metadata
dataset_info:
features:
- name: query
dtype: string
- name: document
dtype: string
- name: query_embedding
sequence: float32
- name: document_embedding
sequence: float32
- name: cluster_assignment
dtype: int64
splits:
- name: train
num_bytes: 11459809916
num_examples: 1695819
download_size: 13131535081
dataset_size: 11459809916
configs:
- config_name: default
data_files:
- split: train
path: data/train-*