metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: qid
dtype: int64
- name: id
dtype: int64
splits:
- name: train
num_bytes: 112947613
num_examples: 36972
- name: test
num_bytes: 23272551
num_examples: 6119
download_size: 57350449
dataset_size: 136220164
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*