metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
splits:
- name: train
num_bytes: 6110047
num_examples: 8000
- name: test
num_bytes: 1526979
num_examples: 2000
download_size: 4526491
dataset_size: 7637026
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*