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": 7.113146781921387,
      "learning_rate": 4e-05,
      "loss": 1.2329,
      "step": 1787
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5976304411888123,
      "eval_rouge1": 0.3912,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3916,
      "eval_rougeLsum": 0.392,
      "eval_runtime": 114.4182,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 0.21,
      "step": 1787
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9403228759765625,
      "learning_rate": 3e-05,
      "loss": 0.7952,
      "step": 3574
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5580247044563293,
      "eval_rouge1": 0.3919,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3921,
      "eval_rougeLsum": 0.3921,
      "eval_runtime": 112.1899,
      "eval_samples_per_second": 6.658,
      "eval_steps_per_second": 0.214,
      "step": 3574
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7163584232330322,
      "learning_rate": 2e-05,
      "loss": 0.7407,
      "step": 5361
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5366058945655823,
      "eval_rouge1": 0.3893,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3879,
      "eval_rougeLsum": 0.3866,
      "eval_runtime": 112.7826,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 0.213,
      "step": 5361
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.9489381313323975,
      "learning_rate": 1e-05,
      "loss": 0.7152,
      "step": 7148
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5401661992073059,
      "eval_rouge1": 0.354,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3512,
      "eval_rougeLsum": 0.3523,
      "eval_runtime": 112.9681,
      "eval_samples_per_second": 6.612,
      "eval_steps_per_second": 0.212,
      "step": 7148
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.7675414085388184,
      "learning_rate": 0.0,
      "loss": 0.7029,
      "step": 8935
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5351393818855286,
      "eval_rouge1": 0.3585,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3555,
      "eval_rougeLsum": 0.357,
      "eval_runtime": 113.5022,
      "eval_samples_per_second": 6.581,
      "eval_steps_per_second": 0.211,
      "step": 8935
    },
    {
      "epoch": 5.0,
      "step": 8935,
      "total_flos": 7.531884760915354e+16,
      "train_loss": 0.8373744595297635,
      "train_runtime": 5174.0873,
      "train_samples_per_second": 13.809,
      "train_steps_per_second": 1.727
    }
  ],
  "logging_steps": 500,
  "max_steps": 8935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 7.531884760915354e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}