mindspore.mint.nn.functional.mish

mindspore.mint.nn.functional.mish(input)[source]

Computes MISH (A Self Regularized Non-Monotonic Neural Activation Function) of input tensors element-wise.

The formula is defined as follows:

\[\text{mish}(input) = input * \tanh(softplus(\text{input}))\]

See more details in A Self Regularized Non-Monotonic Neural Activation Function.

Mish Activation Function Graph:

../../_images/Mish.png
Parameters

input (Tensor) –

The input of MISH. Supported dtypes:

  • Ascend: float16, float32.

Returns

Tensor, has the same type and shape as the input.

Raises

  • TypeError – If input is not a Tensor.

  • TypeError – If dtype of input is not float16 or float32.

Supported Platforms:

Ascend

Examples

>>> import mindspore
>>> from mindspore import Tensor, mint
>>> import numpy as np
>>> x = Tensor(np.array([[-1.1, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)
>>> output = mint.nn.functional.mish(x)
>>> print(output)
[[-3.0764845e-01 3.9974124e+00 -2.6832507e-03]
 [ 1.9439589e+00 -3.3576239e-02 8.9999990e+00]]