Yeah, yeah, this is not a coffee machine.
This LoRA is meant to help generating 2.5D NSFW pics in SDXL.
I've uploaded the ComfyUI workflow (actually Capybara's workflow with some tweaks) used to generate the preview pics. It is tagged as training data (there should be an option to upload workflows together with models).
In the workflow, just input your "1girl..." prompt to the blue box.
The model is NOT as obedient as I would like it to be. I'm going to improve it.
Perpetual coffee request: Keep me motivated and caffeinated:
Description
FAQ
Comments (11)
been waiting to see what your expertise would do on the dark side. looking forwards to it
Why is being nude dark?
any chance of prompts on your pics so we can check if this works or not ?????????
I've uploaded PNGs with workflows embedded. But CivitAI converted them to jpgs for some reason. Maybe I could zip them and attach as "training data"... will do that later.
@konyconi alot of us still dont use comfy so we need prompts to see if these kind of things work before we commit.
I try to use it with juggernautXL_version3 model but it breakes the realistick style. Images moved to 2.5D style.
Yes.. it is meant for 2.5D. Should have stated it the description. Will add it.
哈哈哈哈 伟大的咖啡机之王~
now, that's my kinda coffee
Very awesome work! This is the best 2.5d I've been able to get out of SDXL. Im hoping for a good 2.5d checkpoint to come out at some point, but the results I get with this have been exceptional.
Most agreed!