文档HOT
下载
中文
注册

aclnnMultiScaleDeformableAttentionGrad

支持的产品型号

  • Atlas A2训练系列产品/Atlas 800I A2推理产品。

接口原型

每个算子分为两段式接口，必须先调用“aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize”接口获取计算所需workspace大小以及包含了算子计算流程的执行器，再调用“aclnnMultiScaleDeformableAttentionGrad”接口执行计算。

  • aclnnStatus aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize(const aclTensor *value, const aclTensor *spatialShape, const aclTensor *levelStartIndex, const aclTensor *location, const aclTensor *attnWeight, const aclTensor *gradOutput, aclTensor *gradValue, aclTensor *gradLocation, aclTensor *gradAttnWeight, uint64_t *workspaceSize, aclOpExecutor **executor)
  • aclnnStatus aclnnMultiScaleDeformableAttentionGrad(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)

功能描述

算子功能： MultiScaleDeformableAttention正向算子功能主要通过采样位置（sample location）、注意力权重（attention weights）、映射后的value特征、多尺度特征起始索引位置、多尺度特征图的空间大小（便于将采样位置由归一化的值变成绝对位置）等参数来遍历不同尺寸特征图的不同采样点。而反向算子的功能为根据正向的输入对输出的贡献及初始梯度求出输入对应的梯度。

aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize

  • 参数说明：

    • value（aclTensor*, 计算输入）：特征图的特征值，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, spatial_size, mum_heads, channels），支持非连续的Tensor，数据格式支持ND
    • spatialShape（aclTensor*, 计算输入）：存储每个尺度特征图的高和宽，Device侧的aclTensor，数据类型支持INT32、INT64，shape为（num_levels, 2），支持非连续的Tensor，数据格式支持ND
    • levelStartIndex（aclTensor*, 计算输入）：每张特征图的起始索引，Device侧的aclTensor，数据类型支持INT32、INT64，shape为（num_levels,），支持非连续的Tensor，数据格式支持ND
    • location（aclTensor*, 计算输入）：采样点位置tensor，存储每个采样点的坐标位置，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, num_queries, num_heads, num_levels, num_points, 2），支持非连续的Tensor，数据格式支持ND
    • attnWeight（aclTensor, 计算输入）：采样点权重tensor，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, num_queries, num_heads, num_levels, num_points），支持非连续的Tensor，数据格式支持ND
    • gradOutput（aclTensor*, 计算输入）：正向输出梯度，也是反向算子的初始梯度，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, num_queries, num_heads, channels），支持非连续的Tensor，数据格式支持ND
    • gradValue（aclTensor, 计算输出）：输入value对应的梯度，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, spatial_size, mum_heads, channels），支持非连续的Tensor，数据格式支持ND
    • gradLocation（aclTensor*, 计算输出）：输入location对应的梯度，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, num_queries, num_heads, num_levels, num_points, 2），支持非连续的Tensor，数据格式支持ND
    • gradAttnWeight（aclTensor*, 计算输出）：输入attnWeight对应的梯度，Device侧的aclTensor，数据类型支持FLOAT、FLOAT16、BFLOAT16，shape为（bs, num_queries, num_heads, num_levels, num_points），支持非连续的Tensor，数据格式支持ND
    • workspaceSize（uint64_t*, 出参）：返回需要在Device侧申请的workspace大小。
    • executor（aclOpExecutor**, 出参）：返回op执行器，包含了算子计算流程。

  • 返回值：

    aclnnStatus：返回状态码，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）: 1. 传入的输入或输出是空指针。
返回161002（ACLNN_ERR_PARAM_INVALID）: 1. 输入和输出的数据类型不在支持的范围之内。
                                      2. 输入输出数据类型不一致。

aclnnMultiScaleDeformableAttentionGrad

  • 参数说明：

    • workspace（void*, 入参）：在Device侧申请的workspace内存地址。
    • workspaceSize（uint64_t, 入参）：在Device侧申请的workspace大小，由第一段接口aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize获取。
    • executor（aclOpExecutor*, 入参）：op执行器，包含了算子计算流程。
    • stream（aclrtStream, 入参）：指定执行任务的AscendCL Stream流。

  • 返回值：

    aclnnStatus：返回状态码，具体参见aclnn返回码。

约束与限制

channels + num_points + num_level <= 512 channels % 8 == 0，且channels<=256 num_queries < 500000 num_level < 10 bs <= 32

调用示例

示例代码如下，仅供参考，具体编译和执行过程请参考编译与运行样例。

#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_multi_scale_deformable_attention_grad.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) {
    // 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 = 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);
    // 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<int64_t> 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/stream初始化, 参考acl对外接口列表
    // 根据自己的实际device填写deviceId
    int32_t deviceId = 0;
    aclrtStream stream;
    auto ret = Init(deviceId, &stream);
    // check根据自己的需要处理
    CHECK_RET(ret == 0, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);
    // 2.构造输入与输出，需要根据API的接口自定义构造
    std::vector<int64_t> valueShape = {1, 1, 1, 8};
    std::vector<int64_t> spatialShapeShape = {1, 2};
    std::vector<int64_t> levelStartIndexShape = {1};
    std::vector<int64_t> locationShape = {1, 1, 1, 1, 1, 2};
    std::vector<int64_t> attnWeightShape = {1, 1, 1, 1, 1};
    std::vector<int64_t> gradOutputShape = {1, 1, 8};
    std::vector<int64_t> gradValueShape = {1, 1, 1, 8};
    std::vector<int64_t> gradLocationShape = {1, 1, 1, 1, 1, 2};
    std::vector<int64_t> gradAttnWeightShape = {1, 1, 1, 1, 1};
    void* valueDeviceAddr = nullptr;
    void* spatialShapeDeviceAddr = nullptr;
    void* levelStartIndexDeviceAddr = nullptr;
    void* locationDeviceAddr = nullptr;
    void* attnWeightDeviceAddr = nullptr;
    void* gradOutputDeviceAddr = nullptr;
    void* gradValueDeviceAddr = nullptr;
    void* gradLocationDeviceAddr = nullptr;
    void* gradAttnWeightDeviceAddr = nullptr;
    aclTensor* value = nullptr;
    aclTensor* spatialShape = nullptr;
    aclTensor* levelStartIndex = nullptr;
    aclTensor* location = nullptr;
    aclTensor* attnWeight = nullptr;
    aclTensor* gradOutput = nullptr;
    aclTensor* gradValue = nullptr;
    aclTensor* gradLocation = nullptr;
    aclTensor* gradAttnWeight = nullptr;
    std::vector<float> valueHostData = {1, 1, 1, 1, 1, 1, 1, 1};
    std::vector<float> spatialShapeHostData = {1, 1};
    std::vector<float> levelStartIndexHostData = {0};
    std::vector<float> locationHostData = {1, 1};
    std::vector<float> attnWeightHostData = {1};
    std::vector<float> gradOutputHostData = {1, 1, 1, 1, 1, 1, 1, 1};
    std::vector<float> gradValueHostData = {0, 0, 0, 0, 0, 0, 0, 0};
    std::vector<float> gradLocationHostData = {0, 0};
    std::vector<float> gradAttnWeightHostData = {0};
    // value aclTensor
    ret = CreateAclTensor(valueHostData, valueShape, &valueDeviceAddr, aclDataType::ACL_FLOAT, &value);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建spatialShape aclTensor
    ret = CreateAclTensor(spatialShapeHostData, spatialShapeShape, &spatialShapeDeviceAddr, aclDataType::ACL_INT32, &spatialShape);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建levelStartIndex aclTensor
    ret = CreateAclTensor(levelStartIndexHostData, levelStartIndexShape, &levelStartIndexDeviceAddr, aclDataType::ACL_INT32, &levelStartIndex);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建location aclTensor
    ret = CreateAclTensor(locationHostData, locationShape, &locationDeviceAddr, aclDataType::ACL_FLOAT, &location);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建attnWeight aclTensor
    ret = CreateAclTensor(attnWeightHostData, attnWeightShape, &attnWeightDeviceAddr, aclDataType::ACL_FLOAT, &attnWeight);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // gradOutput aclTensor
    ret = CreateAclTensor(gradOutputHostData, gradOutputShape, &gradOutputDeviceAddr, aclDataType::ACL_FLOAT, &gradOutput);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建gradValue aclTensor
    ret = CreateAclTensor(gradValueHostData, gradValueShape, &gradValueDeviceAddr, aclDataType::ACL_FLOAT, &gradValue);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建gradLocation aclTensor
    ret = CreateAclTensor(gradLocationHostData, gradLocationShape, &gradLocationDeviceAddr, aclDataType::ACL_FLOAT, &gradLocation);
    CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建gradAttnWeight aclTensor
    ret = CreateAclTensor(gradAttnWeightHostData, gradAttnWeightShape, &gradAttnWeightDeviceAddr, aclDataType::ACL_FLOAT, &gradAttnWeight);
    CHECK_RET(ret == ACL_SUCCESS, return ret);

    // 3.调用CANN算子库API，需要修改为具体的API
    uint64_t workspaceSize = 0;
    aclOpExecutor* executor;
    // 调用aclnnAdd第一段接口
    ret = aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize(value, spatialShape, levelStartIndex, location, attnWeight, gradOutput, gradValue, gradLocation, gradAttnWeight, &workspaceSize, &executor);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnMultiScaleDeformableAttentionGradGetWorkspaceSize 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aclnnAdd第二段接口
    ret = aclnnMultiScaleDeformableAttentionGrad(workspaceAddr, workspaceSize, executor, stream);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnMultiScaleDeformableAttentionGrad 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
    auto size = GetShapeSize(gradValueShape);
    std::vector<float> resultData(size, 0);
    ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), gradValueDeviceAddr, 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]);
    }

    // 6.释放aclTensor和aclScalar，需要根据具体API的接口定义修改
    aclDestroyTensor(value);
    aclDestroyTensor(spatialShape);
    aclDestroyTensor(levelStartIndex);
    aclDestroyTensor(location);
    aclDestroyTensor(attnWeight);
    aclDestroyTensor(gradOutput);
    aclDestroyTensor(gradValue);
    aclDestroyTensor(gradLocation);
    aclDestroyTensor(gradAttnWeight);

    // 7.释放device资源，需要根据具体API的接口定义修改
    aclrtFree(valueDeviceAddr);
    aclrtFree(spatialShapeDeviceAddr);
    aclrtFree(levelStartIndexDeviceAddr);
    aclrtFree(locationDeviceAddr);
    aclrtFree(attnWeightDeviceAddr);
    aclrtFree(gradOutputDeviceAddr);
    aclrtFree(gradValueDeviceAddr);
    aclrtFree(gradLocationDeviceAddr);
    aclrtFree(gradAttnWeightDeviceAddr);
    if (workspaceSize > 0) {
        aclrtFree(workspaceAddr);
    }
    aclrtDestroyStream(stream);
    aclrtResetDevice(deviceId);
    aclFinalize();
    return 0;
}