{
  "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": 2915.373291015625,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 152.2409,
      "step": 100
    },
    {
      "epoch": 0.035423308537017355,
      "grad_norm": 1864.051025390625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 221.198,
      "step": 200
    },
    {
      "epoch": 0.053134962805526036,
      "grad_norm": 1426.2816162109375,
      "learning_rate": 3e-06,
      "loss": 363.9534,
      "step": 300
    },
    {
      "epoch": 0.07084661707403471,
      "grad_norm": 573.576904296875,
      "learning_rate": 4.000000000000001e-06,
      "loss": 157.6679,
      "step": 400
    },
    {
      "epoch": 0.08855827134254339,
      "grad_norm": 23269.814453125,
      "learning_rate": 5e-06,
      "loss": 145.8333,
      "step": 500
    },
    {
      "epoch": 0.10626992561105207,
      "grad_norm": 378.4466247558594,
      "learning_rate": 6e-06,
      "loss": 225.834,
      "step": 600
    },
    {
      "epoch": 0.12398157987956075,
      "grad_norm": 560.156494140625,
      "learning_rate": 7e-06,
      "loss": 177.84,
      "step": 700
    },
    {
      "epoch": 0.14169323414806942,
      "grad_norm": 389.28936767578125,
      "learning_rate": 8.000000000000001e-06,
      "loss": 172.5718,
      "step": 800
    },
    {
      "epoch": 0.1594048884165781,
      "grad_norm": 290.60308837890625,
      "learning_rate": 9e-06,
      "loss": 118.532,
      "step": 900
    },
    {
      "epoch": 0.17711654268508678,
      "grad_norm": 708.8328857421875,
      "learning_rate": 1e-05,
      "loss": 102.5411,
      "step": 1000
    },
    {
      "epoch": 0.19482819695359546,
      "grad_norm": 733.8017578125,
      "learning_rate": 9.99695413509548e-06,
      "loss": 203.6356,
      "step": 1100
    },
    {
      "epoch": 0.21253985122210414,
      "grad_norm": 1235.48583984375,
      "learning_rate": 9.987820251299121e-06,
      "loss": 148.6089,
      "step": 1200
    },
    {
      "epoch": 0.23025150549061282,
      "grad_norm": 612.6553955078125,
      "learning_rate": 9.972609476841368e-06,
      "loss": 120.5858,
      "step": 1300
    },
    {
      "epoch": 0.2479631597591215,
      "grad_norm": 539.2703247070312,
      "learning_rate": 9.951340343707852e-06,
      "loss": 127.0271,
      "step": 1400
    },
    {
      "epoch": 0.26567481402763016,
      "grad_norm": 362.9602966308594,
      "learning_rate": 9.924038765061042e-06,
      "loss": 69.1379,
      "step": 1500
    },
    {
      "epoch": 0.28338646829613884,
      "grad_norm": 1124.81591796875,
      "learning_rate": 9.890738003669029e-06,
      "loss": 251.2696,
      "step": 1600
    },
    {
      "epoch": 0.3010981225646475,
      "grad_norm": 2207.07421875,
      "learning_rate": 9.851478631379982e-06,
      "loss": 138.321,
      "step": 1700
    },
    {
      "epoch": 0.3188097768331562,
      "grad_norm": 341.0426940917969,
      "learning_rate": 9.806308479691595e-06,
      "loss": 56.6384,
      "step": 1800
    },
    {
      "epoch": 0.3365214311016649,
      "grad_norm": 735.7184448242188,
      "learning_rate": 9.755282581475769e-06,
      "loss": 130.0931,
      "step": 1900
    },
    {
      "epoch": 0.35423308537017356,
      "grad_norm": 549.526611328125,
      "learning_rate": 9.698463103929542e-06,
      "loss": 231.677,
      "step": 2000
    },
    {
      "epoch": 0.37194473963868224,
      "grad_norm": 607.9448852539062,
      "learning_rate": 9.635919272833938e-06,
      "loss": 97.9832,
      "step": 2100
    },
    {
      "epoch": 0.3896563939071909,
      "grad_norm": 666.7014770507812,
      "learning_rate": 9.567727288213005e-06,
      "loss": 138.039,
      "step": 2200
    },
    {
      "epoch": 0.4073680481756996,
      "grad_norm": 442.1828918457031,
      "learning_rate": 9.493970231495836e-06,
      "loss": 109.4226,
      "step": 2300
    },
    {
      "epoch": 0.4250797024442083,
      "grad_norm": 174.03724670410156,
      "learning_rate": 9.414737964294636e-06,
      "loss": 70.5078,
      "step": 2400
    },
    {
      "epoch": 0.44279135671271697,
      "grad_norm": 456.3837585449219,
      "learning_rate": 9.330127018922195e-06,
      "loss": 88.6896,
      "step": 2500
    },
    {
      "epoch": 0.46050301098122565,
      "grad_norm": 354.64373779296875,
      "learning_rate": 9.24024048078213e-06,
      "loss": 106.9264,
      "step": 2600
    },
    {
      "epoch": 0.4782146652497343,
      "grad_norm": 144.5575408935547,
      "learning_rate": 9.145187862775208e-06,
      "loss": 75.4347,
      "step": 2700
    },
    {
      "epoch": 0.495926319518243,
      "grad_norm": 458.1535339355469,
      "learning_rate": 9.045084971874738e-06,
      "loss": 108.9999,
      "step": 2800
    },
    {
      "epoch": 0.5136379737867517,
      "grad_norm": 1575.434814453125,
      "learning_rate": 8.94005376803361e-06,
      "loss": 111.8511,
      "step": 2900
    },
    {
      "epoch": 0.5313496280552603,
      "grad_norm": 302.79241943359375,
      "learning_rate": 8.83022221559489e-06,
      "loss": 172.8595,
      "step": 3000
    },
    {
      "epoch": 0.549061282323769,
      "grad_norm": 1860.42919921875,
      "learning_rate": 8.715724127386971e-06,
      "loss": 143.5497,
      "step": 3100
    },
    {
      "epoch": 0.5667729365922777,
      "grad_norm": 297.4872131347656,
      "learning_rate": 8.596699001693257e-06,
      "loss": 126.711,
      "step": 3200
    },
    {
      "epoch": 0.5844845908607864,
      "grad_norm": 2128.590087890625,
      "learning_rate": 8.473291852294986e-06,
      "loss": 72.2636,
      "step": 3300
    },
    {
      "epoch": 0.602196245129295,
      "grad_norm": 436.14593505859375,
      "learning_rate": 8.345653031794292e-06,
      "loss": 169.4072,
      "step": 3400
    },
    {
      "epoch": 0.6199078993978038,
      "grad_norm": 1767.817138671875,
      "learning_rate": 8.213938048432697e-06,
      "loss": 134.9774,
      "step": 3500
    },
    {
      "epoch": 0.6376195536663124,
      "grad_norm": 298.466796875,
      "learning_rate": 8.078307376628292e-06,
      "loss": 91.2486,
      "step": 3600
    },
    {
      "epoch": 0.6553312079348211,
      "grad_norm": 641.1651000976562,
      "learning_rate": 7.938926261462366e-06,
      "loss": 145.7362,
      "step": 3700
    },
    {
      "epoch": 0.6730428622033298,
      "grad_norm": 14466.2060546875,
      "learning_rate": 7.795964517353734e-06,
      "loss": 169.7186,
      "step": 3800
    },
    {
      "epoch": 0.6907545164718385,
      "grad_norm": 161.69337463378906,
      "learning_rate": 7.649596321166024e-06,
      "loss": 81.239,
      "step": 3900
    },
    {
      "epoch": 0.7084661707403471,
      "grad_norm": 474.57958984375,
      "learning_rate": 7.500000000000001e-06,
      "loss": 114.6176,
      "step": 4000
    },
    {
      "epoch": 0.7261778250088559,
      "grad_norm": 1847.68359375,
      "learning_rate": 7.347357813929455e-06,
      "loss": 111.0047,
      "step": 4100
    },
    {
      "epoch": 0.7438894792773645,
      "grad_norm": 1322.276611328125,
      "learning_rate": 7.191855733945388e-06,
      "loss": 91.6528,
      "step": 4200
    },
    {
      "epoch": 0.7616011335458732,
      "grad_norm": 1325.048583984375,
      "learning_rate": 7.033683215379002e-06,
      "loss": 94.0535,
      "step": 4300
    },
    {
      "epoch": 0.7793127878143818,
      "grad_norm": 382.36859130859375,
      "learning_rate": 6.873032967079562e-06,
      "loss": 74.0244,
      "step": 4400
    },
    {
      "epoch": 0.7970244420828906,
      "grad_norm": 876.57080078125,
      "learning_rate": 6.710100716628345e-06,
      "loss": 96.6452,
      "step": 4500
    },
    {
      "epoch": 0.8147360963513992,
      "grad_norm": 1393.6138916015625,
      "learning_rate": 6.545084971874738e-06,
      "loss": 118.802,
      "step": 4600
    },
    {
      "epoch": 0.832447750619908,
      "grad_norm": 406.4674377441406,
      "learning_rate": 6.378186779084996e-06,
      "loss": 55.8678,
      "step": 4700
    },
    {
      "epoch": 0.8501594048884166,
      "grad_norm": 1019.0061645507812,
      "learning_rate": 6.209609477998339e-06,
      "loss": 123.8752,
      "step": 4800
    },
    {
      "epoch": 0.8678710591569253,
      "grad_norm": 987.013916015625,
      "learning_rate": 6.039558454088796e-06,
      "loss": 78.0407,
      "step": 4900
    },
    {
      "epoch": 0.8855827134254339,
      "grad_norm": 849.7422485351562,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 93.2876,
      "step": 5000
    },
    {
      "epoch": 0.9032943676939427,
      "grad_norm": 451.16192626953125,
      "learning_rate": 5.695865504800328e-06,
      "loss": 99.8985,
      "step": 5100
    },
    {
      "epoch": 0.9210060219624513,
      "grad_norm": 1357.7943115234375,
      "learning_rate": 5.522642316338268e-06,
      "loss": 106.1713,
      "step": 5200
    },
    {
      "epoch": 0.9387176762309599,
      "grad_norm": 452.63299560546875,
      "learning_rate": 5.348782368720627e-06,
      "loss": 72.7826,
      "step": 5300
    },
    {
      "epoch": 0.9564293304994687,
      "grad_norm": 382.8103942871094,
      "learning_rate": 5.174497483512506e-06,
      "loss": 108.6426,
      "step": 5400
    },
    {
      "epoch": 0.9741409847679773,
      "grad_norm": 589.746826171875,
      "learning_rate": 5e-06,
      "loss": 93.5595,
      "step": 5500
    },
    {
      "epoch": 0.991852639036486,
      "grad_norm": 719.3682250976562,
      "learning_rate": 4.825502516487497e-06,
      "loss": 79.8001,
      "step": 5600
    },
    {
      "epoch": 1.0,
      "eval_loss": 260.0443420410156,
      "eval_runtime": 57.7368,
      "eval_samples_per_second": 43.473,
      "eval_steps_per_second": 10.877,
      "step": 5646
    },
    {
      "epoch": 1.0095642933049946,
      "grad_norm": 344.79461669921875,
      "learning_rate": 4.651217631279374e-06,
      "loss": 87.4869,
      "step": 5700
    },
    {
      "epoch": 1.0272759475735034,
      "grad_norm": 387.0537414550781,
      "learning_rate": 4.477357683661734e-06,
      "loss": 70.2096,
      "step": 5800
    },
    {
      "epoch": 1.0449876018420121,
      "grad_norm": 1872.5281982421875,
      "learning_rate": 4.304134495199675e-06,
      "loss": 70.9989,
      "step": 5900
    },
    {
      "epoch": 1.0626992561105206,
      "grad_norm": 941.2898559570312,
      "learning_rate": 4.131759111665349e-06,
      "loss": 83.4253,
      "step": 6000
    },
    {
      "epoch": 1.0804109103790294,
      "grad_norm": 2921.8916015625,
      "learning_rate": 3.960441545911205e-06,
      "loss": 74.6986,
      "step": 6100
    },
    {
      "epoch": 1.098122564647538,
      "grad_norm": 317.5779113769531,
      "learning_rate": 3.790390522001662e-06,
      "loss": 56.6294,
      "step": 6200
    },
    {
      "epoch": 1.1158342189160468,
      "grad_norm": 610.6322021484375,
      "learning_rate": 3.6218132209150047e-06,
      "loss": 96.5304,
      "step": 6300
    },
    {
      "epoch": 1.1335458731845554,
      "grad_norm": 536.9466552734375,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 88.2391,
      "step": 6400
    },
    {
      "epoch": 1.151257527453064,
      "grad_norm": 707.88330078125,
      "learning_rate": 3.289899283371657e-06,
      "loss": 88.9819,
      "step": 6500
    },
    {
      "epoch": 1.1689691817215728,
      "grad_norm": 2059.515380859375,
      "learning_rate": 3.12696703292044e-06,
      "loss": 57.1338,
      "step": 6600
    },
    {
      "epoch": 1.1866808359900816,
      "grad_norm": 310.9707946777344,
      "learning_rate": 2.966316784621e-06,
      "loss": 107.5339,
      "step": 6700
    },
    {
      "epoch": 1.20439249025859,
      "grad_norm": 511.1593933105469,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 71.3359,
      "step": 6800
    },
    {
      "epoch": 1.2221041445270988,
      "grad_norm": 1355.5145263671875,
      "learning_rate": 2.6526421860705474e-06,
      "loss": 62.7688,
      "step": 6900
    },
    {
      "epoch": 1.2398157987956075,
      "grad_norm": 180.27313232421875,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 138.8191,
      "step": 7000
    },
    {
      "epoch": 1.257527453064116,
      "grad_norm": 477.2627868652344,
      "learning_rate": 2.3504036788339763e-06,
      "loss": 68.2528,
      "step": 7100
    },
    {
      "epoch": 1.2752391073326248,
      "grad_norm": 62.73986053466797,
      "learning_rate": 2.204035482646267e-06,
      "loss": 54.5104,
      "step": 7200
    },
    {
      "epoch": 1.2929507616011335,
      "grad_norm": 164.81520080566406,
      "learning_rate": 2.061073738537635e-06,
      "loss": 103.5922,
      "step": 7300
    },
    {
      "epoch": 1.3106624158696423,
      "grad_norm": 962.6326293945312,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 61.3437,
      "step": 7400
    },
    {
      "epoch": 1.328374070138151,
      "grad_norm": 413.64935302734375,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 102.4006,
      "step": 7500
    },
    {
      "epoch": 1.3460857244066595,
      "grad_norm": 271.7372741699219,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 92.6463,
      "step": 7600
    },
    {
      "epoch": 1.3637973786751683,
      "grad_norm": 776.993896484375,
      "learning_rate": 1.5267081477050132e-06,
      "loss": 131.4324,
      "step": 7700
    },
    {
      "epoch": 1.381509032943677,
      "grad_norm": 412.1511535644531,
      "learning_rate": 1.4033009983067454e-06,
      "loss": 55.6395,
      "step": 7800
    },
    {
      "epoch": 1.3992206872121855,
      "grad_norm": 664.7413940429688,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 137.3694,
      "step": 7900
    },
    {
      "epoch": 1.4169323414806942,
      "grad_norm": 350.5196228027344,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 107.8133,
      "step": 8000
    },
    {
      "epoch": 1.434643995749203,
      "grad_norm": 445.6170349121094,
      "learning_rate": 1.0599462319663906e-06,
      "loss": 51.7356,
      "step": 8100
    },
    {
      "epoch": 1.4523556500177117,
      "grad_norm": 953.5679931640625,
      "learning_rate": 9.549150281252633e-07,
      "loss": 84.7836,
      "step": 8200
    },
    {
      "epoch": 1.4700673042862205,
      "grad_norm": 952.9735107421875,
      "learning_rate": 8.54812137224792e-07,
      "loss": 92.5589,
      "step": 8300
    },
    {
      "epoch": 1.487778958554729,
      "grad_norm": 492.47198486328125,
      "learning_rate": 7.597595192178702e-07,
      "loss": 55.5006,
      "step": 8400
    },
    {
      "epoch": 1.5054906128232377,
      "grad_norm": 1805.4322509765625,
      "learning_rate": 6.698729810778065e-07,
      "loss": 96.7545,
      "step": 8500
    },
    {
      "epoch": 1.5232022670917464,
      "grad_norm": 300.4207458496094,
      "learning_rate": 5.852620357053651e-07,
      "loss": 86.5161,
      "step": 8600
    },
    {
      "epoch": 1.540913921360255,
      "grad_norm": 1440.614990234375,
      "learning_rate": 5.06029768504166e-07,
      "loss": 78.0504,
      "step": 8700
    },
    {
      "epoch": 1.5586255756287637,
      "grad_norm": 405.5276184082031,
      "learning_rate": 4.322727117869951e-07,
      "loss": 76.5312,
      "step": 8800
    },
    {
      "epoch": 1.5763372298972724,
      "grad_norm": 950.4790649414062,
      "learning_rate": 3.6408072716606346e-07,
      "loss": 89.3738,
      "step": 8900
    },
    {
      "epoch": 1.594048884165781,
      "grad_norm": 64.55994415283203,
      "learning_rate": 3.015368960704584e-07,
      "loss": 97.9065,
      "step": 9000
    },
    {
      "epoch": 1.61176053843429,
      "grad_norm": 259.1815490722656,
      "learning_rate": 2.447174185242324e-07,
      "loss": 93.1754,
      "step": 9100
    },
    {
      "epoch": 1.6294721927027984,
      "grad_norm": 366.2008056640625,
      "learning_rate": 1.9369152030840553e-07,
      "loss": 73.6778,
      "step": 9200
    },
    {
      "epoch": 1.6471838469713072,
      "grad_norm": 383.7484436035156,
      "learning_rate": 1.4852136862001766e-07,
      "loss": 48.6865,
      "step": 9300
    },
    {
      "epoch": 1.664895501239816,
      "grad_norm": 392.061767578125,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 118.0324,
      "step": 9400
    },
    {
      "epoch": 1.6826071555083244,
      "grad_norm": 8477.83984375,
      "learning_rate": 7.59612349389599e-08,
      "loss": 109.5857,
      "step": 9500
    },
    {
      "epoch": 1.7003188097768331,
      "grad_norm": 452.080078125,
      "learning_rate": 4.865965629214819e-08,
      "loss": 95.1972,
      "step": 9600
    },
    {
      "epoch": 1.7180304640453419,
      "grad_norm": 142.1215057373047,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 59.5104,
      "step": 9700
    },
    {
      "epoch": 1.7357421183138504,
      "grad_norm": 450.3463134765625,
      "learning_rate": 1.2179748700879013e-08,
      "loss": 79.6601,
      "step": 9800
    },
    {
      "epoch": 1.7534537725823593,
      "grad_norm": 964.9075927734375,
      "learning_rate": 3.0458649045211897e-09,
      "loss": 82.6603,
      "step": 9900
    },
    {
      "epoch": 1.7711654268508679,
      "grad_norm": 203.1498260498047,
      "learning_rate": 0.0,
      "loss": 94.6798,
      "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
}