Matthew Orlando
Matthew Orlando
I am having this problem and I have not customized any of the settings for vscode-clangd. Downgrading to 0.1.6 seems to have resolved it.
> After fixing this function (fairly trivial) we completed the connection perfectly fine. > PS. I'm not providing the patch directly since it's fairly easy to fix and it's not...
> Maybe it might be the case that the compiler can't prove that `rand16()` doesn't touch memory and potentially `i`, so it thinks it needs to store X and reload...
Maybe they mean to do a `git submodule update`?
You need to set a minimum architecture with `TCNN_MIN_GPU_ARCH`. E.g. I used `-DTCNN_MIN_GPU_ARCH=75` to enable features on RTX 2000 series and later cards. And then make sure your cuda compiler...
"[51 lines of output]" is probably what tells where the problem is. Figuring out how to not truncate that would be a first step
As far as I can tell, the intended use is to just locally build and install this using `setup.py` in bindings/torch. Then you can use it as part of pytorch...
> The C++ side of things is a bit more sparse. I basically had to duplicate the functionality of `modules.py` and `bindings.cpp` using libtorch. It would be much more convenient...
> `error: command 'E:\VS_BuildTool\IDE\VC\Tools\MSVC\14.40.33807\bin\Hostx64\x64\cl.exe' failed with exit code 2` I think you're on a pretty old visual studio version. NVCC uses MSVC by default. You should upgrade to VS2022 I...