{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 2590,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.9305019305019306,
      "grad_norm": 3.1227238178253174,
      "learning_rate": 8.06949806949807e-06,
      "loss": 0.0861,
      "step": 500
    },
    {
      "epoch": 1.9305019305019306,
      "eval_runtime": 661.7783,
      "eval_samples_per_second": 266.734,
      "eval_steps_per_second": 16.672,
      "step": 500
    },
    {
      "epoch": 3.861003861003861,
      "grad_norm": 4.147549152374268,
      "learning_rate": 6.13899613899614e-06,
      "loss": 0.0554,
      "step": 1000
    },
    {
      "epoch": 3.861003861003861,
      "eval_runtime": 639.2021,
      "eval_samples_per_second": 276.155,
      "eval_steps_per_second": 17.261,
      "step": 1000
    },
    {
      "epoch": 5.7915057915057915,
      "grad_norm": 4.684039115905762,
      "learning_rate": 4.208494208494209e-06,
      "loss": 0.0428,
      "step": 1500
    },
    {
      "epoch": 5.7915057915057915,
      "eval_runtime": 638.2108,
      "eval_samples_per_second": 276.584,
      "eval_steps_per_second": 17.287,
      "step": 1500
    },
    {
      "epoch": 7.722007722007722,
      "grad_norm": 3.6964356899261475,
      "learning_rate": 2.2779922779922782e-06,
      "loss": 0.0351,
      "step": 2000
    },
    {
      "epoch": 7.722007722007722,
      "eval_runtime": 637.2615,
      "eval_samples_per_second": 276.996,
      "eval_steps_per_second": 17.313,
      "step": 2000
    },
    {
      "epoch": 9.652509652509652,
      "grad_norm": 0.3747519552707672,
      "learning_rate": 3.474903474903475e-07,
      "loss": 0.0314,
      "step": 2500
    },
    {
      "epoch": 9.652509652509652,
      "eval_runtime": 643.7118,
      "eval_samples_per_second": 274.221,
      "eval_steps_per_second": 17.14,
      "step": 2500
    },
    {
      "epoch": 10.0,
      "step": 2590,
      "total_flos": 1.1089078720895386e+19,
      "train_loss": 0.04947091428469507,
      "train_runtime": 17619.1199,
      "train_samples_per_second": 300.557,
      "train_steps_per_second": 0.147
    }
  ],
  "logging_steps": 500,
  "max_steps": 2590,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "total_flos": 1.1089078720895386e+19,
  "train_batch_size": 2048,
  "trial_name": null,
  "trial_params": null
}