Implementation:Tensorflow Serving Model Service pb2 grpc
| Knowledge Sources | |
|---|---|
| Domains | gRPC, API |
| Last Updated | 2026-02-13 00:00 GMT |
Overview
Auto-generated Python gRPC stubs and servicer classes for the TensorFlow Serving ModelService, providing methods for querying model status and reloading model configurations.
Description
This module is generated by the protobuf compiler from the model_service.proto definition. It provides three components: ModelServiceStub is a client-side stub class that takes a gRPC channel and creates unary-unary method stubs for GetModelStatus (sends GetModelStatusRequest, receives GetModelStatusResponse) and HandleReloadConfigRequest (sends ReloadConfigRequest, receives ReloadConfigResponse). ModelServiceServicer is an abstract server-side servicer class with stub implementations of both methods that raise NotImplementedError; server implementations must subclass this and provide real implementations. add_ModelServiceServicer_to_server() is a registration function that creates gRPC method handlers with the appropriate serializers/deserializers and adds them to a gRPC server under the service name "tensorflow.serving.ModelService". Both methods use the unary-unary RPC pattern (single request, single response).
Usage
Use ModelServiceStub on the client side to query model status or trigger configuration reloads. Subclass ModelServiceServicer on the server side to implement the model management service.
Code Reference
Source Location
- Repository: Tensorflow_Serving
- File:
tensorflow_serving/apis/model_service_pb2_grpc.py - Lines: 1-103
Signature
class ModelServiceStub(object):
def __init__(self, channel):
self.GetModelStatus = channel.unary_unary(...)
self.HandleReloadConfigRequest = channel.unary_unary(...)
class ModelServiceServicer(object):
def GetModelStatus(self, request, context): ...
def HandleReloadConfigRequest(self, request, context): ...
def add_ModelServiceServicer_to_server(servicer, server): ...
Import
from tensorflow_serving.apis import model_service_pb2_grpc
I/O Contract
Inputs
| Name | Type | Required | Description |
|---|---|---|---|
| channel | grpc.Channel |
Yes (Stub) | A gRPC channel connected to the serving server |
| request (GetModelStatus) | GetModelStatusRequest |
Yes | Request specifying which model/version to query |
| request (HandleReloadConfigRequest) | ReloadConfigRequest |
Yes | Request containing the new model server configuration |
Outputs
| Name | Type | Description |
|---|---|---|
| GetModelStatus | GetModelStatusResponse |
Status information about the requested model/versions |
| HandleReloadConfigRequest | ReloadConfigResponse |
Result of the configuration reload operation |
Usage Examples
Querying Model Status
import grpc
from tensorflow_serving.apis import model_service_pb2_grpc
from tensorflow_serving.apis import get_model_status_pb2
channel = grpc.insecure_channel('localhost:8500')
stub = model_service_pb2_grpc.ModelServiceStub(channel)
request = get_model_status_pb2.GetModelStatusRequest()
request.model_spec.name = 'my_model'
response = stub.GetModelStatus(request)
Implementing the Servicer
class MyModelServiceServicer(model_service_pb2_grpc.ModelServiceServicer):
def GetModelStatus(self, request, context):
# Implement model status logic
response = get_model_status_pb2.GetModelStatusResponse()
return response
server = grpc.server(futures.ThreadPoolExecutor(max_workers=10))
model_service_pb2_grpc.add_ModelServiceServicer_to_server(
MyModelServiceServicer(), server)