metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 9187152
num_examples: 100000
- name: test
num_bytes: 458795
num_examples: 5000
download_size: 2765841
dataset_size: 9645947
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*