trainer_state.json

{
  "best_metric": 0.9333333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0224/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": 6.006006006006006e-05,
      "loss": 0.7145,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.45127323269844055,
      "eval_runtime": 4.0524,
      "eval_samples_per_second": 925.378,
      "eval_steps_per_second": 14.559,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5.009009009009009e-05,
      "loss": 0.1923,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.3910045623779297,
      "eval_runtime": 4.0245,
      "eval_samples_per_second": 931.799,
      "eval_steps_per_second": 14.66,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6432070732116699,
      "learning_rate": 4.009009009009009e-05,
      "loss": 0.0751,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.39542651176452637,
      "eval_runtime": 4.8455,
      "eval_samples_per_second": 773.911,
      "eval_steps_per_second": 12.176,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5810712575912476,
      "learning_rate": 3.0090090090090087e-05,
      "loss": 0.0328,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9069333333333334,
      "eval_loss": 0.3686295449733734,
      "eval_runtime": 4.1617,
      "eval_samples_per_second": 901.075,
      "eval_steps_per_second": 14.177,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0017875133780762553,
      "learning_rate": 2.009009009009009e-05,
      "loss": 0.0102,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.916,
      "eval_loss": 0.34997624158859253,
      "eval_runtime": 4.0626,
      "eval_samples_per_second": 923.063,
      "eval_steps_per_second": 14.523,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0006678011850453913,
      "learning_rate": 1.0090090090090088e-05,
      "loss": 0.0014,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.3092302083969116,
      "eval_runtime": 4.0556,
      "eval_samples_per_second": 924.657,
      "eval_steps_per_second": 14.548,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.008986072614789009,
      "learning_rate": 9.009009009009008e-08,
      "loss": 0.0002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9333333333333333,
      "eval_loss": 0.2946105897426605,
      "eval_runtime": 4.0867,
      "eval_samples_per_second": 917.611,
      "eval_steps_per_second": 14.437,
      "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
}