trainer_state.json

{
  "best_metric": 0.9816007359705612,
  "best_model_checkpoint": "data/train-test/xlnet-base-cased//model/checkpoint-874",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 874,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": null,
      "eval_f1": 0.9355716878402904,
      "eval_loss": 0.025213107466697693,
      "eval_precision": 0.9230080572963295,
      "eval_recall": 0.9484820607175714,
      "eval_runtime": 1.8189,
      "eval_samples_per_second": 533.284,
      "eval_steps_per_second": 17.043,
      "step": 437
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.2998504638671875,
      "learning_rate": 8.558352402745997e-06,
      "loss": 0.0831,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": null,
      "eval_f1": 0.9775538250114522,
      "eval_loss": 0.015987424179911613,
      "eval_precision": 0.9735401459854015,
      "eval_recall": 0.9816007359705612,
      "eval_runtime": 1.8215,
      "eval_samples_per_second": 532.541,
      "eval_steps_per_second": 17.019,
      "step": 874
    }
  ],
  "logging_steps": 500,
  "max_steps": 874,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 841186234475568.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}