This started as a purely experimental merge, separate from my GonzaLomo models, but has now become my main SDXL source of images. I think the name Mixture of Perverts says it all.
My primary purpose in creating this model is to generate images that I can use for WAN I2V. I wanted something that had a more amateur feel while still being able to produce good color, composition and lighting.
The v5.1 models are among my favorites and offer a wide variety in speed and usage while giving the best photorealistic results.. See the version details for the recommended settings for each of those.
v6.1 and v7.1 are closely related and offer a decent variety of styles and facial types for DMD models.
Most of these versions are DMD models, intended to produce images very rapidly. I recommend using my SDXL DMD workflow or you can choose to use the workflow embedded in the showcase images.
The recommended settings for the DMD models are:
Sampler: LCM
Scheduler: karras, exponential or beta
Steps: 6-14
CFG: 1.0-1.3
Clip Skip: 1-5 (1 doesn't work with some of the newer models)
Fooocus Setup Instructions
1) Click the Advanced checkbox.
2) On the Settings tab, leave the Preset as 'initial' and the Performance as 'Speed'.
3) On the Models tab, set both the Base and Refiner models to MoP.
4) Change the Refiner Switch At setting to anywhere from 0.75 to 0.90. I like 0.85.
5) On the Advanced tab, change the Guidance Scale to 1.
6) Change the Positive ADM Guidance Scaler to 1.
7) Change the Sampler to 'lcm'. You can leave the Scheduler as Karras or Exponential.
8) Set the Forced Overwrite of Sampling Step to 7 to 11. I like 9.
Description
A continuation of v1.0 Alt as I thought that version was better. This version has better lighting and is not quite as prone to unwanted NSFW elements.
FAQ
Comments (12)
i tried to train a lora on civitai with your model, it fails completely , like it was missing a lot of steps or the lora dmd, i have not this problem with big love ( do not include the dmd lora on the checkpoint),
I've heard from multiple people that DMD checkpoints don't work well with LoRA training. Big Love is excellent though. I highly recommend using it.
First time using this Checkpoint and I love it - very impressive results and flexibility - haven't done much with DMD and I am now a convert. Great work!
Awesome, thanks!
Is there a way to not get nipple exposure when wearing a shirt in v2.0?
Do you have an example of your prompt?
THIS + MY 4KWANLORA = absolutely detailed masterpiece
Awesome! That's a very interesting lora. I'll check it out.
what's a MY 4KWANLORA?
@felixxauk110 I assume it's this one: https://civitai.com/models/1776101/luisap-4k-wan-21-to-sdxl-4k-super-resolution?modelVersionId=2020474
I've started using this model and it looks very close to BigLove model for realism and I like that. I don't know if this model was trained on all the same looking skins/breasts but... I just have 2 things I cannot adjust.
the skins always have moles/birth marks on her body. I tried multiple negative prompts and it doesnt work. Oh and there are lots of bikini tans I can't remove too.
and the breasts always look like "the same". I try to find a prompt to make them more roundish, perky youth looking, change areola doesn't work neither.
Any ideas?
Thanks for the valid feedback. Unwanted tan lines are certainly a problem. I've struggled with that as well. The blemishes on the skin are kind of inherent in the realism in the model but I can understand why you sometimes would not want those as well.
You can try putting "(tan lines:1.5), (freckles, blemishes:1.5)" in the negative prompt and changing the CFG to 1.3. That should at least diminish those things.
As for breast type, there is probably a default type/size but I don't seem to have any issue prompting for different types such as "small perky breasts", "tiny sagging breasts", "large round breasts" or "medium pointy breasts". The "size style breasts" type of prompt seems to work for me.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.