trainer_state.json

{
  "best_metric": 0.7375,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224/checkpoint-24",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_loss": 0.5740450024604797,
      "eval_runtime": 1.7884,
      "eval_samples_per_second": 134.201,
      "eval_steps_per_second": 4.473,
      "step": 8
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6033,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_loss": 0.5640280246734619,
      "eval_runtime": 1.8333,
      "eval_samples_per_second": 130.909,
      "eval_steps_per_second": 4.364,
      "step": 16
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.523809523809523e-06,
      "loss": 0.5751,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7375,
      "eval_loss": 0.5567867159843445,
      "eval_runtime": 1.9734,
      "eval_samples_per_second": 121.615,
      "eval_steps_per_second": 4.054,
      "step": 24
    },
    {
      "epoch": 3.0,
      "step": 24,
      "total_flos": 7.4567966957568e+16,
      "train_loss": 0.5813349982102712,
      "train_runtime": 56.867,
      "train_samples_per_second": 52.755,
      "train_steps_per_second": 0.422
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.73,
      "eval_loss": 0.5407183766365051,
      "eval_runtime": 70.3286,
      "eval_samples_per_second": 4.266,
      "eval_steps_per_second": 0.142,
      "step": 24
    }
  ],
  "logging_steps": 10,
  "max_steps": 24,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 7.4567966957568e+16,
  "trial_name": null,
  "trial_params": null
}