metadata
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: chunk_metadata
struct:
- name: chunk_type
dtype: string
splits:
- name: train
num_bytes: 26828352
num_examples: 12292
download_size: 12484712
dataset_size: 26828352
configs:
- config_name: default
data_files:
- split: train
path: data/train-*