{
  "best_metric": 1.9584715366363525,
  "best_model_checkpoint": "./outputs/checkpoint-4100",
  "epoch": 2.987249544626594,
  "eval_steps": 100,
  "global_step": 4100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 2.7198,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.6175436973571777,
      "eval_runtime": 206.1344,
      "eval_samples_per_second": 30.436,
      "eval_steps_per_second": 3.808,
      "step": 100
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 2.5808,
      "step": 200
    },
    {
      "epoch": 0.15,
      "eval_loss": 2.5663881301879883,
      "eval_runtime": 206.2235,
      "eval_samples_per_second": 30.423,
      "eval_steps_per_second": 3.807,
      "step": 200
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 2.5395,
      "step": 300
    },
    {
      "epoch": 0.22,
      "eval_loss": 2.52937912940979,
      "eval_runtime": 205.8891,
      "eval_samples_per_second": 30.473,
      "eval_steps_per_second": 3.813,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 2.5126,
      "step": 400
    },
    {
      "epoch": 0.29,
      "eval_loss": 2.496067762374878,
      "eval_runtime": 206.0161,
      "eval_samples_per_second": 30.454,
      "eval_steps_per_second": 3.81,
      "step": 400
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 2.4697,
      "step": 500
    },
    {
      "epoch": 0.36,
      "eval_loss": 2.4705872535705566,
      "eval_runtime": 206.051,
      "eval_samples_per_second": 30.449,
      "eval_steps_per_second": 3.81,
      "step": 500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 2.445,
      "step": 600
    },
    {
      "epoch": 0.44,
      "eval_loss": 2.445857048034668,
      "eval_runtime": 206.3699,
      "eval_samples_per_second": 30.402,
      "eval_steps_per_second": 3.804,
      "step": 600
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 2.4272,
      "step": 700
    },
    {
      "epoch": 0.51,
      "eval_loss": 2.4212183952331543,
      "eval_runtime": 205.7781,
      "eval_samples_per_second": 30.489,
      "eval_steps_per_second": 3.815,
      "step": 700
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 2.4046,
      "step": 800
    },
    {
      "epoch": 0.58,
      "eval_loss": 2.399564027786255,
      "eval_runtime": 205.9969,
      "eval_samples_per_second": 30.457,
      "eval_steps_per_second": 3.811,
      "step": 800
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 2.3686,
      "step": 900
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.3772318363189697,
      "eval_runtime": 206.4179,
      "eval_samples_per_second": 30.395,
      "eval_steps_per_second": 3.803,
      "step": 900
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 2.3654,
      "step": 1000
    },
    {
      "epoch": 0.73,
      "eval_loss": 2.3593590259552,
      "eval_runtime": 206.0361,
      "eval_samples_per_second": 30.451,
      "eval_steps_per_second": 3.81,
      "step": 1000
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 2.3585,
      "step": 1100
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.33705472946167,
      "eval_runtime": 206.2411,
      "eval_samples_per_second": 30.421,
      "eval_steps_per_second": 3.806,
      "step": 1100
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 2.3156,
      "step": 1200
    },
    {
      "epoch": 0.87,
      "eval_loss": 2.318722724914551,
      "eval_runtime": 205.7818,
      "eval_samples_per_second": 30.489,
      "eval_steps_per_second": 3.815,
      "step": 1200
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 2.3265,
      "step": 1300
    },
    {
      "epoch": 0.95,
      "eval_loss": 2.3025875091552734,
      "eval_runtime": 206.4689,
      "eval_samples_per_second": 30.387,
      "eval_steps_per_second": 3.802,
      "step": 1300
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 2.2785,
      "step": 1400
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.2806317806243896,
      "eval_runtime": 206.1627,
      "eval_samples_per_second": 30.432,
      "eval_steps_per_second": 3.808,
      "step": 1400
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 2.2404,
      "step": 1500
    },
    {
      "epoch": 1.09,
      "eval_loss": 2.264216899871826,
      "eval_runtime": 206.3537,
      "eval_samples_per_second": 30.404,
      "eval_steps_per_second": 3.804,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 2.232,
      "step": 1600
    },
    {
      "epoch": 1.17,
      "eval_loss": 2.2468743324279785,
      "eval_runtime": 206.622,
      "eval_samples_per_second": 30.365,
      "eval_steps_per_second": 3.799,
      "step": 1600
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 2.2326,
      "step": 1700
    },
    {
      "epoch": 1.24,
      "eval_loss": 2.231027126312256,
      "eval_runtime": 206.9085,
      "eval_samples_per_second": 30.323,
      "eval_steps_per_second": 3.794,
      "step": 1700
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 2.1938,
      "step": 1800
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.214425563812256,
      "eval_runtime": 294.4379,
      "eval_samples_per_second": 21.308,
      "eval_steps_per_second": 2.666,
      "step": 1800
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 2.1771,
      "step": 1900
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.1998817920684814,
      "eval_runtime": 206.8568,
      "eval_samples_per_second": 30.33,
      "eval_steps_per_second": 3.795,
      "step": 1900
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 2.1718,
      "step": 2000
    },
    {
      "epoch": 1.46,
      "eval_loss": 2.185276985168457,
      "eval_runtime": 206.8061,
      "eval_samples_per_second": 30.338,
      "eval_steps_per_second": 3.796,
      "step": 2000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 2.1536,
      "step": 2100
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.173062562942505,
      "eval_runtime": 207.2032,
      "eval_samples_per_second": 30.279,
      "eval_steps_per_second": 3.789,
      "step": 2100
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 2.1373,
      "step": 2200
    },
    {
      "epoch": 1.6,
      "eval_loss": 2.155045986175537,
      "eval_runtime": 206.5769,
      "eval_samples_per_second": 30.371,
      "eval_steps_per_second": 3.8,
      "step": 2200
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 2.1267,
      "step": 2300
    },
    {
      "epoch": 1.68,
      "eval_loss": 2.143136501312256,
      "eval_runtime": 206.6375,
      "eval_samples_per_second": 30.362,
      "eval_steps_per_second": 3.799,
      "step": 2300
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 2.1334,
      "step": 2400
    },
    {
      "epoch": 1.75,
      "eval_loss": 2.130755662918091,
      "eval_runtime": 206.6096,
      "eval_samples_per_second": 30.366,
      "eval_steps_per_second": 3.799,
      "step": 2400
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 2.101,
      "step": 2500
    },
    {
      "epoch": 1.82,
      "eval_loss": 2.1189475059509277,
      "eval_runtime": 218.5869,
      "eval_samples_per_second": 28.703,
      "eval_steps_per_second": 3.591,
      "step": 2500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 2.1023,
      "step": 2600
    },
    {
      "epoch": 1.89,
      "eval_loss": 2.1056134700775146,
      "eval_runtime": 206.4995,
      "eval_samples_per_second": 30.383,
      "eval_steps_per_second": 3.801,
      "step": 2600
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 2.0858,
      "step": 2700
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.091339588165283,
      "eval_runtime": 350.1545,
      "eval_samples_per_second": 17.918,
      "eval_steps_per_second": 2.242,
      "step": 2700
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 2.0422,
      "step": 2800
    },
    {
      "epoch": 2.04,
      "eval_loss": 2.0826971530914307,
      "eval_runtime": 350.5431,
      "eval_samples_per_second": 17.898,
      "eval_steps_per_second": 2.239,
      "step": 2800
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 2.013,
      "step": 2900
    },
    {
      "epoch": 2.11,
      "eval_loss": 2.070052146911621,
      "eval_runtime": 417.9713,
      "eval_samples_per_second": 15.011,
      "eval_steps_per_second": 1.878,
      "step": 2900
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 2.0363,
      "step": 3000
    },
    {
      "epoch": 2.19,
      "eval_loss": 2.062880277633667,
      "eval_runtime": 324.7358,
      "eval_samples_per_second": 19.32,
      "eval_steps_per_second": 2.417,
      "step": 3000
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 2.0072,
      "step": 3100
    },
    {
      "epoch": 2.26,
      "eval_loss": 2.051217794418335,
      "eval_runtime": 269.9913,
      "eval_samples_per_second": 23.238,
      "eval_steps_per_second": 2.908,
      "step": 3100
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0002,
      "loss": 2.02,
      "step": 3200
    },
    {
      "epoch": 2.33,
      "eval_loss": 2.039947509765625,
      "eval_runtime": 354.1313,
      "eval_samples_per_second": 17.717,
      "eval_steps_per_second": 2.217,
      "step": 3200
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0002,
      "loss": 1.9872,
      "step": 3300
    },
    {
      "epoch": 2.4,
      "eval_loss": 2.029729127883911,
      "eval_runtime": 353.5701,
      "eval_samples_per_second": 17.745,
      "eval_steps_per_second": 2.22,
      "step": 3300
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002,
      "loss": 1.9815,
      "step": 3400
    },
    {
      "epoch": 2.48,
      "eval_loss": 2.0244295597076416,
      "eval_runtime": 360.2966,
      "eval_samples_per_second": 17.413,
      "eval_steps_per_second": 2.179,
      "step": 3400
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002,
      "loss": 1.9772,
      "step": 3500
    },
    {
      "epoch": 2.55,
      "eval_loss": 2.010683536529541,
      "eval_runtime": 353.864,
      "eval_samples_per_second": 17.73,
      "eval_steps_per_second": 2.218,
      "step": 3500
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002,
      "loss": 1.9742,
      "step": 3600
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.001563310623169,
      "eval_runtime": 352.348,
      "eval_samples_per_second": 17.806,
      "eval_steps_per_second": 2.228,
      "step": 3600
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0002,
      "loss": 1.9551,
      "step": 3700
    },
    {
      "epoch": 2.7,
      "eval_loss": 1.9941232204437256,
      "eval_runtime": 353.227,
      "eval_samples_per_second": 17.762,
      "eval_steps_per_second": 2.222,
      "step": 3700
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0002,
      "loss": 1.9469,
      "step": 3800
    },
    {
      "epoch": 2.77,
      "eval_loss": 1.9865385293960571,
      "eval_runtime": 353.5789,
      "eval_samples_per_second": 17.744,
      "eval_steps_per_second": 2.22,
      "step": 3800
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0002,
      "loss": 1.9346,
      "step": 3900
    },
    {
      "epoch": 2.84,
      "eval_loss": 1.9748302698135376,
      "eval_runtime": 353.4471,
      "eval_samples_per_second": 17.751,
      "eval_steps_per_second": 2.221,
      "step": 3900
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0002,
      "loss": 1.937,
      "step": 4000
    },
    {
      "epoch": 2.91,
      "eval_loss": 1.9734764099121094,
      "eval_runtime": 353.802,
      "eval_samples_per_second": 17.733,
      "eval_steps_per_second": 2.219,
      "step": 4000
    },
    {
      "epoch": 2.99,
      "learning_rate": 0.0002,
      "loss": 1.9379,
      "step": 4100
    },
    {
      "epoch": 2.99,
      "eval_loss": 1.9584715366363525,
      "eval_runtime": 354.4627,
      "eval_samples_per_second": 17.7,
      "eval_steps_per_second": 2.215,
      "step": 4100
    }
  ],
  "logging_steps": 100,
  "max_steps": 4116,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 1.2037527679500288e+17,
  "trial_name": null,
  "trial_params": null
}