metadata
dataset_info:
features:
- name: article
dtype: string
- name: scenario
dtype: string
- name: use_case
dtype: int64
- name: compliance_score
dtype: int64
- name: compliance_explanation
dtype: string
- name: plausibility_score
dtype: int64
- name: plausibility_explanation
dtype: string
- name: annotator
dtype: string
splits:
- name: train
num_bytes: 120024
num_examples: 300
download_size: 18616
dataset_size: 120024
configs:
- config_name: default
data_files:
- split: train
path: data/train-*