metadata
dataset_info:
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 5045107916
num_examples: 185137
download_size: 3018478553
dataset_size: 5045107916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*