metadata
dataset_info:
features:
- name: Instruction
dtype: string
- name: Response
dtype: string
splits:
- name: train
num_bytes: 3682863
num_examples: 2000
download_size: 959900
dataset_size: 3682863
configs:
- config_name: default
data_files:
- split: train
path: data/train-*