stable-diffusion anime aiart

This model is trained for characters from Oshi no Ko (推しの子) I plan to update it in real time along the season.

The model is trained at clip skip 1 on ACertainty You can however use the extracted locon on a lot of different models as I show in the examples on CivitAI. Otherwise you can perform (block) merge and (block) extraction yourself using the full model.

For examples check CivitAI https://civitai.com/models/40182

Screenshots and fan arts are respectively tagged with aniscreen and fanart. Using aniscreen gives both stronger anime style and better resemblance. Using aniscreen on style models may cause weird effects.

Characters

For now, we have

According to my estimate, the following characters will be added in this season

Dataset composition

Version notes

EP01

Up to this point, Aqua, Ruby, Kana, and Akane are only trained with limited fan arts, so the quality of these characters will be largely improved later after more episodes come out.

The model is trained for 42647 steps and there are in total 4 checkpoints here

The LoCon is extracted from the 32783 step one and I also recommend using this one if you prefer to use the full model.

Actually, if you want to put the star in the correct eye for Aqua and Ruby you would have higher chance to achieve it with 32783 and 42647 step checkpoints. However the last one seems to be quite overfitted. It would be much more difficult with earlier checkpoints and extracted LoCons. For comparison please see the last four example images on CivitAI.

On the other hand, for some characters with few training images, TendojiSarina notably, you may want to use early checkpoints or LoCon with lower weight to gain more flexibility.