trainer_state.json

{
  "best_metric": 0.9963201471941122,
  "best_model_checkpoint": "data/train-test/bert-base-output//model/checkpoint-1284",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1284,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.78,
      "grad_norm": 0.0749111995100975,
      "learning_rate": 1.221183800623053e-05,
      "loss": 0.0374,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": null,
      "eval_f1": 0.9917355371900826,
      "eval_loss": 0.008172398433089256,
      "eval_precision": 0.9899175068744271,
      "eval_recall": 0.9935602575896965,
      "eval_runtime": 1.4927,
      "eval_samples_per_second": 649.834,
      "eval_steps_per_second": 20.768,
      "step": 642
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.17515258491039276,
      "learning_rate": 4.42367601246106e-06,
      "loss": 0.0047,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": null,
      "eval_f1": 0.9954044117647058,
      "eval_loss": 0.0051086777821183205,
      "eval_precision": 0.9944903581267218,
      "eval_recall": 0.9963201471941122,
      "eval_runtime": 1.4819,
      "eval_samples_per_second": 654.581,
      "eval_steps_per_second": 20.92,
      "step": 1284
    }
  ],
  "logging_steps": 500,
  "max_steps": 1284,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1114678566004932.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}