metadata
dataset_info:
features:
- name: query
dtype: string
- name: pos
sequence: string
- name: neg
sequence: string
- name: pos_scores
sequence: float64
- name: neg_scores
sequence: float64
splits:
- name: train
num_bytes: 6124024410
num_examples: 502939
download_size: 3225800184
dataset_size: 6124024410
configs:
- config_name: default
data_files:
- split: train
path: data/train-*