---
dataset_info:
  features:
  - name: image_path
    dtype: string
  - name: text
    dtype: string
  - name: chunks
    list:
    - name: bbox
      list: int64
    - name: content
      dtype: string
    - name: label
      dtype: string
  splits:
  - name: train
    num_bytes: 343318763
    num_examples: 39728
  - name: eval
    num_bytes: 1728346
    num_examples: 200
  download_size: 239429918
  dataset_size: 345047109
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
---