metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: meta
dtype: string
- name: question_type
dtype: string
- name: fact
dtype: string
- name: rationale
dtype: string
splits:
- name: train
num_bytes: 528107
num_examples: 800
download_size: 138039
dataset_size: 528107
configs:
- config_name: default
data_files:
- split: train
path: data/train-*