{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.819185320199062,
  "eval_steps": 500,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 9.965867278325039e-05,
      "loss": 1.8363,
      "step": 500
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.931734556650079e-05,
      "loss": 1.2927,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.897601834975117e-05,
      "loss": 1.0069,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.863469113300157e-05,
      "loss": 0.8362,
      "step": 2000
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.829336391625195e-05,
      "loss": 0.7217,
      "step": 2500
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.795203669950235e-05,
      "loss": 0.6462,
      "step": 3000
    },
    {
      "epoch": 0.07,
      "learning_rate": 9.761070948275274e-05,
      "loss": 0.5914,
      "step": 3500
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.726938226600313e-05,
      "loss": 0.5497,
      "step": 4000
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.692805504925352e-05,
      "loss": 0.5154,
      "step": 4500
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.658672783250392e-05,
      "loss": 0.4867,
      "step": 5000
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.62454006157543e-05,
      "loss": 0.4628,
      "step": 5500
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.590407339900468e-05,
      "loss": 0.4393,
      "step": 6000
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.556274618225508e-05,
      "loss": 0.4199,
      "step": 6500
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.522141896550547e-05,
      "loss": 0.3996,
      "step": 7000
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.488009174875586e-05,
      "loss": 0.383,
      "step": 7500
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.453876453200626e-05,
      "loss": 0.37,
      "step": 8000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.419743731525666e-05,
      "loss": 0.3577,
      "step": 8500
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.385611009850704e-05,
      "loss": 0.3468,
      "step": 9000
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.351478288175744e-05,
      "loss": 0.3364,
      "step": 9500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.317345566500783e-05,
      "loss": 0.3281,
      "step": 10000
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.283212844825821e-05,
      "loss": 0.3198,
      "step": 10500
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.249080123150861e-05,
      "loss": 0.3124,
      "step": 11000
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.214947401475899e-05,
      "loss": 0.3061,
      "step": 11500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.180814679800939e-05,
      "loss": 0.2999,
      "step": 12000
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.146681958125977e-05,
      "loss": 0.2938,
      "step": 12500
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.112549236451017e-05,
      "loss": 0.2886,
      "step": 13000
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.078416514776056e-05,
      "loss": 0.2834,
      "step": 13500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.044283793101095e-05,
      "loss": 0.2791,
      "step": 14000
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.010151071426134e-05,
      "loss": 0.2748,
      "step": 14500
    },
    {
      "epoch": 0.31,
      "learning_rate": 8.976018349751174e-05,
      "loss": 0.2708,
      "step": 15000
    },
    {
      "epoch": 0.32,
      "learning_rate": 8.941885628076212e-05,
      "loss": 0.2673,
      "step": 15500
    },
    {
      "epoch": 0.33,
      "learning_rate": 8.90775290640125e-05,
      "loss": 0.2634,
      "step": 16000
    },
    {
      "epoch": 0.34,
      "learning_rate": 8.87362018472629e-05,
      "loss": 0.2604,
      "step": 16500
    },
    {
      "epoch": 0.35,
      "learning_rate": 8.839487463051329e-05,
      "loss": 0.2575,
      "step": 17000
    },
    {
      "epoch": 0.36,
      "learning_rate": 8.805354741376368e-05,
      "loss": 0.2542,
      "step": 17500
    },
    {
      "epoch": 0.37,
      "learning_rate": 8.771222019701407e-05,
      "loss": 0.251,
      "step": 18000
    },
    {
      "epoch": 0.38,
      "learning_rate": 8.737089298026447e-05,
      "loss": 0.2483,
      "step": 18500
    },
    {
      "epoch": 0.39,
      "learning_rate": 8.702956576351485e-05,
      "loss": 0.2458,
      "step": 19000
    },
    {
      "epoch": 0.4,
      "learning_rate": 8.668823854676525e-05,
      "loss": 0.2429,
      "step": 19500
    },
    {
      "epoch": 0.41,
      "learning_rate": 8.634691133001563e-05,
      "loss": 0.2405,
      "step": 20000
    },
    {
      "epoch": 0.42,
      "learning_rate": 8.600558411326603e-05,
      "loss": 0.2391,
      "step": 20500
    },
    {
      "epoch": 0.43,
      "learning_rate": 8.566425689651641e-05,
      "loss": 0.237,
      "step": 21000
    },
    {
      "epoch": 0.44,
      "learning_rate": 8.532292967976681e-05,
      "loss": 0.2342,
      "step": 21500
    },
    {
      "epoch": 0.45,
      "learning_rate": 8.49816024630172e-05,
      "loss": 0.2326,
      "step": 22000
    },
    {
      "epoch": 0.46,
      "learning_rate": 8.464027524626758e-05,
      "loss": 0.2304,
      "step": 22500
    },
    {
      "epoch": 0.47,
      "learning_rate": 8.429894802951798e-05,
      "loss": 0.2286,
      "step": 23000
    },
    {
      "epoch": 0.48,
      "learning_rate": 8.395762081276836e-05,
      "loss": 0.2271,
      "step": 23500
    },
    {
      "epoch": 0.49,
      "learning_rate": 8.361629359601877e-05,
      "loss": 0.2252,
      "step": 24000
    },
    {
      "epoch": 0.5,
      "learning_rate": 8.327496637926916e-05,
      "loss": 0.2231,
      "step": 24500
    },
    {
      "epoch": 0.51,
      "learning_rate": 8.293363916251956e-05,
      "loss": 0.2219,
      "step": 25000
    },
    {
      "epoch": 0.52,
      "learning_rate": 8.259231194576994e-05,
      "loss": 0.2199,
      "step": 25500
    },
    {
      "epoch": 0.53,
      "learning_rate": 8.225098472902032e-05,
      "loss": 0.219,
      "step": 26000
    },
    {
      "epoch": 0.54,
      "learning_rate": 8.190965751227072e-05,
      "loss": 0.2169,
      "step": 26500
    },
    {
      "epoch": 0.55,
      "learning_rate": 8.15683302955211e-05,
      "loss": 0.2157,
      "step": 27000
    },
    {
      "epoch": 0.56,
      "learning_rate": 8.12270030787715e-05,
      "loss": 0.214,
      "step": 27500
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.088567586202189e-05,
      "loss": 0.2129,
      "step": 28000
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.054434864527229e-05,
      "loss": 0.2117,
      "step": 28500
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.020302142852267e-05,
      "loss": 0.2111,
      "step": 29000
    },
    {
      "epoch": 0.6,
      "learning_rate": 7.986169421177307e-05,
      "loss": 0.2091,
      "step": 29500
    },
    {
      "epoch": 0.61,
      "learning_rate": 7.952036699502345e-05,
      "loss": 0.2084,
      "step": 30000
    },
    {
      "epoch": 0.62,
      "learning_rate": 7.917903977827385e-05,
      "loss": 0.2071,
      "step": 30500
    },
    {
      "epoch": 0.63,
      "learning_rate": 7.883771256152423e-05,
      "loss": 0.2057,
      "step": 31000
    },
    {
      "epoch": 0.65,
      "learning_rate": 7.849638534477462e-05,
      "loss": 0.2051,
      "step": 31500
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.815505812802502e-05,
      "loss": 0.2036,
      "step": 32000
    },
    {
      "epoch": 0.67,
      "learning_rate": 7.78137309112754e-05,
      "loss": 0.2027,
      "step": 32500
    },
    {
      "epoch": 0.68,
      "learning_rate": 7.74724036945258e-05,
      "loss": 0.2017,
      "step": 33000
    },
    {
      "epoch": 0.69,
      "learning_rate": 7.713107647777618e-05,
      "loss": 0.2009,
      "step": 33500
    },
    {
      "epoch": 0.7,
      "learning_rate": 7.678974926102658e-05,
      "loss": 0.2003,
      "step": 34000
    },
    {
      "epoch": 0.71,
      "learning_rate": 7.644842204427697e-05,
      "loss": 0.1991,
      "step": 34500
    },
    {
      "epoch": 0.72,
      "learning_rate": 7.610709482752736e-05,
      "loss": 0.198,
      "step": 35000
    },
    {
      "epoch": 0.73,
      "learning_rate": 7.576576761077775e-05,
      "loss": 0.1974,
      "step": 35500
    },
    {
      "epoch": 0.74,
      "learning_rate": 7.542444039402814e-05,
      "loss": 0.1964,
      "step": 36000
    },
    {
      "epoch": 0.75,
      "learning_rate": 7.508311317727853e-05,
      "loss": 0.1958,
      "step": 36500
    },
    {
      "epoch": 0.76,
      "learning_rate": 7.474178596052893e-05,
      "loss": 0.1949,
      "step": 37000
    },
    {
      "epoch": 0.77,
      "learning_rate": 7.440045874377931e-05,
      "loss": 0.1937,
      "step": 37500
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.40591315270297e-05,
      "loss": 0.1931,
      "step": 38000
    },
    {
      "epoch": 0.79,
      "learning_rate": 7.37178043102801e-05,
      "loss": 0.1926,
      "step": 38500
    },
    {
      "epoch": 0.8,
      "learning_rate": 7.337647709353048e-05,
      "loss": 0.1919,
      "step": 39000
    },
    {
      "epoch": 0.81,
      "learning_rate": 7.303514987678088e-05,
      "loss": 0.1909,
      "step": 39500
    },
    {
      "epoch": 0.82,
      "learning_rate": 7.269382266003127e-05,
      "loss": 0.1902,
      "step": 40000
    }
  ],
  "logging_steps": 500,
  "max_steps": 146487,
  "num_train_epochs": 3,
  "save_steps": 10000,
  "total_flos": 2.527620759552e+17,
  "trial_name": null,
  "trial_params": null
}