metadata
dataset_info:
features:
- name: query
dtype: string
- name: passage
dtype: string
- name: label
dtype: float64
splits:
- name: train
num_bytes: 420151955
num_examples: 1000000
download_size: 173869686
dataset_size: 420151955
configs:
- config_name: default
data_files:
- split: train
path: data/train-*