---
dataset_info:
  features:
  - name: image_path
    dtype: string
  - name: caption
    dtype: string
  - name: question
    dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 100028841.0
    num_examples: 999
  download_size: 99402588
  dataset_size: 100028841.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---