metadata

configs:
  - config_name: default
    data_files:
      - split: full
        path: data/full-*
      - split: eval
        path: data/eval-*
dataset_info:
  features:
    - name: prompt
      dtype: string
    - name: subject
      dtype: string
    - name: uid
      dtype: int32
    - name: subject_id
      dtype: int32
  splits:
    - name: full
      num_bytes: 15084
      num_examples: 215
    - name: eval
      num_bytes: 9252
      num_examples: 150
  download_size: 15136
  dataset_size: 24336

Dataset Card

Dataset in ImagenHub.

Citation

Please kindly cite our paper if you use our code, data, models or results:

@article{ku2023imagenhub,
  title={ImagenHub: Standardizing the evaluation of conditional image generation models},
  author={Max Ku, Tianle Li, Kai Zhang, Yujie Lu, Xingyu Fu, Wenwen Zhuang, Wenhu Chen},
  journal={arXiv preprint arXiv:2310.01596},
  year={2023}
}