Tool/software:
I would like to use the YOLOv5n (YOLO12s) object detection model provided by Ultralytics for deployment on the EVM TDA4VM platform. My plan is to retrain the model using our custom dataset and then deploy it to the TDA4VM board. Therefore, I am looking for a detailed set of steps that need to be followed after training the model — including pruning(if required), compilation (generating compiled artifacts, binary files,) and any necessary optimizations — to ensure the model runs smoothly on the board without any performance lag. Can I just get a detailed steps for above process???