Reverse cowgirl anal Lora. Also it can create a template for gifs.
I am not that happy with the result, but it works.
Tags
1girl - +lower weight, for one girl reverse cowgirl anal
4girls - for 4 fames with the same girl
I haven't prepared pictures for real sex, for now it's only dildos. I'll retrain it later with more pictures.
Tends to create disgusting feet, so put it (feet) to the negative prompt, or inpaint it later.
I recomend try it on 512x512 and then upscale using imgToimg.
768x768 and 1024x1024 tends to create weird sets of images, but you may find it interesting, so feel free to experiement.
Versions
v1
Trained on perfectDeliberate, photorealistic, more flexible with 1girl tag
v2
Trained on anythingv4.5, photorealistic, less flexible, but give more consistency
v3
Retrained on bigger dataset and better quality pictures. For frames picture use square canvas and 4girls, 2x2frames tags.
Use 1girl and don't use 2x2frames if you want to get one image of dildo riding.
v3r
Basically it's a different model, which is trained on sex pictures only. So no toys here. Works in the best way with my dpgd model, but usable with any other. May not work with anime models though. These one use 3x3frames (768x768) and 1frame (any resolution) tags.
My guide is still usable.
Enjoy)
Feel free to ask any questions here in comments, or in my discord channel.
Also I am making games with AI arts. They will be free in the future, but if you want to participate in making or have an early access - you can support me on patreon)
Also my wet hair Lora was banned here, I dunno why. You can download the last version of it from my patreon for free.
Description
Flexible
FAQ
Comments (4)
Did you train these on frames of video? Because there are weird artifacts in your sample pictures that look like frames of video.
@Icanhazusername it appears not in every model but yeah, you are right) I tried to retrain it using noise artifacts and lowres tag, it didn't help) So I guess I need to rebuild my dataset from scratch using better images)
Here is a simple Python script to convert a PNG image (with any number of tiles) to a GIF.
The parameters are: H tiles, W tiles, image name, FPS Example: python GenGif.py 2 2 input.png 4
Does this only work for windows? (I am a linux user). And does this take one image and generate the frames? Or does the image already have to have frames? I am asking cause its pretty difficult for me to generate the grid of frames in the first place with my models.