metadata
dataset_info:
features:
- name: input_ids
list: int32
- name: attention_mask
list: int8
splits:
- name: train
num_bytes: 696988347
num_examples: 1093644
download_size: 1112212019
dataset_size: 696988347
configs:
- config_name: default
data_files:
- split: train
path: data/train-*