trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_bleu": 56.4458,
      "eval_gen_len": 36.4625,
      "eval_loss": 0.7646628618240356,
      "eval_runtime": 5.0083,
      "eval_samples_per_second": 15.973,
      "eval_steps_per_second": 1.997,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_bleu": 55.8074,
      "eval_gen_len": 36.4625,
      "eval_loss": 0.7441360354423523,
      "eval_runtime": 4.9818,
      "eval_samples_per_second": 16.059,
      "eval_steps_per_second": 2.007,
      "step": 80
    },
    {
      "epoch": 3.0,
      "eval_bleu": 57.269,
      "eval_gen_len": 36.1375,
      "eval_loss": 0.7390245199203491,
      "eval_runtime": 4.9371,
      "eval_samples_per_second": 16.204,
      "eval_steps_per_second": 2.025,
      "step": 120
    },
    {
      "epoch": 3.0,
      "step": 120,
      "total_flos": 14405745770496.0,
      "train_loss": 0.6587849299112956,
      "train_runtime": 25.0068,
      "train_samples_per_second": 38.27,
      "train_steps_per_second": 4.799
    }
  ],
  "max_steps": 120,
  "num_train_epochs": 3,
  "total_flos": 14405745770496.0,
  "trial_name": null,
  "trial_params": null
}