Data type, dtype, CPU tensor, GPU tensor. 32-bit floating point, torch.float32 or torch.float, torch.FloatTensor, torch.cuda.FloatTensor.
確定! 回上一頁