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

• Description: In the training scenario, the inputs [object Object], [object Object], [object Object], [object Object], and [object Object] are required to implement [object Object] backward computation. The backward computation extracts the [object Object] sequence from [object Object] based on [object Object] in forward computation to reduce the MatMul computation workload.

• Formula: The formula for [object Object] backward computation is as follows:

  $$S = Relu(Matmul(Query, Gather(Key, Indices)))$$ $$Y = Dy*Weights$$ $$dW = Reduce(S * dy)$$ $$dQ = Matmul(ReluGrad(Y, S), Gather(Key, Indices))$$ $$dK = ScatterAdd(Matmul(ReluGrad(Y, S), Q), Indices)$$

[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 validation. The following error codes may be returned.

  [object Object]

[object Object]

• Parameters

  [object Object]

• Returns:

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

[object Object]

• Deterministic computation:

  • [object Object] defaults to non-deterministic implementation. You can call [object Object] to enable deterministic computation.

• When this API is used together with PyTorch, ensure that the CANN package versions match the PyTorch package versions.

• [object Object] can be TND or BSND.

• The following uses the BSND layout as an example to describe the restrictions on the data shape. Where:

  • [object Object] (Batchsize): The value ranges from 1 to 1024.
  • [object Object] (Head-Num): The value is [object Object].
  • [object Object] (Group): The value is [object Object].
  • [object Object] (Seq-LengthQ): The value ranges from 1 to 128K.
  • [object Object] (Seq-LengthK): The value ranges from topK to 128K.
  • [object Object] (Head-Dim): The value is [object Object].
  • [object Object]: The value is [object Object].

[object Object]

The following is an example of aclnn single-operator calling. For details about the compilation and execution process, see .