metadata
dataset_info:
features:
- name: image
dtype: image
- name: markdown
dtype: string
- name: html_source
dtype: string
- name: id
dtype: string
- name: layout_detection
list:
- name: category_id
dtype: int64
- name: latex
dtype: string
- name: poly
sequence: float64
- name: score
dtype: float64
- name: text
dtype: string
- name: reading_order
sequence: 'null'
- name: latex_extracted
sequence: string
- name: image_caption_pairs
list:
- name: caption
sequence: string
- name: footnote
sequence: 'null'
- name: image_path
dtype: string
- name: vqa
sequence: 'null'
- name: page_number
dtype: int64
- name: image_dim
sequence: int64
splits:
- name: train
num_bytes: 9547714
num_examples: 16
download_size: 8303690
dataset_size: 9547714
configs:
- config_name: default
data_files:
- split: train
path: data/train-*