{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.269938650306749,
  "eval_steps": 10000000,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12269938650306748,
      "grad_norm": 24.408997453689647,
      "learning_rate": 1.2345679012345678e-08,
      "loss": 2.7299,
      "step": 10
    },
    {
      "epoch": 0.24539877300613497,
      "grad_norm": 24.80765178012853,
      "learning_rate": 2.4691358024691355e-08,
      "loss": 2.7355,
      "step": 20
    },
    {
      "epoch": 0.36809815950920244,
      "grad_norm": 25.187726534643133,
      "learning_rate": 3.7037037037037036e-08,
      "loss": 2.7705,
      "step": 30
    },
    {
      "epoch": 0.49079754601226994,
      "grad_norm": 25.323501806850057,
      "learning_rate": 4.938271604938271e-08,
      "loss": 2.7413,
      "step": 40
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 24.247097730244068,
      "learning_rate": 6.172839506172839e-08,
      "loss": 2.7371,
      "step": 50
    },
    {
      "epoch": 0.7361963190184049,
      "grad_norm": 24.31447909409825,
      "learning_rate": 7.407407407407407e-08,
      "loss": 2.7364,
      "step": 60
    },
    {
      "epoch": 0.8588957055214724,
      "grad_norm": 23.244236630258634,
      "learning_rate": 8.641975308641975e-08,
      "loss": 2.6991,
      "step": 70
    },
    {
      "epoch": 0.9815950920245399,
      "grad_norm": 23.476416023688525,
      "learning_rate": 9.876543209876542e-08,
      "loss": 2.7081,
      "step": 80
    },
    {
      "epoch": 1.1042944785276074,
      "grad_norm": 20.366176427963136,
      "learning_rate": 1.111111111111111e-07,
      "loss": 2.6605,
      "step": 90
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 19.455018752097676,
      "learning_rate": 1.2345679012345677e-07,
      "loss": 2.6013,
      "step": 100
    },
    {
      "epoch": 1.3496932515337423,
      "grad_norm": 19.61610178323797,
      "learning_rate": 1.3580246913580246e-07,
      "loss": 2.5733,
      "step": 110
    },
    {
      "epoch": 1.4723926380368098,
      "grad_norm": 10.050332977637998,
      "learning_rate": 1.4814814814814815e-07,
      "loss": 2.4569,
      "step": 120
    },
    {
      "epoch": 1.5950920245398774,
      "grad_norm": 8.220625543189085,
      "learning_rate": 1.604938271604938e-07,
      "loss": 2.4254,
      "step": 130
    },
    {
      "epoch": 1.7177914110429446,
      "grad_norm": 6.393668406476926,
      "learning_rate": 1.728395061728395e-07,
      "loss": 2.4087,
      "step": 140
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 4.395664172815528,
      "learning_rate": 1.8518518518518516e-07,
      "loss": 2.316,
      "step": 150
    },
    {
      "epoch": 1.9631901840490797,
      "grad_norm": 3.372590478708616,
      "learning_rate": 1.9753086419753084e-07,
      "loss": 2.3029,
      "step": 160
    },
    {
      "epoch": 2.085889570552147,
      "grad_norm": 3.0108879202152203,
      "learning_rate": 2.0987654320987653e-07,
      "loss": 2.2702,
      "step": 170
    },
    {
      "epoch": 2.208588957055215,
      "grad_norm": 2.7225559520418483,
      "learning_rate": 2.222222222222222e-07,
      "loss": 2.253,
      "step": 180
    },
    {
      "epoch": 2.331288343558282,
      "grad_norm": 2.4432492884647794,
      "learning_rate": 2.3456790123456788e-07,
      "loss": 2.2483,
      "step": 190
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 2.3400196487113116,
      "learning_rate": 2.4691358024691354e-07,
      "loss": 2.223,
      "step": 200
    },
    {
      "epoch": 2.5766871165644174,
      "grad_norm": 2.1352206898241612,
      "learning_rate": 2.5925925925925923e-07,
      "loss": 2.2025,
      "step": 210
    },
    {
      "epoch": 2.6993865030674846,
      "grad_norm": 2.023643903641845,
      "learning_rate": 2.716049382716049e-07,
      "loss": 2.1908,
      "step": 220
    },
    {
      "epoch": 2.8220858895705523,
      "grad_norm": 1.9335965899623562,
      "learning_rate": 2.839506172839506e-07,
      "loss": 2.1824,
      "step": 230
    },
    {
      "epoch": 2.9447852760736195,
      "grad_norm": 1.9077831623847687,
      "learning_rate": 2.962962962962963e-07,
      "loss": 2.1608,
      "step": 240
    },
    {
      "epoch": 3.067484662576687,
      "grad_norm": 1.8162929105709213,
      "learning_rate": 3.086419753086419e-07,
      "loss": 2.1647,
      "step": 250
    },
    {
      "epoch": 3.190184049079755,
      "grad_norm": 1.7624936746874469,
      "learning_rate": 3.209876543209876e-07,
      "loss": 2.141,
      "step": 260
    },
    {
      "epoch": 3.312883435582822,
      "grad_norm": 1.7925134660582491,
      "learning_rate": 3.333333333333333e-07,
      "loss": 2.1508,
      "step": 270
    },
    {
      "epoch": 3.4355828220858897,
      "grad_norm": 1.6707105797889426,
      "learning_rate": 3.45679012345679e-07,
      "loss": 2.1824,
      "step": 280
    },
    {
      "epoch": 3.558282208588957,
      "grad_norm": 1.8016949939412934,
      "learning_rate": 3.580246913580247e-07,
      "loss": 2.153,
      "step": 290
    },
    {
      "epoch": 3.6809815950920246,
      "grad_norm": 1.7179134496371937,
      "learning_rate": 3.703703703703703e-07,
      "loss": 2.1517,
      "step": 300
    },
    {
      "epoch": 3.8036809815950923,
      "grad_norm": 1.644436248853206,
      "learning_rate": 3.82716049382716e-07,
      "loss": 2.1282,
      "step": 310
    },
    {
      "epoch": 3.9263803680981595,
      "grad_norm": 1.6481513302515154,
      "learning_rate": 3.950617283950617e-07,
      "loss": 2.1479,
      "step": 320
    },
    {
      "epoch": 4.049079754601227,
      "grad_norm": 1.5834405350588614,
      "learning_rate": 4.0740740740740737e-07,
      "loss": 2.1205,
      "step": 330
    },
    {
      "epoch": 4.171779141104294,
      "grad_norm": 1.5774683971395849,
      "learning_rate": 4.1975308641975306e-07,
      "loss": 2.0829,
      "step": 340
    },
    {
      "epoch": 4.294478527607362,
      "grad_norm": 1.5942216093543309,
      "learning_rate": 4.320987654320987e-07,
      "loss": 2.1374,
      "step": 350
    },
    {
      "epoch": 4.41717791411043,
      "grad_norm": 1.9297965568116624,
      "learning_rate": 4.444444444444444e-07,
      "loss": 2.1496,
      "step": 360
    },
    {
      "epoch": 4.539877300613497,
      "grad_norm": 1.6010753048632336,
      "learning_rate": 4.5679012345679007e-07,
      "loss": 2.1165,
      "step": 370
    },
    {
      "epoch": 4.662576687116564,
      "grad_norm": 2.0277456221645247,
      "learning_rate": 4.6913580246913576e-07,
      "loss": 2.0951,
      "step": 380
    },
    {
      "epoch": 4.785276073619632,
      "grad_norm": 1.6100359846439383,
      "learning_rate": 4.814814814814814e-07,
      "loss": 2.1221,
      "step": 390
    },
    {
      "epoch": 4.9079754601226995,
      "grad_norm": 1.5913594422754647,
      "learning_rate": 4.938271604938271e-07,
      "loss": 2.124,
      "step": 400
    },
    {
      "epoch": 5.030674846625767,
      "grad_norm": 1.543916416223568,
      "learning_rate": 5.061728395061729e-07,
      "loss": 2.093,
      "step": 410
    },
    {
      "epoch": 5.153374233128835,
      "grad_norm": 1.6314296110841167,
      "learning_rate": 5.185185185185185e-07,
      "loss": 2.0833,
      "step": 420
    },
    {
      "epoch": 5.276073619631902,
      "grad_norm": 1.4616507419385325,
      "learning_rate": 5.308641975308642e-07,
      "loss": 2.1095,
      "step": 430
    },
    {
      "epoch": 5.398773006134969,
      "grad_norm": 1.5844834668769874,
      "learning_rate": 5.432098765432098e-07,
      "loss": 2.094,
      "step": 440
    },
    {
      "epoch": 5.521472392638037,
      "grad_norm": 1.558323789647778,
      "learning_rate": 5.555555555555555e-07,
      "loss": 2.0901,
      "step": 450
    },
    {
      "epoch": 5.644171779141105,
      "grad_norm": 1.526929915237913,
      "learning_rate": 5.679012345679012e-07,
      "loss": 2.0844,
      "step": 460
    },
    {
      "epoch": 5.766871165644172,
      "grad_norm": 1.4896565468847545,
      "learning_rate": 5.802469135802469e-07,
      "loss": 2.0774,
      "step": 470
    },
    {
      "epoch": 5.889570552147239,
      "grad_norm": 1.4983542070251128,
      "learning_rate": 5.925925925925926e-07,
      "loss": 2.0979,
      "step": 480
    },
    {
      "epoch": 6.012269938650307,
      "grad_norm": 1.474608208975854,
      "learning_rate": 6.049382716049383e-07,
      "loss": 2.087,
      "step": 490
    },
    {
      "epoch": 6.134969325153374,
      "grad_norm": 1.6014533478490705,
      "learning_rate": 6.172839506172839e-07,
      "loss": 2.0547,
      "step": 500
    },
    {
      "epoch": 6.257668711656442,
      "grad_norm": 1.454642285240305,
      "learning_rate": 6.296296296296296e-07,
      "loss": 2.0449,
      "step": 510
    },
    {
      "epoch": 6.38036809815951,
      "grad_norm": 1.6436891729462102,
      "learning_rate": 6.419753086419752e-07,
      "loss": 2.0696,
      "step": 520
    },
    {
      "epoch": 6.5030674846625764,
      "grad_norm": 1.4807234984332847,
      "learning_rate": 6.54320987654321e-07,
      "loss": 2.0742,
      "step": 530
    },
    {
      "epoch": 6.625766871165644,
      "grad_norm": 1.5356652480116542,
      "learning_rate": 6.666666666666666e-07,
      "loss": 2.0696,
      "step": 540
    },
    {
      "epoch": 6.748466257668712,
      "grad_norm": 1.6579289531016874,
      "learning_rate": 6.790123456790123e-07,
      "loss": 2.0551,
      "step": 550
    },
    {
      "epoch": 6.871165644171779,
      "grad_norm": 1.55170120881645,
      "learning_rate": 6.91358024691358e-07,
      "loss": 2.0657,
      "step": 560
    },
    {
      "epoch": 6.993865030674847,
      "grad_norm": 1.4899273802926296,
      "learning_rate": 7.037037037037037e-07,
      "loss": 2.078,
      "step": 570
    },
    {
      "epoch": 7.116564417177914,
      "grad_norm": 1.44595351107914,
      "learning_rate": 7.160493827160494e-07,
      "loss": 2.0238,
      "step": 580
    },
    {
      "epoch": 7.2392638036809815,
      "grad_norm": 1.453250943181157,
      "learning_rate": 7.28395061728395e-07,
      "loss": 2.0446,
      "step": 590
    },
    {
      "epoch": 7.361963190184049,
      "grad_norm": 1.598197750855362,
      "learning_rate": 7.407407407407406e-07,
      "loss": 2.0297,
      "step": 600
    },
    {
      "epoch": 7.484662576687117,
      "grad_norm": 1.6244958519175605,
      "learning_rate": 7.530864197530864e-07,
      "loss": 2.0399,
      "step": 610
    },
    {
      "epoch": 7.6073619631901845,
      "grad_norm": 1.690696721697701,
      "learning_rate": 7.65432098765432e-07,
      "loss": 2.0291,
      "step": 620
    },
    {
      "epoch": 7.730061349693251,
      "grad_norm": 1.540010541664445,
      "learning_rate": 7.777777777777778e-07,
      "loss": 2.047,
      "step": 630
    },
    {
      "epoch": 7.852760736196319,
      "grad_norm": 1.4784188621578072,
      "learning_rate": 7.901234567901234e-07,
      "loss": 2.0237,
      "step": 640
    },
    {
      "epoch": 7.975460122699387,
      "grad_norm": 1.5806193808808893,
      "learning_rate": 8.024691358024691e-07,
      "loss": 2.0349,
      "step": 650
    },
    {
      "epoch": 8.098159509202453,
      "grad_norm": 1.5066964330658228,
      "learning_rate": 8.148148148148147e-07,
      "loss": 2.019,
      "step": 660
    },
    {
      "epoch": 8.220858895705522,
      "grad_norm": 1.4657046782549297,
      "learning_rate": 8.271604938271604e-07,
      "loss": 2.0146,
      "step": 670
    },
    {
      "epoch": 8.343558282208589,
      "grad_norm": 1.5129566513929427,
      "learning_rate": 8.395061728395061e-07,
      "loss": 1.9937,
      "step": 680
    },
    {
      "epoch": 8.466257668711656,
      "grad_norm": 1.6073531745364882,
      "learning_rate": 8.518518518518518e-07,
      "loss": 2.0201,
      "step": 690
    },
    {
      "epoch": 8.588957055214724,
      "grad_norm": 1.465742853126048,
      "learning_rate": 8.641975308641974e-07,
      "loss": 2.0263,
      "step": 700
    },
    {
      "epoch": 8.71165644171779,
      "grad_norm": 1.4713574884471192,
      "learning_rate": 8.765432098765432e-07,
      "loss": 2.0074,
      "step": 710
    },
    {
      "epoch": 8.83435582822086,
      "grad_norm": 1.4818223515949642,
      "learning_rate": 8.888888888888888e-07,
      "loss": 1.9803,
      "step": 720
    },
    {
      "epoch": 8.957055214723926,
      "grad_norm": 1.4960768789879928,
      "learning_rate": 9.012345679012346e-07,
      "loss": 1.9988,
      "step": 730
    },
    {
      "epoch": 9.079754601226995,
      "grad_norm": 1.5034869146088043,
      "learning_rate": 9.135802469135801e-07,
      "loss": 1.9713,
      "step": 740
    },
    {
      "epoch": 9.202453987730062,
      "grad_norm": 1.743667247923265,
      "learning_rate": 9.259259259259259e-07,
      "loss": 1.9839,
      "step": 750
    },
    {
      "epoch": 9.325153374233128,
      "grad_norm": 1.4121136174696463,
      "learning_rate": 9.382716049382715e-07,
      "loss": 1.986,
      "step": 760
    },
    {
      "epoch": 9.447852760736197,
      "grad_norm": 1.5213247914456218,
      "learning_rate": 9.506172839506172e-07,
      "loss": 1.9643,
      "step": 770
    },
    {
      "epoch": 9.570552147239264,
      "grad_norm": 1.701165501505246,
      "learning_rate": 9.629629629629628e-07,
      "loss": 1.944,
      "step": 780
    },
    {
      "epoch": 9.69325153374233,
      "grad_norm": 1.7772160828783827,
      "learning_rate": 9.753086419753086e-07,
      "loss": 1.9757,
      "step": 790
    },
    {
      "epoch": 9.815950920245399,
      "grad_norm": 1.5247152271180484,
      "learning_rate": 9.876543209876542e-07,
      "loss": 1.964,
      "step": 800
    },
    {
      "epoch": 9.938650306748466,
      "grad_norm": 1.591870678195776,
      "learning_rate": 1e-06,
      "loss": 1.9758,
      "step": 810
    },
    {
      "epoch": 10.061349693251534,
      "grad_norm": 1.8151248184194306,
      "learning_rate": 9.999953571567085e-07,
      "loss": 1.9689,
      "step": 820
    },
    {
      "epoch": 10.184049079754601,
      "grad_norm": 1.5659661977407868,
      "learning_rate": 9.999814287130579e-07,
      "loss": 1.9584,
      "step": 830
    },
    {
      "epoch": 10.30674846625767,
      "grad_norm": 1.515025607589833,
      "learning_rate": 9.999582149277185e-07,
      "loss": 1.9602,
      "step": 840
    },
    {
      "epoch": 10.429447852760736,
      "grad_norm": 1.6025743323173112,
      "learning_rate": 9.999257162318025e-07,
      "loss": 1.936,
      "step": 850
    },
    {
      "epoch": 10.552147239263803,
      "grad_norm": 1.622286529441025,
      "learning_rate": 9.99883933228855e-07,
      "loss": 1.9014,
      "step": 860
    },
    {
      "epoch": 10.674846625766872,
      "grad_norm": 1.9193284429523172,
      "learning_rate": 9.998328666948437e-07,
      "loss": 1.9282,
      "step": 870
    },
    {
      "epoch": 10.797546012269938,
      "grad_norm": 1.4435929643824292,
      "learning_rate": 9.997725175781443e-07,
      "loss": 1.9131,
      "step": 880
    },
    {
      "epoch": 10.920245398773005,
      "grad_norm": 1.5657198611737813,
      "learning_rate": 9.99702886999523e-07,
      "loss": 1.9444,
      "step": 890
    },
    {
      "epoch": 11.042944785276074,
      "grad_norm": 1.5303385836526135,
      "learning_rate": 9.99623976252115e-07,
      "loss": 1.9228,
      "step": 900
    },
    {
      "epoch": 11.16564417177914,
      "grad_norm": 1.5388464555082484,
      "learning_rate": 9.995357868014013e-07,
      "loss": 1.8934,
      "step": 910
    },
    {
      "epoch": 11.28834355828221,
      "grad_norm": 1.6428643084524577,
      "learning_rate": 9.994383202851812e-07,
      "loss": 1.9055,
      "step": 920
    },
    {
      "epoch": 11.411042944785276,
      "grad_norm": 1.4935758108698727,
      "learning_rate": 9.993315785135416e-07,
      "loss": 1.9118,
      "step": 930
    },
    {
      "epoch": 11.533742331288344,
      "grad_norm": 1.4752185060017586,
      "learning_rate": 9.992155634688238e-07,
      "loss": 1.8953,
      "step": 940
    },
    {
      "epoch": 11.656441717791411,
      "grad_norm": 1.5678765737976184,
      "learning_rate": 9.990902773055865e-07,
      "loss": 1.8996,
      "step": 950
    },
    {
      "epoch": 11.779141104294478,
      "grad_norm": 1.5181431605841578,
      "learning_rate": 9.98955722350566e-07,
      "loss": 1.8932,
      "step": 960
    },
    {
      "epoch": 11.901840490797547,
      "grad_norm": 1.8570007259349715,
      "learning_rate": 9.988119011026324e-07,
      "loss": 1.8856,
      "step": 970
    },
    {
      "epoch": 12.024539877300613,
      "grad_norm": 1.5488809335108225,
      "learning_rate": 9.986588162327434e-07,
      "loss": 1.8918,
      "step": 980
    },
    {
      "epoch": 12.14723926380368,
      "grad_norm": 1.5533160872341008,
      "learning_rate": 9.98496470583896e-07,
      "loss": 1.8707,
      "step": 990
    },
    {
      "epoch": 12.269938650306749,
      "grad_norm": 1.5667441114509637,
      "learning_rate": 9.983248671710714e-07,
      "loss": 1.8489,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 8100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 67376603725824.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}