dataset_info:
features:
- name: prompt
dtype: string - name: content
dtype: string - name: label
dtype: string - name: content_lang
dtype: string - name: scaffold_lang
dtype: string - name: attack_vector
dtype: string - name: attack_family
dtype: string - name: config
dtype: string - name: script
dtype: string - name: resource_tier
dtype: string - name: category
dtype: string
splits: - name: train
num_bytes: 9970960
num_examples: 20720 - name: test
num_bytes: 12456704
num_examples: 20720
download_size: 2811419
dataset_size: 22427664
configs: - config_name: default
data_files:- split: train
path: data/train-* - split: test
path: data/test-*
- split: train