trainer_state.json

{
  "best_metric": 0.8402666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0635/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": 0.0001,
      "loss": 1.4595,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7986666666666666,
      "eval_loss": 0.6715376973152161,
      "eval_runtime": 6.348,
      "eval_samples_per_second": 590.737,
      "eval_steps_per_second": 9.294,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.680118560791016,
      "learning_rate": 0.0001,
      "loss": 0.4783,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8357333333333333,
      "eval_loss": 0.5418625473976135,
      "eval_runtime": 6.4222,
      "eval_samples_per_second": 583.912,
      "eval_steps_per_second": 9.187,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.1511128693819046,
      "learning_rate": 0.0001,
      "loss": 0.3021,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8402666666666667,
      "eval_loss": 0.5223037600517273,
      "eval_runtime": 6.3376,
      "eval_samples_per_second": 591.707,
      "eval_steps_per_second": 9.31,
      "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
}