same notes as v3 but:
i tested this on SD2.1 base and the results are really bad, this LORA was fine tuned using Illuminati_Diffusion_v1.0 wich is almost mandatory to use this LORA.
and most of the samples were generated with the help of midjourney emmbeding.
keep it at 0.75 for better style
Alisa Bosconovitch, from tekken series