trainer_state.json

{
  "best_metric": 0.9032,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0917/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": 8.757507507507508e-05,
      "loss": 0.9387,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8218666666666666,
      "eval_loss": 0.589030921459198,
      "eval_runtime": 12.0504,
      "eval_samples_per_second": 311.192,
      "eval_steps_per_second": 4.896,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6624715328216553,
      "learning_rate": 7.507507507507507e-05,
      "loss": 0.293,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.5254966616630554,
      "eval_runtime": 11.6617,
      "eval_samples_per_second": 321.566,
      "eval_steps_per_second": 5.059,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05322173237800598,
      "learning_rate": 6.257507507507508e-05,
      "loss": 0.141,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.5384753346443176,
      "eval_runtime": 12.0289,
      "eval_samples_per_second": 311.75,
      "eval_steps_per_second": 4.905,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 5.011261261261262e-05,
      "loss": 0.0754,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.5376426577568054,
      "eval_runtime": 11.9291,
      "eval_samples_per_second": 314.357,
      "eval_steps_per_second": 4.946,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.14532636106014252,
      "learning_rate": 3.7612612612612614e-05,
      "loss": 0.035,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.5332834720611572,
      "eval_runtime": 11.8126,
      "eval_samples_per_second": 317.458,
      "eval_steps_per_second": 4.995,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.005736450664699078,
      "learning_rate": 2.5112612612612612e-05,
      "loss": 0.0107,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.47782573103904724,
      "eval_runtime": 12.0804,
      "eval_samples_per_second": 310.419,
      "eval_steps_per_second": 4.884,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.211816824157722e-05,
      "learning_rate": 1.2612612612612611e-05,
      "loss": 0.003,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9032,
      "eval_loss": 0.4486350119113922,
      "eval_runtime": 11.8312,
      "eval_samples_per_second": 316.958,
      "eval_steps_per_second": 4.987,
      "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
}