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": 8.617597838392205,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 1.9812,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 8.59600293568971,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 1.9616,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.700760114166933,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.9581,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.474702241243037,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 1.9587,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 8.14265607188293,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 1.9475,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.905018753825991,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 1.9428,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.405741197461577,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 1.9013,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.948894237039982,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 1.8534,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.883298242894116,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 1.8954,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 3.3788104583059217,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 1.8863,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 3.110278023544174,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 1.8546,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 6.050914508980096,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.8424,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 6.444492255572725,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 1.87,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 6.0842287400931925,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 1.8556,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 5.66417842454959,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 1.7961,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 3.6244519066658234,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 1.8097,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 3.118097699979137,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 1.7498,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.366017106889007,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 1.7231,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.9780396446079023,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 1.7755,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.6220503623277316,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 1.6861,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.6225264992676074,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 1.6357,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.7987864916581213,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 1.6783,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.795471486429829,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 1.6392,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 1.5396053534882497,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 1.6297,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.3881980579140738,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 1.6121,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 1.1794071271540405,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 1.6244,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 1.4377597935377107,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.6218,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 1.4284694623552487,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 1.5945,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 1.379952884884446,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 1.5973,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 1.3479981983897353,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.6101,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 1.111443192592021,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 1.5776,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 1.0087728865293633,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 1.5569,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 1.0817846050360103,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 1.5996,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 1.2754183506816692,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 1.567,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 1.117772771058001,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 1.5759,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.8523615450485009,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 1.5357,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 1.125787711305252,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 1.5321,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 1.0136078205836054,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 1.5821,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.8805433867540483,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 1.5659,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 1.0683619804561089,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 1.5947,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.9187801603908706,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.5663,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.823572751882817,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 1.5262,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.9020057600891004,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 1.5509,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.7817535754769971,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 1.5456,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.9652011694278385,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 1.5129,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.9623083870472557,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 1.5619,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.7517905309481996,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 1.5069,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 1.099193669329019,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 1.511,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.7887802513795341,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 1.5359,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.123796345605809,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 1.5609,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.8032772546625709,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 1.5192,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 1.1590602518791306,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 1.534,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.8235466316513813,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 1.5317,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.8444983623590966,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 1.5603,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.8379042216818946,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 1.4851,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.8599484637250474,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 1.5283,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.813300453921697,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 1.5164,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.8998392839975261,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 1.5114,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 1.0601234051261872,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 1.5284,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.818315563372061,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 1.4973,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.8549250006809715,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 1.4625,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.9381026580126054,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 1.4795,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.9107369230037778,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 1.4938,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.8222529920476536,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 1.5001,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 1.172777565192398,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 1.494,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.9591692250059369,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 1.5303,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 1.0253273711579007,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 1.5003,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 1.4577999566962934,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 1.4935,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 1.4394495527474696,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 1.4929,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 1.2398923482433348,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 1.496,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 1.0359111924013935,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 1.5397,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 1.4371161409195676,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 1.5129,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 1.1728115106909236,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 1.4955,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 1.3850116738087552,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 1.5126,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.0474457975234626,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 1.5425,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 1.234951274039603,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 1.5151,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 1.2457955969853516,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 1.4961,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 1.1818893574900255,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 1.4933,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8729011212292087,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 1.5262,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 1.0796769855777228,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 1.4894,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 1.293505252780895,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 1.4985,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 1.0877279804443873,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.5365,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 1.1090455986942158,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 1.5094,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 1.2451234470603796,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 1.476,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.9355469748332176,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 1.4864,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 1.6189245377105002,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 1.5069,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.995403087251463,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 1.5004,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 1.382593360976765,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 1.5018,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 1.479321362163701,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 1.5007,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 1.1874674846471172,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 1.5184,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 1.453339432341642,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 1.4736,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 1.2514834786724582,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 1.4925,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 1.3500983558555222,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 1.4803,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 1.1949626483194833,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 1.5236,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 1.5909900982953062,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 1.498,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.1555631984524903,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 1.4927,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 1.3977247738152612,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 1.4757,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 1.1520860753722832,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 1.4996,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 1.7603059254365199,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 1.4845,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.421554934626368,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 1.5039,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 1.5914793518097214,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 1.5237,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 1.7608706250737685,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 1.4531,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 1.3998456095078922,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 1.4787,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.3935447402942522,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 1.4787,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 1.6650788335642808,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 1.5279,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 1.4900626273335558,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 1.4999,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 1.2202403945823441,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 1.5052,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 1.3851669964827278,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 1.4721,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 1.025669899156198,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 1.4933,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 1.212138127891812,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 1.4842,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 1.3067909369047384,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 1.4708,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 1.4524615891887496,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 1.5086,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 1.403659590005546,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 1.5269,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 1.0221953619506279,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 1.4671,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 1.4222987266099283,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 1.4936,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 1.4451309228242861,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 1.5107,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 1.419767329328415,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 1.446,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 1.377158692208645,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 1.498,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 1.5187481369515436,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 1.4757,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 1.0997557334829153,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 1.4703,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 1.8555296106483312,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 1.5128,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 1.1965169412097565,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 1.4592,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 1.2768921161807498,
+      "learning_rate": 4e-05,
+      "loss": 1.4857,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 1.3510119397573794,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 1.4576,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.6877696792648427,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 1.4903,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 1.3427996095023798,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 1.4802,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 1.7391629392366474,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 1.5106,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 1.4682201139535385,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 1.511,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 1.568105199308296,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 1.4918,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 1.2214669171997263,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 1.4536,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 1.3895206600754282,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 1.5106,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 1.4677913509263572,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 1.502,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 1.1462620541876913,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 1.4553,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 1.5568297465529448,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 1.4475,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 1.1414201632593375,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 1.4952,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 1.7783076912388363,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 1.4602,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 1.3412681510603668,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 1.5005,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 1.7263940997271072,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 1.4616,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 1.489740593703162,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 1.4742,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 1.9415926920016693,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 1.469,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 1.3328440135666995,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 1.4702,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 1.847158488474377,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 1.4631,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 1.3192393757720386,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 1.4412,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 1.6836195610655533,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 1.4812,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 1.6604750305659146,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 1.464,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 1.7383054327569032,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 1.4945,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 1.6897216009145357,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 1.4487,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 1.5161918548773172,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 1.4844,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 1.6651493822766084,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 1.4858,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.392469589205136,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 1.4672,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 1.4977699069722925,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 1.4831,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 1.772787829803169,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 1.4839,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 1.5159233134090517,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 1.484,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 1.1603792198280451,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 1.489,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 1.7010840263194253,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 1.5094,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 1.2736135215861935,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 1.4547,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 1.5893709526401791,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 1.4982,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.48348672182819,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 1.4962,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 1.2884769301117731,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 1.4859,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 1.8674049606711507,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 1.4368,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 1.1195484210467739,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 1.4908,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 1.8092425695561174,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 1.4437,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 1.633526839175801,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 1.5151,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 1.644418214867434,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 1.428,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 1.6707704742409615,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 1.4708,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 1.4977393537999641,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 1.4294,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 1.3550364602006515,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 1.5065,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 1.1759942071422687,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 1.4825,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 1.2954323431825738,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 1.4322,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 1.3511092520562662,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 1.425,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 1.130251440332502,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 1.4461,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 1.464682599413448,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 1.4289,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 1.1487355574689648,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 1.4313,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 1.3660266596042128,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 1.4722,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.4534338695200502,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 1.4983,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 1.117156776526443,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 1.4818,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 1.658950536634302,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 1.4737,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 1.2847278762434056,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 1.4785,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 1.5695064484909362,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 1.3873,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 1.2255525980373352,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 1.4504,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 1.4815807370675584,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 1.4609,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 1.045247506267775,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 1.4652,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 1.300644860753016,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 1.4783,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 1.0130389656396452,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 1.4522,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 1.383263889226605,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 1.4823,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 1.0092347252748597,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 1.4601,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 1.3780117418422326,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 1.4567,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 1.1547244132427796,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 1.4601,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 1.169755716855418,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 1.5149,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 1.275534010017718,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 1.4757,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 1.3052649604266426,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 1.4546,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 1.1387198661803462,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 1.453,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 1.1299859442744742,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 1.4436,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.9620636481575733,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 1.4679,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 1.5352867892171986,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 1.4606,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.8917970107025449,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 1.4238,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 1.3049146834777776,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 1.4585,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.9741253630002761,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 1.4602,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 1.4086959105020578,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 1.4936,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.0260403459581102,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 1.4762,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 1.2944869137799337,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 1.4682,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 1.2339968247441937,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 1.4665,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 1.016150741247741,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 1.4541,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 1.189544180430917,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 1.4577,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.775087505001017,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 1.4509,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 1.1843006466106425,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 1.4763,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.822964135918417,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 1.4817,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 1.1834353435948706,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 1.4874,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.9174695049369864,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 1.4247,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.9421215827436382,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 1.4633,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 1.0661485856428425,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 1.4491,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.8985265589577408,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 1.4764,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 1.5726812718838665,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 1.4675,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 1.1863405849928925,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 1.4855,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 1.5177386902057526,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 1.48,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 1.4603477260622475,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 1.4652,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 1.2375352136228746,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 1.4499,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 1.1315341340971774,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 1.4518,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 1.4047603331920815,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 1.4126,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.9394547025083132,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 1.438,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 1.094511791077764,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 1.4689,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 1.1546290723374852,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 1.4398,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 1.4089202436928072,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 1.4669,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.7832575468023459,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 1.4895,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 1.0104170475305374,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 1.5021,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 1.0748569880628447,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 1.4775,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 1.3123940469411655,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 1.4506,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.9571083151333442,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 1.4576,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.9595589796707975,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 1.4328,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.9965441472420481,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 1.4271,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.83073902745805,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 1.4584,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.8980860271058129,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 1.4282,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 1.0270182012515834,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 1.4489,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 1.340208575105287,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 1.4238,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.8104382247700033,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 1.4571,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.8360075471206514,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 1.4076,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0086894370189143,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 1.4611,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 1.081340902910155,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 1.5031,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 1.1834626329799445,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 1.4175,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 1.1220130156547763,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 1.4382,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.8146384759794331,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 1.472,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 1.1533866100902783,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 1.5015,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.9166098903215822,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 1.4406,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 1.045217833242825,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 1.4868,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.8809892917731114,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 1.4939,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.8852208932699429,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 1.4385,
+      "step": 246
+    },
+    {
+      "epoch": 1.000506329113924,
+      "grad_norm": 0.8729730171044652,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 1.417,
+      "step": 247
+    },
+    {
+      "epoch": 1.0045569620253165,
+      "grad_norm": 0.8791486902873684,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 1.403,
+      "step": 248
+    },
+    {
+      "epoch": 1.0086075949367088,
+      "grad_norm": 0.9474266298352538,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 1.4095,
+      "step": 249
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.1060329307630068,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 1.3559,
+      "step": 250
+    },
+    {
+      "epoch": 1.0167088607594936,
+      "grad_norm": 1.3033668093154922,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 1.3494,
+      "step": 251
+    },
+    {
+      "epoch": 1.0207594936708861,
+      "grad_norm": 0.8960589096130064,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 1.3866,
+      "step": 252
+    },
+    {
+      "epoch": 1.0248101265822784,
+      "grad_norm": 1.1015764234918468,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 1.4157,
+      "step": 253
+    },
+    {
+      "epoch": 1.028860759493671,
+      "grad_norm": 0.9924903197746014,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 1.4074,
+      "step": 254
+    },
+    {
+      "epoch": 1.0329113924050632,
+      "grad_norm": 1.1477276023263911,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 1.3805,
+      "step": 255
+    },
+    {
+      "epoch": 1.0369620253164558,
+      "grad_norm": 0.9438477533899338,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 1.3788,
+      "step": 256
+    },
+    {
+      "epoch": 1.041012658227848,
+      "grad_norm": 0.795138592428724,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 1.4218,
+      "step": 257
+    },
+    {
+      "epoch": 1.0450632911392406,
+      "grad_norm": 0.892306300289459,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 1.382,
+      "step": 258
+    },
+    {
+      "epoch": 1.0491139240506329,
+      "grad_norm": 0.839515171221026,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 1.4044,
+      "step": 259
+    },
+    {
+      "epoch": 1.0531645569620254,
+      "grad_norm": 0.9738082568059102,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 1.3872,
+      "step": 260
+    },
+    {
+      "epoch": 1.0572151898734177,
+      "grad_norm": 1.244677163407347,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 1.4105,
+      "step": 261
+    },
+    {
+      "epoch": 1.0612658227848102,
+      "grad_norm": 0.8393258643347646,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 1.3826,
+      "step": 262
+    },
+    {
+      "epoch": 1.0653164556962025,
+      "grad_norm": 0.9775247272950657,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 1.3799,
+      "step": 263
+    },
+    {
+      "epoch": 1.069367088607595,
+      "grad_norm": 1.114195230292501,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 1.3671,
+      "step": 264
+    },
+    {
+      "epoch": 1.0734177215189873,
+      "grad_norm": 1.4254540085451792,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 1.3905,
+      "step": 265
+    },
+    {
+      "epoch": 1.0774683544303798,
+      "grad_norm": 0.7704534122897447,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 1.3841,
+      "step": 266
+    },
+    {
+      "epoch": 1.081518987341772,
+      "grad_norm": 1.088659382392286,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 1.3682,
+      "step": 267
+    },
+    {
+      "epoch": 1.0855696202531646,
+      "grad_norm": 1.1585068776138794,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 1.3428,
+      "step": 268
+    },
+    {
+      "epoch": 1.089620253164557,
+      "grad_norm": 0.771634608181161,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 1.3927,
+      "step": 269
+    },
+    {
+      "epoch": 1.0936708860759494,
+      "grad_norm": 0.9198481354399831,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 1.4325,
+      "step": 270
+    },
+    {
+      "epoch": 1.0977215189873417,
+      "grad_norm": 1.0691877340921385,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 1.3591,
+      "step": 271
+    },
+    {
+      "epoch": 1.1017721518987342,
+      "grad_norm": 0.9675630595168563,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 1.3823,
+      "step": 272
+    },
+    {
+      "epoch": 1.1058227848101265,
+      "grad_norm": 0.9989678844071618,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 1.4302,
+      "step": 273
+    },
+    {
+      "epoch": 1.109873417721519,
+      "grad_norm": 0.7371187884375405,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 1.4102,
+      "step": 274
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 1.0373189441713764,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 1.4101,
+      "step": 275
+    },
+    {
+      "epoch": 1.1179746835443038,
+      "grad_norm": 0.9274735820183361,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 1.3952,
+      "step": 276
+    },
+    {
+      "epoch": 1.1220253164556961,
+      "grad_norm": 0.7544335476452133,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 1.382,
+      "step": 277
+    },
+    {
+      "epoch": 1.1260759493670887,
+      "grad_norm": 0.7772889806493559,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 1.3757,
+      "step": 278
+    },
+    {
+      "epoch": 1.130126582278481,
+      "grad_norm": 0.9434641036110949,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 1.3672,
+      "step": 279
+    },
+    {
+      "epoch": 1.1341772151898735,
+      "grad_norm": 1.2028056753869438,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 1.3464,
+      "step": 280
+    },
+    {
+      "epoch": 1.1382278481012658,
+      "grad_norm": 0.8908384389280063,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 1.3892,
+      "step": 281
+    },
+    {
+      "epoch": 1.1422784810126583,
+      "grad_norm": 0.943858513711046,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 1.3841,
+      "step": 282
+    },
+    {
+      "epoch": 1.1463291139240506,
+      "grad_norm": 1.307515994297152,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 1.3657,
+      "step": 283
+    },
+    {
+      "epoch": 1.150379746835443,
+      "grad_norm": 0.861983171452164,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 1.379,
+      "step": 284
+    },
+    {
+      "epoch": 1.1544303797468354,
+      "grad_norm": 0.8194910280648899,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 1.3829,
+      "step": 285
+    },
+    {
+      "epoch": 1.158481012658228,
+      "grad_norm": 0.8513519949962763,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 1.3682,
+      "step": 286
+    },
+    {
+      "epoch": 1.1625316455696202,
+      "grad_norm": 1.1399177582396554,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 1.3581,
+      "step": 287
+    },
+    {
+      "epoch": 1.1665822784810127,
+      "grad_norm": 0.7663462425358679,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 1.355,
+      "step": 288
+    },
+    {
+      "epoch": 1.170632911392405,
+      "grad_norm": 0.7760075336317216,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 1.4024,
+      "step": 289
+    },
+    {
+      "epoch": 1.1746835443037975,
+      "grad_norm": 0.9647443955754902,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 1.4076,
+      "step": 290
+    },
+    {
+      "epoch": 1.1787341772151898,
+      "grad_norm": 0.7328552201414937,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 1.38,
+      "step": 291
+    },
+    {
+      "epoch": 1.1827848101265823,
+      "grad_norm": 0.7400096401036265,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 1.3835,
+      "step": 292
+    },
+    {
+      "epoch": 1.1868354430379746,
+      "grad_norm": 0.917589781937192,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 1.3423,
+      "step": 293
+    },
+    {
+      "epoch": 1.1908860759493671,
+      "grad_norm": 1.2527145294041306,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 1.3569,
+      "step": 294
+    },
+    {
+      "epoch": 1.1949367088607594,
+      "grad_norm": 0.9012467112734274,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 1.3566,
+      "step": 295
+    },
+    {
+      "epoch": 1.198987341772152,
+      "grad_norm": 0.8273803410093903,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 1.3475,
+      "step": 296
+    },
+    {
+      "epoch": 1.2030379746835442,
+      "grad_norm": 1.225816413877141,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 1.4047,
+      "step": 297
+    },
+    {
+      "epoch": 1.2070886075949367,
+      "grad_norm": 0.8720275613986102,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 1.3729,
+      "step": 298
+    },
+    {
+      "epoch": 1.211139240506329,
+      "grad_norm": 1.157641009695598,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 1.4291,
+      "step": 299
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 1.003384426270564,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 1.4035,
+      "step": 300
+    },
+    {
+      "epoch": 1.2192405063291138,
+      "grad_norm": 0.733433394757214,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 1.3705,
+      "step": 301
+    },
+    {
+      "epoch": 1.2232911392405064,
+      "grad_norm": 0.8890962806770986,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 1.4096,
+      "step": 302
+    },
+    {
+      "epoch": 1.2273417721518987,
+      "grad_norm": 0.9496386838291333,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 1.377,
+      "step": 303
+    },
+    {
+      "epoch": 1.2313924050632912,
+      "grad_norm": 1.1478516456870556,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 1.3919,
+      "step": 304
+    },
+    {
+      "epoch": 1.2354430379746835,
+      "grad_norm": 0.6861533286397372,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 1.3416,
+      "step": 305
+    },
+    {
+      "epoch": 1.239493670886076,
+      "grad_norm": 0.9677071410264394,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 1.354,
+      "step": 306
+    },
+    {
+      "epoch": 1.2435443037974683,
+      "grad_norm": 1.1003185561166802,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 1.3357,
+      "step": 307
+    },
+    {
+      "epoch": 1.2475949367088608,
+      "grad_norm": 0.7406374924908882,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 1.3511,
+      "step": 308
+    },
+    {
+      "epoch": 1.251645569620253,
+      "grad_norm": 0.9892836519024093,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 1.4275,
+      "step": 309
+    },
+    {
+      "epoch": 1.2556962025316456,
+      "grad_norm": 1.6574271485617718,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 1.3895,
+      "step": 310
+    },
+    {
+      "epoch": 1.259746835443038,
+      "grad_norm": 0.8695293339330724,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 1.3659,
+      "step": 311
+    },
+    {
+      "epoch": 1.2637974683544304,
+      "grad_norm": 2.042284710476943,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 1.3739,
+      "step": 312
+    },
+    {
+      "epoch": 1.2678481012658227,
+      "grad_norm": 1.556666694842682,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 1.394,
+      "step": 313
+    },
+    {
+      "epoch": 1.2718987341772152,
+      "grad_norm": 1.9672277671653209,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 1.4185,
+      "step": 314
+    },
+    {
+      "epoch": 1.2759493670886077,
+      "grad_norm": 1.8392439360929365,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 1.3853,
+      "step": 315
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1443553647877507,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 1.3935,
+      "step": 316
+    },
+    {
+      "epoch": 1.2840506329113923,
+      "grad_norm": 1.1901410882003,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 1.3664,
+      "step": 317
+    },
+    {
+      "epoch": 1.2881012658227848,
+      "grad_norm": 1.5024821630780905,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 1.3707,
+      "step": 318
+    },
+    {
+      "epoch": 1.2921518987341774,
+      "grad_norm": 0.9363552198165448,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 1.3374,
+      "step": 319
+    },
+    {
+      "epoch": 1.2962025316455696,
+      "grad_norm": 1.700214962692222,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 1.4168,
+      "step": 320
+    },
+    {
+      "epoch": 1.300253164556962,
+      "grad_norm": 1.1655209115256535,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 1.3777,
+      "step": 321
+    },
+    {
+      "epoch": 1.3043037974683545,
+      "grad_norm": 1.9435332132761853,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 1.4103,
+      "step": 322
+    },
+    {
+      "epoch": 1.308354430379747,
+      "grad_norm": 1.7058302669313086,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 1.4163,
+      "step": 323
+    },
+    {
+      "epoch": 1.3124050632911393,
+      "grad_norm": 1.2193697249213518,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 1.3667,
+      "step": 324
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 1.3090422197366365,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 1.3844,
+      "step": 325
+    },
+    {
+      "epoch": 1.320506329113924,
+      "grad_norm": 1.1485804668302109,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 1.4176,
+      "step": 326
+    },
+    {
+      "epoch": 1.3245569620253166,
+      "grad_norm": 1.1156319460703294,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 1.388,
+      "step": 327
+    },
+    {
+      "epoch": 1.3286075949367089,
+      "grad_norm": 1.2989756123804224,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 1.3832,
+      "step": 328
+    },
+    {
+      "epoch": 1.3326582278481012,
+      "grad_norm": 1.147128569370826,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 1.38,
+      "step": 329
+    },
+    {
+      "epoch": 1.3367088607594937,
+      "grad_norm": 1.111597483477992,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 1.3841,
+      "step": 330
+    },
+    {
+      "epoch": 1.3407594936708862,
+      "grad_norm": 1.1455941117147048,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 1.3944,
+      "step": 331
+    },
+    {
+      "epoch": 1.3448101265822785,
+      "grad_norm": 0.804891184224588,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 1.4054,
+      "step": 332
+    },
+    {
+      "epoch": 1.3488607594936708,
+      "grad_norm": 1.1543182275728239,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 1.385,
+      "step": 333
+    },
+    {
+      "epoch": 1.3529113924050633,
+      "grad_norm": 0.8423701976502794,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 1.3838,
+      "step": 334
+    },
+    {
+      "epoch": 1.3569620253164558,
+      "grad_norm": 0.8351731100061353,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 1.3537,
+      "step": 335
+    },
+    {
+      "epoch": 1.3610126582278481,
+      "grad_norm": 0.7304048596620231,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 1.3418,
+      "step": 336
+    },
+    {
+      "epoch": 1.3650632911392404,
+      "grad_norm": 0.8460712068154276,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 1.3645,
+      "step": 337
+    },
+    {
+      "epoch": 1.369113924050633,
+      "grad_norm": 0.8195708215810311,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 1.3695,
+      "step": 338
+    },
+    {
+      "epoch": 1.3731645569620254,
+      "grad_norm": 0.8610666916367373,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 1.3603,
+      "step": 339
+    },
+    {
+      "epoch": 1.3772151898734177,
+      "grad_norm": 0.859055317809985,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 1.3704,
+      "step": 340
+    },
+    {
+      "epoch": 1.38126582278481,
+      "grad_norm": 0.8362129194858037,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 1.3292,
+      "step": 341
+    },
+    {
+      "epoch": 1.3853164556962025,
+      "grad_norm": 0.9702174526494254,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 1.4165,
+      "step": 342
+    },
+    {
+      "epoch": 1.389367088607595,
+      "grad_norm": 0.6801379013885195,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 1.4172,
+      "step": 343
+    },
+    {
+      "epoch": 1.3934177215189874,
+      "grad_norm": 1.0065531932551037,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 1.3746,
+      "step": 344
+    },
+    {
+      "epoch": 1.3974683544303796,
+      "grad_norm": 0.8124826304159695,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 1.4216,
+      "step": 345
+    },
+    {
+      "epoch": 1.4015189873417722,
+      "grad_norm": 1.1061793755550202,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 1.3959,
+      "step": 346
+    },
+    {
+      "epoch": 1.4055696202531647,
+      "grad_norm": 0.9098580926973467,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 1.3804,
+      "step": 347
+    },
+    {
+      "epoch": 1.409620253164557,
+      "grad_norm": 1.0058051524815628,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 1.3757,
+      "step": 348
+    },
+    {
+      "epoch": 1.4136708860759493,
+      "grad_norm": 0.8676237769219604,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 1.3749,
+      "step": 349
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 0.9902604918024219,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 1.3949,
+      "step": 350
+    },
+    {
+      "epoch": 1.4217721518987343,
+      "grad_norm": 0.8048909185702791,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 1.3877,
+      "step": 351
+    },
+    {
+      "epoch": 1.4258227848101266,
+      "grad_norm": 0.9163428403134137,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 1.3606,
+      "step": 352
+    },
+    {
+      "epoch": 1.4298734177215189,
+      "grad_norm": 0.6932311665090758,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 1.382,
+      "step": 353
+    },
+    {
+      "epoch": 1.4339240506329114,
+      "grad_norm": 0.7719416245304481,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 1.3815,
+      "step": 354
+    },
+    {
+      "epoch": 1.437974683544304,
+      "grad_norm": 0.7225187175705734,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 1.3799,
+      "step": 355
+    },
+    {
+      "epoch": 1.4420253164556962,
+      "grad_norm": 0.7124681607188053,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 1.3881,
+      "step": 356
+    },
+    {
+      "epoch": 1.4460759493670885,
+      "grad_norm": 0.8018052628361483,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 1.3361,
+      "step": 357
+    },
+    {
+      "epoch": 1.450126582278481,
+      "grad_norm": 0.7932808376994491,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 1.4031,
+      "step": 358
+    },
+    {
+      "epoch": 1.4541772151898735,
+      "grad_norm": 0.9339392033820825,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 1.4199,
+      "step": 359
+    },
+    {
+      "epoch": 1.4582278481012658,
+      "grad_norm": 0.7509407822635216,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 1.3687,
+      "step": 360
+    },
+    {
+      "epoch": 1.4622784810126581,
+      "grad_norm": 0.5934224236576059,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 1.337,
+      "step": 361
+    },
+    {
+      "epoch": 1.4663291139240506,
+      "grad_norm": 0.8078513521430044,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 1.4228,
+      "step": 362
+    },
+    {
+      "epoch": 1.4703797468354431,
+      "grad_norm": 0.599316375020281,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 1.3708,
+      "step": 363
+    },
+    {
+      "epoch": 1.4744303797468354,
+      "grad_norm": 0.6439076616059256,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 1.3987,
+      "step": 364
+    },
+    {
+      "epoch": 1.4784810126582277,
+      "grad_norm": 0.6828543059791569,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 1.3384,
+      "step": 365
+    },
+    {
+      "epoch": 1.4825316455696202,
+      "grad_norm": 0.601509598899569,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 1.3741,
+      "step": 366
+    },
+    {
+      "epoch": 1.4865822784810128,
+      "grad_norm": 0.6484875606737167,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 1.3341,
+      "step": 367
+    },
+    {
+      "epoch": 1.490632911392405,
+      "grad_norm": 0.6616979697866873,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 1.3535,
+      "step": 368
+    },
+    {
+      "epoch": 1.4946835443037974,
+      "grad_norm": 0.6124128161881657,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 1.4016,
+      "step": 369
+    },
+    {
+      "epoch": 1.4987341772151899,
+      "grad_norm": 0.5816269574695356,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 1.3428,
+      "step": 370
+    },
+    {
+      "epoch": 1.5027848101265824,
+      "grad_norm": 0.5989294481450299,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 1.3838,
+      "step": 371
+    },
+    {
+      "epoch": 1.5068354430379747,
+      "grad_norm": 0.59749260043039,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 1.3862,
+      "step": 372
+    },
+    {
+      "epoch": 1.510886075949367,
+      "grad_norm": 0.5520061686769593,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 1.3827,
+      "step": 373
+    },
+    {
+      "epoch": 1.5149367088607595,
+      "grad_norm": 0.6499806547118789,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 1.3441,
+      "step": 374
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 0.6314651412474097,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 1.3768,
+      "step": 375
+    },
+    {
+      "epoch": 1.5230379746835443,
+      "grad_norm": 0.6990138966820639,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 1.3863,
+      "step": 376
+    },
+    {
+      "epoch": 1.5270886075949366,
+      "grad_norm": 0.6447304157754126,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 1.3646,
+      "step": 377
+    },
+    {
+      "epoch": 1.531139240506329,
+      "grad_norm": 0.6189472291990576,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 1.3608,
+      "step": 378
+    },
+    {
+      "epoch": 1.5351898734177216,
+      "grad_norm": 0.5665590617703401,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 1.3636,
+      "step": 379
+    },
+    {
+      "epoch": 1.539240506329114,
+      "grad_norm": 0.5920435001530503,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 1.3885,
+      "step": 380
+    },
+    {
+      "epoch": 1.5432911392405062,
+      "grad_norm": 0.6414989918181626,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 1.3783,
+      "step": 381
+    },
+    {
+      "epoch": 1.5473417721518987,
+      "grad_norm": 0.681314257700305,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 1.3469,
+      "step": 382
+    },
+    {
+      "epoch": 1.5513924050632912,
+      "grad_norm": 0.6663762622110113,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 1.3613,
+      "step": 383
+    },
+    {
+      "epoch": 1.5554430379746835,
+      "grad_norm": 0.6911868001273052,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 1.3941,
+      "step": 384
+    },
+    {
+      "epoch": 1.5594936708860758,
+      "grad_norm": 0.6192660111314118,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 1.4061,
+      "step": 385
+    },
+    {
+      "epoch": 1.5635443037974683,
+      "grad_norm": 0.6399087859820104,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 1.3912,
+      "step": 386
+    },
+    {
+      "epoch": 1.5675949367088609,
+      "grad_norm": 0.6521913630577976,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 1.3695,
+      "step": 387
+    },
+    {
+      "epoch": 1.5716455696202531,
+      "grad_norm": 0.6713845200737153,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 1.4105,
+      "step": 388
+    },
+    {
+      "epoch": 1.5756962025316454,
+      "grad_norm": 0.560219774852791,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 1.3881,
+      "step": 389
+    },
+    {
+      "epoch": 1.579746835443038,
+      "grad_norm": 0.640193654176339,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 1.4114,
+      "step": 390
+    },
+    {
+      "epoch": 1.5837974683544305,
+      "grad_norm": 0.6268024733216648,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 1.3658,
+      "step": 391
+    },
+    {
+      "epoch": 1.5878481012658228,
+      "grad_norm": 0.5974471270233714,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 1.3541,
+      "step": 392
+    },
+    {
+      "epoch": 1.591898734177215,
+      "grad_norm": 0.5365002050064086,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 1.3874,
+      "step": 393
+    },
+    {
+      "epoch": 1.5959493670886076,
+      "grad_norm": 0.6639295921191062,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 1.4027,
+      "step": 394
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6006530330546203,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 1.3854,
+      "step": 395
+    },
+    {
+      "epoch": 1.6040506329113924,
+      "grad_norm": 0.5441531279543648,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 1.3543,
+      "step": 396
+    },
+    {
+      "epoch": 1.6081012658227847,
+      "grad_norm": 0.6271027818391129,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 1.4008,
+      "step": 397
+    },
+    {
+      "epoch": 1.6121518987341772,
+      "grad_norm": 0.6325044693225313,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 1.3845,
+      "step": 398
+    },
+    {
+      "epoch": 1.6162025316455697,
+      "grad_norm": 0.6577309680811805,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 1.3648,
+      "step": 399
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 0.7330865909849685,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 1.3784,
+      "step": 400
+    },
+    {
+      "epoch": 1.6243037974683543,
+      "grad_norm": 0.7114206040788229,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 1.3858,
+      "step": 401
+    },
+    {
+      "epoch": 1.6283544303797468,
+      "grad_norm": 0.7424348670251691,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 1.3817,
+      "step": 402
+    },
+    {
+      "epoch": 1.6324050632911393,
+      "grad_norm": 0.7695587738130143,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 1.4258,
+      "step": 403
+    },
+    {
+      "epoch": 1.6364556962025316,
+      "grad_norm": 0.6917862118367338,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 1.3464,
+      "step": 404
+    },
+    {
+      "epoch": 1.640506329113924,
+      "grad_norm": 0.7212927198428326,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 1.3853,
+      "step": 405
+    },
+    {
+      "epoch": 1.6445569620253164,
+      "grad_norm": 0.6342569392452794,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 1.4573,
+      "step": 406
+    },
+    {
+      "epoch": 1.648607594936709,
+      "grad_norm": 0.8400717658440706,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 1.4314,
+      "step": 407
+    },
+    {
+      "epoch": 1.6526582278481012,
+      "grad_norm": 0.6877076942254335,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 1.4234,
+      "step": 408
+    },
+    {
+      "epoch": 1.6567088607594935,
+      "grad_norm": 0.7374952951580164,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 1.4046,
+      "step": 409
+    },
+    {
+      "epoch": 1.660759493670886,
+      "grad_norm": 0.7422462871154119,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 1.3832,
+      "step": 410
+    },
+    {
+      "epoch": 1.6648101265822786,
+      "grad_norm": 0.6566595201866184,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 1.3757,
+      "step": 411
+    },
+    {
+      "epoch": 1.6688607594936709,
+      "grad_norm": 0.7158688166818095,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 1.3982,
+      "step": 412
+    },
+    {
+      "epoch": 1.6729113924050631,
+      "grad_norm": 0.6344971814365823,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 1.4215,
+      "step": 413
+    },
+    {
+      "epoch": 1.6769620253164557,
+      "grad_norm": 0.7546447256195772,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 1.3905,
+      "step": 414
+    },
+    {
+      "epoch": 1.6810126582278482,
+      "grad_norm": 0.5921033896543951,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 1.3695,
+      "step": 415
+    },
+    {
+      "epoch": 1.6850632911392405,
+      "grad_norm": 0.6446441278295897,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 1.3557,
+      "step": 416
+    },
+    {
+      "epoch": 1.6891139240506328,
+      "grad_norm": 0.6875797461692409,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 1.3595,
+      "step": 417
+    },
+    {
+      "epoch": 1.6931645569620253,
+      "grad_norm": 0.696191232287149,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 1.3876,
+      "step": 418
+    },
+    {
+      "epoch": 1.6972151898734178,
+      "grad_norm": 0.7231955988324984,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 1.4023,
+      "step": 419
+    },
+    {
+      "epoch": 1.70126582278481,
+      "grad_norm": 0.7790594407812972,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 1.4052,
+      "step": 420
+    },
+    {
+      "epoch": 1.7053164556962024,
+      "grad_norm": 0.7761460791601564,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 1.3956,
+      "step": 421
+    },
+    {
+      "epoch": 1.709367088607595,
+      "grad_norm": 0.7323010765881495,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 1.4028,
+      "step": 422
+    },
+    {
+      "epoch": 1.7134177215189874,
+      "grad_norm": 0.6118255682086515,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 1.3589,
+      "step": 423
+    },
+    {
+      "epoch": 1.7174683544303797,
+      "grad_norm": 0.6955538732183427,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 1.387,
+      "step": 424
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 0.6698316241620522,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 1.3954,
+      "step": 425
+    },
+    {
+      "epoch": 1.7255696202531645,
+      "grad_norm": 0.5264691518045982,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 1.3824,
+      "step": 426
+    },
+    {
+      "epoch": 1.729620253164557,
+      "grad_norm": 0.7475105919297811,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 1.387,
+      "step": 427
+    },
+    {
+      "epoch": 1.7336708860759493,
+      "grad_norm": 0.5830624067938677,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 1.3938,
+      "step": 428
+    },
+    {
+      "epoch": 1.7377215189873416,
+      "grad_norm": 0.76490538262703,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 1.3751,
+      "step": 429
+    },
+    {
+      "epoch": 1.7417721518987341,
+      "grad_norm": 0.6471816863586256,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 1.3531,
+      "step": 430
+    },
+    {
+      "epoch": 1.7458227848101266,
+      "grad_norm": 0.701768148205461,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 1.3482,
+      "step": 431
+    },
+    {
+      "epoch": 1.749873417721519,
+      "grad_norm": 0.6773460772946405,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 1.3872,
+      "step": 432
+    },
+    {
+      "epoch": 1.7539240506329112,
+      "grad_norm": 0.703600398986762,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 1.4076,
+      "step": 433
+    },
+    {
+      "epoch": 1.7579746835443038,
+      "grad_norm": 0.6358567003558843,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 1.364,
+      "step": 434
+    },
+    {
+      "epoch": 1.7620253164556963,
+      "grad_norm": 0.7476534173203234,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 1.3869,
+      "step": 435
+    },
+    {
+      "epoch": 1.7660759493670886,
+      "grad_norm": 0.6246716172908963,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 1.3681,
+      "step": 436
+    },
+    {
+      "epoch": 1.7701265822784809,
+      "grad_norm": 0.7603346011658595,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 1.3847,
+      "step": 437
+    },
+    {
+      "epoch": 1.7741772151898734,
+      "grad_norm": 0.705528255520313,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 1.3461,
+      "step": 438
+    },
+    {
+      "epoch": 1.7782278481012659,
+      "grad_norm": 0.7257783598014371,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 1.3629,
+      "step": 439
+    },
+    {
+      "epoch": 1.7822784810126582,
+      "grad_norm": 0.7320379245343865,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 1.3685,
+      "step": 440
+    },
+    {
+      "epoch": 1.7863291139240505,
+      "grad_norm": 0.7547936682068282,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 1.4069,
+      "step": 441
+    },
+    {
+      "epoch": 1.790379746835443,
+      "grad_norm": 0.6972440991983664,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 1.3628,
+      "step": 442
+    },
+    {
+      "epoch": 1.7944303797468355,
+      "grad_norm": 0.6836097896479717,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 1.3736,
+      "step": 443
+    },
+    {
+      "epoch": 1.7984810126582278,
+      "grad_norm": 0.7002062496784981,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 1.353,
+      "step": 444
+    },
+    {
+      "epoch": 1.80253164556962,
+      "grad_norm": 0.6275991941565772,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 1.3429,
+      "step": 445
+    },
+    {
+      "epoch": 1.8065822784810126,
+      "grad_norm": 0.7266443938914576,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 1.3581,
+      "step": 446
+    },
+    {
+      "epoch": 1.8106329113924051,
+      "grad_norm": 0.536074890109397,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 1.383,
+      "step": 447
+    },
+    {
+      "epoch": 1.8146835443037974,
+      "grad_norm": 0.7281299615534467,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 1.3913,
+      "step": 448
+    },
+    {
+      "epoch": 1.81873417721519,
+      "grad_norm": 0.6105667508411966,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 1.3761,
+      "step": 449
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 0.6699026626133171,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 1.3634,
+      "step": 450
+    },
+    {
+      "epoch": 1.8268354430379747,
+      "grad_norm": 0.6248962517736966,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 1.3831,
+      "step": 451
+    },
+    {
+      "epoch": 1.8308860759493673,
+      "grad_norm": 0.6531972269278807,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 1.3592,
+      "step": 452
+    },
+    {
+      "epoch": 1.8349367088607595,
+      "grad_norm": 0.6358217579104104,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 1.4032,
+      "step": 453
+    },
+    {
+      "epoch": 1.8389873417721518,
+      "grad_norm": 0.6726144372776083,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 1.3863,
+      "step": 454
+    },
+    {
+      "epoch": 1.8430379746835444,
+      "grad_norm": 0.6535878275758301,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 1.3935,
+      "step": 455
+    },
+    {
+      "epoch": 1.8470886075949369,
+      "grad_norm": 0.7168792958361156,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 1.4035,
+      "step": 456
+    },
+    {
+      "epoch": 1.8511392405063292,
+      "grad_norm": 0.6919959703603651,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 1.3647,
+      "step": 457
+    },
+    {
+      "epoch": 1.8551898734177215,
+      "grad_norm": 0.6452630287763563,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 1.3813,
+      "step": 458
+    },
+    {
+      "epoch": 1.859240506329114,
+      "grad_norm": 0.7125411850216663,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 1.3793,
+      "step": 459
+    },
+    {
+      "epoch": 1.8632911392405065,
+      "grad_norm": 0.5965316617077292,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 1.3484,
+      "step": 460
+    },
+    {
+      "epoch": 1.8673417721518988,
+      "grad_norm": 0.7308845450239035,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 1.3821,
+      "step": 461
+    },
+    {
+      "epoch": 1.871392405063291,
+      "grad_norm": 0.630463397970247,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 1.4104,
+      "step": 462
+    },
+    {
+      "epoch": 1.8754430379746836,
+      "grad_norm": 0.6782860529349658,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 1.4005,
+      "step": 463
+    },
+    {
+      "epoch": 1.879493670886076,
+      "grad_norm": 0.573676820803225,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 1.3798,
+      "step": 464
+    },
+    {
+      "epoch": 1.8835443037974684,
+      "grad_norm": 0.7204434215157538,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 1.3916,
+      "step": 465
+    },
+    {
+      "epoch": 1.8875949367088607,
+      "grad_norm": 0.6488229505363499,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 1.3948,
+      "step": 466
+    },
+    {
+      "epoch": 1.8916455696202532,
+      "grad_norm": 0.7219707922651428,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 1.3768,
+      "step": 467
+    },
+    {
+      "epoch": 1.8956962025316457,
+      "grad_norm": 0.6478300677633296,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 1.3557,
+      "step": 468
+    },
+    {
+      "epoch": 1.899746835443038,
+      "grad_norm": 0.6647006707868139,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 1.3872,
+      "step": 469
+    },
+    {
+      "epoch": 1.9037974683544303,
+      "grad_norm": 0.6511469663945285,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 1.3432,
+      "step": 470
+    },
+    {
+      "epoch": 1.9078481012658228,
+      "grad_norm": 0.6023094460979235,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 1.3947,
+      "step": 471
+    },
+    {
+      "epoch": 1.9118987341772153,
+      "grad_norm": 0.7704292217839646,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 1.3716,
+      "step": 472
+    },
+    {
+      "epoch": 1.9159493670886076,
+      "grad_norm": 0.5603822966185269,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 1.3894,
+      "step": 473
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6283701975720616,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 1.4002,
+      "step": 474
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 0.6147853133554295,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 1.3857,
+      "step": 475
+    },
+    {
+      "epoch": 1.928101265822785,
+      "grad_norm": 0.621405180355382,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 1.384,
+      "step": 476
+    },
+    {
+      "epoch": 1.9321518987341773,
+      "grad_norm": 0.6470218204634333,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 1.392,
+      "step": 477
+    },
+    {
+      "epoch": 1.9362025316455695,
+      "grad_norm": 0.6215921966837888,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 1.4192,
+      "step": 478
+    },
+    {
+      "epoch": 1.940253164556962,
+      "grad_norm": 0.6945673739249115,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 1.3878,
+      "step": 479
+    },
+    {
+      "epoch": 1.9443037974683546,
+      "grad_norm": 0.6321282588527939,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 1.4103,
+      "step": 480
+    },
+    {
+      "epoch": 1.9483544303797469,
+      "grad_norm": 0.6305623562984357,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 1.3888,
+      "step": 481
+    },
+    {
+      "epoch": 1.9524050632911392,
+      "grad_norm": 0.6976646430362864,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 1.3929,
+      "step": 482
+    },
+    {
+      "epoch": 1.9564556962025317,
+      "grad_norm": 0.5429775280433221,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 1.3699,
+      "step": 483
+    },
+    {
+      "epoch": 1.9605063291139242,
+      "grad_norm": 0.612106003601725,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 1.3852,
+      "step": 484
+    },
+    {
+      "epoch": 1.9645569620253165,
+      "grad_norm": 0.6764495963655826,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 1.4021,
+      "step": 485
+    },
+    {
+      "epoch": 1.9686075949367088,
+      "grad_norm": 0.7121669940977892,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 1.363,
+      "step": 486
+    },
+    {
+      "epoch": 1.9726582278481013,
+      "grad_norm": 0.7223651303075119,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 1.4041,
+      "step": 487
+    },
+    {
+      "epoch": 1.9767088607594938,
+      "grad_norm": 0.6119141962059248,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 1.3508,
+      "step": 488
+    },
+    {
+      "epoch": 1.980759493670886,
+      "grad_norm": 0.5850366450510703,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 1.3601,
+      "step": 489
+    },
+    {
+      "epoch": 1.9848101265822784,
+      "grad_norm": 0.5824910012092132,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 1.391,
+      "step": 490
+    },
+    {
+      "epoch": 1.988860759493671,
+      "grad_norm": 0.6737059013240063,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 1.3802,
+      "step": 491
+    },
+    {
+      "epoch": 1.9929113924050634,
+      "grad_norm": 0.7191624457189331,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.3925,
+      "step": 492
+    },
+    {
+      "epoch": 1.9969620253164557,
+      "grad_norm": 0.7100626340083893,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 1.3793,
+      "step": 493
+    },
+    {
+      "epoch": 2.001012658227848,
+      "grad_norm": 0.6724106229195012,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 1.3738,
+      "step": 494
+    },
+    {
+      "epoch": 2.0050632911392405,
+      "grad_norm": 0.7826889281860347,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 1.2562,
+      "step": 495
+    },
+    {
+      "epoch": 2.009113924050633,
+      "grad_norm": 0.9378554252064403,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 1.2129,
+      "step": 496
+    },
+    {
+      "epoch": 2.013164556962025,
+      "grad_norm": 0.8755281610037413,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 1.2549,
+      "step": 497
+    },
+    {
+      "epoch": 2.0172151898734176,
+      "grad_norm": 0.9184239446202259,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 1.2353,
+      "step": 498
+    },
+    {
+      "epoch": 2.02126582278481,
+      "grad_norm": 0.7634020553374967,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 1.2352,
+      "step": 499
+    },
+    {
+      "epoch": 2.0253164556962027,
+      "grad_norm": 0.7610101696561771,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 1.2622,
+      "step": 500
+    },
+    {
+      "epoch": 2.0293670886075947,
+      "grad_norm": 0.834189951017623,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 1.2231,
+      "step": 501
+    },
+    {
+      "epoch": 2.0334177215189873,
+      "grad_norm": 0.6259035787999802,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 1.2171,
+      "step": 502
+    },
+    {
+      "epoch": 2.0374683544303798,
+      "grad_norm": 0.9324542921609609,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 1.2322,
+      "step": 503
+    },
+    {
+      "epoch": 2.0415189873417723,
+      "grad_norm": 0.7446554973530657,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 1.27,
+      "step": 504
+    },
+    {
+      "epoch": 2.0455696202531644,
+      "grad_norm": 0.7908926992385323,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 1.2323,
+      "step": 505
+    },
+    {
+      "epoch": 2.049620253164557,
+      "grad_norm": 0.8423357027923903,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 1.2808,
+      "step": 506
+    },
+    {
+      "epoch": 2.0536708860759494,
+      "grad_norm": 0.7455763551519468,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 1.2731,
+      "step": 507
+    },
+    {
+      "epoch": 2.057721518987342,
+      "grad_norm": 0.9549222760837461,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 1.2324,
+      "step": 508
+    },
+    {
+      "epoch": 2.061772151898734,
+      "grad_norm": 0.746708561843258,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 1.2444,
+      "step": 509
+    },
+    {
+      "epoch": 2.0658227848101265,
+      "grad_norm": 0.7570785416973325,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 1.2633,
+      "step": 510
+    },
+    {
+      "epoch": 2.069873417721519,
+      "grad_norm": 0.6973815214546586,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 1.2068,
+      "step": 511
+    },
+    {
+      "epoch": 2.0739240506329115,
+      "grad_norm": 0.7782796156272275,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 1.2398,
+      "step": 512
+    },
+    {
+      "epoch": 2.0779746835443036,
+      "grad_norm": 0.7741504492910508,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 1.2375,
+      "step": 513
+    },
+    {
+      "epoch": 2.082025316455696,
+      "grad_norm": 0.745079732427537,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 1.2588,
+      "step": 514
+    },
+    {
+      "epoch": 2.0860759493670886,
+      "grad_norm": 0.6349570989565555,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 1.2874,
+      "step": 515
+    },
+    {
+      "epoch": 2.090126582278481,
+      "grad_norm": 0.7091846365035692,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 1.2498,
+      "step": 516
+    },
+    {
+      "epoch": 2.094177215189873,
+      "grad_norm": 0.6482636381929426,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 1.2584,
+      "step": 517
+    },
+    {
+      "epoch": 2.0982278481012657,
+      "grad_norm": 0.6020920946172179,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 1.2545,
+      "step": 518
+    },
+    {
+      "epoch": 2.1022784810126582,
+      "grad_norm": 0.6422571755316594,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 1.2751,
+      "step": 519
+    },
+    {
+      "epoch": 2.1063291139240508,
+      "grad_norm": 0.6502946937712737,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 1.2577,
+      "step": 520
+    },
+    {
+      "epoch": 2.110379746835443,
+      "grad_norm": 0.5972821772085207,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 1.2572,
+      "step": 521
+    },
+    {
+      "epoch": 2.1144303797468353,
+      "grad_norm": 0.664914538708513,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 1.2591,
+      "step": 522
+    },
+    {
+      "epoch": 2.118481012658228,
+      "grad_norm": 0.7662137487229386,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 1.2674,
+      "step": 523
+    },
+    {
+      "epoch": 2.1225316455696204,
+      "grad_norm": 0.6576877881642054,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 1.2214,
+      "step": 524
+    },
+    {
+      "epoch": 2.1265822784810124,
+      "grad_norm": 0.7745018450131503,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 1.2475,
+      "step": 525
+    },
+    {
+      "epoch": 2.130632911392405,
+      "grad_norm": 0.7673456497284491,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 1.2711,
+      "step": 526
+    },
+    {
+      "epoch": 2.1346835443037975,
+      "grad_norm": 0.8327530499784673,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 1.2689,
+      "step": 527
+    },
+    {
+      "epoch": 2.13873417721519,
+      "grad_norm": 0.7962358636613416,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 1.2551,
+      "step": 528
+    },
+    {
+      "epoch": 2.1427848101265825,
+      "grad_norm": 0.9627499720606982,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 1.2109,
+      "step": 529
+    },
+    {
+      "epoch": 2.1468354430379746,
+      "grad_norm": 0.6339515501759836,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 1.2277,
+      "step": 530
+    },
+    {
+      "epoch": 2.150886075949367,
+      "grad_norm": 0.7338293136705182,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 1.2459,
+      "step": 531
+    },
+    {
+      "epoch": 2.1549367088607596,
+      "grad_norm": 0.6809360945348591,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 1.2407,
+      "step": 532
+    },
+    {
+      "epoch": 2.1589873417721517,
+      "grad_norm": 0.6143894928285416,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 1.2325,
+      "step": 533
+    },
+    {
+      "epoch": 2.163037974683544,
+      "grad_norm": 0.652749731457271,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 1.2822,
+      "step": 534
+    },
+    {
+      "epoch": 2.1670886075949367,
+      "grad_norm": 0.6285371237431211,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 1.2682,
+      "step": 535
+    },
+    {
+      "epoch": 2.1711392405063292,
+      "grad_norm": 0.715725355136172,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 1.2417,
+      "step": 536
+    },
+    {
+      "epoch": 2.1751898734177217,
+      "grad_norm": 0.5539095003537862,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 1.3037,
+      "step": 537
+    },
+    {
+      "epoch": 2.179240506329114,
+      "grad_norm": 0.7260162419742868,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 1.258,
+      "step": 538
+    },
+    {
+      "epoch": 2.1832911392405063,
+      "grad_norm": 0.5539164588389405,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 1.2745,
+      "step": 539
+    },
+    {
+      "epoch": 2.187341772151899,
+      "grad_norm": 0.7353978137764573,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 1.2894,
+      "step": 540
+    },
+    {
+      "epoch": 2.191392405063291,
+      "grad_norm": 0.5018876035790611,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 1.2313,
+      "step": 541
+    },
+    {
+      "epoch": 2.1954430379746834,
+      "grad_norm": 0.6831785491400044,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 1.2391,
+      "step": 542
+    },
+    {
+      "epoch": 2.199493670886076,
+      "grad_norm": 0.5989702816511421,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 1.2547,
+      "step": 543
+    },
+    {
+      "epoch": 2.2035443037974685,
+      "grad_norm": 0.5773146763724137,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 1.2432,
+      "step": 544
+    },
+    {
+      "epoch": 2.207594936708861,
+      "grad_norm": 0.7966634165515225,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 1.2886,
+      "step": 545
+    },
+    {
+      "epoch": 2.211645569620253,
+      "grad_norm": 0.5558261667821713,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 1.2852,
+      "step": 546
+    },
+    {
+      "epoch": 2.2156962025316456,
+      "grad_norm": 0.7159723724646254,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 1.2307,
+      "step": 547
+    },
+    {
+      "epoch": 2.219746835443038,
+      "grad_norm": 0.6030275135391212,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 1.2161,
+      "step": 548
+    },
+    {
+      "epoch": 2.22379746835443,
+      "grad_norm": 0.6489850649140613,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 1.2764,
+      "step": 549
+    },
+    {
+      "epoch": 2.2278481012658227,
+      "grad_norm": 0.6452967184888317,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 1.2803,
+      "step": 550
+    },
+    {
+      "epoch": 2.231898734177215,
+      "grad_norm": 0.5749348479373442,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 1.2518,
+      "step": 551
+    },
+    {
+      "epoch": 2.2359493670886077,
+      "grad_norm": 0.6357165239099947,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 1.2503,
+      "step": 552
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5988199174566513,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 1.2229,
+      "step": 553
+    },
+    {
+      "epoch": 2.2440506329113923,
+      "grad_norm": 0.6145852701258994,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 1.2373,
+      "step": 554
+    },
+    {
+      "epoch": 2.248101265822785,
+      "grad_norm": 0.6068018197208809,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 1.2639,
+      "step": 555
+    },
+    {
+      "epoch": 2.2521518987341773,
+      "grad_norm": 0.5666294159838443,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 1.2965,
+      "step": 556
+    },
+    {
+      "epoch": 2.2562025316455694,
+      "grad_norm": 0.6307183892894034,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 1.2676,
+      "step": 557
+    },
+    {
+      "epoch": 2.260253164556962,
+      "grad_norm": 0.4867008647509768,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 1.2837,
+      "step": 558
+    },
+    {
+      "epoch": 2.2643037974683544,
+      "grad_norm": 0.5798576781548518,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 1.248,
+      "step": 559
+    },
+    {
+      "epoch": 2.268354430379747,
+      "grad_norm": 0.5420899030427285,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 1.2532,
+      "step": 560
+    },
+    {
+      "epoch": 2.2724050632911394,
+      "grad_norm": 0.5711168167220559,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 1.2638,
+      "step": 561
+    },
+    {
+      "epoch": 2.2764556962025315,
+      "grad_norm": 0.5144272639744324,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 1.2511,
+      "step": 562
+    },
+    {
+      "epoch": 2.280506329113924,
+      "grad_norm": 0.5176037151758793,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 1.2407,
+      "step": 563
+    },
+    {
+      "epoch": 2.2845569620253166,
+      "grad_norm": 0.5746882994067076,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 1.3008,
+      "step": 564
+    },
+    {
+      "epoch": 2.2886075949367086,
+      "grad_norm": 0.5061641479839007,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 1.2581,
+      "step": 565
+    },
+    {
+      "epoch": 2.292658227848101,
+      "grad_norm": 0.5370576111871693,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 1.295,
+      "step": 566
+    },
+    {
+      "epoch": 2.2967088607594937,
+      "grad_norm": 0.5746312474749452,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 1.2148,
+      "step": 567
+    },
+    {
+      "epoch": 2.300759493670886,
+      "grad_norm": 0.5844530949161575,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 1.264,
+      "step": 568
+    },
+    {
+      "epoch": 2.3048101265822787,
+      "grad_norm": 0.5912650322860356,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 1.2652,
+      "step": 569
+    },
+    {
+      "epoch": 2.3088607594936708,
+      "grad_norm": 0.5686690458156755,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 1.2006,
+      "step": 570
+    },
+    {
+      "epoch": 2.3129113924050633,
+      "grad_norm": 0.6609787364004669,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 1.2387,
+      "step": 571
+    },
+    {
+      "epoch": 2.316962025316456,
+      "grad_norm": 0.5830559183750063,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 1.2688,
+      "step": 572
+    },
+    {
+      "epoch": 2.321012658227848,
+      "grad_norm": 0.5895755108199916,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 1.2472,
+      "step": 573
+    },
+    {
+      "epoch": 2.3250632911392404,
+      "grad_norm": 0.6176597272979331,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 1.2791,
+      "step": 574
+    },
+    {
+      "epoch": 2.329113924050633,
+      "grad_norm": 0.5105560769186503,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 1.2533,
+      "step": 575
+    },
+    {
+      "epoch": 2.3331645569620254,
+      "grad_norm": 0.6581460808136933,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 1.2361,
+      "step": 576
+    },
+    {
+      "epoch": 2.337215189873418,
+      "grad_norm": 0.5769609313918609,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 1.2855,
+      "step": 577
+    },
+    {
+      "epoch": 2.34126582278481,
+      "grad_norm": 0.6133055501267154,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 1.291,
+      "step": 578
+    },
+    {
+      "epoch": 2.3453164556962025,
+      "grad_norm": 0.6948368127658118,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 1.223,
+      "step": 579
+    },
+    {
+      "epoch": 2.349367088607595,
+      "grad_norm": 0.5811936276722878,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 1.2528,
+      "step": 580
+    },
+    {
+      "epoch": 2.353417721518987,
+      "grad_norm": 0.7242930560491645,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 1.2391,
+      "step": 581
+    },
+    {
+      "epoch": 2.3574683544303796,
+      "grad_norm": 0.5606100499492793,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 1.2993,
+      "step": 582
+    },
+    {
+      "epoch": 2.361518987341772,
+      "grad_norm": 0.6696559357123385,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 1.2621,
+      "step": 583
+    },
+    {
+      "epoch": 2.3655696202531646,
+      "grad_norm": 0.6661059741195469,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 1.2732,
+      "step": 584
+    },
+    {
+      "epoch": 2.369620253164557,
+      "grad_norm": 0.5896168954861648,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 1.2452,
+      "step": 585
+    },
+    {
+      "epoch": 2.3736708860759492,
+      "grad_norm": 0.505937308608468,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 1.2378,
+      "step": 586
+    },
+    {
+      "epoch": 2.3777215189873417,
+      "grad_norm": 0.5724231596867861,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 1.2531,
+      "step": 587
+    },
+    {
+      "epoch": 2.3817721518987343,
+      "grad_norm": 0.5191283656436099,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 1.2357,
+      "step": 588
+    },
+    {
+      "epoch": 2.3858227848101268,
+      "grad_norm": 0.6312050017175861,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 1.2535,
+      "step": 589
+    },
+    {
+      "epoch": 2.389873417721519,
+      "grad_norm": 0.5733199936926758,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 1.2412,
+      "step": 590
+    },
+    {
+      "epoch": 2.3939240506329114,
+      "grad_norm": 0.5820044508240174,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 1.286,
+      "step": 591
+    },
+    {
+      "epoch": 2.397974683544304,
+      "grad_norm": 0.5635290661248097,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 1.3141,
+      "step": 592
+    },
+    {
+      "epoch": 2.4020253164556964,
+      "grad_norm": 0.5868003773066136,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 1.2768,
+      "step": 593
+    },
+    {
+      "epoch": 2.4060759493670885,
+      "grad_norm": 0.5828468310909419,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 1.243,
+      "step": 594
+    },
+    {
+      "epoch": 2.410126582278481,
+      "grad_norm": 0.6048886387020264,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 1.2508,
+      "step": 595
+    },
+    {
+      "epoch": 2.4141772151898735,
+      "grad_norm": 0.7002032588907113,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 1.2484,
+      "step": 596
+    },
+    {
+      "epoch": 2.418227848101266,
+      "grad_norm": 0.5530470332336507,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 1.2967,
+      "step": 597
+    },
+    {
+      "epoch": 2.422278481012658,
+      "grad_norm": 0.5714387413402713,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 1.2627,
+      "step": 598
+    },
+    {
+      "epoch": 2.4263291139240506,
+      "grad_norm": 0.614141723895819,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 1.2329,
+      "step": 599
+    },
+    {
+      "epoch": 2.430379746835443,
+      "grad_norm": 0.5097516062933038,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 1.2373,
+      "step": 600
+    },
+    {
+      "epoch": 2.4344303797468356,
+      "grad_norm": 0.5122601861785979,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 1.2244,
+      "step": 601
+    },
+    {
+      "epoch": 2.4384810126582277,
+      "grad_norm": 0.5931601563780774,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 1.2831,
+      "step": 602
+    },
+    {
+      "epoch": 2.44253164556962,
+      "grad_norm": 0.579017954715373,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 1.2505,
+      "step": 603
+    },
+    {
+      "epoch": 2.4465822784810127,
+      "grad_norm": 0.5030211044882575,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 1.2608,
+      "step": 604
+    },
+    {
+      "epoch": 2.4506329113924052,
+      "grad_norm": 0.5829030732937696,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 1.273,
+      "step": 605
+    },
+    {
+      "epoch": 2.4546835443037973,
+      "grad_norm": 0.5954961649606004,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 1.3187,
+      "step": 606
+    },
+    {
+      "epoch": 2.45873417721519,
+      "grad_norm": 0.562129575355707,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 1.2922,
+      "step": 607
+    },
+    {
+      "epoch": 2.4627848101265823,
+      "grad_norm": 0.5701612780940113,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 1.2682,
+      "step": 608
+    },
+    {
+      "epoch": 2.466835443037975,
+      "grad_norm": 0.5590999940841528,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 1.2597,
+      "step": 609
+    },
+    {
+      "epoch": 2.470886075949367,
+      "grad_norm": 0.5081446011592508,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 1.2347,
+      "step": 610
+    },
+    {
+      "epoch": 2.4749367088607594,
+      "grad_norm": 0.6228707141380453,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 1.2255,
+      "step": 611
+    },
+    {
+      "epoch": 2.478987341772152,
+      "grad_norm": 0.5125134292735664,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 1.2591,
+      "step": 612
+    },
+    {
+      "epoch": 2.4830379746835445,
+      "grad_norm": 0.5626938535883754,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 1.2804,
+      "step": 613
+    },
+    {
+      "epoch": 2.4870886075949366,
+      "grad_norm": 0.5724138813407502,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 1.2738,
+      "step": 614
+    },
+    {
+      "epoch": 2.491139240506329,
+      "grad_norm": 0.5379634701785702,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 1.2692,
+      "step": 615
+    },
+    {
+      "epoch": 2.4951898734177216,
+      "grad_norm": 0.5807699782916635,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 1.2581,
+      "step": 616
+    },
+    {
+      "epoch": 2.499240506329114,
+      "grad_norm": 0.5328358958186455,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 1.247,
+      "step": 617
+    },
+    {
+      "epoch": 2.503291139240506,
+      "grad_norm": 0.5381048717204003,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 1.2921,
+      "step": 618
+    },
+    {
+      "epoch": 2.5073417721518987,
+      "grad_norm": 0.5963546685490935,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 1.2761,
+      "step": 619
+    },
+    {
+      "epoch": 2.511392405063291,
+      "grad_norm": 0.4717318411195807,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 1.2389,
+      "step": 620
+    },
+    {
+      "epoch": 2.5154430379746833,
+      "grad_norm": 0.5600625070949186,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 1.3124,
+      "step": 621
+    },
+    {
+      "epoch": 2.519493670886076,
+      "grad_norm": 0.5015449994241477,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 1.2496,
+      "step": 622
+    },
+    {
+      "epoch": 2.5235443037974683,
+      "grad_norm": 0.5392031832332682,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 1.2325,
+      "step": 623
+    },
+    {
+      "epoch": 2.527594936708861,
+      "grad_norm": 0.5503708178150819,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 1.2379,
+      "step": 624
+    },
+    {
+      "epoch": 2.5316455696202533,
+      "grad_norm": 0.5929900290802681,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 1.2493,
+      "step": 625
+    },
+    {
+      "epoch": 2.5356962025316454,
+      "grad_norm": 0.5679583610785148,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 1.2795,
+      "step": 626
+    },
+    {
+      "epoch": 2.539746835443038,
+      "grad_norm": 0.625248921751712,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 1.2602,
+      "step": 627
+    },
+    {
+      "epoch": 2.5437974683544304,
+      "grad_norm": 0.5566109203383982,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 1.2245,
+      "step": 628
+    },
+    {
+      "epoch": 2.547848101265823,
+      "grad_norm": 0.5896628344118073,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 1.2156,
+      "step": 629
+    },
+    {
+      "epoch": 2.5518987341772155,
+      "grad_norm": 0.5360658187345876,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 1.2434,
+      "step": 630
+    },
+    {
+      "epoch": 2.5559493670886075,
+      "grad_norm": 0.6292219707764348,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 1.2397,
+      "step": 631
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.538105869472175,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 1.2453,
+      "step": 632
+    },
+    {
+      "epoch": 2.5640506329113926,
+      "grad_norm": 0.5283498208381575,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 1.266,
+      "step": 633
+    },
+    {
+      "epoch": 2.5681012658227846,
+      "grad_norm": 0.5629359029566833,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 1.2768,
+      "step": 634
+    },
+    {
+      "epoch": 2.572151898734177,
+      "grad_norm": 0.5707638864108977,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 1.224,
+      "step": 635
+    },
+    {
+      "epoch": 2.5762025316455697,
+      "grad_norm": 0.7015979202836846,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 1.2506,
+      "step": 636
+    },
+    {
+      "epoch": 2.580253164556962,
+      "grad_norm": 0.5754494254592747,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 1.3146,
+      "step": 637
+    },
+    {
+      "epoch": 2.5843037974683547,
+      "grad_norm": 0.6449243634133233,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 1.2495,
+      "step": 638
+    },
+    {
+      "epoch": 2.5883544303797468,
+      "grad_norm": 0.5528069118027605,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 1.2851,
+      "step": 639
+    },
+    {
+      "epoch": 2.5924050632911393,
+      "grad_norm": 0.6348143362901731,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 1.241,
+      "step": 640
+    },
+    {
+      "epoch": 2.596455696202532,
+      "grad_norm": 0.5646591579971185,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 1.2462,
+      "step": 641
+    },
+    {
+      "epoch": 2.600506329113924,
+      "grad_norm": 0.6918430021358567,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 1.2408,
+      "step": 642
+    },
+    {
+      "epoch": 2.6045569620253164,
+      "grad_norm": 0.5967335676722725,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 1.2571,
+      "step": 643
+    },
+    {
+      "epoch": 2.608607594936709,
+      "grad_norm": 0.5425981145716322,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 1.2643,
+      "step": 644
+    },
+    {
+      "epoch": 2.6126582278481014,
+      "grad_norm": 0.6605416929978356,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 1.2902,
+      "step": 645
+    },
+    {
+      "epoch": 2.616708860759494,
+      "grad_norm": 0.6362582717706048,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 1.2854,
+      "step": 646
+    },
+    {
+      "epoch": 2.620759493670886,
+      "grad_norm": 0.6032123171011724,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 1.2652,
+      "step": 647
+    },
+    {
+      "epoch": 2.6248101265822785,
+      "grad_norm": 0.5891878835294176,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 1.291,
+      "step": 648
+    },
+    {
+      "epoch": 2.628860759493671,
+      "grad_norm": 0.5937713583710792,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 1.2535,
+      "step": 649
+    },
+    {
+      "epoch": 2.632911392405063,
+      "grad_norm": 0.5369008932302479,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 1.2466,
+      "step": 650
+    },
+    {
+      "epoch": 2.6369620253164556,
+      "grad_norm": 0.5462547530592464,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 1.2361,
+      "step": 651
+    },
+    {
+      "epoch": 2.641012658227848,
+      "grad_norm": 0.5906990713143542,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 1.2218,
+      "step": 652
+    },
+    {
+      "epoch": 2.6450632911392407,
+      "grad_norm": 0.5468559964617224,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 1.2626,
+      "step": 653
+    },
+    {
+      "epoch": 2.649113924050633,
+      "grad_norm": 0.5557694865199243,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 1.254,
+      "step": 654
+    },
+    {
+      "epoch": 2.6531645569620252,
+      "grad_norm": 0.5477323677084566,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 1.2311,
+      "step": 655
+    },
+    {
+      "epoch": 2.6572151898734178,
+      "grad_norm": 0.49682495559102774,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 1.238,
+      "step": 656
+    },
+    {
+      "epoch": 2.6612658227848103,
+      "grad_norm": 0.5620619533888844,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 1.2667,
+      "step": 657
+    },
+    {
+      "epoch": 2.6653164556962023,
+      "grad_norm": 0.5573759519460361,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 1.2559,
+      "step": 658
+    },
+    {
+      "epoch": 2.669367088607595,
+      "grad_norm": 0.5156738319616055,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 1.2579,
+      "step": 659
+    },
+    {
+      "epoch": 2.6734177215189874,
+      "grad_norm": 0.5426795052523825,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 1.2215,
+      "step": 660
+    },
+    {
+      "epoch": 2.67746835443038,
+      "grad_norm": 0.5233438379585631,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 1.2783,
+      "step": 661
+    },
+    {
+      "epoch": 2.6815189873417724,
+      "grad_norm": 0.5397349409317129,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 1.2908,
+      "step": 662
+    },
+    {
+      "epoch": 2.6855696202531645,
+      "grad_norm": 0.5477597014254264,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 1.2881,
+      "step": 663
+    },
+    {
+      "epoch": 2.689620253164557,
+      "grad_norm": 0.5925614862666557,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 1.2525,
+      "step": 664
+    },
+    {
+      "epoch": 2.6936708860759495,
+      "grad_norm": 0.5922414008546718,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 1.2444,
+      "step": 665
+    },
+    {
+      "epoch": 2.6977215189873416,
+      "grad_norm": 0.5309506919594217,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 1.2538,
+      "step": 666
+    },
+    {
+      "epoch": 2.701772151898734,
+      "grad_norm": 0.6734599201019353,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 1.2439,
+      "step": 667
+    },
+    {
+      "epoch": 2.7058227848101266,
+      "grad_norm": 0.5506841943878555,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 1.2802,
+      "step": 668
+    },
+    {
+      "epoch": 2.709873417721519,
+      "grad_norm": 0.5667260434715875,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 1.2346,
+      "step": 669
+    },
+    {
+      "epoch": 2.7139240506329116,
+      "grad_norm": 0.5612014310890174,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 1.2739,
+      "step": 670
+    },
+    {
+      "epoch": 2.7179746835443037,
+      "grad_norm": 0.6737483413304547,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 1.3034,
+      "step": 671
+    },
+    {
+      "epoch": 2.7220253164556962,
+      "grad_norm": 0.5282277874082163,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 1.2629,
+      "step": 672
+    },
+    {
+      "epoch": 2.7260759493670887,
+      "grad_norm": 0.5131443148945389,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 1.2772,
+      "step": 673
+    },
+    {
+      "epoch": 2.730126582278481,
+      "grad_norm": 0.5956407233853273,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 1.271,
+      "step": 674
+    },
+    {
+      "epoch": 2.7341772151898733,
+      "grad_norm": 0.5446999610272517,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 1.2641,
+      "step": 675
+    },
+    {
+      "epoch": 2.738227848101266,
+      "grad_norm": 0.5220673536793304,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 1.2371,
+      "step": 676
+    },
+    {
+      "epoch": 2.7422784810126584,
+      "grad_norm": 0.5843742055901926,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 1.238,
+      "step": 677
+    },
+    {
+      "epoch": 2.746329113924051,
+      "grad_norm": 0.564174200964846,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 1.2435,
+      "step": 678
+    },
+    {
+      "epoch": 2.750379746835443,
+      "grad_norm": 0.6007388775170953,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 1.27,
+      "step": 679
+    },
+    {
+      "epoch": 2.7544303797468355,
+      "grad_norm": 0.5021613833658165,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 1.2636,
+      "step": 680
+    },
+    {
+      "epoch": 2.758481012658228,
+      "grad_norm": 0.4993857174548721,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 1.2637,
+      "step": 681
+    },
+    {
+      "epoch": 2.76253164556962,
+      "grad_norm": 0.5222741866281077,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 1.2016,
+      "step": 682
+    },
+    {
+      "epoch": 2.7665822784810126,
+      "grad_norm": 0.5475335194497338,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 1.2449,
+      "step": 683
+    },
+    {
+      "epoch": 2.770632911392405,
+      "grad_norm": 0.5652944030829185,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 1.2514,
+      "step": 684
+    },
+    {
+      "epoch": 2.7746835443037976,
+      "grad_norm": 0.5114296240904409,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 1.2679,
+      "step": 685
+    },
+    {
+      "epoch": 2.77873417721519,
+      "grad_norm": 0.5531819504183846,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 1.2745,
+      "step": 686
+    },
+    {
+      "epoch": 2.782784810126582,
+      "grad_norm": 0.506270977463915,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 1.2469,
+      "step": 687
+    },
+    {
+      "epoch": 2.7868354430379747,
+      "grad_norm": 0.49973533605433174,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 1.3056,
+      "step": 688
+    },
+    {
+      "epoch": 2.790886075949367,
+      "grad_norm": 0.5028263413642563,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 1.255,
+      "step": 689
+    },
+    {
+      "epoch": 2.7949367088607593,
+      "grad_norm": 0.5942200421275363,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 1.2877,
+      "step": 690
+    },
+    {
+      "epoch": 2.798987341772152,
+      "grad_norm": 0.5261141425895168,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 1.1928,
+      "step": 691
+    },
+    {
+      "epoch": 2.8030379746835443,
+      "grad_norm": 0.5650080911995586,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 1.252,
+      "step": 692
+    },
+    {
+      "epoch": 2.807088607594937,
+      "grad_norm": 0.5077484602921568,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 1.2593,
+      "step": 693
+    },
+    {
+      "epoch": 2.8111392405063294,
+      "grad_norm": 0.5447290943557781,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 1.2698,
+      "step": 694
+    },
+    {
+      "epoch": 2.8151898734177214,
+      "grad_norm": 0.6230840918927645,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 1.2611,
+      "step": 695
+    },
+    {
+      "epoch": 2.819240506329114,
+      "grad_norm": 0.46223104578465557,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 1.2805,
+      "step": 696
+    },
+    {
+      "epoch": 2.8232911392405065,
+      "grad_norm": 0.5164139975390251,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 1.2577,
+      "step": 697
+    },
+    {
+      "epoch": 2.8273417721518985,
+      "grad_norm": 0.5906572400143807,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 1.2658,
+      "step": 698
+    },
+    {
+      "epoch": 2.831392405063291,
+      "grad_norm": 0.4769897894958626,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 1.2443,
+      "step": 699
+    },
+    {
+      "epoch": 2.8354430379746836,
+      "grad_norm": 0.5263373299052656,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 1.2331,
+      "step": 700
+    },
+    {
+      "epoch": 2.839493670886076,
+      "grad_norm": 0.5241071303522153,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 1.2385,
+      "step": 701
+    },
+    {
+      "epoch": 2.8435443037974686,
+      "grad_norm": 0.4946296006358129,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 1.2936,
+      "step": 702
+    },
+    {
+      "epoch": 2.8475949367088607,
+      "grad_norm": 0.5041306853716313,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 1.2233,
+      "step": 703
+    },
+    {
+      "epoch": 2.851645569620253,
+      "grad_norm": 0.4807842480347295,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 1.2248,
+      "step": 704
+    },
+    {
+      "epoch": 2.8556962025316457,
+      "grad_norm": 0.5474856019618206,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 1.2546,
+      "step": 705
+    },
+    {
+      "epoch": 2.8597468354430378,
+      "grad_norm": 0.6615831320178128,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 1.2529,
+      "step": 706
+    },
+    {
+      "epoch": 2.8637974683544303,
+      "grad_norm": 0.5206491460209098,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 1.2299,
+      "step": 707
+    },
+    {
+      "epoch": 2.867848101265823,
+      "grad_norm": 0.5201945474413302,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 1.2566,
+      "step": 708
+    },
+    {
+      "epoch": 2.8718987341772153,
+      "grad_norm": 0.5082825414723147,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 1.2674,
+      "step": 709
+    },
+    {
+      "epoch": 2.875949367088608,
+      "grad_norm": 0.5071419783168416,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 1.2594,
+      "step": 710
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5215442274766623,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 1.2528,
+      "step": 711
+    },
+    {
+      "epoch": 2.8840506329113924,
+      "grad_norm": 0.5969920479058002,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 1.2865,
+      "step": 712
+    },
+    {
+      "epoch": 2.888101265822785,
+      "grad_norm": 0.5109942277545163,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 1.2726,
+      "step": 713
+    },
+    {
+      "epoch": 2.892151898734177,
+      "grad_norm": 0.4786209166268272,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 1.2276,
+      "step": 714
+    },
+    {
+      "epoch": 2.8962025316455695,
+      "grad_norm": 0.48143019802453996,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 1.2659,
+      "step": 715
+    },
+    {
+      "epoch": 2.900253164556962,
+      "grad_norm": 0.5456512184508195,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 1.2344,
+      "step": 716
+    },
+    {
+      "epoch": 2.9043037974683545,
+      "grad_norm": 0.5148220967209761,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 1.2358,
+      "step": 717
+    },
+    {
+      "epoch": 2.908354430379747,
+      "grad_norm": 0.4834958496843285,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 1.2283,
+      "step": 718
+    },
+    {
+      "epoch": 2.912405063291139,
+      "grad_norm": 0.5111534732201822,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 1.2741,
+      "step": 719
+    },
+    {
+      "epoch": 2.9164556962025316,
+      "grad_norm": 0.5536534152574998,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 1.262,
+      "step": 720
+    },
+    {
+      "epoch": 2.920506329113924,
+      "grad_norm": 0.4949200143733224,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 1.272,
+      "step": 721
+    },
+    {
+      "epoch": 2.9245569620253162,
+      "grad_norm": 0.5203143924001232,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 1.2549,
+      "step": 722
+    },
+    {
+      "epoch": 2.9286075949367087,
+      "grad_norm": 0.525365070877558,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 1.2168,
+      "step": 723
+    },
+    {
+      "epoch": 2.9326582278481013,
+      "grad_norm": 0.465256951001496,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 1.2567,
+      "step": 724
+    },
+    {
+      "epoch": 2.9367088607594938,
+      "grad_norm": 0.5669620583704078,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 1.2512,
+      "step": 725
+    },
+    {
+      "epoch": 2.9407594936708863,
+      "grad_norm": 0.5589525615411228,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 1.2871,
+      "step": 726
+    },
+    {
+      "epoch": 2.9448101265822784,
+      "grad_norm": 0.5290559105851819,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 1.2272,
+      "step": 727
+    },
+    {
+      "epoch": 2.948860759493671,
+      "grad_norm": 0.5240546410139328,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 1.2386,
+      "step": 728
+    },
+    {
+      "epoch": 2.9529113924050634,
+      "grad_norm": 0.48031814120403954,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 1.2707,
+      "step": 729
+    },
+    {
+      "epoch": 2.9569620253164555,
+      "grad_norm": 0.6054500141233649,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 1.3028,
+      "step": 730
+    },
+    {
+      "epoch": 2.961012658227848,
+      "grad_norm": 0.47643745281770444,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 1.2485,
+      "step": 731
+    },
+    {
+      "epoch": 2.9650632911392405,
+      "grad_norm": 0.5047853363518067,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 1.2615,
+      "step": 732
+    },
+    {
+      "epoch": 2.969113924050633,
+      "grad_norm": 0.57281943692091,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 1.279,
+      "step": 733
+    },
+    {
+      "epoch": 2.9731645569620255,
+      "grad_norm": 0.5258262206628763,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 1.2101,
+      "step": 734
+    },
+    {
+      "epoch": 2.9772151898734176,
+      "grad_norm": 0.4965944745802932,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 1.2703,
+      "step": 735
+    },
+    {
+      "epoch": 2.98126582278481,
+      "grad_norm": 0.5322616387001319,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 1.284,
+      "step": 736
+    },
+    {
+      "epoch": 2.9853164556962026,
+      "grad_norm": 0.6076348249985659,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 1.2716,
+      "step": 737
+    },
+    {
+      "epoch": 2.9893670886075947,
+      "grad_norm": 0.5073565146974499,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 1.2725,
+      "step": 738
+    },
+    {
+      "epoch": 2.993417721518987,
+      "grad_norm": 0.547869166917642,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 1.2757,
+      "step": 739
+    },
+    {
+      "epoch": 2.9974683544303797,
+      "grad_norm": 0.5288085195569453,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 1.2665,
+      "step": 740
+    },
+    {
+      "epoch": 3.0015189873417722,
+      "grad_norm": 0.5870850788216498,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 1.2044,
+      "step": 741
+    },
+    {
+      "epoch": 3.0055696202531648,
+      "grad_norm": 0.7380011147157715,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 1.0832,
+      "step": 742
+    },
+    {
+      "epoch": 3.009620253164557,
+      "grad_norm": 1.4834040795930528,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 1.1005,
+      "step": 743
+    },
+    {
+      "epoch": 3.0136708860759494,
+      "grad_norm": 0.747599473105271,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 1.0736,
+      "step": 744
+    },
+    {
+      "epoch": 3.017721518987342,
+      "grad_norm": 0.9857361791554035,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 1.0958,
+      "step": 745
+    },
+    {
+      "epoch": 3.0217721518987344,
+      "grad_norm": 0.7984591963063422,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 1.0701,
+      "step": 746
+    },
+    {
+      "epoch": 3.0258227848101265,
+      "grad_norm": 0.7555862061802602,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 1.1348,
+      "step": 747
+    },
+    {
+      "epoch": 3.029873417721519,
+      "grad_norm": 0.8293935502972536,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 1.1096,
+      "step": 748
+    },
+    {
+      "epoch": 3.0339240506329115,
+      "grad_norm": 0.7916188523472516,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 1.1356,
+      "step": 749
+    },
+    {
+      "epoch": 3.037974683544304,
+      "grad_norm": 0.8653297148174526,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 1.1277,
+      "step": 750
+    },
+    {
+      "epoch": 3.042025316455696,
+      "grad_norm": 0.6899218805436081,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 1.0896,
+      "step": 751
+    },
+    {
+      "epoch": 3.0460759493670886,
+      "grad_norm": 0.8246702061228901,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 1.0732,
+      "step": 752
+    },
+    {
+      "epoch": 3.050126582278481,
+      "grad_norm": 0.8097510079202286,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 1.0577,
+      "step": 753
+    },
+    {
+      "epoch": 3.0541772151898736,
+      "grad_norm": 0.6032467772207977,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 1.1068,
+      "step": 754
+    },
+    {
+      "epoch": 3.0582278481012657,
+      "grad_norm": 0.7361439752353873,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 1.1179,
+      "step": 755
+    },
+    {
+      "epoch": 3.062278481012658,
+      "grad_norm": 0.7319729425277411,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 1.0976,
+      "step": 756
+    },
+    {
+      "epoch": 3.0663291139240507,
+      "grad_norm": 0.6991963791306958,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 1.0962,
+      "step": 757
+    },
+    {
+      "epoch": 3.0703797468354432,
+      "grad_norm": 0.7853231243178512,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 1.1053,
+      "step": 758
+    },
+    {
+      "epoch": 3.0744303797468353,
+      "grad_norm": 0.6371054966469066,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 1.0737,
+      "step": 759
+    },
+    {
+      "epoch": 3.078481012658228,
+      "grad_norm": 0.6357740280243205,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 1.0681,
+      "step": 760
+    },
+    {
+      "epoch": 3.0825316455696203,
+      "grad_norm": 0.6423042882554525,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 1.0895,
+      "step": 761
+    },
+    {
+      "epoch": 3.086582278481013,
+      "grad_norm": 0.6360634219157251,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 1.0794,
+      "step": 762
+    },
+    {
+      "epoch": 3.090632911392405,
+      "grad_norm": 0.7238065287677878,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 1.1033,
+      "step": 763
+    },
+    {
+      "epoch": 3.0946835443037974,
+      "grad_norm": 0.5540827569181441,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 1.0773,
+      "step": 764
+    },
+    {
+      "epoch": 3.09873417721519,
+      "grad_norm": 0.5844894677547803,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 1.0726,
+      "step": 765
+    },
+    {
+      "epoch": 3.1027848101265825,
+      "grad_norm": 0.6474596883173726,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 1.1038,
+      "step": 766
+    },
+    {
+      "epoch": 3.1068354430379745,
+      "grad_norm": 0.5732152005494472,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 1.091,
+      "step": 767
+    },
+    {
+      "epoch": 3.110886075949367,
+      "grad_norm": 0.5613534027593551,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 1.0801,
+      "step": 768
+    },
+    {
+      "epoch": 3.1149367088607596,
+      "grad_norm": 0.5624405485290622,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 1.1139,
+      "step": 769
+    },
+    {
+      "epoch": 3.118987341772152,
+      "grad_norm": 0.626532338666901,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 1.0932,
+      "step": 770
+    },
+    {
+      "epoch": 3.123037974683544,
+      "grad_norm": 0.49648492715457826,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 1.0872,
+      "step": 771
+    },
+    {
+      "epoch": 3.1270886075949367,
+      "grad_norm": 0.5557187717351948,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 1.1256,
+      "step": 772
+    },
+    {
+      "epoch": 3.131139240506329,
+      "grad_norm": 0.5066529376009189,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 1.1011,
+      "step": 773
+    },
+    {
+      "epoch": 3.1351898734177217,
+      "grad_norm": 0.5010632904059582,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 1.0544,
+      "step": 774
+    },
+    {
+      "epoch": 3.1392405063291138,
+      "grad_norm": 0.5487367440301106,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 1.0901,
+      "step": 775
+    },
+    {
+      "epoch": 3.1432911392405063,
+      "grad_norm": 0.5210000813737828,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 1.0941,
+      "step": 776
+    },
+    {
+      "epoch": 3.147341772151899,
+      "grad_norm": 0.5162999982739483,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 1.0797,
+      "step": 777
+    },
+    {
+      "epoch": 3.1513924050632913,
+      "grad_norm": 0.45082689489606936,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 1.0868,
+      "step": 778
+    },
+    {
+      "epoch": 3.1554430379746834,
+      "grad_norm": 0.4792944655242515,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 1.0829,
+      "step": 779
+    },
+    {
+      "epoch": 3.159493670886076,
+      "grad_norm": 0.5450396350701551,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 1.1125,
+      "step": 780
+    },
+    {
+      "epoch": 3.1635443037974684,
+      "grad_norm": 0.46094587530199155,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 1.07,
+      "step": 781
+    },
+    {
+      "epoch": 3.167594936708861,
+      "grad_norm": 0.5166723325369489,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 1.0978,
+      "step": 782
+    },
+    {
+      "epoch": 3.171645569620253,
+      "grad_norm": 0.5341164429540807,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 1.1058,
+      "step": 783
+    },
+    {
+      "epoch": 3.1756962025316455,
+      "grad_norm": 0.4980623075560834,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 1.1275,
+      "step": 784
+    },
+    {
+      "epoch": 3.179746835443038,
+      "grad_norm": 0.5152366222589686,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 1.0961,
+      "step": 785
+    },
+    {
+      "epoch": 3.1837974683544306,
+      "grad_norm": 0.5229980245890773,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 1.1194,
+      "step": 786
+    },
+    {
+      "epoch": 3.1878481012658226,
+      "grad_norm": 0.5151803586499513,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 1.0698,
+      "step": 787
+    },
+    {
+      "epoch": 3.191898734177215,
+      "grad_norm": 0.4752443328205465,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 1.1083,
+      "step": 788
+    },
+    {
+      "epoch": 3.1959493670886077,
+      "grad_norm": 0.5393391035799078,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 1.1006,
+      "step": 789
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.5423619959630889,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 1.0923,
+      "step": 790
+    },
+    {
+      "epoch": 3.2040506329113922,
+      "grad_norm": 0.49899156756056057,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 1.116,
+      "step": 791
+    },
+    {
+      "epoch": 3.2081012658227848,
+      "grad_norm": 0.5077201364562629,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 1.0771,
+      "step": 792
+    },
+    {
+      "epoch": 3.2121518987341773,
+      "grad_norm": 0.5304462273703419,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 1.0796,
+      "step": 793
+    },
+    {
+      "epoch": 3.21620253164557,
+      "grad_norm": 0.519606028884392,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 1.1,
+      "step": 794
+    },
+    {
+      "epoch": 3.220253164556962,
+      "grad_norm": 0.5172811033147627,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 1.1199,
+      "step": 795
+    },
+    {
+      "epoch": 3.2243037974683544,
+      "grad_norm": 0.5297029210712721,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 1.0861,
+      "step": 796
+    },
+    {
+      "epoch": 3.228354430379747,
+      "grad_norm": 0.5388718576981099,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 1.0899,
+      "step": 797
+    },
+    {
+      "epoch": 3.2324050632911394,
+      "grad_norm": 0.49027908162355177,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 1.1007,
+      "step": 798
+    },
+    {
+      "epoch": 3.2364556962025315,
+      "grad_norm": 0.5328271073091836,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 1.0985,
+      "step": 799
+    },
+    {
+      "epoch": 3.240506329113924,
+      "grad_norm": 0.5843962630367245,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 1.0835,
+      "step": 800
+    },
+    {
+      "epoch": 3.2445569620253165,
+      "grad_norm": 0.5291191200841259,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 1.0767,
+      "step": 801
+    },
+    {
+      "epoch": 3.248607594936709,
+      "grad_norm": 0.5735889512782648,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 1.1248,
+      "step": 802
+    },
+    {
+      "epoch": 3.252658227848101,
+      "grad_norm": 0.5821363106651172,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 1.088,
+      "step": 803
+    },
+    {
+      "epoch": 3.2567088607594936,
+      "grad_norm": 0.5124719966855544,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 1.1254,
+      "step": 804
+    },
+    {
+      "epoch": 3.260759493670886,
+      "grad_norm": 0.5416681076962963,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 1.0992,
+      "step": 805
+    },
+    {
+      "epoch": 3.2648101265822786,
+      "grad_norm": 0.5220838645294046,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 1.1145,
+      "step": 806
+    },
+    {
+      "epoch": 3.2688607594936707,
+      "grad_norm": 0.5480456205057328,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 1.0742,
+      "step": 807
+    },
+    {
+      "epoch": 3.2729113924050632,
+      "grad_norm": 0.48002175488432086,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 1.0529,
+      "step": 808
+    },
+    {
+      "epoch": 3.2769620253164558,
+      "grad_norm": 0.5636182394807844,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 1.0932,
+      "step": 809
+    },
+    {
+      "epoch": 3.2810126582278483,
+      "grad_norm": 0.5331228013733605,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 1.0755,
+      "step": 810
+    },
+    {
+      "epoch": 3.2850632911392403,
+      "grad_norm": 0.5112820138878884,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 1.0559,
+      "step": 811
+    },
+    {
+      "epoch": 3.289113924050633,
+      "grad_norm": 0.5371959099913656,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 1.0949,
+      "step": 812
+    },
+    {
+      "epoch": 3.2931645569620254,
+      "grad_norm": 0.5375246763555377,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 1.0972,
+      "step": 813
+    },
+    {
+      "epoch": 3.297215189873418,
+      "grad_norm": 0.5769627557371743,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 1.1158,
+      "step": 814
+    },
+    {
+      "epoch": 3.30126582278481,
+      "grad_norm": 0.4851980037638188,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 1.0737,
+      "step": 815
+    },
+    {
+      "epoch": 3.3053164556962025,
+      "grad_norm": 0.500393162993338,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 1.0771,
+      "step": 816
+    },
+    {
+      "epoch": 3.309367088607595,
+      "grad_norm": 0.5502972634055959,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 1.113,
+      "step": 817
+    },
+    {
+      "epoch": 3.3134177215189875,
+      "grad_norm": 0.5644794589214842,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 1.0598,
+      "step": 818
+    },
+    {
+      "epoch": 3.3174683544303796,
+      "grad_norm": 0.5597911893687817,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 1.099,
+      "step": 819
+    },
+    {
+      "epoch": 3.321518987341772,
+      "grad_norm": 0.5335395142781278,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 1.0766,
+      "step": 820
+    },
+    {
+      "epoch": 3.3255696202531646,
+      "grad_norm": 0.5704429938996292,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 1.0878,
+      "step": 821
+    },
+    {
+      "epoch": 3.329620253164557,
+      "grad_norm": 0.5389740906467743,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 1.088,
+      "step": 822
+    },
+    {
+      "epoch": 3.333670886075949,
+      "grad_norm": 0.5217828173594223,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 1.1054,
+      "step": 823
+    },
+    {
+      "epoch": 3.3377215189873417,
+      "grad_norm": 0.5993759007180443,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 1.0912,
+      "step": 824
+    },
+    {
+      "epoch": 3.3417721518987342,
+      "grad_norm": 0.5628178014448703,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 1.1216,
+      "step": 825
+    },
+    {
+      "epoch": 3.3458227848101267,
+      "grad_norm": 0.5159350076200965,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 1.1038,
+      "step": 826
+    },
+    {
+      "epoch": 3.349873417721519,
+      "grad_norm": 0.5490627514570068,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 1.0938,
+      "step": 827
+    },
+    {
+      "epoch": 3.3539240506329113,
+      "grad_norm": 0.5382693720421776,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 1.0921,
+      "step": 828
+    },
+    {
+      "epoch": 3.357974683544304,
+      "grad_norm": 0.5332867016810717,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 1.0949,
+      "step": 829
+    },
+    {
+      "epoch": 3.3620253164556964,
+      "grad_norm": 0.5664925631992158,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 1.0897,
+      "step": 830
+    },
+    {
+      "epoch": 3.3660759493670884,
+      "grad_norm": 0.47338464519671075,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 1.1555,
+      "step": 831
+    },
+    {
+      "epoch": 3.370126582278481,
+      "grad_norm": 0.5576384935776653,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 1.0908,
+      "step": 832
+    },
+    {
+      "epoch": 3.3741772151898735,
+      "grad_norm": 0.5180548601345136,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 1.1134,
+      "step": 833
+    },
+    {
+      "epoch": 3.378227848101266,
+      "grad_norm": 0.48742265041991956,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 1.0926,
+      "step": 834
+    },
+    {
+      "epoch": 3.382278481012658,
+      "grad_norm": 0.5067900038157551,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 1.0907,
+      "step": 835
+    },
+    {
+      "epoch": 3.3863291139240506,
+      "grad_norm": 0.5130186836588717,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 1.0847,
+      "step": 836
+    },
+    {
+      "epoch": 3.390379746835443,
+      "grad_norm": 0.5929944295481807,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 1.0869,
+      "step": 837
+    },
+    {
+      "epoch": 3.3944303797468356,
+      "grad_norm": 0.5436303422026426,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 1.1031,
+      "step": 838
+    },
+    {
+      "epoch": 3.3984810126582277,
+      "grad_norm": 0.49124853195247226,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 1.0993,
+      "step": 839
+    },
+    {
+      "epoch": 3.40253164556962,
+      "grad_norm": 0.5553944856099985,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 1.1381,
+      "step": 840
+    },
+    {
+      "epoch": 3.4065822784810127,
+      "grad_norm": 0.5518249252856855,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 1.0803,
+      "step": 841
+    },
+    {
+      "epoch": 3.410632911392405,
+      "grad_norm": 0.514444419804712,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 1.1298,
+      "step": 842
+    },
+    {
+      "epoch": 3.4146835443037973,
+      "grad_norm": 0.4876584178811673,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 1.0905,
+      "step": 843
+    },
+    {
+      "epoch": 3.41873417721519,
+      "grad_norm": 0.5137986064904352,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 1.0607,
+      "step": 844
+    },
+    {
+      "epoch": 3.4227848101265823,
+      "grad_norm": 0.5221244200319632,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 1.1051,
+      "step": 845
+    },
+    {
+      "epoch": 3.426835443037975,
+      "grad_norm": 0.5126526049607916,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 1.0687,
+      "step": 846
+    },
+    {
+      "epoch": 3.430886075949367,
+      "grad_norm": 0.4824383284591642,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 1.1126,
+      "step": 847
+    },
+    {
+      "epoch": 3.4349367088607594,
+      "grad_norm": 0.4848388825342168,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 1.1148,
+      "step": 848
+    },
+    {
+      "epoch": 3.438987341772152,
+      "grad_norm": 0.4902995196518547,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 1.054,
+      "step": 849
+    },
+    {
+      "epoch": 3.4430379746835444,
+      "grad_norm": 0.45996706619062205,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 1.0907,
+      "step": 850
+    },
+    {
+      "epoch": 3.4470886075949365,
+      "grad_norm": 0.49211202028379075,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 1.1392,
+      "step": 851
+    },
+    {
+      "epoch": 3.451139240506329,
+      "grad_norm": 0.47768484358816643,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 1.0813,
+      "step": 852
+    },
+    {
+      "epoch": 3.4551898734177215,
+      "grad_norm": 0.4780202547487013,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 1.1027,
+      "step": 853
+    },
+    {
+      "epoch": 3.459240506329114,
+      "grad_norm": 0.4952145939296536,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 1.0566,
+      "step": 854
+    },
+    {
+      "epoch": 3.463291139240506,
+      "grad_norm": 0.47302645108414026,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 1.1074,
+      "step": 855
+    },
+    {
+      "epoch": 3.4673417721518986,
+      "grad_norm": 0.487502060471696,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 1.1397,
+      "step": 856
+    },
+    {
+      "epoch": 3.471392405063291,
+      "grad_norm": 0.4905002049511962,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 1.0835,
+      "step": 857
+    },
+    {
+      "epoch": 3.4754430379746837,
+      "grad_norm": 0.5296559070658476,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 1.0981,
+      "step": 858
+    },
+    {
+      "epoch": 3.479493670886076,
+      "grad_norm": 0.4834415919003545,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 1.0999,
+      "step": 859
+    },
+    {
+      "epoch": 3.4835443037974683,
+      "grad_norm": 0.4851778719155166,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 1.1059,
+      "step": 860
+    },
+    {
+      "epoch": 3.487594936708861,
+      "grad_norm": 0.46496441758706775,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 1.0845,
+      "step": 861
+    },
+    {
+      "epoch": 3.4916455696202533,
+      "grad_norm": 0.495847182691014,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 1.1032,
+      "step": 862
+    },
+    {
+      "epoch": 3.4956962025316454,
+      "grad_norm": 0.48240600984014903,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 1.0989,
+      "step": 863
+    },
+    {
+      "epoch": 3.499746835443038,
+      "grad_norm": 0.5060572487098134,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 1.0919,
+      "step": 864
+    },
+    {
+      "epoch": 3.5037974683544304,
+      "grad_norm": 0.455926443373757,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 1.0336,
+      "step": 865
+    },
+    {
+      "epoch": 3.507848101265823,
+      "grad_norm": 0.49295332838592465,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 1.0988,
+      "step": 866
+    },
+    {
+      "epoch": 3.5118987341772154,
+      "grad_norm": 0.49089110649723205,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 1.0793,
+      "step": 867
+    },
+    {
+      "epoch": 3.5159493670886075,
+      "grad_norm": 0.46725474404593986,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 1.0914,
+      "step": 868
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.47356108891334187,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 1.0961,
+      "step": 869
+    },
+    {
+      "epoch": 3.5240506329113925,
+      "grad_norm": 0.4610916804814865,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 1.1038,
+      "step": 870
+    },
+    {
+      "epoch": 3.5281012658227846,
+      "grad_norm": 0.46957883756748064,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 1.0982,
+      "step": 871
+    },
+    {
+      "epoch": 3.532151898734177,
+      "grad_norm": 0.4656345738936295,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 1.0919,
+      "step": 872
+    },
+    {
+      "epoch": 3.5362025316455696,
+      "grad_norm": 0.4804455182146021,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 1.1083,
+      "step": 873
+    },
+    {
+      "epoch": 3.540253164556962,
+      "grad_norm": 0.4696860636605175,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 1.0802,
+      "step": 874
+    },
+    {
+      "epoch": 3.5443037974683547,
+      "grad_norm": 0.48485750376003667,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 1.1129,
+      "step": 875
+    },
+    {
+      "epoch": 3.5483544303797467,
+      "grad_norm": 0.435531955991696,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 1.0855,
+      "step": 876
+    },
+    {
+      "epoch": 3.5524050632911393,
+      "grad_norm": 0.47375110166388884,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 1.0747,
+      "step": 877
+    },
+    {
+      "epoch": 3.5564556962025318,
+      "grad_norm": 0.49616904605403245,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 1.1027,
+      "step": 878
+    },
+    {
+      "epoch": 3.560506329113924,
+      "grad_norm": 0.5034421235859694,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 1.1235,
+      "step": 879
+    },
+    {
+      "epoch": 3.5645569620253164,
+      "grad_norm": 0.46626583508191544,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 1.0808,
+      "step": 880
+    },
+    {
+      "epoch": 3.568607594936709,
+      "grad_norm": 0.4732981683547378,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 1.0987,
+      "step": 881
+    },
+    {
+      "epoch": 3.5726582278481014,
+      "grad_norm": 0.4980526172187316,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 1.0732,
+      "step": 882
+    },
+    {
+      "epoch": 3.576708860759494,
+      "grad_norm": 0.4763235129825611,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 1.0564,
+      "step": 883
+    },
+    {
+      "epoch": 3.580759493670886,
+      "grad_norm": 0.46084736953297584,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 1.0849,
+      "step": 884
+    },
+    {
+      "epoch": 3.5848101265822785,
+      "grad_norm": 0.474399603580335,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 1.0766,
+      "step": 885
+    },
+    {
+      "epoch": 3.588860759493671,
+      "grad_norm": 0.5171871884273997,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 1.0978,
+      "step": 886
+    },
+    {
+      "epoch": 3.592911392405063,
+      "grad_norm": 0.5146189326617108,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 1.0856,
+      "step": 887
+    },
+    {
+      "epoch": 3.5969620253164556,
+      "grad_norm": 0.4509032579153856,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 1.0916,
+      "step": 888
+    },
+    {
+      "epoch": 3.601012658227848,
+      "grad_norm": 0.500100177560375,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 1.1045,
+      "step": 889
+    },
+    {
+      "epoch": 3.6050632911392406,
+      "grad_norm": 0.5033985244326453,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 1.1344,
+      "step": 890
+    },
+    {
+      "epoch": 3.609113924050633,
+      "grad_norm": 0.5022822023975303,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 1.0943,
+      "step": 891
+    },
+    {
+      "epoch": 3.613164556962025,
+      "grad_norm": 0.4780077227986291,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 1.0888,
+      "step": 892
+    },
+    {
+      "epoch": 3.6172151898734177,
+      "grad_norm": 0.47200947567907486,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 1.0793,
+      "step": 893
+    },
+    {
+      "epoch": 3.6212658227848102,
+      "grad_norm": 0.517955862300238,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 1.1378,
+      "step": 894
+    },
+    {
+      "epoch": 3.6253164556962023,
+      "grad_norm": 0.4847698041521539,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 1.1122,
+      "step": 895
+    },
+    {
+      "epoch": 3.629367088607595,
+      "grad_norm": 0.5227983386047254,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 1.0896,
+      "step": 896
+    },
+    {
+      "epoch": 3.6334177215189873,
+      "grad_norm": 0.45258703246939425,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 1.1056,
+      "step": 897
+    },
+    {
+      "epoch": 3.63746835443038,
+      "grad_norm": 0.4435234549028214,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 1.1309,
+      "step": 898
+    },
+    {
+      "epoch": 3.6415189873417724,
+      "grad_norm": 0.455220935662557,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 1.0934,
+      "step": 899
+    },
+    {
+      "epoch": 3.6455696202531644,
+      "grad_norm": 0.5401640897757641,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 1.0613,
+      "step": 900
+    },
+    {
+      "epoch": 3.649620253164557,
+      "grad_norm": 0.4906980042237514,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 1.0456,
+      "step": 901
+    },
+    {
+      "epoch": 3.6536708860759495,
+      "grad_norm": 0.43434213722812004,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 1.0692,
+      "step": 902
+    },
+    {
+      "epoch": 3.6577215189873415,
+      "grad_norm": 0.4999406302329833,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 1.0993,
+      "step": 903
+    },
+    {
+      "epoch": 3.661772151898734,
+      "grad_norm": 0.5226389952121681,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 1.0983,
+      "step": 904
+    },
+    {
+      "epoch": 3.6658227848101266,
+      "grad_norm": 0.5181828791569407,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 1.1228,
+      "step": 905
+    },
+    {
+      "epoch": 3.669873417721519,
+      "grad_norm": 0.49749807909556326,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 1.0865,
+      "step": 906
+    },
+    {
+      "epoch": 3.6739240506329116,
+      "grad_norm": 0.47320277516752524,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 1.1088,
+      "step": 907
+    },
+    {
+      "epoch": 3.6779746835443037,
+      "grad_norm": 0.5543429965399255,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 1.0793,
+      "step": 908
+    },
+    {
+      "epoch": 3.682025316455696,
+      "grad_norm": 0.48079421622139723,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 1.1168,
+      "step": 909
+    },
+    {
+      "epoch": 3.6860759493670887,
+      "grad_norm": 0.4597720345819124,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 1.0514,
+      "step": 910
+    },
+    {
+      "epoch": 3.690126582278481,
+      "grad_norm": 0.4910552516692228,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 1.1636,
+      "step": 911
+    },
+    {
+      "epoch": 3.6941772151898733,
+      "grad_norm": 0.49500011502515967,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 1.0972,
+      "step": 912
+    },
+    {
+      "epoch": 3.698227848101266,
+      "grad_norm": 0.49241743415652767,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 1.1003,
+      "step": 913
+    },
+    {
+      "epoch": 3.7022784810126583,
+      "grad_norm": 0.4942506975609191,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 1.0679,
+      "step": 914
+    },
+    {
+      "epoch": 3.706329113924051,
+      "grad_norm": 0.48905961420326843,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 1.1084,
+      "step": 915
+    },
+    {
+      "epoch": 3.710379746835443,
+      "grad_norm": 0.4440566045434787,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 1.1049,
+      "step": 916
+    },
+    {
+      "epoch": 3.7144303797468354,
+      "grad_norm": 0.4781243819679991,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 1.104,
+      "step": 917
+    },
+    {
+      "epoch": 3.718481012658228,
+      "grad_norm": 0.46846731634558364,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 1.1089,
+      "step": 918
+    },
+    {
+      "epoch": 3.72253164556962,
+      "grad_norm": 0.4801950643042868,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 1.1291,
+      "step": 919
+    },
+    {
+      "epoch": 3.7265822784810125,
+      "grad_norm": 0.4547184910846464,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 1.1002,
+      "step": 920
+    },
+    {
+      "epoch": 3.730632911392405,
+      "grad_norm": 0.5136934348233948,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 1.0902,
+      "step": 921
+    },
+    {
+      "epoch": 3.7346835443037976,
+      "grad_norm": 0.4394563066572403,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 1.0857,
+      "step": 922
+    },
+    {
+      "epoch": 3.73873417721519,
+      "grad_norm": 0.4850755626817249,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 1.1101,
+      "step": 923
+    },
+    {
+      "epoch": 3.742784810126582,
+      "grad_norm": 0.48425749159705656,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 1.1298,
+      "step": 924
+    },
+    {
+      "epoch": 3.7468354430379747,
+      "grad_norm": 0.49757990688038684,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 1.0739,
+      "step": 925
+    },
+    {
+      "epoch": 3.750886075949367,
+      "grad_norm": 0.4690301569116768,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 1.1331,
+      "step": 926
+    },
+    {
+      "epoch": 3.7549367088607593,
+      "grad_norm": 0.45454079876305425,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 1.0947,
+      "step": 927
+    },
+    {
+      "epoch": 3.7589873417721518,
+      "grad_norm": 0.4872450182132625,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 1.0842,
+      "step": 928
+    },
+    {
+      "epoch": 3.7630379746835443,
+      "grad_norm": 0.5009019605370861,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 1.0686,
+      "step": 929
+    },
+    {
+      "epoch": 3.767088607594937,
+      "grad_norm": 0.5320860333375942,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 1.0504,
+      "step": 930
+    },
+    {
+      "epoch": 3.7711392405063293,
+      "grad_norm": 0.4724350292599559,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 1.0878,
+      "step": 931
+    },
+    {
+      "epoch": 3.7751898734177214,
+      "grad_norm": 0.4641152917993967,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 1.0782,
+      "step": 932
+    },
+    {
+      "epoch": 3.779240506329114,
+      "grad_norm": 0.4966549660511162,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 1.1276,
+      "step": 933
+    },
+    {
+      "epoch": 3.7832911392405064,
+      "grad_norm": 0.5212387413443806,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 1.0761,
+      "step": 934
+    },
+    {
+      "epoch": 3.7873417721518985,
+      "grad_norm": 0.47395823769820733,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 1.1317,
+      "step": 935
+    },
+    {
+      "epoch": 3.791392405063291,
+      "grad_norm": 0.4623480180703169,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 1.0704,
+      "step": 936
+    },
+    {
+      "epoch": 3.7954430379746835,
+      "grad_norm": 0.530024170027408,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 1.0571,
+      "step": 937
+    },
+    {
+      "epoch": 3.799493670886076,
+      "grad_norm": 0.4488262982718847,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 1.0973,
+      "step": 938
+    },
+    {
+      "epoch": 3.8035443037974686,
+      "grad_norm": 0.4861187926775796,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 1.0878,
+      "step": 939
+    },
+    {
+      "epoch": 3.8075949367088606,
+      "grad_norm": 0.4598374115167874,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 1.081,
+      "step": 940
+    },
+    {
+      "epoch": 3.811645569620253,
+      "grad_norm": 0.46043784677627875,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 1.1,
+      "step": 941
+    },
+    {
+      "epoch": 3.8156962025316457,
+      "grad_norm": 0.4675102008147449,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 1.0771,
+      "step": 942
+    },
+    {
+      "epoch": 3.8197468354430377,
+      "grad_norm": 0.5129952256486336,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 1.1042,
+      "step": 943
+    },
+    {
+      "epoch": 3.8237974683544302,
+      "grad_norm": 0.501060643470693,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 1.083,
+      "step": 944
+    },
+    {
+      "epoch": 3.8278481012658228,
+      "grad_norm": 0.4971899100671292,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 1.0842,
+      "step": 945
+    },
+    {
+      "epoch": 3.8318987341772153,
+      "grad_norm": 0.4694466887289864,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 1.1073,
+      "step": 946
+    },
+    {
+      "epoch": 3.835949367088608,
+      "grad_norm": 0.47220741600972127,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 1.095,
+      "step": 947
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.49833113038754423,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 1.1173,
+      "step": 948
+    },
+    {
+      "epoch": 3.8440506329113924,
+      "grad_norm": 0.48178746802227995,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 1.0588,
+      "step": 949
+    },
+    {
+      "epoch": 3.848101265822785,
+      "grad_norm": 0.4453406005444569,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 1.091,
+      "step": 950
+    },
+    {
+      "epoch": 3.852151898734177,
+      "grad_norm": 0.4843794014241086,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 1.097,
+      "step": 951
+    },
+    {
+      "epoch": 3.8562025316455695,
+      "grad_norm": 0.48418517773428466,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 1.1177,
+      "step": 952
+    },
+    {
+      "epoch": 3.860253164556962,
+      "grad_norm": 0.4459126373068495,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 1.0693,
+      "step": 953
+    },
+    {
+      "epoch": 3.8643037974683545,
+      "grad_norm": 0.473684739718878,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 1.1229,
+      "step": 954
+    },
+    {
+      "epoch": 3.868354430379747,
+      "grad_norm": 0.4823285163177484,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 1.0803,
+      "step": 955
+    },
+    {
+      "epoch": 3.872405063291139,
+      "grad_norm": 0.4539591679116137,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 1.1085,
+      "step": 956
+    },
+    {
+      "epoch": 3.8764556962025316,
+      "grad_norm": 0.4378576667434754,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 1.0804,
+      "step": 957
+    },
+    {
+      "epoch": 3.880506329113924,
+      "grad_norm": 0.4513514616093181,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 1.0939,
+      "step": 958
+    },
+    {
+      "epoch": 3.884556962025316,
+      "grad_norm": 0.45848668425840855,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 1.089,
+      "step": 959
+    },
+    {
+      "epoch": 3.8886075949367087,
+      "grad_norm": 0.5072109489527027,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 1.0629,
+      "step": 960
+    },
+    {
+      "epoch": 3.8926582278481012,
+      "grad_norm": 0.4672902387344472,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 1.12,
+      "step": 961
+    },
+    {
+      "epoch": 3.8967088607594937,
+      "grad_norm": 0.4844979657977251,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 1.0773,
+      "step": 962
+    },
+    {
+      "epoch": 3.9007594936708863,
+      "grad_norm": 0.4681244380637179,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 1.0652,
+      "step": 963
+    },
+    {
+      "epoch": 3.9048101265822783,
+      "grad_norm": 0.4460341483836583,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 1.0875,
+      "step": 964
+    },
+    {
+      "epoch": 3.908860759493671,
+      "grad_norm": 0.49037906234850653,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 1.0942,
+      "step": 965
+    },
+    {
+      "epoch": 3.9129113924050634,
+      "grad_norm": 0.46213210746145345,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 1.1093,
+      "step": 966
+    },
+    {
+      "epoch": 3.9169620253164554,
+      "grad_norm": 0.47114692250692775,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 1.0975,
+      "step": 967
+    },
+    {
+      "epoch": 3.921012658227848,
+      "grad_norm": 0.43654636786285256,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 1.058,
+      "step": 968
+    },
+    {
+      "epoch": 3.9250632911392405,
+      "grad_norm": 0.46912489142652736,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 1.0691,
+      "step": 969
+    },
+    {
+      "epoch": 3.929113924050633,
+      "grad_norm": 0.44176819138029394,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 1.0957,
+      "step": 970
+    },
+    {
+      "epoch": 3.9331645569620255,
+      "grad_norm": 0.4616708836514654,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 1.0621,
+      "step": 971
+    },
+    {
+      "epoch": 3.9372151898734176,
+      "grad_norm": 0.4658612226146692,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 1.1221,
+      "step": 972
+    },
+    {
+      "epoch": 3.94126582278481,
+      "grad_norm": 0.5069367082089369,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 1.1093,
+      "step": 973
+    },
+    {
+      "epoch": 3.9453164556962026,
+      "grad_norm": 0.4541327927425851,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 1.0819,
+      "step": 974
+    },
+    {
+      "epoch": 3.9493670886075947,
+      "grad_norm": 0.4639044359455014,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 1.0597,
+      "step": 975
+    },
+    {
+      "epoch": 3.953417721518987,
+      "grad_norm": 0.45246642454603236,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 1.0987,
+      "step": 976
+    },
+    {
+      "epoch": 3.9574683544303797,
+      "grad_norm": 0.4726573282842486,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 1.0939,
+      "step": 977
+    },
+    {
+      "epoch": 3.961518987341772,
+      "grad_norm": 0.46707616056118684,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 1.0828,
+      "step": 978
+    },
+    {
+      "epoch": 3.9655696202531647,
+      "grad_norm": 0.45018167790487235,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 1.0909,
+      "step": 979
+    },
+    {
+      "epoch": 3.969620253164557,
+      "grad_norm": 0.43688540298760814,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 1.1092,
+      "step": 980
+    },
+    {
+      "epoch": 3.9736708860759493,
+      "grad_norm": 0.44394351485441036,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 1.098,
+      "step": 981
+    },
+    {
+      "epoch": 3.977721518987342,
+      "grad_norm": 0.4468511555347179,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 1.077,
+      "step": 982
+    },
+    {
+      "epoch": 3.981772151898734,
+      "grad_norm": 0.4482571585233651,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 1.0997,
+      "step": 983
+    },
+    {
+      "epoch": 3.9858227848101264,
+      "grad_norm": 0.44221224501868706,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 1.0616,
+      "step": 984
+    },
+    {
+      "epoch": 3.989873417721519,
+      "grad_norm": 0.47149962153373565,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 1.1136,
+      "step": 985
+    },
+    {
+      "epoch": 3.9939240506329114,
+      "grad_norm": 0.4796085754126796,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 1.0787,
+      "step": 986
+    },
+    {
+      "epoch": 3.997974683544304,
+      "grad_norm": 0.4609732068262811,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 1.0967,
+      "step": 987
+    },
+    {
+      "epoch": 4.002025316455696,
+      "grad_norm": 0.648050374118586,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 1.0423,
+      "step": 988
+    },
+    {
+      "epoch": 4.006075949367089,
+      "grad_norm": 0.9765028301228841,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.9618,
+      "step": 989
+    },
+    {
+      "epoch": 4.010126582278481,
+      "grad_norm": 0.7507910219701858,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.9073,
+      "step": 990
+    },
+    {
+      "epoch": 4.014177215189873,
+      "grad_norm": 0.7585646517585802,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.9528,
+      "step": 991
+    },
+    {
+      "epoch": 4.018227848101266,
+      "grad_norm": 1.3045020066696909,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.9531,
+      "step": 992
+    },
+    {
+      "epoch": 4.022278481012658,
+      "grad_norm": 1.3148356786179427,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.936,
+      "step": 993
+    },
+    {
+      "epoch": 4.02632911392405,
+      "grad_norm": 0.9069690070530437,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.9157,
+      "step": 994
+    },
+    {
+      "epoch": 4.030379746835443,
+      "grad_norm": 0.8297530187674871,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.9413,
+      "step": 995
+    },
+    {
+      "epoch": 4.034430379746835,
+      "grad_norm": 0.8442406373210726,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.9332,
+      "step": 996
+    },
+    {
+      "epoch": 4.038481012658228,
+      "grad_norm": 0.8378646869121956,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.958,
+      "step": 997
+    },
+    {
+      "epoch": 4.04253164556962,
+      "grad_norm": 0.7834517548874151,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.9108,
+      "step": 998
+    },
+    {
+      "epoch": 4.046582278481012,
+      "grad_norm": 0.6646833270936053,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.939,
+      "step": 999
+    },
+    {
+      "epoch": 4.050632911392405,
+      "grad_norm": 0.684103805170203,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.9165,
+      "step": 1000
+    },
+    {
+      "epoch": 4.054683544303797,
+      "grad_norm": 0.7654976662438706,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.965,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0587341772151895,
+      "grad_norm": 0.7497404871511678,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.9566,
+      "step": 1002
+    },
+    {
+      "epoch": 4.062784810126582,
+      "grad_norm": 0.6832015502295545,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.9639,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0668354430379745,
+      "grad_norm": 0.6411737308221198,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.935,
+      "step": 1004
+    },
+    {
+      "epoch": 4.0708860759493675,
+      "grad_norm": 0.6628640000057907,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.9461,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0749367088607595,
+      "grad_norm": 0.6656772832100786,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.948,
+      "step": 1006
+    },
+    {
+      "epoch": 4.078987341772152,
+      "grad_norm": 0.6402990285732642,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.9387,
+      "step": 1007
+    },
+    {
+      "epoch": 4.083037974683545,
+      "grad_norm": 0.5934089595617562,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.9453,
+      "step": 1008
+    },
+    {
+      "epoch": 4.087088607594937,
+      "grad_norm": 0.5986024213323308,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.9343,
+      "step": 1009
+    },
+    {
+      "epoch": 4.091139240506329,
+      "grad_norm": 0.6634285268688027,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.9491,
+      "step": 1010
+    },
+    {
+      "epoch": 4.095189873417722,
+      "grad_norm": 0.6183734097510527,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.9016,
+      "step": 1011
+    },
+    {
+      "epoch": 4.099240506329114,
+      "grad_norm": 0.6117508676713921,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.9156,
+      "step": 1012
+    },
+    {
+      "epoch": 4.103291139240507,
+      "grad_norm": 0.5729575625625669,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.973,
+      "step": 1013
+    },
+    {
+      "epoch": 4.107341772151899,
+      "grad_norm": 0.584071706856258,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.9466,
+      "step": 1014
+    },
+    {
+      "epoch": 4.111392405063291,
+      "grad_norm": 0.6125995435519289,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.963,
+      "step": 1015
+    },
+    {
+      "epoch": 4.115443037974684,
+      "grad_norm": 0.5673046043989746,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.9451,
+      "step": 1016
+    },
+    {
+      "epoch": 4.119493670886076,
+      "grad_norm": 0.586555919977371,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.9296,
+      "step": 1017
+    },
+    {
+      "epoch": 4.123544303797468,
+      "grad_norm": 0.5317387576143573,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.9236,
+      "step": 1018
+    },
+    {
+      "epoch": 4.127594936708861,
+      "grad_norm": 0.5808699177466091,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.962,
+      "step": 1019
+    },
+    {
+      "epoch": 4.131645569620253,
+      "grad_norm": 0.542796379372556,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.9336,
+      "step": 1020
+    },
+    {
+      "epoch": 4.135696202531646,
+      "grad_norm": 0.5323320967925554,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.975,
+      "step": 1021
+    },
+    {
+      "epoch": 4.139746835443038,
+      "grad_norm": 0.5379547971600078,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.927,
+      "step": 1022
+    },
+    {
+      "epoch": 4.14379746835443,
+      "grad_norm": 0.5469377384906556,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.9359,
+      "step": 1023
+    },
+    {
+      "epoch": 4.147848101265823,
+      "grad_norm": 0.5625267343630442,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.9271,
+      "step": 1024
+    },
+    {
+      "epoch": 4.151898734177215,
+      "grad_norm": 0.5267011401718313,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.9482,
+      "step": 1025
+    },
+    {
+      "epoch": 4.155949367088607,
+      "grad_norm": 0.5244750017697006,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.9662,
+      "step": 1026
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.5157052480565005,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.9179,
+      "step": 1027
+    },
+    {
+      "epoch": 4.164050632911392,
+      "grad_norm": 0.5354047650754842,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.9284,
+      "step": 1028
+    },
+    {
+      "epoch": 4.168101265822785,
+      "grad_norm": 0.5074306147812987,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.9441,
+      "step": 1029
+    },
+    {
+      "epoch": 4.172151898734177,
+      "grad_norm": 0.5109421406641945,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.9644,
+      "step": 1030
+    },
+    {
+      "epoch": 4.176202531645569,
+      "grad_norm": 0.5378670890478536,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.9158,
+      "step": 1031
+    },
+    {
+      "epoch": 4.180253164556962,
+      "grad_norm": 0.5191640704456416,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.9209,
+      "step": 1032
+    },
+    {
+      "epoch": 4.184303797468354,
+      "grad_norm": 0.5250252074068925,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.9762,
+      "step": 1033
+    },
+    {
+      "epoch": 4.188354430379746,
+      "grad_norm": 0.509466661578754,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.9402,
+      "step": 1034
+    },
+    {
+      "epoch": 4.192405063291139,
+      "grad_norm": 0.4881954541912392,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.9204,
+      "step": 1035
+    },
+    {
+      "epoch": 4.1964556962025314,
+      "grad_norm": 0.5166631971964696,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.9427,
+      "step": 1036
+    },
+    {
+      "epoch": 4.200506329113924,
+      "grad_norm": 0.5138135438095145,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.9562,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2045569620253165,
+      "grad_norm": 0.5351815656053412,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.9432,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2086075949367086,
+      "grad_norm": 0.49603287353303116,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.932,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2126582278481015,
+      "grad_norm": 0.5214433381069242,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.9094,
+      "step": 1040
+    },
+    {
+      "epoch": 4.216708860759494,
+      "grad_norm": 0.5181412093053606,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.9443,
+      "step": 1041
+    },
+    {
+      "epoch": 4.220759493670886,
+      "grad_norm": 0.508607448735455,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.9305,
+      "step": 1042
+    },
+    {
+      "epoch": 4.224810126582279,
+      "grad_norm": 0.5073130066343973,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.9453,
+      "step": 1043
+    },
+    {
+      "epoch": 4.228860759493671,
+      "grad_norm": 0.5003748373092213,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.9616,
+      "step": 1044
+    },
+    {
+      "epoch": 4.232911392405064,
+      "grad_norm": 0.5260857511089219,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.944,
+      "step": 1045
+    },
+    {
+      "epoch": 4.236962025316456,
+      "grad_norm": 0.5192729849536488,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.927,
+      "step": 1046
+    },
+    {
+      "epoch": 4.241012658227848,
+      "grad_norm": 0.5066739575178736,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.9354,
+      "step": 1047
+    },
+    {
+      "epoch": 4.245063291139241,
+      "grad_norm": 0.5078759985045176,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.9413,
+      "step": 1048
+    },
+    {
+      "epoch": 4.249113924050633,
+      "grad_norm": 0.49480277548890844,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.9153,
+      "step": 1049
+    },
+    {
+      "epoch": 4.253164556962025,
+      "grad_norm": 0.4858211758223285,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.9367,
+      "step": 1050
+    },
+    {
+      "epoch": 4.257215189873418,
+      "grad_norm": 0.4985969541870145,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.9373,
+      "step": 1051
+    },
+    {
+      "epoch": 4.26126582278481,
+      "grad_norm": 0.49680790006930026,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.9337,
+      "step": 1052
+    },
+    {
+      "epoch": 4.265316455696203,
+      "grad_norm": 0.4893402517212852,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.9266,
+      "step": 1053
+    },
+    {
+      "epoch": 4.269367088607595,
+      "grad_norm": 0.4917451583923229,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.9389,
+      "step": 1054
+    },
+    {
+      "epoch": 4.273417721518987,
+      "grad_norm": 0.49962102649707424,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.9178,
+      "step": 1055
+    },
+    {
+      "epoch": 4.27746835443038,
+      "grad_norm": 0.485773534789845,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.9297,
+      "step": 1056
+    },
+    {
+      "epoch": 4.281518987341772,
+      "grad_norm": 0.4890635906127962,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.9292,
+      "step": 1057
+    },
+    {
+      "epoch": 4.285569620253165,
+      "grad_norm": 0.4810547562036584,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.9292,
+      "step": 1058
+    },
+    {
+      "epoch": 4.289620253164557,
+      "grad_norm": 0.49916776867212515,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.9482,
+      "step": 1059
+    },
+    {
+      "epoch": 4.293670886075949,
+      "grad_norm": 0.48051676568772966,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.9065,
+      "step": 1060
+    },
+    {
+      "epoch": 4.297721518987342,
+      "grad_norm": 0.48864491286250095,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.9222,
+      "step": 1061
+    },
+    {
+      "epoch": 4.301772151898734,
+      "grad_norm": 0.48649769500829054,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.9156,
+      "step": 1062
+    },
+    {
+      "epoch": 4.305822784810126,
+      "grad_norm": 0.49080651071318415,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.9518,
+      "step": 1063
+    },
+    {
+      "epoch": 4.309873417721519,
+      "grad_norm": 0.5062054013197526,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.928,
+      "step": 1064
+    },
+    {
+      "epoch": 4.313924050632911,
+      "grad_norm": 0.5040998120094453,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.8963,
+      "step": 1065
+    },
+    {
+      "epoch": 4.317974683544303,
+      "grad_norm": 0.5108006975082584,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.9148,
+      "step": 1066
+    },
+    {
+      "epoch": 4.322025316455696,
+      "grad_norm": 0.5148599613065958,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.9544,
+      "step": 1067
+    },
+    {
+      "epoch": 4.326075949367088,
+      "grad_norm": 0.4985873288466356,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.9565,
+      "step": 1068
+    },
+    {
+      "epoch": 4.330126582278481,
+      "grad_norm": 0.5050366017429366,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.9104,
+      "step": 1069
+    },
+    {
+      "epoch": 4.334177215189873,
+      "grad_norm": 0.5040634688073438,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.9291,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3382278481012655,
+      "grad_norm": 0.4927747794494807,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.952,
+      "step": 1071
+    },
+    {
+      "epoch": 4.3422784810126585,
+      "grad_norm": 0.4887016711622718,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.9733,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3463291139240505,
+      "grad_norm": 0.49142568941320935,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.9452,
+      "step": 1073
+    },
+    {
+      "epoch": 4.3503797468354435,
+      "grad_norm": 0.4862688857820641,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.9585,
+      "step": 1074
+    },
+    {
+      "epoch": 4.3544303797468356,
+      "grad_norm": 0.49325913277074696,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.9588,
+      "step": 1075
+    },
+    {
+      "epoch": 4.358481012658228,
+      "grad_norm": 0.49263110566178403,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.9731,
+      "step": 1076
+    },
+    {
+      "epoch": 4.362531645569621,
+      "grad_norm": 0.5008175583399537,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.9314,
+      "step": 1077
+    },
+    {
+      "epoch": 4.366582278481013,
+      "grad_norm": 0.4995908112702826,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.9567,
+      "step": 1078
+    },
+    {
+      "epoch": 4.370632911392405,
+      "grad_norm": 0.48202855232439973,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.9198,
+      "step": 1079
+    },
+    {
+      "epoch": 4.374683544303798,
+      "grad_norm": 0.4979812143728046,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.9247,
+      "step": 1080
+    },
+    {
+      "epoch": 4.37873417721519,
+      "grad_norm": 0.5218427844312082,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.9397,
+      "step": 1081
+    },
+    {
+      "epoch": 4.382784810126582,
+      "grad_norm": 0.5028490101984154,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.9281,
+      "step": 1082
+    },
+    {
+      "epoch": 4.386835443037975,
+      "grad_norm": 0.5124824785670358,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.9705,
+      "step": 1083
+    },
+    {
+      "epoch": 4.390886075949367,
+      "grad_norm": 0.5071219451777795,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.9419,
+      "step": 1084
+    },
+    {
+      "epoch": 4.39493670886076,
+      "grad_norm": 0.5049938890668648,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.9047,
+      "step": 1085
+    },
+    {
+      "epoch": 4.398987341772152,
+      "grad_norm": 0.48566033643055945,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.9146,
+      "step": 1086
+    },
+    {
+      "epoch": 4.403037974683544,
+      "grad_norm": 0.4857281330098013,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.914,
+      "step": 1087
+    },
+    {
+      "epoch": 4.407088607594937,
+      "grad_norm": 0.48043914955541256,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.9532,
+      "step": 1088
+    },
+    {
+      "epoch": 4.411139240506329,
+      "grad_norm": 0.48863046913581965,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.9006,
+      "step": 1089
+    },
+    {
+      "epoch": 4.415189873417722,
+      "grad_norm": 0.4871617327177497,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.9138,
+      "step": 1090
+    },
+    {
+      "epoch": 4.419240506329114,
+      "grad_norm": 0.48097564808393944,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.9658,
+      "step": 1091
+    },
+    {
+      "epoch": 4.423291139240506,
+      "grad_norm": 0.48954691504237247,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.9317,
+      "step": 1092
+    },
+    {
+      "epoch": 4.427341772151899,
+      "grad_norm": 0.4934592568356795,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.9544,
+      "step": 1093
+    },
+    {
+      "epoch": 4.431392405063291,
+      "grad_norm": 0.4801373341553957,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.9101,
+      "step": 1094
+    },
+    {
+      "epoch": 4.435443037974683,
+      "grad_norm": 0.49319334388267994,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.9023,
+      "step": 1095
+    },
+    {
+      "epoch": 4.439493670886076,
+      "grad_norm": 0.4980345251537769,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.9468,
+      "step": 1096
+    },
+    {
+      "epoch": 4.443544303797468,
+      "grad_norm": 0.4824973242894245,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.95,
+      "step": 1097
+    },
+    {
+      "epoch": 4.44759493670886,
+      "grad_norm": 0.4844178281457297,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.9053,
+      "step": 1098
+    },
+    {
+      "epoch": 4.451645569620253,
+      "grad_norm": 0.5121567101364193,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.9072,
+      "step": 1099
+    },
+    {
+      "epoch": 4.455696202531645,
+      "grad_norm": 0.4831367363042172,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.945,
+      "step": 1100
+    },
+    {
+      "epoch": 4.459746835443038,
+      "grad_norm": 0.4929492505350186,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.9872,
+      "step": 1101
+    },
+    {
+      "epoch": 4.46379746835443,
+      "grad_norm": 0.48679769422995534,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.9588,
+      "step": 1102
+    },
+    {
+      "epoch": 4.467848101265822,
+      "grad_norm": 0.4859841364171691,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.9434,
+      "step": 1103
+    },
+    {
+      "epoch": 4.471898734177215,
+      "grad_norm": 0.49211966280110125,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.9453,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4759493670886075,
+      "grad_norm": 0.5017844840013742,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.9719,
+      "step": 1105
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.5026907132617215,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.9038,
+      "step": 1106
+    },
+    {
+      "epoch": 4.4840506329113925,
+      "grad_norm": 0.4938616739696184,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.9405,
+      "step": 1107
+    },
+    {
+      "epoch": 4.488101265822785,
+      "grad_norm": 0.4951040433785826,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.96,
+      "step": 1108
+    },
+    {
+      "epoch": 4.4921518987341775,
+      "grad_norm": 0.4965444654759281,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.949,
+      "step": 1109
+    },
+    {
+      "epoch": 4.49620253164557,
+      "grad_norm": 0.4898056306254701,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.925,
+      "step": 1110
+    },
+    {
+      "epoch": 4.500253164556962,
+      "grad_norm": 0.4961293578099939,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.9201,
+      "step": 1111
+    },
+    {
+      "epoch": 4.504303797468355,
+      "grad_norm": 0.48097521205192956,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.9667,
+      "step": 1112
+    },
+    {
+      "epoch": 4.508354430379747,
+      "grad_norm": 0.49926702327629485,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.9405,
+      "step": 1113
+    },
+    {
+      "epoch": 4.512405063291139,
+      "grad_norm": 0.4987097994495273,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.9383,
+      "step": 1114
+    },
+    {
+      "epoch": 4.516455696202532,
+      "grad_norm": 0.5041454926153422,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.9458,
+      "step": 1115
+    },
+    {
+      "epoch": 4.520506329113924,
+      "grad_norm": 0.49359446950957225,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.9542,
+      "step": 1116
+    },
+    {
+      "epoch": 4.524556962025317,
+      "grad_norm": 0.4828827358660661,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.9277,
+      "step": 1117
+    },
+    {
+      "epoch": 4.528607594936709,
+      "grad_norm": 0.4929697295964308,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.9191,
+      "step": 1118
+    },
+    {
+      "epoch": 4.532658227848101,
+      "grad_norm": 0.4978480857937944,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.9369,
+      "step": 1119
+    },
+    {
+      "epoch": 4.536708860759494,
+      "grad_norm": 0.4882769392842841,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.9017,
+      "step": 1120
+    },
+    {
+      "epoch": 4.540759493670886,
+      "grad_norm": 0.4784889345883871,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.9349,
+      "step": 1121
+    },
+    {
+      "epoch": 4.544810126582279,
+      "grad_norm": 0.4811014627143806,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.9383,
+      "step": 1122
+    },
+    {
+      "epoch": 4.548860759493671,
+      "grad_norm": 0.48431437229636604,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.9148,
+      "step": 1123
+    },
+    {
+      "epoch": 4.552911392405063,
+      "grad_norm": 0.5010036761442731,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.9461,
+      "step": 1124
+    },
+    {
+      "epoch": 4.556962025316456,
+      "grad_norm": 0.49454277914532546,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.9073,
+      "step": 1125
+    },
+    {
+      "epoch": 4.561012658227848,
+      "grad_norm": 0.4884278253849063,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.9642,
+      "step": 1126
+    },
+    {
+      "epoch": 4.56506329113924,
+      "grad_norm": 0.48296551245571295,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.9549,
+      "step": 1127
+    },
+    {
+      "epoch": 4.569113924050633,
+      "grad_norm": 0.4863300629041033,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.9348,
+      "step": 1128
+    },
+    {
+      "epoch": 4.573164556962025,
+      "grad_norm": 0.49528829181042416,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.9213,
+      "step": 1129
+    },
+    {
+      "epoch": 4.577215189873417,
+      "grad_norm": 0.4830239987372704,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.8982,
+      "step": 1130
+    },
+    {
+      "epoch": 4.58126582278481,
+      "grad_norm": 0.4910141103629591,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.9586,
+      "step": 1131
+    },
+    {
+      "epoch": 4.585316455696202,
+      "grad_norm": 0.47592183054452647,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.9293,
+      "step": 1132
+    },
+    {
+      "epoch": 4.589367088607595,
+      "grad_norm": 0.4910504320417566,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.9255,
+      "step": 1133
+    },
+    {
+      "epoch": 4.593417721518987,
+      "grad_norm": 0.4913579489796948,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.9403,
+      "step": 1134
+    },
+    {
+      "epoch": 4.597468354430379,
+      "grad_norm": 0.47839215782001865,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.926,
+      "step": 1135
+    },
+    {
+      "epoch": 4.601518987341772,
+      "grad_norm": 0.48907994167043434,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.9372,
+      "step": 1136
+    },
+    {
+      "epoch": 4.605569620253164,
+      "grad_norm": 0.4877741434564088,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.9279,
+      "step": 1137
+    },
+    {
+      "epoch": 4.609620253164557,
+      "grad_norm": 0.475228077332297,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.944,
+      "step": 1138
+    },
+    {
+      "epoch": 4.613670886075949,
+      "grad_norm": 0.4831930449800069,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.9424,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6177215189873415,
+      "grad_norm": 0.4846237446653407,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.9501,
+      "step": 1140
+    },
+    {
+      "epoch": 4.6217721518987345,
+      "grad_norm": 0.479094050170948,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.9126,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6258227848101265,
+      "grad_norm": 0.480919103253861,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.9508,
+      "step": 1142
+    },
+    {
+      "epoch": 4.629873417721519,
+      "grad_norm": 0.48503378641264955,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.9369,
+      "step": 1143
+    },
+    {
+      "epoch": 4.633924050632912,
+      "grad_norm": 0.4734039180018858,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.9493,
+      "step": 1144
+    },
+    {
+      "epoch": 4.637974683544304,
+      "grad_norm": 0.4787910034232217,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.9553,
+      "step": 1145
+    },
+    {
+      "epoch": 4.642025316455696,
+      "grad_norm": 0.4694267664328505,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.9504,
+      "step": 1146
+    },
+    {
+      "epoch": 4.646075949367089,
+      "grad_norm": 0.48204893873640786,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.9384,
+      "step": 1147
+    },
+    {
+      "epoch": 4.650126582278481,
+      "grad_norm": 0.4968172018965474,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.9785,
+      "step": 1148
+    },
+    {
+      "epoch": 4.654177215189874,
+      "grad_norm": 0.48740684385605804,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.9597,
+      "step": 1149
+    },
+    {
+      "epoch": 4.658227848101266,
+      "grad_norm": 0.48356535556864916,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.9626,
+      "step": 1150
+    },
+    {
+      "epoch": 4.662278481012658,
+      "grad_norm": 0.4786316558755156,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.9531,
+      "step": 1151
+    },
+    {
+      "epoch": 4.666329113924051,
+      "grad_norm": 0.4773773118433581,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.9188,
+      "step": 1152
+    },
+    {
+      "epoch": 4.670379746835443,
+      "grad_norm": 0.47131163368379564,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.9288,
+      "step": 1153
+    },
+    {
+      "epoch": 4.674430379746836,
+      "grad_norm": 0.4738107187884081,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.9258,
+      "step": 1154
+    },
+    {
+      "epoch": 4.678481012658228,
+      "grad_norm": 0.48590552136173715,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.9202,
+      "step": 1155
+    },
+    {
+      "epoch": 4.68253164556962,
+      "grad_norm": 0.48781756293686634,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.9568,
+      "step": 1156
+    },
+    {
+      "epoch": 4.686582278481013,
+      "grad_norm": 0.47528023399988817,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.9677,
+      "step": 1157
+    },
+    {
+      "epoch": 4.690632911392405,
+      "grad_norm": 0.4876200248324055,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.9677,
+      "step": 1158
+    },
+    {
+      "epoch": 4.694683544303797,
+      "grad_norm": 0.4920946332016008,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.9658,
+      "step": 1159
+    },
+    {
+      "epoch": 4.69873417721519,
+      "grad_norm": 0.4734118473037155,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.9426,
+      "step": 1160
+    },
+    {
+      "epoch": 4.702784810126582,
+      "grad_norm": 0.47441581037688735,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.9181,
+      "step": 1161
+    },
+    {
+      "epoch": 4.706835443037974,
+      "grad_norm": 0.4934878165484954,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.9184,
+      "step": 1162
+    },
+    {
+      "epoch": 4.710886075949367,
+      "grad_norm": 0.48344780565567047,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.9048,
+      "step": 1163
+    },
+    {
+      "epoch": 4.714936708860759,
+      "grad_norm": 0.47737741036508824,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.968,
+      "step": 1164
+    },
+    {
+      "epoch": 4.718987341772152,
+      "grad_norm": 0.4940994644584805,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.9468,
+      "step": 1165
+    },
+    {
+      "epoch": 4.723037974683544,
+      "grad_norm": 0.4817661444172113,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.9548,
+      "step": 1166
+    },
+    {
+      "epoch": 4.727088607594936,
+      "grad_norm": 0.4789280770406628,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.9531,
+      "step": 1167
+    },
+    {
+      "epoch": 4.731139240506329,
+      "grad_norm": 0.4846417919210948,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.9171,
+      "step": 1168
+    },
+    {
+      "epoch": 4.735189873417721,
+      "grad_norm": 0.4882304056578239,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.9387,
+      "step": 1169
+    },
+    {
+      "epoch": 4.739240506329114,
+      "grad_norm": 0.4831107074439678,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.9182,
+      "step": 1170
+    },
+    {
+      "epoch": 4.743291139240506,
+      "grad_norm": 0.48207854726243055,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.9381,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7473417721518985,
+      "grad_norm": 0.47905626194945083,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.9475,
+      "step": 1172
+    },
+    {
+      "epoch": 4.751392405063291,
+      "grad_norm": 0.47368206501336024,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.9355,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7554430379746835,
+      "grad_norm": 0.48282617975558423,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.9572,
+      "step": 1174
+    },
+    {
+      "epoch": 4.759493670886076,
+      "grad_norm": 0.47332126111169587,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.9056,
+      "step": 1175
+    },
+    {
+      "epoch": 4.7635443037974685,
+      "grad_norm": 0.476769695000552,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.9583,
+      "step": 1176
+    },
+    {
+      "epoch": 4.767594936708861,
+      "grad_norm": 0.4745472312242688,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.9302,
+      "step": 1177
+    },
+    {
+      "epoch": 4.7716455696202535,
+      "grad_norm": 0.481161027979775,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.949,
+      "step": 1178
+    },
+    {
+      "epoch": 4.775696202531646,
+      "grad_norm": 0.49964596652195314,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.9725,
+      "step": 1179
+    },
+    {
+      "epoch": 4.779746835443038,
+      "grad_norm": 0.4783084505642796,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.9425,
+      "step": 1180
+    },
+    {
+      "epoch": 4.783797468354431,
+      "grad_norm": 0.47704966590293063,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.9884,
+      "step": 1181
+    },
+    {
+      "epoch": 4.787848101265823,
+      "grad_norm": 0.470384290743609,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.9566,
+      "step": 1182
+    },
+    {
+      "epoch": 4.791898734177215,
+      "grad_norm": 0.4742985895471405,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.9356,
+      "step": 1183
+    },
+    {
+      "epoch": 4.795949367088608,
+      "grad_norm": 0.4653199879091189,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.9317,
+      "step": 1184
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.47375456082854167,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.9618,
+      "step": 1185
+    },
+    {
+      "epoch": 4.804050632911393,
+      "grad_norm": 0.4825561510224702,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.9583,
+      "step": 1186
+    },
+    {
+      "epoch": 4.808101265822785,
+      "grad_norm": 0.47528987185851446,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.9473,
+      "step": 1187
+    },
+    {
+      "epoch": 4.812151898734177,
+      "grad_norm": 0.4806209662884066,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.9611,
+      "step": 1188
+    },
+    {
+      "epoch": 4.81620253164557,
+      "grad_norm": 0.4902099609390355,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.9391,
+      "step": 1189
+    },
+    {
+      "epoch": 4.820253164556962,
+      "grad_norm": 0.4854500652328227,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.9622,
+      "step": 1190
+    },
+    {
+      "epoch": 4.824303797468355,
+      "grad_norm": 0.4663433764657648,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.9236,
+      "step": 1191
+    },
+    {
+      "epoch": 4.828354430379747,
+      "grad_norm": 0.4937426649828731,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.9922,
+      "step": 1192
+    },
+    {
+      "epoch": 4.832405063291139,
+      "grad_norm": 0.47972908528688835,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.9394,
+      "step": 1193
+    },
+    {
+      "epoch": 4.836455696202532,
+      "grad_norm": 0.46283873626900873,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.9497,
+      "step": 1194
+    },
+    {
+      "epoch": 4.840506329113924,
+      "grad_norm": 0.48070998048943314,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.936,
+      "step": 1195
+    },
+    {
+      "epoch": 4.844556962025316,
+      "grad_norm": 0.46991487480162614,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.9658,
+      "step": 1196
+    },
+    {
+      "epoch": 4.848607594936709,
+      "grad_norm": 0.48338183318881306,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.9574,
+      "step": 1197
+    },
+    {
+      "epoch": 4.852658227848101,
+      "grad_norm": 0.4701449757488293,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.9268,
+      "step": 1198
+    },
+    {
+      "epoch": 4.856708860759493,
+      "grad_norm": 0.47451111425115755,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.9582,
+      "step": 1199
+    },
+    {
+      "epoch": 4.860759493670886,
+      "grad_norm": 0.47813070298071403,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.9411,
+      "step": 1200
+    },
+    {
+      "epoch": 4.864810126582278,
+      "grad_norm": 0.48105663359364736,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.9109,
+      "step": 1201
+    },
+    {
+      "epoch": 4.868860759493671,
+      "grad_norm": 0.4709202988722081,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.9319,
+      "step": 1202
+    },
+    {
+      "epoch": 4.872911392405063,
+      "grad_norm": 0.4776609529534184,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.9293,
+      "step": 1203
+    },
+    {
+      "epoch": 4.876962025316455,
+      "grad_norm": 0.4763220974642139,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.934,
+      "step": 1204
+    },
+    {
+      "epoch": 4.881012658227848,
+      "grad_norm": 0.46751878240288797,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.9514,
+      "step": 1205
+    },
+    {
+      "epoch": 4.88506329113924,
+      "grad_norm": 0.4632964918279024,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.9626,
+      "step": 1206
+    },
+    {
+      "epoch": 4.889113924050633,
+      "grad_norm": 0.4830308011368329,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.9224,
+      "step": 1207
+    },
+    {
+      "epoch": 4.8931645569620255,
+      "grad_norm": 0.49813987958643313,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.9343,
+      "step": 1208
+    },
+    {
+      "epoch": 4.8972151898734175,
+      "grad_norm": 0.48728471987701505,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.9145,
+      "step": 1209
+    },
+    {
+      "epoch": 4.9012658227848105,
+      "grad_norm": 0.4809190923026788,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.9355,
+      "step": 1210
+    },
+    {
+      "epoch": 4.905316455696203,
+      "grad_norm": 0.47188600824541743,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.9482,
+      "step": 1211
+    },
+    {
+      "epoch": 4.909367088607595,
+      "grad_norm": 0.48145653574684527,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.9746,
+      "step": 1212
+    },
+    {
+      "epoch": 4.913417721518988,
+      "grad_norm": 0.47509542734551913,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.9459,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91746835443038,
+      "grad_norm": 0.4779962281802159,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.9448,
+      "step": 1214
+    },
+    {
+      "epoch": 4.921518987341772,
+      "grad_norm": 0.474312234304907,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.9174,
+      "step": 1215
+    },
+    {
+      "epoch": 4.925569620253165,
+      "grad_norm": 0.48125519333841077,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.9656,
+      "step": 1216
+    },
+    {
+      "epoch": 4.929620253164557,
+      "grad_norm": 0.4947870106473267,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.9333,
+      "step": 1217
+    },
+    {
+      "epoch": 4.93367088607595,
+      "grad_norm": 0.4773686088919652,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.9268,
+      "step": 1218
+    },
+    {
+      "epoch": 4.937721518987342,
+      "grad_norm": 0.4823400313939124,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.9545,
+      "step": 1219
+    },
+    {
+      "epoch": 4.941772151898734,
+      "grad_norm": 0.4570724607752987,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.9249,
+      "step": 1220
+    },
+    {
+      "epoch": 4.945822784810127,
+      "grad_norm": 0.4814720274234706,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.9122,
+      "step": 1221
+    },
+    {
+      "epoch": 4.949873417721519,
+      "grad_norm": 0.47787743504755187,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.9384,
+      "step": 1222
+    },
+    {
+      "epoch": 4.953924050632912,
+      "grad_norm": 0.47179891345241365,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.8777,
+      "step": 1223
+    },
+    {
+      "epoch": 4.957974683544304,
+      "grad_norm": 0.46831135867276397,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.9283,
+      "step": 1224
+    },
+    {
+      "epoch": 4.962025316455696,
+      "grad_norm": 0.48251367700766434,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.9321,
+      "step": 1225
+    },
+    {
+      "epoch": 4.966075949367089,
+      "grad_norm": 0.4753911328015591,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.9192,
+      "step": 1226
+    },
+    {
+      "epoch": 4.970126582278481,
+      "grad_norm": 0.4804985731642478,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.9177,
+      "step": 1227
+    },
+    {
+      "epoch": 4.974177215189873,
+      "grad_norm": 0.4799082826957382,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.9575,
+      "step": 1228
+    },
+    {
+      "epoch": 4.978227848101266,
+      "grad_norm": 0.48282080996557025,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.9095,
+      "step": 1229
+    },
+    {
+      "epoch": 4.982278481012658,
+      "grad_norm": 0.48401689387334246,
+      "learning_rate": 0.0,
+      "loss": 0.9778,
+      "step": 1230
+    },
+    {
+      "epoch": 4.982278481012658,
+      "step": 1230,
+      "total_flos": 2.1651221758317036e+18,
+      "train_loss": 1.2396520211444637,
+      "train_runtime": 15504.8061,
+      "train_samples_per_second": 10.19,
+      "train_steps_per_second": 0.079
+    }
+  ],
+  "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": 2.1651221758317036e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}