This is a series dedicated to bondage. I wanted a model that could handle creating complex scenes of bondage ropes, straps, chains, and cinematic ambience. This turned into a much more intensive process than I expected. But it paid off in the end.
Many models will have ropes & chains coming from nowhere, or randomly floating in the air. I was tired of seeing a mess of unorganized restraints just slapped on a person. So I made my own checkpoint that can handle clean looking bondage scenes.
ApexBDSM is highly experimental and hasn't been tested with loras or merges with other models.
This checkpoint is mixed with different architectures: Krea2, Flux2, MiniMax, & Ernie.
Text encoders have been replaced with Long Clip G & Predict Clip L.
For bondage ideas, download XBound wildcards:
If you like it post your creations!
Donations to my Whop account will exclusively go towards hardware upgrades. Civitai Buzz will be used for site generation and promotions.
Description
Settings I use:
DPM++ 3M SDE karras
3.5 CFG
40 Steps
FAQ
Comments (5)
@ironically_irascible ApexBDSM XL Chimera v1 wouldn't load at all for me after many attempts. Below is the error message returned, if that's of any help to you. All of your other SDXL-based models have always loaded just fine for me.
raise Exception(f"Could not load the {model_type} model! Reason: " + e)
Exception: Could not load the stable-diffusion model!
Reason: 'Embedding' object has no attribute 'wrapped'
This isn't surprising. What app are you using? Apex is highly experimental because it is a merge of multiple arch types. XL is the base. Maybe if I know what diffusion software you are using I can plan version 2 around that. I use Forge Classic without an issue so far.
I'm using v3 of Easy Diffusion. It's fairly old by today's standards, but generally runs ~99% of every XL-based model I've tried the past few years (well over 1,000 checkpoints I think). I've had a handful of models give similar errors when loading. I know 100% that it doesn't run smaller, compressed FP8 checkpoints for some reason, but it seems to handle FP16, BF16 and FP32 anywhere from ~6GB to >12GB, as long as they're XL-based (it doesn't run Flux, Anima, Qwen and other newer models).
Please don't go to any trouble on my behalf, especially since ApexBDSM is an experimental branch. Everyone's results are looking great and that's far more important than changing something purely for a minor use-case compatibility issue imo. Just thought I'd mention it, in case anyone else had a similar experience or if perhaps it was just something simple.
@AFD_0 I do recommend Forge Classic if you have Intel graphics. It is excellent at low GPU management with anything 4GB or larger. https://github.com/Haoming02/sd-webui-forge-classic
My new realistic favorite
Thx!