AndrewNLauder
AndrewNLauder
What would it take to get this going?
> @mik-laj Can I volunteer for this one? Looks like nobody has taken a stab at this one. Are you still interested in it?
> Unfortunately, I cannot speak to Vulkan on Android. It could be a matter of proper configuration to get it working, but I will refrain from more speculation. If I...
Is llama.cpp planning/able to support SpinQuant? According to meta, SpinQuant + QLora are enabling really great things, and it would be great to not have to use meta's llama-stack to...