metadata
dataset_info:
features:
- name: question_id
dtype: int32
- name: question
dtype: string
- name: choices
sequence: string
- name: answer
dtype: string
- name: documents
dtype: string
splits:
- name: train
num_bytes: 56092157
num_examples: 11679
download_size: 17633520
dataset_size: 56092157
configs:
- config_name: default
data_files:
- split: train
path: data/train-*