metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 113452738560
num_examples: 11070720
- name: test
num_bytes: 12605870088
num_examples: 1230081
download_size: 44320963424
dataset_size: 126058608648
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*