---
dataset_info:
  features:
  - name: sample_id
    dtype: string
  - name: img_path
    dtype: binary
  - name: dataset_name
    dtype: string
  - name: ocr_bboxes
    struct:
    - name: word
      sequence:
        sequence: int64
  - name: ocr_predictions
    struct:
    - name: word
      sequence: string
  - name: source_text
    dtype: string
  - name: summary_text
    dtype: string
  - name: table_xml_path
    dtype: string
  splits:
  - name: train
    num_bytes: 287281984784.0
    num_examples: 223008
  download_size: 281987898601
  dataset_size: 287281984784.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---