There is probably less motivation, as these models are much smaller. An M1 Mac can run inference in under a minute. An GPU in as little as 3-5 seconds. This is

There is probably less motivation, as these models are much smaller. An M1 Mac can run inference in under a minute. An GPU in as little as 3-5 seconds. This is supposed to be as much as 20x faster when the distilled Stable diffusion model is released.