Thanks for the confirmation.I installed Optimized Stable Diffusion, reputed to be lighter weight version, on my desktop (OpenSUSE Tumbleweed). To keep things tidy, I installed it in its own login.
I have a Ryzen 5 5600 with 16GB. What really limited me is that I only have an Nvidia 1650 Super with 4G of RAM. In order to be able to generate images I could not login graphically. Rather than run from the console, it was easier to run from another desktop and then use NFS/gwenview to view the results remotely.
It works. This is a full resolution result of "a man walking in a forest in dawn light" (prints available to purchase ):
View attachment 14367
I got bored after a day or two.
GUI? I had to run it from the command line:Thanks for the confirmation.
Is the GUI an all white background?
python optimizedSD/optimized_txt2img.py --prompt 'man walking in a forest in dawn light' --H 256 --W 256 --seed 27 --n_iter 2 --n_samples 10 --ddim_steps 50 --precision full
VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Cape Verde PRO [Radeon HD 7750/8740 / R7 250E]
I suspect it needs a Nvidia or CUDA capable device, but my level of knowledge is pretty minimal, I could be wrong. I bit of googling might be worth doing.Got it, thanks!
I'll install Diffusion when I have more time and let you know how it went.
I've got a Radeon GPU I installed about 7 years ago so I hope to have a good resolution-
Code:VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Cape Verde PRO [Radeon HD 7750/8740 / R7 250E]