Hello everyone,
this is my first attempt at a checkpoint merge.
It's a mix of three of my favorite realistic models.
I'm still a complete beginner when it comes to this whole topic, but I plan to delve deeper into it in my spare time in the future.
That's why I would greatly appreciate any feedback!
The models I merged were:
RealCartoon-Realistic by 7whitefire7
Thank you for creating these amazing models!
I have used these settings almost exclusively. They are a good starting point and can be adjusted depending on what you want to generate.
Sampling method:
DPM++ 2M Karras
DPM++ 2M SDE
DPM++ 2M SDE Karras
DPM++ 3M SDE Karras
Sampling Steps:
20-40 (I use 30-38 most of the times)
CFG Scale: 5-7 (Most of the times)
Example For Negative Prompt (more realistic): (deformed iris, deformed pupils, cgi, 3d, render, sketch, cartoon, drawing, anime), text, cropped, out of frame, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck, 3d, illustration, cartoon
Example For Negative Prompt (more semi-realistic): cartoon, painting, illustration, (worst quality, low quality, normal quality:2), blurry, bad quality
clip skip 1 or 2
Recommended Upscaler Settings:
Hires. fix Upscaler:
4x_NMKD-Superscale-SP_178000_G
8x_NMKD-Superscale_150000_G
4x_foolhardy_Remacri
Hires steps: 10-20 (Most of the times I use half the sampling steps)
Denoising strength: ~ 0.3-0.5
If you can use Hires.fix for better Results!
I don't use restore faces. use ADetailer for a better face!
If you like this model, I would greatly appreciate a positive review along with constructive criticism, suggestions for improvement, or anything similar!
Description
I added a little bit more spice to it. A bit more realistic. And overall I like V2.0 on the aesthetic side more now. I hope you enjoy it.
FAQ
Comments (13)
From my experience the facial expressions sometimes are a little bit stiff and it's pretty hard to change without enough weight (or maybe cuz I'm a noob :v ). Apart from that, everything else is very decent. Thank you for your work!
You're welcome! Thank you for your feedback!
All your uploaded merges are top notch! Outstanding work!
100% agreed ! and the images @Fusch showcases to promote his merges are outstanding. Since I'm here, thanks @olaz for your spicy Loras ;-)
@BretChampagne My pleasure ;)
@olaz @BretChampagne Thank you very much for your kind words. I appreciate that! I'll try my best. ❤️
ESTE ES UN MODELO muy bueno por que tiene todos los detalles de los modelos asiaticos de anime
Thank you 👍🏻 Your work is really nice and gives me alternative ways to create. 😀
I've had a lot of great luck with this model. It handles natural language prompts really well, though admittedly my imagination and vocabulary are lacking in that area. But if you just want to specify details, you can do that just as well.
Thank you for the feedback! I'm glad that you like it!
IMO the new king of sexy time checkpoints. More realistic than Animerge and less inherently censored than Absolute Reality. And understands what you're telling it to do far better than any photorealism models I've messed with. I also haven't seen any reason to actually use clip skip 2 and I think that's a sign of a strong checkpoint.
Great work OP.
100% agreed ! I told Fusch a few weeks ago that i wasn t sure to use realSpice very often (because of its cartoonesque vibe in comparison with realLife) but today i would be sad without testing my prompts on realSpice ;-)
(Thanks for the tip about the clipskip thing, i ll make tests and comparisons soon...)
Dude, jesus, i got over 1000 hours of SD under my belt, i have never found a better semi-real model than yours! (posted a few in gallery)
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.







