weight could be 0.6 - 0.1( makes it flexible for styles )
was trained off almost every picture in the game / artbook - not the 1st person POV
may be two versions with half training and full training
////////////////////////////////////////
face/body
was trained on all characters that was available - obviously certain Styles will work better with this - could perhaps do robotic, or cyborg characters not too sure
may need use of neg due to several characters and being items trained on
may spawn random text on propaganda
or could perhaps merge this style with other and get interesting results
locations
i would assume anything related to futuristic or cyberpunk or 80s/90s vibes
its usually night, city, or neon, indoors, outdoors, etc
1980s (style),