run-15/checkpoint-958/trainer_state.json

{
  "best_metric": 0.8574040219378428,
  "best_model_checkpoint": "berttiny-hate_speech18-bothpretrained\\run-15\\checkpoint-958",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 958,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 2.1822041543780283e-05,
      "loss": 0.4852,
      "step": 958
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8574040219378428,
      "eval_loss": 0.47587475180625916,
      "eval_runtime": 10.7372,
      "eval_samples_per_second": 203.778,
      "eval_steps_per_second": 1.676,
      "step": 958
    }
  ],
  "logging_steps": 500,
  "max_steps": 2874,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 983811512832.0,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 3.2733062315670426e-05,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 8,
    "seed": 22
  }
}