YOLO map(mean avverage precision) implementation

I’m implementing YOLOv1 paper from scratch using PyTorch, I managed to implement the model and define its loss function correctly, and I trained it, it converges and doing very well. The thing is I want to calculate the map(mean average precision) of the model which I’m stuck on, because I don’t know how to calculate the true positives and false positives from the predictions in a reasonable way. So any help on how to compute the true and false positives given that the labels and the predictions are in shape (N,7,7,30) where N = number of examples, 7x7 = number of grids , 30 = (2*5+20) where 2 = number of bounding boxes, 5 = (confidence,x,y,w,h) , 20 = number of classes ??

https://github.com/imessam/YOLO-object-detction/