A WIP animated style NSFW checkpoint foundation. The preview images are HEAVILY cherry-picked. Most images of this models are anatomical monsters.
This is a highly experimental model. It's not yet stable and if you are considering supporting it in early access, please do so knowingly. This model takes a lot of generations to get something half-decent, and only then if you squint.
Place the model in the /models/checkpoints directory to use in ComfyUI.
Can do NSFW, but outputs a lot of broken bodies and anatomy. Especially when it comes to male + female.
Outputs can be good every now and then.
Recommended Settings
Sampler is dpmpp_2m (DPM++ 2M) and the scheduler is beta
No VAE or CLIP is baked in. Use separate sources for those using Dual Clip Loader.
Each sample/preview image contains the used workflow. Here's a quick article with simpler more beginner-friendly workflow. This is a recommended starting point.
Description
FAQ
Comments (2)
I respect the honesty in your warning description, you have integrity.
For 22 ish gig, it really should show more face variation, does it do that?
Appreciated.
Hmm, I think not so great yet. There's something about it that produces quite photograph-like characters, which means that we do get the variety from the "human-ness".
But the style gets a lot of influence from an Arcane style/dataset, which means that it has much less character variation when this knowledge is drawn upon.
If I can't solve the model better, I'll try to extract a LoRA from it eventually. I still want to make it work well with this animation style though.
Details
Files
Available On (3 platforms)
Same model published on other platforms. May have additional downloads or version variants.