---
dataset_info:
  features:
  - name: pdf_path
    dtype: string
  - name: page_number
    dtype: int64
  - name: bbox
    sequence: float64
  - name: text
    dtype: string
  - name: bbox_type
    dtype: string
  - name: image_bbox
    struct:
    - name: bytes
      dtype: binary
    - name: path
      dtype: 'null'
  splits:
  - name: train
    num_bytes: 2100910
    num_examples: 129
  download_size: 2036174
  dataset_size: 2100910
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---