resnet50_os16 forward pass error

[AmeeraBawazir]

I'm trying to use resent 50 as an encoder, however it gives me this error. The code works fine with reset 18 and 34. Any idea ?

conv3d() received an invalid combination of arguments - got (Tensor, Parameter, NoneType, tuple, tuple, tuple, int), but expected one of:

• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, tuple of ints padding, tuple of ints dilation, int groups) didn't match because some of the arguments have invalid types: (Tensor, Parameter, NoneType, tuple, tuple, tuple, int)
• (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, str padding, tuple of ints dilation, int groups) didn't match because some of the arguments have invalid types: (Tensor, Parameter, NoneType, tuple, tuple, tuple, int)

[LucaLumetti]

I noticed this problem too, this problem is given by a double implementation of the class Bottleneck in resnet.py. Have you managed to solve that?

[JEpark-hutom]

I solved this issue! Thanks @LucaLumetti

[green-hand-shandong]

Hi, i'm a new coder in this field, could you please give some details？ @JEpark-hutom

[hackkhai]

@JEpark-hutom Can you help me with the fix?

[benj888]

Bottleneck and BasicBlock should be divided into two groups for modification.

[zhuhaoran666]

I solved this issue. You should put the if type(dilation) != type(1): dilation = 1 in the class Bottleneck.