trainer_state.json

{
  "best_metric": 0.9050666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0044/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": 8.730227815298347e-05,
      "loss": 2.1093,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7749333333333334,
      "eval_loss": 0.7932878136634827,
      "eval_runtime": 10.4502,
      "eval_samples_per_second": 358.843,
      "eval_steps_per_second": 5.646,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7.953256608588366e-05,
      "loss": 0.5474,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8426666666666667,
      "eval_loss": 0.5026018023490906,
      "eval_runtime": 11.6843,
      "eval_samples_per_second": 320.944,
      "eval_steps_per_second": 5.05,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.0253734588623047,
      "learning_rate": 6.758165316736547e-05,
      "loss": 0.2545,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.41560977697372437,
      "eval_runtime": 10.5664,
      "eval_samples_per_second": 354.897,
      "eval_steps_per_second": 5.584,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.11438139528036118,
      "learning_rate": 5.2907059607156464e-05,
      "loss": 0.1211,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.41038277745246887,
      "eval_runtime": 10.6618,
      "eval_samples_per_second": 351.722,
      "eval_steps_per_second": 5.534,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.30288928747177124,
      "learning_rate": 3.727875796255304e-05,
      "loss": 0.0524,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.3860125243663788,
      "eval_runtime": 10.9957,
      "eval_samples_per_second": 341.044,
      "eval_steps_per_second": 5.366,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05355275049805641,
      "learning_rate": 2.258175206106183e-05,
      "loss": 0.0193,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8984,
      "eval_loss": 0.379838764667511,
      "eval_runtime": 11.0025,
      "eval_samples_per_second": 340.832,
      "eval_steps_per_second": 5.362,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0965563952922821,
      "learning_rate": 1.0588717719108745e-05,
      "loss": 0.0056,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9050666666666667,
      "eval_loss": 0.3612273633480072,
      "eval_runtime": 10.3997,
      "eval_samples_per_second": 360.588,
      "eval_steps_per_second": 5.673,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}