I have a very large stable diffusion model(~3G). I want to load it faster, so I’m thinking about multithreading/multiprocessing. But I don’t see any such param in torch.load. Am I thinking wrong?
I have a very large stable diffusion model(~3G). I want to load it faster, so I’m thinking about multithreading/multiprocessing. But I don’t see any such param in torch.load. Am I thinking wrong?