trainer_state.json

{
  "best_metric": 0.7946428571428571,
  "best_model_checkpoint": "convnext-tiny-224-jvadlamudi2/checkpoint-14",
  "epoch": 3.0,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7678571428571429,
      "eval_loss": 0.6085006594657898,
      "eval_runtime": 0.7752,
      "eval_samples_per_second": 144.48,
      "eval_steps_per_second": 5.16,
      "step": 7
    },
    {
      "epoch": 1.43,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.6388,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7946428571428571,
      "eval_loss": 0.5984492897987366,
      "eval_runtime": 0.82,
      "eval_samples_per_second": 136.58,
      "eval_steps_per_second": 4.878,
      "step": 14
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.6218,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7857142857142857,
      "eval_loss": 0.5960062742233276,
      "eval_runtime": 0.8803,
      "eval_samples_per_second": 127.229,
      "eval_steps_per_second": 4.544,
      "step": 21
    },
    {
      "epoch": 3.0,
      "step": 21,
      "total_flos": 6.716617754447462e+16,
      "train_loss": 0.630237170628139,
      "train_runtime": 87.2731,
      "train_samples_per_second": 30.628,
      "train_steps_per_second": 0.241
    }
  ],
  "max_steps": 21,
  "num_train_epochs": 3,
  "total_flos": 6.716617754447462e+16,
  "trial_name": null,
  "trial_params": null
}