metadata
dataset_info:
features:
- name: instructions
dtype: string
splits:
- name: train
num_bytes: 372401
num_examples: 4000
download_size: 96796
dataset_size: 372401
configs:
- config_name: default
data_files:
- split: train
path: data/train-*