metadata
dataset_info:
features:
- name: input
dtype: string
- name: output
list:
- name: label
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 14568438.352396294
num_examples: 4717
- name: test
num_bytes: 769035.6476037052
num_examples: 249
download_size: 8228046
dataset_size: 15337474
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*