Elu - version 6#

This page documents version 6 of operator Elu. See Elu for the latest version (since version 22).

Domain: ai.onnx
Since version: 6

Elu takes one input data (Tensor ) and produces one output data (Tensor ) where the function f(x) = alpha * (exp(x) - 1.) for x = 0., is applied to the tensor elementwise.

Inputs

X (T): Input tensor

Outputs

Y (T): Output tensor

Attributes

alpha (float): Coefficient of ELU.

Type Constraints

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

Examples#

test_cc_elu

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

Inputs:
  X: shape=(2, 3), dtype=float32
    [[-2. , -1. , -0.5],
     [ 0.5,  1. ,  2. ]]

Outputs:
  Y: shape=(2, 3), dtype=float32
    [[-1.7293295 , -1.2642411 , -0.78693867],
     [ 0.5       ,  1.        ,  2.        ]]

test_cc_elu_bfloat16

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

Inputs:
  X: shape=(2, 3), dtype=bfloat16
    [[-2, -1, -0.5],
     [0.5, 1, 2]]

Outputs:
  Y: shape=(2, 3), dtype=bfloat16
    [[-0.863281, -0.632812, -0.392578],
     [0.5, 1, 2]]

test_cc_elu_default

Node:
  Elu(X) -> (Y)

Inputs:
  X: shape=(2, 3), dtype=float32
    [[-2. , -1. , -0.5],
     [ 0.5,  1. ,  2. ]]

Outputs:
  Y: shape=(2, 3), dtype=float32
    [[-0.86466473, -0.63212055, -0.39346933],
     [ 0.5       ,  1.        ,  2.        ]]

test_cc_elu_example

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

Inputs:
  X: shape=(3,), dtype=float32
    [-1.,  0.,  1.]

Outputs:
  Y: shape=(3,), dtype=float32
    [-1.2642411,  0.       ,  1.       ]

test_cc_elu_float16

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

Inputs:
  X: shape=(2, 3), dtype=float16
    [[-2. , -1. , -0.5],
     [ 0.5,  1. ,  2. ]]

Outputs:
  Y: shape=(2, 3), dtype=float16
    [[-1.7295, -1.265 , -0.787 ],
     [ 0.5   ,  1.    ,  2.    ]]

Differences with previous version (1)#

SchemaDiff: Elu (domain 'ai.onnx')

old version: 1
new version: 6
breaking: no