trainer_state.json

{
  "best_metric": 0.5238095238095238,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-12",
  "epoch": 2.6666666666666665,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8888888888888888,
      "eval_accuracy": 0.4444444444444444,
      "eval_loss": 1.740828037261963,
      "eval_runtime": 17.227,
      "eval_samples_per_second": 3.657,
      "eval_steps_per_second": 0.116,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.49206349206349204,
      "eval_loss": 1.4351359605789185,
      "eval_runtime": 16.2702,
      "eval_samples_per_second": 3.872,
      "eval_steps_per_second": 0.123,
      "step": 9
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.383830547332764,
      "learning_rate": 1e-05,
      "loss": 1.7494,
      "step": 10
    },
    {
      "epoch": 2.6666666666666665,
      "eval_accuracy": 0.5238095238095238,
      "eval_loss": 1.3751540184020996,
      "eval_runtime": 16.3345,
      "eval_samples_per_second": 3.857,
      "eval_steps_per_second": 0.122,
      "step": 12
    },
    {
      "epoch": 2.6666666666666665,
      "step": 12,
      "total_flos": 3.768799680376013e+16,
      "train_loss": 1.699290931224823,
      "train_runtime": 1294.0597,
      "train_samples_per_second": 1.312,
      "train_steps_per_second": 0.009
    }
  ],
  "logging_steps": 10,
  "max_steps": 12,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3.768799680376013e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}