MCPcopy Create free account
hub / github.com/PaddlePaddle/Paddle / standard_normal

Function standard_normal

python/paddle/tensor/random.py:808–934  ·  view source on GitHub ↗

Returns a Tensor filled with random values sampled from a standard normal distribution with mean 0 and standard deviation 1, with ``shape`` and ``dtype``. Args: shape (tuple|list|Tensor): Shape of the Tensor to be created. The data type is ``int32`` or ``int64`` .

(
    shape: ShapeLike,
    dtype: DTypeLike | None = None,
    name: str | None = None,
    *,
    out: paddle.Tensor | None = None,
    device: PlaceLike | None = None,
    requires_grad: bool = False,
)

Source from the content-addressed store, hash-verified

806
807
808def standard_normal(
809 shape: ShapeLike,
810 dtype: DTypeLike | None = None,
811 name: str | None = None,
812 *,
813 out: paddle.Tensor | None = None,
814 device: PlaceLike | None = None,
815 requires_grad: bool = False,
816) -> Tensor:
817 """
818 Returns a Tensor filled with random values sampled from a standard
819 normal distribution with mean 0 and standard deviation 1, with ``shape``
820 and ``dtype``.
821
822 Args:
823 shape (tuple|list|Tensor): Shape of the Tensor to be created. The data type is ``int32`` or ``int64`` .
824 If ``shape`` is a list or tuple, each element of it should be integer or 0-D Tensor with shape [].
825 If ``shape`` is an Tensor, it should be an 1-D Tensor which represents a list.
826 dtype (str|np.dtype|paddle.dtype|None, optional): The data type of the output Tensor.
827 Supported data types: float16, bfloat16, float32, float64, complex64, complex128.
828 Default is None, use global default dtype (see ``get_default_dtype``
829 for details).
830 name (str|None, optional): Name for the operation (optional, default is None).
831 For more information, please refer to :ref:`api_guide_Name`.
832 out(Tensor, optional): The output tensor.
833 device(PlaceLike|None, optional): The desired device of returned tensor.
834 if None, uses the current device for the default tensor type (see paddle.device.set_device()).
835 device will be the CPU for CPU tensor types and the current CUDA device for CUDA tensor types. Default: None.
836 requires_grad(bool, optional): If autograd should record operations on the returned tensor. Default: False.
837
838 Returns:
839 Tensor, A Tensor filled with random values sampled from a standard
840 normal distribution with mean 0 and standard deviation 1, with
841 ``shape`` and ``dtype``.
842
843 Examples:
844 .. code-block:: pycon
845
846 >>> import paddle
847
848 >>> # doctest: +SKIP("Random output")
849 >>> # example 1: attr shape is a list which doesn't contain Tensor.
850 >>> out1 = paddle.standard_normal(shape=[2, 3])
851 >>> print(out1)
852 >>> # doctest: +SKIP("Random output")
853 Tensor(shape=[2, 3], dtype=float32, place=Place(cpu), stop_gradient=True,
854 [[-0.33719197, -0.25688133, -0.42868865],
855 [-0.27804616, -0.25058213, -0.28209466]])
856 >>> # doctest: -SKIP
857
858 >>> # example 2: attr shape is a list which contains Tensor.
859 >>> dim1 = paddle.to_tensor(2, 'int64')
860 >>> dim2 = paddle.to_tensor(3, 'int32')
861 >>> out2 = paddle.standard_normal(shape=[dim1, dim2, 2])
862 >>> print(out2)
863 >>> # doctest: +SKIP("Random output")
864 Tensor(shape=[2, 3, 2], dtype=float32, place=Place(cpu), stop_gradient=True,
865 [[[ 0.81888396, -0.64831746],

Callers 2

randnFunction · 0.85
normalFunction · 0.85

Calls 2

gaussianFunction · 0.85

Tested by

no test coverage detected