metadata
dataset_info:
features:
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 13581607992
num_examples: 1657102
download_size: 2953466988
dataset_size: 13581607992