metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: def_completion
dtype: string
- name: adv_completion
dtype: string
splits:
- name: train
num_bytes: 16024143
num_examples: 31323
download_size: 7768917
dataset_size: 16024143
configs:
- config_name: default
data_files:
- split: train
path: data/train-*