ZIT_Center_V1

trained two Loras merged them in together at different levels.

Result: less Asian bias, pro contrast, more details, often feels sharper.

I use it like the original either with euler or dpm++ 2s a RF in Forge classic neo, usually 10 steps