metadata
dataset_info:
features:
- name: output
dtype: string
- name: input
dtype: string
- name: instruction
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 702642
num_examples: 780
download_size: 324630
dataset_size: 702642
configs:
- config_name: default
data_files:
- split: train
path: data/train-*