metadata
dataset_info:
features:
- name: question
dtype: string
- name: choices
sequence: string
- name: rationale
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 42333059
num_examples: 97467
- name: test
num_bytes: 116759
num_examples: 254
- name: validation
num_bytes: 118616
num_examples: 254
download_size: 25485942
dataset_size: 42568434
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*