trainer_state.json

{
  "best_metric": 71.7161,
  "best_model_checkpoint": "bin/indosum-base-0/checkpoint-17830",
  "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": 2.044853687286377,
      "learning_rate": 0.0008,
      "loss": 1.5249,
      "step": 3566
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 103.82933333333334,
      "eval_loss": 0.8857953548431396,
      "eval_rouge1": 65.9377,
      "eval_rouge2": 58.0172,
      "eval_rougeL": 63.1,
      "eval_rougeLsum": 65.0124,
      "eval_runtime": 640.0795,
      "eval_samples_per_second": 1.172,
      "eval_steps_per_second": 0.147,
      "step": 3566
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6789307594299316,
      "learning_rate": 0.0006,
      "loss": 0.8202,
      "step": 7132
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 101.31733333333334,
      "eval_loss": 0.7492297291755676,
      "eval_rouge1": 69.7595,
      "eval_rouge2": 62.2139,
      "eval_rougeL": 66.6576,
      "eval_rougeLsum": 68.9014,
      "eval_runtime": 623.9837,
      "eval_samples_per_second": 1.202,
      "eval_steps_per_second": 0.151,
      "step": 7132
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6821009516716003,
      "learning_rate": 0.0004,
      "loss": 0.6232,
      "step": 10698
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 97.70933333333333,
      "eval_loss": 0.6953277587890625,
      "eval_rouge1": 70.7507,
      "eval_rouge2": 63.4429,
      "eval_rougeL": 67.7582,
      "eval_rougeLsum": 69.8526,
      "eval_runtime": 605.7402,
      "eval_samples_per_second": 1.238,
      "eval_steps_per_second": 0.155,
      "step": 10698
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5773214101791382,
      "learning_rate": 0.0002,
      "loss": 0.4728,
      "step": 14264
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 97.78933333333333,
      "eval_loss": 0.6717060804367065,
      "eval_rouge1": 70.8339,
      "eval_rouge2": 63.5583,
      "eval_rougeL": 67.7546,
      "eval_rougeLsum": 69.8695,
      "eval_runtime": 607.3846,
      "eval_samples_per_second": 1.235,
      "eval_steps_per_second": 0.155,
      "step": 14264
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1052696704864502,
      "learning_rate": 0.0,
      "loss": 0.3238,
      "step": 17830
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 99.53733333333334,
      "eval_loss": 0.7148926258087158,
      "eval_rouge1": 71.7161,
      "eval_rouge2": 64.386,
      "eval_rougeL": 68.5898,
      "eval_rougeLsum": 70.7721,
      "eval_runtime": 612.7477,
      "eval_samples_per_second": 1.224,
      "eval_steps_per_second": 0.153,
      "step": 17830
    },
    {
      "epoch": 5.0,
      "step": 17830,
      "total_flos": 4.883004680306688e+16,
      "train_loss": 0.7530090023945772,
      "train_runtime": 12824.4301,
      "train_samples_per_second": 5.56,
      "train_steps_per_second": 1.39
    }
  ],
  "logging_steps": 500,
  "max_steps": 17830,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 4.883004680306688e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}