View Single Post
Brad
Selfish Heathen
 
Join Date: May 2004
Location: Zone of Pain
 
2022-09-11, 00:22

Those are some clever prompts! Let's see how my version stacks up...

Alf shaped bong only one of these looks like my boy Gordon Shumway
Courtroom sketch of Elmo testifying against Osama Bin Laden a few good ones here
Duran Duran eating durian
Ukiyo-e painting of Alfred E. Neuman not the same MAD lad I grew up with, but still interesting!
Hand jobs of the future surprisingly safe for work
LoFi Golden Corral to relax to I definitely could have had dinner as a kid in one of these in an alternate universe
Ronald McDonald giving birth in a bathtub with a midwife I guess these are NSFW
Bill Cosby caught on trail cam

Not a ton of great results across the board, but I certainly see the potential for greatness.

The model I downloaded and used is Hugging Face's Stable Diffusion v1.4 which is 4GB. That feels pretty tiny considering all it contains. This model is trained on the LAION-5B data set which contains 5.85 billion image-text pairs. That's a lot of input! It takes folks like Hugging Face a very long time and an absurd amount of computation to train a good model on this much data, but the longer they let it train, the better the results you tend to get out. Supposedly newer versions soon trained for even longer will be released in the not-too-distant future. I'll try to keep an eye out for that news and update my system when it arrives.

The quality of this board depends on the quality of the posts. The only way to guarantee thoughtful, informative discussion is to write thoughtful, informative posts. AppleNova is not a real-time chat forum. You have time to compose messages and edit them before and after posting.
  quote