metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: concept
dtype: string
splits:
- name: train
num_bytes: 597540
num_examples: 2225
download_size: 192780
dataset_size: 597540
configs:
- config_name: default
data_files:
- split: train
path: data/train-*