How to write YOLO or U-Net Deep Learning code which is integrated with RasterFrames?

Hi,
I have read the RasterFrames supervised machine learning on the page of "https://rasterframes.io/supervised-learning.html", now I would like to implement Deep Learning similar to “supervised machine learning”, such as "YOLO object detection" and "U-Net semantic segmentation".
However, I am so confused about how to write YOLO or U-Net Deep Learning code which is integrated with RasterFrames? So, would you pls help to give me some suggestions? Many Thanks!