checkpoint-188/trainer_state.json

{
  "best_metric": 2.708937406539917,
  "best_model_checkpoint": "./LongT5-Large-NSPCC/checkpoint-188",
  "epoch": 0.9960264900662251,
  "eval_steps": 500,
  "global_step": 188,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9960264900662251,
      "grad_norm": 1.2397888898849487,
      "learning_rate": 0.00029303322414305795,
      "loss": 6.0401,
      "step": 188
    },
    {
      "epoch": 0.9960264900662251,
      "eval_gen_len": 151.7021,
      "eval_loss": 2.708937406539917,
      "eval_rouge1": 0.2766,
      "eval_rouge2": 0.0617,
      "eval_rougeL": 0.1655,
      "eval_rougeLsum": 0.1657,
      "eval_runtime": 1372.78,
      "eval_samples_per_second": 0.068,
      "eval_steps_per_second": 0.068,
      "step": 188
    }
  ],
  "logging_steps": 500,
  "max_steps": 1504,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 5.56850177114112e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}