Also posted to youtube
It's all about those sequentially numbered controlnet images, getting them to line up with your prompts!
Mine just kept screaming about Gypsies and curses and was basically hysterical. I though maybe I was hallucinating so I ate it real slow and yup, screamed for a while.
Sure, take the job and die, or watch your family be.. given.. kittens and candy.. I think.
That comparison is unfair to idiots.
Definitely! Think I need to get on the discord train, this is still half witchcraft to me. I seriously got lucky with this one.
We have 'em beat by miles, we're fen to f up whole other planets soon!!
He's a method actor and he's playing Doctor Rockso in the upcoming live action Dethklok miniseries
Jason, Francis, this lil guy is going to be Sven (get it?), I'm thinking Giuseppe for him..
"Taco stands on every corner! Sex demons!!"
collective blinking "Soo... do we sign up or..."
If Ann Coulter was a receptionist.
I believe it came from https://huggingface.co/lllyasviel/ControlNet-v1-1/tree/main
Not on my pc so I'm not 100%.. but there's 'models' folder inside the controlnet extension folder, put them there.
Fuck you, Glad. Team Hefty.
Thank you so much for sharing your stuff! It's amazing, and inspiring.
Check out the description in the video, there's a few youtubers whose stuff I watched to learn how.
Basically deforum with parseq, and chatgpt to help generate four minute's worth of people.
spits
What? I'm just offering you some hot tea!
I don't know how, but it seems possible. You'd generate a noisy image, use control net on an image with depth preprocessor and those would be the pieces you could probably do it with. They'd be tricky because going through a batch and trying to adjust your eyes over and over to see which ones work would be pretty taxing.
"Freaking paparazzi, just let me get my groceries"
You sound like the surgeon who worked on my lower back.
Generated using DynavisionXL at 1344x768, Topaz & Resolve for upscaling and post-processing, generated at 15fps and interpolated to 30.
Food always seems to come out really good so I figured I'd go with the flow. It cycles through 50 different food items, randomized movement.
Deforum animation using DynavisionXL - some food for thought
StableDiffusion