After all the content I’ve posted about #AI and art, and generative AI, a friend asked me to help with a photographic project. We’ll see how far we can push #stablediffusion. Soon (?) in a gallery here in London.

@vivaldibrowser I spoke too quickly. Sadly, the issue has re-emerged. With the primary profile. And this time “Reset Download Settings to Default” didn’t fix the issue.

Not sure what changed. Maybe I downloaded an update?

Probably best to discuss over DM or email.

Two years after. https://t.co/rZ2Kq7jhGQ

One thing is clear by listening to @EMostaque AMA on Discord: in the next few years, generative #AI will revolutionize the world of entertainment forever. Not just at the production studio level. Individual “creators” (terrible term) will be able to do things never seen before.

Two of my favourite startups ever working together. Awesome. https://t.co/luosKpahMJ

Impressive 2.0 release for #stablediffusion fork kickstarted by @lincolndstein and evolved into InvokeAI. I’m trying the new web interface on a MBP M1 Pro and it’s great. I can’t wait to see in/outpainting and especially the node manager.

https://t.co/gaKI8vhg2s

RT @EMostaque: To put this in context we usually use 28-50 sampling steps for image creation, this also works for video diffusion.

Everyth…

RT @yarotrof: Russian cruise missile strikes on cities all over Ukraine, timed to maximize civilian casualties. The biggest series of salvo…

@Magnus919 5/ I’m also using a custom script for AUTOMATIC1111 WebUI that generates random prompts using GPT-2. But, like my own study generator, these are exclusively to understand how prompts are evaluated by #stablediffusion

@Magnus919 6/ One resource that I plan to use to better understand prompts is the open prompts database put together by @krea_ai. It should be possible to do some interesting data analysis on that database.