trainer_state.json

{
  "best_metric": 51.1569,
  "best_model_checkpoint": "bin/indosum-pt-pl5-0/checkpoint-4460",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 4460,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0623373985290527,
      "learning_rate": 0.0008,
      "loss": 3.3529,
      "step": 892
    },
    {
      "epoch": 1.0,
      "eval_gen_len": 95.03466666666667,
      "eval_loss": 2.403062343597412,
      "eval_rouge1": 38.6954,
      "eval_rouge2": 16.1544,
      "eval_rougeL": 31.8171,
      "eval_rougeLsum": 36.2119,
      "eval_runtime": 232.3708,
      "eval_samples_per_second": 3.228,
      "eval_steps_per_second": 0.103,
      "step": 892
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9969884157180786,
      "learning_rate": 0.0006,
      "loss": 2.9862,
      "step": 1784
    },
    {
      "epoch": 2.0,
      "eval_gen_len": 92.37066666666666,
      "eval_loss": 2.1683199405670166,
      "eval_rouge1": 43.9573,
      "eval_rouge2": 21.3807,
      "eval_rougeL": 37.3906,
      "eval_rougeLsum": 41.5485,
      "eval_runtime": 224.6137,
      "eval_samples_per_second": 3.339,
      "eval_steps_per_second": 0.107,
      "step": 1784
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0779091119766235,
      "learning_rate": 0.0004,
      "loss": 2.7929,
      "step": 2676
    },
    {
      "epoch": 3.0,
      "eval_gen_len": 100.268,
      "eval_loss": 1.9973465204238892,
      "eval_rouge1": 44.5846,
      "eval_rouge2": 23.2132,
      "eval_rougeL": 38.5252,
      "eval_rougeLsum": 42.4754,
      "eval_runtime": 239.0067,
      "eval_samples_per_second": 3.138,
      "eval_steps_per_second": 0.1,
      "step": 2676
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0467666387557983,
      "learning_rate": 0.0002,
      "loss": 2.6146,
      "step": 3568
    },
    {
      "epoch": 4.0,
      "eval_gen_len": 98.78133333333334,
      "eval_loss": 1.8517733812332153,
      "eval_rouge1": 48.6507,
      "eval_rouge2": 27.6554,
      "eval_rougeL": 42.9432,
      "eval_rougeLsum": 46.628,
      "eval_runtime": 225.9342,
      "eval_samples_per_second": 3.32,
      "eval_steps_per_second": 0.106,
      "step": 3568
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9933086037635803,
      "learning_rate": 0.0,
      "loss": 2.4702,
      "step": 4460
    },
    {
      "epoch": 5.0,
      "eval_gen_len": 96.54666666666667,
      "eval_loss": 1.7467247247695923,
      "eval_rouge1": 51.1569,
      "eval_rouge2": 30.4226,
      "eval_rougeL": 45.5936,
      "eval_rougeLsum": 49.2619,
      "eval_runtime": 228.2543,
      "eval_samples_per_second": 3.286,
      "eval_steps_per_second": 0.105,
      "step": 4460
    },
    {
      "epoch": 5.0,
      "step": 4460,
      "total_flos": 5.530296634048512e+16,
      "train_loss": 2.843380504659473,
      "train_runtime": 6617.6699,
      "train_samples_per_second": 10.776,
      "train_steps_per_second": 0.674
    }
  ],
  "logging_steps": 500,
  "max_steps": 4460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 5.530296634048512e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}