metadata
dataset_info:
features:
- name: prompt
list:
- name: content
dtype: string
- name: role
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 14656
num_examples: 30
download_size: 10439
dataset_size: 14656
configs:
- config_name: default
data_files:
- split: train
path: data/train-*