metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
splits:
- name: train
num_bytes: 4765434.212180747
num_examples: 1832
- name: test
num_bytes: 265324.3939096267
num_examples: 102
- name: val
num_bytes: 265324.3939096267
num_examples: 102
download_size: 159531
dataset_size: 5296083
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: val
path: data/val-*