metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: neg
sequence: string
- name: pos
sequence: string
splits:
- name: train
num_bytes: 3808558
num_examples: 3036
- name: test
num_bytes: 443700
num_examples: 363
- name: dev
num_bytes: 446361
num_examples: 367
download_size: 2717895
dataset_size: 4698619
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: dev
path: data/dev-*