Fidēlis = faithful, loyal, true, trustworthy, dependable
Note: Preview images are generated without the use of LORAs.
About
This model is aimed for photorealism but at the same time versatile and flexible with its generations. It is mostly geared towards generating muscular men with defined muscles so it might not be as good with other body types. Also, while this model is mainly created for male-oriented NSFW stuff, it can still be used to generate SFW images..
My Settings
These are just some examples of the settings I usually use and it's not really a hard requirement so you're free to experiment and see which setting works best for you:
Sampling Method: DPM++ 2M Karras or DPM++ SDE Karras
Sampling Steps: 25 - 50
Hires. fix: On
Upscaler: ESRGAN 4x+ or 4x_NMKD-Siax_200k or 8x_NMKD-Superscale_150000_G (if you have it)
Hires Steps: I usually put it at 0 so it'll be the same as Sampling Steps
Denoising Strength: 0.25 - 0.35
Upscale: 1.5 - 2
CFG Scale: 4 - 8
Clip Skip: 1-2
Extra Tips
Try to keep the prompts short and concise. No need to fill up the prompts with unnecessary wording that does the same thing like high detail, ultra detail, absurdres, hires, photorealistic, hyperrealistic, etc. Keep it simple.
ADetailer is key if you want to get good looking faces. If you're generating a close up shot or portrait, most of the time you won't need it. But if its a full body or far shot, ADetailer is absolutely necessary unless you wanna get distorted or blurry looking faces.
Hires fix is highly recommended, especially for far and wide shots where a lot of details are getting lost. And if you're using LORAs, it is mandatory to use it because without it, the output might look a little blurry and overcooked.
Use negative embeddings instead if writing paragraphs of words into the negative prompts. Avoid using too much at once. Here are some examples of the ones I usually use:
CyberRealistic Negative (Personal fav, seems to work best for me)
BadDream & UnrealisticDream (Less realistic but more creative result)
realisticvision-negative-embedding (Pretty good as well)