---
dataset_info:
  features:
  - name: image_input
    sequence:
      sequence:
        sequence: float32
  - name: text_input
    sequence: int64
  - name: _parse
    struct:
    - name: num_pages
      dtype: int64
    - name: page_indices
      sequence: int64
    - name: page_text
      sequence:
        sequence: string
  - name: text_target
    sequence: int64
  splits:
  - name: train
    num_bytes: 19008551.25
    num_examples: 6
  - name: test
    num_bytes: 6336183.75
    num_examples: 2
  download_size: 907924
  dataset_size: 25344735.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---