trainer_state.json

{
  "best_metric": 0.3168,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0796/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": 0.0005,
      "loss": 3.4546,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.16373333333333334,
      "eval_loss": 3.1187498569488525,
      "eval_runtime": 14.6379,
      "eval_samples_per_second": 256.185,
      "eval_steps_per_second": 4.031,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.946212768554688,
      "learning_rate": 0.0005,
      "loss": 3.0127,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.22453333333333333,
      "eval_loss": 2.945435047149658,
      "eval_runtime": 12.2821,
      "eval_samples_per_second": 305.323,
      "eval_steps_per_second": 4.804,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.294095993041992,
      "learning_rate": 0.0005,
      "loss": 2.8629,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.23573333333333332,
      "eval_loss": 2.8501713275909424,
      "eval_runtime": 13.3298,
      "eval_samples_per_second": 281.324,
      "eval_steps_per_second": 4.426,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.421351432800293,
      "learning_rate": 0.0005,
      "loss": 2.7301,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.25466666666666665,
      "eval_loss": 2.768507719039917,
      "eval_runtime": 11.7403,
      "eval_samples_per_second": 319.411,
      "eval_steps_per_second": 5.025,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 15.811654090881348,
      "learning_rate": 0.0005,
      "loss": 2.6346,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.28986666666666666,
      "eval_loss": 2.597670078277588,
      "eval_runtime": 11.0674,
      "eval_samples_per_second": 338.834,
      "eval_steps_per_second": 5.331,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.5402,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.2981333333333333,
      "eval_loss": 2.582683563232422,
      "eval_runtime": 11.2384,
      "eval_samples_per_second": 333.679,
      "eval_steps_per_second": 5.25,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 14.36782169342041,
      "learning_rate": 0.0005,
      "loss": 2.4688,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3168,
      "eval_loss": 2.505415201187134,
      "eval_runtime": 10.9984,
      "eval_samples_per_second": 340.959,
      "eval_steps_per_second": 5.364,
      "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
}