trainer_state.json

{
  "best_metric": 0.9963201471941122,
  "best_model_checkpoint": "data/train-test/roberta-base-output//model/checkpoint-1676",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1676,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6,
      "grad_norm": 1.825121283531189,
      "learning_rate": 1.4033412887828164e-05,
      "loss": 0.0322,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": null,
      "eval_f1": 0.9926538108356291,
      "eval_loss": 0.00715277437120676,
      "eval_precision": 0.9908340971585701,
      "eval_recall": 0.9944802207911684,
      "eval_runtime": 1.6458,
      "eval_samples_per_second": 589.386,
      "eval_steps_per_second": 18.836,
      "step": 838
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.7001124620437622,
      "learning_rate": 8.066825775656326e-06,
      "loss": 0.0046,
      "step": 1000
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.1312304437160492,
      "learning_rate": 2.100238663484487e-06,
      "loss": 0.0027,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": null,
      "eval_f1": 0.9963201471941122,
      "eval_loss": 0.004037069622427225,
      "eval_precision": 0.9963201471941122,
      "eval_recall": 0.9963201471941122,
      "eval_runtime": 1.6277,
      "eval_samples_per_second": 595.928,
      "eval_steps_per_second": 19.045,
      "step": 1676
    }
  ],
  "logging_steps": 500,
  "max_steps": 1676,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 2013781275950328.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}