{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 566,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0088339222614841,
      "grad_norm": 0.318359375,
      "learning_rate": 8.620689655172414e-06,
      "loss": 0.145,
      "step": 5
    },
    {
      "epoch": 0.0176678445229682,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 0.1351,
      "step": 10
    },
    {
      "epoch": 0.026501766784452298,
      "grad_norm": 0.1513671875,
      "learning_rate": 2.5862068965517244e-05,
      "loss": 0.1249,
      "step": 15
    },
    {
      "epoch": 0.0353356890459364,
      "grad_norm": 0.12109375,
      "learning_rate": 3.4482758620689657e-05,
      "loss": 0.1189,
      "step": 20
    },
    {
      "epoch": 0.044169611307420496,
      "grad_norm": 0.09619140625,
      "learning_rate": 4.3103448275862066e-05,
      "loss": 0.1181,
      "step": 25
    },
    {
      "epoch": 0.053003533568904596,
      "grad_norm": 0.0908203125,
      "learning_rate": 4.999961496300632e-05,
      "loss": 0.1127,
      "step": 30
    },
    {
      "epoch": 0.061837455830388695,
      "grad_norm": 0.0830078125,
      "learning_rate": 4.9986140051876094e-05,
      "loss": 0.1102,
      "step": 35
    },
    {
      "epoch": 0.0706713780918728,
      "grad_norm": 0.08447265625,
      "learning_rate": 4.995342646712217e-05,
      "loss": 0.1093,
      "step": 40
    },
    {
      "epoch": 0.07950530035335689,
      "grad_norm": 0.07958984375,
      "learning_rate": 4.9901502197807084e-05,
      "loss": 0.109,
      "step": 45
    },
    {
      "epoch": 0.08833922261484099,
      "grad_norm": 0.07958984375,
      "learning_rate": 4.9830411669255416e-05,
      "loss": 0.107,
      "step": 50
    },
    {
      "epoch": 0.09717314487632508,
      "grad_norm": 0.07763671875,
      "learning_rate": 4.974021570504443e-05,
      "loss": 0.1041,
      "step": 55
    },
    {
      "epoch": 0.10600706713780919,
      "grad_norm": 0.07421875,
      "learning_rate": 4.963099147496465e-05,
      "loss": 0.1029,
      "step": 60
    },
    {
      "epoch": 0.11484098939929328,
      "grad_norm": 0.0810546875,
      "learning_rate": 4.9502832428995005e-05,
      "loss": 0.1047,
      "step": 65
    },
    {
      "epoch": 0.12367491166077739,
      "grad_norm": 0.0771484375,
      "learning_rate": 4.935584821734901e-05,
      "loss": 0.1044,
      "step": 70
    },
    {
      "epoch": 0.13250883392226148,
      "grad_norm": 0.07568359375,
      "learning_rate": 4.919016459666026e-05,
      "loss": 0.1063,
      "step": 75
    },
    {
      "epoch": 0.1413427561837456,
      "grad_norm": 0.0830078125,
      "learning_rate": 4.9005923322387706e-05,
      "loss": 0.1025,
      "step": 80
    },
    {
      "epoch": 0.1501766784452297,
      "grad_norm": 0.0859375,
      "learning_rate": 4.880328202753264e-05,
      "loss": 0.1022,
      "step": 85
    },
    {
      "epoch": 0.15901060070671377,
      "grad_norm": 0.0791015625,
      "learning_rate": 4.858241408777117e-05,
      "loss": 0.1031,
      "step": 90
    },
    {
      "epoch": 0.16784452296819788,
      "grad_norm": 0.07421875,
      "learning_rate": 4.834350847311758e-05,
      "loss": 0.1016,
      "step": 95
    },
    {
      "epoch": 0.17667844522968199,
      "grad_norm": 0.076171875,
      "learning_rate": 4.8086769586245554e-05,
      "loss": 0.1008,
      "step": 100
    },
    {
      "epoch": 0.1855123674911661,
      "grad_norm": 0.0751953125,
      "learning_rate": 4.7812417087605456e-05,
      "loss": 0.1043,
      "step": 105
    },
    {
      "epoch": 0.19434628975265017,
      "grad_norm": 0.07275390625,
      "learning_rate": 4.752068570748746e-05,
      "loss": 0.1038,
      "step": 110
    },
    {
      "epoch": 0.20318021201413428,
      "grad_norm": 0.078125,
      "learning_rate": 4.721182504519118e-05,
      "loss": 0.1014,
      "step": 115
    },
    {
      "epoch": 0.21201413427561838,
      "grad_norm": 0.0712890625,
      "learning_rate": 4.688609935547371e-05,
      "loss": 0.1004,
      "step": 120
    },
    {
      "epoch": 0.22084805653710246,
      "grad_norm": 0.07177734375,
      "learning_rate": 4.654378732245869e-05,
      "loss": 0.1017,
      "step": 125
    },
    {
      "epoch": 0.22968197879858657,
      "grad_norm": 0.07421875,
      "learning_rate": 4.618518182120011e-05,
      "loss": 0.1006,
      "step": 130
    },
    {
      "epoch": 0.23851590106007067,
      "grad_norm": 0.0751953125,
      "learning_rate": 4.5810589667104347e-05,
      "loss": 0.1008,
      "step": 135
    },
    {
      "epoch": 0.24734982332155478,
      "grad_norm": 0.08447265625,
      "learning_rate": 4.542033135342537e-05,
      "loss": 0.1012,
      "step": 140
    },
    {
      "epoch": 0.25618374558303886,
      "grad_norm": 0.07275390625,
      "learning_rate": 4.5014740777057405e-05,
      "loss": 0.1026,
      "step": 145
    },
    {
      "epoch": 0.26501766784452296,
      "grad_norm": 0.07861328125,
      "learning_rate": 4.45941649528596e-05,
      "loss": 0.1017,
      "step": 150
    },
    {
      "epoch": 0.27385159010600707,
      "grad_norm": 0.0771484375,
      "learning_rate": 4.4158963716757444e-05,
      "loss": 0.099,
      "step": 155
    },
    {
      "epoch": 0.2826855123674912,
      "grad_norm": 0.07666015625,
      "learning_rate": 4.370950941787456e-05,
      "loss": 0.1021,
      "step": 160
    },
    {
      "epoch": 0.2915194346289753,
      "grad_norm": 0.0712890625,
      "learning_rate": 4.324618659995855e-05,
      "loss": 0.1006,
      "step": 165
    },
    {
      "epoch": 0.3003533568904594,
      "grad_norm": 0.076171875,
      "learning_rate": 4.27693916723734e-05,
      "loss": 0.1006,
      "step": 170
    },
    {
      "epoch": 0.30918727915194344,
      "grad_norm": 0.07275390625,
      "learning_rate": 4.227953257093985e-05,
      "loss": 0.0995,
      "step": 175
    },
    {
      "epoch": 0.31802120141342755,
      "grad_norm": 0.07763671875,
      "learning_rate": 4.1777028408913985e-05,
      "loss": 0.1005,
      "step": 180
    },
    {
      "epoch": 0.32685512367491165,
      "grad_norm": 0.07373046875,
      "learning_rate": 4.126230911840269e-05,
      "loss": 0.0995,
      "step": 185
    },
    {
      "epoch": 0.33568904593639576,
      "grad_norm": 0.07666015625,
      "learning_rate": 4.07358150825226e-05,
      "loss": 0.0984,
      "step": 190
    },
    {
      "epoch": 0.34452296819787986,
      "grad_norm": 0.07568359375,
      "learning_rate": 4.0197996758617594e-05,
      "loss": 0.0979,
      "step": 195
    },
    {
      "epoch": 0.35335689045936397,
      "grad_norm": 0.076171875,
      "learning_rate": 3.964931429285675e-05,
      "loss": 0.0998,
      "step": 200
    },
    {
      "epoch": 0.3621908127208481,
      "grad_norm": 0.07958984375,
      "learning_rate": 3.909023712654291e-05,
      "loss": 0.1012,
      "step": 205
    },
    {
      "epoch": 0.3710247349823322,
      "grad_norm": 0.07373046875,
      "learning_rate": 3.852124359446845e-05,
      "loss": 0.0987,
      "step": 210
    },
    {
      "epoch": 0.37985865724381623,
      "grad_norm": 0.0732421875,
      "learning_rate": 3.794282051566199e-05,
      "loss": 0.0982,
      "step": 215
    },
    {
      "epoch": 0.38869257950530034,
      "grad_norm": 0.0751953125,
      "learning_rate": 3.7355462776876184e-05,
      "loss": 0.0984,
      "step": 220
    },
    {
      "epoch": 0.39752650176678445,
      "grad_norm": 0.06982421875,
      "learning_rate": 3.6759672909172846e-05,
      "loss": 0.0973,
      "step": 225
    },
    {
      "epoch": 0.40636042402826855,
      "grad_norm": 0.07080078125,
      "learning_rate": 3.615596065796791e-05,
      "loss": 0.1007,
      "step": 230
    },
    {
      "epoch": 0.41519434628975266,
      "grad_norm": 0.07421875,
      "learning_rate": 3.554484254690379e-05,
      "loss": 0.0972,
      "step": 235
    },
    {
      "epoch": 0.42402826855123676,
      "grad_norm": 0.07080078125,
      "learning_rate": 3.492684143592252e-05,
      "loss": 0.0974,
      "step": 240
    },
    {
      "epoch": 0.43286219081272087,
      "grad_norm": 0.076171875,
      "learning_rate": 3.4302486073917686e-05,
      "loss": 0.0991,
      "step": 245
    },
    {
      "epoch": 0.4416961130742049,
      "grad_norm": 0.076171875,
      "learning_rate": 3.3672310646347844e-05,
      "loss": 0.0979,
      "step": 250
    },
    {
      "epoch": 0.450530035335689,
      "grad_norm": 0.07275390625,
      "learning_rate": 3.3036854318198575e-05,
      "loss": 0.0987,
      "step": 255
    },
    {
      "epoch": 0.45936395759717313,
      "grad_norm": 0.07470703125,
      "learning_rate": 3.2396660772684114e-05,
      "loss": 0.0999,
      "step": 260
    },
    {
      "epoch": 0.46819787985865724,
      "grad_norm": 0.07373046875,
      "learning_rate": 3.1752277746083325e-05,
      "loss": 0.0979,
      "step": 265
    },
    {
      "epoch": 0.47703180212014135,
      "grad_norm": 0.07373046875,
      "learning_rate": 3.110425655910795e-05,
      "loss": 0.0983,
      "step": 270
    },
    {
      "epoch": 0.48586572438162545,
      "grad_norm": 0.07080078125,
      "learning_rate": 3.045315164520405e-05,
      "loss": 0.0981,
      "step": 275
    },
    {
      "epoch": 0.49469964664310956,
      "grad_norm": 0.07568359375,
      "learning_rate": 2.9799520076190268e-05,
      "loss": 0.0987,
      "step": 280
    },
    {
      "epoch": 0.5035335689045937,
      "grad_norm": 0.07080078125,
      "learning_rate": 2.914392108563883e-05,
      "loss": 0.0963,
      "step": 285
    },
    {
      "epoch": 0.5123674911660777,
      "grad_norm": 0.07470703125,
      "learning_rate": 2.848691559040687e-05,
      "loss": 0.0977,
      "step": 290
    },
    {
      "epoch": 0.5212014134275619,
      "grad_norm": 0.072265625,
      "learning_rate": 2.7829065710727682e-05,
      "loss": 0.0959,
      "step": 295
    },
    {
      "epoch": 0.5300353356890459,
      "grad_norm": 0.07275390625,
      "learning_rate": 2.7170934289272327e-05,
      "loss": 0.0983,
      "step": 300
    },
    {
      "epoch": 0.5388692579505301,
      "grad_norm": 0.07177734375,
      "learning_rate": 2.6513084409593137e-05,
      "loss": 0.0981,
      "step": 305
    },
    {
      "epoch": 0.5477031802120141,
      "grad_norm": 0.07373046875,
      "learning_rate": 2.585607891436118e-05,
      "loss": 0.0972,
      "step": 310
    },
    {
      "epoch": 0.5565371024734982,
      "grad_norm": 0.078125,
      "learning_rate": 2.5200479923809738e-05,
      "loss": 0.0964,
      "step": 315
    },
    {
      "epoch": 0.5653710247349824,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.4546848354795954e-05,
      "loss": 0.0966,
      "step": 320
    },
    {
      "epoch": 0.5742049469964664,
      "grad_norm": 0.068359375,
      "learning_rate": 2.3895743440892053e-05,
      "loss": 0.0983,
      "step": 325
    },
    {
      "epoch": 0.5830388692579506,
      "grad_norm": 0.0712890625,
      "learning_rate": 2.3247722253916677e-05,
      "loss": 0.0983,
      "step": 330
    },
    {
      "epoch": 0.5918727915194346,
      "grad_norm": 0.07177734375,
      "learning_rate": 2.2603339227315902e-05,
      "loss": 0.0982,
      "step": 335
    },
    {
      "epoch": 0.6007067137809188,
      "grad_norm": 0.06787109375,
      "learning_rate": 2.1963145681801434e-05,
      "loss": 0.0968,
      "step": 340
    },
    {
      "epoch": 0.6095406360424028,
      "grad_norm": 0.0712890625,
      "learning_rate": 2.132768935365215e-05,
      "loss": 0.0976,
      "step": 345
    },
    {
      "epoch": 0.6183745583038869,
      "grad_norm": 0.0693359375,
      "learning_rate": 2.069751392608232e-05,
      "loss": 0.0974,
      "step": 350
    },
    {
      "epoch": 0.627208480565371,
      "grad_norm": 0.0712890625,
      "learning_rate": 2.0073158564077483e-05,
      "loss": 0.0992,
      "step": 355
    },
    {
      "epoch": 0.6360424028268551,
      "grad_norm": 0.0703125,
      "learning_rate": 1.9455157453096225e-05,
      "loss": 0.0992,
      "step": 360
    },
    {
      "epoch": 0.6448763250883393,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.8844039342032095e-05,
      "loss": 0.0961,
      "step": 365
    },
    {
      "epoch": 0.6537102473498233,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.8240327090827153e-05,
      "loss": 0.097,
      "step": 370
    },
    {
      "epoch": 0.6625441696113075,
      "grad_norm": 0.068359375,
      "learning_rate": 1.764453722312383e-05,
      "loss": 0.0979,
      "step": 375
    },
    {
      "epoch": 0.6713780918727915,
      "grad_norm": 0.0693359375,
      "learning_rate": 1.705717948433801e-05,
      "loss": 0.0963,
      "step": 380
    },
    {
      "epoch": 0.6802120141342756,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.6478756405531564e-05,
      "loss": 0.0969,
      "step": 385
    },
    {
      "epoch": 0.6890459363957597,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.5909762873457096e-05,
      "loss": 0.0963,
      "step": 390
    },
    {
      "epoch": 0.6978798586572438,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.5350685707143258e-05,
      "loss": 0.0973,
      "step": 395
    },
    {
      "epoch": 0.7067137809187279,
      "grad_norm": 0.06591796875,
      "learning_rate": 1.4802003241382406e-05,
      "loss": 0.0963,
      "step": 400
    },
    {
      "epoch": 0.715547703180212,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.4264184917477397e-05,
      "loss": 0.0964,
      "step": 405
    },
    {
      "epoch": 0.7243816254416962,
      "grad_norm": 0.0673828125,
      "learning_rate": 1.3737690881597321e-05,
      "loss": 0.0981,
      "step": 410
    },
    {
      "epoch": 0.7332155477031802,
      "grad_norm": 0.068359375,
      "learning_rate": 1.3222971591086014e-05,
      "loss": 0.0977,
      "step": 415
    },
    {
      "epoch": 0.7420494699646644,
      "grad_norm": 0.06689453125,
      "learning_rate": 1.2720467429060156e-05,
      "loss": 0.0975,
      "step": 420
    },
    {
      "epoch": 0.7508833922261484,
      "grad_norm": 0.06787109375,
      "learning_rate": 1.2230608327626608e-05,
      "loss": 0.0978,
      "step": 425
    },
    {
      "epoch": 0.7597173144876325,
      "grad_norm": 0.0712890625,
      "learning_rate": 1.1753813400041453e-05,
      "loss": 0.0954,
      "step": 430
    },
    {
      "epoch": 0.7685512367491166,
      "grad_norm": 0.06640625,
      "learning_rate": 1.1290490582125454e-05,
      "loss": 0.0952,
      "step": 435
    },
    {
      "epoch": 0.7773851590106007,
      "grad_norm": 0.068359375,
      "learning_rate": 1.0841036283242558e-05,
      "loss": 0.0971,
      "step": 440
    },
    {
      "epoch": 0.7862190812720848,
      "grad_norm": 0.068359375,
      "learning_rate": 1.0405835047140401e-05,
      "loss": 0.0982,
      "step": 445
    },
    {
      "epoch": 0.7950530035335689,
      "grad_norm": 0.06689453125,
      "learning_rate": 9.985259222942602e-06,
      "loss": 0.0952,
      "step": 450
    },
    {
      "epoch": 0.803886925795053,
      "grad_norm": 0.0703125,
      "learning_rate": 9.57966864657463e-06,
      "loss": 0.0978,
      "step": 455
    },
    {
      "epoch": 0.8127208480565371,
      "grad_norm": 0.06982421875,
      "learning_rate": 9.189410332895662e-06,
      "loss": 0.0989,
      "step": 460
    },
    {
      "epoch": 0.8215547703180212,
      "grad_norm": 0.06787109375,
      "learning_rate": 8.814818178799892e-06,
      "loss": 0.0981,
      "step": 465
    },
    {
      "epoch": 0.8303886925795053,
      "grad_norm": 0.06591796875,
      "learning_rate": 8.456212677541312e-06,
      "loss": 0.0945,
      "step": 470
    },
    {
      "epoch": 0.8392226148409894,
      "grad_norm": 0.0673828125,
      "learning_rate": 8.113900644526301e-06,
      "loss": 0.0988,
      "step": 475
    },
    {
      "epoch": 0.8480565371024735,
      "grad_norm": 0.06689453125,
      "learning_rate": 7.788174954808826e-06,
      "loss": 0.0973,
      "step": 480
    },
    {
      "epoch": 0.8568904593639576,
      "grad_norm": 0.06689453125,
      "learning_rate": 7.479314292512542e-06,
      "loss": 0.0972,
      "step": 485
    },
    {
      "epoch": 0.8657243816254417,
      "grad_norm": 0.06689453125,
      "learning_rate": 7.187582912394548e-06,
      "loss": 0.0977,
      "step": 490
    },
    {
      "epoch": 0.8745583038869258,
      "grad_norm": 0.06591796875,
      "learning_rate": 6.913230413754452e-06,
      "loss": 0.0962,
      "step": 495
    },
    {
      "epoch": 0.8833922261484098,
      "grad_norm": 0.06982421875,
      "learning_rate": 6.656491526882422e-06,
      "loss": 0.0958,
      "step": 500
    },
    {
      "epoch": 0.892226148409894,
      "grad_norm": 0.06494140625,
      "learning_rate": 6.417585912228833e-06,
      "loss": 0.0959,
      "step": 505
    },
    {
      "epoch": 0.901060070671378,
      "grad_norm": 0.0703125,
      "learning_rate": 6.196717972467361e-06,
      "loss": 0.0978,
      "step": 510
    },
    {
      "epoch": 0.9098939929328622,
      "grad_norm": 0.0673828125,
      "learning_rate": 5.994076677612297e-06,
      "loss": 0.0982,
      "step": 515
    },
    {
      "epoch": 0.9187279151943463,
      "grad_norm": 0.06640625,
      "learning_rate": 5.809835403339747e-06,
      "loss": 0.0971,
      "step": 520
    },
    {
      "epoch": 0.9275618374558304,
      "grad_norm": 0.06787109375,
      "learning_rate": 5.644151782650993e-06,
      "loss": 0.0953,
      "step": 525
    },
    {
      "epoch": 0.9363957597173145,
      "grad_norm": 0.06982421875,
      "learning_rate": 5.497167571004998e-06,
      "loss": 0.0956,
      "step": 530
    },
    {
      "epoch": 0.9452296819787986,
      "grad_norm": 0.06689453125,
      "learning_rate": 5.36900852503536e-06,
      "loss": 0.0963,
      "step": 535
    },
    {
      "epoch": 0.9540636042402827,
      "grad_norm": 0.06494140625,
      "learning_rate": 5.259784294955576e-06,
      "loss": 0.0985,
      "step": 540
    },
    {
      "epoch": 0.9628975265017667,
      "grad_norm": 0.06884765625,
      "learning_rate": 5.169588330744585e-06,
      "loss": 0.0982,
      "step": 545
    },
    {
      "epoch": 0.9717314487632509,
      "grad_norm": 0.068359375,
      "learning_rate": 5.098497802192923e-06,
      "loss": 0.0979,
      "step": 550
    },
    {
      "epoch": 0.980565371024735,
      "grad_norm": 0.06591796875,
      "learning_rate": 5.046573532877835e-06,
      "loss": 0.0977,
      "step": 555
    },
    {
      "epoch": 0.9893992932862191,
      "grad_norm": 0.06689453125,
      "learning_rate": 5.013859948123909e-06,
      "loss": 0.0976,
      "step": 560
    },
    {
      "epoch": 0.9982332155477032,
      "grad_norm": 0.06591796875,
      "learning_rate": 5.000385036993684e-06,
      "loss": 0.0977,
      "step": 565
    },
    {
      "epoch": 1.0,
      "step": 566,
      "total_flos": 5.828408442271826e+17,
      "train_loss": 0.10050818478102819,
      "train_runtime": 2227.4528,
      "train_samples_per_second": 32.502,
      "train_steps_per_second": 0.254
    }
  ],
  "logging_steps": 5,
  "max_steps": 566,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.828408442271826e+17,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}