diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 8.205563605523626,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.9826,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.520575626107934,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.0245,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.531589483073382,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.0153,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.815733570718443,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.0378,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 7.800544041203901,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.9895,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.656481094026679,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.9873,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.07128875746122,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.9478,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.8699931499539835,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.9416,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.917398750267831,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.8777,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.7323876659844273,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.8686,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.3890464189689933,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.8434,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 4.933304610948769,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.8247,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 5.172978911374976,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.8252,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 4.559331322281286,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.772,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 4.00463824996656,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.795,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.0371495377952753,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.7128,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.257391842290354,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.7072,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.68412941417381,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.6841,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.0461725876484937,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.6889,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.5477297510662222,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.6487,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.3246384239165625,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.6429,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.706059460246355,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.6242,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.4537667222948356,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.6127,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.1158005632256833,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.6042,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.1382990304618488,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.6143,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1833776157522509,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.6018,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.1969725206739448,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.5773,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.0748823324320453,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.575,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.1074757253413123,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.5624,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.8526542016907144,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.5389,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.9488852066767993,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.5599,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 1.0267568139047563,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.5547,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.0226499534668898,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.5467,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.7931677396127635,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.5481,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 1.0411038325780866,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.5528,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 1.0674053440950175,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.5311,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.7602972992759987,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.5163,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.8528369136932196,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.5361,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8536141068931459,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.5349,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.7811007061103901,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.5236,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.7652842609655697,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5074,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8427953106311229,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.4995,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7983777882826968,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.5323,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.8496302607218493,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.513,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7955592443937491,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.5118,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.7243832278149757,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.5107,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7305776200024143,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.5031,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.7357205895964949,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.4993,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.8659368175381161,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.5234,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.7466431689328721,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.5144,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7179398647233523,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.4682,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.725805382855159,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.5141,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8107773106893855,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.5002,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.7702702909724286,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.4976,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.9177771231874979,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.5034,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.7588912666621611,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.4913,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.864541752034926,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.5139,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8701756473537474,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.493,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.1366424021752426,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.4779,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.0452180973819707,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.4819,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.8711996224719698,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.4903,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.2299442986583096,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.4922,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8112293237496551,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.4697,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.9356305503035498,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.4781,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.0768693116005401,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.4801,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7796797011188623,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.4766,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.364701728817263,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.4797,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.0568242609724223,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.4879,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.8560541077868588,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.4885,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.0587213325619305,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.4758,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.8687116271064114,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.4817,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8928923336333813,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.479,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.9823244574659463,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.4762,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.9156606460719805,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.4947,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.0512389032106149,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.4709,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.0686643247068681,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.4836,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.8361236014914584,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.465,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.8902521468360812,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.4823,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9687726701222227,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.4703,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.0731647974838499,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.4781,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.9648881280873751,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.4812,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.232028541250671,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4745,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.9484723376430446,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.4677,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.1715196811969844,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.481,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.0811446008822014,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.4642,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.849420258239314,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.4687,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.2174747926097895,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.487,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.8743140371687586,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.4667,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.8459787082606486,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.4683,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.2769376761378912,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.4666,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.7656773006939884,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.4765,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.1404386828469237,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.457,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.1572915574447904,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.4783,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.8281700477045351,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.4595,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.052436674157122,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.4565,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 0.966896755211345,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.4542,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.9588946079956988,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.47,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.2615724294484987,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.472,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.1260717217693306,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.4614,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.8707306620537176,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.4723,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.405144697090571,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.4748,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.016777232856155,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.4841,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.2847851794579868,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.4745,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.102244708065243,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.4805,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.175835975302144,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.4725,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.8094436722515507,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.4473,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.007725732840739,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.4592,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8840970586392084,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.4565,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.9868247230508853,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.4504,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.1244943444007136,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.4685,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.1264726301742622,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.4694,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.7553915240648305,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.4771,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.030548149386525,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.4542,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.104549899046257,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.4701,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.8487880976061655,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.4656,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.8669760760988977,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.4581,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.0070635589070795,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.4654,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.9419398597633974,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.4616,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.123781339812278,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.4803,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.9442795163106134,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.4661,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.878255899393012,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.4645,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.9644493248797653,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.4809,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.9709919663288659,
+      "learning_rate": 4e-05,
+      "loss": 0.445,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.1864093584230193,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.4718,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.8253774449322091,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.4632,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.0054158692907014,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.4614,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.002989117602263,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.4624,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.8635329735241931,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.457,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.9080657530780644,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.4546,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.885821457771768,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.4594,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.8778310850401578,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.4539,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.7707344778272056,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.4653,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.041112474942178,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.4516,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.9211018269553279,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.4738,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.0104172384412176,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.4545,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.8752131116388303,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.4688,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.9471533693904083,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.4618,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.6709055915487836,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.454,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.8465903271244486,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.4464,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.6544186984234165,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.4572,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.7065754248038563,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.4818,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.6541187545812097,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.4447,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.6645000233010177,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.4497,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.6707812753554334,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.4469,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7574452614621905,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.4462,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.7637520653986111,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.46,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.8234399945452502,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.4483,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.7982074967516908,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.4607,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.8171190644853485,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.447,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.7044076361514297,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.4649,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.6610797114903746,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.45,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.7531233919347446,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.4546,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.7889692222311209,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.4481,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.5872815713619045,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.4502,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.6891905718172654,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.4652,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.8156376676502983,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.466,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.6156265287563699,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.4512,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8011120579662531,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.4513,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.5340877344706761,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.4512,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.783793242859297,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.4486,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6046195387695634,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.4429,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.565191804624553,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.456,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.5659660433705522,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.4471,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.5975982055856516,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.4458,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.5636025521719215,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.4579,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.5507847163965235,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.4248,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.6423338706209568,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.4452,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6162655237237198,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.4327,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.7315687775829846,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.4675,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.5993934990565606,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.4465,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.8695135819081058,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.4454,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.6219663065405475,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.4615,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.754804060719734,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.4401,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.6573206756588172,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.4537,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.6933847925692165,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.4501,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.6592588815585356,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.4377,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.6338736396270657,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.4435,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.6922444857823782,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.4618,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.8640380271451539,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.4472,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.645330386524199,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.4429,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.6051255085923929,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.4454,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.8554968328185691,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.4503,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.7249768071519656,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.4563,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.548481713871832,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.462,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.6358262217600824,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.4339,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.709430208892155,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.4443,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.6932035620121949,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.4541,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.6689990220850016,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.4416,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.698395370677472,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.4513,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.7510682103892015,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.4357,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.6555741492204371,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.4395,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.7405582591896879,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.4503,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.6602564948927093,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.4473,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.6174932439255586,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.4614,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.7521198325328702,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.4601,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.5200807695030266,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.4413,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.6666807895842612,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.4285,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6422350321080126,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.4343,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6567176047110885,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.4444,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.6280760055277653,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.4284,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.6931104447283385,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.4419,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.6073225608881315,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.4449,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5009015896753968,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.4461,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.5285106162954648,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.4417,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.47937668542544576,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.4373,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.6299272978793117,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.4456,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.5265665860028103,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.4429,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.6333015602969901,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.4237,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.5061260308160164,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.436,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.5666190341258476,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.4391,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.5963538588433019,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.4385,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.5947442558658302,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.4437,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.49174281068846154,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.4212,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.6106019493095122,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.4468,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.5328323470788556,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.436,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.6391264114447044,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.4436,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.6826743592737889,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.4336,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.670334935424927,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.4277,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.8990776161277816,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.4489,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.6224724135477777,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.443,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.6944743985611724,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.4409,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.6596326319402427,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.4257,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.6270790043705198,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.4492,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.6008864953119004,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.4428,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.6990388491875826,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.4277,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.6039298302657041,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.4237,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.8291483628287114,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.4413,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.4849440680366992,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.433,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.6869807213545654,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.4355,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.6594030007647461,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.4259,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.6624204799595027,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.4537,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.49263995571913904,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.4278,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.6367710011151443,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.438,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.6416317285255059,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.4301,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.6537470613606488,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.4274,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.6027960494396617,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.4299,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6057420855029728,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.4361,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.5849368915351931,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.4288,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.6034911506292389,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.432,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.643763970704869,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.4321,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.5564287198888858,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.4407,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.6235900913426907,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.4361,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.5536121736000429,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.4317,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.6507137777564832,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.4275,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.5778145971567855,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.4229,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.6503977002161673,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.4423,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.525787840762158,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.4057,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.5828269988082703,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.4004,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.5033336431516809,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.3987,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.5127768048158733,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.3866,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.6244217166781865,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.3886,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.5096569233339372,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.3973,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.5047426386230759,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.3877,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.5225507594426145,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.3994,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.4920525334423321,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.3785,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.5340957323193196,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.3938,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.45820633311466963,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.3939,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.5958924465920712,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.3988,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.535155076574753,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.3926,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.45891233934869946,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.3897,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.5139537999576951,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.3902,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.5010587111422404,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.3837,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.49185601365704384,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.3918,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.571083066252806,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.3923,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.4886076629574032,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.3877,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.5683318193480936,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.3974,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.5238165379084174,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.4125,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.465318540510975,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.3905,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.5302107417075062,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.3983,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.5064039780824413,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.3972,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.594354974165014,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.3986,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.5431905652201978,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.3961,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.5113370604590248,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.3694,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.5829566053531753,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.3744,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.498313757829123,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.39,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.5819724774382438,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.3904,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.4995866583343578,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.3827,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.5888558173167969,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.3817,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.5649823412628839,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.3814,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.4751292718421678,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.3938,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.7089829193191904,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.3851,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.532704443895478,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.3871,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.5127737026363255,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.3774,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.5537579399951169,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.3846,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.49039104360740376,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.3899,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.47887650529667497,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.3944,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.5679905994449338,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.4034,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.5849712151001694,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.4047,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.5249862740567374,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.3852,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.6743876235623733,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.3951,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.5606395013350859,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.4006,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.518834564049066,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.3959,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.5958364900705619,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.3878,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.5049929494434245,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.3823,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.5800207016599819,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.3894,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.49312217718269313,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.3878,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.5260761388178883,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.385,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.5168375530757567,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.3799,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.5407358806713677,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.3956,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.4896421739716556,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.3948,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.5629672037943392,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.4029,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.4566988990028118,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.3977,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.5936137324192661,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.3956,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.514335652599778,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.396,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.5723245881261749,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.4042,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.5983662023437094,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.3729,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.5920293262632803,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.3833,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.518343540660621,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.3835,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.6415071807533029,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.3908,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.5284294893819116,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.3936,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.6120469765474384,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.3928,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.5305619840623916,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4019,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.5082981535992244,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.3797,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.5006330656305775,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.4024,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.45707123827839374,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.377,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.5306059038121849,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.4042,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.48660814711723505,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.4019,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.5206631812779264,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.3834,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.45284371018383346,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.3815,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.5599901990713267,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.391,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.4855434102822341,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.3957,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.4751910934256773,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.4015,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.4907414056646788,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.3794,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.4848116379684677,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.4009,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.48588231759322487,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.3775,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.5343039510164631,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.3926,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.49998535856153736,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.3947,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.5079410277116974,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.3757,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.48576467097791465,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.3717,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.5190152993180762,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.394,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.5021528997406164,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.3932,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.5608883236261658,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.3948,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.4794538817100655,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.4059,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.563800985705449,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.3906,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.5127349591541009,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.3937,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.6315858396827547,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.3968,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.4956114135478616,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.3844,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.6271823310574456,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.3831,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.4732196348303393,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.3807,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.5658273946478825,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.3843,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.6169424863275222,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.4151,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.5216409686512313,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.3729,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.6468893049539661,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.3799,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.5340569595726953,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.3909,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.5580952069080541,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.3902,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.567543001161031,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.3887,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.5247984473850064,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.3987,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.5875650207360172,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.3816,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.54903951580874,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.3849,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.5962314751826311,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.4027,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.570060711219328,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.3985,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.5906344300852827,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.3979,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.4952677795538844,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.3936,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.5100375931700892,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.3939,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.487197216868086,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.3814,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.5226666518078672,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.3932,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.566034091054618,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.3973,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.6644756139211706,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.4056,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.5066084578075929,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.3932,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.5895325867141599,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.3879,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.48057195922136053,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.3761,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.5327897963466052,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.3879,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.5456250468801924,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.4015,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.5320434447695954,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.3975,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5046424961762135,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.3901,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.5036716132896111,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4003,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.4436689561882577,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.3831,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.5497461520900675,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.3937,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.45890551180833333,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.3828,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.493662177394805,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.3921,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.4729309497184521,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.3966,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.4311861273839032,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.391,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.42490897850004467,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.3831,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.4123811664496445,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.3826,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.4811030680026501,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.3912,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.38932466716035524,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.3932,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.45617646465758405,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.3802,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.447037402977748,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.3817,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.4287666563511986,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.3858,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.5229954544165956,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.3972,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.4309056401056063,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.3892,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.5308602750824283,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.3947,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.41840410490119273,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.4029,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.5035610525338285,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.3967,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.4683567963034917,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.3951,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.4991857507937191,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.381,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.4999961126215551,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.3878,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.5183859539263345,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.3798,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.48926914524919635,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.3961,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.4873679162821492,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.388,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.5120084853799863,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.3836,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.46903261319896894,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.3829,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.46374786440611443,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4068,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.5056334731308223,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.3923,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.4483263643750139,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.3984,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.4939746859081739,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.3993,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.4025493651365061,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.3864,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.5341795389626486,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.3864,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.3922803835847682,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.4117,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.49600812648937737,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.3915,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.4631869407701392,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.4032,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.4625639289676811,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.3942,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.4716238734550143,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.3964,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.40620771436879316,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.3964,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.45305793869760547,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.3956,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.45346108039306954,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.3873,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.45511010502569815,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.3789,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.47942290087624945,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.3947,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.42739252266758865,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.3821,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.46031590891093954,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.3969,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.43216258556784476,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.3899,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.5067846748485102,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.3869,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.472268234106378,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.3802,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.4659254731639721,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.3836,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.47067400608277715,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.3901,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.4455731433730483,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.3852,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.4713585920506693,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.4015,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.4385221902243143,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.3921,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.4902032811207186,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.3835,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.4420529649143202,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.3865,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.4569651151133509,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.3768,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.43594779454961113,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.3958,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.4499904613599221,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.405,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.38398653485269574,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.3915,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.4164106728641908,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.3729,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.3824129674550446,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.391,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.41587640560172323,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.4003,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.4231855081600333,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.3957,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.4127675769779056,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.3753,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.4707394233715805,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.396,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.42113482851147843,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.3829,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.5049385892460551,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.3908,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.4385696461436774,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.3911,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.456575554237019,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.3788,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.4650896961313856,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.39,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.4717725352910925,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.3963,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.475568720880946,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.3838,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.49872501833394145,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.3881,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.4256282237140198,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.3903,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.4470466267326493,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.3876,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.4455602543850904,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.3866,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.4149007484882896,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.3767,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.4576546206317219,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.3878,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.39712138200382674,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.3843,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.4085004995951105,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.39,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.40350934194622357,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.3934,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.47877237542728973,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.3925,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.4441050222983409,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.3912,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.422066683584006,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.3738,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.536536898658822,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.3905,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.4422551257111069,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.392,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.4635561795114512,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.3909,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.4230788767635726,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.3745,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.4449361952482121,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.4028,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.4670730184143889,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.3874,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.444108733379471,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.3831,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.4587280781392004,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.3734,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.4455254016010154,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.3796,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.4779258628076881,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.3755,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.4771079831722194,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.403,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.46953603794567983,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.3748,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.5029075951296426,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.3789,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.5151732564506124,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.3825,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.4803495162466444,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.3901,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.47213984207041876,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.3974,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.4820941134538517,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.3794,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.4864398442110171,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.3853,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.46337249312372975,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.4086,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.4285357815462205,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.3945,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.4581473661422968,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.3947,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.3892545310998907,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.381,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.46932972537390355,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.3964,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.46123927646072777,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.3803,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.5520615544528047,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.3908,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.4584971337902706,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.3838,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.4912223001009565,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.4032,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.4152529863189953,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.3762,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.38300916971715626,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.3884,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.38942746065525397,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.3866,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.41543957536773024,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.3954,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.41802992763305097,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.3851,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.38478972605040174,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.3757,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.41894251852935843,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.3882,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.39121803799874544,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.3749,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.4127883595334638,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.3877,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.389797596242296,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.3944,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.40065830168747096,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.3859,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.43412954251736147,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.3893,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.37362017086240523,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.4,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.42960670036953874,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.3883,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.4380929214605192,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.3875,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.428993917687156,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3796,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.4405001380898375,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.3744,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.5165015713292306,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.3293,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.41248329388375626,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.3321,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 0.5798487590264401,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.3226,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.4633194422904303,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.3252,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.520032923221634,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.3368,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.44354329336440973,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.3302,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.519903474255455,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.3181,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.46790994535719066,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.3187,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.4586231808550119,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.3213,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.46524427525505146,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.329,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.5510924366817987,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.3167,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.47350859234334064,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.3068,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.5481773561910597,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.3168,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.4434503221357812,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.3193,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.5258127205215504,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.3269,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.41341830774795013,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.3109,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.45121188405195817,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.3203,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.39833024873623774,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.3199,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.4387647491257373,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.3185,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.4282686927079149,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.3117,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.4142243876241788,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.3207,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.4450909106225386,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.3294,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.4943994718305774,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.3242,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.4277133155891224,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.3348,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.4483171631161261,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.3201,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.43569267015320157,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.3415,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.500834860032526,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.3251,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.46000661201546295,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.3296,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.4549487466452408,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.3262,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.3925689390783989,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.3382,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.47190047407032915,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.3243,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.4041498555055065,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.3148,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.4649023476872481,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.3261,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.45098759310951286,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.313,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.42268636964551887,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.3241,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.4710187040298196,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.3297,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.4470459634446489,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.3296,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.4553325432353688,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.3212,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.4828399713001321,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.325,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.43239206390258006,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.3134,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.5144626157847212,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.3256,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.42895765365728766,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.3102,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.485300577758171,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.3266,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.4498961767645679,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.3276,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.3706684030419022,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.3307,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.4555832944369845,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.3231,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.4074186007172831,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.3309,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.45493736390098094,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.3447,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.4101963342091739,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.3405,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.4314122716446805,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.327,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.40530834195991183,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.3271,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.43588728078150046,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.3125,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.4156236097714724,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.3161,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.4325846653452494,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.3272,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.4236257368332141,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.3172,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.3984762985218149,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.3223,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.42234961091783213,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.3192,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.41111465727777335,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.3245,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.4067404401232882,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.3265,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.40870093095929205,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.3188,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.37792018839561553,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.3244,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.40158787908224336,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.3243,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.38860438964934846,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.3017,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.4086086387230658,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.3235,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.46416968242472306,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.3104,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.40348334802013414,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.3276,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.4167341559837983,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.3257,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.4267656200926468,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.3085,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.3980931810723937,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.3256,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.43852873637665046,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.3306,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.4314332297750394,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.3195,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.4880143170222034,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.3214,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.4643693037857683,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.3375,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.4406865048198543,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.3177,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.4915020688834961,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.3382,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.42452648775348745,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.3303,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.523344740087288,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.3155,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.47250904403176924,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.3339,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.40724408160919134,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.3196,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.4394205483264067,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.3475,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.43123382738386673,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.3203,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.4179416488176803,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.3486,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.47295640555058927,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.3223,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.43103225433506,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.3338,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.46651315324379716,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.3136,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.4326185523784092,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.3256,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.44290511080391803,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.3402,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.43856099458620856,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.3199,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.4577677185314889,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.3299,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.41114459295226896,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.3205,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.44196329551461033,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.3232,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.388505649856442,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.3299,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.4187073885755892,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.3251,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.3911541886370923,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.3279,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.36601832785061733,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.3198,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.394682575132587,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.3433,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.3759369973791244,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.3367,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.4312747052036933,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.3317,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.3886009562833558,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.3281,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.4102843213763093,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.3243,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.40673543431897685,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.334,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.3917921088233052,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.3328,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.40917669105864063,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.3313,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.42318753945683046,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.3348,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.390817382027071,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.3438,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.3618432558817738,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.3311,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.38880800159094653,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.3258,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.41822013956644283,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.3296,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.4147812281780374,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.3289,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.45048471419795344,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.322,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.40694627110778475,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.3325,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.3975135418227337,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.3329,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.4074359667191951,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.313,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.3828663183519221,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.3204,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.37630025134597517,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.3366,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.3866374521933737,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.3306,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.39683356345887505,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.3035,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.39896899845596895,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.3258,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.4031601094416578,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.3313,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.3824062413725956,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.3267,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.40205346309697715,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.3343,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.39717380699665056,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.3391,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.4520367996447986,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.3147,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.37879527569765564,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.3272,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.43725994972757803,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.3277,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.4065799035765681,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.3245,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.43472136048714705,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.3235,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.37969476431995824,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.3241,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.38864992316857366,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.3424,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.39921174111573826,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.3197,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.3402224874984125,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.3234,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.3837974070464876,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.3262,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.4003433719034119,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.3234,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.4029638857822448,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.3245,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.37674371597622724,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.3052,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.3977571552587639,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.3228,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.36455170456730207,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.3305,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.38784356521419344,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.3336,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.4224209180750657,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.3185,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.3948690807579659,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.3207,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.43358160027872306,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.3217,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.4435527985469044,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.3267,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.40043993958972174,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.328,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.4434166497929061,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.3274,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.4092235747015409,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.3254,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.40741104389634325,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.3428,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.40786433612636763,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.3329,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.38793691805891894,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.3292,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.3945053736214257,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.3384,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.4206436593154239,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.3372,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.3725538368310547,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.3309,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.39800311612955597,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.3226,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.3563826991161623,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.3276,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.39086677101817563,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.3387,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.35837159990258227,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.3291,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.4240188404974845,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.3224,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.39376608855053813,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.3071,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.39372896834805005,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.3148,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.41165123892114897,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.3241,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.4006973967490716,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.3289,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.36971304320050535,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.3292,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.37183507998694526,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.3297,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.33414984804749076,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.3325,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.32299796689900934,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.3371,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.35962034205645,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.3321,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.38431521869064356,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.3423,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.3776655700669433,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.3417,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.3552385188830119,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.3345,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.4109378244991988,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.3439,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.33804775712494917,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.321,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.41134273312695274,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.3132,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.40217173074668444,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.3313,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.36907943083705785,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.3214,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.3821659098010914,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.329,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.36988099016602904,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.3288,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.3739892190722599,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.3359,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.3929333136040789,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.3225,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.42854233293457394,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.3296,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.36835320323009596,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.323,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.35538583097984394,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.3151,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.37484659163731004,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.3441,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.3751274741102272,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.3321,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.3761176775600002,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.3219,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.37436314873656945,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.3475,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.34635641078255136,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.3193,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.36006631060414296,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.3338,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.3996720574593829,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.3332,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.40501950390870145,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.3436,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.42656464212456635,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.3183,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.39117148826642123,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.3248,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.374836832859297,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.3317,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.42028372025754523,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.3272,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.33698284461979294,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.3353,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.37945271743151504,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.3194,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.36416567548909506,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.3078,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.3583734756280559,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.3166,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.3628206887928804,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.3324,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.355431833268649,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.3169,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.39924615390594687,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.3268,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.34496574768819427,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.3182,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.3852524200875301,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.3338,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.3576477028444599,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.3276,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.3667358722262798,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.3263,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.38252093426847567,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.3325,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.3552020087188636,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.3207,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.3532700896833334,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.3392,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.3538558027380506,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.3239,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.3768244276047353,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.329,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.41791264193627486,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.3276,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.33893654496855524,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.3395,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.3697921863981474,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.3283,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.40329091140341844,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.3122,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.36719720495943114,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.3314,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.38985241750891186,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.3238,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.3789652081372114,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.3251,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.3773342763510798,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.3178,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.37841466530676704,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.3156,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.3881531398270112,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.3243,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.36331916728874486,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.327,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.377385422657216,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.3255,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.36887927160784223,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.3355,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.3853204650671251,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.3179,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.3625115649569671,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.3331,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.3507784948999875,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.3295,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.36656704062620526,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.3081,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.3447592513265888,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.3168,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.3705994326726648,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.3192,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.3512266903399639,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.3249,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.3591969441523364,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.3299,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.34784665759902217,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.3289,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.3736033365993856,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.3246,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.3432493849809737,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.3187,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.3744316819609724,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.343,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.3657546793545248,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.3218,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.35364221146480346,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.3347,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.3729453905487406,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.3285,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.365349310266993,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.3331,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.3843708305881172,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.3302,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.39830157131941085,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.3282,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.3546204375556284,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.3327,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.37033806239919453,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.3259,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.35685054473621847,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.3332,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.3954207502248193,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.3164,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.3969802025366344,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.3232,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.3465319834685161,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.3176,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.40417413925345214,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.328,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.4347678845713465,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.3019,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.4274096653587594,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.2527,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.626166714927738,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.2605,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.4887941145207756,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.2568,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.4841002381593789,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.2685,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.5415684230751536,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.2577,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.41182437752824114,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.2546,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.4673044269300216,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.272,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.4433982846756233,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.261,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.41232688023639985,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.2614,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.3650674418352713,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.2601,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.4323212187290974,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.2327,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.4022672564631653,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.2574,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.38202444258528023,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.2651,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.4077291279627586,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.265,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.4180833726099066,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.2569,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.4263593903854492,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.2502,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.40103766909882516,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.2388,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.39951140545809416,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.2522,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.4373896379572844,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.2493,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.3609871280052099,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.2571,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.4222715758693231,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.2624,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.3719169565473931,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.2558,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.35415947873659226,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.2533,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.4086076275733911,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.2618,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.4082759722282498,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.271,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.37580477903902126,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.2552,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.4155085418175544,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.2463,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.43849123369402415,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.2417,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.3787594965514273,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.2717,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.3966795161729525,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.2632,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.4205779900024952,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.2596,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.4032624074635962,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.2661,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.4226453489545353,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.2553,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.39690010038644846,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.2558,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.3681050384988563,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.2643,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.41246162036679285,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.2401,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.38985778490799183,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.2713,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.3955541171206844,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.2707,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.38611512820423816,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.2458,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3613327754730746,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.258,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.35852609003145464,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.2572,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.3279402725857431,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.266,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.3724537778663484,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.2529,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.36853707745375325,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.2566,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.3613550138053133,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.2538,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.3874952941456213,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.2568,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.37650523773360056,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.2562,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.341420274905444,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.2564,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.3528844635187211,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.2728,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.3746665501440213,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.2638,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.35747965378220653,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.2426,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.3808515134709718,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.2629,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.33532483973580846,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.2572,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.36278292422486097,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.2615,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.3416663373132735,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.2469,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.3600591871400527,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.2367,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.3637742925328942,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.244,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.3643599265923125,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.2573,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.37758617778769493,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.2635,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.34867822006705085,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.2616,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.36283088362887694,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.2506,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.35743027701987506,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.2727,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.44131125162795287,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.2645,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.3419339899850311,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.256,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.3523688495182964,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.2496,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.35836498991766325,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.2608,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.35620412250883027,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.2578,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.35471484807086695,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.2579,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.35574924520494783,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.2595,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.3295499569442109,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.2615,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.3585118192053597,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.2684,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.34009773881891064,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.2566,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.35904909467845036,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.2475,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.3562599631183793,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.2407,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.3497715309255977,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.2546,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.3765202231536961,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.264,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.3966416917914038,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.2451,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.33672764550882706,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.2524,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.37429739866995293,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.27,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.3485774990486502,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.2471,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.3683738581498369,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.2697,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.37908543434362124,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.2572,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.3525664035844728,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.272,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.34081801691453173,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.2631,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.381439239497038,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.2636,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.354992727198453,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.2661,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.34403055454897147,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.252,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.3729339728372235,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.255,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.36128529847229973,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.2399,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.36008780219312236,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.2416,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.3394153973781896,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.2674,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.346239971989748,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.2645,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.3720795397777903,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.251,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.35626492983663416,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.2729,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.3350047453049649,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.2805,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.3487187828762305,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.2602,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.3601396342732347,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.2661,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.3419892671582826,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.2507,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.3658391787187104,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.251,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.3687373637723782,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.2682,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.3436687464001311,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.2551,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.34301286428757677,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.2556,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.33025564444766403,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.2594,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.36163514654335427,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.2592,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.3446815842396628,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.2562,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.33690873993473086,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.2579,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.3400747941950399,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.2583,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.36630969918786405,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.2399,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.3674159458284236,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.2585,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.36608805143426687,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.2561,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.3730588677597406,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.259,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.36141229603546526,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.2516,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.3529575242557733,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.258,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.34842306957628366,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.2744,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.35779897641988484,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.245,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.37221818997059813,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.2543,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.3351066930397617,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.2695,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.36359059934965027,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.2433,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.3532466784583059,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.2502,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.32131587774532766,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.2688,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.36119925075829795,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.2539,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.37384372320278236,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.2419,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.3509534854980787,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.244,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.34612368265263904,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.2539,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.383865867939247,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.2466,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.35311088908490756,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.2656,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.33986671809401187,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.2614,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.364624144605724,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.2562,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.3393385181646463,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.2602,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.3474766232971165,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.2651,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.34129926444080083,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.257,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.3357112645613674,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.2561,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.3563380825580113,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.2598,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.34526605114553616,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.2559,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.3260140476710736,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.2586,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.3226227355822354,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.2601,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.331282944899843,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.2467,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.34476548626069603,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.2533,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.3364004170782678,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.2607,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.37639053282730806,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.249,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.36030466456768523,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.2516,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.350828424328476,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.2423,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.34308105368651554,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.2648,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.34178164478175627,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.2594,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3566643357759728,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.2341,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.3624485061104329,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.2495,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.34149110295834845,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.2522,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.3387733395187681,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.2498,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.34492378491644815,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.2664,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.3857365056293913,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.2559,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.3388501604055655,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.2556,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.35644761628757077,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.2533,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.331290378999541,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.2597,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.34655286244626154,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.2684,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.3343531113693397,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.27,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.3340748942446601,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.2559,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.3420798739053651,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.261,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.3656292913907304,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.2426,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.3416345779578955,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.2513,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.34573102874049755,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.2631,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.34245727887485783,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.2567,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.34507093283619067,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.2595,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.34123525790171794,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.2568,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.3619115972322085,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.254,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.33204290796814373,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.2646,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.3284750803415239,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.2587,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.3442225559871413,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.2383,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.35402945297900285,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.2544,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.332159972849458,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.255,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.3231336877512513,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.2485,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.31420393669015295,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.2576,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.3165790287902733,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.263,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.3251125458103973,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.265,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.3196107975091023,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.2588,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.3418889696083659,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.2557,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.325388447627888,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.2534,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.32632291617101616,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.2516,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.3230147795628839,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.2528,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.35235573266681214,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.2506,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.34298256796024856,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.2571,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.31393864000400473,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.2593,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.32278049567274486,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.278,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.3199725233930355,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.2561,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.3244394048049322,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.2502,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.33192961036699287,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.2568,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.31693842329506683,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.2545,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.32126855526648174,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.2655,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.33280926806664324,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.2621,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.349893225995229,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.2574,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.33845440238705865,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.2639,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.32734561605866924,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.2536,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.3168437070087709,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.2572,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.30006270706630767,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.2732,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.3200000389690007,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.2556,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.34979146805632516,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.2482,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.3485384521119159,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.2425,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.3492939754874566,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.2636,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.3332354805474273,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.2502,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.32185488131989515,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.2476,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.32475346193907156,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.2482,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.3194438476270994,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.2558,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.32420317711737234,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.2657,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.33236573191577995,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.2706,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.3130657663948901,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.2619,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.30949724113607624,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.2538,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.3179652267941201,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.2581,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.33439602703048066,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.2685,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.3206175573139513,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.2674,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.3071292414077877,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.2699,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.3455490213110419,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.2401,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.30959810556953216,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.2611,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.34559586542847787,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.2348,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.32583388084078113,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.2564,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.31503817075753915,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.2411,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.3166138855290963,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.2528,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.3038966224550739,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.2674,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.34288806225168444,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.2373,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.3233658294969492,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.2441,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.3273167628304312,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.2622,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.34539233127653257,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.2427,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.33686731101279677,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.2409,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.3262225190158217,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.2596,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.30671619830226704,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.2359,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.3422361479288116,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.2475,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.3206450206345261,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.2558,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.32007345386307234,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.2491,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.31848163323947054,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.2471,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.3088856745871696,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.2581,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.32760065751959583,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.26,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.33290468916354543,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.2542,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.32352621017308514,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.2398,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.3371443583236499,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.2528,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.3241835050445812,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.25,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.3185447305029052,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.2591,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.3070894691080085,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.2669,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.3388927447603726,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.2489,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.34702198130528833,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.2472,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.32227638373032214,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.2501,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.32868811861580255,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.2494,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.36035786644318507,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.245,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.3218528086663044,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.251,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.32707856718362255,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.2483,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.31096157453317536,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.2506,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.3392469010382586,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.2464,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.3228874373908619,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.25,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.32325600769615004,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.2353,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.4475319115909968,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.2108,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.434562236101355,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.2001,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.32827687048585386,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.195,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.39451956221097345,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.1924,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.5904272003043195,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.2064,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.5012158546486669,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.2123,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3358405085848931,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.2092,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.42634689156262817,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.2003,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.4687602655757464,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.201,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.4210711755289249,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.2147,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.32078231561710935,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.2027,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.35086860345529,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.199,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.39997931594044367,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.1983,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.4218185901083457,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.1954,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.35444283754868583,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.1901,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.2991907919620863,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.1906,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.3125244680244546,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.2238,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.34329804610233866,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.2008,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.3445431082375558,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.2045,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.3121996389652113,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.1903,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.31345529412802603,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.211,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.3391035350053192,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.2108,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.3445610004599466,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.2088,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.31070823111882595,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.2108,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.3113401288678123,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.1877,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.32206602587828725,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.207,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.3035240085785448,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.2103,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.3251129354086264,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.17,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.3035208875746999,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.1879,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.28487265297809233,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.2184,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.29979347510818966,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.2102,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.3096330372720498,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.2133,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.3184956574961,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.1933,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.3028934067527914,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.2073,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.3269552084357353,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.198,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.33713404544810993,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.1987,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.3092469953005195,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.2081,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.33101564347869095,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.1842,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.3392230679744529,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.1893,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.3104919473281154,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.1888,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.32903644751846706,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.2039,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.3670072436397523,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.1995,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.31749936053528366,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.2015,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.2970239194716737,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.2168,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.31022516315682697,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.212,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.2986185339851442,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.2122,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.2945590657235354,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.2095,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2974070424685937,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.1961,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.29845690031134586,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.1942,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.3093743445252386,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.1981,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.3134754505600547,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.1968,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.35571384077099466,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.1766,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.30849481900310205,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.2067,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2932523938955792,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.2043,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2870022602256085,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2132,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.3137133821255253,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.1866,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.31445636193240967,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.2044,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.29039968908058117,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.2202,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.3363212717663999,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.202,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.3101982167902651,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.2001,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.29771878978728017,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.2152,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.3082704792199714,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.2104,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.29562856293049894,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.2056,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.320013438649321,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.1993,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.2985032901442859,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.199,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.29130888407167427,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.203,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.30067864501268937,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.199,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.3026082516051964,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.1944,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.3002029782717318,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.1826,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.3002132801975866,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.2138,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.28871860826154205,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.201,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.3021237817303886,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.1903,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.3097234625881696,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.1935,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.2941668592096005,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.2002,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.31535645326580053,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.198,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.29900353822587145,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.203,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.28698529253747196,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.2074,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.2910119864940706,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.1928,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.287527811953485,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.2085,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.2958577067392289,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.2067,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.32012315042353634,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.2017,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.31543318382944846,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2004,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.31907962008203783,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.1908,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.29105516821158917,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.2064,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.2863381141670873,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.1956,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.28533608883942596,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.209,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.28697127566312264,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.2109,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.29956105327392796,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.1851,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.28607482298097126,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.2046,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.31717067667602483,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.1917,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2968109762268536,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.2145,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.2857989913320283,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.2282,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.30363933606588894,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.1956,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.2951546797618115,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.2155,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.30617981994930377,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.1964,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.2856332863001102,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.1993,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.3021306505438966,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.1893,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.2911336379086679,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.1917,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.2998310224674077,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.208,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.3021788698450773,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.202,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.28633766384281134,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.189,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.28708223709972164,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.2043,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.28035322685852104,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.2108,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.2984201621686926,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.198,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2897846628968472,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.1896,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.29429554154651766,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.207,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.30732935189506283,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.2127,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.29548312548560257,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.2017,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.3035669261991647,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.1776,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.3173238504208833,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.1918,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.2893266251233025,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.204,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.2922233039736376,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.2016,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.31013951891703934,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.1895,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.30070708870051316,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.1899,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.297552536396352,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.2167,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.30163732045305086,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.2013,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.3192761659175723,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.2098,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.29804539102720035,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.1901,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.2825936467314246,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.198,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.3024380665327541,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.1979,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.2782871385211981,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.2112,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.2834413724853452,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.194,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.28529989638542197,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.2107,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.30546440224328103,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.1897,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.3047179981531733,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.1865,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.3521858175177294,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.1992,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.2994892630707705,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.2105,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.29353309240121295,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.1886,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.30173593905823304,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.1969,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.31861451170555116,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.2031,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.27023839752783807,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.2207,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.29483876721387153,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.2092,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.31127522561940807,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.2085,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.2790133122831481,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.2079,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.30768639694306527,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.1874,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.2883710667731071,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.1978,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.2906332137420836,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.1914,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.3111468135505344,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.1938,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.2992318219726616,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.1904,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.3032603467764455,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.2108,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.29113162611992577,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.1986,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.3599431116419085,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.2141,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.30908367396783976,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.1862,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.29613645353107626,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.2126,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.29138512635295766,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.1885,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.28887342292013896,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.1888,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.3053359630937725,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.1895,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.2926758039440732,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.2111,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.2761429693228098,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.1998,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.2857148219913376,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.1939,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.28671707077622305,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.2193,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.3072268451315383,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.1906,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.29015467269120004,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.2189,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.30062276556652584,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.1934,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.2942368980698579,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.1996,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.2948489162075722,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.1936,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.27992227509692186,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.1988,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.289232029045811,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.2047,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.28792965784897845,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.1944,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.28013408970221243,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.1842,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.2907800879772052,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.2181,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.31574652680408366,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.1709,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.2961670317136961,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.206,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.29635801782911547,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.2132,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.31124773726254823,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.1922,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.29141643518359067,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.1956,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.2788675628613129,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.2138,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.28429972573008844,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.2134,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.3048581869837047,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.1754,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.3027628456214814,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.1867,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.31470916330074905,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.1947,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.26873668733356815,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.2106,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.29654406221238166,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.1911,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.2836274665773814,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.2008,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.27152528459698855,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.2028,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.3089897835635458,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.1841,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.33667986347506934,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.203,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.26660264231299374,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.2042,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.3208795224907666,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.193,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.2949516571262024,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.2005,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.2862795965902875,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.2066,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.27728424901599236,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.1885,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.2954119619560151,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.1986,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.2896205609215128,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.2192,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.2879959384602161,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.199,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.2993469436499246,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.2085,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.31277712355660503,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.2004,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.3107066733506445,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.1939,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.28872968035926877,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.1971,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.3063528796292699,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2209,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.2920169603196883,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.1873,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.2742472882477973,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.2077,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.30063941999333477,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.1859,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.2926660123052196,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.2044,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.29086099134342186,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.1892,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.2892574989568626,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.2078,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.29239061520985177,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.2005,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.2842862536167715,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.2097,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.28560863992578644,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.1968,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.2647850647143467,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.215,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.2872151329865877,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.2132,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.2923473585172447,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.2021,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.27498364637726846,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.2157,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.28970432815853486,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.2132,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.2940606718394336,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.1973,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.2979187022036254,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.1897,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.29195767055941246,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.202,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.2847388845366988,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.2047,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.281485150785538,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.228,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.29907639136685454,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.189,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.28045236869338247,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.2049,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.28128915098181184,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.1981,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.28934929524619984,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.1948,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.28776466706400844,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.1964,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.2692448038974924,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.223,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.3048078236684754,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.1972,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.2900819252788125,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.206,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.2932020349952676,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.1934,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.272288866540082,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.2132,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2717444899061312,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.2043,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2879720931370659,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.2128,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.2714174988602119,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.2156,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.2808880620068754,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.1953,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.28568766961067016,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.2014,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.2983427528044573,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.2024,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.279084757821102,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.1995,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.2889283359896976,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.2044,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.2929538566809365,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.188,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.299715010489614,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.1835,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.29193423797807494,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.1855,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.30110237373368554,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.2015,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.29334535805649725,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.2097,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.3065812477604011,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.2009,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.29915060264470333,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.1952,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.3107592515443552,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.1923,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.29356337914704056,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.2063,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2931159492487582,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.1986,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.3035325211127472,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.2039,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.2823500713285082,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.2056,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.29371239841117314,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.212,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.262352855561308,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.212,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.28797105178538607,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.1741,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.3204619541177409,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.1899,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.2871784236261053,
+      "learning_rate": 0.0,
+      "loss": 0.1931,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 2.2440850871277322e+18,
+      "train_loss": 0.33498664268148626,
+      "train_runtime": 81407.8422,
+      "train_samples_per_second": 1.941,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2440850871277322e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}