trainer_state.json

{
  "best_metric": 0.8714666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0300/checkpoints/checkpoint-666",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 666,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.9731466770172119,
      "learning_rate": 9e-05,
      "loss": 0.7276,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.540214478969574,
      "eval_runtime": 11.1016,
      "eval_samples_per_second": 337.789,
      "eval_steps_per_second": 5.315,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.2648,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.4666679799556732,
      "eval_runtime": 11.3409,
      "eval_samples_per_second": 330.66,
      "eval_steps_per_second": 5.202,
      "step": 666
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.29482641245184e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}