trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 17865,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.893374919891357,
      "learning_rate": 4e-05,
      "loss": 0.6352,
      "step": 3573
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.46136122941970825,
      "eval_rouge1": 0.3871,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3846,
      "eval_rougeLsum": 0.3884,
      "eval_runtime": 336.5055,
      "eval_samples_per_second": 2.22,
      "eval_steps_per_second": 0.279,
      "step": 3573
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.110297679901123,
      "learning_rate": 3e-05,
      "loss": 0.4361,
      "step": 7146
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.4357260465621948,
      "eval_rouge1": 0.3574,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3543,
      "eval_rougeLsum": 0.3544,
      "eval_runtime": 327.4958,
      "eval_samples_per_second": 2.281,
      "eval_steps_per_second": 0.287,
      "step": 7146
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.471202850341797,
      "learning_rate": 2e-05,
      "loss": 0.3391,
      "step": 10719
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.4478732943534851,
      "eval_rouge1": 0.3973,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3975,
      "eval_rougeLsum": 0.4009,
      "eval_runtime": 333.0142,
      "eval_samples_per_second": 2.243,
      "eval_steps_per_second": 0.282,
      "step": 10719
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.917109966278076,
      "learning_rate": 1e-05,
      "loss": 0.2686,
      "step": 14292
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.4638811945915222,
      "eval_rouge1": 0.4113,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4102,
      "eval_rougeLsum": 0.4115,
      "eval_runtime": 334.1059,
      "eval_samples_per_second": 2.236,
      "eval_steps_per_second": 0.281,
      "step": 14292
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.981204092502594,
      "learning_rate": 0.0,
      "loss": 0.2221,
      "step": 17865
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.48620468378067017,
      "eval_rouge1": 0.3867,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.3833,
      "eval_rougeLsum": 0.386,
      "eval_runtime": 329.3468,
      "eval_samples_per_second": 2.268,
      "eval_steps_per_second": 0.285,
      "step": 17865
    },
    {
      "epoch": 5.0,
      "step": 17865,
      "total_flos": 6.55200095561257e+16,
      "train_loss": 0.38020934802358836,
      "train_runtime": 9310.5317,
      "train_samples_per_second": 7.674,
      "train_steps_per_second": 1.919
    }
  ],
  "logging_steps": 500,
  "max_steps": 17865,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 6.55200095561257e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}