Tensorflow 2.0 serving .Net Core 5 React/Redux client (GRPC/Rest)

Tensor Flow 2.0 Serving C# client example with gRPC and Rest. There are MNIST CNN prediction model example and ASP.NET Core 5.0 client application with ReactJS/Redux/Typescript front end.

Repository contains the following content:

• learning - python scripts with MNIST CNN model preparing using Keras and Tensorflow 2.0.
• ClientBaseLib - base library with TF Serving gRPC generated classes and utils classes to create Tensors on C# .NET Core 5 . (Code compatible with TF 1.0)
• ASP.NET Core 5.0/ ReactJS/Redux Client - SPA application gRPC/Rest clients for MNIST prediction TensorFlow 2.0 Serving

Environment requirements for running example:

• Docker
• Windows 10
• MSVS 2019
• VS Code

How to run example

• Run command promt in root directory in repository

cd src

• Init TensorFlowServing docker image with code (Required only first time)

init_tensorflow_serving_in_docker.bat

• Run trained model in TensorFlowServing docker container

start run_all_models_serving_in_docker.bat

• Build and Run .Net Client Application

start run_client.bat

• Navigate to http://localhost:5000/ and test prediction app

Available examples

Number prediction example - MNIST, based on simple CNN

Cats and Dogs classification example, based on TensorFlow tutorials and MobileNetV2