metadata
dataset_info:
features:
- name: deceive_instruction
dtype: string
- name: normal_instruction
dtype: string
- name: question
dtype: string
- name: question_prefix
dtype: string
- name: answer_prefix
dtype: string
splits:
- name: train
num_bytes: 13175
num_examples: 27
download_size: 14256
dataset_size: 13175
configs:
- config_name: default
data_files:
- split: train
path: data/train-*