metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: response
list:
- name: content
dtype: string
- name: role
dtype: string
- name: helpfulness
dtype: int64
- name: correctness
dtype: int64
- name: coherence
dtype: int64
- name: complexity
dtype: int64
- name: verbosity
dtype: int64
splits:
- name: train
num_bytes: 60799143
num_examples: 20324
download_size: 21755989
dataset_size: 60799143
configs:
- config_name: default
data_files:
- split: train
path: data/train-*