Hi HN, we're releasing weights for our latest text to image model and publishing this writeup on how we trained it in quite a bit of depth.
I hope there is something in the report for everyone, we included a fair bit on the actual training and data infrastructure usually not written about much, that I think will be interesting to people here. There's more that didn't fit, happy to answer questions!
This is a massive technical report for an open weights image gen model. As someone who has followed this space closely, it’s really cool to read about the behind-the-scenes experimentation and effort that went into the final product. I hope you will release some of the find tuning tools so the community can experiment with them as well and really push what the model’s capable of.
You can find some links and details in the GitHub readme for finetuning / LoRA support. Ostiris, musubi tuner, fal and hugging face diffusers are all day-0 supported :)
https://github.com/krea-ai/krea-2
We recommend training off the undistilled, Raw checkpoint, and then applying the LoRA to the Turbo model for inference.
Krea 2 Large (on the website and api) was trained with the FLUX 2 VAE, if you want to test it out and push realism. After working with both I think the flux VAE has a slight edge in learning realistic textures but it's smaller than you might think, the Qwen VAE was overall very good in ablations and good at learning to produce a diverse set of styles.
I hope there is something in the report for everyone, we included a fair bit on the actual training and data infrastructure usually not written about much, that I think will be interesting to people here. There's more that didn't fit, happy to answer questions!
We recommend training off the undistilled, Raw checkpoint, and then applying the LoRA to the Turbo model for inference.
May I ask how much did the training cost you?