{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 625,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.0092181354761123,
      "epoch": 0.016,
      "grad_norm": 1.508150339126587,
      "learning_rate": 4.736842105263158e-05,
      "loss": 2.2481,
      "mean_token_accuracy": 0.5437881387770176,
      "num_tokens": 10218.0,
      "step": 10
    },
    {
      "entropy": 2.0736056357622146,
      "epoch": 0.032,
      "grad_norm": 1.2320566177368164,
      "learning_rate": 0.0001,
      "loss": 2.0578,
      "mean_token_accuracy": 0.5636145301163197,
      "num_tokens": 20480.0,
      "step": 20
    },
    {
      "entropy": 1.958076971769333,
      "epoch": 0.048,
      "grad_norm": 0.867118239402771,
      "learning_rate": 0.00015263157894736845,
      "loss": 1.9242,
      "mean_token_accuracy": 0.5743255637586117,
      "num_tokens": 30543.0,
      "step": 30
    },
    {
      "entropy": 1.9567744851112365,
      "epoch": 0.064,
      "grad_norm": 0.7917498350143433,
      "learning_rate": 0.00019999966405802826,
      "loss": 1.9526,
      "mean_token_accuracy": 0.5756677143275738,
      "num_tokens": 40658.0,
      "step": 40
    },
    {
      "entropy": 1.8826423227787017,
      "epoch": 0.08,
      "grad_norm": 0.8305525183677673,
      "learning_rate": 0.00019995935375248606,
      "loss": 1.8509,
      "mean_token_accuracy": 0.5996200643479824,
      "num_tokens": 50792.0,
      "step": 50
    },
    {
      "entropy": 1.9008120864629745,
      "epoch": 0.096,
      "grad_norm": 0.9713864326477051,
      "learning_rate": 0.000199851886084842,
      "loss": 1.9109,
      "mean_token_accuracy": 0.5911510705947876,
      "num_tokens": 60899.0,
      "step": 60
    },
    {
      "entropy": 1.880544126033783,
      "epoch": 0.112,
      "grad_norm": 0.9007127285003662,
      "learning_rate": 0.0001996773332568941,
      "loss": 1.8743,
      "mean_token_accuracy": 0.5929530031979084,
      "num_tokens": 71221.0,
      "step": 70
    },
    {
      "entropy": 1.952101269364357,
      "epoch": 0.128,
      "grad_norm": 0.78743577003479,
      "learning_rate": 0.0001994358125413841,
      "loss": 1.9528,
      "mean_token_accuracy": 0.5812688320875168,
      "num_tokens": 81172.0,
      "step": 80
    },
    {
      "entropy": 1.9294454216957093,
      "epoch": 0.144,
      "grad_norm": 0.7737237215042114,
      "learning_rate": 0.00019912748620320794,
      "loss": 1.9409,
      "mean_token_accuracy": 0.5753811992704868,
      "num_tokens": 91264.0,
      "step": 90
    },
    {
      "entropy": 1.9249590784311295,
      "epoch": 0.16,
      "grad_norm": 0.8270857334136963,
      "learning_rate": 0.00019875256139039902,
      "loss": 1.9009,
      "mean_token_accuracy": 0.5813618630170823,
      "num_tokens": 101877.0,
      "step": 100
    },
    {
      "epoch": 0.16,
      "eval_entropy": 1.8299660499279315,
      "eval_loss": 1.7842161655426025,
      "eval_mean_token_accuracy": 0.6030518962786748,
      "eval_num_tokens": 101877.0,
      "eval_runtime": 0.8408,
      "eval_samples_per_second": 118.935,
      "eval_steps_per_second": 15.462,
      "step": 100
    },
    {
      "entropy": 1.8542533993721009,
      "epoch": 0.176,
      "grad_norm": 0.8512797355651855,
      "learning_rate": 0.00019831128999495606,
      "loss": 1.8219,
      "mean_token_accuracy": 0.5989311248064041,
      "num_tokens": 112259.0,
      "step": 110
    },
    {
      "entropy": 1.7832964926958084,
      "epoch": 0.192,
      "grad_norm": 0.7618836760520935,
      "learning_rate": 0.0001978039684836106,
      "loss": 1.7925,
      "mean_token_accuracy": 0.6045975714921952,
      "num_tokens": 122294.0,
      "step": 120
    },
    {
      "entropy": 1.880309948325157,
      "epoch": 0.208,
      "grad_norm": 0.7805933952331543,
      "learning_rate": 0.00019723093769864663,
      "loss": 1.8632,
      "mean_token_accuracy": 0.5855595611035824,
      "num_tokens": 132251.0,
      "step": 130
    },
    {
      "entropy": 1.8512971937656402,
      "epoch": 0.224,
      "grad_norm": 0.926387369632721,
      "learning_rate": 0.00019659258262890683,
      "loss": 1.8367,
      "mean_token_accuracy": 0.5898601979017257,
      "num_tokens": 142542.0,
      "step": 140
    },
    {
      "entropy": 1.9222039520740508,
      "epoch": 0.24,
      "grad_norm": 0.8940110802650452,
      "learning_rate": 0.00019588933215113926,
      "loss": 1.9201,
      "mean_token_accuracy": 0.5759184911847115,
      "num_tokens": 152806.0,
      "step": 150
    },
    {
      "entropy": 1.9089818447828293,
      "epoch": 0.256,
      "grad_norm": 0.7893499732017517,
      "learning_rate": 0.00019512165874185767,
      "loss": 1.8895,
      "mean_token_accuracy": 0.5838948510587215,
      "num_tokens": 162706.0,
      "step": 160
    },
    {
      "entropy": 1.8353556275367737,
      "epoch": 0.272,
      "grad_norm": 1.2676271200180054,
      "learning_rate": 0.00019429007815990993,
      "loss": 1.7889,
      "mean_token_accuracy": 0.5981082618236542,
      "num_tokens": 173501.0,
      "step": 170
    },
    {
      "entropy": 1.823873969912529,
      "epoch": 0.288,
      "grad_norm": 0.7941910028457642,
      "learning_rate": 0.00019339514909996706,
      "loss": 1.8137,
      "mean_token_accuracy": 0.597452136874199,
      "num_tokens": 183521.0,
      "step": 180
    },
    {
      "entropy": 1.7840645104646682,
      "epoch": 0.304,
      "grad_norm": 0.8442012071609497,
      "learning_rate": 0.000192437472817166,
      "loss": 1.8134,
      "mean_token_accuracy": 0.6012439027428627,
      "num_tokens": 193753.0,
      "step": 190
    },
    {
      "entropy": 1.8204753398895264,
      "epoch": 0.32,
      "grad_norm": 0.636513888835907,
      "learning_rate": 0.00019141769272315858,
      "loss": 1.7482,
      "mean_token_accuracy": 0.6097412109375,
      "num_tokens": 204430.0,
      "step": 200
    },
    {
      "epoch": 0.32,
      "eval_entropy": 1.7530423127687895,
      "eval_loss": 1.7683205604553223,
      "eval_mean_token_accuracy": 0.60484268115117,
      "eval_num_tokens": 204430.0,
      "eval_runtime": 0.5699,
      "eval_samples_per_second": 175.459,
      "eval_steps_per_second": 22.81,
      "step": 200
    },
    {
      "entropy": 1.8239326030015945,
      "epoch": 0.336,
      "grad_norm": 0.8120871782302856,
      "learning_rate": 0.00019033649395383702,
      "loss": 1.8557,
      "mean_token_accuracy": 0.5853650733828545,
      "num_tokens": 214577.0,
      "step": 210
    },
    {
      "entropy": 1.8361420631408691,
      "epoch": 0.352,
      "grad_norm": 0.9607922434806824,
      "learning_rate": 0.00018919460290902826,
      "loss": 1.8244,
      "mean_token_accuracy": 0.5930058181285858,
      "num_tokens": 224970.0,
      "step": 220
    },
    {
      "entropy": 1.922423180937767,
      "epoch": 0.368,
      "grad_norm": 0.9304735064506531,
      "learning_rate": 0.00018799278676446423,
      "loss": 1.8917,
      "mean_token_accuracy": 0.5863452732563019,
      "num_tokens": 234866.0,
      "step": 230
    },
    {
      "entropy": 1.7584299921989441,
      "epoch": 0.384,
      "grad_norm": 0.8348444104194641,
      "learning_rate": 0.0001867318529563574,
      "loss": 1.7554,
      "mean_token_accuracy": 0.6109944336116314,
      "num_tokens": 244776.0,
      "step": 240
    },
    {
      "entropy": 1.816238895058632,
      "epoch": 0.4,
      "grad_norm": 0.6481600999832153,
      "learning_rate": 0.00018541264863892754,
      "loss": 1.8336,
      "mean_token_accuracy": 0.5937890991568565,
      "num_tokens": 255035.0,
      "step": 250
    },
    {
      "entropy": 1.8480711460113526,
      "epoch": 0.416,
      "grad_norm": 0.781286895275116,
      "learning_rate": 0.000184036060115244,
      "loss": 1.8261,
      "mean_token_accuracy": 0.5947530373930932,
      "num_tokens": 265072.0,
      "step": 260
    },
    {
      "entropy": 1.8989493697881699,
      "epoch": 0.432,
      "grad_norm": 0.8746325969696045,
      "learning_rate": 0.00018260301224176558,
      "loss": 1.8927,
      "mean_token_accuracy": 0.5771610237658024,
      "num_tokens": 275332.0,
      "step": 270
    },
    {
      "entropy": 1.859290638566017,
      "epoch": 0.448,
      "grad_norm": 0.7687608599662781,
      "learning_rate": 0.00018111446780697929,
      "loss": 1.8285,
      "mean_token_accuracy": 0.5923414275050163,
      "num_tokens": 285283.0,
      "step": 280
    },
    {
      "entropy": 1.7830641388893127,
      "epoch": 0.464,
      "grad_norm": 0.8059773445129395,
      "learning_rate": 0.00017957142688455362,
      "loss": 1.7907,
      "mean_token_accuracy": 0.5985377252101898,
      "num_tokens": 295429.0,
      "step": 290
    },
    {
      "entropy": 1.7804506599903107,
      "epoch": 0.48,
      "grad_norm": 0.9017696380615234,
      "learning_rate": 0.00017797492616144256,
      "loss": 1.7849,
      "mean_token_accuracy": 0.6041601061820984,
      "num_tokens": 305778.0,
      "step": 300
    },
    {
      "epoch": 0.48,
      "eval_entropy": 1.7502480378517737,
      "eval_loss": 1.7552363872528076,
      "eval_mean_token_accuracy": 0.6089264658781198,
      "eval_num_tokens": 305778.0,
      "eval_runtime": 0.5719,
      "eval_samples_per_second": 174.868,
      "eval_steps_per_second": 22.733,
      "step": 300
    },
    {
      "entropy": 1.70162413418293,
      "epoch": 0.496,
      "grad_norm": 0.6979867219924927,
      "learning_rate": 0.00017632603824139085,
      "loss": 1.7008,
      "mean_token_accuracy": 0.6215453043580055,
      "num_tokens": 316331.0,
      "step": 310
    },
    {
      "entropy": 1.815586867928505,
      "epoch": 0.512,
      "grad_norm": 0.8245130777359009,
      "learning_rate": 0.00017462587092430875,
      "loss": 1.8283,
      "mean_token_accuracy": 0.6016101613640785,
      "num_tokens": 326634.0,
      "step": 320
    },
    {
      "entropy": 1.9167057573795319,
      "epoch": 0.528,
      "grad_norm": 0.8274908065795898,
      "learning_rate": 0.00017287556646200018,
      "loss": 1.885,
      "mean_token_accuracy": 0.5891454972326755,
      "num_tokens": 336975.0,
      "step": 330
    },
    {
      "entropy": 1.806039023399353,
      "epoch": 0.544,
      "grad_norm": 1.0455838441848755,
      "learning_rate": 0.00017107630079074478,
      "loss": 1.7743,
      "mean_token_accuracy": 0.5969359435141086,
      "num_tokens": 347323.0,
      "step": 340
    },
    {
      "entropy": 1.8411190688610077,
      "epoch": 0.56,
      "grad_norm": 1.0417566299438477,
      "learning_rate": 0.00016922928274124886,
      "loss": 1.861,
      "mean_token_accuracy": 0.5893909811973572,
      "num_tokens": 357656.0,
      "step": 350
    },
    {
      "entropy": 1.868698662519455,
      "epoch": 0.576,
      "grad_norm": 0.7351287603378296,
      "learning_rate": 0.00016733575322649657,
      "loss": 1.8244,
      "mean_token_accuracy": 0.5993935190141201,
      "num_tokens": 367692.0,
      "step": 360
    },
    {
      "entropy": 1.8023948460817336,
      "epoch": 0.592,
      "grad_norm": 0.8541484475135803,
      "learning_rate": 0.00016539698440804661,
      "loss": 1.818,
      "mean_token_accuracy": 0.5955550998449326,
      "num_tokens": 377994.0,
      "step": 370
    },
    {
      "entropy": 1.9064868450164796,
      "epoch": 0.608,
      "grad_norm": 0.8737623691558838,
      "learning_rate": 0.0001634142788413346,
      "loss": 1.8982,
      "mean_token_accuracy": 0.5769180566072464,
      "num_tokens": 387844.0,
      "step": 380
    },
    {
      "entropy": 1.848566684126854,
      "epoch": 0.624,
      "grad_norm": 0.7124635577201843,
      "learning_rate": 0.00016138896860055555,
      "loss": 1.8218,
      "mean_token_accuracy": 0.5997274614870548,
      "num_tokens": 398009.0,
      "step": 390
    },
    {
      "entropy": 1.7892296761274338,
      "epoch": 0.64,
      "grad_norm": 1.0577216148376465,
      "learning_rate": 0.0001593224143837142,
      "loss": 1.8346,
      "mean_token_accuracy": 0.5980690255761146,
      "num_tokens": 408155.0,
      "step": 400
    },
    {
      "epoch": 0.64,
      "eval_entropy": 1.7474457025527954,
      "eval_loss": 1.7562793493270874,
      "eval_mean_token_accuracy": 0.6087833138612601,
      "eval_num_tokens": 408155.0,
      "eval_runtime": 0.556,
      "eval_samples_per_second": 179.852,
      "eval_steps_per_second": 23.381,
      "step": 400
    },
    {
      "entropy": 1.765562105178833,
      "epoch": 0.656,
      "grad_norm": 0.8375977873802185,
      "learning_rate": 0.00015721600459844468,
      "loss": 1.7165,
      "mean_token_accuracy": 0.6151046514511108,
      "num_tokens": 418669.0,
      "step": 410
    },
    {
      "entropy": 1.8461785107851028,
      "epoch": 0.672,
      "grad_norm": 0.8062698841094971,
      "learning_rate": 0.0001550711544292131,
      "loss": 1.8239,
      "mean_token_accuracy": 0.5912051677703858,
      "num_tokens": 428586.0,
      "step": 420
    },
    {
      "entropy": 1.724032711982727,
      "epoch": 0.688,
      "grad_norm": 0.9207695126533508,
      "learning_rate": 0.00015288930488653094,
      "loss": 1.7222,
      "mean_token_accuracy": 0.6095151364803314,
      "num_tokens": 438852.0,
      "step": 430
    },
    {
      "entropy": 1.8236116498708725,
      "epoch": 0.704,
      "grad_norm": 0.8827204704284668,
      "learning_rate": 0.00015067192183881658,
      "loss": 1.8165,
      "mean_token_accuracy": 0.5982112549245358,
      "num_tokens": 448800.0,
      "step": 440
    },
    {
      "entropy": 1.8432465463876724,
      "epoch": 0.72,
      "grad_norm": 0.8164684772491455,
      "learning_rate": 0.0001484204950275565,
      "loss": 1.822,
      "mean_token_accuracy": 0.6006790846586227,
      "num_tokens": 459455.0,
      "step": 450
    },
    {
      "entropy": 1.819486153125763,
      "epoch": 0.736,
      "grad_norm": 0.9052107930183411,
      "learning_rate": 0.0001461365370664276,
      "loss": 1.8213,
      "mean_token_accuracy": 0.594074235111475,
      "num_tokens": 470403.0,
      "step": 460
    },
    {
      "entropy": 1.825221198797226,
      "epoch": 0.752,
      "grad_norm": 0.9059354662895203,
      "learning_rate": 0.00014382158242505234,
      "loss": 1.8018,
      "mean_token_accuracy": 0.5954498581588268,
      "num_tokens": 480402.0,
      "step": 470
    },
    {
      "entropy": 1.8063141599297523,
      "epoch": 0.768,
      "grad_norm": 0.8374266028404236,
      "learning_rate": 0.0001414771863980707,
      "loss": 1.7355,
      "mean_token_accuracy": 0.6036714628338814,
      "num_tokens": 490673.0,
      "step": 480
    },
    {
      "entropy": 1.7501532822847365,
      "epoch": 0.784,
      "grad_norm": 0.880260169506073,
      "learning_rate": 0.00013910492406022033,
      "loss": 1.7516,
      "mean_token_accuracy": 0.6063065364956856,
      "num_tokens": 501198.0,
      "step": 490
    },
    {
      "entropy": 1.7936234265565871,
      "epoch": 0.8,
      "grad_norm": 1.014539361000061,
      "learning_rate": 0.000136706389208128,
      "loss": 1.7999,
      "mean_token_accuracy": 0.5939958959817886,
      "num_tokens": 511342.0,
      "step": 500
    },
    {
      "epoch": 0.8,
      "eval_entropy": 1.755663092319782,
      "eval_loss": 1.732832670211792,
      "eval_mean_token_accuracy": 0.6125743526678818,
      "eval_num_tokens": 511342.0,
      "eval_runtime": 0.5702,
      "eval_samples_per_second": 175.363,
      "eval_steps_per_second": 22.797,
      "step": 500
    },
    {
      "entropy": 1.7916649729013443,
      "epoch": 0.816,
      "grad_norm": 0.8876123428344727,
      "learning_rate": 0.00013428319328952253,
      "loss": 1.7903,
      "mean_token_accuracy": 0.6008384607732296,
      "num_tokens": 521786.0,
      "step": 510
    },
    {
      "entropy": 1.7850607708096504,
      "epoch": 0.832,
      "grad_norm": 0.8096923828125,
      "learning_rate": 0.00013183696432058888,
      "loss": 1.7517,
      "mean_token_accuracy": 0.602414033561945,
      "num_tokens": 532299.0,
      "step": 520
    },
    {
      "entropy": 1.8719337031245231,
      "epoch": 0.848,
      "grad_norm": 0.9576754570007324,
      "learning_rate": 0.00012936934579219094,
      "loss": 1.8609,
      "mean_token_accuracy": 0.5931283295154571,
      "num_tokens": 542782.0,
      "step": 530
    },
    {
      "entropy": 1.9048831015825272,
      "epoch": 0.864,
      "grad_norm": 0.7667029500007629,
      "learning_rate": 0.00012688199556569753,
      "loss": 1.8724,
      "mean_token_accuracy": 0.5880887821316719,
      "num_tokens": 552890.0,
      "step": 540
    },
    {
      "entropy": 1.7590193003416061,
      "epoch": 0.88,
      "grad_norm": 0.9162896275520325,
      "learning_rate": 0.00012437658475915377,
      "loss": 1.7175,
      "mean_token_accuracy": 0.6152773804962635,
      "num_tokens": 563354.0,
      "step": 550
    },
    {
      "entropy": 1.808611199259758,
      "epoch": 0.896,
      "grad_norm": 0.8291792273521423,
      "learning_rate": 0.00012185479662454595,
      "loss": 1.7864,
      "mean_token_accuracy": 0.5958017274737358,
      "num_tokens": 573757.0,
      "step": 560
    },
    {
      "entropy": 1.7611112475395203,
      "epoch": 0.912,
      "grad_norm": 0.909171998500824,
      "learning_rate": 0.00011931832541691418,
      "loss": 1.7679,
      "mean_token_accuracy": 0.602883157134056,
      "num_tokens": 583662.0,
      "step": 570
    },
    {
      "entropy": 1.781042018532753,
      "epoch": 0.928,
      "grad_norm": 0.9129692912101746,
      "learning_rate": 0.00011676887525607271,
      "loss": 1.8436,
      "mean_token_accuracy": 0.5947870224714279,
      "num_tokens": 593910.0,
      "step": 580
    },
    {
      "entropy": 1.836389735341072,
      "epoch": 0.944,
      "grad_norm": 0.826946496963501,
      "learning_rate": 0.0001142081589817027,
      "loss": 1.802,
      "mean_token_accuracy": 0.5981609359383583,
      "num_tokens": 604072.0,
      "step": 590
    },
    {
      "entropy": 1.7494287639856339,
      "epoch": 0.96,
      "grad_norm": 0.8292899131774902,
      "learning_rate": 0.00011163789700258655,
      "loss": 1.7124,
      "mean_token_accuracy": 0.6123683467507363,
      "num_tokens": 614766.0,
      "step": 600
    },
    {
      "epoch": 0.96,
      "eval_entropy": 1.7274914796535785,
      "eval_loss": 1.7254457473754883,
      "eval_mean_token_accuracy": 0.6123642738048847,
      "eval_num_tokens": 614766.0,
      "eval_runtime": 0.561,
      "eval_samples_per_second": 178.238,
      "eval_steps_per_second": 23.171,
      "step": 600
    },
    {
      "entropy": 1.8243329256772995,
      "epoch": 0.976,
      "grad_norm": 0.874477207660675,
      "learning_rate": 0.00010905981614075693,
      "loss": 1.8293,
      "mean_token_accuracy": 0.5974094405770302,
      "num_tokens": 624956.0,
      "step": 610
    },
    {
      "entropy": 1.7830285251140594,
      "epoch": 0.992,
      "grad_norm": 0.8535038232803345,
      "learning_rate": 0.000106475648471337,
      "loss": 1.741,
      "mean_token_accuracy": 0.6073737636208534,
      "num_tokens": 635175.0,
      "step": 620
    }
  ],
  "logging_steps": 10,
  "max_steps": 1250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.465314573775667e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}