metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: chosen_intent
dtype: string
- name: rejected_intent
dtype: string
- name: data_from
dtype: string
- name: is_synthesized
dtype: bool
splits:
- name: train
num_bytes: 674
num_examples: 1
download_size: 7668
dataset_size: 674
configs:
- config_name: default
data_files:
- split: train
path: data/train-*