this LoRA will make your result become incredible while using Animagine XL V3.
Weight recommended is 0.7 (personal prefence), you can adjust it depend on your preference too.
for Sampling Method Euler A is recommended, no try another sampling and compare it by yourself.. because again, this is only my personal preference.
My personal Hires. fix setting:
Upscaler: 4xUltrasharp
Hires Step: 10-15
Denoising: 0.4
Description
FAQ
Comments (3)
表記のベースモデルがSD1.5になっております。
SDXLに変更されると良いでしょう。
ご報告ありがとうございます。
Hi, how to train such a detail-adder, Many thanks for some tips.

