{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.0,
  "global_step": 75124,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5e-09,
      "loss": 10.4893,
      "step": 1
    },
    {
      "epoch": 0.09,
      "learning_rate": 2.5e-06,
      "loss": 9.3442,
      "step": 500
    },
    {
      "epoch": 0.19,
      "learning_rate": 5e-06,
      "loss": 7.4232,
      "step": 1000
    },
    {
      "epoch": 0.28,
      "learning_rate": 7.5e-06,
      "loss": 6.0986,
      "step": 1500
    },
    {
      "epoch": 0.37,
      "learning_rate": 1e-05,
      "loss": 5.8257,
      "step": 2000
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.25e-05,
      "loss": 5.7081,
      "step": 2500
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.5e-05,
      "loss": 5.6336,
      "step": 3000
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.75e-05,
      "loss": 5.5724,
      "step": 3500
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 5.529,
      "step": 4000
    },
    {
      "epoch": 0.84,
      "learning_rate": 2.25e-05,
      "loss": 5.4913,
      "step": 4500
    },
    {
      "epoch": 0.93,
      "learning_rate": 2.5e-05,
      "loss": 5.4578,
      "step": 5000
    },
    {
      "epoch": 1.02,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 5.4299,
      "step": 5500
    },
    {
      "epoch": 1.12,
      "learning_rate": 3e-05,
      "loss": 5.4036,
      "step": 6000
    },
    {
      "epoch": 1.21,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 5.3821,
      "step": 6500
    },
    {
      "epoch": 1.3,
      "learning_rate": 3.5e-05,
      "loss": 5.3594,
      "step": 7000
    },
    {
      "epoch": 1.4,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 5.3419,
      "step": 7500
    },
    {
      "epoch": 1.49,
      "learning_rate": 4e-05,
      "loss": 5.321,
      "step": 8000
    },
    {
      "epoch": 1.58,
      "learning_rate": 4.2495e-05,
      "loss": 5.3034,
      "step": 8500
    },
    {
      "epoch": 1.68,
      "learning_rate": 4.4995000000000005e-05,
      "loss": 5.2938,
      "step": 9000
    },
    {
      "epoch": 1.77,
      "learning_rate": 4.7495e-05,
      "loss": 5.2774,
      "step": 9500
    },
    {
      "epoch": 1.86,
      "learning_rate": 4.9995000000000005e-05,
      "loss": 5.2669,
      "step": 10000
    },
    {
      "epoch": 1.96,
      "learning_rate": 4.99883448792361e-05,
      "loss": 5.2542,
      "step": 10500
    },
    {
      "epoch": 2.05,
      "learning_rate": 4.997664295075829e-05,
      "loss": 5.2418,
      "step": 11000
    },
    {
      "epoch": 2.14,
      "learning_rate": 4.9964941022280475e-05,
      "loss": 5.231,
      "step": 11500
    },
    {
      "epoch": 2.24,
      "learning_rate": 4.995323909380266e-05,
      "loss": 5.2201,
      "step": 12000
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.994156056918181e-05,
      "loss": 5.2114,
      "step": 12500
    },
    {
      "epoch": 2.42,
      "learning_rate": 4.992985864070399e-05,
      "loss": 5.2043,
      "step": 13000
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.9918156712226175e-05,
      "loss": 5.194,
      "step": 13500
    },
    {
      "epoch": 2.61,
      "learning_rate": 4.9906454783748366e-05,
      "loss": 5.1832,
      "step": 14000
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.989475285527055e-05,
      "loss": 5.1801,
      "step": 14500
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.988307433064969e-05,
      "loss": 5.1721,
      "step": 15000
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.987137240217188e-05,
      "loss": 5.1657,
      "step": 15500
    },
    {
      "epoch": 2.98,
      "learning_rate": 4.9859670473694066e-05,
      "loss": 5.16,
      "step": 16000
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.984796854521626e-05,
      "loss": 5.1548,
      "step": 16500
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.983629002059539e-05,
      "loss": 5.1467,
      "step": 17000
    },
    {
      "epoch": 3.26,
      "learning_rate": 4.982458809211759e-05,
      "loss": 5.1421,
      "step": 17500
    },
    {
      "epoch": 3.35,
      "learning_rate": 4.981288616363977e-05,
      "loss": 5.1356,
      "step": 18000
    },
    {
      "epoch": 3.45,
      "learning_rate": 4.980118423516196e-05,
      "loss": 5.1325,
      "step": 18500
    },
    {
      "epoch": 3.54,
      "learning_rate": 4.97895057105411e-05,
      "loss": 5.1272,
      "step": 19000
    },
    {
      "epoch": 3.63,
      "learning_rate": 4.977780378206329e-05,
      "loss": 5.1207,
      "step": 19500
    },
    {
      "epoch": 3.73,
      "learning_rate": 4.976610185358547e-05,
      "loss": 5.1182,
      "step": 20000
    },
    {
      "epoch": 3.82,
      "learning_rate": 4.975439992510766e-05,
      "loss": 5.1137,
      "step": 20500
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.974269799662985e-05,
      "loss": 5.1099,
      "step": 21000
    },
    {
      "epoch": 4.01,
      "learning_rate": 4.973101947200899e-05,
      "loss": 5.1054,
      "step": 21500
    },
    {
      "epoch": 4.1,
      "learning_rate": 4.971931754353117e-05,
      "loss": 5.0999,
      "step": 22000
    },
    {
      "epoch": 4.19,
      "learning_rate": 4.9707615615053363e-05,
      "loss": 5.0948,
      "step": 22500
    },
    {
      "epoch": 4.29,
      "learning_rate": 4.9695913686575554e-05,
      "loss": 5.0925,
      "step": 23000
    },
    {
      "epoch": 4.38,
      "learning_rate": 4.968423516195469e-05,
      "loss": 5.0874,
      "step": 23500
    },
    {
      "epoch": 4.47,
      "learning_rate": 4.967253323347688e-05,
      "loss": 5.0848,
      "step": 24000
    },
    {
      "epoch": 4.57,
      "learning_rate": 4.966083130499907e-05,
      "loss": 5.0815,
      "step": 24500
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.9649129376521254e-05,
      "loss": 5.0802,
      "step": 25000
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.9637450851900395e-05,
      "loss": 5.0777,
      "step": 25500
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.962574892342258e-05,
      "loss": 5.0732,
      "step": 26000
    },
    {
      "epoch": 4.94,
      "learning_rate": 4.961404699494477e-05,
      "loss": 5.0705,
      "step": 26500
    },
    {
      "epoch": 5.03,
      "learning_rate": 4.9602345066466954e-05,
      "loss": 5.0673,
      "step": 27000
    },
    {
      "epoch": 5.12,
      "learning_rate": 4.9590666541846095e-05,
      "loss": 5.0608,
      "step": 27500
    },
    {
      "epoch": 5.22,
      "learning_rate": 4.9578964613368286e-05,
      "loss": 5.0599,
      "step": 28000
    },
    {
      "epoch": 5.31,
      "learning_rate": 4.956726268489047e-05,
      "loss": 5.0567,
      "step": 28500
    },
    {
      "epoch": 5.4,
      "learning_rate": 4.955556075641266e-05,
      "loss": 5.0523,
      "step": 29000
    },
    {
      "epoch": 5.5,
      "learning_rate": 4.95438822317918e-05,
      "loss": 5.051,
      "step": 29500
    },
    {
      "epoch": 5.59,
      "learning_rate": 4.953218030331399e-05,
      "loss": 5.0469,
      "step": 30000
    },
    {
      "epoch": 5.68,
      "learning_rate": 4.952047837483618e-05,
      "loss": 5.0424,
      "step": 30500
    },
    {
      "epoch": 5.78,
      "learning_rate": 4.950877644635836e-05,
      "loss": 5.0405,
      "step": 31000
    },
    {
      "epoch": 5.87,
      "learning_rate": 4.949709792173751e-05,
      "loss": 5.0073,
      "step": 31500
    },
    {
      "epoch": 5.96,
      "learning_rate": 4.948539599325969e-05,
      "loss": 4.6646,
      "step": 32000
    },
    {
      "epoch": 6.06,
      "learning_rate": 4.9473694064781877e-05,
      "loss": 4.3483,
      "step": 32500
    },
    {
      "epoch": 6.15,
      "learning_rate": 4.946199213630406e-05,
      "loss": 4.0878,
      "step": 33000
    },
    {
      "epoch": 6.24,
      "learning_rate": 4.945031361168321e-05,
      "loss": 3.8246,
      "step": 33500
    },
    {
      "epoch": 6.34,
      "learning_rate": 4.943861168320539e-05,
      "loss": 3.221,
      "step": 34000
    },
    {
      "epoch": 6.43,
      "learning_rate": 4.9426909754727577e-05,
      "loss": 2.7026,
      "step": 34500
    },
    {
      "epoch": 6.52,
      "learning_rate": 4.9415207826249774e-05,
      "loss": 2.3592,
      "step": 35000
    },
    {
      "epoch": 6.62,
      "learning_rate": 4.940352930162891e-05,
      "loss": 1.9468,
      "step": 35500
    },
    {
      "epoch": 6.71,
      "learning_rate": 4.939182737315109e-05,
      "loss": 1.6962,
      "step": 36000
    },
    {
      "epoch": 6.8,
      "learning_rate": 4.938012544467328e-05,
      "loss": 1.5455,
      "step": 36500
    },
    {
      "epoch": 6.9,
      "learning_rate": 4.9368423516195474e-05,
      "loss": 1.4404,
      "step": 37000
    },
    {
      "epoch": 6.99,
      "learning_rate": 4.9356744991574615e-05,
      "loss": 1.3671,
      "step": 37500
    },
    {
      "epoch": 7.08,
      "learning_rate": 4.93450430630968e-05,
      "loss": 1.3047,
      "step": 38000
    },
    {
      "epoch": 7.17,
      "learning_rate": 4.933334113461899e-05,
      "loss": 1.242,
      "step": 38500
    },
    {
      "epoch": 7.27,
      "learning_rate": 4.9321639206141174e-05,
      "loss": 1.1857,
      "step": 39000
    },
    {
      "epoch": 7.36,
      "learning_rate": 4.930993727766336e-05,
      "loss": 1.1364,
      "step": 39500
    },
    {
      "epoch": 7.45,
      "learning_rate": 4.9298258753042506e-05,
      "loss": 1.0976,
      "step": 40000
    },
    {
      "epoch": 7.55,
      "learning_rate": 4.928655682456469e-05,
      "loss": 1.062,
      "step": 40500
    },
    {
      "epoch": 7.64,
      "learning_rate": 4.9274854896086874e-05,
      "loss": 1.0284,
      "step": 41000
    },
    {
      "epoch": 7.73,
      "learning_rate": 4.9263152967609065e-05,
      "loss": 1.0023,
      "step": 41500
    },
    {
      "epoch": 7.83,
      "learning_rate": 4.9251474442988206e-05,
      "loss": 0.98,
      "step": 42000
    },
    {
      "epoch": 7.92,
      "learning_rate": 4.9239772514510397e-05,
      "loss": 0.9582,
      "step": 42500
    },
    {
      "epoch": 8.01,
      "learning_rate": 4.922809398988954e-05,
      "loss": 0.9424,
      "step": 43000
    },
    {
      "epoch": 8.11,
      "learning_rate": 4.921639206141173e-05,
      "loss": 0.9232,
      "step": 43500
    },
    {
      "epoch": 8.2,
      "learning_rate": 4.920469013293391e-05,
      "loss": 0.908,
      "step": 44000
    },
    {
      "epoch": 8.29,
      "learning_rate": 4.9192988204456097e-05,
      "loss": 0.8941,
      "step": 44500
    },
    {
      "epoch": 8.39,
      "learning_rate": 4.918128627597828e-05,
      "loss": 0.8833,
      "step": 45000
    },
    {
      "epoch": 8.48,
      "learning_rate": 4.916958434750047e-05,
      "loss": 0.8697,
      "step": 45500
    },
    {
      "epoch": 8.57,
      "learning_rate": 4.9157882419022655e-05,
      "loss": 0.8558,
      "step": 46000
    },
    {
      "epoch": 8.67,
      "learning_rate": 4.9146203894401796e-05,
      "loss": 0.8465,
      "step": 46500
    },
    {
      "epoch": 8.76,
      "learning_rate": 4.913450196592399e-05,
      "loss": 0.8352,
      "step": 47000
    },
    {
      "epoch": 8.85,
      "learning_rate": 4.912280003744618e-05,
      "loss": 0.8253,
      "step": 47500
    },
    {
      "epoch": 8.95,
      "learning_rate": 4.911109810896836e-05,
      "loss": 0.8135,
      "step": 48000
    },
    {
      "epoch": 9.04,
      "learning_rate": 4.9099396180490546e-05,
      "loss": 0.8064,
      "step": 48500
    },
    {
      "epoch": 9.13,
      "learning_rate": 4.908769425201274e-05,
      "loss": 0.7971,
      "step": 49000
    },
    {
      "epoch": 9.22,
      "learning_rate": 4.907601572739188e-05,
      "loss": 0.7846,
      "step": 49500
    },
    {
      "epoch": 9.32,
      "learning_rate": 4.906431379891406e-05,
      "loss": 0.779,
      "step": 50000
    },
    {
      "epoch": 9.41,
      "learning_rate": 4.9052611870436246e-05,
      "loss": 0.7717,
      "step": 50500
    },
    {
      "epoch": 9.5,
      "learning_rate": 4.904090994195844e-05,
      "loss": 0.7618,
      "step": 51000
    },
    {
      "epoch": 9.6,
      "learning_rate": 4.902920801348063e-05,
      "loss": 0.7573,
      "step": 51500
    },
    {
      "epoch": 9.69,
      "learning_rate": 4.901752948885976e-05,
      "loss": 0.7505,
      "step": 52000
    },
    {
      "epoch": 9.78,
      "learning_rate": 4.900582756038195e-05,
      "loss": 0.7445,
      "step": 52500
    },
    {
      "epoch": 9.88,
      "learning_rate": 4.8994125631904144e-05,
      "loss": 0.7389,
      "step": 53000
    },
    {
      "epoch": 9.97,
      "learning_rate": 4.898242370342633e-05,
      "loss": 0.7314,
      "step": 53500
    },
    {
      "epoch": 10.06,
      "learning_rate": 4.897072177494851e-05,
      "loss": 0.7245,
      "step": 54000
    },
    {
      "epoch": 10.16,
      "learning_rate": 4.895904325032766e-05,
      "loss": 0.7193,
      "step": 54500
    },
    {
      "epoch": 10.25,
      "learning_rate": 4.8947341321849843e-05,
      "loss": 0.7137,
      "step": 55000
    },
    {
      "epoch": 10.34,
      "learning_rate": 4.893563939337203e-05,
      "loss": 0.7083,
      "step": 55500
    },
    {
      "epoch": 10.44,
      "learning_rate": 4.892393746489422e-05,
      "loss": 0.7039,
      "step": 56000
    },
    {
      "epoch": 10.53,
      "learning_rate": 4.891223553641641e-05,
      "loss": 0.6995,
      "step": 56500
    },
    {
      "epoch": 10.62,
      "learning_rate": 4.890053360793859e-05,
      "loss": 0.6941,
      "step": 57000
    },
    {
      "epoch": 10.72,
      "learning_rate": 4.888885508331773e-05,
      "loss": 0.6904,
      "step": 57500
    },
    {
      "epoch": 10.81,
      "learning_rate": 4.8877153154839925e-05,
      "loss": 0.6846,
      "step": 58000
    },
    {
      "epoch": 10.9,
      "learning_rate": 4.886545122636211e-05,
      "loss": 0.6806,
      "step": 58500
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.885374929788429e-05,
      "loss": 0.6771,
      "step": 59000
    },
    {
      "epoch": 11.09,
      "learning_rate": 4.884204736940648e-05,
      "loss": 0.6726,
      "step": 59500
    },
    {
      "epoch": 11.18,
      "learning_rate": 4.8830368844785625e-05,
      "loss": 0.6679,
      "step": 60000
    },
    {
      "epoch": 11.27,
      "learning_rate": 4.881866691630781e-05,
      "loss": 0.6634,
      "step": 60500
    },
    {
      "epoch": 11.37,
      "learning_rate": 4.880696498782999e-05,
      "loss": 0.6607,
      "step": 61000
    },
    {
      "epoch": 11.46,
      "learning_rate": 4.8795263059352184e-05,
      "loss": 0.6568,
      "step": 61500
    },
    {
      "epoch": 11.55,
      "learning_rate": 4.8783561130874375e-05,
      "loss": 0.6546,
      "step": 62000
    },
    {
      "epoch": 11.65,
      "learning_rate": 4.877188260625351e-05,
      "loss": 0.6527,
      "step": 62500
    },
    {
      "epoch": 11.74,
      "learning_rate": 4.87601806777757e-05,
      "loss": 0.6455,
      "step": 63000
    },
    {
      "epoch": 11.83,
      "learning_rate": 4.874847874929789e-05,
      "loss": 0.6437,
      "step": 63500
    },
    {
      "epoch": 11.93,
      "learning_rate": 4.8736776820820074e-05,
      "loss": 0.6408,
      "step": 64000
    },
    {
      "epoch": 12.02,
      "learning_rate": 4.872507489234226e-05,
      "loss": 0.6369,
      "step": 64500
    },
    {
      "epoch": 12.11,
      "learning_rate": 4.871337296386444e-05,
      "loss": 0.6336,
      "step": 65000
    },
    {
      "epoch": 12.21,
      "learning_rate": 4.870167103538664e-05,
      "loss": 0.63,
      "step": 65500
    },
    {
      "epoch": 12.3,
      "learning_rate": 4.8689992510765774e-05,
      "loss": 0.6277,
      "step": 66000
    },
    {
      "epoch": 12.39,
      "learning_rate": 4.867829058228796e-05,
      "loss": 0.6265,
      "step": 66500
    },
    {
      "epoch": 12.49,
      "learning_rate": 4.866658865381015e-05,
      "loss": 0.6238,
      "step": 67000
    },
    {
      "epoch": 12.58,
      "learning_rate": 4.865488672533234e-05,
      "loss": 0.6192,
      "step": 67500
    },
    {
      "epoch": 12.67,
      "learning_rate": 4.864320820071148e-05,
      "loss": 0.6165,
      "step": 68000
    },
    {
      "epoch": 12.77,
      "learning_rate": 4.8631506272233665e-05,
      "loss": 0.613,
      "step": 68500
    },
    {
      "epoch": 12.86,
      "learning_rate": 4.8619804343755856e-05,
      "loss": 0.6124,
      "step": 69000
    },
    {
      "epoch": 12.95,
      "learning_rate": 4.860810241527804e-05,
      "loss": 0.6111,
      "step": 69500
    },
    {
      "epoch": 13.05,
      "learning_rate": 4.8596400486800224e-05,
      "loss": 0.607,
      "step": 70000
    },
    {
      "epoch": 13.14,
      "learning_rate": 4.858469855832241e-05,
      "loss": 0.602,
      "step": 70500
    },
    {
      "epoch": 13.23,
      "learning_rate": 4.8573020033701556e-05,
      "loss": 0.602,
      "step": 71000
    },
    {
      "epoch": 13.32,
      "learning_rate": 4.856131810522374e-05,
      "loss": 0.6008,
      "step": 71500
    },
    {
      "epoch": 13.42,
      "learning_rate": 4.854961617674593e-05,
      "loss": 0.5974,
      "step": 72000
    },
    {
      "epoch": 13.51,
      "learning_rate": 4.853791424826812e-05,
      "loss": 0.5939,
      "step": 72500
    },
    {
      "epoch": 13.6,
      "learning_rate": 4.8526212319790305e-05,
      "loss": 0.5942,
      "step": 73000
    },
    {
      "epoch": 13.7,
      "learning_rate": 4.851451039131249e-05,
      "loss": 0.5902,
      "step": 73500
    },
    {
      "epoch": 13.79,
      "learning_rate": 4.850283186669163e-05,
      "loss": 0.5893,
      "step": 74000
    },
    {
      "epoch": 13.88,
      "learning_rate": 4.849112993821382e-05,
      "loss": 0.5889,
      "step": 74500
    },
    {
      "epoch": 13.98,
      "learning_rate": 4.8479428009736005e-05,
      "loss": 0.5864,
      "step": 75000
    }
  ],
  "max_steps": 2146400,
  "num_train_epochs": 400,
  "total_flos": 2.024730999544978e+19,
  "trial_name": null,
  "trial_params": null
}