trainer_state.json

{
  "best_metric": 0.8816,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0993/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 6.676676676676677e-05,
      "loss": 1.5731,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8072,
      "eval_loss": 0.6853334903717041,
      "eval_runtime": 7.6761,
      "eval_samples_per_second": 488.527,
      "eval_steps_per_second": 7.686,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.4755859375,
      "learning_rate": 3.3433433433433435e-05,
      "loss": 0.4209,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.4759966731071472,
      "eval_runtime": 6.6715,
      "eval_samples_per_second": 562.096,
      "eval_steps_per_second": 8.844,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.357871651649475,
      "learning_rate": 1.001001001001001e-07,
      "loss": 0.1454,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.39465025067329407,
      "eval_runtime": 6.8627,
      "eval_samples_per_second": 546.433,
      "eval_steps_per_second": 8.597,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.94223961867776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}