{
  "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": 5928.86572265625,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 167.7101,
      "step": 100
    },
    {
      "epoch": 0.035423308537017355,
      "grad_norm": 1039.6475830078125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 150.4972,
      "step": 200
    },
    {
      "epoch": 0.053134962805526036,
      "grad_norm": 2417.1572265625,
      "learning_rate": 3e-06,
      "loss": 118.4022,
      "step": 300
    },
    {
      "epoch": 0.07084661707403471,
      "grad_norm": 658.3015747070312,
      "learning_rate": 4.000000000000001e-06,
      "loss": 81.2197,
      "step": 400
    },
    {
      "epoch": 0.08855827134254339,
      "grad_norm": 11006.369140625,
      "learning_rate": 5e-06,
      "loss": 191.8971,
      "step": 500
    },
    {
      "epoch": 0.10626992561105207,
      "grad_norm": 387.76507568359375,
      "learning_rate": 6e-06,
      "loss": 238.5224,
      "step": 600
    },
    {
      "epoch": 0.12398157987956075,
      "grad_norm": 520.3596801757812,
      "learning_rate": 7e-06,
      "loss": 140.7788,
      "step": 700
    },
    {
      "epoch": 0.14169323414806942,
      "grad_norm": 711.9447631835938,
      "learning_rate": 8.000000000000001e-06,
      "loss": 240.6655,
      "step": 800
    },
    {
      "epoch": 0.1594048884165781,
      "grad_norm": 685.3497314453125,
      "learning_rate": 9e-06,
      "loss": 142.9791,
      "step": 900
    },
    {
      "epoch": 0.17711654268508678,
      "grad_norm": 0.0,
      "learning_rate": 1e-05,
      "loss": 98.8305,
      "step": 1000
    },
    {
      "epoch": 0.19482819695359546,
      "grad_norm": 672.6516723632812,
      "learning_rate": 9.99695413509548e-06,
      "loss": 93.7172,
      "step": 1100
    },
    {
      "epoch": 0.21253985122210414,
      "grad_norm": 670.732666015625,
      "learning_rate": 9.987820251299121e-06,
      "loss": 108.1393,
      "step": 1200
    },
    {
      "epoch": 0.23025150549061282,
      "grad_norm": 2993.84228515625,
      "learning_rate": 9.972609476841368e-06,
      "loss": 112.0507,
      "step": 1300
    },
    {
      "epoch": 0.2479631597591215,
      "grad_norm": 494.7082214355469,
      "learning_rate": 9.951340343707852e-06,
      "loss": 87.2083,
      "step": 1400
    },
    {
      "epoch": 0.26567481402763016,
      "grad_norm": 9774.228515625,
      "learning_rate": 9.924038765061042e-06,
      "loss": 83.2371,
      "step": 1500
    },
    {
      "epoch": 0.28338646829613884,
      "grad_norm": 1228.3587646484375,
      "learning_rate": 9.890738003669029e-06,
      "loss": 132.57,
      "step": 1600
    },
    {
      "epoch": 0.3010981225646475,
      "grad_norm": 1199.085693359375,
      "learning_rate": 9.851478631379982e-06,
      "loss": 84.5867,
      "step": 1700
    },
    {
      "epoch": 0.3188097768331562,
      "grad_norm": 4537.9521484375,
      "learning_rate": 9.806308479691595e-06,
      "loss": 96.3146,
      "step": 1800
    },
    {
      "epoch": 0.3365214311016649,
      "grad_norm": 1682.6268310546875,
      "learning_rate": 9.755282581475769e-06,
      "loss": 122.0458,
      "step": 1900
    },
    {
      "epoch": 0.35423308537017356,
      "grad_norm": 474.109619140625,
      "learning_rate": 9.698463103929542e-06,
      "loss": 82.0088,
      "step": 2000
    },
    {
      "epoch": 0.37194473963868224,
      "grad_norm": 550.49072265625,
      "learning_rate": 9.635919272833938e-06,
      "loss": 102.8292,
      "step": 2100
    },
    {
      "epoch": 0.3896563939071909,
      "grad_norm": 5476.99560546875,
      "learning_rate": 9.567727288213005e-06,
      "loss": 146.4771,
      "step": 2200
    },
    {
      "epoch": 0.4073680481756996,
      "grad_norm": 962.66064453125,
      "learning_rate": 9.493970231495836e-06,
      "loss": 113.7458,
      "step": 2300
    },
    {
      "epoch": 0.4250797024442083,
      "grad_norm": 176.8959197998047,
      "learning_rate": 9.414737964294636e-06,
      "loss": 79.8068,
      "step": 2400
    },
    {
      "epoch": 0.44279135671271697,
      "grad_norm": 0.0,
      "learning_rate": 9.330127018922195e-06,
      "loss": 95.6945,
      "step": 2500
    },
    {
      "epoch": 0.46050301098122565,
      "grad_norm": 739.9681396484375,
      "learning_rate": 9.24024048078213e-06,
      "loss": 127.2645,
      "step": 2600
    },
    {
      "epoch": 0.4782146652497343,
      "grad_norm": 892.9599609375,
      "learning_rate": 9.145187862775208e-06,
      "loss": 67.8974,
      "step": 2700
    },
    {
      "epoch": 0.495926319518243,
      "grad_norm": 478.8652038574219,
      "learning_rate": 9.045084971874738e-06,
      "loss": 125.3699,
      "step": 2800
    },
    {
      "epoch": 0.5136379737867517,
      "grad_norm": 176.39854431152344,
      "learning_rate": 8.94005376803361e-06,
      "loss": 145.9999,
      "step": 2900
    },
    {
      "epoch": 0.5313496280552603,
      "grad_norm": 738.6134643554688,
      "learning_rate": 8.83022221559489e-06,
      "loss": 161.4359,
      "step": 3000
    },
    {
      "epoch": 0.549061282323769,
      "grad_norm": 389.0908203125,
      "learning_rate": 8.715724127386971e-06,
      "loss": 63.7037,
      "step": 3100
    },
    {
      "epoch": 0.5667729365922777,
      "grad_norm": 622.670654296875,
      "learning_rate": 8.596699001693257e-06,
      "loss": 66.3127,
      "step": 3200
    },
    {
      "epoch": 0.5844845908607864,
      "grad_norm": 332.5394592285156,
      "learning_rate": 8.473291852294986e-06,
      "loss": 78.8835,
      "step": 3300
    },
    {
      "epoch": 0.602196245129295,
      "grad_norm": 334.5383605957031,
      "learning_rate": 8.345653031794292e-06,
      "loss": 74.8007,
      "step": 3400
    },
    {
      "epoch": 0.6199078993978038,
      "grad_norm": 2048.822021484375,
      "learning_rate": 8.213938048432697e-06,
      "loss": 105.3306,
      "step": 3500
    },
    {
      "epoch": 0.6376195536663124,
      "grad_norm": 418.65618896484375,
      "learning_rate": 8.078307376628292e-06,
      "loss": 92.522,
      "step": 3600
    },
    {
      "epoch": 0.6553312079348211,
      "grad_norm": 588.8889770507812,
      "learning_rate": 7.938926261462366e-06,
      "loss": 93.5883,
      "step": 3700
    },
    {
      "epoch": 0.6730428622033298,
      "grad_norm": 213.2515411376953,
      "learning_rate": 7.795964517353734e-06,
      "loss": 81.18,
      "step": 3800
    },
    {
      "epoch": 0.6907545164718385,
      "grad_norm": 502.90399169921875,
      "learning_rate": 7.649596321166024e-06,
      "loss": 91.3637,
      "step": 3900
    },
    {
      "epoch": 0.7084661707403471,
      "grad_norm": 37754.41015625,
      "learning_rate": 7.500000000000001e-06,
      "loss": 84.6845,
      "step": 4000
    },
    {
      "epoch": 0.7261778250088559,
      "grad_norm": 620.5169677734375,
      "learning_rate": 7.347357813929455e-06,
      "loss": 76.8845,
      "step": 4100
    },
    {
      "epoch": 0.7438894792773645,
      "grad_norm": 607.29541015625,
      "learning_rate": 7.191855733945388e-06,
      "loss": 136.6091,
      "step": 4200
    },
    {
      "epoch": 0.7616011335458732,
      "grad_norm": 1504.0474853515625,
      "learning_rate": 7.033683215379002e-06,
      "loss": 87.2499,
      "step": 4300
    },
    {
      "epoch": 0.7793127878143818,
      "grad_norm": 351.63482666015625,
      "learning_rate": 6.873032967079562e-06,
      "loss": 70.2709,
      "step": 4400
    },
    {
      "epoch": 0.7970244420828906,
      "grad_norm": 608.7922973632812,
      "learning_rate": 6.710100716628345e-06,
      "loss": 60.6958,
      "step": 4500
    },
    {
      "epoch": 0.8147360963513992,
      "grad_norm": 501.0238037109375,
      "learning_rate": 6.545084971874738e-06,
      "loss": 78.5794,
      "step": 4600
    },
    {
      "epoch": 0.832447750619908,
      "grad_norm": 667.3199462890625,
      "learning_rate": 6.378186779084996e-06,
      "loss": 71.3733,
      "step": 4700
    },
    {
      "epoch": 0.8501594048884166,
      "grad_norm": 280.9075927734375,
      "learning_rate": 6.209609477998339e-06,
      "loss": 62.0313,
      "step": 4800
    },
    {
      "epoch": 0.8678710591569253,
      "grad_norm": 257.11114501953125,
      "learning_rate": 6.039558454088796e-06,
      "loss": 61.1799,
      "step": 4900
    },
    {
      "epoch": 0.8855827134254339,
      "grad_norm": 878.3767700195312,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 91.728,
      "step": 5000
    },
    {
      "epoch": 0.9032943676939427,
      "grad_norm": 1279.4716796875,
      "learning_rate": 5.695865504800328e-06,
      "loss": 78.0071,
      "step": 5100
    },
    {
      "epoch": 0.9210060219624513,
      "grad_norm": 360.426025390625,
      "learning_rate": 5.522642316338268e-06,
      "loss": 63.6748,
      "step": 5200
    },
    {
      "epoch": 0.9387176762309599,
      "grad_norm": 670.2398071289062,
      "learning_rate": 5.348782368720627e-06,
      "loss": 87.3999,
      "step": 5300
    },
    {
      "epoch": 0.9564293304994687,
      "grad_norm": 642.230712890625,
      "learning_rate": 5.174497483512506e-06,
      "loss": 100.3238,
      "step": 5400
    },
    {
      "epoch": 0.9741409847679773,
      "grad_norm": 462.1632080078125,
      "learning_rate": 5e-06,
      "loss": 95.8226,
      "step": 5500
    },
    {
      "epoch": 0.991852639036486,
      "grad_norm": 1005.881591796875,
      "learning_rate": 4.825502516487497e-06,
      "loss": 67.1857,
      "step": 5600
    },
    {
      "epoch": 1.0,
      "eval_loss": 292.22265625,
      "eval_runtime": 58.4771,
      "eval_samples_per_second": 42.923,
      "eval_steps_per_second": 10.739,
      "step": 5646
    },
    {
      "epoch": 1.0095642933049946,
      "grad_norm": 342.584716796875,
      "learning_rate": 4.651217631279374e-06,
      "loss": 61.0431,
      "step": 5700
    },
    {
      "epoch": 1.0272759475735034,
      "grad_norm": 10458.5888671875,
      "learning_rate": 4.477357683661734e-06,
      "loss": 100.5392,
      "step": 5800
    },
    {
      "epoch": 1.0449876018420121,
      "grad_norm": 266.7583923339844,
      "learning_rate": 4.304134495199675e-06,
      "loss": 40.7089,
      "step": 5900
    },
    {
      "epoch": 1.0626992561105206,
      "grad_norm": 156.3075408935547,
      "learning_rate": 4.131759111665349e-06,
      "loss": 49.5499,
      "step": 6000
    },
    {
      "epoch": 1.0804109103790294,
      "grad_norm": 255.9273223876953,
      "learning_rate": 3.960441545911205e-06,
      "loss": 70.7226,
      "step": 6100
    },
    {
      "epoch": 1.098122564647538,
      "grad_norm": 571.5280151367188,
      "learning_rate": 3.790390522001662e-06,
      "loss": 91.4461,
      "step": 6200
    },
    {
      "epoch": 1.1158342189160468,
      "grad_norm": 947.069091796875,
      "learning_rate": 3.6218132209150047e-06,
      "loss": 83.3039,
      "step": 6300
    },
    {
      "epoch": 1.1335458731845554,
      "grad_norm": 182.46408081054688,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 53.0065,
      "step": 6400
    },
    {
      "epoch": 1.151257527453064,
      "grad_norm": 3747.530517578125,
      "learning_rate": 3.289899283371657e-06,
      "loss": 55.4958,
      "step": 6500
    },
    {
      "epoch": 1.1689691817215728,
      "grad_norm": 186.1503448486328,
      "learning_rate": 3.12696703292044e-06,
      "loss": 76.359,
      "step": 6600
    },
    {
      "epoch": 1.1866808359900816,
      "grad_norm": 213.66128540039062,
      "learning_rate": 2.966316784621e-06,
      "loss": 45.9871,
      "step": 6700
    },
    {
      "epoch": 1.20439249025859,
      "grad_norm": 817.01806640625,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 50.9655,
      "step": 6800
    },
    {
      "epoch": 1.2221041445270988,
      "grad_norm": 236.37423706054688,
      "learning_rate": 2.6526421860705474e-06,
      "loss": 65.4025,
      "step": 6900
    },
    {
      "epoch": 1.2398157987956075,
      "grad_norm": 470.6351013183594,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 53.6276,
      "step": 7000
    },
    {
      "epoch": 1.257527453064116,
      "grad_norm": 839.293701171875,
      "learning_rate": 2.3504036788339763e-06,
      "loss": 50.0294,
      "step": 7100
    },
    {
      "epoch": 1.2752391073326248,
      "grad_norm": 404.825927734375,
      "learning_rate": 2.204035482646267e-06,
      "loss": 113.8194,
      "step": 7200
    },
    {
      "epoch": 1.2929507616011335,
      "grad_norm": 590.0635986328125,
      "learning_rate": 2.061073738537635e-06,
      "loss": 57.679,
      "step": 7300
    },
    {
      "epoch": 1.3106624158696423,
      "grad_norm": 131.95318603515625,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 53.4347,
      "step": 7400
    },
    {
      "epoch": 1.328374070138151,
      "grad_norm": 245.02337646484375,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 78.5544,
      "step": 7500
    },
    {
      "epoch": 1.3460857244066595,
      "grad_norm": 321.28204345703125,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 110.1398,
      "step": 7600
    },
    {
      "epoch": 1.3637973786751683,
      "grad_norm": 367.9349365234375,
      "learning_rate": 1.5267081477050132e-06,
      "loss": 56.705,
      "step": 7700
    },
    {
      "epoch": 1.381509032943677,
      "grad_norm": 278.4571533203125,
      "learning_rate": 1.4033009983067454e-06,
      "loss": 46.6718,
      "step": 7800
    },
    {
      "epoch": 1.3992206872121855,
      "grad_norm": 1197.47265625,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 88.5932,
      "step": 7900
    },
    {
      "epoch": 1.4169323414806942,
      "grad_norm": 403.6693420410156,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 56.4835,
      "step": 8000
    },
    {
      "epoch": 1.434643995749203,
      "grad_norm": 774.799072265625,
      "learning_rate": 1.0599462319663906e-06,
      "loss": 80.3962,
      "step": 8100
    },
    {
      "epoch": 1.4523556500177117,
      "grad_norm": 250.503662109375,
      "learning_rate": 9.549150281252633e-07,
      "loss": 67.517,
      "step": 8200
    },
    {
      "epoch": 1.4700673042862205,
      "grad_norm": 750.6965942382812,
      "learning_rate": 8.54812137224792e-07,
      "loss": 49.6351,
      "step": 8300
    },
    {
      "epoch": 1.487778958554729,
      "grad_norm": 292.1385498046875,
      "learning_rate": 7.597595192178702e-07,
      "loss": 43.521,
      "step": 8400
    },
    {
      "epoch": 1.5054906128232377,
      "grad_norm": 148.10813903808594,
      "learning_rate": 6.698729810778065e-07,
      "loss": 75.3037,
      "step": 8500
    },
    {
      "epoch": 1.5232022670917464,
      "grad_norm": 754.1968994140625,
      "learning_rate": 5.852620357053651e-07,
      "loss": 56.1155,
      "step": 8600
    },
    {
      "epoch": 1.540913921360255,
      "grad_norm": 329.13482666015625,
      "learning_rate": 5.06029768504166e-07,
      "loss": 56.4271,
      "step": 8700
    },
    {
      "epoch": 1.5586255756287637,
      "grad_norm": 142.98875427246094,
      "learning_rate": 4.322727117869951e-07,
      "loss": 55.9837,
      "step": 8800
    },
    {
      "epoch": 1.5763372298972724,
      "grad_norm": 2218.41845703125,
      "learning_rate": 3.6408072716606346e-07,
      "loss": 55.8351,
      "step": 8900
    },
    {
      "epoch": 1.594048884165781,
      "grad_norm": 483.5868225097656,
      "learning_rate": 3.015368960704584e-07,
      "loss": 78.6791,
      "step": 9000
    },
    {
      "epoch": 1.61176053843429,
      "grad_norm": 507.6253356933594,
      "learning_rate": 2.447174185242324e-07,
      "loss": 52.5828,
      "step": 9100
    },
    {
      "epoch": 1.6294721927027984,
      "grad_norm": 839.6124267578125,
      "learning_rate": 1.9369152030840553e-07,
      "loss": 60.3558,
      "step": 9200
    },
    {
      "epoch": 1.6471838469713072,
      "grad_norm": 725.627685546875,
      "learning_rate": 1.4852136862001766e-07,
      "loss": 72.146,
      "step": 9300
    },
    {
      "epoch": 1.664895501239816,
      "grad_norm": 287.16595458984375,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 66.3919,
      "step": 9400
    },
    {
      "epoch": 1.6826071555083244,
      "grad_norm": 305.81866455078125,
      "learning_rate": 7.59612349389599e-08,
      "loss": 38.9066,
      "step": 9500
    },
    {
      "epoch": 1.7003188097768331,
      "grad_norm": 787.4326782226562,
      "learning_rate": 4.865965629214819e-08,
      "loss": 81.1311,
      "step": 9600
    },
    {
      "epoch": 1.7180304640453419,
      "grad_norm": 564.6408081054688,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 81.2833,
      "step": 9700
    },
    {
      "epoch": 1.7357421183138504,
      "grad_norm": 750.64111328125,
      "learning_rate": 1.2179748700879013e-08,
      "loss": 36.9215,
      "step": 9800
    },
    {
      "epoch": 1.7534537725823593,
      "grad_norm": 477.2333068847656,
      "learning_rate": 3.0458649045211897e-09,
      "loss": 74.1448,
      "step": 9900
    },
    {
      "epoch": 1.7711654268508679,
      "grad_norm": 722.1519775390625,
      "learning_rate": 0.0,
      "loss": 60.6096,
      "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
}