metadata
dataset_info:
features:
- name: choices
struct:
- name: label
sequence: string
- name: text
sequence: string
- name: answerKey
sequence: string
- name: question
dtype: string
- name: rationale
dtype: string
- name: dataset
dtype: string
splits:
- name: train
num_bytes: 916620
num_examples: 2138
download_size: 429951
dataset_size: 916620
configs:
- config_name: default
data_files:
- split: train
path: data/train-*