metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: task
dtype: string
- name: answer_choice
dtype: string
splits:
- name: train
num_bytes: 11282092
num_examples: 10691
- name: test
num_bytes: 434388
num_examples: 750
download_size: 4419319
dataset_size: 11716480
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*