metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: requirements
dtype: string
- name: response_0
dtype: string
- name: response_1
dtype: string
- name: check
dtype: string
- name: importance
dtype: int64
- name: response_0_1_judged_preference
sequence: float64
splits:
- name: train
num_bytes: 33330
num_examples: 10
download_size: 35763
dataset_size: 33330
configs:
- config_name: default
data_files:
- split: train
path: data/train-*