News for hackers. Stuff that matters.
Re:FLUX.1 Kontext by liuliu on Thursday May 29, 2025 @03:11PM (#44129202)
Seems implementation is straightforward (very similar to everyone else, HiDream-E1, ICEdit, DreamO etc.), the magic is on data curation (which details are lightly shared).
Re:FLUX.1 Kontext by krackers on Thursday May 29, 2025 @04:45PM (#44130235)
I haven't been following image generation models closely, at a high level is this new Flux model still diffusion based, or have they moved to block autoregressive (possibly with diffusion for upscaling) similar to 4o?
Re:FLUX.1 Kontext by liuliu on Thursday May 29, 2025 @05:54PM (#44130759)
Diffusion based. There is no point to move to auto-regressive if you are not also training a multimodality LLM, which these companies are not doing that.
Re:FLUX.1 Kontext by anotherpaul on Friday May 30, 2025 @09:58AM (#44136304)
Well it's a "generative flow matching model"

That's not the same as a diffusion model.

Here is a post about the difference that seems right at first glance: https://diffusionflow.github.io/