WongKinYiu
commented
2 years ago it due to --img-size 640 with --augment will generate the image with size not be 64x as input.
you could add check size of s in https://github.com/WongKinYiu/yolor/blob/main/models/models.py#L544 to solve the problem.
When I'm running inference with the following command: python detect.py --source ../input/ --cfg cfg/yolor_p6.cfg --weights runs/train/yolor_multy6/weights/best.pt --img-size 640 --device 3 --augment --names data/deers.names
and getting the error:
File "/home/yolor/detect.py", line 186, in
detect()
File "/home/yolor/detect.py", line 85, in detect
pred = model(img, augment=opt.augment)[0]
File "/home/miniconda3/envs/rn/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, *kwargs)
File "/home/yolor/models/models.py", line 553, in forward
y.append(self.forward_once(xi)[0])
File "/home/yolor/models/models.py", line 594, in forward_once
x = module(x, out) # WeightedFeatureFusion(), FeatureConcat()
File "/home/miniconda3/envs/rn/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(input, **kwargs)
File "/home/yolor/utils/layers.py", line 69, in forward
return torch.cat([outputs[i] for i in self.layers], 1) if self.multiple else outputs[self.layers[0]]
RuntimeError: Sizes of tensors must match except in dimension 3. Got 18 and 17 (The offending index is 0)