trainer_state.json

{
  "best_metric": 0.6875,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-jvadlamudi2/checkpoint-21",
  "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": 0.5892857142857143,
      "eval_loss": 0.6659354567527771,
      "eval_runtime": 0.8649,
      "eval_samples_per_second": 129.489,
      "eval_steps_per_second": 4.625,
      "step": 7
    },
    {
      "epoch": 1.43,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.6765,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6517857142857143,
      "eval_loss": 0.6479455232620239,
      "eval_runtime": 0.8944,
      "eval_samples_per_second": 125.219,
      "eval_steps_per_second": 4.472,
      "step": 14
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.6502,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6875,
      "eval_loss": 0.6389141082763672,
      "eval_runtime": 0.8926,
      "eval_samples_per_second": 125.472,
      "eval_steps_per_second": 4.481,
      "step": 21
    },
    {
      "epoch": 3.0,
      "step": 21,
      "total_flos": 6.651462652615066e+16,
      "train_loss": 0.6625232639766875,
      "train_runtime": 75.1281,
      "train_samples_per_second": 35.619,
      "train_steps_per_second": 0.28
    }
  ],
  "max_steps": 21,
  "num_train_epochs": 3,
  "total_flos": 6.651462652615066e+16,
  "trial_name": null,
  "trial_params": null
}