metadata
dataset_info:
features:
- name: wrong
dtype: string
- name: right
dtype: string
splits:
- name: train
num_bytes: 91564
num_examples: 635
download_size: 68151
dataset_size: 91564
configs:
- config_name: default
data_files:
- split: train
path: data/train-*