metadata
dataset_info:
features:
- name: query
dtype: string
- name: negatives
sequence: string
- name: positive
dtype: string
splits:
- name: train
num_bytes: 50688651
num_examples: 15408
download_size: 28696501
dataset_size: 50688651
configs:
- config_name: default
data_files:
- split: train
path: data/train-*