---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: qa
    list:
    - name: question
      dtype: string
    - name: answer
      dtype: string
  splits:
  - name: train
    num_bytes: 1745870758.3
    num_examples: 22075
  download_size: 1773396776
  dataset_size: 1745870758.3
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---