I’ve taken my image generation completely local. No more oversensitive nannybots!
Here’s how:
(scroll down to “Installing Forge”)
Download one of these
and place it in the folder \webui_forge_cu121_torch231\webui\models\Stable-diffusion.
NB: you’ll need some 13 gigabytes for the whole package, including a checkpoint.
Locally-generated image based on the image of Dolly Parton posted earlier.
ETA: dunno why it’s squashed. Displays properly at imgur.
You are linking to the “album”, not the individual image.
Ah, thanks. Haven’t linked there in a while.
I have a support ticket in with the Night Café peeps, asking to be allowed access to my created images so I can download some. I did point out that my suspension was likely the result of nannybot error(s); at least I saved the videos before being suspended.
I had been downloading upon creation of the old images, never occurred to me to do the same with the new stuff.
Whatever PhotoPad uses is much more accurate for generating images from scratch and does okay for using AI to infill unwanted portions of existing images or making minor alterations such as swapping clothing on SFW subjects. And each minor alteration counts toward the 50/month limit.
Here’s a PhotoPad reimagining
There’s approximately 17 kajillion SD variants on Civit.ai as well. Or you get into Flux, Qwen and all the rest of it.
I usually use Google Flow for online stuff when I’m just dinking around but do a lot of stuff locally as well.
PhotoPad doesn’t really have a nannybot, rather it simply refuses to process some prompts by claiming to have an error. Copyrighted characters seem to be no problem, as long as you’re not trying to have them do something naughty. Here’s PhotoPad’s version of Jessica Rabbit as a parade balloon.
