I tried yolov4-csp,yolov4-mish pretrained weights & anchor boxes are not proper.

[akashAD98]

This is result if i used pre-trained model of mish & csp on deepsort code.

Also i tried to run same code on my custom data using yolov4-mish,csp , im getting this error ` File "main.py", line 115, in vdo_trk.run() File "main.py", line 61, in run boxes = do_detect(self.detector, sized, 0.5, 0.4, 1) File "D:\Object_tracking_deep_sort\n\deepsort_yolov4-master\deepsort_yolov4-master\utils\utils.py", line 449, in do_detect boxes.append(get_region_boxes1(list_boxes[i].data.cpu().numpy(), 0.6, 80, masked_anchors, len(anchor_masks[i]))) File "D:\Object_tracking_deep_sort\n\deepsort_yolov4-master\deepsort_yolov4-master\utils\utils.py", line 210, in get_region_boxes1 assert (output.shape[1] == (5 + num_classes) * num_anchors) AssertionError

` Hope you can help me to solve this issue. Thanks in advance. @guhuozhengling