metadata
dataset_info:
features:
- name: predictions
dtype: string
- name: references
dtype: string
splits:
- name: train
num_bytes: 22303
num_examples: 161
download_size: 7614
dataset_size: 22303
configs:
- config_name: default
data_files:
- split: train
path: data/train-*