create

engine

Engine

Required

Parameter under kwargs (see Class Introduction). The parameter must be passed; otherwise, KeyError is reported.

index_type

String

Optional

Vector retrieval type. Currently, IVFFLAT and HNSW (default) are supported. This parameter is valid for dense vectors in dense and hybrid retrieval modes. HNSW is used for sparse vector retrieval and cannot be changed.

metric_type

String

Optional

Vector distance calculation mode, which can be IP (default), L2, and COSINE.

collection_name

String

Optional

Collection name, which cannot be empty. The maximum length is 1024 characters. The value must be a valid Python identifier. The default value is vectorstore.

search_mode

SearchMode

Optional

Retrieval mode. Currently, three modes are supported: DENSE for dense retrieval (default), SPARSE for sparse retrieval, and HYBRID for hybrid retrieval.

For more details, see SearchMode.

dense_dim

Integer

Optional

Dense vector dimension.

sparse_dim

Integer

Optional

Sparse vector dimension. The default value is 100000. Set this parameter based on the vocabulary size of a sparse vector model. For example, the vocabulary size of the BGE-M3 model is 250002.

params

Dict

Optional

Additional parameter of the index type. The default value is None. If None is provided, an empty dictionary is used. It corresponds to params in the add_index method. For details, see Vector Index Introduction of openGauss. For dictionary verification, the dictionary, its strings, and all iterable sequences must not exceed 1024 in length, and the number of nested dictionary layers cannot exceed 2. Parameter configurations now include "sparse" and "dense" types to specify whether they apply to sparse or dense retrieval modes. A configuration example is as follows:

{

"dense": {"lists": 200},

"sparse": {"m": 16, "ef_construction": 64}

}

OpenGaussDB

OpenGaussDB object.