Create funny AI image fails
Experiment with AI image generation and create your own bizarre and surreal images for free!
When Stable Diffusion Works (part 1)
Created 3 years ago · 29 comments· 0 likes
Stable Diffusion 1.52x
This image follows the same principle as the previous one --> [
I passed a prompt that here doesn't work into a WORKING Stable Diffusion model named RPG40. The start image is visible, made with that, look at it.
I used noise at 10% to limit the inevitable damage that THIS DEVASTATING model of Stable Diffusion does to ANY image.
Result is up here. Bad if compared to the original, but good if compared to the usual results we get.
In Italy exists a phrase, pronounced by the fabled Uccello Gargiulo, that luckily I can't translate here. 😉
🔗.=.=.=.=. LINKS .=.=.=.=.🔗
🧚♀️ My latest visual short story, "The Faerie Expedition" -->
🖼My creations available as prints on my shop -->
🎨My favourite creations in my gallery -->
Progress Images
29 Comments
Join the conversation
Very nice work on the Stable diffusion part.... sorry You Doublecrash are awesome
Just to be clear: these Stable Diffusion models are out there and they are completely FREE. You don't have to pay, you don't have to run a local version that your computer can't handle, you don't have to do ANYTHING.
They're out there. Just as that.
Creation Settings
- Text Prompts
Doublecrash has hidden the prompt
- Model
- Initial Resolution
Thumb
- Aspect Ratio
1:1
- Runtime
Short
- Overall Prompt Weight
70%
- Noise Weight
10%
- Sampling method
K_EULER_ANCESTRAL
Wonderful !