---
dataset_info:
  features:
  - name: markdown_text
    dtype: string
  - name: pdf_metadata
    struct:
    - name: id
      dtype: string
    - name: categories
      sequence: string
  - name: chunk_metadata
    struct:
    - name: file_path
      dtype: string
    - name: page_count
      dtype: int32
    - name: page_num
      dtype: int32
    - name: table_of_content_items
      sequence:
      - name: level
        dtype: int32
      - name: title
        dtype: string
      - name: page
        dtype: int32
    - name: tables
      sequence: string
    - name: images
      sequence: string
    - name: graphics
      sequence: string
  splits:
  - name: train
    num_bytes: 537609
    num_examples: 120
  download_size: 301351
  dataset_size: 537609
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---