metadata
dataset_info:
features:
- name: sentence1
dtype: string
- name: sentence2
dtype: string
splits:
- name: train
num_bytes: 6408086238.602618
num_examples: 19624898
download_size: 4771685123
dataset_size: 6408086238.602618
configs:
- config_name: default
data_files:
- split: train
path: data/train-*