Lora to generate a person wearing a, sort of, "Slutty" Pop Idol outfit, if I dare call it that. I couldn't come up with something better.
Use a Dynamic Prompt, like this one, for pushing the boundaries of variety:
wearing Idol type fashion with cute design and lots of accessories {|, __view__} {|, __pose__} {|, full body}, indoors, <lora:Fashion_Slutty_Idol_000003000:1> Fashion_Slutty_Idol. Outfit style: __idol-prompts__
Use generative AI to extend the idol-prompts wildcard file further.
I know some would say: "Why the LoRA? Can't we just use the wildcards? (OR) Why the wildcards? Can't we just use the LoRA?" <= the reality is YES to both... but you can't beat the quality and the variety you get when you use them together. They reinforce each other. The LoRA is trained on a relatively small set of images, and the output is random to a degree within the learned style. But the prompt (that's where the wildcard comes in) is the part that you can turn that randomness into deliberate controlled styling.
Note: Wildcard files added as extra for your convenience: view, pose, color and idol-prompts (place in your wildcard folder). The color wildcard file is used by the idol-prompts, to extend the variability of the outfits color design.
Description
FAQ
Comments (2)
This is straight up my alley. I use a whole lot of wildcards myself, so noticing someone who not only sets up an awesome LoRA but also uses wildcards is a big yay.
I've been running it for a short while now, and so much good stuff is being put out.
(I'll be sure to submit some).
Glad you enjoy.



















