metadata
dataset_info:
features:
- name: sentence1
dtype: string
- name: sentence2
dtype: string
- name: score
dtype: float32
splits:
- name: train
num_bytes: 30168028
num_examples: 112696
download_size: 20067804
dataset_size: 30168028
configs:
- config_name: default
data_files:
- split: train
path: data/train-*