{
  "best_metric": 0.645846426486969,
  "best_model_checkpoint": "models/checkpoints/checkpoint-9000",
  "epoch": 16.2748643761302,
  "global_step": 9000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "learning_rate": 1.9819168173598556e-05,
      "loss": 1.9232,
      "step": 100
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.9638336347197107e-05,
      "loss": 1.3768,
      "step": 200
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.9457504520795662e-05,
      "loss": 1.269,
      "step": 300
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.9276672694394213e-05,
      "loss": 1.2114,
      "step": 400
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.9095840867992768e-05,
      "loss": 1.1765,
      "step": 500
    },
    {
      "epoch": 0.9,
      "eval_loss": 1.0532047748565674,
      "eval_runtime": 12.8374,
      "eval_samples_per_second": 612.197,
      "eval_steps_per_second": 4.83,
      "step": 500
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.8915009041591322e-05,
      "loss": 1.1195,
      "step": 600
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.8734177215189874e-05,
      "loss": 1.1068,
      "step": 700
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.8553345388788428e-05,
      "loss": 1.0338,
      "step": 800
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.8372513562386983e-05,
      "loss": 1.0417,
      "step": 900
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.8191681735985537e-05,
      "loss": 1.0382,
      "step": 1000
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.964276909828186,
      "eval_runtime": 12.8331,
      "eval_samples_per_second": 612.399,
      "eval_steps_per_second": 4.831,
      "step": 1000
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.801084990958409e-05,
      "loss": 1.0084,
      "step": 1100
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.783001808318264e-05,
      "loss": 0.9976,
      "step": 1200
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.7649186256781194e-05,
      "loss": 0.9962,
      "step": 1300
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.746835443037975e-05,
      "loss": 0.953,
      "step": 1400
    },
    {
      "epoch": 2.71,
      "learning_rate": 1.72875226039783e-05,
      "loss": 0.9687,
      "step": 1500
    },
    {
      "epoch": 2.71,
      "eval_loss": 0.9130759239196777,
      "eval_runtime": 12.8023,
      "eval_samples_per_second": 613.873,
      "eval_steps_per_second": 4.843,
      "step": 1500
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.7106690777576855e-05,
      "loss": 0.9551,
      "step": 1600
    },
    {
      "epoch": 3.07,
      "learning_rate": 1.692585895117541e-05,
      "loss": 0.9156,
      "step": 1700
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.6745027124773964e-05,
      "loss": 0.9072,
      "step": 1800
    },
    {
      "epoch": 3.44,
      "learning_rate": 1.6564195298372515e-05,
      "loss": 0.9038,
      "step": 1900
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.6383363471971066e-05,
      "loss": 0.9157,
      "step": 2000
    },
    {
      "epoch": 3.62,
      "eval_loss": 0.8786566853523254,
      "eval_runtime": 12.8067,
      "eval_samples_per_second": 613.662,
      "eval_steps_per_second": 4.841,
      "step": 2000
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.620253164556962e-05,
      "loss": 0.8994,
      "step": 2100
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.6021699819168176e-05,
      "loss": 0.8963,
      "step": 2200
    },
    {
      "epoch": 4.16,
      "learning_rate": 1.584086799276673e-05,
      "loss": 0.8878,
      "step": 2300
    },
    {
      "epoch": 4.34,
      "learning_rate": 1.566003616636528e-05,
      "loss": 0.8739,
      "step": 2400
    },
    {
      "epoch": 4.52,
      "learning_rate": 1.5479204339963836e-05,
      "loss": 0.8671,
      "step": 2500
    },
    {
      "epoch": 4.52,
      "eval_loss": 0.8628956079483032,
      "eval_runtime": 12.8156,
      "eval_samples_per_second": 613.239,
      "eval_steps_per_second": 4.838,
      "step": 2500
    },
    {
      "epoch": 4.7,
      "learning_rate": 1.5298372513562387e-05,
      "loss": 0.8676,
      "step": 2600
    },
    {
      "epoch": 4.88,
      "learning_rate": 1.5117540687160942e-05,
      "loss": 0.8329,
      "step": 2700
    },
    {
      "epoch": 5.06,
      "learning_rate": 1.4936708860759495e-05,
      "loss": 0.8438,
      "step": 2800
    },
    {
      "epoch": 5.24,
      "learning_rate": 1.4755877034358048e-05,
      "loss": 0.8294,
      "step": 2900
    },
    {
      "epoch": 5.42,
      "learning_rate": 1.4575045207956602e-05,
      "loss": 0.8259,
      "step": 3000
    },
    {
      "epoch": 5.42,
      "eval_loss": 0.825766921043396,
      "eval_runtime": 12.7989,
      "eval_samples_per_second": 614.036,
      "eval_steps_per_second": 4.844,
      "step": 3000
    },
    {
      "epoch": 5.61,
      "learning_rate": 1.4394213381555155e-05,
      "loss": 0.8065,
      "step": 3100
    },
    {
      "epoch": 5.79,
      "learning_rate": 1.421338155515371e-05,
      "loss": 0.8193,
      "step": 3200
    },
    {
      "epoch": 5.97,
      "learning_rate": 1.403254972875226e-05,
      "loss": 0.8152,
      "step": 3300
    },
    {
      "epoch": 6.15,
      "learning_rate": 1.3851717902350814e-05,
      "loss": 0.784,
      "step": 3400
    },
    {
      "epoch": 6.33,
      "learning_rate": 1.3670886075949368e-05,
      "loss": 0.7941,
      "step": 3500
    },
    {
      "epoch": 6.33,
      "eval_loss": 0.8114328384399414,
      "eval_runtime": 12.8036,
      "eval_samples_per_second": 613.81,
      "eval_steps_per_second": 4.842,
      "step": 3500
    },
    {
      "epoch": 6.51,
      "learning_rate": 1.3490054249547921e-05,
      "loss": 0.7952,
      "step": 3600
    },
    {
      "epoch": 6.69,
      "learning_rate": 1.3309222423146476e-05,
      "loss": 0.8168,
      "step": 3700
    },
    {
      "epoch": 6.87,
      "learning_rate": 1.3128390596745029e-05,
      "loss": 0.7917,
      "step": 3800
    },
    {
      "epoch": 7.05,
      "learning_rate": 1.2947558770343582e-05,
      "loss": 0.7582,
      "step": 3900
    },
    {
      "epoch": 7.23,
      "learning_rate": 1.2766726943942136e-05,
      "loss": 0.7605,
      "step": 4000
    },
    {
      "epoch": 7.23,
      "eval_loss": 0.7806207537651062,
      "eval_runtime": 12.8016,
      "eval_samples_per_second": 613.906,
      "eval_steps_per_second": 4.843,
      "step": 4000
    },
    {
      "epoch": 7.41,
      "learning_rate": 1.2585895117540687e-05,
      "loss": 0.7651,
      "step": 4100
    },
    {
      "epoch": 7.59,
      "learning_rate": 1.240506329113924e-05,
      "loss": 0.7716,
      "step": 4200
    },
    {
      "epoch": 7.78,
      "learning_rate": 1.2224231464737795e-05,
      "loss": 0.7704,
      "step": 4300
    },
    {
      "epoch": 7.96,
      "learning_rate": 1.2043399638336348e-05,
      "loss": 0.7483,
      "step": 4400
    },
    {
      "epoch": 8.14,
      "learning_rate": 1.1862567811934902e-05,
      "loss": 0.7323,
      "step": 4500
    },
    {
      "epoch": 8.14,
      "eval_loss": 0.7683702707290649,
      "eval_runtime": 12.8031,
      "eval_samples_per_second": 613.836,
      "eval_steps_per_second": 4.843,
      "step": 4500
    },
    {
      "epoch": 8.32,
      "learning_rate": 1.1681735985533455e-05,
      "loss": 0.7432,
      "step": 4600
    },
    {
      "epoch": 8.5,
      "learning_rate": 1.150090415913201e-05,
      "loss": 0.7486,
      "step": 4700
    },
    {
      "epoch": 8.68,
      "learning_rate": 1.1320072332730561e-05,
      "loss": 0.7479,
      "step": 4800
    },
    {
      "epoch": 8.86,
      "learning_rate": 1.1139240506329114e-05,
      "loss": 0.7192,
      "step": 4900
    },
    {
      "epoch": 9.04,
      "learning_rate": 1.0958408679927669e-05,
      "loss": 0.7292,
      "step": 5000
    },
    {
      "epoch": 9.04,
      "eval_loss": 0.7645160555839539,
      "eval_runtime": 12.802,
      "eval_samples_per_second": 613.888,
      "eval_steps_per_second": 4.843,
      "step": 5000
    },
    {
      "epoch": 9.22,
      "learning_rate": 1.0777576853526221e-05,
      "loss": 0.7055,
      "step": 5100
    },
    {
      "epoch": 9.4,
      "learning_rate": 1.0596745027124774e-05,
      "loss": 0.7094,
      "step": 5200
    },
    {
      "epoch": 9.58,
      "learning_rate": 1.0415913200723329e-05,
      "loss": 0.7297,
      "step": 5300
    },
    {
      "epoch": 9.76,
      "learning_rate": 1.0235081374321882e-05,
      "loss": 0.7096,
      "step": 5400
    },
    {
      "epoch": 9.95,
      "learning_rate": 1.0054249547920433e-05,
      "loss": 0.7259,
      "step": 5500
    },
    {
      "epoch": 9.95,
      "eval_loss": 0.7333822846412659,
      "eval_runtime": 12.806,
      "eval_samples_per_second": 613.699,
      "eval_steps_per_second": 4.841,
      "step": 5500
    },
    {
      "epoch": 10.13,
      "learning_rate": 9.87341772151899e-06,
      "loss": 0.7138,
      "step": 5600
    },
    {
      "epoch": 10.31,
      "learning_rate": 9.69258589511754e-06,
      "loss": 0.7118,
      "step": 5700
    },
    {
      "epoch": 10.49,
      "learning_rate": 9.511754068716095e-06,
      "loss": 0.7091,
      "step": 5800
    },
    {
      "epoch": 10.67,
      "learning_rate": 9.330922242314648e-06,
      "loss": 0.6839,
      "step": 5900
    },
    {
      "epoch": 10.85,
      "learning_rate": 9.150090415913203e-06,
      "loss": 0.7045,
      "step": 6000
    },
    {
      "epoch": 10.85,
      "eval_loss": 0.7454974055290222,
      "eval_runtime": 12.7967,
      "eval_samples_per_second": 614.144,
      "eval_steps_per_second": 4.845,
      "step": 6000
    },
    {
      "epoch": 11.03,
      "learning_rate": 8.969258589511754e-06,
      "loss": 0.7059,
      "step": 6100
    },
    {
      "epoch": 11.21,
      "learning_rate": 8.788426763110308e-06,
      "loss": 0.6759,
      "step": 6200
    },
    {
      "epoch": 11.39,
      "learning_rate": 8.607594936708861e-06,
      "loss": 0.6806,
      "step": 6300
    },
    {
      "epoch": 11.57,
      "learning_rate": 8.426763110307414e-06,
      "loss": 0.6734,
      "step": 6400
    },
    {
      "epoch": 11.75,
      "learning_rate": 8.245931283905967e-06,
      "loss": 0.677,
      "step": 6500
    },
    {
      "epoch": 11.75,
      "eval_loss": 0.7277078032493591,
      "eval_runtime": 12.8036,
      "eval_samples_per_second": 613.809,
      "eval_steps_per_second": 4.842,
      "step": 6500
    },
    {
      "epoch": 11.93,
      "learning_rate": 8.065099457504522e-06,
      "loss": 0.6811,
      "step": 6600
    },
    {
      "epoch": 12.12,
      "learning_rate": 7.884267631103075e-06,
      "loss": 0.669,
      "step": 6700
    },
    {
      "epoch": 12.3,
      "learning_rate": 7.703435804701628e-06,
      "loss": 0.6586,
      "step": 6800
    },
    {
      "epoch": 12.48,
      "learning_rate": 7.522603978300181e-06,
      "loss": 0.6428,
      "step": 6900
    },
    {
      "epoch": 12.66,
      "learning_rate": 7.341772151898735e-06,
      "loss": 0.6618,
      "step": 7000
    },
    {
      "epoch": 12.66,
      "eval_loss": 0.7127390503883362,
      "eval_runtime": 12.8054,
      "eval_samples_per_second": 613.727,
      "eval_steps_per_second": 4.842,
      "step": 7000
    },
    {
      "epoch": 12.84,
      "learning_rate": 7.160940325497288e-06,
      "loss": 0.6703,
      "step": 7100
    },
    {
      "epoch": 13.02,
      "learning_rate": 6.980108499095841e-06,
      "loss": 0.6577,
      "step": 7200
    },
    {
      "epoch": 13.2,
      "learning_rate": 6.799276672694395e-06,
      "loss": 0.6431,
      "step": 7300
    },
    {
      "epoch": 13.38,
      "learning_rate": 6.618444846292948e-06,
      "loss": 0.6525,
      "step": 7400
    },
    {
      "epoch": 13.56,
      "learning_rate": 6.437613019891501e-06,
      "loss": 0.6817,
      "step": 7500
    },
    {
      "epoch": 13.56,
      "eval_loss": 0.7023378014564514,
      "eval_runtime": 12.8161,
      "eval_samples_per_second": 613.215,
      "eval_steps_per_second": 4.838,
      "step": 7500
    },
    {
      "epoch": 13.74,
      "learning_rate": 6.256781193490055e-06,
      "loss": 0.6523,
      "step": 7600
    },
    {
      "epoch": 13.92,
      "learning_rate": 6.075949367088608e-06,
      "loss": 0.6679,
      "step": 7700
    },
    {
      "epoch": 14.1,
      "learning_rate": 5.895117540687162e-06,
      "loss": 0.6368,
      "step": 7800
    },
    {
      "epoch": 14.29,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.641,
      "step": 7900
    },
    {
      "epoch": 14.47,
      "learning_rate": 5.533453887884268e-06,
      "loss": 0.6337,
      "step": 8000
    },
    {
      "epoch": 14.47,
      "eval_loss": 0.7140629887580872,
      "eval_runtime": 12.8021,
      "eval_samples_per_second": 613.882,
      "eval_steps_per_second": 4.843,
      "step": 8000
    },
    {
      "epoch": 14.65,
      "learning_rate": 5.352622061482822e-06,
      "loss": 0.6359,
      "step": 8100
    },
    {
      "epoch": 14.83,
      "learning_rate": 5.171790235081374e-06,
      "loss": 0.6298,
      "step": 8200
    },
    {
      "epoch": 15.01,
      "learning_rate": 4.990958408679928e-06,
      "loss": 0.6444,
      "step": 8300
    },
    {
      "epoch": 15.19,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 0.639,
      "step": 8400
    },
    {
      "epoch": 15.37,
      "learning_rate": 4.6292947558770344e-06,
      "loss": 0.6182,
      "step": 8500
    },
    {
      "epoch": 15.37,
      "eval_loss": 0.7226015329360962,
      "eval_runtime": 12.7692,
      "eval_samples_per_second": 615.466,
      "eval_steps_per_second": 4.855,
      "step": 8500
    },
    {
      "epoch": 15.55,
      "learning_rate": 4.448462929475588e-06,
      "loss": 0.6425,
      "step": 8600
    },
    {
      "epoch": 15.73,
      "learning_rate": 4.267631103074141e-06,
      "loss": 0.6255,
      "step": 8700
    },
    {
      "epoch": 15.91,
      "learning_rate": 4.086799276672695e-06,
      "loss": 0.6223,
      "step": 8800
    },
    {
      "epoch": 16.09,
      "learning_rate": 3.905967450271248e-06,
      "loss": 0.6303,
      "step": 8900
    },
    {
      "epoch": 16.27,
      "learning_rate": 3.7251356238698015e-06,
      "loss": 0.6367,
      "step": 9000
    },
    {
      "epoch": 16.27,
      "eval_loss": 0.645846426486969,
      "eval_runtime": 12.785,
      "eval_samples_per_second": 614.705,
      "eval_steps_per_second": 4.849,
      "step": 9000
    }
  ],
  "max_steps": 11060,
  "num_train_epochs": 20,
  "total_flos": 5.947714605600499e+16,
  "trial_name": null,
  "trial_params": null
}