MoP Mixes are a series of checkpoints that merge the MoP (Mixture of Perverts) model with one other model that I find interesting. These are completely separate from the normal MoP releases because those are a much more complicated merge that requires a lot of testing. MoP Mixes are much simpler to produce with the goal of creating a MoP variation that hopefully some people may find useful or just fun.
To be clear, these MoP mixes should in no way be considered better than or a replacement for the checkpoint being merged with MoP. I strongly encourage you to download and test out the checkpoint for yourself and use it in the way that its author intended.
Each checkpoint will be set up in the manner that I think works best. So some of them will have the DMD2 LoRA fully integrated, some partially and some not at all. The recommended settings for each mix will most likely be different so check the About This Version section for all the details, as well as the recommended workflow.
Note: If you want to convert any of these DMD checkpoints to a non-DMD checkpoint then all you need to do is add the DMD LoRA and set the strength to a negative value. I would start with -0.85 and adjust from there. That will allow you to use a higher CFG and different samplers than LCM.
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 a merge between MoP v5.1 DMD and WAI-illustrious with a little Zynthra thrown in for some extra color spice. Use one of my MoP Mix workflows for best results.
This is a DMD-enabled checkpoint so the recommended settings are:
Sampler: LCM
Scheduler: karras, exponential or beta
Steps: 8-14
CFG: 0.8-1.2
FAQ
Comments (25)
Bravo, you just keep hitting it outta the park! Really cool look to this one for anime.
Thanks, I'm beginning to appreciate the different cartoon/comic/anime styles more and would like to explore that further.
Ah thx für the note of converting DMD in non_DMD!
Could someone help a me a bit. If I want to use this cp like non-dmd, which dmd lora should I use with it? And should I put that lora in positive prompt or negative and then add that -0.85?
can you make non dmd for the natvis model
I have instructions in the model description for how you can remove the DMD effect yourself by adding the DMD2 lora with a negative strength.
@GBRX As a master model maker, can you briefly describe what is happening when we do this? I thought DMD was for acceleration, but I don't have to add any steps if I subtract out the DMD. Is it more about the diffusion? Really love your models...thx.
@Kid_Glitch_63 I appreciate the kind words but I'm not really a model maker, more of a model merger. I take other models that people have fine-tuned, smash them together and hope for the best.
That said, I can't say I understand all the math behind the adding and removing of the DMD lora influence on the checkpoint but it does much more than speed up inference. DMD is a highly-trained image lora itself so it can add a lot of detail too. It can also override or muddy the detail from the checkpoint so there's a tradeoff.
Your experience of removing the DMD but then not having to add steps is also something I don't fully understand. Obviously something else is going on there. My MoP v5.1 Flexible model works like this. It kind of gives the best of both worlds but there must be some drawback otherwise everyone would be doing it I think.
You nailed again. 'MoP-Mix@Babes' another awesome release. Keep it up, Bro.
MOP-MIX-BABES😍😁 I love it 👌BIG THANKS
MoP Babes... HELL YEAH! Very nice and smooth! You know, I've not really gotten the @Stable_Yogi checkpoints to work out for me in the past, but this is going to make me go back and try a little harder with them! Thanks to you both!! I posted a few images as well... Pussy shots, of course! :-P
Are we taking requests? How about a little Perfection Amateur mix from @6tZ ? I like to mix it in to some of my merges... it works pretty well for me! :-D - https://civitai.com/models/1466262?modelVersionId=2402838
I do have Perfection Cinematic merged as part of the MoP Illustrious model. Not quite the same though. I like everything from 6tz so I'll sure there will be some kind of MoP Mix eventually.
@GBRX Cool! Thanks!!
would be good if it didnt only generate Caucasian women
Which model are you referring to? I don't think I've had any problems with generating Asian, African, Hispanic, etc.
@GBRX natvis refuses to generate anything but a woman with pale skin for me
You must have something in your prompt that is causing an issue. I just created African, Hispanic and Asian images with no problem - https://civitai.com/posts/24869146.
african woman, dark skin, black curly hair, white bikini, pool, blue sky
Try starting with a simple prompt and build up from there.
@GBRX retried it and got it to work eventually!
Nah, it can create every ethnicity. Granted, I tricked it a bit with the help of the skin color Lora to shift it where I wanted it, but the results are just awesome.
Hi @GBRX
Could you please share the workflow for the video clips you just posted on this page?
Sorry, that's the only workflow I don't publish freely. It's for sale on my Patreon, https://www.patreon.com/posts/wan-2-2-image-to-141336142?source=storefront
@GBRX
Ah, I understand! No worries – I figured it might be WAN-based. Unfortunately, I don't have a powerful enough PC to run WAN 2.2 at the moment. I'm planning to upgrade my system in February, so I'll likely subscribe to your Patreon and reach out to you then. Thanks for letting me know!
@trikathik581 WAN runs surpisingly well on older hardware and small VRAM. So, give it a shot!
@efv7792
Dude, I tried with FP8 and gguf and end up with Low VRam warnings, I have just 6GB vram.






