metadata
dataset_info:
features:
- name: heading
dtype: string
- name: paragraph
dtype: string
splits:
- name: test
num_bytes: 1381214
num_examples: 2185
- name: train
num_bytes: 574854413
num_examples: 859965
- name: validation
num_bytes: 1219110
num_examples: 1841
download_size: 313662462
dataset_size: 577454737
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: train
path: data/train-*
- split: validation
path: data/validation-*