How to use OpenVINO as EP for inference modeling on NPUs

[Gusha-nye]

Hello big guys, I have downloaded a project from Github(https://github.com/microsoft/ai-powered-notes-winui3-sample?tab=readme-ov-file) and now I want to adjust the EP of the project to OPENVINO. As shown below.

But currently there is no way to enable NPU and still perform the inference process on GPU. I followed the way in that document (https://onnxruntime.ai/docs/build/eps.html#openvino) to build different EPs for ONNX Runtime, but both failed.
Currently when I run it through the following code, the output is shown below.

DmlExecutionProvider
CPUExecutionProvider

I would like to know how I can configure this to allow the project to reason on the NPU?