this is a lora trained on 512 x 512 (get's good results even on higher resolution, just try not to surpass 768) with clip skip 2 using the anything v4.5 model, try to use thick lineart in the prompt if you want a new coffeelove68 style, or if :1 strength isn't enought try to add Coffeelove68 style in the prompt at the start (sometimes it helps sometimes it dosn't) if your having any problem with the colors try install the Vae model, i added a lot of tags that you can use in the image generation (you can see all of it in the last example images), and yes it can generate even ??????? (and sometimes but harder to get, pussy)