trainer_state.json

{
  "best_metric": 0.904,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0377/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": Infinity,
      "learning_rate": 6.003003003003003e-05,
      "loss": 1.6337,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7973333333333333,
      "eval_loss": 0.7328707575798035,
      "eval_runtime": 6.9038,
      "eval_samples_per_second": 543.181,
      "eval_steps_per_second": 8.546,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.359355926513672,
      "learning_rate": 5.0030030030030024e-05,
      "loss": 0.4782,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8576,
      "eval_loss": 0.4941653907299042,
      "eval_runtime": 6.9459,
      "eval_samples_per_second": 539.885,
      "eval_steps_per_second": 8.494,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1959383487701416,
      "learning_rate": 4.003003003003003e-05,
      "loss": 0.2633,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8794666666666666,
      "eval_loss": 0.4000539779663086,
      "eval_runtime": 7.3096,
      "eval_samples_per_second": 513.024,
      "eval_steps_per_second": 8.072,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.45420756936073303,
      "learning_rate": 3.003003003003003e-05,
      "loss": 0.1462,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.3847833275794983,
      "eval_runtime": 6.9369,
      "eval_samples_per_second": 540.586,
      "eval_steps_per_second": 8.505,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.0186386108398438,
      "learning_rate": 2.003003003003003e-05,
      "loss": 0.0782,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8898666666666667,
      "eval_loss": 0.3895341157913208,
      "eval_runtime": 7.0459,
      "eval_samples_per_second": 532.222,
      "eval_steps_per_second": 8.374,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.26239752769470215,
      "learning_rate": 1.003003003003003e-05,
      "loss": 0.0373,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9,
      "eval_loss": 0.36605390906333923,
      "eval_runtime": 6.8736,
      "eval_samples_per_second": 545.568,
      "eval_steps_per_second": 8.584,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.07380411773920059,
      "learning_rate": 3.0030030030030026e-08,
      "loss": 0.0166,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.3544391393661499,
      "eval_runtime": 7.3816,
      "eval_samples_per_second": 508.02,
      "eval_steps_per_second": 7.993,
      "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
}