LatexBondageRealVisionNoise

I´ve trained a 1.5 model on the sample images provided from _v11 with about the same amount of images from latexperiment and reflective desire ( images mostly, with some content)! that "basic model" has been merged with realisticvision 5.1 and then I´ve merged noiseoffset in, to make it "blacker" ;)

I haven´t experimented much with it, yet, just some test images that came out quite good (as you can see in the previews) so I´d say it´s obviously more a beta then a final product

Would be nice to see some renderings of others! the overall goal was to produce the images i already made without using a rather huge amount of loras (before i needed armbinders, clothes and at least a decent lora)

I also tried to train the model to depict "swedish collar" explicitely, while it still has the ability to depict metal collar as a different more general collar! you have to try around and play around with activation words!