metadata
dataset_info:
features:
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 1479981056
num_examples: 9999872
download_size: 96514416
dataset_size: 1479981056
configs:
- config_name: default
data_files:
- split: train
path: data/train-*