---
dataset_info:
  features:
  - name: markdown_text
    dtype: string
  - name: pdf_metadata
    struct:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: categories
      sequence: string
  - name: header_metadata
    struct:
    - name: Header 1
      dtype: string
    - name: Header 2
      dtype: string
    - name: Header 3
      dtype: string
    - name: Header 4
      dtype: string
    - name: Header 5
      dtype: string
  - name: chunk_metadata
    struct:
    - name: chunk_type
      dtype: string
  splits:
  - name: train
    num_bytes: 27722461
    num_examples: 12985
  download_size: 12604704
  dataset_size: 27722461
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---