This lora can give you more detail, like face, background, skin, etc, deppend on what detail on your promt is.
i've tested this lora on many model, and i "think", i dont see any problem, :)
u can ajust the weight from -2 to 2
but,
FOR BETTER RESULT, USE IT BEETWEN 0.6 up to 1.4,
feel free to adjust as you want,
here's the setting i use on stable diffusion:
sampler : euler_a, DPM++ 2M karras family
step : 20 up to 40 (good)
scale : 5 to 7 (best)
clip skip : 2
Always use Hires Fix For better quality Image
Description
Details
Downloads
1,266
Platform
CivitAI
Platform Status
Available
Created
7/15/2024
Updated
9/27/2025
Deleted
-
Files
detail_booster_v1.safetensors
Mirrors
Huggingface (1 mirrors)
CivitAI (1 mirrors)