metadata
dataset_info:
features:
- name: question
dtype: string
- name: choices
sequence: string
- name: answer
dtype: string
- name: rationale
dtype: string
- name: split
dtype: string
- name: source_dataset
dtype: string
splits:
- name: train
num_bytes: 6849837
num_examples: 11679
- name: test
num_bytes: 588927
num_examples: 1000
- name: validation
num_bytes: 585120
num_examples: 1000
download_size: 4543068
dataset_size: 8023884
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*