metadata
dataset_info:
- config_name: default
features:
- name: headline
dtype: string
- name: news
dtype: string
- name: score
dtype: float64
splits:
- name: train
configs:
- config_name: default
data_files:
- split: train
path: data.jsonl