when load model set inplace=False, error is:RuntimeError: The size of tensor a (0) must match the size of tensor b (34) at non-singleton dimension 4

[whzz101]

1. start run 'python detect.py' demo is ok
2. when I set load model para inplace=False，then run 'python detect.py' error

model = attempt_load(weights, map_location=device, inplace=False)  # load FP32 model

image 1/2 /data/zhou.zhou/project/46_YiXin/1-edgeai-yolov5/data/images/bus.jpg: Traceback (most recent call last):
  File "detect_abc.py", line 203, in <module>
    detect(opt=opt)
  File "detect_abc.py", line 77, in detect
    pred = model(img, augment=opt.augment)
  File "/data/zhou.zhou/project_envs/P27_yolov5s61/lib/python3.6/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/data/zhou.zhou/project/46_YiXin/1-edgeai-yolov5/models/yolo.py", line 159, in forward
    return self.forward_once(x, profile)  # single-scale inference, train
  File "/data/zhou.zhou/project/46_YiXin/1-edgeai-yolov5/models/yolo.py", line 190, in forward_once
    x = m(x)  # run
  File "/data/zhou.zhou/project_envs/P27_yolov5s61/lib/python3.6/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
    result = self.forward(*input, **kwargs)
  File "/data/zhou.zhou/project/46_YiXin/1-edgeai-yolov5/models/yolo.py", line 100, in forward
    y[..., 6:] = (y[..., 6:] * 2. - 0.5 + self.grid[i].repeat((1,1,1,1,self.nkpt))) * self.stride[i]  # xy
RuntimeError: The size of tensor a (0) must match the size of tensor b (34) at non-singleton dimension 4

it is not support inplace set False? or has bug