metadata
dataset_info:
features:
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 124429674772
num_examples: 7913769
- name: validation
num_bytes: 1574209176
num_examples: 100000
download_size: 19685906443
dataset_size: 126003883948
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*