{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 1070,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 2.035820722579956,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.0122,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 2.3610706329345703,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.3526,
      "step": 20
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 1.186436653137207,
      "learning_rate": 5.370370370370371e-05,
      "loss": 0.2888,
      "step": 30
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 2.2752561569213867,
      "learning_rate": 7.222222222222222e-05,
      "loss": 0.2957,
      "step": 40
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 1.9183247089385986,
      "learning_rate": 9.074074074074075e-05,
      "loss": 0.2282,
      "step": 50
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 1.8604793548583984,
      "learning_rate": 9.999402437003975e-05,
      "loss": 0.169,
      "step": 60
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 1.4165388345718384,
      "learning_rate": 9.99462278999732e-05,
      "loss": 0.1587,
      "step": 70
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 1.3570247888565063,
      "learning_rate": 9.985068065535225e-05,
      "loss": 0.1176,
      "step": 80
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 0.8269003033638,
      "learning_rate": 9.970747398351445e-05,
      "loss": 0.1031,
      "step": 90
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 1.9058301448822021,
      "learning_rate": 9.951674479629056e-05,
      "loss": 0.0991,
      "step": 100
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 1.9390472173690796,
      "learning_rate": 9.927867543911091e-05,
      "loss": 0.0873,
      "step": 110
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 0.9300522208213806,
      "learning_rate": 9.899349351667522e-05,
      "loss": 0.0946,
      "step": 120
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 0.9239804148674011,
      "learning_rate": 9.866147167535254e-05,
      "loss": 0.0794,
      "step": 130
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 1.4692777395248413,
      "learning_rate": 9.828292734251944e-05,
      "loss": 0.0898,
      "step": 140
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 0.8343071937561035,
      "learning_rate": 9.785822242308562e-05,
      "loss": 0.0819,
      "step": 150
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 0.7223255634307861,
      "learning_rate": 9.738776295349687e-05,
      "loss": 0.0803,
      "step": 160
    },
    {
      "epoch": 1.588785046728972,
      "grad_norm": 0.7587630748748779,
      "learning_rate": 9.687199871354669e-05,
      "loss": 0.0745,
      "step": 170
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 1.022656798362732,
      "learning_rate": 9.631142279636706e-05,
      "loss": 0.0894,
      "step": 180
    },
    {
      "epoch": 1.7757009345794392,
      "grad_norm": 0.8869289755821228,
      "learning_rate": 9.570657113700985e-05,
      "loss": 0.0859,
      "step": 190
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 0.9170877933502197,
      "learning_rate": 9.50580220000696e-05,
      "loss": 0.0737,
      "step": 200
    },
    {
      "epoch": 1.9626168224299065,
      "grad_norm": 1.0185060501098633,
      "learning_rate": 9.436639542683727e-05,
      "loss": 0.0696,
      "step": 210
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 1.1690869331359863,
      "learning_rate": 9.363235264251369e-05,
      "loss": 0.0635,
      "step": 220
    },
    {
      "epoch": 2.149532710280374,
      "grad_norm": 0.940513551235199,
      "learning_rate": 9.285659542404941e-05,
      "loss": 0.072,
      "step": 230
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 0.5256121158599854,
      "learning_rate": 9.203986542921532e-05,
      "loss": 0.065,
      "step": 240
    },
    {
      "epoch": 2.336448598130841,
      "grad_norm": 0.7171943783760071,
      "learning_rate": 9.11829434875454e-05,
      "loss": 0.0594,
      "step": 250
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 0.6012755036354065,
      "learning_rate": 9.02866488538296e-05,
      "loss": 0.0657,
      "step": 260
    },
    {
      "epoch": 2.5233644859813085,
      "grad_norm": 0.496242493391037,
      "learning_rate": 8.93518384248705e-05,
      "loss": 0.0712,
      "step": 270
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 0.7729496955871582,
      "learning_rate": 8.837940592025257e-05,
      "loss": 0.0581,
      "step": 280
    },
    {
      "epoch": 2.710280373831776,
      "grad_norm": 0.6967052817344666,
      "learning_rate": 8.737028102790723e-05,
      "loss": 0.065,
      "step": 290
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 1.2209193706512451,
      "learning_rate": 8.632542851529051e-05,
      "loss": 0.0673,
      "step": 300
    },
    {
      "epoch": 2.897196261682243,
      "grad_norm": 0.8657803535461426,
      "learning_rate": 8.524584730702339e-05,
      "loss": 0.0778,
      "step": 310
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 0.6701989769935608,
      "learning_rate": 8.413256952987611e-05,
      "loss": 0.0617,
      "step": 320
    },
    {
      "epoch": 3.0841121495327104,
      "grad_norm": 0.46317973732948303,
      "learning_rate": 8.298665952600999e-05,
      "loss": 0.061,
      "step": 330
    },
    {
      "epoch": 3.177570093457944,
      "grad_norm": 0.549863338470459,
      "learning_rate": 8.180921283541986e-05,
      "loss": 0.0608,
      "step": 340
    },
    {
      "epoch": 3.2710280373831777,
      "grad_norm": 0.5789127349853516,
      "learning_rate": 8.060135514854994e-05,
      "loss": 0.0588,
      "step": 350
    },
    {
      "epoch": 3.364485981308411,
      "grad_norm": 0.5358679294586182,
      "learning_rate": 7.936424123008464e-05,
      "loss": 0.0584,
      "step": 360
    },
    {
      "epoch": 3.457943925233645,
      "grad_norm": 0.682597815990448,
      "learning_rate": 7.809905381494316e-05,
      "loss": 0.0585,
      "step": 370
    },
    {
      "epoch": 3.5514018691588785,
      "grad_norm": 0.7718446254730225,
      "learning_rate": 7.68070024775332e-05,
      "loss": 0.0525,
      "step": 380
    },
    {
      "epoch": 3.6448598130841123,
      "grad_norm": 0.5267370343208313,
      "learning_rate": 7.548932247534506e-05,
      "loss": 0.0537,
      "step": 390
    },
    {
      "epoch": 3.7383177570093458,
      "grad_norm": 0.8042123913764954,
      "learning_rate": 7.414727356799154e-05,
      "loss": 0.0613,
      "step": 400
    },
    {
      "epoch": 3.831775700934579,
      "grad_norm": 0.5799658298492432,
      "learning_rate": 7.27821388128227e-05,
      "loss": 0.0535,
      "step": 410
    },
    {
      "epoch": 3.925233644859813,
      "grad_norm": 0.8124668598175049,
      "learning_rate": 7.139522333826707e-05,
      "loss": 0.0493,
      "step": 420
    },
    {
      "epoch": 4.018691588785047,
      "grad_norm": 0.8771295547485352,
      "learning_rate": 6.99878530960719e-05,
      "loss": 0.0615,
      "step": 430
    },
    {
      "epoch": 4.11214953271028,
      "grad_norm": 0.7655673027038574,
      "learning_rate": 6.856137359363533e-05,
      "loss": 0.0605,
      "step": 440
    },
    {
      "epoch": 4.205607476635514,
      "grad_norm": 0.6487105488777161,
      "learning_rate": 6.711714860764266e-05,
      "loss": 0.0559,
      "step": 450
    },
    {
      "epoch": 4.299065420560748,
      "grad_norm": 0.7439901232719421,
      "learning_rate": 6.565655888023618e-05,
      "loss": 0.0562,
      "step": 460
    },
    {
      "epoch": 4.392523364485982,
      "grad_norm": 0.566832423210144,
      "learning_rate": 6.418100079896556e-05,
      "loss": 0.0535,
      "step": 470
    },
    {
      "epoch": 4.485981308411215,
      "grad_norm": 0.5109242796897888,
      "learning_rate": 6.269188506178019e-05,
      "loss": 0.0558,
      "step": 480
    },
    {
      "epoch": 4.579439252336448,
      "grad_norm": 0.5645363926887512,
      "learning_rate": 6.11906353283405e-05,
      "loss": 0.0445,
      "step": 490
    },
    {
      "epoch": 4.672897196261682,
      "grad_norm": 0.7012343406677246,
      "learning_rate": 5.967868685893715e-05,
      "loss": 0.0496,
      "step": 500
    },
    {
      "epoch": 4.766355140186916,
      "grad_norm": 0.47865885496139526,
      "learning_rate": 5.815748514231944e-05,
      "loss": 0.0486,
      "step": 510
    },
    {
      "epoch": 4.859813084112149,
      "grad_norm": 0.38202741742134094,
      "learning_rate": 5.6628484513745e-05,
      "loss": 0.0412,
      "step": 520
    },
    {
      "epoch": 4.953271028037383,
      "grad_norm": 0.7451114058494568,
      "learning_rate": 5.5093146764571866e-05,
      "loss": 0.0561,
      "step": 530
    },
    {
      "epoch": 5.046728971962617,
      "grad_norm": 0.29692840576171875,
      "learning_rate": 5.355293974472197e-05,
      "loss": 0.0432,
      "step": 540
    },
    {
      "epoch": 5.140186915887851,
      "grad_norm": 0.5470758080482483,
      "learning_rate": 5.2009335959352666e-05,
      "loss": 0.0456,
      "step": 550
    },
    {
      "epoch": 5.233644859813084,
      "grad_norm": 0.3530406057834625,
      "learning_rate": 5.046381116107742e-05,
      "loss": 0.05,
      "step": 560
    },
    {
      "epoch": 5.327102803738318,
      "grad_norm": 0.5904394388198853,
      "learning_rate": 4.891784293908192e-05,
      "loss": 0.0479,
      "step": 570
    },
    {
      "epoch": 5.420560747663552,
      "grad_norm": 0.6387467384338379,
      "learning_rate": 4.7372909306484276e-05,
      "loss": 0.0419,
      "step": 580
    },
    {
      "epoch": 5.5140186915887845,
      "grad_norm": 0.3460151255130768,
      "learning_rate": 4.5830487287289966e-05,
      "loss": 0.0457,
      "step": 590
    },
    {
      "epoch": 5.607476635514018,
      "grad_norm": 0.6098641157150269,
      "learning_rate": 4.429205150429241e-05,
      "loss": 0.038,
      "step": 600
    },
    {
      "epoch": 5.700934579439252,
      "grad_norm": 0.5131182670593262,
      "learning_rate": 4.275907276926918e-05,
      "loss": 0.0518,
      "step": 610
    },
    {
      "epoch": 5.794392523364486,
      "grad_norm": 0.5069635510444641,
      "learning_rate": 4.123301667682171e-05,
      "loss": 0.0447,
      "step": 620
    },
    {
      "epoch": 5.88785046728972,
      "grad_norm": 0.4461626708507538,
      "learning_rate": 3.971534220320291e-05,
      "loss": 0.0465,
      "step": 630
    },
    {
      "epoch": 5.981308411214953,
      "grad_norm": 0.5642176866531372,
      "learning_rate": 3.820750031147211e-05,
      "loss": 0.0415,
      "step": 640
    },
    {
      "epoch": 6.074766355140187,
      "grad_norm": 0.525589108467102,
      "learning_rate": 3.67109325643111e-05,
      "loss": 0.0398,
      "step": 650
    },
    {
      "epoch": 6.168224299065421,
      "grad_norm": 0.5913786292076111,
      "learning_rate": 3.522706974582717e-05,
      "loss": 0.0421,
      "step": 660
    },
    {
      "epoch": 6.261682242990654,
      "grad_norm": 0.44272592663764954,
      "learning_rate": 3.375733049366115e-05,
      "loss": 0.0377,
      "step": 670
    },
    {
      "epoch": 6.355140186915888,
      "grad_norm": 0.5991699695587158,
      "learning_rate": 3.2303119942707796e-05,
      "loss": 0.0401,
      "step": 680
    },
    {
      "epoch": 6.4485981308411215,
      "grad_norm": 0.4788666069507599,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0344,
      "step": 690
    },
    {
      "epoch": 6.542056074766355,
      "grad_norm": 0.2347191423177719,
      "learning_rate": 2.944682992425959e-05,
      "loss": 0.0431,
      "step": 700
    },
    {
      "epoch": 6.635514018691588,
      "grad_norm": 0.4982701241970062,
      "learning_rate": 2.804748119472969e-05,
      "loss": 0.0431,
      "step": 710
    },
    {
      "epoch": 6.728971962616822,
      "grad_norm": 0.3474225103855133,
      "learning_rate": 2.6669120031637663e-05,
      "loss": 0.0312,
      "step": 720
    },
    {
      "epoch": 6.822429906542056,
      "grad_norm": 0.484762042760849,
      "learning_rate": 2.5313064208435423e-05,
      "loss": 0.0402,
      "step": 730
    },
    {
      "epoch": 6.91588785046729,
      "grad_norm": 0.3601996898651123,
      "learning_rate": 2.3980610173696255e-05,
      "loss": 0.0353,
      "step": 740
    },
    {
      "epoch": 7.009345794392523,
      "grad_norm": 0.4373137056827545,
      "learning_rate": 2.2673031811653034e-05,
      "loss": 0.0515,
      "step": 750
    },
    {
      "epoch": 7.102803738317757,
      "grad_norm": 0.4686773121356964,
      "learning_rate": 2.139157922430956e-05,
      "loss": 0.042,
      "step": 760
    },
    {
      "epoch": 7.196261682242991,
      "grad_norm": 0.4924180507659912,
      "learning_rate": 2.01374775362883e-05,
      "loss": 0.0396,
      "step": 770
    },
    {
      "epoch": 7.289719626168225,
      "grad_norm": 0.3904656767845154,
      "learning_rate": 1.8911925723557806e-05,
      "loss": 0.0399,
      "step": 780
    },
    {
      "epoch": 7.383177570093458,
      "grad_norm": 0.283329039812088,
      "learning_rate": 1.7716095467159393e-05,
      "loss": 0.0402,
      "step": 790
    },
    {
      "epoch": 7.4766355140186915,
      "grad_norm": 0.4036600887775421,
      "learning_rate": 1.6551130033028827e-05,
      "loss": 0.0373,
      "step": 800
    },
    {
      "epoch": 7.570093457943925,
      "grad_norm": 0.3786483407020569,
      "learning_rate": 1.541814317898425e-05,
      "loss": 0.0399,
      "step": 810
    },
    {
      "epoch": 7.663551401869158,
      "grad_norm": 0.34531161189079285,
      "learning_rate": 1.4318218089924962e-05,
      "loss": 0.04,
      "step": 820
    },
    {
      "epoch": 7.757009345794392,
      "grad_norm": 0.3333149254322052,
      "learning_rate": 1.3252406342259527e-05,
      "loss": 0.0328,
      "step": 830
    },
    {
      "epoch": 7.850467289719626,
      "grad_norm": 0.35911786556243896,
      "learning_rate": 1.2221726898552665e-05,
      "loss": 0.0359,
      "step": 840
    },
    {
      "epoch": 7.94392523364486,
      "grad_norm": 0.37050867080688477,
      "learning_rate": 1.122716513335262e-05,
      "loss": 0.0468,
      "step": 850
    },
    {
      "epoch": 8.037383177570094,
      "grad_norm": 0.27867504954338074,
      "learning_rate": 1.0269671891130123e-05,
      "loss": 0.0308,
      "step": 860
    },
    {
      "epoch": 8.130841121495328,
      "grad_norm": 0.36876147985458374,
      "learning_rate": 9.350162577229432e-06,
      "loss": 0.0379,
      "step": 870
    },
    {
      "epoch": 8.22429906542056,
      "grad_norm": 0.3114546835422516,
      "learning_rate": 8.46951628270098e-06,
      "loss": 0.0354,
      "step": 880
    },
    {
      "epoch": 8.317757009345794,
      "grad_norm": 0.26373207569122314,
      "learning_rate": 7.628574943851852e-06,
      "loss": 0.0296,
      "step": 890
    },
    {
      "epoch": 8.411214953271028,
      "grad_norm": 0.3879351019859314,
      "learning_rate": 6.82814253731801e-06,
      "loss": 0.0288,
      "step": 900
    },
    {
      "epoch": 8.504672897196262,
      "grad_norm": 0.3110659718513489,
      "learning_rate": 6.06898431142745e-06,
      "loss": 0.0293,
      "step": 910
    },
    {
      "epoch": 8.598130841121495,
      "grad_norm": 0.21370220184326172,
      "learning_rate": 5.351826054589393e-06,
      "loss": 0.0294,
      "step": 920
    },
    {
      "epoch": 8.69158878504673,
      "grad_norm": 0.304156094789505,
      "learning_rate": 4.677353401408974e-06,
      "loss": 0.0323,
      "step": 930
    },
    {
      "epoch": 8.785046728971963,
      "grad_norm": 0.30655941367149353,
      "learning_rate": 4.04621117719049e-06,
      "loss": 0.0356,
      "step": 940
    },
    {
      "epoch": 8.878504672897197,
      "grad_norm": 0.43658676743507385,
      "learning_rate": 3.459002781456344e-06,
      "loss": 0.0332,
      "step": 950
    },
    {
      "epoch": 8.97196261682243,
      "grad_norm": 0.486379474401474,
      "learning_rate": 2.9162896110707163e-06,
      "loss": 0.0336,
      "step": 960
    },
    {
      "epoch": 9.065420560747663,
      "grad_norm": 0.35680046677589417,
      "learning_rate": 2.418590523519687e-06,
      "loss": 0.0313,
      "step": 970
    },
    {
      "epoch": 9.158878504672897,
      "grad_norm": 0.2002888023853302,
      "learning_rate": 1.9663813408607845e-06,
      "loss": 0.0357,
      "step": 980
    },
    {
      "epoch": 9.25233644859813,
      "grad_norm": 0.27867501974105835,
      "learning_rate": 1.5600943948163527e-06,
      "loss": 0.0336,
      "step": 990
    },
    {
      "epoch": 9.345794392523365,
      "grad_norm": 0.2867244780063629,
      "learning_rate": 1.2001181134455475e-06,
      "loss": 0.0325,
      "step": 1000
    },
    {
      "epoch": 9.439252336448599,
      "grad_norm": 0.284397691488266,
      "learning_rate": 8.867966497901282e-07,
      "loss": 0.0278,
      "step": 1010
    },
    {
      "epoch": 9.532710280373832,
      "grad_norm": 0.24502161145210266,
      "learning_rate": 6.204295528491555e-07,
      "loss": 0.0377,
      "step": 1020
    },
    {
      "epoch": 9.626168224299064,
      "grad_norm": 0.3403957784175873,
      "learning_rate": 4.012714811970464e-07,
      "loss": 0.0294,
      "step": 1030
    },
    {
      "epoch": 9.719626168224298,
      "grad_norm": 0.311471551656723,
      "learning_rate": 2.295319595188805e-07,
      "loss": 0.0277,
      "step": 1040
    },
    {
      "epoch": 9.813084112149532,
      "grad_norm": 0.26186737418174744,
      "learning_rate": 1.0537517829562472e-07,
      "loss": 0.0281,
      "step": 1050
    },
    {
      "epoch": 9.906542056074766,
      "grad_norm": 0.20402012765407562,
      "learning_rate": 2.8919836830887392e-08,
      "loss": 0.0305,
      "step": 1060
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.8117493391036987,
      "learning_rate": 2.3902976920009423e-10,
      "loss": 0.0289,
      "step": 1070
    },
    {
      "epoch": 10.0,
      "step": 1070,
      "total_flos": 0.0,
      "train_loss": 0.07138736554395372,
      "train_runtime": 1117.3117,
      "train_samples_per_second": 46.513,
      "train_steps_per_second": 0.958
    }
  ],
  "logging_steps": 10,
  "max_steps": 1070,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 49,
  "trial_name": null,
  "trial_params": null
}