trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9967417865870214,
  "eval_steps": 500,
  "global_step": 3680,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.54,
      "learning_rate": 1.731521739130435e-05,
      "loss": 1.9542,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.9149,
      "eval_gen_len": 18.56723716381418,
      "eval_loss": 1.534995198249817,
      "eval_precision": 0.9229,
      "eval_recall": 0.9074,
      "eval_rouge1": 0.4928,
      "eval_rouge2": 0.2436,
      "eval_rougeL": 0.4085,
      "eval_rougeLsum": 0.4086,
      "eval_runtime": 246.8012,
      "eval_samples_per_second": 3.314,
      "eval_steps_per_second": 0.831,
      "step": 920
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4603260869565218e-05,
      "loss": 1.7174,
      "step": 1000
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.1885869565217392e-05,
      "loss": 1.6331,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.9166,
      "eval_gen_len": 18.815403422982886,
      "eval_loss": 1.491409182548523,
      "eval_precision": 0.9246,
      "eval_recall": 0.9092,
      "eval_rouge1": 0.5037,
      "eval_rouge2": 0.257,
      "eval_rougeL": 0.4202,
      "eval_rougeLsum": 0.4206,
      "eval_runtime": 247.8681,
      "eval_samples_per_second": 3.3,
      "eval_steps_per_second": 0.827,
      "step": 1841
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.179347826086958e-06,
      "loss": 1.5996,
      "step": 2000
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.472826086956522e-06,
      "loss": 1.5694,
      "step": 2500
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.917,
      "eval_gen_len": 19.448655256723715,
      "eval_loss": 1.4760992527008057,
      "eval_precision": 0.9241,
      "eval_recall": 0.9103,
      "eval_rouge1": 0.5071,
      "eval_rouge2": 0.259,
      "eval_rougeL": 0.4212,
      "eval_rougeLsum": 0.4214,
      "eval_runtime": 249.775,
      "eval_samples_per_second": 3.275,
      "eval_steps_per_second": 0.821,
      "step": 2762
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.7554347826086963e-06,
      "loss": 1.5609,
      "step": 3000
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.048913043478261e-06,
      "loss": 1.5374,
      "step": 3500
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.917,
      "eval_gen_len": 19.1479217603912,
      "eval_loss": 1.4708906412124634,
      "eval_precision": 0.9247,
      "eval_recall": 0.9099,
      "eval_rouge1": 0.5072,
      "eval_rouge2": 0.2631,
      "eval_rougeL": 0.4243,
      "eval_rougeLsum": 0.4244,
      "eval_runtime": 248.4801,
      "eval_samples_per_second": 3.292,
      "eval_steps_per_second": 0.825,
      "step": 3680
    },
    {
      "epoch": 4.0,
      "step": 3680,
      "total_flos": 3.997817815616717e+16,
      "train_loss": 1.6477281238721764,
      "train_runtime": 4221.2735,
      "train_samples_per_second": 13.96,
      "train_steps_per_second": 0.872
    }
  ],
  "logging_steps": 500,
  "max_steps": 3680,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 3.997817815616717e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}