aclnnAdaLayerNorm

Applicable Products

Product	Supported
Ascend 950PR/Ascend 950DT	×
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	×
Atlas inference products	×
Atlas training products	×

Function

Description: The AdaLayerNorm operator fuses LayerNorm with downstream Add and Mul operations, and adjusts the normalization process through adaptive parameters scale and shift.
Formula:
$out = LayerNorm(x) * (1 + scale) + shift$
LayerNorm formula:
$LayerNorm(x) = {{x-E(x)}\over\sqrt {Var(x)+epsilon}} * weightOptional + biasOptional$
Where E(x) represents the mean of the input, and Var(x) represents the variance of the input.

Function Prototype

Each operator is divided into a two-phase API. The "aclnnAdaLayerNormGetWorkspaceSize" API must be called first to obtain the workspace size required for computation and the executor that encapsulates the operator computation flow. Then, the "aclnnAdaLayerNorm" API is called to perform the computation.

aclnnStatus aclnnAdaLayerNormGetWorkspaceSize(
  const aclTensor* x,
  const aclTensor* scale,
  const aclTensor* shift,
  const aclTensor* weightOptional,
  const aclTensor* biasOptional,
  double           epsilon,
  aclTensor*       out,
  uint64_t*        workspaceSize,
  aclOpExecutor**  executor)

aclnnStatus aclnnAdaLayerNorm(
  void          *workspace,
  uint64_t       workspaceSize,
  aclOpExecutor *executor,
  aclrtStream    stream)

aclnnAdaLayerNormGetWorkspaceSize

Parameters

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (shape)	Non-contiguous Tensor
x (aclTensor*)	Input	Represents the input tensor for computation. Corresponds to `x` in the formula.	Null tensor is not supported. The shape is [B, S, H], where B supports 0 to 6 dimensions.	FLOAT32, FLOAT16, BFLOAT16	ND	2-8	√
scale (aclTensor*)	Input	Represents the adaptive scaling parameter. Corresponds to `scale` in the formula.	Null tensor not supported. The data type remains consistent with that of the input parameter `x`. The shape is [B, H] or [B, 1, H], where B supports 0 to 6 dimensions. The number and size of dimensions remain consistent with B in `x`, and H maintains dimension consistency with the H dimension in `x`.	FLOAT32, FLOAT16, BFLOAT16	ND	1-8	√
shift (aclTensor*)	Input	Represents the adaptive shift parameter. Corresponds to `shift` in the formula.	Null tensor not supported. The data type remains consistent with that of the input `x`. The shape is [B, H] or [B, 1, H], where B supports 0 to 6 dimensions. The number and size of dimensions remain consistent with B in `x`, and H maintains dimension consistency with the H dimension in `x`.	FLOAT32, FLOAT16, BFLOAT16	ND	1-8	√
weightOptional (aclTensor*)	Input	Optional input parameter, representing the normalization scaling parameter. Corresponds to `weightOptional` in the corresponding formula.	Null tensor not supported. The data type remains consistent with that of the input parameter `x`. The shape is [H], where H maintains dimension consistency with the H dimension in `x`.	FLOAT32, FLOAT16, BFLOAT16	ND	1	√
biasOptional (aclTensor*)	Input	Optional input parameter, representing the normalization bias parameter. Corresponds to `biasOptional` in the corresponding formula.	Null tensor not supported. The data type remains consistent with that of the input `x`. The shape is [H], where H maintains dimension consistency with the H dimension in `x`.	FLOAT32, FLOAT16, BFLOAT16	ND	1	√
epsilon (double)	Input.	The value added to the denominator to ensure numerical stability, corresponding to `epsilon` in the formula.	-	-	-	-	-
out (aclTensor*)	Output	Represents the output tensor of the computation. Corresponds to `out` in the formula.	Null tensor not supported. The data type remains consistent with that of the input `x`. The shape remains consistent with that of `x`.	FLOAT32, FLOAT16, BFLOAT16	ND	2-8	√
workspaceSize (uint64_t*)	Output	Returns the workspace size to be applied on the Device side.	-	-	-	-	-
executor (aclOpExecutor**)	Output	Returns the operator executor, which contains the operator computation process.	-	-	-	-	-

Return Value

aclnnStatus: return code. For details, see aclnn Return Code.

The first-phase API performs input parameter validation and returns an error in the following scenarios:

Return Code	Error Code	Description
ACLNN_ERR_PARAM_NULLPTR	161001	Null pointers for x, scale, shift, or out.
ACLNN_ERR_PARAM_INVALID	161002	Data type or data format of x, scale, shift, or out not within the supported range.
		When weightOptional is not a null pointer, the data type or data format is not within the supported range.
		When biasOptional is not a null pointer, the data type or data format is not within the supported range.
		The data types of scale, shift, weightOptional, biasOptional, out, and x are not consistent.
		The shapes of x, scale, shift, weightOptional, biasOptional, and out are not consistent with those specified in Parameters.

aclnnAdaLayerNorm

Parameters

Parameter	Input/Output	Description
workspace	Input	Device-side workspace memory address.
workspaceSize	Input	Device-side workspace size, obtained from the first-phase API aclnnAdaLayerNormGetWorkspaceSize.
executor	Input	Operator executor that contains the operator computation flow.
stream	Input	Specifies the stream that executes the task.

Return Value

aclnnStatus: return code. For details, see aclnn Return Code.

Constraints

Deterministic computation:
- aclnnAdaLayerNorm defaults to a deterministic implementation.

Example

The following provides example code for reference only. For details about compilation and running, see Compile and Run Samples.

#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_ada_layer_norm.h"

#define CHECK_RET(cond, return_expr) \
  do {                               \
    if (!(cond)) {                   \
      return_expr;                   \
    }                                \
  } while (0)

#define LOG_PRINT(message, ...)     \
  do {                              \
    printf(message, ##__VA_ARGS__); \
  } while (0)

int64_t GetShapeSize(const std::vector<int64_t>& shape) {
    int64_t shape_size = 1;
    for (auto i : shape) {
        shape_size *= i;
    }
    return shape_size;
}

int Init(int32_t deviceId, aclrtStream* stream) {
    // Boilerplate for resource initialization.
    auto ret = aclInit(nullptr);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclInit failed. ERROR: %d\n", ret); return ret);
    ret = aclrtSetDevice(deviceId);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSetDevice failed. ERROR: %d\n", ret); return ret);
    ret = aclrtCreateStream(stream);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtCreateStream failed. ERROR: %d\n", ret); return ret);
    return 0;
}

template <typename T>
int CreateAclTensor(const std::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                    aclDataType dataType, aclTensor** tensor) {
    auto size = GetShapeSize(shape) * sizeof(T);
    // Call aclrtMalloc to apply for device-side memory.
    auto ret = aclrtMalloc(deviceAddr, size, ACL_MEM_MALLOC_HUGE_FIRST);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMalloc failed. ERROR: %d\n", ret); return ret);

    // Copy host-side data to device-side memory by calling aclrtMemcpy.
    ret = aclrtMemcpy(*deviceAddr, size, hostData.data(), size, ACL_MEMCPY_HOST_TO_DEVICE);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMemcpy failed. ERROR: %d\n", ret); return ret);

    // Compute the strides of a contiguous tensor.
    std::vector<int64_t> strides(shape.size(), 1);
    for (int64_t i = shape.size() - 2; i >= 0; i--) {
        strides[i] = shape[i + 1] * strides[i + 1];
    }

    // Create an aclTensor by calling aclCreateTensor.
    *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, ACL_FORMAT_ND,
                              shape.data(), shape.size(), *deviceAddr);
    return 0;
}

int main() {
    // 1. (Boilerplate) Initialize the device/stream. See the acl API manual.
    // Set deviceId based on the actual device.
    int32_t deviceId = 0;
    aclrtStream stream;
    auto ret = Init(deviceId, &stream);
    // Handle based on your own needs.
    CHECK_RET(ret == 0, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);
    // 2. Construct the input and output. Customize the construction based on the API.
    std::vector<int64_t> xShape = {2, 4, 8};
    std::vector<int64_t> scaleShape = {2, 8};
    std::vector<int64_t> shiftShape = {2, 8};
    std::vector<int64_t> weightShape = {8};
    std::vector<int64_t> biasShape = {8};
    std::vector<int64_t> outShape = {2, 4, 8};
    double epsilon = 1e-5;
    void* xDeviceAddr = nullptr;
    void* scaleDeviceAddr = nullptr;
    void* shiftDeviceAddr = nullptr;
    void* weightDeviceAddr = nullptr;
    void* biasDeviceAddr = nullptr;
    void* outDeviceAddr = nullptr;
    aclTensor* x = nullptr;
    aclTensor* scale = nullptr;
    aclTensor* shift = nullptr;
    aclTensor* weight = nullptr;
    aclTensor* bias = nullptr;
    aclTensor* out = nullptr;
    std::vector<float> xHostData(2 * 4 * 8, 1);
    std::vector<float> scaleHostData(2 * 8, 1);
    std::vector<float> shiftHostData(2 * 8, 1);
    std::vector<float> weightHostData(8, 1);
    std::vector<float> biasHostData(8, 1);
    std::vector<float> outHostData(2 * 4 * 8, 0);
    // Create the x aclTensor.
    ret = CreateAclTensor(xHostData, xShape, &xDeviceAddr, aclDataType::ACL_FLOAT, &x);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create the scale aclTensor.
    ret = CreateAclTensor(scaleHostData, scaleShape, &scaleDeviceAddr, aclDataType::ACL_FLOAT, &scale);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create the shift aclTensor.
    ret = CreateAclTensor(shiftHostData, shiftShape, &shiftDeviceAddr, aclDataType::ACL_FLOAT, &shift);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create weight aclTensor.
    ret = CreateAclTensor(weightHostData, weightShape, &weightDeviceAddr, aclDataType::ACL_FLOAT, &weight);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create bias aclTensor.
    ret = CreateAclTensor(biasHostData, biasShape, &biasDeviceAddr, aclDataType::ACL_FLOAT, &bias);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create out aclTensor.
    ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT, &out);
    CHECK_RET(ret == ACL_SUCCESS, return ret);

    // 3. Call the CANN operator library API. Replace it with the specific API.
    uint64_t workspaceSize = 0;
    aclOpExecutor* executor;
    // Call the first-phase API of aclnnAdaLayerNorm.
    ret = aclnnAdaLayerNormGetWorkspaceSize(x, scale, shift, weight, bias, epsilon, out, &workspaceSize, &executor);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAdaLayerNormGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
    // Apply for device memory based on the workspaceSize calculated by the first-phase API.
    void* workspaceAddr = nullptr;
    if (workspaceSize > 0) {
        ret = aclrtMalloc(&workspaceAddr, workspaceSize, ACL_MEM_MALLOC_HUGE_FIRST);
        CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("allocate workspace failed. ERROR: %d\n", ret); return ret;);
    }
    // Call the second-phase API of aclnnAdaLayerNorm.
    ret = aclnnAdaLayerNorm(workspaceAddr, workspaceSize, executor, stream);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAdaLayerNorm failed. ERROR: %d\n", ret); return ret);
    // 4. (Boilerplate) Synchronize and wait for task completion.
    ret = aclrtSynchronizeStream(stream);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);
    // 5. Obtain the output value and copy the result from the device side to the host side. Modify this based on the specific API definition.
    auto size = GetShapeSize(outShape);
    std::vector<float> resultData(size, 0);
    ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), outDeviceAddr, size * sizeof(float),
                      ACL_MEMCPY_DEVICE_TO_HOST);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("copy result from device to host failed. ERROR: %d\n", ret); return ret);
    for (int64_t i = 0; i < size; i++) {
        LOG_PRINT("result[%ld] is: %f\n", i, resultData[i]);
    }

    // 6. Release aclTensor and aclScalar. Modify this based on the specific API definition.
    aclDestroyTensor(x);
    aclDestroyTensor(scale);
    aclDestroyTensor(shift);
    aclDestroyTensor(weight);
    aclDestroyTensor(bias);
    aclDestroyTensor(out);

    // 7. Release device resources. Modify based on the specific API definition.
    aclrtFree(xDeviceAddr);
    aclrtFree(scaleDeviceAddr);
    aclrtFree(shiftDeviceAddr);
    aclrtFree(weightDeviceAddr);
    aclrtFree(biasDeviceAddr);
    aclrtFree(outDeviceAddr);
    if (workspaceSize > 0) {
        aclrtFree(workspaceAddr);
    }
    aclrtDestroyStream(stream);
    aclrtResetDevice(deviceId);
    aclFinalize();
    return 0;
}