trainer_state.json

{
  "best_metric": 0.8712,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0538/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 13.552969932556152,
      "learning_rate": 7e-05,
      "loss": 1.6283,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7962666666666667,
      "eval_loss": 0.7473523616790771,
      "eval_runtime": 7.4644,
      "eval_samples_per_second": 502.386,
      "eval_steps_per_second": 7.904,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.5205,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8424,
      "eval_loss": 0.5379555225372314,
      "eval_runtime": 7.4093,
      "eval_samples_per_second": 506.123,
      "eval_steps_per_second": 7.963,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 27.032501220703125,
      "learning_rate": 7e-05,
      "loss": 0.3144,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.5022523403167725,
      "eval_runtime": 7.3689,
      "eval_samples_per_second": 508.897,
      "eval_steps_per_second": 8.007,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.495960533618927,
      "learning_rate": 7e-05,
      "loss": 0.2066,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.49607589840888977,
      "eval_runtime": 7.5024,
      "eval_samples_per_second": 499.841,
      "eval_steps_per_second": 7.864,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5330110192298889,
      "learning_rate": 7e-05,
      "loss": 0.1429,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.5520705580711365,
      "eval_runtime": 7.5491,
      "eval_samples_per_second": 496.749,
      "eval_steps_per_second": 7.816,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04824044927954674,
      "learning_rate": 7e-05,
      "loss": 0.1015,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.5107033252716064,
      "eval_runtime": 7.5898,
      "eval_samples_per_second": 494.083,
      "eval_steps_per_second": 7.774,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1427602469921112,
      "learning_rate": 7e-05,
      "loss": 0.0866,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.49833711981773376,
      "eval_runtime": 7.8982,
      "eval_samples_per_second": 474.792,
      "eval_steps_per_second": 7.47,
      "step": 2331
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}