tessl/pypi-google-cloud-aiplatform
Comprehensive Python client library for Google Cloud Vertex AI, offering machine learning tools, generative AI models, and MLOps capabilities
—
Pending
vector-search.mddocs/
Vector Search
High-performance vector similarity search with approximate nearest neighbor capabilities for embedding-based applications.
Capabilities
Matching Engine Index
Create and manage vector indices for similarity search with configurable algorithms and performance settings.
class MatchingEngineIndex:
@classmethod
def create(
cls,
display_name: str,
contents_delta_uri: str,
config: Optional[Dict] = None,
labels: Optional[Dict[str, str]] = None,
description: Optional[str] = None,
**kwargs
) -> 'MatchingEngineIndex': ...
def update_embeddings(
self,
contents_delta_uri: str,
is_complete_overwrite: bool = False,
**kwargs
) -> None: ...
def upsert_datapoints(
self,
datapoints: List[Dict[str, Any]],
update_mask: Optional[str] = None,
**kwargs
) -> None: ...
def remove_datapoints(
self,
datapoint_ids: List[str],
**kwargs
) -> None: ...Index Endpoints
Deploy indices to endpoints for serving similarity queries with traffic management.
class MatchingEngineIndexEndpoint:
@classmethod
def create(
cls,
display_name: str,
network: Optional[str] = None,
public_endpoint_enabled: bool = False,
labels: Optional[Dict[str, str]] = None,
description: Optional[str] = None,
**kwargs
) -> 'MatchingEngineIndexEndpoint': ...
def deploy_index(
self,
index: MatchingEngineIndex,
deployed_index_id: str,
display_name: Optional[str] = None,
machine_type: str = 'e2-standard-2',
min_replica_count: int = 1,
max_replica_count: int = 1,
enable_access_logging: bool = False,
**kwargs
) -> None: ...
def match(
self,
deployed_index_id: str,
queries: List[List[float]],
num_neighbors: int = 1,
filter: Optional[List[Dict[str, Any]]] = None,
**kwargs
) -> List[List[MatchNeighbor]]: ...
def batch_get_embeddings(
self,
requests: List[Dict[str, Any]],
**kwargs
) -> List[List[float]]: ...Install with Tessl CLI
npx tessl i tessl/pypi-google-cloud-aiplatform