metadata
dataset_info:
features:
- name: tokens
dtype: string
splits:
- name: train
num_bytes: 3933937033
num_examples: 1000000
download_size: 2537026815
dataset_size: 3933937033
configs:
- config_name: default
data_files:
- split: train
path: data/train-*