[object Object]

[object Object][object Object]undefined[object Object]

• Description:

  The [object Object] function is a variant of the standard ReLU function. It is mainly used to square the output of the ReLU function and is often used as the activation function of a model.

• Formula:

  $$y_i=(ReLU(x_i))^2$$

  The formula of ReLU is as follows:

  $$ReLU(x_i)=max(0,x_i)$$

  x is the input tensor, y is the output tensor, and i is the index of an element in the tensor.

[object Object]

Each operator has calls. First, aclnnSquaredReluGetWorkspaceSize is called to obtain the input parameters and compute the required workspace size based on the process. Then, aclnnSquaredRelu is called to perform computation.

[object Object]

• Parameters:

  [object Object]

• Returns:

  [object Object]: status code. For details, see .

  The first-phase API implements input parameter verification. The following errors may be thrown:

  [object Object]

[object Object]

• Parameters:

  [object Object]

• Returns:

  [object Object]: status code. For details, see .

[object Object]

• Deterministic compute:

  • aclnnSquaredRelu defaults to a deterministic implementation.

• If the input is NaN, the output is also NaN. If the input is Inf, the output is also Inf.

• If the input is -Inf, the output is 0.

• The input shape supports only one to eight dimensions. If the input shape is out of the range, an error is reported.

[object Object]

The following example is for reference only. For details, see .