metadata
dataset_info:
features:
- name: functionality
dtype: string
- name: case_id
dtype: int64
- name: test_case
dtype: string
- name: label_gold
dtype: string
- name: target_ident
dtype: string
- name: direction
dtype: string
- name: focus_words
dtype: string
- name: focus_lemma
dtype: string
- name: ref_case_id
dtype: float64
- name: ref_templ_id
dtype: float64
- name: templ_id
dtype: int64
- name: case_templ
dtype: string
- name: predicted_label_llamaguard1b
dtype: int64
splits:
- name: train
num_bytes: 1455387
num_examples: 6291
download_size: 270838
dataset_size: 1455387
configs:
- config_name: default
data_files:
- split: train
path: data/train-*