Testing new stable diffusion model. 👀
Btw if you run Amethyst or any client that allows zapping DMS, you can try it out. DM if interested.
Discussion
This one https://stability.ai/blog/stable-diffusion-sdxl-1-announcement
Are you running it localy or linking to an online service like https://clipdrop.co/stable-diffusion
Just tested sdxl locally and added to the dvm. Takes a bit longer and here's a result with the same prompt:
Another one
Looking at these images I got an idea for a game: 'Guess the prompt' - GM generates first image, others post prompts to try to get similar outputs as the original. One could use a visual similarity metric to rate outputs relative to the original. Each post has to contribute sats — winner takes all after some condition is fulfielled 😅
A cute girl but her right hand looks a bit off…
DM this man. Haven't tried yet but it's cool for 3min thing later. Generate yet again might as well try 🤷♂️
I'm interested in this stable diffusion model.
It should now work with any client now. DM if you want to beta test :)