{
  "best_global_step": 904,
  "best_metric": 0.10195600241422653,
  "best_model_checkpoint": "saves_stability/prefix-tuning/llama-3-8b-instruct/train_cb_1757340264/checkpoint-904",
  "epoch": 20.0,
  "eval_steps": 113,
  "global_step": 2260,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04424778761061947,
      "grad_norm": 216.9446258544922,
      "learning_rate": 8.849557522123894e-07,
      "loss": 9.803,
      "num_input_tokens_seen": 1600,
      "step": 5
    },
    {
      "epoch": 0.08849557522123894,
      "grad_norm": 173.4165496826172,
      "learning_rate": 1.991150442477876e-06,
      "loss": 8.9927,
      "num_input_tokens_seen": 3104,
      "step": 10
    },
    {
      "epoch": 0.13274336283185842,
      "grad_norm": 125.74637603759766,
      "learning_rate": 3.097345132743363e-06,
      "loss": 8.6905,
      "num_input_tokens_seen": 4416,
      "step": 15
    },
    {
      "epoch": 0.17699115044247787,
      "grad_norm": 99.64391326904297,
      "learning_rate": 4.2035398230088504e-06,
      "loss": 7.4647,
      "num_input_tokens_seen": 5856,
      "step": 20
    },
    {
      "epoch": 0.22123893805309736,
      "grad_norm": 109.86019134521484,
      "learning_rate": 5.3097345132743365e-06,
      "loss": 5.4612,
      "num_input_tokens_seen": 7328,
      "step": 25
    },
    {
      "epoch": 0.26548672566371684,
      "grad_norm": 72.39360809326172,
      "learning_rate": 6.415929203539823e-06,
      "loss": 4.304,
      "num_input_tokens_seen": 8864,
      "step": 30
    },
    {
      "epoch": 0.30973451327433627,
      "grad_norm": 56.51565170288086,
      "learning_rate": 7.52212389380531e-06,
      "loss": 2.0983,
      "num_input_tokens_seen": 10400,
      "step": 35
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 53.66331100463867,
      "learning_rate": 8.628318584070797e-06,
      "loss": 1.404,
      "num_input_tokens_seen": 11584,
      "step": 40
    },
    {
      "epoch": 0.39823008849557523,
      "grad_norm": 33.16804122924805,
      "learning_rate": 9.734513274336284e-06,
      "loss": 0.4689,
      "num_input_tokens_seen": 12816,
      "step": 45
    },
    {
      "epoch": 0.4424778761061947,
      "grad_norm": 97.21773529052734,
      "learning_rate": 1.0840707964601771e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 14464,
      "step": 50
    },
    {
      "epoch": 0.48672566371681414,
      "grad_norm": 25.265501022338867,
      "learning_rate": 1.1946902654867258e-05,
      "loss": 0.3246,
      "num_input_tokens_seen": 15648,
      "step": 55
    },
    {
      "epoch": 0.5309734513274337,
      "grad_norm": 12.91411018371582,
      "learning_rate": 1.3053097345132745e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 17008,
      "step": 60
    },
    {
      "epoch": 0.5752212389380531,
      "grad_norm": 15.694816589355469,
      "learning_rate": 1.415929203539823e-05,
      "loss": 0.7632,
      "num_input_tokens_seen": 18256,
      "step": 65
    },
    {
      "epoch": 0.6194690265486725,
      "grad_norm": 25.84539794921875,
      "learning_rate": 1.5265486725663717e-05,
      "loss": 0.3145,
      "num_input_tokens_seen": 19360,
      "step": 70
    },
    {
      "epoch": 0.6637168141592921,
      "grad_norm": 27.247764587402344,
      "learning_rate": 1.6371681415929206e-05,
      "loss": 0.19,
      "num_input_tokens_seen": 20784,
      "step": 75
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 21.10613441467285,
      "learning_rate": 1.747787610619469e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 21776,
      "step": 80
    },
    {
      "epoch": 0.7522123893805309,
      "grad_norm": 16.608076095581055,
      "learning_rate": 1.858407079646018e-05,
      "loss": 0.2901,
      "num_input_tokens_seen": 23344,
      "step": 85
    },
    {
      "epoch": 0.7964601769911505,
      "grad_norm": 16.492979049682617,
      "learning_rate": 1.9690265486725665e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 24416,
      "step": 90
    },
    {
      "epoch": 0.8407079646017699,
      "grad_norm": 14.89802074432373,
      "learning_rate": 2.079646017699115e-05,
      "loss": 0.6743,
      "num_input_tokens_seen": 25472,
      "step": 95
    },
    {
      "epoch": 0.8849557522123894,
      "grad_norm": 22.4132022857666,
      "learning_rate": 2.190265486725664e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 27120,
      "step": 100
    },
    {
      "epoch": 0.9292035398230089,
      "grad_norm": 53.07675552368164,
      "learning_rate": 2.3008849557522124e-05,
      "loss": 0.3154,
      "num_input_tokens_seen": 28464,
      "step": 105
    },
    {
      "epoch": 0.9734513274336283,
      "grad_norm": 31.527185440063477,
      "learning_rate": 2.411504424778761e-05,
      "loss": 0.4167,
      "num_input_tokens_seen": 29632,
      "step": 110
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.2356422245502472,
      "eval_runtime": 0.7139,
      "eval_samples_per_second": 35.018,
      "eval_steps_per_second": 18.209,
      "num_input_tokens_seen": 30240,
      "step": 113
    },
    {
      "epoch": 1.0176991150442478,
      "grad_norm": 8.958460807800293,
      "learning_rate": 2.5221238938053098e-05,
      "loss": 0.2027,
      "num_input_tokens_seen": 30672,
      "step": 115
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 13.822118759155273,
      "learning_rate": 2.6327433628318586e-05,
      "loss": 0.7973,
      "num_input_tokens_seen": 32304,
      "step": 120
    },
    {
      "epoch": 1.1061946902654867,
      "grad_norm": 7.096332550048828,
      "learning_rate": 2.743362831858407e-05,
      "loss": 0.1341,
      "num_input_tokens_seen": 33568,
      "step": 125
    },
    {
      "epoch": 1.1504424778761062,
      "grad_norm": 4.537890434265137,
      "learning_rate": 2.853982300884956e-05,
      "loss": 0.1614,
      "num_input_tokens_seen": 35280,
      "step": 130
    },
    {
      "epoch": 1.1946902654867257,
      "grad_norm": 15.707793235778809,
      "learning_rate": 2.964601769911505e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 36560,
      "step": 135
    },
    {
      "epoch": 1.238938053097345,
      "grad_norm": 27.812833786010742,
      "learning_rate": 3.075221238938053e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 38080,
      "step": 140
    },
    {
      "epoch": 1.2831858407079646,
      "grad_norm": 29.91482162475586,
      "learning_rate": 3.185840707964602e-05,
      "loss": 0.3061,
      "num_input_tokens_seen": 39360,
      "step": 145
    },
    {
      "epoch": 1.3274336283185841,
      "grad_norm": 54.62297439575195,
      "learning_rate": 3.296460176991151e-05,
      "loss": 0.3754,
      "num_input_tokens_seen": 40480,
      "step": 150
    },
    {
      "epoch": 1.3716814159292037,
      "grad_norm": 38.17515182495117,
      "learning_rate": 3.407079646017699e-05,
      "loss": 0.642,
      "num_input_tokens_seen": 41872,
      "step": 155
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 51.05940246582031,
      "learning_rate": 3.517699115044248e-05,
      "loss": 0.7989,
      "num_input_tokens_seen": 43184,
      "step": 160
    },
    {
      "epoch": 1.4601769911504425,
      "grad_norm": 16.191553115844727,
      "learning_rate": 3.628318584070797e-05,
      "loss": 0.5834,
      "num_input_tokens_seen": 44288,
      "step": 165
    },
    {
      "epoch": 1.504424778761062,
      "grad_norm": 8.015981674194336,
      "learning_rate": 3.7389380530973455e-05,
      "loss": 0.2561,
      "num_input_tokens_seen": 45584,
      "step": 170
    },
    {
      "epoch": 1.5486725663716814,
      "grad_norm": 41.44169235229492,
      "learning_rate": 3.849557522123894e-05,
      "loss": 0.8934,
      "num_input_tokens_seen": 46848,
      "step": 175
    },
    {
      "epoch": 1.592920353982301,
      "grad_norm": 4.978911399841309,
      "learning_rate": 3.9601769911504426e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 48384,
      "step": 180
    },
    {
      "epoch": 1.6371681415929205,
      "grad_norm": 3.805788278579712,
      "learning_rate": 4.0707964601769914e-05,
      "loss": 0.135,
      "num_input_tokens_seen": 50176,
      "step": 185
    },
    {
      "epoch": 1.6814159292035398,
      "grad_norm": 15.465529441833496,
      "learning_rate": 4.1814159292035396e-05,
      "loss": 0.2251,
      "num_input_tokens_seen": 51792,
      "step": 190
    },
    {
      "epoch": 1.7256637168141593,
      "grad_norm": 0.04187662526965141,
      "learning_rate": 4.2920353982300885e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 52768,
      "step": 195
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 79.34677124023438,
      "learning_rate": 4.4026548672566373e-05,
      "loss": 1.0914,
      "num_input_tokens_seen": 54256,
      "step": 200
    },
    {
      "epoch": 1.8141592920353982,
      "grad_norm": 17.46209716796875,
      "learning_rate": 4.5132743362831855e-05,
      "loss": 0.6367,
      "num_input_tokens_seen": 55888,
      "step": 205
    },
    {
      "epoch": 1.8584070796460177,
      "grad_norm": 86.74883270263672,
      "learning_rate": 4.6238938053097344e-05,
      "loss": 0.9824,
      "num_input_tokens_seen": 57200,
      "step": 210
    },
    {
      "epoch": 1.9026548672566372,
      "grad_norm": 1.0989741086959839,
      "learning_rate": 4.734513274336283e-05,
      "loss": 0.6968,
      "num_input_tokens_seen": 58528,
      "step": 215
    },
    {
      "epoch": 1.9469026548672566,
      "grad_norm": 34.84535217285156,
      "learning_rate": 4.845132743362832e-05,
      "loss": 0.3508,
      "num_input_tokens_seen": 59888,
      "step": 220
    },
    {
      "epoch": 1.991150442477876,
      "grad_norm": 20.072399139404297,
      "learning_rate": 4.955752212389381e-05,
      "loss": 0.6586,
      "num_input_tokens_seen": 61392,
      "step": 225
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.3422188460826874,
      "eval_runtime": 0.7198,
      "eval_samples_per_second": 34.732,
      "eval_steps_per_second": 18.061,
      "num_input_tokens_seen": 61600,
      "step": 226
    },
    {
      "epoch": 2.0353982300884956,
      "grad_norm": 0.32449668645858765,
      "learning_rate": 4.9999731620342936e-05,
      "loss": 0.2724,
      "num_input_tokens_seen": 62640,
      "step": 230
    },
    {
      "epoch": 2.079646017699115,
      "grad_norm": 14.44190788269043,
      "learning_rate": 4.9998091543305845e-05,
      "loss": 0.6147,
      "num_input_tokens_seen": 63984,
      "step": 235
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 49.907142639160156,
      "learning_rate": 4.999496058673635e-05,
      "loss": 0.6559,
      "num_input_tokens_seen": 65408,
      "step": 240
    },
    {
      "epoch": 2.168141592920354,
      "grad_norm": 20.732376098632812,
      "learning_rate": 4.999033893736386e-05,
      "loss": 1.0433,
      "num_input_tokens_seen": 66784,
      "step": 245
    },
    {
      "epoch": 2.2123893805309733,
      "grad_norm": 9.330907821655273,
      "learning_rate": 4.99842268708223e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 68016,
      "step": 250
    },
    {
      "epoch": 2.256637168141593,
      "grad_norm": 10.19569206237793,
      "learning_rate": 4.9976624751633725e-05,
      "loss": 0.2034,
      "num_input_tokens_seen": 69504,
      "step": 255
    },
    {
      "epoch": 2.3008849557522124,
      "grad_norm": 11.039266586303711,
      "learning_rate": 4.996753303318648e-05,
      "loss": 0.3076,
      "num_input_tokens_seen": 70688,
      "step": 260
    },
    {
      "epoch": 2.3451327433628317,
      "grad_norm": 21.324434280395508,
      "learning_rate": 4.995695225770825e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 71968,
      "step": 265
    },
    {
      "epoch": 2.3893805309734515,
      "grad_norm": 0.44300082325935364,
      "learning_rate": 4.994488305623365e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 73632,
      "step": 270
    },
    {
      "epoch": 2.433628318584071,
      "grad_norm": 5.458860874176025,
      "learning_rate": 4.993132614856666e-05,
      "loss": 0.5822,
      "num_input_tokens_seen": 75216,
      "step": 275
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 10.641469955444336,
      "learning_rate": 4.991628234323765e-05,
      "loss": 0.2626,
      "num_input_tokens_seen": 76496,
      "step": 280
    },
    {
      "epoch": 2.52212389380531,
      "grad_norm": 0.15004612505435944,
      "learning_rate": 4.9899752537455166e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 78288,
      "step": 285
    },
    {
      "epoch": 2.566371681415929,
      "grad_norm": 0.10617993026971817,
      "learning_rate": 4.9881737717052436e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 79504,
      "step": 290
    },
    {
      "epoch": 2.6106194690265485,
      "grad_norm": 12.52049732208252,
      "learning_rate": 4.9862238956428556e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 80944,
      "step": 295
    },
    {
      "epoch": 2.6548672566371683,
      "grad_norm": 11.246700286865234,
      "learning_rate": 4.984125741848441e-05,
      "loss": 0.3715,
      "num_input_tokens_seen": 82432,
      "step": 300
    },
    {
      "epoch": 2.6991150442477876,
      "grad_norm": 4.463377475738525,
      "learning_rate": 4.981879435455336e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 83840,
      "step": 305
    },
    {
      "epoch": 2.7433628318584073,
      "grad_norm": 15.015288352966309,
      "learning_rate": 4.9794851104326554e-05,
      "loss": 0.1499,
      "num_input_tokens_seen": 85056,
      "step": 310
    },
    {
      "epoch": 2.7876106194690267,
      "grad_norm": 11.237860679626465,
      "learning_rate": 4.976942909577307e-05,
      "loss": 0.934,
      "num_input_tokens_seen": 86304,
      "step": 315
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 20.591720581054688,
      "learning_rate": 4.974252984505475e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 87360,
      "step": 320
    },
    {
      "epoch": 2.8761061946902657,
      "grad_norm": 3.167823553085327,
      "learning_rate": 4.971415495643574e-05,
      "loss": 0.3231,
      "num_input_tokens_seen": 88560,
      "step": 325
    },
    {
      "epoch": 2.920353982300885,
      "grad_norm": 24.052827835083008,
      "learning_rate": 4.968430612218687e-05,
      "loss": 0.7549,
      "num_input_tokens_seen": 90128,
      "step": 330
    },
    {
      "epoch": 2.9646017699115044,
      "grad_norm": 5.836208343505859,
      "learning_rate": 4.965298512248466e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 91744,
      "step": 335
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.18477772176265717,
      "eval_runtime": 0.7189,
      "eval_samples_per_second": 34.776,
      "eval_steps_per_second": 18.084,
      "num_input_tokens_seen": 92552,
      "step": 339
    },
    {
      "epoch": 3.0088495575221237,
      "grad_norm": 5.814078330993652,
      "learning_rate": 4.962019382530521e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 92776,
      "step": 340
    },
    {
      "epoch": 3.0530973451327434,
      "grad_norm": 0.12010909616947174,
      "learning_rate": 4.958593418631275e-05,
      "loss": 0.2039,
      "num_input_tokens_seen": 93896,
      "step": 345
    },
    {
      "epoch": 3.0973451327433628,
      "grad_norm": 0.23024114966392517,
      "learning_rate": 4.955020824874307e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 95288,
      "step": 350
    },
    {
      "epoch": 3.1415929203539825,
      "grad_norm": 4.606919765472412,
      "learning_rate": 4.951301814328157e-05,
      "loss": 0.1304,
      "num_input_tokens_seen": 96712,
      "step": 355
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 1.6433827877044678,
      "learning_rate": 4.947436608793624e-05,
      "loss": 0.1061,
      "num_input_tokens_seen": 98120,
      "step": 360
    },
    {
      "epoch": 3.230088495575221,
      "grad_norm": 0.8425772190093994,
      "learning_rate": 4.9434254387905395e-05,
      "loss": 0.0062,
      "num_input_tokens_seen": 99608,
      "step": 365
    },
    {
      "epoch": 3.274336283185841,
      "grad_norm": 0.2277369648218155,
      "learning_rate": 4.9392685435440154e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 100872,
      "step": 370
    },
    {
      "epoch": 3.3185840707964602,
      "grad_norm": 0.9869487881660461,
      "learning_rate": 4.93496617097018e-05,
      "loss": 0.2861,
      "num_input_tokens_seen": 102312,
      "step": 375
    },
    {
      "epoch": 3.3628318584070795,
      "grad_norm": 1.492474913597107,
      "learning_rate": 4.930518577661388e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 103448,
      "step": 380
    },
    {
      "epoch": 3.4070796460176993,
      "grad_norm": 0.12841834127902985,
      "learning_rate": 4.925926028870923e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 104984,
      "step": 385
    },
    {
      "epoch": 3.4513274336283186,
      "grad_norm": 1.710552453994751,
      "learning_rate": 4.921188798497173e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 106568,
      "step": 390
    },
    {
      "epoch": 3.495575221238938,
      "grad_norm": 7.591513156890869,
      "learning_rate": 4.9163071690672973e-05,
      "loss": 0.2732,
      "num_input_tokens_seen": 108184,
      "step": 395
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 7.629326343536377,
      "learning_rate": 4.911281431720378e-05,
      "loss": 0.5439,
      "num_input_tokens_seen": 109528,
      "step": 400
    },
    {
      "epoch": 3.584070796460177,
      "grad_norm": 3.547006845474243,
      "learning_rate": 4.9061118861900537e-05,
      "loss": 0.1592,
      "num_input_tokens_seen": 110968,
      "step": 405
    },
    {
      "epoch": 3.6283185840707963,
      "grad_norm": 7.460524082183838,
      "learning_rate": 4.900798840786645e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 112104,
      "step": 410
    },
    {
      "epoch": 3.672566371681416,
      "grad_norm": 6.08279275894165,
      "learning_rate": 4.8953426123787674e-05,
      "loss": 0.7284,
      "num_input_tokens_seen": 113416,
      "step": 415
    },
    {
      "epoch": 3.7168141592920354,
      "grad_norm": 0.7806785702705383,
      "learning_rate": 4.889743526374432e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 114680,
      "step": 420
    },
    {
      "epoch": 3.7610619469026547,
      "grad_norm": 0.05194513499736786,
      "learning_rate": 4.884001916701639e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 116264,
      "step": 425
    },
    {
      "epoch": 3.8053097345132745,
      "grad_norm": 4.3158488273620605,
      "learning_rate": 4.878118125788462e-05,
      "loss": 0.0841,
      "num_input_tokens_seen": 117624,
      "step": 430
    },
    {
      "epoch": 3.849557522123894,
      "grad_norm": 0.07794106006622314,
      "learning_rate": 4.872092504542629e-05,
      "loss": 0.0741,
      "num_input_tokens_seen": 118808,
      "step": 435
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 10.949700355529785,
      "learning_rate": 4.865925412330586e-05,
      "loss": 0.3256,
      "num_input_tokens_seen": 120632,
      "step": 440
    },
    {
      "epoch": 3.938053097345133,
      "grad_norm": 0.07506046444177628,
      "learning_rate": 4.859617216956074e-05,
      "loss": 0.8648,
      "num_input_tokens_seen": 122040,
      "step": 445
    },
    {
      "epoch": 3.982300884955752,
      "grad_norm": 3.3355536460876465,
      "learning_rate": 4.8531682946381874e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 123592,
      "step": 450
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.29289668798446655,
      "eval_runtime": 0.718,
      "eval_samples_per_second": 34.818,
      "eval_steps_per_second": 18.105,
      "num_input_tokens_seen": 123976,
      "step": 452
    },
    {
      "epoch": 4.0265486725663715,
      "grad_norm": 0.1712837815284729,
      "learning_rate": 4.846579029988939e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 124792,
      "step": 455
    },
    {
      "epoch": 4.070796460176991,
      "grad_norm": 0.7876209020614624,
      "learning_rate": 4.8398498159903194e-05,
      "loss": 0.2165,
      "num_input_tokens_seen": 126488,
      "step": 460
    },
    {
      "epoch": 4.115044247787611,
      "grad_norm": 8.310274124145508,
      "learning_rate": 4.8329810539708625e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 127752,
      "step": 465
    },
    {
      "epoch": 4.15929203539823,
      "grad_norm": 0.25518226623535156,
      "learning_rate": 4.825973153581709e-05,
      "loss": 0.3121,
      "num_input_tokens_seen": 128872,
      "step": 470
    },
    {
      "epoch": 4.20353982300885,
      "grad_norm": 7.439866542816162,
      "learning_rate": 4.818826532772174e-05,
      "loss": 0.2336,
      "num_input_tokens_seen": 130536,
      "step": 475
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 2.1193883419036865,
      "learning_rate": 4.8115416177648234e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 131640,
      "step": 480
    },
    {
      "epoch": 4.292035398230088,
      "grad_norm": 0.17700505256652832,
      "learning_rate": 4.804118843030049e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 133304,
      "step": 485
    },
    {
      "epoch": 4.336283185840708,
      "grad_norm": 10.366985321044922,
      "learning_rate": 4.796558651260165e-05,
      "loss": 0.1367,
      "num_input_tokens_seen": 134744,
      "step": 490
    },
    {
      "epoch": 4.380530973451328,
      "grad_norm": 0.09531063586473465,
      "learning_rate": 4.7888614933429955e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 136072,
      "step": 495
    },
    {
      "epoch": 4.424778761061947,
      "grad_norm": 1.0017696619033813,
      "learning_rate": 4.781027828334994e-05,
      "loss": 0.1347,
      "num_input_tokens_seen": 137368,
      "step": 500
    },
    {
      "epoch": 4.469026548672566,
      "grad_norm": 0.32628941535949707,
      "learning_rate": 4.773058123433857e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 138584,
      "step": 505
    },
    {
      "epoch": 4.513274336283186,
      "grad_norm": 0.013911600224673748,
      "learning_rate": 4.7649528539506673e-05,
      "loss": 0.0264,
      "num_input_tokens_seen": 139928,
      "step": 510
    },
    {
      "epoch": 4.557522123893805,
      "grad_norm": 0.8271692991256714,
      "learning_rate": 4.7567125032815394e-05,
      "loss": 0.7486,
      "num_input_tokens_seen": 141368,
      "step": 515
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 4.551598072052002,
      "learning_rate": 4.7483375628787975e-05,
      "loss": 0.2631,
      "num_input_tokens_seen": 142632,
      "step": 520
    },
    {
      "epoch": 4.646017699115045,
      "grad_norm": 0.38087373971939087,
      "learning_rate": 4.739828532221661e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 143800,
      "step": 525
    },
    {
      "epoch": 4.6902654867256635,
      "grad_norm": 4.788417816162109,
      "learning_rate": 4.731185918786453e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 145176,
      "step": 530
    },
    {
      "epoch": 4.734513274336283,
      "grad_norm": 3.977271556854248,
      "learning_rate": 4.722410238016343e-05,
      "loss": 0.3341,
      "num_input_tokens_seen": 146424,
      "step": 535
    },
    {
      "epoch": 4.778761061946903,
      "grad_norm": 4.327799320220947,
      "learning_rate": 4.7135020132905985e-05,
      "loss": 0.2965,
      "num_input_tokens_seen": 147640,
      "step": 540
    },
    {
      "epoch": 4.823008849557522,
      "grad_norm": 0.4887475073337555,
      "learning_rate": 4.7044617758933714e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 148936,
      "step": 545
    },
    {
      "epoch": 4.867256637168142,
      "grad_norm": 3.0246636867523193,
      "learning_rate": 4.695290064982018e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 150984,
      "step": 550
    },
    {
      "epoch": 4.911504424778761,
      "grad_norm": 0.023805344477295876,
      "learning_rate": 4.6859874275549376e-05,
      "loss": 0.3103,
      "num_input_tokens_seen": 152664,
      "step": 555
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 5.441285133361816,
      "learning_rate": 4.676554418418953e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 153848,
      "step": 560
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8133242726325989,
      "learning_rate": 4.66699160015622e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 155224,
      "step": 565
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.28866487741470337,
      "eval_runtime": 0.7205,
      "eval_samples_per_second": 34.697,
      "eval_steps_per_second": 18.042,
      "num_input_tokens_seen": 155224,
      "step": 565
    },
    {
      "epoch": 5.04424778761062,
      "grad_norm": 2.2284185886383057,
      "learning_rate": 4.6572995430906784e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 156584,
      "step": 570
    },
    {
      "epoch": 5.088495575221239,
      "grad_norm": 8.456231117248535,
      "learning_rate": 4.6474788252540323e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 157832,
      "step": 575
    },
    {
      "epoch": 5.132743362831858,
      "grad_norm": 14.935659408569336,
      "learning_rate": 4.637530032351284e-05,
      "loss": 0.1777,
      "num_input_tokens_seen": 158968,
      "step": 580
    },
    {
      "epoch": 5.176991150442478,
      "grad_norm": 18.7364501953125,
      "learning_rate": 4.627453757725796e-05,
      "loss": 0.3454,
      "num_input_tokens_seen": 160424,
      "step": 585
    },
    {
      "epoch": 5.221238938053097,
      "grad_norm": 0.39791396260261536,
      "learning_rate": 4.617250602323907e-05,
      "loss": 0.2104,
      "num_input_tokens_seen": 161784,
      "step": 590
    },
    {
      "epoch": 5.265486725663717,
      "grad_norm": 1.050893783569336,
      "learning_rate": 4.6069211746590926e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 163368,
      "step": 595
    },
    {
      "epoch": 5.3097345132743365,
      "grad_norm": 0.061371274292469025,
      "learning_rate": 4.596466090775672e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 164648,
      "step": 600
    },
    {
      "epoch": 5.353982300884955,
      "grad_norm": 1.0668625831604004,
      "learning_rate": 4.585885974212068e-05,
      "loss": 0.0137,
      "num_input_tokens_seen": 166008,
      "step": 605
    },
    {
      "epoch": 5.398230088495575,
      "grad_norm": 7.902926921844482,
      "learning_rate": 4.575181455963619e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 167320,
      "step": 610
    },
    {
      "epoch": 5.442477876106195,
      "grad_norm": 0.5615270137786865,
      "learning_rate": 4.5643531744449474e-05,
      "loss": 0.1468,
      "num_input_tokens_seen": 168968,
      "step": 615
    },
    {
      "epoch": 5.486725663716814,
      "grad_norm": 8.79254150390625,
      "learning_rate": 4.553401775451882e-05,
      "loss": 0.3513,
      "num_input_tokens_seen": 170712,
      "step": 620
    },
    {
      "epoch": 5.530973451327434,
      "grad_norm": 0.07487445324659348,
      "learning_rate": 4.542327912122949e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 172216,
      "step": 625
    },
    {
      "epoch": 5.575221238938053,
      "grad_norm": 7.967508316040039,
      "learning_rate": 4.531132244900411e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 173640,
      "step": 630
    },
    {
      "epoch": 5.619469026548672,
      "grad_norm": 0.08340954035520554,
      "learning_rate": 4.519815441490884e-05,
      "loss": 0.1772,
      "num_input_tokens_seen": 174840,
      "step": 635
    },
    {
      "epoch": 5.663716814159292,
      "grad_norm": 1.82449471950531,
      "learning_rate": 4.508378176825516e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 176312,
      "step": 640
    },
    {
      "epoch": 5.707964601769912,
      "grad_norm": 2.036057233810425,
      "learning_rate": 4.496821133019728e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 177784,
      "step": 645
    },
    {
      "epoch": 5.752212389380531,
      "grad_norm": 3.3237433433532715,
      "learning_rate": 4.485144999332541e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 179256,
      "step": 650
    },
    {
      "epoch": 5.79646017699115,
      "grad_norm": 0.14721624553203583,
      "learning_rate": 4.4733504721254625e-05,
      "loss": 0.359,
      "num_input_tokens_seen": 180312,
      "step": 655
    },
    {
      "epoch": 5.84070796460177,
      "grad_norm": 3.8793351650238037,
      "learning_rate": 4.461438254820959e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 182024,
      "step": 660
    },
    {
      "epoch": 5.88495575221239,
      "grad_norm": 9.448125839233398,
      "learning_rate": 4.449409057860504e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 183496,
      "step": 665
    },
    {
      "epoch": 5.929203539823009,
      "grad_norm": 0.10135721415281296,
      "learning_rate": 4.4372635986622044e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 184584,
      "step": 670
    },
    {
      "epoch": 5.9734513274336285,
      "grad_norm": 0.3512079417705536,
      "learning_rate": 4.425002601578017e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 185720,
      "step": 675
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.2238500863313675,
      "eval_runtime": 0.719,
      "eval_samples_per_second": 34.77,
      "eval_steps_per_second": 18.081,
      "num_input_tokens_seen": 186368,
      "step": 678
    },
    {
      "epoch": 6.017699115044247,
      "grad_norm": 0.03489464521408081,
      "learning_rate": 4.4126267978505486e-05,
      "loss": 0.2593,
      "num_input_tokens_seen": 187152,
      "step": 680
    },
    {
      "epoch": 6.061946902654867,
      "grad_norm": 0.9035913944244385,
      "learning_rate": 4.4001369255694416e-05,
      "loss": 0.1324,
      "num_input_tokens_seen": 188560,
      "step": 685
    },
    {
      "epoch": 6.106194690265487,
      "grad_norm": 1.0599687099456787,
      "learning_rate": 4.387533729627359e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 189680,
      "step": 690
    },
    {
      "epoch": 6.150442477876107,
      "grad_norm": 0.013446489349007607,
      "learning_rate": 4.374817961675553e-05,
      "loss": 0.0169,
      "num_input_tokens_seen": 191296,
      "step": 695
    },
    {
      "epoch": 6.1946902654867255,
      "grad_norm": 0.04751233756542206,
      "learning_rate": 4.3619903800790465e-05,
      "loss": 0.1421,
      "num_input_tokens_seen": 192512,
      "step": 700
    },
    {
      "epoch": 6.238938053097345,
      "grad_norm": 19.853158950805664,
      "learning_rate": 4.3490517498713924e-05,
      "loss": 0.3854,
      "num_input_tokens_seen": 193728,
      "step": 705
    },
    {
      "epoch": 6.283185840707965,
      "grad_norm": 4.788447380065918,
      "learning_rate": 4.336002842709057e-05,
      "loss": 0.2173,
      "num_input_tokens_seen": 195136,
      "step": 710
    },
    {
      "epoch": 6.327433628318584,
      "grad_norm": 1.004189133644104,
      "learning_rate": 4.3228444368253925e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 196464,
      "step": 715
    },
    {
      "epoch": 6.371681415929204,
      "grad_norm": 5.019634246826172,
      "learning_rate": 4.309577316984228e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 197872,
      "step": 720
    },
    {
      "epoch": 6.415929203539823,
      "grad_norm": 0.14478158950805664,
      "learning_rate": 4.2962022744330616e-05,
      "loss": 0.1154,
      "num_input_tokens_seen": 199104,
      "step": 725
    },
    {
      "epoch": 6.460176991150442,
      "grad_norm": 0.2269548922777176,
      "learning_rate": 4.282720106855876e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 200624,
      "step": 730
    },
    {
      "epoch": 6.504424778761062,
      "grad_norm": 0.25746700167655945,
      "learning_rate": 4.269131618325559e-05,
      "loss": 0.0016,
      "num_input_tokens_seen": 201712,
      "step": 735
    },
    {
      "epoch": 6.548672566371682,
      "grad_norm": 0.010229190811514854,
      "learning_rate": 4.255437619255955e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 202896,
      "step": 740
    },
    {
      "epoch": 6.592920353982301,
      "grad_norm": 0.8397567272186279,
      "learning_rate": 4.241638926353526e-05,
      "loss": 0.3707,
      "num_input_tokens_seen": 204496,
      "step": 745
    },
    {
      "epoch": 6.6371681415929205,
      "grad_norm": 0.018929218873381615,
      "learning_rate": 4.2277363625686475e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 206016,
      "step": 750
    },
    {
      "epoch": 6.68141592920354,
      "grad_norm": 0.006839994341135025,
      "learning_rate": 4.213730757046528e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 207392,
      "step": 755
    },
    {
      "epoch": 6.725663716814159,
      "grad_norm": 0.009381293319165707,
      "learning_rate": 4.199622945077755e-05,
      "loss": 0.1264,
      "num_input_tokens_seen": 208624,
      "step": 760
    },
    {
      "epoch": 6.769911504424779,
      "grad_norm": 0.14909829199314117,
      "learning_rate": 4.185413768048483e-05,
      "loss": 0.0039,
      "num_input_tokens_seen": 210096,
      "step": 765
    },
    {
      "epoch": 6.814159292035399,
      "grad_norm": 6.23675537109375,
      "learning_rate": 4.1711040733902526e-05,
      "loss": 0.0143,
      "num_input_tokens_seen": 211328,
      "step": 770
    },
    {
      "epoch": 6.8584070796460175,
      "grad_norm": 3.918607234954834,
      "learning_rate": 4.1566947145294474e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 212736,
      "step": 775
    },
    {
      "epoch": 6.902654867256637,
      "grad_norm": 1.782132863998413,
      "learning_rate": 4.142186550836399e-05,
      "loss": 0.1733,
      "num_input_tokens_seen": 214448,
      "step": 780
    },
    {
      "epoch": 6.946902654867257,
      "grad_norm": 0.9768128991127014,
      "learning_rate": 4.127580447574131e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 215680,
      "step": 785
    },
    {
      "epoch": 6.991150442477876,
      "grad_norm": 2.9292213916778564,
      "learning_rate": 4.1128772758467604e-05,
      "loss": 0.0136,
      "num_input_tokens_seen": 217168,
      "step": 790
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.1438879668712616,
      "eval_runtime": 0.7235,
      "eval_samples_per_second": 34.553,
      "eval_steps_per_second": 17.968,
      "num_input_tokens_seen": 217280,
      "step": 791
    },
    {
      "epoch": 7.035398230088496,
      "grad_norm": 0.896648108959198,
      "learning_rate": 4.098077912547536e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 218192,
      "step": 795
    },
    {
      "epoch": 7.079646017699115,
      "grad_norm": 0.006593635305762291,
      "learning_rate": 4.0831832403065526e-05,
      "loss": 0.0054,
      "num_input_tokens_seen": 219664,
      "step": 800
    },
    {
      "epoch": 7.123893805309734,
      "grad_norm": 0.016005270183086395,
      "learning_rate": 4.068194147438101e-05,
      "loss": 0.2851,
      "num_input_tokens_seen": 220784,
      "step": 805
    },
    {
      "epoch": 7.168141592920354,
      "grad_norm": 0.06200814247131348,
      "learning_rate": 4.0531115278876934e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 222048,
      "step": 810
    },
    {
      "epoch": 7.212389380530974,
      "grad_norm": 9.109589576721191,
      "learning_rate": 4.0379362811787504e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 223664,
      "step": 815
    },
    {
      "epoch": 7.256637168141593,
      "grad_norm": 0.003705468960106373,
      "learning_rate": 4.022669312358949e-05,
      "loss": 0.0068,
      "num_input_tokens_seen": 225216,
      "step": 820
    },
    {
      "epoch": 7.300884955752212,
      "grad_norm": 0.0026334961876273155,
      "learning_rate": 4.007311531946252e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 226592,
      "step": 825
    },
    {
      "epoch": 7.345132743362832,
      "grad_norm": 0.006368690170347691,
      "learning_rate": 3.9918638558745966e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 227808,
      "step": 830
    },
    {
      "epoch": 7.389380530973451,
      "grad_norm": 1.219059944152832,
      "learning_rate": 3.976327205439279e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 229344,
      "step": 835
    },
    {
      "epoch": 7.433628318584071,
      "grad_norm": 0.6809638142585754,
      "learning_rate": 3.9607025072419986e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 230464,
      "step": 840
    },
    {
      "epoch": 7.477876106194691,
      "grad_norm": 0.046488989144563675,
      "learning_rate": 3.9449906931356005e-05,
      "loss": 0.0061,
      "num_input_tokens_seen": 231968,
      "step": 845
    },
    {
      "epoch": 7.522123893805309,
      "grad_norm": 0.01232304610311985,
      "learning_rate": 3.929192700168501e-05,
      "loss": 0.0197,
      "num_input_tokens_seen": 233248,
      "step": 850
    },
    {
      "epoch": 7.566371681415929,
      "grad_norm": 0.006855678744614124,
      "learning_rate": 3.9133094705287984e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 234640,
      "step": 855
    },
    {
      "epoch": 7.610619469026549,
      "grad_norm": 0.8517729043960571,
      "learning_rate": 3.897341951488087e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 236192,
      "step": 860
    },
    {
      "epoch": 7.654867256637168,
      "grad_norm": 0.005257199052721262,
      "learning_rate": 3.8812910953449555e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 237696,
      "step": 865
    },
    {
      "epoch": 7.699115044247788,
      "grad_norm": 1.7694588899612427,
      "learning_rate": 3.865157859368196e-05,
      "loss": 0.0038,
      "num_input_tokens_seen": 239152,
      "step": 870
    },
    {
      "epoch": 7.743362831858407,
      "grad_norm": 0.004542375449091196,
      "learning_rate": 3.848943205739711e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 240576,
      "step": 875
    },
    {
      "epoch": 7.787610619469026,
      "grad_norm": 0.0026545259170234203,
      "learning_rate": 3.832648101497134e-05,
      "loss": 1.0045,
      "num_input_tokens_seen": 241824,
      "step": 880
    },
    {
      "epoch": 7.831858407079646,
      "grad_norm": 0.014892376959323883,
      "learning_rate": 3.8162735184761476e-05,
      "loss": 0.005,
      "num_input_tokens_seen": 243136,
      "step": 885
    },
    {
      "epoch": 7.876106194690266,
      "grad_norm": 0.07414080947637558,
      "learning_rate": 3.799820433252529e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 244448,
      "step": 890
    },
    {
      "epoch": 7.920353982300885,
      "grad_norm": 0.5333564877510071,
      "learning_rate": 3.783289827083905e-05,
      "loss": 0.1678,
      "num_input_tokens_seen": 245696,
      "step": 895
    },
    {
      "epoch": 7.964601769911504,
      "grad_norm": 0.10257107019424438,
      "learning_rate": 3.766682685851234e-05,
      "loss": 0.0123,
      "num_input_tokens_seen": 247088,
      "step": 900
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.10195600241422653,
      "eval_runtime": 0.719,
      "eval_samples_per_second": 34.771,
      "eval_steps_per_second": 18.081,
      "num_input_tokens_seen": 248064,
      "step": 904
    },
    {
      "epoch": 8.008849557522124,
      "grad_norm": 0.010762057267129421,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 248224,
      "step": 905
    },
    {
      "epoch": 8.053097345132743,
      "grad_norm": 0.991465151309967,
      "learning_rate": 3.733242764481154e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 249600,
      "step": 910
    },
    {
      "epoch": 8.097345132743364,
      "grad_norm": 0.0036801989190280437,
      "learning_rate": 3.716411978691766e-05,
      "loss": 0.0244,
      "num_input_tokens_seen": 250896,
      "step": 915
    },
    {
      "epoch": 8.141592920353983,
      "grad_norm": 2.2528061866760254,
      "learning_rate": 3.699508646415424e-05,
      "loss": 0.0119,
      "num_input_tokens_seen": 252192,
      "step": 920
    },
    {
      "epoch": 8.185840707964601,
      "grad_norm": 0.3149694800376892,
      "learning_rate": 3.6825337757623696e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 253632,
      "step": 925
    },
    {
      "epoch": 8.230088495575222,
      "grad_norm": 0.09691987186670303,
      "learning_rate": 3.665488379109377e-05,
      "loss": 0.0067,
      "num_input_tokens_seen": 254912,
      "step": 930
    },
    {
      "epoch": 8.274336283185841,
      "grad_norm": 0.01924165152013302,
      "learning_rate": 3.648373473039368e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 256608,
      "step": 935
    },
    {
      "epoch": 8.31858407079646,
      "grad_norm": 0.006609531119465828,
      "learning_rate": 3.631190078280791e-05,
      "loss": 0.0016,
      "num_input_tokens_seen": 257984,
      "step": 940
    },
    {
      "epoch": 8.36283185840708,
      "grad_norm": 0.009501444175839424,
      "learning_rate": 3.613939219646739e-05,
      "loss": 0.0047,
      "num_input_tokens_seen": 259456,
      "step": 945
    },
    {
      "epoch": 8.4070796460177,
      "grad_norm": 0.0061758230440318584,
      "learning_rate": 3.596621925973835e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 260992,
      "step": 950
    },
    {
      "epoch": 8.451327433628318,
      "grad_norm": 0.009280437603592873,
      "learning_rate": 3.579239230060867e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 262256,
      "step": 955
    },
    {
      "epoch": 8.495575221238939,
      "grad_norm": 2.6710667610168457,
      "learning_rate": 3.5617921686071995e-05,
      "loss": 0.0038,
      "num_input_tokens_seen": 263616,
      "step": 960
    },
    {
      "epoch": 8.539823008849558,
      "grad_norm": 0.7105626463890076,
      "learning_rate": 3.544281782150936e-05,
      "loss": 0.0016,
      "num_input_tokens_seen": 264768,
      "step": 965
    },
    {
      "epoch": 8.584070796460177,
      "grad_norm": 0.012485911138355732,
      "learning_rate": 3.526709115006871e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 266224,
      "step": 970
    },
    {
      "epoch": 8.628318584070797,
      "grad_norm": 8.606228828430176,
      "learning_rate": 3.5090752152041975e-05,
      "loss": 0.0161,
      "num_input_tokens_seen": 267648,
      "step": 975
    },
    {
      "epoch": 8.672566371681416,
      "grad_norm": 0.11525888741016388,
      "learning_rate": 3.491381134424012e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 268912,
      "step": 980
    },
    {
      "epoch": 8.716814159292035,
      "grad_norm": 0.005239203572273254,
      "learning_rate": 3.4736279279365876e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 270304,
      "step": 985
    },
    {
      "epoch": 8.761061946902656,
      "grad_norm": 12.8204984664917,
      "learning_rate": 3.455816654538438e-05,
      "loss": 0.0978,
      "num_input_tokens_seen": 271632,
      "step": 990
    },
    {
      "epoch": 8.805309734513274,
      "grad_norm": 0.1226809099316597,
      "learning_rate": 3.437948376489172e-05,
      "loss": 0.0027,
      "num_input_tokens_seen": 272880,
      "step": 995
    },
    {
      "epoch": 8.849557522123893,
      "grad_norm": 0.04560926556587219,
      "learning_rate": 3.420024159448142e-05,
      "loss": 0.0024,
      "num_input_tokens_seen": 274176,
      "step": 1000
    },
    {
      "epoch": 8.893805309734514,
      "grad_norm": 0.05620700120925903,
      "learning_rate": 3.402045072410886e-05,
      "loss": 0.0171,
      "num_input_tokens_seen": 275456,
      "step": 1005
    },
    {
      "epoch": 8.938053097345133,
      "grad_norm": 0.005380111746490002,
      "learning_rate": 3.3840121876453734e-05,
      "loss": 0.0058,
      "num_input_tokens_seen": 276912,
      "step": 1010
    },
    {
      "epoch": 8.982300884955752,
      "grad_norm": 0.025887787342071533,
      "learning_rate": 3.365926580628057e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 278272,
      "step": 1015
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.2387761026620865,
      "eval_runtime": 0.7189,
      "eval_samples_per_second": 34.776,
      "eval_steps_per_second": 18.084,
      "num_input_tokens_seen": 278576,
      "step": 1017
    },
    {
      "epoch": 9.026548672566372,
      "grad_norm": 0.004052956588566303,
      "learning_rate": 3.3477893299797304e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 279424,
      "step": 1020
    },
    {
      "epoch": 9.070796460176991,
      "grad_norm": 27.44761085510254,
      "learning_rate": 3.3296015174011984e-05,
      "loss": 0.1137,
      "num_input_tokens_seen": 280592,
      "step": 1025
    },
    {
      "epoch": 9.11504424778761,
      "grad_norm": 0.0014455394120886922,
      "learning_rate": 3.311364227608768e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 281872,
      "step": 1030
    },
    {
      "epoch": 9.15929203539823,
      "grad_norm": 0.002056779805570841,
      "learning_rate": 3.293078548269553e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 283200,
      "step": 1035
    },
    {
      "epoch": 9.20353982300885,
      "grad_norm": 0.0055632153525948524,
      "learning_rate": 3.2747455699366056e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 284736,
      "step": 1040
    },
    {
      "epoch": 9.247787610619469,
      "grad_norm": 0.02439134195446968,
      "learning_rate": 3.256366385983879e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 285760,
      "step": 1045
    },
    {
      "epoch": 9.29203539823009,
      "grad_norm": 0.13755837082862854,
      "learning_rate": 3.237942092541018e-05,
      "loss": 0.0097,
      "num_input_tokens_seen": 287120,
      "step": 1050
    },
    {
      "epoch": 9.336283185840708,
      "grad_norm": 1.6239471435546875,
      "learning_rate": 3.219473788427984e-05,
      "loss": 0.0048,
      "num_input_tokens_seen": 288256,
      "step": 1055
    },
    {
      "epoch": 9.380530973451327,
      "grad_norm": 0.055729396641254425,
      "learning_rate": 3.2009625750895224e-05,
      "loss": 0.0096,
      "num_input_tokens_seen": 289440,
      "step": 1060
    },
    {
      "epoch": 9.424778761061948,
      "grad_norm": 0.005138949956744909,
      "learning_rate": 3.182409556529476e-05,
      "loss": 0.0076,
      "num_input_tokens_seen": 290832,
      "step": 1065
    },
    {
      "epoch": 9.469026548672566,
      "grad_norm": 0.002471720101311803,
      "learning_rate": 3.163815839244937e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 292672,
      "step": 1070
    },
    {
      "epoch": 9.513274336283185,
      "grad_norm": 0.0025828483048826456,
      "learning_rate": 3.14518253216026e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 293872,
      "step": 1075
    },
    {
      "epoch": 9.557522123893806,
      "grad_norm": 0.041861437261104584,
      "learning_rate": 3.126510746560925e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 295168,
      "step": 1080
    },
    {
      "epoch": 9.601769911504425,
      "grad_norm": 0.011501029133796692,
      "learning_rate": 3.107801596027261e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 296336,
      "step": 1085
    },
    {
      "epoch": 9.646017699115044,
      "grad_norm": 0.003055627690628171,
      "learning_rate": 3.0890561963680306e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 297904,
      "step": 1090
    },
    {
      "epoch": 9.690265486725664,
      "grad_norm": 0.014435662887990475,
      "learning_rate": 3.0702756655538835e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 299536,
      "step": 1095
    },
    {
      "epoch": 9.734513274336283,
      "grad_norm": 0.002437986433506012,
      "learning_rate": 3.051461123650685e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 300800,
      "step": 1100
    },
    {
      "epoch": 9.778761061946902,
      "grad_norm": 0.006748379208147526,
      "learning_rate": 3.032613692752711e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 302288,
      "step": 1105
    },
    {
      "epoch": 9.823008849557523,
      "grad_norm": 0.0018612981075420976,
      "learning_rate": 3.0137344969157284e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 303760,
      "step": 1110
    },
    {
      "epoch": 9.867256637168142,
      "grad_norm": 0.0028278871905058622,
      "learning_rate": 2.9948246620899557e-05,
      "loss": 0.0043,
      "num_input_tokens_seen": 305296,
      "step": 1115
    },
    {
      "epoch": 9.91150442477876,
      "grad_norm": 0.0011935485526919365,
      "learning_rate": 2.9758853160529148e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 306704,
      "step": 1120
    },
    {
      "epoch": 9.955752212389381,
      "grad_norm": 0.0016115003963932395,
      "learning_rate": 2.9569175883421672e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 308256,
      "step": 1125
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0005373014719225466,
      "learning_rate": 2.93792261018795e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 309584,
      "step": 1130
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.17034229636192322,
      "eval_runtime": 0.7168,
      "eval_samples_per_second": 34.876,
      "eval_steps_per_second": 18.136,
      "num_input_tokens_seen": 309584,
      "step": 1130
    },
    {
      "epoch": 10.044247787610619,
      "grad_norm": 0.0006830533384345472,
      "learning_rate": 2.9189015144457087e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 311040,
      "step": 1135
    },
    {
      "epoch": 10.08849557522124,
      "grad_norm": 0.0070992279797792435,
      "learning_rate": 2.8998554355285355e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 312496,
      "step": 1140
    },
    {
      "epoch": 10.132743362831858,
      "grad_norm": 0.0010292730294167995,
      "learning_rate": 2.8807855093395126e-05,
      "loss": 0.0041,
      "num_input_tokens_seen": 313680,
      "step": 1145
    },
    {
      "epoch": 10.176991150442477,
      "grad_norm": 0.5420211553573608,
      "learning_rate": 2.8616928732039684e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 315248,
      "step": 1150
    },
    {
      "epoch": 10.221238938053098,
      "grad_norm": 0.0005968459299765527,
      "learning_rate": 2.8425786658016423e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 316704,
      "step": 1155
    },
    {
      "epoch": 10.265486725663717,
      "grad_norm": 0.020497633144259453,
      "learning_rate": 2.8234440270987837e-05,
      "loss": 0.0015,
      "num_input_tokens_seen": 318144,
      "step": 1160
    },
    {
      "epoch": 10.309734513274336,
      "grad_norm": 0.00072331796400249,
      "learning_rate": 2.804290098280155e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 319408,
      "step": 1165
    },
    {
      "epoch": 10.353982300884956,
      "grad_norm": 0.06552136689424515,
      "learning_rate": 2.7851180216809796e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 320832,
      "step": 1170
    },
    {
      "epoch": 10.398230088495575,
      "grad_norm": 0.0033154201228171587,
      "learning_rate": 2.765928940718806e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 322048,
      "step": 1175
    },
    {
      "epoch": 10.442477876106194,
      "grad_norm": 0.0014622806338593364,
      "learning_rate": 2.7467239998253214e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 323440,
      "step": 1180
    },
    {
      "epoch": 10.486725663716815,
      "grad_norm": 0.04821912199258804,
      "learning_rate": 2.7275043443780934e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 324688,
      "step": 1185
    },
    {
      "epoch": 10.530973451327434,
      "grad_norm": 0.029812661930918694,
      "learning_rate": 2.708271120632262e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 325808,
      "step": 1190
    },
    {
      "epoch": 10.575221238938052,
      "grad_norm": 0.012087096460163593,
      "learning_rate": 2.6890254756521778e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 327408,
      "step": 1195
    },
    {
      "epoch": 10.619469026548673,
      "grad_norm": 0.000868955219630152,
      "learning_rate": 2.6697685572429886e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 328816,
      "step": 1200
    },
    {
      "epoch": 10.663716814159292,
      "grad_norm": 0.0009623424848541617,
      "learning_rate": 2.65050151388219e-05,
      "loss": 0.0079,
      "num_input_tokens_seen": 330128,
      "step": 1205
    },
    {
      "epoch": 10.70796460176991,
      "grad_norm": 0.0017418720526620746,
      "learning_rate": 2.6312254946511217e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 331696,
      "step": 1210
    },
    {
      "epoch": 10.752212389380531,
      "grad_norm": 0.009964069351553917,
      "learning_rate": 2.6119416491664472e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 333344,
      "step": 1215
    },
    {
      "epoch": 10.79646017699115,
      "grad_norm": 0.0011894770432263613,
      "learning_rate": 2.5926511275115827e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 334784,
      "step": 1220
    },
    {
      "epoch": 10.84070796460177,
      "grad_norm": 0.009204513393342495,
      "learning_rate": 2.57335508016811e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 336128,
      "step": 1225
    },
    {
      "epoch": 10.88495575221239,
      "grad_norm": 0.002534894971176982,
      "learning_rate": 2.5540546579471624e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 337392,
      "step": 1230
    },
    {
      "epoch": 10.929203539823009,
      "grad_norm": 0.001946152769960463,
      "learning_rate": 2.5347510119207878e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 338688,
      "step": 1235
    },
    {
      "epoch": 10.973451327433628,
      "grad_norm": 0.0021194510627537966,
      "learning_rate": 2.515445293353304e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 340096,
      "step": 1240
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.21472612023353577,
      "eval_runtime": 0.7168,
      "eval_samples_per_second": 34.876,
      "eval_steps_per_second": 18.136,
      "num_input_tokens_seen": 340752,
      "step": 1243
    },
    {
      "epoch": 11.017699115044248,
      "grad_norm": 0.015722211450338364,
      "learning_rate": 2.4961386536326307e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 341328,
      "step": 1245
    },
    {
      "epoch": 11.061946902654867,
      "grad_norm": 0.002159221563488245,
      "learning_rate": 2.4768322442016278e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 342704,
      "step": 1250
    },
    {
      "epoch": 11.106194690265486,
      "grad_norm": 0.009501132182776928,
      "learning_rate": 2.457527216489421e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 344000,
      "step": 1255
    },
    {
      "epoch": 11.150442477876107,
      "grad_norm": 0.0016542176017537713,
      "learning_rate": 2.438224721842728e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 345216,
      "step": 1260
    },
    {
      "epoch": 11.194690265486726,
      "grad_norm": 0.0008815676555968821,
      "learning_rate": 2.4189259114571984e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 346736,
      "step": 1265
    },
    {
      "epoch": 11.238938053097344,
      "grad_norm": 0.0017127166502177715,
      "learning_rate": 2.39963193630875e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 348000,
      "step": 1270
    },
    {
      "epoch": 11.283185840707965,
      "grad_norm": 0.013254914432764053,
      "learning_rate": 2.3803439470849335e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 349344,
      "step": 1275
    },
    {
      "epoch": 11.327433628318584,
      "grad_norm": 0.010147323831915855,
      "learning_rate": 2.361063094116293e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 351056,
      "step": 1280
    },
    {
      "epoch": 11.371681415929203,
      "grad_norm": 0.0026426322292536497,
      "learning_rate": 2.3417905273077756e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 352768,
      "step": 1285
    },
    {
      "epoch": 11.415929203539823,
      "grad_norm": 0.0006548184319399297,
      "learning_rate": 2.32252739607014e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 354224,
      "step": 1290
    },
    {
      "epoch": 11.460176991150442,
      "grad_norm": 0.0005002717371098697,
      "learning_rate": 2.3032748492514116e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 355648,
      "step": 1295
    },
    {
      "epoch": 11.504424778761061,
      "grad_norm": 0.0006029328797012568,
      "learning_rate": 2.2840340350683622e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 356896,
      "step": 1300
    },
    {
      "epoch": 11.548672566371682,
      "grad_norm": 0.005663470830768347,
      "learning_rate": 2.2648061010380346e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 358224,
      "step": 1305
    },
    {
      "epoch": 11.5929203539823,
      "grad_norm": 0.0008107605972327292,
      "learning_rate": 2.2455921939093e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 359632,
      "step": 1310
    },
    {
      "epoch": 11.63716814159292,
      "grad_norm": 0.0007355871493928134,
      "learning_rate": 2.2263934595944716e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 361024,
      "step": 1315
    },
    {
      "epoch": 11.68141592920354,
      "grad_norm": 0.003707675961777568,
      "learning_rate": 2.207211043100958e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 362464,
      "step": 1320
    },
    {
      "epoch": 11.725663716814159,
      "grad_norm": 0.017738239839673042,
      "learning_rate": 2.188046088462979e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 363680,
      "step": 1325
    },
    {
      "epoch": 11.769911504424778,
      "grad_norm": 0.0035451988223940134,
      "learning_rate": 2.1688997386733316e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 364896,
      "step": 1330
    },
    {
      "epoch": 11.814159292035399,
      "grad_norm": 0.0006249516154639423,
      "learning_rate": 2.1497731356152286e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 366224,
      "step": 1335
    },
    {
      "epoch": 11.858407079646017,
      "grad_norm": 0.0005412474274635315,
      "learning_rate": 2.1306674199941872e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 367808,
      "step": 1340
    },
    {
      "epoch": 11.902654867256636,
      "grad_norm": 0.0006882218876853585,
      "learning_rate": 2.1115837312700088e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 369392,
      "step": 1345
    },
    {
      "epoch": 11.946902654867257,
      "grad_norm": 0.0007479353225789964,
      "learning_rate": 2.0925232075888143e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 370784,
      "step": 1350
    },
    {
      "epoch": 11.991150442477876,
      "grad_norm": 0.0007245749584399164,
      "learning_rate": 2.0734869857151666e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 372128,
      "step": 1355
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.26409417390823364,
      "eval_runtime": 0.7181,
      "eval_samples_per_second": 34.814,
      "eval_steps_per_second": 18.104,
      "num_input_tokens_seen": 372240,
      "step": 1356
    },
    {
      "epoch": 12.035398230088495,
      "grad_norm": 0.006367493886500597,
      "learning_rate": 2.054476200964278e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 373312,
      "step": 1360
    },
    {
      "epoch": 12.079646017699115,
      "grad_norm": 0.0037162748631089926,
      "learning_rate": 2.035491987134294e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 374592,
      "step": 1365
    },
    {
      "epoch": 12.123893805309734,
      "grad_norm": 0.0011322189820930362,
      "learning_rate": 2.0165354764386807e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 375888,
      "step": 1370
    },
    {
      "epoch": 12.168141592920353,
      "grad_norm": 0.000572588003706187,
      "learning_rate": 1.997607799438694e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 377072,
      "step": 1375
    },
    {
      "epoch": 12.212389380530974,
      "grad_norm": 0.01743907295167446,
      "learning_rate": 1.978710084975959e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 378336,
      "step": 1380
    },
    {
      "epoch": 12.256637168141593,
      "grad_norm": 0.0007353649707511067,
      "learning_rate": 1.9598434601051386e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 380016,
      "step": 1385
    },
    {
      "epoch": 12.300884955752213,
      "grad_norm": 0.002723166486248374,
      "learning_rate": 1.941009050026726e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 381264,
      "step": 1390
    },
    {
      "epoch": 12.345132743362832,
      "grad_norm": 0.0006109076202847064,
      "learning_rate": 1.922207978019928e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 382672,
      "step": 1395
    },
    {
      "epoch": 12.389380530973451,
      "grad_norm": 0.0009360491530969739,
      "learning_rate": 1.903441365375681e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 384576,
      "step": 1400
    },
    {
      "epoch": 12.43362831858407,
      "grad_norm": 0.000474228902021423,
      "learning_rate": 1.884710331329772e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 386032,
      "step": 1405
    },
    {
      "epoch": 12.47787610619469,
      "grad_norm": 0.00048585364129394293,
      "learning_rate": 1.8660159929960914e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 387456,
      "step": 1410
    },
    {
      "epoch": 12.52212389380531,
      "grad_norm": 0.0010553976753726602,
      "learning_rate": 1.847359465300006e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 388896,
      "step": 1415
    },
    {
      "epoch": 12.56637168141593,
      "grad_norm": 0.0011701054172590375,
      "learning_rate": 1.828741860911867e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 390016,
      "step": 1420
    },
    {
      "epoch": 12.610619469026549,
      "grad_norm": 0.000516147818416357,
      "learning_rate": 1.8101642901806486e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 391504,
      "step": 1425
    },
    {
      "epoch": 12.654867256637168,
      "grad_norm": 0.009110220707952976,
      "learning_rate": 1.791627861067731e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 392720,
      "step": 1430
    },
    {
      "epoch": 12.699115044247787,
      "grad_norm": 0.00038364005740731955,
      "learning_rate": 1.7731336790808146e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 393792,
      "step": 1435
    },
    {
      "epoch": 12.743362831858407,
      "grad_norm": 0.0007843800121918321,
      "learning_rate": 1.7546828472079992e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 395024,
      "step": 1440
    },
    {
      "epoch": 12.787610619469026,
      "grad_norm": 0.0008192975074052811,
      "learning_rate": 1.7362764658519877e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 396448,
      "step": 1445
    },
    {
      "epoch": 12.831858407079647,
      "grad_norm": 0.014924599789083004,
      "learning_rate": 1.7179156327644724e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 397664,
      "step": 1450
    },
    {
      "epoch": 12.876106194690266,
      "grad_norm": 0.001769204973243177,
      "learning_rate": 1.699601442980655e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 398960,
      "step": 1455
    },
    {
      "epoch": 12.920353982300885,
      "grad_norm": 0.0009972075931727886,
      "learning_rate": 1.6813349887539443e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 400400,
      "step": 1460
    },
    {
      "epoch": 12.964601769911503,
      "grad_norm": 0.000636521028354764,
      "learning_rate": 1.663117359490814e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 401616,
      "step": 1465
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.2672768533229828,
      "eval_runtime": 0.7192,
      "eval_samples_per_second": 34.76,
      "eval_steps_per_second": 18.075,
      "num_input_tokens_seen": 402976,
      "step": 1469
    },
    {
      "epoch": 13.008849557522124,
      "grad_norm": 0.00043239217484369874,
      "learning_rate": 1.6449496416858284e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 403232,
      "step": 1470
    },
    {
      "epoch": 13.053097345132743,
      "grad_norm": 0.0005190548836253583,
      "learning_rate": 1.6268329188568468e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 404368,
      "step": 1475
    },
    {
      "epoch": 13.097345132743364,
      "grad_norm": 0.01273986604064703,
      "learning_rate": 1.6087682714804002e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 405680,
      "step": 1480
    },
    {
      "epoch": 13.141592920353983,
      "grad_norm": 0.0007397303124889731,
      "learning_rate": 1.5907567769272568e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 406896,
      "step": 1485
    },
    {
      "epoch": 13.185840707964601,
      "grad_norm": 0.0006049118819646537,
      "learning_rate": 1.5727995093981598e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 408208,
      "step": 1490
    },
    {
      "epoch": 13.230088495575222,
      "grad_norm": 0.0008805468096397817,
      "learning_rate": 1.5548975398597718e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 409888,
      "step": 1495
    },
    {
      "epoch": 13.274336283185841,
      "grad_norm": 0.0004492686130106449,
      "learning_rate": 1.537051935980794e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 411136,
      "step": 1500
    },
    {
      "epoch": 13.31858407079646,
      "grad_norm": 0.0007866571540944278,
      "learning_rate": 1.5192637620682981e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 412432,
      "step": 1505
    },
    {
      "epoch": 13.36283185840708,
      "grad_norm": 0.00045057214447297156,
      "learning_rate": 1.5015340790042446e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 413648,
      "step": 1510
    },
    {
      "epoch": 13.4070796460177,
      "grad_norm": 0.0005917617818340659,
      "learning_rate": 1.4838639441822183e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 414752,
      "step": 1515
    },
    {
      "epoch": 13.451327433628318,
      "grad_norm": 0.013488112017512321,
      "learning_rate": 1.46625441144436e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 416256,
      "step": 1520
    },
    {
      "epoch": 13.495575221238939,
      "grad_norm": 0.008368285372853279,
      "learning_rate": 1.4487065310185202e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 417440,
      "step": 1525
    },
    {
      "epoch": 13.539823008849558,
      "grad_norm": 0.0006658629863522947,
      "learning_rate": 1.4312213494556218e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 419104,
      "step": 1530
    },
    {
      "epoch": 13.584070796460177,
      "grad_norm": 0.008035008795559406,
      "learning_rate": 1.4137999095672444e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 420656,
      "step": 1535
    },
    {
      "epoch": 13.628318584070797,
      "grad_norm": 0.0008954021614044905,
      "learning_rate": 1.3964432503634281e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 422320,
      "step": 1540
    },
    {
      "epoch": 13.672566371681416,
      "grad_norm": 0.003024559933692217,
      "learning_rate": 1.3791524069907141e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 423808,
      "step": 1545
    },
    {
      "epoch": 13.716814159292035,
      "grad_norm": 0.003669372759759426,
      "learning_rate": 1.361928410670403e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 424976,
      "step": 1550
    },
    {
      "epoch": 13.761061946902656,
      "grad_norm": 0.0005071991472505033,
      "learning_rate": 1.3447722886370565e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 426176,
      "step": 1555
    },
    {
      "epoch": 13.805309734513274,
      "grad_norm": 0.00048327608965337276,
      "learning_rate": 1.3276850640772288e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 427968,
      "step": 1560
    },
    {
      "epoch": 13.849557522123893,
      "grad_norm": 0.001241826917976141,
      "learning_rate": 1.3106677560684494e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 429040,
      "step": 1565
    },
    {
      "epoch": 13.893805309734514,
      "grad_norm": 0.0008833057363517582,
      "learning_rate": 1.2937213795184434e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 430480,
      "step": 1570
    },
    {
      "epoch": 13.938053097345133,
      "grad_norm": 0.0005532755749300122,
      "learning_rate": 1.2768469451046029e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 431648,
      "step": 1575
    },
    {
      "epoch": 13.982300884955752,
      "grad_norm": 0.0006302348920144141,
      "learning_rate": 1.2600454592137062e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 433200,
      "step": 1580
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.2683930993080139,
      "eval_runtime": 0.7175,
      "eval_samples_per_second": 34.843,
      "eval_steps_per_second": 18.118,
      "num_input_tokens_seen": 433800,
      "step": 1582
    },
    {
      "epoch": 14.026548672566372,
      "grad_norm": 0.0006576204905286431,
      "learning_rate": 1.2433179238819077e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 434712,
      "step": 1585
    },
    {
      "epoch": 14.070796460176991,
      "grad_norm": 0.0016585837583988905,
      "learning_rate": 1.2266653367349657e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 436104,
      "step": 1590
    },
    {
      "epoch": 14.11504424778761,
      "grad_norm": 0.0008089800830930471,
      "learning_rate": 1.2100886909287478e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 437240,
      "step": 1595
    },
    {
      "epoch": 14.15929203539823,
      "grad_norm": 0.0009019608842208982,
      "learning_rate": 1.1935889750900034e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 438712,
      "step": 1600
    },
    {
      "epoch": 14.20353982300885,
      "grad_norm": 0.0007121966918930411,
      "learning_rate": 1.1771671732573976e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 440072,
      "step": 1605
    },
    {
      "epoch": 14.247787610619469,
      "grad_norm": 0.007135974708944559,
      "learning_rate": 1.1608242648228257e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 441368,
      "step": 1610
    },
    {
      "epoch": 14.29203539823009,
      "grad_norm": 0.00034434202825650573,
      "learning_rate": 1.1445612244729984e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 442664,
      "step": 1615
    },
    {
      "epoch": 14.336283185840708,
      "grad_norm": 0.0006468449137173593,
      "learning_rate": 1.1283790221313208e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 444296,
      "step": 1620
    },
    {
      "epoch": 14.380530973451327,
      "grad_norm": 0.0005794033058919013,
      "learning_rate": 1.1122786229000356e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 445832,
      "step": 1625
    },
    {
      "epoch": 14.424778761061948,
      "grad_norm": 0.0014206281630322337,
      "learning_rate": 1.0962609870026724e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 447016,
      "step": 1630
    },
    {
      "epoch": 14.469026548672566,
      "grad_norm": 0.0006809905171394348,
      "learning_rate": 1.0803270697267764e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 448392,
      "step": 1635
    },
    {
      "epoch": 14.513274336283185,
      "grad_norm": 0.0003818897239398211,
      "learning_rate": 1.0644778213669385e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 449592,
      "step": 1640
    },
    {
      "epoch": 14.557522123893806,
      "grad_norm": 0.016543259844183922,
      "learning_rate": 1.0487141871681142e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 450776,
      "step": 1645
    },
    {
      "epoch": 14.601769911504425,
      "grad_norm": 0.00043906690552830696,
      "learning_rate": 1.0330371072692565e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 452440,
      "step": 1650
    },
    {
      "epoch": 14.646017699115044,
      "grad_norm": 0.007470273412764072,
      "learning_rate": 1.0174475166472417e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 453624,
      "step": 1655
    },
    {
      "epoch": 14.690265486725664,
      "grad_norm": 0.0004087320703547448,
      "learning_rate": 1.0019463450611103e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 455656,
      "step": 1660
    },
    {
      "epoch": 14.734513274336283,
      "grad_norm": 0.0008214203990064561,
      "learning_rate": 9.865345169966114e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 457064,
      "step": 1665
    },
    {
      "epoch": 14.778761061946902,
      "grad_norm": 0.00039667729288339615,
      "learning_rate": 9.71212951611074e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 458648,
      "step": 1670
    },
    {
      "epoch": 14.823008849557523,
      "grad_norm": 0.0007470098207704723,
      "learning_rate": 9.559825626785837e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 460232,
      "step": 1675
    },
    {
      "epoch": 14.867256637168142,
      "grad_norm": 0.0009791249176487327,
      "learning_rate": 9.40844258535487e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 461480,
      "step": 1680
    },
    {
      "epoch": 14.91150442477876,
      "grad_norm": 0.0028466489166021347,
      "learning_rate": 9.257989420262151e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 462840,
      "step": 1685
    },
    {
      "epoch": 14.955752212389381,
      "grad_norm": 0.0005519299884326756,
      "learning_rate": 9.108475104494475e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 463992,
      "step": 1690
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.0005300536286085844,
      "learning_rate": 8.959908555045846e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 465096,
      "step": 1695
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.2670702338218689,
      "eval_runtime": 0.7189,
      "eval_samples_per_second": 34.774,
      "eval_steps_per_second": 18.083,
      "num_input_tokens_seen": 465096,
      "step": 1695
    },
    {
      "epoch": 15.044247787610619,
      "grad_norm": 0.0009419475682079792,
      "learning_rate": 8.812298632385784e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 466440,
      "step": 1700
    },
    {
      "epoch": 15.08849557522124,
      "grad_norm": 0.000824528920929879,
      "learning_rate": 8.66565413993082e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 467640,
      "step": 1705
    },
    {
      "epoch": 15.132743362831858,
      "grad_norm": 0.005965210031718016,
      "learning_rate": 8.519983823519496e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 468936,
      "step": 1710
    },
    {
      "epoch": 15.176991150442477,
      "grad_norm": 0.007382329553365707,
      "learning_rate": 8.375296370890749e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 470152,
      "step": 1715
    },
    {
      "epoch": 15.221238938053098,
      "grad_norm": 0.008001770824193954,
      "learning_rate": 8.231600411165757e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 471928,
      "step": 1720
    },
    {
      "epoch": 15.265486725663717,
      "grad_norm": 0.0007758951396681368,
      "learning_rate": 8.088904514333384e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 473272,
      "step": 1725
    },
    {
      "epoch": 15.309734513274336,
      "grad_norm": 0.001055500004440546,
      "learning_rate": 7.947217190738945e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 474728,
      "step": 1730
    },
    {
      "epoch": 15.353982300884956,
      "grad_norm": 0.0006287958822213113,
      "learning_rate": 7.806546890576753e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 476040,
      "step": 1735
    },
    {
      "epoch": 15.398230088495575,
      "grad_norm": 0.0005976333632133901,
      "learning_rate": 7.666902003386104e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 477304,
      "step": 1740
    },
    {
      "epoch": 15.442477876106194,
      "grad_norm": 0.0008192738168872893,
      "learning_rate": 7.528290857550943e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 478472,
      "step": 1745
    },
    {
      "epoch": 15.486725663716815,
      "grad_norm": 0.0008036806830205023,
      "learning_rate": 7.390721719803137e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 479976,
      "step": 1750
    },
    {
      "epoch": 15.530973451327434,
      "grad_norm": 0.0006847975309938192,
      "learning_rate": 7.254202794729484e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 481448,
      "step": 1755
    },
    {
      "epoch": 15.575221238938052,
      "grad_norm": 0.00041765187052078545,
      "learning_rate": 7.11874222428238e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 482952,
      "step": 1760
    },
    {
      "epoch": 15.619469026548673,
      "grad_norm": 0.00045087214675731957,
      "learning_rate": 6.9843480872942294e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 484696,
      "step": 1765
    },
    {
      "epoch": 15.663716814159292,
      "grad_norm": 0.0011875146301463246,
      "learning_rate": 6.851028398995607e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 485912,
      "step": 1770
    },
    {
      "epoch": 15.70796460176991,
      "grad_norm": 0.0016977775376290083,
      "learning_rate": 6.718791110537287e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 487144,
      "step": 1775
    },
    {
      "epoch": 15.752212389380531,
      "grad_norm": 0.0003741350956261158,
      "learning_rate": 6.587644108515986e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 488808,
      "step": 1780
    },
    {
      "epoch": 15.79646017699115,
      "grad_norm": 0.0010718172416090965,
      "learning_rate": 6.457595214504042e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 490152,
      "step": 1785
    },
    {
      "epoch": 15.84070796460177,
      "grad_norm": 0.004870334640145302,
      "learning_rate": 6.328652184582884e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 491160,
      "step": 1790
    },
    {
      "epoch": 15.88495575221239,
      "grad_norm": 0.003318296978250146,
      "learning_rate": 6.200822708880563e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 492872,
      "step": 1795
    },
    {
      "epoch": 15.929203539823009,
      "grad_norm": 0.006690215785056353,
      "learning_rate": 6.074114411112997e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 494024,
      "step": 1800
    },
    {
      "epoch": 15.973451327433628,
      "grad_norm": 0.0025137620978057384,
      "learning_rate": 5.948534848129378e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 495432,
      "step": 1805
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.2720543146133423,
      "eval_runtime": 0.7185,
      "eval_samples_per_second": 34.793,
      "eval_steps_per_second": 18.092,
      "num_input_tokens_seen": 496184,
      "step": 1808
    },
    {
      "epoch": 16.01769911504425,
      "grad_norm": 0.00030161571339704096,
      "learning_rate": 5.824091509461449e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 496600,
      "step": 1810
    },
    {
      "epoch": 16.061946902654867,
      "grad_norm": 0.0003936327120754868,
      "learning_rate": 5.7007918168768405e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 498200,
      "step": 1815
    },
    {
      "epoch": 16.106194690265486,
      "grad_norm": 0.0003133926074951887,
      "learning_rate": 5.5786431239364365e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 499752,
      "step": 1820
    },
    {
      "epoch": 16.150442477876105,
      "grad_norm": 0.0003800129343289882,
      "learning_rate": 5.457652715555781e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 501464,
      "step": 1825
    },
    {
      "epoch": 16.194690265486727,
      "grad_norm": 0.0003382577560842037,
      "learning_rate": 5.337827807570689e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 502760,
      "step": 1830
    },
    {
      "epoch": 16.238938053097346,
      "grad_norm": 0.007132966071367264,
      "learning_rate": 5.219175546306784e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 504008,
      "step": 1835
    },
    {
      "epoch": 16.283185840707965,
      "grad_norm": 0.010155999101698399,
      "learning_rate": 5.1017030081533914e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 505272,
      "step": 1840
    },
    {
      "epoch": 16.327433628318584,
      "grad_norm": 0.0009144521318376064,
      "learning_rate": 4.985417199141443e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 506600,
      "step": 1845
    },
    {
      "epoch": 16.371681415929203,
      "grad_norm": 0.0014653526013717055,
      "learning_rate": 4.870325054525673e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 508088,
      "step": 1850
    },
    {
      "epoch": 16.41592920353982,
      "grad_norm": 0.00045606083585880697,
      "learning_rate": 4.7564334383709745e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 509576,
      "step": 1855
    },
    {
      "epoch": 16.460176991150444,
      "grad_norm": 0.00035226400359533727,
      "learning_rate": 4.6437491431430556e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 511176,
      "step": 1860
    },
    {
      "epoch": 16.504424778761063,
      "grad_norm": 0.0005832520546391606,
      "learning_rate": 4.5322788893033155e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 512264,
      "step": 1865
    },
    {
      "epoch": 16.548672566371682,
      "grad_norm": 0.0013271081261336803,
      "learning_rate": 4.422029324908061e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 513512,
      "step": 1870
    },
    {
      "epoch": 16.5929203539823,
      "grad_norm": 0.000609318376518786,
      "learning_rate": 4.313007025211985e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 515272,
      "step": 1875
    },
    {
      "epoch": 16.63716814159292,
      "grad_norm": 0.000544449663721025,
      "learning_rate": 4.205218492276055e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 516568,
      "step": 1880
    },
    {
      "epoch": 16.68141592920354,
      "grad_norm": 0.000308368937112391,
      "learning_rate": 4.098670154579715e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 517720,
      "step": 1885
    },
    {
      "epoch": 16.72566371681416,
      "grad_norm": 0.0004592226177919656,
      "learning_rate": 3.9933683666374986e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 518840,
      "step": 1890
    },
    {
      "epoch": 16.76991150442478,
      "grad_norm": 0.002832420403137803,
      "learning_rate": 3.889319408620021e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 520312,
      "step": 1895
    },
    {
      "epoch": 16.8141592920354,
      "grad_norm": 0.0008214914705604315,
      "learning_rate": 3.7865294859794926e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 521768,
      "step": 1900
    },
    {
      "epoch": 16.858407079646017,
      "grad_norm": 0.000486409873701632,
      "learning_rate": 3.68500472907955e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 523368,
      "step": 1905
    },
    {
      "epoch": 16.902654867256636,
      "grad_norm": 0.0006678001955151558,
      "learning_rate": 3.584751192829705e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 524680,
      "step": 1910
    },
    {
      "epoch": 16.946902654867255,
      "grad_norm": 0.008256644010543823,
      "learning_rate": 3.4857748563242006e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 525864,
      "step": 1915
    },
    {
      "epoch": 16.991150442477878,
      "grad_norm": 0.0005858740769326687,
      "learning_rate": 3.388081622485431e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 527304,
      "step": 1920
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.2716752290725708,
      "eval_runtime": 0.7215,
      "eval_samples_per_second": 34.65,
      "eval_steps_per_second": 18.018,
      "num_input_tokens_seen": 527400,
      "step": 1921
    },
    {
      "epoch": 17.035398230088497,
      "grad_norm": 0.0004789613885805011,
      "learning_rate": 3.2916773177118778e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 528344,
      "step": 1925
    },
    {
      "epoch": 17.079646017699115,
      "grad_norm": 0.0009302425314672291,
      "learning_rate": 3.1965676915306384e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 529608,
      "step": 1930
    },
    {
      "epoch": 17.123893805309734,
      "grad_norm": 0.001780318678356707,
      "learning_rate": 3.102758416254545e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 531032,
      "step": 1935
    },
    {
      "epoch": 17.168141592920353,
      "grad_norm": 0.0004699099517893046,
      "learning_rate": 3.010255086643818e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 532264,
      "step": 1940
    },
    {
      "epoch": 17.212389380530972,
      "grad_norm": 0.00032142718555405736,
      "learning_rate": 2.919063219572438e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 533816,
      "step": 1945
    },
    {
      "epoch": 17.256637168141594,
      "grad_norm": 0.00035525771090760827,
      "learning_rate": 2.829188253699111e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 535048,
      "step": 1950
    },
    {
      "epoch": 17.300884955752213,
      "grad_norm": 0.0004667598695959896,
      "learning_rate": 2.7406355491429086e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 536408,
      "step": 1955
    },
    {
      "epoch": 17.345132743362832,
      "grad_norm": 0.015801800414919853,
      "learning_rate": 2.653410387163574e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 537656,
      "step": 1960
    },
    {
      "epoch": 17.38938053097345,
      "grad_norm": 0.0002979618147946894,
      "learning_rate": 2.567517969846575e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 538968,
      "step": 1965
    },
    {
      "epoch": 17.43362831858407,
      "grad_norm": 0.0018873842200264335,
      "learning_rate": 2.482963419792844e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 540280,
      "step": 1970
    },
    {
      "epoch": 17.47787610619469,
      "grad_norm": 0.000713904039002955,
      "learning_rate": 2.399751779813264e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 541816,
      "step": 1975
    },
    {
      "epoch": 17.52212389380531,
      "grad_norm": 0.0015280984807759523,
      "learning_rate": 2.317888012627914e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 543096,
      "step": 1980
    },
    {
      "epoch": 17.56637168141593,
      "grad_norm": 0.00048026355216279626,
      "learning_rate": 2.2373770005700955e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 544424,
      "step": 1985
    },
    {
      "epoch": 17.61061946902655,
      "grad_norm": 0.0003930071252398193,
      "learning_rate": 2.1582235452951682e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 546088,
      "step": 1990
    },
    {
      "epoch": 17.654867256637168,
      "grad_norm": 0.0005723510985262692,
      "learning_rate": 2.0804323674941563e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 547416,
      "step": 1995
    },
    {
      "epoch": 17.699115044247787,
      "grad_norm": 0.0012349956668913364,
      "learning_rate": 2.0040081066122043e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 548648,
      "step": 2000
    },
    {
      "epoch": 17.743362831858406,
      "grad_norm": 0.0034373090602457523,
      "learning_rate": 1.9289553205719317e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 549944,
      "step": 2005
    },
    {
      "epoch": 17.787610619469028,
      "grad_norm": 0.0003913345281034708,
      "learning_rate": 1.8552784855015215e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 551176,
      "step": 2010
    },
    {
      "epoch": 17.831858407079647,
      "grad_norm": 0.0026745900977402925,
      "learning_rate": 1.7829819954678361e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 552680,
      "step": 2015
    },
    {
      "epoch": 17.876106194690266,
      "grad_norm": 0.0003091230464633554,
      "learning_rate": 1.7120701622143132e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 554536,
      "step": 2020
    },
    {
      "epoch": 17.920353982300885,
      "grad_norm": 0.0006508951773867011,
      "learning_rate": 1.6425472149038361e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 555800,
      "step": 2025
    },
    {
      "epoch": 17.964601769911503,
      "grad_norm": 0.0014680474996566772,
      "learning_rate": 1.5744172998664902e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 557400,
      "step": 2030
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.2720074951648712,
      "eval_runtime": 0.7208,
      "eval_samples_per_second": 34.682,
      "eval_steps_per_second": 18.035,
      "num_input_tokens_seen": 558656,
      "step": 2034
    },
    {
      "epoch": 18.008849557522122,
      "grad_norm": 0.0168607160449028,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 559008,
      "step": 2035
    },
    {
      "epoch": 18.053097345132745,
      "grad_norm": 0.00042135766125284135,
      "learning_rate": 1.4423527362888546e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 560208,
      "step": 2040
    },
    {
      "epoch": 18.097345132743364,
      "grad_norm": 0.0004008494142908603,
      "learning_rate": 1.3784259640440279e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 562208,
      "step": 2045
    },
    {
      "epoch": 18.141592920353983,
      "grad_norm": 0.00046455985284410417,
      "learning_rate": 1.3159079761934923e-06,
      "loss": 0.0001,
      "num_input_tokens_seen": 563680,
      "step": 2050
    },
    {
      "epoch": 18.1858407079646,
      "grad_norm": 0.0004232655337546021,
      "learning_rate": 1.2548025012934367e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 564960,
      "step": 2055
    },
    {
      "epoch": 18.23008849557522,
      "grad_norm": 0.00264491094276309,
      "learning_rate": 1.195113183658131e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 566512,
      "step": 2060
    },
    {
      "epoch": 18.27433628318584,
      "grad_norm": 0.0003135623992420733,
      "learning_rate": 1.1368435831426021e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 567808,
      "step": 2065
    },
    {
      "epoch": 18.31858407079646,
      "grad_norm": 0.00040393025847151875,
      "learning_rate": 1.0799971749303333e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 568992,
      "step": 2070
    },
    {
      "epoch": 18.36283185840708,
      "grad_norm": 0.0007846535299904644,
      "learning_rate": 1.0245773493259946e-06,
      "loss": 0.0,
      "num_input_tokens_seen": 570336,
      "step": 2075
    },
    {
      "epoch": 18.4070796460177,
      "grad_norm": 0.006305730901658535,
      "learning_rate": 9.705874115532532e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 571600,
      "step": 2080
    },
    {
      "epoch": 18.451327433628318,
      "grad_norm": 0.0009637536131776869,
      "learning_rate": 9.180305815576301e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 573024,
      "step": 2085
    },
    {
      "epoch": 18.495575221238937,
      "grad_norm": 0.00033186632208526134,
      "learning_rate": 8.669099938144992e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 574432,
      "step": 2090
    },
    {
      "epoch": 18.539823008849556,
      "grad_norm": 0.008162146434187889,
      "learning_rate": 8.172286971421167e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 575744,
      "step": 2095
    },
    {
      "epoch": 18.58407079646018,
      "grad_norm": 0.0004633248026948422,
      "learning_rate": 7.689896545198111e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 577488,
      "step": 2100
    },
    {
      "epoch": 18.628318584070797,
      "grad_norm": 0.0005262296181172132,
      "learning_rate": 7.221957429112469e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 578784,
      "step": 2105
    },
    {
      "epoch": 18.672566371681416,
      "grad_norm": 0.0003769147442653775,
      "learning_rate": 6.768497530928785e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 580064,
      "step": 2110
    },
    {
      "epoch": 18.716814159292035,
      "grad_norm": 0.0010795833077281713,
      "learning_rate": 6.329543894874779e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 581232,
      "step": 2115
    },
    {
      "epoch": 18.761061946902654,
      "grad_norm": 0.0012543437769636512,
      "learning_rate": 5.905122700028576e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 582320,
      "step": 2120
    },
    {
      "epoch": 18.805309734513273,
      "grad_norm": 0.0014014469925314188,
      "learning_rate": 5.49525925875738e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 583584,
      "step": 2125
    },
    {
      "epoch": 18.849557522123895,
      "grad_norm": 0.0025279626715928316,
      "learning_rate": 5.099978015207868e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 584864,
      "step": 2130
    },
    {
      "epoch": 18.893805309734514,
      "grad_norm": 0.0004394982533995062,
      "learning_rate": 4.719302543848225e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 586144,
      "step": 2135
    },
    {
      "epoch": 18.938053097345133,
      "grad_norm": 0.0008687974186614156,
      "learning_rate": 4.3532555480624295e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 587888,
      "step": 2140
    },
    {
      "epoch": 18.98230088495575,
      "grad_norm": 0.007426698226481676,
      "learning_rate": 4.001858858795893e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 589312,
      "step": 2145
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.27254992723464966,
      "eval_runtime": 0.7201,
      "eval_samples_per_second": 34.72,
      "eval_steps_per_second": 18.054,
      "num_input_tokens_seen": 589928,
      "step": 2147
    },
    {
      "epoch": 19.02654867256637,
      "grad_norm": 0.0005953509826213121,
      "learning_rate": 3.665133433253809e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 590712,
      "step": 2150
    },
    {
      "epoch": 19.07079646017699,
      "grad_norm": 0.0003601981734391302,
      "learning_rate": 3.34309935365093e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 591976,
      "step": 2155
    },
    {
      "epoch": 19.115044247787612,
      "grad_norm": 0.01488531194627285,
      "learning_rate": 3.03577582601422e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 593352,
      "step": 2160
    },
    {
      "epoch": 19.15929203539823,
      "grad_norm": 0.00042196488357149065,
      "learning_rate": 2.743181179037047e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 594792,
      "step": 2165
    },
    {
      "epoch": 19.20353982300885,
      "grad_norm": 0.00048617066931910813,
      "learning_rate": 2.465332862986447e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 595992,
      "step": 2170
    },
    {
      "epoch": 19.24778761061947,
      "grad_norm": 0.0003980123728979379,
      "learning_rate": 2.2022474486620427e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 597480,
      "step": 2175
    },
    {
      "epoch": 19.292035398230087,
      "grad_norm": 0.0006522561307065189,
      "learning_rate": 1.953940626408024e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 598984,
      "step": 2180
    },
    {
      "epoch": 19.336283185840706,
      "grad_norm": 0.0062956251204013824,
      "learning_rate": 1.720427205177233e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 600456,
      "step": 2185
    },
    {
      "epoch": 19.38053097345133,
      "grad_norm": 0.0003871411317959428,
      "learning_rate": 1.5017211116479802e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 601768,
      "step": 2190
    },
    {
      "epoch": 19.424778761061948,
      "grad_norm": 0.0006148723769001663,
      "learning_rate": 1.297835389393598e-07,
      "loss": 0.0001,
      "num_input_tokens_seen": 603032,
      "step": 2195
    },
    {
      "epoch": 19.469026548672566,
      "grad_norm": 0.0004471699066925794,
      "learning_rate": 1.1087821981042856e-07,
      "loss": 0.0,
      "num_input_tokens_seen": 604184,
      "step": 2200
    },
    {
      "epoch": 19.513274336283185,
      "grad_norm": 0.0004819620226044208,
      "learning_rate": 9.345728128621611e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 605848,
      "step": 2205
    },
    {
      "epoch": 19.557522123893804,
      "grad_norm": 0.001638778019696474,
      "learning_rate": 7.752176234685771e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 607128,
      "step": 2210
    },
    {
      "epoch": 19.601769911504427,
      "grad_norm": 0.00036322633968666196,
      "learning_rate": 6.307261338246718e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 608552,
      "step": 2215
    },
    {
      "epoch": 19.646017699115045,
      "grad_norm": 0.0004453824076335877,
      "learning_rate": 5.011069613644892e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 610152,
      "step": 2220
    },
    {
      "epoch": 19.690265486725664,
      "grad_norm": 0.0014128797920420766,
      "learning_rate": 3.8636783654100174e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 611448,
      "step": 2225
    },
    {
      "epoch": 19.734513274336283,
      "grad_norm": 0.0008530104532837868,
      "learning_rate": 2.865156023650617e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 612968,
      "step": 2230
    },
    {
      "epoch": 19.778761061946902,
      "grad_norm": 0.0013214597711339593,
      "learning_rate": 2.0155621399742254e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 614280,
      "step": 2235
    },
    {
      "epoch": 19.82300884955752,
      "grad_norm": 0.0017563271103426814,
      "learning_rate": 1.31494738393384e-08,
      "loss": 0.0,
      "num_input_tokens_seen": 616056,
      "step": 2240
    },
    {
      "epoch": 19.86725663716814,
      "grad_norm": 0.0018662940710783005,
      "learning_rate": 7.633535400070057e-09,
      "loss": 0.0,
      "num_input_tokens_seen": 617320,
      "step": 2245
    },
    {
      "epoch": 19.911504424778762,
      "grad_norm": 0.0032873074524104595,
      "learning_rate": 3.6081350510447365e-09,
      "loss": 0.0,
      "num_input_tokens_seen": 618440,
      "step": 2250
    },
    {
      "epoch": 19.95575221238938,
      "grad_norm": 0.0003584111691452563,
      "learning_rate": 1.0735128660649406e-09,
      "loss": 0.0001,
      "num_input_tokens_seen": 619544,
      "step": 2255
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0007038831245154142,
      "learning_rate": 2.982000932294504e-11,
      "loss": 0.0,
      "num_input_tokens_seen": 621040,
      "step": 2260
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.2709304988384247,
      "eval_runtime": 0.7228,
      "eval_samples_per_second": 34.586,
      "eval_steps_per_second": 17.985,
      "num_input_tokens_seen": 621040,
      "step": 2260
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 621040,
      "step": 2260,
      "total_flos": 2.796515050979328e+16,
      "train_loss": 0.21139920982813262,
      "train_runtime": 252.4417,
      "train_samples_per_second": 17.826,
      "train_steps_per_second": 8.953
    }
  ],
  "logging_steps": 5,
  "max_steps": 2260,
  "num_input_tokens_seen": 621040,
  "num_train_epochs": 20,
  "save_steps": 113,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.796515050979328e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}