metadata
dataset_info:
features:
- name: question
dtype: string
- name: choice
sequence: string
- name: answer
dtype: string
- name: input
dtype: string
splits:
- name: train
num_bytes: 20791639
num_examples: 54753
download_size: 9694394
dataset_size: 20791639
configs:
- config_name: default
data_files:
- split: train
path: data/train-*