Z-Image is here, and it's pretty damn great. Let's make it more NSFW!
Still a work in progress. It can output good genitals, but not always. About 3/5 images are good, working on penis stability specifically for future versions.
ComfyUI: Models should be placed in diffusion_models, and uses the official workflow with the Load Diffusion Model loader, and loading the VAE/CLIP models separately. Models named "_AIO" should go in the checkpoints folder, and uses the normal checkpoint loaders.
Recommended Settings
Steps: 12
CFG: 1.0
Sampler: dpmpp_sde
Scheduler: simple
Description
FAQ
Comments (9)
z image base checkpoint also needed ... thanks a lot
Then go make one!
Best nsfw ZIT imo! Great quality and the AIO is very efficient on a 16gb system, though the Diffusion Model seems a little more flexible and handles Loras better so i use that.
There shouldn't be any difference with the AIO or diffusion. It should give identical or near identical results, provided you use the same models outside of the AIO ^^
But who knows, our Comfy lord work in mysterious ways.
@6tZ Good to know, checked my comparison and the seed was unlocked.
So are the diffusion models the type we use to train a character? Instead of the typical checkpoint?
From what i gathered is that you dont use the turbo checkpoints for training, you want to train it on the Z-image base model. Can confirm that they work like magic on also the turbo variants.
@MrToon Yeah i was able to successfully train on the base model but I was hoping someone would eventually release a tutorial or the appropriate checkpoint (turbo or base)to train my character so that it is more finetuned to NSFW. Can't seem to find anyone who has successfully done it.
This is trained on Turbo still. I have started training Base too, but not gotten to work with that yet. The key to making NSFW work, is training a lot. And by a lot, I mean 150-300k steps. Enjoy!