metadata
dataset_info:
features:
- name: query
dtype: string
- name: pos
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 4711923052.981122
num_examples: 4017114
download_size: 2782912795
dataset_size: 4711923052.981122
configs:
- config_name: default
data_files:
- split: train
path: data/train-*