trainer_state.json

{
  "best_metric": 0.6106666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0839/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": 0.0004810825448943271,
      "loss": 3.0643,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.29706666666666665,
      "eval_loss": 2.5800931453704834,
      "eval_runtime": 6.6287,
      "eval_samples_per_second": 565.726,
      "eval_steps_per_second": 8.901,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0004271931404387096,
      "loss": 2.3306,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.40613333333333335,
      "eval_loss": 2.166173219680786,
      "eval_runtime": 7.1625,
      "eval_samples_per_second": 523.562,
      "eval_steps_per_second": 8.237,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003464873897488322,
      "loss": 1.8785,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.46853333333333336,
      "eval_loss": 1.857843041419983,
      "eval_runtime": 7.4846,
      "eval_samples_per_second": 501.029,
      "eval_steps_per_second": 7.883,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.149824142456055,
      "learning_rate": 0.00025088445554620134,
      "loss": 1.4348,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5397333333333333,
      "eval_loss": 1.6192086935043335,
      "eval_runtime": 7.1724,
      "eval_samples_per_second": 522.837,
      "eval_steps_per_second": 8.226,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 21.50592803955078,
      "learning_rate": 0.00015514687100427082,
      "loss": 0.9526,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5736,
      "eval_loss": 1.5202983617782593,
      "eval_runtime": 6.8488,
      "eval_samples_per_second": 547.54,
      "eval_steps_per_second": 8.615,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.375426292419434,
      "learning_rate": 7.384981550618446e-05,
      "loss": 0.4532,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6,
      "eval_loss": 1.4901623725891113,
      "eval_runtime": 6.8454,
      "eval_samples_per_second": 547.814,
      "eval_steps_per_second": 8.619,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3777942955493927,
      "learning_rate": 1.9370028791882848e-05,
      "loss": 0.13,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6106666666666667,
      "eval_loss": 1.5406244993209839,
      "eval_runtime": 7.3161,
      "eval_samples_per_second": 512.565,
      "eval_steps_per_second": 8.064,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}