README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: image
      dtype:
        image:
          decode: false
    - name: prompt
      dtype: string
    - name: score
      dtype: float32
  splits:
    - name: train
      num_bytes: 163555377.4
      num_examples: 3200
  download_size: 143266620
  dataset_size: 163555377.4
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*