Sign In

HomoVeritas

762
5.3k
168
Verified:
SafeTensor
Type
Checkpoint Merge
Stats
1,411
Reviews
Published
Feb 18, 2024
Base Model
SD 1.5
Usage Tips
Clip Skip: 2
Hash
AutoV2
FD7CE2EEFD
default creator card background decoration
Kairen92's Avatar
Kairen92

Vēritās = truth, verity, reality, true nature, real life

Note: Preview images are generated without the use of LORAs. Some inpainting might be present.

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++ 3M SDE Karras

  • Sampling Steps: 20 - 50

  • Hires. fix: On (optional)

  • Upscaler: ESRGAN 4x+ or 4x_NMKD-Siax_200k

  • Hires Steps: I usually put it at 0 so it'll be the same as Sampling Steps

  • Denoising Strength: 0.2 - 0.3

  • Upscale: 1.5 - 2

  • CFG Scale: 4.5 - 6

Extra Tips

  1. 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.

  2. ADetailer is kinda optional for this one unless your faces are really far away and distorted. But it does help adding extra details to your character's faces with things like lines, shading, eye details, eye color, face contour, etc. Use either face_yolov8n or mediapipe_face_mesh..

  3. Use negative embeddings instead if writing paragraphs of words into the negative prompts. Here are some examples of the ones I usually use:

Credits

Models included in the merge are listed below. Credits to the original creators, go check them out:

v3 alt

What's new:

  • So I had to make a small update because I noticed that the v3 version has a lot of issues mainly its incompatibility with a lot of LORAs. Because of that I had to rebuild it again with some tweaks.

  • Most models used in the merge are basically the same except I had to remove some LORAs and reduce some of the weights included in the merge.

  • Reduced the realism because I thought the v3.0 was a little too realistic for my taste.

  • Also reduced the contrast and color grading because I thought it made some of the output a little too dark.

v3

What's new:

  • Rebuilt from scratch base off of my latest HomoFidelis v3 model.

  • Slight increase in realism, mostly things like the textures of the skin, the details on objects, etc.

  • In a attempt to make the body more manageable, I accidentally made it beefier and while penis generation is a lot better, it also tends to be bigger and thicker. Personally I don't mind it but if you aren't into the new look, the v2 version is still there.

v2

What's new:

Rebuilt from scratch base off of my latest HomoFidelis v2 model and my HomoSimile model. I realized that the previous version has a very toned down and dimmed color which I really starting to dislike. So I made this new merge with a more vibrant tone, plus with a slight improvement to realism. But not so much that it became a realistic model.