Introducing Krea2/Moss-Image Manga Alpha Image Lora
retained text
Hi, After a weekend experimenting with RGBA generation using MOSS-Image, I found myself training a LoRA for a manga project I have in mind. I was wondering if the same approach would work with Krea 2, and after doing some research with Claude, I found that there was a VAE compatible with Krea 2. So after that, it was just a matter of training with the same dataset I used for MOSS. Here are the results of both trainings: **MOSS-Image:** [https://huggingface.co/paralaif/MOSS-Manga-Alpha-Image-LoRA](https://huggingface.co/paralaif/MOSS-Manga-Alpha-Image-LoRA) **Krea 2:** [https://huggingface.co/paralaif/Krea2-Manga-Alpha-Image-LoRA](https://huggingface.co/paralaif/Krea2-Manga-Alpha-Image-LoRA) In the case of Krea 2, sadly it doesn't work with Kroma 😢, but I tried it with Kreamax Edit and got good results. However, I recommend using Base or Turbo for the best results. Is not perfect, but i think is better than i expected. Hope you find useful.
document fetched 2026-10-27 00:32 · canonical https://reddit.com/r/StableDiffusion/comments/1xfonos/introducing_krea2mossimage_manga_alpha_image_lora