Home Artists Posts Import Register

Downloads

Content

◀️Shameless to Spreading▶️

🖼️ See some examples 🖼️

LoRA prompt weight: 0.5~0.7
Model used for training: CyberRealisticV3

🔍Recomendations:

Activation Prompt: jkpussyspreading

Start your prompts with:

jkpussyspreading, beautiful_face, makeup, spread_pussy, lying, hands_below_legs, pussy, (anus:1.1), nipples, [woman characteristics]...

Recommended starting neg with:

(close-up:1.4), CyberRealistic_Negative-neg, easynegative, bad-hands-5...

🌟Tutorial for improving results:

Use extra seed if you want improve results. Use hires to improve details, I recommend using 0.35-0.55 denoise and 4x ultrasharp upscaler. Look at example images metadata to find more details.

LoRA details and tendencies (+resolution):

Inside the folder I left some example images.

Comments

Neal Gafter

For those on a mac, would you please consider publishing a LoRA?

Patricia

Hello what does extra seed mean. Thank you

JKCHSTR

Do you use auto1111/similar or do you do your prompts directly in the console? Because if you use auto1111 on mac it is possible to install extensions for it.

Neal Gafter

I use the Draw Anything app, which has been optimized for running on Apple silicon. Apparently Automatic1111 is poorly optimized for Apple silicon and its neural engine.

JKCHSTR

I made a tutorial explaining how to use that. https://www.patreon.com/posts/improving-image-84641571

Neal Gafter

See: 1. https://apps.apple.com/us/app/draw-things-ai-generation/id6444050820 2. https://huggingface.co/blog/diffusers-coreml 3. https://arstechnica.com/information-technology/2022/12/apple-slices-its-ai-image-synthesis-times-in-half-with-new-stable-diffusion-fix/ 4. https://forums.macrumors.com/threads/apple-silicon-in-ai-2023.2383074/page-4?post=32056404#post-32056404 5. https://wandb.ai/capecape/stable_diffusions/reports/Speed-Up-Stable-Diffusion-on-Your-M1Pro-Macbook-Pro--VmlldzoyNjY0ODYz The neural implementation is fast enough to do a decent job on an iPhone.