metadata
dataset_info:
features:
- name: query
dtype: string
- name: document
dtype: string
- name: negative
sequence: string
- name: dataset
dtype: string
splits:
- name: train
num_bytes: 57826178362
num_examples: 1695819
download_size: 29325347390
dataset_size: 57826178362
configs:
- config_name: default
data_files:
- split: train
path: data/train-*