metadata
language:
- ko
dataset_info:
features:
- name: file
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 819282
num_examples: 2465
download_size: 281561
dataset_size: 819282
configs:
- config_name: default
data_files:
- split: train
path: data/train-*