checkpoint-752/trainer_state.json

{
  "best_metric": 0.6807650327682495,
  "best_model_checkpoint": "./LongT5-XLarge-NSPCC/checkpoint-566",
  "epoch": 3.984105960264901,
  "eval_steps": 500,
  "global_step": 752,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9960264900662251,
      "grad_norm": 0.6866635680198669,
      "learning_rate": 0.00026365056951661766,
      "loss": 3.6911,
      "step": 188
    },
    {
      "epoch": 0.9960264900662251,
      "eval_gen_len": 360.7021,
      "eval_loss": 0.7291953563690186,
      "eval_rouge1": 0.4665,
      "eval_rouge2": 0.1826,
      "eval_rougeL": 0.2611,
      "eval_rougeLsum": 0.2611,
      "eval_runtime": 6342.3266,
      "eval_samples_per_second": 0.015,
      "eval_steps_per_second": 0.015,
      "step": 188
    },
    {
      "epoch": 1.9973509933774833,
      "grad_norm": 0.37886831164360046,
      "learning_rate": 0.00015678507557860595,
      "loss": 0.8701,
      "step": 377
    },
    {
      "epoch": 1.9973509933774833,
      "eval_gen_len": 365.3298,
      "eval_loss": 0.6966970562934875,
      "eval_rouge1": 0.4886,
      "eval_rouge2": 0.2073,
      "eval_rougeL": 0.2805,
      "eval_rougeLsum": 0.2799,
      "eval_runtime": 6456.3838,
      "eval_samples_per_second": 0.015,
      "eval_steps_per_second": 0.015,
      "step": 377
    },
    {
      "epoch": 2.9986754966887417,
      "grad_norm": 0.5063174366950989,
      "learning_rate": 4.566183323963369e-05,
      "loss": 0.7849,
      "step": 566
    },
    {
      "epoch": 2.9986754966887417,
      "eval_gen_len": 332.3191,
      "eval_loss": 0.6807650327682495,
      "eval_rouge1": 0.5116,
      "eval_rouge2": 0.2302,
      "eval_rougeL": 0.2995,
      "eval_rougeLsum": 0.2997,
      "eval_runtime": 5946.7561,
      "eval_samples_per_second": 0.016,
      "eval_steps_per_second": 0.016,
      "step": 566
    },
    {
      "epoch": 3.984105960264901,
      "grad_norm": 0.3437921106815338,
      "learning_rate": 0.0,
      "loss": 0.7769,
      "step": 752
    },
    {
      "epoch": 3.984105960264901,
      "eval_gen_len": 337.6809,
      "eval_loss": 0.6843053102493286,
      "eval_rouge1": 0.5138,
      "eval_rouge2": 0.2297,
      "eval_rougeL": 0.2999,
      "eval_rougeLsum": 0.2995,
      "eval_runtime": 6050.9414,
      "eval_samples_per_second": 0.016,
      "eval_steps_per_second": 0.016,
      "step": 752
    }
  ],
  "logging_steps": 500,
  "max_steps": 752,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 8.260168119704617e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}