In NVIDIA's new paper on #Diffusion Models, they show how more denoisers (for each stage) and more embeddings (text, image) helps with quality!
TL;DR: If you buy more GPUs, you get correct spelling too.
deepimagination.cc/eDiffi/ #AI #ML
TL;DR: If you buy more GPUs, you get correct spelling too.
deepimagination.cc/eDiffi/ #AI #ML
With so many different labs rushing to research and deploy this kind of technology, this will quickly turn into a race for more efficiency as different providers compete on costs too.
The paper is a bit evasive on the dataset (LAION?) — I presume for legal reasons. But the good news is that it's "only" 1B text/image pairs... although they are highly filtered.
IMHO there's much more room to improve quality with the current datasets.
IMHO there's much more room to improve quality with the current datasets.
Note that in the first tweet, the "Trending On ArtStation" prompt engineering hack that's equivalent for photos is "4K DSLR"!
It still can't get fingers right though! (eDiffi is on the right, DALLE-2 in the middle, Stable on left.)
It still can't get fingers right though! (eDiffi is on the right, DALLE-2 in the middle, Stable on left.)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
