Bill Bowles
Bill Bowles
Same story here. Running on M1 Ultra with 128 GB Ram. Python quits every time I run reference_only or reference_adain+attn. Curiously, reference_adain works just fine.
as @shawnington pointed out, it does work when "Upcast cross attention layer to float32" is switched off in Stable Diffusion settings. I did that, used it for awhile, and then...
Same problem here. Reference only within ControlNet. Crashes Python every time and gives above error.
@jwooldridge234 I have the same error that you had last week. UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown warnings.warn('resource_tracker: There appear to...
Same problem here. Looks like we need to follow the above instructions: "please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU." I've also found instructions elsewhere advising:...
I just updated to Torch 2.0.1. Still have the same error message. "AssertionError: Torch not compiled with CUDA enabled"
I'm running on Apple M1, and SD-CN worked well for me 8 days ago before the most recent update. There's clearly a solution, as we already had it in the...
I'm still having this problem. Sure, uninstalling SD-CN does fix the bug for the rest of SD. But I would love to actually use SD-CN to it's full potential. And...