---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: bboxes
    list:
      list: float64
  - name: category_id
    list: int64
  - name: segmentation
    list:
      list:
        list: float64
  - name: area
    list: float64
  - name: pdf_cells
    list:
      list:
      - name: bbox
        list: float64
      - name: font
        struct:
        - name: color
          list: int64
        - name: name
          dtype: string
        - name: size
          dtype: float64
      - name: text
        dtype: string
  - name: metadata
    struct:
    - name: coco_height
      dtype: int64
    - name: coco_width
      dtype: int64
    - name: collection
      dtype: string
    - name: doc_category
      dtype: string
    - name: image_id
      dtype: int64
    - name: num_pages
      dtype: int64
    - name: original_filename
      dtype: string
    - name: original_height
      dtype: float64
    - name: original_width
      dtype: float64
    - name: page_hash
      dtype: string
    - name: page_no
      dtype: int64
  - name: pdf
    dtype: binary
  - name: modalities
    list: string
  splits:
  - name: train
    num_bytes: 55851712
    num_examples: 100
  - name: validation
    num_bytes: 45316427
    num_examples: 100
  - name: test
    num_bytes: 57504045
    num_examples: 100
  download_size: 154258619
  dataset_size: 158672184
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---