I've begun fiddling with this a bit, but I'm curious what kinds of prompts people are using to get their results, or whether they're primarily being driven by loras? How vulgar or clinical in the descriptions of the actions, et cetera, haven't seen any real discussion of that in the guide or the threads. So far I've stuck close to what appears to be the preferred prompting method for wan 2.1 in general, mildly punched up with LLMs, but have had what I would call only limited success.

E.g., this was

>Cartoon style, a charming and adorable southern girl with a scattering of freckles across her nose and cheeks, playfully winks at the viewer. She’s dressed casually in a fitted white crop-top t-shirt and classic short denim jeans. The scene is brightly and naturally lit, casting distinct, playful shadows around her. With a cheerful expression, she reaches down with a lighthearted gesture, subtly pulling up the bottom of her white t-shirt to reveal firm, pert breasts with small pink nipples. She stands confidently, radiating youthful energy. Medium shot, eye-level perspective, with a slight dynamic camera pan to follow her movement. Vibrant, colorful, and expressive animation style.