FLUX.1 Krea: post-trained text-to-image model from Black Forest Labs and Krea

9 dvrp 6 7/31/2025, 5:38:28 PM github.com ↗

Comments (6)

sangwulee · 10h ago
Hello HackerNews. My name is Sangwu Lee . I work for Krea and I led the research efforts around the post-training for this model. I'll try to answer any questions you may have, but I recommend you read the technical report I wrote on our site (https://www.krea.ai/blog/flux-krea-open-source-release).

I also see that my colleagues already commented here, but I'll try to answer questions you may have.

oompty · 7h ago
The model looks incredible!

Regarding this part: > Since flux-dev-raw is a guidance distilled model, we devise a custom loss to finetune the model directly on a classifier-free guided distribution.

Could you go more into detail on the loss used for this and other possible tips for finetuning those? I remember the general open source ai art community had a hard time with finetuning the original distilled flux-dev so I'm very curious about that.

gnabgib · 6h ago
Best you comment on the bigger discussion (106 points, 41 comments) https://news.ycombinator.com/item?id=44745555
dvrp · 10h ago
vipermu · 10h ago
hey hn! I'm one of the founders at Krea.

we prepared a blogpost about how we trained FLUX Krea if you're interested in learning more: https://www.krea.ai/blog/flux-krea-open-source-release

ehsanakhgari · 7h ago
can you explain the TPO technique or perhaps point me to a reference please?