I found a new AI tool Shakker, a best image to image tool. You can try it via <a target="_blank" rel="ugc" href="https://www.shakker.ai">https://www.shakker.ai</a> This LORA (Linguistic Oriented Response Algorithm) was trained using promotional images from "Azur Lane" and is specialized for generating character images on a white background. It is an early practice work by the author.You can try the online demo to experience it.In the future, the author plans to try creating characters from mobile games like "Genshin Impact" and "Code: Live."Since this is a practice project, characters do not have specific tags and their appearance is rather casual.Using the basic sampler should yield good results. If you've played certain games, you can also try using in-game character images as prompts in SD (Semi-Detailed) mode to generate new images.This model can even create entirely new characters based on the specified style.

Azur Lane poster V1.0

Azur Lane poster.safetensors

Azur Lane poster

Beauty Model V1.0

This is a LORA using the Walker Dan Furnace for an alchemy test. The unexpected results were good. The author is used to training for ten hours or even tens of hours, and rarely uses less than an hour and a half as training time. I'll upload some pictures to show the effects of this Dan. Well, you can upload it as a work at all. Trigger words are randomly filled in during training, even without effects. I recommend everyone try the Walker Dan Burner too.

Files with SHA256: cfca532b9f5901d990bc240764ea85e33c61f72caa625349ac9637f2d0f08d46