That's not the case with PyTorch. Our data (tensors) should be 'sent' to the GPU device in order to be executed on it. Let's create multiply 1000x1000 ...
確定! 回上一頁