workbench-bridges 0.2.11

DEPRECATED — merged into the `workbench` package. Migrate to workbench.endpoints.*

Workbench Bridges — DEPRECATED

⚠️ This package is deprecated and no longer maintained.

All functionality has been merged into the main Workbench package. New code should depend on workbench directly. Existing deployments will continue to work — this PyPI release is frozen, not removed.

Migration

workbench_bridges import	workbench replacement
from workbench_bridges.endpoints.fast_inference import fast_inference	from workbench.endpoints.fast_inference import fast_inference
from workbench_bridges.endpoints.async_inference import async_inference	from workbench.endpoints.async_inference import async_inference
from workbench_bridges.api import ParameterStore	from workbench.endpoints.parameter_store import ParameterStore
from workbench_bridges.api import DFStore	from workbench.endpoints.df_store import DFStore
from workbench_bridges.api import InferenceStore	from workbench.api import InferenceStore
from workbench_bridges.api import PublicData	from workbench.api import PublicData
from workbench_bridges.utils.execution_environment import running_as_service	from workbench.utils.execution_environment import running_as_service
from workbench_bridges.aws.sagemaker_session import get_boto3_session	from workbench.core.cloud_platform.aws.boto_session import get_boto3_session

Lambda / model-script code that previously installed workbench-bridges for its lightweight AWS surface should now pip install workbench — the core install is endpoint-safe (boto3 / pandas / sklearn / scipy / rdkit) with heavy orchestration deps moved to optional extras (workbench[aws], workbench[modeling], workbench[ui]).

The workbench.endpoints.* namespace is the contract for code running inside SageMaker endpoint containers — see Workbench's scripts/endpoint_import_smoke.py for the enforced surface.

Historical context

This package originally existed to give end-user applications a small, fast-installing way to invoke deployed Workbench endpoints and read/write AWS Parameter Store, DataFrame Store, etc., without pulling in Workbench's full orchestration dependencies. The Workbench package now provides that same lightweight surface via workbench.endpoints.* and its restructured optional-extras layout, so a separate bridges package is no longer needed.