metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 490401450
num_examples: 318115
- name: test
num_bytes: 55345869
num_examples: 35983
download_size: 97753464
dataset_size: 545747319
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*