metadata
dataset_info:
features:
- name: choice1
dtype: string
- name: choice2
dtype: string
- name: idx
dtype: int64
- name: label
dtype: int64
- name: premise
dtype: string
- name: question
dtype: string
splits:
- name: train
num_bytes: 49638
num_examples: 400
download_size: 35370
dataset_size: 49638
configs:
- config_name: default
data_files:
- split: train
path: data/train-*