How to convert yolov3 model to libtorch model

It’s really a problem for me now. when I try https://github.com/eriklindernoren/PyTorch-YOLOv3. this project and convert yolov3.weights to yolov3.pt by torch.jit.trace(model.example), errors issued:
magic_number = pickle_module.load(f, **pickle_load_args)
_pickle.UnpicklingError: invalid load key, ‘\x00’.

my code is as below:

model = Darknet(’/home/yao/Documents/PyTorch-YOLOv3-master/config/yolov3.cfg’, 416)
model.load_darknet_weights(torch.load(’/home/yao/Documents/PyTorch-YOLOv3-master/weights/yolov3.weights’))
model.eval()

example_img = torch.rand(1,3,416,416)
with torch.jit.optimized_execution(True):
traced_script_module = torch.jit.trace(model, example_img)# save the converted model
traced_script_module.save(“yolov3.pt”)
print(‘ok’)

if any body can help me , really appreciate it!