metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: answer
dtype: string
- name: idx
dtype: int64
splits:
- name: train
num_bytes: 21707
num_examples: 800
- name: validation
num_bytes: 10853
num_examples: 400
- name: test
num_bytes: 5434
num_examples: 200
download_size: 31318
dataset_size: 37994
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*