metadata
dataset_info:
features:
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 1478210437
num_examples: 13560
- name: test
num_bytes: 372102436
num_examples: 3390
download_size: 656124125
dataset_size: 1850312873
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*