metadata
dataset_info:
features:
- name: sentence1
dtype: string
- name: sentence2
dtype: string
- name: score
dtype: float64
- name: similarity
dtype: float64
- name: type
dtype: string
- name: hard
dtype: int64
splits:
- name: train
num_bytes: 19963933317.00798
num_examples: 7598619
- name: test
num_bytes: 98293090.34918484
num_examples: 30000
- name: eval
num_bytes: 32764363.44972828
num_examples: 10000
download_size: 13437283463
dataset_size: 20094990770.806896
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: eval
path: data/eval-*