Huang Haiduo
Huang Haiduo
> Hi @haiduo, > > In our implementation actually stride > 1 changes the **resolution** of the coordinates, instead of the **scale** of the coordinates. After n downsampling layers, all...
Thanks for your Reply. First of all, I found the error by Vit_QAT training, and reporting: `'MatMul' object has no attribute 'input_quantizer'`, so I did this, to exclude this error...
solve?
> Since this topic is still open. This is how I was able to install Minkowski on 3090 Ti using Anaconda: > > ``` > conda create -n py3-mink python=3.8...
Who can help me take a look? I would be very grateful. The process is as followsh [issue](https://github.com/NVIDIA/MinkowskiEngine/issues/208#issuecomment-1246672195)
分类数num_class看写对了没？！
> Didn't see fully connected or channel fully connected (CFC) in your code? It's different from what you stated in your paper. > > “z = t * self.cfc[None, :,...
> @ysj9909 Thanks for your interests. I can add it by the end of May since I'm really busy recently. I will let you know once added. 大佬，您好，我有点不懂如何根据不同的query points(不同的position patch)去得到对应的attention...
> @haiduo 你好，就是算出对应位置的attention。 您好，您方便分享下这块生成attention的代码不，我还是不理解“不同的query points(不同的position patch)去得到对应的attention map”，可以的话加我的微信 abc1056530546，或者发我的邮箱 [email protected], 谢谢您的回复。