{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 71754,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.9651587367951616e-05,
      "loss": 2.9853,
      "step": 500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.930317473590323e-05,
      "loss": 2.2101,
      "step": 1000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.895476210385484e-05,
      "loss": 1.9807,
      "step": 1500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.860634947180645e-05,
      "loss": 1.8067,
      "step": 2000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.825793683975807e-05,
      "loss": 1.7071,
      "step": 2500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.7909524207709675e-05,
      "loss": 1.6716,
      "step": 3000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.756111157566129e-05,
      "loss": 1.6077,
      "step": 3500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.72126989436129e-05,
      "loss": 1.5357,
      "step": 4000
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.6864286311564515e-05,
      "loss": 1.4992,
      "step": 4500
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.651587367951613e-05,
      "loss": 1.4536,
      "step": 5000
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.616746104746774e-05,
      "loss": 1.4247,
      "step": 5500
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.5819048415419355e-05,
      "loss": 1.4124,
      "step": 6000
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.547063578337096e-05,
      "loss": 1.4125,
      "step": 6500
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.512222315132258e-05,
      "loss": 1.358,
      "step": 7000
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.477381051927419e-05,
      "loss": 1.3406,
      "step": 7500
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.44253978872258e-05,
      "loss": 1.3251,
      "step": 8000
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.4076985255177414e-05,
      "loss": 1.3072,
      "step": 8500
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.372857262312903e-05,
      "loss": 1.3088,
      "step": 9000
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.338015999108064e-05,
      "loss": 1.2935,
      "step": 9500
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.303174735903225e-05,
      "loss": 1.2937,
      "step": 10000
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.268333472698387e-05,
      "loss": 1.2535,
      "step": 10500
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.2334922094935473e-05,
      "loss": 1.253,
      "step": 11000
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.198650946288709e-05,
      "loss": 1.229,
      "step": 11500
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.16380968308387e-05,
      "loss": 1.2517,
      "step": 12000
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.128968419879031e-05,
      "loss": 1.2161,
      "step": 12500
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.0941271566741926e-05,
      "loss": 1.2112,
      "step": 13000
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.059285893469354e-05,
      "loss": 1.2249,
      "step": 13500
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.024444630264515e-05,
      "loss": 1.1943,
      "step": 14000
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.989603367059676e-05,
      "loss": 1.1875,
      "step": 14500
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.954762103854838e-05,
      "loss": 1.1844,
      "step": 15000
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9199208406499986e-05,
      "loss": 1.1732,
      "step": 15500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.88507957744516e-05,
      "loss": 1.161,
      "step": 16000
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.850238314240321e-05,
      "loss": 1.1668,
      "step": 16500
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.8153970510354826e-05,
      "loss": 1.1866,
      "step": 17000
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.780555787830644e-05,
      "loss": 1.1433,
      "step": 17500
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.745714524625805e-05,
      "loss": 1.1662,
      "step": 18000
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.7108732614209665e-05,
      "loss": 1.1432,
      "step": 18500
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.676031998216127e-05,
      "loss": 1.113,
      "step": 19000
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.641190735011289e-05,
      "loss": 1.1083,
      "step": 19500
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.60634947180645e-05,
      "loss": 1.1354,
      "step": 20000
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.571508208601611e-05,
      "loss": 1.1009,
      "step": 20500
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.5366669453967725e-05,
      "loss": 1.0959,
      "step": 21000
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.501825682191934e-05,
      "loss": 1.1142,
      "step": 21500
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.466984418987095e-05,
      "loss": 1.0885,
      "step": 22000
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.4321431557822564e-05,
      "loss": 1.1067,
      "step": 22500
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.397301892577418e-05,
      "loss": 1.0969,
      "step": 23000
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3624606293725784e-05,
      "loss": 1.0864,
      "step": 23500
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.3276193661677404e-05,
      "loss": 1.0521,
      "step": 24000
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.292778102962901e-05,
      "loss": 0.9488,
      "step": 24500
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.2579368397580624e-05,
      "loss": 0.9548,
      "step": 25000
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.223095576553224e-05,
      "loss": 0.9526,
      "step": 25500
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.188254313348385e-05,
      "loss": 0.9632,
      "step": 26000
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.1534130501435464e-05,
      "loss": 0.973,
      "step": 26500
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.118571786938708e-05,
      "loss": 0.9313,
      "step": 27000
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.083730523733869e-05,
      "loss": 0.9518,
      "step": 27500
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.0488892605290297e-05,
      "loss": 0.958,
      "step": 28000
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.014047997324191e-05,
      "loss": 0.9608,
      "step": 28500
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9792067341193523e-05,
      "loss": 0.9505,
      "step": 29000
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.9443654709145136e-05,
      "loss": 0.9459,
      "step": 29500
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.9095242077096746e-05,
      "loss": 0.9714,
      "step": 30000
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.8746829445048363e-05,
      "loss": 0.9428,
      "step": 30500
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.8398416812999973e-05,
      "loss": 0.9437,
      "step": 31000
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.8050004180951583e-05,
      "loss": 0.9598,
      "step": 31500
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.77015915489032e-05,
      "loss": 0.9452,
      "step": 32000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.735317891685481e-05,
      "loss": 0.9371,
      "step": 32500
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7004766284806422e-05,
      "loss": 0.937,
      "step": 33000
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.6656353652758035e-05,
      "loss": 0.929,
      "step": 33500
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.630794102070965e-05,
      "loss": 0.9305,
      "step": 34000
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.595952838866126e-05,
      "loss": 0.9396,
      "step": 34500
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5611115756612875e-05,
      "loss": 0.9248,
      "step": 35000
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5262703124564485e-05,
      "loss": 0.9366,
      "step": 35500
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.49142904925161e-05,
      "loss": 0.9199,
      "step": 36000
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4565877860467708e-05,
      "loss": 0.9239,
      "step": 36500
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.421746522841932e-05,
      "loss": 0.9351,
      "step": 37000
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3869052596370935e-05,
      "loss": 0.916,
      "step": 37500
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3520639964322548e-05,
      "loss": 0.919,
      "step": 38000
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.317222733227416e-05,
      "loss": 0.8994,
      "step": 38500
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.2823814700225774e-05,
      "loss": 0.9181,
      "step": 39000
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2475402068177384e-05,
      "loss": 0.9098,
      "step": 39500
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2126989436128997e-05,
      "loss": 0.908,
      "step": 40000
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.177857680408061e-05,
      "loss": 0.9131,
      "step": 40500
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.143016417203222e-05,
      "loss": 0.8964,
      "step": 41000
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.1081751539983834e-05,
      "loss": 0.9017,
      "step": 41500
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.0733338907935447e-05,
      "loss": 0.8914,
      "step": 42000
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.038492627588706e-05,
      "loss": 0.893,
      "step": 42500
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.0036513643838674e-05,
      "loss": 0.8932,
      "step": 43000
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.9688101011790283e-05,
      "loss": 0.9003,
      "step": 43500
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.9339688379741897e-05,
      "loss": 0.8818,
      "step": 44000
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.899127574769351e-05,
      "loss": 0.8941,
      "step": 44500
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.864286311564512e-05,
      "loss": 0.8694,
      "step": 45000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.8294450483596733e-05,
      "loss": 0.8887,
      "step": 45500
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.7946037851548346e-05,
      "loss": 0.8857,
      "step": 46000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.759762521949996e-05,
      "loss": 0.8746,
      "step": 46500
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7249212587451573e-05,
      "loss": 0.8668,
      "step": 47000
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.6900799955403186e-05,
      "loss": 0.8817,
      "step": 47500
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.6552387323354796e-05,
      "loss": 0.8397,
      "step": 48000
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.620397469130641e-05,
      "loss": 0.7568,
      "step": 48500
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5855562059258022e-05,
      "loss": 0.7493,
      "step": 49000
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.5507149427209632e-05,
      "loss": 0.7577,
      "step": 49500
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5158736795161247e-05,
      "loss": 0.7413,
      "step": 50000
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.481032416311286e-05,
      "loss": 0.7521,
      "step": 50500
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.446191153106447e-05,
      "loss": 0.7405,
      "step": 51000
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4113498899016083e-05,
      "loss": 0.7474,
      "step": 51500
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3765086266967697e-05,
      "loss": 0.7649,
      "step": 52000
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3416673634919308e-05,
      "loss": 0.7551,
      "step": 52500
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3068261002870921e-05,
      "loss": 0.7677,
      "step": 53000
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2719848370822531e-05,
      "loss": 0.7214,
      "step": 53500
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.2371435738774145e-05,
      "loss": 0.7513,
      "step": 54000
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.202302310672576e-05,
      "loss": 0.7361,
      "step": 54500
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1674610474677371e-05,
      "loss": 0.7561,
      "step": 55000
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1326197842628983e-05,
      "loss": 0.7495,
      "step": 55500
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.0977785210580594e-05,
      "loss": 0.7528,
      "step": 56000
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0629372578532207e-05,
      "loss": 0.7488,
      "step": 56500
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.028095994648382e-05,
      "loss": 0.7446,
      "step": 57000
    },
    {
      "epoch": 2.4,
      "learning_rate": 9.932547314435432e-06,
      "loss": 0.7346,
      "step": 57500
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.584134682387045e-06,
      "loss": 0.7392,
      "step": 58000
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.235722050338657e-06,
      "loss": 0.7414,
      "step": 58500
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.88730941829027e-06,
      "loss": 0.7366,
      "step": 59000
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.538896786241882e-06,
      "loss": 0.7376,
      "step": 59500
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.190484154193495e-06,
      "loss": 0.7466,
      "step": 60000
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.842071522145107e-06,
      "loss": 0.743,
      "step": 60500
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.493658890096719e-06,
      "loss": 0.7447,
      "step": 61000
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.145246258048332e-06,
      "loss": 0.7335,
      "step": 61500
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.796833625999945e-06,
      "loss": 0.7277,
      "step": 62000
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.448420993951557e-06,
      "loss": 0.745,
      "step": 62500
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.100008361903169e-06,
      "loss": 0.7365,
      "step": 63000
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.751595729854782e-06,
      "loss": 0.7334,
      "step": 63500
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.403183097806394e-06,
      "loss": 0.7392,
      "step": 64000
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.054770465758007e-06,
      "loss": 0.7392,
      "step": 64500
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.70635783370962e-06,
      "loss": 0.7318,
      "step": 65000
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.357945201661231e-06,
      "loss": 0.7208,
      "step": 65500
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.009532569612845e-06,
      "loss": 0.7268,
      "step": 66000
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.661119937564456e-06,
      "loss": 0.734,
      "step": 66500
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.312707305516069e-06,
      "loss": 0.7354,
      "step": 67000
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.9642946734676814e-06,
      "loss": 0.733,
      "step": 67500
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.615882041419294e-06,
      "loss": 0.7167,
      "step": 68000
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.267469409370906e-06,
      "loss": 0.7181,
      "step": 68500
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.919056777322519e-06,
      "loss": 0.7399,
      "step": 69000
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.5706441452741312e-06,
      "loss": 0.7239,
      "step": 69500
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.2222315132257436e-06,
      "loss": 0.7194,
      "step": 70000
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.73818881177356e-07,
      "loss": 0.7319,
      "step": 70500
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.254062491289685e-07,
      "loss": 0.7162,
      "step": 71000
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.7699361708058088e-07,
      "loss": 0.7239,
      "step": 71500
    },
    {
      "epoch": 3.0,
      "step": 71754,
      "total_flos": 8.271547272486912e+16,
      "train_loss": 1.0012687716917485,
      "train_runtime": 19490.5067,
      "train_samples_per_second": 36.815,
      "train_steps_per_second": 3.681
    }
  ],
  "max_steps": 71754,
  "num_train_epochs": 3,
  "total_flos": 8.271547272486912e+16,
  "trial_name": null,
  "trial_params": null
}