metadata
dataset_info:
features:
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
- name: prompt
list:
- name: content
dtype: string
- name: role
dtype: string
- name: question
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 1890529
num_examples: 369
- name: test
num_bytes: 638257
num_examples: 123
download_size: 157177
dataset_size: 2528786
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*