Hello! Due to many requests, I decided to make my LoRA publicly available for everyone. I hope you enjoy the style, and I’m excited to see what you will create with it.

Update v3

I spent a lot of time refining my understanding of LoRA training, and for this version I pushed everything further.

This update focuses on improving style consistency, overall image quality, and cleaner outputs.

The dataset was carefully adjusted and refined to reduce errors and improve visual coherence.

You should notice more stable results, better linework, and improved color handling compared to previous versions.

No trigger word is required.

Recommended LoRA strength: 0.7 – 1.0

Hi-Res Fix and ADetailer are still recommended for best results.