---
dataset_info:
  features:
  - name: source_images
    sequence: image
  - name: source_captions
    sequence: string
  - name: prompt
    dtype: string
  - name: target_image
    dtype: image
  - name: target_caption
    dtype: string
  splits:
  - name: train
    num_bytes: 168173014864.0
    num_examples: 360036
  download_size: 90350985520
  dataset_size: 168173014864.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---