metadata
dataset_info:
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 13976268494
num_examples: 398792
download_size: 7376188746
dataset_size: 13976268494
configs:
- config_name: default
data_files:
- split: train
path: data/train-*