labeling images for object detection model: add black region?

Question

When doing a labeling job for an object detection model, (e.g. YOLO). what options do I have when an object is partially shown in the frame? specifically, if I have half a person (let's say, the bottom half) should I tag it? or leave it be? Can I, for example, remove that fraction of the frame by adding a black region covering the partial object?.