metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: completion
dtype: string
- name: meta
struct:
- name: source
dtype: string
splits:
- name: train
num_bytes: 15980
num_examples: 16
download_size: 16092
dataset_size: 15980
configs:
- config_name: default
data_files:
- split: train
path: data/train-*