[object Object]

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

• Description: Performs backpropagation of . It is used to compute the gradient of the input tensor, so that the model parameters can be updated during backpropagation.
• Formula:$$gradBetaOut = \sum_{i=1}^n gradOut$$ $$gradGammaOut = \sum_{i=1}^n (gradOut \cdot \hat{x})$$ $$gradInput = mean \cdot rstd \cdot gamma \begin{bmatrix} gradOut - \frac{1}{N} (gradBetaOut + \hat{x} \cdot gradGammaOut) \end{bmatrix}$$

[object Object]

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

[object Object]

• Parameters

  [object Object]

  • [object Object]Atlas training products[object Object]:

    • The data types of [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], and [object Object] cannot be BFLOAT16.
    • The data types of [object Object] and [object Object] must be the same.

  • [object Object]Atlas A2 training products/Atlas A2 inference products[object Object] and [object Object]Atlas A3 training products/Atlas A3 inference products[object Object]:

    The data types of [object Object] and [object Object] must be the same.

  • Ascend 950PR/Ascend 950DT:

    The mapping between the data types supported by the [object Object] and [object Object] parameters is as follows:

    [object Object]undefined

• 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:
  • aclnnGroupNormBackward defaults to a non-deterministic implementation. You can call aclrtCtxSetSysParamOpt to enable deterministic computation.

[object Object]

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