{
  "best_metric": 1.9712703227996826,
  "best_model_checkpoint": "./outputs/checkpoint-4000",
  "epoch": 2.9143897996357016,
  "eval_steps": 100,
  "global_step": 4000,
  "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": 2.7202,
      "step": 100
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.6173269748687744,
      "eval_runtime": 210.4944,
      "eval_samples_per_second": 29.806,
      "eval_steps_per_second": 3.729,
      "step": 100
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 2.5807,
      "step": 200
    },
    {
      "epoch": 0.15,
      "eval_loss": 2.566875696182251,
      "eval_runtime": 206.05,
      "eval_samples_per_second": 30.449,
      "eval_steps_per_second": 3.81,
      "step": 200
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 2.5403,
      "step": 300
    },
    {
      "epoch": 0.22,
      "eval_loss": 2.531456708908081,
      "eval_runtime": 205.7527,
      "eval_samples_per_second": 30.493,
      "eval_steps_per_second": 3.815,
      "step": 300
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 2.5135,
      "step": 400
    },
    {
      "epoch": 0.29,
      "eval_loss": 2.4987542629241943,
      "eval_runtime": 205.7584,
      "eval_samples_per_second": 30.492,
      "eval_steps_per_second": 3.815,
      "step": 400
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 2.4703,
      "step": 500
    },
    {
      "epoch": 0.36,
      "eval_loss": 2.4730148315429688,
      "eval_runtime": 205.8153,
      "eval_samples_per_second": 30.484,
      "eval_steps_per_second": 3.814,
      "step": 500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 2.4449,
      "step": 600
    },
    {
      "epoch": 0.44,
      "eval_loss": 2.4445066452026367,
      "eval_runtime": 205.9496,
      "eval_samples_per_second": 30.464,
      "eval_steps_per_second": 3.812,
      "step": 600
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 2.4265,
      "step": 700
    },
    {
      "epoch": 0.51,
      "eval_loss": 2.421610116958618,
      "eval_runtime": 206.1548,
      "eval_samples_per_second": 30.433,
      "eval_steps_per_second": 3.808,
      "step": 700
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 2.4059,
      "step": 800
    },
    {
      "epoch": 0.58,
      "eval_loss": 2.401268720626831,
      "eval_runtime": 206.1413,
      "eval_samples_per_second": 30.435,
      "eval_steps_per_second": 3.808,
      "step": 800
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 2.3694,
      "step": 900
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.3795337677001953,
      "eval_runtime": 206.144,
      "eval_samples_per_second": 30.435,
      "eval_steps_per_second": 3.808,
      "step": 900
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 2.3664,
      "step": 1000
    },
    {
      "epoch": 0.73,
      "eval_loss": 2.3618950843811035,
      "eval_runtime": 205.9336,
      "eval_samples_per_second": 30.466,
      "eval_steps_per_second": 3.812,
      "step": 1000
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 2.3599,
      "step": 1100
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.338806390762329,
      "eval_runtime": 205.8736,
      "eval_samples_per_second": 30.475,
      "eval_steps_per_second": 3.813,
      "step": 1100
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 2.3162,
      "step": 1200
    },
    {
      "epoch": 0.87,
      "eval_loss": 2.3209476470947266,
      "eval_runtime": 206.0155,
      "eval_samples_per_second": 30.454,
      "eval_steps_per_second": 3.81,
      "step": 1200
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 2.3272,
      "step": 1300
    },
    {
      "epoch": 0.95,
      "eval_loss": 2.3050100803375244,
      "eval_runtime": 205.8768,
      "eval_samples_per_second": 30.475,
      "eval_steps_per_second": 3.813,
      "step": 1300
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 2.2787,
      "step": 1400
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.2802278995513916,
      "eval_runtime": 206.0479,
      "eval_samples_per_second": 30.449,
      "eval_steps_per_second": 3.81,
      "step": 1400
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 2.2387,
      "step": 1500
    },
    {
      "epoch": 1.09,
      "eval_loss": 2.263134479522705,
      "eval_runtime": 205.9604,
      "eval_samples_per_second": 30.462,
      "eval_steps_per_second": 3.811,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 2.2303,
      "step": 1600
    },
    {
      "epoch": 1.17,
      "eval_loss": 2.246272325515747,
      "eval_runtime": 206.1382,
      "eval_samples_per_second": 30.436,
      "eval_steps_per_second": 3.808,
      "step": 1600
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 2.2323,
      "step": 1700
    },
    {
      "epoch": 1.24,
      "eval_loss": 2.231445789337158,
      "eval_runtime": 205.9926,
      "eval_samples_per_second": 30.457,
      "eval_steps_per_second": 3.811,
      "step": 1700
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 2.193,
      "step": 1800
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.2156155109405518,
      "eval_runtime": 217.8526,
      "eval_samples_per_second": 28.799,
      "eval_steps_per_second": 3.603,
      "step": 1800
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 2.1776,
      "step": 1900
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.202601909637451,
      "eval_runtime": 529.3371,
      "eval_samples_per_second": 11.853,
      "eval_steps_per_second": 1.483,
      "step": 1900
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 2.1728,
      "step": 2000
    },
    {
      "epoch": 1.46,
      "eval_loss": 2.1862785816192627,
      "eval_runtime": 232.6763,
      "eval_samples_per_second": 26.965,
      "eval_steps_per_second": 3.374,
      "step": 2000
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 2.1562,
      "step": 2100
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.174529552459717,
      "eval_runtime": 205.9758,
      "eval_samples_per_second": 30.46,
      "eval_steps_per_second": 3.811,
      "step": 2100
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 2.1386,
      "step": 2200
    },
    {
      "epoch": 1.6,
      "eval_loss": 2.1577847003936768,
      "eval_runtime": 206.4236,
      "eval_samples_per_second": 30.394,
      "eval_steps_per_second": 3.803,
      "step": 2200
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 2.1296,
      "step": 2300
    },
    {
      "epoch": 1.68,
      "eval_loss": 2.147472858428955,
      "eval_runtime": 206.3739,
      "eval_samples_per_second": 30.401,
      "eval_steps_per_second": 3.804,
      "step": 2300
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 2.1374,
      "step": 2400
    },
    {
      "epoch": 1.75,
      "eval_loss": 2.135169506072998,
      "eval_runtime": 206.2542,
      "eval_samples_per_second": 30.419,
      "eval_steps_per_second": 3.806,
      "step": 2400
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 2.1035,
      "step": 2500
    },
    {
      "epoch": 1.82,
      "eval_loss": 2.1208178997039795,
      "eval_runtime": 205.9676,
      "eval_samples_per_second": 30.461,
      "eval_steps_per_second": 3.811,
      "step": 2500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 2.1034,
      "step": 2600
    },
    {
      "epoch": 1.89,
      "eval_loss": 2.1102426052093506,
      "eval_runtime": 217.9461,
      "eval_samples_per_second": 28.787,
      "eval_steps_per_second": 3.602,
      "step": 2600
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 2.0912,
      "step": 2700
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.0947265625,
      "eval_runtime": 732.5061,
      "eval_samples_per_second": 8.565,
      "eval_steps_per_second": 1.072,
      "step": 2700
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 2.0451,
      "step": 2800
    },
    {
      "epoch": 2.04,
      "eval_loss": 2.0868542194366455,
      "eval_runtime": 206.5812,
      "eval_samples_per_second": 30.371,
      "eval_steps_per_second": 3.8,
      "step": 2800
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 2.0179,
      "step": 2900
    },
    {
      "epoch": 2.11,
      "eval_loss": 2.074641227722168,
      "eval_runtime": 206.384,
      "eval_samples_per_second": 30.4,
      "eval_steps_per_second": 3.804,
      "step": 2900
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 2.0386,
      "step": 3000
    },
    {
      "epoch": 2.19,
      "eval_loss": 2.0661864280700684,
      "eval_runtime": 206.2191,
      "eval_samples_per_second": 30.424,
      "eval_steps_per_second": 3.807,
      "step": 3000
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 2.0078,
      "step": 3100
    },
    {
      "epoch": 2.26,
      "eval_loss": 2.05454158782959,
      "eval_runtime": 206.3207,
      "eval_samples_per_second": 30.409,
      "eval_steps_per_second": 3.805,
      "step": 3100
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0002,
      "loss": 2.0222,
      "step": 3200
    },
    {
      "epoch": 2.33,
      "eval_loss": 2.0429155826568604,
      "eval_runtime": 207.289,
      "eval_samples_per_second": 30.267,
      "eval_steps_per_second": 3.787,
      "step": 3200
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0002,
      "loss": 1.9896,
      "step": 3300
    },
    {
      "epoch": 2.4,
      "eval_loss": 2.031524896621704,
      "eval_runtime": 254.9126,
      "eval_samples_per_second": 24.612,
      "eval_steps_per_second": 3.079,
      "step": 3300
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002,
      "loss": 1.9847,
      "step": 3400
    },
    {
      "epoch": 2.48,
      "eval_loss": 2.02608060836792,
      "eval_runtime": 206.1634,
      "eval_samples_per_second": 30.432,
      "eval_steps_per_second": 3.808,
      "step": 3400
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002,
      "loss": 1.9801,
      "step": 3500
    },
    {
      "epoch": 2.55,
      "eval_loss": 2.013916492462158,
      "eval_runtime": 215.475,
      "eval_samples_per_second": 29.117,
      "eval_steps_per_second": 3.643,
      "step": 3500
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002,
      "loss": 1.9764,
      "step": 3600
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.0044028759002686,
      "eval_runtime": 208.9999,
      "eval_samples_per_second": 30.019,
      "eval_steps_per_second": 3.756,
      "step": 3600
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0002,
      "loss": 1.9572,
      "step": 3700
    },
    {
      "epoch": 2.7,
      "eval_loss": 1.994431972503662,
      "eval_runtime": 209.7442,
      "eval_samples_per_second": 29.913,
      "eval_steps_per_second": 3.743,
      "step": 3700
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0002,
      "loss": 1.9505,
      "step": 3800
    },
    {
      "epoch": 2.77,
      "eval_loss": 1.9890899658203125,
      "eval_runtime": 225.3887,
      "eval_samples_per_second": 27.836,
      "eval_steps_per_second": 3.483,
      "step": 3800
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0002,
      "loss": 1.9371,
      "step": 3900
    },
    {
      "epoch": 2.84,
      "eval_loss": 1.9771522283554077,
      "eval_runtime": 216.1751,
      "eval_samples_per_second": 29.023,
      "eval_steps_per_second": 3.631,
      "step": 3900
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0002,
      "loss": 1.9399,
      "step": 4000
    },
    {
      "epoch": 2.91,
      "eval_loss": 1.9712703227996826,
      "eval_runtime": 210.5398,
      "eval_samples_per_second": 29.8,
      "eval_steps_per_second": 3.729,
      "step": 4000
    }
  ],
  "logging_steps": 100,
  "max_steps": 4116,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 1.1744177835220992e+17,
  "trial_name": null,
  "trial_params": null
}