--- title: aclnnBinaryCrossEntropyWithLogits description: "每个算子有两段接口，必须先调用“aclnnXxxGetWorkspaceSize”接口获取入参并根据计算流程计算所需workspace大小，再调用“aclnnXxx”接口执行计算。两段式接口如下：" url: https://www.hiascend.com/document/detail/zh/Atlas200IDKA2DeveloperKit/latest/Application%20Development%20Guide/aadgc/aclcppdevg_03_0235.html sourcePath: /source/zh/Atlas200IDKA2DeveloperKit/23.0.RC2/Application Development Guide/aadgc/aclcppdevg_03_0235.html indexId: 9df4639e581a001d9242551d13aec7db056a75fa6f5fc2b3fe6ac3fc1a388215105 --- # aclnnBinaryCrossEntropyWithLogits #### 接口原型 每个算子有两段接口，必须先调用“aclnnXxxGetWorkspaceSize”接口获取入参并根据计算流程计算所需workspace大小，再调用“aclnnXxx”接口执行计算。两段式接口如下： - 第一段接口：aclnnStatus aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize(const aclTensor* self, const aclTensor* target, const aclTensor* weightOptional, const aclTensor* posWeightOptional, int64_t reduction, aclTensor* out, uint64_t* workspaceSize, aclOpExecutor** executor) - 第二段接口：aclnnstatus aclnnBinaryCrossEntropyWithLogits(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, const aclrtStream stream) #### 功能描述 - 算子功能：计算logits输出self与标签target之间的BCE损失函数（Binary CrossEntropy Loss）。 - 计算公式： - 单标签场景 - 多标签场景 #### aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize - 接口定义： aclnnStatus aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize(const aclTensor* self, const aclTensor* target, const aclTensor* weightOptional, const aclTensor* posWeightOptional, int64_t reduction, aclTensor* out, uint64_t* workspaceSize, aclOpExecutor** executor) - 参数说明： - self：连接层输出，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16，支持非连续的Tensor，数据格式支持ND。 - target：label标签值，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16，支持非连续的Tensor，数据格式支持ND。 - weightOptional：可选入参，二分交叉熵权重，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16，支持非连续的Tensor，数据格式支持ND。 - posWeightOptional：可选入参，各类的正类权重，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16，支持非连续的Tensor，数据格式支持ND。 - reduction：可选入参，输出结果计算方式，Host侧的int类型值。 取值支持0、1、2，0代表None不操作，1代表Mean求损失函数均值，2代表Sum求损失函数的和。 - out：输出误差，Device侧的aclTensor，数据类型支持FLOAT，支持非连续的Tensor，数据格式支持ND。 - workspaceSize：返回用户需要在Device侧申请的workspace大小。 - executor：返回op执行器，包含了算子计算流程。 - 返回值： 返回aclnnStatus状态码，具体参见aclnn返回码。 第一段接口完成入参校验，出现以下场景时报错： 返回161001（ACLNN_ERR_PARAM_NULLPTR）：传入的tensor或out是空指针。 返回161002（ACLNN_ERR_PARAM_INVALID）： self、target、weightOptional和posWeightOptional的数据类型和数据格式不在支持的范围之内。 self和target维度不一致。 weightOptional、posWeightOptional不能扩展成self、target形状。 reduction取值非0、1、2三值之一。 #### aclnnBinaryCrossEntropyWithLogits - 接口定义： aclnnstatus aclnnBinaryCrossEntropyWithLogits(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, const aclrtStream stream) - 参数说明： - workspace：在Device侧申请的workspace内存起址。 - workspaceSize：在Device侧申请的workspace大小，由第一段接口aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize获取。 - executor：op执行器，包含了算子计算流程。 - stream：指定执行任务的AscendCL stream流。 - 返回值： 返回aclnnStatus状态码，具体参见aclnn返回码。 #### 调用示例 ``` #include #include #include "acl/acl.h" #include "aclnnop/aclnn_binary_cross_entropy_with_logits.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& shape) { int64_t shape_size = 1; for (auto i : shape) { shape_size *= i; } return shape_size; } void PrintOutResult(std::vector &shape, void** deviceAddr) { auto size = GetShapeSize(shape); std::vector resultData(size, 0); auto ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), *deviceAddr, 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); for (int64_t i = 0; i < size; i++) { LOG_PRINT("result[%ld] is: %f\n", i, resultData[i]); } } int Init(int32_t deviceId, aclrtContext* context, aclrtStream* stream) { // 固定写法，AscendCL初始化 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 = aclrtCreateContext(context, deviceId); CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtCreateContext failed. ERROR: %d\n", ret); return ret); ret = aclrtSetCurrentContext(*context); CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSetCurrentContext 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 int CreateAclTensor(const std::vector& hostData, const std::vector& shape, void** deviceAddr, aclDataType dataType, aclTensor** tensor) { auto size = GetShapeSize(shape) * sizeof(T); // 调用aclrtMalloc申请device侧内存 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); // 调用aclrtMemcpy将Host侧数据拷贝到device侧内存上 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); // 计算连续tensor的strides std::vector strides(shape.size(), 1); for (int64_t i = shape.size() - 2; i >= 0; i--) { strides[i] = shape[i + 1] * strides[i + 1]; } // 调用aclCreateTensor接口创建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. （固定写法）device/context/stream初始化, 参考AscendCL对外接口列表 // 根据自己的实际device填写deviceId int32_t deviceId = 0; aclrtContext context; aclrtStream stream; auto ret = Init(deviceId, &context, &stream); CHECK_RET(ret == 0, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret); // check根据自己的需要处理 // 2. 构造输入与输出，需要根据API的接口自定义构造 std::vector inputShape = {4, 2}; std::vector targetShape = {4, 2}; std::vector weightShape = {4, 2}; std::vector posWeightShape = {4, 2}; std::vector outShape = {4, 2}; void* inputDeviceAddr = nullptr; void* targetDeviceAddr = nullptr; void* weightDeviceAddr = nullptr; void* posWeightDeviceAddr = nullptr; void* outDeviceAddr = nullptr; aclTensor* input = nullptr; aclTensor* target = nullptr; aclTensor* weight = nullptr; aclTensor* posWeight = nullptr; aclTensor* out = nullptr; std::vector inputHostData = {0.1, 0.1, 0.2, 0.2, 0.3, 0.3, 0.4, 0.4}; std::vector targetHostData = {0.2, 0.2, 0.1, 0.1, 0.2, 0.2, 0.1, 0.1}; std::vector weightHostData = {0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5}; std::vector posWeightHostData = {0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5, 0.5}; std::vector outHostData = {0, 0, 0, 0, 0, 0, 0, 0}; ret = CreateAclTensor(inputHostData, inputShape, &inputDeviceAddr, aclDataType::ACL_FLOAT, &input); CHECK_RET(ret == ACL_SUCCESS, return ret); ret = CreateAclTensor(targetHostData, targetShape, &targetDeviceAddr, aclDataType::ACL_FLOAT, &target); CHECK_RET(ret == ACL_SUCCESS, return ret); ret = CreateAclTensor(weightHostData, weightShape, &weightDeviceAddr, aclDataType::ACL_FLOAT, &weight); CHECK_RET(ret == ACL_SUCCESS, return ret); ret = CreateAclTensor(posWeightHostData, posWeightShape, &posWeightDeviceAddr, aclDataType::ACL_FLOAT, &posWeight); CHECK_RET(ret == ACL_SUCCESS, return ret); ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT, &out); CHECK_RET(ret == ACL_SUCCESS, return ret); int64_t reduction = 0; // 3.调用CANN算子库API，需要修改为具体的算子接口 uint64_t workspaceSize = 0; aclOpExecutor* executor; // 调用aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize第一段接口 ret = aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize(input, target, weight, posWeight, reduction, out, &workspaceSize, &executor); CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBinaryCrossEntropyWithLogitsGetWorkspaceSize failed. ERROR: %d\n", ret); return ret); // 根据第一段接口计算出的workspaceSize申请device内存 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;); } // 调用aclnnBinaryCrossEntropyWithLogits第二段接口 ret = aclnnBinaryCrossEntropyWithLogits(workspaceAddr, workspaceSize, executor, stream); CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBinaryCrossEntropyWithLogits failed. ERROR: %d\n", ret); return ret); // 4. （固定写法）同步等待任务执行结束 ret = aclrtSynchronizeStream(stream); CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret); // 5. 获取输出的值，将device侧内存上的结果拷贝至Host侧，需要根据具体API的接口定义修改 PrintOutResult(outShape, &outDeviceAddr); // 6. 释放aclTensor和aclScalar，需要根据具体API的接口定义修改 aclDestroyTensor(input); aclDestroyTensor(target); aclDestroyTensor(weight); aclDestroyTensor(posWeight); aclDestroyTensor(out); return 0; } ```