metadata
dataset_info:
features:
- name: act
dtype: string
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 9816
num_examples: 170
download_size: 9196
dataset_size: 9816
configs:
- config_name: default
data_files:
- split: train
path: data/train-*