still no word on how to convert/train other finetuned models into their format :(

In case you get stuck again and need more games:

• DevilutionX (free, open source, needs gamedata) lets you play Diablo1 on Android, very good time killer (you might need to fetch the gamedata somewhere)
• Out There: Omega (paid but one time purchase) is a relaxed starship roguelite
• Battle For Wesnoth (free, open source) fantasy style tactical game
• Jagged Alliance 2 Stracciatella (free, open source, needs gamedata) - Jagged Alliance 2 on Android, tactical RPG, great timekiller like classic UFO or the old Fallout games.

Notable mentions: WorldOfGoo, Human Resource Machine

Just install ComfyUI and start it with the --cpu flag. Ensure you have enough system RAM and a swap partition (preferably on nvme/ssd).

SSDs are not really good for long lasting backups. They hold data by electric charge, if you unplug your SSD and store it, then it might loose its data after just a couple of years. HDD “spinning rust” still has its merits when it comes to long term data storage, they hold their magnetic data longer without fresh power.

locally, always.

I even got it to run without GPU on a pure old i5 CPU with 8GB system RAM (not VRAM) paired with 32GB swap. SD1.5 takes 4-10 minutes per image, SDXL about 2 hours. But it works. With GPU its between 7 and 90 seconds per image, depending on model and settings.

I first tried it a few days ago, I’m still a bit lost. Inpainting, which is the major part of my workflow, feels not as swift as in automatic1111 and I’m still searching for the only-masked-area inpainting in ComfyUI.

But I can confirm it is much faster and uses less VRAM. And I somehow love the ability to save the entire workflow into a json. I’m missing my prompt-autocomplete plugin the most.

Yes, I tested it and although it works in its current state it takes 2-3 hours per picture on Pi and 20 minutes per picture on my Desktop CPU.

But…isn’t unsupervised backfeeding the same as simply overtraining the same dataset? We already know overtraining causes broken models.

Besides, the next AI models will be fed with the interactions from humans with AI, not just it’s own content. ChatGPT already works like this, it learns with every interaction, every chat.

And the generative image models will be fed with AI-assisted images where humans will have fixed flaws like anatomy (the famous hands) or other glitches.

So as interesting as this is, as long as humans interact with AI the hybrid output used for training will contain enough new “input” to keep the models on track. There are already refined image generators trained with their own but human-assisted output that are better than their predecessor.

smoked bell pepper powder, chilli powder, MSG, salt and maybe some powdered onion/garlic.

tortilla flavoured popcorn, low on calories if made with a hot air popcorn machine.

Over at lemmy.dbzer0.com there is a StableDiffusion community that is a bit more active. !stable_diffusion@lemmy.dbzer0.com

May I suggest to get a foldable potty instead? They are super light and the size of an original Gameboy when folded. Combined with plastic bags and silicapowder (do not eat) it will help you stay healthy instead of dehydrated and constipated. If amazon is not fast enough, check the nearest baby mart. If you are just shy, there are also foldable privacy shields (camping section) available to let you do your business with privacy.