!40703 modify the supported platforms in py files

Merge pull request !40703 from 宦晓玲/code_docs_0822

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.atan2.rst

@@ -11,7 +11,7 @@ mindspore.Tensor.atan2
     输入 `x` 和 `y` 会通过隐式数据类型转换使数据类型保持一致。如果数据类型不同，低精度的数据类型会被转换到高精度的数据类型。
 
     参数：
-        - **y** (Tensor) - 输入Tensor，shape应能在广播后与 `x` 相同，或 `x` 的shape在广播后与 `y` 相同。
+        - **y** (Tensor) - 输入Tensor。shape应能在广播后与 `x` 相同，或 `x` 的shape在广播后与 `y` 相同。
 
     返回：
         Tensor，与广播后的输入shape相同，和 `x` 数据类型相同。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.bernoulli.rst

@@ -10,15 +10,15 @@ mindspore.Tensor.bernoulli
         out_{i} \sim Bernoulli(p_{i})
 
     参数：
-        - **p** (Union[Tensor, float], 可选) - shape需要可以被广播到当前Tensor。其数据类型为float32或float64。`p` 中每个值代表输出Tensor中对应广播位置为1的概率，数值范围在0到1之间。默认值：0.5。
+        - **p** (Union[Tensor, float], 可选) - shape需要可以被广播到当前Tensor。其数据类型为float32或float64。`p` 中每个值代表输出Tensor中对应广播位置为1的概率。数值范围在0到1之间。默认值：0.5。
-        - **seed** (int, 可选) - 随机种子，用于生成随机数，数值范围是-1或正整数。默认值：-1，代表取当前时间戳。
+        - **seed** (int, 可选) - 随机种子，用于生成随机数。数值范围是-1或正整数。默认值：-1，代表取当前时间戳。
 
     返回：
         Tensor，shape和数据类型与当前Tensor相同。
 
     异常：
-        - **TypeError** - 当前Tensor的数据类型不在int8, uint8, int16, int32, int64, bool, float32和float64中。
+        - **TypeError** - 当前Tensor的数据类型不在int8、uint8、int16、int32、int64、bool、float32和float64中。
         - **TypeError** - `p` 的数据类型既不是float32也不是float64。
-        - **TypeError** - `seed` 不是int。
+        - **TypeError** - `seed` 的数据类型不是int。
-        - **ValueError** - `seed` 是负数且不为-1。
+        - **ValueError** - `p` 数值范围不在0到1之间。
-        - **ValueError** - `p` 数值范围不在0到1之间。
+        - **ValueError** - `seed` 是负数且不为-1。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.inv.rst

@@ -3,7 +3,7 @@ mindspore.Tensor.inv
 
 .. py:method:: mindspore.Tensor.inv()
 
-    计算当前Tensor的倒数。
+    逐元素计算当前Tensor的倒数。
 
     .. math::
         out_i = \frac{1}{x_{i} }

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.invert.rst

@@ -3,7 +3,7 @@ mindspore.Tensor.invert
 
 .. py:method:: mindspore.Tensor.invert()
 
-    按位翻转当前Tensor。
+    逐元素按位翻转当前Tensor。
 
     .. math::
         out_i = \sim x_{i}

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.lerp.rst

@@ -3,13 +3,13 @@ mindspore.Tensor.lerp
 
 .. py:method:: mindspore.Tensor.lerp(end, weight)
 
-    基于某个浮点数Scalar或权重Tensor的值， 计算当前Tensor和 `end` Tensor之间的线性插值。
+    基于某个浮点数Scalar或权重Tensor的值，计算当前Tensor和 `end` Tensor之间的线性插值。
 
     如果参数 `weight` 是一个Tensor，那么另两个输入的维度信息可以被广播到当前Tensor。
-    如果参数 `weight` 是一个Scalar， 那么 `end` 的维度信息可以被广播到当前Tensor。
+    如果参数 `weight` 是一个Scalar，那么 `end` 的维度信息可以被广播到当前Tensor。
 
     参数：
-        - **end** (Tensor) - 进行线性插值的Tensor结束点，其数据类型必须为float16或者float32。
+        - **end** (Tensor) - 进行线性插值的Tensor结束点。数据类型必须为float16或者float32。
         - **weight** (Union[float, Tensor]) - 线性插值公式的权重参数。当为Scalar时，其数据类型为float，当为Tensor时，其数据类型为float16或者float32。
 
     返回：

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.log1p.rst

@@ -5,6 +5,8 @@ mindspore.Tensor.log1p
 
     对当前Tensor逐元素加一后计算自然对数。
 
+    其中 `x` 表示当前Tensor。
+
     .. math::
         out_i = {log_e}(x_i + 1)
 

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.logit.rst

@@ -3,7 +3,7 @@ mindspore.Tensor.logit
 
 .. py:method:: mindspore.Tensor.logit(eps=None)
 
-    逐元素计算张量的logit值，当 eps 不是 None 时， `x` 中的元素被截断到范围[eps, 1-eps]内。
+    逐元素计算张量的logit值。当 eps 不是 None 时， `x` 中的元素被截断到范围[eps, 1-eps]内。
     当 eps 为 None 时，输入 `x` 不进行数值截断。
 
     `x` 指的当前 Tensor。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.std.rst

@@ -6,7 +6,7 @@ mindspore.Tensor.std
     计算指定维度的标准差。
     标准差是方差的算术平方根，如：:math:`std = sqrt(mean(abs(x - x.mean())**2))` 。
 
-    返回标准差。默认情况下计算展开数组的标准差，否则在指定维度上计算。
+    返回标准差，默认情况下计算展开数组的标准差，否则在指定维度上计算。
 
     .. note::
         不支持NumPy参数 `dtype` 、 `out` 和 `where` 。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.svd.rst

@@ -8,8 +8,8 @@ mindspore.Tensor.svd
     更多参考详见 :func:`mindspore.ops.svd`。
 
     参数：
-        - **full_matrices** (bool, optional) - 如果这个参数为True，则计算完整的 :math:`U` 和 :math:`V` 。否则 :math:`U` 和 :math:`V` 的shape和P有关，P是M和N的较小值, M和N是输入矩阵的行和列。默认值：False。
+        - **full_matrices** (bool, optional) - 如果这个参数为True，则计算完整的 :math:`U` 和 :math:`V` 。否则 :math:`U` 和 :math:`V` 的shape和P有关。P是M和N的较小值。M和N是输入矩阵的行和列。默认值：False。
-        - **compute_uv** (bool, optional) - 如果这个参数为True，则计算 :math:`U` 和 :math:`V` ，否则只计算 :math:`S` 。默认值：True。
+        - **compute_uv** (bool, optional) - 如果这个参数为True，则计算 :math:`U` 和 :math:`V` 。如果为false，只计算 :math:`S` 。默认值：True。
 
     返回：
         - **s** (Tensor) - 奇异值。shape为 :math:`(*, P)`。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.tan.rst

@@ -10,7 +10,7 @@ mindspore.Tensor.tan
         out_i = tan(x_i)
 
     返回：
-        Tensor。
+        Tensor，和当前输入的shape一样。
 
     异常：
         - **TypeError** - 当前输入不是Tensor。

docs/api/api_python/mindspore/Tensor/mindspore.Tensor.var.rst

@@ -7,7 +7,7 @@ mindspore.Tensor.var
 
     方差是平均值的平方偏差的平均值，即：:math:`var = mean(abs(x - x.mean())**2)` 。
 
-    返回方差值。默认情况下计算展开Tensor的方差，否则在指定维度上计算。
+    返回方差值，默认情况下计算展开Tensor的方差，否则在指定维度上计算。
 
     .. note::
         不支持NumPy参数 `dtype` 、 `out` 和 `where` 。

mindspore/python/mindspore/common/tensor.py

@@ -421,6 +421,7 @@ class Tensor(Tensor_):
         Convert numpy array to Tensor.
         If the data is not C contiguous, the data will be copied to C contiguous to construct the tensor.
         Otherwise, The tensor will be constructed using this numpy array without copy.
+
         Args:
             array (numpy.array): The input array.
 
@@ -696,7 +697,7 @@ class Tensor(Tensor_):
         r"""
         Returns arctangent of x/y element-wise.
 
-        `x` refer to self tensor.
+        `x` refers to self tensor.
 
         It returns :math:`\theta\ \in\ [-\pi, \pi]`
         such that :math:`x = r*\sin(\theta), y = r*\cos(\theta)`, where :math:`r = \sqrt{x^2 + y^2}`.
@@ -706,10 +707,11 @@ class Tensor(Tensor_):
         the relatively highest precision data type.
 
         Args:
-            y (Tensor): The input tensor. It has the same shape with `x`.
+            y (Tensor): The input tensor. It has the same shape with `x` after broadcasting,
+                or the shape of `x` is the same as `y` after broadcasting.
 
         Returns:
-            Tensor, the shape is the same as the one after broadcasting,and the data type is same as `x`.
+            Tensor, the shape is the same as the one after broadcasting, and the data type is same as `x`.
 
         Raises:
             TypeError: If `x` or `y` is not a Tensor.
@@ -1246,7 +1248,7 @@ class Tensor(Tensor_):
         """
         Does a linear interpolation of two tensors start and end based on a float or tensor weight.
 
-        If `weight` is a tensor, the shapes of two inputs need to be broadcast;
+        If `weight` is a tensor, the shapes of two inputs need to be broadcast.
         If `weight` is a float, the shapes of `end` need to be broadcast.
 
         Args:
@@ -1407,7 +1409,7 @@ class Tensor(Tensor_):
         r"""
         Computes the determinant of one or more square matrices.
 
-        `x` refer to self tensor.
+        `x` refers to self tensor.
 
         Returns:
 
@@ -1435,7 +1437,7 @@ class Tensor(Tensor_):
         r"""
         Returns the natural logarithm of one plus the input tensor element-wise.
 
-        `x` refer to self tensor.
+        `x` refers to self tensor.
 
         .. math::
             out_i = {log_e}(x_i + 1)
@@ -1465,7 +1467,7 @@ class Tensor(Tensor_):
         Calculate the logit of a tensor element-wise. When eps is not None, element in 'x' is clamped to [eps, 1-eps].
         When eps is None, input 'x' is not clamped.
 
-        `x` refer to self tensor.
+        `x` refers to self tensor.
 
         .. math::
             \begin{align}
@@ -1508,7 +1510,7 @@ class Tensor(Tensor_):
         r"""
         Computes the sign and the log of the absolute value of the determinant of one or more square matrices.
 
-        `x` refer to self tensor.
+        `x` refers to self tensor.
 
         Returns:
             Tensor, The signs of the log determinants. The shape is :math:`x\_shape[:-2]`, the dtype is same as `x`.
@@ -1601,6 +1603,8 @@ class Tensor(Tensor_):
         .. math::
             out_i = \frac{1}{x_{i} }
 
+        where `x` refers to self Tensor.
+
         Returns:
             Tensor, has the same type and shape as self Tensor.
 
@@ -1626,6 +1630,8 @@ class Tensor(Tensor_):
         .. math::
             out_i = \sim x_{i}
 
+        where `x` refers to self Tensor.
+
         Returns:
             Tensor, has the same shape as as self Tensor.
 
@@ -2191,9 +2197,6 @@ class Tensor(Tensor_):
         Returns:
             Tensor, has the same data type as input.
 
-        Supported Platforms:
-            ``Ascend`` ``GPU`` ``CPU``
-
         Raises:
             TypeError: If `order` is not string type.
             ValueError: If `order` is string type, but not 'C' or 'F'.
@@ -2203,6 +2206,9 @@ class Tensor(Tensor_):
 
             :func:`mindspore.Tensor.ravel`: Return a contiguous flattened tensor.
 
+        Supported Platforms:
+            ``Ascend`` ``GPU`` ``CPU``
+
         Examples:
             >>> import numpy as np
             >>> from mindspore import Tensor
@@ -2582,12 +2588,12 @@ class Tensor(Tensor_):
         Returns:
             Tensor.
 
-        Supported Platforms:
-            ``Ascend`` ``GPU`` ``CPU``
-
         See also:
             :func:`mindspore.Tensor.sum`: Return sum of tensor elements over a given axis.
 
+        Supported Platforms:
+            ``Ascend`` ``GPU`` ``CPU``
+
         Examples:
             >>> import numpy as np
             >>> from mindspore import Tensor
@@ -3773,12 +3779,12 @@ class Tensor(Tensor_):
         Returns:
             Tensor, the merged result.
 
-        Supported Platforms:
-            ``Ascend`` ``GPU`` ``CPU``
-
         Raises:
             ValueError: If the input tensor and any of the `choices` cannot be broadcast.
 
+        Supported Platforms:
+            ``Ascend`` ``GPU`` ``CPU``
+
         Examples:
             >>> import numpy as np
             >>> from mindspore import Tensor
@@ -4572,7 +4578,7 @@ class Tensor(Tensor_):
             Only 2-D tensor is supported for now.
 
         Returns:
-            COOTensor, a sparse representation of the original dense tensor, containing:
+            COOTensor, a sparse representation of the original dense tensor, containing the following parts.
 
             - indices (Tensor): 2-D integer tensor, indicates the positions of `values` of the dense tensor.
             - values (Tensor): 1-D tensor, indicates the non-zero values of the dense tensor.
@@ -4606,7 +4612,7 @@ class Tensor(Tensor_):
             Only 2-D tensor is supported for now.
 
         Returns:
-            CSRTensor, a sparse representation of the original dense tensor, containing:
+            CSRTensor, a sparse representation of the original dense tensor, containing the following parts.
 
             - indptr (Tensor): 1-D integer tensor, indicates the start and end point for `values` in each row.
             - indices (Tensor): 1-D integer tensor, indicates the column positions of all non-zero values of the input.
@@ -4764,7 +4770,7 @@ class Tensor(Tensor_):
             RuntimeError: If `axis` is not in the range of :math:`[-ndim, ndim-1]`.
 
         Supported Platforms:
-            ``GPU``
+            ``Ascend`` ``GPU``
 
         Examples:
             >>> import numpy as np
@@ -4861,20 +4867,20 @@ class Tensor(Tensor_):
         return tensor_operator_registry.get('diag')()(self)
 
     def xdivy(self, y):
-        """
+        r"""
         Divides self tensor by the input tensor element-wise. Returns zero when self is zero. The dtype of
         original Tensor must be one of float, complex or bool. For simplicity, denote the original Tensor by x.
 
         .. math::
 
-            out_i = x_{i} / {y_{i}}
+            out_i = x_{i}\y_{i}
 
         `x` and `y` comply with the implicit type conversion rules to make the data types consistent.
         'y' must be tensor or scalar, when y is tensor, dtypes of x and y cannot be bool at the same time,
         and the shapes of them could be broadcast. When y is scalar, the scalar can only be a constant.
 
         Args:
-            - **y** (Union[Tensor, Number, bool]) - The second input y is a Number,
+            y (Union[Tensor, number.Number, bool]): The second input y is a Number,
               or a bool when the first input x is a tensor, or a tensor whose data type is float16,
               float32, float64, complex64, complex128 or bool.
 
@@ -4964,7 +4970,7 @@ class Tensor(Tensor_):
             - On Ascend, the data type of `x` and `y` must be float16 or float32.
 
         Args:
-            - **y** (Union[Tensor, number.Number, bool]) - The `y` input is a number.Number or
+            y (Union[Tensor, number.Number, bool]): The `y` input is a number.Number or
               a bool or a tensor whose data type is number or bool.
 
         Returns:
@@ -5039,7 +5045,7 @@ class Tensor(Tensor_):
         .. warning::
             - If sorted is set to 'False', it will use the aicpu operator, the performance may be reduced.
 
-        `input_x` refer to self tensor.
+        `input_x` refers to self tensor.
 
         If the `input_x` is a one-dimensional Tensor, finds the `k` largest entries in the Tensor,
         and outputs its value and index as a Tensor. Therefore, values[`k`] is the `k` largest item in `input_x`,

mindspore/python/mindspore/nn/layer/activation.py

@@ -90,7 +90,7 @@ class CELU(Cell):
         TypeError: If the dtype of 'input_x' is neither float16 nor float32.
 
     Supported Platforms:
-        ``Ascend``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> x = Tensor(np.array([-2.0, -1.0, 1.0, 2.0]), mindspore.float32)
@@ -967,7 +967,7 @@ class Softsign(Cell):
     Refer to :func:`mindspore.ops.softsign` for more details.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> x = Tensor(np.array([0, -1, 2, 30, -30]), mindspore.float32)
@@ -1409,7 +1409,7 @@ class Mish(Cell):
     Refer to :func:`mindspore.ops.mish` for more details.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)

mindspore/python/mindspore/nn/layer/conv.py

@@ -839,9 +839,6 @@ class Conv3dTranspose(_Conv):
                 \text{kernel_size[2]} - 1 }{\text{stride[2]}} + 1} \right \rfloor \\
             \end{array}
 
-    Supported Platforms:
-        ``Ascend`` ``GPU``
-
     Raises:
         TypeError: If `in_channels`, `out_channels` or `group` is not an int.
         TypeError: If `kernel_size`, `stride`, `padding` , `dilation` or `output_padding`
@@ -854,6 +851,9 @@ class Conv3dTranspose(_Conv):
         ValueError: If `pad_mode` is not equal to 'pad' and `padding` is not equal to (0, 0, 0, 0, 0, 0).
         ValueError: If `data_format` is not 'NCDHW'.
 
+    Supported Platforms:
+        ``Ascend`` ``GPU`` ``CPU``
+
     Examples:
         >>> x = Tensor(np.ones([32, 16, 10, 32, 32]), mindspore.float32)
         >>> conv3d_transpose = nn.Conv3dTranspose(in_channels=16, out_channels=3, kernel_size=(4, 6, 2),

mindspore/python/mindspore/ops/function/math_func.py

@@ -409,7 +409,7 @@ def argmin(x, axis=-1):
         TypeError: If `axis` is not an int.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> input_x = Tensor(np.array([2.0, 3.1, 1.2]), mindspore.float32)
@@ -1998,7 +1998,7 @@ def bessel_i0e(x):
         TypeError: If dtype of `x` is not float16, float32 or float64.
 
     Supported Platforms:
-        ``CPU``
+        ``GPU`` ``CPU``
 
     Examples:
         >>> x = Tensor(np.array([-1, -0.5, 0.5, 1]), mindspore.float32)
@@ -2405,7 +2405,7 @@ def trunc(input_x):
         TypeError: If `input_x` is not a Tensor.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``CPU``
 
     Examples:
         >>> input_x = Tensor(np.array([3.4742, 0.5466, -0.8008, -3.9079]),mindspore.float32)

mindspore/python/mindspore/ops/function/nn_func.py

@@ -851,7 +851,7 @@ def softsign(x):
         TypeError: If dtype of `x` is neither float16 nor float32.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> from mindspore.ops import functional as F
@@ -1729,7 +1729,7 @@ def mish(x):
         TypeError: If dtype of `x` is neither float16 nor float32.
 
     Supported Platforms:
-        ``Ascend`` ``CPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> input_x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)

mindspore/python/mindspore/ops/operations/array_ops.py

@@ -588,8 +588,8 @@ class Im2Col(Primitive):
             Default: 0.
 
     Inputs:
-        - **x** (Tensor) : input tensor, only 4-D input tensors (batched image-like tensors) are supported.
+        - **x** (Tensor) - input tensor, only 4-D input tensors (batched image-like tensors) are supported.
-            support all real number data type.
+          support all real number data type.
 
     Outputs:
         Tensor, a 4-D Tensor with same type of input `x`.
@@ -3348,7 +3348,7 @@ class StridedSlice(PrimitiveWithInfer):
     `x[2:,...]` is equivalent to `x[2:5,:,:,:]`.
 
     If the ith bit of `new_axis_mask` is set, `begin`, `end` and `strides` are ignored and a new length 1
-    dimension is added at the specified position in tthe output tensor.
+    dimension is added at the specified position in the output tensor.
 
     As for a 5*6*7 tensor, `x[:2, newaxis, :6]` will produce a tensor with shape :math:`(2, 1, 7)` .
 
@@ -6460,7 +6460,7 @@ class SearchSorted(PrimitiveWithInfer):
 
     Inputs:
         - **sequence** (Tensor) - The shape of tensor is :math:`(x_1, x_2, ..., x_R-1, x_R)` or `(x_1)`.
-                                  It must contain monitonically increasing sequence on the innermost dimension.
+          It must contain monitonically increasing sequence on the innermost dimension.
         - **values** (Tensor) - The shape of tensor is : math:`(x_1, x_2, ..., x_R-1, x_S)`.
 
     Outputs:
@@ -8120,6 +8120,9 @@ class PopulationCount(Primitive):
     Computes element-wise population count(a.k.a bitsum, bitcount).
 
     Refer to :func:`mindspore.ops.population_count` for more detail.
+
+    Supported Platforms:
+        ``Ascend`` ``GPU`` ``CPU``
     """
 
     @prim_attr_register

mindspore/python/mindspore/ops/operations/math_ops.py

@@ -1859,6 +1859,9 @@ class InplaceAdd(PrimitiveWithInfer):
 
     Refer to :func:`mindspore.ops.inplace_add` for more detail.
 
+    Supported Platforms:
+        ``Ascend`` ``CPU``
+
     Examples:
         >>> import numpy as np
         >>> import mindspore
@@ -1912,6 +1915,9 @@ class InplaceSub(PrimitiveWithInfer):
 
     Refer to :func:`mindspore.ops.inplace_sub` for more detail.
 
+    Supported Platforms:
+        ``Ascend`` ``CPU``
+
     Examples:
         >>> import numpy as np
         >>> import mindspore
@@ -4923,7 +4929,7 @@ class BesselI0e(Primitive):
         TypeError: If dtype of `x` is not float16, float32 or float64.
 
     Supported Platforms:
-        ``Ascend`` ``CPU`` ``GPU``
+        ``GPU`` ``CPU``
 
     Examples:
         >>> bessel_i0e = ops.BesselI0e()
@@ -4963,7 +4969,7 @@ class BesselI1e(Primitive):
         TypeError: If dtype of `x` is not float16, float32 or float64.
 
     Supported Platforms:
-        ``Ascend`` ``CPU`` ``GPU``
+        ``CPU``
 
     Examples:
         >>> bessel_i1e = ops.BesselI1e()
@@ -5773,6 +5779,9 @@ class Trunc(Primitive):
     Returns a new tensor with the truncated integer values of the elements of input.
 
     Refer to :func:`mindspore.ops.trunc` for more detail.
+
+    Supported Platforms:
+        ``CPU``
     """
 
     @prim_attr_register

mindspore/python/mindspore/ops/operations/nn_ops.py

@@ -7253,6 +7253,9 @@ class CTCGreedyDecoder(Primitive):
     Performs greedy decoding on the logits given in inputs.
 
     Refer to :func:`mindspore.ops.ctc_greedy_decoder` for more detail.
+
+    Supported Platforms:
+        ``Ascend`` ``CPU``
     """
 
     @prim_attr_register
@@ -8585,7 +8588,7 @@ class Conv3DTranspose(Primitive):
         ValueError: If bias is not none. The rank of dout and weight is not 5.
 
     Supported Platforms:
-        ``Ascend`` ``GPU``
+        ``Ascend`` ``GPU`` ``CPU``
 
     Examples:
         >>> dout = Tensor(np.ones([32, 16, 10, 32, 32]), mindspore.float16)