.. _op_ai_onnx_Celu:

Celu
====

- **Domain**: ``ai.onnx``
- **Since version**: 28

Continuously Differentiable Exponential Linear Units:
Perform the linear unit element-wise on the input tensor X
using formula:

.. code-block::

    max(0,x) + min(0,alpha*(exp(x/alpha)-1))


**Inputs**

- **X** (*T*): Input tensor

**Outputs**

- **Y** (*T*): Output tensor

**Attributes**

- **alpha** (*float*): The Alpha value in Celu formula which control the shape of the unit. The default value is 1.0.

**Type Constraints**

- **T**: Constrain input and output types to float tensors.
  Allowed types: tensor(bfloat16), tensor(double), tensor(float), tensor(float16).

Examples
--------

**test_cc_celu_bfloat16**

.. code-block:: text

    Node:
      Celu(X) -> (Y)
      Attributes:
        alpha = 1.0

.. code-block:: text

    Inputs:
      X: shape=(2, 3), dtype=bfloat16
        [[-3, -1, 0],
         [1, 2, 3]]

    Outputs:
      Y: shape=(2, 3), dtype=bfloat16
        [[-0.949219, -0.632812, 0],
         [1, 2, 3]]

**test_cc_celu_float16**

.. code-block:: text

    Node:
      Celu(X) -> (Y)
      Attributes:
        alpha = 2.0

.. code-block:: text

    Inputs:
      X: shape=(2, 3), dtype=float16
        [[-3., -1.,  0.],
         [ 1.,  2.,  3.]]

    Outputs:
      Y: shape=(2, 3), dtype=float16
        [[-1.554, -0.787,  0.   ],
         [ 1.   ,  2.   ,  3.   ]]

Differences with previous version (12)
--------------------------------------

**SchemaDiff**: ``Celu`` (domain ``'ai.onnx'``)

* old version: 12
* new version: 28
* breaking: no

**Type constraints:**

* changed 'T': added types: ['tensor(bfloat16)', 'tensor(double)', 'tensor(float16)']

Version History
---------------

- :doc:`Version 12 <Celu-12>`