{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.20630892698727074,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "grad_norm": 12.003094673156738,
      "learning_rate": 0.00001,
      "loss": 2.4799,
      "step": 100
    },
    {
      "epoch": 0,
      "grad_norm": 2.0890064239501953,
      "learning_rate": 0.00002,
      "loss": 0.6329,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.033069133758545,
      "learning_rate": 0.00003,
      "loss": 0.5035,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.252149820327759,
      "learning_rate": 0.00004,
      "loss": 0.4943,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.118387460708618,
      "learning_rate": 0.00005,
      "loss": 0.5105,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.954121708869934,
      "learning_rate": 0.000049473684210526315,
      "loss": 0.5008,
      "step": 600
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.036804676055908,
      "learning_rate": 0.000048947368421052635,
      "loss": 0.4793,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.824407696723938,
      "learning_rate": 0.00004842105263157895,
      "loss": 0.469,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.724145531654358,
      "learning_rate": 0.00004789473684210526,
      "loss": 0.4633,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3427892923355103,
      "learning_rate": 0.00004736842105263158,
      "loss": 0.4614,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0959577560424805,
      "learning_rate": 0.0000468421052631579,
      "loss": 0.4605,
      "step": 1100
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5479639768600464,
      "learning_rate": 0.000046315789473684214,
      "loss": 0.46,
      "step": 1200
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.535975456237793,
      "learning_rate": 0.000045789473684210527,
      "loss": 0.4252,
      "step": 1300
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5321663618087769,
      "learning_rate": 0.000045263157894736846,
      "loss": 0.4369,
      "step": 1400
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5459933280944824,
      "learning_rate": 0.00004473684210526316,
      "loss": 0.4276,
      "step": 1500
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.7711249589920044,
      "learning_rate": 0.00004421052631578947,
      "loss": 0.4474,
      "step": 1600
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.376570701599121,
      "learning_rate": 0.00004368421052631579,
      "loss": 0.4064,
      "step": 1700
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.74457585811615,
      "learning_rate": 0.000043157894736842105,
      "loss": 0.4093,
      "step": 1800
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.4628654718399048,
      "learning_rate": 0.000042631578947368425,
      "loss": 0.4236,
      "step": 1900
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0637551546096802,
      "learning_rate": 0.00004210526315789474,
      "loss": 0.4137,
      "step": 2000
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.385728359222412,
      "learning_rate": 0.00004157894736842106,
      "loss": 0.4071,
      "step": 2100
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.3536651134490967,
      "learning_rate": 0.00004105263157894737,
      "loss": 0.4215,
      "step": 2200
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5586481094360352,
      "learning_rate": 0.000040526315789473684,
      "loss": 0.4041,
      "step": 2300
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0527902841567993,
      "learning_rate": 0.00004,
      "loss": 0.3984,
      "step": 2400
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5508724451065063,
      "learning_rate": 0.000039473684210526316,
      "loss": 0.406,
      "step": 2500
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7268626689910889,
      "learning_rate": 0.00003894736842105263,
      "loss": 0.3978,
      "step": 2600
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.259959101676941,
      "learning_rate": 0.00003842105263157895,
      "loss": 0.4111,
      "step": 2700
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5187910795211792,
      "learning_rate": 0.00003789473684210527,
      "loss": 0.3862,
      "step": 2800
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.4480005502700806,
      "learning_rate": 0.00003736842105263158,
      "loss": 0.4021,
      "step": 2900
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.2096664905548096,
      "learning_rate": 0.000036842105263157895,
      "loss": 0.4092,
      "step": 3000
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0258468389511108,
      "learning_rate": 0.000036315789473684214,
      "loss": 0.3829,
      "step": 3100
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.336549162864685,
      "learning_rate": 0.00003578947368421053,
      "loss": 0.4082,
      "step": 3200
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1610171794891357,
      "learning_rate": 0.00003526315789473684,
      "loss": 0.4216,
      "step": 3300
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1225395202636719,
      "learning_rate": 0.00003473684210526316,
      "loss": 0.3866,
      "step": 3400
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2079445123672485,
      "learning_rate": 0.00003421052631578947,
      "loss": 0.3827,
      "step": 3500
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8150144815444946,
      "learning_rate": 0.00003368421052631579,
      "loss": 0.3833,
      "step": 3600
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1681067943572998,
      "learning_rate": 0.000033157894736842106,
      "loss": 0.3953,
      "step": 3700
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1494988203048706,
      "learning_rate": 0.000032631578947368426,
      "loss": 0.3785,
      "step": 3800
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8699579238891602,
      "learning_rate": 0.00003210526315789474,
      "loss": 0.3802,
      "step": 3900
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.560229778289795,
      "learning_rate": 0.00003157894736842105,
      "loss": 0.3837,
      "step": 4000
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1917518377304077,
      "learning_rate": 0.00003105263157894737,
      "loss": 0.3775,
      "step": 4100
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.050404667854309,
      "learning_rate": 0.000030526315789473684,
      "loss": 0.3781,
      "step": 4200
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8547882437705994,
      "learning_rate": 0.00003,
      "loss": 0.3629,
      "step": 4300
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0195717811584473,
      "learning_rate": 0.000029473684210526314,
      "loss": 0.3644,
      "step": 4400
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0551220178604126,
      "learning_rate": 0.000028947368421052634,
      "loss": 0.3774,
      "step": 4500
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9689894318580627,
      "learning_rate": 0.00002842105263157895,
      "loss": 0.3384,
      "step": 4600
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3785911798477173,
      "learning_rate": 0.000027894736842105263,
      "loss": 0.3725,
      "step": 4700
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1647924184799194,
      "learning_rate": 0.000027368421052631583,
      "loss": 0.3716,
      "step": 4800
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0317350625991821,
      "learning_rate": 0.000026842105263157896,
      "loss": 0.3454,
      "step": 4900
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2156306505203247,
      "learning_rate": 0.000026315789473684212,
      "loss": 0.3414,
      "step": 5000
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.0330926179885864,
      "learning_rate": 0.00002578947368421053,
      "loss": 0.3658,
      "step": 5100
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1999331712722778,
      "learning_rate": 0.000025263157894736845,
      "loss": 0.3659,
      "step": 5200
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8645588159561157,
      "learning_rate": 0.000024736842105263158,
      "loss": 0.3567,
      "step": 5300
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3966509103775024,
      "learning_rate": 0.000024210526315789474,
      "loss": 0.3593,
      "step": 5400
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.9758176207542419,
      "learning_rate": 0.00002368421052631579,
      "loss": 0.3491,
      "step": 5500
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0079020261764526,
      "learning_rate": 0.000023157894736842107,
      "loss": 0.3403,
      "step": 5600
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0078275203704834,
      "learning_rate": 0.000022631578947368423,
      "loss": 0.3396,
      "step": 5700
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2329459190368652,
      "learning_rate": 0.000022105263157894736,
      "loss": 0.3323,
      "step": 5800
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5663444995880127,
      "learning_rate": 0.000021578947368421053,
      "loss": 0.332,
      "step": 5900
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.104218602180481,
      "learning_rate": 0.00002105263157894737,
      "loss": 0.346,
      "step": 6000
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1730228662490845,
      "learning_rate": 0.000020526315789473685,
      "loss": 0.3485,
      "step": 6100
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2409980297088623,
      "learning_rate": 0.00002,
      "loss": 0.3539,
      "step": 6200
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.933708906173706,
      "learning_rate": 0.000019473684210526315,
      "loss": 0.3177,
      "step": 6300
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.0058486461639404,
      "learning_rate": 0.000018947368421052634,
      "loss": 0.3394,
      "step": 6400
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2994540929794312,
      "learning_rate": 0.000018421052631578947,
      "loss": 0.3447,
      "step": 6500
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1113885641098022,
      "learning_rate": 0.000017894736842105264,
      "loss": 0.3572,
      "step": 6600
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1501245498657227,
      "learning_rate": 0.00001736842105263158,
      "loss": 0.3468,
      "step": 6700
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5754578113555908,
      "learning_rate": 0.000016842105263157896,
      "loss": 0.3461,
      "step": 6800
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.214449167251587,
      "learning_rate": 0.000016315789473684213,
      "loss": 0.3463,
      "step": 6900
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.38958740234375,
      "learning_rate": 0.000015789473684210526,
      "loss": 0.3432,
      "step": 7000
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1833761930465698,
      "learning_rate": 0.000015263157894736842,
      "loss": 0.336,
      "step": 7100
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2316564321517944,
      "learning_rate": 0.000014736842105263157,
      "loss": 0.3266,
      "step": 7200
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5909250974655151,
      "learning_rate": 0.000014210526315789475,
      "loss": 0.3354,
      "step": 7300
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1195193529129028,
      "learning_rate": 0.000013684210526315791,
      "loss": 0.3388,
      "step": 7400
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1586064100265503,
      "learning_rate": 0.000013157894736842106,
      "loss": 0.3413,
      "step": 7500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.935364305973053,
      "learning_rate": 0.000012631578947368422,
      "loss": 0.3359,
      "step": 7600
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3456859588623047,
      "learning_rate": 0.000012105263157894737,
      "loss": 0.3328,
      "step": 7700
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5139777660369873,
      "learning_rate": 0.000011578947368421053,
      "loss": 0.3499,
      "step": 7800
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.0874484777450562,
      "learning_rate": 0.000011052631578947368,
      "loss": 0.3123,
      "step": 7900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9059338569641113,
      "learning_rate": 0.000010526315789473684,
      "loss": 0.3305,
      "step": 8000
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1520755290985107,
      "learning_rate": 0.00001,
      "loss": 0.3235,
      "step": 8100
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2957168817520142,
      "learning_rate": 0.000009473684210526317,
      "loss": 0.3204,
      "step": 8200
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.0766379833221436,
      "learning_rate": 0.000008947368421052632,
      "loss": 0.3239,
      "step": 8300
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1667858362197876,
      "learning_rate": 0.000008421052631578948,
      "loss": 0.3183,
      "step": 8400
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.364989161491394,
      "learning_rate": 0.000007894736842105263,
      "loss": 0.3326,
      "step": 8500
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2203367948532104,
      "learning_rate": 0.0000073684210526315784,
      "loss": 0.3205,
      "step": 8600
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.254922866821289,
      "learning_rate": 0.000006842105263157896,
      "loss": 0.3448,
      "step": 8700
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.471166729927063,
      "learning_rate": 0.000006315789473684211,
      "loss": 0.3307,
      "step": 8800
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1350233554840088,
      "learning_rate": 0.000005789473684210527,
      "loss": 0.3174,
      "step": 8900
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.953658401966095,
      "learning_rate": 0.000005263157894736842,
      "loss": 0.3327,
      "step": 9000
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2155749797821045,
      "learning_rate": 0.000004736842105263159,
      "loss": 0.3251,
      "step": 9100
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.106632947921753,
      "learning_rate": 0.000004210526315789474,
      "loss": 0.3011,
      "step": 9200
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2169803380966187,
      "learning_rate": 0.0000036842105263157892,
      "loss": 0.3331,
      "step": 9300
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7549806237220764,
      "learning_rate": 0.0000031578947368421056,
      "loss": 0.319,
      "step": 9400
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2598366737365723,
      "learning_rate": 0.000002631578947368421,
      "loss": 0.3201,
      "step": 9500
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.991303563117981,
      "learning_rate": 0.000002105263157894737,
      "loss": 0.3144,
      "step": 9600
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.120094656944275,
      "learning_rate": 0.0000015789473684210528,
      "loss": 0.2997,
      "step": 9700
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.9906071424484253,
      "learning_rate": 0.0000010526315789473685,
      "loss": 0.3244,
      "step": 9800
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.083082675933838,
      "learning_rate": 5.263157894736843e-7,
      "loss": 0.3382,
      "step": 9900
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6393320560455322,
      "learning_rate": 0,
      "loss": 0.3064,
      "step": 10000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "total_flos": 21671046021120000,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}