Suppoese I have the following nested loops using the GPU
Data set up.
datTensor = [torch.from_numpy(i) for i in dat] deviceType = torch.device(f\'cuda: