Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Fooocus: OSS for image generation by ControlNet author

github.com

Post reply on HN
Print view View original

1–10 of 59 posts

Re: Fooocus: OSS for image generation by ControlNet author

#2

Post by erwannmillon »

"Native refiner swap inside one single k-sampler. The advantage is that now the refiner model can reuse the base model's momentum (or ODE's history parameters) collected from k-sampling to achieve more coherent sampling. In Automatic1111's high-res fix and ComfyUI's node system, the base model and refiner use two independent k-samplers, which means the momentum is largely wasted, and the sampling continuity is broken. Fooocus uses its own advanced k-diffusion sampling that ensures seamless, native, and continuous swap in a refiner setup."

This is so interesting and seems obvious in retrospect, but super impressive! The code is simple too, going to hack around with this over the weekend :)

Re: Fooocus: OSS for image generation by ControlNet author

#4

Post by liuliu »

erwannmillon wrote: post #2

"Native refiner swap inside one single k-sampler. The advantage is that now the refiner model can reuse the base model's momentum (or ODE's history parameters) collected from k-sampling to achieve more coherent sampling. In Automatic1111's high-res fix and ComfyUI's node system, the base model and refiner use two independent k-samplers, which means the momentum is largely wasted, and the sampling continuity is broken…

Definitely interesting. It seems that I accidentally implemented that in Draw Things.

Re: Fooocus: OSS for image generation by ControlNet author

#7

Post by davely »

liuliu wrote: post #4
erwannmillon wrote: post #2

"Native refiner swap inside one single k-sampler. The advantage is that now the refiner model can reuse the base model's momentum (or ODE's history parameters) collected from k-sampling to achieve more coherent sampling. In Automatic1111's high-res fix and ComfyUI's node system, the base model and refiner use two independent k-samplers, which means the momentum is largely wasted, and the sampling continuity is broken…

Definitely interesting. It seems that I accidentally implemented that in Draw Things.

Nice! I love that app. It’s wild that I can generate images natively on my iPhone.

I’ve used some of your ideas for inspiration for ArtBot (mobile friendly web front-end for the AI Horde distributed Stable Diffusion project).

I love seeing updates to it. Keep it up! (SDXL is so much fun)

Post reply on HN