The pony version of a semi-realistic merge designed for generating good results from NSFW concept LoRAs and transformations. I personally use this model to test out my pony LoRAs. I've been sticking with this one for a while so I felt it's time to post it so that the reference images for some of my LoRAs have the model linked and available.
Generating
I found the best results were starting with 864x1152, clip skip 2, cfg 7-8, Euler A. Then img2img: resize by 2x, denoise 0.15, Script: SD Upscale -> R-ESRGAN 4x+ @ scale factor 2. (I use forge to help with vram).
Prompts
Suggested general prompt:
score_9, score_8_up, score_7_up, realistic,
Suggested img2img prompt additions:
ultra realistic, (high quality, detailed, beautiful), detailed beautiful eyes, realistic, outstanding,
<lora:add-detail-xl:0.7> <lora:xl_more_art-full_v1:0.1>, <lora:Realistic000003_1_2.5DAnime_0.05_Merge1:0.5>
Tips
This is a pony model so it will mostly generate nudity unless you specify clothing or things like "covered nipples".
Using "anthro", "furry", or "scalie" will make the results less human.
Specifying a hair color, facial features, or expressions (like "ahegao") will make the results more human.
All of the example images were generated without the "ultra realistic" tag onwards. The following tags were added for img2img. I'd suggest leaving realism and detail LoRAs out of the initial generation since they usually dampen concepts.
It has a slight watermark bias, you can add "watermark" and "signature" to the negative prompt.
Description
FAQ
Comments (15)
Is this model merged with your tf-related Loras?
if so, do I still need your loras, like your tf assist when using it?
I did not merge my loras in, but have found them to work better with this (for the style I'm looking for). I've never been successful doing lora merges but maybe I can give it another try once I'm done with v2 of the tf-assist LoRA.
@_NSFW_ Thanks for your confirmation, i did some generations after I downloaded it, and used the legtf lora, and it does need the lora to work. but I find using your model gives better result
this model is very good at unusual backgrounds, I especially like how it handles "dirty apartment, grungy, dirty walls"
Thank you for sharing this work of art and for including the generation data in the images you posted.
For those of us with learning disability, having a template to work with and tweak is a godsend.
What would you recommend to avoid black images?
if you are using A111 I would recommend switching to comfyUI which does not have that problem
One of the other best pony real checkpoints can be blended with this for good results... get duchaiten pony real, do the cyber fix on it (add cyberrealistic xl subtract sd_xl_base at 1.0 strength) then blend the unet only 50% with this, it seems to improve it
I agree! The current version I use has a bit of that thrown in :)
I am totally new on AI image generation and installed ComfyUI a few days ago. This is exactly what i came for! But since i am an absolute noob i could use a little help with the trigger words and workflow. Is there any manual for this checkpoint?
If you haven't figured some things out, DM me. I can throw out come example workflows to get you started.
@Builder001 That would be really nice, i had some nice outcomes but i still don't know what i am actually doing.
for example i want to make pictures with a realistic (not comic style) girl with cow features (not full cow but like the yawning cowgirl in the barn) but with more explicit content.
@noreg254 Realistic and weird are my jam. I'll smash together a starter tutorial for you.
@Builder001 Thanks! I really appreciate your help :-)
thank you
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.