I am quite new to gRPC. I created my model serving Docker image. The model has a unique pre-processing pipeline (hence not using the TensorFlow serving).

I was able to create an inference.proto:

// The request message containing the images. repeated keyword for list
message InferenceRequest {
    repeated int32 length = 1;
    repeated int32 width = 2;
    repeated int32 type = 3;
    repeated bytes image = 4;
}


message InferenceReply {
    repeated bytes pred = 1;
}

I followed a tutorial online to create a simple stub and send requests in Python which works fine. I need to change the client side to C++ now. I saw limited information about using it with TensorFlow Serving and C++ clients but nothing generic.

Any help is appreciated.