VoltaML

[sauvage7]

Hey guys, did you take a look at:

https://github.com/VoltaML/voltaML-fast-stable-diffusion

Seems to be a valid way for speeding up generation by a lot (3x faster)... it optimizes the model using TensorRT for nvidia GPUs.