Yolov5x-seg custom model deployment on huggingface.co and use via API URL in CVAT.ai Models

Dear Community
I’m new here and would like to ask if somebody can help me to deploy my custom yolov5x-seg.pt model in hugginface.co, so that everybody could use it via huggingface api and CVAT.ai Models URL or can give me example Code or Tutorials or Community Links.
I try to understand the huggingface documentation, but I can’t follow up.

Thank you very much.

1 Like

I am actually hoping someone to help as I am facing the same issue. and if we can use the GPU for a live production app?

2 Likes

There are several ways to use GPUs with Hugging Face.

If you want a robust endpoint that can be accessed by multiple users, Inference Endpoints would be a good choice. If you are using YOLO, you may need to customize it a bit.

For more casual use, it is easy to create a Space as a demo and use it via Gradio API.

To use Hugging Face’s GPU

BTW, CVAT can call external model URLs, including Hugging Face.