Class ModelControlContext

Class Documentation

class ModelControlContext

A ModelControlContext object is used to control the model loading / unloading on the inference server.

Once created a ModelControlContext object can be used repeatedly.

A ModelControlContext object can use either HTTP protocol or GRPC protocol depending on the Create function (ModelControlHttpContext::Create or ModelControlGrpcContext::Create). For example:

std::unique_ptr<ModelControlContext> ctx;
ModelControlGrpcContext::Create(&ctx, "localhost:8000");
std::string model_name = "model";
ctx->Load(model_name);
...
ctx->Unload(model_name);
...

Public Functions

virtual ~ModelControlContext() = 0
virtual Error Load(const std::string &model_name) = 0

Load a model on the inference server.

If the model is already loaded, it will be reloaded to use the latest configuration.

Return

Error object indicating success or failure.

Parameters
  • model_name: The name of the model to be loaded.

virtual Error Unload(const std::string &model_name) = 0

Unload a model from the inference server.

Unloading a model that is not loaded on server has no affect and success code will be returned.

Return

Error object indicating success or failure.

Parameters
  • model_name: The name of the model to be unloaded.