metadata
dataset_info:
features:
- name: system
dtype: string
- name: question
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: chosen_score
dtype: int64
- name: rejected_score
dtype: int64
- name: winner
dtype: string
- name: topic
dtype: string
splits:
- name: train
num_bytes: 25488402
num_examples: 7987
download_size: 13993628
dataset_size: 25488402
configs:
- config_name: default
data_files:
- split: train
path: data/train-*