How to train my yolov8 model on a ppe dataset using transfer learning

Hi,
I am trying to train my yolov8 model which is pretrained on coco dataset to detect humans to detect ppe objects. Essentially, I want to differentiate between the person wearing a safety suit vs a person not wearing a safety suit and accordingly put the boundary box for the person.
I want to use Transfer Learning from the TAO Toolkit to do this. On top of the person detection, i want to train the model to detect the safety suit as well.
Can you help me with the step-by-step solution to do this since YOLO model is not compatible with the TAO Toolkit

Why not use Ultralytics to train? It uses transfer learning by default

Hi @meghana.v , I replied to you as well in Unable to train YOLOv6 model with TAO - #8 by Morganh.