下载
中文
注册

aclnnDynamicQuantV2

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“aclnnDynamicQuantV2GetWorkspaceSize”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“aclnnDynamicQuantV2”接口执行计算。

  • aclnnstatus aclnnDynamicQuantV2GetWorkspaceSize(const aclTensor *x, aclTensor *smoothScalesOptional, aclTensor *groupIndexesOptional, int64_t dstType, aclTensor *yOut, aclTensor *scaleOut, uint64_t *workspaceSize, aclOpExecutor **executor)
  • aclnnstatus aclnnDynamicQuantV2(void *workspace, uint64_t workspaceSize, aclOpExecutor *executor, aclrtStream stream)

功能描述

  • 算子功能:为输入张量进行pre-token对称/非对称动态量化。MOE场景,每个专家的smoothScale是不同的,根据输入的groupIndexes进行区分

  • 计算公式:

    • 对称量化:
    • 若不输入smoothScale,则
    scaleOut=row_max(abs(x))/127scaleOut=row\_max(abs(x))/127 yOut=round(x/scalOut)yOut=round(x/scalOut)
    • 若输入smoothScales,则scaleOut=row_max(abs(xsmoothScalesOptional))/127scaleOut=row\_max(abs(x\cdot smoothScalesOptional))/127
    yOut=round(x/scalOut)yOut=round(x/scalOut)
    • 非对称量化:
    • 若不输入smoothScale,则scaleOut=(row_max(x)row_min(x))/scale_optscaleOut=(row\_max(x) - row\_min(x))/scale\_opt offset=offset_optrow_max(x)/scaleOutoffset=offset\_opt-row\_max(x)/scaleOut yOut=round(x/scaleOut+offset)yOut=round(x/scaleOut+offset)
    • 若输入smoothScales,则input=xsmoothScalesOptionalinput = x\cdot smoothScalesOptional scaleOut=(row_max(input)row_min(input))/scale_optscaleOut=(row\_max(input) - row\_min(input))/scale\_opt offset=offset_optrow_max(input)/scaleOutoffset=offset\_opt-row\_max(input)/scaleOut yOut=round(input/scaleOut+offset)yOut=round(input/scaleOut+offset) 其中row_max代表每行求最大值,row_min代表每行求最小值。当输出yOut类型为INT8时,scale_opt为255.0,offset_opt为127.0;yOut类型为INT4时,scale_opt为15.0,offset_opt为7.0。

aclnnDynamicQuantV2GetWorkspaceSize

  • 参数说明:

    • x(aclTensor*, 计算输入):必选参数,算子输入的Tensor,shape维度要大于1,Device侧的aclTensor,数据类型支持FLOAT16、BFLOAT16,支持非连续的Tensor数据格式支持ND。
    • smoothScalesOptional(aclTensor*, 计算输入):可选参数,算子输入的smoothScales,当没有groupIndexsOptional输入时shape维度时x的最后一维,有groupIndexsOptional输入时shape是两维,第一维大小是专家数,不超过1024,第二维大小是x的最后一维,Device侧的aclTensor,数据类型支持FLOAT16、BFLOAT16,并且数据类型要和x保持一致,支持非连续的Tensor数据格式支持ND。
    • groupIndexesOptional(aclTensor*, 计算输入):可选参数,算子输入的groupIndexes,shape只有一维,Device侧的aclTensor,数据类型支持INT32,支持非连续的Tensor数据格式支持ND。
    • dstType (int64_t, 计算输入):可选参数,输出y的类型对应的枚举值,Host侧的int,如果输出y类型为INT8,则为2;y类型为INT4时,则为29,默认为2。
    • yOut(aclTensor*, 计算输出):量化后的输出Tensor,shape维度和x保持一致,Device侧的aclTensor,数据类型支持INT4,INT8,暂不支持非连续的Tensor,数据格式支持ND。
    • scaleOut(aclTensor*, 计算输出):量化使用的scale,shape维度为x的shape剔除最后一维,Device侧的aclTensor,数据类型支持FLOAT,暂不支持非连续的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x或out参数是空指针。
    返回561002 (ACLNN_ERR_PARAM_INVALID):1. 输入或输出数据格式不在支持的范围之内
    返回561003 (ACLNN_ERR_INNER_FIND_KERNEL_ERROR):1. 输入或输出的数据类型不在支持范围之内

aclnnDynamicQuantV2

  • 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aclnnAbsGetWorkspaceSize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

针对Atlas A2训练系列产品/Atlas 800I A2推理产品,groupIndexesOptional的维度不超过1024。

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_dynamic_quant_v2.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;
}

void PrintOutResult(std::vector<int64_t> &shape, void** deviceAddr) {
  auto size = GetShapeSize(shape);
  std::vector<float> resultData(size, 0);
  auto ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]),
                         *deviceAddr, 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);
  for (int64_t i = 0; i < size; i++) {
    LOG_PRINT("mean result[%ld] is: %f\n", i, resultData[i]);
  }
}

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AscendCL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_RET(ret == ACL_SUCCESS, 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
  int rowNum = 4;
  int rowLen = 2;
  int groupNum = 2;
  std::vector<int64_t> xShape = {4, 2};
  std::vector<int64_t> smoothShape = {groupNum, rowLen};
  std::vector<int64_t> groupShape = {groupNum};
  std::vector<int64_t> yShape = {4, 2};
  std::vector<int64_t> scaleShape = {4};
  std::vector<int64_t> offsetShape = {4};

  void* xDeviceAddr = nullptr;
  void* smoothDeviceAddr = nullptr;
  void* groupDeviceAddr = nullptr;
  void* yDeviceAddr = nullptr;
  void* scaleDeviceAddr = nullptr;
  void* offsetDeviceAddr = nullptr;

  aclTensor* x = nullptr;
  aclTensor* smooth = nullptr;
  aclTensor* group = nullptr;
  aclTensor* y = nullptr;
  aclTensor* scale = nullptr;
  aclTensor* offset = nullptr;

  std::vector<aclFloat16> xHostData;
  std::vector<aclFloat16> smoothHostData;
  std::vector<int32_t> groupHostData = {2, rowNum};
  std::vector<int8_t> yHostData;
  std::vector<float> scaleHostData;
  std::vector<float> offsetHostData;
  for (int i = 0; i < rowNum; ++i) {
    for (int j = 0; j < rowLen; ++j) {
      float value1 = i * rowLen + j;
      xHostData.push_back(aclFloatToFloat16(value1));
      yHostData.push_back(0);
    }
    scaleHostData.push_back(0);
    offsetHostData.push_back(0);
  }

  for (int m = 0; m < groupNum; ++m) {
    for (int n = 0; n < rowLen; ++n) {
      float value2 = m * rowLen + n;
      smoothHostData.push_back(aclFloatToFloat16(value2));
    }
  }

  // 创建x aclTensor
  ret = CreateAclTensor(xHostData, xShape, &xDeviceAddr, aclDataType::ACL_FLOAT16, &x);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建smooth aclTensor
  ret = CreateAclTensor(smoothHostData, smoothShape, &smoothDeviceAddr, aclDataType::ACL_FLOAT16, &smooth);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建group aclTensor
  ret = CreateAclTensor(groupHostData, groupShape, &groupDeviceAddr, aclDataType::ACL_INT32, &group);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建y aclTensor
  ret = CreateAclTensor(yHostData, yShape, &yDeviceAddr, aclDataType::ACL_INT8, &y);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // 创建scale aclTensor
  ret = CreateAclTensor(scaleHostData, scaleShape, &scaleDeviceAddr, aclDataType::ACL_FLOAT, &scale);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
    // 创建offset aclTensor
  ret = CreateAclTensor(offsetHostData, offsetShape, &offsetDeviceAddr, aclDataType::ACL_FLOAT, &offset);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. 调用CANN算子库API,需要修改为具体的Api名称
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;

  // 调用aclnnDynamicQuantV2第一段接口
  ret = aclnnDynamicQuantV2GetWorkspaceSize(x, smooth, group, 2,  y, scale, offset, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnDynamicQuantV2GetWorkspaceSize 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aclnnDynamicQuantV2第二段接口
  ret = aclnnDynamicQuantV2(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnDynamicQuantV2 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(yShape, &yDeviceAddr);

  // 6. 释放aclTensor和aclScalar,需要根据具体API的接口定义修改
  aclDestroyTensor(x);
  aclDestroyTensor(smooth);
  aclDestroyTensor(y);
  aclDestroyTensor(scale);
  aclDestroyTensor(offset);

  // 7. 释放device资源
  aclrtFree(xDeviceAddr);
  aclrtFree(smoothDeviceAddr);
  aclrtFree(yDeviceAddr);
  aclrtFree(scaleDeviceAddr);
  aclrtFree(offsetDeviceAddr);
  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();

  return 0;
}