metadata
dataset_info:
features:
- name: question1
dtype: string
- name: question2
dtype: string
- name: label
dtype: int32
splits:
- name: train
num_bytes: 2984640
num_examples: 22926
download_size: 1972799
dataset_size: 2984640
configs:
- config_name: default
data_files:
- split: train
path: data/train-*