aclnnNLLLoss

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: Computes the negative log-likelihood loss value.
Formula:

When reduction is none,

\ell(x, y) = L = \{l_1,\dots,l_N\}^\top, \quad l_n = - w_{y_n} x_{n,y_n}, \quad w_{c} = \text{weight}[c] \cdot \mathbb{1}\{c \not= \text{ignoreIndex}\},

where $x$ is self, $y$ is target, $w$ is weight, and $N$ is the batch size. If reduction is not 'none', then:

\ell(x, y) = \begin{cases} \sum_{n=1}^N \frac{1}{\sum_{n=1}^N w_{y_n}} l_n, & \text{if reduction} = \text{`mean'}\\ \sum_{n=1}^N l_n, & \text{if reduction} = \text{`sum'} \end{cases}

totalWeight = \sum_{n=1}^N w_{y_n}

Function Prototype

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

aclnnStatus aclnnNLLLossGetWorkspaceSize(
    const aclTensor *self,
    const aclTensor *target,
    const aclTensor *weight,
    int64_t          reduction,
    int64_t          ignoreIndex,
    aclTensor       *out,
    aclTensor       *totalWeightOut,
    uint64_t        *workspaceSize,
    aclOpExecutor  **executor)

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

aclnnNLLLossGetWorkspaceSize

Parameters

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (Shape)	Non-Contiguous Tensor
self (aclTensor*)	Input	Represents the input tensor.	-	FLOAT, FLOAT16, BFLOAT16	ND	(N, C) or (C) where N indicates the batch size and C indicates the number of classes.	√
target (aclTensor*)	Input	Indicates the ground-truth labels.	When the shape of self is (N, C), the shape of target is (N). When the shape of self is (C), the shape of target is (), where each element value ranges from [0, C - 1].	INT64, UINT8, INT32	ND	-	√
weight (aclTensor*)	Input	Represents the scaling weight for each class.	-	FLOAT, FLOAT16, BFLOAT16	ND	(C)	√
reduction (int64_t)	Input	Reduction to apply to the output.	Supports 0 (none), 1 (mean), or 2 (sum). 'none' indicates no reduction is applied. 'mean' indicates the sum of the output is divided by the number of elements in the output. 'sum' indicates the output is summed.	INT64	-	-	-
ignoreIndex (int64_t)	Input	Specifies a target value that is ignored and does not affect the input gradient.	-	INT64	-	-	-
out (aclTensor*)	Output	The out in the formula.	When reduction is 0 (none) and self has a 2D shape, the out shape is (N,); otherwise, it is (1,).	FLOAT, FLOAT16, BFLOAT16	ND	-	-
totalWeightOut (aclTensor*)	Output	totalWeightOut in the formula.	The output value is valid when reduction is non-zero (non-none), with a shape of (1,).	FLOAT, FLOAT16, BFLOAT16	ND	-	-
workspaceSize (uint64_t*)	Output	Returns the workspace size to be applied for on the Device side.	-	-	-	-	-
executor (aclOpExecutor**)	Output	Returns the operator executor, which contains the operator computation flow.	-	-	-	-	-

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	self, target, weight, out, or totalWeightOut is a null pointer.
ACLNN_ERR_PARAM_INVALID	161002	The data type of self, target, or weight is not within the supported range.
		Inconsistent data types of self and weight.
		Incorrect shapes of self, weight, out, and totalWeightOut.
		The reduction value is not within the range of 0 to 2.

aclnnNLLLoss

Parameters

Parameter	Input/Output	Description
workspace	Input	Workspace memory address applied for on the Device side.
workspaceSize	Input	Workspace size applied for on the Device side, obtained from the first-phase API aclnnNLLLossGetWorkspaceSize.
executor	Input	Operator executor that 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:
- aclnnNLLLoss defaults to a non-deterministic implementation. You can enable 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_nll_loss.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 data from the host to the device 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 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, 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);
  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.
  std::vector<int64_t> selfShape = {2, 3};
  std::vector<int64_t> targetShape = {2};
  std::vector<int64_t> weightShape = {3};
  std::vector<int64_t> outShape = {2};
  std::vector<int64_t> totalWeightOutShape = {1};
  void* selfDeviceAddr = nullptr;
  void* targetDeviceAddr = nullptr;
  void* weightDeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  void* totalWeightOutDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* target = nullptr;
  aclTensor* weight = nullptr;
  aclTensor* out = nullptr;
  aclTensor* totalWeightOut = nullptr;
  std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5};
  std::vector<int32_t> targetHostData = {0, 2};
  std::vector<float> weightHostData = {1.1, 1.2, 1.3};
  std::vector<float> outHostData(2, 0);
  std::vector<float> totalWeightOutHostData(1, 0);
  int64_t reduction = 0;
  int64_t ignoreIndex = -100;
  // Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the target aclTensor.
  ret = CreateAclTensor(targetHostData, targetShape, &targetDeviceAddr, aclDataType::ACL_INT32, &target);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the weight aclTensor.
  ret = CreateAclTensor(weightHostData, weightShape, &weightDeviceAddr, aclDataType::ACL_FLOAT, &weight);
  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);
  // Create the totalWeightOut aclTensor.
  ret = CreateAclTensor(totalWeightOutHostData, totalWeightOutShape, &totalWeightOutDeviceAddr, aclDataType::ACL_FLOAT,
                        &totalWeightOut);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. Call the CANN operator library API. Replace with the specific API name.
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // Call the aclnnNLLLoss first-phase API.
  ret = aclnnNLLLossGetWorkspaceSize(self, target, weight, reduction, ignoreIndex, out, totalWeightOut, &workspaceSize,
                                     &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnNLLLossGetWorkspaceSize 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 aclnnNLLLoss second-phase API.
  ret = aclnnNLLLoss(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnNLLLoss failed. ERROR: %d\n", ret); return ret);

  // 4. (Boilerplate) Synchronously wait for task execution to finish.
  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 memory to the host. 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(self);
  aclDestroyTensor(target);
  aclDestroyTensor(weight);
  aclDestroyTensor(out);
  aclDestroyTensor(totalWeightOut);

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