Q4_K_S possible

#1
by Hugs4Llamas - opened

Hey thanks for doing gguf on your own :-)
I'm running on mobile so speed is a limiting factor for me, I would appreciate Q4_K_S which is faster according to the numbers which thebloke once released about the different Quantization strategies.

Owner

No problem, I'll start that!

Owner

Done

Thanks, turns out that the speed is great for my usecase.

Sign up or log in to comment