metadata
dataset_info:
features:
- name: question
dtype: string
- name: ref_answer
dtype: string
- name: gen_answer
dtype: string
- name: feedback
struct:
- name: rating
dtype: int64
- name: explanation
dtype: string
splits:
- name: train
num_bytes: 16378
num_examples: 40
download_size: 13614
dataset_size: 16378
configs:
- config_name: default
data_files:
- split: train
path: data/train-*