Datasets:

Modalities:
Image
Formats:
parquet
Libraries:
Datasets
Dask
processed_celeba / README.md
giulio98's picture
Update README.md
b39a705 verified
metadata
dataset_info:
  features:
    - name: image
      dtype: image
    - name: male
      dtype:
        class_label:
          names:
            '0': female
            '1': male
    - name: young
      dtype:
        class_label:
          names:
            '0': not young
            '1': young
    - name: narrow_eyes
      dtype:
        class_label:
          names:
            '0': no narrow eyes
            '1': narrow eyes
    - name: eyeglasses
      dtype:
        class_label:
          names:
            '0': no eyeglasses
            '1': eyeglasses
    - name: hair_color
      dtype:
        class_label:
          names:
            '0': black
            '1': blond
            '2': brown
            '3': gray
            '4': other
  splits:
    - name: train
      num_bytes: 1364510021.75
      num_examples: 162770
    - name: validation
      num_bytes: 167226678.625
      num_examples: 19867
    - name: test
      num_bytes: 167357727.75
      num_examples: 19962
  download_size: 1695572347
  dataset_size: 1699094428.125
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*