Open Pedrodilamuerte opened 1 year ago
Try removing --no-half or --no-half-vae from cmdline args
yep it works but I get black images ^^'
yep it works but I get black images ^^'
You're on AMD hardware?
yep it works but I get black images ^^'
Remove --disable-nan-check,and try again
yep it works but I get black images ^^'
You're on AMD hardware?
No , Nvidia GPU and intel core
yep it works but I get black images ^^'
Remove --disable-nan-check,and try again
didn't work sadly ...
Thats odd you shouldn't get NaNs with Nvidia hardware, which GPU? Please check that you haven't set --precision full
I have a RTX 3060 , and no I didn't. I made a fresh install but now I get the error
OutOfMemoryError: CUDA out of memory. Tried to allocate 26.00 MiB (GPU 0; 12.00 GiB total capacity; 11.06 GiB already allocated; 0 bytes free; 11.30 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
I didn't even get it before ...
It does work now, but only at 1024X1024, not more , but that's regardless of the use of the extension ...
Try --medvram
Hello,
I'm trying to use the extension but I get this error when the refiner kicks in. Any idea ?
RuntimeError: mat1 and mat2 must have the same dtype
Best regards !