metadata
dataset_info:
features:
- name: ref
dtype: string
- name: cob
dtype: string
splits:
- name: train
num_bytes: 207060
num_examples: 370
download_size: 86090
dataset_size: 207060
configs:
- config_name: default
data_files:
- split: train
path: data/train-*