metadata
dataset_info:
features:
- name: Instruction
dtype: string
- name: Input
dtype: string
- name: Response
dtype: string
splits:
- name: train
num_bytes: 16863
num_examples: 20
download_size: 12335
dataset_size: 16863
configs:
- config_name: default
data_files:
- split: train
path: data/train-*