metadata
dataset_info:
features:
- name: query
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
splits:
- name: train
num_bytes: 7437224726.5145
num_examples: 10000000
download_size: 4735377439
dataset_size: 7437224726.5145
configs:
- config_name: default
data_files:
- split: train
path: data/train-*