metadata
dataset_info:
features:
- name: language
dtype: string
- name: anchor
dtype: string
- name: positive
dtype: string
splits:
- name: train
num_bytes: 94153117
num_examples: 628630
download_size: 54859797
dataset_size: 94153117
configs:
- config_name: default
data_files:
- split: train
path: data/train-*