ops softmin

docs/api/api_python/mindspore.ops.rst

@@ -92,6 +92,7 @@ mindspore.ops
     mindspore.ops.selu
     mindspore.ops.soft_shrink
     mindspore.ops.softmax
+    mindspore.ops.softmin
     mindspore.ops.softsign
     mindspore.ops.tanh
 

docs/api/api_python/ops/mindspore.ops.func_softmin.rst

@@ -0,0 +1,24 @@
+mindspore.ops.softmin
+=====================
+
+.. py:function:: mindspore.ops.softmin(x, axis=-1)
+
+    在指定轴上对输入Tensor执行Softmin函数做归一化操作。假设指定轴 :math:`x` 上有切片，那么每个元素 :math:`x_i` 所对应的Softmin函数如下所示：
+
+    .. math::
+        \text{output}(x_i) = \frac{exp(-x_i)}{\sum_{j = 0}^{N-1}\exp(-x_j)},
+
+    其中 :math:`N` 代表Tensor的长度。
+
+    参数：
+        - **x** (Tensor) - Softmin的输入，任意维度的Tensor。其数据类型为float16或float32。
+        - **axis** (Union[int, tuple[int]], 可选) - 指定Softmin操作的轴。默认值：-1。
+
+    返回：
+        Tensor，数据类型和shape与 `x` 相同。
+
+    异常：
+        - **TypeError** - `axis` 不是int或者tuple。
+        - **TypeError** - `x` 的数据类型既不是float16也不是float32。
+        - **ValueError** - `axis` 是长度小于1的tuple。
+        - **ValueError** - `axis` 是一个tuple，其元素不全在[-len(x.shape), len(x.shape))范围中。

docs/api/api_python_en/mindspore.ops.rst

@@ -94,6 +94,7 @@ Activation Functions
     mindspore.ops.softsign
     mindspore.ops.soft_shrink
     mindspore.ops.softmax
+    mindspore.ops.softmin
     mindspore.ops.tanh
 
 Distance Functions

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

@@ -27,6 +27,7 @@ from mindspore.ops import operations as P
 from mindspore.ops.operations import nn_ops as NN_OPS
 from mindspore.ops.primitive import constexpr
 from mindspore.nn.cell import Cell
+from mindspore import ops
 
 __all__ = ['Softmin',
            'Softmax',
@@ -156,11 +157,10 @@ class Softmin(Cell):
     def __init__(self, axis=-1):
         """Initialize Softmin."""
         super(Softmin, self).__init__()
-        self.softmax = P.Softmax(axis)
+        self.axis = axis
 
     def construct(self, x):
-        x = -1 * x
-        return self.softmax(x)
+        return ops.function.softmin(x, self.axis)
 
 
 class Softmax2d(Cell):

mindspore/python/mindspore/ops/function/__init__.py

@@ -415,6 +415,7 @@ from .nn_func import (
     softsign,
     selu,
     softmax,
+    softmin,
     pdist,
     pad,
     prelu,

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

@@ -2135,6 +2135,45 @@ def softmax(x, axis=-1):
     return softmax_(x)
 
 
+def softmin(x, axis=-1):
+    r"""
+    Applies the Softmin operation to the input tensor on the specified axis.
+    Suppose a slice in the given axis :math:`x`, then for each element :math:`x_i`,
+    the Softmin function is shown as follows:
+
+    .. math::
+        \text{output}(x_i) = \frac{exp(-x_i)}{\sum_{j = 0}^{N-1}\exp(-x_j)},
+
+    where :math:`N` is the length of the tensor.
+
+    Args:
+        axis (Union[int, tuple[int]], optional): The axis to perform the Softmin operation. Default: -1.
+        x (Tensor): Tensor of shape :math:`(N, *)`, where :math:`*` means, any number of
+          additional dimensions, with float16 or float32 data type.
+
+    Returns:
+        Tensor, with the same type and shape as the logits.
+
+    Raises:
+        TypeError: If `axis` is not an int or a tuple.
+        TypeError: If dtype of `x` is neither float16 nor float32.
+        ValueError: If `axis` is a tuple whose length is less than 1.
+        ValueError: If `axis` is a tuple whose elements are not all in range [-len(logits.shape), len(logits.shape)).
+
+    Supported Platforms:
+        ``Ascend`` ``GPU`` ``CPU``
+
+    Examples:
+        >>> x = Tensor(np.array([-1, -2, 0, 2, 1]), mindspore.float16)
+        >>> output = ops.softmin(x)
+        >>> print(output)
+        [0.2341  0.636  0.0862  0.01165  0.03168 ]
+    """
+
+    softmax_ = _get_cache_prim(P.Softmax)(axis=axis)
+    return softmax_(-x)
+
+
 def soft_shrink(x, lambd=0.5):
     r"""
     Applies the SoftShrink function element-wise.
@@ -5155,6 +5194,7 @@ __all__ = [
     'softsign',
     'selu',
     'softmax',
+    'softmin',
     'pdist',
     'pad',
     'prelu',

tests/st/ops/test_softmin.py

@@ -0,0 +1,47 @@
+# Copyright 2022 Huawei Technologies Co., Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+import numpy as np
+import pytest
+
+import mindspore as ms
+from mindspore import Tensor, nn
+import mindspore.ops.function as F
+
+
+class NetSoftmin(nn.Cell):
+    def construct(self, x, axis=-1):
+        return F.softmin(x, axis)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_x86_cpu
+@pytest.mark.platform_arm_cpu
+@pytest.mark.platform_x86_gpu_training
+@pytest.mark.platform_arm_ascend_training
+@pytest.mark.platform_x86_ascend_training
+@pytest.mark.env_onecard
+@pytest.mark.parametrize('mode', [ms.GRAPH_MODE, ms.PYNATIVE_MODE])
+def test_net(mode):
+    """
+    Feature: test softmin op
+    Description: verify the result of softmin
+    Expectation: assertion success
+    """
+    ms.set_context(mode=mode)
+    x = Tensor(np.array([1, 2, 3, 4, 5]), ms.float32)
+    softmin = NetSoftmin()
+    output = softmin(x)
+    np_out = np.array([0.6364086, 0.23412165, 0.08612854, 0.03168492, 0.01165623])
+    assert np.allclose(output.asnumpy(), np_out)