Asmodeus - the prince of demons; representing Lust...
This checkpoint is capable to produce nearly everything you can imagine... It's completely uncensored. The main focus here was creating photorealistic pictures but it can create nearly every style.
Asmodeus is NOT created to produce just porn! (you can; but don't share it here; nude art is ok though)
If you like it, feel free to give me some feedback. Also if you want to support me you can do this here.
https://ko-fi.com/afroman4peace
Version 3.0:
Well, this is the final version of this checkpoint. Asmodeus is now merges with DEMONCORE "Research" https://tensor.art/models/724032586164281309 and Midgard 1.3 https://civarchive.com/models/463847/midgard-thl-hybrid. (I highly recommend to check it out too.) Thanks to freek22.
This model combines creativity with photorealism. This model is made for things like creating nude art. You can of course use it as normal general checkpoint but be careful with it.
Don't let your kids play with this thing! NSFW images can be shown without prompting it.
This model works on a high range of cfg and steps. ( I will add more details soon).
Use about 10-16 steps.. (DPM++ SDE)
I recommend using DPM++ SDE (Karras or normal) at 3-4 cfg
Don't use DPM 2M or Heun!
A workflow will be added soon.
Known bugs:
This checkpoint can produce skimpy or even nude images by accident. If this happens (mostly when you describe the skin our breast size). Add nude, tits, NSFW.. to the negative prompt to prevent this. Be careful with the word "girl"
Do not attempt to produce naked man; seriosly don't! It will not produce male genitals
Version 2.2:
This Version is a very complex merge between several my Hephaistos (many of them are not public) and Asmodeus 1.7. This version behaves a bit better and the textures eyes and faces are sharper.
Feel free to try all versions and tell me which one do you prefer.
Why version it began with V1.7?
This checkpoint is based on my XL6 Hephaistos model V3.30 (which isn't avaible yet). The goal was to create a much more capable model regarding realism. I did this by adding several Loras and my own training data. By adding NSFW Loras I improved the anatomy of humans a lot (regarding faces and other things). The first result was terrible at first; just body parts and "human blobs". So I added some parts of my other model I am working on... this resulted in good pictures. But mostly naked woman and no art. So I remerged it with Hephaistos 2.1 and freek22s FenrisXL. (I highly recommend this checkpoint he created on base of my Hephaistos model). This stabilized it and now it works - with some flaws.. but it works.
Disclaimer and warning:
This checkpoint is fully uncensored. Only your imagination is the limit. Beware of this.. It's not intented to create just porn but it can be very horny. I would advice to consider this and maybe add NSFW in negative prompts... Don't use this to harm others... Please use this responsible! You are fully responsible for your actions!
And most important:
Don't use this to create CP! If I see such pictures here in the gallery I will flag it and you might get banned from here..
Useful tricks:
Don't use the refiner! It seems to causes only bad results.
On standard this model trys to create photos. So there is no prompt needed for this but it can enhance the results. play with different cameras and settings.
You (almost) don't need negative prompts. I mostly use the prompt "boring". To handle the skimpy behaviour of this model try to use words like tits, nude, NSFW in the negative. Sometimes especially for creatures like cats or mythical creatures add woman to the negative prompt.
(V2.2 and V1.7) I mostly use the 2M SDE Karras for my prompts at around 40-100+ steps with a cfg from 7.5-10. In my tests it showed that more steps (up to 300 steps) can increase the quality. https://civarchive.com/posts/541429
Feel free to experiment here. I haven't tested it very long.. only about a week.
Asmodeus can handle resolutions up to 1200x1200 very well. You can go above but it can increase your failure rate
To upscale your results you can just use img to img. use around 0.2-0.3 denoise. (this can vary depending on the prompt) feel free to experiment with it.
This is a workflow for Comfy UI https://civarchive.com/articles/1978/workflow-for-xl-13-asmodeus
Have fun
Credits:
Midgard by freek22
Finenude V.02 by finenude
Hella Babes, a SDXL LoRA for better nudity by vfx_ai
Film Like XL by getjinxed
Fenris XL by freek22
Description
first published version;
Do NOT use the refiner!
Known bugs:
- sometimes it will create skimpy images by accident; I try fix this behaviour in the next versions
- It can't create nude man; at least no male genitals.. Don't try it!
FAQ
Comments (29)
Nice -- I'm glad I saw the announcement of this. I just started a large X/Y/Z run to look at some Lovecraftian horror (based on some feedback on an earlier review). Considering the quality of your other models, I'm excited to be able to abort that run and include this in the experiment.
👍. Can't wait for the results. This was a nerve-racking release. A thunderstorm is raging outside and I panicked about my rig.
@Afroman4peace ... Perhaps now is a good time to unplug it? :)
@Flibertygibbet Yes I did this. I am editing the description with my smartphone :-)
just make sure to read the instructions. This Checkpoint is different than other models out there. it needs nearly no negative prompt. Don't use the refiner
@Afroman4peace Yeah, I've dispensed with generic negatives in my generation due to my earlier investigations. (Did you see my CFG post?) And Automatic1111 doesn't have refiner-step. It will be interesting to see how well it deals. My negatives right now are for age (oooof) and for anime.
@Afroman4peace Good. Remember the updated Villain's motto: "Power! Quite limited power!"
@Afroman4peace Not having the best results, honestly. It feels a touch overtuned -- it's ignored both art styles I've tried to throw at it. Let me know if you want more details.
@Afroman4peace yep, and subsequent runs also have a very odd grainy texture not present in other checkpoints.
@Flibertygibbet yes of course. what are the settings? which Sampler did you use? then CFG and step count. In my tests I found out that it only runs great with very specific settings. What styles did it ignore? This really helps me to improve it ;-) Thanks for testing it out.
@Afroman4peace Lets see here:
cute undergraduate student wearing latex sitting at a university desk listening attentively, oil painting, thick impasto
Negative prompt: child,
Steps: 50, Sampler: DPM++ 2M Karras, CFG scale: 9, Seed: 4278723622, Size: 1152x896, Model hash: 8114939f71, Model: xl13Asmodeus_v17BakedVAE, NGMS: 0.03, Version: v1.5.2, Hashes: {"model": "8114939f71"}
Just... completely disregarded in all attempts. Bumping the LoRA up to 2 made things worse.
cute undergraduate student wearing latex sitting at a university desk listening attentively, oil painting, thick impasto Negative prompt: child, Steps: 50, Sampler: DPM++ 2M Karras, CFG scale: 9, Seed: 4278723622, Size: 1152x896, Model hash: 8114939f71, Model: xl13Asmodeus_v17BakedVAE, NGMS: 0.03, Version: v1.5.2, Hashes: {"model": "8114939f71"}Had a painted face but a more realistic body. Very disconcerting.
cute Australian 20 year old undergraduate student wearing tight latex thigh highs sitting at a university desk listening attentively, cyberpunk classroom with neon and fancy computers, shiny latex, (lineart:1.5), line art, (flat colors:1.5), underboob, highly detailed background, highly detailed clothing, <lora:lineartSDXL:1.00>, <lora:SDXLHighDetail_v5:1.00>, <lora:UnderboobXL:1.00>, Negative prompt: child, photo, paintings on the walls, people in background, Steps: 50, Sampler: DPM++ 2M Karras, CFG scale: 9, Seed: 4084234636, Size: 896x1152, Model hash: 8114939f71, Model: xl13Asmodeus_v17BakedVAE, NGMS: 0.03, Lora hashes: "lineartSDXL: 8f6c79c447e5, SDXLHighDetail_v5: b6c758219f29, UnderboobXL: b44dd934dc44", Version: v1.5.2, Hashes: {"lora:lineartSDXL": "b747719129", "lora:SDXLHighDetail_v5": "a93ab595b4", "lora:UnderboobXL": "61bc40ab89", "model": "8114939f71"}Just made me try to figure out what was going on. Lot of LoRA there, so might not have been you -- but multiple attempts at Lineart have produced bleh results.
cute Australian 20 year old undergraduate student wearing tight latex thigh highs sitting at a university desk, cyberpunk student hacker caught hacking her grades, cyberpunk megacorp classroom with glowing panels and fancy computers, three fourth perspective, shiny latex, (lineart:1.5), line art, (flat colors:1.5), flat color, futuristic, highly detailed background, highly detailed cyberpunk latex schoolgirl clothing, <lora:lineartSDXL:1.00>, Negative prompt: child, photo, paintings on the walls, people in background, Steps: 40, Sampler: DPM++ 2M Karras, CFG scale: 8, Seed: 2127267132, Size: 896x1152, Model hash: 8114939f71, Model: xl13Asmodeus_v17BakedVAE, NGMS: 0.03, ADetailer model: person_yolov8n-seg.pt, ADetailer confidence: 0.3, ADetailer dilate/erode: 4, ADetailer mask blur: 4, ADetailer denoising strength: 0.4, ADetailer inpaint only masked: False, ADetailer model 2nd: face_yolov8s.pt, ADetailer confidence 2nd: 0.3, ADetailer dilate/erode 2nd: 4, ADetailer mask blur 2nd: 4, ADetailer denoising strength 2nd: 0.4, ADetailer inpaint only masked 2nd: True, ADetailer inpaint padding 2nd: 32, ADetailer version: 23.8.1, Lora hashes: "lineartSDXL: 8f6c79c447e5", Version: v1.5.2, Hashes: {"lora:lineartSDXL": "b747719129", "model": "8114939f71"}Was the one with odd spotting. (You can see a similiar spotty pattern in the Cthulhu post of your model I made yesterday: (https://civitai.com/images/2208562?period=AllTime&periodMode=published&sort=Newest&view=categories&username=flibertygibbet&withTags=true skin texture on leg)
@Flibertygibbet Thanks for testing this out. I will test your settings.Try to test it with the Sampler DPM++ SDE Karras or DPM++ 2M SDE Karras. Yesterday we tested it with DPM++ SDE Karras. a cfg from 8-10 seemed to worked fine at 100+ steps.
Here is the post. https://civitai.com/posts/541429
I need to test this further with more sampler.
@Afroman4peace Ok -- it looks like Lineart is a bit tetchy. It really likes Copax Cute but isn't behaving well on the rest. 100+ really? I'll give it a go.
@Flibertygibbet yes. It's something I never thought of either. I just vising my family so I can't do further testing at the moment. maybe add static, noisy to the negative prompt
@Afroman4peace No worries at all. Trying a dedicated run with no Adetailer at 100+ steps and your sampler.
@Afroman4peace Ok, aborted early in the run because instead of a cyberpunk hacking scene in flat art with latex -- I got a complete naked scene in a forest. I am so confused. Reducing words.
@Flibertygibbet What? It completely misbehaved? Need to improve this.. Strange thing... Well I think there is much room for improvement :-)
@Afroman4peace Yeah. It's really fighting me on this one. The negative prompt is more of a suggestion, and the positive isn't much better. I'll post a few of the ones that kind of worked, but I think it's the next version for me.
@Flibertygibbet Asmodeus has a lot of potential but it was a good decision to set it as different model instead of updating Hephaistos. But errors can be more helpful than you think.
@Afroman4peace Oh, errors are very useful. Just... kind of frustrating right now when I want to play with this line art style. :)
Very good call about this being a new model.
@Afroman4peace Posted: https://civitai.com/posts/541316 -- had to dramatically simplify the prompt and it still ignored the legs. Still, a decent result by the end. Spent way too much power going for it.
@Flibertygibbet Thanks for all your work. I think in the terms of art Hephaistos is much better. I really need to improve the model in this regard. You helped me a lot. Really thank you very much.
@Afroman4peace It was my pleasure, mate. Glad to have helped.
Hi all. Please read the instructions ;-)
Asmodeus is a odd Checkpoint.Here are some basics.Do not use the refiner! Use almost no negative prompts..
I had some good results with Dpm++ 2M SDE Karras@ 40 steps with 7cfg
I just created a workflow for Comfy UI. Here is the link: https://civitai.com/articles/1978/workflow-for-xl-13-asmodeus
Some example pictures are posted in the gallery
InvokeAI (3.0) does not want to load this model, which from all accounts is amazing. Any ideas (other than using A'1111) how to get it working?
I am not at home right now. But I will test this when I got home. Thanks for telling me this. I have tested it successfully with A1111 and comfyUI. In about two days I can test it.
I just stumbled on this on the GitHub page. https://github.com/invoke-ai/InvokeAI/issues/4348 Maybe this is helpful for you.
yesterday I did some tests. Here is a range test using the Sampler DPM++ SDE Karras.