metadata
dataset_info:
features:
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 2436318372
num_examples: 297257
download_size: 501062424
dataset_size: 2436318372