Update v5 (Final)
For everyone who downloaded the version from April 26 — you may want to re-download the updated version, but feel free to do whatever you prefer.
This is the final version — no further updates will follow.
I’ve pushed the quality and style as far as possible, delivering the most consistent and refined results I could achieve.
This LoRA represents the finished version of my work — enjoy creating with it.
No trigger word is required.
Recommended strength: 0.7 – 1.0
Hi-Res Fix and ADetailer are recommended.
P.S.: Images were generated using Liquid111VAE.
Description
Added 50+ new training images and switched to a different checkpoint.
Improved overall quality, consistency, and style accuracy.
Pre-final version before V5.
FAQ
Comments (4)
Looks incredibly crisp, exactly what i prefer in most cases.
Amazing job!