site banner

Friday Fun Thread for October 25, 2024

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

3
Jump in the discussion.

No email address required.

Let's say I have a folder with many pictures of clothed humans drawn by the same artist, a good GPU and literally zero experience with artificial neural networks.

What is the correct sequence of steps to train a model that:

  • draws pictures of clothed humans in the same style?
  • draws pictures of unclothed humans in the same style?
  • draws pictures of unclothed humans engaged in various forms of intercourse in the same style?

I don't know the sequence of steps, but it seems like you'd want to train a LORA, which I believe can be done using the popular Automatic1111 UI. As best as I can tell, Stable Diffusion 1.5 is still the most popular base model on which to build LORAs, but I think the tech exists for the more recent versions too. There are a decent amount of resources in the Stable Diffusion subreddit about this sort of stuff.

4chan is probably a better place to ask this, given their prodigious amount of, um, art of this kind.

Pieter Levels mentioned how much porn helped in training models kek

Hasn't image processing been using a crop from Playboy in Academic papers from the very beginning?

https://en.wikipedia.org/wiki/Lenna

lol i had no clue