metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 149463040
num_examples: 20320
- name: test
num_bytes: 7942128
num_examples: 1070
download_size: 64735073
dataset_size: 157405168
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*