>>90794845NTA. but doesn't look like you're using an Illustrious based model?
Illustrious is trained to know any Danbooru tag with around 200+ images, so you can do things like;
>(anchor:1.8), holding object, object focus,Or
>(anchor:1.8), holding weapon, weapon over shoulder, Or like pic related (sorry for the repost)
>(anchor:1.8), pole dancing, oversized object, sitting on object, 