OhMen Lifelike v2

(2024/02/02) OhMen Toontastic is online ! Link here

>>> Please, make sure to check out the model description, especially the "best settings" and "to go prompts" sections <<<

OhMen Lifelike is the enhanced version of OhMen Origins, designed to create realistic and lifelike images
Check my previous mixes here.

The OhMen Lifelike mix is expertly crafted to enhance male characteristics by combining different base models and my trained mix "OhMen Origins".
It simplifies the creation of lifelike / photorealistic images of handsome and beefy males, without depending solely on LoRAs.
However, if preferred, LoRAs can still be utilized as the mix offers great versatility.
The OhMen Lifelike mix also provides a wide range of sub-styles offering even more options for artistic expression.

If you're interested in bringing OhMen Lifelike to life through merging, you can refer to the "Merge a OhMen model" section. This will guide you on how to combine and blend different elements to create your desired result.

╭───────────────────────╮

│My to go prompts

╰───────────────────────╯

Positive prompt : photo (1man), muscular, brown eyes, casual, bulge,

Positive prompt (????) : (1man),flaccid ?????, saggy ?????????, muscular, brown eyes,nude 

Negative prompt : (render, cartoon, cgi, render, illustration, painting, drawing,deviantart), FastNegativeV2, deformityv6,  bad-hands-5,

Fastnegative / Deformityv6

╭───────────────────────╮

│Best Settings

╰───────────────────────╯

• Clip 1 is highly recommended, while Clip 2 will likely generate pics with less details.

• DPM++ 2M Karras, DPM++ 2M SDE Karras, or my personal favorite, DPM++ 3M SDE Exponential Sampler. Restart works well too

• For sampling, try using 35-60 steps.

• To get the best results, set your CFG scale at 6. It works great between 3.5 and 8.

• Make sure to uncheck the Face restore option.

• Recommended resolutions fall between 512x512 and 768x768.
  You can go for 1024x1024, but sometimes it can generate some weird pics.

• Keep in mind that upscaling can sometimes reduce the level of details instead of increasing it.
  To counter that, I recommend using a denoising strength of 0.35 or 0.4.
  Also, try using 15-20 Hires steps and either "NMKD-Superscale-SP_178000_G" for most of your gens or "8x NMKD Faces" for close-ups and portraits as your upscaler.

• If you want your faces to look super clean and even more realistic, you should totally give the AfterDetailer extension a try. It's not mandatory since OhMen usually does a really great job with portrait, but using it can add that extra touch of realism to your composition. Same for ???????, but you can check out this ????? model for the AfterDetailer extension ;

• The Vae-ft-mse-840000-ema-pruned is included but you may need to set your VAE on 'automatic';

• If you wanna gen some ???? pictures, add the "flaccid ?????, ????, saggy ?????, ..." tags at the beginning of your prompt.
  

╭───────────────────────╮

│Known Issues

╰───────────────────────╯
V1

• Lack of color in my gens: add the following in your negative prompt as suggested by @boomerang: "(colorless, HDR, flat, lens flare)."

• Skin appears 3D generated: (render, cartoon, CGI, illustration, painting, drawing, deviantart).

• Feet aren't properly generated : SD doesn't generate feet properly most of the time. However, you can solve this issue by adding the "Neg Feet" embedding in your negative prompt or using a LORA such as the "Male Feet Pose" one.

• You don't like beefy or super hairy guys ? Add "(Muscular:1.5) or "(hairy:1.5)" in your negative prompt, I'm gonna try to fix that in the V2. This issue has been fixed in V2.

V2 (No known issues reported yet, please report them in the "Discussion" section)

╭───────────────────────╮

│Merging a OhMen model

╰───────────────────────╯

A OhMen model represents the culmination of countless hours dedicated to make it stable and easy to work with. So, I recommend the following:

• Set OhMen at 0.5-0.6 as Checkpoint A (if you plan to use it as a base model) or 0.35-0.5 if OhMen is your Checkpoint B.

• If you plan to merge LoRAs, merge them with higher values before adding OhMen to your mix.

• As OhMen already includes HQ built-in ?????, set a lower value if you are merging one.

In case you are block merging and only want a specific feature from the model (as checkpoint B) :

• Strong overall style, textures affected: 0,0,0.5,0.5,1,1,1,0.5,1,0.5,0.5,0.5,0,0,1,1,0,0,0,0.5,0.5,0.5,0.5,1,1

• Strong overall style and composition, ???????? untouched (from a Rentry post): 1, 0.9, 0.7, 0.5, 0.3, 0.1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0.1, 0.3, 0.5, 0.7, 0.9, 1

• Alternative overall style and composition: 0,0.9,0.7,0.5,0.3,0.1,1,1,0.35,1,1,1,0,0.45,0.25,0.25,0.5,0.35,0.5,0.25,0.3,0.5,0.7,0.9,1

• strong overall style and mid-weak composition: 0,0,0.7,0.5,1,1,1,1,1,1,1,1,0,0,1,1,0,0,0,0.1,0.15,0.3,0.5,0.9,1

• Mid overall style and composition, ???????? slightly affected : 0,0.5,0.5,0.5,0.5,0.5,0.5,0.5,1,0.5,0.5,0.5,0,0,1,1,0.3,0.5,0.3,1,0.5,0.5,0.5,0.8,0.8

• weak overall style and composition, ???????? untouched: 0.2,0.5,0.5,0.5,0,0,0,0.3,0.45,1,1,1,0,0,0.45,0.4,0,0,0,0.25,0.35,0.4,0.35,0.7,0.7

• ????????, slightly affects facial features: 0,0,0,0,0,0,0,0,0,0,0,0,0,0.3,0,0,1,0.55,1,0.25,0,0,0,0,0

• Strong ????????, skin textures and faces affected : 0,0,0,0,1,1,0,0,0.5,0,0,0,0,0.5,0.5,0.5,1,0.55,1,0.45,0,0,0,0,0

• hand repair (based on a Rentry post)(may requires a top-rated or well trained model): 1,0.6,0.40,0.20,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
  

╭───────────────────────╮

│Credits

╰───────────────────────╯
A HUGE thanks to my beta testers from the Men-Only community (Unstable diffusion discord ) :

• DigitalGhosts https://civitai.com/user/digitalghosts

• Resonance https://civitai.com/user/resonance_ai

• Boomerang https://civitai.com/user/b00merang

• Revelancy https://civitai.com/user/Relevancy

• Jayar https://civitai.com/user/crazypaco
  

Checkpoints :
- Stocked Reality : https://civitai.com/models/86526/stoked-reality
- OhMen Origins V2 (coming up soon) : https://civitai.com/models/207007/ohmen-origins
- Elegant Realism : https://civitai.com/models/115920/elegant-realism
- Absolute Reality : https://civitai.com/models/81458/absolutereality
- Humans : https://civitai.com/models/98755/humans
- CyberRealistic : https://civitai.com/models/15003/cyberrealistic
V2 was 95% reworked upon the updated base models from V1

Loras:
- DPO Lora : https://civitai.com/models/242825?modelVersionId=273993
- Skin HQ : Personal Lora trained on skins
- Furthermore : https://civitai.com/models/157105/furthermore-detail-enhancer
- Contrast Control : https://civitai.com/models/155496/contrast-control-sd15
- Film Grain control : https://civitai.com/models/155526/film-grain-control-sd15
- Advanced Enhancer : https://civitai.com/models/106717/advanced-enhancer-lora
- Stock Male MJ : Personal Lora , trained on male stock-like photos sourced from my Midjourney RAW prompts
- Anymen RD : Personal Lora , from a combination of Anymen and Midjourney pics

Please note that with OhMen Lifelike, you have the ability to sell the images generated using this model. It provides you the opportunity to monetize your creations and AI artworks :)

Airfuck , KingApollo and Muserato ❤️💙