---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: canny_edges
    dtype: image
  - name: caption
    dtype: string
  - name: captions
    sequence: string
  splits:
  - name: train
    num_bytes: 2656560746.5
    num_examples: 20638
  download_size: 2648738157
  dataset_size: 2656560746.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---