Welcome to the aether weary Traveller.

I offer this:
A lora that helps create multiple views consistently, with a specific sketchy aesthetic to it.

I suggest using ch4r, multiple views, portrait, full body view, as your starting point.
emphasis helps, but the amount of views you will get is random. from_side, from_below etc which are known booru tags should help create variety.
you may also try emphasis on specific body parts and ad view to it, eg. "(groin view)" or "(??????? view)"

the recommended strength is 1.0, I have not tested any higher.
this was not trained on any known characters, just a couple of sketches.

i suggest using adetailer. and to catch multiple faces/anatomy, drop "detection model confidence threshold" to 0.05-0.2 (this is a setting in adetailer A1111/forge) - also, I did not use adetailer in these examples, nor did I hi-res fix. I expect generations to be even better with these.