Documentation
Download
文档中心
aclnnBatchNormElemt

aclnnBatchNormElemt

📄 View Source Code

Applicable Products

ProductSupported
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: Takes the global mean and the reciprocal of the standard deviation as operator inputs, and performs BatchNorm computation on x. This operator is an element-wise BatchNorm operation function used to normalize input data in certain specific scenarios. Compared with aclnnBatchNorm, aclnnBatchNormElemt may be adjusted for specific hardware or optimization requirements.

  • Formula:

    y=(x−E[x])Var(x)+ε∗weight+biasy = \frac{(x-E[x])}{\sqrt{Var(x)+ ε}} * weight + bias

    The relationship between standard deviation and variance is as follows:

    1S=1Var(x)+eps\frac{1}{S} = \frac{1}{\sqrt{Var(x) + eps}}

Function Prototype

Each operator is divided into a two-phase API. You must first call the "aclnnBatchNormElemtGetWorkspaceSize" API to obtain the input parameters and calculate the required workspace size based on the process, and then call the "aclnnBatchNormElemt" API to perform the computation.

aclnnStatus aclnnBatchNormElemtGetWorkspaceSize(
  const aclTensor* input,
  const aclTensor* weight,
  const aclTensor* bias,
  aclTensor*       mean,
  aclTensor*       invstd,
  double           eps,
  aclTensor*       output,
  uint64_t*        workspaceSize,
  aclOpExecutor**  executor)
aclnnStatus aclnnBatchNormElemt(
  void                *workspace,
  uint64_t             workspaceSize,
  aclOpExecutor       *executor,
  const aclrtStream    stream)

aclnnBatchNormElemtGetWorkspaceSize

  • Parameters

    Parameter Input/Output Description Instruction Data Type Data Format Shape Non-contiguous Tensor
    input (aclTensor*) Input Indicates the input for BatchNorm computation, corresponding to `x` in the formula.
    • Supports empty tensors.
    • Supported shapes and formats include: 2D (corresponding format: NC), 3D (corresponding format: NCL), 4D (corresponding format: NCHW), 5D (corresponding format: NCDHW), and 6D to 8D (corresponding format: ND, where the second dimension is fixed as the channel axis).
    		 FLOAT32, FLOAT16, BFLOAT16 NC, NCL, NCHW, NCDHW, ND 2-8 √
    weight (aclTensor*) Input Represents the weight tensor for BatchNorm computation, corresponding to `weight` in the formula.
    • Supports empty tensors.
    • The data type is consistent with that of input.
    • The shape length is equal to the length of the channel axis in the input parameter input.
    		 FLOAT32, FLOAT16, BFLOAT16 ND 1 √
    bias (aclTensor*) Input Indicates the bias tensor for BatchNorm computation, corresponding to `bias` in the formula.
    • Supports empty tensors.
    • The data type is consistent with that of `input`.
    • The shape length is equal to the length of the channel axis in the input parameter `input`.
    		 FLOAT32, FLOAT16, BFLOAT16 ND 1 √
    mean (aclTensor*) Input Indicates the mean of the input data, corresponding to `E(x)` in the formula.
    • Supports empty tensors.
    • The data type is consistent with that of `input`.
    • The shape length is equal to the length of the channel axis in the input parameter `input`.
    		 FLOAT32, FLOAT16, BFLOAT16 ND 1 √
    invstd (aclTensor*) Input. Indicates the reciprocal of the standard deviation of the input data, corresponding to the reciprocal of the square root of `Var(x) + eps` in the formula.
    • Supports empty tensors.
    • Supports scenarios where all element values are greater than 0.
    • The data type is consistent with the data type of `input`.
    • The shape length is equal to the length of the channel axis in the input parameter `input`.
    		 FLOAT32, FLOAT16, BFLOAT16 ND 1 √
    eps (double) Input Indicates the value added to the variance to prevent division by zero. Corresponds to `eps` in the formula. - - - - -
    output (aclTensor*) Output Indicates the final output result, corresponding to `y` in the formula.
    • Supports empty tensors.
    • The data type and shape are consistent with those of the input input.
    • The shape is the same as that of the input parameter input. Supported shapes and formats include: 2D (corresponding format: NC), 3D (corresponding format: NCL), 4D (corresponding format: NCHW), 5D (corresponding format: NCDHW), and 6-8D (corresponding format: ND, where the second dimension is fixed as the channel axis).
    		 FLOAT32, FLOAT16, BFLOAT16 NC, NCL, NCHW, NCDHW, ND 2-8 √
    workspaceSize (uint64_t*) Output Returns the workspace size that the user needs to apply for on the Device side. - - - - -
    executor (aclOpExecutor**) Output Returns the operator executor, which contains the operator computation process. - - - - -
    • Atlas training products, Atlas inference products: The data types of the parameters input, weight, bias, mean, invstd, and output do not support BFLOAT16.

  • 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 A pointer-type input parameter is a null pointer.
    ACLNN_ERR_PARAM_INVALID 161002 The data types and data formats of input and output are not within the supported range.
    The shapes of input and output are not within the supported range.
    An empty tensor with input C axis equal to 0 is not supported.
    The dimension of weight/bias/mean/invstd is not one-dimensional, or the shape is not equal to the length of the input C axis.
    The data formats of input and output are inconsistent.
    Inconsistent data types among input, weight, bias, mean, invstd, and output.
    Inconsistent shapes between input and output.

aclnnBatchNormElemt

  • Parameters

    Parameter Input/Output Description
    workspace Input The workspace memory address applied for on the Device side.
    workspaceSize Input The workspace size applied for on the Device side, obtained through the first-phase API aclnnBatchNormElemtGetWorkspaceSize.
    executor Input Operator executor, which contains the operator computation flow.
    stream Input Specifies the stream for task execution.

  • Return Value

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

Constraints

  • Deterministic computation:

    • aclnnBatchNormElemt defaults to a non-deterministic implementation. It supports enabling deterministic computation through aclrtCtxSetSysParamOpt.

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_batch_norm_elemt.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 shapeSize = 1;
    for (auto i : shape) {
        shapeSize *= i;
    }
    return shapeSize;
}

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 the host-side data to the 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);

    // Calculate the strides of the 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 the aclCreateTensor API.
    *tensor = aclCreateTensor(
        shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::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 the check based on your needs.
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);

    // 2. Construct the input and output. Customize the construction based on the API interface.
    std::vector<int64_t> inputShape = {2, 4, 2};
    std::vector<int64_t> meanShape = {4};
    std::vector<int64_t> invstdShape = {4};
    std::vector<int64_t> outShape = {2, 4, 2};
    double eps = 1e-2;
    void* inputDeviceAddr = nullptr;
    void* meanDeviceAddr = nullptr;
    void* invstdDeviceAddr = nullptr;
    void* outDeviceAddr = nullptr;
    aclTensor* input = nullptr;
    aclTensor* mean = nullptr;
    aclTensor* invstd = nullptr;
    aclTensor* out = nullptr;
    std::vector<float> inputHostData = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15};
    std::vector<float> meanHostData = {1, 2, 3, 4};
    std::vector<float> invstdHostData = {5, 6, 7, 8};
    std::vector<float> outHostData(16, 0);
    // Create the input aclTensor.
    ret = CreateAclTensor(inputHostData, inputShape, &inputDeviceAddr, aclDataType::ACL_FLOAT, &input);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create the mean aclTensor.
    ret = CreateAclTensor(meanHostData, meanShape, &meanDeviceAddr, aclDataType::ACL_FLOAT, &mean);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create the invstd aclTensor.
    ret = CreateAclTensor(invstdHostData, invstdShape, &invstdDeviceAddr, aclDataType::ACL_FLOAT, &invstd);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // Create the out aclTensor.
    ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT, &out);
    CHECK_RET(ret == ACL_SUCCESS, return ret);

    uint64_t workspaceSize = 0;
    aclOpExecutor* executor;

    // Example of aclnnBatchNormElemt API call.
    // 3. Call the CANN operator library API. Replace with the specific API name.
    // Call the first-phase API of aclnnBatchNormElemt.
    ret =
        aclnnBatchNormElemtGetWorkspaceSize(input, nullptr, nullptr, mean, invstd, eps, out, &workspaceSize, &executor);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchNormElemtGetWorkspaceSize 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 aclnnBatchNormElemt.
    ret = aclnnBatchNormElemt(workspaceAddr, workspaceSize, executor, stream);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchNormElemt failed. ERROR: %d\n", ret); return ret);

    // 4. (Boilerplate) Wait for the task execution to complete synchronously.
    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 memory to the Host side. Modify 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(resultData[0]),
        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 based on the specific API definition.
    aclDestroyTensor(input);
    aclDestroyTensor(mean);
    aclDestroyTensor(invstd);
    aclDestroyTensor(out);

    // 7. Release Device resources. Modify based on the specific API definition.
    aclrtFree(inputDeviceAddr);
    aclrtFree(meanDeviceAddr);
    aclrtFree(invstdDeviceAddr);
    aclrtFree(outDeviceAddr);
    if (workspaceSize > 0) {
        aclrtFree(workspaceAddr);
    }
    aclrtDestroyStream(stream);
    aclrtResetDevice(deviceId);
    aclFinalize();
    return 0;
}
This site uses cookies to improve your browsing experience. By continuing to browse this site, you agree to our use of cookies.View Details