Add support for running models on NPU of `RPi5 with the Hailo 8L` and `Orange Pi 5 Pro (RK3588S)`

Search before asking

• I have searched the Inference issues and found no similar feature requests.

Description

User requested a feature as described below:

Hi folks,

Been loving the interface, but having issues with the inference module required on Edge devices. I am working on accessing the NPU of two of my devices a RPi5 with the Hailo 8L, and an Orange Pi 5 Pro (RK3588S). For example, the Orange Pi 5 Pro has a 6TOPS NPU that can accelerate my work quite a bit vs the CPU, Yolov8 is supported, but it needs the .pt file (or ONNX) to convert into it’s own optimized format. The Hailo for RPi 5 has a similar (albeit simpler) process and from the documentation it seems that there is built in support already for Hailo via inference. For the life of me I cannot figure out how to properly get a model export going so I can get things converted for the RK3588s’ npu. There have been past forum posts asking about this but its been a while. I really want to avoid being hardware/ecosystem locked if I can avoid it.

Any tips, tricks, thoughts?

User asks to add support for running models on NPU of RPi5 with the Hailo 8L and Orange Pi 5 Pro (RK3588S)

Use case

As explained by the user:

Orange Pi 5 Pro has a 6TOPS NPU that can accelerate my work quite a bit vs the CPU

Providing models that can be loaded onto user hardware will result in faster inference time

Additional

User states:

The Hailo for RPi 5 has a similar (albeit simpler) process and from the documentation it seems that there is built in support already for Hailo via inference.

Maybe it would be possible to achieve what user asks for through ONNXRUNTIME_EXECUTION_PROVIDERS?

Are you willing to submit a PR?

• Yes I'd like to help by submitting a PR!

No files or tests are named. Start by tracing the existing Hailo support and the ONNXRUNTIME_EXECUTION_PROVIDERS path, then determine the integration requirements for the Raspberry Pi 5 and Orange Pi 5 Pro. Done should mean models can be exported or loaded and run on both devices' NPUs with documented usage.