yolov7 trains best.pt through its own data set, but when detecting the target picture when executing detect.py, the anchor box does not appear. But I can use the yolov7.pt that comes with it, so I would like to ask everyone.

 

There is no marked frame in the picture above. How to set it so that the marked frame appears. I found related solutions on the Internet and tried to fail.

I change best.pt to yolov7.pt and it will appear.

parser.add_argument('--weights', nargs='+', type=str, default='weights/best.pt', help='model.pt path(s)')

 

 Changed to yolov7.pt, the anchor box appears on the test picture in run/detect/exp8, as shown in the figure;

 Please advise, thank you