metadata
dataset_info:
features:
- name: file
dtype: string
- name: tokens
sequence: int64
splits:
- name: train
num_bytes: 161594930
num_examples: 1020
- name: test
num_bytes: 17779984
num_examples: 128
- name: valid
num_bytes: 25546000
num_examples: 128
download_size: 25622788
dataset_size: 204920914
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*