trainer_state.json

{
  "best_metric": 0.7948717948717948,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-pueba1/checkpoint-15",
  "epoch": 2.7272727272727275,
  "eval_steps": 500,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9090909090909091,
      "eval_accuracy": 0.6410256410256411,
      "eval_loss": 0.6238493919372559,
      "eval_runtime": 79.8353,
      "eval_samples_per_second": 0.977,
      "eval_steps_per_second": 0.038,
      "step": 5
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 4.14777135848999,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.653,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7692307692307693,
      "eval_loss": 0.5155723094940186,
      "eval_runtime": 1.7882,
      "eval_samples_per_second": 43.618,
      "eval_steps_per_second": 1.678,
      "step": 11
    },
    {
      "epoch": 2.7272727272727275,
      "eval_accuracy": 0.7948717948717948,
      "eval_loss": 0.4784446656703949,
      "eval_runtime": 0.5085,
      "eval_samples_per_second": 153.379,
      "eval_steps_per_second": 5.899,
      "step": 15
    },
    {
      "epoch": 2.7272727272727275,
      "step": 15,
      "total_flos": 4.757436291892838e+16,
      "train_loss": 0.6080072561899821,
      "train_runtime": 822.3142,
      "train_samples_per_second": 2.557,
      "train_steps_per_second": 0.018
    }
  ],
  "logging_steps": 10,
  "max_steps": 15,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 4.757436291892838e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}