metadata
dataset_info:
features:
- name: bibcode
dtype: string
- name: pdf_url
dtype: string
- name: content
dtype: string
- name: sections
dtype: string
splits:
- name: train
num_bytes: 848868244
num_examples: 9950
download_size: 361613303
dataset_size: 848868244
configs:
- config_name: default
data_files:
- split: train
path: data/train-*