Function create_tensor

python/paddle/tensor/creation.py:289–330

Create a variable, which will hold a Tensor with data type dtype. Args: dtype(str|paddle.dtype|np.dtype, optional): the data type of Tensor to be created, the data type is bool, float16, float32, float64, int8, int16, int32 and int64. name(string, optional): The

(
    dtype: DTypeLike, name: str | None = None, persistable: bool = False
)

def create_tensor(
    dtype: DTypeLike, name: str | None = None, persistable: bool = False
) -> paddle.Tensor:
    """
    Create a variable, which will hold a Tensor with data type dtype.

    Args:
        dtype(str|paddle.dtype|np.dtype, optional): the data type of Tensor to be created, the
            data type is bool, float16, float32, float64, int8, int16, int32 and int64.
        name(string, optional): The default value is None.  Normally there is no need for
            user to set this property.  For more information, please refer to :ref:`api_guide_Name`
        persistable(bool): Set the persistable flag of the create tensor.
            default value is False.

    Returns:
        Variable: The tensor to be created according to dtype.

    Examples:
        .. code-block:: pycon

            >>> import paddle
            >>> tensor = paddle.tensor.create_tensor(dtype='float32')
    """
    check_dtype(
        dtype,
        'dtype',
        [
            'bool',
            'float16',
            'float32',
            'float64',
            'int8',
            'int32',
            'int32',
            'int64',
        ],
        'create_tensor',
    )
    helper = LayerHelper("create_tensor", **locals())
    return helper.create_variable(
        name=helper.name, dtype=dtype, persistable=persistable
    )


@param_two_alias(["stop", "end"], ["num", "steps"])