[object Object]

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

• Description: Performs the Swish Gated Linear Unit (SwiGlu) activation function, which implements the SwiGlu computation of x.

• Formula:

  [object Object]

  A[object Object]i[object Object] indicates the first part of the tensor after x[object Object]i[object Object] is divided into two parts based on the specified dim, and B[object Object]i[object Object] indicates the second part of the tensor after x[object Object]i[object Object] is divided into two parts based on the specified dimension.

[object Object]

Each operator has calls. First, aclnnSwiGluGetWorkspaceSize is called to obtain the workspace size required for computation and the executor that contains the operator computation process. Then, aclnnSwiGlu is called to perform computation.

[object Object]

• Parameters:

  [object Object]
  • [object Object]Atlas inference series products[object Object]: The data type can be FLOAT16 or FLOAT32. The shape must be 64-byte aligned. Empty tensors are not supported.

• Returns:

  aclnnStatus: 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:

  aclnnStatus: status code. For details, see .

[object Object]

• Deterministic compute:
  • aclnnSwiGlu defaults to a deterministic implementation.

[object Object]

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