trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 8935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.201539516448975,
      "learning_rate": 4e-05,
      "loss": 1.3998,
      "step": 1787
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.6384263634681702,
      "eval_rouge1": 0.3848,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3845,
      "eval_rougeLsum": 0.3857,
      "eval_runtime": 115.2126,
      "eval_samples_per_second": 6.484,
      "eval_steps_per_second": 0.208,
      "step": 1787
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.993088722229004,
      "learning_rate": 3e-05,
      "loss": 0.8601,
      "step": 3574
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5792745351791382,
      "eval_rouge1": 0.391,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.391,
      "eval_rougeLsum": 0.3923,
      "eval_runtime": 112.3658,
      "eval_samples_per_second": 6.648,
      "eval_steps_per_second": 0.214,
      "step": 3574
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.5980020761489868,
      "learning_rate": 2e-05,
      "loss": 0.7972,
      "step": 5361
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5575756430625916,
      "eval_rouge1": 0.3776,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3767,
      "eval_rougeLsum": 0.3768,
      "eval_runtime": 112.6817,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 0.213,
      "step": 5361
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.543159246444702,
      "learning_rate": 1e-05,
      "loss": 0.7698,
      "step": 7148
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5619614720344543,
      "eval_rouge1": 0.3579,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3565,
      "eval_rougeLsum": 0.3592,
      "eval_runtime": 112.2967,
      "eval_samples_per_second": 6.652,
      "eval_steps_per_second": 0.214,
      "step": 7148
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.1870193481445312,
      "learning_rate": 0.0,
      "loss": 0.7566,
      "step": 8935
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5532283186912537,
      "eval_rouge1": 0.3696,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3694,
      "eval_rougeLsum": 0.3713,
      "eval_runtime": 112.5191,
      "eval_samples_per_second": 6.639,
      "eval_steps_per_second": 0.213,
      "step": 8935
    },
    {
      "epoch": 5.0,
      "step": 8935,
      "total_flos": 7.502224981312512e+16,
      "train_loss": 0.9166960628191452,
      "train_runtime": 5169.5099,
      "train_samples_per_second": 13.821,
      "train_steps_per_second": 1.728
    }
  ],
  "logging_steps": 500,
  "max_steps": 8935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7.502224981312512e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}