{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.13000095839500783,
  "eval_steps": 500,
  "global_step": 3052,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001320455344117052,
      "grad_norm": 29.63249969482422,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 9.114,
      "step": 31
    },
    {
      "epoch": 0.002640910688234104,
      "grad_norm": 14.638819694519043,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.7274,
      "step": 62
    },
    {
      "epoch": 0.003961366032351156,
      "grad_norm": 12.396350860595703,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 6.3242,
      "step": 93
    },
    {
      "epoch": 0.005281821376468208,
      "grad_norm": 14.456698417663574,
      "learning_rate": 4.062909567496723e-06,
      "loss": 5.0889,
      "step": 124
    },
    {
      "epoch": 0.00660227672058526,
      "grad_norm": 22.106542587280273,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.5309,
      "step": 155
    },
    {
      "epoch": 0.007922732064702311,
      "grad_norm": 17.991788864135742,
      "learning_rate": 6.094364351245085e-06,
      "loss": 4.2146,
      "step": 186
    },
    {
      "epoch": 0.009243187408819364,
      "grad_norm": 15.457059860229492,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.9509,
      "step": 217
    },
    {
      "epoch": 0.010563642752936416,
      "grad_norm": 16.144941329956055,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.7073,
      "step": 248
    },
    {
      "epoch": 0.011884098097053469,
      "grad_norm": 18.386638641357422,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.5325,
      "step": 279
    },
    {
      "epoch": 0.01320455344117052,
      "grad_norm": 21.711469650268555,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.4011,
      "step": 310
    },
    {
      "epoch": 0.014525008785287572,
      "grad_norm": 14.871952056884766,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.2887,
      "step": 341
    },
    {
      "epoch": 0.015845464129404623,
      "grad_norm": 16.011245727539062,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.1892,
      "step": 372
    },
    {
      "epoch": 0.017165919473521675,
      "grad_norm": 12.8394193649292,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 3.0572,
      "step": 403
    },
    {
      "epoch": 0.018486374817638727,
      "grad_norm": 9.7935791015625,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 2.9882,
      "step": 434
    },
    {
      "epoch": 0.01980683016175578,
      "grad_norm": 11.32317066192627,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 2.9312,
      "step": 465
    },
    {
      "epoch": 0.021127285505872832,
      "grad_norm": 9.307639122009277,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.8479,
      "step": 496
    },
    {
      "epoch": 0.022447740849989885,
      "grad_norm": 9.315092086791992,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.8021,
      "step": 527
    },
    {
      "epoch": 0.023768196194106937,
      "grad_norm": 9.456792831420898,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.7374,
      "step": 558
    },
    {
      "epoch": 0.025088651538223986,
      "grad_norm": 8.998442649841309,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.7135,
      "step": 589
    },
    {
      "epoch": 0.02640910688234104,
      "grad_norm": 7.467686176300049,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.6459,
      "step": 620
    },
    {
      "epoch": 0.02772956222645809,
      "grad_norm": 7.3969831466674805,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.6074,
      "step": 651
    },
    {
      "epoch": 0.029050017570575144,
      "grad_norm": 7.884100914001465,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.5871,
      "step": 682
    },
    {
      "epoch": 0.030370472914692196,
      "grad_norm": 6.4375224113464355,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.5069,
      "step": 713
    },
    {
      "epoch": 0.031690928258809245,
      "grad_norm": 6.661777973175049,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.4812,
      "step": 744
    },
    {
      "epoch": 0.0330113836029263,
      "grad_norm": 7.4887566566467285,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.4623,
      "step": 775
    },
    {
      "epoch": 0.03433183894704335,
      "grad_norm": 6.085121154785156,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.4202,
      "step": 806
    },
    {
      "epoch": 0.0356522942911604,
      "grad_norm": 5.930830478668213,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.3901,
      "step": 837
    },
    {
      "epoch": 0.036972749635277455,
      "grad_norm": 6.05289363861084,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.3408,
      "step": 868
    },
    {
      "epoch": 0.03829320497939451,
      "grad_norm": 5.337353229522705,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.335,
      "step": 899
    },
    {
      "epoch": 0.03961366032351156,
      "grad_norm": 6.452489376068115,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.2955,
      "step": 930
    },
    {
      "epoch": 0.04093411566762861,
      "grad_norm": 5.795267581939697,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.2812,
      "step": 961
    },
    {
      "epoch": 0.042254571011745665,
      "grad_norm": 5.245680809020996,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.2487,
      "step": 992
    },
    {
      "epoch": 0.04357502635586272,
      "grad_norm": 4.897157192230225,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.2226,
      "step": 1023
    },
    {
      "epoch": 0.04489548169997977,
      "grad_norm": 4.6457743644714355,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.1903,
      "step": 1054
    },
    {
      "epoch": 0.04621593704409682,
      "grad_norm": 4.918821334838867,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.1739,
      "step": 1085
    },
    {
      "epoch": 0.047536392388213874,
      "grad_norm": 4.948458194732666,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.1354,
      "step": 1116
    },
    {
      "epoch": 0.04885684773233092,
      "grad_norm": 4.321451663970947,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.1236,
      "step": 1147
    },
    {
      "epoch": 0.05017730307644797,
      "grad_norm": 4.225420951843262,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.109,
      "step": 1178
    },
    {
      "epoch": 0.051497758420565025,
      "grad_norm": 3.955733299255371,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.0956,
      "step": 1209
    },
    {
      "epoch": 0.05281821376468208,
      "grad_norm": 4.185420036315918,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.0712,
      "step": 1240
    },
    {
      "epoch": 0.05413866910879913,
      "grad_norm": 4.096803665161133,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.0668,
      "step": 1271
    },
    {
      "epoch": 0.05545912445291618,
      "grad_norm": 4.0685133934021,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.0522,
      "step": 1302
    },
    {
      "epoch": 0.056779579797033235,
      "grad_norm": 3.7675204277038574,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.0196,
      "step": 1333
    },
    {
      "epoch": 0.05810003514115029,
      "grad_norm": 3.8732528686523438,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.008,
      "step": 1364
    },
    {
      "epoch": 0.05942049048526734,
      "grad_norm": 3.6281332969665527,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.0225,
      "step": 1395
    },
    {
      "epoch": 0.06074094582938439,
      "grad_norm": 3.630474805831909,
      "learning_rate": 4.672346002621232e-05,
      "loss": 1.9959,
      "step": 1426
    },
    {
      "epoch": 0.062061401173501445,
      "grad_norm": 3.5610458850860596,
      "learning_rate": 4.77391874180865e-05,
      "loss": 1.9743,
      "step": 1457
    },
    {
      "epoch": 0.06338185651761849,
      "grad_norm": 3.3346333503723145,
      "learning_rate": 4.875491480996068e-05,
      "loss": 1.9779,
      "step": 1488
    },
    {
      "epoch": 0.06470231186173554,
      "grad_norm": 3.3958165645599365,
      "learning_rate": 4.977064220183487e-05,
      "loss": 1.9418,
      "step": 1519
    },
    {
      "epoch": 0.0660227672058526,
      "grad_norm": 3.1751558780670166,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 1.9516,
      "step": 1550
    },
    {
      "epoch": 0.06734322254996965,
      "grad_norm": 3.2260396480560303,
      "learning_rate": 4.999955597496219e-05,
      "loss": 1.9547,
      "step": 1581
    },
    {
      "epoch": 0.0686636778940867,
      "grad_norm": 3.0978410243988037,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 1.956,
      "step": 1612
    },
    {
      "epoch": 0.06998413323820375,
      "grad_norm": 3.0465667247772217,
      "learning_rate": 4.999799067923527e-05,
      "loss": 1.9154,
      "step": 1643
    },
    {
      "epoch": 0.0713045885823208,
      "grad_norm": 2.839390277862549,
      "learning_rate": 4.999678487776908e-05,
      "loss": 1.9295,
      "step": 1674
    },
    {
      "epoch": 0.07262504392643786,
      "grad_norm": 2.9113218784332275,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 1.9182,
      "step": 1705
    },
    {
      "epoch": 0.07394549927055491,
      "grad_norm": 3.088271141052246,
      "learning_rate": 4.999352703566763e-05,
      "loss": 1.8987,
      "step": 1736
    },
    {
      "epoch": 0.07526595461467196,
      "grad_norm": 2.7890849113464355,
      "learning_rate": 4.999147503179668e-05,
      "loss": 1.8665,
      "step": 1767
    },
    {
      "epoch": 0.07658640995878901,
      "grad_norm": 2.795407295227051,
      "learning_rate": 4.998914100252672e-05,
      "loss": 1.896,
      "step": 1798
    },
    {
      "epoch": 0.07790686530290607,
      "grad_norm": 2.7121315002441406,
      "learning_rate": 4.998652497419696e-05,
      "loss": 1.8766,
      "step": 1829
    },
    {
      "epoch": 0.07922732064702312,
      "grad_norm": 2.75180983543396,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 1.8708,
      "step": 1860
    },
    {
      "epoch": 0.08054777599114017,
      "grad_norm": 2.46042799949646,
      "learning_rate": 4.998044704162613e-05,
      "loss": 1.8606,
      "step": 1891
    },
    {
      "epoch": 0.08186823133525722,
      "grad_norm": 2.579169511795044,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 1.8314,
      "step": 1922
    },
    {
      "epoch": 0.08318868667937428,
      "grad_norm": 2.867847442626953,
      "learning_rate": 4.997324150843799e-05,
      "loss": 1.8452,
      "step": 1953
    },
    {
      "epoch": 0.08450914202349133,
      "grad_norm": 2.783097505569458,
      "learning_rate": 4.99692159912661e-05,
      "loss": 1.8513,
      "step": 1984
    },
    {
      "epoch": 0.08582959736760838,
      "grad_norm": 2.689948320388794,
      "learning_rate": 4.996490869988546e-05,
      "loss": 1.8078,
      "step": 2015
    },
    {
      "epoch": 0.08715005271172543,
      "grad_norm": 2.5084853172302246,
      "learning_rate": 4.996031968290326e-05,
      "loss": 1.8036,
      "step": 2046
    },
    {
      "epoch": 0.08847050805584249,
      "grad_norm": 2.3911800384521484,
      "learning_rate": 4.995544899210594e-05,
      "loss": 1.8146,
      "step": 2077
    },
    {
      "epoch": 0.08979096339995954,
      "grad_norm": 2.5737738609313965,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 1.7868,
      "step": 2108
    },
    {
      "epoch": 0.09111141874407659,
      "grad_norm": 2.506517171859741,
      "learning_rate": 4.994486281210429e-05,
      "loss": 1.7947,
      "step": 2139
    },
    {
      "epoch": 0.09243187408819364,
      "grad_norm": 2.4341087341308594,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 1.7864,
      "step": 2170
    },
    {
      "epoch": 0.0937523294323107,
      "grad_norm": 2.530733585357666,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 1.778,
      "step": 2201
    },
    {
      "epoch": 0.09507278477642775,
      "grad_norm": 2.351055145263672,
      "learning_rate": 4.992687246588743e-05,
      "loss": 1.7821,
      "step": 2232
    },
    {
      "epoch": 0.09639324012054479,
      "grad_norm": 2.3601133823394775,
      "learning_rate": 4.992031299767347e-05,
      "loss": 1.7765,
      "step": 2263
    },
    {
      "epoch": 0.09771369546466184,
      "grad_norm": 2.4816064834594727,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 1.7855,
      "step": 2294
    },
    {
      "epoch": 0.09903415080877889,
      "grad_norm": 2.3871512413024902,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 1.7343,
      "step": 2325
    },
    {
      "epoch": 0.10035460615289594,
      "grad_norm": 2.357987642288208,
      "learning_rate": 4.989894757091861e-05,
      "loss": 1.7511,
      "step": 2356
    },
    {
      "epoch": 0.101675061497013,
      "grad_norm": 2.2945756912231445,
      "learning_rate": 4.989126368919158e-05,
      "loss": 1.7602,
      "step": 2387
    },
    {
      "epoch": 0.10299551684113005,
      "grad_norm": 2.372497081756592,
      "learning_rate": 4.988329891293693e-05,
      "loss": 1.7485,
      "step": 2418
    },
    {
      "epoch": 0.1043159721852471,
      "grad_norm": 2.2850053310394287,
      "learning_rate": 4.987505333203608e-05,
      "loss": 1.749,
      "step": 2449
    },
    {
      "epoch": 0.10563642752936415,
      "grad_norm": 2.311553955078125,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.7613,
      "step": 2480
    },
    {
      "epoch": 0.10695688287348121,
      "grad_norm": 2.346198558807373,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 1.7412,
      "step": 2511
    },
    {
      "epoch": 0.10827733821759826,
      "grad_norm": 4.81157112121582,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 1.7303,
      "step": 2542
    },
    {
      "epoch": 0.10959779356171531,
      "grad_norm": 2.1670663356781006,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.7273,
      "step": 2573
    },
    {
      "epoch": 0.11091824890583236,
      "grad_norm": 2.2016870975494385,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.7468,
      "step": 2604
    },
    {
      "epoch": 0.11223870424994942,
      "grad_norm": 2.3066468238830566,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.7506,
      "step": 2635
    },
    {
      "epoch": 0.11355915959406647,
      "grad_norm": 2.3891100883483887,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.7037,
      "step": 2666
    },
    {
      "epoch": 0.11487961493818352,
      "grad_norm": 2.234248399734497,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.74,
      "step": 2697
    },
    {
      "epoch": 0.11620007028230057,
      "grad_norm": 2.1144893169403076,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.7356,
      "step": 2728
    },
    {
      "epoch": 0.11752052562641763,
      "grad_norm": 2.2365314960479736,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.7121,
      "step": 2759
    },
    {
      "epoch": 0.11884098097053468,
      "grad_norm": 2.1084237098693848,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.7213,
      "step": 2790
    },
    {
      "epoch": 0.12016143631465173,
      "grad_norm": 2.123991012573242,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.7181,
      "step": 2821
    },
    {
      "epoch": 0.12148189165876878,
      "grad_norm": 2.23928165435791,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.6927,
      "step": 2852
    },
    {
      "epoch": 0.12280234700288584,
      "grad_norm": 2.3638722896575928,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.739,
      "step": 2883
    },
    {
      "epoch": 0.12412280234700289,
      "grad_norm": 2.0232160091400146,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.6869,
      "step": 2914
    },
    {
      "epoch": 0.12544325769111994,
      "grad_norm": 2.0534439086914062,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.6657,
      "step": 2945
    },
    {
      "epoch": 0.12676371303523698,
      "grad_norm": 2.1189608573913574,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.6935,
      "step": 2976
    },
    {
      "epoch": 0.12808416837935405,
      "grad_norm": 2.021146535873413,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.6653,
      "step": 3007
    },
    {
      "epoch": 0.12940462372347108,
      "grad_norm": 2.048736095428467,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.6953,
      "step": 3038
    }
  ],
  "logging_steps": 31,
  "max_steps": 30517,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 3052,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.263945054022271e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}