---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: answers
    sequence: string
  - name: ocr_tokens
    sequence: string
  splits:
  - name: train
    num_bytes: 234101029.0
    num_examples: 2000
  download_size: 233665330
  dataset_size: 234101029.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---