metadata
dataset_info:
features:
- name: source_paper
dtype: string
- name: source_section
dtype: string
- name: premise
dtype: string
- name: target_doi
dtype: string
- name: target_paper
dtype: string
splits:
- name: train
num_bytes: 26085341
num_examples: 77103
download_size: 8085835
dataset_size: 26085341
configs:
- config_name: default
data_files:
- split: train
path: data/train-*