metadata
dataset_info:
features:
- name: sentence1
dtype: string
- name: sentence2
dtype: string
splits:
- name: train
num_bytes: 20906926837
num_examples: 26083582
download_size: 13976694888
dataset_size: 20906926837
configs:
- config_name: default
data_files:
- split: train
path: data/train-*