metadata
dataset_info:
features:
- name: sentence1
dtype: string
- name: sentence2
dtype: string
- name: score
dtype: float32
splits:
- name: train
num_bytes: 96944086
num_examples: 828395
download_size: 26601520
dataset_size: 96944086
configs:
- config_name: default
data_files:
- split: train
path: data/train-*