aclnnEmbeddingRenorm

📄 View Source Code

Applicable Products

Function

• Description: Returns the corrected result of the input tensor at the specified indices based on the given maxNorm and normType.

• Formula: The norm of a vector is calculated as follows, where p is the norm value specified by normType:

  $$||X||_{p}=\sqrt[p]{\sum_{i=1}^nx_{i}^p}$$ $$where X=(x_{1}, x_{2}, ... , x_{n})$$

  For scenarios where the computed norm exceeds maxNorm, normalization is required: the 0-dimension elements specified by indices are multiplied by a coefficient:

  $$scalar = \frac{maxNorm}{currentNorm+1e^{-7}}$$

Function Prototype

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

aclnnStatus aclnnEmbeddingRenormGetWorkspaceSize(
  aclTensor       *selfRef,
  const aclTensor *indices,
  double           maxNorm,
  double           normType,
  uint64_t        *workspaceSize,
  aclOpExecutor  **executor)

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

aclnnEmbeddingRenormGetWorkspaceSize

• Parameters:

  Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (Shape)	Non-contiguous Tensor
  selfRef (aclTensor*)	Input	Specifies the input for the renormalization computation, corresponding to x in the formula.	-	FLOAT32, FLOAT16, BFLOAT16	ND	2	√
  indices (aclTensor*)	Input	Indices in dimension 0 of selfRef for renormalization.	Index data in indices must not be out of bounds.	INT32, INT64	ND	0-8	√
  maxNorm (double)	Input	Specifies the maximum value of the norm. If the norm exceeds this value, the embedding result needs to be renormalized.	-	-	-	-	-
  normType (double)	Input	Specifies the type of L_P norm, i.e., p in the formula.	-	-	-	-	-
  workspaceSize (uint64_t*)	Output	Returns the workspace size to be allocated on the Device side.	-	-	-	-	-
  executor (aclOpExecutor**)	Output	Returns the operator executor, which contains the operator computation flow.	-	-	-	-	-

  • Atlas inference products, Atlas training products: The BFLOAT16 data type is not supported.

• 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 Value	Error Code	Description
  ACLNN_ERR_PARAM_NULLPTR	161001	The passed selfRef or indices is a null pointer.
  ACLNN_ERR_PARAM_INVALID	161002	The data type and data format of selfRef and indices are not within the supported range.
  		The data type of maxNorm and normType is not within the supported range.
  		The shape dimension of selfRef is not 2.
  		The shape dimension of indices exceeds 8 dimensions.

aclnnEmbeddingRenorm

• Parameters:

  Parameter	Input/Output	Description
  workspace	Input	Address of the workspace memory applied for on the Device side.
  workspaceSize	Input	Size of the workspace applied for on the Device side, obtained through the first-phase API aclnnEmbeddingRenormGetWorkspaceSize.
  executor	Input	Operator executor, which 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:

  • aclnnEmbeddingRenorm 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_embedding_renorm.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. Initialize resources.
  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 memory using 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 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() {
  int32_t deviceId = 0;
  aclrtStream stream;
  auto ret = Init(deviceId, &stream);
  CHECK_RET(ret == 0, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);
  std::vector<int64_t> selfShape = {4, 2};
  std::vector<int64_t> indicesShape = {4, 2};
  void* selfDeviceAddr = nullptr;
  void* indicesDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* indices = nullptr;
  std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
  std::vector<int> indicesHostData = {1, 1, 1, 1, 0, 0, 0, 0};
  float normType = 1.0f;
  float maxNorm = 2.0f;
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  ret = CreateAclTensor(indicesHostData, indicesShape, &indicesDeviceAddr, aclDataType::ACL_INT32, &indices);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  ret = aclnnEmbeddingRenormGetWorkspaceSize(self, indices, maxNorm, normType, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnEmbeddingRenormGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);

  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);
  }

  ret = aclnnEmbeddingRenorm(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnEmbeddingRenorm failed. ERROR: %d\n", ret); return ret);

  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  auto size = GetShapeSize(selfShape);
  std::vector<float> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), selfDeviceAddr, 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]);
  }

  aclDestroyTensor(self);
  aclDestroyTensor(indices);
  aclrtFree(selfDeviceAddr);
  aclrtFree(indicesDeviceAddr);
  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();
  return 0;
}