metadata
dataset_info:
features:
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 131790400
num_examples: 32144
download_size: 14459619
dataset_size: 131790400
configs:
- config_name: default
data_files:
- split: train
path: data/train-*