metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: inputs
dtype: string
- name: response
dtype: string
splits:
- name: train
num_bytes: 757586
num_examples: 2308
download_size: 103121
dataset_size: 757586
configs:
- config_name: default
data_files:
- split: train
path: data/train-*