metadata
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-*