metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: positive
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 32148559
num_examples: 10000
download_size: 19887098
dataset_size: 32148559
configs:
- config_name: default
data_files:
- split: train
path: data/train-*