mindspore 2.3.0rc1__cp37-cp37m-manylinux1_x86_64.whl → 2.3.0rc2__cp37-cp37m-manylinux1_x86_64.whl

mindspore/ops/operations/array_ops.py

@@ -35,12 +35,14 @@ from mindspore.common import Tensor, CSRTensor, COOTensor
 from mindspore._c_expression import Tensor as Tensor_
 from mindspore._c_expression import CSRTensor as CSRTensor_
 from mindspore._c_expression import COOTensor as COOTensor_
-from ..auto_generate import (ExpandDims, Reshape, TensorShape, Transpose, Gather, OnesLike, ZerosLike, Argmax,
-                             ReverseV2, Diag, Eye, ScatterNd, ResizeNearestNeighborV2, GatherNd, GatherD,
-                             Range, MaskedFill, RightShift, NonZero, ResizeNearestNeighbor, Identity, Split,
-                             CumSum, CumProd, Cummax, Cummin, Argmin, Concat, UnsortedSegmentSum, ScalarToTensor,
-                             BroadcastTo, StridedSlice, Select)
-from .manually_defined import Rank, Shape, Tile, Cast
+from ..auto_generate import (ExpandDims, Reshape, TensorShape, Transpose, Gather,
+                             OnesLike, ZerosLike, Argmax, ArgMaxExt,
+                             ReverseV2, Diag, Eye, ScatterNd, ResizeNearestNeighborV2,
+                             GatherNd, GatherD, Range, MaskedFill, RightShift, NonZero,
+                             ResizeNearestNeighbor, Identity, Split, CumSum, CumProd,
+                             Cummax, Cummin, Argmin, Concat, UnsortedSegmentSum, ScalarToTensor,
+                             Tril, Triu, BroadcastTo, StridedSlice, Select, TopkExt)
+from .manually_defined import Rank, Shape, Tile, Cast, Ones, Zeros
 from ..auto_generate import ArgMaxWithValue, ArgMinWithValue
 
 class _ScatterOp(PrimitiveWithInfer):
@@ -817,7 +819,7 @@ class Size(Primitive):
 
     Inputs:
         - **input_x** (Tensor) - Input parameters, the shape of tensor is :math:`(x_1, x_2, ..., x_R)`. The data type is
-          `number <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/mindspore.html#mindspore.dtype>`_.
+          `number <https://www.mindspore.cn/docs/en/master/api_python/mindspore.html#mindspore.dtype>`_.
 
     Outputs:
         int. A scalar representing the elements' size of `input_x`, tensor is the number of elements
@@ -1289,81 +1291,6 @@ class FillV2(PrimitiveWithCheck):
         return Tensor(np.full(dims, x.asnumpy()))
 
 
-class Ones(Primitive):
-    r"""
-    Creates a tensor filled with value ones.
-
-    Refer to :func:`mindspore.ops.ones` for more details.
-
-    Inputs:
-        - **shape** (Union[tuple[int], int]) - The specified shape of output tensor.
-        - **type** (:class:`mindspore.dtype`) - The specified type of output tensor.
-
-    Outputs:
-        Tensor, has the same type and shape as input shape value.
-
-    Supported Platforms:
-        ``Ascend`` ``GPU`` ``CPU``
-
-    Examples:
-        >>> import mindspore
-        >>> from mindspore import ops
-        >>> ones = ops.Ones()
-        >>> output = ones((2, 2), mindspore.float32)
-        >>> print(output)
-        [[1. 1.]
-         [1. 1.]]
-        >>> output = ones((3, 3), mindspore.float32)
-        >>> print(output)
-        [[1. 1. 1.]
-         [1. 1. 1.]
-         [1. 1. 1.]]
-    """
-
-    @prim_attr_register
-    def __init__(self):
-        """Initialize Ones"""
-
-
-class Zeros(Primitive):
-    r"""
-    Zeros will be deprecated in the future. Please use class `mindspore.ops.zeros` instead.
-
-    Creates a tensor filled with value zeros.
-
-    Creates a tensor with shape described by the first argument and
-    fills it with value zeros in type of the second argument.
-
-    Inputs:
-        - **shape** (Union[tuple[int], int]) - The specified shape of output tensor.
-        - **type** (mindspore.dtype) - The specified type of output tensor.
-
-    Outputs:
-        Tensor, has the same type and shape as input shape value.
-
-    Raises:
-        TypeError: If `shape` is neither int nor tuple.
-        TypeError: If `shape` is a tuple whose elements are not all int.
-
-    Supported Platforms:
-        Deprecated
-
-    Examples:
-        >>> import mindspore
-        >>> from mindspore import ops
-        >>> zeros = ops.Zeros()
-        >>> output = zeros((2, 2), mindspore.float32)
-        >>> print(output)
-        [[0. 0.]
-         [0. 0.]]
-
-    """
-
-    @prim_attr_register
-    def __init__(self):
-        """Initialize Zeros"""
-
-
 class TupleToArray(PrimitiveWithInfer):
     """
     Converts a tuple to a tensor.
@@ -2836,80 +2763,6 @@ class ScatterSub(Primitive):
         self.add_prim_attr('side_effect_mem', True)
 
 
-class Triu(Primitive):
-    """
-    Returns the upper triangular portion of the 2-D matrix or the set of matrices
-    in a batch. The remaining elements of the resulting Tensor are assigned a value of 0.
-    The upper triangular section of the matrix comprises of the
-    elements present on and above the main diagonal.
-
-    .. warning::
-        This is an experimental API that is subject to change or deletion.
-
-    Args:
-        diagonal (int, optional): The index of diagonal. Default: ``0`` , indicating the main diagonal.
-
-    Inputs:
-        - **x** (Tensor) - The input tensor with shape :math:`(M, N, *)`
-          where :math:`*` means any number of additional dimensions.
-
-    Outputs:
-        - **y** (Tensor) - A tensor has the same shape and data type as input.
-
-    Raises:
-        TypeError: If `x` is not a Tensor.
-        TypeError: If `diagonal` is not an int.
-        ValueError: If the dimension of `input` is less than 2.
-
-    Supported Platforms:
-        ``Ascend`` ``GPU`` ``CPU``
-
-    Examples:
-        >>> import numpy as np
-        >>> from mindspore import Tensor, ops
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> triu = ops.Triu()
-        >>> result = triu(x)
-        >>> print(result)
-        [[ 1  2  3  4]
-         [ 0  6  7  8]
-         [ 0  0 12 13]
-         [ 0  0  0 17]]
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> triu = ops.Triu(diagonal=1)
-        >>> result = triu(x)
-        >>> print(result)
-        [[ 0  2  3  4]
-         [ 0  0  7  8]
-         [ 0  0  0 13]
-         [ 0  0  0  0]]
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> triu = ops.Triu(diagonal=-1)
-        >>> result = triu(x)
-        >>> print(result)
-        [[ 1  2  3  4]
-         [ 5  6  7  8]
-         [ 0 11 12 13]
-         [ 0  0 16 17]]
-    """
-
-    @prim_attr_register
-    def __init__(self, diagonal=0):
-        """Initialize Triu"""
-        validator.check_value_type("diagonal", diagonal, [int], self.name)
-        self.diagonal = diagonal
-        self.init_prim_io_names(inputs=['x'], outputs=['y'])
-
-
 class ScatterMul(_ScatterOpDynamic):
     r"""
     Updates the value of the input tensor through the multiply operation.
@@ -5190,8 +5043,6 @@ class TensorScatterElements(Primitive):
         - **indices** (Tensor) - The index of `input_x` to do scatter operation whose data type must be int32 or
           int64. It has the same rank as `data`. And accepted range is [-s, s) where s is the size along axis.
         - **updates** (Tensor) - The tensor doing the scatter operation with `data`,
-          it has the same shape and type as `data`.
-        - **update** (Tensor) - The tensor doing the scatter operation with `data`,
           it has the same type as `data` and the same shape as `indices`.
 
     Outputs:
@@ -5560,80 +5411,6 @@ class LogSpace(Primitive):
         self.init_prim_io_names(inputs=['start', 'end'], outputs=['y'])
 
 
-class Tril(Primitive):
-    """
-    Returns the lower triangular portion of the 2-D matrix or the set of matrices
-    in a batch. The remaining elements of the resulting Tensor are assigned a value of 0.
-    The lower triangular section of the matrix comprises of the
-    elements present on and below the main diagonal.
-
-    .. warning::
-        This is an experimental API that is subject to change or deletion.
-
-    Args:
-        diagonal (int, optional): An optional attribute indicates the diagonal to consider, default: ``0`` ,
-            indicating the main diagonal.
-
-    Inputs:
-        - **x** (Tensor) - The input tensor with shape :math:`(M, N, *)`
-          where :math:`*` means any number of additional dimensions.
-
-    Outputs:
-        Tensor, the same shape and data type as the input `x`.
-
-    Raises:
-        TypeError: If `x` is not a Tensor.
-        TypeError: If `diagonal` is not an int.
-        ValueError: If the rank of `x` is less than 2.
-
-    Supported Platforms:
-        ``Ascend`` ``GPU`` ``CPU``
-
-    Examples:
-        >>> import numpy as np
-        >>> from mindspore import Tensor, ops
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> tril = ops.Tril()
-        >>> result = tril(x)
-        >>> print(result)
-        [[ 1  0  0  0]
-         [ 5  6  0  0]
-         [10 11 12  0]
-         [14 15 16 17]]
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> tril = ops.Tril(diagonal=1)
-        >>> result = tril(x)
-        >>> print(result)
-        [[ 1  2  0  0]
-         [ 5  6  7  0]
-         [10 11 12 13]
-         [14 15 16 17]]
-        >>> x = Tensor(np.array([[ 1,  2,  3,  4],
-        ...                      [ 5,  6,  7,  8],
-        ...                      [10, 11, 12, 13],
-        ...                      [14, 15, 16, 17]]))
-        >>> tril = ops.Tril(diagonal=-1)
-        >>> result = tril(x)
-        >>> print(result)
-        [[ 0  0  0  0]
-         [ 5  0  0  0]
-         [10 11  0  0]
-         [14 15 16  0]]
-    """
-
-    @prim_attr_register
-    def __init__(self, diagonal=0):
-        """Initialize Tril."""
-        self.init_prim_io_names(inputs=["x"], outputs=["y"])
-        validator.check_value_type("diagonal", diagonal, [int], self.name)
-
-
 class IndexFill(Primitive):
     """
     Fills the elements under the `dim` dimension of the input Tensor `x` with the input `value`
@@ -6338,7 +6115,8 @@ class TopK(Primitive):
           - GPU: float16, float32.
           - CPU: all numeric types.
 
-        - **k** (int) - The number of top elements to be computed along the last dimension, constant input is needed.
+        - **k** (Union(Tensor, int)) - The number of top elements to be computed along the last dimension.
+          If `k` is a Tensor, the supported dtype is int32 and it should be 0-D or 1-D with shape :math:`(1, )` .
 
     Outputs:
         A tuple consisting of `values` and `indexes`.

mindspore/ops/operations/comm_ops.py

@@ -53,14 +53,14 @@ class ReduceOp:
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with multiple devices.
 
@@ -144,14 +144,14 @@ class AllReduce(Primitive):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with 2 devices.
 
@@ -180,7 +180,7 @@ class AllReduce(Primitive):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - AllReduce
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#allreduce>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#allreduce>`_
 
     """
 
@@ -233,14 +233,14 @@ class AllGather(PrimitiveWithInfer):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with 2 devices.
 
@@ -272,7 +272,7 @@ class AllGather(PrimitiveWithInfer):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - AllGather
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#allgather>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#allgather>`_
 
     """
 
@@ -458,14 +458,14 @@ class ReduceScatter(Primitive):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with 2 devices.
 
@@ -498,7 +498,7 @@ class ReduceScatter(Primitive):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - ReduceScatter
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#reducescatter>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#reducescatter>`_
 
     """
 
@@ -600,14 +600,14 @@ class Broadcast(PrimitiveWithInfer):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with multiple devices.
 
@@ -638,7 +638,7 @@ class Broadcast(PrimitiveWithInfer):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - Broadcast
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#broadcast>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#broadcast>`_
 
     """
 
@@ -718,11 +718,11 @@ class NeighborExchange(Primitive):
         The user needs to preset
         communication environment variables before running the following example, please check the details on the
         official website of `MindSpore \
-        <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/mindspore.ops.primitive.html#communication-operator>`_.
+        <https://www.mindspore.cn/docs/en/master/api_python/mindspore.ops.primitive.html#communication-operator>`_.
 
         This operator requires a full-mesh network topology, each device has the same vlan id, and the ip & mask are
         in the same subnet, please check the `details \
-        <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#notes>`_.
+        <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#notes>`_.
 
     Args:
         send_rank_ids (list(int)): Ranks which the data is sent to.
@@ -771,7 +771,7 @@ class NeighborExchange(Primitive):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - NeighborExchange
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#neighborexchange>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#neighborexchange>`_
 
     """
 
@@ -804,7 +804,7 @@ class AlltoAll(PrimitiveWithInfer):
     Note:
         This operator requires a full-mesh network topology, each device has the same vlan id, and the ip & mask are
         in the same subnet, please check the `details \
-        <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#notes>`_.
+        <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#notes>`_.
 
     Args:
         split_count (int): On each process, divide blocks into split_count number.
@@ -835,14 +835,14 @@ class AlltoAll(PrimitiveWithInfer):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with 8 devices.
 
@@ -873,7 +873,7 @@ class AlltoAll(PrimitiveWithInfer):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - AlltoAll
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#alltoall>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#alltoall>`_
 
     """
 
@@ -921,7 +921,7 @@ class NeighborExchangeV2(Primitive):
     Note:
         This operator requires a full-mesh network topology, each device has the same vlan id, and the ip & mask are
         in the same subnet, please check the `details \
-        <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#notes>`_.
+        <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#notes>`_.
 
     Args:
         send_rank_ids (list(int)): Ranks which the data is sent to. 8 rank_ids represents 8 directions, if one
@@ -959,14 +959,14 @@ class NeighborExchangeV2(Primitive):
 
             For the Ascend devices, users need to prepare the rank table, set rank_id and device_id.
             Please see the `rank table Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/rank_table.html>`_
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/rank_table.html>`_
             for more details.
 
             For the GPU devices, users need to prepare the host file and mpi, please see the `mpirun Startup
-            <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/mpirun.html>`_ .
+            <https://www.mindspore.cn/tutorials/experts/en/master/parallel/mpirun.html>`_ .
 
             For the CPU device, users need to write a dynamic cluster startup script, please see the `Dynamic Cluster
-            Startup <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/parallel/dynamic_cluster.html>`_ .
+            Startup <https://www.mindspore.cn/tutorials/experts/en/master/parallel/dynamic_cluster.html>`_ .
 
             This example should be run with 2 devices.
 
@@ -1017,7 +1017,7 @@ class NeighborExchangeV2(Primitive):
 
     Tutorial Examples:
         - `Distributed Set Communication Primitives - NeighborExchangeV2
-          <https://www.mindspore.cn/docs/en/r2.3.q1/api_python/samples/ops/communicate_ops.html#neighborexchangev2>`_
+          <https://www.mindspore.cn/docs/en/master/api_python/samples/ops/communicate_ops.html#neighborexchangev2>`_
 
     """
 

mindspore/ops/operations/custom_ops.py

@@ -164,7 +164,7 @@ class Custom(ops.PrimitiveWithInfer):
     function if needed. Then these `Custom` objects can be directly used in neural networks.
     Detailed description and introduction of user-defined operators, including correct writing of parameters,
     please refer to `Custom Operators Tutorial
-    <https://www.mindspore.cn/tutorials/experts/en/r2.3.q1/operation/op_custom.html>`_ .
+    <https://www.mindspore.cn/tutorials/experts/en/master/operation/op_custom.html>`_ .
 
     .. warning::
         - This is an experimental API that is subject to change.
@@ -172,19 +172,11 @@ class Custom(ops.PrimitiveWithInfer):
     .. note::
         The supported platforms are determined by the input `func_type`. The supported platforms are as follows:
 
-        - "hybrid": supports ["Ascend", "GPU", "CPU"].
-        - "akg": supports ["Ascend", "GPU", "CPU"].
-        - "tbe": supports ["Ascend"].
+        - "hybrid": supports ["GPU", "CPU"].
+        - "akg": supports ["GPU", "CPU"].
         - "aot": supports ["GPU", "CPU"].
         - "pyfunc": supports ["CPU"].
         - "julia": supports ["CPU"].
-        - "aicpu": supports ["Ascend"].
-
-        If run on ge backend, use `CustomRegOp` to generate the registration information of "aicpu" and "tbe" operator,
-        use `custom_info_register` to bind the registration information to the `func` of the "tbe" operator,
-        then save the registration information of "aicpu" operator and the `func` implementation of "tbe" operator to
-        a file or separate files, keep these files in a separate directory, and set the absolute path of this directory
-        to environment variable "MS_DEV_CUSTOM_OPP_PATH" before running the network.
 
     Args:
         func (Union[function, str]):
@@ -193,9 +185,8 @@ class Custom(ops.PrimitiveWithInfer):
               computation logic of a user defined operator. The function can be one of the following:
 
               1. A AKG operator implementation function, which can use ir builder/tvm compute/hybrid grammar.
-              2. A TBE operator implementation function.
-              3. A pure python function
-              4. An kernel decorated function written by the Hybrid DSL.
+              2. A pure python function
+              3. An kernel decorated function written by the Hybrid DSL.
 
             - str: If func is of str type, then str should be a path of file along with a function name.
               This could be used when func_type is "aot" or "julia".
@@ -299,18 +290,7 @@ class Custom(ops.PrimitiveWithInfer):
 
         func_type (str): The implementation type of `func`, should be one of
 
-            [ ``"hybrid"`` , ``"akg"`` , ``"tbe"`` , ``"aot"`` , ``"pyfunc"`` , ``"julia"`` , ``"aicpu"`` ].
-
-            Each `func_type` only supports specific platforms(targets). Default: ``"hybrid"`` .
-            The supported platforms of `func_type`:
-
-            - ``"hybrid"``: supports ["Ascend", "GPU", "CPU"].
-            - ``"akg"``: supports ["Ascend", "GPU", "CPU"].
-            - ``"tbe"``: supports ["Ascend"].
-            - ``"aot"``: supports ["GPU", "CPU"].
-            - ``"pyfunc"``: supports ["CPU"].
-            - ``"julia"``: supports ["CPU"].
-            - ``"aicpu"``: supports ["Ascend"].
+            [ ``"hybrid"`` , ``"akg"`` , ``"aot"`` , ``"pyfunc"`` , ``"julia"`` ].
 
         bprop (function): The back propagation function of `func`. Default: ``None`` .
         reg_info (Union[str, dict, list, tuple]): Represents the registration information(reg info) of `func` with
@@ -343,7 +323,7 @@ class Custom(ops.PrimitiveWithInfer):
             or the attributes of `func` differs in different targets.
 
     Supported Platforms:
-        ``Ascend`` ``GPU`` ``CPU``
+        ``GPU`` ``CPU``
 
     Examples:
         >>> import numpy as np
@@ -372,68 +352,6 @@ class Custom(ops.PrimitiveWithInfer):
         >>> # the result will be a 16 * 16 tensor with all elements 2
         >>> print(output.shape)
         (16, 16)
-        >>> # Example, func_type = "tbe"
-        >>> square_with_bias_op_info = CustomRegOp() \
-        ...     .fusion_type("OPAQUE") \
-        ...     .attr("bias", "required", "float") \
-        ...     .input(0, "x") \
-        ...     .output(0, "y") \
-        ...     .dtype_format(DataType.F32_Default, DataType.F32_Default) \
-        ...     .dtype_format(DataType.F16_Default, DataType.F16_Default) \
-        ...     .target("Ascend") \
-        ...     .get_op_info()
-        >>>
-        >>> @custom_info_register(square_with_bias_op_info)
-        ... def square_with_bias(input_x, output_y, bias=0.0, kernel_name="square_with_bias"):
-        ...     import te.lang.cce
-        ...     from te import tvm
-        ...     from topi.cce import util
-        ...
-        ...     shape = input_x.get("shape")
-        ...     dtype = input_x.get("dtype").lower()
-        ...
-        ...     shape = util.shape_refine(shape)
-        ...     data = tvm.placeholder(shape, name="data", dtype=dtype)
-        ...
-        ...     with tvm.target.cce():
-        ...         res0 = te.lang.cce.vmul(data, data)
-        ...         res = te.lang.cce.vadds(res0, bias)
-        ...         sch = te.lang.cce.auto_schedule(res)
-        ...
-        ...     config = {"print_ir": False,
-        ...               "name": kernel_name,
-        ...               "tensor_list": [data, res]}
-        ...
-        ...     te.lang.cce.cce_build_code(sch, config)
-        >>>
-        >>> def test_tbe():
-        ...     square_with_bias = ops.Custom(square_with_bias, out_shape=lambda x, _: x, \
-        ...                                   out_dtype=lambda x, _: x, func_type="tbe")
-        ...     res = self.square_with_bias(input_x, 1.0)
-        ...     return res
-        >>>
-        >>> # Example, func_type = "aicpu"
-        >>> resize_bilinear_op_info = CustomRegOp("ResizeBilinear") \
-        ...     .fusion_type("OPAQUE") \
-        ...     .input(0, "input", "required") \
-        ...     .output(1, "output", "required") \
-        ...     .attr("align_corners", "required", "bool") \
-        ...     .attr("cust_aicpu", "optional", "str", "aicpu_kernels") \
-        ...     .dtype_format(DataType.F32_Default, DataType.F32_Default) \
-        ...     .dtype_format(DataType.F16_Default, DataType.F32_Default) \
-        ...     .target("Ascend") \
-        ...     .get_op_info()
-        >>>
-        >>> @custom_info_register(resize_bilinear_op_info)
-        ... def resize_bilinear_aicpu():
-        ...     return
-        >>>
-        >>> def test_aicpu(x):
-        ...     resize_bilinear_op = ops.Custom(resize_bilinear_aicpu, out_shape=[1, 1, 9, 9], \
-        ...                                     out_dtype=mstype.float32, func_type="aicpu")
-        ...     res = resize_bilinear_op(x, True, "aicpu_kernels")
-        ...     return res
-        >>>
         >>> # Example, func_type = "aot"
         >>> def test_aot(x, y, out_shapes, out_types):
         ...     program = ops.Custom("./reorganize.so:CustomReorganize", out_shapes, out_types, "aot")