jd926 comments

Results 6 comments of


                                            jd926

评估Culane数据集

> > > > 您好，请问您是如何在WIndow环境下安装Opencv C++的呢？谢谢请问这个问题你解决了嘛

RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

> > > 大佬，您说的将列表 '[]' 替换为 nn.ModuleList()具体怎么实现啊 > > 大佬，请问你最终解决了吗？怎么解决这个问题啊大佬，这个问题请问你解决了嘛

AttributeError: module 'my_interp' has no attribute 'run'

> @wjywjy14 have you successfully installed the CUDA extension? > > ``` > cd my_interp > > sh build.sh > # If this fails, you might need to upgrade your...

AttributeError: module 'my_interp' has no attribute 'run'

我查看了gcc的版本 gcc version 9.4.0(Ubuntu 9.4.0-1ubuntu1-20.04.1)

AttributeError: module 'my_interp' has no attribute 'run'

> @jd926 如果出现这个问题，肯定是你编译my_interp的CUDA和你pytorch的CUDA不一样导致的。你的pytorch可能用的是通过conda安装的libcudatoolkit，而你编译my_interp的是你本地的cuda。你可以通过如下方式查看： > > * libcudatoolkit: `conda list | grep cuda` > * 本地cuda: `ls -al /usr/local/cuda` > 如果他们版本不一致，那么你需要安装一个和pytorch的libcudatoolkit版本一致的本地cuda编译my_inter 好的，非常感谢您！

Some problem about my_interp in training

> I have solved this problem. The reason why this problem arise is I have compiled my_interp with CUDA10.1 first, I just delete the old version my_interp and then compile...