.7325564535677532,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.5271,
+      "step": 298
+    },
+    {
+      "epoch": 1.2113924050632912,
+      "grad_norm": 0.7628311419306474,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.5514,
+      "step": 299
+    },
+    {
+      "epoch": 1.2154430379746834,
+      "grad_norm": 0.7079645279405029,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.5216,
+      "step": 300
+    },
+    {
+      "epoch": 1.219493670886076,
+      "grad_norm": 0.6179398153613539,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.5231,
+      "step": 301
+    },
+    {
+      "epoch": 1.2235443037974683,
+      "grad_norm": 0.782592807988237,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.5656,
+      "step": 302
+    },
+    {
+      "epoch": 1.2275949367088608,
+      "grad_norm": 0.6289218065165803,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.4968,
+      "step": 303
+    },
+    {
+      "epoch": 1.231645569620253,
+      "grad_norm": 0.7497492765282757,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.6238,
+      "step": 304
+    },
+    {
+      "epoch": 1.2356962025316456,
+      "grad_norm": 0.7461821944857912,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4796,
+      "step": 305
+    },
+    {
+      "epoch": 1.239746835443038,
+      "grad_norm": 0.7818311181532216,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.5606,
+      "step": 306
+    },
+    {
+      "epoch": 1.2437974683544304,
+      "grad_norm": 0.6659066386688293,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.5097,
+      "step": 307
+    },
+    {
+      "epoch": 1.2478481012658227,
+      "grad_norm": 0.7425212594202908,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.5513,
+      "step": 308
+    },
+    {
+      "epoch": 1.2518987341772152,
+      "grad_norm": 0.6912479803768411,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.5256,
+      "step": 309
+    },
+    {
+      "epoch": 1.2559493670886077,
+      "grad_norm": 0.7385710268644965,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.557,
+      "step": 310
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6485731667621707,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.5406,
+      "step": 311
+    },
+    {
+      "epoch": 1.2640506329113923,
+      "grad_norm": 0.6410179332168048,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.5024,
+      "step": 312
+    },
+    {
+      "epoch": 1.2681012658227848,
+      "grad_norm": 0.7772476788126488,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.5584,
+      "step": 313
+    },
+    {
+      "epoch": 1.2721518987341773,
+      "grad_norm": 0.6701116268060914,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4975,
+      "step": 314
+    },
+    {
+      "epoch": 1.2762025316455696,
+      "grad_norm": 0.7154042254396068,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.5546,
+      "step": 315
+    },
+    {
+      "epoch": 1.280253164556962,
+      "grad_norm": 0.8096252649782184,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.5268,
+      "step": 316
+    },
+    {
+      "epoch": 1.2843037974683544,
+      "grad_norm": 0.7673995341666917,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.5586,
+      "step": 317
+    },
+    {
+      "epoch": 1.288354430379747,
+      "grad_norm": 0.6680255101434694,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.5169,
+      "step": 318
+    },
+    {
+      "epoch": 1.2924050632911392,
+      "grad_norm": 0.519261338824872,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.4875,
+      "step": 319
+    },
+    {
+      "epoch": 1.2964556962025315,
+      "grad_norm": 0.7953709135081325,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.5947,
+      "step": 320
+    },
+    {
+      "epoch": 1.300506329113924,
+      "grad_norm": 0.5158012911832226,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.5273,
+      "step": 321
+    },
+    {
+      "epoch": 1.3045569620253166,
+      "grad_norm": 0.7278128360926331,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.5438,
+      "step": 322
+    },
+    {
+      "epoch": 1.3086075949367089,
+      "grad_norm": 0.7358432376349991,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.5204,
+      "step": 323
+    },
+    {
+      "epoch": 1.3126582278481012,
+      "grad_norm": 0.7637597844284224,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.5242,
+      "step": 324
+    },
+    {
+      "epoch": 1.3167088607594937,
+      "grad_norm": 0.7478727899034836,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.5575,
+      "step": 325
+    },
+    {
+      "epoch": 1.3207594936708862,
+      "grad_norm": 0.7782192392141688,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.5347,
+      "step": 326
+    },
+    {
+      "epoch": 1.3248101265822785,
+      "grad_norm": 0.7807495727404549,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.5179,
+      "step": 327
+    },
+    {
+      "epoch": 1.3288607594936708,
+      "grad_norm": 0.7775124959092382,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.5464,
+      "step": 328
+    },
+    {
+      "epoch": 1.3329113924050633,
+      "grad_norm": 0.7901677098335301,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.5726,
+      "step": 329
+    },
+    {
+      "epoch": 1.3369620253164558,
+      "grad_norm": 0.7475655876222138,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.5024,
+      "step": 330
+    },
+    {
+      "epoch": 1.341012658227848,
+      "grad_norm": 0.8149120442684487,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.5332,
+      "step": 331
+    },
+    {
+      "epoch": 1.3450632911392404,
+      "grad_norm": 0.8939349270223009,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.5139,
+      "step": 332
+    },
+    {
+      "epoch": 1.349113924050633,
+      "grad_norm": 1.048261110975364,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.5756,
+      "step": 333
+    },
+    {
+      "epoch": 1.3531645569620254,
+      "grad_norm": 0.8905279689523212,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.4995,
+      "step": 334
+    },
+    {
+      "epoch": 1.3572151898734177,
+      "grad_norm": 1.0943134766500848,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.5729,
+      "step": 335
+    },
+    {
+      "epoch": 1.36126582278481,
+      "grad_norm": 0.896179214605804,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.5442,
+      "step": 336
+    },
+    {
+      "epoch": 1.3653164556962025,
+      "grad_norm": 0.8629632136201635,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.5397,
+      "step": 337
+    },
+    {
+      "epoch": 1.369367088607595,
+      "grad_norm": 0.7128834196185092,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.5473,
+      "step": 338
+    },
+    {
+      "epoch": 1.3734177215189873,
+      "grad_norm": 0.7993657201833773,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5227,
+      "step": 339
+    },
+    {
+      "epoch": 1.3774683544303796,
+      "grad_norm": 0.7429877571683161,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.526,
+      "step": 340
+    },
+    {
+      "epoch": 1.3815189873417721,
+      "grad_norm": 0.8158397219728912,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.5269,
+      "step": 341
+    },
+    {
+      "epoch": 1.3855696202531647,
+      "grad_norm": 0.763925680984283,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.5535,
+      "step": 342
+    },
+    {
+      "epoch": 1.389620253164557,
+      "grad_norm": 0.8164728705747548,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.5074,
+      "step": 343
+    },
+    {
+      "epoch": 1.3936708860759492,
+      "grad_norm": 0.5982531873978788,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.5448,
+      "step": 344
+    },
+    {
+      "epoch": 1.3977215189873418,
+      "grad_norm": 0.8406570050186799,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.5309,
+      "step": 345
+    },
+    {
+      "epoch": 1.4017721518987343,
+      "grad_norm": 0.7188483269768406,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.5466,
+      "step": 346
+    },
+    {
+      "epoch": 1.4058227848101266,
+      "grad_norm": 0.7355198570177172,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.5247,
+      "step": 347
+    },
+    {
+      "epoch": 1.4098734177215189,
+      "grad_norm": 0.9070012544967203,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.5518,
+      "step": 348
+    },
+    {
+      "epoch": 1.4139240506329114,
+      "grad_norm": 0.8053420645672132,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.5114,
+      "step": 349
+    },
+    {
+      "epoch": 1.417974683544304,
+      "grad_norm": 0.7172107469358631,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.5665,
+      "step": 350
+    },
+    {
+      "epoch": 1.4220253164556962,
+      "grad_norm": 0.8668334688653698,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.5428,
+      "step": 351
+    },
+    {
+      "epoch": 1.4260759493670885,
+      "grad_norm": 0.7062853417722891,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.5096,
+      "step": 352
+    },
+    {
+      "epoch": 1.430126582278481,
+      "grad_norm": 0.7171562259025508,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.5339,
+      "step": 353
+    },
+    {
+      "epoch": 1.4341772151898735,
+      "grad_norm": 0.7143425070615537,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.5236,
+      "step": 354
+    },
+    {
+      "epoch": 1.4382278481012658,
+      "grad_norm": 0.8734677621493264,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.5419,
+      "step": 355
+    },
+    {
+      "epoch": 1.442278481012658,
+      "grad_norm": 0.6654257823819231,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.5218,
+      "step": 356
+    },
+    {
+      "epoch": 1.4463291139240506,
+      "grad_norm": 0.8771206413013289,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.5663,
+      "step": 357
+    },
+    {
+      "epoch": 1.4503797468354431,
+      "grad_norm": 0.8280025228340774,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4775,
+      "step": 358
+    },
+    {
+      "epoch": 1.4544303797468354,
+      "grad_norm": 0.8613088292236915,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.5176,
+      "step": 359
+    },
+    {
+      "epoch": 1.4584810126582277,
+      "grad_norm": 0.8869421010766709,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.5635,
+      "step": 360
+    },
+    {
+      "epoch": 1.4625316455696202,
+      "grad_norm": 0.8078946982822428,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.5411,
+      "step": 361
+    },
+    {
+      "epoch": 1.4665822784810127,
+      "grad_norm": 0.9527052198102269,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.5279,
+      "step": 362
+    },
+    {
+      "epoch": 1.470632911392405,
+      "grad_norm": 0.7379611889308306,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4948,
+      "step": 363
+    },
+    {
+      "epoch": 1.4746835443037973,
+      "grad_norm": 1.037536832374824,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.5615,
+      "step": 364
+    },
+    {
+      "epoch": 1.4787341772151898,
+      "grad_norm": 1.0693374991753821,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.5444,
+      "step": 365
+    },
+    {
+      "epoch": 1.4827848101265824,
+      "grad_norm": 0.7138031816922276,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.5548,
+      "step": 366
+    },
+    {
+      "epoch": 1.4868354430379747,
+      "grad_norm": 1.3170934039355493,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.5065,
+      "step": 367
+    },
+    {
+      "epoch": 1.490886075949367,
+      "grad_norm": 0.8200775846675421,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.5513,
+      "step": 368
+    },
+    {
+      "epoch": 1.4949367088607595,
+      "grad_norm": 1.6124921248441304,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.5321,
+      "step": 369
+    },
+    {
+      "epoch": 1.498987341772152,
+      "grad_norm": 1.1346409082267763,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.4799,
+      "step": 370
+    },
+    {
+      "epoch": 1.5030379746835443,
+      "grad_norm": 1.4339834045853836,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.5424,
+      "step": 371
+    },
+    {
+      "epoch": 1.5070886075949366,
+      "grad_norm": 1.5250631610232088,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.6127,
+      "step": 372
+    },
+    {
+      "epoch": 1.511139240506329,
+      "grad_norm": 1.082972026084195,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.5136,
+      "step": 373
+    },
+    {
+      "epoch": 1.5151898734177216,
+      "grad_norm": 1.2295060702908684,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.5643,
+      "step": 374
+    },
+    {
+      "epoch": 1.519240506329114,
+      "grad_norm": 1.1285295609201969,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.5344,
+      "step": 375
+    },
+    {
+      "epoch": 1.5232911392405062,
+      "grad_norm": 0.9033979036969477,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.5179,
+      "step": 376
+    },
+    {
+      "epoch": 1.5273417721518987,
+      "grad_norm": 1.2815545144400042,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.5404,
+      "step": 377
+    },
+    {
+      "epoch": 1.5313924050632912,
+      "grad_norm": 0.8754576548827399,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.5252,
+      "step": 378
+    },
+    {
+      "epoch": 1.5354430379746835,
+      "grad_norm": 1.1824708259500432,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.5085,
+      "step": 379
+    },
+    {
+      "epoch": 1.5394936708860758,
+      "grad_norm": 0.8337980191058465,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.5192,
+      "step": 380
+    },
+    {
+      "epoch": 1.5435443037974683,
+      "grad_norm": 1.255239915007499,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.5515,
+      "step": 381
+    },
+    {
+      "epoch": 1.5475949367088608,
+      "grad_norm": 0.802745331443439,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.5048,
+      "step": 382
+    },
+    {
+      "epoch": 1.5516455696202531,
+      "grad_norm": 1.3700122073883318,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.5635,
+      "step": 383
+    },
+    {
+      "epoch": 1.5556962025316454,
+      "grad_norm": 0.8267956444699474,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.5003,
+      "step": 384
+    },
+    {
+      "epoch": 1.559746835443038,
+      "grad_norm": 1.586334408935424,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.5723,
+      "step": 385
+    },
+    {
+      "epoch": 1.5637974683544305,
+      "grad_norm": 1.0006891398748612,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.5362,
+      "step": 386
+    },
+    {
+      "epoch": 1.5678481012658227,
+      "grad_norm": 1.3721405586236546,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.4872,
+      "step": 387
+    },
+    {
+      "epoch": 1.571898734177215,
+      "grad_norm": 1.6702956322971638,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.5897,
+      "step": 388
+    },
+    {
+      "epoch": 1.5759493670886076,
+      "grad_norm": 0.8034602163346699,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.4987,
+      "step": 389
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4948796979048882,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.5368,
+      "step": 390
+    },
+    {
+      "epoch": 1.5840506329113924,
+      "grad_norm": 0.6712868309282561,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.5213,
+      "step": 391
+    },
+    {
+      "epoch": 1.5881012658227847,
+      "grad_norm": 1.4902244299098586,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.5492,
+      "step": 392
+    },
+    {
+      "epoch": 1.5921518987341772,
+      "grad_norm": 1.1788229748503278,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.5563,
+      "step": 393
+    },
+    {
+      "epoch": 1.5962025316455697,
+      "grad_norm": 1.0962935069258886,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.4841,
+      "step": 394
+    },
+    {
+      "epoch": 1.600253164556962,
+      "grad_norm": 1.2827731023765487,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.5517,
+      "step": 395
+    },
+    {
+      "epoch": 1.6043037974683543,
+      "grad_norm": 0.8593718505449981,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.5328,
+      "step": 396
+    },
+    {
+      "epoch": 1.6083544303797468,
+      "grad_norm": 0.9861967973750696,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.5378,
+      "step": 397
+    },
+    {
+      "epoch": 1.6124050632911393,
+      "grad_norm": 0.7329773454588082,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.473,
+      "step": 398
+    },
+    {
+      "epoch": 1.6164556962025316,
+      "grad_norm": 1.0463843620838313,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.5424,
+      "step": 399
+    },
+    {
+      "epoch": 1.620506329113924,
+      "grad_norm": 0.8542830938780757,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.5633,
+      "step": 400
+    },
+    {
+      "epoch": 1.6245569620253164,
+      "grad_norm": 0.8173024575268236,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.4836,
+      "step": 401
+    },
+    {
+      "epoch": 1.628607594936709,
+      "grad_norm": 0.6466132951606368,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.5346,
+      "step": 402
+    },
+    {
+      "epoch": 1.6326582278481012,
+      "grad_norm": 0.8444118498149114,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.5625,
+      "step": 403
+    },
+    {
+      "epoch": 1.6367088607594935,
+      "grad_norm": 0.6234632699296221,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.5366,
+      "step": 404
+    },
+    {
+      "epoch": 1.640759493670886,
+      "grad_norm": 0.7859143697415626,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.5421,
+      "step": 405
+    },
+    {
+      "epoch": 1.6448101265822785,
+      "grad_norm": 0.6451752432758884,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.5048,
+      "step": 406
+    },
+    {
+      "epoch": 1.6488607594936708,
+      "grad_norm": 0.8490154068029508,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.545,
+      "step": 407
+    },
+    {
+      "epoch": 1.6529113924050631,
+      "grad_norm": 0.6651407785115806,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.5303,
+      "step": 408
+    },
+    {
+      "epoch": 1.6569620253164556,
+      "grad_norm": 0.7876377459377151,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.5362,
+      "step": 409
+    },
+    {
+      "epoch": 1.6610126582278482,
+      "grad_norm": 0.5386576944935724,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.5227,
+      "step": 410
+    },
+    {
+      "epoch": 1.6650632911392405,
+      "grad_norm": 0.8408366515495476,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.5746,
+      "step": 411
+    },
+    {
+      "epoch": 1.6691139240506327,
+      "grad_norm": 0.6528637092171089,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.5078,
+      "step": 412
+    },
+    {
+      "epoch": 1.6731645569620253,
+      "grad_norm": 0.7498749417735426,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.5418,
+      "step": 413
+    },
+    {
+      "epoch": 1.6772151898734178,
+      "grad_norm": 0.6525914263058754,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.5412,
+      "step": 414
+    },
+    {
+      "epoch": 1.68126582278481,
+      "grad_norm": 0.6446934840865008,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.5221,
+      "step": 415
+    },
+    {
+      "epoch": 1.6853164556962026,
+      "grad_norm": 0.6671260437359778,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.4896,
+      "step": 416
+    },
+    {
+      "epoch": 1.6893670886075949,
+      "grad_norm": 0.657120516319114,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.578,
+      "step": 417
+    },
+    {
+      "epoch": 1.6934177215189874,
+      "grad_norm": 0.6191300011832233,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.5039,
+      "step": 418
+    },
+    {
+      "epoch": 1.69746835443038,
+      "grad_norm": 0.696398367403582,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.5314,
+      "step": 419
+    },
+    {
+      "epoch": 1.7015189873417722,
+      "grad_norm": 0.5702118591989692,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5166,
+      "step": 420
+    },
+    {
+      "epoch": 1.7055696202531645,
+      "grad_norm": 0.6919246993681929,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.5615,
+      "step": 421
+    },
+    {
+      "epoch": 1.709620253164557,
+      "grad_norm": 0.6055434786653293,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.5376,
+      "step": 422
+    },
+    {
+      "epoch": 1.7136708860759495,
+      "grad_norm": 0.7329285640587678,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.5845,
+      "step": 423
+    },
+    {
+      "epoch": 1.7177215189873418,
+      "grad_norm": 0.6827633423355121,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.5085,
+      "step": 424
+    },
+    {
+      "epoch": 1.7217721518987341,
+      "grad_norm": 0.5533513988811852,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.5317,
+      "step": 425
+    },
+    {
+      "epoch": 1.7258227848101266,
+      "grad_norm": 0.5557611331384712,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.4868,
+      "step": 426
+    },
+    {
+      "epoch": 1.7298734177215191,
+      "grad_norm": 0.6515161778773387,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.5291,
+      "step": 427
+    },
+    {
+      "epoch": 1.7339240506329114,
+      "grad_norm": 0.548235080196943,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.5124,
+      "step": 428
+    },
+    {
+      "epoch": 1.7379746835443037,
+      "grad_norm": 0.6422666558080213,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.5252,
+      "step": 429
+    },
+    {
+      "epoch": 1.7420253164556962,
+      "grad_norm": 0.6463530863952022,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.5255,
+      "step": 430
+    },
+    {
+      "epoch": 1.7460759493670888,
+      "grad_norm": 0.6129207275947827,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.5533,
+      "step": 431
+    },
+    {
+      "epoch": 1.750126582278481,
+      "grad_norm": 0.6041943265661842,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.5223,
+      "step": 432
+    },
+    {
+      "epoch": 1.7541772151898734,
+      "grad_norm": 0.6293625219699406,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.4969,
+      "step": 433
+    },
+    {
+      "epoch": 1.7582278481012659,
+      "grad_norm": 0.6051251727962021,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.5915,
+      "step": 434
+    },
+    {
+      "epoch": 1.7622784810126584,
+      "grad_norm": 0.5803773565732196,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.4675,
+      "step": 435
+    },
+    {
+      "epoch": 1.7663291139240507,
+      "grad_norm": 0.6611714090537727,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.538,
+      "step": 436
+    },
+    {
+      "epoch": 1.770379746835443,
+      "grad_norm": 0.9164302376294405,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.5442,
+      "step": 437
+    },
+    {
+      "epoch": 1.7744303797468355,
+      "grad_norm": 0.660529992730085,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.5176,
+      "step": 438
+    },
+    {
+      "epoch": 1.778481012658228,
+      "grad_norm": 0.7219912638653426,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.5102,
+      "step": 439
+    },
+    {
+      "epoch": 1.7825316455696203,
+      "grad_norm": 0.5456527223371624,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.5216,
+      "step": 440
+    },
+    {
+      "epoch": 1.7865822784810126,
+      "grad_norm": 0.7188805441524913,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.5151,
+      "step": 441
+    },
+    {
+      "epoch": 1.790632911392405,
+      "grad_norm": 0.7881440981364776,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.5771,
+      "step": 442
+    },
+    {
+      "epoch": 1.7946835443037976,
+      "grad_norm": 0.5857343140429639,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.5028,
+      "step": 443
+    },
+    {
+      "epoch": 1.79873417721519,
+      "grad_norm": 0.7189572571672036,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.5291,
+      "step": 444
+    },
+    {
+      "epoch": 1.8027848101265822,
+      "grad_norm": 0.6731676938115837,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.5198,
+      "step": 445
+    },
+    {
+      "epoch": 1.8068354430379747,
+      "grad_norm": 0.6505047149857615,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.5491,
+      "step": 446
+    },
+    {
+      "epoch": 1.8108860759493672,
+      "grad_norm": 0.6961239245600651,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.498,
+      "step": 447
+    },
+    {
+      "epoch": 1.8149367088607595,
+      "grad_norm": 0.6006940619338451,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.537,
+      "step": 448
+    },
+    {
+      "epoch": 1.8189873417721518,
+      "grad_norm": 0.6072928759056202,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.5134,
+      "step": 449
+    },
+    {
+      "epoch": 1.8230379746835443,
+      "grad_norm": 0.7509347171910691,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.5594,
+      "step": 450
+    },
+    {
+      "epoch": 1.8270886075949369,
+      "grad_norm": 0.5462004767276315,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.5103,
+      "step": 451
+    },
+    {
+      "epoch": 1.8311392405063291,
+      "grad_norm": 0.7401160079373621,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.5362,
+      "step": 452
+    },
+    {
+      "epoch": 1.8351898734177214,
+      "grad_norm": 0.8324607521737274,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.5393,
+      "step": 453
+    },
+    {
+      "epoch": 1.839240506329114,
+      "grad_norm": 0.551488193585642,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.5197,
+      "step": 454
+    },
+    {
+      "epoch": 1.8432911392405065,
+      "grad_norm": 0.7830603835369453,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.4942,
+      "step": 455
+    },
+    {
+      "epoch": 1.8473417721518988,
+      "grad_norm": 0.5783266141423138,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.49,
+      "step": 456
+    },
+    {
+      "epoch": 1.851392405063291,
+      "grad_norm": 0.6400544027725846,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.5464,
+      "step": 457
+    },
+    {
+      "epoch": 1.8554430379746836,
+      "grad_norm": 0.8369562212066275,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.5269,
+      "step": 458
+    },
+    {
+      "epoch": 1.859493670886076,
+      "grad_norm": 0.5856264576567846,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.5317,
+      "step": 459
+    },
+    {
+      "epoch": 1.8635443037974684,
+      "grad_norm": 0.6945156690654839,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.4919,
+      "step": 460
+    },
+    {
+      "epoch": 1.8675949367088607,
+      "grad_norm": 0.7468991026233852,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.5585,
+      "step": 461
+    },
+    {
+      "epoch": 1.8716455696202532,
+      "grad_norm": 0.5931798966489197,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.5648,
+      "step": 462
+    },
+    {
+      "epoch": 1.8756962025316457,
+      "grad_norm": 0.6380303717176617,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.5056,
+      "step": 463
+    },
+    {
+      "epoch": 1.879746835443038,
+      "grad_norm": 0.6404848675497172,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.5193,
+      "step": 464
+    },
+    {
+      "epoch": 1.8837974683544303,
+      "grad_norm": 0.6095747686339011,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.5351,
+      "step": 465
+    },
+    {
+      "epoch": 1.8878481012658228,
+      "grad_norm": 0.7113563185806772,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.5103,
+      "step": 466
+    },
+    {
+      "epoch": 1.8918987341772153,
+      "grad_norm": 0.6586780019617671,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.5616,
+      "step": 467
+    },
+    {
+      "epoch": 1.8959493670886076,
+      "grad_norm": 0.68424999453921,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.5621,
+      "step": 468
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5866570130780453,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.5331,
+      "step": 469
+    },
+    {
+      "epoch": 1.9040506329113924,
+      "grad_norm": 0.7049678346378531,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.5551,
+      "step": 470
+    },
+    {
+      "epoch": 1.908101265822785,
+      "grad_norm": 0.5573047248480321,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.5159,
+      "step": 471
+    },
+    {
+      "epoch": 1.9121518987341772,
+      "grad_norm": 0.6152448790186686,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.4909,
+      "step": 472
+    },
+    {
+      "epoch": 1.9162025316455695,
+      "grad_norm": 0.5725696450057538,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.5401,
+      "step": 473
+    },
+    {
+      "epoch": 1.920253164556962,
+      "grad_norm": 0.9174494751608109,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5229,
+      "step": 474
+    },
+    {
+      "epoch": 1.9243037974683546,
+      "grad_norm": 0.6820056903748684,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.5222,
+      "step": 475
+    },
+    {
+      "epoch": 1.9283544303797469,
+      "grad_norm": 0.6531775914330018,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.5548,
+      "step": 476
+    },
+    {
+      "epoch": 1.9324050632911391,
+      "grad_norm": 0.7219375197681888,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.5349,
+      "step": 477
+    },
+    {
+      "epoch": 1.9364556962025317,
+      "grad_norm": 0.45830651996684696,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.5079,
+      "step": 478
+    },
+    {
+      "epoch": 1.9405063291139242,
+      "grad_norm": 0.766580151973448,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.5313,
+      "step": 479
+    },
+    {
+      "epoch": 1.9445569620253165,
+      "grad_norm": 0.6849201662198069,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.5208,
+      "step": 480
+    },
+    {
+      "epoch": 1.9486075949367088,
+      "grad_norm": 0.5058876423464879,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.5068,
+      "step": 481
+    },
+    {
+      "epoch": 1.9526582278481013,
+      "grad_norm": 0.6866379614602172,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.5408,
+      "step": 482
+    },
+    {
+      "epoch": 1.9567088607594938,
+      "grad_norm": 0.5374063687876747,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.5503,
+      "step": 483
+    },
+    {
+      "epoch": 1.960759493670886,
+      "grad_norm": 0.6688932201293888,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.4961,
+      "step": 484
+    },
+    {
+      "epoch": 1.9648101265822784,
+      "grad_norm": 0.6820234915996668,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.5225,
+      "step": 485
+    },
+    {
+      "epoch": 1.968860759493671,
+      "grad_norm": 0.5260888237593319,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.5742,
+      "step": 486
+    },
+    {
+      "epoch": 1.9729113924050634,
+      "grad_norm": 0.637750016180435,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.5204,
+      "step": 487
+    },
+    {
+      "epoch": 1.9769620253164557,
+      "grad_norm": 0.6059491605210939,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.5375,
+      "step": 488
+    },
+    {
+      "epoch": 1.981012658227848,
+      "grad_norm": 0.5073902709569702,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.5088,
+      "step": 489
+    },
+    {
+      "epoch": 1.9850632911392405,
+      "grad_norm": 0.5974956505847179,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.5077,
+      "step": 490
+    },
+    {
+      "epoch": 1.989113924050633,
+      "grad_norm": 0.6135404872694891,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.5752,
+      "step": 491
+    },
+    {
+      "epoch": 1.9931645569620253,
+      "grad_norm": 0.5957821478548706,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.5259,
+      "step": 492
+    },
+    {
+      "epoch": 1.9972151898734176,
+      "grad_norm": 0.5588008264838369,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.5772,
+      "step": 493
+    },
+    {
+      "epoch": 2.0015189873417722,
+      "grad_norm": 0.7142109061370712,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.547,
+      "step": 494
+    },
+    {
+      "epoch": 2.0055696202531648,
+      "grad_norm": 0.5971778243167354,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.4335,
+      "step": 495
+    },
+    {
+      "epoch": 2.009620253164557,
+      "grad_norm": 1.0858711741321507,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.4607,
+      "step": 496
+    },
+    {
+      "epoch": 2.0136708860759494,
+      "grad_norm": 0.89346863876639,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.4635,
+      "step": 497
+    },
+    {
+      "epoch": 2.017721518987342,
+      "grad_norm": 0.7474366873719537,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3921,
+      "step": 498
+    },
+    {
+      "epoch": 2.0217721518987344,
+      "grad_norm": 0.8142073960057717,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.441,
+      "step": 499
+    },
+    {
+      "epoch": 2.0258227848101265,
+      "grad_norm": 0.7103985867494478,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.4124,
+      "step": 500
+    },
+    {
+      "epoch": 2.029873417721519,
+      "grad_norm": 0.6985428659913748,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.429,
+      "step": 501
+    },
+    {
+      "epoch": 2.0339240506329115,
+      "grad_norm": 0.7242246572253824,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.4562,
+      "step": 502
+    },
+    {
+      "epoch": 2.037974683544304,
+      "grad_norm": 0.6757313556469043,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.4225,
+      "step": 503
+    },
+    {
+      "epoch": 2.042025316455696,
+      "grad_norm": 0.7235850959150214,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.4187,
+      "step": 504
+    },
+    {
+      "epoch": 2.0460759493670886,
+      "grad_norm": 0.6765018777019174,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.4564,
+      "step": 505
+    },
+    {
+      "epoch": 2.050126582278481,
+      "grad_norm": 0.8803289454097694,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.4109,
+      "step": 506
+    },
+    {
+      "epoch": 2.0541772151898736,
+      "grad_norm": 0.6445947719593863,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.4295,
+      "step": 507
+    },
+    {
+      "epoch": 2.0582278481012657,
+      "grad_norm": 0.7926558531439836,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.4049,
+      "step": 508
+    },
+    {
+      "epoch": 2.062278481012658,
+      "grad_norm": 0.6923578634644288,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.398,
+      "step": 509
+    },
+    {
+      "epoch": 2.0663291139240507,
+      "grad_norm": 0.7916672752848571,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.4776,
+      "step": 510
+    },
+    {
+      "epoch": 2.0703797468354432,
+      "grad_norm": 0.762249511147647,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.4267,
+      "step": 511
+    },
+    {
+      "epoch": 2.0744303797468353,
+      "grad_norm": 0.6734172160623457,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.4031,
+      "step": 512
+    },
+    {
+      "epoch": 2.078481012658228,
+      "grad_norm": 0.666569069754466,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.426,
+      "step": 513
+    },
+    {
+      "epoch": 2.0825316455696203,
+      "grad_norm": 0.7318548727127212,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.4111,
+      "step": 514
+    },
+    {
+      "epoch": 2.086582278481013,
+      "grad_norm": 0.6531571069439703,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.4265,
+      "step": 515
+    },
+    {
+      "epoch": 2.090632911392405,
+      "grad_norm": 0.7370893744193006,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.4272,
+      "step": 516
+    },
+    {
+      "epoch": 2.0946835443037974,
+      "grad_norm": 0.5161302678630806,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3866,
+      "step": 517
+    },
+    {
+      "epoch": 2.09873417721519,
+      "grad_norm": 0.6188172550670268,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.4439,
+      "step": 518
+    },
+    {
+      "epoch": 2.1027848101265825,
+      "grad_norm": 0.5277330823290988,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.4479,
+      "step": 519
+    },
+    {
+      "epoch": 2.1068354430379745,
+      "grad_norm": 0.526826164303142,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.4297,
+      "step": 520
+    },
+    {
+      "epoch": 2.110886075949367,
+      "grad_norm": 0.5659587153434418,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.4094,
+      "step": 521
+    },
+    {
+      "epoch": 2.1149367088607596,
+      "grad_norm": 0.4776812632179298,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.4089,
+      "step": 522
+    },
+    {
+      "epoch": 2.118987341772152,
+      "grad_norm": 0.6206550439700692,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.4559,
+      "step": 523
+    },
+    {
+      "epoch": 2.123037974683544,
+      "grad_norm": 0.4823479689667362,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.417,
+      "step": 524
+    },
+    {
+      "epoch": 2.1270886075949367,
+      "grad_norm": 0.5714407009562481,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.4472,
+      "step": 525
+    },
+    {
+      "epoch": 2.131139240506329,
+      "grad_norm": 0.571463495193429,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3874,
+      "step": 526
+    },
+    {
+      "epoch": 2.1351898734177217,
+      "grad_norm": 0.6023320897030572,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.4463,
+      "step": 527
+    },
+    {
+      "epoch": 2.1392405063291138,
+      "grad_norm": 0.49290501699125205,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4256,
+      "step": 528
+    },
+    {
+      "epoch": 2.1432911392405063,
+      "grad_norm": 0.5852095470665444,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.4156,
+      "step": 529
+    },
+    {
+      "epoch": 2.147341772151899,
+      "grad_norm": 0.506198241345646,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.4307,
+      "step": 530
+    },
+    {
+      "epoch": 2.1513924050632913,
+      "grad_norm": 0.5068506337891088,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.441,
+      "step": 531
+    },
+    {
+      "epoch": 2.1554430379746834,
+      "grad_norm": 0.620293821013748,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.4195,
+      "step": 532
+    },
+    {
+      "epoch": 2.159493670886076,
+      "grad_norm": 0.48372368335928134,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.4291,
+      "step": 533
+    },
+    {
+      "epoch": 2.1635443037974684,
+      "grad_norm": 0.5971576552761663,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.4434,
+      "step": 534
+    },
+    {
+      "epoch": 2.167594936708861,
+      "grad_norm": 0.4124363523330016,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3556,
+      "step": 535
+    },
+    {
+      "epoch": 2.171645569620253,
+      "grad_norm": 0.7354336830833796,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.451,
+      "step": 536
+    },
+    {
+      "epoch": 2.1756962025316455,
+      "grad_norm": 0.4710450335947975,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.4245,
+      "step": 537
+    },
+    {
+      "epoch": 2.179746835443038,
+      "grad_norm": 0.6120176303324681,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.4264,
+      "step": 538
+    },
+    {
+      "epoch": 2.1837974683544306,
+      "grad_norm": 0.5163584324799755,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.4201,
+      "step": 539
+    },
+    {
+      "epoch": 2.1878481012658226,
+      "grad_norm": 0.4609868715747531,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.419,
+      "step": 540
+    },
+    {
+      "epoch": 2.191898734177215,
+      "grad_norm": 0.4648557032534407,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.4142,
+      "step": 541
+    },
+    {
+      "epoch": 2.1959493670886077,
+      "grad_norm": 0.5026750025968879,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.425,
+      "step": 542
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.5571685900250907,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.452,
+      "step": 543
+    },
+    {
+      "epoch": 2.2040506329113922,
+      "grad_norm": 0.48602034022959506,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.4307,
+      "step": 544
+    },
+    {
+      "epoch": 2.2081012658227848,
+      "grad_norm": 0.6387928780281421,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.4448,
+      "step": 545
+    },
+    {
+      "epoch": 2.2121518987341773,
+      "grad_norm": 0.4237518111640032,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.386,
+      "step": 546
+    },
+    {
+      "epoch": 2.21620253164557,
+      "grad_norm": 0.5736309940471025,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.4411,
+      "step": 547
+    },
+    {
+      "epoch": 2.220253164556962,
+      "grad_norm": 0.5235452022368923,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.4384,
+      "step": 548
+    },
+    {
+      "epoch": 2.2243037974683544,
+      "grad_norm": 0.442924183391874,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3976,
+      "step": 549
+    },
+    {
+      "epoch": 2.228354430379747,
+      "grad_norm": 0.5099239972272517,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.4179,
+      "step": 550
+    },
+    {
+      "epoch": 2.2324050632911394,
+      "grad_norm": 0.5052375545176062,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.4145,
+      "step": 551
+    },
+    {
+      "epoch": 2.2364556962025315,
+      "grad_norm": 0.5136145316142368,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.433,
+      "step": 552
+    },
+    {
+      "epoch": 2.240506329113924,
+      "grad_norm": 0.5188721338607946,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3991,
+      "step": 553
+    },
+    {
+      "epoch": 2.2445569620253165,
+      "grad_norm": 0.4929423824079526,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.4147,
+      "step": 554
+    },
+    {
+      "epoch": 2.248607594936709,
+      "grad_norm": 0.43869134529789583,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.453,
+      "step": 555
+    },
+    {
+      "epoch": 2.252658227848101,
+      "grad_norm": 0.5305562669135508,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.4201,
+      "step": 556
+    },
+    {
+      "epoch": 2.2567088607594936,
+      "grad_norm": 0.4291103077478918,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.4028,
+      "step": 557
+    },
+    {
+      "epoch": 2.260759493670886,
+      "grad_norm": 0.5496072446140529,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.4343,
+      "step": 558
+    },
+    {
+      "epoch": 2.2648101265822786,
+      "grad_norm": 0.5919851622624096,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.4441,
+      "step": 559
+    },
+    {
+      "epoch": 2.2688607594936707,
+      "grad_norm": 0.46002331496180143,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3791,
+      "step": 560
+    },
+    {
+      "epoch": 2.2729113924050632,
+      "grad_norm": 0.515820405101862,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.4657,
+      "step": 561
+    },
+    {
+      "epoch": 2.2769620253164558,
+      "grad_norm": 0.46693235972055935,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.4117,
+      "step": 562
+    },
+    {
+      "epoch": 2.2810126582278483,
+      "grad_norm": 0.4481069296209866,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.4014,
+      "step": 563
+    },
+    {
+      "epoch": 2.2850632911392403,
+      "grad_norm": 0.5103204234702331,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.4247,
+      "step": 564
+    },
+    {
+      "epoch": 2.289113924050633,
+      "grad_norm": 0.5191398562160927,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.4528,
+      "step": 565
+    },
+    {
+      "epoch": 2.2931645569620254,
+      "grad_norm": 0.5554501843420067,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.4448,
+      "step": 566
+    },
+    {
+      "epoch": 2.297215189873418,
+      "grad_norm": 0.4794067972691308,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.4175,
+      "step": 567
+    },
+    {
+      "epoch": 2.30126582278481,
+      "grad_norm": 0.5284770682154801,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.4028,
+      "step": 568
+    },
+    {
+      "epoch": 2.3053164556962025,
+      "grad_norm": 0.4576955902960689,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.423,
+      "step": 569
+    },
+    {
+      "epoch": 2.309367088607595,
+      "grad_norm": 0.6256445694633666,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.4063,
+      "step": 570
+    },
+    {
+      "epoch": 2.3134177215189875,
+      "grad_norm": 0.4527702438182646,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.4564,
+      "step": 571
+    },
+    {
+      "epoch": 2.3174683544303796,
+      "grad_norm": 0.5754409197138856,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.417,
+      "step": 572
+    },
+    {
+      "epoch": 2.321518987341772,
+      "grad_norm": 0.49702441585757606,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.4266,
+      "step": 573
+    },
+    {
+      "epoch": 2.3255696202531646,
+      "grad_norm": 0.5276407858371107,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.4179,
+      "step": 574
+    },
+    {
+      "epoch": 2.329620253164557,
+      "grad_norm": 0.5713634802683929,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.4089,
+      "step": 575
+    },
+    {
+      "epoch": 2.333670886075949,
+      "grad_norm": 0.4879887662635231,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.4528,
+      "step": 576
+    },
+    {
+      "epoch": 2.3377215189873417,
+      "grad_norm": 0.45883825825353874,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.4525,
+      "step": 577
+    },
+    {
+      "epoch": 2.3417721518987342,
+      "grad_norm": 0.5679453294461484,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.4064,
+      "step": 578
+    },
+    {
+      "epoch": 2.3458227848101267,
+      "grad_norm": 0.5050617920831095,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.4176,
+      "step": 579
+    },
+    {
+      "epoch": 2.349873417721519,
+      "grad_norm": 0.510755999919126,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.4203,
+      "step": 580
+    },
+    {
+      "epoch": 2.3539240506329113,
+      "grad_norm": 0.48253723890290046,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.4252,
+      "step": 581
+    },
+    {
+      "epoch": 2.357974683544304,
+      "grad_norm": 0.5314685702242034,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4394,
+      "step": 582
+    },
+    {
+      "epoch": 2.3620253164556964,
+      "grad_norm": 0.40695718955638666,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3726,
+      "step": 583
+    },
+    {
+      "epoch": 2.3660759493670884,
+      "grad_norm": 0.463752190236461,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.422,
+      "step": 584
+    },
+    {
+      "epoch": 2.370126582278481,
+      "grad_norm": 0.45216031924339467,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3975,
+      "step": 585
+    },
+    {
+      "epoch": 2.3741772151898735,
+      "grad_norm": 0.5167662798309056,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.475,
+      "step": 586
+    },
+    {
+      "epoch": 2.378227848101266,
+      "grad_norm": 0.5146550704974726,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.4438,
+      "step": 587
+    },
+    {
+      "epoch": 2.382278481012658,
+      "grad_norm": 0.5053675117234435,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.4122,
+      "step": 588
+    },
+    {
+      "epoch": 2.3863291139240506,
+      "grad_norm": 0.5175711074939241,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.4047,
+      "step": 589
+    },
+    {
+      "epoch": 2.390379746835443,
+      "grad_norm": 0.547897820146105,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.4383,
+      "step": 590
+    },
+    {
+      "epoch": 2.3944303797468356,
+      "grad_norm": 0.5484487791080884,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.426,
+      "step": 591
+    },
+    {
+      "epoch": 2.3984810126582277,
+      "grad_norm": 0.44364461317309306,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.4327,
+      "step": 592
+    },
+    {
+      "epoch": 2.40253164556962,
+      "grad_norm": 0.5135478723402688,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.4815,
+      "step": 593
+    },
+    {
+      "epoch": 2.4065822784810127,
+      "grad_norm": 0.38481456487375365,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.3865,
+      "step": 594
+    },
+    {
+      "epoch": 2.410632911392405,
+      "grad_norm": 0.4777893427885791,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.415,
+      "step": 595
+    },
+    {
+      "epoch": 2.4146835443037973,
+      "grad_norm": 0.5314096595686632,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.4543,
+      "step": 596
+    },
+    {
+      "epoch": 2.41873417721519,
+      "grad_norm": 0.41476186903595336,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.4089,
+      "step": 597
+    },
+    {
+      "epoch": 2.4227848101265823,
+      "grad_norm": 0.563867337576708,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.4291,
+      "step": 598
+    },
+    {
+      "epoch": 2.426835443037975,
+      "grad_norm": 0.46611160929941864,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.4224,
+      "step": 599
+    },
+    {
+      "epoch": 2.430886075949367,
+      "grad_norm": 0.5245990958557222,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.4184,
+      "step": 600
+    },
+    {
+      "epoch": 2.4349367088607594,
+      "grad_norm": 0.5218662026813128,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.4533,
+      "step": 601
+    },
+    {
+      "epoch": 2.438987341772152,
+      "grad_norm": 0.43294194093606214,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.4109,
+      "step": 602
+    },
+    {
+      "epoch": 2.4430379746835444,
+      "grad_norm": 0.5254308606399737,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.4385,
+      "step": 603
+    },
+    {
+      "epoch": 2.4470886075949365,
+      "grad_norm": 0.538540636307333,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.4416,
+      "step": 604
+    },
+    {
+      "epoch": 2.451139240506329,
+      "grad_norm": 0.6294670588310088,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.4083,
+      "step": 605
+    },
+    {
+      "epoch": 2.4551898734177215,
+      "grad_norm": 0.5503499155494198,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.4399,
+      "step": 606
+    },
+    {
+      "epoch": 2.459240506329114,
+      "grad_norm": 0.5976174071257218,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3977,
+      "step": 607
+    },
+    {
+      "epoch": 2.463291139240506,
+      "grad_norm": 0.5474029485588765,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.4586,
+      "step": 608
+    },
+    {
+      "epoch": 2.4673417721518986,
+      "grad_norm": 0.505896637112045,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4123,
+      "step": 609
+    },
+    {
+      "epoch": 2.471392405063291,
+      "grad_norm": 0.5299157803399326,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.4132,
+      "step": 610
+    },
+    {
+      "epoch": 2.4754430379746837,
+      "grad_norm": 0.5178584460148602,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.4547,
+      "step": 611
+    },
+    {
+      "epoch": 2.479493670886076,
+      "grad_norm": 0.45346216954914387,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.4125,
+      "step": 612
+    },
+    {
+      "epoch": 2.4835443037974683,
+      "grad_norm": 0.50329324502656,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.4664,
+      "step": 613
+    },
+    {
+      "epoch": 2.487594936708861,
+      "grad_norm": 0.478586565685544,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.4052,
+      "step": 614
+    },
+    {
+      "epoch": 2.4916455696202533,
+      "grad_norm": 0.5387775050728358,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.4401,
+      "step": 615
+    },
+    {
+      "epoch": 2.4956962025316454,
+      "grad_norm": 0.5005048245577457,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.4172,
+      "step": 616
+    },
+    {
+      "epoch": 2.499746835443038,
+      "grad_norm": 0.5229740264246426,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.4227,
+      "step": 617
+    },
+    {
+      "epoch": 2.5037974683544304,
+      "grad_norm": 0.6302167772320204,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.4447,
+      "step": 618
+    },
+    {
+      "epoch": 2.507848101265823,
+      "grad_norm": 0.5436111576599202,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.4296,
+      "step": 619
+    },
+    {
+      "epoch": 2.5118987341772154,
+      "grad_norm": 0.5476475608217171,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.4399,
+      "step": 620
+    },
+    {
+      "epoch": 2.5159493670886075,
+      "grad_norm": 0.5397928279106226,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.4329,
+      "step": 621
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.5160156913624833,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.4102,
+      "step": 622
+    },
+    {
+      "epoch": 2.5240506329113925,
+      "grad_norm": 0.5224245667818569,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.4424,
+      "step": 623
+    },
+    {
+      "epoch": 2.5281012658227846,
+      "grad_norm": 0.5915114445268438,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.4546,
+      "step": 624
+    },
+    {
+      "epoch": 2.532151898734177,
+      "grad_norm": 0.4915564057627853,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.4013,
+      "step": 625
+    },
+    {
+      "epoch": 2.5362025316455696,
+      "grad_norm": 0.49364306545983133,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.4164,
+      "step": 626
+    },
+    {
+      "epoch": 2.540253164556962,
+      "grad_norm": 0.6894660383449082,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.4588,
+      "step": 627
+    },
+    {
+      "epoch": 2.5443037974683547,
+      "grad_norm": 0.43836373241938154,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.4143,
+      "step": 628
+    },
+    {
+      "epoch": 2.5483544303797467,
+      "grad_norm": 0.6938059429711972,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.4428,
+      "step": 629
+    },
+    {
+      "epoch": 2.5524050632911393,
+      "grad_norm": 0.4645166808771536,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.427,
+      "step": 630
+    },
+    {
+      "epoch": 2.5564556962025318,
+      "grad_norm": 0.4948625857349155,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3948,
+      "step": 631
+    },
+    {
+      "epoch": 2.560506329113924,
+      "grad_norm": 0.5284966188277499,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.4463,
+      "step": 632
+    },
+    {
+      "epoch": 2.5645569620253164,
+      "grad_norm": 0.4374004123293572,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.4215,
+      "step": 633
+    },
+    {
+      "epoch": 2.568607594936709,
+      "grad_norm": 0.5418965532293888,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.426,
+      "step": 634
+    },
+    {
+      "epoch": 2.5726582278481014,
+      "grad_norm": 0.5048948405659673,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.4287,
+      "step": 635
+    },
+    {
+      "epoch": 2.576708860759494,
+      "grad_norm": 0.5561873751256334,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4308,
+      "step": 636
+    },
+    {
+      "epoch": 2.580759493670886,
+      "grad_norm": 0.5534074959689884,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.4361,
+      "step": 637
+    },
+    {
+      "epoch": 2.5848101265822785,
+      "grad_norm": 0.46835569965110785,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.4018,
+      "step": 638
+    },
+    {
+      "epoch": 2.588860759493671,
+      "grad_norm": 0.5929654329132523,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.4592,
+      "step": 639
+    },
+    {
+      "epoch": 2.592911392405063,
+      "grad_norm": 0.4811937995143609,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.4381,
+      "step": 640
+    },
+    {
+      "epoch": 2.5969620253164556,
+      "grad_norm": 0.49060233113020635,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.431,
+      "step": 641
+    },
+    {
+      "epoch": 2.601012658227848,
+      "grad_norm": 0.45777530609931627,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.4005,
+      "step": 642
+    },
+    {
+      "epoch": 2.6050632911392406,
+      "grad_norm": 0.4695746149222098,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.4425,
+      "step": 643
+    },
+    {
+      "epoch": 2.609113924050633,
+      "grad_norm": 0.4115787397126848,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.4201,
+      "step": 644
+    },
+    {
+      "epoch": 2.613164556962025,
+      "grad_norm": 0.5078763118274502,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.4069,
+      "step": 645
+    },
+    {
+      "epoch": 2.6172151898734177,
+      "grad_norm": 0.4273954417678827,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.4195,
+      "step": 646
+    },
+    {
+      "epoch": 2.6212658227848102,
+      "grad_norm": 0.44591613714345696,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.422,
+      "step": 647
+    },
+    {
+      "epoch": 2.6253164556962023,
+      "grad_norm": 0.4362523859939479,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.4694,
+      "step": 648
+    },
+    {
+      "epoch": 2.629367088607595,
+      "grad_norm": 0.5523928625232588,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.4125,
+      "step": 649
+    },
+    {
+      "epoch": 2.6334177215189873,
+      "grad_norm": 0.4262504859483176,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.4147,
+      "step": 650
+    },
+    {
+      "epoch": 2.63746835443038,
+      "grad_norm": 0.5246692545073564,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.4216,
+      "step": 651
+    },
+    {
+      "epoch": 2.6415189873417724,
+      "grad_norm": 0.4187169384117014,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.4254,
+      "step": 652
+    },
+    {
+      "epoch": 2.6455696202531644,
+      "grad_norm": 0.5520031985944907,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.413,
+      "step": 653
+    },
+    {
+      "epoch": 2.649620253164557,
+      "grad_norm": 0.4364832571112007,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.4911,
+      "step": 654
+    },
+    {
+      "epoch": 2.6536708860759495,
+      "grad_norm": 0.5837304367617978,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.4335,
+      "step": 655
+    },
+    {
+      "epoch": 2.6577215189873415,
+      "grad_norm": 0.42182818870440225,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3969,
+      "step": 656
+    },
+    {
+      "epoch": 2.661772151898734,
+      "grad_norm": 0.5223140262289049,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3929,
+      "step": 657
+    },
+    {
+      "epoch": 2.6658227848101266,
+      "grad_norm": 0.4263713527540944,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.418,
+      "step": 658
+    },
+    {
+      "epoch": 2.669873417721519,
+      "grad_norm": 0.50615674764645,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.4461,
+      "step": 659
+    },
+    {
+      "epoch": 2.6739240506329116,
+      "grad_norm": 0.46493055477756295,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3913,
+      "step": 660
+    },
+    {
+      "epoch": 2.6779746835443037,
+      "grad_norm": 0.6454122836252628,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.4771,
+      "step": 661
+    },
+    {
+      "epoch": 2.682025316455696,
+      "grad_norm": 0.44038115804161615,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.4447,
+      "step": 662
+    },
+    {
+      "epoch": 2.6860759493670887,
+      "grad_norm": 0.49278326324220073,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.3963,
+      "step": 663
+    },
+    {
+      "epoch": 2.690126582278481,
+      "grad_norm": 0.48054995367015046,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.4461,
+      "step": 664
+    },
+    {
+      "epoch": 2.6941772151898733,
+      "grad_norm": 0.46431890372480084,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.4214,
+      "step": 665
+    },
+    {
+      "epoch": 2.698227848101266,
+      "grad_norm": 0.4616051784241901,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.4365,
+      "step": 666
+    },
+    {
+      "epoch": 2.7022784810126583,
+      "grad_norm": 0.45363804098972893,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.4318,
+      "step": 667
+    },
+    {
+      "epoch": 2.706329113924051,
+      "grad_norm": 0.43506862238000593,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3989,
+      "step": 668
+    },
+    {
+      "epoch": 2.710379746835443,
+      "grad_norm": 0.44200371636145447,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.4181,
+      "step": 669
+    },
+    {
+      "epoch": 2.7144303797468354,
+      "grad_norm": 0.4109495179049366,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.415,
+      "step": 670
+    },
+    {
+      "epoch": 2.718481012658228,
+      "grad_norm": 0.40890495081279765,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.4393,
+      "step": 671
+    },
+    {
+      "epoch": 2.72253164556962,
+      "grad_norm": 0.4350327409499441,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.4195,
+      "step": 672
+    },
+    {
+      "epoch": 2.7265822784810125,
+      "grad_norm": 0.4205484729539491,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.4136,
+      "step": 673
+    },
+    {
+      "epoch": 2.730632911392405,
+      "grad_norm": 0.45423270266954274,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.4305,
+      "step": 674
+    },
+    {
+      "epoch": 2.7346835443037976,
+      "grad_norm": 0.5089851027866287,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.442,
+      "step": 675
+    },
+    {
+      "epoch": 2.73873417721519,
+      "grad_norm": 0.43368480244491125,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.4111,
+      "step": 676
+    },
+    {
+      "epoch": 2.742784810126582,
+      "grad_norm": 0.48041327277393836,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.4357,
+      "step": 677
+    },
+    {
+      "epoch": 2.7468354430379747,
+      "grad_norm": 0.4590330285510384,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.4044,
+      "step": 678
+    },
+    {
+      "epoch": 2.750886075949367,
+      "grad_norm": 0.5170465456530927,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.4382,
+      "step": 679
+    },
+    {
+      "epoch": 2.7549367088607593,
+      "grad_norm": 0.430985054945564,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.4207,
+      "step": 680
+    },
+    {
+      "epoch": 2.7589873417721518,
+      "grad_norm": 0.44900709883801937,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.4257,
+      "step": 681
+    },
+    {
+      "epoch": 2.7630379746835443,
+      "grad_norm": 0.4222542201533496,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.4302,
+      "step": 682
+    },
+    {
+      "epoch": 2.767088607594937,
+      "grad_norm": 0.4686751275774582,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.4411,
+      "step": 683
+    },
+    {
+      "epoch": 2.7711392405063293,
+      "grad_norm": 0.41717342228974236,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.4363,
+      "step": 684
+    },
+    {
+      "epoch": 2.7751898734177214,
+      "grad_norm": 0.44236149557329285,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.4263,
+      "step": 685
+    },
+    {
+      "epoch": 2.779240506329114,
+      "grad_norm": 0.44673292216211224,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.4333,
+      "step": 686
+    },
+    {
+      "epoch": 2.7832911392405064,
+      "grad_norm": 0.5201178283697823,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.4697,
+      "step": 687
+    },
+    {
+      "epoch": 2.7873417721518985,
+      "grad_norm": 0.41882997985859777,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3508,
+      "step": 688
+    },
+    {
+      "epoch": 2.791392405063291,
+      "grad_norm": 0.3943793784690471,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.4108,
+      "step": 689
+    },
+    {
+      "epoch": 2.7954430379746835,
+      "grad_norm": 0.5110511694167242,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4396,
+      "step": 690
+    },
+    {
+      "epoch": 2.799493670886076,
+      "grad_norm": 0.46838798967898804,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.429,
+      "step": 691
+    },
+    {
+      "epoch": 2.8035443037974686,
+      "grad_norm": 0.45694500079209743,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3628,
+      "step": 692
+    },
+    {
+      "epoch": 2.8075949367088606,
+      "grad_norm": 0.5353832657560298,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.5053,
+      "step": 693
+    },
+    {
+      "epoch": 2.811645569620253,
+      "grad_norm": 0.3818593107768754,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3723,
+      "step": 694
+    },
+    {
+      "epoch": 2.8156962025316457,
+      "grad_norm": 0.516333215692529,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.4507,
+      "step": 695
+    },
+    {
+      "epoch": 2.8197468354430377,
+      "grad_norm": 0.4566475555077562,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.447,
+      "step": 696
+    },
+    {
+      "epoch": 2.8237974683544302,
+      "grad_norm": 0.4220317964350597,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3643,
+      "step": 697
+    },
+    {
+      "epoch": 2.8278481012658228,
+      "grad_norm": 0.3991236463589033,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.4225,
+      "step": 698
+    },
+    {
+      "epoch": 2.8318987341772153,
+      "grad_norm": 0.48203555049805974,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.4452,
+      "step": 699
+    },
+    {
+      "epoch": 2.835949367088608,
+      "grad_norm": 0.3911662757767854,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.4129,
+      "step": 700
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.4599643238616371,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.4303,
+      "step": 701
+    },
+    {
+      "epoch": 2.8440506329113924,
+      "grad_norm": 0.4768862889654696,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.4509,
+      "step": 702
+    },
+    {
+      "epoch": 2.848101265822785,
+      "grad_norm": 0.4191040578074366,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.4104,
+      "step": 703
+    },
+    {
+      "epoch": 2.852151898734177,
+      "grad_norm": 0.43007727269554547,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.4249,
+      "step": 704
+    },
+    {
+      "epoch": 2.8562025316455695,
+      "grad_norm": 0.3936398909142753,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3869,
+      "step": 705
+    },
+    {
+      "epoch": 2.860253164556962,
+      "grad_norm": 0.46440397631152613,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.4482,
+      "step": 706
+    },
+    {
+      "epoch": 2.8643037974683545,
+      "grad_norm": 0.46841118573607593,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.4251,
+      "step": 707
+    },
+    {
+      "epoch": 2.868354430379747,
+      "grad_norm": 0.4786769601565617,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.4287,
+      "step": 708
+    },
+    {
+      "epoch": 2.872405063291139,
+      "grad_norm": 0.5083024534331989,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.4321,
+      "step": 709
+    },
+    {
+      "epoch": 2.8764556962025316,
+      "grad_norm": 0.39844698142372803,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3913,
+      "step": 710
+    },
+    {
+      "epoch": 2.880506329113924,
+      "grad_norm": 0.5110178760120426,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.4722,
+      "step": 711
+    },
+    {
+      "epoch": 2.884556962025316,
+      "grad_norm": 0.3740285037848518,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.4103,
+      "step": 712
+    },
+    {
+      "epoch": 2.8886075949367087,
+      "grad_norm": 0.3939554318050098,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.4194,
+      "step": 713
+    },
+    {
+      "epoch": 2.8926582278481012,
+      "grad_norm": 0.45997200521586756,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.4364,
+      "step": 714
+    },
+    {
+      "epoch": 2.8967088607594937,
+      "grad_norm": 0.4374193953402831,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.4346,
+      "step": 715
+    },
+    {
+      "epoch": 2.9007594936708863,
+      "grad_norm": 0.4289640134921434,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.4055,
+      "step": 716
+    },
+    {
+      "epoch": 2.9048101265822783,
+      "grad_norm": 0.4332326575462663,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4455,
+      "step": 717
+    },
+    {
+      "epoch": 2.908860759493671,
+      "grad_norm": 0.44525928426212763,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.4345,
+      "step": 718
+    },
+    {
+      "epoch": 2.9129113924050634,
+      "grad_norm": 0.4403732438043824,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.4412,
+      "step": 719
+    },
+    {
+      "epoch": 2.9169620253164554,
+      "grad_norm": 0.40606386814783024,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.4123,
+      "step": 720
+    },
+    {
+      "epoch": 2.921012658227848,
+      "grad_norm": 0.4778179138949842,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.4284,
+      "step": 721
+    },
+    {
+      "epoch": 2.9250632911392405,
+      "grad_norm": 0.44245927491376297,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.4465,
+      "step": 722
+    },
+    {
+      "epoch": 2.929113924050633,
+      "grad_norm": 0.4001547520208834,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3922,
+      "step": 723
+    },
+    {
+      "epoch": 2.9331645569620255,
+      "grad_norm": 0.43694988727088135,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.43,
+      "step": 724
+    },
+    {
+      "epoch": 2.9372151898734176,
+      "grad_norm": 0.45338070460280505,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.4413,
+      "step": 725
+    },
+    {
+      "epoch": 2.94126582278481,
+      "grad_norm": 0.38553607813193075,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.4221,
+      "step": 726
+    },
+    {
+      "epoch": 2.9453164556962026,
+      "grad_norm": 0.4318648928711469,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.4043,
+      "step": 727
+    },
+    {
+      "epoch": 2.9493670886075947,
+      "grad_norm": 0.45386906381116193,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.4452,
+      "step": 728
+    },
+    {
+      "epoch": 2.953417721518987,
+      "grad_norm": 0.4239722502822553,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.4506,
+      "step": 729
+    },
+    {
+      "epoch": 2.9574683544303797,
+      "grad_norm": 0.36884728762611646,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.415,
+      "step": 730
+    },
+    {
+      "epoch": 2.961518987341772,
+      "grad_norm": 0.4516428249567874,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.4218,
+      "step": 731
+    },
+    {
+      "epoch": 2.9655696202531647,
+      "grad_norm": 0.4362085473574307,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.406,
+      "step": 732
+    },
+    {
+      "epoch": 2.969620253164557,
+      "grad_norm": 0.41409491220705485,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.4281,
+      "step": 733
+    },
+    {
+      "epoch": 2.9736708860759493,
+      "grad_norm": 0.3920203791761346,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3778,
+      "step": 734
+    },
+    {
+      "epoch": 2.977721518987342,
+      "grad_norm": 0.47155656621277625,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.4488,
+      "step": 735
+    },
+    {
+      "epoch": 2.981772151898734,
+      "grad_norm": 0.4296912676968355,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.4273,
+      "step": 736
+    },
+    {
+      "epoch": 2.9858227848101264,
+      "grad_norm": 0.4480172463797977,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.4276,
+      "step": 737
+    },
+    {
+      "epoch": 2.989873417721519,
+      "grad_norm": 0.4337399689862748,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.4166,
+      "step": 738
+    },
+    {
+      "epoch": 2.9939240506329114,
+      "grad_norm": 0.4492701388484977,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.4232,
+      "step": 739
+    },
+    {
+      "epoch": 2.997974683544304,
+      "grad_norm": 0.5279432349461561,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.4507,
+      "step": 740
+    },
+    {
+      "epoch": 3.002278481012658,
+      "grad_norm": 0.5435228850813848,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.4137,
+      "step": 741
+    },
+    {
+      "epoch": 3.0063291139240507,
+      "grad_norm": 0.604414606583873,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.3232,
+      "step": 742
+    },
+    {
+      "epoch": 3.010379746835443,
+      "grad_norm": 0.48163801111220456,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.3124,
+      "step": 743
+    },
+    {
+      "epoch": 3.0144303797468353,
+      "grad_norm": 0.9851153042026485,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.3626,
+      "step": 744
+    },
+    {
+      "epoch": 3.0184810126582278,
+      "grad_norm": 0.5982138701774492,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2963,
+      "step": 745
+    },
+    {
+      "epoch": 3.0225316455696203,
+      "grad_norm": 0.5821064449073874,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2883,
+      "step": 746
+    },
+    {
+      "epoch": 3.026582278481013,
+      "grad_norm": 0.614472316780645,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.3449,
+      "step": 747
+    },
+    {
+      "epoch": 3.030632911392405,
+      "grad_norm": 0.5747630732909984,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.3046,
+      "step": 748
+    },
+    {
+      "epoch": 3.0346835443037974,
+      "grad_norm": 0.5544277401745573,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.3288,
+      "step": 749
+    },
+    {
+      "epoch": 3.03873417721519,
+      "grad_norm": 0.5258749435121917,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2985,
+      "step": 750
+    },
+    {
+      "epoch": 3.0427848101265824,
+      "grad_norm": 0.5496764025334598,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.3313,
+      "step": 751
+    },
+    {
+      "epoch": 3.0468354430379745,
+      "grad_norm": 0.5701222179147744,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.3414,
+      "step": 752
+    },
+    {
+      "epoch": 3.050886075949367,
+      "grad_norm": 0.5735600634097834,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.3385,
+      "step": 753
+    },
+    {
+      "epoch": 3.0549367088607595,
+      "grad_norm": 0.4739633390701903,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.3173,
+      "step": 754
+    },
+    {
+      "epoch": 3.058987341772152,
+      "grad_norm": 0.5191012310730974,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2903,
+      "step": 755
+    },
+    {
+      "epoch": 3.063037974683544,
+      "grad_norm": 0.4477710795677636,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.3032,
+      "step": 756
+    },
+    {
+      "epoch": 3.0670886075949366,
+      "grad_norm": 0.4656777586635711,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.302,
+      "step": 757
+    },
+    {
+      "epoch": 3.071139240506329,
+      "grad_norm": 0.4429864297109438,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2898,
+      "step": 758
+    },
+    {
+      "epoch": 3.0751898734177217,
+      "grad_norm": 0.42844195129136176,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.3098,
+      "step": 759
+    },
+    {
+      "epoch": 3.0792405063291137,
+      "grad_norm": 0.4635168330171718,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.3473,
+      "step": 760
+    },
+    {
+      "epoch": 3.0832911392405062,
+      "grad_norm": 0.4322689450631132,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.3136,
+      "step": 761
+    },
+    {
+      "epoch": 3.0873417721518988,
+      "grad_norm": 0.41787830266911113,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2859,
+      "step": 762
+    },
+    {
+      "epoch": 3.0913924050632913,
+      "grad_norm": 0.422515499496248,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.3295,
+      "step": 763
+    },
+    {
+      "epoch": 3.0954430379746833,
+      "grad_norm": 0.5221909278069066,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.3313,
+      "step": 764
+    },
+    {
+      "epoch": 3.099493670886076,
+      "grad_norm": 0.41427544720144466,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.3099,
+      "step": 765
+    },
+    {
+      "epoch": 3.1035443037974684,
+      "grad_norm": 0.4743981221066582,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2962,
+      "step": 766
+    },
+    {
+      "epoch": 3.107594936708861,
+      "grad_norm": 0.5494119848697887,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.3562,
+      "step": 767
+    },
+    {
+      "epoch": 3.111645569620253,
+      "grad_norm": 0.4153540928529232,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.3085,
+      "step": 768
+    },
+    {
+      "epoch": 3.1156962025316455,
+      "grad_norm": 0.559706929841413,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.3242,
+      "step": 769
+    },
+    {
+      "epoch": 3.119746835443038,
+      "grad_norm": 0.36634190684195883,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2768,
+      "step": 770
+    },
+    {
+      "epoch": 3.1237974683544305,
+      "grad_norm": 0.4879025480186631,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.3187,
+      "step": 771
+    },
+    {
+      "epoch": 3.1278481012658226,
+      "grad_norm": 0.41376143800475335,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.3039,
+      "step": 772
+    },
+    {
+      "epoch": 3.131898734177215,
+      "grad_norm": 0.35342875939225266,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2762,
+      "step": 773
+    },
+    {
+      "epoch": 3.1359493670886076,
+      "grad_norm": 0.4438032153245839,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.3485,
+      "step": 774
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.35241390747136775,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.3091,
+      "step": 775
+    },
+    {
+      "epoch": 3.1440506329113926,
+      "grad_norm": 0.4065875013137187,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.3075,
+      "step": 776
+    },
+    {
+      "epoch": 3.1481012658227847,
+      "grad_norm": 0.4110147622112397,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.3245,
+      "step": 777
+    },
+    {
+      "epoch": 3.1521518987341772,
+      "grad_norm": 0.3944143900112115,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.3105,
+      "step": 778
+    },
+    {
+      "epoch": 3.1562025316455697,
+      "grad_norm": 0.4109490086005844,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.3374,
+      "step": 779
+    },
+    {
+      "epoch": 3.160253164556962,
+      "grad_norm": 0.39422070971417655,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.2846,
+      "step": 780
+    },
+    {
+      "epoch": 3.1643037974683543,
+      "grad_norm": 0.45685276487464277,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.3729,
+      "step": 781
+    },
+    {
+      "epoch": 3.168354430379747,
+      "grad_norm": 0.39226580253722354,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.307,
+      "step": 782
+    },
+    {
+      "epoch": 3.1724050632911394,
+      "grad_norm": 0.36862089990628044,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2833,
+      "step": 783
+    },
+    {
+      "epoch": 3.176455696202532,
+      "grad_norm": 0.36724871314143875,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2937,
+      "step": 784
+    },
+    {
+      "epoch": 3.180506329113924,
+      "grad_norm": 0.4028911299417396,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.3311,
+      "step": 785
+    },
+    {
+      "epoch": 3.1845569620253165,
+      "grad_norm": 0.3752694837143601,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.275,
+      "step": 786
+    },
+    {
+      "epoch": 3.188607594936709,
+      "grad_norm": 0.45484344922502,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.3679,
+      "step": 787
+    },
+    {
+      "epoch": 3.192658227848101,
+      "grad_norm": 0.3996383400086033,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.3016,
+      "step": 788
+    },
+    {
+      "epoch": 3.1967088607594936,
+      "grad_norm": 0.353012788320879,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2824,
+      "step": 789
+    },
+    {
+      "epoch": 3.200759493670886,
+      "grad_norm": 0.5240495935767944,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.3503,
+      "step": 790
+    },
+    {
+      "epoch": 3.2048101265822786,
+      "grad_norm": 0.41257741995548286,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.314,
+      "step": 791
+    },
+    {
+      "epoch": 3.208860759493671,
+      "grad_norm": 0.45252183755387493,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.3111,
+      "step": 792
+    },
+    {
+      "epoch": 3.212911392405063,
+      "grad_norm": 0.46874710378796675,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2772,
+      "step": 793
+    },
+    {
+      "epoch": 3.2169620253164557,
+      "grad_norm": 0.40497819652850203,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.3201,
+      "step": 794
+    },
+    {
+      "epoch": 3.221012658227848,
+      "grad_norm": 0.47615234825371283,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.3005,
+      "step": 795
+    },
+    {
+      "epoch": 3.2250632911392403,
+      "grad_norm": 0.4721333297752663,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.3455,
+      "step": 796
+    },
+    {
+      "epoch": 3.229113924050633,
+      "grad_norm": 0.40553171925494313,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.3053,
+      "step": 797
+    },
+    {
+      "epoch": 3.2331645569620253,
+      "grad_norm": 0.4306875564715964,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3099,
+      "step": 798
+    },
+    {
+      "epoch": 3.237215189873418,
+      "grad_norm": 0.3943308454107528,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.3265,
+      "step": 799
+    },
+    {
+      "epoch": 3.2412658227848103,
+      "grad_norm": 0.41881553475907457,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2826,
+      "step": 800
+    },
+    {
+      "epoch": 3.2453164556962024,
+      "grad_norm": 0.44267610267618235,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.3351,
+      "step": 801
+    },
+    {
+      "epoch": 3.249367088607595,
+      "grad_norm": 0.3506535452045794,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2785,
+      "step": 802
+    },
+    {
+      "epoch": 3.2534177215189874,
+      "grad_norm": 0.43581929890756543,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.3382,
+      "step": 803
+    },
+    {
+      "epoch": 3.2574683544303795,
+      "grad_norm": 0.37154687117216706,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.3116,
+      "step": 804
+    },
+    {
+      "epoch": 3.261518987341772,
+      "grad_norm": 0.3430543423212978,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2939,
+      "step": 805
+    },
+    {
+      "epoch": 3.2655696202531646,
+      "grad_norm": 0.3843894861392431,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.3218,
+      "step": 806
+    },
+    {
+      "epoch": 3.269620253164557,
+      "grad_norm": 0.40872920198922646,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.343,
+      "step": 807
+    },
+    {
+      "epoch": 3.2736708860759496,
+      "grad_norm": 0.3703859728727641,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.3153,
+      "step": 808
+    },
+    {
+      "epoch": 3.2777215189873417,
+      "grad_norm": 0.38962731108117654,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2894,
+      "step": 809
+    },
+    {
+      "epoch": 3.281772151898734,
+      "grad_norm": 0.38742007118472005,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.3228,
+      "step": 810
+    },
+    {
+      "epoch": 3.2858227848101267,
+      "grad_norm": 0.42676293577047186,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.3121,
+      "step": 811
+    },
+    {
+      "epoch": 3.2898734177215188,
+      "grad_norm": 0.4631551480685098,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.3385,
+      "step": 812
+    },
+    {
+      "epoch": 3.2939240506329113,
+      "grad_norm": 0.3530408609142979,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2843,
+      "step": 813
+    },
+    {
+      "epoch": 3.297974683544304,
+      "grad_norm": 0.4306928763529447,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.3054,
+      "step": 814
+    },
+    {
+      "epoch": 3.3020253164556963,
+      "grad_norm": 0.42125717438572935,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.3156,
+      "step": 815
+    },
+    {
+      "epoch": 3.306075949367089,
+      "grad_norm": 0.38672961769253583,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.3139,
+      "step": 816
+    },
+    {
+      "epoch": 3.310126582278481,
+      "grad_norm": 0.4874364841461359,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.3387,
+      "step": 817
+    },
+    {
+      "epoch": 3.3141772151898734,
+      "grad_norm": 0.34802671044669065,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2925,
+      "step": 818
+    },
+    {
+      "epoch": 3.318227848101266,
+      "grad_norm": 0.4471969470085418,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.3293,
+      "step": 819
+    },
+    {
+      "epoch": 3.322278481012658,
+      "grad_norm": 0.40145283221051326,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.3275,
+      "step": 820
+    },
+    {
+      "epoch": 3.3263291139240505,
+      "grad_norm": 0.33572566019107836,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.3028,
+      "step": 821
+    },
+    {
+      "epoch": 3.330379746835443,
+      "grad_norm": 0.41685112647713757,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.3321,
+      "step": 822
+    },
+    {
+      "epoch": 3.3344303797468355,
+      "grad_norm": 0.36536020070251285,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.3133,
+      "step": 823
+    },
+    {
+      "epoch": 3.338481012658228,
+      "grad_norm": 0.362118971664496,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.309,
+      "step": 824
+    },
+    {
+      "epoch": 3.34253164556962,
+      "grad_norm": 0.43366573402606745,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3479,
+      "step": 825
+    },
+    {
+      "epoch": 3.3465822784810126,
+      "grad_norm": 0.36743147794730624,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.3159,
+      "step": 826
+    },
+    {
+      "epoch": 3.350632911392405,
+      "grad_norm": 0.3981477096902416,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.2813,
+      "step": 827
+    },
+    {
+      "epoch": 3.3546835443037972,
+      "grad_norm": 0.37006047587597124,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.317,
+      "step": 828
+    },
+    {
+      "epoch": 3.3587341772151897,
+      "grad_norm": 0.39765301722274793,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.3247,
+      "step": 829
+    },
+    {
+      "epoch": 3.3627848101265823,
+      "grad_norm": 0.3831017490315007,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.327,
+      "step": 830
+    },
+    {
+      "epoch": 3.3668354430379748,
+      "grad_norm": 0.39629617902177805,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.3142,
+      "step": 831
+    },
+    {
+      "epoch": 3.3708860759493673,
+      "grad_norm": 0.35314845223528935,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.302,
+      "step": 832
+    },
+    {
+      "epoch": 3.3749367088607594,
+      "grad_norm": 0.4043235655209997,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.3306,
+      "step": 833
+    },
+    {
+      "epoch": 3.378987341772152,
+      "grad_norm": 0.3670605369435142,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.3127,
+      "step": 834
+    },
+    {
+      "epoch": 3.3830379746835444,
+      "grad_norm": 0.4164210139476333,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.3522,
+      "step": 835
+    },
+    {
+      "epoch": 3.3870886075949365,
+      "grad_norm": 0.35487967151598554,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2827,
+      "step": 836
+    },
+    {
+      "epoch": 3.391139240506329,
+      "grad_norm": 0.37753239860578636,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.324,
+      "step": 837
+    },
+    {
+      "epoch": 3.3951898734177215,
+      "grad_norm": 0.37260895301908614,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.3073,
+      "step": 838
+    },
+    {
+      "epoch": 3.399240506329114,
+      "grad_norm": 0.41335809618853975,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.3078,
+      "step": 839
+    },
+    {
+      "epoch": 3.4032911392405065,
+      "grad_norm": 0.3953321969776827,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.3564,
+      "step": 840
+    },
+    {
+      "epoch": 3.4073417721518986,
+      "grad_norm": 0.3623353364640113,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2744,
+      "step": 841
+    },
+    {
+      "epoch": 3.411392405063291,
+      "grad_norm": 0.40110626009611683,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.3266,
+      "step": 842
+    },
+    {
+      "epoch": 3.4154430379746836,
+      "grad_norm": 0.36797986497696783,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2984,
+      "step": 843
+    },
+    {
+      "epoch": 3.419493670886076,
+      "grad_norm": 0.3897141737167115,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.3103,
+      "step": 844
+    },
+    {
+      "epoch": 3.423544303797468,
+      "grad_norm": 0.3773502712546857,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.3167,
+      "step": 845
+    },
+    {
+      "epoch": 3.4275949367088607,
+      "grad_norm": 0.37284197823784093,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.3288,
+      "step": 846
+    },
+    {
+      "epoch": 3.4316455696202532,
+      "grad_norm": 0.39583666436785653,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2982,
+      "step": 847
+    },
+    {
+      "epoch": 3.4356962025316458,
+      "grad_norm": 0.37812322066277687,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2779,
+      "step": 848
+    },
+    {
+      "epoch": 3.439746835443038,
+      "grad_norm": 0.4175768986539499,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.3384,
+      "step": 849
+    },
+    {
+      "epoch": 3.4437974683544303,
+      "grad_norm": 0.36434646102486656,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2964,
+      "step": 850
+    },
+    {
+      "epoch": 3.447848101265823,
+      "grad_norm": 0.37278923534295577,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.3284,
+      "step": 851
+    },
+    {
+      "epoch": 3.4518987341772154,
+      "grad_norm": 0.3720135351582702,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.3069,
+      "step": 852
+    },
+    {
+      "epoch": 3.4559493670886074,
+      "grad_norm": 0.33226935443853306,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.2929,
+      "step": 853
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.3612729426716603,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.3229,
+      "step": 854
+    },
+    {
+      "epoch": 3.4640506329113925,
+      "grad_norm": 0.3989475476963846,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.3257,
+      "step": 855
+    },
+    {
+      "epoch": 3.468101265822785,
+      "grad_norm": 0.34102247885466036,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2895,
+      "step": 856
+    },
+    {
+      "epoch": 3.472151898734177,
+      "grad_norm": 0.38596000775074185,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.3343,
+      "step": 857
+    },
+    {
+      "epoch": 3.4762025316455696,
+      "grad_norm": 0.3468688240776255,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2893,
+      "step": 858
+    },
+    {
+      "epoch": 3.480253164556962,
+      "grad_norm": 0.35977254087850075,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.3097,
+      "step": 859
+    },
+    {
+      "epoch": 3.4843037974683546,
+      "grad_norm": 0.36956624968646595,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.3041,
+      "step": 860
+    },
+    {
+      "epoch": 3.4883544303797467,
+      "grad_norm": 0.42250224017958676,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.3396,
+      "step": 861
+    },
+    {
+      "epoch": 3.492405063291139,
+      "grad_norm": 0.32913679660059836,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2966,
+      "step": 862
+    },
+    {
+      "epoch": 3.4964556962025317,
+      "grad_norm": 0.4083367860679121,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.3268,
+      "step": 863
+    },
+    {
+      "epoch": 3.5005063291139242,
+      "grad_norm": 0.4017224425650291,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.3133,
+      "step": 864
+    },
+    {
+      "epoch": 3.5045569620253163,
+      "grad_norm": 0.38760982866095317,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.3331,
+      "step": 865
+    },
+    {
+      "epoch": 3.508607594936709,
+      "grad_norm": 0.3413876708230215,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2936,
+      "step": 866
+    },
+    {
+      "epoch": 3.5126582278481013,
+      "grad_norm": 0.3770148775037438,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.3401,
+      "step": 867
+    },
+    {
+      "epoch": 3.5167088607594934,
+      "grad_norm": 0.37355081538418694,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.3157,
+      "step": 868
+    },
+    {
+      "epoch": 3.520759493670886,
+      "grad_norm": 0.36022453790279296,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.3102,
+      "step": 869
+    },
+    {
+      "epoch": 3.5248101265822784,
+      "grad_norm": 0.3846289424593832,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.3397,
+      "step": 870
+    },
+    {
+      "epoch": 3.528860759493671,
+      "grad_norm": 0.3471139541017973,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.3077,
+      "step": 871
+    },
+    {
+      "epoch": 3.5329113924050635,
+      "grad_norm": 0.37936388129542836,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.3097,
+      "step": 872
+    },
+    {
+      "epoch": 3.5369620253164555,
+      "grad_norm": 0.419304396446441,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.3256,
+      "step": 873
+    },
+    {
+      "epoch": 3.541012658227848,
+      "grad_norm": 0.36961429300197923,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.3214,
+      "step": 874
+    },
+    {
+      "epoch": 3.5450632911392406,
+      "grad_norm": 0.37452798196218196,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.3062,
+      "step": 875
+    },
+    {
+      "epoch": 3.5491139240506326,
+      "grad_norm": 0.3847151434752731,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.3306,
+      "step": 876
+    },
+    {
+      "epoch": 3.553164556962025,
+      "grad_norm": 0.3722974081286379,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.3188,
+      "step": 877
+    },
+    {
+      "epoch": 3.5572151898734177,
+      "grad_norm": 0.38793903265357,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.3161,
+      "step": 878
+    },
+    {
+      "epoch": 3.56126582278481,
+      "grad_norm": 0.38692851469718026,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.3275,
+      "step": 879
+    },
+    {
+      "epoch": 3.5653164556962027,
+      "grad_norm": 0.3380821525730518,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.2895,
+      "step": 880
+    },
+    {
+      "epoch": 3.5693670886075948,
+      "grad_norm": 0.394868362243374,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.3414,
+      "step": 881
+    },
+    {
+      "epoch": 3.5734177215189873,
+      "grad_norm": 0.36875011153359694,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.311,
+      "step": 882
+    },
+    {
+      "epoch": 3.57746835443038,
+      "grad_norm": 0.3621645591121142,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.3019,
+      "step": 883
+    },
+    {
+      "epoch": 3.581518987341772,
+      "grad_norm": 0.3438886690677529,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2957,
+      "step": 884
+    },
+    {
+      "epoch": 3.5855696202531644,
+      "grad_norm": 0.39941548934287285,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.3595,
+      "step": 885
+    },
+    {
+      "epoch": 3.589620253164557,
+      "grad_norm": 0.3455785906620442,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.3056,
+      "step": 886
+    },
+    {
+      "epoch": 3.5936708860759494,
+      "grad_norm": 0.3461373476211002,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.3002,
+      "step": 887
+    },
+    {
+      "epoch": 3.597721518987342,
+      "grad_norm": 0.39761643110749745,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.3578,
+      "step": 888
+    },
+    {
+      "epoch": 3.601772151898734,
+      "grad_norm": 0.3685788036110618,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.3266,
+      "step": 889
+    },
+    {
+      "epoch": 3.6058227848101265,
+      "grad_norm": 0.3282117732505965,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2739,
+      "step": 890
+    },
+    {
+      "epoch": 3.609873417721519,
+      "grad_norm": 0.3518373996703747,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.3208,
+      "step": 891
+    },
+    {
+      "epoch": 3.6139240506329116,
+      "grad_norm": 0.3796651623613062,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.3612,
+      "step": 892
+    },
+    {
+      "epoch": 3.617974683544304,
+      "grad_norm": 0.34894299136907087,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.3089,
+      "step": 893
+    },
+    {
+      "epoch": 3.622025316455696,
+      "grad_norm": 0.34465673844835265,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.3029,
+      "step": 894
+    },
+    {
+      "epoch": 3.6260759493670887,
+      "grad_norm": 0.3469102415865181,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.3099,
+      "step": 895
+    },
+    {
+      "epoch": 3.630126582278481,
+      "grad_norm": 0.3367891855843339,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.3011,
+      "step": 896
+    },
+    {
+      "epoch": 3.6341772151898732,
+      "grad_norm": 0.364264016562065,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.3051,
+      "step": 897
+    },
+    {
+      "epoch": 3.6382278481012658,
+      "grad_norm": 0.35443956623320216,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.3169,
+      "step": 898
+    },
+    {
+      "epoch": 3.6422784810126583,
+      "grad_norm": 0.38299149940593485,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.3245,
+      "step": 899
+    },
+    {
+      "epoch": 3.646329113924051,
+      "grad_norm": 0.371103548429182,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.3051,
+      "step": 900
+    },
+    {
+      "epoch": 3.6503797468354433,
+      "grad_norm": 0.3607474318870085,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.3236,
+      "step": 901
+    },
+    {
+      "epoch": 3.6544303797468354,
+      "grad_norm": 0.34945818896197534,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.3272,
+      "step": 902
+    },
+    {
+      "epoch": 3.658481012658228,
+      "grad_norm": 0.41038566664586423,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.3383,
+      "step": 903
+    },
+    {
+      "epoch": 3.6625316455696204,
+      "grad_norm": 0.34240718675119697,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.2915,
+      "step": 904
+    },
+    {
+      "epoch": 3.6665822784810125,
+      "grad_norm": 0.35409653136469155,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.3013,
+      "step": 905
+    },
+    {
+      "epoch": 3.670632911392405,
+      "grad_norm": 0.3512799218823998,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3242,
+      "step": 906
+    },
+    {
+      "epoch": 3.6746835443037975,
+      "grad_norm": 0.3312592092347123,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2887,
+      "step": 907
+    },
+    {
+      "epoch": 3.67873417721519,
+      "grad_norm": 0.3954496832819805,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.3565,
+      "step": 908
+    },
+    {
+      "epoch": 3.6827848101265825,
+      "grad_norm": 0.3530958531820086,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.3225,
+      "step": 909
+    },
+    {
+      "epoch": 3.6868354430379746,
+      "grad_norm": 0.3436895964655284,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.298,
+      "step": 910
+    },
+    {
+      "epoch": 3.690886075949367,
+      "grad_norm": 0.3790292180460926,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.3261,
+      "step": 911
+    },
+    {
+      "epoch": 3.6949367088607596,
+      "grad_norm": 0.342491723876838,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.3043,
+      "step": 912
+    },
+    {
+      "epoch": 3.6989873417721517,
+      "grad_norm": 0.3830978931887183,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.3387,
+      "step": 913
+    },
+    {
+      "epoch": 3.7030379746835442,
+      "grad_norm": 0.3626569000928498,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.3145,
+      "step": 914
+    },
+    {
+      "epoch": 3.7070886075949367,
+      "grad_norm": 0.37271934417394925,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.285,
+      "step": 915
+    },
+    {
+      "epoch": 3.7111392405063293,
+      "grad_norm": 0.3740012578875894,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.3216,
+      "step": 916
+    },
+    {
+      "epoch": 3.7151898734177218,
+      "grad_norm": 0.3386650654219099,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.3146,
+      "step": 917
+    },
+    {
+      "epoch": 3.719240506329114,
+      "grad_norm": 0.35936083847618155,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.3098,
+      "step": 918
+    },
+    {
+      "epoch": 3.7232911392405064,
+      "grad_norm": 0.3578340254813407,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.3468,
+      "step": 919
+    },
+    {
+      "epoch": 3.727341772151899,
+      "grad_norm": 0.3418403006499396,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2897,
+      "step": 920
+    },
+    {
+      "epoch": 3.731392405063291,
+      "grad_norm": 0.39424186149549156,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.353,
+      "step": 921
+    },
+    {
+      "epoch": 3.7354430379746835,
+      "grad_norm": 0.3637692557930696,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.2909,
+      "step": 922
+    },
+    {
+      "epoch": 3.739493670886076,
+      "grad_norm": 0.36127877073041004,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2857,
+      "step": 923
+    },
+    {
+      "epoch": 3.7435443037974685,
+      "grad_norm": 0.4157545299042238,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.3277,
+      "step": 924
+    },
+    {
+      "epoch": 3.747594936708861,
+      "grad_norm": 0.39605144590681857,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.3571,
+      "step": 925
+    },
+    {
+      "epoch": 3.751645569620253,
+      "grad_norm": 0.34424741922132124,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2908,
+      "step": 926
+    },
+    {
+      "epoch": 3.7556962025316456,
+      "grad_norm": 0.41117987079207474,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.3226,
+      "step": 927
+    },
+    {
+      "epoch": 3.759746835443038,
+      "grad_norm": 0.37085192498579583,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2981,
+      "step": 928
+    },
+    {
+      "epoch": 3.76379746835443,
+      "grad_norm": 0.3540674754528945,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.3074,
+      "step": 929
+    },
+    {
+      "epoch": 3.7678481012658227,
+      "grad_norm": 0.3426017479203356,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.3,
+      "step": 930
+    },
+    {
+      "epoch": 3.771898734177215,
+      "grad_norm": 0.38162057930726984,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.3133,
+      "step": 931
+    },
+    {
+      "epoch": 3.7759493670886077,
+      "grad_norm": 0.3543189363499177,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2908,
+      "step": 932
+    },
+    {
+      "epoch": 3.7800000000000002,
+      "grad_norm": 0.40598561510851805,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3437,
+      "step": 933
+    },
+    {
+      "epoch": 3.7840506329113923,
+      "grad_norm": 0.33616958876454023,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2704,
+      "step": 934
+    },
+    {
+      "epoch": 3.788101265822785,
+      "grad_norm": 0.36691573032945707,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.3066,
+      "step": 935
+    },
+    {
+      "epoch": 3.7921518987341774,
+      "grad_norm": 0.3513386163169949,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2893,
+      "step": 936
+    },
+    {
+      "epoch": 3.7962025316455694,
+      "grad_norm": 0.3985552312103252,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.3565,
+      "step": 937
+    },
+    {
+      "epoch": 3.800253164556962,
+      "grad_norm": 0.3564688826137191,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.3132,
+      "step": 938
+    },
+    {
+      "epoch": 3.8043037974683545,
+      "grad_norm": 0.3350339944270921,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2896,
+      "step": 939
+    },
+    {
+      "epoch": 3.808354430379747,
+      "grad_norm": 0.3521375340572988,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.3393,
+      "step": 940
+    },
+    {
+      "epoch": 3.8124050632911395,
+      "grad_norm": 0.30918826626239604,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2464,
+      "step": 941
+    },
+    {
+      "epoch": 3.8164556962025316,
+      "grad_norm": 0.37195050749953695,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.3427,
+      "step": 942
+    },
+    {
+      "epoch": 3.820506329113924,
+      "grad_norm": 0.34129797025188296,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.3057,
+      "step": 943
+    },
+    {
+      "epoch": 3.8245569620253166,
+      "grad_norm": 0.344990402749119,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.3204,
+      "step": 944
+    },
+    {
+      "epoch": 3.8286075949367087,
+      "grad_norm": 0.34909152674242444,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.3068,
+      "step": 945
+    },
+    {
+      "epoch": 3.832658227848101,
+      "grad_norm": 0.33182979740067925,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.3099,
+      "step": 946
+    },
+    {
+      "epoch": 3.8367088607594937,
+      "grad_norm": 0.34373207425147884,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.3082,
+      "step": 947
+    },
+    {
+      "epoch": 3.840759493670886,
+      "grad_norm": 0.3344720089303944,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.3014,
+      "step": 948
+    },
+    {
+      "epoch": 3.8448101265822787,
+      "grad_norm": 0.34809347145891784,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.3233,
+      "step": 949
+    },
+    {
+      "epoch": 3.848860759493671,
+      "grad_norm": 0.3319320962763181,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.315,
+      "step": 950
+    },
+    {
+      "epoch": 3.8529113924050633,
+      "grad_norm": 0.3355841209997152,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2897,
+      "step": 951
+    },
+    {
+      "epoch": 3.856962025316456,
+      "grad_norm": 0.35980161312685716,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.3239,
+      "step": 952
+    },
+    {
+      "epoch": 3.861012658227848,
+      "grad_norm": 0.334301034379734,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.3008,
+      "step": 953
+    },
+    {
+      "epoch": 3.8650632911392404,
+      "grad_norm": 0.6453249542893541,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.343,
+      "step": 954
+    },
+    {
+      "epoch": 3.869113924050633,
+      "grad_norm": 0.34904240383566104,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.3149,
+      "step": 955
+    },
+    {
+      "epoch": 3.8731645569620254,
+      "grad_norm": 0.3729049057541835,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.3114,
+      "step": 956
+    },
+    {
+      "epoch": 3.877215189873418,
+      "grad_norm": 0.3538326016458772,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.3093,
+      "step": 957
+    },
+    {
+      "epoch": 3.88126582278481,
+      "grad_norm": 0.3417131959566994,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.3099,
+      "step": 958
+    },
+    {
+      "epoch": 3.8853164556962025,
+      "grad_norm": 0.3374388597297284,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2956,
+      "step": 959
+    },
+    {
+      "epoch": 3.889367088607595,
+      "grad_norm": 0.33799814114938437,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.315,
+      "step": 960
+    },
+    {
+      "epoch": 3.893417721518987,
+      "grad_norm": 0.3507166255518697,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.3083,
+      "step": 961
+    },
+    {
+      "epoch": 3.8974683544303796,
+      "grad_norm": 0.33127588202804703,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.3345,
+      "step": 962
+    },
+    {
+      "epoch": 3.901518987341772,
+      "grad_norm": 0.3322391280102111,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.3273,
+      "step": 963
+    },
+    {
+      "epoch": 3.9055696202531647,
+      "grad_norm": 0.31519306495645705,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2855,
+      "step": 964
+    },
+    {
+      "epoch": 3.909620253164557,
+      "grad_norm": 0.34257308482421905,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.3295,
+      "step": 965
+    },
+    {
+      "epoch": 3.9136708860759493,
+      "grad_norm": 0.32185405342675394,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2998,
+      "step": 966
+    },
+    {
+      "epoch": 3.9177215189873418,
+      "grad_norm": 0.33510601474334656,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.3314,
+      "step": 967
+    },
+    {
+      "epoch": 3.9217721518987343,
+      "grad_norm": 0.32477076542704353,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.3116,
+      "step": 968
+    },
+    {
+      "epoch": 3.9258227848101264,
+      "grad_norm": 0.3245634444986896,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.2976,
+      "step": 969
+    },
+    {
+      "epoch": 3.929873417721519,
+      "grad_norm": 0.3739114786902045,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.3138,
+      "step": 970
+    },
+    {
+      "epoch": 3.9339240506329114,
+      "grad_norm": 0.3601634194345993,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.3148,
+      "step": 971
+    },
+    {
+      "epoch": 3.937974683544304,
+      "grad_norm": 0.33760906751704245,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.314,
+      "step": 972
+    },
+    {
+      "epoch": 3.9420253164556964,
+      "grad_norm": 0.3508393726943037,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.3123,
+      "step": 973
+    },
+    {
+      "epoch": 3.9460759493670885,
+      "grad_norm": 0.3419178379613408,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2975,
+      "step": 974
+    },
+    {
+      "epoch": 3.950126582278481,
+      "grad_norm": 0.35610405428408864,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.3468,
+      "step": 975
+    },
+    {
+      "epoch": 3.9541772151898735,
+      "grad_norm": 0.3545144265467179,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.3191,
+      "step": 976
+    },
+    {
+      "epoch": 3.9582278481012656,
+      "grad_norm": 0.3200277980737422,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.3151,
+      "step": 977
+    },
+    {
+      "epoch": 3.962278481012658,
+      "grad_norm": 0.35718405266882414,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.3109,
+      "step": 978
+    },
+    {
+      "epoch": 3.9663291139240506,
+      "grad_norm": 0.35187490686973494,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2958,
+      "step": 979
+    },
+    {
+      "epoch": 3.970379746835443,
+      "grad_norm": 0.3394591805475553,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.3237,
+      "step": 980
+    },
+    {
+      "epoch": 3.9744303797468357,
+      "grad_norm": 0.30419665440408017,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.2795,
+      "step": 981
+    },
+    {
+      "epoch": 3.9784810126582277,
+      "grad_norm": 0.35306253697358303,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.3257,
+      "step": 982
+    },
+    {
+      "epoch": 3.9825316455696202,
+      "grad_norm": 0.35905718018793054,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.3039,
+      "step": 983
+    },
+    {
+      "epoch": 3.9865822784810128,
+      "grad_norm": 0.34127567596716135,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.3028,
+      "step": 984
+    },
+    {
+      "epoch": 3.990632911392405,
+      "grad_norm": 0.3152574782961082,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.3237,
+      "step": 985
+    },
+    {
+      "epoch": 3.9946835443037974,
+      "grad_norm": 0.3168410799359329,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2925,
+      "step": 986
+    },
+    {
+      "epoch": 3.99873417721519,
+      "grad_norm": 0.42544466368381784,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3523,
+      "step": 987
+    },
+    {
+      "epoch": 4.0030379746835445,
+      "grad_norm": 0.5079446472917329,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.3011,
+      "step": 988
+    },
+    {
+      "epoch": 4.007088607594937,
+      "grad_norm": 0.48591735550267373,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.1981,
+      "step": 989
+    },
+    {
+      "epoch": 4.0111392405063295,
+      "grad_norm": 0.46848555995876806,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.2521,
+      "step": 990
+    },
+    {
+      "epoch": 4.015189873417722,
+      "grad_norm": 0.35405827792303946,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2355,
+      "step": 991
+    },
+    {
+      "epoch": 4.019240506329114,
+      "grad_norm": 0.44912507152009523,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2417,
+      "step": 992
+    },
+    {
+      "epoch": 4.023291139240507,
+      "grad_norm": 0.6557946297419023,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2745,
+      "step": 993
+    },
+    {
+      "epoch": 4.027341772151899,
+      "grad_norm": 0.5521002909442486,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2273,
+      "step": 994
+    },
+    {
+      "epoch": 4.031392405063291,
+      "grad_norm": 0.43552665728868706,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.2139,
+      "step": 995
+    },
+    {
+      "epoch": 4.035443037974684,
+      "grad_norm": 0.3978534096225862,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2426,
+      "step": 996
+    },
+    {
+      "epoch": 4.039493670886076,
+      "grad_norm": 0.4073483653891952,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2226,
+      "step": 997
+    },
+    {
+      "epoch": 4.043544303797469,
+      "grad_norm": 0.43821482225215613,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.2232,
+      "step": 998
+    },
+    {
+      "epoch": 4.047594936708861,
+      "grad_norm": 0.4604833556158873,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.2602,
+      "step": 999
+    },
+    {
+      "epoch": 4.051645569620253,
+      "grad_norm": 0.38160818268391683,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.2339,
+      "step": 1000
+    },
+    {
+      "epoch": 4.055696202531646,
+      "grad_norm": 0.3581220866527235,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.2282,
+      "step": 1001
+    },
+    {
+      "epoch": 4.059746835443038,
+      "grad_norm": 0.3717822450574522,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.2239,
+      "step": 1002
+    },
+    {
+      "epoch": 4.06379746835443,
+      "grad_norm": 0.41611350408267544,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2347,
+      "step": 1003
+    },
+    {
+      "epoch": 4.067848101265823,
+      "grad_norm": 0.4082106230774672,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2355,
+      "step": 1004
+    },
+    {
+      "epoch": 4.071898734177215,
+      "grad_norm": 0.3894466861788688,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.221,
+      "step": 1005
+    },
+    {
+      "epoch": 4.075949367088608,
+      "grad_norm": 0.36398985187983623,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.2183,
+      "step": 1006
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.35973245626163153,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.2282,
+      "step": 1007
+    },
+    {
+      "epoch": 4.084050632911392,
+      "grad_norm": 0.3985729200413975,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.1935,
+      "step": 1008
+    },
+    {
+      "epoch": 4.088101265822785,
+      "grad_norm": 0.35544450557095547,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2375,
+      "step": 1009
+    },
+    {
+      "epoch": 4.092151898734177,
+      "grad_norm": 0.36194096027865713,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2585,
+      "step": 1010
+    },
+    {
+      "epoch": 4.096202531645569,
+      "grad_norm": 0.34233061205269777,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.2106,
+      "step": 1011
+    },
+    {
+      "epoch": 4.100253164556962,
+      "grad_norm": 0.3336448844295868,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.221,
+      "step": 1012
+    },
+    {
+      "epoch": 4.104303797468354,
+      "grad_norm": 0.3583042088662754,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2474,
+      "step": 1013
+    },
+    {
+      "epoch": 4.108354430379747,
+      "grad_norm": 0.3304132695000972,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.2135,
+      "step": 1014
+    },
+    {
+      "epoch": 4.112405063291139,
+      "grad_norm": 0.3539932511170129,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.2473,
+      "step": 1015
+    },
+    {
+      "epoch": 4.116455696202531,
+      "grad_norm": 0.3407178453096992,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.258,
+      "step": 1016
+    },
+    {
+      "epoch": 4.120506329113924,
+      "grad_norm": 0.34336345545555275,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2242,
+      "step": 1017
+    },
+    {
+      "epoch": 4.124556962025316,
+      "grad_norm": 0.34147928906050723,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2209,
+      "step": 1018
+    },
+    {
+      "epoch": 4.1286075949367085,
+      "grad_norm": 0.34204361257591404,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.2144,
+      "step": 1019
+    },
+    {
+      "epoch": 4.132658227848101,
+      "grad_norm": 0.338021718454069,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2299,
+      "step": 1020
+    },
+    {
+      "epoch": 4.1367088607594935,
+      "grad_norm": 0.34590779947732264,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.2473,
+      "step": 1021
+    },
+    {
+      "epoch": 4.1407594936708865,
+      "grad_norm": 0.3331253811393305,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.2364,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1448101265822785,
+      "grad_norm": 0.3116764499253776,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2094,
+      "step": 1023
+    },
+    {
+      "epoch": 4.148860759493671,
+      "grad_norm": 0.3349520441350794,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.242,
+      "step": 1024
+    },
+    {
+      "epoch": 4.152911392405064,
+      "grad_norm": 0.3295304247506509,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.2085,
+      "step": 1025
+    },
+    {
+      "epoch": 4.156962025316456,
+      "grad_norm": 0.3453211576036429,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.232,
+      "step": 1026
+    },
+    {
+      "epoch": 4.161012658227848,
+      "grad_norm": 0.33145964626630053,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2152,
+      "step": 1027
+    },
+    {
+      "epoch": 4.165063291139241,
+      "grad_norm": 0.34499963450122123,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.2416,
+      "step": 1028
+    },
+    {
+      "epoch": 4.169113924050633,
+      "grad_norm": 0.29605977202234407,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2059,
+      "step": 1029
+    },
+    {
+      "epoch": 4.173164556962026,
+      "grad_norm": 0.34293855994720357,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2487,
+      "step": 1030
+    },
+    {
+      "epoch": 4.177215189873418,
+      "grad_norm": 0.3506169498359619,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.2513,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18126582278481,
+      "grad_norm": 0.3437392701106812,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2328,
+      "step": 1032
+    },
+    {
+      "epoch": 4.185316455696203,
+      "grad_norm": 0.32018805517515914,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2218,
+      "step": 1033
+    },
+    {
+      "epoch": 4.189367088607595,
+      "grad_norm": 0.3181966567845084,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.2267,
+      "step": 1034
+    },
+    {
+      "epoch": 4.193417721518987,
+      "grad_norm": 0.31632380308508223,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.2436,
+      "step": 1035
+    },
+    {
+      "epoch": 4.19746835443038,
+      "grad_norm": 0.3095439421345031,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.2104,
+      "step": 1036
+    },
+    {
+      "epoch": 4.201518987341772,
+      "grad_norm": 0.3404672613303262,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.2371,
+      "step": 1037
+    },
+    {
+      "epoch": 4.205569620253165,
+      "grad_norm": 0.3279491657605266,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.2355,
+      "step": 1038
+    },
+    {
+      "epoch": 4.209620253164557,
+      "grad_norm": 0.34138367726433144,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2512,
+      "step": 1039
+    },
+    {
+      "epoch": 4.213670886075949,
+      "grad_norm": 0.3190784136598891,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2256,
+      "step": 1040
+    },
+    {
+      "epoch": 4.217721518987342,
+      "grad_norm": 0.32109930526320646,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2315,
+      "step": 1041
+    },
+    {
+      "epoch": 4.221772151898734,
+      "grad_norm": 0.31766548396877464,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.2437,
+      "step": 1042
+    },
+    {
+      "epoch": 4.225822784810126,
+      "grad_norm": 0.3170148705923161,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.2384,
+      "step": 1043
+    },
+    {
+      "epoch": 4.229873417721519,
+      "grad_norm": 0.3249249091189558,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2317,
+      "step": 1044
+    },
+    {
+      "epoch": 4.233924050632911,
+      "grad_norm": 0.3466453032359142,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.2355,
+      "step": 1045
+    },
+    {
+      "epoch": 4.237974683544304,
+      "grad_norm": 0.3034524136870784,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2163,
+      "step": 1046
+    },
+    {
+      "epoch": 4.242025316455696,
+      "grad_norm": 0.33775626231413175,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2387,
+      "step": 1047
+    },
+    {
+      "epoch": 4.246075949367088,
+      "grad_norm": 0.3155863134689609,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.2231,
+      "step": 1048
+    },
+    {
+      "epoch": 4.250126582278481,
+      "grad_norm": 0.3354759123251955,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2491,
+      "step": 1049
+    },
+    {
+      "epoch": 4.254177215189873,
+      "grad_norm": 0.32495953517164294,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.217,
+      "step": 1050
+    },
+    {
+      "epoch": 4.258227848101265,
+      "grad_norm": 0.3193506754937909,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.2523,
+      "step": 1051
+    },
+    {
+      "epoch": 4.262278481012658,
+      "grad_norm": 0.3176717626652151,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.2354,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2663291139240505,
+      "grad_norm": 0.298249410672444,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.2026,
+      "step": 1053
+    },
+    {
+      "epoch": 4.270379746835443,
+      "grad_norm": 0.34515905313024625,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.239,
+      "step": 1054
+    },
+    {
+      "epoch": 4.2744303797468355,
+      "grad_norm": 0.32867697398738444,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.2312,
+      "step": 1055
+    },
+    {
+      "epoch": 4.2784810126582276,
+      "grad_norm": 0.29212763757381394,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.1912,
+      "step": 1056
+    },
+    {
+      "epoch": 4.2825316455696205,
+      "grad_norm": 0.3245547619698615,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.2504,
+      "step": 1057
+    },
+    {
+      "epoch": 4.286582278481013,
+      "grad_norm": 0.31248559982527035,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.2337,
+      "step": 1058
+    },
+    {
+      "epoch": 4.290632911392405,
+      "grad_norm": 0.34262130840148647,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.2537,
+      "step": 1059
+    },
+    {
+      "epoch": 4.294683544303798,
+      "grad_norm": 0.3431128508840366,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2286,
+      "step": 1060
+    },
+    {
+      "epoch": 4.29873417721519,
+      "grad_norm": 0.32257689761664654,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.2467,
+      "step": 1061
+    },
+    {
+      "epoch": 4.302784810126583,
+      "grad_norm": 0.32486990653464815,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.2372,
+      "step": 1062
+    },
+    {
+      "epoch": 4.306835443037975,
+      "grad_norm": 0.3083400737028036,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2085,
+      "step": 1063
+    },
+    {
+      "epoch": 4.310886075949367,
+      "grad_norm": 0.332057007592239,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.2355,
+      "step": 1064
+    },
+    {
+      "epoch": 4.31493670886076,
+      "grad_norm": 0.31885629489654344,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2209,
+      "step": 1065
+    },
+    {
+      "epoch": 4.318987341772152,
+      "grad_norm": 0.32253470917366195,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2275,
+      "step": 1066
+    },
+    {
+      "epoch": 4.323037974683544,
+      "grad_norm": 0.31519492587792414,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2212,
+      "step": 1067
+    },
+    {
+      "epoch": 4.327088607594937,
+      "grad_norm": 0.3262900142494503,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.229,
+      "step": 1068
+    },
+    {
+      "epoch": 4.331139240506329,
+      "grad_norm": 0.31742288467555463,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.2267,
+      "step": 1069
+    },
+    {
+      "epoch": 4.335189873417722,
+      "grad_norm": 0.31614735342274686,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.2459,
+      "step": 1070
+    },
+    {
+      "epoch": 4.339240506329114,
+      "grad_norm": 0.32365147265856875,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.2397,
+      "step": 1071
+    },
+    {
+      "epoch": 4.343291139240506,
+      "grad_norm": 0.3127732912592606,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.218,
+      "step": 1072
+    },
+    {
+      "epoch": 4.347341772151899,
+      "grad_norm": 0.3342590461128827,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2464,
+      "step": 1073
+    },
+    {
+      "epoch": 4.351392405063291,
+      "grad_norm": 0.3071935707035216,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.2184,
+      "step": 1074
+    },
+    {
+      "epoch": 4.355443037974683,
+      "grad_norm": 0.303000305625921,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.2201,
+      "step": 1075
+    },
+    {
+      "epoch": 4.359493670886076,
+      "grad_norm": 0.29775756614384974,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.2193,
+      "step": 1076
+    },
+    {
+      "epoch": 4.363544303797468,
+      "grad_norm": 0.30820239396781407,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.2304,
+      "step": 1077
+    },
+    {
+      "epoch": 4.367594936708861,
+      "grad_norm": 0.3020638237193348,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.2041,
+      "step": 1078
+    },
+    {
+      "epoch": 4.371645569620253,
+      "grad_norm": 0.3253902771506906,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.256,
+      "step": 1079
+    },
+    {
+      "epoch": 4.375696202531645,
+      "grad_norm": 0.31042696749026594,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2503,
+      "step": 1080
+    },
+    {
+      "epoch": 4.379746835443038,
+      "grad_norm": 0.3116225250053105,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.2155,
+      "step": 1081
+    },
+    {
+      "epoch": 4.38379746835443,
+      "grad_norm": 0.30655624662225417,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.2107,
+      "step": 1082
+    },
+    {
+      "epoch": 4.387848101265822,
+      "grad_norm": 0.32341589469313037,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.2234,
+      "step": 1083
+    },
+    {
+      "epoch": 4.391898734177215,
+      "grad_norm": 0.3302027233873041,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.2288,
+      "step": 1084
+    },
+    {
+      "epoch": 4.395949367088607,
+      "grad_norm": 0.29849157538121274,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.2256,
+      "step": 1085
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.34546771697443024,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.2592,
+      "step": 1086
+    },
+    {
+      "epoch": 4.404050632911392,
+      "grad_norm": 0.3088003825597379,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.2177,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4081012658227845,
+      "grad_norm": 0.3144112473003213,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2357,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4121518987341775,
+      "grad_norm": 0.3011080191029588,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2152,
+      "step": 1089
+    },
+    {
+      "epoch": 4.4162025316455695,
+      "grad_norm": 0.3364853148360307,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.2404,
+      "step": 1090
+    },
+    {
+      "epoch": 4.4202531645569625,
+      "grad_norm": 0.3074940188453271,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.2211,
+      "step": 1091
+    },
+    {
+      "epoch": 4.424303797468355,
+      "grad_norm": 0.31616263616502766,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.2246,
+      "step": 1092
+    },
+    {
+      "epoch": 4.428354430379747,
+      "grad_norm": 0.3067886307700066,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2442,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43240506329114,
+      "grad_norm": 0.2953526484387316,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2035,
+      "step": 1094
+    },
+    {
+      "epoch": 4.436455696202532,
+      "grad_norm": 0.3159878798450858,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2325,
+      "step": 1095
+    },
+    {
+      "epoch": 4.440506329113924,
+      "grad_norm": 0.3301863152340416,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.2516,
+      "step": 1096
+    },
+    {
+      "epoch": 4.444556962025317,
+      "grad_norm": 0.31076213447323425,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.2071,
+      "step": 1097
+    },
+    {
+      "epoch": 4.448607594936709,
+      "grad_norm": 0.3095805873210312,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2101,
+      "step": 1098
+    },
+    {
+      "epoch": 4.452658227848101,
+      "grad_norm": 0.3380813878403558,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.2594,
+      "step": 1099
+    },
+    {
+      "epoch": 4.456708860759494,
+      "grad_norm": 0.31001554222577443,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.2141,
+      "step": 1100
+    },
+    {
+      "epoch": 4.460759493670886,
+      "grad_norm": 0.31667484129312723,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2407,
+      "step": 1101
+    },
+    {
+      "epoch": 4.464810126582279,
+      "grad_norm": 0.3430885546343663,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2285,
+      "step": 1102
+    },
+    {
+      "epoch": 4.468860759493671,
+      "grad_norm": 0.3044416177295222,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2117,
+      "step": 1103
+    },
+    {
+      "epoch": 4.472911392405063,
+      "grad_norm": 0.3180842490310364,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.2393,
+      "step": 1104
+    },
+    {
+      "epoch": 4.476962025316456,
+      "grad_norm": 0.32651385754680534,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.2494,
+      "step": 1105
+    },
+    {
+      "epoch": 4.481012658227848,
+      "grad_norm": 0.2913253122012983,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.2078,
+      "step": 1106
+    },
+    {
+      "epoch": 4.485063291139241,
+      "grad_norm": 0.3354544253680717,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2495,
+      "step": 1107
+    },
+    {
+      "epoch": 4.489113924050633,
+      "grad_norm": 0.3169338876557915,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.2375,
+      "step": 1108
+    },
+    {
+      "epoch": 4.493164556962025,
+      "grad_norm": 0.30273373932017783,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2148,
+      "step": 1109
+    },
+    {
+      "epoch": 4.497215189873418,
+      "grad_norm": 0.3218480414890918,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.2325,
+      "step": 1110
+    },
+    {
+      "epoch": 4.50126582278481,
+      "grad_norm": 0.334803791916304,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.2464,
+      "step": 1111
+    },
+    {
+      "epoch": 4.505316455696202,
+      "grad_norm": 0.3122386818531078,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.236,
+      "step": 1112
+    },
+    {
+      "epoch": 4.509367088607595,
+      "grad_norm": 0.30807736087196874,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2258,
+      "step": 1113
+    },
+    {
+      "epoch": 4.513417721518987,
+      "grad_norm": 0.31076768866106047,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.2257,
+      "step": 1114
+    },
+    {
+      "epoch": 4.517468354430379,
+      "grad_norm": 0.32802762576815847,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.2618,
+      "step": 1115
+    },
+    {
+      "epoch": 4.521518987341772,
+      "grad_norm": 0.32579635807344726,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2248,
+      "step": 1116
+    },
+    {
+      "epoch": 4.525569620253164,
+      "grad_norm": 0.3220599660092697,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2363,
+      "step": 1117
+    },
+    {
+      "epoch": 4.529620253164557,
+      "grad_norm": 0.29942997295553764,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2228,
+      "step": 1118
+    },
+    {
+      "epoch": 4.533670886075949,
+      "grad_norm": 0.32408645964663035,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2275,
+      "step": 1119
+    },
+    {
+      "epoch": 4.537721518987341,
+      "grad_norm": 0.3186218458585043,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.2283,
+      "step": 1120
+    },
+    {
+      "epoch": 4.541772151898734,
+      "grad_norm": 0.312784360717046,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.2271,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5458227848101265,
+      "grad_norm": 0.32069465252098084,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2382,
+      "step": 1122
+    },
+    {
+      "epoch": 4.549873417721519,
+      "grad_norm": 0.29652914526583973,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.2181,
+      "step": 1123
+    },
+    {
+      "epoch": 4.5539240506329115,
+      "grad_norm": 0.3295795346735293,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.2478,
+      "step": 1124
+    },
+    {
+      "epoch": 4.557974683544304,
+      "grad_norm": 0.3132448559406914,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.2397,
+      "step": 1125
+    },
+    {
+      "epoch": 4.5620253164556965,
+      "grad_norm": 0.3091039941062537,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2397,
+      "step": 1126
+    },
+    {
+      "epoch": 4.566075949367089,
+      "grad_norm": 0.29111887021385624,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.2171,
+      "step": 1127
+    },
+    {
+      "epoch": 4.570126582278481,
+      "grad_norm": 0.2927043510013831,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.2055,
+      "step": 1128
+    },
+    {
+      "epoch": 4.574177215189874,
+      "grad_norm": 0.33886222165268953,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.2682,
+      "step": 1129
+    },
+    {
+      "epoch": 4.578227848101266,
+      "grad_norm": 0.30446442339679847,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2086,
+      "step": 1130
+    },
+    {
+      "epoch": 4.582278481012658,
+      "grad_norm": 0.31620373980696137,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.2559,
+      "step": 1131
+    },
+    {
+      "epoch": 4.586329113924051,
+      "grad_norm": 0.3078258416863651,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.2315,
+      "step": 1132
+    },
+    {
+      "epoch": 4.590379746835443,
+      "grad_norm": 0.3365108458402454,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.2279,
+      "step": 1133
+    },
+    {
+      "epoch": 4.594430379746836,
+      "grad_norm": 0.31444736424234104,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.203,
+      "step": 1134
+    },
+    {
+      "epoch": 4.598481012658228,
+      "grad_norm": 0.3266085290876586,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.2643,
+      "step": 1135
+    },
+    {
+      "epoch": 4.60253164556962,
+      "grad_norm": 0.3214273051515812,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.2186,
+      "step": 1136
+    },
+    {
+      "epoch": 4.606582278481013,
+      "grad_norm": 0.31347816330548894,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2146,
+      "step": 1137
+    },
+    {
+      "epoch": 4.610632911392405,
+      "grad_norm": 0.3075523933923044,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.2265,
+      "step": 1138
+    },
+    {
+      "epoch": 4.614683544303798,
+      "grad_norm": 0.3183296319627729,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2541,
+      "step": 1139
+    },
+    {
+      "epoch": 4.61873417721519,
+      "grad_norm": 0.30938684380910697,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.2259,
+      "step": 1140
+    },
+    {
+      "epoch": 4.622784810126582,
+      "grad_norm": 0.3830715607402039,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.2401,
+      "step": 1141
+    },
+    {
+      "epoch": 4.626835443037975,
+      "grad_norm": 0.3247177356364176,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.2413,
+      "step": 1142
+    },
+    {
+      "epoch": 4.630886075949367,
+      "grad_norm": 0.3065317992703947,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.2158,
+      "step": 1143
+    },
+    {
+      "epoch": 4.634936708860759,
+      "grad_norm": 0.3104542093882322,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.253,
+      "step": 1144
+    },
+    {
+      "epoch": 4.638987341772152,
+      "grad_norm": 0.307200753547952,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.2296,
+      "step": 1145
+    },
+    {
+      "epoch": 4.643037974683544,
+      "grad_norm": 0.33312387614770717,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.2206,
+      "step": 1146
+    },
+    {
+      "epoch": 4.647088607594936,
+      "grad_norm": 0.31648243754201516,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2392,
+      "step": 1147
+    },
+    {
+      "epoch": 4.651139240506329,
+      "grad_norm": 0.30520454259274693,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.2354,
+      "step": 1148
+    },
+    {
+      "epoch": 4.655189873417721,
+      "grad_norm": 0.2939122317350534,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.2037,
+      "step": 1149
+    },
+    {
+      "epoch": 4.659240506329114,
+      "grad_norm": 0.33931610614614355,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.2663,
+      "step": 1150
+    },
+    {
+      "epoch": 4.663291139240506,
+      "grad_norm": 0.32928578137312636,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.228,
+      "step": 1151
+    },
+    {
+      "epoch": 4.667341772151898,
+      "grad_norm": 0.3034247566710897,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.2222,
+      "step": 1152
+    },
+    {
+      "epoch": 4.671392405063291,
+      "grad_norm": 0.3088414350440436,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2461,
+      "step": 1153
+    },
+    {
+      "epoch": 4.675443037974683,
+      "grad_norm": 0.3151104636532443,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.2316,
+      "step": 1154
+    },
+    {
+      "epoch": 4.679493670886076,
+      "grad_norm": 0.31440747155009957,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.2459,
+      "step": 1155
+    },
+    {
+      "epoch": 4.6835443037974684,
+      "grad_norm": 0.3082633876565363,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.2093,
+      "step": 1156
+    },
+    {
+      "epoch": 4.6875949367088605,
+      "grad_norm": 0.320926650183746,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.2379,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6916455696202535,
+      "grad_norm": 0.30907391317376687,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2168,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6956962025316455,
+      "grad_norm": 0.3205278197568248,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.26,
+      "step": 1159
+    },
+    {
+      "epoch": 4.699746835443038,
+      "grad_norm": 0.30711817378349,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2398,
+      "step": 1160
+    },
+    {
+      "epoch": 4.703797468354431,
+      "grad_norm": 0.3145593412047012,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.2275,
+      "step": 1161
+    },
+    {
+      "epoch": 4.707848101265823,
+      "grad_norm": 0.33214044996459674,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.2337,
+      "step": 1162
+    },
+    {
+      "epoch": 4.711898734177215,
+      "grad_norm": 0.30813754159720574,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.2447,
+      "step": 1163
+    },
+    {
+      "epoch": 4.715949367088608,
+      "grad_norm": 0.29606314360788527,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2301,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.30864347993821173,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.2521,
+      "step": 1165
+    },
+    {
+      "epoch": 4.724050632911393,
+      "grad_norm": 0.3187845215983957,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2346,
+      "step": 1166
+    },
+    {
+      "epoch": 4.728101265822785,
+      "grad_norm": 0.3045077623974581,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2424,
+      "step": 1167
+    },
+    {
+      "epoch": 4.732151898734177,
+      "grad_norm": 0.29352436946662647,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.2347,
+      "step": 1168
+    },
+    {
+      "epoch": 4.73620253164557,
+      "grad_norm": 0.3213587794740573,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.2366,
+      "step": 1169
+    },
+    {
+      "epoch": 4.740253164556962,
+      "grad_norm": 0.3047735871453943,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2328,
+      "step": 1170
+    },
+    {
+      "epoch": 4.744303797468355,
+      "grad_norm": 0.29569939270750684,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.2252,
+      "step": 1171
+    },
+    {
+      "epoch": 4.748354430379747,
+      "grad_norm": 0.31715841469755546,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2488,
+      "step": 1172
+    },
+    {
+      "epoch": 4.752405063291139,
+      "grad_norm": 0.30945943467494474,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2176,
+      "step": 1173
+    },
+    {
+      "epoch": 4.756455696202532,
+      "grad_norm": 0.32455906763710307,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.2374,
+      "step": 1174
+    },
+    {
+      "epoch": 4.760506329113924,
+      "grad_norm": 0.3252555515825012,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.2487,
+      "step": 1175
+    },
+    {
+      "epoch": 4.764556962025316,
+      "grad_norm": 0.31492645788936324,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2205,
+      "step": 1176
+    },
+    {
+      "epoch": 4.768607594936709,
+      "grad_norm": 0.34083088066551387,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.2344,
+      "step": 1177
+    },
+    {
+      "epoch": 4.772658227848101,
+      "grad_norm": 0.31313761717784216,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2421,
+      "step": 1178
+    },
+    {
+      "epoch": 4.776708860759493,
+      "grad_norm": 0.2952093332637497,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.2005,
+      "step": 1179
+    },
+    {
+      "epoch": 4.780759493670886,
+      "grad_norm": 0.3193947001836654,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2211,
+      "step": 1180
+    },
+    {
+      "epoch": 4.784810126582278,
+      "grad_norm": 0.3345875925677834,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.2341,
+      "step": 1181
+    },
+    {
+      "epoch": 4.788860759493671,
+      "grad_norm": 0.3258772697092736,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2168,
+      "step": 1182
+    },
+    {
+      "epoch": 4.792911392405063,
+      "grad_norm": 0.30216475862425457,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2085,
+      "step": 1183
+    },
+    {
+      "epoch": 4.796962025316455,
+      "grad_norm": 0.3110905799931299,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2558,
+      "step": 1184
+    },
+    {
+      "epoch": 4.801012658227848,
+      "grad_norm": 0.3184004923259186,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.2349,
+      "step": 1185
+    },
+    {
+      "epoch": 4.80506329113924,
+      "grad_norm": 0.308148303842863,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.2124,
+      "step": 1186
+    },
+    {
+      "epoch": 4.809113924050633,
+      "grad_norm": 0.3097948563084637,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2262,
+      "step": 1187
+    },
+    {
+      "epoch": 4.813164556962025,
+      "grad_norm": 0.32670313641358384,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2505,
+      "step": 1188
+    },
+    {
+      "epoch": 4.8172151898734175,
+      "grad_norm": 0.30104393067622326,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.252,
+      "step": 1189
+    },
+    {
+      "epoch": 4.82126582278481,
+      "grad_norm": 0.3019504910052267,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2109,
+      "step": 1190
+    },
+    {
+      "epoch": 4.8253164556962025,
+      "grad_norm": 0.32065757038133985,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.2494,
+      "step": 1191
+    },
+    {
+      "epoch": 4.829367088607595,
+      "grad_norm": 0.30102668075288636,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.2173,
+      "step": 1192
+    },
+    {
+      "epoch": 4.8334177215189875,
+      "grad_norm": 0.3123157177572655,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.2279,
+      "step": 1193
+    },
+    {
+      "epoch": 4.83746835443038,
+      "grad_norm": 0.3157469625663789,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2306,
+      "step": 1194
+    },
+    {
+      "epoch": 4.841518987341772,
+      "grad_norm": 0.33132122688129206,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.2537,
+      "step": 1195
+    },
+    {
+      "epoch": 4.845569620253165,
+      "grad_norm": 0.2934831489349099,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.2128,
+      "step": 1196
+    },
+    {
+      "epoch": 4.849620253164557,
+      "grad_norm": 0.31719061567959217,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.2337,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85367088607595,
+      "grad_norm": 0.30046457283962047,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.2196,
+      "step": 1198
+    },
+    {
+      "epoch": 4.857721518987342,
+      "grad_norm": 0.3372224091097723,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.2546,
+      "step": 1199
+    },
+    {
+      "epoch": 4.861772151898734,
+      "grad_norm": 0.3056808105667589,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1988,
+      "step": 1200
+    },
+    {
+      "epoch": 4.865822784810127,
+      "grad_norm": 0.3248271832754149,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.2449,
+      "step": 1201
+    },
+    {
+      "epoch": 4.869873417721519,
+      "grad_norm": 0.2989891384785881,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.2257,
+      "step": 1202
+    },
+    {
+      "epoch": 4.873924050632912,
+      "grad_norm": 0.2971827837710027,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2128,
+      "step": 1203
+    },
+    {
+      "epoch": 4.877974683544304,
+      "grad_norm": 0.3256911743399988,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.2522,
+      "step": 1204
+    },
+    {
+      "epoch": 4.882025316455696,
+      "grad_norm": 0.31335315529233415,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2368,
+      "step": 1205
+    },
+    {
+      "epoch": 4.886075949367089,
+      "grad_norm": 0.3050969317332606,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.2152,
+      "step": 1206
+    },
+    {
+      "epoch": 4.890126582278481,
+      "grad_norm": 0.30508863582784757,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2119,
+      "step": 1207
+    },
+    {
+      "epoch": 4.894177215189873,
+      "grad_norm": 0.32683304168872856,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2529,
+      "step": 1208
+    },
+    {
+      "epoch": 4.898227848101266,
+      "grad_norm": 0.2984320695984988,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.2218,
+      "step": 1209
+    },
+    {
+      "epoch": 4.902278481012658,
+      "grad_norm": 0.31198257852375655,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.2382,
+      "step": 1210
+    },
+    {
+      "epoch": 4.90632911392405,
+      "grad_norm": 0.31772663730430417,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2339,
+      "step": 1211
+    },
+    {
+      "epoch": 4.910379746835443,
+      "grad_norm": 0.3136481097421503,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.2217,
+      "step": 1212
+    },
+    {
+      "epoch": 4.914430379746835,
+      "grad_norm": 0.3237310930280893,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.244,
+      "step": 1213
+    },
+    {
+      "epoch": 4.918481012658228,
+      "grad_norm": 0.30830687143814045,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.2201,
+      "step": 1214
+    },
+    {
+      "epoch": 4.92253164556962,
+      "grad_norm": 0.28859669981602387,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.2214,
+      "step": 1215
+    },
+    {
+      "epoch": 4.926582278481012,
+      "grad_norm": 0.3125673914997827,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.2321,
+      "step": 1216
+    },
+    {
+      "epoch": 4.930632911392405,
+      "grad_norm": 0.30212304122944694,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2204,
+      "step": 1217
+    },
+    {
+      "epoch": 4.934683544303797,
+      "grad_norm": 0.3007868437834777,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2229,
+      "step": 1218
+    },
+    {
+      "epoch": 4.93873417721519,
+      "grad_norm": 0.3251508315171844,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2402,
+      "step": 1219
+    },
+    {
+      "epoch": 4.942784810126582,
+      "grad_norm": 0.3257649242557537,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.2485,
+      "step": 1220
+    },
+    {
+      "epoch": 4.946835443037974,
+      "grad_norm": 0.29797245578476583,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.2085,
+      "step": 1221
+    },
+    {
+      "epoch": 4.950886075949367,
+      "grad_norm": 0.3064023037990366,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2279,
+      "step": 1222
+    },
+    {
+      "epoch": 4.954936708860759,
+      "grad_norm": 0.3177673300477984,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.2397,
+      "step": 1223
+    },
+    {
+      "epoch": 4.958987341772152,
+      "grad_norm": 0.3194018552377612,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2108,
+      "step": 1224
+    },
+    {
+      "epoch": 4.9630379746835445,
+      "grad_norm": 0.33571415442118485,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2306,
+      "step": 1225
+    },
+    {
+      "epoch": 4.9670886075949365,
+      "grad_norm": 0.31294789991593425,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.2195,
+      "step": 1226
+    },
+    {
+      "epoch": 4.9711392405063295,
+      "grad_norm": 0.3054999014245937,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.2315,
+      "step": 1227
+    },
+    {
+      "epoch": 4.975189873417722,
+      "grad_norm": 0.3380929794220337,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.2301,
+      "step": 1228
+    },
+    {
+      "epoch": 4.979240506329114,
+      "grad_norm": 0.3321144544759669,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.241,
+      "step": 1229
+    },
+    {
+      "epoch": 4.983291139240507,
+      "grad_norm": 0.2915411264613831,
+      "learning_rate": 0.0,
+      "loss": 0.2115,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983291139240507,
+      "step": 1230,
+      "total_flos": 2.3904400115949896e+18,
+      "train_loss": 0.4335456542489005,
+      "train_runtime": 120917.786,
+      "train_samples_per_second": 1.307,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "logging_steps": 1.0,
+  "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": 2.3904400115949896e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}