metadata
dataset_info:
features:
- name: query
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
- name: label
dtype: float32
splits:
- name: train
num_bytes: 70025081.4
num_examples: 95000
- name: test
num_bytes: 3685530.6
num_examples: 5000
download_size: 46492122
dataset_size: 73710612
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*