trainer_state.json

{
  "best_metric": 0.9026666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0181/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": 11.313546180725098,
      "learning_rate": 3e-05,
      "loss": 0.7835,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.884,
      "eval_loss": 0.37668731808662415,
      "eval_runtime": 7.0221,
      "eval_samples_per_second": 534.028,
      "eval_steps_per_second": 8.402,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04086972028017044,
      "learning_rate": 3e-05,
      "loss": 0.1784,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8976,
      "eval_loss": 0.3577789068222046,
      "eval_runtime": 6.7331,
      "eval_samples_per_second": 556.954,
      "eval_steps_per_second": 8.763,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.07929404079914093,
      "learning_rate": 3e-05,
      "loss": 0.0793,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.3566098213195801,
      "eval_runtime": 7.3776,
      "eval_samples_per_second": 508.293,
      "eval_steps_per_second": 7.997,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.12470819056034088,
      "learning_rate": 3e-05,
      "loss": 0.0537,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9008,
      "eval_loss": 0.38357973098754883,
      "eval_runtime": 7.2697,
      "eval_samples_per_second": 515.843,
      "eval_steps_per_second": 8.116,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1567668467760086,
      "learning_rate": 3e-05,
      "loss": 0.0391,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9013333333333333,
      "eval_loss": 0.38224485516548157,
      "eval_runtime": 6.8817,
      "eval_samples_per_second": 544.922,
      "eval_steps_per_second": 8.573,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.7451837062835693,
      "learning_rate": 3e-05,
      "loss": 0.0347,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8994666666666666,
      "eval_loss": 0.42451655864715576,
      "eval_runtime": 6.858,
      "eval_samples_per_second": 546.81,
      "eval_steps_per_second": 8.603,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3251921534538269,
      "learning_rate": 3e-05,
      "loss": 0.0271,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9026666666666666,
      "eval_loss": 0.43378138542175293,
      "eval_runtime": 7.8096,
      "eval_samples_per_second": 480.179,
      "eval_steps_per_second": 7.555,
      "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
}