metadata
language:
- ko
- en
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*
dataset_info:
features:
- name: text
dtype: string
- name: ignore_loss
sequence: int64
splits:
- name: train
num_bytes: 26306600
num_examples: 10939
- name: val
num_bytes: 3872937
num_examples: 1639
download_size: 13652180
dataset_size: 30179537