{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.99290780141844,
  "eval_steps": 1000,
  "global_step": 951,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03152088258471237,
      "grad_norm": 30.522348403930664,
      "learning_rate": 9.375000000000001e-06,
      "loss": 8.3857,
      "step": 10
    },
    {
      "epoch": 0.06304176516942474,
      "grad_norm": 9.116207122802734,
      "learning_rate": 1.9791666666666665e-05,
      "loss": 5.2184,
      "step": 20
    },
    {
      "epoch": 0.09456264775413711,
      "grad_norm": 15.578251838684082,
      "learning_rate": 3.0208333333333334e-05,
      "loss": 4.6933,
      "step": 30
    },
    {
      "epoch": 0.12608353033884948,
      "grad_norm": 8.683645248413086,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 4.4078,
      "step": 40
    },
    {
      "epoch": 0.15760441292356187,
      "grad_norm": 6.897328853607178,
      "learning_rate": 4.999984870177963e-05,
      "loss": 4.3808,
      "step": 50
    },
    {
      "epoch": 0.18912529550827423,
      "grad_norm": 6.390584468841553,
      "learning_rate": 4.9981695131093794e-05,
      "loss": 4.2976,
      "step": 60
    },
    {
      "epoch": 0.22064617809298662,
      "grad_norm": 4.875730991363525,
      "learning_rate": 4.9933307091588796e-05,
      "loss": 4.1837,
      "step": 70
    },
    {
      "epoch": 0.25216706067769895,
      "grad_norm": 5.509742736816406,
      "learning_rate": 4.985474314561054e-05,
      "loss": 4.165,
      "step": 80
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 36.74412536621094,
      "learning_rate": 4.974609837634544e-05,
      "loss": 4.4715,
      "step": 90
    },
    {
      "epoch": 0.31520882584712373,
      "grad_norm": 4.898118495941162,
      "learning_rate": 4.9607504272744575e-05,
      "loss": 4.3442,
      "step": 100
    },
    {
      "epoch": 0.3467297084318361,
      "grad_norm": 5.879344940185547,
      "learning_rate": 4.943912857038719e-05,
      "loss": 4.0963,
      "step": 110
    },
    {
      "epoch": 0.37825059101654845,
      "grad_norm": 5.230091094970703,
      "learning_rate": 4.9241175048476275e-05,
      "loss": 4.2056,
      "step": 120
    },
    {
      "epoch": 0.4097714736012608,
      "grad_norm": 4.855470180511475,
      "learning_rate": 4.90138832832117e-05,
      "loss": 4.066,
      "step": 130
    },
    {
      "epoch": 0.44129235618597323,
      "grad_norm": 5.235150337219238,
      "learning_rate": 4.8757528357839564e-05,
      "loss": 4.2,
      "step": 140
    },
    {
      "epoch": 0.4728132387706856,
      "grad_norm": 5.098632335662842,
      "learning_rate": 4.847242052972859e-05,
      "loss": 4.1053,
      "step": 150
    },
    {
      "epoch": 0.5043341213553979,
      "grad_norm": 5.152740001678467,
      "learning_rate": 4.8158904854876555e-05,
      "loss": 4.2672,
      "step": 160
    },
    {
      "epoch": 0.5358550039401103,
      "grad_norm": 4.3414082527160645,
      "learning_rate": 4.781736077030106e-05,
      "loss": 3.9562,
      "step": 170
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 4.447638511657715,
      "learning_rate": 4.744820163482029e-05,
      "loss": 3.829,
      "step": 180
    },
    {
      "epoch": 0.598896769109535,
      "grad_norm": 4.406827926635742,
      "learning_rate": 4.70518742287793e-05,
      "loss": 3.7732,
      "step": 190
    },
    {
      "epoch": 0.6304176516942475,
      "grad_norm": 4.291316986083984,
      "learning_rate": 4.66288582133275e-05,
      "loss": 3.748,
      "step": 200
    },
    {
      "epoch": 0.6619385342789598,
      "grad_norm": 5.285233974456787,
      "learning_rate": 4.6179665549901506e-05,
      "loss": 3.7187,
      "step": 210
    },
    {
      "epoch": 0.6934594168636722,
      "grad_norm": 4.025429725646973,
      "learning_rate": 4.5704839880616296e-05,
      "loss": 3.8554,
      "step": 220
    },
    {
      "epoch": 0.7249802994483846,
      "grad_norm": 4.376532554626465,
      "learning_rate": 4.520495587031414e-05,
      "loss": 3.6829,
      "step": 230
    },
    {
      "epoch": 0.7565011820330969,
      "grad_norm": 4.8744306564331055,
      "learning_rate": 4.468061851106797e-05,
      "loss": 3.7738,
      "step": 240
    },
    {
      "epoch": 0.7880220646178093,
      "grad_norm": 20.5545597076416,
      "learning_rate": 4.413246238998069e-05,
      "loss": 3.7111,
      "step": 250
    },
    {
      "epoch": 0.8195429472025216,
      "grad_norm": 4.567636489868164,
      "learning_rate": 4.356115092116668e-05,
      "loss": 3.5261,
      "step": 260
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 4.672543048858643,
      "learning_rate": 4.2967375542844974e-05,
      "loss": 3.5459,
      "step": 270
    },
    {
      "epoch": 0.8825847123719465,
      "grad_norm": 4.592697620391846,
      "learning_rate": 4.235185488051585e-05,
      "loss": 3.6955,
      "step": 280
    },
    {
      "epoch": 0.9141055949566588,
      "grad_norm": 4.082319259643555,
      "learning_rate": 4.171533387723362e-05,
      "loss": 3.6012,
      "step": 290
    },
    {
      "epoch": 0.9456264775413712,
      "grad_norm": 4.013556003570557,
      "learning_rate": 4.1058582892028175e-05,
      "loss": 3.4834,
      "step": 300
    },
    {
      "epoch": 0.9771473601260835,
      "grad_norm": 4.339940547943115,
      "learning_rate": 4.0382396767566536e-05,
      "loss": 3.687,
      "step": 310
    },
    {
      "epoch": 1.0063041765169425,
      "grad_norm": 4.115149021148682,
      "learning_rate": 3.968759386818259e-05,
      "loss": 3.016,
      "step": 320
    },
    {
      "epoch": 1.037825059101655,
      "grad_norm": 3.8200573921203613,
      "learning_rate": 3.897501508943955e-05,
      "loss": 2.0546,
      "step": 330
    },
    {
      "epoch": 1.0693459416863673,
      "grad_norm": 3.4586374759674072,
      "learning_rate": 3.824552284042351e-05,
      "loss": 1.9928,
      "step": 340
    },
    {
      "epoch": 1.1008668242710795,
      "grad_norm": 4.013545513153076,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 2.0627,
      "step": 350
    },
    {
      "epoch": 1.132387706855792,
      "grad_norm": 3.6197803020477295,
      "learning_rate": 3.673934884829673e-05,
      "loss": 1.9238,
      "step": 360
    },
    {
      "epoch": 1.1639085894405043,
      "grad_norm": 3.7971880435943604,
      "learning_rate": 3.5964489974705553e-05,
      "loss": 1.9624,
      "step": 370
    },
    {
      "epoch": 1.1954294720252168,
      "grad_norm": 3.554382562637329,
      "learning_rate": 3.517636116372546e-05,
      "loss": 2.0095,
      "step": 380
    },
    {
      "epoch": 1.226950354609929,
      "grad_norm": 4.342985153198242,
      "learning_rate": 3.437591625999497e-05,
      "loss": 1.8905,
      "step": 390
    },
    {
      "epoch": 1.2584712371946414,
      "grad_norm": 3.6071176528930664,
      "learning_rate": 3.356412401388732e-05,
      "loss": 2.0062,
      "step": 400
    },
    {
      "epoch": 1.2899921197793538,
      "grad_norm": 4.067968845367432,
      "learning_rate": 3.274196690906602e-05,
      "loss": 1.9529,
      "step": 410
    },
    {
      "epoch": 1.3215130023640662,
      "grad_norm": 3.6475672721862793,
      "learning_rate": 3.1910439973419294e-05,
      "loss": 1.9173,
      "step": 420
    },
    {
      "epoch": 1.3530338849487786,
      "grad_norm": 3.6837363243103027,
      "learning_rate": 3.107054957481271e-05,
      "loss": 1.9451,
      "step": 430
    },
    {
      "epoch": 1.384554767533491,
      "grad_norm": 4.136275768280029,
      "learning_rate": 3.022331220311756e-05,
      "loss": 1.9187,
      "step": 440
    },
    {
      "epoch": 1.4160756501182032,
      "grad_norm": 3.433151960372925,
      "learning_rate": 2.93697532399888e-05,
      "loss": 1.7977,
      "step": 450
    },
    {
      "epoch": 1.4475965327029157,
      "grad_norm": 3.2098097801208496,
      "learning_rate": 2.8510905717881614e-05,
      "loss": 1.8138,
      "step": 460
    },
    {
      "epoch": 1.479117415287628,
      "grad_norm": 3.2374401092529297,
      "learning_rate": 2.76478090698085e-05,
      "loss": 1.8966,
      "step": 470
    },
    {
      "epoch": 1.5106382978723403,
      "grad_norm": 3.700331449508667,
      "learning_rate": 2.6781507871349993e-05,
      "loss": 2.0714,
      "step": 480
    },
    {
      "epoch": 1.5421591804570527,
      "grad_norm": 4.613447666168213,
      "learning_rate": 2.5913050576441477e-05,
      "loss": 1.79,
      "step": 490
    },
    {
      "epoch": 1.573680063041765,
      "grad_norm": 3.2395973205566406,
      "learning_rate": 2.5043488248466184e-05,
      "loss": 1.8355,
      "step": 500
    },
    {
      "epoch": 1.6052009456264775,
      "grad_norm": 3.447166681289673,
      "learning_rate": 2.4173873288190114e-05,
      "loss": 1.8917,
      "step": 510
    },
    {
      "epoch": 1.63672182821119,
      "grad_norm": 3.5781800746917725,
      "learning_rate": 2.3305258160078274e-05,
      "loss": 1.8153,
      "step": 520
    },
    {
      "epoch": 1.6682427107959024,
      "grad_norm": 3.1937813758850098,
      "learning_rate": 2.2438694118533875e-05,
      "loss": 1.8061,
      "step": 530
    },
    {
      "epoch": 1.6997635933806148,
      "grad_norm": 3.4112651348114014,
      "learning_rate": 2.1575229935602086e-05,
      "loss": 1.7042,
      "step": 540
    },
    {
      "epoch": 1.731284475965327,
      "grad_norm": 3.6018483638763428,
      "learning_rate": 2.0715910631677968e-05,
      "loss": 1.6401,
      "step": 550
    },
    {
      "epoch": 1.7628053585500394,
      "grad_norm": 3.27825927734375,
      "learning_rate": 1.986177621075499e-05,
      "loss": 1.7344,
      "step": 560
    },
    {
      "epoch": 1.7943262411347518,
      "grad_norm": 3.4082956314086914,
      "learning_rate": 1.9013860401744716e-05,
      "loss": 1.8261,
      "step": 570
    },
    {
      "epoch": 1.825847123719464,
      "grad_norm": 3.231468439102173,
      "learning_rate": 1.817318940739098e-05,
      "loss": 1.6267,
      "step": 580
    },
    {
      "epoch": 1.8573680063041764,
      "grad_norm": 3.3175694942474365,
      "learning_rate": 1.7340780662292677e-05,
      "loss": 1.7841,
      "step": 590
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 3.1403086185455322,
      "learning_rate": 1.651764160153844e-05,
      "loss": 1.7242,
      "step": 600
    },
    {
      "epoch": 1.9204097714736013,
      "grad_norm": 3.0145950317382812,
      "learning_rate": 1.570476844144329e-05,
      "loss": 1.7107,
      "step": 610
    },
    {
      "epoch": 1.9519306540583137,
      "grad_norm": 3.2950940132141113,
      "learning_rate": 1.4903144973862973e-05,
      "loss": 1.5581,
      "step": 620
    },
    {
      "epoch": 1.983451536643026,
      "grad_norm": 3.2230618000030518,
      "learning_rate": 1.4113741375545222e-05,
      "loss": 1.5261,
      "step": 630
    },
    {
      "epoch": 2.012608353033885,
      "grad_norm": 2.3006889820098877,
      "learning_rate": 1.3337513033958904e-05,
      "loss": 1.0825,
      "step": 640
    },
    {
      "epoch": 2.0441292356185974,
      "grad_norm": 2.841254711151123,
      "learning_rate": 1.2575399391022061e-05,
      "loss": 0.6505,
      "step": 650
    },
    {
      "epoch": 2.07565011820331,
      "grad_norm": 2.3877882957458496,
      "learning_rate": 1.1828322806128373e-05,
      "loss": 0.7065,
      "step": 660
    },
    {
      "epoch": 2.107171000788022,
      "grad_norm": 3.346599578857422,
      "learning_rate": 1.1097187439847939e-05,
      "loss": 0.6801,
      "step": 670
    },
    {
      "epoch": 2.1386918833727346,
      "grad_norm": 2.6883623600006104,
      "learning_rate": 1.0382878159653447e-05,
      "loss": 0.7106,
      "step": 680
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 2.5878891944885254,
      "learning_rate": 9.686259468996153e-06,
      "loss": 0.6787,
      "step": 690
    },
    {
      "epoch": 2.201733648542159,
      "grad_norm": 2.9516632556915283,
      "learning_rate": 9.008174461027724e-06,
      "loss": 0.7173,
      "step": 700
    },
    {
      "epoch": 2.2332545311268714,
      "grad_norm": 2.9583373069763184,
      "learning_rate": 8.349443798234116e-06,
      "loss": 0.6442,
      "step": 710
    },
    {
      "epoch": 2.264775413711584,
      "grad_norm": 2.751403331756592,
      "learning_rate": 7.710864719216637e-06,
      "loss": 0.6828,
      "step": 720
    },
    {
      "epoch": 2.2962962962962963,
      "grad_norm": 2.852689743041992,
      "learning_rate": 7.0932100738220265e-06,
      "loss": 0.703,
      "step": 730
    },
    {
      "epoch": 2.3278171788810087,
      "grad_norm": 2.892123222351074,
      "learning_rate": 6.497227387789392e-06,
      "loss": 0.6609,
      "step": 740
    },
    {
      "epoch": 2.359338061465721,
      "grad_norm": 2.5285375118255615,
      "learning_rate": 5.923637958046058e-06,
      "loss": 0.6635,
      "step": 750
    },
    {
      "epoch": 2.3908589440504335,
      "grad_norm": 2.6019787788391113,
      "learning_rate": 5.373135979747227e-06,
      "loss": 0.6354,
      "step": 760
    },
    {
      "epoch": 2.422379826635146,
      "grad_norm": 2.3159821033477783,
      "learning_rate": 4.846387706115932e-06,
      "loss": 0.5731,
      "step": 770
    },
    {
      "epoch": 2.453900709219858,
      "grad_norm": 2.7990407943725586,
      "learning_rate": 4.344030642100133e-06,
      "loss": 0.6754,
      "step": 780
    },
    {
      "epoch": 2.4854215918045703,
      "grad_norm": 2.336843967437744,
      "learning_rate": 3.866672772822863e-06,
      "loss": 0.6382,
      "step": 790
    },
    {
      "epoch": 2.5169424743892828,
      "grad_norm": 2.7701914310455322,
      "learning_rate": 3.4148918277592005e-06,
      "loss": 0.6316,
      "step": 800
    },
    {
      "epoch": 2.548463356973995,
      "grad_norm": 2.583923101425171,
      "learning_rate": 2.989234581530509e-06,
      "loss": 0.631,
      "step": 810
    },
    {
      "epoch": 2.5799842395587076,
      "grad_norm": 2.658202648162842,
      "learning_rate": 2.5902161921623454e-06,
      "loss": 0.6692,
      "step": 820
    },
    {
      "epoch": 2.61150512214342,
      "grad_norm": 3.21925687789917,
      "learning_rate": 2.218319577606778e-06,
      "loss": 0.6858,
      "step": 830
    },
    {
      "epoch": 2.6430260047281324,
      "grad_norm": 2.6975653171539307,
      "learning_rate": 1.8739948312837014e-06,
      "loss": 0.6282,
      "step": 840
    },
    {
      "epoch": 2.674546887312845,
      "grad_norm": 2.680405855178833,
      "learning_rate": 1.5576586773486195e-06,
      "loss": 0.6061,
      "step": 850
    },
    {
      "epoch": 2.7060677698975573,
      "grad_norm": 2.4803316593170166,
      "learning_rate": 1.2696939663459973e-06,
      "loss": 0.6601,
      "step": 860
    },
    {
      "epoch": 2.7375886524822697,
      "grad_norm": 3.0240488052368164,
      "learning_rate": 1.0104492118586773e-06,
      "loss": 0.6088,
      "step": 870
    },
    {
      "epoch": 2.769109535066982,
      "grad_norm": 2.943082809448242,
      "learning_rate": 7.802381687141535e-07,
      "loss": 0.66,
      "step": 880
    },
    {
      "epoch": 2.8006304176516945,
      "grad_norm": 2.768092632293701,
      "learning_rate": 5.793394532580765e-07,
      "loss": 0.6184,
      "step": 890
    },
    {
      "epoch": 2.8321513002364065,
      "grad_norm": 2.631650924682617,
      "learning_rate": 4.079962061546955e-07,
      "loss": 0.5972,
      "step": 900
    },
    {
      "epoch": 2.863672182821119,
      "grad_norm": 2.7756528854370117,
      "learning_rate": 2.664157981222437e-07,
      "loss": 0.6151,
      "step": 910
    },
    {
      "epoch": 2.8951930654058313,
      "grad_norm": 2.6522891521453857,
      "learning_rate": 1.5476957895942946e-07,
      "loss": 0.6093,
      "step": 920
    },
    {
      "epoch": 2.9267139479905437,
      "grad_norm": 2.9670333862304688,
      "learning_rate": 7.319267016678288e-08,
      "loss": 0.6333,
      "step": 930
    },
    {
      "epoch": 2.958234830575256,
      "grad_norm": 2.2578587532043457,
      "learning_rate": 2.1783801413866046e-08,
      "loss": 0.5926,
      "step": 940
    },
    {
      "epoch": 2.9897557131599686,
      "grad_norm": 2.546121120452881,
      "learning_rate": 6.05191050198628e-10,
      "loss": 0.6528,
      "step": 950
    },
    {
      "epoch": 2.99290780141844,
      "step": 951,
      "total_flos": 6.011828060016345e+18,
      "train_loss": 2.2139622458774837,
      "train_runtime": 221179.8915,
      "train_samples_per_second": 0.551,
      "train_steps_per_second": 0.004
    }
  ],
  "logging_steps": 10,
  "max_steps": 951,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 6.011828060016345e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}