Class AlloyDBEmbeddings (0.17.0) 

AlloyDBEmbeddings(
    key: object, engine: langchain_google_alloydb_pg.engine.AlloyDBEngine, model_id: str
)

Google AlloyDB Embeddings available via Model Endpoint Management.

Methods

AlloyDBEmbeddings

AlloyDBEmbeddings(
    key: object, engine: langchain_google_alloydb_pg.engine.AlloyDBEngine, model_id: str
)

AlloyDBEmbeddings constructor.

Parameters
Name Description
key object

Prevent direct constructor usage.
engine AlloyDBEngine

Connection pool engine for managing connections to Postgres database.
model_id str

The model id used for generating embeddings.
Exceptions
Type Description
ValueError if model does not exist. Use AlloyDBModelManager to create the model.

__aembed_image

__aembed_image(image_uri: str) -> list[float]

Coroutine for generating embeddings for a given image.

Parameter
Name Description
image_uri str

Cloud Storage URI (gs://...) or base64-encoded image string to embed.
Returns
Type Description
list[float] Embedding vector.

__aembed_query

__aembed_query(query: str) -> list[float]

Coroutine for generating embeddings for a given query.

Parameter
Name Description
query str

Text to embed.
Returns
Type Description
list[float] Embedding.

__amodel_exists

__amodel_exists() -> bool

Checks if the embedding model exists.

Returns
Type Description
Bool True if a model with the given name exists, False otherwise.

aembed_documents

aembed_documents(texts: list[str]) -> list[list[float]]

Asynchronous Embed search docs.

aembed_image

aembed_image(image_uri: str) -> list[float]

Asynchronous Embed image using google_ml.image_embedding.

Requires the google_ml_integration extension with the google_ml_integration.enable_model_support database flag set to on and a registered multimodal embedding model such as multimodalembedding@001.

Parameter
Name Description
image_uri str

Cloud Storage URI (gs://...) or base64-encoded image string to embed.
Exceptions
Type Description
ValueError If image_uri is empty or whitespace.
sqlalchemy.exc.DBAPIError If the database query fails.
Returns
Type Description
list[float] Embedding vector.

aembed_query

aembed_query(text: str) -> list[float]

Asynchronous Embed query text.

Parameter
Name Description
query str

Text to embed.
Returns
Type Description
list[float] Embedding.

amodel_exists

amodel_exists() -> bool

Checks if the embedding model exists.

Returns
Type Description
Bool True if a model with the given name exists, False otherwise.

create

create(
    engine: langchain_google_alloydb_pg.engine.AlloyDBEngine, model_id: str
) -> langchain_google_alloydb_pg.embeddings.AlloyDBEmbeddings

Create AlloyDBEmbeddings instance.

Parameters
Name Description
key object

Prevent direct constructor usage.
engine AlloyDBEngine

Connection pool engine for managing connections to Postgres database.
model_id str

The model id used for generating embeddings.
Returns
Type Description
AlloyDBEmbeddings Instance of AlloyDBEmbeddings.

create_sync

create_sync(
    engine: langchain_google_alloydb_pg.engine.AlloyDBEngine, model_id: str
) -> langchain_google_alloydb_pg.embeddings.AlloyDBEmbeddings

Create AlloyDBEmbeddings instance.

Parameters
Name Description
key object

Prevent direct constructor usage.
engine AlloyDBEngine

Connection pool engine for managing connections to Postgres database.
model_id str

The model id used for generating embeddings.
Returns
Type Description
AlloyDBEmbeddings Instance of AlloyDBEmbeddings.

embed_documents

embed_documents(texts: list[str]) -> list[list[float]]

Embed search docs.

embed_image

embed_image(image_uri: str) -> list[float]

Embed image using google_ml.image_embedding.

Requires the google_ml_integration extension with the google_ml_integration.enable_model_support database flag set to on and a registered multimodal embedding model such as multimodalembedding@001.

Parameter
Name Description
image_uri str

Cloud Storage URI (gs://...) or base64-encoded image string to embed.
Exceptions
Type Description
ValueError If image_uri is empty or whitespace.
sqlalchemy.exc.DBAPIError If the database query fails.
Returns
Type Description
list[float] Embedding vector.

embed_query

embed_query(text: str) -> list[float]

Embed query text.

Parameter
Name Description
query str

Text to embed.
Returns
Type Description
list[float] Embedding.

model_exists

model_exists() -> bool

Checks if the embedding model exists.

Returns
Type Description
Bool True if a model with the given name exists, False otherwise.