7dc35380Add on 17 days agoHistory

aclnnInplaceNormalTensor

Applicable Products

Ascend 950PR&950DT products: support

Atlas A3 products: support

Atlas A2 products: support

Atlas 200I/500 A2 inference products: not supported

Atlas inference products: not supported

Atlas training products: not supported

Function

Fills the tensor with elements from a normal distribution parameterized by mean and std. Random numbers are drawn from a discrete normal distribution with the given mean and standard deviation std to fill the selfRef tensor. The mean is a base type float that contains the mean of the normal distribution associated with each output element. std is a base type float that contains the standard deviation of the normal distribution associated with each output element. The random number offset of this API is the sum of the offsetTensor value and offset.

Function Prototype

Each operator uses a two-phase API. You must first call "aclnnInplaceNormalTensorGetWorkspaceSize" to obtain the workspace size required for computation and the executor that includes calculation of the operator, and then call "aclnnInplaceNormalTensor" to perform the computation.

aclnnStatus aclnnInplaceNormalTensorGetWorkspaceSize(
  const aclTensor* selfRef,
  float            mean,
  float            std,
  const aclTensor* seedTensor,
  const aclTensor* offsetTensor,
  int64_t          offset,
  uint64_t*        workspaceSize,
  aclOpExecutor**  executor)

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

aclnnInplaceNormalTensorGetWorkspaceSize

Parameters:

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (shape)	Non-contiguous Tensor
selfRef	Input/Output	Input/output tensor.	If selfRef is an empty tensor, success is returned directly and no calculation is performed.	FLOAT32, FLOAT16, BFLOAT16, INT32, INT16, INT8, UINT8, INT64, BOOL, DOUBLE	ND	0-8 dimensions	√
mean	Input	Random mean. Host-side scalar.	-	FLOAT	-	-	-
std	Input	Standard deviation of random numbers. Host-side scalar. Must be greater than or equal to 0.	-	FLOAT	-	-	-
seedTensor	Input	Sets the seed value of the random number generator.	-	INT64	ND	[1]	√
offsetTensor	Input	The accumulation result with the scalar offset is used as the offset of the random number operator.	-	INT64	ND	[1]	√
offset	Input	Accumulation amount for offsetTensor. Host-side scalar.	-	INT64	-	-	-
workspaceSize	Output	Returns the Device-allocated workspace size required by the user.	-	-	-	-	-
executor	Output	Returns the op executor, which includes calculation of the operator computation flow.	-	-	-	-	-

Return value:

aclnnStatus: return status code. For details, see aclnn Return Codes.

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

Return Value	Error Code	Description
ACLNN_ERR_PARAM_NULLPTR	161001	The incoming selfRef is a null pointer.
ACLNN_ERR_PARAM_NULLPTR	161001	The incoming seedTensor is a null pointer.
ACLNN_ERR_PARAM_NULLPTR	161001	The incoming offsetTensor is a null pointer.
ACLNN_ERR_PARAM_INVALID	161002	The incoming std is less than 0.
ACLNN_ERR_PARAM_INVALID	161002	The data type of selfRef is not in the supported range.
		The data types of std and mean do not meet the API input parameter requirements.
		The shape of selfRef exceeds 8 dimensions.

aclnnInplaceNormalTensor

Parameters:

Parameter	Input/Output	Description
workspace	Input	Workspace memory address side-applied on the Device.
workspaceSize	Input	Size of the workspace side-applied on the Device, obtained by the first-phase API aclnnInplaceNormalTensorGetWorkspaceSize.
executor	Input	Op executor. It includes the calculation process of the operator.
stream	Input	Stream used to execute tasks.

Return value

aclnnStatus: Returned status code. For details, see aclnn return codes.

Constraints

Deterministic computation:
- aclnnInplaceNormal defaults to a deterministic implementation.

Ascend 950PR&950DT products: offset must be a multiple of 4.

Example

The following sample code is for reference only. For details about compilation and execution, see Compile and Run Samples.

#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_normal.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, 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 allocate 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);
  // Call aclrtMemcpy to copy host-side data to device-side memory.
  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];
  }

  // Call the aclCreateTensor API to create an aclTensor.
  *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 device/stream. Refer to the acl API manual.
  // Set deviceId based on your actual device.
  int32_t deviceId = 0;
  aclrtStream stream;
  auto ret = Init(deviceId, &stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);

  // 2. Construct the input and output. Custom construction is required based on the API.
  std::vector<int64_t> selfShape = {2, 2};
  std::vector<int64_t> seedShape = {1};
  std::vector<int64_t> offsetShape = {1};
  float mean = 2.;
  float std = 1.;
  void* seedDeviceAddr = nullptr;
  aclTensor* seed = nullptr;
  void* offsetDeviceAddr = nullptr;
  aclTensor* offset = nullptr;
  int64_t offset2 = 100;
  void* selfDeviceAddr = nullptr;
  aclTensor* selfRef = nullptr;
  std::vector<float> selfHostData = {0.0, 0.0, 0.0, 0.0};
  std::vector<int64_t> seedHostData = {0};
  std::vector<int64_t> offsetHostData = {392};

  // Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &selfRef);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the seed aclTensor.
  ret = CreateAclTensor(seedHostData, seedShape, &seedDeviceAddr, aclDataType::ACL_INT64, &seed);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the offset aclTensor.
  ret = CreateAclTensor(offsetHostData, offsetShape, &offsetDeviceAddr, aclDataType::ACL_INT64, &offset);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. Call the CANN operator library API. Change it to the specific API name.
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // Call the first-phase API of aclnnInplaceNormalTensor.
  ret = aclnnInplaceNormalTensorGetWorkspaceSize(selfRef, mean, std, seed, offset, offset2, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceNormalTensorGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
  // Allocate 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 aclnnInplaceNormalTensor.
  ret = aclnnInplaceNormalTensor(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceNormalTensor failed. ERROR: %d\n", ret); return ret);

  // 4. (boilerplate) Synchronously wait for the task execution to complete.
  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 on device-side memory to the host side. Change it based on the API definition.
  auto size = GetShapeSize(selfShape);
  std::vector<float> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), selfDeviceAddr,
                    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. Change it based on the API definition.
  aclDestroyTensor(selfRef);
  aclDestroyTensor(seed);
  aclDestroyTensor(offset);

  // 7. Release device resources. Change it based on the API definition.
  aclrtFree(selfDeviceAddr);
  aclrtFree(seedDeviceAddr);
  aclrtFree(offsetDeviceAddr);
  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();

  return 0;
}