Text-to-Image
Diffusers
Safetensors
StableDiffusionPipeline
stable-diffusion
Inference Endpoints

Feedback on Training

#17
by LaughterOnWater - opened

Just a thought... Many of the images being used to train stable diffusion come from getty images.
How do I know?
Prompt: japanese temple garden with blooming sakura and kami
Half of the images generated have something that looks more than vaguely like the grey getty images stamp on it.
Even when you use the negative prompt "text", some form of ghost proof stamp shows up.
Is there a way to train that out of the next version?

00832-2085360646-japanese temple garden with blooming sakura and kami.png

00829-2085360643-japanese temple garden with blooming sakura and kami.png

did you try "gettImages" in negative prompt?

did you try "gettImages" in negative prompt?

Yeah, it makes no difference. :/

Sign up or log in to comment