Sam Xu
@DRosemei Any progress?
> Thanks for bringing this to our attention, would you be able to add the output of `nvidia-smi` to this issue? I'm trying to narrow down an nvidia library version....
> Hey specify the gpu > > `echo $CUDA_VISIBLE_DEVICES` _blank_ > > `export CUDA_VISIBLE_DEVICES=0` > > and reinstall Still the same error...
So I tried to set `export CUDA_VISIBLE_DEVICES=0` and `CMAKE_ARGS="-DLLAMA_CUBLAS=on"` none of it worked. I am going to install a new wsl and see if it's something wrong with my wsl...
> I had the same issue with Titan RTX. I tried several of the suggestions in this post, but none worked for me. I ended up compiling from the source,...
> building it from source fixed this for me I built it from the source, it did not fix it for me. Could be cuda-toolkit version?
Is there any update? I would love to see official QMK and especially VIA support for this fantastic board!