Aidan Hughes art style LoRA.
(Set LoRA weight to 1.0), trigger word is: asdKMFDM, though I don't think it is actually necessary.
Used in the album cover art of the band "KMFDM" as well as other media.
Can be used for NSFW art or not.
Trained on 35 images, example of art this file is trained on (link may contain NSFW illustrations)
Trained resolution: 512x512, learning rate: 5e-05, epoch: 1, standard LoRA type, based on SD 1.5 (recommended to try out different base models for better results)
Example prompt: asdKMFDM, strong guy aiming a large sniper rifle, colorful illustration, sharp focus, hard lines, stylized stencil effect
Negative prompt: blending colors, blended colors
Steps: 20, Sampler: Euler a, CFG scale: 7, experiment with these. ♥️
Description
Initial release
FAQ
Comments (8)
You get an upvote just for KMFDM.
Thank you. :)
and you get an upvote for upvoting just for kmfdm :)
Upvote for you, upvote for you, upvotes for everyone !
This is really fucking rad, never thought this could be real.
If you don't mind me asking, do you have a recommended model at all?
No problem. :) I've primarily use it with crescentwonder_v2 and yurrrlmix_v11, which are models made to generate the more realistic images. But I liked the results it gave, so I keep using them. Almost every base model I've tried works well with it, with the exception of the official SD 1.5 model. It also doesn't work well when being used with SOME other lora models at the same time for some reason.
@nodiyet419561 Alright, thanks a lot man!
amazing to see this any chance this could be made for SDXL or even better FLUX?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.