diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.988354430379747,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004050632911392405,
+      "grad_norm": 8.05819599507268,
+      "learning_rate": 3.2520325203252037e-07,
+      "loss": 0.8324,
+      "step": 1
+    },
+    {
+      "epoch": 0.00810126582278481,
+      "grad_norm": 7.848467195962461,
+      "learning_rate": 6.504065040650407e-07,
+      "loss": 0.8217,
+      "step": 2
+    },
+    {
+      "epoch": 0.012151898734177215,
+      "grad_norm": 8.653118560915889,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.876,
+      "step": 3
+    },
+    {
+      "epoch": 0.01620253164556962,
+      "grad_norm": 8.48835919274855,
+      "learning_rate": 1.3008130081300815e-06,
+      "loss": 0.8797,
+      "step": 4
+    },
+    {
+      "epoch": 0.020253164556962026,
+      "grad_norm": 7.965502883555725,
+      "learning_rate": 1.6260162601626018e-06,
+      "loss": 0.8389,
+      "step": 5
+    },
+    {
+      "epoch": 0.02430379746835443,
+      "grad_norm": 7.470873741941392,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.8282,
+      "step": 6
+    },
+    {
+      "epoch": 0.028354430379746835,
+      "grad_norm": 6.264087527041432,
+      "learning_rate": 2.2764227642276426e-06,
+      "loss": 0.7765,
+      "step": 7
+    },
+    {
+      "epoch": 0.03240506329113924,
+      "grad_norm": 5.5361690587656796,
+      "learning_rate": 2.601626016260163e-06,
+      "loss": 0.7528,
+      "step": 8
+    },
+    {
+      "epoch": 0.03645569620253165,
+      "grad_norm": 3.2753838999534475,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.6888,
+      "step": 9
+    },
+    {
+      "epoch": 0.04050632911392405,
+      "grad_norm": 2.923177122401346,
+      "learning_rate": 3.2520325203252037e-06,
+      "loss": 0.6805,
+      "step": 10
+    },
+    {
+      "epoch": 0.044556962025316456,
+      "grad_norm": 2.893921352377869,
+      "learning_rate": 3.577235772357724e-06,
+      "loss": 0.6949,
+      "step": 11
+    },
+    {
+      "epoch": 0.04860759493670886,
+      "grad_norm": 3.7906886775683977,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.6366,
+      "step": 12
+    },
+    {
+      "epoch": 0.052658227848101265,
+      "grad_norm": 3.937286167363296,
+      "learning_rate": 4.227642276422765e-06,
+      "loss": 0.6532,
+      "step": 13
+    },
+    {
+      "epoch": 0.05670886075949367,
+      "grad_norm": 3.498634014430776,
+      "learning_rate": 4.552845528455285e-06,
+      "loss": 0.6114,
+      "step": 14
+    },
+    {
+      "epoch": 0.060759493670886074,
+      "grad_norm": 3.261845169800002,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.6239,
+      "step": 15
+    },
+    {
+      "epoch": 0.06481012658227848,
+      "grad_norm": 2.7743393948250437,
+      "learning_rate": 5.203252032520326e-06,
+      "loss": 0.5644,
+      "step": 16
+    },
+    {
+      "epoch": 0.06886075949367089,
+      "grad_norm": 2.5918671247048604,
+      "learning_rate": 5.528455284552846e-06,
+      "loss": 0.5417,
+      "step": 17
+    },
+    {
+      "epoch": 0.0729113924050633,
+      "grad_norm": 2.374541787563303,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.5387,
+      "step": 18
+    },
+    {
+      "epoch": 0.0769620253164557,
+      "grad_norm": 1.6510121902937769,
+      "learning_rate": 6.178861788617887e-06,
+      "loss": 0.4883,
+      "step": 19
+    },
+    {
+      "epoch": 0.0810126582278481,
+      "grad_norm": 1.3842239821910804,
+      "learning_rate": 6.504065040650407e-06,
+      "loss": 0.5378,
+      "step": 20
+    },
+    {
+      "epoch": 0.08506329113924051,
+      "grad_norm": 1.1546787604788677,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.4983,
+      "step": 21
+    },
+    {
+      "epoch": 0.08911392405063291,
+      "grad_norm": 1.2542848814539684,
+      "learning_rate": 7.154471544715448e-06,
+      "loss": 0.4805,
+      "step": 22
+    },
+    {
+      "epoch": 0.09316455696202532,
+      "grad_norm": 1.2767730266852426,
+      "learning_rate": 7.4796747967479676e-06,
+      "loss": 0.4789,
+      "step": 23
+    },
+    {
+      "epoch": 0.09721518987341772,
+      "grad_norm": 0.9751890409469309,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.4583,
+      "step": 24
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.0389417170673476,
+      "learning_rate": 8.130081300813009e-06,
+      "loss": 0.4616,
+      "step": 25
+    },
+    {
+      "epoch": 0.10531645569620253,
+      "grad_norm": 0.9437719434044721,
+      "learning_rate": 8.45528455284553e-06,
+      "loss": 0.4337,
+      "step": 26
+    },
+    {
+      "epoch": 0.10936708860759493,
+      "grad_norm": 0.8451542460378371,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.446,
+      "step": 27
+    },
+    {
+      "epoch": 0.11341772151898734,
+      "grad_norm": 0.8575026253637806,
+      "learning_rate": 9.10569105691057e-06,
+      "loss": 0.4446,
+      "step": 28
+    },
+    {
+      "epoch": 0.11746835443037974,
+      "grad_norm": 0.7819957525958288,
+      "learning_rate": 9.43089430894309e-06,
+      "loss": 0.4413,
+      "step": 29
+    },
+    {
+      "epoch": 0.12151898734177215,
+      "grad_norm": 0.7254271585869795,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.4176,
+      "step": 30
+    },
+    {
+      "epoch": 0.12556962025316457,
+      "grad_norm": 0.7835795074246176,
+      "learning_rate": 1.008130081300813e-05,
+      "loss": 0.4196,
+      "step": 31
+    },
+    {
+      "epoch": 0.12962025316455697,
+      "grad_norm": 0.7245663106011094,
+      "learning_rate": 1.0406504065040652e-05,
+      "loss": 0.4125,
+      "step": 32
+    },
+    {
+      "epoch": 0.13367088607594937,
+      "grad_norm": 0.7537121594377464,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.4303,
+      "step": 33
+    },
+    {
+      "epoch": 0.13772151898734178,
+      "grad_norm": 0.6629732783711294,
+      "learning_rate": 1.1056910569105692e-05,
+      "loss": 0.429,
+      "step": 34
+    },
+    {
+      "epoch": 0.14177215189873418,
+      "grad_norm": 0.7643722160053215,
+      "learning_rate": 1.1382113821138213e-05,
+      "loss": 0.4239,
+      "step": 35
+    },
+    {
+      "epoch": 0.1458227848101266,
+      "grad_norm": 0.6398070053399993,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 0.41,
+      "step": 36
+    },
+    {
+      "epoch": 0.149873417721519,
+      "grad_norm": 0.5969500160625146,
+      "learning_rate": 1.2032520325203254e-05,
+      "loss": 0.3979,
+      "step": 37
+    },
+    {
+      "epoch": 0.1539240506329114,
+      "grad_norm": 0.6919559161315522,
+      "learning_rate": 1.2357723577235774e-05,
+      "loss": 0.4128,
+      "step": 38
+    },
+    {
+      "epoch": 0.1579746835443038,
+      "grad_norm": 0.5776504220190729,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.3997,
+      "step": 39
+    },
+    {
+      "epoch": 0.1620253164556962,
+      "grad_norm": 0.6748323645995529,
+      "learning_rate": 1.3008130081300815e-05,
+      "loss": 0.399,
+      "step": 40
+    },
+    {
+      "epoch": 0.1660759493670886,
+      "grad_norm": 0.5815598087275042,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.3849,
+      "step": 41
+    },
+    {
+      "epoch": 0.17012658227848101,
+      "grad_norm": 0.5991575095338975,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 0.3974,
+      "step": 42
+    },
+    {
+      "epoch": 0.17417721518987342,
+      "grad_norm": 0.7022112672527474,
+      "learning_rate": 1.3983739837398376e-05,
+      "loss": 0.376,
+      "step": 43
+    },
+    {
+      "epoch": 0.17822784810126582,
+      "grad_norm": 0.6582769215760502,
+      "learning_rate": 1.4308943089430896e-05,
+      "loss": 0.3923,
+      "step": 44
+    },
+    {
+      "epoch": 0.18227848101265823,
+      "grad_norm": 0.7994750079590498,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 0.3845,
+      "step": 45
+    },
+    {
+      "epoch": 0.18632911392405063,
+      "grad_norm": 0.6439414824081495,
+      "learning_rate": 1.4959349593495935e-05,
+      "loss": 0.3766,
+      "step": 46
+    },
+    {
+      "epoch": 0.19037974683544304,
+      "grad_norm": 0.6698828774032329,
+      "learning_rate": 1.528455284552846e-05,
+      "loss": 0.3842,
+      "step": 47
+    },
+    {
+      "epoch": 0.19443037974683544,
+      "grad_norm": 0.6327177927793102,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 0.3956,
+      "step": 48
+    },
+    {
+      "epoch": 0.19848101265822785,
+      "grad_norm": 0.6504673340498772,
+      "learning_rate": 1.5934959349593496e-05,
+      "loss": 0.3827,
+      "step": 49
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.573901785599904,
+      "learning_rate": 1.6260162601626018e-05,
+      "loss": 0.3929,
+      "step": 50
+    },
+    {
+      "epoch": 0.20658227848101265,
+      "grad_norm": 0.7268006057380433,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.3864,
+      "step": 51
+    },
+    {
+      "epoch": 0.21063291139240506,
+      "grad_norm": 0.7373048116206145,
+      "learning_rate": 1.691056910569106e-05,
+      "loss": 0.3728,
+      "step": 52
+    },
+    {
+      "epoch": 0.21468354430379746,
+      "grad_norm": 0.6697875042143181,
+      "learning_rate": 1.7235772357723578e-05,
+      "loss": 0.3808,
+      "step": 53
+    },
+    {
+      "epoch": 0.21873417721518987,
+      "grad_norm": 0.934325831873399,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 0.3565,
+      "step": 54
+    },
+    {
+      "epoch": 0.22278481012658227,
+      "grad_norm": 0.6792931146971879,
+      "learning_rate": 1.788617886178862e-05,
+      "loss": 0.3728,
+      "step": 55
+    },
+    {
+      "epoch": 0.22683544303797468,
+      "grad_norm": 0.8730089305316302,
+      "learning_rate": 1.821138211382114e-05,
+      "loss": 0.3736,
+      "step": 56
+    },
+    {
+      "epoch": 0.23088607594936708,
+      "grad_norm": 0.743032528527125,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 0.3734,
+      "step": 57
+    },
+    {
+      "epoch": 0.23493670886075949,
+      "grad_norm": 0.7870983132582742,
+      "learning_rate": 1.886178861788618e-05,
+      "loss": 0.3756,
+      "step": 58
+    },
+    {
+      "epoch": 0.2389873417721519,
+      "grad_norm": 0.724694431348685,
+      "learning_rate": 1.91869918699187e-05,
+      "loss": 0.3842,
+      "step": 59
+    },
+    {
+      "epoch": 0.2430379746835443,
+      "grad_norm": 0.8882248369212769,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 0.3783,
+      "step": 60
+    },
+    {
+      "epoch": 0.2470886075949367,
+      "grad_norm": 0.9875058718047589,
+      "learning_rate": 1.983739837398374e-05,
+      "loss": 0.3797,
+      "step": 61
+    },
+    {
+      "epoch": 0.25113924050632913,
+      "grad_norm": 0.6818193479219504,
+      "learning_rate": 2.016260162601626e-05,
+      "loss": 0.3624,
+      "step": 62
+    },
+    {
+      "epoch": 0.25518987341772154,
+      "grad_norm": 0.8959359449077375,
+      "learning_rate": 2.048780487804878e-05,
+      "loss": 0.364,
+      "step": 63
+    },
+    {
+      "epoch": 0.25924050632911394,
+      "grad_norm": 0.7578240324233603,
+      "learning_rate": 2.0813008130081303e-05,
+      "loss": 0.3711,
+      "step": 64
+    },
+    {
+      "epoch": 0.26329113924050634,
+      "grad_norm": 0.8208007853059349,
+      "learning_rate": 2.1138211382113822e-05,
+      "loss": 0.3723,
+      "step": 65
+    },
+    {
+      "epoch": 0.26734177215189875,
+      "grad_norm": 0.7816777040640615,
+      "learning_rate": 2.1463414634146344e-05,
+      "loss": 0.3726,
+      "step": 66
+    },
+    {
+      "epoch": 0.27139240506329115,
+      "grad_norm": 0.9063112419209278,
+      "learning_rate": 2.1788617886178863e-05,
+      "loss": 0.3565,
+      "step": 67
+    },
+    {
+      "epoch": 0.27544303797468356,
+      "grad_norm": 0.6108572313783478,
+      "learning_rate": 2.2113821138211385e-05,
+      "loss": 0.352,
+      "step": 68
+    },
+    {
+      "epoch": 0.27949367088607596,
+      "grad_norm": 0.9430404488018518,
+      "learning_rate": 2.2439024390243907e-05,
+      "loss": 0.3555,
+      "step": 69
+    },
+    {
+      "epoch": 0.28354430379746837,
+      "grad_norm": 0.7542665122812747,
+      "learning_rate": 2.2764227642276426e-05,
+      "loss": 0.3603,
+      "step": 70
+    },
+    {
+      "epoch": 0.28759493670886077,
+      "grad_norm": 0.724653538668058,
+      "learning_rate": 2.3089430894308948e-05,
+      "loss": 0.3683,
+      "step": 71
+    },
+    {
+      "epoch": 0.2916455696202532,
+      "grad_norm": 0.7722941461749725,
+      "learning_rate": 2.3414634146341463e-05,
+      "loss": 0.3577,
+      "step": 72
+    },
+    {
+      "epoch": 0.2956962025316456,
+      "grad_norm": 0.667553937848939,
+      "learning_rate": 2.3739837398373985e-05,
+      "loss": 0.3633,
+      "step": 73
+    },
+    {
+      "epoch": 0.299746835443038,
+      "grad_norm": 0.8178721420510607,
+      "learning_rate": 2.4065040650406507e-05,
+      "loss": 0.3494,
+      "step": 74
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.6109902319434565,
+      "learning_rate": 2.4390243902439026e-05,
+      "loss": 0.3505,
+      "step": 75
+    },
+    {
+      "epoch": 0.3078481012658228,
+      "grad_norm": 0.6737152692854489,
+      "learning_rate": 2.4715447154471548e-05,
+      "loss": 0.3549,
+      "step": 76
+    },
+    {
+      "epoch": 0.3118987341772152,
+      "grad_norm": 0.5874631813425014,
+      "learning_rate": 2.5040650406504066e-05,
+      "loss": 0.3673,
+      "step": 77
+    },
+    {
+      "epoch": 0.3159493670886076,
+      "grad_norm": 0.6828317849771783,
+      "learning_rate": 2.536585365853659e-05,
+      "loss": 0.337,
+      "step": 78
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6576040813743189,
+      "learning_rate": 2.569105691056911e-05,
+      "loss": 0.3677,
+      "step": 79
+    },
+    {
+      "epoch": 0.3240506329113924,
+      "grad_norm": 0.8555145288476081,
+      "learning_rate": 2.601626016260163e-05,
+      "loss": 0.3476,
+      "step": 80
+    },
+    {
+      "epoch": 0.3281012658227848,
+      "grad_norm": 0.6990333963417039,
+      "learning_rate": 2.634146341463415e-05,
+      "loss": 0.3436,
+      "step": 81
+    },
+    {
+      "epoch": 0.3321518987341772,
+      "grad_norm": 0.7731419790188844,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.3389,
+      "step": 82
+    },
+    {
+      "epoch": 0.3362025316455696,
+      "grad_norm": 0.7584818022422869,
+      "learning_rate": 2.699186991869919e-05,
+      "loss": 0.3534,
+      "step": 83
+    },
+    {
+      "epoch": 0.34025316455696203,
+      "grad_norm": 0.6465995281844283,
+      "learning_rate": 2.731707317073171e-05,
+      "loss": 0.3315,
+      "step": 84
+    },
+    {
+      "epoch": 0.34430379746835443,
+      "grad_norm": 0.9381392526567806,
+      "learning_rate": 2.764227642276423e-05,
+      "loss": 0.3641,
+      "step": 85
+    },
+    {
+      "epoch": 0.34835443037974684,
+      "grad_norm": 0.733992997350353,
+      "learning_rate": 2.796747967479675e-05,
+      "loss": 0.3556,
+      "step": 86
+    },
+    {
+      "epoch": 0.35240506329113924,
+      "grad_norm": 0.8216004721250206,
+      "learning_rate": 2.829268292682927e-05,
+      "loss": 0.3517,
+      "step": 87
+    },
+    {
+      "epoch": 0.35645569620253165,
+      "grad_norm": 0.7011938227356519,
+      "learning_rate": 2.8617886178861792e-05,
+      "loss": 0.3457,
+      "step": 88
+    },
+    {
+      "epoch": 0.36050632911392405,
+      "grad_norm": 0.6447682298431209,
+      "learning_rate": 2.8943089430894314e-05,
+      "loss": 0.3642,
+      "step": 89
+    },
+    {
+      "epoch": 0.36455696202531646,
+      "grad_norm": 0.6244969777835045,
+      "learning_rate": 2.926829268292683e-05,
+      "loss": 0.3486,
+      "step": 90
+    },
+    {
+      "epoch": 0.36860759493670886,
+      "grad_norm": 0.6525230802129197,
+      "learning_rate": 2.959349593495935e-05,
+      "loss": 0.3477,
+      "step": 91
+    },
+    {
+      "epoch": 0.37265822784810126,
+      "grad_norm": 0.6415180707087147,
+      "learning_rate": 2.991869918699187e-05,
+      "loss": 0.3538,
+      "step": 92
+    },
+    {
+      "epoch": 0.37670886075949367,
+      "grad_norm": 0.5735058976818895,
+      "learning_rate": 3.0243902439024392e-05,
+      "loss": 0.3476,
+      "step": 93
+    },
+    {
+      "epoch": 0.3807594936708861,
+      "grad_norm": 0.6338745886511703,
+      "learning_rate": 3.056910569105692e-05,
+      "loss": 0.3419,
+      "step": 94
+    },
+    {
+      "epoch": 0.3848101265822785,
+      "grad_norm": 0.9660817362399025,
+      "learning_rate": 3.089430894308943e-05,
+      "loss": 0.348,
+      "step": 95
+    },
+    {
+      "epoch": 0.3888607594936709,
+      "grad_norm": 1.0209502057048343,
+      "learning_rate": 3.1219512195121955e-05,
+      "loss": 0.3383,
+      "step": 96
+    },
+    {
+      "epoch": 0.3929113924050633,
+      "grad_norm": 0.621002876440081,
+      "learning_rate": 3.154471544715447e-05,
+      "loss": 0.3558,
+      "step": 97
+    },
+    {
+      "epoch": 0.3969620253164557,
+      "grad_norm": 0.6313641150234258,
+      "learning_rate": 3.186991869918699e-05,
+      "loss": 0.331,
+      "step": 98
+    },
+    {
+      "epoch": 0.4010126582278481,
+      "grad_norm": 0.8993281228121466,
+      "learning_rate": 3.2195121951219514e-05,
+      "loss": 0.3484,
+      "step": 99
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.7654231074605067,
+      "learning_rate": 3.2520325203252037e-05,
+      "loss": 0.3562,
+      "step": 100
+    },
+    {
+      "epoch": 0.4091139240506329,
+      "grad_norm": 0.5552662280574909,
+      "learning_rate": 3.284552845528456e-05,
+      "loss": 0.3472,
+      "step": 101
+    },
+    {
+      "epoch": 0.4131645569620253,
+      "grad_norm": 0.6079173634960368,
+      "learning_rate": 3.3170731707317074e-05,
+      "loss": 0.3273,
+      "step": 102
+    },
+    {
+      "epoch": 0.4172151898734177,
+      "grad_norm": 0.8407671211299851,
+      "learning_rate": 3.3495934959349596e-05,
+      "loss": 0.3509,
+      "step": 103
+    },
+    {
+      "epoch": 0.4212658227848101,
+      "grad_norm": 1.1347544741560678,
+      "learning_rate": 3.382113821138212e-05,
+      "loss": 0.3484,
+      "step": 104
+    },
+    {
+      "epoch": 0.4253164556962025,
+      "grad_norm": 0.6403248370782134,
+      "learning_rate": 3.414634146341463e-05,
+      "loss": 0.3461,
+      "step": 105
+    },
+    {
+      "epoch": 0.4293670886075949,
+      "grad_norm": 0.8986319679737554,
+      "learning_rate": 3.4471544715447155e-05,
+      "loss": 0.3445,
+      "step": 106
+    },
+    {
+      "epoch": 0.43341772151898733,
+      "grad_norm": 0.935753216246839,
+      "learning_rate": 3.479674796747968e-05,
+      "loss": 0.3478,
+      "step": 107
+    },
+    {
+      "epoch": 0.43746835443037974,
+      "grad_norm": 0.8929136040536524,
+      "learning_rate": 3.51219512195122e-05,
+      "loss": 0.3476,
+      "step": 108
+    },
+    {
+      "epoch": 0.44151898734177214,
+      "grad_norm": 0.6966339510931092,
+      "learning_rate": 3.544715447154472e-05,
+      "loss": 0.3538,
+      "step": 109
+    },
+    {
+      "epoch": 0.44556962025316454,
+      "grad_norm": 0.8742501785694772,
+      "learning_rate": 3.577235772357724e-05,
+      "loss": 0.3414,
+      "step": 110
+    },
+    {
+      "epoch": 0.44962025316455695,
+      "grad_norm": 0.8619665660698809,
+      "learning_rate": 3.609756097560976e-05,
+      "loss": 0.3424,
+      "step": 111
+    },
+    {
+      "epoch": 0.45367088607594935,
+      "grad_norm": 0.9019318898920907,
+      "learning_rate": 3.642276422764228e-05,
+      "loss": 0.3337,
+      "step": 112
+    },
+    {
+      "epoch": 0.45772151898734176,
+      "grad_norm": 0.6578112244877221,
+      "learning_rate": 3.67479674796748e-05,
+      "loss": 0.3481,
+      "step": 113
+    },
+    {
+      "epoch": 0.46177215189873416,
+      "grad_norm": 0.931978065189565,
+      "learning_rate": 3.7073170731707325e-05,
+      "loss": 0.3444,
+      "step": 114
+    },
+    {
+      "epoch": 0.46582278481012657,
+      "grad_norm": 0.7706288319300265,
+      "learning_rate": 3.739837398373984e-05,
+      "loss": 0.3426,
+      "step": 115
+    },
+    {
+      "epoch": 0.46987341772151897,
+      "grad_norm": 0.7369190304781595,
+      "learning_rate": 3.772357723577236e-05,
+      "loss": 0.347,
+      "step": 116
+    },
+    {
+      "epoch": 0.4739240506329114,
+      "grad_norm": 0.665885423543691,
+      "learning_rate": 3.804878048780488e-05,
+      "loss": 0.3392,
+      "step": 117
+    },
+    {
+      "epoch": 0.4779746835443038,
+      "grad_norm": 0.7695883862567001,
+      "learning_rate": 3.83739837398374e-05,
+      "loss": 0.3501,
+      "step": 118
+    },
+    {
+      "epoch": 0.4820253164556962,
+      "grad_norm": 0.5533995716517147,
+      "learning_rate": 3.869918699186992e-05,
+      "loss": 0.3463,
+      "step": 119
+    },
+    {
+      "epoch": 0.4860759493670886,
+      "grad_norm": 0.6663490560930994,
+      "learning_rate": 3.9024390243902444e-05,
+      "loss": 0.3519,
+      "step": 120
+    },
+    {
+      "epoch": 0.490126582278481,
+      "grad_norm": 0.7576054567359516,
+      "learning_rate": 3.9349593495934966e-05,
+      "loss": 0.3445,
+      "step": 121
+    },
+    {
+      "epoch": 0.4941772151898734,
+      "grad_norm": 0.7623332685427542,
+      "learning_rate": 3.967479674796748e-05,
+      "loss": 0.3533,
+      "step": 122
+    },
+    {
+      "epoch": 0.4982278481012658,
+      "grad_norm": 0.6699833019648852,
+      "learning_rate": 4e-05,
+      "loss": 0.3499,
+      "step": 123
+    },
+    {
+      "epoch": 0.5022784810126583,
+      "grad_norm": 0.9225579699625168,
+      "learning_rate": 3.999991946137476e-05,
+      "loss": 0.3511,
+      "step": 124
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.6565473918220055,
+      "learning_rate": 3.999967784614766e-05,
+      "loss": 0.35,
+      "step": 125
+    },
+    {
+      "epoch": 0.5103797468354431,
+      "grad_norm": 0.7821075219045001,
+      "learning_rate": 3.9999275156264656e-05,
+      "loss": 0.3503,
+      "step": 126
+    },
+    {
+      "epoch": 0.5144303797468355,
+      "grad_norm": 0.8135545888519947,
+      "learning_rate": 3.999871139496895e-05,
+      "loss": 0.337,
+      "step": 127
+    },
+    {
+      "epoch": 0.5184810126582279,
+      "grad_norm": 0.6921826317626489,
+      "learning_rate": 3.9997986566800995e-05,
+      "loss": 0.3589,
+      "step": 128
+    },
+    {
+      "epoch": 0.5225316455696203,
+      "grad_norm": 0.6665166855339147,
+      "learning_rate": 3.999710067759846e-05,
+      "loss": 0.3303,
+      "step": 129
+    },
+    {
+      "epoch": 0.5265822784810127,
+      "grad_norm": 0.6860967890619778,
+      "learning_rate": 3.999605373449617e-05,
+      "loss": 0.3473,
+      "step": 130
+    },
+    {
+      "epoch": 0.5306329113924051,
+      "grad_norm": 0.5304297019059846,
+      "learning_rate": 3.9994845745926075e-05,
+      "loss": 0.3589,
+      "step": 131
+    },
+    {
+      "epoch": 0.5346835443037975,
+      "grad_norm": 0.5865422006715262,
+      "learning_rate": 3.999347672161713e-05,
+      "loss": 0.347,
+      "step": 132
+    },
+    {
+      "epoch": 0.5387341772151899,
+      "grad_norm": 0.5624027932122773,
+      "learning_rate": 3.999194667259528e-05,
+      "loss": 0.3388,
+      "step": 133
+    },
+    {
+      "epoch": 0.5427848101265823,
+      "grad_norm": 0.5701764241932465,
+      "learning_rate": 3.999025561118334e-05,
+      "loss": 0.3508,
+      "step": 134
+    },
+    {
+      "epoch": 0.5468354430379747,
+      "grad_norm": 0.5575987665800516,
+      "learning_rate": 3.998840355100086e-05,
+      "loss": 0.327,
+      "step": 135
+    },
+    {
+      "epoch": 0.5508860759493671,
+      "grad_norm": 0.5858998302552779,
+      "learning_rate": 3.998639050696409e-05,
+      "loss": 0.3312,
+      "step": 136
+    },
+    {
+      "epoch": 0.5549367088607595,
+      "grad_norm": 0.6135412088086432,
+      "learning_rate": 3.998421649528582e-05,
+      "loss": 0.3513,
+      "step": 137
+    },
+    {
+      "epoch": 0.5589873417721519,
+      "grad_norm": 0.6111561611682713,
+      "learning_rate": 3.9981881533475234e-05,
+      "loss": 0.3506,
+      "step": 138
+    },
+    {
+      "epoch": 0.5630379746835443,
+      "grad_norm": 0.5692404995173551,
+      "learning_rate": 3.997938564033779e-05,
+      "loss": 0.3325,
+      "step": 139
+    },
+    {
+      "epoch": 0.5670886075949367,
+      "grad_norm": 0.7464314268799813,
+      "learning_rate": 3.9976728835975064e-05,
+      "loss": 0.3366,
+      "step": 140
+    },
+    {
+      "epoch": 0.5711392405063291,
+      "grad_norm": 0.6075755636603308,
+      "learning_rate": 3.9973911141784605e-05,
+      "loss": 0.3456,
+      "step": 141
+    },
+    {
+      "epoch": 0.5751898734177215,
+      "grad_norm": 0.7708062158285452,
+      "learning_rate": 3.997093258045973e-05,
+      "loss": 0.3217,
+      "step": 142
+    },
+    {
+      "epoch": 0.579240506329114,
+      "grad_norm": 0.7745644745002094,
+      "learning_rate": 3.996779317598936e-05,
+      "loss": 0.3386,
+      "step": 143
+    },
+    {
+      "epoch": 0.5832911392405064,
+      "grad_norm": 0.5817372694819892,
+      "learning_rate": 3.996449295365782e-05,
+      "loss": 0.3335,
+      "step": 144
+    },
+    {
+      "epoch": 0.5873417721518988,
+      "grad_norm": 0.7782727162735549,
+      "learning_rate": 3.996103194004467e-05,
+      "loss": 0.3415,
+      "step": 145
+    },
+    {
+      "epoch": 0.5913924050632912,
+      "grad_norm": 0.6864691526633074,
+      "learning_rate": 3.995741016302441e-05,
+      "loss": 0.3412,
+      "step": 146
+    },
+    {
+      "epoch": 0.5954430379746836,
+      "grad_norm": 0.6159522924950384,
+      "learning_rate": 3.9953627651766364e-05,
+      "loss": 0.3464,
+      "step": 147
+    },
+    {
+      "epoch": 0.599493670886076,
+      "grad_norm": 0.6978699335183439,
+      "learning_rate": 3.9949684436734325e-05,
+      "loss": 0.3373,
+      "step": 148
+    },
+    {
+      "epoch": 0.6035443037974684,
+      "grad_norm": 0.5277086517096553,
+      "learning_rate": 3.994558054968643e-05,
+      "loss": 0.3217,
+      "step": 149
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.6455264361779735,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.3399,
+      "step": 150
+    },
+    {
+      "epoch": 0.6116455696202532,
+      "grad_norm": 0.5225164800418022,
+      "learning_rate": 3.9936890893045376e-05,
+      "loss": 0.3173,
+      "step": 151
+    },
+    {
+      "epoch": 0.6156962025316456,
+      "grad_norm": 0.570423755397257,
+      "learning_rate": 3.993230519343752e-05,
+      "loss": 0.3452,
+      "step": 152
+    },
+    {
+      "epoch": 0.619746835443038,
+      "grad_norm": 0.5547024308626956,
+      "learning_rate": 3.992755896178383e-05,
+      "loss": 0.347,
+      "step": 153
+    },
+    {
+      "epoch": 0.6237974683544304,
+      "grad_norm": 0.5200087525600763,
+      "learning_rate": 3.992265223630981e-05,
+      "loss": 0.3275,
+      "step": 154
+    },
+    {
+      "epoch": 0.6278481012658228,
+      "grad_norm": 0.5436541088288269,
+      "learning_rate": 3.991758505653355e-05,
+      "loss": 0.3436,
+      "step": 155
+    },
+    {
+      "epoch": 0.6318987341772152,
+      "grad_norm": 0.5683569910599022,
+      "learning_rate": 3.991235746326543e-05,
+      "loss": 0.3327,
+      "step": 156
+    },
+    {
+      "epoch": 0.6359493670886076,
+      "grad_norm": 0.4968861312180156,
+      "learning_rate": 3.9906969498607745e-05,
+      "loss": 0.3217,
+      "step": 157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6083456856859718,
+      "learning_rate": 3.990142120595444e-05,
+      "loss": 0.3403,
+      "step": 158
+    },
+    {
+      "epoch": 0.6440506329113924,
+      "grad_norm": 0.660628395066163,
+      "learning_rate": 3.98957126299907e-05,
+      "loss": 0.3356,
+      "step": 159
+    },
+    {
+      "epoch": 0.6481012658227848,
+      "grad_norm": 0.5157575561047221,
+      "learning_rate": 3.9889843816692596e-05,
+      "loss": 0.3359,
+      "step": 160
+    },
+    {
+      "epoch": 0.6521518987341772,
+      "grad_norm": 0.6849899335402262,
+      "learning_rate": 3.9883814813326766e-05,
+      "loss": 0.3269,
+      "step": 161
+    },
+    {
+      "epoch": 0.6562025316455696,
+      "grad_norm": 0.6219046105514563,
+      "learning_rate": 3.9877625668449956e-05,
+      "loss": 0.3331,
+      "step": 162
+    },
+    {
+      "epoch": 0.660253164556962,
+      "grad_norm": 0.6057025422690239,
+      "learning_rate": 3.98712764319087e-05,
+      "loss": 0.3354,
+      "step": 163
+    },
+    {
+      "epoch": 0.6643037974683544,
+      "grad_norm": 0.6037119903910835,
+      "learning_rate": 3.9864767154838864e-05,
+      "loss": 0.3337,
+      "step": 164
+    },
+    {
+      "epoch": 0.6683544303797468,
+      "grad_norm": 0.4898906352083855,
+      "learning_rate": 3.9858097889665277e-05,
+      "loss": 0.3169,
+      "step": 165
+    },
+    {
+      "epoch": 0.6724050632911392,
+      "grad_norm": 0.6186075700169921,
+      "learning_rate": 3.985126869010129e-05,
+      "loss": 0.321,
+      "step": 166
+    },
+    {
+      "epoch": 0.6764556962025317,
+      "grad_norm": 0.5512816520992717,
+      "learning_rate": 3.984427961114833e-05,
+      "loss": 0.3399,
+      "step": 167
+    },
+    {
+      "epoch": 0.6805063291139241,
+      "grad_norm": 0.6354084840250838,
+      "learning_rate": 3.9837130709095475e-05,
+      "loss": 0.333,
+      "step": 168
+    },
+    {
+      "epoch": 0.6845569620253165,
+      "grad_norm": 0.5125497101063142,
+      "learning_rate": 3.982982204151901e-05,
+      "loss": 0.3318,
+      "step": 169
+    },
+    {
+      "epoch": 0.6886075949367089,
+      "grad_norm": 0.6625645134149334,
+      "learning_rate": 3.982235366728193e-05,
+      "loss": 0.3411,
+      "step": 170
+    },
+    {
+      "epoch": 0.6926582278481013,
+      "grad_norm": 0.49731408840664887,
+      "learning_rate": 3.9814725646533505e-05,
+      "loss": 0.3364,
+      "step": 171
+    },
+    {
+      "epoch": 0.6967088607594937,
+      "grad_norm": 0.5631046778166326,
+      "learning_rate": 3.9806938040708746e-05,
+      "loss": 0.3451,
+      "step": 172
+    },
+    {
+      "epoch": 0.7007594936708861,
+      "grad_norm": 0.455230209913656,
+      "learning_rate": 3.9798990912527976e-05,
+      "loss": 0.3376,
+      "step": 173
+    },
+    {
+      "epoch": 0.7048101265822785,
+      "grad_norm": 0.551211721229534,
+      "learning_rate": 3.979088432599627e-05,
+      "loss": 0.3308,
+      "step": 174
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.43870605706147836,
+      "learning_rate": 3.9782618346402964e-05,
+      "loss": 0.3244,
+      "step": 175
+    },
+    {
+      "epoch": 0.7129113924050633,
+      "grad_norm": 0.47354578977987677,
+      "learning_rate": 3.977419304032111e-05,
+      "loss": 0.3336,
+      "step": 176
+    },
+    {
+      "epoch": 0.7169620253164557,
+      "grad_norm": 0.4624659842635099,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.3285,
+      "step": 177
+    },
+    {
+      "epoch": 0.7210126582278481,
+      "grad_norm": 0.5440661616247529,
+      "learning_rate": 3.9756864721399456e-05,
+      "loss": 0.3311,
+      "step": 178
+    },
+    {
+      "epoch": 0.7250632911392405,
+      "grad_norm": 0.4847557245710079,
+      "learning_rate": 3.974796184811956e-05,
+      "loss": 0.3281,
+      "step": 179
+    },
+    {
+      "epoch": 0.7291139240506329,
+      "grad_norm": 0.557945074332379,
+      "learning_rate": 3.973889992746979e-05,
+      "loss": 0.3295,
+      "step": 180
+    },
+    {
+      "epoch": 0.7331645569620253,
+      "grad_norm": 0.5214941236410273,
+      "learning_rate": 3.972967903243361e-05,
+      "loss": 0.3415,
+      "step": 181
+    },
+    {
+      "epoch": 0.7372151898734177,
+      "grad_norm": 0.4450640289151871,
+      "learning_rate": 3.972029923727486e-05,
+      "loss": 0.3141,
+      "step": 182
+    },
+    {
+      "epoch": 0.7412658227848101,
+      "grad_norm": 0.4467521757682597,
+      "learning_rate": 3.971076061753709e-05,
+      "loss": 0.3397,
+      "step": 183
+    },
+    {
+      "epoch": 0.7453164556962025,
+      "grad_norm": 0.46520401088326674,
+      "learning_rate": 3.9701063250043066e-05,
+      "loss": 0.3177,
+      "step": 184
+    },
+    {
+      "epoch": 0.7493670886075949,
+      "grad_norm": 0.4809445448624303,
+      "learning_rate": 3.969120721289402e-05,
+      "loss": 0.3186,
+      "step": 185
+    },
+    {
+      "epoch": 0.7534177215189873,
+      "grad_norm": 0.461709875447719,
+      "learning_rate": 3.9681192585469146e-05,
+      "loss": 0.3571,
+      "step": 186
+    },
+    {
+      "epoch": 0.7574683544303797,
+      "grad_norm": 0.48711212756336286,
+      "learning_rate": 3.9671019448424865e-05,
+      "loss": 0.3159,
+      "step": 187
+    },
+    {
+      "epoch": 0.7615189873417721,
+      "grad_norm": 0.4799412103112035,
+      "learning_rate": 3.966068788369422e-05,
+      "loss": 0.3285,
+      "step": 188
+    },
+    {
+      "epoch": 0.7655696202531646,
+      "grad_norm": 0.4542101140680477,
+      "learning_rate": 3.965019797448622e-05,
+      "loss": 0.3253,
+      "step": 189
+    },
+    {
+      "epoch": 0.769620253164557,
+      "grad_norm": 0.4969223275655312,
+      "learning_rate": 3.963954980528515e-05,
+      "loss": 0.3325,
+      "step": 190
+    },
+    {
+      "epoch": 0.7736708860759494,
+      "grad_norm": 0.4913279350318146,
+      "learning_rate": 3.9628743461849905e-05,
+      "loss": 0.3303,
+      "step": 191
+    },
+    {
+      "epoch": 0.7777215189873418,
+      "grad_norm": 0.5276721212540879,
+      "learning_rate": 3.961777903121329e-05,
+      "loss": 0.3371,
+      "step": 192
+    },
+    {
+      "epoch": 0.7817721518987342,
+      "grad_norm": 0.49985987293176637,
+      "learning_rate": 3.960665660168131e-05,
+      "loss": 0.3483,
+      "step": 193
+    },
+    {
+      "epoch": 0.7858227848101266,
+      "grad_norm": 0.48199732872217704,
+      "learning_rate": 3.9595376262832485e-05,
+      "loss": 0.3414,
+      "step": 194
+    },
+    {
+      "epoch": 0.789873417721519,
+      "grad_norm": 0.5159617735392672,
+      "learning_rate": 3.9583938105517127e-05,
+      "loss": 0.3299,
+      "step": 195
+    },
+    {
+      "epoch": 0.7939240506329114,
+      "grad_norm": 0.6042136499412439,
+      "learning_rate": 3.957234222185657e-05,
+      "loss": 0.3058,
+      "step": 196
+    },
+    {
+      "epoch": 0.7979746835443038,
+      "grad_norm": 0.4473868351032443,
+      "learning_rate": 3.9560588705242474e-05,
+      "loss": 0.3387,
+      "step": 197
+    },
+    {
+      "epoch": 0.8020253164556962,
+      "grad_norm": 0.6368849344453956,
+      "learning_rate": 3.954867765033605e-05,
+      "loss": 0.3297,
+      "step": 198
+    },
+    {
+      "epoch": 0.8060759493670886,
+      "grad_norm": 0.6279002386080006,
+      "learning_rate": 3.953660915306728e-05,
+      "loss": 0.3381,
+      "step": 199
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.5014000843421582,
+      "learning_rate": 3.952438331063419e-05,
+      "loss": 0.3178,
+      "step": 200
+    },
+    {
+      "epoch": 0.8141772151898734,
+      "grad_norm": 0.5290991731397322,
+      "learning_rate": 3.951200022150205e-05,
+      "loss": 0.3314,
+      "step": 201
+    },
+    {
+      "epoch": 0.8182278481012658,
+      "grad_norm": 0.5004984037787601,
+      "learning_rate": 3.949945998540253e-05,
+      "loss": 0.3196,
+      "step": 202
+    },
+    {
+      "epoch": 0.8222784810126582,
+      "grad_norm": 0.5153869522452682,
+      "learning_rate": 3.9486762703332993e-05,
+      "loss": 0.3161,
+      "step": 203
+    },
+    {
+      "epoch": 0.8263291139240506,
+      "grad_norm": 0.4844528551618672,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.3254,
+      "step": 204
+    },
+    {
+      "epoch": 0.830379746835443,
+      "grad_norm": 0.4330016578259714,
+      "learning_rate": 3.946089741159648e-05,
+      "loss": 0.3221,
+      "step": 205
+    },
+    {
+      "epoch": 0.8344303797468354,
+      "grad_norm": 0.45689618239284946,
+      "learning_rate": 3.944772961024501e-05,
+      "loss": 0.3177,
+      "step": 206
+    },
+    {
+      "epoch": 0.8384810126582278,
+      "grad_norm": 0.3975999917347927,
+      "learning_rate": 3.943440517955285e-05,
+      "loss": 0.3264,
+      "step": 207
+    },
+    {
+      "epoch": 0.8425316455696202,
+      "grad_norm": 0.4542828743210102,
+      "learning_rate": 3.9420924226833126e-05,
+      "loss": 0.3227,
+      "step": 208
+    },
+    {
+      "epoch": 0.8465822784810126,
+      "grad_norm": 0.41126256598210037,
+      "learning_rate": 3.9407286860659566e-05,
+      "loss": 0.3236,
+      "step": 209
+    },
+    {
+      "epoch": 0.850632911392405,
+      "grad_norm": 0.4850321747458276,
+      "learning_rate": 3.9393493190865657e-05,
+      "loss": 0.3303,
+      "step": 210
+    },
+    {
+      "epoch": 0.8546835443037974,
+      "grad_norm": 0.4302186670664923,
+      "learning_rate": 3.937954332854371e-05,
+      "loss": 0.3295,
+      "step": 211
+    },
+    {
+      "epoch": 0.8587341772151899,
+      "grad_norm": 0.4392147591332327,
+      "learning_rate": 3.9365437386044016e-05,
+      "loss": 0.3151,
+      "step": 212
+    },
+    {
+      "epoch": 0.8627848101265823,
+      "grad_norm": 0.46396742033727406,
+      "learning_rate": 3.935117547697387e-05,
+      "loss": 0.3356,
+      "step": 213
+    },
+    {
+      "epoch": 0.8668354430379747,
+      "grad_norm": 0.4469421550314178,
+      "learning_rate": 3.933675771619675e-05,
+      "loss": 0.3199,
+      "step": 214
+    },
+    {
+      "epoch": 0.8708860759493671,
+      "grad_norm": 0.49544283547472595,
+      "learning_rate": 3.932218421983131e-05,
+      "loss": 0.3249,
+      "step": 215
+    },
+    {
+      "epoch": 0.8749367088607595,
+      "grad_norm": 0.44249651099734577,
+      "learning_rate": 3.9307455105250484e-05,
+      "loss": 0.3274,
+      "step": 216
+    },
+    {
+      "epoch": 0.8789873417721519,
+      "grad_norm": 0.49099556501795777,
+      "learning_rate": 3.929257049108054e-05,
+      "loss": 0.316,
+      "step": 217
+    },
+    {
+      "epoch": 0.8830379746835443,
+      "grad_norm": 0.3919055897422822,
+      "learning_rate": 3.927753049720011e-05,
+      "loss": 0.3323,
+      "step": 218
+    },
+    {
+      "epoch": 0.8870886075949367,
+      "grad_norm": 0.44144615831803713,
+      "learning_rate": 3.9262335244739234e-05,
+      "loss": 0.3378,
+      "step": 219
+    },
+    {
+      "epoch": 0.8911392405063291,
+      "grad_norm": 0.42851713793832064,
+      "learning_rate": 3.92469848560784e-05,
+      "loss": 0.321,
+      "step": 220
+    },
+    {
+      "epoch": 0.8951898734177215,
+      "grad_norm": 0.42418990508889104,
+      "learning_rate": 3.923147945484751e-05,
+      "loss": 0.3253,
+      "step": 221
+    },
+    {
+      "epoch": 0.8992405063291139,
+      "grad_norm": 0.41719357796552464,
+      "learning_rate": 3.9215819165924956e-05,
+      "loss": 0.3142,
+      "step": 222
+    },
+    {
+      "epoch": 0.9032911392405063,
+      "grad_norm": 0.44065460944909335,
+      "learning_rate": 3.920000411543654e-05,
+      "loss": 0.3186,
+      "step": 223
+    },
+    {
+      "epoch": 0.9073417721518987,
+      "grad_norm": 0.49624809044687074,
+      "learning_rate": 3.9184034430754495e-05,
+      "loss": 0.3183,
+      "step": 224
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.49633453301313496,
+      "learning_rate": 3.916791024049648e-05,
+      "loss": 0.3227,
+      "step": 225
+    },
+    {
+      "epoch": 0.9154430379746835,
+      "grad_norm": 0.46958358592192717,
+      "learning_rate": 3.91516316745245e-05,
+      "loss": 0.3174,
+      "step": 226
+    },
+    {
+      "epoch": 0.9194936708860759,
+      "grad_norm": 0.54622454445768,
+      "learning_rate": 3.913519886394389e-05,
+      "loss": 0.3341,
+      "step": 227
+    },
+    {
+      "epoch": 0.9235443037974683,
+      "grad_norm": 0.4545749743004333,
+      "learning_rate": 3.911861194110225e-05,
+      "loss": 0.329,
+      "step": 228
+    },
+    {
+      "epoch": 0.9275949367088607,
+      "grad_norm": 0.49365065351482296,
+      "learning_rate": 3.910187103958837e-05,
+      "loss": 0.3161,
+      "step": 229
+    },
+    {
+      "epoch": 0.9316455696202531,
+      "grad_norm": 0.40525387429736165,
+      "learning_rate": 3.908497629423117e-05,
+      "loss": 0.318,
+      "step": 230
+    },
+    {
+      "epoch": 0.9356962025316455,
+      "grad_norm": 0.40269595742994646,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.3344,
+      "step": 231
+    },
+    {
+      "epoch": 0.9397468354430379,
+      "grad_norm": 0.3973965612400986,
+      "learning_rate": 3.9050725817496594e-05,
+      "loss": 0.3098,
+      "step": 232
+    },
+    {
+      "epoch": 0.9437974683544303,
+      "grad_norm": 0.4193327815776918,
+      "learning_rate": 3.9033370361967844e-05,
+      "loss": 0.3159,
+      "step": 233
+    },
+    {
+      "epoch": 0.9478481012658228,
+      "grad_norm": 0.4086124025416428,
+      "learning_rate": 3.901586161429081e-05,
+      "loss": 0.3275,
+      "step": 234
+    },
+    {
+      "epoch": 0.9518987341772152,
+      "grad_norm": 0.43545197592005735,
+      "learning_rate": 3.8998199715478545e-05,
+      "loss": 0.3265,
+      "step": 235
+    },
+    {
+      "epoch": 0.9559493670886076,
+      "grad_norm": 0.4193770021339919,
+      "learning_rate": 3.8980384807777564e-05,
+      "loss": 0.3162,
+      "step": 236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4061663220846516,
+      "learning_rate": 3.896241703466667e-05,
+      "loss": 0.3045,
+      "step": 237
+    },
+    {
+      "epoch": 0.9640506329113924,
+      "grad_norm": 0.45858103541458706,
+      "learning_rate": 3.894429654085585e-05,
+      "loss": 0.3154,
+      "step": 238
+    },
+    {
+      "epoch": 0.9681012658227848,
+      "grad_norm": 0.43237318515930534,
+      "learning_rate": 3.892602347228505e-05,
+      "loss": 0.3256,
+      "step": 239
+    },
+    {
+      "epoch": 0.9721518987341772,
+      "grad_norm": 0.39257687313111916,
+      "learning_rate": 3.890759797612307e-05,
+      "loss": 0.3103,
+      "step": 240
+    },
+    {
+      "epoch": 0.9762025316455696,
+      "grad_norm": 0.45955667294125796,
+      "learning_rate": 3.888902020076632e-05,
+      "loss": 0.3275,
+      "step": 241
+    },
+    {
+      "epoch": 0.980253164556962,
+      "grad_norm": 0.4042917437357599,
+      "learning_rate": 3.887029029583764e-05,
+      "loss": 0.318,
+      "step": 242
+    },
+    {
+      "epoch": 0.9843037974683544,
+      "grad_norm": 0.4562087639081624,
+      "learning_rate": 3.8851408412185125e-05,
+      "loss": 0.3157,
+      "step": 243
+    },
+    {
+      "epoch": 0.9883544303797468,
+      "grad_norm": 0.42741694961909676,
+      "learning_rate": 3.8832374701880855e-05,
+      "loss": 0.3197,
+      "step": 244
+    },
+    {
+      "epoch": 0.9924050632911392,
+      "grad_norm": 0.4334573902411466,
+      "learning_rate": 3.881318931821972e-05,
+      "loss": 0.3277,
+      "step": 245
+    },
+    {
+      "epoch": 0.9964556962025316,
+      "grad_norm": 0.48489869357738075,
+      "learning_rate": 3.879385241571817e-05,
+      "loss": 0.3355,
+      "step": 246
+    },
+    {
+      "epoch": 1.0020253164556963,
+      "grad_norm": 0.46617882908524233,
+      "learning_rate": 3.8774364150112955e-05,
+      "loss": 0.3078,
+      "step": 247
+    },
+    {
+      "epoch": 1.0060759493670886,
+      "grad_norm": 0.39785348245426805,
+      "learning_rate": 3.8754724678359884e-05,
+      "loss": 0.2627,
+      "step": 248
+    },
+    {
+      "epoch": 1.010126582278481,
+      "grad_norm": 0.5600374106757209,
+      "learning_rate": 3.873493415863256e-05,
+      "loss": 0.2779,
+      "step": 249
+    },
+    {
+      "epoch": 1.0141772151898734,
+      "grad_norm": 0.47804746471206877,
+      "learning_rate": 3.871499275032111e-05,
+      "loss": 0.2714,
+      "step": 250
+    },
+    {
+      "epoch": 1.0182278481012659,
+      "grad_norm": 0.42968695779041827,
+      "learning_rate": 3.869490061403091e-05,
+      "loss": 0.2631,
+      "step": 251
+    },
+    {
+      "epoch": 1.0222784810126582,
+      "grad_norm": 0.4309592985419051,
+      "learning_rate": 3.867465791158124e-05,
+      "loss": 0.2676,
+      "step": 252
+    },
+    {
+      "epoch": 1.0263291139240507,
+      "grad_norm": 0.48113031925222693,
+      "learning_rate": 3.865426480600407e-05,
+      "loss": 0.2688,
+      "step": 253
+    },
+    {
+      "epoch": 1.030379746835443,
+      "grad_norm": 0.42111569459161213,
+      "learning_rate": 3.863372146154264e-05,
+      "loss": 0.2765,
+      "step": 254
+    },
+    {
+      "epoch": 1.0344303797468355,
+      "grad_norm": 0.3683798657550155,
+      "learning_rate": 3.861302804365024e-05,
+      "loss": 0.2536,
+      "step": 255
+    },
+    {
+      "epoch": 1.0384810126582278,
+      "grad_norm": 0.3959787313441287,
+      "learning_rate": 3.85921847189888e-05,
+      "loss": 0.2578,
+      "step": 256
+    },
+    {
+      "epoch": 1.0425316455696203,
+      "grad_norm": 0.38784350829426345,
+      "learning_rate": 3.85711916554276e-05,
+      "loss": 0.2685,
+      "step": 257
+    },
+    {
+      "epoch": 1.0465822784810126,
+      "grad_norm": 0.4354298559310261,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.2637,
+      "step": 258
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.41166407556105095,
+      "learning_rate": 3.852875698911154e-05,
+      "loss": 0.2662,
+      "step": 259
+    },
+    {
+      "epoch": 1.0546835443037974,
+      "grad_norm": 0.4094678869376007,
+      "learning_rate": 3.850731572811963e-05,
+      "loss": 0.283,
+      "step": 260
+    },
+    {
+      "epoch": 1.05873417721519,
+      "grad_norm": 0.40171046539378696,
+      "learning_rate": 3.848572541175116e-05,
+      "loss": 0.2823,
+      "step": 261
+    },
+    {
+      "epoch": 1.0627848101265822,
+      "grad_norm": 0.3946008094105798,
+      "learning_rate": 3.846398621389154e-05,
+      "loss": 0.2704,
+      "step": 262
+    },
+    {
+      "epoch": 1.0668354430379747,
+      "grad_norm": 0.4228527106196662,
+      "learning_rate": 3.84420983096253e-05,
+      "loss": 0.2683,
+      "step": 263
+    },
+    {
+      "epoch": 1.070886075949367,
+      "grad_norm": 0.4401732730857555,
+      "learning_rate": 3.8420061875234606e-05,
+      "loss": 0.2732,
+      "step": 264
+    },
+    {
+      "epoch": 1.0749367088607595,
+      "grad_norm": 0.40559904715791234,
+      "learning_rate": 3.839787708819787e-05,
+      "loss": 0.2641,
+      "step": 265
+    },
+    {
+      "epoch": 1.0789873417721518,
+      "grad_norm": 0.43824852392402075,
+      "learning_rate": 3.8375544127188325e-05,
+      "loss": 0.2655,
+      "step": 266
+    },
+    {
+      "epoch": 1.0830379746835443,
+      "grad_norm": 0.3823081213010939,
+      "learning_rate": 3.8353063172072564e-05,
+      "loss": 0.2831,
+      "step": 267
+    },
+    {
+      "epoch": 1.0870886075949366,
+      "grad_norm": 0.38839419694726374,
+      "learning_rate": 3.8330434403909105e-05,
+      "loss": 0.2663,
+      "step": 268
+    },
+    {
+      "epoch": 1.0911392405063292,
+      "grad_norm": 0.3864863747366155,
+      "learning_rate": 3.8307658004946934e-05,
+      "loss": 0.2615,
+      "step": 269
+    },
+    {
+      "epoch": 1.0951898734177214,
+      "grad_norm": 0.43484191372079933,
+      "learning_rate": 3.8284734158624046e-05,
+      "loss": 0.2752,
+      "step": 270
+    },
+    {
+      "epoch": 1.099240506329114,
+      "grad_norm": 0.402487377699847,
+      "learning_rate": 3.826166304956594e-05,
+      "loss": 0.2621,
+      "step": 271
+    },
+    {
+      "epoch": 1.1032911392405063,
+      "grad_norm": 0.3866685552372165,
+      "learning_rate": 3.8238444863584164e-05,
+      "loss": 0.2726,
+      "step": 272
+    },
+    {
+      "epoch": 1.1073417721518988,
+      "grad_norm": 0.37543905728571625,
+      "learning_rate": 3.821507978767479e-05,
+      "loss": 0.2694,
+      "step": 273
+    },
+    {
+      "epoch": 1.111392405063291,
+      "grad_norm": 0.4148643339351974,
+      "learning_rate": 3.819156801001693e-05,
+      "loss": 0.259,
+      "step": 274
+    },
+    {
+      "epoch": 1.1154430379746836,
+      "grad_norm": 0.41491104032202436,
+      "learning_rate": 3.816790971997121e-05,
+      "loss": 0.2758,
+      "step": 275
+    },
+    {
+      "epoch": 1.1194936708860759,
+      "grad_norm": 0.4261971090493251,
+      "learning_rate": 3.8144105108078246e-05,
+      "loss": 0.2655,
+      "step": 276
+    },
+    {
+      "epoch": 1.1235443037974684,
+      "grad_norm": 0.40772967765905377,
+      "learning_rate": 3.81201543660571e-05,
+      "loss": 0.2615,
+      "step": 277
+    },
+    {
+      "epoch": 1.1275949367088607,
+      "grad_norm": 0.4062729790650647,
+      "learning_rate": 3.809605768680377e-05,
+      "loss": 0.264,
+      "step": 278
+    },
+    {
+      "epoch": 1.1316455696202532,
+      "grad_norm": 0.39028890595153015,
+      "learning_rate": 3.807181526438958e-05,
+      "loss": 0.2725,
+      "step": 279
+    },
+    {
+      "epoch": 1.1356962025316455,
+      "grad_norm": 0.4256580489694506,
+      "learning_rate": 3.8047427294059697e-05,
+      "loss": 0.2705,
+      "step": 280
+    },
+    {
+      "epoch": 1.139746835443038,
+      "grad_norm": 0.38583123805859454,
+      "learning_rate": 3.802289397223145e-05,
+      "loss": 0.2584,
+      "step": 281
+    },
+    {
+      "epoch": 1.1437974683544303,
+      "grad_norm": 0.38432872985219596,
+      "learning_rate": 3.7998215496492854e-05,
+      "loss": 0.2748,
+      "step": 282
+    },
+    {
+      "epoch": 1.1478481012658228,
+      "grad_norm": 0.3915143036447229,
+      "learning_rate": 3.797339206560096e-05,
+      "loss": 0.2511,
+      "step": 283
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.40548070758539423,
+      "learning_rate": 3.794842387948027e-05,
+      "loss": 0.2614,
+      "step": 284
+    },
+    {
+      "epoch": 1.1559493670886076,
+      "grad_norm": 0.40819404267609594,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.282,
+      "step": 285
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3753120493046467,
+      "learning_rate": 3.7898054047078054e-05,
+      "loss": 0.2798,
+      "step": 286
+    },
+    {
+      "epoch": 1.1640506329113924,
+      "grad_norm": 0.3963404729523744,
+      "learning_rate": 3.787265280646825e-05,
+      "loss": 0.2799,
+      "step": 287
+    },
+    {
+      "epoch": 1.1681012658227847,
+      "grad_norm": 0.4065711788938437,
+      "learning_rate": 3.7847107621969786e-05,
+      "loss": 0.2649,
+      "step": 288
+    },
+    {
+      "epoch": 1.1721518987341772,
+      "grad_norm": 0.3777282283635587,
+      "learning_rate": 3.7821418699320064e-05,
+      "loss": 0.283,
+      "step": 289
+    },
+    {
+      "epoch": 1.1762025316455695,
+      "grad_norm": 0.41340439318477046,
+      "learning_rate": 3.7795586245414145e-05,
+      "loss": 0.2701,
+      "step": 290
+    },
+    {
+      "epoch": 1.180253164556962,
+      "grad_norm": 0.39227166027540544,
+      "learning_rate": 3.776961046830306e-05,
+      "loss": 0.2658,
+      "step": 291
+    },
+    {
+      "epoch": 1.1843037974683543,
+      "grad_norm": 0.40054640342578135,
+      "learning_rate": 3.774349157719215e-05,
+      "loss": 0.2683,
+      "step": 292
+    },
+    {
+      "epoch": 1.1883544303797469,
+      "grad_norm": 0.4877059162096422,
+      "learning_rate": 3.7717229782439365e-05,
+      "loss": 0.2683,
+      "step": 293
+    },
+    {
+      "epoch": 1.1924050632911392,
+      "grad_norm": 0.39003937130953686,
+      "learning_rate": 3.769082529555359e-05,
+      "loss": 0.2628,
+      "step": 294
+    },
+    {
+      "epoch": 1.1964556962025317,
+      "grad_norm": 0.4565081523746762,
+      "learning_rate": 3.766427832919294e-05,
+      "loss": 0.2788,
+      "step": 295
+    },
+    {
+      "epoch": 1.200506329113924,
+      "grad_norm": 0.44969578305724367,
+      "learning_rate": 3.7637589097163024e-05,
+      "loss": 0.2809,
+      "step": 296
+    },
+    {
+      "epoch": 1.2045569620253165,
+      "grad_norm": 0.4089905456656113,
+      "learning_rate": 3.761075781441526e-05,
+      "loss": 0.2675,
+      "step": 297
+    },
+    {
+      "epoch": 1.2086075949367088,
+      "grad_norm": 0.46299126986160044,
+      "learning_rate": 3.75837846970451e-05,
+      "loss": 0.2706,
+      "step": 298
+    },
+    {
+      "epoch": 1.2126582278481013,
+      "grad_norm": 0.4228358989749569,
+      "learning_rate": 3.755666996229032e-05,
+      "loss": 0.2746,
+      "step": 299
+    },
+    {
+      "epoch": 1.2167088607594936,
+      "grad_norm": 0.4084813528360387,
+      "learning_rate": 3.752941382852927e-05,
+      "loss": 0.2762,
+      "step": 300
+    },
+    {
+      "epoch": 1.220759493670886,
+      "grad_norm": 0.46110403748396367,
+      "learning_rate": 3.7502016515279115e-05,
+      "loss": 0.2694,
+      "step": 301
+    },
+    {
+      "epoch": 1.2248101265822784,
+      "grad_norm": 0.3743856865427884,
+      "learning_rate": 3.7474478243194043e-05,
+      "loss": 0.2742,
+      "step": 302
+    },
+    {
+      "epoch": 1.228860759493671,
+      "grad_norm": 0.4256262995255064,
+      "learning_rate": 3.744679923406351e-05,
+      "loss": 0.2575,
+      "step": 303
+    },
+    {
+      "epoch": 1.2329113924050632,
+      "grad_norm": 0.3391715107952419,
+      "learning_rate": 3.741897971081043e-05,
+      "loss": 0.2688,
+      "step": 304
+    },
+    {
+      "epoch": 1.2369620253164557,
+      "grad_norm": 0.4071001268860348,
+      "learning_rate": 3.739101989748946e-05,
+      "loss": 0.2673,
+      "step": 305
+    },
+    {
+      "epoch": 1.241012658227848,
+      "grad_norm": 0.4085732288349824,
+      "learning_rate": 3.7362920019285066e-05,
+      "loss": 0.2688,
+      "step": 306
+    },
+    {
+      "epoch": 1.2450632911392405,
+      "grad_norm": 0.38066282474191376,
+      "learning_rate": 3.73346803025098e-05,
+      "loss": 0.2674,
+      "step": 307
+    },
+    {
+      "epoch": 1.2491139240506328,
+      "grad_norm": 0.3862980409040882,
+      "learning_rate": 3.730630097460247e-05,
+      "loss": 0.2783,
+      "step": 308
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.3832947654106215,
+      "learning_rate": 3.727778226412628e-05,
+      "loss": 0.2768,
+      "step": 309
+    },
+    {
+      "epoch": 1.2572151898734178,
+      "grad_norm": 0.3811054746925965,
+      "learning_rate": 3.7249124400767006e-05,
+      "loss": 0.2851,
+      "step": 310
+    },
+    {
+      "epoch": 1.2612658227848101,
+      "grad_norm": 0.37251865623193187,
+      "learning_rate": 3.722032761533114e-05,
+      "loss": 0.2685,
+      "step": 311
+    },
+    {
+      "epoch": 1.2653164556962024,
+      "grad_norm": 0.4285890974567403,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.2638,
+      "step": 312
+    },
+    {
+      "epoch": 1.269367088607595,
+      "grad_norm": 0.3733642726878229,
+      "learning_rate": 3.7162318207048006e-05,
+      "loss": 0.2736,
+      "step": 313
+    },
+    {
+      "epoch": 1.2734177215189875,
+      "grad_norm": 0.4469325270749346,
+      "learning_rate": 3.713310605140055e-05,
+      "loss": 0.2789,
+      "step": 314
+    },
+    {
+      "epoch": 1.2774683544303798,
+      "grad_norm": 0.38657717461728064,
+      "learning_rate": 3.710375590807233e-05,
+      "loss": 0.2672,
+      "step": 315
+    },
+    {
+      "epoch": 1.281518987341772,
+      "grad_norm": 0.40764238329466485,
+      "learning_rate": 3.7074268013445365e-05,
+      "loss": 0.2791,
+      "step": 316
+    },
+    {
+      "epoch": 1.2855696202531646,
+      "grad_norm": 0.40645954057500855,
+      "learning_rate": 3.7044642605011114e-05,
+      "loss": 0.2784,
+      "step": 317
+    },
+    {
+      "epoch": 1.289620253164557,
+      "grad_norm": 0.407342599375613,
+      "learning_rate": 3.701487992136854e-05,
+      "loss": 0.2601,
+      "step": 318
+    },
+    {
+      "epoch": 1.2936708860759494,
+      "grad_norm": 0.4141079340000889,
+      "learning_rate": 3.69849802022222e-05,
+      "loss": 0.2625,
+      "step": 319
+    },
+    {
+      "epoch": 1.2977215189873417,
+      "grad_norm": 0.37033843924124266,
+      "learning_rate": 3.6954943688380334e-05,
+      "loss": 0.2695,
+      "step": 320
+    },
+    {
+      "epoch": 1.3017721518987342,
+      "grad_norm": 0.41014736482772113,
+      "learning_rate": 3.692477062175289e-05,
+      "loss": 0.2653,
+      "step": 321
+    },
+    {
+      "epoch": 1.3058227848101267,
+      "grad_norm": 0.3978252942391586,
+      "learning_rate": 3.689446124534958e-05,
+      "loss": 0.2669,
+      "step": 322
+    },
+    {
+      "epoch": 1.309873417721519,
+      "grad_norm": 0.37824861937168835,
+      "learning_rate": 3.686401580327799e-05,
+      "loss": 0.2644,
+      "step": 323
+    },
+    {
+      "epoch": 1.3139240506329113,
+      "grad_norm": 0.44812101009289307,
+      "learning_rate": 3.683343454074149e-05,
+      "loss": 0.2733,
+      "step": 324
+    },
+    {
+      "epoch": 1.3179746835443038,
+      "grad_norm": 0.39767929187411294,
+      "learning_rate": 3.6802717704037386e-05,
+      "loss": 0.2513,
+      "step": 325
+    },
+    {
+      "epoch": 1.3220253164556963,
+      "grad_norm": 0.4869594478056333,
+      "learning_rate": 3.6771865540554855e-05,
+      "loss": 0.2679,
+      "step": 326
+    },
+    {
+      "epoch": 1.3260759493670886,
+      "grad_norm": 0.34351966236337333,
+      "learning_rate": 3.674087829877297e-05,
+      "loss": 0.2694,
+      "step": 327
+    },
+    {
+      "epoch": 1.330126582278481,
+      "grad_norm": 0.4132073807051578,
+      "learning_rate": 3.6709756228258735e-05,
+      "loss": 0.2738,
+      "step": 328
+    },
+    {
+      "epoch": 1.3341772151898734,
+      "grad_norm": 0.39742303231763215,
+      "learning_rate": 3.667849957966501e-05,
+      "loss": 0.2676,
+      "step": 329
+    },
+    {
+      "epoch": 1.338227848101266,
+      "grad_norm": 0.4458089919004806,
+      "learning_rate": 3.6647108604728546e-05,
+      "loss": 0.2517,
+      "step": 330
+    },
+    {
+      "epoch": 1.3422784810126582,
+      "grad_norm": 0.4519113469029214,
+      "learning_rate": 3.661558355626795e-05,
+      "loss": 0.2727,
+      "step": 331
+    },
+    {
+      "epoch": 1.3463291139240505,
+      "grad_norm": 0.48223891852301476,
+      "learning_rate": 3.658392468818163e-05,
+      "loss": 0.2851,
+      "step": 332
+    },
+    {
+      "epoch": 1.350379746835443,
+      "grad_norm": 0.4506159911574278,
+      "learning_rate": 3.655213225544574e-05,
+      "loss": 0.2612,
+      "step": 333
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 0.5002192414067483,
+      "learning_rate": 3.652020651411218e-05,
+      "loss": 0.2638,
+      "step": 334
+    },
+    {
+      "epoch": 1.3584810126582278,
+      "grad_norm": 0.38679880524691984,
+      "learning_rate": 3.6488147721306474e-05,
+      "loss": 0.2776,
+      "step": 335
+    },
+    {
+      "epoch": 1.3625316455696201,
+      "grad_norm": 0.5815240283879434,
+      "learning_rate": 3.645595613522574e-05,
+      "loss": 0.2679,
+      "step": 336
+    },
+    {
+      "epoch": 1.3665822784810127,
+      "grad_norm": 0.3924626244715886,
+      "learning_rate": 3.642363201513657e-05,
+      "loss": 0.2795,
+      "step": 337
+    },
+    {
+      "epoch": 1.3706329113924052,
+      "grad_norm": 0.5330881638494749,
+      "learning_rate": 3.6391175621373006e-05,
+      "loss": 0.2666,
+      "step": 338
+    },
+    {
+      "epoch": 1.3746835443037975,
+      "grad_norm": 0.3729574396648099,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.256,
+      "step": 339
+    },
+    {
+      "epoch": 1.3787341772151898,
+      "grad_norm": 0.4578182737898028,
+      "learning_rate": 3.632586705948318e-05,
+      "loss": 0.2718,
+      "step": 340
+    },
+    {
+      "epoch": 1.3827848101265823,
+      "grad_norm": 0.4255267740064292,
+      "learning_rate": 3.629301541734311e-05,
+      "loss": 0.2768,
+      "step": 341
+    },
+    {
+      "epoch": 1.3868354430379748,
+      "grad_norm": 0.45303179212560934,
+      "learning_rate": 3.626003255349676e-05,
+      "loss": 0.2771,
+      "step": 342
+    },
+    {
+      "epoch": 1.390886075949367,
+      "grad_norm": 0.4216847632973556,
+      "learning_rate": 3.622691873358357e-05,
+      "loss": 0.2719,
+      "step": 343
+    },
+    {
+      "epoch": 1.3949367088607594,
+      "grad_norm": 0.4783631596274916,
+      "learning_rate": 3.61936742242977e-05,
+      "loss": 0.2762,
+      "step": 344
+    },
+    {
+      "epoch": 1.398987341772152,
+      "grad_norm": 0.3809328359780469,
+      "learning_rate": 3.6160299293385864e-05,
+      "loss": 0.2617,
+      "step": 345
+    },
+    {
+      "epoch": 1.4030379746835444,
+      "grad_norm": 0.4381904891755294,
+      "learning_rate": 3.612679420964516e-05,
+      "loss": 0.2758,
+      "step": 346
+    },
+    {
+      "epoch": 1.4070886075949367,
+      "grad_norm": 0.43769483797730857,
+      "learning_rate": 3.609315924292092e-05,
+      "loss": 0.2845,
+      "step": 347
+    },
+    {
+      "epoch": 1.411139240506329,
+      "grad_norm": 0.37681591886761584,
+      "learning_rate": 3.6059394664104554e-05,
+      "loss": 0.2716,
+      "step": 348
+    },
+    {
+      "epoch": 1.4151898734177215,
+      "grad_norm": 0.4181102651465381,
+      "learning_rate": 3.602550074513133e-05,
+      "loss": 0.2873,
+      "step": 349
+    },
+    {
+      "epoch": 1.419240506329114,
+      "grad_norm": 0.3601062949949995,
+      "learning_rate": 3.599147775897822e-05,
+      "loss": 0.2817,
+      "step": 350
+    },
+    {
+      "epoch": 1.4232911392405063,
+      "grad_norm": 0.39682968360779863,
+      "learning_rate": 3.595732597966167e-05,
+      "loss": 0.2704,
+      "step": 351
+    },
+    {
+      "epoch": 1.4273417721518986,
+      "grad_norm": 0.37697636789641753,
+      "learning_rate": 3.592304568223542e-05,
+      "loss": 0.2732,
+      "step": 352
+    },
+    {
+      "epoch": 1.4313924050632911,
+      "grad_norm": 0.4304989777597985,
+      "learning_rate": 3.588863714278826e-05,
+      "loss": 0.2529,
+      "step": 353
+    },
+    {
+      "epoch": 1.4354430379746836,
+      "grad_norm": 0.425487346297316,
+      "learning_rate": 3.585410063844186e-05,
+      "loss": 0.2591,
+      "step": 354
+    },
+    {
+      "epoch": 1.439493670886076,
+      "grad_norm": 0.40754171905617403,
+      "learning_rate": 3.581943644734846e-05,
+      "loss": 0.2741,
+      "step": 355
+    },
+    {
+      "epoch": 1.4435443037974682,
+      "grad_norm": 0.4424851007079983,
+      "learning_rate": 3.578464484868869e-05,
+      "loss": 0.2696,
+      "step": 356
+    },
+    {
+      "epoch": 1.4475949367088607,
+      "grad_norm": 0.407776475723211,
+      "learning_rate": 3.5749726122669316e-05,
+      "loss": 0.2649,
+      "step": 357
+    },
+    {
+      "epoch": 1.4516455696202533,
+      "grad_norm": 0.3525480933737716,
+      "learning_rate": 3.5714680550520943e-05,
+      "loss": 0.2756,
+      "step": 358
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.41029703128098965,
+      "learning_rate": 3.5679508414495794e-05,
+      "loss": 0.2752,
+      "step": 359
+    },
+    {
+      "epoch": 1.4597468354430378,
+      "grad_norm": 0.36652947222728066,
+      "learning_rate": 3.564420999786543e-05,
+      "loss": 0.2714,
+      "step": 360
+    },
+    {
+      "epoch": 1.4637974683544304,
+      "grad_norm": 0.46525243279214623,
+      "learning_rate": 3.560878558491842e-05,
+      "loss": 0.2807,
+      "step": 361
+    },
+    {
+      "epoch": 1.4678481012658229,
+      "grad_norm": 0.3656903347845267,
+      "learning_rate": 3.5573235460958145e-05,
+      "loss": 0.2779,
+      "step": 362
+    },
+    {
+      "epoch": 1.4718987341772152,
+      "grad_norm": 0.4278605020421252,
+      "learning_rate": 3.553755991230039e-05,
+      "loss": 0.2751,
+      "step": 363
+    },
+    {
+      "epoch": 1.4759493670886075,
+      "grad_norm": 0.4845672157469642,
+      "learning_rate": 3.5501759226271144e-05,
+      "loss": 0.2766,
+      "step": 364
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3575688560349677,
+      "learning_rate": 3.546583369120419e-05,
+      "loss": 0.2592,
+      "step": 365
+    },
+    {
+      "epoch": 1.4840506329113925,
+      "grad_norm": 0.4303934478049915,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.277,
+      "step": 366
+    },
+    {
+      "epoch": 1.4881012658227848,
+      "grad_norm": 0.40057995388519896,
+      "learning_rate": 3.539360923231766e-05,
+      "loss": 0.2745,
+      "step": 367
+    },
+    {
+      "epoch": 1.492151898734177,
+      "grad_norm": 0.4138796547385809,
+      "learning_rate": 3.535731089018394e-05,
+      "loss": 0.271,
+      "step": 368
+    },
+    {
+      "epoch": 1.4962025316455696,
+      "grad_norm": 0.40169726836925174,
+      "learning_rate": 3.532088886237956e-05,
+      "loss": 0.27,
+      "step": 369
+    },
+    {
+      "epoch": 1.5002531645569621,
+      "grad_norm": 0.4506603352661273,
+      "learning_rate": 3.528434344224253e-05,
+      "loss": 0.2771,
+      "step": 370
+    },
+    {
+      "epoch": 1.5043037974683544,
+      "grad_norm": 0.432125782771888,
+      "learning_rate": 3.524767492410464e-05,
+      "loss": 0.2884,
+      "step": 371
+    },
+    {
+      "epoch": 1.5083544303797467,
+      "grad_norm": 0.4235632432424522,
+      "learning_rate": 3.521088360328908e-05,
+      "loss": 0.2763,
+      "step": 372
+    },
+    {
+      "epoch": 1.5124050632911392,
+      "grad_norm": 0.4167388963658851,
+      "learning_rate": 3.517396977610811e-05,
+      "loss": 0.258,
+      "step": 373
+    },
+    {
+      "epoch": 1.5164556962025317,
+      "grad_norm": 0.43412967086750814,
+      "learning_rate": 3.5136933739860595e-05,
+      "loss": 0.2664,
+      "step": 374
+    },
+    {
+      "epoch": 1.520506329113924,
+      "grad_norm": 0.4114461653200382,
+      "learning_rate": 3.509977579282971e-05,
+      "loss": 0.267,
+      "step": 375
+    },
+    {
+      "epoch": 1.5245569620253163,
+      "grad_norm": 0.43095098312718466,
+      "learning_rate": 3.5062496234280424e-05,
+      "loss": 0.268,
+      "step": 376
+    },
+    {
+      "epoch": 1.5286075949367088,
+      "grad_norm": 0.39515148200530337,
+      "learning_rate": 3.502509536445719e-05,
+      "loss": 0.2772,
+      "step": 377
+    },
+    {
+      "epoch": 1.5326582278481014,
+      "grad_norm": 0.40251742091715464,
+      "learning_rate": 3.498757348458147e-05,
+      "loss": 0.2686,
+      "step": 378
+    },
+    {
+      "epoch": 1.5367088607594936,
+      "grad_norm": 0.4564182888608486,
+      "learning_rate": 3.4949930896849324e-05,
+      "loss": 0.2792,
+      "step": 379
+    },
+    {
+      "epoch": 1.540759493670886,
+      "grad_norm": 0.38750369660217693,
+      "learning_rate": 3.491216790442899e-05,
+      "loss": 0.2818,
+      "step": 380
+    },
+    {
+      "epoch": 1.5448101265822785,
+      "grad_norm": 0.433588295507515,
+      "learning_rate": 3.487428481145839e-05,
+      "loss": 0.2732,
+      "step": 381
+    },
+    {
+      "epoch": 1.548860759493671,
+      "grad_norm": 0.38735107945276825,
+      "learning_rate": 3.483628192304278e-05,
+      "loss": 0.2729,
+      "step": 382
+    },
+    {
+      "epoch": 1.5529113924050633,
+      "grad_norm": 0.44626201502844093,
+      "learning_rate": 3.479815954525219e-05,
+      "loss": 0.2915,
+      "step": 383
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.4639129537719396,
+      "learning_rate": 3.475991798511899e-05,
+      "loss": 0.2817,
+      "step": 384
+    },
+    {
+      "epoch": 1.561012658227848,
+      "grad_norm": 0.4284583587024216,
+      "learning_rate": 3.4721557550635464e-05,
+      "loss": 0.2806,
+      "step": 385
+    },
+    {
+      "epoch": 1.5650632911392406,
+      "grad_norm": 0.4645363248768687,
+      "learning_rate": 3.468307855075128e-05,
+      "loss": 0.2762,
+      "step": 386
+    },
+    {
+      "epoch": 1.5691139240506329,
+      "grad_norm": 0.34373012338157843,
+      "learning_rate": 3.4644481295371005e-05,
+      "loss": 0.2664,
+      "step": 387
+    },
+    {
+      "epoch": 1.5731645569620252,
+      "grad_norm": 0.3981298278152739,
+      "learning_rate": 3.460576609535163e-05,
+      "loss": 0.2785,
+      "step": 388
+    },
+    {
+      "epoch": 1.5772151898734177,
+      "grad_norm": 0.3578996770401982,
+      "learning_rate": 3.456693326250006e-05,
+      "loss": 0.2687,
+      "step": 389
+    },
+    {
+      "epoch": 1.5812658227848102,
+      "grad_norm": 0.3795912390793899,
+      "learning_rate": 3.452798310957058e-05,
+      "loss": 0.2709,
+      "step": 390
+    },
+    {
+      "epoch": 1.5853164556962025,
+      "grad_norm": 0.3749911839238573,
+      "learning_rate": 3.4488915950262386e-05,
+      "loss": 0.2641,
+      "step": 391
+    },
+    {
+      "epoch": 1.5893670886075948,
+      "grad_norm": 0.4358881956832152,
+      "learning_rate": 3.4449732099216985e-05,
+      "loss": 0.2674,
+      "step": 392
+    },
+    {
+      "epoch": 1.5934177215189873,
+      "grad_norm": 0.3544731314850457,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.2751,
+      "step": 393
+    },
+    {
+      "epoch": 1.5974683544303798,
+      "grad_norm": 0.440154796692243,
+      "learning_rate": 3.437101558517728e-05,
+      "loss": 0.2672,
+      "step": 394
+    },
+    {
+      "epoch": 1.6015189873417721,
+      "grad_norm": 0.3602617265884372,
+      "learning_rate": 3.433148355615496e-05,
+      "loss": 0.2663,
+      "step": 395
+    },
+    {
+      "epoch": 1.6055696202531644,
+      "grad_norm": 0.4106409685553198,
+      "learning_rate": 3.4291836103334294e-05,
+      "loss": 0.2838,
+      "step": 396
+    },
+    {
+      "epoch": 1.609620253164557,
+      "grad_norm": 0.34847687341181194,
+      "learning_rate": 3.425207354603043e-05,
+      "loss": 0.2839,
+      "step": 397
+    },
+    {
+      "epoch": 1.6136708860759494,
+      "grad_norm": 0.450996161731124,
+      "learning_rate": 3.421219620448553e-05,
+      "loss": 0.272,
+      "step": 398
+    },
+    {
+      "epoch": 1.6177215189873417,
+      "grad_norm": 0.3648158546015473,
+      "learning_rate": 3.417220439986623e-05,
+      "loss": 0.2695,
+      "step": 399
+    },
+    {
+      "epoch": 1.621772151898734,
+      "grad_norm": 0.41179075379198626,
+      "learning_rate": 3.4132098454261024e-05,
+      "loss": 0.2685,
+      "step": 400
+    },
+    {
+      "epoch": 1.6258227848101265,
+      "grad_norm": 0.3628481426970466,
+      "learning_rate": 3.4091878690677676e-05,
+      "loss": 0.2765,
+      "step": 401
+    },
+    {
+      "epoch": 1.629873417721519,
+      "grad_norm": 0.36262200906194564,
+      "learning_rate": 3.405154543304065e-05,
+      "loss": 0.2726,
+      "step": 402
+    },
+    {
+      "epoch": 1.6339240506329114,
+      "grad_norm": 0.37325361920611705,
+      "learning_rate": 3.401109900618843e-05,
+      "loss": 0.2699,
+      "step": 403
+    },
+    {
+      "epoch": 1.6379746835443036,
+      "grad_norm": 0.38436662920357584,
+      "learning_rate": 3.3970539735870996e-05,
+      "loss": 0.2842,
+      "step": 404
+    },
+    {
+      "epoch": 1.6420253164556962,
+      "grad_norm": 0.31382485990835757,
+      "learning_rate": 3.392986794874714e-05,
+      "loss": 0.2604,
+      "step": 405
+    },
+    {
+      "epoch": 1.6460759493670887,
+      "grad_norm": 0.3509698180614754,
+      "learning_rate": 3.388908397238184e-05,
+      "loss": 0.2681,
+      "step": 406
+    },
+    {
+      "epoch": 1.650126582278481,
+      "grad_norm": 0.3717814918850923,
+      "learning_rate": 3.384818813524362e-05,
+      "loss": 0.2671,
+      "step": 407
+    },
+    {
+      "epoch": 1.6541772151898733,
+      "grad_norm": 0.39025488699140487,
+      "learning_rate": 3.380718076670195e-05,
+      "loss": 0.2711,
+      "step": 408
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.36024286656648535,
+      "learning_rate": 3.376606219702454e-05,
+      "loss": 0.2727,
+      "step": 409
+    },
+    {
+      "epoch": 1.6622784810126583,
+      "grad_norm": 0.378434179590298,
+      "learning_rate": 3.372483275737468e-05,
+      "loss": 0.2767,
+      "step": 410
+    },
+    {
+      "epoch": 1.6663291139240506,
+      "grad_norm": 0.36077644454896196,
+      "learning_rate": 3.368349277980861e-05,
+      "loss": 0.2647,
+      "step": 411
+    },
+    {
+      "epoch": 1.6703797468354429,
+      "grad_norm": 0.35002498889241157,
+      "learning_rate": 3.3642042597272844e-05,
+      "loss": 0.2781,
+      "step": 412
+    },
+    {
+      "epoch": 1.6744303797468354,
+      "grad_norm": 0.32837531395524044,
+      "learning_rate": 3.360048254360144e-05,
+      "loss": 0.2669,
+      "step": 413
+    },
+    {
+      "epoch": 1.678481012658228,
+      "grad_norm": 0.32295929642556315,
+      "learning_rate": 3.355881295351336e-05,
+      "loss": 0.259,
+      "step": 414
+    },
+    {
+      "epoch": 1.6825316455696202,
+      "grad_norm": 0.36642837850765075,
+      "learning_rate": 3.351703416260975e-05,
+      "loss": 0.2643,
+      "step": 415
+    },
+    {
+      "epoch": 1.6865822784810125,
+      "grad_norm": 0.3248595866605338,
+      "learning_rate": 3.347514650737126e-05,
+      "loss": 0.2645,
+      "step": 416
+    },
+    {
+      "epoch": 1.690632911392405,
+      "grad_norm": 0.34416009951240795,
+      "learning_rate": 3.3433150325155295e-05,
+      "loss": 0.2685,
+      "step": 417
+    },
+    {
+      "epoch": 1.6946835443037975,
+      "grad_norm": 0.3250472885373944,
+      "learning_rate": 3.339104595419334e-05,
+      "loss": 0.272,
+      "step": 418
+    },
+    {
+      "epoch": 1.6987341772151898,
+      "grad_norm": 0.37338026789098194,
+      "learning_rate": 3.3348833733588204e-05,
+      "loss": 0.2847,
+      "step": 419
+    },
+    {
+      "epoch": 1.7027848101265821,
+      "grad_norm": 0.35913625701284707,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.2748,
+      "step": 420
+    },
+    {
+      "epoch": 1.7068354430379746,
+      "grad_norm": 0.36819751611718726,
+      "learning_rate": 3.326408710419996e-05,
+      "loss": 0.2806,
+      "step": 421
+    },
+    {
+      "epoch": 1.7108860759493671,
+      "grad_norm": 0.3389437581016939,
+      "learning_rate": 3.322155337795454e-05,
+      "loss": 0.2593,
+      "step": 422
+    },
+    {
+      "epoch": 1.7149367088607594,
+      "grad_norm": 0.38238374171708617,
+      "learning_rate": 3.317891316713587e-05,
+      "loss": 0.283,
+      "step": 423
+    },
+    {
+      "epoch": 1.7189873417721517,
+      "grad_norm": 0.3577739373987797,
+      "learning_rate": 3.313616681516231e-05,
+      "loss": 0.2624,
+      "step": 424
+    },
+    {
+      "epoch": 1.7230379746835442,
+      "grad_norm": 0.4314297028652508,
+      "learning_rate": 3.309331466630713e-05,
+      "loss": 0.2732,
+      "step": 425
+    },
+    {
+      "epoch": 1.7270886075949368,
+      "grad_norm": 0.3590055801698935,
+      "learning_rate": 3.305035706569563e-05,
+      "loss": 0.2623,
+      "step": 426
+    },
+    {
+      "epoch": 1.7311392405063293,
+      "grad_norm": 0.41202399066468826,
+      "learning_rate": 3.3007294359302433e-05,
+      "loss": 0.2755,
+      "step": 427
+    },
+    {
+      "epoch": 1.7351898734177216,
+      "grad_norm": 0.41533785860359085,
+      "learning_rate": 3.296412689394864e-05,
+      "loss": 0.2795,
+      "step": 428
+    },
+    {
+      "epoch": 1.7392405063291139,
+      "grad_norm": 0.34243863372965266,
+      "learning_rate": 3.292085501729909e-05,
+      "loss": 0.2762,
+      "step": 429
+    },
+    {
+      "epoch": 1.7432911392405064,
+      "grad_norm": 0.3745072843594452,
+      "learning_rate": 3.2877479077859534e-05,
+      "loss": 0.2753,
+      "step": 430
+    },
+    {
+      "epoch": 1.747341772151899,
+      "grad_norm": 0.3594293619341717,
+      "learning_rate": 3.283399942497381e-05,
+      "loss": 0.2524,
+      "step": 431
+    },
+    {
+      "epoch": 1.7513924050632912,
+      "grad_norm": 0.36954440644353176,
+      "learning_rate": 3.279041640882108e-05,
+      "loss": 0.2662,
+      "step": 432
+    },
+    {
+      "epoch": 1.7554430379746835,
+      "grad_norm": 0.3499785589053878,
+      "learning_rate": 3.2746730380412964e-05,
+      "loss": 0.2568,
+      "step": 433
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 0.3524812135247928,
+      "learning_rate": 3.2702941691590726e-05,
+      "loss": 0.2598,
+      "step": 434
+    },
+    {
+      "epoch": 1.7635443037974685,
+      "grad_norm": 0.38225382271187425,
+      "learning_rate": 3.265905069502244e-05,
+      "loss": 0.2687,
+      "step": 435
+    },
+    {
+      "epoch": 1.7675949367088608,
+      "grad_norm": 0.36531856906416743,
+      "learning_rate": 3.261505774420016e-05,
+      "loss": 0.2738,
+      "step": 436
+    },
+    {
+      "epoch": 1.771645569620253,
+      "grad_norm": 0.3298918056610023,
+      "learning_rate": 3.257096319343707e-05,
+      "loss": 0.2677,
+      "step": 437
+    },
+    {
+      "epoch": 1.7756962025316456,
+      "grad_norm": 0.3625949281057629,
+      "learning_rate": 3.2526767397864614e-05,
+      "loss": 0.2709,
+      "step": 438
+    },
+    {
+      "epoch": 1.7797468354430381,
+      "grad_norm": 0.3824958529596506,
+      "learning_rate": 3.248247071342966e-05,
+      "loss": 0.259,
+      "step": 439
+    },
+    {
+      "epoch": 1.7837974683544304,
+      "grad_norm": 0.3875038889112726,
+      "learning_rate": 3.243807349689161e-05,
+      "loss": 0.2696,
+      "step": 440
+    },
+    {
+      "epoch": 1.7878481012658227,
+      "grad_norm": 0.39460798131642516,
+      "learning_rate": 3.2393576105819544e-05,
+      "loss": 0.2641,
+      "step": 441
+    },
+    {
+      "epoch": 1.7918987341772152,
+      "grad_norm": 0.3838243118394281,
+      "learning_rate": 3.2348978898589333e-05,
+      "loss": 0.2718,
+      "step": 442
+    },
+    {
+      "epoch": 1.7959493670886078,
+      "grad_norm": 0.40460125116946677,
+      "learning_rate": 3.230428223438075e-05,
+      "loss": 0.2675,
+      "step": 443
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.36028046009020576,
+      "learning_rate": 3.225948647317459e-05,
+      "loss": 0.2584,
+      "step": 444
+    },
+    {
+      "epoch": 1.8040506329113923,
+      "grad_norm": 0.36311137097111484,
+      "learning_rate": 3.2214591975749745e-05,
+      "loss": 0.2619,
+      "step": 445
+    },
+    {
+      "epoch": 1.8081012658227849,
+      "grad_norm": 0.337513068420752,
+      "learning_rate": 3.216959910368034e-05,
+      "loss": 0.2627,
+      "step": 446
+    },
+    {
+      "epoch": 1.8121518987341774,
+      "grad_norm": 0.34859508704392567,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.2501,
+      "step": 447
+    },
+    {
+      "epoch": 1.8162025316455697,
+      "grad_norm": 0.3828143022322927,
+      "learning_rate": 3.207931968586281e-05,
+      "loss": 0.2929,
+      "step": 448
+    },
+    {
+      "epoch": 1.820253164556962,
+      "grad_norm": 0.37868625745198814,
+      "learning_rate": 3.203403386721272e-05,
+      "loss": 0.2666,
+      "step": 449
+    },
+    {
+      "epoch": 1.8243037974683545,
+      "grad_norm": 0.3880583919385183,
+      "learning_rate": 3.1988651128108245e-05,
+      "loss": 0.262,
+      "step": 450
+    },
+    {
+      "epoch": 1.828354430379747,
+      "grad_norm": 0.3804379913365177,
+      "learning_rate": 3.194317183405573e-05,
+      "loss": 0.2812,
+      "step": 451
+    },
+    {
+      "epoch": 1.8324050632911393,
+      "grad_norm": 0.38206413283177676,
+      "learning_rate": 3.189759635133914e-05,
+      "loss": 0.2646,
+      "step": 452
+    },
+    {
+      "epoch": 1.8364556962025316,
+      "grad_norm": 0.3877381477716973,
+      "learning_rate": 3.185192504701718e-05,
+      "loss": 0.2757,
+      "step": 453
+    },
+    {
+      "epoch": 1.840506329113924,
+      "grad_norm": 0.3536810553727718,
+      "learning_rate": 3.1806158288920234e-05,
+      "loss": 0.2624,
+      "step": 454
+    },
+    {
+      "epoch": 1.8445569620253166,
+      "grad_norm": 0.37039979476875257,
+      "learning_rate": 3.1760296445647477e-05,
+      "loss": 0.2754,
+      "step": 455
+    },
+    {
+      "epoch": 1.848607594936709,
+      "grad_norm": 0.3471379017857986,
+      "learning_rate": 3.1714339886563896e-05,
+      "loss": 0.2746,
+      "step": 456
+    },
+    {
+      "epoch": 1.8526582278481012,
+      "grad_norm": 0.33646246969232146,
+      "learning_rate": 3.166828898179731e-05,
+      "loss": 0.2791,
+      "step": 457
+    },
+    {
+      "epoch": 1.8567088607594937,
+      "grad_norm": 0.40441523854874667,
+      "learning_rate": 3.162214410223536e-05,
+      "loss": 0.2793,
+      "step": 458
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.3522384314025848,
+      "learning_rate": 3.157590561952257e-05,
+      "loss": 0.2757,
+      "step": 459
+    },
+    {
+      "epoch": 1.8648101265822785,
+      "grad_norm": 0.3701022884595472,
+      "learning_rate": 3.152957390605732e-05,
+      "loss": 0.2669,
+      "step": 460
+    },
+    {
+      "epoch": 1.8688607594936708,
+      "grad_norm": 0.3260260153842614,
+      "learning_rate": 3.148314933498886e-05,
+      "loss": 0.2578,
+      "step": 461
+    },
+    {
+      "epoch": 1.8729113924050633,
+      "grad_norm": 0.36367921187042007,
+      "learning_rate": 3.143663228021431e-05,
+      "loss": 0.2728,
+      "step": 462
+    },
+    {
+      "epoch": 1.8769620253164558,
+      "grad_norm": 0.3382428343668073,
+      "learning_rate": 3.1390023116375624e-05,
+      "loss": 0.2776,
+      "step": 463
+    },
+    {
+      "epoch": 1.8810126582278481,
+      "grad_norm": 0.3212645631382389,
+      "learning_rate": 3.134332221885661e-05,
+      "loss": 0.2702,
+      "step": 464
+    },
+    {
+      "epoch": 1.8850632911392404,
+      "grad_norm": 0.33323077401524304,
+      "learning_rate": 3.129652996377987e-05,
+      "loss": 0.2599,
+      "step": 465
+    },
+    {
+      "epoch": 1.889113924050633,
+      "grad_norm": 0.35713773149166567,
+      "learning_rate": 3.12496467280038e-05,
+      "loss": 0.2643,
+      "step": 466
+    },
+    {
+      "epoch": 1.8931645569620255,
+      "grad_norm": 0.3225869159208159,
+      "learning_rate": 3.120267288911952e-05,
+      "loss": 0.2568,
+      "step": 467
+    },
+    {
+      "epoch": 1.8972151898734178,
+      "grad_norm": 0.34610198409764964,
+      "learning_rate": 3.11556088254479e-05,
+      "loss": 0.2787,
+      "step": 468
+    },
+    {
+      "epoch": 1.90126582278481,
+      "grad_norm": 0.3362720930325975,
+      "learning_rate": 3.11084549160364e-05,
+      "loss": 0.2681,
+      "step": 469
+    },
+    {
+      "epoch": 1.9053164556962026,
+      "grad_norm": 0.3516176992461195,
+      "learning_rate": 3.106121154065615e-05,
+      "loss": 0.2651,
+      "step": 470
+    },
+    {
+      "epoch": 1.909367088607595,
+      "grad_norm": 0.34988094803777076,
+      "learning_rate": 3.1013879079798805e-05,
+      "loss": 0.2635,
+      "step": 471
+    },
+    {
+      "epoch": 1.9134177215189874,
+      "grad_norm": 0.3473185360408808,
+      "learning_rate": 3.096645791467348e-05,
+      "loss": 0.2621,
+      "step": 472
+    },
+    {
+      "epoch": 1.9174683544303797,
+      "grad_norm": 0.3221122210834881,
+      "learning_rate": 3.091894842720373e-05,
+      "loss": 0.2739,
+      "step": 473
+    },
+    {
+      "epoch": 1.9215189873417722,
+      "grad_norm": 0.34967339926655305,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.267,
+      "step": 474
+    },
+    {
+      "epoch": 1.9255696202531647,
+      "grad_norm": 0.35417511809935553,
+      "learning_rate": 3.0823666016478716e-05,
+      "loss": 0.2784,
+      "step": 475
+    },
+    {
+      "epoch": 1.929620253164557,
+      "grad_norm": 0.362650554958712,
+      "learning_rate": 3.0775893860614896e-05,
+      "loss": 0.258,
+      "step": 476
+    },
+    {
+      "epoch": 1.9336708860759493,
+      "grad_norm": 0.4077973551205214,
+      "learning_rate": 3.0728034917183336e-05,
+      "loss": 0.2466,
+      "step": 477
+    },
+    {
+      "epoch": 1.9377215189873418,
+      "grad_norm": 0.3428197722819325,
+      "learning_rate": 3.06800895716334e-05,
+      "loss": 0.2715,
+      "step": 478
+    },
+    {
+      "epoch": 1.9417721518987343,
+      "grad_norm": 0.4448717468087317,
+      "learning_rate": 3.063205821011029e-05,
+      "loss": 0.2808,
+      "step": 479
+    },
+    {
+      "epoch": 1.9458227848101266,
+      "grad_norm": 0.36159693931759995,
+      "learning_rate": 3.0583941219452016e-05,
+      "loss": 0.2748,
+      "step": 480
+    },
+    {
+      "epoch": 1.949873417721519,
+      "grad_norm": 0.3617494825089683,
+      "learning_rate": 3.053573898718618e-05,
+      "loss": 0.2571,
+      "step": 481
+    },
+    {
+      "epoch": 1.9539240506329114,
+      "grad_norm": 0.4042060158347555,
+      "learning_rate": 3.0487451901526956e-05,
+      "loss": 0.2805,
+      "step": 482
+    },
+    {
+      "epoch": 1.957974683544304,
+      "grad_norm": 0.4031904740187985,
+      "learning_rate": 3.0439080351371875e-05,
+      "loss": 0.2755,
+      "step": 483
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.37399962410893606,
+      "learning_rate": 3.0390624726298764e-05,
+      "loss": 0.2643,
+      "step": 484
+    },
+    {
+      "epoch": 1.9660759493670885,
+      "grad_norm": 0.4593321518174994,
+      "learning_rate": 3.034208541656255e-05,
+      "loss": 0.2782,
+      "step": 485
+    },
+    {
+      "epoch": 1.970126582278481,
+      "grad_norm": 0.32819354675368384,
+      "learning_rate": 3.029346281309218e-05,
+      "loss": 0.2559,
+      "step": 486
+    },
+    {
+      "epoch": 1.9741772151898735,
+      "grad_norm": 0.3697594847531436,
+      "learning_rate": 3.0244757307487415e-05,
+      "loss": 0.2702,
+      "step": 487
+    },
+    {
+      "epoch": 1.9782278481012658,
+      "grad_norm": 0.40609700862321335,
+      "learning_rate": 3.019596929201569e-05,
+      "loss": 0.2688,
+      "step": 488
+    },
+    {
+      "epoch": 1.9822784810126581,
+      "grad_norm": 0.3510762073535274,
+      "learning_rate": 3.0147099159608985e-05,
+      "loss": 0.2807,
+      "step": 489
+    },
+    {
+      "epoch": 1.9863291139240506,
+      "grad_norm": 0.4507182114316109,
+      "learning_rate": 3.0098147303860616e-05,
+      "loss": 0.259,
+      "step": 490
+    },
+    {
+      "epoch": 1.9903797468354432,
+      "grad_norm": 0.3217423093381456,
+      "learning_rate": 3.0049114119022117e-05,
+      "loss": 0.2682,
+      "step": 491
+    },
+    {
+      "epoch": 1.9944303797468355,
+      "grad_norm": 0.4559395783889565,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.2637,
+      "step": 492
+    },
+    {
+      "epoch": 1.9984810126582278,
+      "grad_norm": 0.34459261903539634,
+      "learning_rate": 2.995080534235264e-05,
+      "loss": 0.2653,
+      "step": 493
+    },
+    {
+      "epoch": 2.0040506329113925,
+      "grad_norm": 0.4665725350090339,
+      "learning_rate": 2.9901530542287044e-05,
+      "loss": 0.2026,
+      "step": 494
+    },
+    {
+      "epoch": 2.008101265822785,
+      "grad_norm": 0.3356210787089712,
+      "learning_rate": 2.9852175996655676e-05,
+      "loss": 0.1895,
+      "step": 495
+    },
+    {
+      "epoch": 2.012151898734177,
+      "grad_norm": 0.5517523351334922,
+      "learning_rate": 2.980274210295326e-05,
+      "loss": 0.2044,
+      "step": 496
+    },
+    {
+      "epoch": 2.0162025316455696,
+      "grad_norm": 0.378721286324737,
+      "learning_rate": 2.9753229259313578e-05,
+      "loss": 0.207,
+      "step": 497
+    },
+    {
+      "epoch": 2.020253164556962,
+      "grad_norm": 0.326603336110659,
+      "learning_rate": 2.9703637864506274e-05,
+      "loss": 0.1877,
+      "step": 498
+    },
+    {
+      "epoch": 2.0243037974683546,
+      "grad_norm": 0.38885611104460355,
+      "learning_rate": 2.965396831793362e-05,
+      "loss": 0.2118,
+      "step": 499
+    },
+    {
+      "epoch": 2.0283544303797467,
+      "grad_norm": 0.32983590316935846,
+      "learning_rate": 2.9604221019627316e-05,
+      "loss": 0.2012,
+      "step": 500
+    },
+    {
+      "epoch": 2.0324050632911392,
+      "grad_norm": 0.38176033558661965,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.1962,
+      "step": 501
+    },
+    {
+      "epoch": 2.0364556962025318,
+      "grad_norm": 0.3369625324063912,
+      "learning_rate": 2.9504494771068334e-05,
+      "loss": 0.1917,
+      "step": 502
+    },
+    {
+      "epoch": 2.0405063291139243,
+      "grad_norm": 0.3637252938061798,
+      "learning_rate": 2.9454516623997156e-05,
+      "loss": 0.1901,
+      "step": 503
+    },
+    {
+      "epoch": 2.0445569620253163,
+      "grad_norm": 0.3379402256463755,
+      "learning_rate": 2.9404462331548847e-05,
+      "loss": 0.2085,
+      "step": 504
+    },
+    {
+      "epoch": 2.048607594936709,
+      "grad_norm": 0.3088655244128315,
+      "learning_rate": 2.93543322968538e-05,
+      "loss": 0.1964,
+      "step": 505
+    },
+    {
+      "epoch": 2.0526582278481014,
+      "grad_norm": 0.3544111971695929,
+      "learning_rate": 2.9304126923652428e-05,
+      "loss": 0.2055,
+      "step": 506
+    },
+    {
+      "epoch": 2.056708860759494,
+      "grad_norm": 0.3270008687095874,
+      "learning_rate": 2.9253846616291896e-05,
+      "loss": 0.2096,
+      "step": 507
+    },
+    {
+      "epoch": 2.060759493670886,
+      "grad_norm": 0.316848418760831,
+      "learning_rate": 2.9203491779722896e-05,
+      "loss": 0.186,
+      "step": 508
+    },
+    {
+      "epoch": 2.0648101265822785,
+      "grad_norm": 0.3272487453887169,
+      "learning_rate": 2.9153062819496357e-05,
+      "loss": 0.1916,
+      "step": 509
+    },
+    {
+      "epoch": 2.068860759493671,
+      "grad_norm": 0.34820959913703065,
+      "learning_rate": 2.9102560141760178e-05,
+      "loss": 0.2102,
+      "step": 510
+    },
+    {
+      "epoch": 2.0729113924050635,
+      "grad_norm": 0.3242245816811273,
+      "learning_rate": 2.9051984153256004e-05,
+      "loss": 0.1988,
+      "step": 511
+    },
+    {
+      "epoch": 2.0769620253164556,
+      "grad_norm": 0.30062489949628063,
+      "learning_rate": 2.900133526131588e-05,
+      "loss": 0.197,
+      "step": 512
+    },
+    {
+      "epoch": 2.081012658227848,
+      "grad_norm": 0.35553879851586756,
+      "learning_rate": 2.8950613873859025e-05,
+      "loss": 0.2049,
+      "step": 513
+    },
+    {
+      "epoch": 2.0850632911392406,
+      "grad_norm": 0.338205903178864,
+      "learning_rate": 2.8899820399388515e-05,
+      "loss": 0.2084,
+      "step": 514
+    },
+    {
+      "epoch": 2.089113924050633,
+      "grad_norm": 0.33707960470095344,
+      "learning_rate": 2.8848955246988012e-05,
+      "loss": 0.1973,
+      "step": 515
+    },
+    {
+      "epoch": 2.093164556962025,
+      "grad_norm": 0.33222852149937315,
+      "learning_rate": 2.879801882631847e-05,
+      "loss": 0.2096,
+      "step": 516
+    },
+    {
+      "epoch": 2.0972151898734177,
+      "grad_norm": 0.3213761793297667,
+      "learning_rate": 2.8747011547614808e-05,
+      "loss": 0.1902,
+      "step": 517
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.35746812714896625,
+      "learning_rate": 2.8695933821682635e-05,
+      "loss": 0.1932,
+      "step": 518
+    },
+    {
+      "epoch": 2.1053164556962027,
+      "grad_norm": 0.31521426852559564,
+      "learning_rate": 2.864478605989494e-05,
+      "loss": 0.1972,
+      "step": 519
+    },
+    {
+      "epoch": 2.109367088607595,
+      "grad_norm": 0.3913485946901582,
+      "learning_rate": 2.8593568674188765e-05,
+      "loss": 0.1953,
+      "step": 520
+    },
+    {
+      "epoch": 2.1134177215189873,
+      "grad_norm": 0.3347639045875019,
+      "learning_rate": 2.8542282077061892e-05,
+      "loss": 0.1963,
+      "step": 521
+    },
+    {
+      "epoch": 2.11746835443038,
+      "grad_norm": 0.4108179878253096,
+      "learning_rate": 2.8490926681569523e-05,
+      "loss": 0.196,
+      "step": 522
+    },
+    {
+      "epoch": 2.1215189873417724,
+      "grad_norm": 0.3074875830900616,
+      "learning_rate": 2.8439502901320956e-05,
+      "loss": 0.2002,
+      "step": 523
+    },
+    {
+      "epoch": 2.1255696202531644,
+      "grad_norm": 0.3720291840709936,
+      "learning_rate": 2.8388011150476237e-05,
+      "loss": 0.1924,
+      "step": 524
+    },
+    {
+      "epoch": 2.129620253164557,
+      "grad_norm": 0.3077447027060388,
+      "learning_rate": 2.8336451843742866e-05,
+      "loss": 0.1875,
+      "step": 525
+    },
+    {
+      "epoch": 2.1336708860759495,
+      "grad_norm": 0.35542587166198397,
+      "learning_rate": 2.8284825396372387e-05,
+      "loss": 0.2006,
+      "step": 526
+    },
+    {
+      "epoch": 2.137721518987342,
+      "grad_norm": 0.32693955216562326,
+      "learning_rate": 2.8233132224157132e-05,
+      "loss": 0.1883,
+      "step": 527
+    },
+    {
+      "epoch": 2.141772151898734,
+      "grad_norm": 0.35554226010315526,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.1856,
+      "step": 528
+    },
+    {
+      "epoch": 2.1458227848101266,
+      "grad_norm": 0.3321208159869709,
+      "learning_rate": 2.8129547371045128e-05,
+      "loss": 0.1943,
+      "step": 529
+    },
+    {
+      "epoch": 2.149873417721519,
+      "grad_norm": 0.4176868717591123,
+      "learning_rate": 2.8077656524406534e-05,
+      "loss": 0.2063,
+      "step": 530
+    },
+    {
+      "epoch": 2.1539240506329116,
+      "grad_norm": 0.32523610959013094,
+      "learning_rate": 2.802570062143278e-05,
+      "loss": 0.1946,
+      "step": 531
+    },
+    {
+      "epoch": 2.1579746835443037,
+      "grad_norm": 0.36841987256034064,
+      "learning_rate": 2.7973680080569555e-05,
+      "loss": 0.1909,
+      "step": 532
+    },
+    {
+      "epoch": 2.162025316455696,
+      "grad_norm": 0.3559056427533516,
+      "learning_rate": 2.792159532078314e-05,
+      "loss": 0.1918,
+      "step": 533
+    },
+    {
+      "epoch": 2.1660759493670887,
+      "grad_norm": 0.35669708972966335,
+      "learning_rate": 2.7869446761557033e-05,
+      "loss": 0.1919,
+      "step": 534
+    },
+    {
+      "epoch": 2.170126582278481,
+      "grad_norm": 0.340002498666834,
+      "learning_rate": 2.781723482288857e-05,
+      "loss": 0.1924,
+      "step": 535
+    },
+    {
+      "epoch": 2.1741772151898733,
+      "grad_norm": 0.34436820971195903,
+      "learning_rate": 2.7764959925285517e-05,
+      "loss": 0.194,
+      "step": 536
+    },
+    {
+      "epoch": 2.178227848101266,
+      "grad_norm": 0.3953670498625473,
+      "learning_rate": 2.771262248976272e-05,
+      "loss": 0.2068,
+      "step": 537
+    },
+    {
+      "epoch": 2.1822784810126583,
+      "grad_norm": 0.3025575621329444,
+      "learning_rate": 2.7660222937838677e-05,
+      "loss": 0.1932,
+      "step": 538
+    },
+    {
+      "epoch": 2.186329113924051,
+      "grad_norm": 0.39043080501165595,
+      "learning_rate": 2.7607761691532186e-05,
+      "loss": 0.1917,
+      "step": 539
+    },
+    {
+      "epoch": 2.190379746835443,
+      "grad_norm": 0.3072351277090597,
+      "learning_rate": 2.7555239173358916e-05,
+      "loss": 0.2028,
+      "step": 540
+    },
+    {
+      "epoch": 2.1944303797468354,
+      "grad_norm": 0.3511861222402343,
+      "learning_rate": 2.7502655806328e-05,
+      "loss": 0.196,
+      "step": 541
+    },
+    {
+      "epoch": 2.198481012658228,
+      "grad_norm": 0.3207567981216629,
+      "learning_rate": 2.7450012013938648e-05,
+      "loss": 0.1898,
+      "step": 542
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.3227363582421811,
+      "learning_rate": 2.739730822017673e-05,
+      "loss": 0.1971,
+      "step": 543
+    },
+    {
+      "epoch": 2.2065822784810125,
+      "grad_norm": 0.3402163283748372,
+      "learning_rate": 2.7344544849511355e-05,
+      "loss": 0.1982,
+      "step": 544
+    },
+    {
+      "epoch": 2.210632911392405,
+      "grad_norm": 0.3362338331351972,
+      "learning_rate": 2.7291722326891456e-05,
+      "loss": 0.2037,
+      "step": 545
+    },
+    {
+      "epoch": 2.2146835443037975,
+      "grad_norm": 0.31093887217029825,
+      "learning_rate": 2.723884107774236e-05,
+      "loss": 0.1944,
+      "step": 546
+    },
+    {
+      "epoch": 2.21873417721519,
+      "grad_norm": 0.34414668711420443,
+      "learning_rate": 2.718590152796239e-05,
+      "loss": 0.2008,
+      "step": 547
+    },
+    {
+      "epoch": 2.222784810126582,
+      "grad_norm": 0.3317701651077702,
+      "learning_rate": 2.71329041039194e-05,
+      "loss": 0.1973,
+      "step": 548
+    },
+    {
+      "epoch": 2.2268354430379746,
+      "grad_norm": 0.3369317199630825,
+      "learning_rate": 2.7079849232447357e-05,
+      "loss": 0.1975,
+      "step": 549
+    },
+    {
+      "epoch": 2.230886075949367,
+      "grad_norm": 0.3280787673211415,
+      "learning_rate": 2.7026737340842895e-05,
+      "loss": 0.1875,
+      "step": 550
+    },
+    {
+      "epoch": 2.2349367088607597,
+      "grad_norm": 0.3301199229455708,
+      "learning_rate": 2.697356885686189e-05,
+      "loss": 0.1989,
+      "step": 551
+    },
+    {
+      "epoch": 2.2389873417721518,
+      "grad_norm": 0.3239931295716303,
+      "learning_rate": 2.6920344208716014e-05,
+      "loss": 0.1876,
+      "step": 552
+    },
+    {
+      "epoch": 2.2430379746835443,
+      "grad_norm": 0.3168184718628148,
+      "learning_rate": 2.6867063825069252e-05,
+      "loss": 0.1837,
+      "step": 553
+    },
+    {
+      "epoch": 2.247088607594937,
+      "grad_norm": 0.359692631623311,
+      "learning_rate": 2.6813728135034494e-05,
+      "loss": 0.2012,
+      "step": 554
+    },
+    {
+      "epoch": 2.2511392405063293,
+      "grad_norm": 0.3477050454498739,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.203,
+      "step": 555
+    },
+    {
+      "epoch": 2.2551898734177214,
+      "grad_norm": 0.34352232256919807,
+      "learning_rate": 2.6706892554476226e-05,
+      "loss": 0.202,
+      "step": 556
+    },
+    {
+      "epoch": 2.259240506329114,
+      "grad_norm": 0.3431307055030208,
+      "learning_rate": 2.6653393524391795e-05,
+      "loss": 0.196,
+      "step": 557
+    },
+    {
+      "epoch": 2.2632911392405064,
+      "grad_norm": 0.33587670966242417,
+      "learning_rate": 2.6599840908790592e-05,
+      "loss": 0.2022,
+      "step": 558
+    },
+    {
+      "epoch": 2.267341772151899,
+      "grad_norm": 0.3294890519583124,
+      "learning_rate": 2.6546235138978028e-05,
+      "loss": 0.2071,
+      "step": 559
+    },
+    {
+      "epoch": 2.271392405063291,
+      "grad_norm": 0.3088655912611178,
+      "learning_rate": 2.6492576646687597e-05,
+      "loss": 0.2016,
+      "step": 560
+    },
+    {
+      "epoch": 2.2754430379746835,
+      "grad_norm": 0.34694857965927567,
+      "learning_rate": 2.6438865864077425e-05,
+      "loss": 0.2075,
+      "step": 561
+    },
+    {
+      "epoch": 2.279493670886076,
+      "grad_norm": 0.3425683412604654,
+      "learning_rate": 2.6385103223726766e-05,
+      "loss": 0.211,
+      "step": 562
+    },
+    {
+      "epoch": 2.2835443037974685,
+      "grad_norm": 0.3009254063943637,
+      "learning_rate": 2.6331289158632537e-05,
+      "loss": 0.2095,
+      "step": 563
+    },
+    {
+      "epoch": 2.2875949367088606,
+      "grad_norm": 0.3461559818165035,
+      "learning_rate": 2.6277424102205817e-05,
+      "loss": 0.2031,
+      "step": 564
+    },
+    {
+      "epoch": 2.291645569620253,
+      "grad_norm": 0.32525910675334513,
+      "learning_rate": 2.6223508488268374e-05,
+      "loss": 0.1952,
+      "step": 565
+    },
+    {
+      "epoch": 2.2956962025316456,
+      "grad_norm": 0.34777287118232136,
+      "learning_rate": 2.6169542751049148e-05,
+      "loss": 0.2066,
+      "step": 566
+    },
+    {
+      "epoch": 2.299746835443038,
+      "grad_norm": 0.3568148393044382,
+      "learning_rate": 2.6115527325180754e-05,
+      "loss": 0.1913,
+      "step": 567
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.34235208756561436,
+      "learning_rate": 2.606146264569603e-05,
+      "loss": 0.2021,
+      "step": 568
+    },
+    {
+      "epoch": 2.3078481012658227,
+      "grad_norm": 0.32448929029449114,
+      "learning_rate": 2.6007349148024447e-05,
+      "loss": 0.199,
+      "step": 569
+    },
+    {
+      "epoch": 2.3118987341772153,
+      "grad_norm": 0.3671640403969276,
+      "learning_rate": 2.5953187267988694e-05,
+      "loss": 0.2003,
+      "step": 570
+    },
+    {
+      "epoch": 2.3159493670886078,
+      "grad_norm": 0.3273581343020394,
+      "learning_rate": 2.5898977441801097e-05,
+      "loss": 0.1909,
+      "step": 571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.37274863944289643,
+      "learning_rate": 2.584472010606015e-05,
+      "loss": 0.1914,
+      "step": 572
+    },
+    {
+      "epoch": 2.3240506329113924,
+      "grad_norm": 0.35997610272824465,
+      "learning_rate": 2.5790415697746976e-05,
+      "loss": 0.1908,
+      "step": 573
+    },
+    {
+      "epoch": 2.328101265822785,
+      "grad_norm": 0.36343286974420896,
+      "learning_rate": 2.5736064654221808e-05,
+      "loss": 0.1912,
+      "step": 574
+    },
+    {
+      "epoch": 2.3321518987341774,
+      "grad_norm": 0.35138755472990757,
+      "learning_rate": 2.568166741322048e-05,
+      "loss": 0.1979,
+      "step": 575
+    },
+    {
+      "epoch": 2.3362025316455695,
+      "grad_norm": 0.36190940805357685,
+      "learning_rate": 2.56272244128509e-05,
+      "loss": 0.1874,
+      "step": 576
+    },
+    {
+      "epoch": 2.340253164556962,
+      "grad_norm": 0.3834635173372978,
+      "learning_rate": 2.55727360915895e-05,
+      "loss": 0.1905,
+      "step": 577
+    },
+    {
+      "epoch": 2.3443037974683545,
+      "grad_norm": 0.32305338846536025,
+      "learning_rate": 2.5518202888277734e-05,
+      "loss": 0.197,
+      "step": 578
+    },
+    {
+      "epoch": 2.348354430379747,
+      "grad_norm": 0.3389535413321349,
+      "learning_rate": 2.5463625242118523e-05,
+      "loss": 0.2037,
+      "step": 579
+    },
+    {
+      "epoch": 2.352405063291139,
+      "grad_norm": 0.32320686221314576,
+      "learning_rate": 2.5409003592672723e-05,
+      "loss": 0.1979,
+      "step": 580
+    },
+    {
+      "epoch": 2.3564556962025316,
+      "grad_norm": 0.32098370333112813,
+      "learning_rate": 2.535433837985559e-05,
+      "loss": 0.1883,
+      "step": 581
+    },
+    {
+      "epoch": 2.360506329113924,
+      "grad_norm": 0.3657704139908037,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.2055,
+      "step": 582
+    },
+    {
+      "epoch": 2.3645569620253166,
+      "grad_norm": 0.33692715095620746,
+      "learning_rate": 2.524487902551908e-05,
+      "loss": 0.204,
+      "step": 583
+    },
+    {
+      "epoch": 2.3686075949367087,
+      "grad_norm": 0.3525443285754283,
+      "learning_rate": 2.519008576557029e-05,
+      "loss": 0.21,
+      "step": 584
+    },
+    {
+      "epoch": 2.372658227848101,
+      "grad_norm": 0.33764077012599775,
+      "learning_rate": 2.5135250705384254e-05,
+      "loss": 0.1962,
+      "step": 585
+    },
+    {
+      "epoch": 2.3767088607594937,
+      "grad_norm": 0.329522493190511,
+      "learning_rate": 2.5080374286595007e-05,
+      "loss": 0.1954,
+      "step": 586
+    },
+    {
+      "epoch": 2.3807594936708862,
+      "grad_norm": 0.3563595102685283,
+      "learning_rate": 2.5025456951169677e-05,
+      "loss": 0.2034,
+      "step": 587
+    },
+    {
+      "epoch": 2.3848101265822783,
+      "grad_norm": 0.3239996241574444,
+      "learning_rate": 2.4970499141404942e-05,
+      "loss": 0.2003,
+      "step": 588
+    },
+    {
+      "epoch": 2.388860759493671,
+      "grad_norm": 0.33636033403274357,
+      "learning_rate": 2.491550129992345e-05,
+      "loss": 0.1929,
+      "step": 589
+    },
+    {
+      "epoch": 2.3929113924050633,
+      "grad_norm": 0.3155324646064635,
+      "learning_rate": 2.486046386967024e-05,
+      "loss": 0.206,
+      "step": 590
+    },
+    {
+      "epoch": 2.396962025316456,
+      "grad_norm": 0.33084309855288785,
+      "learning_rate": 2.4805387293909214e-05,
+      "loss": 0.2051,
+      "step": 591
+    },
+    {
+      "epoch": 2.401012658227848,
+      "grad_norm": 0.32906844762669607,
+      "learning_rate": 2.4750272016219552e-05,
+      "loss": 0.192,
+      "step": 592
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.36501586273474845,
+      "learning_rate": 2.4695118480492114e-05,
+      "loss": 0.2013,
+      "step": 593
+    },
+    {
+      "epoch": 2.409113924050633,
+      "grad_norm": 0.33322099474954187,
+      "learning_rate": 2.4639927130925898e-05,
+      "loss": 0.1974,
+      "step": 594
+    },
+    {
+      "epoch": 2.4131645569620255,
+      "grad_norm": 0.3305457946938421,
+      "learning_rate": 2.458469841202444e-05,
+      "loss": 0.1898,
+      "step": 595
+    },
+    {
+      "epoch": 2.4172151898734175,
+      "grad_norm": 0.38605034294207946,
+      "learning_rate": 2.452943276859226e-05,
+      "loss": 0.1863,
+      "step": 596
+    },
+    {
+      "epoch": 2.42126582278481,
+      "grad_norm": 0.336359493538504,
+      "learning_rate": 2.447413064573125e-05,
+      "loss": 0.208,
+      "step": 597
+    },
+    {
+      "epoch": 2.4253164556962026,
+      "grad_norm": 0.34024657040137163,
+      "learning_rate": 2.4418792488837095e-05,
+      "loss": 0.2011,
+      "step": 598
+    },
+    {
+      "epoch": 2.429367088607595,
+      "grad_norm": 0.3159981723197647,
+      "learning_rate": 2.4363418743595713e-05,
+      "loss": 0.2039,
+      "step": 599
+    },
+    {
+      "epoch": 2.433417721518987,
+      "grad_norm": 0.3518322369181373,
+      "learning_rate": 2.430800985597963e-05,
+      "loss": 0.1966,
+      "step": 600
+    },
+    {
+      "epoch": 2.4374683544303797,
+      "grad_norm": 0.37686357049818,
+      "learning_rate": 2.4252566272244415e-05,
+      "loss": 0.1954,
+      "step": 601
+    },
+    {
+      "epoch": 2.441518987341772,
+      "grad_norm": 0.3483281893061969,
+      "learning_rate": 2.4197088438925063e-05,
+      "loss": 0.1939,
+      "step": 602
+    },
+    {
+      "epoch": 2.4455696202531647,
+      "grad_norm": 0.4059010679068205,
+      "learning_rate": 2.4141576802832417e-05,
+      "loss": 0.2072,
+      "step": 603
+    },
+    {
+      "epoch": 2.449620253164557,
+      "grad_norm": 0.32744366714259265,
+      "learning_rate": 2.408603181104957e-05,
+      "loss": 0.216,
+      "step": 604
+    },
+    {
+      "epoch": 2.4536708860759493,
+      "grad_norm": 0.3915506293243557,
+      "learning_rate": 2.4030453910928245e-05,
+      "loss": 0.2168,
+      "step": 605
+    },
+    {
+      "epoch": 2.457721518987342,
+      "grad_norm": 0.3400760152190279,
+      "learning_rate": 2.397484355008521e-05,
+      "loss": 0.2246,
+      "step": 606
+    },
+    {
+      "epoch": 2.4617721518987343,
+      "grad_norm": 0.3406308199316695,
+      "learning_rate": 2.3919201176398662e-05,
+      "loss": 0.2068,
+      "step": 607
+    },
+    {
+      "epoch": 2.4658227848101264,
+      "grad_norm": 0.3310174274898972,
+      "learning_rate": 2.3863527238004633e-05,
+      "loss": 0.2047,
+      "step": 608
+    },
+    {
+      "epoch": 2.469873417721519,
+      "grad_norm": 0.333601040175897,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.1893,
+      "step": 609
+    },
+    {
+      "epoch": 2.4739240506329114,
+      "grad_norm": 0.35667213686671373,
+      "learning_rate": 2.3752086460905725e-05,
+      "loss": 0.2022,
+      "step": 610
+    },
+    {
+      "epoch": 2.477974683544304,
+      "grad_norm": 0.35368160424760337,
+      "learning_rate": 2.3696320519729544e-05,
+      "loss": 0.2028,
+      "step": 611
+    },
+    {
+      "epoch": 2.482025316455696,
+      "grad_norm": 0.3372564174652635,
+      "learning_rate": 2.3640524808896045e-05,
+      "loss": 0.1977,
+      "step": 612
+    },
+    {
+      "epoch": 2.4860759493670885,
+      "grad_norm": 0.31413909923623434,
+      "learning_rate": 2.3584699777776222e-05,
+      "loss": 0.1982,
+      "step": 613
+    },
+    {
+      "epoch": 2.490126582278481,
+      "grad_norm": 0.3262813099618204,
+      "learning_rate": 2.3528845875977195e-05,
+      "loss": 0.1932,
+      "step": 614
+    },
+    {
+      "epoch": 2.4941772151898736,
+      "grad_norm": 0.3521312529757842,
+      "learning_rate": 2.3472963553338614e-05,
+      "loss": 0.2099,
+      "step": 615
+    },
+    {
+      "epoch": 2.4982278481012656,
+      "grad_norm": 0.3386803464042975,
+      "learning_rate": 2.341705325992901e-05,
+      "loss": 0.1963,
+      "step": 616
+    },
+    {
+      "epoch": 2.502278481012658,
+      "grad_norm": 0.34368335723981425,
+      "learning_rate": 2.336111544604222e-05,
+      "loss": 0.1983,
+      "step": 617
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.36334289173613143,
+      "learning_rate": 2.33051505621937e-05,
+      "loss": 0.1992,
+      "step": 618
+    },
+    {
+      "epoch": 2.510379746835443,
+      "grad_norm": 0.34236582051163816,
+      "learning_rate": 2.324915905911693e-05,
+      "loss": 0.2038,
+      "step": 619
+    },
+    {
+      "epoch": 2.5144303797468357,
+      "grad_norm": 0.3397046241352714,
+      "learning_rate": 2.319314138775977e-05,
+      "loss": 0.1964,
+      "step": 620
+    },
+    {
+      "epoch": 2.5184810126582278,
+      "grad_norm": 0.3515074381688505,
+      "learning_rate": 2.3137097999280856e-05,
+      "loss": 0.2078,
+      "step": 621
+    },
+    {
+      "epoch": 2.5225316455696203,
+      "grad_norm": 0.30677506914401326,
+      "learning_rate": 2.308102934504593e-05,
+      "loss": 0.1899,
+      "step": 622
+    },
+    {
+      "epoch": 2.526582278481013,
+      "grad_norm": 0.3320062745236185,
+      "learning_rate": 2.3024935876624222e-05,
+      "loss": 0.1898,
+      "step": 623
+    },
+    {
+      "epoch": 2.530632911392405,
+      "grad_norm": 0.3254749854279839,
+      "learning_rate": 2.2968818045784813e-05,
+      "loss": 0.1885,
+      "step": 624
+    },
+    {
+      "epoch": 2.5346835443037974,
+      "grad_norm": 0.32656721326826027,
+      "learning_rate": 2.2912676304493006e-05,
+      "loss": 0.1998,
+      "step": 625
+    },
+    {
+      "epoch": 2.53873417721519,
+      "grad_norm": 0.33875861127963736,
+      "learning_rate": 2.2856511104906668e-05,
+      "loss": 0.1921,
+      "step": 626
+    },
+    {
+      "epoch": 2.5427848101265824,
+      "grad_norm": 0.3418999389357287,
+      "learning_rate": 2.2800322899372586e-05,
+      "loss": 0.1962,
+      "step": 627
+    },
+    {
+      "epoch": 2.546835443037975,
+      "grad_norm": 0.33073314039253093,
+      "learning_rate": 2.2744112140422844e-05,
+      "loss": 0.2014,
+      "step": 628
+    },
+    {
+      "epoch": 2.550886075949367,
+      "grad_norm": 0.3096689029518842,
+      "learning_rate": 2.2687879280771177e-05,
+      "loss": 0.205,
+      "step": 629
+    },
+    {
+      "epoch": 2.5549367088607595,
+      "grad_norm": 0.3500933056351304,
+      "learning_rate": 2.26316247733093e-05,
+      "loss": 0.2053,
+      "step": 630
+    },
+    {
+      "epoch": 2.558987341772152,
+      "grad_norm": 0.32437162227055294,
+      "learning_rate": 2.257534907110328e-05,
+      "loss": 0.1902,
+      "step": 631
+    },
+    {
+      "epoch": 2.563037974683544,
+      "grad_norm": 0.32881147546400236,
+      "learning_rate": 2.2519052627389882e-05,
+      "loss": 0.2004,
+      "step": 632
+    },
+    {
+      "epoch": 2.5670886075949366,
+      "grad_norm": 0.3230805172119517,
+      "learning_rate": 2.246273589557294e-05,
+      "loss": 0.1928,
+      "step": 633
+    },
+    {
+      "epoch": 2.571139240506329,
+      "grad_norm": 0.3244115725208516,
+      "learning_rate": 2.240639932921966e-05,
+      "loss": 0.1993,
+      "step": 634
+    },
+    {
+      "epoch": 2.5751898734177217,
+      "grad_norm": 0.3287769521132742,
+      "learning_rate": 2.2350043382056995e-05,
+      "loss": 0.2005,
+      "step": 635
+    },
+    {
+      "epoch": 2.579240506329114,
+      "grad_norm": 0.3211112210442208,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.2006,
+      "step": 636
+    },
+    {
+      "epoch": 2.5832911392405062,
+      "grad_norm": 0.3360142673578109,
+      "learning_rate": 2.2237275160988186e-05,
+      "loss": 0.1926,
+      "step": 637
+    },
+    {
+      "epoch": 2.5873417721518988,
+      "grad_norm": 0.33618537277308935,
+      "learning_rate": 2.2180863795301787e-05,
+      "loss": 0.2054,
+      "step": 638
+    },
+    {
+      "epoch": 2.5913924050632913,
+      "grad_norm": 0.3188167522865972,
+      "learning_rate": 2.212443486523819e-05,
+      "loss": 0.1927,
+      "step": 639
+    },
+    {
+      "epoch": 2.5954430379746833,
+      "grad_norm": 0.35570866975821724,
+      "learning_rate": 2.2067988825268243e-05,
+      "loss": 0.1977,
+      "step": 640
+    },
+    {
+      "epoch": 2.599493670886076,
+      "grad_norm": 0.33085944416382435,
+      "learning_rate": 2.2011526130000596e-05,
+      "loss": 0.2079,
+      "step": 641
+    },
+    {
+      "epoch": 2.6035443037974684,
+      "grad_norm": 0.3102323807520179,
+      "learning_rate": 2.1955047234178038e-05,
+      "loss": 0.2148,
+      "step": 642
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.3347110949406634,
+      "learning_rate": 2.1898552592673825e-05,
+      "loss": 0.215,
+      "step": 643
+    },
+    {
+      "epoch": 2.6116455696202534,
+      "grad_norm": 0.3430066215145567,
+      "learning_rate": 2.184204266048803e-05,
+      "loss": 0.1946,
+      "step": 644
+    },
+    {
+      "epoch": 2.6156962025316455,
+      "grad_norm": 0.3390920885487507,
+      "learning_rate": 2.1785517892743887e-05,
+      "loss": 0.2005,
+      "step": 645
+    },
+    {
+      "epoch": 2.619746835443038,
+      "grad_norm": 0.32044723130429975,
+      "learning_rate": 2.17289787446841e-05,
+      "loss": 0.1968,
+      "step": 646
+    },
+    {
+      "epoch": 2.6237974683544305,
+      "grad_norm": 0.33070269139983455,
+      "learning_rate": 2.1672425671667198e-05,
+      "loss": 0.1919,
+      "step": 647
+    },
+    {
+      "epoch": 2.6278481012658226,
+      "grad_norm": 0.319963706900542,
+      "learning_rate": 2.161585912916385e-05,
+      "loss": 0.2125,
+      "step": 648
+    },
+    {
+      "epoch": 2.631898734177215,
+      "grad_norm": 0.31255485293918556,
+      "learning_rate": 2.1559279572753214e-05,
+      "loss": 0.1858,
+      "step": 649
+    },
+    {
+      "epoch": 2.6359493670886076,
+      "grad_norm": 0.33302026745545166,
+      "learning_rate": 2.1502687458119268e-05,
+      "loss": 0.1963,
+      "step": 650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.33557752596410434,
+      "learning_rate": 2.1446083241047116e-05,
+      "loss": 0.1961,
+      "step": 651
+    },
+    {
+      "epoch": 2.6440506329113926,
+      "grad_norm": 0.35489643713149105,
+      "learning_rate": 2.1389467377419333e-05,
+      "loss": 0.1901,
+      "step": 652
+    },
+    {
+      "epoch": 2.6481012658227847,
+      "grad_norm": 0.3254129565791354,
+      "learning_rate": 2.133284032321232e-05,
+      "loss": 0.1917,
+      "step": 653
+    },
+    {
+      "epoch": 2.6521518987341772,
+      "grad_norm": 0.3612850074247574,
+      "learning_rate": 2.1276202534492566e-05,
+      "loss": 0.205,
+      "step": 654
+    },
+    {
+      "epoch": 2.6562025316455697,
+      "grad_norm": 0.3447458783070852,
+      "learning_rate": 2.121955446741306e-05,
+      "loss": 0.1996,
+      "step": 655
+    },
+    {
+      "epoch": 2.660253164556962,
+      "grad_norm": 0.3421435707062175,
+      "learning_rate": 2.1162896578209517e-05,
+      "loss": 0.1965,
+      "step": 656
+    },
+    {
+      "epoch": 2.6643037974683543,
+      "grad_norm": 0.34552642428463376,
+      "learning_rate": 2.1106229323196813e-05,
+      "loss": 0.2038,
+      "step": 657
+    },
+    {
+      "epoch": 2.668354430379747,
+      "grad_norm": 0.3424805252596436,
+      "learning_rate": 2.1049553158765214e-05,
+      "loss": 0.1929,
+      "step": 658
+    },
+    {
+      "epoch": 2.6724050632911394,
+      "grad_norm": 0.30681468669104367,
+      "learning_rate": 2.0992868541376764e-05,
+      "loss": 0.1965,
+      "step": 659
+    },
+    {
+      "epoch": 2.676455696202532,
+      "grad_norm": 0.3485653679267567,
+      "learning_rate": 2.093617592756158e-05,
+      "loss": 0.1893,
+      "step": 660
+    },
+    {
+      "epoch": 2.680506329113924,
+      "grad_norm": 0.322008785581921,
+      "learning_rate": 2.0879475773914167e-05,
+      "loss": 0.1985,
+      "step": 661
+    },
+    {
+      "epoch": 2.6845569620253165,
+      "grad_norm": 0.323040667063144,
+      "learning_rate": 2.082276853708978e-05,
+      "loss": 0.1986,
+      "step": 662
+    },
+    {
+      "epoch": 2.688607594936709,
+      "grad_norm": 0.3058178103836208,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.2006,
+      "step": 663
+    },
+    {
+      "epoch": 2.692658227848101,
+      "grad_norm": 0.3120458674692138,
+      "learning_rate": 2.0709334640812613e-05,
+      "loss": 0.1912,
+      "step": 664
+    },
+    {
+      "epoch": 2.6967088607594936,
+      "grad_norm": 0.3121647754400163,
+      "learning_rate": 2.0652608894940824e-05,
+      "loss": 0.2096,
+      "step": 665
+    },
+    {
+      "epoch": 2.700759493670886,
+      "grad_norm": 0.3023851883214024,
+      "learning_rate": 2.0595877893046722e-05,
+      "loss": 0.202,
+      "step": 666
+    },
+    {
+      "epoch": 2.7048101265822786,
+      "grad_norm": 0.31236086311836275,
+      "learning_rate": 2.0539142092033985e-05,
+      "loss": 0.2096,
+      "step": 667
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.33476167610743013,
+      "learning_rate": 2.048240194884496e-05,
+      "loss": 0.1933,
+      "step": 668
+    },
+    {
+      "epoch": 2.712911392405063,
+      "grad_norm": 0.3149104237364329,
+      "learning_rate": 2.042565792045695e-05,
+      "loss": 0.1904,
+      "step": 669
+    },
+    {
+      "epoch": 2.7169620253164557,
+      "grad_norm": 0.33096508309257455,
+      "learning_rate": 2.036891046387857e-05,
+      "loss": 0.1985,
+      "step": 670
+    },
+    {
+      "epoch": 2.721012658227848,
+      "grad_norm": 0.3300769049079634,
+      "learning_rate": 2.0312160036146036e-05,
+      "loss": 0.2179,
+      "step": 671
+    },
+    {
+      "epoch": 2.7250632911392403,
+      "grad_norm": 0.29609667729068895,
+      "learning_rate": 2.025540709431948e-05,
+      "loss": 0.2071,
+      "step": 672
+    },
+    {
+      "epoch": 2.729113924050633,
+      "grad_norm": 0.3256952188601732,
+      "learning_rate": 2.0198652095479298e-05,
+      "loss": 0.2051,
+      "step": 673
+    },
+    {
+      "epoch": 2.7331645569620253,
+      "grad_norm": 0.2960809977824228,
+      "learning_rate": 2.014189549672245e-05,
+      "loss": 0.1898,
+      "step": 674
+    },
+    {
+      "epoch": 2.737215189873418,
+      "grad_norm": 0.3242667984570822,
+      "learning_rate": 2.0085137755158776e-05,
+      "loss": 0.2045,
+      "step": 675
+    },
+    {
+      "epoch": 2.7412658227848103,
+      "grad_norm": 0.3189774505478913,
+      "learning_rate": 2.0028379327907327e-05,
+      "loss": 0.2052,
+      "step": 676
+    },
+    {
+      "epoch": 2.7453164556962024,
+      "grad_norm": 0.29799657686323505,
+      "learning_rate": 1.9971620672092676e-05,
+      "loss": 0.217,
+      "step": 677
+    },
+    {
+      "epoch": 2.749367088607595,
+      "grad_norm": 0.29871726871597004,
+      "learning_rate": 1.991486224484123e-05,
+      "loss": 0.1975,
+      "step": 678
+    },
+    {
+      "epoch": 2.7534177215189874,
+      "grad_norm": 0.3148566148846412,
+      "learning_rate": 1.985810450327756e-05,
+      "loss": 0.1993,
+      "step": 679
+    },
+    {
+      "epoch": 2.7574683544303795,
+      "grad_norm": 0.31442886270229525,
+      "learning_rate": 1.9801347904520706e-05,
+      "loss": 0.2064,
+      "step": 680
+    },
+    {
+      "epoch": 2.761518987341772,
+      "grad_norm": 0.27706901362171094,
+      "learning_rate": 1.974459290568053e-05,
+      "loss": 0.1968,
+      "step": 681
+    },
+    {
+      "epoch": 2.7655696202531646,
+      "grad_norm": 0.31777561906350066,
+      "learning_rate": 1.968783996385397e-05,
+      "loss": 0.1979,
+      "step": 682
+    },
+    {
+      "epoch": 2.769620253164557,
+      "grad_norm": 0.31852102345204336,
+      "learning_rate": 1.963108953612143e-05,
+      "loss": 0.2022,
+      "step": 683
+    },
+    {
+      "epoch": 2.7736708860759496,
+      "grad_norm": 0.30342593505504123,
+      "learning_rate": 1.9574342079543056e-05,
+      "loss": 0.1981,
+      "step": 684
+    },
+    {
+      "epoch": 2.7777215189873417,
+      "grad_norm": 0.2890998586530124,
+      "learning_rate": 1.9517598051155046e-05,
+      "loss": 0.1952,
+      "step": 685
+    },
+    {
+      "epoch": 2.781772151898734,
+      "grad_norm": 0.30260125498058643,
+      "learning_rate": 1.9460857907966025e-05,
+      "loss": 0.1983,
+      "step": 686
+    },
+    {
+      "epoch": 2.7858227848101267,
+      "grad_norm": 0.32455561891069057,
+      "learning_rate": 1.9404122106953285e-05,
+      "loss": 0.1857,
+      "step": 687
+    },
+    {
+      "epoch": 2.7898734177215188,
+      "grad_norm": 0.3189375295524177,
+      "learning_rate": 1.9347391105059176e-05,
+      "loss": 0.1985,
+      "step": 688
+    },
+    {
+      "epoch": 2.7939240506329113,
+      "grad_norm": 0.320895064889201,
+      "learning_rate": 1.92906653591874e-05,
+      "loss": 0.1977,
+      "step": 689
+    },
+    {
+      "epoch": 2.797974683544304,
+      "grad_norm": 0.307646227808254,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.1975,
+      "step": 690
+    },
+    {
+      "epoch": 2.8020253164556963,
+      "grad_norm": 0.3348932852479888,
+      "learning_rate": 1.917723146291022e-05,
+      "loss": 0.1874,
+      "step": 691
+    },
+    {
+      "epoch": 2.806075949367089,
+      "grad_norm": 0.3194431472477189,
+      "learning_rate": 1.912052422608584e-05,
+      "loss": 0.2035,
+      "step": 692
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.3217529634828901,
+      "learning_rate": 1.9063824072438428e-05,
+      "loss": 0.1973,
+      "step": 693
+    },
+    {
+      "epoch": 2.8141772151898734,
+      "grad_norm": 0.3405324859644441,
+      "learning_rate": 1.9007131458623246e-05,
+      "loss": 0.202,
+      "step": 694
+    },
+    {
+      "epoch": 2.818227848101266,
+      "grad_norm": 0.30139959194270854,
+      "learning_rate": 1.895044684123479e-05,
+      "loss": 0.1906,
+      "step": 695
+    },
+    {
+      "epoch": 2.822278481012658,
+      "grad_norm": 0.3291510934337519,
+      "learning_rate": 1.8893770676803194e-05,
+      "loss": 0.2083,
+      "step": 696
+    },
+    {
+      "epoch": 2.8263291139240505,
+      "grad_norm": 0.3308137240025994,
+      "learning_rate": 1.8837103421790486e-05,
+      "loss": 0.2048,
+      "step": 697
+    },
+    {
+      "epoch": 2.830379746835443,
+      "grad_norm": 0.3090196287075821,
+      "learning_rate": 1.8780445532586952e-05,
+      "loss": 0.2002,
+      "step": 698
+    },
+    {
+      "epoch": 2.8344303797468355,
+      "grad_norm": 0.3186398421283876,
+      "learning_rate": 1.872379746550743e-05,
+      "loss": 0.193,
+      "step": 699
+    },
+    {
+      "epoch": 2.838481012658228,
+      "grad_norm": 0.31875084954404237,
+      "learning_rate": 1.866715967678769e-05,
+      "loss": 0.2046,
+      "step": 700
+    },
+    {
+      "epoch": 2.84253164556962,
+      "grad_norm": 0.29960821588248854,
+      "learning_rate": 1.861053262258067e-05,
+      "loss": 0.1963,
+      "step": 701
+    },
+    {
+      "epoch": 2.8465822784810126,
+      "grad_norm": 0.323979264041461,
+      "learning_rate": 1.8553916758952897e-05,
+      "loss": 0.2047,
+      "step": 702
+    },
+    {
+      "epoch": 2.850632911392405,
+      "grad_norm": 0.3142193162420491,
+      "learning_rate": 1.8497312541880735e-05,
+      "loss": 0.1897,
+      "step": 703
+    },
+    {
+      "epoch": 2.8546835443037972,
+      "grad_norm": 0.31916618773626393,
+      "learning_rate": 1.8440720427246786e-05,
+      "loss": 0.1959,
+      "step": 704
+    },
+    {
+      "epoch": 2.8587341772151897,
+      "grad_norm": 0.33050339165455367,
+      "learning_rate": 1.8384140870836157e-05,
+      "loss": 0.19,
+      "step": 705
+    },
+    {
+      "epoch": 2.8627848101265823,
+      "grad_norm": 0.3183430293635347,
+      "learning_rate": 1.8327574328332806e-05,
+      "loss": 0.211,
+      "step": 706
+    },
+    {
+      "epoch": 2.8668354430379748,
+      "grad_norm": 0.33822822057585605,
+      "learning_rate": 1.8271021255315906e-05,
+      "loss": 0.2117,
+      "step": 707
+    },
+    {
+      "epoch": 2.8708860759493673,
+      "grad_norm": 0.32464184299921345,
+      "learning_rate": 1.8214482107256117e-05,
+      "loss": 0.1888,
+      "step": 708
+    },
+    {
+      "epoch": 2.8749367088607594,
+      "grad_norm": 0.3122466497889213,
+      "learning_rate": 1.8157957339511968e-05,
+      "loss": 0.2055,
+      "step": 709
+    },
+    {
+      "epoch": 2.878987341772152,
+      "grad_norm": 0.35965723956248236,
+      "learning_rate": 1.8101447407326182e-05,
+      "loss": 0.2013,
+      "step": 710
+    },
+    {
+      "epoch": 2.8830379746835444,
+      "grad_norm": 0.28932911118878707,
+      "learning_rate": 1.8044952765821966e-05,
+      "loss": 0.1977,
+      "step": 711
+    },
+    {
+      "epoch": 2.8870886075949365,
+      "grad_norm": 0.3002999316944365,
+      "learning_rate": 1.7988473869999407e-05,
+      "loss": 0.193,
+      "step": 712
+    },
+    {
+      "epoch": 2.891139240506329,
+      "grad_norm": 0.3318882102487256,
+      "learning_rate": 1.7932011174731764e-05,
+      "loss": 0.1982,
+      "step": 713
+    },
+    {
+      "epoch": 2.8951898734177215,
+      "grad_norm": 0.3013508885439591,
+      "learning_rate": 1.7875565134761817e-05,
+      "loss": 0.207,
+      "step": 714
+    },
+    {
+      "epoch": 2.899240506329114,
+      "grad_norm": 0.3179845609479174,
+      "learning_rate": 1.7819136204698226e-05,
+      "loss": 0.19,
+      "step": 715
+    },
+    {
+      "epoch": 2.9032911392405065,
+      "grad_norm": 0.321260900893894,
+      "learning_rate": 1.776272483901182e-05,
+      "loss": 0.1892,
+      "step": 716
+    },
+    {
+      "epoch": 2.9073417721518986,
+      "grad_norm": 0.3102736347101959,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.1953,
+      "step": 717
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.28080805294832417,
+      "learning_rate": 1.764995661794301e-05,
+      "loss": 0.1897,
+      "step": 718
+    },
+    {
+      "epoch": 2.9154430379746836,
+      "grad_norm": 0.3372050797733071,
+      "learning_rate": 1.759360067078035e-05,
+      "loss": 0.199,
+      "step": 719
+    },
+    {
+      "epoch": 2.9194936708860757,
+      "grad_norm": 0.3162673119453784,
+      "learning_rate": 1.7537264104427064e-05,
+      "loss": 0.2063,
+      "step": 720
+    },
+    {
+      "epoch": 2.923544303797468,
+      "grad_norm": 0.3231629814452705,
+      "learning_rate": 1.748094737261012e-05,
+      "loss": 0.2071,
+      "step": 721
+    },
+    {
+      "epoch": 2.9275949367088607,
+      "grad_norm": 0.3139933588230321,
+      "learning_rate": 1.7424650928896726e-05,
+      "loss": 0.1985,
+      "step": 722
+    },
+    {
+      "epoch": 2.9316455696202532,
+      "grad_norm": 0.2940268116025611,
+      "learning_rate": 1.7368375226690712e-05,
+      "loss": 0.1949,
+      "step": 723
+    },
+    {
+      "epoch": 2.9356962025316458,
+      "grad_norm": 0.30627806472462227,
+      "learning_rate": 1.731212071922883e-05,
+      "loss": 0.2043,
+      "step": 724
+    },
+    {
+      "epoch": 2.939746835443038,
+      "grad_norm": 0.3134443249929089,
+      "learning_rate": 1.7255887859577156e-05,
+      "loss": 0.192,
+      "step": 725
+    },
+    {
+      "epoch": 2.9437974683544303,
+      "grad_norm": 0.30729040784015443,
+      "learning_rate": 1.7199677100627427e-05,
+      "loss": 0.2065,
+      "step": 726
+    },
+    {
+      "epoch": 2.947848101265823,
+      "grad_norm": 0.2842068636411367,
+      "learning_rate": 1.7143488895093343e-05,
+      "loss": 0.1928,
+      "step": 727
+    },
+    {
+      "epoch": 2.951898734177215,
+      "grad_norm": 0.3113063349657297,
+      "learning_rate": 1.7087323695506994e-05,
+      "loss": 0.2118,
+      "step": 728
+    },
+    {
+      "epoch": 2.9559493670886074,
+      "grad_norm": 0.31408640309340813,
+      "learning_rate": 1.7031181954215194e-05,
+      "loss": 0.1918,
+      "step": 729
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3025629885351385,
+      "learning_rate": 1.6975064123375788e-05,
+      "loss": 0.1921,
+      "step": 730
+    },
+    {
+      "epoch": 2.9640506329113925,
+      "grad_norm": 0.3114886029816687,
+      "learning_rate": 1.6918970654954084e-05,
+      "loss": 0.2006,
+      "step": 731
+    },
+    {
+      "epoch": 2.968101265822785,
+      "grad_norm": 0.29181736675682396,
+      "learning_rate": 1.686290200071915e-05,
+      "loss": 0.1793,
+      "step": 732
+    },
+    {
+      "epoch": 2.972151898734177,
+      "grad_norm": 0.2942888611269011,
+      "learning_rate": 1.6806858612240234e-05,
+      "loss": 0.1966,
+      "step": 733
+    },
+    {
+      "epoch": 2.9762025316455696,
+      "grad_norm": 0.31979070304837603,
+      "learning_rate": 1.6750840940883078e-05,
+      "loss": 0.2014,
+      "step": 734
+    },
+    {
+      "epoch": 2.980253164556962,
+      "grad_norm": 0.31518577193379116,
+      "learning_rate": 1.6694849437806305e-05,
+      "loss": 0.2058,
+      "step": 735
+    },
+    {
+      "epoch": 2.984303797468354,
+      "grad_norm": 0.3071461661771092,
+      "learning_rate": 1.663888455395778e-05,
+      "loss": 0.1985,
+      "step": 736
+    },
+    {
+      "epoch": 2.9883544303797467,
+      "grad_norm": 0.3074028628827926,
+      "learning_rate": 1.6582946740070995e-05,
+      "loss": 0.1985,
+      "step": 737
+    },
+    {
+      "epoch": 2.992405063291139,
+      "grad_norm": 0.3167514507451893,
+      "learning_rate": 1.6527036446661396e-05,
+      "loss": 0.2099,
+      "step": 738
+    },
+    {
+      "epoch": 2.9964556962025317,
+      "grad_norm": 0.30315899760811343,
+      "learning_rate": 1.6471154124022818e-05,
+      "loss": 0.1874,
+      "step": 739
+    },
+    {
+      "epoch": 3.002025316455696,
+      "grad_norm": 0.3062956605444015,
+      "learning_rate": 1.6415300222223788e-05,
+      "loss": 0.1689,
+      "step": 740
+    },
+    {
+      "epoch": 3.0060759493670886,
+      "grad_norm": 0.35864522121570913,
+      "learning_rate": 1.6359475191103958e-05,
+      "loss": 0.1419,
+      "step": 741
+    },
+    {
+      "epoch": 3.010126582278481,
+      "grad_norm": 0.3015164203166217,
+      "learning_rate": 1.6303679480270466e-05,
+      "loss": 0.1272,
+      "step": 742
+    },
+    {
+      "epoch": 3.0141772151898736,
+      "grad_norm": 0.4301621233689881,
+      "learning_rate": 1.624791353909428e-05,
+      "loss": 0.1306,
+      "step": 743
+    },
+    {
+      "epoch": 3.0182278481012657,
+      "grad_norm": 0.48392351429421987,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.1277,
+      "step": 744
+    },
+    {
+      "epoch": 3.022278481012658,
+      "grad_norm": 0.3300731915513254,
+      "learning_rate": 1.6136472761995373e-05,
+      "loss": 0.1395,
+      "step": 745
+    },
+    {
+      "epoch": 3.0263291139240507,
+      "grad_norm": 0.36427187886488505,
+      "learning_rate": 1.608079882360134e-05,
+      "loss": 0.1368,
+      "step": 746
+    },
+    {
+      "epoch": 3.030379746835443,
+      "grad_norm": 0.383109983658657,
+      "learning_rate": 1.60251564499148e-05,
+      "loss": 0.1349,
+      "step": 747
+    },
+    {
+      "epoch": 3.0344303797468353,
+      "grad_norm": 0.3263736403002393,
+      "learning_rate": 1.596954608907176e-05,
+      "loss": 0.121,
+      "step": 748
+    },
+    {
+      "epoch": 3.038481012658228,
+      "grad_norm": 0.3112883582672956,
+      "learning_rate": 1.591396818895043e-05,
+      "loss": 0.1281,
+      "step": 749
+    },
+    {
+      "epoch": 3.0425316455696203,
+      "grad_norm": 0.3335279505016545,
+      "learning_rate": 1.585842319716759e-05,
+      "loss": 0.1324,
+      "step": 750
+    },
+    {
+      "epoch": 3.046582278481013,
+      "grad_norm": 0.36411419138017853,
+      "learning_rate": 1.5802911561074944e-05,
+      "loss": 0.1366,
+      "step": 751
+    },
+    {
+      "epoch": 3.050632911392405,
+      "grad_norm": 0.32437299457231905,
+      "learning_rate": 1.5747433727755595e-05,
+      "loss": 0.1304,
+      "step": 752
+    },
+    {
+      "epoch": 3.0546835443037974,
+      "grad_norm": 0.32087269334326746,
+      "learning_rate": 1.5691990144020376e-05,
+      "loss": 0.1257,
+      "step": 753
+    },
+    {
+      "epoch": 3.05873417721519,
+      "grad_norm": 0.3150770741637823,
+      "learning_rate": 1.5636581256404297e-05,
+      "loss": 0.1329,
+      "step": 754
+    },
+    {
+      "epoch": 3.0627848101265824,
+      "grad_norm": 0.31626443963070494,
+      "learning_rate": 1.558120751116291e-05,
+      "loss": 0.1282,
+      "step": 755
+    },
+    {
+      "epoch": 3.0668354430379745,
+      "grad_norm": 0.3019302872144577,
+      "learning_rate": 1.552586935426876e-05,
+      "loss": 0.1282,
+      "step": 756
+    },
+    {
+      "epoch": 3.070886075949367,
+      "grad_norm": 0.32190000235820304,
+      "learning_rate": 1.547056723140774e-05,
+      "loss": 0.1442,
+      "step": 757
+    },
+    {
+      "epoch": 3.0749367088607595,
+      "grad_norm": 0.3118653769363254,
+      "learning_rate": 1.5415301587975565e-05,
+      "loss": 0.143,
+      "step": 758
+    },
+    {
+      "epoch": 3.078987341772152,
+      "grad_norm": 0.30888286376514995,
+      "learning_rate": 1.536007286907411e-05,
+      "loss": 0.1333,
+      "step": 759
+    },
+    {
+      "epoch": 3.083037974683544,
+      "grad_norm": 0.30631423068166563,
+      "learning_rate": 1.5304881519507896e-05,
+      "loss": 0.1354,
+      "step": 760
+    },
+    {
+      "epoch": 3.0870886075949366,
+      "grad_norm": 0.32496895120247704,
+      "learning_rate": 1.5249727983780453e-05,
+      "loss": 0.1372,
+      "step": 761
+    },
+    {
+      "epoch": 3.091139240506329,
+      "grad_norm": 0.2972977229173278,
+      "learning_rate": 1.5194612706090786e-05,
+      "loss": 0.1204,
+      "step": 762
+    },
+    {
+      "epoch": 3.0951898734177217,
+      "grad_norm": 0.31693298774854817,
+      "learning_rate": 1.5139536130329771e-05,
+      "loss": 0.1368,
+      "step": 763
+    },
+    {
+      "epoch": 3.0992405063291137,
+      "grad_norm": 0.30951515282144026,
+      "learning_rate": 1.508449870007656e-05,
+      "loss": 0.1334,
+      "step": 764
+    },
+    {
+      "epoch": 3.1032911392405063,
+      "grad_norm": 0.30856024185451575,
+      "learning_rate": 1.5029500858595056e-05,
+      "loss": 0.1295,
+      "step": 765
+    },
+    {
+      "epoch": 3.1073417721518988,
+      "grad_norm": 0.29582414502630744,
+      "learning_rate": 1.4974543048830328e-05,
+      "loss": 0.139,
+      "step": 766
+    },
+    {
+      "epoch": 3.1113924050632913,
+      "grad_norm": 0.3060211205466185,
+      "learning_rate": 1.4919625713405e-05,
+      "loss": 0.1319,
+      "step": 767
+    },
+    {
+      "epoch": 3.1154430379746834,
+      "grad_norm": 0.322278382498117,
+      "learning_rate": 1.4864749294615756e-05,
+      "loss": 0.1409,
+      "step": 768
+    },
+    {
+      "epoch": 3.119493670886076,
+      "grad_norm": 0.3060174581050804,
+      "learning_rate": 1.4809914234429716e-05,
+      "loss": 0.1204,
+      "step": 769
+    },
+    {
+      "epoch": 3.1235443037974684,
+      "grad_norm": 0.29362603780412716,
+      "learning_rate": 1.4755120974480923e-05,
+      "loss": 0.1339,
+      "step": 770
+    },
+    {
+      "epoch": 3.127594936708861,
+      "grad_norm": 0.29847795935433724,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.1299,
+      "step": 771
+    },
+    {
+      "epoch": 3.131645569620253,
+      "grad_norm": 0.30299506672681564,
+      "learning_rate": 1.4645661620144413e-05,
+      "loss": 0.1206,
+      "step": 772
+    },
+    {
+      "epoch": 3.1356962025316455,
+      "grad_norm": 0.29534704459643846,
+      "learning_rate": 1.4590996407327284e-05,
+      "loss": 0.1354,
+      "step": 773
+    },
+    {
+      "epoch": 3.139746835443038,
+      "grad_norm": 0.30870420588314784,
+      "learning_rate": 1.4536374757881487e-05,
+      "loss": 0.1251,
+      "step": 774
+    },
+    {
+      "epoch": 3.1437974683544305,
+      "grad_norm": 0.3036145611971784,
+      "learning_rate": 1.4481797111722271e-05,
+      "loss": 0.1183,
+      "step": 775
+    },
+    {
+      "epoch": 3.1478481012658226,
+      "grad_norm": 0.2825989381926883,
+      "learning_rate": 1.4427263908410507e-05,
+      "loss": 0.1307,
+      "step": 776
+    },
+    {
+      "epoch": 3.151898734177215,
+      "grad_norm": 0.3342287380287574,
+      "learning_rate": 1.4372775587149108e-05,
+      "loss": 0.1297,
+      "step": 777
+    },
+    {
+      "epoch": 3.1559493670886076,
+      "grad_norm": 0.3140556344885269,
+      "learning_rate": 1.4318332586779522e-05,
+      "loss": 0.1355,
+      "step": 778
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.2996586085115626,
+      "learning_rate": 1.4263935345778202e-05,
+      "loss": 0.1343,
+      "step": 779
+    },
+    {
+      "epoch": 3.164050632911392,
+      "grad_norm": 0.3346680082703197,
+      "learning_rate": 1.420958430225303e-05,
+      "loss": 0.1326,
+      "step": 780
+    },
+    {
+      "epoch": 3.1681012658227847,
+      "grad_norm": 0.3012627406523601,
+      "learning_rate": 1.415527989393985e-05,
+      "loss": 0.1331,
+      "step": 781
+    },
+    {
+      "epoch": 3.1721518987341772,
+      "grad_norm": 0.33387372358035694,
+      "learning_rate": 1.410102255819891e-05,
+      "loss": 0.1272,
+      "step": 782
+    },
+    {
+      "epoch": 3.1762025316455698,
+      "grad_norm": 0.30996373840610947,
+      "learning_rate": 1.404681273201131e-05,
+      "loss": 0.1319,
+      "step": 783
+    },
+    {
+      "epoch": 3.180253164556962,
+      "grad_norm": 0.3370976626828798,
+      "learning_rate": 1.399265085197556e-05,
+      "loss": 0.123,
+      "step": 784
+    },
+    {
+      "epoch": 3.1843037974683543,
+      "grad_norm": 0.30805456664001835,
+      "learning_rate": 1.393853735430398e-05,
+      "loss": 0.1258,
+      "step": 785
+    },
+    {
+      "epoch": 3.188354430379747,
+      "grad_norm": 0.3075188241158418,
+      "learning_rate": 1.3884472674819246e-05,
+      "loss": 0.131,
+      "step": 786
+    },
+    {
+      "epoch": 3.1924050632911394,
+      "grad_norm": 0.3101211085235056,
+      "learning_rate": 1.3830457248950864e-05,
+      "loss": 0.13,
+      "step": 787
+    },
+    {
+      "epoch": 3.1964556962025314,
+      "grad_norm": 0.28834372349273996,
+      "learning_rate": 1.377649151173163e-05,
+      "loss": 0.1223,
+      "step": 788
+    },
+    {
+      "epoch": 3.200506329113924,
+      "grad_norm": 0.3097366667047267,
+      "learning_rate": 1.3722575897794181e-05,
+      "loss": 0.1383,
+      "step": 789
+    },
+    {
+      "epoch": 3.2045569620253165,
+      "grad_norm": 0.28696478129809244,
+      "learning_rate": 1.3668710841367472e-05,
+      "loss": 0.1328,
+      "step": 790
+    },
+    {
+      "epoch": 3.208607594936709,
+      "grad_norm": 0.3017479552791321,
+      "learning_rate": 1.361489677627324e-05,
+      "loss": 0.1249,
+      "step": 791
+    },
+    {
+      "epoch": 3.212658227848101,
+      "grad_norm": 0.30424097263757377,
+      "learning_rate": 1.3561134135922585e-05,
+      "loss": 0.1346,
+      "step": 792
+    },
+    {
+      "epoch": 3.2167088607594936,
+      "grad_norm": 0.2951324252809513,
+      "learning_rate": 1.350742335331241e-05,
+      "loss": 0.12,
+      "step": 793
+    },
+    {
+      "epoch": 3.220759493670886,
+      "grad_norm": 0.30497115383322515,
+      "learning_rate": 1.345376486102198e-05,
+      "loss": 0.1301,
+      "step": 794
+    },
+    {
+      "epoch": 3.2248101265822786,
+      "grad_norm": 0.31427085072569777,
+      "learning_rate": 1.3400159091209414e-05,
+      "loss": 0.1221,
+      "step": 795
+    },
+    {
+      "epoch": 3.2288607594936707,
+      "grad_norm": 0.31337511592821216,
+      "learning_rate": 1.3346606475608216e-05,
+      "loss": 0.1277,
+      "step": 796
+    },
+    {
+      "epoch": 3.232911392405063,
+      "grad_norm": 0.29351506048821135,
+      "learning_rate": 1.3293107445523781e-05,
+      "loss": 0.1298,
+      "step": 797
+    },
+    {
+      "epoch": 3.2369620253164557,
+      "grad_norm": 0.31890562676123874,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.1367,
+      "step": 798
+    },
+    {
+      "epoch": 3.2410126582278482,
+      "grad_norm": 0.2918245950885417,
+      "learning_rate": 1.3186271864965509e-05,
+      "loss": 0.1258,
+      "step": 799
+    },
+    {
+      "epoch": 3.2450632911392403,
+      "grad_norm": 0.2959254656292249,
+      "learning_rate": 1.3132936174930756e-05,
+      "loss": 0.1383,
+      "step": 800
+    },
+    {
+      "epoch": 3.249113924050633,
+      "grad_norm": 0.30890811991204187,
+      "learning_rate": 1.3079655791283995e-05,
+      "loss": 0.132,
+      "step": 801
+    },
+    {
+      "epoch": 3.2531645569620253,
+      "grad_norm": 0.3033423263857011,
+      "learning_rate": 1.3026431143138108e-05,
+      "loss": 0.1352,
+      "step": 802
+    },
+    {
+      "epoch": 3.257215189873418,
+      "grad_norm": 0.28887772347139207,
+      "learning_rate": 1.2973262659157114e-05,
+      "loss": 0.1316,
+      "step": 803
+    },
+    {
+      "epoch": 3.26126582278481,
+      "grad_norm": 0.2852106280501523,
+      "learning_rate": 1.2920150767552651e-05,
+      "loss": 0.1376,
+      "step": 804
+    },
+    {
+      "epoch": 3.2653164556962024,
+      "grad_norm": 0.3108315176110905,
+      "learning_rate": 1.2867095896080607e-05,
+      "loss": 0.1255,
+      "step": 805
+    },
+    {
+      "epoch": 3.269367088607595,
+      "grad_norm": 0.27871329954885027,
+      "learning_rate": 1.2814098472037612e-05,
+      "loss": 0.1305,
+      "step": 806
+    },
+    {
+      "epoch": 3.2734177215189875,
+      "grad_norm": 0.30289276869891685,
+      "learning_rate": 1.276115892225764e-05,
+      "loss": 0.1305,
+      "step": 807
+    },
+    {
+      "epoch": 3.27746835443038,
+      "grad_norm": 0.31464273155855466,
+      "learning_rate": 1.2708277673108555e-05,
+      "loss": 0.1386,
+      "step": 808
+    },
+    {
+      "epoch": 3.281518987341772,
+      "grad_norm": 0.3328896679250284,
+      "learning_rate": 1.2655455150488649e-05,
+      "loss": 0.1281,
+      "step": 809
+    },
+    {
+      "epoch": 3.2855696202531646,
+      "grad_norm": 0.28953979456954376,
+      "learning_rate": 1.2602691779823272e-05,
+      "loss": 0.1465,
+      "step": 810
+    },
+    {
+      "epoch": 3.289620253164557,
+      "grad_norm": 0.308173337581417,
+      "learning_rate": 1.2549987986061355e-05,
+      "loss": 0.1358,
+      "step": 811
+    },
+    {
+      "epoch": 3.293670886075949,
+      "grad_norm": 0.3044136598780952,
+      "learning_rate": 1.2497344193672005e-05,
+      "loss": 0.1404,
+      "step": 812
+    },
+    {
+      "epoch": 3.2977215189873417,
+      "grad_norm": 0.29465627926861887,
+      "learning_rate": 1.2444760826641092e-05,
+      "loss": 0.1226,
+      "step": 813
+    },
+    {
+      "epoch": 3.301772151898734,
+      "grad_norm": 0.2964643293687634,
+      "learning_rate": 1.2392238308467817e-05,
+      "loss": 0.1332,
+      "step": 814
+    },
+    {
+      "epoch": 3.3058227848101267,
+      "grad_norm": 0.2920750743083762,
+      "learning_rate": 1.2339777062161326e-05,
+      "loss": 0.1256,
+      "step": 815
+    },
+    {
+      "epoch": 3.309873417721519,
+      "grad_norm": 0.3207080800304372,
+      "learning_rate": 1.2287377510237293e-05,
+      "loss": 0.1224,
+      "step": 816
+    },
+    {
+      "epoch": 3.3139240506329113,
+      "grad_norm": 0.3091326894192946,
+      "learning_rate": 1.2235040074714488e-05,
+      "loss": 0.1344,
+      "step": 817
+    },
+    {
+      "epoch": 3.317974683544304,
+      "grad_norm": 0.2822478003766466,
+      "learning_rate": 1.2182765177111434e-05,
+      "loss": 0.127,
+      "step": 818
+    },
+    {
+      "epoch": 3.3220253164556963,
+      "grad_norm": 0.29926832736430103,
+      "learning_rate": 1.213055323844297e-05,
+      "loss": 0.1228,
+      "step": 819
+    },
+    {
+      "epoch": 3.3260759493670884,
+      "grad_norm": 0.28489288805575014,
+      "learning_rate": 1.2078404679216864e-05,
+      "loss": 0.1297,
+      "step": 820
+    },
+    {
+      "epoch": 3.330126582278481,
+      "grad_norm": 0.29912931097272577,
+      "learning_rate": 1.2026319919430458e-05,
+      "loss": 0.1212,
+      "step": 821
+    },
+    {
+      "epoch": 3.3341772151898734,
+      "grad_norm": 0.29071469315869075,
+      "learning_rate": 1.1974299378567227e-05,
+      "loss": 0.1308,
+      "step": 822
+    },
+    {
+      "epoch": 3.338227848101266,
+      "grad_norm": 0.2903724513387322,
+      "learning_rate": 1.1922343475593462e-05,
+      "loss": 0.1332,
+      "step": 823
+    },
+    {
+      "epoch": 3.3422784810126585,
+      "grad_norm": 0.2975119052030087,
+      "learning_rate": 1.187045262895488e-05,
+      "loss": 0.1295,
+      "step": 824
+    },
+    {
+      "epoch": 3.3463291139240505,
+      "grad_norm": 0.2732078074854909,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.1333,
+      "step": 825
+    },
+    {
+      "epoch": 3.350379746835443,
+      "grad_norm": 0.2828951353698092,
+      "learning_rate": 1.1766867775842864e-05,
+      "loss": 0.1182,
+      "step": 826
+    },
+    {
+      "epoch": 3.3544303797468356,
+      "grad_norm": 0.2782452653086802,
+      "learning_rate": 1.1715174603627615e-05,
+      "loss": 0.1197,
+      "step": 827
+    },
+    {
+      "epoch": 3.3584810126582276,
+      "grad_norm": 0.3004699782467339,
+      "learning_rate": 1.1663548156257147e-05,
+      "loss": 0.1395,
+      "step": 828
+    },
+    {
+      "epoch": 3.36253164556962,
+      "grad_norm": 0.2936606260594329,
+      "learning_rate": 1.161198884952377e-05,
+      "loss": 0.134,
+      "step": 829
+    },
+    {
+      "epoch": 3.3665822784810127,
+      "grad_norm": 0.28748051044261835,
+      "learning_rate": 1.1560497098679056e-05,
+      "loss": 0.131,
+      "step": 830
+    },
+    {
+      "epoch": 3.370632911392405,
+      "grad_norm": 0.2924895607185613,
+      "learning_rate": 1.1509073318430479e-05,
+      "loss": 0.1406,
+      "step": 831
+    },
+    {
+      "epoch": 3.3746835443037977,
+      "grad_norm": 0.30055704186543974,
+      "learning_rate": 1.1457717922938116e-05,
+      "loss": 0.1278,
+      "step": 832
+    },
+    {
+      "epoch": 3.3787341772151898,
+      "grad_norm": 0.30088317199949316,
+      "learning_rate": 1.1406431325811233e-05,
+      "loss": 0.131,
+      "step": 833
+    },
+    {
+      "epoch": 3.3827848101265823,
+      "grad_norm": 0.2932622304939579,
+      "learning_rate": 1.135521394010506e-05,
+      "loss": 0.1354,
+      "step": 834
+    },
+    {
+      "epoch": 3.386835443037975,
+      "grad_norm": 0.3028730688582558,
+      "learning_rate": 1.1304066178317367e-05,
+      "loss": 0.1295,
+      "step": 835
+    },
+    {
+      "epoch": 3.390886075949367,
+      "grad_norm": 0.31536782914783235,
+      "learning_rate": 1.1252988452385199e-05,
+      "loss": 0.1177,
+      "step": 836
+    },
+    {
+      "epoch": 3.3949367088607594,
+      "grad_norm": 0.295215784008639,
+      "learning_rate": 1.1201981173681536e-05,
+      "loss": 0.118,
+      "step": 837
+    },
+    {
+      "epoch": 3.398987341772152,
+      "grad_norm": 0.2953852589732501,
+      "learning_rate": 1.1151044753011991e-05,
+      "loss": 0.1226,
+      "step": 838
+    },
+    {
+      "epoch": 3.4030379746835444,
+      "grad_norm": 0.2895435018777221,
+      "learning_rate": 1.1100179600611491e-05,
+      "loss": 0.1265,
+      "step": 839
+    },
+    {
+      "epoch": 3.407088607594937,
+      "grad_norm": 0.29210014093237585,
+      "learning_rate": 1.1049386126140985e-05,
+      "loss": 0.1312,
+      "step": 840
+    },
+    {
+      "epoch": 3.411139240506329,
+      "grad_norm": 0.29135684035727094,
+      "learning_rate": 1.0998664738684128e-05,
+      "loss": 0.1328,
+      "step": 841
+    },
+    {
+      "epoch": 3.4151898734177215,
+      "grad_norm": 0.30799356841412695,
+      "learning_rate": 1.0948015846744e-05,
+      "loss": 0.1372,
+      "step": 842
+    },
+    {
+      "epoch": 3.419240506329114,
+      "grad_norm": 0.3146630783417251,
+      "learning_rate": 1.0897439858239832e-05,
+      "loss": 0.1514,
+      "step": 843
+    },
+    {
+      "epoch": 3.423291139240506,
+      "grad_norm": 0.2915032254265693,
+      "learning_rate": 1.0846937180503652e-05,
+      "loss": 0.1218,
+      "step": 844
+    },
+    {
+      "epoch": 3.4273417721518986,
+      "grad_norm": 0.29026404343346657,
+      "learning_rate": 1.0796508220277117e-05,
+      "loss": 0.118,
+      "step": 845
+    },
+    {
+      "epoch": 3.431392405063291,
+      "grad_norm": 0.3043909009673083,
+      "learning_rate": 1.0746153383708107e-05,
+      "loss": 0.1287,
+      "step": 846
+    },
+    {
+      "epoch": 3.4354430379746836,
+      "grad_norm": 0.30774924581301477,
+      "learning_rate": 1.0695873076347579e-05,
+      "loss": 0.1333,
+      "step": 847
+    },
+    {
+      "epoch": 3.439493670886076,
+      "grad_norm": 0.2907695256270274,
+      "learning_rate": 1.0645667703146205e-05,
+      "loss": 0.122,
+      "step": 848
+    },
+    {
+      "epoch": 3.4435443037974682,
+      "grad_norm": 0.30791686781723593,
+      "learning_rate": 1.0595537668451161e-05,
+      "loss": 0.1228,
+      "step": 849
+    },
+    {
+      "epoch": 3.4475949367088607,
+      "grad_norm": 0.294034288185776,
+      "learning_rate": 1.0545483376002854e-05,
+      "loss": 0.1319,
+      "step": 850
+    },
+    {
+      "epoch": 3.4516455696202533,
+      "grad_norm": 0.27597449319571354,
+      "learning_rate": 1.0495505228931676e-05,
+      "loss": 0.1382,
+      "step": 851
+    },
+    {
+      "epoch": 3.4556962025316453,
+      "grad_norm": 0.2993513112737985,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.1198,
+      "step": 852
+    },
+    {
+      "epoch": 3.459746835443038,
+      "grad_norm": 0.2967751420246856,
+      "learning_rate": 1.0395778980372695e-05,
+      "loss": 0.1095,
+      "step": 853
+    },
+    {
+      "epoch": 3.4637974683544304,
+      "grad_norm": 0.29734541730891156,
+      "learning_rate": 1.0346031682066381e-05,
+      "loss": 0.1313,
+      "step": 854
+    },
+    {
+      "epoch": 3.467848101265823,
+      "grad_norm": 0.2975251397541841,
+      "learning_rate": 1.0296362135493724e-05,
+      "loss": 0.1295,
+      "step": 855
+    },
+    {
+      "epoch": 3.4718987341772154,
+      "grad_norm": 0.28544144505077157,
+      "learning_rate": 1.0246770740686422e-05,
+      "loss": 0.128,
+      "step": 856
+    },
+    {
+      "epoch": 3.4759493670886075,
+      "grad_norm": 0.29918270651852163,
+      "learning_rate": 1.0197257897046743e-05,
+      "loss": 0.1174,
+      "step": 857
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.3050156399763809,
+      "learning_rate": 1.014782400334433e-05,
+      "loss": 0.1336,
+      "step": 858
+    },
+    {
+      "epoch": 3.4840506329113925,
+      "grad_norm": 0.29184243328970194,
+      "learning_rate": 1.009846945771296e-05,
+      "loss": 0.1395,
+      "step": 859
+    },
+    {
+      "epoch": 3.4881012658227846,
+      "grad_norm": 0.2813675148886344,
+      "learning_rate": 1.0049194657647363e-05,
+      "loss": 0.1193,
+      "step": 860
+    },
+    {
+      "epoch": 3.492151898734177,
+      "grad_norm": 0.28931314438302724,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.1348,
+      "step": 861
+    },
+    {
+      "epoch": 3.4962025316455696,
+      "grad_norm": 0.29469496972046616,
+      "learning_rate": 9.950885880977891e-06,
+      "loss": 0.1283,
+      "step": 862
+    },
+    {
+      "epoch": 3.500253164556962,
+      "grad_norm": 0.27557547167667273,
+      "learning_rate": 9.901852696139382e-06,
+      "loss": 0.1159,
+      "step": 863
+    },
+    {
+      "epoch": 3.5043037974683546,
+      "grad_norm": 0.27447706011004325,
+      "learning_rate": 9.852900840391027e-06,
+      "loss": 0.131,
+      "step": 864
+    },
+    {
+      "epoch": 3.5083544303797467,
+      "grad_norm": 0.3065023209085679,
+      "learning_rate": 9.804030707984313e-06,
+      "loss": 0.1215,
+      "step": 865
+    },
+    {
+      "epoch": 3.512405063291139,
+      "grad_norm": 0.28565212785296695,
+      "learning_rate": 9.755242692512599e-06,
+      "loss": 0.1237,
+      "step": 866
+    },
+    {
+      "epoch": 3.5164556962025317,
+      "grad_norm": 0.2825999459082425,
+      "learning_rate": 9.70653718690782e-06,
+      "loss": 0.1348,
+      "step": 867
+    },
+    {
+      "epoch": 3.520506329113924,
+      "grad_norm": 0.28675727887067504,
+      "learning_rate": 9.657914583437454e-06,
+      "loss": 0.1313,
+      "step": 868
+    },
+    {
+      "epoch": 3.5245569620253163,
+      "grad_norm": 0.2907328173736073,
+      "learning_rate": 9.609375273701246e-06,
+      "loss": 0.1308,
+      "step": 869
+    },
+    {
+      "epoch": 3.528607594936709,
+      "grad_norm": 0.2888731560797771,
+      "learning_rate": 9.560919648628133e-06,
+      "loss": 0.1265,
+      "step": 870
+    },
+    {
+      "epoch": 3.5326582278481014,
+      "grad_norm": 0.30093923805615463,
+      "learning_rate": 9.512548098473047e-06,
+      "loss": 0.1346,
+      "step": 871
+    },
+    {
+      "epoch": 3.536708860759494,
+      "grad_norm": 0.28931765437115287,
+      "learning_rate": 9.464261012813825e-06,
+      "loss": 0.1277,
+      "step": 872
+    },
+    {
+      "epoch": 3.540759493670886,
+      "grad_norm": 0.28823011713765206,
+      "learning_rate": 9.416058780547987e-06,
+      "loss": 0.1142,
+      "step": 873
+    },
+    {
+      "epoch": 3.5448101265822785,
+      "grad_norm": 0.28273390713834357,
+      "learning_rate": 9.367941789889714e-06,
+      "loss": 0.1322,
+      "step": 874
+    },
+    {
+      "epoch": 3.548860759493671,
+      "grad_norm": 0.2929438496178239,
+      "learning_rate": 9.319910428366607e-06,
+      "loss": 0.139,
+      "step": 875
+    },
+    {
+      "epoch": 3.552911392405063,
+      "grad_norm": 0.2781056368911867,
+      "learning_rate": 9.271965082816667e-06,
+      "loss": 0.1315,
+      "step": 876
+    },
+    {
+      "epoch": 3.5569620253164556,
+      "grad_norm": 0.29120270713569624,
+      "learning_rate": 9.224106139385111e-06,
+      "loss": 0.1261,
+      "step": 877
+    },
+    {
+      "epoch": 3.561012658227848,
+      "grad_norm": 0.29326279897511975,
+      "learning_rate": 9.176333983521291e-06,
+      "loss": 0.1295,
+      "step": 878
+    },
+    {
+      "epoch": 3.5650632911392406,
+      "grad_norm": 0.2767557650848104,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.1388,
+      "step": 879
+    },
+    {
+      "epoch": 3.569113924050633,
+      "grad_norm": 0.285138516847677,
+      "learning_rate": 9.08105157279628e-06,
+      "loss": 0.1245,
+      "step": 880
+    },
+    {
+      "epoch": 3.573164556962025,
+      "grad_norm": 0.2688291090999992,
+      "learning_rate": 9.03354208532653e-06,
+      "loss": 0.1166,
+      "step": 881
+    },
+    {
+      "epoch": 3.5772151898734177,
+      "grad_norm": 0.28740781714265784,
+      "learning_rate": 8.986120920201205e-06,
+      "loss": 0.1266,
+      "step": 882
+    },
+    {
+      "epoch": 3.58126582278481,
+      "grad_norm": 0.304031166426565,
+      "learning_rate": 8.938788459343852e-06,
+      "loss": 0.1235,
+      "step": 883
+    },
+    {
+      "epoch": 3.5853164556962023,
+      "grad_norm": 0.3064417382552023,
+      "learning_rate": 8.8915450839636e-06,
+      "loss": 0.1318,
+      "step": 884
+    },
+    {
+      "epoch": 3.589367088607595,
+      "grad_norm": 0.3036712526302319,
+      "learning_rate": 8.844391174552116e-06,
+      "loss": 0.1241,
+      "step": 885
+    },
+    {
+      "epoch": 3.5934177215189873,
+      "grad_norm": 0.2797494834608018,
+      "learning_rate": 8.797327110880479e-06,
+      "loss": 0.1185,
+      "step": 886
+    },
+    {
+      "epoch": 3.59746835443038,
+      "grad_norm": 0.2814760686108722,
+      "learning_rate": 8.750353271996206e-06,
+      "loss": 0.1313,
+      "step": 887
+    },
+    {
+      "epoch": 3.6015189873417723,
+      "grad_norm": 0.2770615349372462,
+      "learning_rate": 8.703470036220132e-06,
+      "loss": 0.1257,
+      "step": 888
+    },
+    {
+      "epoch": 3.6055696202531644,
+      "grad_norm": 0.27837664455459077,
+      "learning_rate": 8.656677781143394e-06,
+      "loss": 0.1265,
+      "step": 889
+    },
+    {
+      "epoch": 3.609620253164557,
+      "grad_norm": 0.2798559155654779,
+      "learning_rate": 8.609976883624377e-06,
+      "loss": 0.1237,
+      "step": 890
+    },
+    {
+      "epoch": 3.6136708860759494,
+      "grad_norm": 0.28788193500338466,
+      "learning_rate": 8.563367719785698e-06,
+      "loss": 0.1305,
+      "step": 891
+    },
+    {
+      "epoch": 3.6177215189873415,
+      "grad_norm": 0.28730808349855946,
+      "learning_rate": 8.516850665011138e-06,
+      "loss": 0.1231,
+      "step": 892
+    },
+    {
+      "epoch": 3.621772151898734,
+      "grad_norm": 0.2781342357822305,
+      "learning_rate": 8.47042609394269e-06,
+      "loss": 0.1321,
+      "step": 893
+    },
+    {
+      "epoch": 3.6258227848101265,
+      "grad_norm": 0.2705868359949422,
+      "learning_rate": 8.424094380477432e-06,
+      "loss": 0.1246,
+      "step": 894
+    },
+    {
+      "epoch": 3.629873417721519,
+      "grad_norm": 0.28178369960719796,
+      "learning_rate": 8.37785589776465e-06,
+      "loss": 0.12,
+      "step": 895
+    },
+    {
+      "epoch": 3.6339240506329116,
+      "grad_norm": 0.28899188434421597,
+      "learning_rate": 8.331711018202694e-06,
+      "loss": 0.127,
+      "step": 896
+    },
+    {
+      "epoch": 3.6379746835443036,
+      "grad_norm": 0.2948641291156313,
+      "learning_rate": 8.285660113436104e-06,
+      "loss": 0.1342,
+      "step": 897
+    },
+    {
+      "epoch": 3.642025316455696,
+      "grad_norm": 0.28312734178524235,
+      "learning_rate": 8.239703554352527e-06,
+      "loss": 0.1249,
+      "step": 898
+    },
+    {
+      "epoch": 3.6460759493670887,
+      "grad_norm": 0.2873449112629255,
+      "learning_rate": 8.193841711079775e-06,
+      "loss": 0.1401,
+      "step": 899
+    },
+    {
+      "epoch": 3.6501265822784807,
+      "grad_norm": 0.3054670607080649,
+      "learning_rate": 8.148074952982828e-06,
+      "loss": 0.1211,
+      "step": 900
+    },
+    {
+      "epoch": 3.6541772151898733,
+      "grad_norm": 0.2870706694817431,
+      "learning_rate": 8.102403648660859e-06,
+      "loss": 0.1219,
+      "step": 901
+    },
+    {
+      "epoch": 3.6582278481012658,
+      "grad_norm": 0.27842078702196466,
+      "learning_rate": 8.056828165944282e-06,
+      "loss": 0.1303,
+      "step": 902
+    },
+    {
+      "epoch": 3.6622784810126583,
+      "grad_norm": 0.2789252739890797,
+      "learning_rate": 8.011348871891762e-06,
+      "loss": 0.1271,
+      "step": 903
+    },
+    {
+      "epoch": 3.666329113924051,
+      "grad_norm": 0.281584885656733,
+      "learning_rate": 7.965966132787287e-06,
+      "loss": 0.1206,
+      "step": 904
+    },
+    {
+      "epoch": 3.670379746835443,
+      "grad_norm": 0.2997678620973134,
+      "learning_rate": 7.920680314137189e-06,
+      "loss": 0.1343,
+      "step": 905
+    },
+    {
+      "epoch": 3.6744303797468354,
+      "grad_norm": 0.2946212331428721,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.1126,
+      "step": 906
+    },
+    {
+      "epoch": 3.678481012658228,
+      "grad_norm": 0.29541585928625497,
+      "learning_rate": 7.830400896319667e-06,
+      "loss": 0.1244,
+      "step": 907
+    },
+    {
+      "epoch": 3.68253164556962,
+      "grad_norm": 0.29746551932322934,
+      "learning_rate": 7.785408024250259e-06,
+      "loss": 0.1212,
+      "step": 908
+    },
+    {
+      "epoch": 3.6865822784810125,
+      "grad_norm": 0.27571110877658916,
+      "learning_rate": 7.74051352682542e-06,
+      "loss": 0.1169,
+      "step": 909
+    },
+    {
+      "epoch": 3.690632911392405,
+      "grad_norm": 0.27328110124646515,
+      "learning_rate": 7.695717765619257e-06,
+      "loss": 0.1248,
+      "step": 910
+    },
+    {
+      "epoch": 3.6946835443037975,
+      "grad_norm": 0.2784042005452249,
+      "learning_rate": 7.651021101410673e-06,
+      "loss": 0.1222,
+      "step": 911
+    },
+    {
+      "epoch": 3.69873417721519,
+      "grad_norm": 0.2701436431932897,
+      "learning_rate": 7.606423894180464e-06,
+      "loss": 0.1226,
+      "step": 912
+    },
+    {
+      "epoch": 3.702784810126582,
+      "grad_norm": 0.2875007461126029,
+      "learning_rate": 7.56192650310839e-06,
+      "loss": 0.1251,
+      "step": 913
+    },
+    {
+      "epoch": 3.7068354430379746,
+      "grad_norm": 0.2808464564991589,
+      "learning_rate": 7.517529286570349e-06,
+      "loss": 0.121,
+      "step": 914
+    },
+    {
+      "epoch": 3.710886075949367,
+      "grad_norm": 0.28856159659899394,
+      "learning_rate": 7.473232602135387e-06,
+      "loss": 0.1261,
+      "step": 915
+    },
+    {
+      "epoch": 3.714936708860759,
+      "grad_norm": 0.2923945274733488,
+      "learning_rate": 7.429036806562935e-06,
+      "loss": 0.1298,
+      "step": 916
+    },
+    {
+      "epoch": 3.7189873417721517,
+      "grad_norm": 0.2845134724019234,
+      "learning_rate": 7.3849422557998455e-06,
+      "loss": 0.1344,
+      "step": 917
+    },
+    {
+      "epoch": 3.7230379746835442,
+      "grad_norm": 0.2840758254453205,
+      "learning_rate": 7.340949304977567e-06,
+      "loss": 0.1331,
+      "step": 918
+    },
+    {
+      "epoch": 3.7270886075949368,
+      "grad_norm": 0.27435356345669343,
+      "learning_rate": 7.297058308409282e-06,
+      "loss": 0.1331,
+      "step": 919
+    },
+    {
+      "epoch": 3.7311392405063293,
+      "grad_norm": 0.29295794902549993,
+      "learning_rate": 7.25326961958704e-06,
+      "loss": 0.13,
+      "step": 920
+    },
+    {
+      "epoch": 3.7351898734177214,
+      "grad_norm": 0.28551073946627614,
+      "learning_rate": 7.209583591178921e-06,
+      "loss": 0.1308,
+      "step": 921
+    },
+    {
+      "epoch": 3.739240506329114,
+      "grad_norm": 0.2715683596061035,
+      "learning_rate": 7.1660005750261925e-06,
+      "loss": 0.1333,
+      "step": 922
+    },
+    {
+      "epoch": 3.7432911392405064,
+      "grad_norm": 0.28122571739515106,
+      "learning_rate": 7.1225209221404765e-06,
+      "loss": 0.1298,
+      "step": 923
+    },
+    {
+      "epoch": 3.747341772151899,
+      "grad_norm": 0.284858015957344,
+      "learning_rate": 7.079144982700909e-06,
+      "loss": 0.1157,
+      "step": 924
+    },
+    {
+      "epoch": 3.7513924050632914,
+      "grad_norm": 0.2856461693751234,
+      "learning_rate": 7.0358731060513695e-06,
+      "loss": 0.1222,
+      "step": 925
+    },
+    {
+      "epoch": 3.7554430379746835,
+      "grad_norm": 0.2893111386954869,
+      "learning_rate": 6.99270564069757e-06,
+      "loss": 0.1259,
+      "step": 926
+    },
+    {
+      "epoch": 3.759493670886076,
+      "grad_norm": 0.2710951570909194,
+      "learning_rate": 6.949642934304375e-06,
+      "loss": 0.1312,
+      "step": 927
+    },
+    {
+      "epoch": 3.7635443037974685,
+      "grad_norm": 0.28671167151995586,
+      "learning_rate": 6.906685333692871e-06,
+      "loss": 0.1315,
+      "step": 928
+    },
+    {
+      "epoch": 3.7675949367088606,
+      "grad_norm": 0.26957333310950266,
+      "learning_rate": 6.86383318483769e-06,
+      "loss": 0.1338,
+      "step": 929
+    },
+    {
+      "epoch": 3.771645569620253,
+      "grad_norm": 0.27115033074435824,
+      "learning_rate": 6.821086832864139e-06,
+      "loss": 0.1343,
+      "step": 930
+    },
+    {
+      "epoch": 3.7756962025316456,
+      "grad_norm": 0.2739125508764413,
+      "learning_rate": 6.77844662204546e-06,
+      "loss": 0.1307,
+      "step": 931
+    },
+    {
+      "epoch": 3.779746835443038,
+      "grad_norm": 0.2667350303996891,
+      "learning_rate": 6.7359128958000455e-06,
+      "loss": 0.1402,
+      "step": 932
+    },
+    {
+      "epoch": 3.7837974683544306,
+      "grad_norm": 0.2811757011439706,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.1176,
+      "step": 933
+    },
+    {
+      "epoch": 3.7878481012658227,
+      "grad_norm": 0.26398911116634555,
+      "learning_rate": 6.651166266411801e-06,
+      "loss": 0.1283,
+      "step": 934
+    },
+    {
+      "epoch": 3.7918987341772152,
+      "grad_norm": 0.3234748857414268,
+      "learning_rate": 6.6089540458066725e-06,
+      "loss": 0.1382,
+      "step": 935
+    },
+    {
+      "epoch": 3.7959493670886078,
+      "grad_norm": 0.27257340237875155,
+      "learning_rate": 6.566849674844711e-06,
+      "loss": 0.1153,
+      "step": 936
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.28532345456827324,
+      "learning_rate": 6.524853492628747e-06,
+      "loss": 0.1209,
+      "step": 937
+    },
+    {
+      "epoch": 3.8040506329113923,
+      "grad_norm": 0.2694451508627589,
+      "learning_rate": 6.4829658373902536e-06,
+      "loss": 0.1257,
+      "step": 938
+    },
+    {
+      "epoch": 3.808101265822785,
+      "grad_norm": 0.28907422825213075,
+      "learning_rate": 6.441187046486648e-06,
+      "loss": 0.1199,
+      "step": 939
+    },
+    {
+      "epoch": 3.8121518987341774,
+      "grad_norm": 0.2727006789504044,
+      "learning_rate": 6.399517456398567e-06,
+      "loss": 0.1344,
+      "step": 940
+    },
+    {
+      "epoch": 3.81620253164557,
+      "grad_norm": 0.2778216337147786,
+      "learning_rate": 6.357957402727164e-06,
+      "loss": 0.1192,
+      "step": 941
+    },
+    {
+      "epoch": 3.820253164556962,
+      "grad_norm": 0.2787387717814247,
+      "learning_rate": 6.316507220191395e-06,
+      "loss": 0.1337,
+      "step": 942
+    },
+    {
+      "epoch": 3.8243037974683545,
+      "grad_norm": 0.2677651011334803,
+      "learning_rate": 6.275167242625331e-06,
+      "loss": 0.1163,
+      "step": 943
+    },
+    {
+      "epoch": 3.828354430379747,
+      "grad_norm": 0.29285250014567143,
+      "learning_rate": 6.233937802975471e-06,
+      "loss": 0.1276,
+      "step": 944
+    },
+    {
+      "epoch": 3.832405063291139,
+      "grad_norm": 0.2784712851583435,
+      "learning_rate": 6.192819233298046e-06,
+      "loss": 0.1231,
+      "step": 945
+    },
+    {
+      "epoch": 3.8364556962025316,
+      "grad_norm": 0.2773432808167056,
+      "learning_rate": 6.151811864756383e-06,
+      "loss": 0.1167,
+      "step": 946
+    },
+    {
+      "epoch": 3.840506329113924,
+      "grad_norm": 0.2884141048323366,
+      "learning_rate": 6.1109160276181655e-06,
+      "loss": 0.1236,
+      "step": 947
+    },
+    {
+      "epoch": 3.8445569620253166,
+      "grad_norm": 0.2888952602025546,
+      "learning_rate": 6.070132051252868e-06,
+      "loss": 0.1371,
+      "step": 948
+    },
+    {
+      "epoch": 3.848607594936709,
+      "grad_norm": 0.28138985275471856,
+      "learning_rate": 6.0294602641290034e-06,
+      "loss": 0.1124,
+      "step": 949
+    },
+    {
+      "epoch": 3.852658227848101,
+      "grad_norm": 0.2798421933706704,
+      "learning_rate": 5.988900993811575e-06,
+      "loss": 0.1259,
+      "step": 950
+    },
+    {
+      "epoch": 3.8567088607594937,
+      "grad_norm": 0.27349067163173574,
+      "learning_rate": 5.948454566959363e-06,
+      "loss": 0.1271,
+      "step": 951
+    },
+    {
+      "epoch": 3.8607594936708862,
+      "grad_norm": 0.28122369248785045,
+      "learning_rate": 5.908121309322328e-06,
+      "loss": 0.1315,
+      "step": 952
+    },
+    {
+      "epoch": 3.8648101265822783,
+      "grad_norm": 0.27628379933238184,
+      "learning_rate": 5.867901545738976e-06,
+      "loss": 0.1345,
+      "step": 953
+    },
+    {
+      "epoch": 3.868860759493671,
+      "grad_norm": 0.29681043496312637,
+      "learning_rate": 5.827795600133774e-06,
+      "loss": 0.1256,
+      "step": 954
+    },
+    {
+      "epoch": 3.8729113924050633,
+      "grad_norm": 0.27403461898843834,
+      "learning_rate": 5.787803795514466e-06,
+      "loss": 0.1284,
+      "step": 955
+    },
+    {
+      "epoch": 3.876962025316456,
+      "grad_norm": 0.26729991110370516,
+      "learning_rate": 5.747926453969576e-06,
+      "loss": 0.1211,
+      "step": 956
+    },
+    {
+      "epoch": 3.8810126582278484,
+      "grad_norm": 0.2694292100921135,
+      "learning_rate": 5.708163896665708e-06,
+      "loss": 0.1259,
+      "step": 957
+    },
+    {
+      "epoch": 3.8850632911392404,
+      "grad_norm": 0.2718869464961677,
+      "learning_rate": 5.668516443845047e-06,
+      "loss": 0.1269,
+      "step": 958
+    },
+    {
+      "epoch": 3.889113924050633,
+      "grad_norm": 0.2768512692578364,
+      "learning_rate": 5.6289844148227225e-06,
+      "loss": 0.1137,
+      "step": 959
+    },
+    {
+      "epoch": 3.8931645569620255,
+      "grad_norm": 0.2861141318602218,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.137,
+      "step": 960
+    },
+    {
+      "epoch": 3.8972151898734175,
+      "grad_norm": 0.2646604711852767,
+      "learning_rate": 5.550267900783019e-06,
+      "loss": 0.13,
+      "step": 961
+    },
+    {
+      "epoch": 3.90126582278481,
+      "grad_norm": 0.2769322000989181,
+      "learning_rate": 5.511084049737623e-06,
+      "loss": 0.1247,
+      "step": 962
+    },
+    {
+      "epoch": 3.9053164556962026,
+      "grad_norm": 0.29203115665076285,
+      "learning_rate": 5.4720168904294215e-06,
+      "loss": 0.1306,
+      "step": 963
+    },
+    {
+      "epoch": 3.909367088607595,
+      "grad_norm": 0.26976633537705846,
+      "learning_rate": 5.433066737499948e-06,
+      "loss": 0.1282,
+      "step": 964
+    },
+    {
+      "epoch": 3.9134177215189876,
+      "grad_norm": 0.26744739840735277,
+      "learning_rate": 5.394233904648376e-06,
+      "loss": 0.1308,
+      "step": 965
+    },
+    {
+      "epoch": 3.9174683544303797,
+      "grad_norm": 0.2771221717966821,
+      "learning_rate": 5.355518704628997e-06,
+      "loss": 0.1212,
+      "step": 966
+    },
+    {
+      "epoch": 3.921518987341772,
+      "grad_norm": 0.28821141382016424,
+      "learning_rate": 5.316921449248731e-06,
+      "loss": 0.1297,
+      "step": 967
+    },
+    {
+      "epoch": 3.9255696202531647,
+      "grad_norm": 0.2788875220579584,
+      "learning_rate": 5.278442449364538e-06,
+      "loss": 0.1212,
+      "step": 968
+    },
+    {
+      "epoch": 3.9296202531645568,
+      "grad_norm": 0.2748728254938739,
+      "learning_rate": 5.240082014881016e-06,
+      "loss": 0.12,
+      "step": 969
+    },
+    {
+      "epoch": 3.9336708860759493,
+      "grad_norm": 0.2827073425321622,
+      "learning_rate": 5.201840454747822e-06,
+      "loss": 0.1145,
+      "step": 970
+    },
+    {
+      "epoch": 3.937721518987342,
+      "grad_norm": 0.27006951550513325,
+      "learning_rate": 5.163718076957223e-06,
+      "loss": 0.1331,
+      "step": 971
+    },
+    {
+      "epoch": 3.9417721518987343,
+      "grad_norm": 0.2745238283616329,
+      "learning_rate": 5.125715188541609e-06,
+      "loss": 0.1243,
+      "step": 972
+    },
+    {
+      "epoch": 3.945822784810127,
+      "grad_norm": 0.2779443781668121,
+      "learning_rate": 5.087832095571021e-06,
+      "loss": 0.1214,
+      "step": 973
+    },
+    {
+      "epoch": 3.949873417721519,
+      "grad_norm": 0.28212037796942596,
+      "learning_rate": 5.0500691031506766e-06,
+      "loss": 0.1222,
+      "step": 974
+    },
+    {
+      "epoch": 3.9539240506329114,
+      "grad_norm": 0.2606399689370262,
+      "learning_rate": 5.01242651541854e-06,
+      "loss": 0.1259,
+      "step": 975
+    },
+    {
+      "epoch": 3.957974683544304,
+      "grad_norm": 0.26951529390973505,
+      "learning_rate": 4.974904635542815e-06,
+      "loss": 0.124,
+      "step": 976
+    },
+    {
+      "epoch": 3.962025316455696,
+      "grad_norm": 0.28275428794354857,
+      "learning_rate": 4.937503765719582e-06,
+      "loss": 0.1308,
+      "step": 977
+    },
+    {
+      "epoch": 3.9660759493670885,
+      "grad_norm": 0.2688922771841612,
+      "learning_rate": 4.900224207170299e-06,
+      "loss": 0.1274,
+      "step": 978
+    },
+    {
+      "epoch": 3.970126582278481,
+      "grad_norm": 0.2872461425178586,
+      "learning_rate": 4.8630662601394065e-06,
+      "loss": 0.1293,
+      "step": 979
+    },
+    {
+      "epoch": 3.9741772151898735,
+      "grad_norm": 0.2827303137737239,
+      "learning_rate": 4.8260302238918995e-06,
+      "loss": 0.1195,
+      "step": 980
+    },
+    {
+      "epoch": 3.978227848101266,
+      "grad_norm": 0.26059774537868996,
+      "learning_rate": 4.789116396710924e-06,
+      "loss": 0.1231,
+      "step": 981
+    },
+    {
+      "epoch": 3.982278481012658,
+      "grad_norm": 0.26968051739037197,
+      "learning_rate": 4.752325075895368e-06,
+      "loss": 0.121,
+      "step": 982
+    },
+    {
+      "epoch": 3.9863291139240506,
+      "grad_norm": 0.2661885181816016,
+      "learning_rate": 4.715656557757473e-06,
+      "loss": 0.1318,
+      "step": 983
+    },
+    {
+      "epoch": 3.990379746835443,
+      "grad_norm": 0.2815775579206786,
+      "learning_rate": 4.679111137620442e-06,
+      "loss": 0.1156,
+      "step": 984
+    },
+    {
+      "epoch": 3.9944303797468352,
+      "grad_norm": 0.2945292635819614,
+      "learning_rate": 4.6426891098160585e-06,
+      "loss": 0.1159,
+      "step": 985
+    },
+    {
+      "epoch": 3.9984810126582278,
+      "grad_norm": 0.2767458150691669,
+      "learning_rate": 4.6063907676823474e-06,
+      "loss": 0.1187,
+      "step": 986
+    },
+    {
+      "epoch": 4.004050632911392,
+      "grad_norm": 0.31873368923266243,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.0738,
+      "step": 987
+    },
+    {
+      "epoch": 4.008101265822785,
+      "grad_norm": 0.2952986850674694,
+      "learning_rate": 4.534166308795815e-06,
+      "loss": 0.0842,
+      "step": 988
+    },
+    {
+      "epoch": 4.012151898734177,
+      "grad_norm": 0.254476297156985,
+      "learning_rate": 4.498240773728859e-06,
+      "loss": 0.0892,
+      "step": 989
+    },
+    {
+      "epoch": 4.01620253164557,
+      "grad_norm": 0.2228227232057975,
+      "learning_rate": 4.462440087699609e-06,
+      "loss": 0.0739,
+      "step": 990
+    },
+    {
+      "epoch": 4.020253164556962,
+      "grad_norm": 0.27877468578359826,
+      "learning_rate": 4.426764539041861e-06,
+      "loss": 0.0831,
+      "step": 991
+    },
+    {
+      "epoch": 4.024303797468354,
+      "grad_norm": 0.29667197113947175,
+      "learning_rate": 4.391214415081582e-06,
+      "loss": 0.0852,
+      "step": 992
+    },
+    {
+      "epoch": 4.028354430379747,
+      "grad_norm": 0.3435677273317354,
+      "learning_rate": 4.355790002134579e-06,
+      "loss": 0.0869,
+      "step": 993
+    },
+    {
+      "epoch": 4.032405063291139,
+      "grad_norm": 0.3100642569491875,
+      "learning_rate": 4.320491585504207e-06,
+      "loss": 0.0793,
+      "step": 994
+    },
+    {
+      "epoch": 4.036455696202531,
+      "grad_norm": 0.2696511305782832,
+      "learning_rate": 4.2853194494790615e-06,
+      "loss": 0.0776,
+      "step": 995
+    },
+    {
+      "epoch": 4.040506329113924,
+      "grad_norm": 0.2671299544520489,
+      "learning_rate": 4.250273877330691e-06,
+      "loss": 0.0986,
+      "step": 996
+    },
+    {
+      "epoch": 4.044556962025316,
+      "grad_norm": 0.2795078900750349,
+      "learning_rate": 4.215355151311313e-06,
+      "loss": 0.0807,
+      "step": 997
+    },
+    {
+      "epoch": 4.048607594936709,
+      "grad_norm": 0.2742486620985594,
+      "learning_rate": 4.180563552651542e-06,
+      "loss": 0.0793,
+      "step": 998
+    },
+    {
+      "epoch": 4.052658227848101,
+      "grad_norm": 0.25976755524244444,
+      "learning_rate": 4.145899361558147e-06,
+      "loss": 0.0855,
+      "step": 999
+    },
+    {
+      "epoch": 4.056708860759493,
+      "grad_norm": 0.25407381112918626,
+      "learning_rate": 4.111362857211738e-06,
+      "loss": 0.0798,
+      "step": 1000
+    },
+    {
+      "epoch": 4.060759493670886,
+      "grad_norm": 0.24298029152627781,
+      "learning_rate": 4.076954317764592e-06,
+      "loss": 0.0935,
+      "step": 1001
+    },
+    {
+      "epoch": 4.0648101265822785,
+      "grad_norm": 0.24527508876465012,
+      "learning_rate": 4.042674020338335e-06,
+      "loss": 0.0783,
+      "step": 1002
+    },
+    {
+      "epoch": 4.0688607594936705,
+      "grad_norm": 0.2667161459247534,
+      "learning_rate": 4.0085222410217835e-06,
+      "loss": 0.0826,
+      "step": 1003
+    },
+    {
+      "epoch": 4.0729113924050635,
+      "grad_norm": 0.25760770758522994,
+      "learning_rate": 3.974499254868674e-06,
+      "loss": 0.0971,
+      "step": 1004
+    },
+    {
+      "epoch": 4.076962025316456,
+      "grad_norm": 0.27034830892589373,
+      "learning_rate": 3.940605335895451e-06,
+      "loss": 0.0786,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0810126582278485,
+      "grad_norm": 0.2561277884373583,
+      "learning_rate": 3.90684075707908e-06,
+      "loss": 0.0855,
+      "step": 1006
+    },
+    {
+      "epoch": 4.085063291139241,
+      "grad_norm": 0.27343877009867745,
+      "learning_rate": 3.8732057903548505e-06,
+      "loss": 0.0937,
+      "step": 1007
+    },
+    {
+      "epoch": 4.089113924050633,
+      "grad_norm": 0.23269048952507534,
+      "learning_rate": 3.8397007066141375e-06,
+      "loss": 0.0797,
+      "step": 1008
+    },
+    {
+      "epoch": 4.093164556962026,
+      "grad_norm": 0.23561024461260782,
+      "learning_rate": 3.806325775702304e-06,
+      "loss": 0.086,
+      "step": 1009
+    },
+    {
+      "epoch": 4.097215189873418,
+      "grad_norm": 0.2410431064818257,
+      "learning_rate": 3.773081266416434e-06,
+      "loss": 0.0751,
+      "step": 1010
+    },
+    {
+      "epoch": 4.10126582278481,
+      "grad_norm": 0.26062582944097495,
+      "learning_rate": 3.739967446503245e-06,
+      "loss": 0.083,
+      "step": 1011
+    },
+    {
+      "epoch": 4.105316455696203,
+      "grad_norm": 0.25049455361199763,
+      "learning_rate": 3.706984582656894e-06,
+      "loss": 0.0842,
+      "step": 1012
+    },
+    {
+      "epoch": 4.109367088607595,
+      "grad_norm": 0.264559019840626,
+      "learning_rate": 3.6741329405168237e-06,
+      "loss": 0.0824,
+      "step": 1013
+    },
+    {
+      "epoch": 4.113417721518988,
+      "grad_norm": 0.23462243780077247,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.0874,
+      "step": 1014
+    },
+    {
+      "epoch": 4.11746835443038,
+      "grad_norm": 0.23749804268317135,
+      "learning_rate": 3.608824378627005e-06,
+      "loss": 0.0701,
+      "step": 1015
+    },
+    {
+      "epoch": 4.121518987341772,
+      "grad_norm": 0.26020872308959025,
+      "learning_rate": 3.5763679848634337e-06,
+      "loss": 0.0758,
+      "step": 1016
+    },
+    {
+      "epoch": 4.125569620253165,
+      "grad_norm": 0.25978363336801025,
+      "learning_rate": 3.544043864774269e-06,
+      "loss": 0.0913,
+      "step": 1017
+    },
+    {
+      "epoch": 4.129620253164557,
+      "grad_norm": 0.23546448626189947,
+      "learning_rate": 3.5118522786935282e-06,
+      "loss": 0.0869,
+      "step": 1018
+    },
+    {
+      "epoch": 4.133670886075949,
+      "grad_norm": 0.23195262240244768,
+      "learning_rate": 3.479793485887819e-06,
+      "loss": 0.0816,
+      "step": 1019
+    },
+    {
+      "epoch": 4.137721518987342,
+      "grad_norm": 0.23552661317810383,
+      "learning_rate": 3.4478677445542653e-06,
+      "loss": 0.0948,
+      "step": 1020
+    },
+    {
+      "epoch": 4.141772151898734,
+      "grad_norm": 0.23036419166568,
+      "learning_rate": 3.4160753118183767e-06,
+      "loss": 0.0829,
+      "step": 1021
+    },
+    {
+      "epoch": 4.145822784810127,
+      "grad_norm": 0.23898456121957412,
+      "learning_rate": 3.3844164437320527e-06,
+      "loss": 0.0906,
+      "step": 1022
+    },
+    {
+      "epoch": 4.149873417721519,
+      "grad_norm": 0.24030606993167183,
+      "learning_rate": 3.3528913952714558e-06,
+      "loss": 0.0837,
+      "step": 1023
+    },
+    {
+      "epoch": 4.153924050632911,
+      "grad_norm": 0.2428356620933478,
+      "learning_rate": 3.321500420335e-06,
+      "loss": 0.0788,
+      "step": 1024
+    },
+    {
+      "epoch": 4.157974683544304,
+      "grad_norm": 0.24730055995923386,
+      "learning_rate": 3.290243771741275e-06,
+      "loss": 0.0785,
+      "step": 1025
+    },
+    {
+      "epoch": 4.162025316455696,
+      "grad_norm": 0.24784960808689277,
+      "learning_rate": 3.2591217012270325e-06,
+      "loss": 0.0852,
+      "step": 1026
+    },
+    {
+      "epoch": 4.166075949367088,
+      "grad_norm": 0.22638942351287777,
+      "learning_rate": 3.228134459445149e-06,
+      "loss": 0.0697,
+      "step": 1027
+    },
+    {
+      "epoch": 4.170126582278481,
+      "grad_norm": 0.24896601576809996,
+      "learning_rate": 3.1972822959626205e-06,
+      "loss": 0.0835,
+      "step": 1028
+    },
+    {
+      "epoch": 4.174177215189873,
+      "grad_norm": 0.23951983942549043,
+      "learning_rate": 3.166565459258513e-06,
+      "loss": 0.0806,
+      "step": 1029
+    },
+    {
+      "epoch": 4.178227848101266,
+      "grad_norm": 0.24129371732862342,
+      "learning_rate": 3.1359841967220193e-06,
+      "loss": 0.0833,
+      "step": 1030
+    },
+    {
+      "epoch": 4.182278481012658,
+      "grad_norm": 0.2603581737721094,
+      "learning_rate": 3.105538754650419e-06,
+      "loss": 0.0735,
+      "step": 1031
+    },
+    {
+      "epoch": 4.18632911392405,
+      "grad_norm": 0.23029193870302317,
+      "learning_rate": 3.07522937824712e-06,
+      "loss": 0.0938,
+      "step": 1032
+    },
+    {
+      "epoch": 4.190379746835443,
+      "grad_norm": 0.23060224562466697,
+      "learning_rate": 3.0450563116196697e-06,
+      "loss": 0.0687,
+      "step": 1033
+    },
+    {
+      "epoch": 4.194430379746835,
+      "grad_norm": 0.2349962697004771,
+      "learning_rate": 3.0150197977778008e-06,
+      "loss": 0.0869,
+      "step": 1034
+    },
+    {
+      "epoch": 4.1984810126582275,
+      "grad_norm": 0.23788932274583524,
+      "learning_rate": 2.985120078631465e-06,
+      "loss": 0.0782,
+      "step": 1035
+    },
+    {
+      "epoch": 4.2025316455696204,
+      "grad_norm": 0.24179448871342635,
+      "learning_rate": 2.9553573949888893e-06,
+      "loss": 0.0701,
+      "step": 1036
+    },
+    {
+      "epoch": 4.2065822784810125,
+      "grad_norm": 0.2332606680284559,
+      "learning_rate": 2.9257319865546384e-06,
+      "loss": 0.0732,
+      "step": 1037
+    },
+    {
+      "epoch": 4.2106329113924055,
+      "grad_norm": 0.2377577678170282,
+      "learning_rate": 2.896244091927678e-06,
+      "loss": 0.0822,
+      "step": 1038
+    },
+    {
+      "epoch": 4.2146835443037975,
+      "grad_norm": 0.22788063965220798,
+      "learning_rate": 2.8668939485994584e-06,
+      "loss": 0.091,
+      "step": 1039
+    },
+    {
+      "epoch": 4.21873417721519,
+      "grad_norm": 0.2405778742685026,
+      "learning_rate": 2.837681792951994e-06,
+      "loss": 0.0933,
+      "step": 1040
+    },
+    {
+      "epoch": 4.222784810126583,
+      "grad_norm": 0.2386713129519911,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.0868,
+      "step": 1041
+    },
+    {
+      "epoch": 4.226835443037975,
+      "grad_norm": 0.243480517773955,
+      "learning_rate": 2.7796723846688634e-06,
+      "loss": 0.084,
+      "step": 1042
+    },
+    {
+      "epoch": 4.230886075949367,
+      "grad_norm": 0.22749175094684532,
+      "learning_rate": 2.7508755992329937e-06,
+      "loss": 0.0868,
+      "step": 1043
+    },
+    {
+      "epoch": 4.23493670886076,
+      "grad_norm": 0.2312541659809669,
+      "learning_rate": 2.722217735873718e-06,
+      "loss": 0.0773,
+      "step": 1044
+    },
+    {
+      "epoch": 4.238987341772152,
+      "grad_norm": 0.2430979044405334,
+      "learning_rate": 2.6936990253975315e-06,
+      "loss": 0.0825,
+      "step": 1045
+    },
+    {
+      "epoch": 4.243037974683545,
+      "grad_norm": 0.2423586704652075,
+      "learning_rate": 2.665319697490205e-06,
+      "loss": 0.0785,
+      "step": 1046
+    },
+    {
+      "epoch": 4.247088607594937,
+      "grad_norm": 0.23972628302634463,
+      "learning_rate": 2.637079980714945e-06,
+      "loss": 0.0807,
+      "step": 1047
+    },
+    {
+      "epoch": 4.251139240506329,
+      "grad_norm": 0.2308998073885542,
+      "learning_rate": 2.6089801025105453e-06,
+      "loss": 0.0866,
+      "step": 1048
+    },
+    {
+      "epoch": 4.255189873417722,
+      "grad_norm": 0.23429939916520565,
+      "learning_rate": 2.581020289189571e-06,
+      "loss": 0.0862,
+      "step": 1049
+    },
+    {
+      "epoch": 4.259240506329114,
+      "grad_norm": 0.25613404489894853,
+      "learning_rate": 2.553200765936501e-06,
+      "loss": 0.0886,
+      "step": 1050
+    },
+    {
+      "epoch": 4.263291139240506,
+      "grad_norm": 0.23550225163952596,
+      "learning_rate": 2.525521756805962e-06,
+      "loss": 0.0761,
+      "step": 1051
+    },
+    {
+      "epoch": 4.267341772151899,
+      "grad_norm": 0.2618903094391508,
+      "learning_rate": 2.497983484720885e-06,
+      "loss": 0.0726,
+      "step": 1052
+    },
+    {
+      "epoch": 4.271392405063291,
+      "grad_norm": 0.2407452771668295,
+      "learning_rate": 2.470586171470728e-06,
+      "loss": 0.0846,
+      "step": 1053
+    },
+    {
+      "epoch": 4.275443037974684,
+      "grad_norm": 0.23066264974210773,
+      "learning_rate": 2.4433300377096836e-06,
+      "loss": 0.0768,
+      "step": 1054
+    },
+    {
+      "epoch": 4.279493670886076,
+      "grad_norm": 0.25377653579374,
+      "learning_rate": 2.4162153029549073e-06,
+      "loss": 0.0864,
+      "step": 1055
+    },
+    {
+      "epoch": 4.283544303797468,
+      "grad_norm": 0.23170714067818196,
+      "learning_rate": 2.3892421855847458e-06,
+      "loss": 0.0774,
+      "step": 1056
+    },
+    {
+      "epoch": 4.287594936708861,
+      "grad_norm": 0.2317128201519415,
+      "learning_rate": 2.362410902836978e-06,
+      "loss": 0.0756,
+      "step": 1057
+    },
+    {
+      "epoch": 4.291645569620253,
+      "grad_norm": 0.2360636912494169,
+      "learning_rate": 2.3357216708070653e-06,
+      "loss": 0.086,
+      "step": 1058
+    },
+    {
+      "epoch": 4.295696202531645,
+      "grad_norm": 0.2318820257733153,
+      "learning_rate": 2.309174704446411e-06,
+      "loss": 0.0738,
+      "step": 1059
+    },
+    {
+      "epoch": 4.299746835443038,
+      "grad_norm": 0.2475174099255845,
+      "learning_rate": 2.2827702175606437e-06,
+      "loss": 0.0836,
+      "step": 1060
+    },
+    {
+      "epoch": 4.30379746835443,
+      "grad_norm": 0.24786840026369697,
+      "learning_rate": 2.256508422807855e-06,
+      "loss": 0.0809,
+      "step": 1061
+    },
+    {
+      "epoch": 4.307848101265823,
+      "grad_norm": 0.23759956485889777,
+      "learning_rate": 2.230389531696946e-06,
+      "loss": 0.097,
+      "step": 1062
+    },
+    {
+      "epoch": 4.311898734177215,
+      "grad_norm": 0.2417252795217626,
+      "learning_rate": 2.204413754585857e-06,
+      "loss": 0.0794,
+      "step": 1063
+    },
+    {
+      "epoch": 4.315949367088607,
+      "grad_norm": 0.24047644933349616,
+      "learning_rate": 2.1785813006799406e-06,
+      "loss": 0.0892,
+      "step": 1064
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.23711900594711932,
+      "learning_rate": 2.1528923780302224e-06,
+      "loss": 0.0856,
+      "step": 1065
+    },
+    {
+      "epoch": 4.324050632911392,
+      "grad_norm": 0.22845937731836707,
+      "learning_rate": 2.127347193531757e-06,
+      "loss": 0.076,
+      "step": 1066
+    },
+    {
+      "epoch": 4.328101265822784,
+      "grad_norm": 0.23279860209461653,
+      "learning_rate": 2.101945952921942e-06,
+      "loss": 0.0749,
+      "step": 1067
+    },
+    {
+      "epoch": 4.332151898734177,
+      "grad_norm": 0.23538641766792073,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.0819,
+      "step": 1068
+    },
+    {
+      "epoch": 4.3362025316455695,
+      "grad_norm": 0.24648419997613008,
+      "learning_rate": 2.0515761205197337e-06,
+      "loss": 0.0794,
+      "step": 1069
+    },
+    {
+      "epoch": 4.340253164556962,
+      "grad_norm": 0.2339922303108145,
+      "learning_rate": 2.0266079343990453e-06,
+      "loss": 0.0824,
+      "step": 1070
+    },
+    {
+      "epoch": 4.3443037974683545,
+      "grad_norm": 0.22555762309633848,
+      "learning_rate": 2.0017845035071494e-06,
+      "loss": 0.0881,
+      "step": 1071
+    },
+    {
+      "epoch": 4.348354430379747,
+      "grad_norm": 0.23514056948591727,
+      "learning_rate": 1.9771060277685537e-06,
+      "loss": 0.0746,
+      "step": 1072
+    },
+    {
+      "epoch": 4.3524050632911395,
+      "grad_norm": 0.23355711372502777,
+      "learning_rate": 1.95257270594031e-06,
+      "loss": 0.0779,
+      "step": 1073
+    },
+    {
+      "epoch": 4.356455696202532,
+      "grad_norm": 0.2364652544125728,
+      "learning_rate": 1.9281847356104188e-06,
+      "loss": 0.0807,
+      "step": 1074
+    },
+    {
+      "epoch": 4.360506329113924,
+      "grad_norm": 0.23221342183435734,
+      "learning_rate": 1.9039423131962365e-06,
+      "loss": 0.076,
+      "step": 1075
+    },
+    {
+      "epoch": 4.364556962025317,
+      "grad_norm": 0.2530832143400456,
+      "learning_rate": 1.8798456339429027e-06,
+      "loss": 0.0842,
+      "step": 1076
+    },
+    {
+      "epoch": 4.368607594936709,
+      "grad_norm": 0.2469851684422886,
+      "learning_rate": 1.8558948919217612e-06,
+      "loss": 0.0943,
+      "step": 1077
+    },
+    {
+      "epoch": 4.372658227848102,
+      "grad_norm": 0.2411081327081187,
+      "learning_rate": 1.8320902800287954e-06,
+      "loss": 0.0717,
+      "step": 1078
+    },
+    {
+      "epoch": 4.376708860759494,
+      "grad_norm": 0.2305571978091865,
+      "learning_rate": 1.8084319899830726e-06,
+      "loss": 0.0815,
+      "step": 1079
+    },
+    {
+      "epoch": 4.380759493670886,
+      "grad_norm": 0.23138125807405796,
+      "learning_rate": 1.7849202123252097e-06,
+      "loss": 0.0827,
+      "step": 1080
+    },
+    {
+      "epoch": 4.384810126582279,
+      "grad_norm": 0.23810578698142768,
+      "learning_rate": 1.7615551364158401e-06,
+      "loss": 0.0854,
+      "step": 1081
+    },
+    {
+      "epoch": 4.388860759493671,
+      "grad_norm": 0.22873069285728492,
+      "learning_rate": 1.738336950434061e-06,
+      "loss": 0.0846,
+      "step": 1082
+    },
+    {
+      "epoch": 4.392911392405063,
+      "grad_norm": 0.2565133979388028,
+      "learning_rate": 1.715265841375957e-06,
+      "loss": 0.0778,
+      "step": 1083
+    },
+    {
+      "epoch": 4.396962025316456,
+      "grad_norm": 0.2519967966429847,
+      "learning_rate": 1.6923419950530684e-06,
+      "loss": 0.0918,
+      "step": 1084
+    },
+    {
+      "epoch": 4.401012658227848,
+      "grad_norm": 0.24412663291960943,
+      "learning_rate": 1.6695655960909008e-06,
+      "loss": 0.0848,
+      "step": 1085
+    },
+    {
+      "epoch": 4.405063291139241,
+      "grad_norm": 0.23575429315341861,
+      "learning_rate": 1.646936827927441e-06,
+      "loss": 0.0762,
+      "step": 1086
+    },
+    {
+      "epoch": 4.409113924050633,
+      "grad_norm": 0.23628688873148662,
+      "learning_rate": 1.6244558728116766e-06,
+      "loss": 0.0737,
+      "step": 1087
+    },
+    {
+      "epoch": 4.413164556962025,
+      "grad_norm": 0.2354906317752142,
+      "learning_rate": 1.6021229118021265e-06,
+      "loss": 0.0767,
+      "step": 1088
+    },
+    {
+      "epoch": 4.417215189873418,
+      "grad_norm": 0.24720157124563835,
+      "learning_rate": 1.5799381247653967e-06,
+      "loss": 0.0942,
+      "step": 1089
+    },
+    {
+      "epoch": 4.42126582278481,
+      "grad_norm": 0.23768478667751716,
+      "learning_rate": 1.5579016903747013e-06,
+      "loss": 0.0754,
+      "step": 1090
+    },
+    {
+      "epoch": 4.425316455696202,
+      "grad_norm": 0.2287029666493575,
+      "learning_rate": 1.5360137861084656e-06,
+      "loss": 0.0773,
+      "step": 1091
+    },
+    {
+      "epoch": 4.429367088607595,
+      "grad_norm": 0.23745817889086143,
+      "learning_rate": 1.5142745882488475e-06,
+      "loss": 0.0828,
+      "step": 1092
+    },
+    {
+      "epoch": 4.433417721518987,
+      "grad_norm": 0.2481815953355656,
+      "learning_rate": 1.4926842718803691e-06,
+      "loss": 0.0833,
+      "step": 1093
+    },
+    {
+      "epoch": 4.43746835443038,
+      "grad_norm": 0.2287421231535079,
+      "learning_rate": 1.4712430108884657e-06,
+      "loss": 0.0831,
+      "step": 1094
+    },
+    {
+      "epoch": 4.441518987341772,
+      "grad_norm": 0.22152391009889583,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.0887,
+      "step": 1095
+    },
+    {
+      "epoch": 4.445569620253164,
+      "grad_norm": 0.2299536542287035,
+      "learning_rate": 1.4288083445723988e-06,
+      "loss": 0.0756,
+      "step": 1096
+    },
+    {
+      "epoch": 4.449620253164557,
+      "grad_norm": 0.22751602110425298,
+      "learning_rate": 1.4078152810112045e-06,
+      "loss": 0.0865,
+      "step": 1097
+    },
+    {
+      "epoch": 4.453670886075949,
+      "grad_norm": 0.23130054947260004,
+      "learning_rate": 1.3869719563497697e-06,
+      "loss": 0.0776,
+      "step": 1098
+    },
+    {
+      "epoch": 4.457721518987341,
+      "grad_norm": 0.23333491272142334,
+      "learning_rate": 1.3662785384573663e-06,
+      "loss": 0.092,
+      "step": 1099
+    },
+    {
+      "epoch": 4.461772151898734,
+      "grad_norm": 0.2471567084222378,
+      "learning_rate": 1.3457351939959383e-06,
+      "loss": 0.0853,
+      "step": 1100
+    },
+    {
+      "epoch": 4.465822784810126,
+      "grad_norm": 0.24018651499101354,
+      "learning_rate": 1.3253420884187551e-06,
+      "loss": 0.0728,
+      "step": 1101
+    },
+    {
+      "epoch": 4.469873417721519,
+      "grad_norm": 0.25273208614062415,
+      "learning_rate": 1.3050993859690953e-06,
+      "loss": 0.092,
+      "step": 1102
+    },
+    {
+      "epoch": 4.473924050632911,
+      "grad_norm": 0.2292714260696701,
+      "learning_rate": 1.2850072496788869e-06,
+      "loss": 0.0709,
+      "step": 1103
+    },
+    {
+      "epoch": 4.4779746835443035,
+      "grad_norm": 0.23758171350124635,
+      "learning_rate": 1.2650658413674434e-06,
+      "loss": 0.082,
+      "step": 1104
+    },
+    {
+      "epoch": 4.4820253164556965,
+      "grad_norm": 0.23463564602268425,
+      "learning_rate": 1.2452753216401226e-06,
+      "loss": 0.0788,
+      "step": 1105
+    },
+    {
+      "epoch": 4.4860759493670885,
+      "grad_norm": 0.23672313387140664,
+      "learning_rate": 1.2256358498870503e-06,
+      "loss": 0.084,
+      "step": 1106
+    },
+    {
+      "epoch": 4.490126582278481,
+      "grad_norm": 0.23291093690834586,
+      "learning_rate": 1.2061475842818337e-06,
+      "loss": 0.0819,
+      "step": 1107
+    },
+    {
+      "epoch": 4.494177215189874,
+      "grad_norm": 0.24911412962853877,
+      "learning_rate": 1.1868106817802816e-06,
+      "loss": 0.0669,
+      "step": 1108
+    },
+    {
+      "epoch": 4.498227848101266,
+      "grad_norm": 0.22679869107170525,
+      "learning_rate": 1.1676252981191482e-06,
+      "loss": 0.0869,
+      "step": 1109
+    },
+    {
+      "epoch": 4.502278481012659,
+      "grad_norm": 0.23815103229605777,
+      "learning_rate": 1.1485915878148823e-06,
+      "loss": 0.0817,
+      "step": 1110
+    },
+    {
+      "epoch": 4.506329113924051,
+      "grad_norm": 0.24405346806708358,
+      "learning_rate": 1.1297097041623584e-06,
+      "loss": 0.0829,
+      "step": 1111
+    },
+    {
+      "epoch": 4.510379746835443,
+      "grad_norm": 0.23515618989122336,
+      "learning_rate": 1.1109797992336847e-06,
+      "loss": 0.077,
+      "step": 1112
+    },
+    {
+      "epoch": 4.514430379746836,
+      "grad_norm": 0.2239608642030838,
+      "learning_rate": 1.092402023876933e-06,
+      "loss": 0.0756,
+      "step": 1113
+    },
+    {
+      "epoch": 4.518481012658228,
+      "grad_norm": 0.2318214963733561,
+      "learning_rate": 1.0739765277149527e-06,
+      "loss": 0.0746,
+      "step": 1114
+    },
+    {
+      "epoch": 4.52253164556962,
+      "grad_norm": 0.22986864155849815,
+      "learning_rate": 1.0557034591441596e-06,
+      "loss": 0.0762,
+      "step": 1115
+    },
+    {
+      "epoch": 4.526582278481013,
+      "grad_norm": 0.22943265866166043,
+      "learning_rate": 1.0375829653333324e-06,
+      "loss": 0.0807,
+      "step": 1116
+    },
+    {
+      "epoch": 4.530632911392405,
+      "grad_norm": 0.2398067348140117,
+      "learning_rate": 1.0196151922224385e-06,
+      "loss": 0.0809,
+      "step": 1117
+    },
+    {
+      "epoch": 4.534683544303798,
+      "grad_norm": 0.2492087076688112,
+      "learning_rate": 1.0018002845214526e-06,
+      "loss": 0.0835,
+      "step": 1118
+    },
+    {
+      "epoch": 4.53873417721519,
+      "grad_norm": 0.23636270518718094,
+      "learning_rate": 9.841383857091947e-07,
+      "loss": 0.0823,
+      "step": 1119
+    },
+    {
+      "epoch": 4.542784810126582,
+      "grad_norm": 0.24388078823177178,
+      "learning_rate": 9.666296380321616e-07,
+      "loss": 0.0831,
+      "step": 1120
+    },
+    {
+      "epoch": 4.546835443037975,
+      "grad_norm": 0.23525990919754974,
+      "learning_rate": 9.492741825034124e-07,
+      "loss": 0.0804,
+      "step": 1121
+    },
+    {
+      "epoch": 4.550886075949367,
+      "grad_norm": 0.23580044619948906,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.0754,
+      "step": 1122
+    },
+    {
+      "epoch": 4.55493670886076,
+      "grad_norm": 0.23354410087991356,
+      "learning_rate": 9.150237057688339e-07,
+      "loss": 0.0744,
+      "step": 1123
+    },
+    {
+      "epoch": 4.558987341772152,
+      "grad_norm": 0.21818880866889467,
+      "learning_rate": 8.981289604116328e-07,
+      "loss": 0.0824,
+      "step": 1124
+    },
+    {
+      "epoch": 4.563037974683544,
+      "grad_norm": 0.23227484802076867,
+      "learning_rate": 8.813880588977542e-07,
+      "loss": 0.0864,
+      "step": 1125
+    },
+    {
+      "epoch": 4.567088607594937,
+      "grad_norm": 0.23302216335531303,
+      "learning_rate": 8.648011360561126e-07,
+      "loss": 0.0745,
+      "step": 1126
+    },
+    {
+      "epoch": 4.571139240506329,
+      "grad_norm": 0.22860616519953594,
+      "learning_rate": 8.483683254755037e-07,
+      "loss": 0.0783,
+      "step": 1127
+    },
+    {
+      "epoch": 4.575189873417721,
+      "grad_norm": 0.24360322869123793,
+      "learning_rate": 8.320897595035227e-07,
+      "loss": 0.084,
+      "step": 1128
+    },
+    {
+      "epoch": 4.579240506329114,
+      "grad_norm": 0.24879739337736798,
+      "learning_rate": 8.159655692455093e-07,
+      "loss": 0.0823,
+      "step": 1129
+    },
+    {
+      "epoch": 4.583291139240506,
+      "grad_norm": 0.24105248867471854,
+      "learning_rate": 7.999958845634648e-07,
+      "loss": 0.076,
+      "step": 1130
+    },
+    {
+      "epoch": 4.587341772151898,
+      "grad_norm": 0.24048638879401701,
+      "learning_rate": 7.841808340750478e-07,
+      "loss": 0.083,
+      "step": 1131
+    },
+    {
+      "epoch": 4.591392405063291,
+      "grad_norm": 0.22897214451916048,
+      "learning_rate": 7.685205451524869e-07,
+      "loss": 0.0782,
+      "step": 1132
+    },
+    {
+      "epoch": 4.595443037974683,
+      "grad_norm": 0.22369916000573034,
+      "learning_rate": 7.530151439216027e-07,
+      "loss": 0.0846,
+      "step": 1133
+    },
+    {
+      "epoch": 4.599493670886076,
+      "grad_norm": 0.2420027890552323,
+      "learning_rate": 7.376647552607675e-07,
+      "loss": 0.0823,
+      "step": 1134
+    },
+    {
+      "epoch": 4.603544303797468,
+      "grad_norm": 0.24309388882672223,
+      "learning_rate": 7.224695027998963e-07,
+      "loss": 0.0777,
+      "step": 1135
+    },
+    {
+      "epoch": 4.6075949367088604,
+      "grad_norm": 0.23529759750690807,
+      "learning_rate": 7.07429508919466e-07,
+      "loss": 0.0887,
+      "step": 1136
+    },
+    {
+      "epoch": 4.611645569620253,
+      "grad_norm": 0.23121097506665622,
+      "learning_rate": 6.925448947495206e-07,
+      "loss": 0.0787,
+      "step": 1137
+    },
+    {
+      "epoch": 4.6156962025316455,
+      "grad_norm": 0.25247201610439896,
+      "learning_rate": 6.778157801686936e-07,
+      "loss": 0.0848,
+      "step": 1138
+    },
+    {
+      "epoch": 4.619746835443038,
+      "grad_norm": 0.23043103549658073,
+      "learning_rate": 6.632422838032515e-07,
+      "loss": 0.0766,
+      "step": 1139
+    },
+    {
+      "epoch": 4.6237974683544305,
+      "grad_norm": 0.2289936270586846,
+      "learning_rate": 6.488245230261281e-07,
+      "loss": 0.0724,
+      "step": 1140
+    },
+    {
+      "epoch": 4.627848101265823,
+      "grad_norm": 0.22777523881494885,
+      "learning_rate": 6.345626139559868e-07,
+      "loss": 0.0772,
+      "step": 1141
+    },
+    {
+      "epoch": 4.6318987341772155,
+      "grad_norm": 0.23572747262202934,
+      "learning_rate": 6.204566714562866e-07,
+      "loss": 0.0775,
+      "step": 1142
+    },
+    {
+      "epoch": 4.635949367088608,
+      "grad_norm": 0.23483832556068485,
+      "learning_rate": 6.06506809134344e-07,
+      "loss": 0.0909,
+      "step": 1143
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.22032123552666966,
+      "learning_rate": 5.927131393404373e-07,
+      "loss": 0.0808,
+      "step": 1144
+    },
+    {
+      "epoch": 4.644050632911393,
+      "grad_norm": 0.23897670939096316,
+      "learning_rate": 5.790757731668817e-07,
+      "loss": 0.0727,
+      "step": 1145
+    },
+    {
+      "epoch": 4.648101265822785,
+      "grad_norm": 0.22518754215494804,
+      "learning_rate": 5.655948204471507e-07,
+      "loss": 0.0841,
+      "step": 1146
+    },
+    {
+      "epoch": 4.652151898734177,
+      "grad_norm": 0.23086610568796248,
+      "learning_rate": 5.522703897549875e-07,
+      "loss": 0.0767,
+      "step": 1147
+    },
+    {
+      "epoch": 4.65620253164557,
+      "grad_norm": 0.21950154169593888,
+      "learning_rate": 5.391025884035239e-07,
+      "loss": 0.0865,
+      "step": 1148
+    },
+    {
+      "epoch": 4.660253164556962,
+      "grad_norm": 0.2313290302788973,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.0829,
+      "step": 1149
+    },
+    {
+      "epoch": 4.664303797468355,
+      "grad_norm": 0.23987189093079334,
+      "learning_rate": 5.132372966670129e-07,
+      "loss": 0.0835,
+      "step": 1150
+    },
+    {
+      "epoch": 4.668354430379747,
+      "grad_norm": 0.23556784661945596,
+      "learning_rate": 5.005400145974704e-07,
+      "loss": 0.0887,
+      "step": 1151
+    },
+    {
+      "epoch": 4.672405063291139,
+      "grad_norm": 0.24166117190870579,
+      "learning_rate": 4.879997784979562e-07,
+      "loss": 0.077,
+      "step": 1152
+    },
+    {
+      "epoch": 4.676455696202532,
+      "grad_norm": 0.2360749211222573,
+      "learning_rate": 4.7561668936580984e-07,
+      "loss": 0.0868,
+      "step": 1153
+    },
+    {
+      "epoch": 4.680506329113924,
+      "grad_norm": 0.24513810202286418,
+      "learning_rate": 4.6339084693272306e-07,
+      "loss": 0.0759,
+      "step": 1154
+    },
+    {
+      "epoch": 4.684556962025317,
+      "grad_norm": 0.22533697506361913,
+      "learning_rate": 4.5132234966395847e-07,
+      "loss": 0.064,
+      "step": 1155
+    },
+    {
+      "epoch": 4.688607594936709,
+      "grad_norm": 0.2471830108382937,
+      "learning_rate": 4.3941129475752795e-07,
+      "loss": 0.0881,
+      "step": 1156
+    },
+    {
+      "epoch": 4.692658227848101,
+      "grad_norm": 0.29884631870831324,
+      "learning_rate": 4.27657778143431e-07,
+      "loss": 0.0783,
+      "step": 1157
+    },
+    {
+      "epoch": 4.696708860759494,
+      "grad_norm": 0.23197305957803283,
+      "learning_rate": 4.1606189448287757e-07,
+      "loss": 0.0758,
+      "step": 1158
+    },
+    {
+      "epoch": 4.700759493670886,
+      "grad_norm": 0.2379614296627773,
+      "learning_rate": 4.046237371675177e-07,
+      "loss": 0.0841,
+      "step": 1159
+    },
+    {
+      "epoch": 4.704810126582278,
+      "grad_norm": 0.2508185541676351,
+      "learning_rate": 3.9334339831869963e-07,
+      "loss": 0.0841,
+      "step": 1160
+    },
+    {
+      "epoch": 4.708860759493671,
+      "grad_norm": 0.22982209411749416,
+      "learning_rate": 3.8222096878671955e-07,
+      "loss": 0.0783,
+      "step": 1161
+    },
+    {
+      "epoch": 4.712911392405063,
+      "grad_norm": 0.23006253021023704,
+      "learning_rate": 3.7125653815009545e-07,
+      "loss": 0.0842,
+      "step": 1162
+    },
+    {
+      "epoch": 4.716962025316455,
+      "grad_norm": 0.2441099748539755,
+      "learning_rate": 3.6045019471484974e-07,
+      "loss": 0.083,
+      "step": 1163
+    },
+    {
+      "epoch": 4.721012658227848,
+      "grad_norm": 0.23117801610820454,
+      "learning_rate": 3.498020255137813e-07,
+      "loss": 0.072,
+      "step": 1164
+    },
+    {
+      "epoch": 4.72506329113924,
+      "grad_norm": 0.23815251097114437,
+      "learning_rate": 3.393121163057811e-07,
+      "loss": 0.086,
+      "step": 1165
+    },
+    {
+      "epoch": 4.729113924050633,
+      "grad_norm": 0.232164251877344,
+      "learning_rate": 3.289805515751399e-07,
+      "loss": 0.0778,
+      "step": 1166
+    },
+    {
+      "epoch": 4.733164556962025,
+      "grad_norm": 0.22978219872707484,
+      "learning_rate": 3.188074145308573e-07,
+      "loss": 0.09,
+      "step": 1167
+    },
+    {
+      "epoch": 4.737215189873417,
+      "grad_norm": 0.23698185599333121,
+      "learning_rate": 3.087927871059804e-07,
+      "loss": 0.084,
+      "step": 1168
+    },
+    {
+      "epoch": 4.74126582278481,
+      "grad_norm": 0.22211363484571695,
+      "learning_rate": 2.989367499569418e-07,
+      "loss": 0.0801,
+      "step": 1169
+    },
+    {
+      "epoch": 4.745316455696202,
+      "grad_norm": 0.26102934700551655,
+      "learning_rate": 2.8923938246290917e-07,
+      "loss": 0.0897,
+      "step": 1170
+    },
+    {
+      "epoch": 4.749367088607595,
+      "grad_norm": 0.23197235859671206,
+      "learning_rate": 2.7970076272514804e-07,
+      "loss": 0.0688,
+      "step": 1171
+    },
+    {
+      "epoch": 4.7534177215189874,
+      "grad_norm": 0.22384326146370093,
+      "learning_rate": 2.703209675663887e-07,
+      "loss": 0.0763,
+      "step": 1172
+    },
+    {
+      "epoch": 4.7574683544303795,
+      "grad_norm": 0.22086984827172834,
+      "learning_rate": 2.6110007253021374e-07,
+      "loss": 0.0782,
+      "step": 1173
+    },
+    {
+      "epoch": 4.7615189873417725,
+      "grad_norm": 0.23121823086864485,
+      "learning_rate": 2.520381518804471e-07,
+      "loss": 0.0825,
+      "step": 1174
+    },
+    {
+      "epoch": 4.7655696202531646,
+      "grad_norm": 0.24356393659874193,
+      "learning_rate": 2.4313527860054585e-07,
+      "loss": 0.0875,
+      "step": 1175
+    },
+    {
+      "epoch": 4.769620253164557,
+      "grad_norm": 0.24199463525259446,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.1013,
+      "step": 1176
+    },
+    {
+      "epoch": 4.77367088607595,
+      "grad_norm": 0.23519973760332957,
+      "learning_rate": 2.2580695967889367e-07,
+      "loss": 0.0929,
+      "step": 1177
+    },
+    {
+      "epoch": 4.777721518987342,
+      "grad_norm": 0.23068933136148598,
+      "learning_rate": 2.1738165359704189e-07,
+      "loss": 0.076,
+      "step": 1178
+    },
+    {
+      "epoch": 4.781772151898734,
+      "grad_norm": 0.23185051071155102,
+      "learning_rate": 2.0911567400373257e-07,
+      "loss": 0.0864,
+      "step": 1179
+    },
+    {
+      "epoch": 4.785822784810127,
+      "grad_norm": 0.24454835487535287,
+      "learning_rate": 2.0100908747202607e-07,
+      "loss": 0.0934,
+      "step": 1180
+    },
+    {
+      "epoch": 4.789873417721519,
+      "grad_norm": 0.22904413901752688,
+      "learning_rate": 1.9306195929125638e-07,
+      "loss": 0.0687,
+      "step": 1181
+    },
+    {
+      "epoch": 4.793924050632912,
+      "grad_norm": 0.23726537194404929,
+      "learning_rate": 1.8527435346650247e-07,
+      "loss": 0.0785,
+      "step": 1182
+    },
+    {
+      "epoch": 4.797974683544304,
+      "grad_norm": 0.23462503405322513,
+      "learning_rate": 1.7764633271807108e-07,
+      "loss": 0.0704,
+      "step": 1183
+    },
+    {
+      "epoch": 4.802025316455696,
+      "grad_norm": 0.23947092636847805,
+      "learning_rate": 1.7017795848099262e-07,
+      "loss": 0.072,
+      "step": 1184
+    },
+    {
+      "epoch": 4.806075949367089,
+      "grad_norm": 0.22438127120707696,
+      "learning_rate": 1.6286929090452596e-07,
+      "loss": 0.0851,
+      "step": 1185
+    },
+    {
+      "epoch": 4.810126582278481,
+      "grad_norm": 0.2246420054203143,
+      "learning_rate": 1.557203888516745e-07,
+      "loss": 0.075,
+      "step": 1186
+    },
+    {
+      "epoch": 4.814177215189874,
+      "grad_norm": 0.242776293074344,
+      "learning_rate": 1.487313098987131e-07,
+      "loss": 0.0897,
+      "step": 1187
+    },
+    {
+      "epoch": 4.818227848101266,
+      "grad_norm": 0.23293573213670826,
+      "learning_rate": 1.4190211033472402e-07,
+      "loss": 0.0835,
+      "step": 1188
+    },
+    {
+      "epoch": 4.822278481012658,
+      "grad_norm": 0.2351474025647129,
+      "learning_rate": 1.3523284516113955e-07,
+      "loss": 0.0774,
+      "step": 1189
+    },
+    {
+      "epoch": 4.826329113924051,
+      "grad_norm": 0.2538916979644035,
+      "learning_rate": 1.2872356809130682e-07,
+      "loss": 0.0806,
+      "step": 1190
+    },
+    {
+      "epoch": 4.830379746835443,
+      "grad_norm": 0.23932005340058976,
+      "learning_rate": 1.2237433155004807e-07,
+      "loss": 0.0757,
+      "step": 1191
+    },
+    {
+      "epoch": 4.834430379746835,
+      "grad_norm": 0.22588187666781648,
+      "learning_rate": 1.1618518667323886e-07,
+      "loss": 0.088,
+      "step": 1192
+    },
+    {
+      "epoch": 4.838481012658228,
+      "grad_norm": 0.22566289427260808,
+      "learning_rate": 1.1015618330740385e-07,
+      "loss": 0.0833,
+      "step": 1193
+    },
+    {
+      "epoch": 4.84253164556962,
+      "grad_norm": 0.22383655382513643,
+      "learning_rate": 1.042873700093061e-07,
+      "loss": 0.0724,
+      "step": 1194
+    },
+    {
+      "epoch": 4.846582278481012,
+      "grad_norm": 0.23411692649383614,
+      "learning_rate": 9.857879404556291e-08,
+      "loss": 0.1016,
+      "step": 1195
+    },
+    {
+      "epoch": 4.850632911392405,
+      "grad_norm": 0.23636796219737458,
+      "learning_rate": 9.303050139225722e-08,
+      "loss": 0.0867,
+      "step": 1196
+    },
+    {
+      "epoch": 4.854683544303797,
+      "grad_norm": 0.22872945605436876,
+      "learning_rate": 8.76425367345779e-08,
+      "loss": 0.0791,
+      "step": 1197
+    },
+    {
+      "epoch": 4.85873417721519,
+      "grad_norm": 0.22412375149931102,
+      "learning_rate": 8.241494346644897e-08,
+      "loss": 0.0857,
+      "step": 1198
+    },
+    {
+      "epoch": 4.862784810126582,
+      "grad_norm": 0.22747367061862694,
+      "learning_rate": 7.734776369019204e-08,
+      "loss": 0.0821,
+      "step": 1199
+    },
+    {
+      "epoch": 4.866835443037974,
+      "grad_norm": 0.2354631369179892,
+      "learning_rate": 7.244103821617332e-08,
+      "loss": 0.0758,
+      "step": 1200
+    },
+    {
+      "epoch": 4.870886075949367,
+      "grad_norm": 0.2440389205155095,
+      "learning_rate": 6.769480656248606e-08,
+      "loss": 0.0767,
+      "step": 1201
+    },
+    {
+      "epoch": 4.874936708860759,
+      "grad_norm": 0.23386582913961212,
+      "learning_rate": 6.310910695462635e-08,
+      "loss": 0.0819,
+      "step": 1202
+    },
+    {
+      "epoch": 4.878987341772152,
+      "grad_norm": 0.22431030250370496,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.0696,
+      "step": 1203
+    },
+    {
+      "epoch": 4.883037974683544,
+      "grad_norm": 0.24028066589202357,
+      "learning_rate": 5.4419450313571984e-08,
+      "loss": 0.0808,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8870886075949365,
+      "grad_norm": 0.22674708108536895,
+      "learning_rate": 5.031556326567488e-08,
+      "loss": 0.0826,
+      "step": 1205
+    },
+    {
+      "epoch": 4.891139240506329,
+      "grad_norm": 0.2298330812084933,
+      "learning_rate": 4.637234823364312e-08,
+      "loss": 0.0747,
+      "step": 1206
+    },
+    {
+      "epoch": 4.8951898734177215,
+      "grad_norm": 0.2349717910521498,
+      "learning_rate": 4.258983697558838e-08,
+      "loss": 0.0932,
+      "step": 1207
+    },
+    {
+      "epoch": 4.899240506329114,
+      "grad_norm": 0.23208116702782772,
+      "learning_rate": 3.896805995533548e-08,
+      "loss": 0.0887,
+      "step": 1208
+    },
+    {
+      "epoch": 4.9032911392405065,
+      "grad_norm": 0.23147345990823992,
+      "learning_rate": 3.550704634218028e-08,
+      "loss": 0.0744,
+      "step": 1209
+    },
+    {
+      "epoch": 4.907341772151899,
+      "grad_norm": 0.2269741272670773,
+      "learning_rate": 3.2206824010647676e-08,
+      "loss": 0.0924,
+      "step": 1210
+    },
+    {
+      "epoch": 4.911392405063291,
+      "grad_norm": 0.2397222337328691,
+      "learning_rate": 2.9067419540278476e-08,
+      "loss": 0.0783,
+      "step": 1211
+    },
+    {
+      "epoch": 4.915443037974684,
+      "grad_norm": 0.2294943301011067,
+      "learning_rate": 2.6088858215400638e-08,
+      "loss": 0.0864,
+      "step": 1212
+    },
+    {
+      "epoch": 4.919493670886076,
+      "grad_norm": 0.22720707851365066,
+      "learning_rate": 2.3271164024940564e-08,
+      "loss": 0.0788,
+      "step": 1213
+    },
+    {
+      "epoch": 4.923544303797469,
+      "grad_norm": 0.23997651827057367,
+      "learning_rate": 2.061435966221881e-08,
+      "loss": 0.0882,
+      "step": 1214
+    },
+    {
+      "epoch": 4.927594936708861,
+      "grad_norm": 0.22359710499569385,
+      "learning_rate": 1.811846652477245e-08,
+      "loss": 0.0933,
+      "step": 1215
+    },
+    {
+      "epoch": 4.931645569620253,
+      "grad_norm": 0.2285037247165017,
+      "learning_rate": 1.5783504714184106e-08,
+      "loss": 0.0883,
+      "step": 1216
+    },
+    {
+      "epoch": 4.935696202531646,
+      "grad_norm": 0.23397087042290374,
+      "learning_rate": 1.360949303591097e-08,
+      "loss": 0.0847,
+      "step": 1217
+    },
+    {
+      "epoch": 4.939746835443038,
+      "grad_norm": 0.22556136484863343,
+      "learning_rate": 1.1596448999144916e-08,
+      "loss": 0.0732,
+      "step": 1218
+    },
+    {
+      "epoch": 4.943797468354431,
+      "grad_norm": 0.2326133236994794,
+      "learning_rate": 9.744388816668172e-09,
+      "loss": 0.0861,
+      "step": 1219
+    },
+    {
+      "epoch": 4.947848101265823,
+      "grad_norm": 0.23263367218257147,
+      "learning_rate": 8.05332740472009e-09,
+      "loss": 0.0831,
+      "step": 1220
+    },
+    {
+      "epoch": 4.951898734177215,
+      "grad_norm": 0.22800922761530193,
+      "learning_rate": 6.523278382872811e-09,
+      "loss": 0.0942,
+      "step": 1221
+    },
+    {
+      "epoch": 4.955949367088608,
+      "grad_norm": 0.23343434465812946,
+      "learning_rate": 5.15425407393133e-09,
+      "loss": 0.0747,
+      "step": 1222
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2291786666680766,
+      "learning_rate": 3.94626550383137e-09,
+      "loss": 0.0692,
+      "step": 1223
+    },
+    {
+      "epoch": 4.964050632911392,
+      "grad_norm": 0.23667002913132554,
+      "learning_rate": 2.899322401546112e-09,
+      "loss": 0.0844,
+      "step": 1224
+    },
+    {
+      "epoch": 4.968101265822785,
+      "grad_norm": 0.24758455970902107,
+      "learning_rate": 2.013433199010706e-09,
+      "loss": 0.0764,
+      "step": 1225
+    },
+    {
+      "epoch": 4.972151898734177,
+      "grad_norm": 0.2336192111379587,
+      "learning_rate": 1.2886050310556563e-09,
+      "loss": 0.0854,
+      "step": 1226
+    },
+    {
+      "epoch": 4.976202531645569,
+      "grad_norm": 0.23649862571171182,
+      "learning_rate": 7.248437353468695e-10,
+      "loss": 0.0794,
+      "step": 1227
+    },
+    {
+      "epoch": 4.980253164556962,
+      "grad_norm": 0.23345200117621007,
+      "learning_rate": 3.221538523412449e-10,
+      "loss": 0.0715,
+      "step": 1228
+    },
+    {
+      "epoch": 4.984303797468354,
+      "grad_norm": 0.2392972908835394,
+      "learning_rate": 8.053862524670663e-11,
+      "loss": 0.0779,
+      "step": 1229
+    },
+    {
+      "epoch": 4.988354430379747,
+      "grad_norm": 0.2449693057284251,
+      "learning_rate": 0.0,
+      "loss": 0.0821,
+      "step": 1230
+    },
+    {
+      "epoch": 4.988354430379747,
+      "step": 1230,
+      "total_flos": 1.49639705657344e+18,
+      "train_loss": 0.21159650911160602,
+      "train_runtime": 51604.1878,
+      "train_samples_per_second": 3.062,
+      "train_steps_per_second": 0.024
+    }
+  ],
+  "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": 1.49639705657344e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}