Datasets:
metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: response
dtype: string
splits:
- name: train
num_bytes: 32962041
num_examples: 67714
- name: test
num_bytes: 322359
num_examples: 684
download_size: 21130799
dataset_size: 33284400
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*