Open patrickvonplaten opened 1 year ago
Thank you very much for informing me about this. It's very helpful for me to stay up to date with the latest updates. I have recently been experimenting with IF, and I am amazed by its excellent performance. I will continue to explore and conduct more related experiments.
:wave: from Diffusers,
Just wanted to reach out as I see that you're adding IF :heart_eyes:
We're seeing some nice speed-ups with DeepFloyd's IF on PyTorch 2 in
diffusers
and have also added 8bit T5-XXL inference which allows <10GB GPU RAM inference.Think allowing low memory use cases for IF would be super nice for this repo! Some pointers: