---
dataset_info:
  features:
  - name: chunk_id
    dtype: string
  - name: text
    dtype: string
  - name: metadata
    struct:
    - name: Header 1
      dtype: string
    - name: Header 2
      dtype: string
    - name: image_references
      sequence: string
    - name: images_base64
      sequence: string
    - name: start_index
      dtype: int64
  - name: source_filename
    dtype: string
  splits:
  - name: train
    num_bytes: 3967777
    num_examples: 868
  download_size: 3075318
  dataset_size: 3967777
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---