{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 29.387755102040817,
  "eval_steps": 20,
  "global_step": 360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.82,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 1.7509,
      "step": 10
    },
    {
      "epoch": 1.63,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 1.6967,
      "step": 20
    },
    {
      "epoch": 1.63,
      "eval_loss": 1.6115258932113647,
      "eval_runtime": 30.6688,
      "eval_samples_per_second": 12.619,
      "eval_steps_per_second": 1.598,
      "step": 20
    },
    {
      "epoch": 2.45,
      "learning_rate": 8.999999999999999e-05,
      "loss": 1.5138,
      "step": 30
    },
    {
      "epoch": 3.27,
      "learning_rate": 0.00011999999999999999,
      "loss": 1.0417,
      "step": 40
    },
    {
      "epoch": 3.27,
      "eval_loss": 0.5990618467330933,
      "eval_runtime": 45.282,
      "eval_samples_per_second": 8.546,
      "eval_steps_per_second": 1.082,
      "step": 40
    },
    {
      "epoch": 4.08,
      "learning_rate": 0.00015,
      "loss": 0.4159,
      "step": 50
    },
    {
      "epoch": 4.9,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.3112,
      "step": 60
    },
    {
      "epoch": 4.9,
      "eval_loss": 0.277026891708374,
      "eval_runtime": 30.7302,
      "eval_samples_per_second": 12.593,
      "eval_steps_per_second": 1.595,
      "step": 60
    },
    {
      "epoch": 5.71,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.285,
      "step": 70
    },
    {
      "epoch": 6.53,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.27,
      "step": 80
    },
    {
      "epoch": 6.53,
      "eval_loss": 0.24872899055480957,
      "eval_runtime": 30.6498,
      "eval_samples_per_second": 12.627,
      "eval_steps_per_second": 1.599,
      "step": 80
    },
    {
      "epoch": 7.35,
      "learning_rate": 0.00027,
      "loss": 0.2464,
      "step": 90
    },
    {
      "epoch": 8.16,
      "learning_rate": 0.0003,
      "loss": 0.2276,
      "step": 100
    },
    {
      "epoch": 8.16,
      "eval_loss": 0.22304466366767883,
      "eval_runtime": 30.6362,
      "eval_samples_per_second": 12.632,
      "eval_steps_per_second": 1.599,
      "step": 100
    },
    {
      "epoch": 8.98,
      "learning_rate": 0.00029,
      "loss": 0.1993,
      "step": 110
    },
    {
      "epoch": 9.8,
      "learning_rate": 0.00028,
      "loss": 0.1747,
      "step": 120
    },
    {
      "epoch": 9.8,
      "eval_loss": 0.20949400961399078,
      "eval_runtime": 30.5837,
      "eval_samples_per_second": 12.654,
      "eval_steps_per_second": 1.602,
      "step": 120
    },
    {
      "epoch": 10.61,
      "learning_rate": 0.00027,
      "loss": 0.1408,
      "step": 130
    },
    {
      "epoch": 11.43,
      "learning_rate": 0.00026,
      "loss": 0.1131,
      "step": 140
    },
    {
      "epoch": 11.43,
      "eval_loss": 0.20713694393634796,
      "eval_runtime": 30.5887,
      "eval_samples_per_second": 12.652,
      "eval_steps_per_second": 1.602,
      "step": 140
    },
    {
      "epoch": 12.24,
      "learning_rate": 0.00025,
      "loss": 0.0839,
      "step": 150
    },
    {
      "epoch": 13.06,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.0644,
      "step": 160
    },
    {
      "epoch": 13.06,
      "eval_loss": 0.2342994511127472,
      "eval_runtime": 30.5213,
      "eval_samples_per_second": 12.68,
      "eval_steps_per_second": 1.605,
      "step": 160
    },
    {
      "epoch": 13.88,
      "learning_rate": 0.00023,
      "loss": 0.0466,
      "step": 170
    },
    {
      "epoch": 14.69,
      "learning_rate": 0.00021999999999999995,
      "loss": 0.0365,
      "step": 180
    },
    {
      "epoch": 14.69,
      "eval_loss": 0.263688862323761,
      "eval_runtime": 30.5969,
      "eval_samples_per_second": 12.648,
      "eval_steps_per_second": 1.601,
      "step": 180
    },
    {
      "epoch": 15.51,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.029,
      "step": 190
    },
    {
      "epoch": 16.33,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.0244,
      "step": 200
    },
    {
      "epoch": 16.33,
      "eval_loss": 0.28023940324783325,
      "eval_runtime": 30.5334,
      "eval_samples_per_second": 12.675,
      "eval_steps_per_second": 1.605,
      "step": 200
    },
    {
      "epoch": 17.14,
      "learning_rate": 0.00018999999999999998,
      "loss": 0.0208,
      "step": 210
    },
    {
      "epoch": 17.96,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.0186,
      "step": 220
    },
    {
      "epoch": 17.96,
      "eval_loss": 0.30082279443740845,
      "eval_runtime": 30.6266,
      "eval_samples_per_second": 12.636,
      "eval_steps_per_second": 1.6,
      "step": 220
    },
    {
      "epoch": 18.78,
      "learning_rate": 0.00016999999999999999,
      "loss": 0.0174,
      "step": 230
    },
    {
      "epoch": 19.59,
      "learning_rate": 0.00015999999999999999,
      "loss": 0.0161,
      "step": 240
    },
    {
      "epoch": 19.59,
      "eval_loss": 0.3072655200958252,
      "eval_runtime": 30.5563,
      "eval_samples_per_second": 12.665,
      "eval_steps_per_second": 1.604,
      "step": 240
    },
    {
      "epoch": 20.41,
      "learning_rate": 0.00015,
      "loss": 0.0153,
      "step": 250
    },
    {
      "epoch": 21.22,
      "learning_rate": 0.00014,
      "loss": 0.0147,
      "step": 260
    },
    {
      "epoch": 21.22,
      "eval_loss": 0.32090628147125244,
      "eval_runtime": 30.6058,
      "eval_samples_per_second": 12.645,
      "eval_steps_per_second": 1.601,
      "step": 260
    },
    {
      "epoch": 22.04,
      "learning_rate": 0.00013,
      "loss": 0.0145,
      "step": 270
    },
    {
      "epoch": 22.86,
      "learning_rate": 0.00011999999999999999,
      "loss": 0.0139,
      "step": 280
    },
    {
      "epoch": 22.86,
      "eval_loss": 0.32418766617774963,
      "eval_runtime": 30.6034,
      "eval_samples_per_second": 12.646,
      "eval_steps_per_second": 1.601,
      "step": 280
    },
    {
      "epoch": 23.67,
      "learning_rate": 0.00010999999999999998,
      "loss": 0.0138,
      "step": 290
    },
    {
      "epoch": 24.49,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.0135,
      "step": 300
    },
    {
      "epoch": 24.49,
      "eval_loss": 0.33807095885276794,
      "eval_runtime": 30.7038,
      "eval_samples_per_second": 12.604,
      "eval_steps_per_second": 1.596,
      "step": 300
    },
    {
      "epoch": 25.31,
      "learning_rate": 8.999999999999999e-05,
      "loss": 0.0134,
      "step": 310
    },
    {
      "epoch": 26.12,
      "learning_rate": 7.999999999999999e-05,
      "loss": 0.0133,
      "step": 320
    },
    {
      "epoch": 26.12,
      "eval_loss": 0.33249062299728394,
      "eval_runtime": 30.6095,
      "eval_samples_per_second": 12.643,
      "eval_steps_per_second": 1.601,
      "step": 320
    },
    {
      "epoch": 26.94,
      "learning_rate": 7e-05,
      "loss": 0.0131,
      "step": 330
    },
    {
      "epoch": 27.76,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 0.0131,
      "step": 340
    },
    {
      "epoch": 27.76,
      "eval_loss": 0.3382267951965332,
      "eval_runtime": 30.6161,
      "eval_samples_per_second": 12.64,
      "eval_steps_per_second": 1.6,
      "step": 340
    },
    {
      "epoch": 28.57,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 0.0128,
      "step": 350
    },
    {
      "epoch": 29.39,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 0.0127,
      "step": 360
    },
    {
      "epoch": 29.39,
      "eval_loss": 0.3383706212043762,
      "eval_runtime": 30.6092,
      "eval_samples_per_second": 12.643,
      "eval_steps_per_second": 1.601,
      "step": 360
    }
  ],
  "logging_steps": 10,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 34,
  "save_steps": 20,
  "total_flos": 9.164499307808686e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}