Xinzhe Geng

In my opinion, x_shift = self.gwconv_shift(torch.cat((x2, x1), dim=1)) can achieve conv with dilation 4 (green part in paper section B fig. 4), self.gwconv(x) can achieve conv with dilation 2 (red...

torch.nn里没有BReLU这个激活函数，要把作者自己写的BReLU添加到PyTorch源文件里

> @Scotty1027 > `@weak_module > class BReLU(Hardtanh): > def **init**(self, inplace=False): > super(BReLU, self).**init**(0., 1., inplace) > > ``` > def extra_repr(self): > inplace_str = 'inplace=True' if self.inplace else ''...

> 我没有修改pytorch的源文件，而是直接调用了作者写的BReLU()到网络中，具体如图

I also meet the same problem. Have you successfully trained on the binary semantic segmentation? @gjustin40

> I think now you can try to run inference with a checkpoint @czczup 你好，我按照[这个](https://github.com/czczup/ViT-Adapter/issues/41#issuecomment-1250229196)修改了代码，然后运行了`python test.py`，遇到了同样的问题 > TypeError: ms_deform_attn_backward(): incompatible function arguments. The following argument types are supported: > 1....

> > > I think now you can try to run inference with a checkpoint > > > > > > @czczup 你好，我按照[这个](https://github.com/czczup/ViT-Adapter/issues/41#issuecomment-1250229196)修改了代码，然后运行了`python test.py`，遇到了同样的问题 > > > TypeError: ms_deform_attn_backward(): incompatible...

> > > > > I think now you can try to run inference with a checkpoint > > > > > > > > > > > > @czczup...