metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: positive
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 2990582.1
num_examples: 54000
- name: test
num_bytes: 332286.9
num_examples: 6000
download_size: 2511463
dataset_size: 3322869
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*