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
This is an alternative to the other v1.0 model. It has a bit more of a realistic amateur feel to it. Just like the other model, it's primary purpose is for me to generate good images I can use to create I2V videos. It has the same recommended settings as the other model.
FAQ
Comments (20)
what is the difference between V1.0 Alt and V1.0?
I've added a post showing what I think is a good example of the differences, https://civitai.com/posts/20990335. The primary purpose of this model family is to create images that translate well to WAN I2V videos. I was mostly happy with V1.0 but I thought it was lacking in providing some facial details, lighting and skin textures. I think the Alt version does a little better job at those things. Also, the Alt version is less prone to NSFW than my other models although it is just as NSFW-capable when prompted. e.g. unprompted covered nipples should occur less frequently
If you decided to try out these models I would appreciate any feedback you provide. I have a great deal of respect for the quality of your own content and models.
GBRX Thanks! You certainly generate great stuff! I was planning to try them, so once I will do I will post some tests and thoughts.
GBRX I have had a chance to run the Alt version as both a refiner and for primary generation. Great model, in terms of feedback there is not much to say, I have not run into any issues. I have modified your workflow to add an extra round of up-scaling and refining.
OMG I love it :-) Because of you my hobby is becoming an addiction. Great work.
Thanks! Sorry (not sorry) about your addiction ;)
GBRX
Alt: more details, little bit more makeup, more aggressive light, little bit asian faces, less warm light....and a bunch of other little things, but overall you can’t say that one is better than the other, it really is an alternative.
Oh no, not Asians, please don't do this even a little!!!😖
But in any case, I would like to change her face, it is impossible to change her face everywhere in all checkpoints, it seems to me that everyone who makes checkpoints uses the same model of the girl, I like her but not very much 😁😅
realismenhancer585 i use [BSS] - Russian girls (Illustrious) in 0.45 for negate asian faces xDDD
I haven't noticed any tendency towards Asian faces. I had kind of thought there was a little more overall diversity in the facial types with the Alt version. Aside from that I think all your other observations are dead on. It is just meant to be a slightly different style as I figure out the best version to use for I2V videos.
Do you have a non DMD version?
No, sorry, and it's not really on the agenda to provide one. I did that with GonzaLomo v4.0 and the results weren't great. It also wasn't very popular. Trying to strip DMD2 out of a checkpoint doesn't work well and you wind up with images that are very soft and require extra processing to look decent.
GBRX I see thanks anyways :)
@GBRX I would say the main advantage of a non-DMD2 checkpoint is that it's much better for training LoRAs on. Either way, thank you for your work :)
I have tested out my samples are below check out
You did great job this checkpoint trained very well
and thanks for the WORKFLOW man its amazing
Your images are great - thanks for testing out my model!
This honestly blew me away! my first try at a fully dmd model as I was mostly into dmd loras and wasn't too impressed with the overall results, but this one is the bomb!
Hey, thanks!
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.