{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2,
      "grad_norm": 187.0,
      "learning_rate": 4e-05,
      "loss": 36.0545,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 59.75,
      "learning_rate": 0.0002,
      "loss": 29.9207,
      "step": 5
    },
    {
      "epoch": 1.0,
      "eval_loss": 14.432188034057617,
      "eval_runtime": 0.2474,
      "eval_samples_per_second": 40.424,
      "eval_steps_per_second": 4.042,
      "step": 5
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.75,
      "learning_rate": 0.00019396926207859084,
      "loss": 18.5488,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 10.560973167419434,
      "eval_runtime": 0.2335,
      "eval_samples_per_second": 42.825,
      "eval_steps_per_second": 4.282,
      "step": 10
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.53125,
      "learning_rate": 0.0001766044443118978,
      "loss": 15.0605,
      "step": 15
    },
    {
      "epoch": 3.0,
      "eval_loss": 8.22590160369873,
      "eval_runtime": 0.2341,
      "eval_samples_per_second": 42.711,
      "eval_steps_per_second": 4.271,
      "step": 15
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.3125,
      "learning_rate": 0.00015000000000000001,
      "loss": 13.929,
      "step": 20
    },
    {
      "epoch": 4.0,
      "eval_loss": 7.625598907470703,
      "eval_runtime": 0.2337,
      "eval_samples_per_second": 42.796,
      "eval_steps_per_second": 4.28,
      "step": 20
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.78125,
      "learning_rate": 0.00011736481776669306,
      "loss": 13.2624,
      "step": 25
    },
    {
      "epoch": 5.0,
      "eval_loss": 7.346440315246582,
      "eval_runtime": 0.234,
      "eval_samples_per_second": 42.736,
      "eval_steps_per_second": 4.274,
      "step": 25
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.625,
      "learning_rate": 8.263518223330697e-05,
      "loss": 12.6383,
      "step": 30
    },
    {
      "epoch": 6.0,
      "eval_loss": 7.076879024505615,
      "eval_runtime": 0.2334,
      "eval_samples_per_second": 42.842,
      "eval_steps_per_second": 4.284,
      "step": 30
    },
    {
      "epoch": 7.0,
      "grad_norm": 9.0,
      "learning_rate": 5.000000000000002e-05,
      "loss": 11.963,
      "step": 35
    },
    {
      "epoch": 7.0,
      "eval_loss": 6.880378723144531,
      "eval_runtime": 0.2349,
      "eval_samples_per_second": 42.563,
      "eval_steps_per_second": 4.256,
      "step": 35
    },
    {
      "epoch": 8.0,
      "grad_norm": 10.125,
      "learning_rate": 2.339555568810221e-05,
      "loss": 11.3592,
      "step": 40
    },
    {
      "epoch": 8.0,
      "eval_loss": 6.748847007751465,
      "eval_runtime": 0.2339,
      "eval_samples_per_second": 42.76,
      "eval_steps_per_second": 4.276,
      "step": 40
    },
    {
      "epoch": 9.0,
      "grad_norm": 10.5,
      "learning_rate": 6.030737921409169e-06,
      "loss": 11.0502,
      "step": 45
    },
    {
      "epoch": 9.0,
      "eval_loss": 6.706507682800293,
      "eval_runtime": 0.2365,
      "eval_samples_per_second": 42.29,
      "eval_steps_per_second": 4.229,
      "step": 45
    },
    {
      "epoch": 10.0,
      "grad_norm": 10.625,
      "learning_rate": 0.0,
      "loss": 10.9441,
      "step": 50
    },
    {
      "epoch": 10.0,
      "eval_loss": 6.698519229888916,
      "eval_runtime": 0.2341,
      "eval_samples_per_second": 42.721,
      "eval_steps_per_second": 4.272,
      "step": 50
    },
    {
      "epoch": 10.0,
      "step": 50,
      "total_flos": 1.52458893266518e+17,
      "train_loss": 14.990281143188476,
      "train_runtime": 123.5377,
      "train_samples_per_second": 24.851,
      "train_steps_per_second": 0.405
    }
  ],
  "logging_steps": 5,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 1.52458893266518e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}