metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 136632
num_examples: 250
download_size: 82109
dataset_size: 136632
configs:
- config_name: default
data_files:
- split: train
path: data/train-*