metadata
dataset_info:
features:
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 20571873813
num_examples: 346568426
download_size: 9466929004
dataset_size: 20571873813
configs:
- config_name: default
data_files:
- split: train
path: data/train-*