This LoRA model has been trained on Kohya SS based on Shexyo artstyle on WAI-NSFW-illustrious-SDXL checkpoint. Obtained graphics are really close the original art style. This LoRA model could be use for sexy anime representation of women.
Note : While the training dataset includes works created by Shexyo without direct authorization, this LoRA model's design ensures full compliance with copyright laws. This is achieved through a transformative training process that abstracts and generalizes Shexyo artistic style rather than replicating specific copyrighted works. As a result, this LoRA model is able to generate innovative and original content inspired by Shexyo artistic influences, while remaining within the boundaries of legal and ethical standards.
Description
FAQ
Comments (7)
You and I both know that Note is nothing but hogwash.
We're already hated massively, why try to justify plagiarism?
Plagiarism is the act of taking someone else’s ideas, words, work, or creations and presenting them as one’s own, without properly citing the source or acknowledging the original author.
In my case, I make sure to cite sources accurately and to give credit to the author. I do not claim ownership of the generated images. That's the role of the Note. I simply state that they were produced using a LoRA model derived from the author’s work. This is not plagiarism. It is, on the contrary, a form of appreciation and recognition of the author’s contribution.
If some people dislike me for this, or fail to see the difference between plagiarism and recognition, that is beyond my control. Their opinion does not change the fact that I do not need their approval or consent to pursue what I choose to do in my own life.
@Indecent_Studio You use terms like "Inspired" and "transformative". When, in reality, you just took their art and generalized it. With nothing substantial on your part.
Many people on this site would gladly remove their LoRA if asked, but I'm going to go on a limb here and say you wouldn't because it was done in "good faith" and within "legal boundaries".
In all honesty, I do not care to an extent, it's the verbiage.
@Saikami I am primarily interested in the technical process of creating LoRA models rather than the artistic value of the images produced with them. The images I share are meant solely to demonstrate the model’s capabilities.
The terms “inspired” and “transformative” are appropriate because the generated images are not replicas of an artist’s work. Instead, they represent variations that draw on a similar style (inspiration) while presenting different outcomes (transformations). My actual contribution lies in the training of the LoRA model .This includes image retrieval, cropping, cleaning, upscaling, tagging, and fine-tuning training parameters, but not in the artistic creation of the images themselves.
Your statement about the relationship between artists and LoRA trainers is simply incorrect. I have been training models for over three years, and in that time, no artist has ever asked me to remove my models. On the contrary, those who have contacted me have done so either to thank me for crediting them as the source of inspiration (thus giving them more visibility) or to ask for help in creating models based on their own style. In many cases, artists use these models to accelerate their workflow or to handle steps they are less interested in, such as coloring or generating multiple poses for visual novels for example.
Awesome style!
Thank you !!!!
I wish you could do the art by artist JK_arts07 ...THANKS