trainer_state.json

{
  "best_metric": 0.7333333333333333,
  "best_model_checkpoint": "swinv2-tiny-patch4-window8-256/checkpoint-8",
  "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": 1.4140104055404663,
      "eval_runtime": 2.2803,
      "eval_samples_per_second": 105.251,
      "eval_steps_per_second": 3.508,
      "step": 8
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 5.2877,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_loss": 0.8763239979743958,
      "eval_runtime": 2.3693,
      "eval_samples_per_second": 101.296,
      "eval_steps_per_second": 3.377,
      "step": 16
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.523809523809523e-06,
      "loss": 0.9272,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7333333333333333,
      "eval_loss": 0.6204068064689636,
      "eval_runtime": 2.4851,
      "eval_samples_per_second": 96.577,
      "eval_steps_per_second": 3.219,
      "step": 24
    },
    {
      "epoch": 3.0,
      "step": 24,
      "total_flos": 1.00318990565376e+17,
      "train_loss": 2.710611641407013,
      "train_runtime": 107.4175,
      "train_samples_per_second": 27.928,
      "train_steps_per_second": 0.223
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.73,
      "eval_loss": 1.472348928451538,
      "eval_runtime": 3.3432,
      "eval_samples_per_second": 89.735,
      "eval_steps_per_second": 2.991,
      "step": 24
    }
  ],
  "logging_steps": 10,
  "max_steps": 24,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.00318990565376e+17,
  "trial_name": null,
  "trial_params": null
}