metadata
dataset_info:
features:
- name: question
dtype: string
- name: candidate_answers
sequence: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 4034642
num_examples: 781
download_size: 201856
dataset_size: 4034642
configs:
- config_name: default
data_files:
- split: train
path: data/train-*