Have you converted Wan2.1-14B-VACE to fp8?
#11
by QAdam - opened
Hello @QuantStack,
I'm very impressive with result from model which you have converted to GGUF. Howevers, time inference is very slow. Now I upgrade my hardware, i want to test with other format like: fp8.safetensors to increase inference time.
I would really appreciate if you can public model