trainer_state.json

{
  "best_metric": 0.7678571428571429,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-jvadlamudi2/checkpoint-16",
  "epoch": 3.0,
  "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": 0.75,
      "eval_loss": 0.5683054327964783,
      "eval_runtime": 0.7946,
      "eval_samples_per_second": 140.946,
      "eval_steps_per_second": 5.034,
      "step": 8
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2735,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7678571428571429,
      "eval_loss": 0.5364691615104675,
      "eval_runtime": 0.8091,
      "eval_samples_per_second": 138.419,
      "eval_steps_per_second": 4.944,
      "step": 16
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.523809523809523e-06,
      "loss": 0.3789,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7678571428571429,
      "eval_loss": 0.5417876839637756,
      "eval_runtime": 0.9425,
      "eval_samples_per_second": 118.834,
      "eval_steps_per_second": 4.244,
      "step": 24
    },
    {
      "epoch": 3.0,
      "step": 24,
      "total_flos": 7.471710289148314e+16,
      "train_loss": 0.3398028165102005,
      "train_runtime": 91.8361,
      "train_samples_per_second": 32.732,
      "train_steps_per_second": 0.261
    }
  ],
  "max_steps": 24,
  "num_train_epochs": 3,
  "total_flos": 7.471710289148314e+16,
  "trial_name": null,
  "trial_params": null
}