Swish#

Domain: ai.onnx
Since version: 24

Swish function takes one input data (Tensor ) and produces one output data (Tensor ) of the same shape, where $Swish(x) = x * sigmoid(alpha * x)$.

Inputs

X (T): Input tensor

Outputs

Y (T): Output tensor

Attributes

alpha (float): Coefficient to multiply with input before sigmoid.

Type Constraints

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

Examples#

test_cc_swish

Node:
  Swish(X) -> (Y)

Inputs:
  X: shape=(2, 3), dtype=float32
    [[-4., -1.,  0.],
     [ 1.,  2.,  4.]]

Outputs:
  Y: shape=(2, 3), dtype=float32
    [[-0.07194484, -0.26894143,  0.        ],
     [ 0.7310586 ,  1.761594  ,  3.928055  ]]

test_cc_swish_alpha

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

Inputs:
  X: shape=(2, 3), dtype=float32
    [[-4., -1.,  0.],
     [ 1.,  2.,  4.]]

Outputs:
  Y: shape=(2, 3), dtype=float32
    [[-1.3414006e-03, -1.1920292e-01,  0.0000000e+00],
     [ 8.8079703e-01,  1.9640275e+00,  3.9986587e+00]]