{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.7711654268508679,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017711654268508677,
      "grad_norm": 1103.915771484375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 387.058,
      "step": 100
    },
    {
      "epoch": 0.035423308537017355,
      "grad_norm": 14950.353515625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 706.6331,
      "step": 200
    },
    {
      "epoch": 0.053134962805526036,
      "grad_norm": 500.2298278808594,
      "learning_rate": 3e-06,
      "loss": 183.6831,
      "step": 300
    },
    {
      "epoch": 0.07084661707403471,
      "grad_norm": 874.4174194335938,
      "learning_rate": 4.000000000000001e-06,
      "loss": 374.0061,
      "step": 400
    },
    {
      "epoch": 0.08855827134254339,
      "grad_norm": 772.3428344726562,
      "learning_rate": 5e-06,
      "loss": 253.8809,
      "step": 500
    },
    {
      "epoch": 0.10626992561105207,
      "grad_norm": 241.94956970214844,
      "learning_rate": 6e-06,
      "loss": 148.413,
      "step": 600
    },
    {
      "epoch": 0.12398157987956075,
      "grad_norm": 705.2841186523438,
      "learning_rate": 7e-06,
      "loss": 350.3304,
      "step": 700
    },
    {
      "epoch": 0.14169323414806942,
      "grad_norm": 637.541015625,
      "learning_rate": 8.000000000000001e-06,
      "loss": 215.5345,
      "step": 800
    },
    {
      "epoch": 0.1594048884165781,
      "grad_norm": 637.4180297851562,
      "learning_rate": 9e-06,
      "loss": 260.8833,
      "step": 900
    },
    {
      "epoch": 0.17711654268508678,
      "grad_norm": 2717.751953125,
      "learning_rate": 1e-05,
      "loss": 250.8861,
      "step": 1000
    },
    {
      "epoch": 0.19482819695359546,
      "grad_norm": 513.18896484375,
      "learning_rate": 9.99695413509548e-06,
      "loss": 277.5799,
      "step": 1100
    },
    {
      "epoch": 0.21253985122210414,
      "grad_norm": 979.9168090820312,
      "learning_rate": 9.987820251299121e-06,
      "loss": 140.9566,
      "step": 1200
    },
    {
      "epoch": 0.23025150549061282,
      "grad_norm": 668.2918701171875,
      "learning_rate": 9.972609476841368e-06,
      "loss": 86.1729,
      "step": 1300
    },
    {
      "epoch": 0.2479631597591215,
      "grad_norm": 673.4871215820312,
      "learning_rate": 9.951340343707852e-06,
      "loss": 176.5392,
      "step": 1400
    },
    {
      "epoch": 0.26567481402763016,
      "grad_norm": 2616.958984375,
      "learning_rate": 9.924038765061042e-06,
      "loss": 116.1355,
      "step": 1500
    },
    {
      "epoch": 0.28338646829613884,
      "grad_norm": 335.2275695800781,
      "learning_rate": 9.890738003669029e-06,
      "loss": 100.1151,
      "step": 1600
    },
    {
      "epoch": 0.3010981225646475,
      "grad_norm": 1734.4193115234375,
      "learning_rate": 9.851478631379982e-06,
      "loss": 166.2296,
      "step": 1700
    },
    {
      "epoch": 0.3188097768331562,
      "grad_norm": 1094.8076171875,
      "learning_rate": 9.806308479691595e-06,
      "loss": 179.6621,
      "step": 1800
    },
    {
      "epoch": 0.3365214311016649,
      "grad_norm": 860.8914794921875,
      "learning_rate": 9.755282581475769e-06,
      "loss": 87.479,
      "step": 1900
    },
    {
      "epoch": 0.35423308537017356,
      "grad_norm": 388.6125183105469,
      "learning_rate": 9.698463103929542e-06,
      "loss": 124.4702,
      "step": 2000
    },
    {
      "epoch": 0.37194473963868224,
      "grad_norm": 382.7939758300781,
      "learning_rate": 9.635919272833938e-06,
      "loss": 205.2312,
      "step": 2100
    },
    {
      "epoch": 0.3896563939071909,
      "grad_norm": 210.30653381347656,
      "learning_rate": 9.567727288213005e-06,
      "loss": 158.1804,
      "step": 2200
    },
    {
      "epoch": 0.4073680481756996,
      "grad_norm": 494.79364013671875,
      "learning_rate": 9.493970231495836e-06,
      "loss": 107.7146,
      "step": 2300
    },
    {
      "epoch": 0.4250797024442083,
      "grad_norm": 389.76348876953125,
      "learning_rate": 9.414737964294636e-06,
      "loss": 114.4283,
      "step": 2400
    },
    {
      "epoch": 0.44279135671271697,
      "grad_norm": 3284.485595703125,
      "learning_rate": 9.330127018922195e-06,
      "loss": 89.9879,
      "step": 2500
    },
    {
      "epoch": 0.46050301098122565,
      "grad_norm": 340.7903137207031,
      "learning_rate": 9.24024048078213e-06,
      "loss": 140.386,
      "step": 2600
    },
    {
      "epoch": 0.4782146652497343,
      "grad_norm": 1951.8299560546875,
      "learning_rate": 9.145187862775208e-06,
      "loss": 279.5895,
      "step": 2700
    },
    {
      "epoch": 0.495926319518243,
      "grad_norm": 200.66510009765625,
      "learning_rate": 9.045084971874738e-06,
      "loss": 134.429,
      "step": 2800
    },
    {
      "epoch": 0.5136379737867517,
      "grad_norm": 8959.56640625,
      "learning_rate": 8.94005376803361e-06,
      "loss": 159.9825,
      "step": 2900
    },
    {
      "epoch": 0.5313496280552603,
      "grad_norm": 3806.5439453125,
      "learning_rate": 8.83022221559489e-06,
      "loss": 164.7516,
      "step": 3000
    },
    {
      "epoch": 0.549061282323769,
      "grad_norm": 743.8880004882812,
      "learning_rate": 8.715724127386971e-06,
      "loss": 83.0674,
      "step": 3100
    },
    {
      "epoch": 0.5667729365922777,
      "grad_norm": 372.5606689453125,
      "learning_rate": 8.596699001693257e-06,
      "loss": 80.553,
      "step": 3200
    },
    {
      "epoch": 0.5844845908607864,
      "grad_norm": 6004.2822265625,
      "learning_rate": 8.473291852294986e-06,
      "loss": 95.1305,
      "step": 3300
    },
    {
      "epoch": 0.602196245129295,
      "grad_norm": 676.382080078125,
      "learning_rate": 8.345653031794292e-06,
      "loss": 153.7045,
      "step": 3400
    },
    {
      "epoch": 0.6199078993978038,
      "grad_norm": 493.2138671875,
      "learning_rate": 8.213938048432697e-06,
      "loss": 131.3091,
      "step": 3500
    },
    {
      "epoch": 0.6376195536663124,
      "grad_norm": 946.7998657226562,
      "learning_rate": 8.078307376628292e-06,
      "loss": 100.9372,
      "step": 3600
    },
    {
      "epoch": 0.6553312079348211,
      "grad_norm": 1105.7960205078125,
      "learning_rate": 7.938926261462366e-06,
      "loss": 153.4672,
      "step": 3700
    },
    {
      "epoch": 0.6730428622033298,
      "grad_norm": 372.66217041015625,
      "learning_rate": 7.795964517353734e-06,
      "loss": 141.8139,
      "step": 3800
    },
    {
      "epoch": 0.6907545164718385,
      "grad_norm": 895.3282470703125,
      "learning_rate": 7.649596321166024e-06,
      "loss": 183.8056,
      "step": 3900
    },
    {
      "epoch": 0.7084661707403471,
      "grad_norm": 750.4679565429688,
      "learning_rate": 7.500000000000001e-06,
      "loss": 79.4339,
      "step": 4000
    },
    {
      "epoch": 0.7261778250088559,
      "grad_norm": 718.4357299804688,
      "learning_rate": 7.347357813929455e-06,
      "loss": 137.0742,
      "step": 4100
    },
    {
      "epoch": 0.7438894792773645,
      "grad_norm": 426.361083984375,
      "learning_rate": 7.191855733945388e-06,
      "loss": 190.6817,
      "step": 4200
    },
    {
      "epoch": 0.7616011335458732,
      "grad_norm": 923.95947265625,
      "learning_rate": 7.033683215379002e-06,
      "loss": 162.6752,
      "step": 4300
    },
    {
      "epoch": 0.7793127878143818,
      "grad_norm": 596.4622802734375,
      "learning_rate": 6.873032967079562e-06,
      "loss": 205.6588,
      "step": 4400
    },
    {
      "epoch": 0.7970244420828906,
      "grad_norm": 564.4142456054688,
      "learning_rate": 6.710100716628345e-06,
      "loss": 103.3634,
      "step": 4500
    },
    {
      "epoch": 0.8147360963513992,
      "grad_norm": 1955.7783203125,
      "learning_rate": 6.545084971874738e-06,
      "loss": 162.0502,
      "step": 4600
    },
    {
      "epoch": 0.832447750619908,
      "grad_norm": 268.0696716308594,
      "learning_rate": 6.378186779084996e-06,
      "loss": 107.0621,
      "step": 4700
    },
    {
      "epoch": 0.8501594048884166,
      "grad_norm": 342.3700866699219,
      "learning_rate": 6.209609477998339e-06,
      "loss": 170.9172,
      "step": 4800
    },
    {
      "epoch": 0.8678710591569253,
      "grad_norm": 301.4243469238281,
      "learning_rate": 6.039558454088796e-06,
      "loss": 116.3496,
      "step": 4900
    },
    {
      "epoch": 0.8855827134254339,
      "grad_norm": 263.0123291015625,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 68.4552,
      "step": 5000
    },
    {
      "epoch": 0.9032943676939427,
      "grad_norm": 217.31036376953125,
      "learning_rate": 5.695865504800328e-06,
      "loss": 94.4736,
      "step": 5100
    },
    {
      "epoch": 0.9210060219624513,
      "grad_norm": 489.5201416015625,
      "learning_rate": 5.522642316338268e-06,
      "loss": 193.8479,
      "step": 5200
    },
    {
      "epoch": 0.9387176762309599,
      "grad_norm": 2904.69140625,
      "learning_rate": 5.348782368720627e-06,
      "loss": 84.7795,
      "step": 5300
    },
    {
      "epoch": 0.9564293304994687,
      "grad_norm": 260.4606628417969,
      "learning_rate": 5.174497483512506e-06,
      "loss": 128.8338,
      "step": 5400
    },
    {
      "epoch": 0.9741409847679773,
      "grad_norm": 7584.078125,
      "learning_rate": 5e-06,
      "loss": 149.5825,
      "step": 5500
    },
    {
      "epoch": 0.991852639036486,
      "grad_norm": 299.5451965332031,
      "learning_rate": 4.825502516487497e-06,
      "loss": 108.1543,
      "step": 5600
    },
    {
      "epoch": 1.0,
      "eval_loss": 297.8221740722656,
      "eval_runtime": 18.2951,
      "eval_samples_per_second": 137.195,
      "eval_steps_per_second": 34.326,
      "step": 5646
    },
    {
      "epoch": 1.0095642933049946,
      "grad_norm": 384.34698486328125,
      "learning_rate": 4.651217631279374e-06,
      "loss": 139.0317,
      "step": 5700
    },
    {
      "epoch": 1.0272759475735034,
      "grad_norm": 363.44384765625,
      "learning_rate": 4.477357683661734e-06,
      "loss": 91.1543,
      "step": 5800
    },
    {
      "epoch": 1.0449876018420121,
      "grad_norm": 712.39404296875,
      "learning_rate": 4.304134495199675e-06,
      "loss": 105.6641,
      "step": 5900
    },
    {
      "epoch": 1.0626992561105206,
      "grad_norm": 587.1605224609375,
      "learning_rate": 4.131759111665349e-06,
      "loss": 143.8675,
      "step": 6000
    },
    {
      "epoch": 1.0804109103790294,
      "grad_norm": 151.92108154296875,
      "learning_rate": 3.960441545911205e-06,
      "loss": 73.3581,
      "step": 6100
    },
    {
      "epoch": 1.098122564647538,
      "grad_norm": 362.7684326171875,
      "learning_rate": 3.790390522001662e-06,
      "loss": 139.9421,
      "step": 6200
    },
    {
      "epoch": 1.1158342189160468,
      "grad_norm": 404.06536865234375,
      "learning_rate": 3.6218132209150047e-06,
      "loss": 82.1686,
      "step": 6300
    },
    {
      "epoch": 1.1335458731845554,
      "grad_norm": 226.1641082763672,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 195.2268,
      "step": 6400
    },
    {
      "epoch": 1.151257527453064,
      "grad_norm": 1199.0345458984375,
      "learning_rate": 3.289899283371657e-06,
      "loss": 123.0975,
      "step": 6500
    },
    {
      "epoch": 1.1689691817215728,
      "grad_norm": 1147.0032958984375,
      "learning_rate": 3.12696703292044e-06,
      "loss": 97.2361,
      "step": 6600
    },
    {
      "epoch": 1.1866808359900816,
      "grad_norm": 211.77349853515625,
      "learning_rate": 2.966316784621e-06,
      "loss": 158.7633,
      "step": 6700
    },
    {
      "epoch": 1.20439249025859,
      "grad_norm": 505.9942626953125,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 179.0317,
      "step": 6800
    },
    {
      "epoch": 1.2221041445270988,
      "grad_norm": 592.3914184570312,
      "learning_rate": 2.6526421860705474e-06,
      "loss": 109.1925,
      "step": 6900
    },
    {
      "epoch": 1.2398157987956075,
      "grad_norm": 316.92071533203125,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 114.6693,
      "step": 7000
    },
    {
      "epoch": 1.257527453064116,
      "grad_norm": 208.8818817138672,
      "learning_rate": 2.3504036788339763e-06,
      "loss": 96.6279,
      "step": 7100
    },
    {
      "epoch": 1.2752391073326248,
      "grad_norm": 573.2861938476562,
      "learning_rate": 2.204035482646267e-06,
      "loss": 130.7464,
      "step": 7200
    },
    {
      "epoch": 1.2929507616011335,
      "grad_norm": 27379.625,
      "learning_rate": 2.061073738537635e-06,
      "loss": 140.7592,
      "step": 7300
    },
    {
      "epoch": 1.3106624158696423,
      "grad_norm": 185.75172424316406,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 102.8975,
      "step": 7400
    },
    {
      "epoch": 1.328374070138151,
      "grad_norm": 128.01487731933594,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 120.6439,
      "step": 7500
    },
    {
      "epoch": 1.3460857244066595,
      "grad_norm": 757.961181640625,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 145.0263,
      "step": 7600
    },
    {
      "epoch": 1.3637973786751683,
      "grad_norm": 355.3944091796875,
      "learning_rate": 1.5267081477050132e-06,
      "loss": 98.2204,
      "step": 7700
    },
    {
      "epoch": 1.381509032943677,
      "grad_norm": 413.9019470214844,
      "learning_rate": 1.4033009983067454e-06,
      "loss": 149.2902,
      "step": 7800
    },
    {
      "epoch": 1.3992206872121855,
      "grad_norm": 376.12225341796875,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 83.9799,
      "step": 7900
    },
    {
      "epoch": 1.4169323414806942,
      "grad_norm": 922.9308471679688,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 76.7685,
      "step": 8000
    },
    {
      "epoch": 1.434643995749203,
      "grad_norm": 351.0088806152344,
      "learning_rate": 1.0599462319663906e-06,
      "loss": 80.1955,
      "step": 8100
    },
    {
      "epoch": 1.4523556500177117,
      "grad_norm": 1167.3990478515625,
      "learning_rate": 9.549150281252633e-07,
      "loss": 134.8115,
      "step": 8200
    },
    {
      "epoch": 1.4700673042862205,
      "grad_norm": 305.848876953125,
      "learning_rate": 8.54812137224792e-07,
      "loss": 134.6639,
      "step": 8300
    },
    {
      "epoch": 1.487778958554729,
      "grad_norm": 369.099853515625,
      "learning_rate": 7.597595192178702e-07,
      "loss": 80.5874,
      "step": 8400
    },
    {
      "epoch": 1.5054906128232377,
      "grad_norm": 249.34478759765625,
      "learning_rate": 6.698729810778065e-07,
      "loss": 124.899,
      "step": 8500
    },
    {
      "epoch": 1.5232022670917464,
      "grad_norm": 459.1353759765625,
      "learning_rate": 5.852620357053651e-07,
      "loss": 85.7345,
      "step": 8600
    },
    {
      "epoch": 1.540913921360255,
      "grad_norm": 995.8704833984375,
      "learning_rate": 5.06029768504166e-07,
      "loss": 73.1755,
      "step": 8700
    },
    {
      "epoch": 1.5586255756287637,
      "grad_norm": 1414.3692626953125,
      "learning_rate": 4.322727117869951e-07,
      "loss": 135.3178,
      "step": 8800
    },
    {
      "epoch": 1.5763372298972724,
      "grad_norm": 1469.5264892578125,
      "learning_rate": 3.6408072716606346e-07,
      "loss": 90.5656,
      "step": 8900
    },
    {
      "epoch": 1.594048884165781,
      "grad_norm": 741.5259399414062,
      "learning_rate": 3.015368960704584e-07,
      "loss": 85.5567,
      "step": 9000
    },
    {
      "epoch": 1.61176053843429,
      "grad_norm": 730.26025390625,
      "learning_rate": 2.447174185242324e-07,
      "loss": 91.601,
      "step": 9100
    },
    {
      "epoch": 1.6294721927027984,
      "grad_norm": 227.96351623535156,
      "learning_rate": 1.9369152030840553e-07,
      "loss": 66.1068,
      "step": 9200
    },
    {
      "epoch": 1.6471838469713072,
      "grad_norm": 358.64703369140625,
      "learning_rate": 1.4852136862001766e-07,
      "loss": 72.2795,
      "step": 9300
    },
    {
      "epoch": 1.664895501239816,
      "grad_norm": 464.4366149902344,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 190.3921,
      "step": 9400
    },
    {
      "epoch": 1.6826071555083244,
      "grad_norm": 749.0379028320312,
      "learning_rate": 7.59612349389599e-08,
      "loss": 150.7065,
      "step": 9500
    },
    {
      "epoch": 1.7003188097768331,
      "grad_norm": 188.99282836914062,
      "learning_rate": 4.865965629214819e-08,
      "loss": 69.5602,
      "step": 9600
    },
    {
      "epoch": 1.7180304640453419,
      "grad_norm": 630.560791015625,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 120.8536,
      "step": 9700
    },
    {
      "epoch": 1.7357421183138504,
      "grad_norm": 862.2304077148438,
      "learning_rate": 1.2179748700879013e-08,
      "loss": 90.8576,
      "step": 9800
    },
    {
      "epoch": 1.7534537725823593,
      "grad_norm": 317.0333251953125,
      "learning_rate": 3.0458649045211897e-09,
      "loss": 224.1108,
      "step": 9900
    },
    {
      "epoch": 1.7711654268508679,
      "grad_norm": 965.94580078125,
      "learning_rate": 0.0,
      "loss": 83.882,
      "step": 10000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "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": 4,
  "trial_name": null,
  "trial_params": null
}