aclnnAddRelu&aclnnInplaceAddRelu
Applicable Products
| Product | Supported |
|---|---|
| √ | |
| √ | |
| √ | |
| × | |
| × | |
| √ |
Function
Description: Performs activation on the result after completing the addition computation.
Formula:
Function Prototype
aclnnAddRelu and aclnnInplaceAddRelu implement the same functionality. The differences are as follows. Select the appropriate operator based on your actual scenario.
aclnnAddRelu: Requires creating a new output tensor object to store the computation result.
aclnnInplaceAddRelu: Does not require creating a new output tensor object. The computation result is stored directly in the memory of the input tensor.
Each operator is divided into a two-phase API. You must first call the "aclnnAddReluGetWorkspaceSize" or "aclnnInplaceAddReluGetWorkspaceSize" API to obtain the workspace size required for computation and the executor that includes the operator computation process, and then call the "aclnnAddRelu" or "aclnnInplaceAddRelu" API to perform the computation.
aclnnStatus aclnnAddReluGetWorkspaceSize( const aclTensor *self, const aclTensor *other, aclScalar *alpha, aclTensor *out, uint64_t *workspaceSize, aclOpExecutor **executor)aclnnStatus aclnnAddRelu( void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)aclnnStatus aclnnInplaceAddReluGetWorkspaceSize( aclTensor *selfRef, const aclTensor *other, aclScalar *alpha, uint64_t *workspaceSize, aclOpExecutor **executor)aclnnStatus aclnnInplaceAddRelu( void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)
aclnnAddReluGetWorkspaceSize
Parameters:
Parameter Input/Output Description Instruction Data Type Data Format Dimension (shape) Non-contiguous Tensor self (aclTensor*) Input Input self in the formula, representing the target tensor to be transformed. BFLOAT16, FLOAT16, FLOAT32, INT8, UINT8, INT16, INT32, INT64 ND 0-8 √ other (aclTensor*) Input Input other in the formula. BFLOAT16, FLOAT16, FLOAT32, INT8, UINT8, INT16, INT32, INT64 ND 0-8 √ alpha (aclScalar*) Input alpha in the formula. The data type needs to be convertible to the derived data type of self and other. BFLOAT16, FLOAT16, FLOAT32, INT8, UINT8, INT16, INT32, INT64 - - - out (aclTensor*) Output out in the formula. The data type needs to be a convertible data type after derivation of self and other, and the shape needs to be the shape after broadcasting self and other. BFLOAT16, FLOAT16, FLOAT32, INT8, UINT8, INT16, INT32, INT64 ND 0-8 √ workspaceSize (uint64_t*) Output Returns the workspace size to be applied on the Device side. - - - - - executor (aclOpExecutor**) Output Returns the operator executor, which includes the computation process. - - - - - Atlas training products : The data types ofself,other,alpha, andoutdo 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 The passed self, other, alpha, or out is a null pointer. ACLNN_ERR_PARAM_INVALID 161002 The data types of self and other are not within the supported range. Unable to perform data type deduction on self and other. The derived data type cannot be converted to the specified output out type. The shapes of self and other cannot be broadcast. alpha is unable to convert to the derived data type of self and other.
aclnnAddRelu
Parameters:
Parameter Input/Output Description workspace Input Workspace memory address applied on the device side. workspaceSize Input Workspace size applied on the device side, obtained from the first-phase API aclnnAddReluGetWorkspaceSize. executor Input Operator executor, which includes the computation process of the operator. stream Input Specifies the stream for executing the task. Return value
aclnnStatus: return code. For details, see aclnn Return Code.
aclnnInplaceAddReluGetWorkspaceSize
Parameters:
Atlas training products : The data type of parametersselfRef,other, andalphadoes 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 The passed selfRef, other, or alpha is a null pointer. ACLNN_ERR_PARAM_INVALID 161002 The data types of selfRef and other are not within the supported range. Unable to perform data type deduction for selfRef and other. The derived data type is unable to convert to the type of selfRef. The shapes of selfRef and other cannot be broadcast. alpha is unable to convert to the derived data type of selfRef and other.
aclnnInplaceAddRelu
Parameters:
Parameter Input/Output Description workspace Input Workspace memory address applied on the Device side. workspaceSize Input Workspace size applied on the Device side, obtained from the first-phase API aclnnInplaceAddReluGetWorkspaceSize. executor Input Operator executor, which includes the computation process of the operator. stream Input Specifies the stream for task execution. Return value
aclnnStatus: return code. For details, see aclnn Return Code.
Constraints
Deterministic computation:
- aclnnAddRelu&aclnnInplaceAddRelu default to a deterministic implementation.
For scenarios where the data type of selfRef is INT8 and the data type of other is INT32:
When the cast operator converts INT32 to INT8, precision issues exist (for details, see aclnnCast). Therefore, the precision of the output result cannot be guaranteed in this scenario.
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_add_relu.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 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 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 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];
}
// 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 the device/stream. Refer to 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 = {4, 2};
std::vector<int64_t> otherShape = {4, 2};
std::vector<int64_t> outShape = {4, 2};
void* selfDeviceAddr = nullptr;
void* otherDeviceAddr = nullptr;
void* outDeviceAddr = nullptr;
aclTensor* self = nullptr;
aclTensor* other = nullptr;
aclScalar* alpha = nullptr;
aclTensor* out = nullptr;
std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
std::vector<float> otherHostData = {1, 1, 1, 2, 2, 2, 3, 3};
std::vector<float> outHostData(8, 0);
float alphaValue = 1.2f;
// Create the self aclTensor.
ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// Create the other aclTensor.
ret = CreateAclTensor(otherHostData, otherShape, &otherDeviceAddr, aclDataType::ACL_FLOAT, &other);
CHECK_RET(ret == ACL_SUCCESS, return ret);
// Create the alpha aclScalar.
alpha = aclCreateScalar(&alphaValue, aclDataType::ACL_FLOAT);
CHECK_RET(alpha != nullptr, 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 aclnnAddRelu API call
// 3. Call the CANN operator library API.
// Call the first-phase API of aclnnAddRelu.
ret = aclnnAddReluGetWorkspaceSize(self, other, alpha, out, &workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddReluGetWorkspaceSize 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 aclnnAddRelu.
ret = aclnnAddRelu(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddRelu failed. ERROR: %d\n", ret); return ret);
// 4. (Boilerplate) Wait for the task execution to finish 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 memory to the host memory. 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(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]);
}
// Example of calling the aclnnInplaceAddRelu API
// 3. Call the CANN operator library API.
LOG_PRINT("\ntest aclnnInplaceAddRelu\n");
// Call the first-phase API of aclnnInplaceAddRelu.
ret = aclnnInplaceAddReluGetWorkspaceSize(self, other, alpha, &workspaceSize, &executor);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceAddReluGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
// Apply for device memory based on the workspaceSize calculated by the first-phase API.
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 aclnnInplaceAddRelu.
ret = aclnnInplaceAddRelu(workspaceAddr, workspaceSize, executor, stream);
CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnInplaceAddRelu failed. ERROR: %d\n", ret); return ret);
// 4. (Boilerplate) Synchronously wait for 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 from the device memory to the host memory. Modify based on the specific API definition.
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. Modify based on the specific API definition.
aclDestroyTensor(self);
aclDestroyTensor(other);
aclDestroyScalar(alpha);
aclDestroyTensor(out);
// 7. Release Device resources. Modify based on the specific API definition.
aclrtFree(selfDeviceAddr);
aclrtFree(otherDeviceAddr);
aclrtFree(outDeviceAddr);
if (workspaceSize > 0) {
aclrtFree(workspaceAddr);
}
aclrtDestroyStream(stream);
aclrtResetDevice(deviceId);
aclFinalize();
return 0;
}