{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.8656716417910446,
  "eval_steps": 500,
  "global_step": 48,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 7.837143285917327,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.2354,
      "step": 1
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.311491863709907,
      "learning_rate": 2e-05,
      "loss": 2.1965,
      "step": 5
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.970375942777289,
      "learning_rate": 1.9340161087325483e-05,
      "loss": 2.1045,
      "step": 10
    },
    {
      "epoch": 0.9,
      "grad_norm": 3.5227839335545066,
      "learning_rate": 1.744772182743782e-05,
      "loss": 2.2107,
      "step": 15
    },
    {
      "epoch": 0.96,
      "eval_loss": 1.5384026765823364,
      "eval_runtime": 1.7145,
      "eval_samples_per_second": 4.083,
      "eval_steps_per_second": 0.583,
      "step": 16
    },
    {
      "epoch": 1.19,
      "grad_norm": 4.57009081553093,
      "learning_rate": 1.4572423233046386e-05,
      "loss": 1.8611,
      "step": 20
    },
    {
      "epoch": 1.49,
      "grad_norm": 5.062546542837419,
      "learning_rate": 1.1093712083778748e-05,
      "loss": 1.533,
      "step": 25
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.8164674548597004,
      "learning_rate": 7.470666176083193e-06,
      "loss": 1.4572,
      "step": 30
    },
    {
      "epoch": 1.97,
      "eval_loss": 1.646084189414978,
      "eval_runtime": 1.7409,
      "eval_samples_per_second": 4.021,
      "eval_steps_per_second": 0.574,
      "step": 33
    },
    {
      "epoch": 2.09,
      "grad_norm": 9.370778150439074,
      "learning_rate": 4.181410844420473e-06,
      "loss": 1.342,
      "step": 35
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.2248384357705895,
      "learning_rate": 1.660021821101222e-06,
      "loss": 1.0214,
      "step": 40
    },
    {
      "epoch": 2.69,
      "grad_norm": 4.735590938981405,
      "learning_rate": 2.392412244407294e-07,
      "loss": 0.9565,
      "step": 45
    },
    {
      "epoch": 2.87,
      "eval_loss": 1.8004332780838013,
      "eval_runtime": 1.6722,
      "eval_samples_per_second": 4.186,
      "eval_steps_per_second": 0.598,
      "step": 48
    },
    {
      "epoch": 2.87,
      "step": 48,
      "total_flos": 9997878558720.0,
      "train_loss": 1.592680846651395,
      "train_runtime": 972.6889,
      "train_samples_per_second": 3.3,
      "train_steps_per_second": 0.049
    }
  ],
  "logging_steps": 5,
  "max_steps": 48,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 9997878558720.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}