[object Object]

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

• Description: Performs Softmax operation for the output of [object Object] in MoE computation to obtain the [object Object] result. [object Object] is the topK result of softmax. [object Object] is the index result of the topK value, that is, the corresponding expert sequence number. [object Object] is the column value result with the same shape as [object Object], indicating the original row position corresponding to each position. If a row is finished as [object Object], the expert index can be set to the value of [object Object], that is, the size of the last axis of [object Object].
• Formula:$$softmaxOut=softmax(x,axis=-1)$$ $$yOut,expertIdxOut=topK(softmaxOut,k=k)$$ $$rowIdxRange=arange(expertIdxOut.shape[0]*expertIdxOut.shape[1])$$ $$rowIdxOut=rowIdxRange.reshape([expertIdxOut.shape[1],expertIdxOut.shape[0]]).transpose(1,0)$$

[object Object]

Each operator has calls. First, [object Object] is called to obtain the workspace size required for computation and the executor that contains the operator computation process. Then, [object Object] 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 computation:

  • [object Object] defaults to a deterministic implementation.

• Ascend 950PR/Ascend 950DT:

  • The value range of the size of the last dimension of x (that is, the number of experts) is [1, 2048].

[object Object]

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