metadata
dataset_info:
features:
- name: query
dtype: string
- name: passage
dtype: string
- name: label
dtype: float64
splits:
- name: train
num_bytes: 7968749
num_examples: 20000
download_size: 3431503
dataset_size: 7968749
configs:
- config_name: default
data_files:
- split: train
path: data/train-*