Open wert93 opened 3 years ago
Is there a way to improve the speed performance of the solver by doing the computation on GPU? If so, which tensors should be assigned to 'cuda'?
I find if you use your CPU with this repo, the speed performance would be improved well
Is there a way to improve the speed performance of the solver by doing the computation on GPU? If so, which tensors should be assigned to 'cuda'?