{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.948717948717949,
  "eval_steps": 500,
  "global_step": 970,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020512820512820513,
      "grad_norm": 7.845447540283203,
      "learning_rate": 9.989690721649485e-05,
      "loss": 9.5248,
      "step": 2
    },
    {
      "epoch": 0.041025641025641026,
      "grad_norm": 8.389087677001953,
      "learning_rate": 9.969072164948454e-05,
      "loss": 9.1949,
      "step": 4
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 5.169976711273193,
      "learning_rate": 9.948453608247423e-05,
      "loss": 8.6566,
      "step": 6
    },
    {
      "epoch": 0.08205128205128205,
      "grad_norm": 4.878049850463867,
      "learning_rate": 9.927835051546392e-05,
      "loss": 8.1281,
      "step": 8
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 4.062979698181152,
      "learning_rate": 9.907216494845362e-05,
      "loss": 7.7464,
      "step": 10
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 2.26177716255188,
      "learning_rate": 9.88659793814433e-05,
      "loss": 7.4979,
      "step": 12
    },
    {
      "epoch": 0.14358974358974358,
      "grad_norm": 1.6796530485153198,
      "learning_rate": 9.8659793814433e-05,
      "loss": 7.4101,
      "step": 14
    },
    {
      "epoch": 0.1641025641025641,
      "grad_norm": 1.3691627979278564,
      "learning_rate": 9.845360824742269e-05,
      "loss": 7.2759,
      "step": 16
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 1.8158212900161743,
      "learning_rate": 9.824742268041237e-05,
      "loss": 7.1752,
      "step": 18
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 1.489174246788025,
      "learning_rate": 9.804123711340207e-05,
      "loss": 7.2658,
      "step": 20
    },
    {
      "epoch": 0.22564102564102564,
      "grad_norm": 1.363490104675293,
      "learning_rate": 9.783505154639175e-05,
      "loss": 7.1867,
      "step": 22
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 1.5879547595977783,
      "learning_rate": 9.762886597938145e-05,
      "loss": 7.1374,
      "step": 24
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 1.046538233757019,
      "learning_rate": 9.742268041237114e-05,
      "loss": 7.106,
      "step": 26
    },
    {
      "epoch": 0.28717948717948716,
      "grad_norm": 1.169195532798767,
      "learning_rate": 9.721649484536083e-05,
      "loss": 7.1945,
      "step": 28
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.8861628174781799,
      "learning_rate": 9.701030927835052e-05,
      "loss": 7.0199,
      "step": 30
    },
    {
      "epoch": 0.3282051282051282,
      "grad_norm": 1.4861819744110107,
      "learning_rate": 9.680412371134021e-05,
      "loss": 6.8774,
      "step": 32
    },
    {
      "epoch": 0.3487179487179487,
      "grad_norm": 0.9394097924232483,
      "learning_rate": 9.65979381443299e-05,
      "loss": 6.9136,
      "step": 34
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 1.1648708581924438,
      "learning_rate": 9.639175257731959e-05,
      "loss": 7.121,
      "step": 36
    },
    {
      "epoch": 0.38974358974358975,
      "grad_norm": 0.8142098784446716,
      "learning_rate": 9.618556701030928e-05,
      "loss": 7.1084,
      "step": 38
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 1.0768697261810303,
      "learning_rate": 9.597938144329898e-05,
      "loss": 7.0565,
      "step": 40
    },
    {
      "epoch": 0.4307692307692308,
      "grad_norm": 0.6553409099578857,
      "learning_rate": 9.577319587628867e-05,
      "loss": 6.9963,
      "step": 42
    },
    {
      "epoch": 0.4512820512820513,
      "grad_norm": 0.8519483208656311,
      "learning_rate": 9.556701030927836e-05,
      "loss": 7.0575,
      "step": 44
    },
    {
      "epoch": 0.4717948717948718,
      "grad_norm": 1.1110036373138428,
      "learning_rate": 9.536082474226805e-05,
      "loss": 6.998,
      "step": 46
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 0.8143446445465088,
      "learning_rate": 9.515463917525773e-05,
      "loss": 6.9762,
      "step": 48
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 1.4277548789978027,
      "learning_rate": 9.494845360824743e-05,
      "loss": 6.9832,
      "step": 50
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.6880565881729126,
      "learning_rate": 9.474226804123711e-05,
      "loss": 7.0254,
      "step": 52
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 0.8316001296043396,
      "learning_rate": 9.45360824742268e-05,
      "loss": 6.9704,
      "step": 54
    },
    {
      "epoch": 0.5743589743589743,
      "grad_norm": 0.91054368019104,
      "learning_rate": 9.43298969072165e-05,
      "loss": 6.9184,
      "step": 56
    },
    {
      "epoch": 0.5948717948717949,
      "grad_norm": 0.7943957448005676,
      "learning_rate": 9.412371134020619e-05,
      "loss": 6.9518,
      "step": 58
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 1.0420604944229126,
      "learning_rate": 9.391752577319588e-05,
      "loss": 6.9369,
      "step": 60
    },
    {
      "epoch": 0.6358974358974359,
      "grad_norm": 0.7538536787033081,
      "learning_rate": 9.371134020618557e-05,
      "loss": 6.8748,
      "step": 62
    },
    {
      "epoch": 0.6564102564102564,
      "grad_norm": 1.2202829122543335,
      "learning_rate": 9.350515463917526e-05,
      "loss": 6.8046,
      "step": 64
    },
    {
      "epoch": 0.676923076923077,
      "grad_norm": 1.0980966091156006,
      "learning_rate": 9.329896907216495e-05,
      "loss": 6.8146,
      "step": 66
    },
    {
      "epoch": 0.6974358974358974,
      "grad_norm": 1.5506396293640137,
      "learning_rate": 9.309278350515465e-05,
      "loss": 7.0011,
      "step": 68
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 1.1768453121185303,
      "learning_rate": 9.288659793814434e-05,
      "loss": 6.8345,
      "step": 70
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 1.1223098039627075,
      "learning_rate": 9.268041237113403e-05,
      "loss": 6.8371,
      "step": 72
    },
    {
      "epoch": 0.7589743589743589,
      "grad_norm": 0.7969951629638672,
      "learning_rate": 9.247422680412372e-05,
      "loss": 6.902,
      "step": 74
    },
    {
      "epoch": 0.7794871794871795,
      "grad_norm": 0.8236146569252014,
      "learning_rate": 9.22680412371134e-05,
      "loss": 6.8724,
      "step": 76
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.0809398889541626,
      "learning_rate": 9.206185567010309e-05,
      "loss": 6.8455,
      "step": 78
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 1.0138286352157593,
      "learning_rate": 9.185567010309279e-05,
      "loss": 6.8771,
      "step": 80
    },
    {
      "epoch": 0.841025641025641,
      "grad_norm": 1.5918043851852417,
      "learning_rate": 9.164948453608247e-05,
      "loss": 6.8593,
      "step": 82
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 1.9432188272476196,
      "learning_rate": 9.144329896907217e-05,
      "loss": 6.8676,
      "step": 84
    },
    {
      "epoch": 0.882051282051282,
      "grad_norm": 1.5236680507659912,
      "learning_rate": 9.123711340206186e-05,
      "loss": 6.6954,
      "step": 86
    },
    {
      "epoch": 0.9025641025641026,
      "grad_norm": 1.347008228302002,
      "learning_rate": 9.103092783505155e-05,
      "loss": 6.7739,
      "step": 88
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.8976010680198669,
      "learning_rate": 9.082474226804124e-05,
      "loss": 6.8272,
      "step": 90
    },
    {
      "epoch": 0.9435897435897436,
      "grad_norm": 1.0713074207305908,
      "learning_rate": 9.061855670103093e-05,
      "loss": 6.722,
      "step": 92
    },
    {
      "epoch": 0.9641025641025641,
      "grad_norm": 1.3298990726470947,
      "learning_rate": 9.041237113402063e-05,
      "loss": 6.7779,
      "step": 94
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 1.3666430711746216,
      "learning_rate": 9.020618556701031e-05,
      "loss": 6.6804,
      "step": 96
    },
    {
      "epoch": 1.005128205128205,
      "grad_norm": 1.5196622610092163,
      "learning_rate": 9e-05,
      "loss": 6.7935,
      "step": 98
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 1.0106366872787476,
      "learning_rate": 8.97938144329897e-05,
      "loss": 6.7578,
      "step": 100
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 0.9565912485122681,
      "learning_rate": 8.958762886597939e-05,
      "loss": 6.81,
      "step": 102
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.7427358627319336,
      "learning_rate": 8.938144329896908e-05,
      "loss": 6.7069,
      "step": 104
    },
    {
      "epoch": 1.087179487179487,
      "grad_norm": 1.8422271013259888,
      "learning_rate": 8.917525773195877e-05,
      "loss": 6.7098,
      "step": 106
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 0.5381059050559998,
      "learning_rate": 8.896907216494845e-05,
      "loss": 6.7934,
      "step": 108
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 0.8050350546836853,
      "learning_rate": 8.876288659793815e-05,
      "loss": 6.6336,
      "step": 110
    },
    {
      "epoch": 1.1487179487179486,
      "grad_norm": 0.570678174495697,
      "learning_rate": 8.855670103092783e-05,
      "loss": 6.8265,
      "step": 112
    },
    {
      "epoch": 1.1692307692307693,
      "grad_norm": 0.5224724411964417,
      "learning_rate": 8.835051546391753e-05,
      "loss": 6.9796,
      "step": 114
    },
    {
      "epoch": 1.1897435897435897,
      "grad_norm": 0.6684609651565552,
      "learning_rate": 8.814432989690722e-05,
      "loss": 6.701,
      "step": 116
    },
    {
      "epoch": 1.2102564102564102,
      "grad_norm": 0.9216952919960022,
      "learning_rate": 8.793814432989691e-05,
      "loss": 6.819,
      "step": 118
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.925770103931427,
      "learning_rate": 8.77319587628866e-05,
      "loss": 6.6087,
      "step": 120
    },
    {
      "epoch": 1.2512820512820513,
      "grad_norm": 0.6525481343269348,
      "learning_rate": 8.752577319587629e-05,
      "loss": 6.7305,
      "step": 122
    },
    {
      "epoch": 1.2717948717948717,
      "grad_norm": 0.7587235569953918,
      "learning_rate": 8.731958762886599e-05,
      "loss": 6.6621,
      "step": 124
    },
    {
      "epoch": 1.2923076923076924,
      "grad_norm": 0.7010625600814819,
      "learning_rate": 8.711340206185567e-05,
      "loss": 6.6443,
      "step": 126
    },
    {
      "epoch": 1.3128205128205128,
      "grad_norm": 0.7843243479728699,
      "learning_rate": 8.690721649484537e-05,
      "loss": 6.7056,
      "step": 128
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.8412699699401855,
      "learning_rate": 8.670103092783506e-05,
      "loss": 6.8242,
      "step": 130
    },
    {
      "epoch": 1.353846153846154,
      "grad_norm": 0.899800181388855,
      "learning_rate": 8.649484536082475e-05,
      "loss": 6.7249,
      "step": 132
    },
    {
      "epoch": 1.3743589743589744,
      "grad_norm": 1.413313388824463,
      "learning_rate": 8.628865979381444e-05,
      "loss": 6.6534,
      "step": 134
    },
    {
      "epoch": 1.3948717948717948,
      "grad_norm": 1.53885018825531,
      "learning_rate": 8.608247422680413e-05,
      "loss": 6.6828,
      "step": 136
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 0.7499423027038574,
      "learning_rate": 8.587628865979381e-05,
      "loss": 6.5888,
      "step": 138
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 0.48903796076774597,
      "learning_rate": 8.567010309278351e-05,
      "loss": 6.7217,
      "step": 140
    },
    {
      "epoch": 1.4564102564102563,
      "grad_norm": 0.5164963006973267,
      "learning_rate": 8.546391752577319e-05,
      "loss": 6.7929,
      "step": 142
    },
    {
      "epoch": 1.476923076923077,
      "grad_norm": 0.4488673210144043,
      "learning_rate": 8.525773195876289e-05,
      "loss": 6.7542,
      "step": 144
    },
    {
      "epoch": 1.4974358974358974,
      "grad_norm": 0.9296787977218628,
      "learning_rate": 8.505154639175259e-05,
      "loss": 6.595,
      "step": 146
    },
    {
      "epoch": 1.5179487179487179,
      "grad_norm": 0.48948934674263,
      "learning_rate": 8.484536082474227e-05,
      "loss": 6.6336,
      "step": 148
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 1.3344464302062988,
      "learning_rate": 8.463917525773197e-05,
      "loss": 6.5617,
      "step": 150
    },
    {
      "epoch": 1.558974358974359,
      "grad_norm": 0.8237192034721375,
      "learning_rate": 8.443298969072165e-05,
      "loss": 6.6697,
      "step": 152
    },
    {
      "epoch": 1.5794871794871796,
      "grad_norm": 1.013886570930481,
      "learning_rate": 8.422680412371135e-05,
      "loss": 6.5097,
      "step": 154
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5843191742897034,
      "learning_rate": 8.402061855670103e-05,
      "loss": 6.761,
      "step": 156
    },
    {
      "epoch": 1.6205128205128205,
      "grad_norm": 0.5963282585144043,
      "learning_rate": 8.381443298969073e-05,
      "loss": 6.8233,
      "step": 158
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 0.7131335735321045,
      "learning_rate": 8.360824742268042e-05,
      "loss": 6.5581,
      "step": 160
    },
    {
      "epoch": 1.6615384615384614,
      "grad_norm": 0.5245983004570007,
      "learning_rate": 8.340206185567011e-05,
      "loss": 6.5801,
      "step": 162
    },
    {
      "epoch": 1.682051282051282,
      "grad_norm": 0.6111565232276917,
      "learning_rate": 8.31958762886598e-05,
      "loss": 6.6157,
      "step": 164
    },
    {
      "epoch": 1.7025641025641025,
      "grad_norm": 0.7455165982246399,
      "learning_rate": 8.298969072164949e-05,
      "loss": 6.6316,
      "step": 166
    },
    {
      "epoch": 1.7230769230769232,
      "grad_norm": 0.7735164761543274,
      "learning_rate": 8.278350515463917e-05,
      "loss": 6.6299,
      "step": 168
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 0.5676758885383606,
      "learning_rate": 8.257731958762887e-05,
      "loss": 6.7204,
      "step": 170
    },
    {
      "epoch": 1.764102564102564,
      "grad_norm": 0.8821406364440918,
      "learning_rate": 8.237113402061855e-05,
      "loss": 6.3924,
      "step": 172
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 0.6726123094558716,
      "learning_rate": 8.216494845360825e-05,
      "loss": 6.7632,
      "step": 174
    },
    {
      "epoch": 1.8051282051282052,
      "grad_norm": 0.5056736469268799,
      "learning_rate": 8.195876288659795e-05,
      "loss": 6.5864,
      "step": 176
    },
    {
      "epoch": 1.8256410256410256,
      "grad_norm": 0.5093751549720764,
      "learning_rate": 8.175257731958763e-05,
      "loss": 6.5112,
      "step": 178
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.6652563810348511,
      "learning_rate": 8.154639175257733e-05,
      "loss": 6.5538,
      "step": 180
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 0.6594399213790894,
      "learning_rate": 8.134020618556701e-05,
      "loss": 6.6136,
      "step": 182
    },
    {
      "epoch": 1.8871794871794871,
      "grad_norm": 0.6552271246910095,
      "learning_rate": 8.113402061855671e-05,
      "loss": 6.4301,
      "step": 184
    },
    {
      "epoch": 1.9076923076923076,
      "grad_norm": 1.0746954679489136,
      "learning_rate": 8.092783505154639e-05,
      "loss": 6.3992,
      "step": 186
    },
    {
      "epoch": 1.9282051282051282,
      "grad_norm": 0.5134009122848511,
      "learning_rate": 8.072164948453609e-05,
      "loss": 6.5104,
      "step": 188
    },
    {
      "epoch": 1.9487179487179487,
      "grad_norm": 0.6844182014465332,
      "learning_rate": 8.051546391752578e-05,
      "loss": 6.4957,
      "step": 190
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 0.5666801929473877,
      "learning_rate": 8.030927835051547e-05,
      "loss": 6.679,
      "step": 192
    },
    {
      "epoch": 1.9897435897435898,
      "grad_norm": 0.7489727139472961,
      "learning_rate": 8.010309278350516e-05,
      "loss": 6.5199,
      "step": 194
    },
    {
      "epoch": 2.01025641025641,
      "grad_norm": 0.6453472971916199,
      "learning_rate": 7.989690721649485e-05,
      "loss": 6.5507,
      "step": 196
    },
    {
      "epoch": 2.0307692307692307,
      "grad_norm": 0.5807784199714661,
      "learning_rate": 7.969072164948455e-05,
      "loss": 6.7443,
      "step": 198
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 0.6182963848114014,
      "learning_rate": 7.948453608247423e-05,
      "loss": 6.4959,
      "step": 200
    },
    {
      "epoch": 2.071794871794872,
      "grad_norm": 0.579248309135437,
      "learning_rate": 7.927835051546391e-05,
      "loss": 6.597,
      "step": 202
    },
    {
      "epoch": 2.0923076923076924,
      "grad_norm": 0.5996450781822205,
      "learning_rate": 7.907216494845361e-05,
      "loss": 6.7087,
      "step": 204
    },
    {
      "epoch": 2.112820512820513,
      "grad_norm": 0.4910239577293396,
      "learning_rate": 7.88659793814433e-05,
      "loss": 6.666,
      "step": 206
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.6239397525787354,
      "learning_rate": 7.865979381443299e-05,
      "loss": 6.4757,
      "step": 208
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 0.581991970539093,
      "learning_rate": 7.845360824742269e-05,
      "loss": 6.4388,
      "step": 210
    },
    {
      "epoch": 2.174358974358974,
      "grad_norm": 0.6894115209579468,
      "learning_rate": 7.824742268041237e-05,
      "loss": 6.7599,
      "step": 212
    },
    {
      "epoch": 2.194871794871795,
      "grad_norm": 0.6936022043228149,
      "learning_rate": 7.804123711340207e-05,
      "loss": 6.5268,
      "step": 214
    },
    {
      "epoch": 2.2153846153846155,
      "grad_norm": 0.6604471802711487,
      "learning_rate": 7.783505154639175e-05,
      "loss": 6.5513,
      "step": 216
    },
    {
      "epoch": 2.235897435897436,
      "grad_norm": 0.6073073744773865,
      "learning_rate": 7.762886597938145e-05,
      "loss": 6.5175,
      "step": 218
    },
    {
      "epoch": 2.2564102564102564,
      "grad_norm": 0.5706610679626465,
      "learning_rate": 7.742268041237114e-05,
      "loss": 6.6232,
      "step": 220
    },
    {
      "epoch": 2.276923076923077,
      "grad_norm": 0.6433441042900085,
      "learning_rate": 7.721649484536083e-05,
      "loss": 6.324,
      "step": 222
    },
    {
      "epoch": 2.2974358974358973,
      "grad_norm": 0.6643648743629456,
      "learning_rate": 7.701030927835053e-05,
      "loss": 6.5876,
      "step": 224
    },
    {
      "epoch": 2.3179487179487177,
      "grad_norm": 0.6086193323135376,
      "learning_rate": 7.680412371134021e-05,
      "loss": 6.5513,
      "step": 226
    },
    {
      "epoch": 2.3384615384615386,
      "grad_norm": 0.7945834994316101,
      "learning_rate": 7.65979381443299e-05,
      "loss": 6.4032,
      "step": 228
    },
    {
      "epoch": 2.358974358974359,
      "grad_norm": 0.7115459442138672,
      "learning_rate": 7.639175257731959e-05,
      "loss": 6.7107,
      "step": 230
    },
    {
      "epoch": 2.3794871794871795,
      "grad_norm": 0.6793861985206604,
      "learning_rate": 7.618556701030927e-05,
      "loss": 6.6648,
      "step": 232
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.7627358436584473,
      "learning_rate": 7.597938144329897e-05,
      "loss": 6.5648,
      "step": 234
    },
    {
      "epoch": 2.4205128205128204,
      "grad_norm": 0.6426063179969788,
      "learning_rate": 7.577319587628867e-05,
      "loss": 6.2766,
      "step": 236
    },
    {
      "epoch": 2.4410256410256412,
      "grad_norm": 0.6580905914306641,
      "learning_rate": 7.556701030927835e-05,
      "loss": 6.6025,
      "step": 238
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 0.7617653608322144,
      "learning_rate": 7.536082474226805e-05,
      "loss": 6.5868,
      "step": 240
    },
    {
      "epoch": 2.482051282051282,
      "grad_norm": 0.5646825432777405,
      "learning_rate": 7.515463917525773e-05,
      "loss": 6.6345,
      "step": 242
    },
    {
      "epoch": 2.5025641025641026,
      "grad_norm": 0.637170135974884,
      "learning_rate": 7.494845360824743e-05,
      "loss": 6.4776,
      "step": 244
    },
    {
      "epoch": 2.523076923076923,
      "grad_norm": 0.6873669028282166,
      "learning_rate": 7.474226804123711e-05,
      "loss": 6.5078,
      "step": 246
    },
    {
      "epoch": 2.5435897435897434,
      "grad_norm": 0.5859996676445007,
      "learning_rate": 7.453608247422681e-05,
      "loss": 6.3846,
      "step": 248
    },
    {
      "epoch": 2.564102564102564,
      "grad_norm": 0.760138750076294,
      "learning_rate": 7.43298969072165e-05,
      "loss": 6.4852,
      "step": 250
    },
    {
      "epoch": 2.5846153846153848,
      "grad_norm": 0.650406002998352,
      "learning_rate": 7.412371134020619e-05,
      "loss": 6.57,
      "step": 252
    },
    {
      "epoch": 2.605128205128205,
      "grad_norm": 0.591342031955719,
      "learning_rate": 7.391752577319589e-05,
      "loss": 6.5708,
      "step": 254
    },
    {
      "epoch": 2.6256410256410256,
      "grad_norm": 0.5927435755729675,
      "learning_rate": 7.371134020618557e-05,
      "loss": 6.6123,
      "step": 256
    },
    {
      "epoch": 2.646153846153846,
      "grad_norm": 0.7065881490707397,
      "learning_rate": 7.350515463917527e-05,
      "loss": 6.5514,
      "step": 258
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.4774336516857147,
      "learning_rate": 7.329896907216495e-05,
      "loss": 6.4391,
      "step": 260
    },
    {
      "epoch": 2.6871794871794874,
      "grad_norm": 0.6040366888046265,
      "learning_rate": 7.309278350515463e-05,
      "loss": 6.5542,
      "step": 262
    },
    {
      "epoch": 2.707692307692308,
      "grad_norm": 0.6922281980514526,
      "learning_rate": 7.288659793814433e-05,
      "loss": 6.4446,
      "step": 264
    },
    {
      "epoch": 2.7282051282051283,
      "grad_norm": 0.8469168543815613,
      "learning_rate": 7.268041237113403e-05,
      "loss": 6.4419,
      "step": 266
    },
    {
      "epoch": 2.7487179487179487,
      "grad_norm": 0.6104958057403564,
      "learning_rate": 7.247422680412371e-05,
      "loss": 6.5526,
      "step": 268
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.6809393167495728,
      "learning_rate": 7.226804123711341e-05,
      "loss": 6.6546,
      "step": 270
    },
    {
      "epoch": 2.7897435897435896,
      "grad_norm": 0.7717563509941101,
      "learning_rate": 7.206185567010309e-05,
      "loss": 6.3678,
      "step": 272
    },
    {
      "epoch": 2.81025641025641,
      "grad_norm": 0.7897918820381165,
      "learning_rate": 7.185567010309279e-05,
      "loss": 6.2131,
      "step": 274
    },
    {
      "epoch": 2.830769230769231,
      "grad_norm": 0.6473929286003113,
      "learning_rate": 7.164948453608247e-05,
      "loss": 6.6045,
      "step": 276
    },
    {
      "epoch": 2.8512820512820514,
      "grad_norm": 0.7253041863441467,
      "learning_rate": 7.144329896907217e-05,
      "loss": 6.4355,
      "step": 278
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 0.6646418571472168,
      "learning_rate": 7.123711340206187e-05,
      "loss": 6.4636,
      "step": 280
    },
    {
      "epoch": 2.8923076923076922,
      "grad_norm": 0.6046567559242249,
      "learning_rate": 7.103092783505155e-05,
      "loss": 6.3511,
      "step": 282
    },
    {
      "epoch": 2.9128205128205127,
      "grad_norm": 0.6396191120147705,
      "learning_rate": 7.082474226804125e-05,
      "loss": 6.5544,
      "step": 284
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 0.7889509797096252,
      "learning_rate": 7.061855670103093e-05,
      "loss": 6.5913,
      "step": 286
    },
    {
      "epoch": 2.953846153846154,
      "grad_norm": 0.7650195956230164,
      "learning_rate": 7.041237113402063e-05,
      "loss": 6.4351,
      "step": 288
    },
    {
      "epoch": 2.9743589743589745,
      "grad_norm": 0.7612701654434204,
      "learning_rate": 7.020618556701031e-05,
      "loss": 6.3414,
      "step": 290
    },
    {
      "epoch": 2.994871794871795,
      "grad_norm": 0.7275530695915222,
      "learning_rate": 7e-05,
      "loss": 6.5395,
      "step": 292
    },
    {
      "epoch": 3.0153846153846153,
      "grad_norm": 0.6664620637893677,
      "learning_rate": 6.979381443298969e-05,
      "loss": 6.1818,
      "step": 294
    },
    {
      "epoch": 3.0358974358974358,
      "grad_norm": 0.6935043334960938,
      "learning_rate": 6.958762886597939e-05,
      "loss": 6.2724,
      "step": 296
    },
    {
      "epoch": 3.056410256410256,
      "grad_norm": 0.691132664680481,
      "learning_rate": 6.938144329896907e-05,
      "loss": 6.3618,
      "step": 298
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.6852958798408508,
      "learning_rate": 6.917525773195877e-05,
      "loss": 6.4248,
      "step": 300
    },
    {
      "epoch": 3.0974358974358975,
      "grad_norm": 0.7287474274635315,
      "learning_rate": 6.896907216494845e-05,
      "loss": 6.5134,
      "step": 302
    },
    {
      "epoch": 3.117948717948718,
      "grad_norm": 0.6913236379623413,
      "learning_rate": 6.876288659793815e-05,
      "loss": 6.5583,
      "step": 304
    },
    {
      "epoch": 3.1384615384615384,
      "grad_norm": 0.7557226419448853,
      "learning_rate": 6.855670103092783e-05,
      "loss": 6.3671,
      "step": 306
    },
    {
      "epoch": 3.158974358974359,
      "grad_norm": 0.673149049282074,
      "learning_rate": 6.835051546391753e-05,
      "loss": 6.463,
      "step": 308
    },
    {
      "epoch": 3.1794871794871793,
      "grad_norm": 0.7111195921897888,
      "learning_rate": 6.814432989690723e-05,
      "loss": 6.4296,
      "step": 310
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.7462407946586609,
      "learning_rate": 6.793814432989691e-05,
      "loss": 6.3978,
      "step": 312
    },
    {
      "epoch": 3.2205128205128206,
      "grad_norm": 0.633978009223938,
      "learning_rate": 6.773195876288661e-05,
      "loss": 6.5069,
      "step": 314
    },
    {
      "epoch": 3.241025641025641,
      "grad_norm": 0.5375587940216064,
      "learning_rate": 6.752577319587629e-05,
      "loss": 6.5878,
      "step": 316
    },
    {
      "epoch": 3.2615384615384615,
      "grad_norm": 0.6872283816337585,
      "learning_rate": 6.731958762886599e-05,
      "loss": 6.5541,
      "step": 318
    },
    {
      "epoch": 3.282051282051282,
      "grad_norm": 0.7773441076278687,
      "learning_rate": 6.711340206185567e-05,
      "loss": 6.3327,
      "step": 320
    },
    {
      "epoch": 3.3025641025641024,
      "grad_norm": 0.9842550754547119,
      "learning_rate": 6.690721649484535e-05,
      "loss": 6.4966,
      "step": 322
    },
    {
      "epoch": 3.3230769230769233,
      "grad_norm": 0.5430623292922974,
      "learning_rate": 6.670103092783505e-05,
      "loss": 6.5381,
      "step": 324
    },
    {
      "epoch": 3.3435897435897437,
      "grad_norm": 0.6485368609428406,
      "learning_rate": 6.649484536082475e-05,
      "loss": 6.4691,
      "step": 326
    },
    {
      "epoch": 3.364102564102564,
      "grad_norm": 0.8230199217796326,
      "learning_rate": 6.628865979381443e-05,
      "loss": 6.3189,
      "step": 328
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 0.6916872262954712,
      "learning_rate": 6.608247422680413e-05,
      "loss": 6.3726,
      "step": 330
    },
    {
      "epoch": 3.405128205128205,
      "grad_norm": 0.6563546657562256,
      "learning_rate": 6.587628865979381e-05,
      "loss": 6.532,
      "step": 332
    },
    {
      "epoch": 3.4256410256410255,
      "grad_norm": 0.8116714954376221,
      "learning_rate": 6.567010309278351e-05,
      "loss": 6.3005,
      "step": 334
    },
    {
      "epoch": 3.4461538461538463,
      "grad_norm": 0.8624938726425171,
      "learning_rate": 6.546391752577319e-05,
      "loss": 6.2658,
      "step": 336
    },
    {
      "epoch": 3.466666666666667,
      "grad_norm": 0.8297551274299622,
      "learning_rate": 6.525773195876289e-05,
      "loss": 6.3995,
      "step": 338
    },
    {
      "epoch": 3.4871794871794872,
      "grad_norm": 0.6855469942092896,
      "learning_rate": 6.505154639175259e-05,
      "loss": 6.3061,
      "step": 340
    },
    {
      "epoch": 3.5076923076923077,
      "grad_norm": 0.7813113331794739,
      "learning_rate": 6.484536082474227e-05,
      "loss": 6.3236,
      "step": 342
    },
    {
      "epoch": 3.528205128205128,
      "grad_norm": 0.6506015062332153,
      "learning_rate": 6.463917525773197e-05,
      "loss": 6.4452,
      "step": 344
    },
    {
      "epoch": 3.5487179487179485,
      "grad_norm": 0.7242425084114075,
      "learning_rate": 6.443298969072165e-05,
      "loss": 6.4236,
      "step": 346
    },
    {
      "epoch": 3.569230769230769,
      "grad_norm": 0.7753481864929199,
      "learning_rate": 6.422680412371135e-05,
      "loss": 6.503,
      "step": 348
    },
    {
      "epoch": 3.58974358974359,
      "grad_norm": 0.8971739411354065,
      "learning_rate": 6.402061855670103e-05,
      "loss": 6.2226,
      "step": 350
    },
    {
      "epoch": 3.6102564102564103,
      "grad_norm": 0.719221293926239,
      "learning_rate": 6.381443298969071e-05,
      "loss": 6.4305,
      "step": 352
    },
    {
      "epoch": 3.6307692307692307,
      "grad_norm": 0.7092213034629822,
      "learning_rate": 6.360824742268041e-05,
      "loss": 6.5107,
      "step": 354
    },
    {
      "epoch": 3.651282051282051,
      "grad_norm": 0.69648277759552,
      "learning_rate": 6.340206185567011e-05,
      "loss": 6.1619,
      "step": 356
    },
    {
      "epoch": 3.6717948717948716,
      "grad_norm": 0.7983478903770447,
      "learning_rate": 6.319587628865979e-05,
      "loss": 6.2945,
      "step": 358
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.6972605586051941,
      "learning_rate": 6.298969072164949e-05,
      "loss": 6.3397,
      "step": 360
    },
    {
      "epoch": 3.712820512820513,
      "grad_norm": 0.7316955924034119,
      "learning_rate": 6.278350515463917e-05,
      "loss": 6.5642,
      "step": 362
    },
    {
      "epoch": 3.7333333333333334,
      "grad_norm": 0.819416344165802,
      "learning_rate": 6.257731958762887e-05,
      "loss": 6.5361,
      "step": 364
    },
    {
      "epoch": 3.753846153846154,
      "grad_norm": 0.7884580492973328,
      "learning_rate": 6.237113402061855e-05,
      "loss": 6.1174,
      "step": 366
    },
    {
      "epoch": 3.7743589743589743,
      "grad_norm": 0.7763500809669495,
      "learning_rate": 6.216494845360825e-05,
      "loss": 6.2631,
      "step": 368
    },
    {
      "epoch": 3.7948717948717947,
      "grad_norm": 0.8002706170082092,
      "learning_rate": 6.195876288659795e-05,
      "loss": 6.3999,
      "step": 370
    },
    {
      "epoch": 3.815384615384615,
      "grad_norm": 0.6830271482467651,
      "learning_rate": 6.175257731958763e-05,
      "loss": 6.4109,
      "step": 372
    },
    {
      "epoch": 3.835897435897436,
      "grad_norm": 0.8474668860435486,
      "learning_rate": 6.154639175257733e-05,
      "loss": 6.2914,
      "step": 374
    },
    {
      "epoch": 3.8564102564102565,
      "grad_norm": 0.7036705017089844,
      "learning_rate": 6.134020618556701e-05,
      "loss": 6.5863,
      "step": 376
    },
    {
      "epoch": 3.876923076923077,
      "grad_norm": 0.6333104968070984,
      "learning_rate": 6.113402061855671e-05,
      "loss": 6.5162,
      "step": 378
    },
    {
      "epoch": 3.8974358974358974,
      "grad_norm": 0.5705188512802124,
      "learning_rate": 6.092783505154639e-05,
      "loss": 6.3006,
      "step": 380
    },
    {
      "epoch": 3.917948717948718,
      "grad_norm": 0.634800136089325,
      "learning_rate": 6.072164948453608e-05,
      "loss": 6.4401,
      "step": 382
    },
    {
      "epoch": 3.9384615384615387,
      "grad_norm": 0.6740937232971191,
      "learning_rate": 6.051546391752577e-05,
      "loss": 6.698,
      "step": 384
    },
    {
      "epoch": 3.958974358974359,
      "grad_norm": 0.9393010139465332,
      "learning_rate": 6.030927835051546e-05,
      "loss": 6.1999,
      "step": 386
    },
    {
      "epoch": 3.9794871794871796,
      "grad_norm": 0.6255496740341187,
      "learning_rate": 6.010309278350515e-05,
      "loss": 6.3655,
      "step": 388
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6616120934486389,
      "learning_rate": 5.989690721649484e-05,
      "loss": 6.4234,
      "step": 390
    },
    {
      "epoch": 4.02051282051282,
      "grad_norm": 0.6440649032592773,
      "learning_rate": 5.969072164948454e-05,
      "loss": 6.3687,
      "step": 392
    },
    {
      "epoch": 4.041025641025641,
      "grad_norm": 0.6960980296134949,
      "learning_rate": 5.948453608247423e-05,
      "loss": 6.453,
      "step": 394
    },
    {
      "epoch": 4.061538461538461,
      "grad_norm": 0.8931025266647339,
      "learning_rate": 5.927835051546392e-05,
      "loss": 6.4502,
      "step": 396
    },
    {
      "epoch": 4.082051282051282,
      "grad_norm": 0.6508329510688782,
      "learning_rate": 5.907216494845361e-05,
      "loss": 6.4387,
      "step": 398
    },
    {
      "epoch": 4.102564102564102,
      "grad_norm": 0.7409791946411133,
      "learning_rate": 5.88659793814433e-05,
      "loss": 6.3901,
      "step": 400
    },
    {
      "epoch": 4.123076923076923,
      "grad_norm": 0.6797041893005371,
      "learning_rate": 5.8659793814433e-05,
      "loss": 6.3299,
      "step": 402
    },
    {
      "epoch": 4.143589743589744,
      "grad_norm": 0.7487124800682068,
      "learning_rate": 5.845360824742269e-05,
      "loss": 6.3846,
      "step": 404
    },
    {
      "epoch": 4.164102564102564,
      "grad_norm": 0.6995207071304321,
      "learning_rate": 5.824742268041238e-05,
      "loss": 6.4029,
      "step": 406
    },
    {
      "epoch": 4.184615384615385,
      "grad_norm": 0.920820951461792,
      "learning_rate": 5.804123711340207e-05,
      "loss": 6.398,
      "step": 408
    },
    {
      "epoch": 4.205128205128205,
      "grad_norm": 0.7857652306556702,
      "learning_rate": 5.783505154639176e-05,
      "loss": 6.2116,
      "step": 410
    },
    {
      "epoch": 4.225641025641026,
      "grad_norm": 0.655503511428833,
      "learning_rate": 5.762886597938144e-05,
      "loss": 6.4523,
      "step": 412
    },
    {
      "epoch": 4.246153846153846,
      "grad_norm": 0.7093043327331543,
      "learning_rate": 5.742268041237113e-05,
      "loss": 6.2784,
      "step": 414
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.6597480177879333,
      "learning_rate": 5.721649484536082e-05,
      "loss": 6.4637,
      "step": 416
    },
    {
      "epoch": 4.287179487179487,
      "grad_norm": 0.8404126763343811,
      "learning_rate": 5.7010309278350514e-05,
      "loss": 6.226,
      "step": 418
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 0.7892007827758789,
      "learning_rate": 5.6804123711340204e-05,
      "loss": 6.3117,
      "step": 420
    },
    {
      "epoch": 4.328205128205128,
      "grad_norm": 0.7826732397079468,
      "learning_rate": 5.65979381443299e-05,
      "loss": 6.2826,
      "step": 422
    },
    {
      "epoch": 4.348717948717948,
      "grad_norm": 0.7233083248138428,
      "learning_rate": 5.639175257731959e-05,
      "loss": 6.3743,
      "step": 424
    },
    {
      "epoch": 4.36923076923077,
      "grad_norm": 0.6873509287834167,
      "learning_rate": 5.618556701030928e-05,
      "loss": 6.3368,
      "step": 426
    },
    {
      "epoch": 4.38974358974359,
      "grad_norm": 0.8165833353996277,
      "learning_rate": 5.597938144329897e-05,
      "loss": 6.4183,
      "step": 428
    },
    {
      "epoch": 4.410256410256411,
      "grad_norm": 0.7139970064163208,
      "learning_rate": 5.577319587628866e-05,
      "loss": 6.3427,
      "step": 430
    },
    {
      "epoch": 4.430769230769231,
      "grad_norm": 0.61057448387146,
      "learning_rate": 5.556701030927836e-05,
      "loss": 6.3747,
      "step": 432
    },
    {
      "epoch": 4.4512820512820515,
      "grad_norm": 0.8155067563056946,
      "learning_rate": 5.536082474226805e-05,
      "loss": 6.1935,
      "step": 434
    },
    {
      "epoch": 4.471794871794872,
      "grad_norm": 0.7220977544784546,
      "learning_rate": 5.515463917525774e-05,
      "loss": 6.4346,
      "step": 436
    },
    {
      "epoch": 4.492307692307692,
      "grad_norm": 0.6980003714561462,
      "learning_rate": 5.494845360824743e-05,
      "loss": 6.3355,
      "step": 438
    },
    {
      "epoch": 4.512820512820513,
      "grad_norm": 0.7350332736968994,
      "learning_rate": 5.474226804123712e-05,
      "loss": 6.2946,
      "step": 440
    },
    {
      "epoch": 4.533333333333333,
      "grad_norm": 0.6564266085624695,
      "learning_rate": 5.45360824742268e-05,
      "loss": 6.3044,
      "step": 442
    },
    {
      "epoch": 4.553846153846154,
      "grad_norm": 0.8935516476631165,
      "learning_rate": 5.4329896907216494e-05,
      "loss": 6.0784,
      "step": 444
    },
    {
      "epoch": 4.574358974358974,
      "grad_norm": 0.6622797846794128,
      "learning_rate": 5.4123711340206184e-05,
      "loss": 6.3122,
      "step": 446
    },
    {
      "epoch": 4.5948717948717945,
      "grad_norm": 0.6584466695785522,
      "learning_rate": 5.3917525773195874e-05,
      "loss": 6.4586,
      "step": 448
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.7214877605438232,
      "learning_rate": 5.3711340206185564e-05,
      "loss": 6.4671,
      "step": 450
    },
    {
      "epoch": 4.635897435897435,
      "grad_norm": 0.6560644507408142,
      "learning_rate": 5.350515463917526e-05,
      "loss": 6.2472,
      "step": 452
    },
    {
      "epoch": 4.656410256410257,
      "grad_norm": 0.7048325538635254,
      "learning_rate": 5.329896907216495e-05,
      "loss": 6.5175,
      "step": 454
    },
    {
      "epoch": 4.676923076923077,
      "grad_norm": 0.6315538883209229,
      "learning_rate": 5.309278350515464e-05,
      "loss": 6.0901,
      "step": 456
    },
    {
      "epoch": 4.697435897435898,
      "grad_norm": 0.6339179277420044,
      "learning_rate": 5.288659793814433e-05,
      "loss": 6.3935,
      "step": 458
    },
    {
      "epoch": 4.717948717948718,
      "grad_norm": 0.8235088586807251,
      "learning_rate": 5.268041237113402e-05,
      "loss": 5.9898,
      "step": 460
    },
    {
      "epoch": 4.7384615384615385,
      "grad_norm": 0.7074024081230164,
      "learning_rate": 5.247422680412372e-05,
      "loss": 6.1804,
      "step": 462
    },
    {
      "epoch": 4.758974358974359,
      "grad_norm": 0.8095348477363586,
      "learning_rate": 5.226804123711341e-05,
      "loss": 6.2209,
      "step": 464
    },
    {
      "epoch": 4.779487179487179,
      "grad_norm": 0.641472578048706,
      "learning_rate": 5.20618556701031e-05,
      "loss": 6.3102,
      "step": 466
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7063437700271606,
      "learning_rate": 5.185567010309279e-05,
      "loss": 6.2348,
      "step": 468
    },
    {
      "epoch": 4.82051282051282,
      "grad_norm": 0.7544019818305969,
      "learning_rate": 5.164948453608248e-05,
      "loss": 6.3763,
      "step": 470
    },
    {
      "epoch": 4.841025641025641,
      "grad_norm": 0.827674925327301,
      "learning_rate": 5.1443298969072164e-05,
      "loss": 6.0069,
      "step": 472
    },
    {
      "epoch": 4.861538461538462,
      "grad_norm": 0.8011050224304199,
      "learning_rate": 5.1237113402061854e-05,
      "loss": 6.257,
      "step": 474
    },
    {
      "epoch": 4.8820512820512825,
      "grad_norm": 0.6386663913726807,
      "learning_rate": 5.1030927835051544e-05,
      "loss": 6.3936,
      "step": 476
    },
    {
      "epoch": 4.902564102564103,
      "grad_norm": 0.7228955030441284,
      "learning_rate": 5.0824742268041235e-05,
      "loss": 6.4586,
      "step": 478
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 0.6252913475036621,
      "learning_rate": 5.0618556701030925e-05,
      "loss": 6.3536,
      "step": 480
    },
    {
      "epoch": 4.943589743589744,
      "grad_norm": 0.7947326898574829,
      "learning_rate": 5.041237113402062e-05,
      "loss": 6.3856,
      "step": 482
    },
    {
      "epoch": 4.964102564102564,
      "grad_norm": 0.6296289563179016,
      "learning_rate": 5.020618556701031e-05,
      "loss": 6.4445,
      "step": 484
    },
    {
      "epoch": 4.984615384615385,
      "grad_norm": 0.6455388069152832,
      "learning_rate": 5e-05,
      "loss": 6.4503,
      "step": 486
    },
    {
      "epoch": 5.005128205128205,
      "grad_norm": 0.7692280411720276,
      "learning_rate": 4.979381443298969e-05,
      "loss": 6.4011,
      "step": 488
    },
    {
      "epoch": 5.0256410256410255,
      "grad_norm": 0.6310716867446899,
      "learning_rate": 4.958762886597938e-05,
      "loss": 6.3749,
      "step": 490
    },
    {
      "epoch": 5.046153846153846,
      "grad_norm": 0.8823466897010803,
      "learning_rate": 4.938144329896908e-05,
      "loss": 6.2891,
      "step": 492
    },
    {
      "epoch": 5.066666666666666,
      "grad_norm": 0.7742350101470947,
      "learning_rate": 4.9175257731958763e-05,
      "loss": 6.1306,
      "step": 494
    },
    {
      "epoch": 5.087179487179487,
      "grad_norm": 0.6190592050552368,
      "learning_rate": 4.8969072164948454e-05,
      "loss": 6.3312,
      "step": 496
    },
    {
      "epoch": 5.107692307692307,
      "grad_norm": 0.7244901657104492,
      "learning_rate": 4.8762886597938144e-05,
      "loss": 6.3803,
      "step": 498
    },
    {
      "epoch": 5.128205128205128,
      "grad_norm": 0.7023468017578125,
      "learning_rate": 4.8556701030927834e-05,
      "loss": 6.2492,
      "step": 500
    },
    {
      "epoch": 5.148717948717949,
      "grad_norm": 0.7719900608062744,
      "learning_rate": 4.835051546391753e-05,
      "loss": 6.1033,
      "step": 502
    },
    {
      "epoch": 5.1692307692307695,
      "grad_norm": 0.8199257254600525,
      "learning_rate": 4.814432989690722e-05,
      "loss": 6.1981,
      "step": 504
    },
    {
      "epoch": 5.18974358974359,
      "grad_norm": 0.8238953351974487,
      "learning_rate": 4.793814432989691e-05,
      "loss": 6.2671,
      "step": 506
    },
    {
      "epoch": 5.21025641025641,
      "grad_norm": 0.5926157832145691,
      "learning_rate": 4.77319587628866e-05,
      "loss": 6.3837,
      "step": 508
    },
    {
      "epoch": 5.230769230769231,
      "grad_norm": 0.6979117393493652,
      "learning_rate": 4.7525773195876285e-05,
      "loss": 6.2836,
      "step": 510
    },
    {
      "epoch": 5.251282051282051,
      "grad_norm": 0.6661803126335144,
      "learning_rate": 4.731958762886598e-05,
      "loss": 6.1655,
      "step": 512
    },
    {
      "epoch": 5.271794871794872,
      "grad_norm": 0.6727478504180908,
      "learning_rate": 4.711340206185567e-05,
      "loss": 6.0956,
      "step": 514
    },
    {
      "epoch": 5.292307692307692,
      "grad_norm": 0.665337085723877,
      "learning_rate": 4.690721649484536e-05,
      "loss": 6.3724,
      "step": 516
    },
    {
      "epoch": 5.312820512820513,
      "grad_norm": 0.702816903591156,
      "learning_rate": 4.670103092783505e-05,
      "loss": 6.2241,
      "step": 518
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.6713479161262512,
      "learning_rate": 4.6494845360824743e-05,
      "loss": 6.3077,
      "step": 520
    },
    {
      "epoch": 5.3538461538461535,
      "grad_norm": 0.6276135444641113,
      "learning_rate": 4.628865979381444e-05,
      "loss": 6.4411,
      "step": 522
    },
    {
      "epoch": 5.374358974358975,
      "grad_norm": 0.7353730201721191,
      "learning_rate": 4.6082474226804124e-05,
      "loss": 6.1735,
      "step": 524
    },
    {
      "epoch": 5.394871794871795,
      "grad_norm": 0.7393136620521545,
      "learning_rate": 4.5876288659793814e-05,
      "loss": 6.413,
      "step": 526
    },
    {
      "epoch": 5.415384615384616,
      "grad_norm": 0.5835318565368652,
      "learning_rate": 4.5670103092783504e-05,
      "loss": 6.4997,
      "step": 528
    },
    {
      "epoch": 5.435897435897436,
      "grad_norm": 0.6543262600898743,
      "learning_rate": 4.5463917525773195e-05,
      "loss": 6.3292,
      "step": 530
    },
    {
      "epoch": 5.456410256410257,
      "grad_norm": 0.7140960693359375,
      "learning_rate": 4.525773195876289e-05,
      "loss": 6.3267,
      "step": 532
    },
    {
      "epoch": 5.476923076923077,
      "grad_norm": 0.6978656649589539,
      "learning_rate": 4.505154639175258e-05,
      "loss": 6.4343,
      "step": 534
    },
    {
      "epoch": 5.4974358974358974,
      "grad_norm": 0.7488942742347717,
      "learning_rate": 4.484536082474227e-05,
      "loss": 6.3557,
      "step": 536
    },
    {
      "epoch": 5.517948717948718,
      "grad_norm": 0.753161609172821,
      "learning_rate": 4.463917525773196e-05,
      "loss": 6.2488,
      "step": 538
    },
    {
      "epoch": 5.538461538461538,
      "grad_norm": 0.7196393609046936,
      "learning_rate": 4.4432989690721646e-05,
      "loss": 6.2132,
      "step": 540
    },
    {
      "epoch": 5.558974358974359,
      "grad_norm": 0.681937038898468,
      "learning_rate": 4.422680412371134e-05,
      "loss": 6.1831,
      "step": 542
    },
    {
      "epoch": 5.579487179487179,
      "grad_norm": 0.6542847752571106,
      "learning_rate": 4.402061855670103e-05,
      "loss": 6.3156,
      "step": 544
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.7364455461502075,
      "learning_rate": 4.3814432989690723e-05,
      "loss": 6.3007,
      "step": 546
    },
    {
      "epoch": 5.62051282051282,
      "grad_norm": 0.7742012739181519,
      "learning_rate": 4.3608247422680414e-05,
      "loss": 5.9234,
      "step": 548
    },
    {
      "epoch": 5.641025641025641,
      "grad_norm": 0.7302541136741638,
      "learning_rate": 4.3402061855670104e-05,
      "loss": 6.3115,
      "step": 550
    },
    {
      "epoch": 5.661538461538462,
      "grad_norm": 0.7138387560844421,
      "learning_rate": 4.31958762886598e-05,
      "loss": 6.1719,
      "step": 552
    },
    {
      "epoch": 5.682051282051282,
      "grad_norm": 0.6778889298439026,
      "learning_rate": 4.298969072164949e-05,
      "loss": 6.1886,
      "step": 554
    },
    {
      "epoch": 5.702564102564103,
      "grad_norm": 0.7845616936683655,
      "learning_rate": 4.2783505154639175e-05,
      "loss": 6.1925,
      "step": 556
    },
    {
      "epoch": 5.723076923076923,
      "grad_norm": 0.780630886554718,
      "learning_rate": 4.2577319587628865e-05,
      "loss": 6.3325,
      "step": 558
    },
    {
      "epoch": 5.743589743589744,
      "grad_norm": 0.630280613899231,
      "learning_rate": 4.2371134020618555e-05,
      "loss": 6.3334,
      "step": 560
    },
    {
      "epoch": 5.764102564102564,
      "grad_norm": 0.7802500128746033,
      "learning_rate": 4.216494845360825e-05,
      "loss": 6.2872,
      "step": 562
    },
    {
      "epoch": 5.7846153846153845,
      "grad_norm": 0.6819648146629333,
      "learning_rate": 4.195876288659794e-05,
      "loss": 6.3133,
      "step": 564
    },
    {
      "epoch": 5.805128205128205,
      "grad_norm": 0.7848822474479675,
      "learning_rate": 4.175257731958763e-05,
      "loss": 6.4798,
      "step": 566
    },
    {
      "epoch": 5.825641025641025,
      "grad_norm": 0.7532023787498474,
      "learning_rate": 4.154639175257732e-05,
      "loss": 6.1145,
      "step": 568
    },
    {
      "epoch": 5.846153846153846,
      "grad_norm": 0.679144561290741,
      "learning_rate": 4.1340206185567006e-05,
      "loss": 6.2509,
      "step": 570
    },
    {
      "epoch": 5.866666666666667,
      "grad_norm": 0.7792911529541016,
      "learning_rate": 4.1134020618556704e-05,
      "loss": 6.1544,
      "step": 572
    },
    {
      "epoch": 5.887179487179488,
      "grad_norm": 0.7862560153007507,
      "learning_rate": 4.0927835051546394e-05,
      "loss": 6.2661,
      "step": 574
    },
    {
      "epoch": 5.907692307692308,
      "grad_norm": 0.6902645826339722,
      "learning_rate": 4.0721649484536084e-05,
      "loss": 6.3423,
      "step": 576
    },
    {
      "epoch": 5.9282051282051285,
      "grad_norm": 0.6430610418319702,
      "learning_rate": 4.0515463917525774e-05,
      "loss": 6.3058,
      "step": 578
    },
    {
      "epoch": 5.948717948717949,
      "grad_norm": 0.7760281562805176,
      "learning_rate": 4.0309278350515465e-05,
      "loss": 6.3007,
      "step": 580
    },
    {
      "epoch": 5.969230769230769,
      "grad_norm": 0.7323822975158691,
      "learning_rate": 4.010309278350516e-05,
      "loss": 6.3434,
      "step": 582
    },
    {
      "epoch": 5.98974358974359,
      "grad_norm": 0.7060137391090393,
      "learning_rate": 3.989690721649485e-05,
      "loss": 6.2571,
      "step": 584
    },
    {
      "epoch": 6.01025641025641,
      "grad_norm": 0.7796257734298706,
      "learning_rate": 3.9690721649484535e-05,
      "loss": 6.1228,
      "step": 586
    },
    {
      "epoch": 6.030769230769231,
      "grad_norm": 0.8794940710067749,
      "learning_rate": 3.9484536082474226e-05,
      "loss": 6.3449,
      "step": 588
    },
    {
      "epoch": 6.051282051282051,
      "grad_norm": 0.6452521681785583,
      "learning_rate": 3.9278350515463916e-05,
      "loss": 6.0527,
      "step": 590
    },
    {
      "epoch": 6.0717948717948715,
      "grad_norm": 0.7071980834007263,
      "learning_rate": 3.907216494845361e-05,
      "loss": 6.1828,
      "step": 592
    },
    {
      "epoch": 6.092307692307692,
      "grad_norm": 0.6614767909049988,
      "learning_rate": 3.88659793814433e-05,
      "loss": 6.2702,
      "step": 594
    },
    {
      "epoch": 6.112820512820512,
      "grad_norm": 0.6533995866775513,
      "learning_rate": 3.865979381443299e-05,
      "loss": 6.2421,
      "step": 596
    },
    {
      "epoch": 6.133333333333334,
      "grad_norm": 0.7907997965812683,
      "learning_rate": 3.8453608247422684e-05,
      "loss": 6.4282,
      "step": 598
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 0.709160327911377,
      "learning_rate": 3.824742268041237e-05,
      "loss": 6.0814,
      "step": 600
    },
    {
      "epoch": 6.174358974358975,
      "grad_norm": 0.8180028796195984,
      "learning_rate": 3.8041237113402064e-05,
      "loss": 6.3014,
      "step": 602
    },
    {
      "epoch": 6.194871794871795,
      "grad_norm": 0.6765635013580322,
      "learning_rate": 3.7835051546391754e-05,
      "loss": 6.3844,
      "step": 604
    },
    {
      "epoch": 6.2153846153846155,
      "grad_norm": 0.6009554862976074,
      "learning_rate": 3.7628865979381445e-05,
      "loss": 6.1951,
      "step": 606
    },
    {
      "epoch": 6.235897435897436,
      "grad_norm": 0.708351731300354,
      "learning_rate": 3.7422680412371135e-05,
      "loss": 6.4139,
      "step": 608
    },
    {
      "epoch": 6.256410256410256,
      "grad_norm": 0.7396306395530701,
      "learning_rate": 3.7216494845360825e-05,
      "loss": 6.194,
      "step": 610
    },
    {
      "epoch": 6.276923076923077,
      "grad_norm": 0.772692859172821,
      "learning_rate": 3.701030927835052e-05,
      "loss": 6.2968,
      "step": 612
    },
    {
      "epoch": 6.297435897435897,
      "grad_norm": 0.8002333641052246,
      "learning_rate": 3.680412371134021e-05,
      "loss": 6.2677,
      "step": 614
    },
    {
      "epoch": 6.317948717948718,
      "grad_norm": 1.078007698059082,
      "learning_rate": 3.6597938144329896e-05,
      "loss": 6.2674,
      "step": 616
    },
    {
      "epoch": 6.338461538461538,
      "grad_norm": 0.6645541787147522,
      "learning_rate": 3.6391752577319586e-05,
      "loss": 6.1999,
      "step": 618
    },
    {
      "epoch": 6.358974358974359,
      "grad_norm": 0.6708665490150452,
      "learning_rate": 3.6185567010309276e-05,
      "loss": 6.4461,
      "step": 620
    },
    {
      "epoch": 6.37948717948718,
      "grad_norm": 0.6594524383544922,
      "learning_rate": 3.597938144329897e-05,
      "loss": 6.2954,
      "step": 622
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.9436811804771423,
      "learning_rate": 3.5773195876288664e-05,
      "loss": 6.1473,
      "step": 624
    },
    {
      "epoch": 6.420512820512821,
      "grad_norm": 0.7767890691757202,
      "learning_rate": 3.5567010309278354e-05,
      "loss": 6.2695,
      "step": 626
    },
    {
      "epoch": 6.441025641025641,
      "grad_norm": 0.8223767876625061,
      "learning_rate": 3.5360824742268044e-05,
      "loss": 6.2998,
      "step": 628
    },
    {
      "epoch": 6.461538461538462,
      "grad_norm": 0.753822922706604,
      "learning_rate": 3.515463917525773e-05,
      "loss": 6.1723,
      "step": 630
    },
    {
      "epoch": 6.482051282051282,
      "grad_norm": 0.7371943593025208,
      "learning_rate": 3.4948453608247425e-05,
      "loss": 6.2228,
      "step": 632
    },
    {
      "epoch": 6.5025641025641026,
      "grad_norm": 0.6721214056015015,
      "learning_rate": 3.4742268041237115e-05,
      "loss": 6.06,
      "step": 634
    },
    {
      "epoch": 6.523076923076923,
      "grad_norm": 0.7266871333122253,
      "learning_rate": 3.4536082474226805e-05,
      "loss": 6.3087,
      "step": 636
    },
    {
      "epoch": 6.543589743589743,
      "grad_norm": 0.7427794933319092,
      "learning_rate": 3.4329896907216495e-05,
      "loss": 6.2757,
      "step": 638
    },
    {
      "epoch": 6.564102564102564,
      "grad_norm": 0.7372191548347473,
      "learning_rate": 3.4123711340206186e-05,
      "loss": 6.1809,
      "step": 640
    },
    {
      "epoch": 6.584615384615384,
      "grad_norm": 0.773668646812439,
      "learning_rate": 3.391752577319588e-05,
      "loss": 6.0375,
      "step": 642
    },
    {
      "epoch": 6.605128205128205,
      "grad_norm": 0.8196036219596863,
      "learning_rate": 3.371134020618557e-05,
      "loss": 6.3719,
      "step": 644
    },
    {
      "epoch": 6.625641025641025,
      "grad_norm": 0.6300514936447144,
      "learning_rate": 3.3505154639175256e-05,
      "loss": 6.338,
      "step": 646
    },
    {
      "epoch": 6.6461538461538465,
      "grad_norm": 0.5757874846458435,
      "learning_rate": 3.3298969072164947e-05,
      "loss": 6.4443,
      "step": 648
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.8658593893051147,
      "learning_rate": 3.309278350515464e-05,
      "loss": 6.1216,
      "step": 650
    },
    {
      "epoch": 6.687179487179487,
      "grad_norm": 0.7729750275611877,
      "learning_rate": 3.2886597938144334e-05,
      "loss": 5.9493,
      "step": 652
    },
    {
      "epoch": 6.707692307692308,
      "grad_norm": 0.7950991988182068,
      "learning_rate": 3.2680412371134024e-05,
      "loss": 6.1614,
      "step": 654
    },
    {
      "epoch": 6.728205128205128,
      "grad_norm": 0.7309786081314087,
      "learning_rate": 3.2474226804123714e-05,
      "loss": 6.2484,
      "step": 656
    },
    {
      "epoch": 6.748717948717949,
      "grad_norm": 0.7773774266242981,
      "learning_rate": 3.2268041237113405e-05,
      "loss": 6.1015,
      "step": 658
    },
    {
      "epoch": 6.769230769230769,
      "grad_norm": 0.6493405103683472,
      "learning_rate": 3.2061855670103095e-05,
      "loss": 6.3609,
      "step": 660
    },
    {
      "epoch": 6.78974358974359,
      "grad_norm": 0.7441650032997131,
      "learning_rate": 3.1855670103092785e-05,
      "loss": 6.2702,
      "step": 662
    },
    {
      "epoch": 6.81025641025641,
      "grad_norm": 0.6217365264892578,
      "learning_rate": 3.1649484536082475e-05,
      "loss": 6.2799,
      "step": 664
    },
    {
      "epoch": 6.8307692307692305,
      "grad_norm": 0.84025639295578,
      "learning_rate": 3.1443298969072166e-05,
      "loss": 6.132,
      "step": 666
    },
    {
      "epoch": 6.851282051282051,
      "grad_norm": 0.7524210214614868,
      "learning_rate": 3.1237113402061856e-05,
      "loss": 6.1726,
      "step": 668
    },
    {
      "epoch": 6.871794871794872,
      "grad_norm": 0.6827425956726074,
      "learning_rate": 3.1030927835051546e-05,
      "loss": 6.2313,
      "step": 670
    },
    {
      "epoch": 6.892307692307693,
      "grad_norm": 0.6644515991210938,
      "learning_rate": 3.082474226804124e-05,
      "loss": 6.2285,
      "step": 672
    },
    {
      "epoch": 6.912820512820513,
      "grad_norm": 0.7661225199699402,
      "learning_rate": 3.0618556701030933e-05,
      "loss": 6.3277,
      "step": 674
    },
    {
      "epoch": 6.933333333333334,
      "grad_norm": 0.645756185054779,
      "learning_rate": 3.0412371134020617e-05,
      "loss": 6.2675,
      "step": 676
    },
    {
      "epoch": 6.953846153846154,
      "grad_norm": 0.8538247346878052,
      "learning_rate": 3.020618556701031e-05,
      "loss": 6.0726,
      "step": 678
    },
    {
      "epoch": 6.9743589743589745,
      "grad_norm": 0.6849417686462402,
      "learning_rate": 3e-05,
      "loss": 6.2466,
      "step": 680
    },
    {
      "epoch": 6.994871794871795,
      "grad_norm": 0.7335018515586853,
      "learning_rate": 2.979381443298969e-05,
      "loss": 6.1795,
      "step": 682
    },
    {
      "epoch": 7.015384615384615,
      "grad_norm": 0.7281456589698792,
      "learning_rate": 2.9587628865979385e-05,
      "loss": 6.2103,
      "step": 684
    },
    {
      "epoch": 7.035897435897436,
      "grad_norm": 0.6370302438735962,
      "learning_rate": 2.9381443298969075e-05,
      "loss": 6.2624,
      "step": 686
    },
    {
      "epoch": 7.056410256410256,
      "grad_norm": 0.7385585904121399,
      "learning_rate": 2.9175257731958765e-05,
      "loss": 6.1058,
      "step": 688
    },
    {
      "epoch": 7.076923076923077,
      "grad_norm": 0.7029441595077515,
      "learning_rate": 2.896907216494846e-05,
      "loss": 5.897,
      "step": 690
    },
    {
      "epoch": 7.097435897435897,
      "grad_norm": 0.7929611802101135,
      "learning_rate": 2.8762886597938142e-05,
      "loss": 6.0785,
      "step": 692
    },
    {
      "epoch": 7.1179487179487175,
      "grad_norm": 0.7209746837615967,
      "learning_rate": 2.8556701030927836e-05,
      "loss": 6.3427,
      "step": 694
    },
    {
      "epoch": 7.138461538461539,
      "grad_norm": 0.7581877112388611,
      "learning_rate": 2.8350515463917526e-05,
      "loss": 6.1258,
      "step": 696
    },
    {
      "epoch": 7.158974358974359,
      "grad_norm": 0.7337011694908142,
      "learning_rate": 2.8144329896907216e-05,
      "loss": 6.2402,
      "step": 698
    },
    {
      "epoch": 7.17948717948718,
      "grad_norm": 0.8132451772689819,
      "learning_rate": 2.793814432989691e-05,
      "loss": 6.2399,
      "step": 700
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.6459912061691284,
      "learning_rate": 2.77319587628866e-05,
      "loss": 6.3475,
      "step": 702
    },
    {
      "epoch": 7.220512820512821,
      "grad_norm": 0.7494829893112183,
      "learning_rate": 2.752577319587629e-05,
      "loss": 6.1483,
      "step": 704
    },
    {
      "epoch": 7.241025641025641,
      "grad_norm": 0.7804868221282959,
      "learning_rate": 2.7319587628865977e-05,
      "loss": 5.9649,
      "step": 706
    },
    {
      "epoch": 7.2615384615384615,
      "grad_norm": 0.6681525111198425,
      "learning_rate": 2.711340206185567e-05,
      "loss": 6.1267,
      "step": 708
    },
    {
      "epoch": 7.282051282051282,
      "grad_norm": 0.8127874732017517,
      "learning_rate": 2.690721649484536e-05,
      "loss": 6.1324,
      "step": 710
    },
    {
      "epoch": 7.302564102564102,
      "grad_norm": 0.690651535987854,
      "learning_rate": 2.670103092783505e-05,
      "loss": 6.3486,
      "step": 712
    },
    {
      "epoch": 7.323076923076923,
      "grad_norm": 0.7900473475456238,
      "learning_rate": 2.6494845360824745e-05,
      "loss": 6.2113,
      "step": 714
    },
    {
      "epoch": 7.343589743589743,
      "grad_norm": 0.9337382316589355,
      "learning_rate": 2.6288659793814435e-05,
      "loss": 6.2222,
      "step": 716
    },
    {
      "epoch": 7.364102564102564,
      "grad_norm": 0.6768183708190918,
      "learning_rate": 2.6082474226804126e-05,
      "loss": 6.0611,
      "step": 718
    },
    {
      "epoch": 7.384615384615385,
      "grad_norm": 0.8045716881752014,
      "learning_rate": 2.587628865979382e-05,
      "loss": 6.1869,
      "step": 720
    },
    {
      "epoch": 7.4051282051282055,
      "grad_norm": 0.7803860902786255,
      "learning_rate": 2.5670103092783503e-05,
      "loss": 6.2456,
      "step": 722
    },
    {
      "epoch": 7.425641025641026,
      "grad_norm": 0.6583536863327026,
      "learning_rate": 2.5463917525773196e-05,
      "loss": 6.3823,
      "step": 724
    },
    {
      "epoch": 7.446153846153846,
      "grad_norm": 0.6829738616943359,
      "learning_rate": 2.5257731958762887e-05,
      "loss": 5.9673,
      "step": 726
    },
    {
      "epoch": 7.466666666666667,
      "grad_norm": 0.6171258091926575,
      "learning_rate": 2.5051546391752577e-05,
      "loss": 6.4778,
      "step": 728
    },
    {
      "epoch": 7.487179487179487,
      "grad_norm": 0.9100000858306885,
      "learning_rate": 2.484536082474227e-05,
      "loss": 6.1923,
      "step": 730
    },
    {
      "epoch": 7.507692307692308,
      "grad_norm": 0.6728296279907227,
      "learning_rate": 2.463917525773196e-05,
      "loss": 6.1936,
      "step": 732
    },
    {
      "epoch": 7.528205128205128,
      "grad_norm": 0.6537327766418457,
      "learning_rate": 2.443298969072165e-05,
      "loss": 6.3321,
      "step": 734
    },
    {
      "epoch": 7.5487179487179485,
      "grad_norm": 0.7532163858413696,
      "learning_rate": 2.422680412371134e-05,
      "loss": 6.263,
      "step": 736
    },
    {
      "epoch": 7.569230769230769,
      "grad_norm": 0.8065058588981628,
      "learning_rate": 2.402061855670103e-05,
      "loss": 6.187,
      "step": 738
    },
    {
      "epoch": 7.589743589743589,
      "grad_norm": 0.6982044577598572,
      "learning_rate": 2.3814432989690722e-05,
      "loss": 6.2071,
      "step": 740
    },
    {
      "epoch": 7.61025641025641,
      "grad_norm": 0.7518334984779358,
      "learning_rate": 2.3608247422680412e-05,
      "loss": 6.1698,
      "step": 742
    },
    {
      "epoch": 7.63076923076923,
      "grad_norm": 0.6980510950088501,
      "learning_rate": 2.3402061855670106e-05,
      "loss": 6.1269,
      "step": 744
    },
    {
      "epoch": 7.651282051282052,
      "grad_norm": 0.7211715579032898,
      "learning_rate": 2.3195876288659796e-05,
      "loss": 6.2005,
      "step": 746
    },
    {
      "epoch": 7.671794871794872,
      "grad_norm": 0.7351675033569336,
      "learning_rate": 2.2989690721649483e-05,
      "loss": 6.1096,
      "step": 748
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 0.6142987608909607,
      "learning_rate": 2.2783505154639176e-05,
      "loss": 6.0628,
      "step": 750
    },
    {
      "epoch": 7.712820512820513,
      "grad_norm": 0.6700417995452881,
      "learning_rate": 2.2577319587628867e-05,
      "loss": 6.2528,
      "step": 752
    },
    {
      "epoch": 7.733333333333333,
      "grad_norm": 0.7297637462615967,
      "learning_rate": 2.237113402061856e-05,
      "loss": 6.165,
      "step": 754
    },
    {
      "epoch": 7.753846153846154,
      "grad_norm": 0.6762688755989075,
      "learning_rate": 2.2164948453608247e-05,
      "loss": 6.3651,
      "step": 756
    },
    {
      "epoch": 7.774358974358974,
      "grad_norm": 0.6471550464630127,
      "learning_rate": 2.1958762886597937e-05,
      "loss": 6.142,
      "step": 758
    },
    {
      "epoch": 7.794871794871795,
      "grad_norm": 0.6905655264854431,
      "learning_rate": 2.175257731958763e-05,
      "loss": 6.4471,
      "step": 760
    },
    {
      "epoch": 7.815384615384615,
      "grad_norm": 0.6673563122749329,
      "learning_rate": 2.154639175257732e-05,
      "loss": 6.0799,
      "step": 762
    },
    {
      "epoch": 7.835897435897436,
      "grad_norm": 0.7106631994247437,
      "learning_rate": 2.134020618556701e-05,
      "loss": 6.3385,
      "step": 764
    },
    {
      "epoch": 7.856410256410256,
      "grad_norm": 0.7296525239944458,
      "learning_rate": 2.1134020618556702e-05,
      "loss": 6.1772,
      "step": 766
    },
    {
      "epoch": 7.876923076923077,
      "grad_norm": 0.6923534274101257,
      "learning_rate": 2.0927835051546392e-05,
      "loss": 6.0845,
      "step": 768
    },
    {
      "epoch": 7.897435897435898,
      "grad_norm": 0.6479873061180115,
      "learning_rate": 2.0721649484536086e-05,
      "loss": 6.4237,
      "step": 770
    },
    {
      "epoch": 7.917948717948718,
      "grad_norm": 0.7043232321739197,
      "learning_rate": 2.0515463917525773e-05,
      "loss": 6.2389,
      "step": 772
    },
    {
      "epoch": 7.938461538461539,
      "grad_norm": 0.6955374479293823,
      "learning_rate": 2.0309278350515466e-05,
      "loss": 6.2594,
      "step": 774
    },
    {
      "epoch": 7.958974358974359,
      "grad_norm": 0.6248512864112854,
      "learning_rate": 2.0103092783505157e-05,
      "loss": 6.3424,
      "step": 776
    },
    {
      "epoch": 7.97948717948718,
      "grad_norm": 0.7696142196655273,
      "learning_rate": 1.9896907216494843e-05,
      "loss": 6.0404,
      "step": 778
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.7960737347602844,
      "learning_rate": 1.9690721649484537e-05,
      "loss": 6.2037,
      "step": 780
    },
    {
      "epoch": 8.02051282051282,
      "grad_norm": 0.764180600643158,
      "learning_rate": 1.9484536082474227e-05,
      "loss": 6.0193,
      "step": 782
    },
    {
      "epoch": 8.04102564102564,
      "grad_norm": 0.7745684385299683,
      "learning_rate": 1.927835051546392e-05,
      "loss": 6.1723,
      "step": 784
    },
    {
      "epoch": 8.061538461538461,
      "grad_norm": 0.7468089461326599,
      "learning_rate": 1.9072164948453608e-05,
      "loss": 5.989,
      "step": 786
    },
    {
      "epoch": 8.082051282051282,
      "grad_norm": 0.7076855301856995,
      "learning_rate": 1.8865979381443298e-05,
      "loss": 6.2975,
      "step": 788
    },
    {
      "epoch": 8.102564102564102,
      "grad_norm": 0.750068187713623,
      "learning_rate": 1.865979381443299e-05,
      "loss": 6.2057,
      "step": 790
    },
    {
      "epoch": 8.123076923076923,
      "grad_norm": 0.5782350301742554,
      "learning_rate": 1.8453608247422682e-05,
      "loss": 6.3482,
      "step": 792
    },
    {
      "epoch": 8.143589743589743,
      "grad_norm": 0.7477689981460571,
      "learning_rate": 1.8247422680412372e-05,
      "loss": 6.1132,
      "step": 794
    },
    {
      "epoch": 8.164102564102564,
      "grad_norm": 0.7583786845207214,
      "learning_rate": 1.8041237113402062e-05,
      "loss": 6.3927,
      "step": 796
    },
    {
      "epoch": 8.184615384615384,
      "grad_norm": 1.066887378692627,
      "learning_rate": 1.7835051546391753e-05,
      "loss": 6.3036,
      "step": 798
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 0.7979178428649902,
      "learning_rate": 1.7628865979381446e-05,
      "loss": 6.0357,
      "step": 800
    },
    {
      "epoch": 8.225641025641025,
      "grad_norm": 0.7925027012825012,
      "learning_rate": 1.7422680412371133e-05,
      "loss": 6.1571,
      "step": 802
    },
    {
      "epoch": 8.246153846153845,
      "grad_norm": 0.6301718950271606,
      "learning_rate": 1.7216494845360827e-05,
      "loss": 6.2834,
      "step": 804
    },
    {
      "epoch": 8.266666666666667,
      "grad_norm": 0.8579639196395874,
      "learning_rate": 1.7010309278350517e-05,
      "loss": 6.1507,
      "step": 806
    },
    {
      "epoch": 8.287179487179488,
      "grad_norm": 0.6312888264656067,
      "learning_rate": 1.6804123711340207e-05,
      "loss": 6.4582,
      "step": 808
    },
    {
      "epoch": 8.307692307692308,
      "grad_norm": 0.7432986497879028,
      "learning_rate": 1.6597938144329898e-05,
      "loss": 6.2031,
      "step": 810
    },
    {
      "epoch": 8.328205128205129,
      "grad_norm": 0.6533825993537903,
      "learning_rate": 1.6391752577319588e-05,
      "loss": 6.2383,
      "step": 812
    },
    {
      "epoch": 8.34871794871795,
      "grad_norm": 0.7833907604217529,
      "learning_rate": 1.618556701030928e-05,
      "loss": 6.2307,
      "step": 814
    },
    {
      "epoch": 8.36923076923077,
      "grad_norm": 0.7582297325134277,
      "learning_rate": 1.5979381443298968e-05,
      "loss": 6.2063,
      "step": 816
    },
    {
      "epoch": 8.38974358974359,
      "grad_norm": 0.7262023687362671,
      "learning_rate": 1.577319587628866e-05,
      "loss": 6.2125,
      "step": 818
    },
    {
      "epoch": 8.41025641025641,
      "grad_norm": 0.833846926689148,
      "learning_rate": 1.5567010309278352e-05,
      "loss": 5.9073,
      "step": 820
    },
    {
      "epoch": 8.430769230769231,
      "grad_norm": 0.6528037190437317,
      "learning_rate": 1.5360824742268042e-05,
      "loss": 6.1687,
      "step": 822
    },
    {
      "epoch": 8.451282051282051,
      "grad_norm": 0.715062141418457,
      "learning_rate": 1.5154639175257731e-05,
      "loss": 6.3858,
      "step": 824
    },
    {
      "epoch": 8.471794871794872,
      "grad_norm": 0.6965751647949219,
      "learning_rate": 1.4948453608247423e-05,
      "loss": 6.2247,
      "step": 826
    },
    {
      "epoch": 8.492307692307692,
      "grad_norm": 0.644185483455658,
      "learning_rate": 1.4742268041237115e-05,
      "loss": 6.4232,
      "step": 828
    },
    {
      "epoch": 8.512820512820513,
      "grad_norm": 0.8024539351463318,
      "learning_rate": 1.4536082474226805e-05,
      "loss": 6.2306,
      "step": 830
    },
    {
      "epoch": 8.533333333333333,
      "grad_norm": 0.6547759175300598,
      "learning_rate": 1.4329896907216495e-05,
      "loss": 6.2272,
      "step": 832
    },
    {
      "epoch": 8.553846153846154,
      "grad_norm": 0.8308131694793701,
      "learning_rate": 1.4123711340206186e-05,
      "loss": 6.2751,
      "step": 834
    },
    {
      "epoch": 8.574358974358974,
      "grad_norm": 0.6475526094436646,
      "learning_rate": 1.3917525773195878e-05,
      "loss": 6.3477,
      "step": 836
    },
    {
      "epoch": 8.594871794871795,
      "grad_norm": 0.7138957977294922,
      "learning_rate": 1.371134020618557e-05,
      "loss": 6.2169,
      "step": 838
    },
    {
      "epoch": 8.615384615384615,
      "grad_norm": 0.7940370440483093,
      "learning_rate": 1.3505154639175258e-05,
      "loss": 6.0711,
      "step": 840
    },
    {
      "epoch": 8.635897435897435,
      "grad_norm": 0.710660457611084,
      "learning_rate": 1.3298969072164948e-05,
      "loss": 6.1328,
      "step": 842
    },
    {
      "epoch": 8.656410256410256,
      "grad_norm": 0.6740992069244385,
      "learning_rate": 1.309278350515464e-05,
      "loss": 6.2399,
      "step": 844
    },
    {
      "epoch": 8.676923076923076,
      "grad_norm": 0.6954624652862549,
      "learning_rate": 1.2886597938144329e-05,
      "loss": 5.8876,
      "step": 846
    },
    {
      "epoch": 8.697435897435897,
      "grad_norm": 0.6282210350036621,
      "learning_rate": 1.268041237113402e-05,
      "loss": 6.2823,
      "step": 848
    },
    {
      "epoch": 8.717948717948717,
      "grad_norm": 0.6641991138458252,
      "learning_rate": 1.2474226804123713e-05,
      "loss": 6.0473,
      "step": 850
    },
    {
      "epoch": 8.73846153846154,
      "grad_norm": 0.9313386678695679,
      "learning_rate": 1.2268041237113401e-05,
      "loss": 5.9459,
      "step": 852
    },
    {
      "epoch": 8.75897435897436,
      "grad_norm": 0.6434590816497803,
      "learning_rate": 1.2061855670103093e-05,
      "loss": 6.2093,
      "step": 854
    },
    {
      "epoch": 8.77948717948718,
      "grad_norm": 0.7539594173431396,
      "learning_rate": 1.1855670103092783e-05,
      "loss": 6.1687,
      "step": 856
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.7946255803108215,
      "learning_rate": 1.1649484536082475e-05,
      "loss": 6.1376,
      "step": 858
    },
    {
      "epoch": 8.820512820512821,
      "grad_norm": 0.7640373706817627,
      "learning_rate": 1.1443298969072166e-05,
      "loss": 6.1806,
      "step": 860
    },
    {
      "epoch": 8.841025641025642,
      "grad_norm": 0.6656478643417358,
      "learning_rate": 1.1237113402061856e-05,
      "loss": 6.0972,
      "step": 862
    },
    {
      "epoch": 8.861538461538462,
      "grad_norm": 0.7396649718284607,
      "learning_rate": 1.1030927835051546e-05,
      "loss": 6.0908,
      "step": 864
    },
    {
      "epoch": 8.882051282051282,
      "grad_norm": 0.6619882583618164,
      "learning_rate": 1.0824742268041238e-05,
      "loss": 6.4136,
      "step": 866
    },
    {
      "epoch": 8.902564102564103,
      "grad_norm": 0.7462313175201416,
      "learning_rate": 1.0618556701030928e-05,
      "loss": 6.0462,
      "step": 868
    },
    {
      "epoch": 8.923076923076923,
      "grad_norm": 1.2745590209960938,
      "learning_rate": 1.041237113402062e-05,
      "loss": 6.0757,
      "step": 870
    },
    {
      "epoch": 8.943589743589744,
      "grad_norm": 0.7905880212783813,
      "learning_rate": 1.0206185567010309e-05,
      "loss": 5.9552,
      "step": 872
    },
    {
      "epoch": 8.964102564102564,
      "grad_norm": 0.7253865003585815,
      "learning_rate": 1e-05,
      "loss": 6.2264,
      "step": 874
    },
    {
      "epoch": 8.984615384615385,
      "grad_norm": 0.7785334587097168,
      "learning_rate": 9.793814432989691e-06,
      "loss": 6.195,
      "step": 876
    },
    {
      "epoch": 9.005128205128205,
      "grad_norm": 0.8180606961250305,
      "learning_rate": 9.587628865979383e-06,
      "loss": 5.991,
      "step": 878
    },
    {
      "epoch": 9.025641025641026,
      "grad_norm": 0.7430230379104614,
      "learning_rate": 9.381443298969073e-06,
      "loss": 5.9541,
      "step": 880
    },
    {
      "epoch": 9.046153846153846,
      "grad_norm": 0.6766589879989624,
      "learning_rate": 9.175257731958762e-06,
      "loss": 6.2283,
      "step": 882
    },
    {
      "epoch": 9.066666666666666,
      "grad_norm": 0.7492343187332153,
      "learning_rate": 8.969072164948454e-06,
      "loss": 6.1433,
      "step": 884
    },
    {
      "epoch": 9.087179487179487,
      "grad_norm": 0.787506103515625,
      "learning_rate": 8.762886597938144e-06,
      "loss": 5.7677,
      "step": 886
    },
    {
      "epoch": 9.107692307692307,
      "grad_norm": 0.6319190859794617,
      "learning_rate": 8.556701030927836e-06,
      "loss": 6.2569,
      "step": 888
    },
    {
      "epoch": 9.128205128205128,
      "grad_norm": 0.5653165578842163,
      "learning_rate": 8.350515463917526e-06,
      "loss": 6.3534,
      "step": 890
    },
    {
      "epoch": 9.148717948717948,
      "grad_norm": 1.010986566543579,
      "learning_rate": 8.144329896907216e-06,
      "loss": 6.2486,
      "step": 892
    },
    {
      "epoch": 9.169230769230769,
      "grad_norm": 0.7131024599075317,
      "learning_rate": 7.938144329896907e-06,
      "loss": 6.0887,
      "step": 894
    },
    {
      "epoch": 9.189743589743589,
      "grad_norm": 0.7308672666549683,
      "learning_rate": 7.731958762886599e-06,
      "loss": 6.1484,
      "step": 896
    },
    {
      "epoch": 9.21025641025641,
      "grad_norm": 0.7401643991470337,
      "learning_rate": 7.525773195876289e-06,
      "loss": 6.1046,
      "step": 898
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 0.7252262830734253,
      "learning_rate": 7.31958762886598e-06,
      "loss": 6.2832,
      "step": 900
    },
    {
      "epoch": 9.25128205128205,
      "grad_norm": 0.7304520606994629,
      "learning_rate": 7.11340206185567e-06,
      "loss": 6.2027,
      "step": 902
    },
    {
      "epoch": 9.271794871794873,
      "grad_norm": 0.6221430897712708,
      "learning_rate": 6.907216494845361e-06,
      "loss": 6.3802,
      "step": 904
    },
    {
      "epoch": 9.292307692307693,
      "grad_norm": 0.7544053792953491,
      "learning_rate": 6.701030927835052e-06,
      "loss": 6.1048,
      "step": 906
    },
    {
      "epoch": 9.312820512820513,
      "grad_norm": 0.6470032930374146,
      "learning_rate": 6.494845360824743e-06,
      "loss": 6.3017,
      "step": 908
    },
    {
      "epoch": 9.333333333333334,
      "grad_norm": 0.5889255404472351,
      "learning_rate": 6.288659793814433e-06,
      "loss": 6.2106,
      "step": 910
    },
    {
      "epoch": 9.353846153846154,
      "grad_norm": 0.6223869323730469,
      "learning_rate": 6.082474226804124e-06,
      "loss": 6.1163,
      "step": 912
    },
    {
      "epoch": 9.374358974358975,
      "grad_norm": 0.6783230304718018,
      "learning_rate": 5.876288659793814e-06,
      "loss": 6.1365,
      "step": 914
    },
    {
      "epoch": 9.394871794871795,
      "grad_norm": 0.7757629752159119,
      "learning_rate": 5.670103092783505e-06,
      "loss": 6.0445,
      "step": 916
    },
    {
      "epoch": 9.415384615384616,
      "grad_norm": 0.7045403122901917,
      "learning_rate": 5.4639175257731965e-06,
      "loss": 6.3132,
      "step": 918
    },
    {
      "epoch": 9.435897435897436,
      "grad_norm": 0.7770035862922668,
      "learning_rate": 5.257731958762887e-06,
      "loss": 6.0804,
      "step": 920
    },
    {
      "epoch": 9.456410256410257,
      "grad_norm": 0.6999940276145935,
      "learning_rate": 5.051546391752578e-06,
      "loss": 6.0772,
      "step": 922
    },
    {
      "epoch": 9.476923076923077,
      "grad_norm": 0.7113653421401978,
      "learning_rate": 4.845360824742268e-06,
      "loss": 5.9905,
      "step": 924
    },
    {
      "epoch": 9.497435897435897,
      "grad_norm": 0.6142096519470215,
      "learning_rate": 4.639175257731959e-06,
      "loss": 6.2268,
      "step": 926
    },
    {
      "epoch": 9.517948717948718,
      "grad_norm": 0.6630533933639526,
      "learning_rate": 4.43298969072165e-06,
      "loss": 6.3161,
      "step": 928
    },
    {
      "epoch": 9.538461538461538,
      "grad_norm": 0.8330481052398682,
      "learning_rate": 4.2268041237113405e-06,
      "loss": 6.1675,
      "step": 930
    },
    {
      "epoch": 9.558974358974359,
      "grad_norm": 0.6822926998138428,
      "learning_rate": 4.020618556701032e-06,
      "loss": 6.106,
      "step": 932
    },
    {
      "epoch": 9.57948717948718,
      "grad_norm": 0.7913044095039368,
      "learning_rate": 3.8144329896907223e-06,
      "loss": 6.2236,
      "step": 934
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.8586515188217163,
      "learning_rate": 3.608247422680412e-06,
      "loss": 6.0912,
      "step": 936
    },
    {
      "epoch": 9.62051282051282,
      "grad_norm": 0.6389515399932861,
      "learning_rate": 3.402061855670103e-06,
      "loss": 6.1972,
      "step": 938
    },
    {
      "epoch": 9.64102564102564,
      "grad_norm": 0.647665798664093,
      "learning_rate": 3.195876288659794e-06,
      "loss": 6.462,
      "step": 940
    },
    {
      "epoch": 9.661538461538461,
      "grad_norm": 0.7906177043914795,
      "learning_rate": 2.989690721649485e-06,
      "loss": 6.1017,
      "step": 942
    },
    {
      "epoch": 9.682051282051281,
      "grad_norm": 0.7301990985870361,
      "learning_rate": 2.7835051546391757e-06,
      "loss": 6.1658,
      "step": 944
    },
    {
      "epoch": 9.702564102564102,
      "grad_norm": 0.676216185092926,
      "learning_rate": 2.577319587628866e-06,
      "loss": 6.3423,
      "step": 946
    },
    {
      "epoch": 9.723076923076922,
      "grad_norm": 0.7950716614723206,
      "learning_rate": 2.3711340206185566e-06,
      "loss": 6.1037,
      "step": 948
    },
    {
      "epoch": 9.743589743589745,
      "grad_norm": 0.6446979641914368,
      "learning_rate": 2.1649484536082473e-06,
      "loss": 6.245,
      "step": 950
    },
    {
      "epoch": 9.764102564102565,
      "grad_norm": 0.7083817720413208,
      "learning_rate": 1.9587628865979384e-06,
      "loss": 6.1738,
      "step": 952
    },
    {
      "epoch": 9.784615384615385,
      "grad_norm": 0.611316442489624,
      "learning_rate": 1.752577319587629e-06,
      "loss": 6.3228,
      "step": 954
    },
    {
      "epoch": 9.805128205128206,
      "grad_norm": 0.7233178019523621,
      "learning_rate": 1.5463917525773197e-06,
      "loss": 5.7474,
      "step": 956
    },
    {
      "epoch": 9.825641025641026,
      "grad_norm": 0.6572598814964294,
      "learning_rate": 1.3402061855670102e-06,
      "loss": 6.245,
      "step": 958
    },
    {
      "epoch": 9.846153846153847,
      "grad_norm": 0.7650203108787537,
      "learning_rate": 1.134020618556701e-06,
      "loss": 6.2761,
      "step": 960
    },
    {
      "epoch": 9.866666666666667,
      "grad_norm": 0.780173122882843,
      "learning_rate": 9.278350515463918e-07,
      "loss": 6.2113,
      "step": 962
    },
    {
      "epoch": 9.887179487179488,
      "grad_norm": 0.7613905668258667,
      "learning_rate": 7.216494845360825e-07,
      "loss": 6.106,
      "step": 964
    },
    {
      "epoch": 9.907692307692308,
      "grad_norm": 0.8180409669876099,
      "learning_rate": 5.154639175257732e-07,
      "loss": 6.1713,
      "step": 966
    },
    {
      "epoch": 9.928205128205128,
      "grad_norm": 0.6750867366790771,
      "learning_rate": 3.0927835051546394e-07,
      "loss": 6.1835,
      "step": 968
    },
    {
      "epoch": 9.948717948717949,
      "grad_norm": 0.7007511854171753,
      "learning_rate": 1.0309278350515465e-07,
      "loss": 6.2497,
      "step": 970
    },
    {
      "epoch": 9.948717948717949,
      "step": 970,
      "total_flos": 1620553852147860.0,
      "train_loss": 6.414023149628,
      "train_runtime": 12031.1209,
      "train_samples_per_second": 1.296,
      "train_steps_per_second": 0.081
    }
  ],
  "logging_steps": 2,
  "max_steps": 970,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 1620553852147860.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}