HiUnlike the non-quantized versions of these models, it appears that trying to split it between two GPUs doesn't work
· Sign up or log in to comment