{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22977589191342043,
  "eval_steps": 500,
  "global_step": 7500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003063678558845606,
      "grad_norm": 2.5270798206329346,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 3.1564,
      "step": 100
    },
    {
      "epoch": 0.006127357117691212,
      "grad_norm": 2.604092597961426,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 3.1166,
      "step": 200
    },
    {
      "epoch": 0.009191035676536818,
      "grad_norm": 2.5005924701690674,
      "learning_rate": 3e-06,
      "loss": 3.0905,
      "step": 300
    },
    {
      "epoch": 0.012254714235382424,
      "grad_norm": 2.7525177001953125,
      "learning_rate": 4.000000000000001e-06,
      "loss": 3.0652,
      "step": 400
    },
    {
      "epoch": 0.015318392794228029,
      "grad_norm": 2.5251283645629883,
      "learning_rate": 5e-06,
      "loss": 3.0351,
      "step": 500
    },
    {
      "epoch": 0.018382071353073636,
      "grad_norm": 2.6771860122680664,
      "learning_rate": 6e-06,
      "loss": 3.0028,
      "step": 600
    },
    {
      "epoch": 0.02144574991191924,
      "grad_norm": 2.4799458980560303,
      "learning_rate": 7e-06,
      "loss": 3.0007,
      "step": 700
    },
    {
      "epoch": 0.024509428470764847,
      "grad_norm": 2.8402369022369385,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.9753,
      "step": 800
    },
    {
      "epoch": 0.027573107029610452,
      "grad_norm": 2.446842908859253,
      "learning_rate": 9e-06,
      "loss": 2.9737,
      "step": 900
    },
    {
      "epoch": 0.030636785588456058,
      "grad_norm": 2.3989391326904297,
      "learning_rate": 1e-05,
      "loss": 2.9618,
      "step": 1000
    },
    {
      "epoch": 0.03370046414730166,
      "grad_norm": 2.469160795211792,
      "learning_rate": 9.99695413509548e-06,
      "loss": 2.928,
      "step": 1100
    },
    {
      "epoch": 0.03676414270614727,
      "grad_norm": 2.4107494354248047,
      "learning_rate": 9.987820251299121e-06,
      "loss": 2.9214,
      "step": 1200
    },
    {
      "epoch": 0.039827821264992874,
      "grad_norm": 2.540278911590576,
      "learning_rate": 9.972609476841368e-06,
      "loss": 2.9198,
      "step": 1300
    },
    {
      "epoch": 0.04289149982383848,
      "grad_norm": 2.4398670196533203,
      "learning_rate": 9.951340343707852e-06,
      "loss": 2.9182,
      "step": 1400
    },
    {
      "epoch": 0.04595517838268409,
      "grad_norm": 2.49687123298645,
      "learning_rate": 9.924038765061042e-06,
      "loss": 2.8994,
      "step": 1500
    },
    {
      "epoch": 0.049018856941529694,
      "grad_norm": 2.4655182361602783,
      "learning_rate": 9.890738003669029e-06,
      "loss": 2.8805,
      "step": 1600
    },
    {
      "epoch": 0.0520825355003753,
      "grad_norm": 2.4170656204223633,
      "learning_rate": 9.851478631379982e-06,
      "loss": 2.8829,
      "step": 1700
    },
    {
      "epoch": 0.055146214059220905,
      "grad_norm": 2.362334966659546,
      "learning_rate": 9.806308479691595e-06,
      "loss": 2.8891,
      "step": 1800
    },
    {
      "epoch": 0.058209892618066514,
      "grad_norm": 2.426173210144043,
      "learning_rate": 9.755282581475769e-06,
      "loss": 2.875,
      "step": 1900
    },
    {
      "epoch": 0.061273571176912116,
      "grad_norm": 2.6380629539489746,
      "learning_rate": 9.698463103929542e-06,
      "loss": 2.879,
      "step": 2000
    },
    {
      "epoch": 0.06433724973575772,
      "grad_norm": 2.5123095512390137,
      "learning_rate": 9.635919272833938e-06,
      "loss": 2.8622,
      "step": 2100
    },
    {
      "epoch": 0.06740092829460333,
      "grad_norm": 2.3789472579956055,
      "learning_rate": 9.567727288213005e-06,
      "loss": 2.8669,
      "step": 2200
    },
    {
      "epoch": 0.07046460685344894,
      "grad_norm": 2.387221336364746,
      "learning_rate": 9.493970231495836e-06,
      "loss": 2.8641,
      "step": 2300
    },
    {
      "epoch": 0.07352828541229454,
      "grad_norm": 2.4947221279144287,
      "learning_rate": 9.414737964294636e-06,
      "loss": 2.8525,
      "step": 2400
    },
    {
      "epoch": 0.07659196397114015,
      "grad_norm": 2.390381097793579,
      "learning_rate": 9.330127018922195e-06,
      "loss": 2.8497,
      "step": 2500
    },
    {
      "epoch": 0.07965564252998575,
      "grad_norm": 2.451345682144165,
      "learning_rate": 9.24024048078213e-06,
      "loss": 2.8482,
      "step": 2600
    },
    {
      "epoch": 0.08271932108883136,
      "grad_norm": 2.317512035369873,
      "learning_rate": 9.145187862775208e-06,
      "loss": 2.8486,
      "step": 2700
    },
    {
      "epoch": 0.08578299964767697,
      "grad_norm": 2.620295524597168,
      "learning_rate": 9.045084971874738e-06,
      "loss": 2.8474,
      "step": 2800
    },
    {
      "epoch": 0.08884667820652258,
      "grad_norm": 2.3616020679473877,
      "learning_rate": 8.94005376803361e-06,
      "loss": 2.8283,
      "step": 2900
    },
    {
      "epoch": 0.09191035676536818,
      "grad_norm": 2.517153263092041,
      "learning_rate": 8.83022221559489e-06,
      "loss": 2.8451,
      "step": 3000
    },
    {
      "epoch": 0.09497403532421378,
      "grad_norm": 2.6935179233551025,
      "learning_rate": 8.715724127386971e-06,
      "loss": 2.8472,
      "step": 3100
    },
    {
      "epoch": 0.09803771388305939,
      "grad_norm": 2.4080545902252197,
      "learning_rate": 8.596699001693257e-06,
      "loss": 2.8428,
      "step": 3200
    },
    {
      "epoch": 0.101101392441905,
      "grad_norm": 2.369750499725342,
      "learning_rate": 8.473291852294986e-06,
      "loss": 2.829,
      "step": 3300
    },
    {
      "epoch": 0.1041650710007506,
      "grad_norm": 2.3385934829711914,
      "learning_rate": 8.345653031794292e-06,
      "loss": 2.8349,
      "step": 3400
    },
    {
      "epoch": 0.1072287495595962,
      "grad_norm": 2.2659566402435303,
      "learning_rate": 8.213938048432697e-06,
      "loss": 2.8306,
      "step": 3500
    },
    {
      "epoch": 0.11029242811844181,
      "grad_norm": 2.508164882659912,
      "learning_rate": 8.078307376628292e-06,
      "loss": 2.8225,
      "step": 3600
    },
    {
      "epoch": 0.11335610667728742,
      "grad_norm": 2.302351713180542,
      "learning_rate": 7.938926261462366e-06,
      "loss": 2.808,
      "step": 3700
    },
    {
      "epoch": 0.11641978523613303,
      "grad_norm": 2.3972647190093994,
      "learning_rate": 7.795964517353734e-06,
      "loss": 2.8214,
      "step": 3800
    },
    {
      "epoch": 0.11948346379497864,
      "grad_norm": 2.4006614685058594,
      "learning_rate": 7.649596321166024e-06,
      "loss": 2.8198,
      "step": 3900
    },
    {
      "epoch": 0.12254714235382423,
      "grad_norm": 2.271172285079956,
      "learning_rate": 7.500000000000001e-06,
      "loss": 2.8186,
      "step": 4000
    },
    {
      "epoch": 0.12561082091266984,
      "grad_norm": 2.3733065128326416,
      "learning_rate": 7.347357813929455e-06,
      "loss": 2.8192,
      "step": 4100
    },
    {
      "epoch": 0.12867449947151544,
      "grad_norm": 2.3404409885406494,
      "learning_rate": 7.191855733945388e-06,
      "loss": 2.8235,
      "step": 4200
    },
    {
      "epoch": 0.13173817803036106,
      "grad_norm": 2.417692184448242,
      "learning_rate": 7.033683215379002e-06,
      "loss": 2.8109,
      "step": 4300
    },
    {
      "epoch": 0.13480185658920665,
      "grad_norm": 2.3670148849487305,
      "learning_rate": 6.873032967079562e-06,
      "loss": 2.8204,
      "step": 4400
    },
    {
      "epoch": 0.13786553514805228,
      "grad_norm": 2.336367130279541,
      "learning_rate": 6.710100716628345e-06,
      "loss": 2.8056,
      "step": 4500
    },
    {
      "epoch": 0.14092921370689787,
      "grad_norm": 2.3793513774871826,
      "learning_rate": 6.545084971874738e-06,
      "loss": 2.7987,
      "step": 4600
    },
    {
      "epoch": 0.14399289226574347,
      "grad_norm": 2.3895699977874756,
      "learning_rate": 6.378186779084996e-06,
      "loss": 2.8127,
      "step": 4700
    },
    {
      "epoch": 0.1470565708245891,
      "grad_norm": 2.288458824157715,
      "learning_rate": 6.209609477998339e-06,
      "loss": 2.8111,
      "step": 4800
    },
    {
      "epoch": 0.15012024938343468,
      "grad_norm": 2.299058437347412,
      "learning_rate": 6.039558454088796e-06,
      "loss": 2.7945,
      "step": 4900
    },
    {
      "epoch": 0.1531839279422803,
      "grad_norm": 2.3097774982452393,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 2.7981,
      "step": 5000
    },
    {
      "epoch": 0.1562476065011259,
      "grad_norm": 2.433981418609619,
      "learning_rate": 5.695865504800328e-06,
      "loss": 2.7938,
      "step": 5100
    },
    {
      "epoch": 0.1593112850599715,
      "grad_norm": 2.2754006385803223,
      "learning_rate": 5.522642316338268e-06,
      "loss": 2.7886,
      "step": 5200
    },
    {
      "epoch": 0.16237496361881712,
      "grad_norm": 2.3001139163970947,
      "learning_rate": 5.348782368720627e-06,
      "loss": 2.8008,
      "step": 5300
    },
    {
      "epoch": 0.16543864217766271,
      "grad_norm": 2.365751266479492,
      "learning_rate": 5.174497483512506e-06,
      "loss": 2.7858,
      "step": 5400
    },
    {
      "epoch": 0.16850232073650834,
      "grad_norm": 2.448115348815918,
      "learning_rate": 5e-06,
      "loss": 2.795,
      "step": 5500
    },
    {
      "epoch": 0.17156599929535393,
      "grad_norm": 2.414354085922241,
      "learning_rate": 4.825502516487497e-06,
      "loss": 2.7971,
      "step": 5600
    },
    {
      "epoch": 0.17462967785419953,
      "grad_norm": 2.447974443435669,
      "learning_rate": 4.651217631279374e-06,
      "loss": 2.7803,
      "step": 5700
    },
    {
      "epoch": 0.17769335641304515,
      "grad_norm": 2.38202166557312,
      "learning_rate": 4.477357683661734e-06,
      "loss": 2.784,
      "step": 5800
    },
    {
      "epoch": 0.18075703497189075,
      "grad_norm": 2.391268730163574,
      "learning_rate": 4.304134495199675e-06,
      "loss": 2.7903,
      "step": 5900
    },
    {
      "epoch": 0.18382071353073637,
      "grad_norm": 2.351288318634033,
      "learning_rate": 4.131759111665349e-06,
      "loss": 2.7788,
      "step": 6000
    },
    {
      "epoch": 0.18688439208958196,
      "grad_norm": 2.399247169494629,
      "learning_rate": 3.960441545911205e-06,
      "loss": 2.7868,
      "step": 6100
    },
    {
      "epoch": 0.18994807064842756,
      "grad_norm": 2.426527976989746,
      "learning_rate": 3.790390522001662e-06,
      "loss": 2.8012,
      "step": 6200
    },
    {
      "epoch": 0.19301174920727318,
      "grad_norm": 2.3167338371276855,
      "learning_rate": 3.6218132209150047e-06,
      "loss": 2.774,
      "step": 6300
    },
    {
      "epoch": 0.19607542776611878,
      "grad_norm": 2.2909271717071533,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 2.7715,
      "step": 6400
    },
    {
      "epoch": 0.19913910632496437,
      "grad_norm": 2.3950772285461426,
      "learning_rate": 3.289899283371657e-06,
      "loss": 2.7832,
      "step": 6500
    },
    {
      "epoch": 0.20220278488381,
      "grad_norm": 2.5031042098999023,
      "learning_rate": 3.12696703292044e-06,
      "loss": 2.7801,
      "step": 6600
    },
    {
      "epoch": 0.2052664634426556,
      "grad_norm": 2.3458404541015625,
      "learning_rate": 2.966316784621e-06,
      "loss": 2.7788,
      "step": 6700
    },
    {
      "epoch": 0.2083301420015012,
      "grad_norm": 2.3030614852905273,
      "learning_rate": 2.8081442660546126e-06,
      "loss": 2.772,
      "step": 6800
    },
    {
      "epoch": 0.2113938205603468,
      "grad_norm": 2.2177374362945557,
      "learning_rate": 2.6526421860705474e-06,
      "loss": 2.7706,
      "step": 6900
    },
    {
      "epoch": 0.2144574991191924,
      "grad_norm": 2.2889530658721924,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 2.7721,
      "step": 7000
    },
    {
      "epoch": 0.21752117767803802,
      "grad_norm": 2.288100481033325,
      "learning_rate": 2.3504036788339763e-06,
      "loss": 2.7774,
      "step": 7100
    },
    {
      "epoch": 0.22058485623688362,
      "grad_norm": 2.3013720512390137,
      "learning_rate": 2.204035482646267e-06,
      "loss": 2.7739,
      "step": 7200
    },
    {
      "epoch": 0.22364853479572924,
      "grad_norm": 2.4036357402801514,
      "learning_rate": 2.061073738537635e-06,
      "loss": 2.7759,
      "step": 7300
    },
    {
      "epoch": 0.22671221335457484,
      "grad_norm": 2.35628080368042,
      "learning_rate": 1.9216926233717087e-06,
      "loss": 2.7724,
      "step": 7400
    },
    {
      "epoch": 0.22977589191342043,
      "grad_norm": 2.314425230026245,
      "learning_rate": 1.7860619515673034e-06,
      "loss": 2.7742,
      "step": 7500
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.1748116545536e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}