metadata
language:
- en
dataset_info:
features:
- name: prompt
dtype: string
- name: classification
dtype:
class_label:
names:
'0': bening
'1': jailbreak
splits:
- name: train
num_bytes: 762013517.00123
num_examples: 797365
- name: test
num_bytes: 190504095.99876994
num_examples: 199342
download_size: 598336491
dataset_size: 952517613
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*