* Copyright (c) 2026 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License").
* Please refer to the License for details. You may not use this file except in compliance with the License.
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
* NOTE: Portions of this code were AI-generated and have been
* technically reviewed for functional accuracy and security
*/
* @file test_aclnn_celu_v3.cpp
* @brief CeluV3 算子调用示例
*
* 演示如何通过 aclnn 接口在 NPU 上调用 CeluV3 算子。
*
* 公式: CELU(x) = max(0, x) + min(0, alpha * (exp(x / alpha) - 1))
*
* 编译方法:
* 1. 先编译并安装算子包:
* cd ops/celu_v3 && bash build.sh --pkg --soc=ascend910b
* cd build && ./custom_opp_ubuntu_aarch64.run
*
* 2. 编译本示例:
* g++ -o test_aclnn_celu_v3 test_aclnn_celu_v3.cpp \
* -I${ASCEND_HOME_PATH}/include \
* -L${ASCEND_HOME_PATH}/lib64 \
* -lascendcl -lnnopbase -lopapi -lacl_op_compiler
*
* 运行方法:
* ./test_aclnn_celu_v3
*/
#include <iostream>
#include <vector>
#include <cmath>
#include <cstring>
#include "acl/acl.h"
#include "aclnn_celu_v3.h"
#define CHECK_ACL(expr) \
do { \
auto _ret = (expr); \
if (_ret != ACL_SUCCESS) { \
std::cerr << "ACL Error: " << #expr << " returned " << _ret << " at " << __FILE__ << ":" << __LINE__ \
<< std::endl; \
return 1; \
} \
} while (0)
int main()
{
constexpr int64_t ELEM_COUNT = 8;
const int64_t shape[] = {ELEM_COUNT};
const int64_t strides[] = {1};
constexpr int64_t ndim = 1;
constexpr double alpha = 1.0;
float hostInput[ELEM_COUNT] = {-3.0f, -1.0f, -0.5f, 0.0f, 0.5f, 1.0f, 2.0f, 3.0f};
CHECK_ACL(aclInit(nullptr));
CHECK_ACL(aclrtSetDevice(0));
aclrtStream stream = nullptr;
CHECK_ACL(aclrtCreateStream(&stream));
size_t dataBytes = ELEM_COUNT * sizeof(float);
void* devInput = nullptr;
void* devOutput = nullptr;
CHECK_ACL(aclrtMalloc(&devInput, dataBytes, ACL_MEM_MALLOC_HUGE_FIRST));
CHECK_ACL(aclrtMalloc(&devOutput, dataBytes, ACL_MEM_MALLOC_HUGE_FIRST));
CHECK_ACL(aclrtMemset(devOutput, dataBytes, 0, dataBytes));
CHECK_ACL(aclrtMemcpy(devInput, dataBytes, hostInput, dataBytes, ACL_MEMCPY_HOST_TO_DEVICE));
aclTensor* selfTensor = aclCreateTensor(shape, ndim, ACL_FLOAT, strides, 0, ACL_FORMAT_ND, shape, ndim, devInput);
aclTensor* outTensor = aclCreateTensor(shape, ndim, ACL_FLOAT, strides, 0, ACL_FORMAT_ND, shape, ndim, devOutput);
uint64_t workspaceSize = 0;
aclOpExecutor* executor = nullptr;
CHECK_ACL(aclnnCeluV3GetWorkspaceSize(selfTensor, alpha, outTensor, &workspaceSize, &executor));
void* workspace = nullptr;
if (workspaceSize > 0) {
CHECK_ACL(aclrtMalloc(&workspace, workspaceSize, ACL_MEM_MALLOC_HUGE_FIRST));
}
CHECK_ACL(aclnnCeluV3(workspace, workspaceSize, executor, stream));
CHECK_ACL(aclrtSynchronizeStream(stream));
float hostOutput[ELEM_COUNT] = {};
CHECK_ACL(aclrtMemcpy(hostOutput, dataBytes, devOutput, dataBytes, ACL_MEMCPY_DEVICE_TO_HOST));
std::cout << "CeluV3 Example (alpha=" << alpha << ")" << std::endl;
std::cout << "-------------------------------------------" << std::endl;
std::cout << " Index | Input | Output | Expected" << std::endl;
std::cout << "-------------------------------------------" << std::endl;
for (int i = 0; i < ELEM_COUNT; ++i) {
float x = hostInput[i];
float expected = std::max(0.0f, x) +
std::min(0.0f, static_cast<float>(alpha) * (expf(x / static_cast<float>(alpha)) - 1.0f));
printf(" %5d | %9.4f | %9.4f | %9.4f\n", i, x, hostOutput[i], expected);
}
std::cout << "-------------------------------------------" << std::endl;
aclDestroyTensor(selfTensor);
aclDestroyTensor(outTensor);
if (workspace)
aclrtFree(workspace);
aclrtFree(devInput);
aclrtFree(devOutput);
aclrtDestroyStream(stream);
aclrtResetDevice(0);
aclFinalize();
std::cout << "CeluV3 example finished successfully." << std::endl;
return 0;
}