diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.982278481012658,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 9.715104192838567,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.8634,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 9.911585576660412,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.8422,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 9.622975864409776,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.8501,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 9.476371439556594,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.8368,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.90924205239641,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.7752,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 8.741907492397482,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.7988,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 7.1296712920618175,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.783,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 6.653729293904294,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.7563,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 4.196578861950112,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.7077,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.678511640849096,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.6872,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.533741073353499,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.6927,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 6.162375910061681,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.6724,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 6.4189312512241585,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.6925,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 6.300318803735138,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.6475,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 5.780575281307569,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.6396,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 4.484434683552672,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.6096,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 4.218687934158408,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.5671,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 3.679852773056102,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.5755,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 2.748445413163893,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 1.5183,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 2.1825579334187295,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 1.4938,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.9795669371004145,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.4827,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 2.0206299186852066,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 1.4864,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.9863419154991722,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 1.4639,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.787948657963244,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 1.457,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.3227038495536836,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 1.4597,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.6693426693644067,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 1.4269,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.7206711719560108,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.4157,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.285275735897094,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 1.4278,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.2308164849010794,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 1.4235,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.2382199152826892,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.414,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.1585976951753023,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 1.4275,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 1.1756402813461437,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 1.4174,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.0979197384818362,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 1.3913,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 1.0387104017568338,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 1.3676,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.9455077334163449,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 1.3581,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 1.1731969511738167,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 1.3552,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.8113854899147275,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 1.3607,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 1.0248023463311766,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 1.3551,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.918172586044147,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 1.3738,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.840940978191582,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 1.3678,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 1.01651941979176,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.376,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.8463296763724326,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 1.3688,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.9132829147138898,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 1.3524,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.8923237886173885,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 1.3311,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 1.035880503308946,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 1.3493,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.8475922259865425,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 1.3325,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.9289265447492482,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 1.3787,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 1.1923695239239502,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 1.3663,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.918329521878312,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 1.3414,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.9167005028544347,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 1.3526,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 1.2565633712018978,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 1.3289,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.8927500139545372,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 1.3394,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 1.2695854744505943,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 1.3226,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 1.1988225962809922,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 1.3552,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.81625549040345,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 1.3178,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.9939486324660751,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 1.3283,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 1.1734938016288479,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 1.3587,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8434938890356535,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 1.3069,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.242084052271512,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 1.3541,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 1.245319081014771,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 1.3352,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 1.0781299500526325,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 1.3306,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 1.6038434131982768,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 1.3062,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 1.1222269715679671,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 1.3562,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 1.5645079851713324,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 1.3482,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.8627715323086304,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 1.343,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 1.6501126870353664,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 1.3187,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.1043081299667419,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 1.312,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.4033983840202993,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 1.307,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.1598676365731848,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 1.3181,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.067636758411421,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 1.2947,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.0993668096292555,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 1.309,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.8059351275878893,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 1.3273,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.2263763073778382,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 1.2985,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.0714410350592891,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 1.3228,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.0075717286465191,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 1.2968,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.4509788055110615,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 1.3148,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.2453015758261503,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 1.2883,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.9671786314173614,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 1.3345,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.342049115212934,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 1.2617,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.3927981280238666,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 1.3038,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.0303551856642637,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 1.2992,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.959483734297538,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.3099,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.2155467275672776,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 1.3007,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.7735685101417273,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 1.3142,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 1.5464951648292034,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 1.2985,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.925292222620893,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 1.288,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 1.5279682680751128,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 1.3005,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 2.269836019614837,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 1.3213,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.9965778565528647,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 1.2729,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.9093755725734858,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 1.3043,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 2.066192919997544,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 1.2993,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.490696728894191,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 1.2993,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.53157489952652,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 1.3023,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.415751832618074,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 1.2946,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.4540347124898405,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 1.2856,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.5393845058164741,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 1.2863,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.149443120612095,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 1.2873,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 2.0340705261343275,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 1.3134,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.5693755863816654,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 1.3002,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 2.2877715322491583,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 1.3095,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 2.1468004509887155,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 1.2776,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.575052445671897,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 1.298,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.9863412405799905,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 1.3085,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.48404389243251,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 1.2889,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 2.1615039300456638,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 1.3198,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.5412241533652897,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 1.3025,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.408704491006043,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 1.2777,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.4664512862379147,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 1.2747,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.2018701881971916,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 1.2871,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.4601431141765921,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 1.2818,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.3640546997979381,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 1.3036,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.1435781950400508,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 1.2978,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.828278838502671,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 1.3044,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.2053202581875218,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 1.296,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.2575620567239498,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 1.2708,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.9788632977404473,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 1.3137,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.3887485838143219,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 1.2877,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.7567379361792024,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 1.2732,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.06887918096504,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 1.2814,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 2.578245055267917,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 1.3022,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.7808890015103866,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 1.2914,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 2.027253951198519,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 1.3102,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.642671026265804,
+      "learning_rate": 4e-05,
+      "loss": 1.2966,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.6796435985792892,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 1.2884,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.2097846408160913,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 1.3303,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.3788723192964962,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 1.2934,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.3137112007006764,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 1.2974,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.5334960274489062,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 1.3053,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.3393741809854856,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 1.2845,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.1618529946008669,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 1.2958,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.6983065645897666,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 1.2944,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.157597412455993,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 1.2783,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.3390510572156864,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 1.2602,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.4457620096452073,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 1.2769,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.0610168088050171,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 1.2603,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.6157296786946838,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 1.2628,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.02864260569558,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 1.2753,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.6133931576352036,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 1.2854,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.1077413783208447,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 1.258,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.5383742494913681,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 1.2932,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.4727577991785819,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 1.2809,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.2917829824424811,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 1.2762,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.3670929483908698,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 1.2621,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.2588547656000513,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 1.2837,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.1056674023152175,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 1.3086,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.2535368595908158,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 1.313,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.038903279617118,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 1.2863,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.1640058579848909,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 1.2563,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.4114954649215596,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 1.281,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.0427880981353301,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 1.2552,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.5263822997877614,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 1.2653,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.0361652765643976,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 1.281,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.4209947630556483,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 1.2981,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.3268107697027212,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 1.2817,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.8406202693392815,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 1.2671,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.1186327928537674,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 1.2857,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 2.627939953345451,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 1.2795,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8669426369636115,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 1.2682,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 2.0924928100265725,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 1.2548,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.7284246703441821,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 1.284,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.8862541141212823,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 1.2853,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.4681170256975962,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 1.2821,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.581524185498311,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 1.2799,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.4111259960208553,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 1.2378,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.016158961058217,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 1.2812,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.7326288930349019,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 1.2596,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.0523431956554898,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 1.2777,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 2.484602792934271,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 1.2746,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 2.576387676500367,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 1.2566,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.2877609856649115,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 1.302,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.8189295442736297,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 1.2867,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.8177933355981608,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 1.2763,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.2070847734969763,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 1.2786,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.8073848417649256,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 1.2801,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.5511564595670924,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 1.275,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.50201284121554,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 1.3158,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.2406053749311967,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 1.2761,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.0359951507189182,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 1.2957,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.665717517276046,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 1.2825,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.0939378695729896,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 1.2645,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.8996880865759813,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 1.2622,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.3454410760007605,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 1.2794,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 2.160741060180749,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 1.2891,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.8743489421824067,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 1.2905,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.5413020517580962,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 1.2616,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.6446236275566175,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 1.2739,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.3153917922011138,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 1.2676,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 1.0412985206183407,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 1.2573,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.5979134367949068,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 1.2629,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.0983253471180252,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 1.2681,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.7103901569429236,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 1.2688,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.3682203491555207,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 1.2748,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.535255159601047,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 1.2769,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 1.211219582013992,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 1.2662,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.6593336415271993,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 1.2712,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 1.238067471697759,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 1.2769,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.975570392329248,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 1.2839,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 1.8527007964941744,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 1.2665,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.2281690617369518,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 1.2638,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.734560283863862,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 1.2925,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.0339641451792445,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 1.2478,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 1.4671259237647591,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 1.2783,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.8985669229068189,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 1.2409,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.9888484762384204,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 1.2806,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.9002992508295106,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 1.2689,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.2545358054051694,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 1.2901,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.9317195557611272,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 1.2644,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.3671164037424268,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 1.2692,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.9839238618408674,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 1.2611,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 1.242747950440388,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 1.2602,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 1.0918128562274474,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 1.2979,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.8975029012850251,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 1.2575,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.9461686333690209,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 1.2698,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.8859725723253069,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 1.2569,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.1124002461096512,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 1.2876,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.9606962620228567,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 1.2623,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.9817655846507902,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 1.2836,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.0203290879498232,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 1.2611,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.9034469136827384,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 1.2409,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 1.3692322767579805,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 1.2422,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 1.0044045843951204,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 1.2581,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 1.4176069844607677,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 1.2334,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 1.3524113892173744,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 1.2828,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 1.0772132478270549,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 1.2484,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.8448111510262327,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 1.2746,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.9992110309174586,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 1.2471,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.848566562714919,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 1.2804,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9882320870071724,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 1.2509,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.8155292163815935,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 1.266,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.9884543775765444,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 1.2604,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 1.04211553290313,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 1.2703,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.8667521433901875,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 1.2711,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.8891327835561422,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 1.2756,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.8597267110596573,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 1.2861,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.7286176618228102,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 1.2483,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.9669521292038903,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 1.2778,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8034073567527337,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 1.2599,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.9475511210762485,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 1.2705,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.7514595454491779,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 1.2597,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 1.0119243687176946,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 1.2845,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.944494081932996,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 1.2814,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.2054650480242848,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 1.2989,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 1.0009631464161501,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 1.2556,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.8976846174830309,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 1.2678,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.846125171005309,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 1.284,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 1.0120174766648344,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 1.2756,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 1.0044855005427977,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 1.2595,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.9257934407230873,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 1.1979,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.9385507036792304,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 1.1886,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.1108114591434843,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 1.2033,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 0.9561521176903772,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 1.1716,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.9719985279659247,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 1.1999,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 0.8660983300595189,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 1.1825,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.8455190720254618,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 1.1827,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 1.1286377881514993,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 1.1893,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.9940622239368543,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 1.1719,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.7610931353150544,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 1.1966,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 1.0122367478233243,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 1.1842,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 1.3054962924972184,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 1.2081,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.7737579110256042,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 1.1825,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 1.0574491662795142,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 1.1786,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.8852135279084209,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 1.1896,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.8946246766800718,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 1.179,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 0.8519985192988974,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 1.1882,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 1.1714134603358304,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 1.1768,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.8025793112774113,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 1.1912,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 1.156572975725374,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 1.203,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 0.8940860769139252,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 1.1678,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.9285487311303859,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 1.1811,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.7688735712401837,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 1.1965,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 0.8397542592079714,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 1.1846,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.6822559854906327,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 1.1869,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.7951116872851901,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 1.1668,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7567010439727854,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 1.1763,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 0.9666574047377483,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 1.1378,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.6729665188191238,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 1.1782,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.9025021212934988,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 1.1848,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.6669252907978925,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 1.1954,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.7713831062326728,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 1.1495,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 0.6206272587852482,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 1.1824,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.719331189929342,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 1.1605,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.625819114922838,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 1.2127,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 0.7496072003002134,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 1.2047,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.8149915562870599,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 1.1858,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.7566082065339276,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 1.143,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.7376345508882559,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 1.1898,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 0.8265166058163242,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 1.1772,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.7185951013961922,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 1.198,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.7779607140592167,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 1.1676,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.629575737422095,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 1.2042,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.8120473265885314,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 1.1676,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.7070403330881265,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 1.194,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.7263495405218359,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 1.1786,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 0.8492185562777069,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 1.1684,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.861850360331744,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 1.225,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.8159368292092508,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 1.1945,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 0.7259008306032689,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 1.2045,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8409972906106773,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 1.1928,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 0.8130633264344819,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 1.152,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 0.6247797533251003,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 1.1837,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.7472491232361467,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 1.1674,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8953915172504486,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 1.1897,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.7196164429009246,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 1.2021,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 0.777511972704057,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 1.1809,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.9971950055513531,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 1.1982,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.6454012902616019,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 1.1652,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 0.7635770937103951,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 1.1637,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.6266795624540816,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 1.1847,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.6091835434146365,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 1.187,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 0.7404695279355521,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 1.1858,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.6024666559046816,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 1.1789,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 0.6639626266763324,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 1.1875,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 0.6119334698205803,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 1.1788,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 0.5987317860695961,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 1.1857,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 0.6533482079650624,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 1.1809,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.646615891140646,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 1.1853,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 0.7315245724800346,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 1.2017,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 0.6708311590996542,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 1.1831,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.6232064514911418,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 1.2012,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 0.6464225928921215,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 1.1499,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 0.5573847127739832,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 1.1841,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 0.6864778061137569,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 1.1803,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 0.6426372656672258,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 1.1708,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 0.6592395738587772,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 1.1807,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 0.6957188290683738,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 1.1451,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 0.7226144819392062,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 1.1907,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 0.6961710169431706,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 1.1717,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 0.6167577243586291,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 1.1878,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 0.798519347154792,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 1.2024,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 0.6030729904314637,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 1.1704,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 0.8419938531167678,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 1.2108,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 1.0065151731588795,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 1.1989,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 0.8950582083896245,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 1.1772,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.7910000913532015,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 1.1918,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.8559698444143072,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 1.1887,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 1.0015818991375978,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 1.1899,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.7152195736948364,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 1.2094,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.7925610867796716,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 1.2046,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.9361833068644413,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 1.1578,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.827154824239697,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 1.1671,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.9796061431094188,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 1.1852,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.8843756434771662,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 1.2217,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 1.009485260909672,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 1.1949,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 0.6934655096536804,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 1.1652,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.8633624873456659,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 1.1916,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 0.9708171807922925,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 1.1878,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.5971790241003766,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 1.1777,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 0.8695034029565665,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 1.1794,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.7244994821851654,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 1.1832,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.7583157613019809,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 1.1727,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.8774202751572704,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 1.1786,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.8569471563145515,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 1.1857,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.8944803519678151,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 1.1546,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.7133464038078888,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 1.1763,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.8033216678539494,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 1.1779,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.6430204318187684,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 1.1928,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.6261252371540562,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 1.2003,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.6240346084233178,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 1.1827,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.514534766063522,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 1.1835,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.6692088963162337,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 1.1954,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.5765087742514654,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 1.175,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.715841231765026,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 1.1799,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.6768539653676214,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 1.1742,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6827707049823684,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 1.1783,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.7869192403089184,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 1.1583,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.7232731490894334,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 1.1863,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6539838106173177,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 1.194,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.7895263425617576,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 1.1905,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6163164433955507,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 1.1836,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.8569954815094917,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 1.1518,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.8095276615961724,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 1.1809,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.6728178291968815,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 1.1678,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.7231664132172915,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 1.1669,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.5939282535341054,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 1.1835,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.7529107454954792,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 1.1704,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.6786427537336858,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 1.166,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.641354527243491,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 1.1756,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.7861149278433102,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 1.2025,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.8383130004159864,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 1.2076,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.6028478019350785,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 1.1774,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6589009918381309,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 1.1945,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.6051484027164696,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 1.185,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.5720017204978837,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 1.1868,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.6986426339797741,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 1.1878,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6618833015879825,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 1.1843,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6851124075963936,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 1.221,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.7361538623820955,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 1.1805,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6917111480479332,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 1.1769,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.6928920968874592,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 1.1787,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.7423097498813875,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 1.1736,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6102792703707479,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 1.1812,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.8162103674385514,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 1.1838,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.6315256067820583,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 1.1934,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.7657604056827205,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 1.1764,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.612451714668804,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 1.1757,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.7942117010252752,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 1.1657,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6050426581034491,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 1.1724,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.5951394246680042,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 1.1767,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6593812609526661,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 1.1994,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.5103854380749303,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 1.1724,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.6296937666759395,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 1.1893,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.5690306160581692,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 1.1847,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.599543480602089,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 1.1381,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.5425542071667665,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 1.1831,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.6636590976704811,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 1.1839,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.5953921709209026,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 1.1542,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.71633255368306,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 1.189,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.6224392459325679,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 1.1943,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.6443639557807694,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 1.177,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.5670820570424471,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 1.1608,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6414396268399284,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 1.1915,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.5385273205176623,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 1.1814,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.6302582390815485,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 1.1573,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.5956670190025852,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 1.1804,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.5806894016664604,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 1.1935,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.5941832880614853,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 1.1832,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.643633337803752,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 1.1941,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.5704687163861993,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 1.1841,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.5934108522032502,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 1.1942,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.6464101263975431,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 1.1747,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.5552299275299679,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 1.1633,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.6172121147748008,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 1.1952,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.6041663097945246,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 1.1798,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6996872002847964,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 1.1787,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.7032534532445651,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 1.1973,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.7046706031288906,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 1.1787,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.6959080079465855,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 1.1494,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.7107068090861055,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 1.202,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.7344839209377692,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 1.1912,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.733243115947052,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 1.1763,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.5321138885690903,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 1.1658,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.8427032504109747,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 1.1891,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.7168278131089595,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 1.1833,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.6487034573876749,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 1.1773,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.8851441416867695,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 1.199,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.7612561040051353,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 1.1727,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.6946469656137408,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 1.1706,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.8833852669800081,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 1.1778,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.5822763553034958,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 1.1769,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.8264470184370394,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 1.1658,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.7086519722628639,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 1.1781,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.5701814558719505,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 1.1923,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6075237022505356,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 1.1643,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.5822617472718175,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 1.1585,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.5963879809817774,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 1.168,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.6166755434315879,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 1.212,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.6165232356421468,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 1.1776,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.6598200360629504,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 1.1854,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6630784448428786,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 1.1821,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.5590557047794834,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 1.1774,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.6931680699950684,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 1.2038,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.8571993504405301,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 1.1915,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.5705145833051881,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 1.1872,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.7458666933758776,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 1.1963,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.6875028687205327,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 1.1686,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.6266372830735896,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 1.1806,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.7304492047732604,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 1.1875,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.5467705114234771,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 1.1844,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.6231354312584513,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 1.1726,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.6099939585506633,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 1.1864,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.6606864827729262,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 1.1698,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.5978984711922253,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 1.1762,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.6523347653839395,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 1.1923,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.5281637008387843,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 1.1721,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.7323113275470378,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 1.1801,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.4939478870195709,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 1.1711,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.5383179052450504,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 1.1804,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.5377880692756607,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 1.1901,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.5071515275087702,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 1.1756,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.61571820961762,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 1.1777,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.4325747582555546,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 1.1677,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.695137016280725,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 1.2148,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5188874173359198,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 1.1735,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5433891213862343,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 1.1888,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.5254782381548688,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 1.1701,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.5802460339425025,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 1.1859,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.5650731191083244,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 1.171,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.62454181692941,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 1.1751,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.5832976965537633,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 1.1887,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.6759201195489072,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 1.1828,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.5207823206258737,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 1.1558,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6539150857202253,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 1.1734,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.6029186463217104,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 1.1875,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.5518269363759966,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 1.1808,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.5708745573395265,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 1.1895,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.525665278781853,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 1.1551,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.6470463685239779,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 1.1896,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.47620717625564946,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 1.1711,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.6892868720148013,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 1.1769,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5322359174565467,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 1.1631,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.5423260323309932,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 1.1635,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.5708625397223802,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.1804,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.4914476376401023,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 1.1686,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.544169608524135,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 1.158,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.6924483464859719,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 1.0408,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.8037762203588966,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 1.0298,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.6593077811523268,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 1.0588,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.9811590833349273,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 1.0381,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.6850589336068318,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 1.0158,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.8428435195226353,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 1.0698,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.7898871402574772,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 1.0213,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6881229222321587,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 1.0545,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.7700888258080218,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 1.0212,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.7738320243495719,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 1.0298,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.7751494379837556,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 1.0201,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.9949624797324408,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 1.0686,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.5575494032438036,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 1.0443,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.7748253321322293,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 1.0189,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.5222427918935587,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 1.0428,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.7302439626857106,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 1.0634,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.5771032182407855,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 1.0259,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.7855037501485448,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 1.058,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.634630309001079,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 1.0384,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.6054182303926774,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 1.0448,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.6496032276380671,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 1.0465,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.6754777287023018,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 1.0599,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.6312190730291863,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 1.0791,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.7588028496055121,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 1.0131,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.502745691379509,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 1.0366,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.6811130462295836,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 1.0326,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5039577969890455,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 1.0238,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.6271762105900376,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 1.0237,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.5450609673922266,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 1.0406,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.4943978700146169,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 1.0432,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.6196816620092361,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 1.0337,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.5015888586866524,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 1.0528,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.5870075492507558,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 1.0401,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.5358825100277719,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 1.0418,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.5984788033239272,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 1.0388,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.5674364394043458,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 1.0397,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.562820687407089,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 1.0549,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.6276931212517751,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 1.0481,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.5467251432864145,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 1.0368,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.6352923669435698,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 1.0489,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.6069514881063265,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 1.0568,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.5302644865656403,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 1.034,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.6139741028367314,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 1.0374,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.6312703169097201,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 1.0338,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.607071856655427,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 1.0428,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.672487545226556,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 1.0373,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.6429593981903905,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 1.0325,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.6289578352126414,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 1.043,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5740864093685768,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 1.0413,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.6479615140484318,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 1.0352,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.49307372565902874,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 1.0234,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.6239976144273477,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 1.037,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.5053802624466267,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 1.0374,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.5585567150106538,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 1.0485,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.5268856673739207,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 1.0403,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.5340822413478552,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 1.0528,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.5195926145834207,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 1.0449,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.49390745982655476,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 1.0518,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5686614940284275,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 1.0459,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.5262697314501006,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 1.0387,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.5019911350612026,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 1.0457,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.4641485143349731,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 1.0232,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.5187971988164334,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 1.0519,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.5979885338686404,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 1.0535,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5507807523413634,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 1.0582,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5501824769840786,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 1.0381,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5764243288349461,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 1.0685,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5693247830746182,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 1.053,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5542731067189233,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 1.0762,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.503883279983583,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 1.0437,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5855594964896981,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 1.0424,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5529192677476757,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 1.0537,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5308952728829993,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 1.0421,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5407297140816788,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 1.0271,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.4519611598667168,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 1.0659,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.4927020869972054,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 1.0498,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.47546280811026076,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 1.038,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.4991118472949758,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 1.0259,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.4843537614762413,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 1.055,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.5385278901270503,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 1.0377,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5145119099234804,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 1.0592,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.42589133022348136,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 1.0648,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.49207306903746445,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 1.0338,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.48133059275570916,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 1.0316,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.492352408156703,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 1.0644,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5197510786393145,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 1.0517,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.473318729442859,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 1.0376,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5003525992715746,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 1.0506,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.4916827627722704,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 1.0519,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.533891567498327,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 1.0448,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.44858408621356244,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 1.0516,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.5174507991729026,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 1.042,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.436979940424765,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 1.0501,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.4961050649316023,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 1.0558,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.46684499056904777,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 1.0637,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5798212675087163,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 1.0339,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.4881210110875718,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 1.056,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.6249852936995848,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 1.0511,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.47571354977986663,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 1.0461,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5761207810710964,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 1.057,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.5231834389422628,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 1.0435,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.5537209443681236,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 1.0624,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.6074998997091668,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 1.0501,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.48022198778548186,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 1.0454,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.5853662844911703,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 1.0562,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.54853630639861,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 1.025,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.5323492630950059,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 1.0328,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5721294792273968,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 1.034,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.5325596029568079,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 1.0283,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5097538835957851,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 1.0445,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.538437342754811,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 1.0352,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5047250753912512,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 1.0269,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.5015116429931346,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 1.0373,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.5296468958737209,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 1.0595,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.46567226501890396,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 1.0548,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5616118554953794,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 1.0559,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.499116242126321,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 1.0408,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.516493404458395,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 1.0482,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.4567193973093227,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 1.0789,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5237594096499215,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 1.0369,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.4916556145929289,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 1.0499,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5251403072530468,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 1.0613,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.6107879182215814,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 1.0337,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5279027285908771,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 1.0621,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.6536466884400963,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 1.0572,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.5022776315373937,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 1.0419,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.6035360364948404,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 1.06,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5202130242818387,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 1.0272,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5490795736917288,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 1.023,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5348327276223342,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 1.0253,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.5140167418423081,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 1.0351,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5161268644421597,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 1.0452,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.4846842421747326,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 1.0505,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5692486229857902,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 1.0464,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.49013750133306044,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 1.0608,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5511196376673693,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 1.0599,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.5068607756830831,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 1.0309,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.5831518057373377,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 1.0607,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5063754143234067,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 1.0681,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5272200865578663,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 1.0605,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5512033242001375,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 1.0476,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.4690650588950576,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 1.0683,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.4763259681325161,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 1.0491,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.5506193521628624,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 1.0531,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.4584805676978386,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 1.081,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.4793791041212703,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 1.0206,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.523773254437704,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 1.0789,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.5109479648928673,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 1.0238,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5758945508327388,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 1.0587,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.500962247884293,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 1.0246,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.6096555536192836,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 1.0409,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.5154103500358629,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 1.0406,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.46985465422018735,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 1.039,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.522737974521083,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 1.0439,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.4913762886917353,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 1.031,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.4933112960563552,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 1.0483,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.5451525284653072,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 1.0582,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.5039044059484086,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 1.0752,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.4945161014130651,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 1.0597,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.5089182249626694,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 1.0757,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.46289008925660674,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 1.038,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.4682671348149049,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 1.055,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.46566145955510213,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 1.0535,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5324568206686956,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 1.0537,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.4691139131609931,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 1.0394,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5656947635641009,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 1.0542,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5821728413692226,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 1.0772,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.5026428518003064,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 1.0593,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5008467782420429,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 1.0409,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5491409322432044,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 1.0417,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.49814194274049195,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 1.0439,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.48272419417111917,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 1.0578,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.5016962789791116,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 1.048,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.4738426993155479,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 1.0354,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.4848441835827118,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 1.0279,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5181912124213884,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 1.0605,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.47268421798195653,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 1.0615,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.5203804363089138,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 1.0271,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.45465649961163085,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 1.0226,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.48150141608286606,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 1.0494,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.4884835200755149,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 1.0511,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.4665423750821624,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 1.0384,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.4323165452282403,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 1.0295,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.5470007329329022,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 1.0408,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.46204254786556037,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 1.0643,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.49366440877847784,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 1.0676,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.520635138756991,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 1.0719,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.46526413293410246,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 1.049,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.48232574817834956,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 1.0558,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.4288379347331897,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 1.0386,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.4909730582147011,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 1.0505,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.5101726924141013,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 1.0707,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.46691241594418276,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 1.0609,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.4851987216823345,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 1.0394,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.4486784961414057,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 1.0336,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.5252062285369983,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 1.0605,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.45445082003748405,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 1.0523,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5253134975016287,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 1.055,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.45941614899392236,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 1.0462,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.5932085280312072,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 1.0461,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.42776513903393626,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 1.0516,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.5427406384068965,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 1.0478,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.5545316120209728,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 1.0697,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.46675819996058115,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 1.0492,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.5541444319747849,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 1.0442,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.4810438238893908,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 1.0572,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.5204259006848317,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 1.0182,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.45507059793099564,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 1.0322,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5144681727470357,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 1.0505,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4949435353428202,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 1.059,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.4630445554853246,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 1.0394,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.49898707013576454,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 1.0439,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.4664654992336049,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 1.0345,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.47521855312349126,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 1.0495,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5076155589425728,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 1.0444,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.4747253945859919,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 1.0396,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5100786735464933,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 1.0743,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5326975533058383,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 1.0407,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.47598410206209474,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 1.0499,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.5425991015531546,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 1.0653,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.48412541520415925,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 1.052,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.4650491558127613,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 1.0561,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.529940934907524,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 1.0529,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.49775832860617386,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 1.0502,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.4497561040801792,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 1.0248,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.4963813836780174,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 1.0421,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.49385026797861004,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 1.0531,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.460239754624027,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 1.0347,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.5844775178893551,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 1.0583,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.44854150229199885,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 1.0464,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.46833998872583155,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 1.05,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5132756300576186,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 1.0708,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.4710838360796435,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 1.0159,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.4846545225614132,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 1.0581,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.4966307053034554,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 1.0428,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.4502028470495629,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 1.0369,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.5130105018250534,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 1.0331,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.4564230094385848,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 1.0636,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.4715045606010778,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 1.0412,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.45574673496209206,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 1.0768,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.46330254201404647,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 1.034,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.5311134895919952,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 1.0514,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.4619411411766266,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 1.0291,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.49687443770833506,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 1.0482,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.6364204781573405,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 1.032,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.4787100920995651,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 1.0527,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.6310894830955899,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.9941,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.8180102555889671,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.9041,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 1.3535039606984312,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.9124,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.7358474091613717,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.8864,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.8865789657257809,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.8826,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.7949213143525546,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.9057,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.5824288905085515,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.8995,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.722468973489139,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.8785,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.5897294473467032,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.8824,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.7458184450896463,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.8968,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.6250276347337589,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.872,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.7683624055740472,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.8851,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.618350810175572,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.8745,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.6700331187704023,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.87,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.6327188920031971,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.8816,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.6778364290543175,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.8864,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.5905208146163522,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.865,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.6140692294600693,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.8805,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.5983951912948063,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.8623,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.5864963993985547,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.8659,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.5676153853231466,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.897,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.4956236450819413,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.8675,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.5468724869802108,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.8877,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.5090977990778784,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.8747,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.5244843468313641,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.8948,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.49506801186165467,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.9058,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.6004150433785618,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.8811,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.5136429488102756,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.8902,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.5690067449782942,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.8862,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.6104030589645677,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.8765,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.5086921405172239,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.9,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.5154400629484853,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.8891,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.5178334925860547,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.8945,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.4787889034805734,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.881,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.4864699417620517,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.8843,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.5072982656951146,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.8841,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5204854749523411,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.8576,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.5910070760661782,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.857,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.5309855730976006,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.871,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.5447790564967205,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.8732,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.5353496832451279,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.8918,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.4701684714963658,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.8764,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.5520397023912471,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.8695,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4892786666956534,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.9009,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.5293258637344455,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.8688,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.48894143597485146,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.9008,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.5085108632581762,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.8755,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.4959822496389948,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.8886,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.45987316942271494,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.8739,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.49986957468158294,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.8815,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.46235706037383684,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.861,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.47234679017989956,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.8813,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.44840201464406676,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.8834,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.4982890611023794,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.8795,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.4606465137108056,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.8938,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.4864250188051069,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.8834,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.4829028613475291,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.8745,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.5056197569065203,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.864,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.47623732756093756,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.8744,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.48814166231943545,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.8971,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.5451312005162723,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.893,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.5387799422079101,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.8891,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.48625260286485644,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.8525,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.5213545642094328,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.8788,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.5810509768842562,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.8735,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.5147248688868965,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.8891,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.5097670518171973,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.8933,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.46905604416609226,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.8677,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.4793158641498187,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.8826,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.49659656530960067,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.8871,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.45587318952016986,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.8785,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.5014264902705274,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.8728,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.49983864726897526,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.893,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.47193716478800274,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.9048,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.45017756451716745,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.9001,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.49962178546229047,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.8681,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.4677777345879565,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.8839,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.4967255952611871,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.8673,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.45707081021679347,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.8896,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.48564837281329104,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.8547,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.4600337937594165,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.8739,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.4658195063906593,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.8826,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.5080345225989445,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.873,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.4793796144616891,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.8671,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.5091036913236003,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.8764,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.5294232426798272,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.8972,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.508162275333726,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.8986,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.46456368239114376,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.8964,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.5448423216775264,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.896,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.5209877985818187,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.8907,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.4854416925823396,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.8743,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.47466495247951923,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.8866,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.4944086069214936,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.8886,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.5001860831475555,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.8802,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.4690724760443913,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.8887,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.46725357486368535,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.8648,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.4811414534474131,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.9038,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.47196357427448543,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.8797,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.5219664625870237,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.8512,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.5492679218050823,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.8864,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.4694576384983073,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.8881,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.5156455283617992,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.9033,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.5812756348476723,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.8989,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.5265772599329609,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.8693,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.46823858601328794,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.877,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.6038032847052581,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.8791,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4822567653814655,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.8763,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4735396245530225,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.8804,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.552016528011612,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.8814,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.472263835429021,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.8753,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.4945392422424211,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.8838,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.5179138105518121,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.8884,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.48271694425052974,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.8849,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.49677802299024565,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.8736,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.5008230464190682,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.8792,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.4958007946269214,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.8795,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.49575687054238055,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.8888,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.5049840221123827,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.856,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.47748614901845826,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.8994,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.5342054480212363,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.8836,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.4923173232817806,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.8981,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.543574974923143,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.8876,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.5626183316937887,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.8823,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.4922220314473337,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.8834,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.5720214760201604,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.8869,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.44532497040779045,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.8811,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.46066480781562474,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.886,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.504638531931878,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.88,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.47549301474087885,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.8696,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.5095958375945056,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.8647,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.4877166267500471,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.8811,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.46368144951921675,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.8903,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.5112444999642245,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.8914,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.4524218104301258,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.8805,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.47204061149176785,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.8956,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.46406606693788865,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.896,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.47097566591759504,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.8821,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.4490990525552651,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.8956,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.4435872920435449,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.8784,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.46267021939263314,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.881,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4622727943654197,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.87,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4529544991173733,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.8964,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.47461240110877706,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.8848,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.43697565211427963,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.8658,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.4499473335972394,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.9102,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.4672976027535272,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.8749,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.44112581691393943,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.8658,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.44650691911363133,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.8956,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.4657788553603753,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.8912,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.44794879038198787,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.89,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.4708721884114357,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.8765,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.44630967695679846,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.8823,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.4662044455267001,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.8797,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.4484995613219303,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.8592,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4399776080740483,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.8876,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.5091187228531728,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.8675,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.46786527941164696,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.8745,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4450452057191952,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.8698,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.45159689410451587,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.8734,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.4818584801736891,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.8732,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.46148955470473896,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.8993,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.44018101930362374,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.898,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4527596679027131,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.8964,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.4702223175021084,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.8936,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.4383822571267992,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.8873,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.4266755933611976,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.8887,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.45800668513819115,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.8868,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.47333343013029877,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.8906,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.436349395024753,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.8855,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.4570516760514788,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.884,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.46022979995375624,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.9003,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.4430846744324896,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.854,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.4270354994019947,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.874,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.44255053493765467,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.8719,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.46976341560020335,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.873,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4542807589265525,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.8945,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.4472350107844088,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.8875,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.5182602314514315,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.8986,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.41735570054290566,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.8887,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.43537235403285596,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.8966,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.4512428619723997,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.8891,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.9150158422539177,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.9253,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.5132003543610879,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.8864,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.4683851912404276,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.8875,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.4622614705315623,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.8872,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.47728322013854374,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.8749,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.43308415843033743,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.8628,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4438987842432331,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.9017,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.4325986554854069,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.8795,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.43847954388592075,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.9002,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.4353876740425362,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.8932,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.43893687870752884,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.8864,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.46135865242445556,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.8849,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.4298804476279949,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.856,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.41745971353216005,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.8799,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.43677053657113296,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.88,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.42022969207673955,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.8687,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.44223020342503794,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.8869,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.42756540579868063,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.8698,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.43923288348592465,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.871,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.425993239341892,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.8823,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.43168756995650565,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.8721,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.4395953444930949,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.8988,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.4390422682473557,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.8679,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.43225398469108783,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.8933,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.4441330616357335,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.8671,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.43832020587959936,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.875,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.4449617705162789,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.8804,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.4360625254197982,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.8757,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.4747678073937475,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.8882,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.43597589143908305,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.9149,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.43361279910370437,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.8659,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.44086915487162065,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.8765,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.44664198538093436,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.8572,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.43019572904620423,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.8762,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.4451181107959956,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.882,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.4341992414538811,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.8938,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.45412464875522746,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.8749,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.4687408938958566,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.8743,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.43773412319302646,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.8515,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.41655484978228663,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.8916,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.4456744561904983,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.8968,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4576939818824952,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.8892,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.4287514662505629,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.8891,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.45777900219569573,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.8913,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.42919212940836643,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.8601,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.4317068551753513,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.8796,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.43426751569139266,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.8989,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.4263651685086541,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.8577,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.43828481572916544,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.8934,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.41957228895499626,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.8784,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.4377631999103462,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.8719,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.43522480928320884,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.8969,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.4511358478499723,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.8862,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.44500895074969815,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.8828,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.4114171523746381,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.8779,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.4168790694787328,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.8693,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.4205047897138705,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.8763,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.4253738186332201,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.9022,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.43089497437273344,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.8843,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.43280716403396996,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.8728,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.42906265952991446,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.8811,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.42883670520355016,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.8861,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.41960820424789963,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.8736,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.42525548406935254,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.8893,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.4413838672839164,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.8784,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4174721269769123,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.8456,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.6155083519966948,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.819,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.9561759252968006,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.7529,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.6857576841448072,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.7217,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.861786497350572,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.7553,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 1.2515522045815959,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.7274,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 1.0546368909786605,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.7218,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.8066066231883815,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.7515,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.9767689680564277,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.7184,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.9535248247081104,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.7457,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.7434822844601587,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.7458,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.6235926380750707,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.7267,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.6611421736854836,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.7177,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.7517806205601149,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.7431,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.7573308368851783,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.7327,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.6522667097061046,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.7238,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.5406145812028827,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.7153,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.5754927592497974,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.7451,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.6340776958460811,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.7149,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.5864798077904663,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.7311,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.5757449325967806,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.7375,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.5363677442090738,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.7331,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.5557386838012588,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.7332,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.5316348451379913,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.7056,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.5485126502512537,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.7371,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.5477257281455122,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.7291,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.5551039562088412,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.7622,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.5210679537009392,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.7343,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.5068894728099159,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.7153,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.5013802547543935,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.7465,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.5693299115463599,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.742,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.5679914142985081,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.7453,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.5280455673116544,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.7391,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.5142902799171822,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.7183,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.5592027028979482,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.724,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.5220628189496982,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.7242,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.5166739824397905,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.7216,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.5214491311861654,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.7264,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.5105965394821627,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.7103,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.4829968560460906,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.7504,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.48448798488122874,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.7191,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.5020916366473639,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.7518,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.49205910281066506,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.7286,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.5029721231372624,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.7566,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.5228072448972337,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.7366,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.4721094827064083,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.7207,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.49113204223351137,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.7486,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.48660609712563463,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.7499,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.47336619833916754,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.736,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.46970959715343225,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.7232,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.48464970892767395,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.7325,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.4931303092359112,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.7445,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.4921457293310269,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.7149,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.4917085024620295,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.7287,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.49339672286488734,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.7517,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.47163973152526717,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.7147,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.483536034016903,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.7459,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.4763802971796761,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.7464,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.49351536559467907,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.7498,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.47615321831160484,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.7396,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.47157478041497347,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.7199,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.4928272038858355,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.7259,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.46814573555352623,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.7354,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.4853692081337121,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.707,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.4851336307896996,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.7617,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.4645343436355272,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.7115,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.4872765502631068,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.7362,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.4873655209378128,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.7201,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.4808049293332522,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.7268,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.47491635653061426,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.7175,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.46899421182369033,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.7243,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.48707301246741536,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.7436,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.469643247863743,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.7216,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.4847360858251042,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.7362,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.48591984962338186,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.7287,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.4813501166915706,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.7102,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.48308551753321066,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.7353,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.4715922355681568,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.7292,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.47668288206146975,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.7133,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.48134419994862776,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.7266,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.4958854947176346,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.7453,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.47792092894753,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.718,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.48387917795653934,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.7364,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.4861483723354823,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.7394,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.46906501108375875,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.7376,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.4815898722738489,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.714,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.46427185573575236,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.731,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.46302265093230166,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.729,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.4668051119260838,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.7302,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.46681511412017856,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.7197,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.4678897588133719,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.7245,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.47092950699764463,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.7372,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.48806805119823315,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.7368,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.5569223684522442,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.753,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.4669702832657187,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.736,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.47099499369283515,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.7072,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.4800132637430607,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.742,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.48545301932111057,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.726,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.4840474521689336,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.7145,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.47686364110463414,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.7294,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.47107384698575727,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.7309,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.4791117269035014,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.7176,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.46491600154898316,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.7097,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.4734595792632678,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.7314,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.4790887582565522,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.7307,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.46259878314802905,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.7478,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.47575995507558144,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.7462,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.46852481132322965,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.7153,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.48053858077094974,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.7541,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.4722720745648607,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.7269,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.4625221628513618,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.7426,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.4619872450801662,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.7233,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.46908449411440883,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.7354,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.47458587561058785,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.7456,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.48487994327824785,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.7312,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.46961384996245215,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.7217,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.4668814204033196,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.7515,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.4721884726439901,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.7404,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.4697838813893813,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.7377,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.47119087681821314,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.7347,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.47298245233595215,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.7558,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.47502822119823124,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.7426,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.4819242104690814,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.7428,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.47650229243120656,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.7354,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.4661841263043608,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.7357,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.48513910918147496,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.7091,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.46600353543185574,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.7367,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.4830068196719165,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.7344,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.4760134150251606,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.7385,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.4696630384770401,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.7313,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.4645030070231969,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.744,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.4689877630014468,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.7349,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.477756405898306,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.7366,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.4771275312492012,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.716,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.46285486390431163,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.7319,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.47715584222354557,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.7204,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.4661806612319913,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.7346,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.4598841510574184,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.7157,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.4644815634521155,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.7283,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.4682330701277009,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.7071,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.4752768209079538,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.7417,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.4763051571334349,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.7412,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.4635804775616971,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.7215,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.47330998350548026,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.7354,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.46691311956138876,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.7281,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.4668431153587308,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.7112,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.47379536372009784,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.7321,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.47119378669762274,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.7578,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.46661323219472806,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.7484,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.4615837567228492,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.7295,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.4644460558307983,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.7378,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.46335252597683324,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.7506,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.469902304822145,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.7069,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.47256217431417985,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.7368,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.4656728135205727,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.73,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.46418402881994064,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.7481,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.4589106753965094,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.7302,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.46783340169239723,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.7333,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.46787436219475287,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.7292,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.468645152684087,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.7191,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.46440990308800606,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.7572,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.46464257648249174,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.7506,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.4593508556975793,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.7217,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.49042758253848145,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.7266,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.47647257483966116,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.729,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.47110233598765755,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.7162,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.47031601552075414,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.7301,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.4637918306689836,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.7212,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.4597027283632771,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.7429,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.4695080720452875,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.7111,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.4683299782903068,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.7299,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.4612227785496506,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.7043,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.4684312303488188,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.7105,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.46027218764258754,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.7038,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.468725254263014,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.7315,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.4612818252914972,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.7258,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.4626134845422531,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.732,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.4626591420055856,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.7484,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.4803766048873325,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.7464,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.4754131662889612,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.7365,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.48213910502289414,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.7441,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.47641941469009286,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.7386,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.4760169034607843,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.7317,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.46906477338739977,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.7201,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.46782105861543827,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.7363,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.47007408696813074,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.7198,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.4656422162308048,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.7177,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.46423708242608597,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.7436,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.46595445872434005,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.7008,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.46603819673052616,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.7342,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.4623684870398543,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.7491,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.4747064060135233,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.7553,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.4586054511214838,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.7327,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.46750925219595824,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.7339,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.46397631610061624,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.7559,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.46624681356606,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.7474,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.46357889729795043,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.7222,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.4570132406039567,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.7337,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.4673122239142358,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.7438,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.4569842430490534,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.7357,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.45959718826147633,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.7348,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.4589783973670068,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.7192,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.45890777615098377,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.7367,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.4610788247804613,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.7296,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.4688218647176079,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.7179,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.46215891480950827,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.7252,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.46260754756608485,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.747,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.4650368686412901,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.7426,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.4595905979616877,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.7249,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.4726976849313269,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.7147,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.46094806012171474,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.749,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.46928710771460735,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.7199,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.46132809016742565,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.736,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.4643887280649863,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.7281,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.46668534138782775,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.7512,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.46303256571999946,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.7407,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.45968326294040335,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.7615,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.46561320258145283,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.7354,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.4711355019428249,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.7231,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.457251602020112,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.7579,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.4541225054164393,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.7166,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.46743301020447847,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.7546,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.4614537644684759,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.7074,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.4650597845767091,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.7368,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.4643482602615043,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.7098,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.4603714391234357,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.7319,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.4575610245802634,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.7396,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.47208034354961154,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.7517,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.4666665243092385,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.724,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.46293659496933776,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.7388,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.46426113918658063,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.7501,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.46913534012554514,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.739,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.4657738813909136,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.7511,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.4623191017571186,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.7483,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.45818819355508855,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.7083,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.4650104901620667,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.733,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.45825913132334206,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.7402,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.4647089466500658,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.7471,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.45924557565720975,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.7325,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.4694569856757483,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.7444,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.4754997038922736,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.7131,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.4656274996350381,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.7264,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.46301731042811334,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.7423,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.4613268580190517,
+      "learning_rate": 0.0,
+      "loss": 0.7152,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.1872555195142144e+18,
+      "train_loss": 1.0362177829916883,
+      "train_runtime": 17180.708,
+      "train_samples_per_second": 9.196,
+      "train_steps_per_second": 0.072
+    }
+  ],
+  "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": 2.1872555195142144e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}