metadata
dataset_info:
config_name: m1
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: response
dtype: string
- name: reward_score
dtype: float64
- name: gpt_score
dtype: float64
splits:
- name: default
num_bytes: 276939
num_examples: 64
download_size: 176218
dataset_size: 276939
configs:
- config_name: m1
data_files:
- split: default
path: m1/default-*