checkpoint-1180/trainer_state.json

{
  "best_metric": 0.298869788646698,
  "best_model_checkpoint": "facial_emotions_image_detection/checkpoint-1180",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.42,
      "grad_norm": 16.416099548339844,
      "learning_rate": 9.23076923076923e-08,
      "loss": 0.8215,
      "step": 500
    },
    {
      "epoch": 0.85,
      "grad_norm": 18.07982635498047,
      "learning_rate": 8.376068376068375e-08,
      "loss": 0.8413,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.909217322208979,
      "eval_loss": 0.298869788646698,
      "eval_runtime": 809.7668,
      "eval_samples_per_second": 31.083,
      "eval_steps_per_second": 3.886,
      "step": 1180
    }
  ],
  "logging_steps": 500,
  "max_steps": 5900,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 2.9256861880439747e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}