metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 273763529.5899378
num_examples: 92198
- name: test
num_bytes: 30420479.410062183
num_examples: 10245
download_size: 38666368
dataset_size: 304184009
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*