Comment on BitsFusion: 1.99 bits Weight Quantization of Diffusion Model
db0@lemmy.dbzer0.com 4 weeks ago
8x smaller! That’s pretty bonkers and awesome! I hope this will become the standard as it will allow the AI Horde workers to server 10x more models each!
Even_Adder@lemmy.dbzer0.com 4 weeks ago
Unfortunately, this won’t work with SD3 since it doesn’t use a U-Net. I wonder how many people are still training 1.5 models?
db0@lemmy.dbzer0.com 4 weeks ago
What about sdxl?
Even_Adder@lemmy.dbzer0.com 4 weeks ago
It does, so hopefully this will work on it too.