metadata
dataset_info:
features:
- name: tokens
sequence:
sequence: int64
splits:
- name: train
num_bytes: 3721244000
num_examples: 4605500
download_size: 151837798
dataset_size: 3721244000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*