aclnnAddLora

📄 View Source Code

Applicable Products

Function

• Description:

  Multiplies the input x with the corresponding weightA and weightB based on the input indices, and then accumulates the result onto the input y and outputs it.

• Formula:

  Given an input tensor x whose last dimension has a length of 2d, the AddLora function performs the following computation:

  1. Rearrange x according to the indices in indices, so that x corresponding to the same group of weights are placed together.

  2. Iterate over each LoRA group, and perform matrix multiplication on the corresponding x and weightA respectively:

     $$Z1 = x_{i} \cdot weightA[i, layerIdx, :, :]$$

  3. The resulting Z1 then undergoes matrix multiplication with weightB:

     $$Z2 = Z1 \cdot weightB[i, layerIdx, :, :] \times scale$$

  4. Finally, accumulate the output Z2 onto y:

     $$\text{out} = y[:, yOffset: yOffset+ySliceSize] + Z2$$

Function Prototype

Each operator uses a two-phase API. You must first call the "aclnnAddLoraGetWorkspaceSize" API to obtain the input parameters and calculate the required workspace size based on the computation process, and then call the "aclnnAddLora" API to perform the computation.

aclnnStatus aclnnAddLoraGetWorkspaceSize(
    const aclTensor *y,
    const aclTensor *x,
    const aclTensor *weightB,
    const aclTensor *indices,
    const aclTensor *weightAOptional,
    int64_t          layerIdx,
    double           scale,
    int64_t          yOffset,
    int64_t          ySliceSize,
    const aclTensor *out,
    uint64_t        *workspaceSize,
    aclOpExecutor  **executor)

aclnnStatus aclnnAddLora(
    void          *workspace,
    uint64_t       workspaceSize,
    aclOpExecutor *executor,
    aclrtStream    stream)

aclnnAddLoraGetWorkspaceSize

• Parameters:

  Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (Shape)	Non-contiguous Tensor
  y (aclTensor*)	Input	Represents the tensor to be updated by accumulation, i.e., y in the formula.	The shape has two dimensions: [B, H3], where H3 is an integer multiple of 16 and must be in the range of 1 to 131072. The first dimension must be consistent with the first dimension of x, both represented by `B`. Empty tensor not supported.	FLOAT16	ND	2	√
  x (aclTensor*)	Input	Represents the input tensor before grouping, i.e., x in the formula.	Shape with 2 dimensions: [B, H1], where H1 is an integer multiple of 16. Empty tensor not supported.	FLOAT16	ND	2	√
  weightB (aclTensor*)	Input	Represents the second weight matrix for matrix multiplication, i.e., weightB in the formula.	Shape dimension: 4D [W, L, H2, R]. The third dimension must be less than or equal to the second dimension of y (H2 ≤ H3), H2 must be an integer multiple of 16, and H2 must be in the range of 1 to 131072. R must be in the range of 1 to 128 and must be an integer multiple of 16. Empty tensor not supported.	FLOAT16	ND, NZ	4	√
  indices (aclTensor*)	Input	Indicates the grouping index of input x, corresponding to the input indices in the formula.	The shape has one dimension: [B]. The first dimension must keep consistent with the first dimension of x and y, all represented by `B`. Empty tensor not supported.	INT32	ND	1	√
  weightAOptional (aclTensor*)	Input	Represents the first weight matrix for matrix multiplication. When empty, the first matrix multiplication is skipped. This corresponds to weightA in the formula.	Shape dimension: 4D [W, L, R, H1]. The first two dimensions must be consistent with the first two dimensions of `weightB`, denoted by `W` and `L`, where W ranges from 1 to 32 and L ranges from 1 to 32. The third dimension must keep consistent with the fourth dimension of `weightB`, both denoted by `R`. The fourth dimension must keep consistent with the second dimension of `x`, both denoted by `H1`, and must be an integer multiple of 16. Empty tensor not supported.	FLOAT16	ND, NZ		√
  layerIdx (int64_t)	Input	Represents the layer index, i.e., layerIdx in the formula.	The value must be less than L, the second dimension of weightB.	-	-	-	-
  scale (double)	Input	Represents the scaling factor, i.e., scale in the formula.	-	-	-	-	-
  yOffset (int64_t)	Input	Represents the offset when updating y, i.e., yOffset in the formula.	The value must be less than the second dimension H3 of y.	-	-	-	-
  ySliceSize (int64_t)	Input	Indicates the range when y is updated, i.e., ySliceSize in the formula.	The value must be less than or equal to the second dimension H3 of y.	-	-	-	-
  out (aclTensor*)	Output	Output tensor, which is the output out in the formula.	The data type of the output is consistent with that of the input. The output shape has the same dimensions as the shape of the input y.	FLOAT16	ND	2	√
  workspaceSize (uint64_t*)	Output	Returns the workspace size to be requested on the Device side.	-	-	-	-	-
  executor (aclOpExecutor**)	Output	Returns the op executor, which contains the operator computation process.	-	-	-	-	-

  • Atlas A2 training products/Atlas A2 inference products: weightB and weightAOptional support the ND data format.

• Return value

  aclnnStatus: return code. For details, see aclnn Return Codes.

  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 input parameters (x, y, weightB, indices) or the output parameter out is a null pointer.
  ACLNN_ERR_PARAM_INVALID	161002	The data type of the input parameters (x, y, weightB, indices) or the output parameter is not within the supported range.
  ACLNN_ERR_INNER_TILING_ERROR	561002	The shape information of multiple input tensors does not match (see Parameters for details).
  		The shape information of the input tensor is not supported (see Parameters for details).

aclnnAddLora

• Parameters:

  Parameter	Input/Output	Description
  workspace	Input	Address of the workspace memory applied on the device side.
  workspaceSize	Input	Size of the workspace applied on the device side, obtained by the first-phase API aclnnAddLoraGetWorkspaceSize.
  executor	Input	Operator executor, which contains the operator computation flow.
  stream	Input	Specifies the stream for task execution.

• Return value

  aclnnStatus: return code. For details, see aclnn Return Codes.

Constraints

• Deterministic computation:

  • aclnnAddLora defaults to a deterministic implementation.

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_lora.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) {
  // Boilerplate: initialize.
  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 allocate 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);
  // Call aclrtMemcpy to copy host-side data to the device-side memory.
  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);

  // Compute 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. See the ACL API documentation.
  // Enter the deviceId based on your 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. Custom construction is required based on the API.
  int32_t batchSize = 1;
  int32_t H1 = 16;
  int32_t H2 = 16;
  int32_t R = 16;
  int32_t loraNum = 1;
  int32_t layerNum = 1;

  std::vector<int64_t> xShape = {batchSize, H1};
  std::vector<int64_t> yShape = {batchSize, H2};
  std::vector<int64_t> weightBShape = {loraNum, layerNum, H2, R};
  std::vector<int64_t> indicesShape = {batchSize};
  std::vector<int64_t> weightAShape = {loraNum, layerNum, R, H1};
  std::vector<int64_t> outShape = {batchSize, H2};

  std::vector<float> xHostData(batchSize * H1, 1);
  std::vector<float> yHostData(batchSize * H2, 1);
  std::vector<float> weightBHostData(loraNum * layerNum * H2 * R, 1);
  std::vector<float> indicesHostData(batchSize, 0);
  std::vector<float> weightAHostData(loraNum * layerNum * R * H1, 1);
  std::vector<float> outHostData(batchSize * H2, 1);

  void* xInputDeviceAddr = nullptr;
  void* yInputDeviceAddr = nullptr;
  void* weightBInputDeviceAddr = nullptr;
  void* indicesInputDeviceAddr = nullptr;
  void* weightAInputDeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;

  aclTensor* xInput = nullptr;
  aclTensor* yInput = nullptr;
  aclTensor* weightBInput = nullptr;
  aclTensor* indicesInput = nullptr;
  aclTensor* weightAInput = nullptr;
  aclTensor* out = nullptr;

  // Create input x.
  ret = CreateAclTensor(xHostData, xShape, &xInputDeviceAddr, aclDataType::ACL_FLOAT16, &xInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create input y.
  ret = CreateAclTensor(yHostData, yShape, &yInputDeviceAddr, aclDataType::ACL_FLOAT16, &yInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create input weightB.
  ret = CreateAclTensor(weightBHostData, weightBShape, &weightBInputDeviceAddr, aclDataType::ACL_FLOAT16, &weightBInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create input indices.
  ret = CreateAclTensor(indicesHostData, indicesShape, &indicesInputDeviceAddr, aclDataType::ACL_INT32, &indicesInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create input weightA.
  ret = CreateAclTensor(weightAHostData, weightAShape, &weightAInputDeviceAddr, aclDataType::ACL_FLOAT16, &weightAInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the out aclTensor.
  ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT16, &out);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  int64_t layer_idx = 0;
  double scale = 1.0;
  int64_t y_offset = 0;
  int64_t y_slice_size = H2;

  // 3. Call the CANN operator library API. Replace with the specific API name.
  uint64_t workspaceSize = 16 * 1024 * 1024;
  aclOpExecutor* executor;

  // Call the aclnnAddLora first-phase API.
  ret = aclnnAddLoraGetWorkspaceSize(yInput, xInput, weightBInput, indicesInput, weightAInput, layer_idx, scale, y_offset, y_slice_size, out, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddLoraGetWorkspaceSize 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 > static_cast<uint64_t>(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 aclnnAddLora second-phase API.
  ret = aclnnAddLora(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAddLora 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. Copy the result from the device memory to the host. Modify based on the specific API definition.
  PrintOutResult(outShape, &outDeviceAddr);

  // 6. Release aclTensor. Modify based on the specific API definition.
  aclDestroyTensor(xInput);
  aclDestroyTensor(yInput);
  aclDestroyTensor(weightBInput);
  aclDestroyTensor(indicesInput);
  aclDestroyTensor(weightAInput);
  aclDestroyTensor(out);

  // 7. Release device resources. Modify based on the specific API definition.
  aclrtFree(xInputDeviceAddr);
  aclrtFree(yInputDeviceAddr);
  aclrtFree(weightBInputDeviceAddr);
  aclrtFree(indicesInputDeviceAddr);
  aclrtFree(weightAInputDeviceAddr);
  aclrtFree(outDeviceAddr);
  if (workspaceSize > static_cast<uint64_t>(0)) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();

  return 0;
}