---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_filename
    dtype: string
  - name: query
    dtype: string
  - name: answer
    dtype: string
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 28217643498.125
    num_examples: 39463
  download_size: 7112536377
  dataset_size: 28217643498.125
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
task_categories:
- document-question-answering
- visual-document-retrieval
---