[object Object]

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

• Description: Gaussian error linear unit (GELU) activation gate function. Based on aclnnGeGlu, the direction of the data block on which the activation function is performed can be set.
• Formula: If activateLeft is set to true, activation is performed on the left half of $self$.$$out_{i}=GeGlu(self_{i}) = Gelu(A) \cdot B$$ If activateLeft is set to false, activation is performed on the right half of $self$.$$out_{i}=GeGlu(self_{i}) = A \cdot Gelu(B)$$ $A$ indicates the left half of $self$, and $B$ indicates the right half of $self$.

[object Object]

Each operator has calls. First, aclnnGeGluV3GetWorkspaceSize is called to obtain the workspace size required for computation and the executor that contains the operator computation process. Then, aclnnGeGluV3 is called to perform computation.

[object Object]

• Parameters:

  [object Object]
  • [object Object]Atlas inference products[object Object] and [object Object]Atlas training products[object Object]: The data type can be FLOAT or FLOAT16.

• 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 computation:
  • aclnnGeGluV3 defaults to a deterministic implementation.

[object Object]

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