metadata
language:
- en
dataset_info:
features:
- name: prompt
dtype: string
- name: answerA
dtype: string
- name: answerB
dtype: string
- name: choice
dtype: string
splits:
- name: train
num_bytes: 25216662
num_examples: 6795
download_size: 12914720
dataset_size: 25216662
configs:
- config_name: default
data_files:
- split: train
path: data/train-*