Every tensor has a data type, which is typically float32 in deep learning ... Note that we pass a constructor that modifies the tensor data type by astype .
確定！ 回上一頁
