aclnnAdaptiveMaxPool3d

Applicable Products

Applicable Products	Supported
Ascend 950PR/Ascend 950DT	√
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	×
Atlas inference products	×
Atlas training products	×

Function

Description: Computes the kernel size for each pooling operation based on the input outputSize, performs 3D max pooling on the input self, and outputs the pooled values outputOut and indices indicesOut. The difference between aclnnAdaptiveMaxPool3d and aclnnMaxPool3d is that you only need to specify the outputSize, and the pooling regions are divided according to the outputSize.
Formula:

N (batch) represents the batch size, C (channels) represents the feature map channels, D (depth) represents the feature map depth, H (height) represents the feature map height, and W (width) represents the feature map width. For the scenario where the input self has dimensions $[N,C,D,H,W]$ and the outputSize is $[D_o,H_o,W_o]$ , the output output has dimensions $[N,C,D_o,H_o,W_o]$ , the indices indices have dimensions $[N,C,D_o,H_o,W_o]$ , and the formula for each element $(l,m,n)$ in the corresponding tensor is as follows:
$D^{l}_{left} = \lfloor(l*D)/D_o\rfloor$ $D^{l}_{right} = \lceil((l+1)*D)/D_o\rceil$ $H^{m}_{left} = \lfloor(m*H)/H_o\rfloor$ $H^{m}_{right} = \lceil((m+1)*H)/H_o\rceil$ $W^{n}_{left} = \lfloor(n*W)/W_o\rfloor$ $W^{n}_{right} = \lceil((n+1)*W)/W_o\rceil$ $output(N,C,l,m,n)=\underset {i \in [D^{l}_{left}, D^{l}_{right}],j\in [H^m_{left},H^m_{right}], k \in [W^n_{left},W^n_{right}] }{max} input(N,C,i,j,k)$ $indices(N,C,l,m,n)=\underset {i \in [D^{l}_{left}, D^{l}_{right}],j\in [H^m_{left},H^m_{right}], k \in [W^n_{left},W^n_{right}] }{argmax} input(N,C,i,j,k)$

Function Prototype

Each operator is divided into a two-phase API. The "aclnnAdaptiveMaxPool3dGetWorkspaceSize" API must be called first to obtain the workspace size required for computation and the executor that encapsulates the operator computation flow. Then, the "aclnnAdaptiveMaxPool3d" API is called to perform the computation.

aclnnStatus aclnnAdaptiveMaxPool3dGetWorkspaceSize(
  const aclTensor   *self,
  const aclIntArray *outputSize,
  aclTensor         *outputOut,
  aclTensor         *indicesOut,
  uint64_t          *workspaceSize,
  aclOpExecutor     **executor)

aclnnStatus aclnnAdaptiveMaxPool3d(
  void          *workspace,
  uint64_t       workspaceSize,
  aclOpExecutor *executor,
  aclrtStream    stream)

aclnnAdaptiveMaxPool3dGetWorkspaceSize

Parameters:

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (shape)	Non-contiguous Tensor
self	Input	Tensor to be computed.	The product of the three dimensions D, H, and W (DHW) must not exceed the maximum representation of int32, and must have the same data type as outputOut.	BFLOAT16, FLOAT16, FLOAT32	NCHW, NCDHW	4-5	√
outputSize	input	Output window size.	Indicates the spatial size of the output in the Dout, Hout, and Wout dimensions.	-	INT32	-	-
outputOut	Output	Pooled result.	Has the same data type as self, and the same shape as indicesOut.	BFLOAT16, FLOAT16, FLOAT32	NCHW, NCDHW	4-5	√
indicesOut	Output	Index position of each outputOut element in the input self.	Shape is the same as outputOut.	INT32, INT64	NCHW, NCDHW	4-5	√
workspaceSize	Output	Returns the workspace size to be allocated on the Device side.	-	-	-	-	-
executor	Output	Returns the operator executor, which contains the operator computation flow.	-	-	-	-	-

Atlas A3 training products/Atlas A3 inference products and Atlas A2 training products/Atlas A2 inference products: The data type of the parameter indicesOut does not support INT64,

Return value

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

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 passed self, outputSize, outputOut, or indicesOut is a null pointer.
ACLNN_ERR_PARAM_INVALID	161002	The data type, shape, format, or parameter value of self, outputOut, or indicesOut is out of the supported range.
		The shape or parameter value of outputSize is out of the supported range.
		The data types of self and outputOut are inconsistent.
		The shapes of outputOut and indicesOut are inconsistent.
		The platform is not supported.
		depth * height * width > max int32, which exceeds the representation range of indices.

aclnnAdaptiveMaxPool3d

Parameters:

Parameter	Input/Output	Description
workspace	Input.	Workspace memory address applied for on the Device side.
workspaceSize	Input.	Workspace size applied for on the Device side, obtained through the first-phase API aclnnAdaptiveMaxPool3dGetWorkspaceSize.
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 Code.

Constraints

Deterministic computation:
aclnnAdaptiveMaxPool3d defaults to a deterministic implementation.
Shape description:
self.shape = (N, C, Din, Hin, Win)
- outputSize = [Dout, Hout, Wout]
- outputOut.shape = (N, C, Dout, Hout, Wout)
- indicesOut.shape = (N, C, Dout, Hout, Wout)

Example

The following provides example code for reference only. For details about compilation and running, see Compile and Run Samples.

#include <cstdio>
#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_adaptive_max_pool3d.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) {
  // Boilerplate: resource initialization.
  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 apply for memory on the device side.
  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 data from the host to the device 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 the 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_NCDHW,
                            shape.data(), shape.size(), *deviceAddr);
  return 0;
}

int main() {
  // 1. (Boilerplate) Initialize the device/stream. See the ACL API manual.
  // Set deviceId based on the 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. Customize the construction based on the API interface.
  std::vector<int64_t> selfShape = {1, 1, 1, 4, 4};
  std::vector<int64_t> outShape = {1, 1, 1, 2, 2};
  void* selfDeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  void* indDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* out = nullptr;
  aclTensor* indices = nullptr;
  std::vector<float> selfHostData = {0, 1, 2, 3, 4.1, 5, 6, 7,
                                     8, 9, 10, 11, 12, 13, 14, 15};
  std::vector<float> outHostData = {0, 0, 0, 0.0};
  std::vector<int32_t> indicesHostData = {0, 0, 0, 0};

  //Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  //Create the out aclTensor.
  ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT, &out);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  //Create the indices aclTensor.
  ret = CreateAclTensor(indicesHostData, outShape, &indDeviceAddr, aclDataType::ACL_INT32, &indices);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  std::vector<int64_t> arraySize = {1, 2, 2};
  const aclIntArray *outputSize = aclCreateIntArray(arraySize.data(), arraySize.size());
  CHECK_RET(outputSize != nullptr, return ACL_ERROR_INTERNAL_ERROR);

  // 3. Call the CANN operator library API. Replace with the specific API name.
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // Call the first-phase API of aclnnAdaptiveMaxPool3d.
  ret = aclnnAdaptiveMaxPool3dGetWorkspaceSize(self, outputSize, out, indices, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAdaptiveMaxPool3dGetWorkspaceSize 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 > 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 second-phase API of aclnnAdaptiveMaxPool3d.
  ret = aclnnAdaptiveMaxPool3d(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnAdaptiveMaxPool3d failed. ERROR: %d\n", ret); return ret);

  // 4. (Boilerplate) Synchronize and 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 and copy the result from the device memory to the host. Modify based on the specific API definition.
  auto size = GetShapeSize(outShape);
  std::vector<float> outData(size, 0);
  std::vector<int32_t> indicesData(size, 0);
  ret = aclrtMemcpy(outData.data(), outData.size() * sizeof(outData[0]), outDeviceAddr,
                    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);
  ret = aclrtMemcpy(indicesData.data(), indicesData.size() * sizeof(indicesData[0]), indDeviceAddr,
                    size * sizeof(indicesData[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 ret);

  for (int64_t i = 0; i < size; i++) {
    LOG_PRINT("out[%ld] is: %f\n", i, outData[i]);
  }

  // 6. Release aclTensor. Modify based on the specific API definition.
  aclDestroyTensor(self);
  aclDestroyTensor(out);
  aclDestroyTensor(indices);
  aclDestroyIntArray(outputSize);

  // 7. Release device resources. Modify based on the specific API definition.
  aclrtFree(selfDeviceAddr);
  aclrtFree(outDeviceAddr);
  aclrtFree(indDeviceAddr);
  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();

  return 0;
}