trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 17830,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.377143144607544,
      "learning_rate": 4e-05,
      "loss": 0.6316,
      "step": 3566
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.48070013523101807,
      "eval_rouge1": 0.4602,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4604,
      "eval_rougeLsum": 0.4565,
      "eval_runtime": 339.8998,
      "eval_samples_per_second": 2.204,
      "eval_steps_per_second": 0.277,
      "step": 3566
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.1111056804656982,
      "learning_rate": 3e-05,
      "loss": 0.4336,
      "step": 7132
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.47169604897499084,
      "eval_rouge1": 0.4661,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.466,
      "eval_rougeLsum": 0.4622,
      "eval_runtime": 338.0027,
      "eval_samples_per_second": 2.216,
      "eval_steps_per_second": 0.278,
      "step": 7132
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.34107506275177,
      "learning_rate": 2e-05,
      "loss": 0.3363,
      "step": 10698
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.4722852408885956,
      "eval_rouge1": 0.4799,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.479,
      "eval_rougeLsum": 0.4762,
      "eval_runtime": 338.8999,
      "eval_samples_per_second": 2.21,
      "eval_steps_per_second": 0.277,
      "step": 10698
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.9266836643218994,
      "learning_rate": 1e-05,
      "loss": 0.2656,
      "step": 14264
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.48245567083358765,
      "eval_rouge1": 0.4713,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4703,
      "eval_rougeLsum": 0.4666,
      "eval_runtime": 340.0086,
      "eval_samples_per_second": 2.203,
      "eval_steps_per_second": 0.276,
      "step": 14264
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7729974389076233,
      "learning_rate": 0.0,
      "loss": 0.219,
      "step": 17830
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 1.0,
      "eval_loss": 0.5102769732475281,
      "eval_rouge1": 0.4427,
      "eval_rouge2": 0.0,
      "eval_rougeL": 0.4423,
      "eval_rougeLsum": 0.4403,
      "eval_runtime": 336.8192,
      "eval_samples_per_second": 2.224,
      "eval_steps_per_second": 0.279,
      "step": 17830
    },
    {
      "epoch": 5.0,
      "step": 17830,
      "total_flos": 6.531382403512013e+16,
      "train_loss": 0.37723224017303175,
      "train_runtime": 9322.5764,
      "train_samples_per_second": 7.65,
      "train_steps_per_second": 1.913
    }
  ],
  "logging_steps": 500,
  "max_steps": 17830,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 6.531382403512013e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}