README.md

metadata

size_categories:
  - 10K<n<100K
task_categories:
  - image-segmentation
  - image-to-image
pretty_name: CelebAHQ Mask Dataset
dataset_info:
  features:
    - name: image_id
      dtype: string
    - name: image
      dtype: image
    - name: annotation
      dtype: image
  splits:
    - name: train
      num_bytes: 3003968731
      num_examples: 30000
  download_size: 2990778412
  dataset_size: 3003968731
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*