metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: direction
dtype: string
splits:
- name: train
num_bytes: 8252
num_examples: 80
- name: test
num_bytes: 1078
num_examples: 10
- name: valid
num_bytes: 989
num_examples: 10
download_size: 9492
dataset_size: 10319
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*