Ladies' Night D&D

Ladies' Night Dungeons & Dragons - Now for ZIT!

A LoRA that can keep you from having to ue regional prompting tools in order to generate multiple character images!

While intended to create a group of women playing a tabletop RPG like D&D, this ZIT version has a bit more versatility than the previous version created for Flux.

The image below was generated with no tools used other than the Ladies' Night D&D LoRA and the simple prompt "a photograph of five beautiful women playing Dungeons & Dragons."

...created using ComfyUI, so you can "drag & drop" to open a very simple workflow, to get started, or if you'd like to make more refined images: (AKA pornography), grab the workflow from the image below to create more detailed character interactions, but still doesn't require regional prompting in order to create multiple unique characters:

While traind mostly using landscape dimensions (1536x1024 & 1920x1080), additional dimensions including 896x1152,, 1024x1024 and several others were included in this version. If you want portrait orientation, I'd recommend keeping the number of included characters under 6 for quality purposes, but with landscape orientation like 1536x1024, it can get pretty crowded while still creating highly detailed unique characters (it will create men if the prompt demands, but was trained using only female subjects).

The keywords "Ladies' Night D&D" can be included, but aren't required in order for this LoRA to activate. Also, this updated dataset was generated using only Z-Image Turbo diffusion models and includes quite a bit more NSFW training than the previous version for Flux.

Give it a try and let me know what you think!

Special thanks to slln for the request - I'd been meaning to update this LoRA for quite some time!