---
dataset_info:
  features:
  - name: image_sha256
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 21974810178.04
    num_examples: 153592
  download_size: 15883131179
  dataset_size: 21974810178.04
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---