aclnnDropout

Applicable Products

Product	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: During training, randomly sets elements in the input to zero with probability p, and scales the output by a factor of 1/(1-p).
Formula:
$out_i=\begin{cases}0,&\text { with probability }p \\\frac{1}{1-p}input_i,&\text { with probability }1-p\end{cases}$

Function Prototype

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

aclnnStatus aclnnDropoutGetWorkspaceSize(
  const aclTensor*  input,
  double            p,
  bool              train,
  int64_t           seed,
  int64_t           offset,
  aclTensor*        out,
  aclTensor*        maskOut,
  uint64_t*         workspaceSize,
  aclOpExecutor**   executor)

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

aclnnDropoutGetWorkspaceSize

Parameters:

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Shape	Non-contiguous Tensor
input (aclTensor*)	Input	Indicates the tensor to be processed, input in the formula.	Does not support null tensors.	BFLOAT16, FLOAT16, FLOAT32	ND	1-8	√
p (double)	Input	Probability of an element being set to zero, i.e., p in the formula.	The value range is [0, 1].	-	-	-	-
train (bool)	Input	Indicates whether it is a training scenario.	-	-	-	-	-
seed (int64_t)	Input	A seed for random number generation that affects the generated random number sequence.	-	-	-	-	-
offset (int64_t)	Input	Specifies the offset of the random number, which affects the position in the generated random number sequence.	-	-	-	-	-
out (aclTensor*)	Output	Output tensor, i.e., out in the formula.	Does not support null tensors. The data type must be convertible from the input data type, and the shape must match that of input.	BFLOAT16, FLOAT16, FLOAT32	ND	1-8	√
maskOut (aclTensor*)	Output	Mask data of bit type stored in UINT8 type.	Does not support null tensors. When p=0, p=1, or train=false, no processing is performed on the passed maskOut. In other scenarios, generates a mask with the given p as the zeroing probability. The shape must be (align(number of elements in input, 128)/8), indicating that the number of bits must be aligned to 128. Here, align means rounding up the number of elements in input to a multiple of 128. For example, align(1111, 128) yields 1152.	UINT8	ND	1	√
workspaceSize (uint64_t*)	Output	Returns the workspace size to be applied on the Device side.	-	-	-	-	-
executor (aclOpExecutor**)	Output	Returns the operator executor, which includes the operator computation process.	-	-	-	-	-

Atlas training products: The data type does not support BFLOAT16.

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	Passes null pointers for input, out, and maskOut.
ACLNN_ERR_PARAM_INVALID	161002	The data types of input, out, and maskOut are not within the supported range.
		The value of p is not between 0 and 1.
		The input has more than 8 dimensions.
		The shapes of input and out are inconsistent.
		The shape of maskOut does not meet the condition.

aclnnDropout

Parameters:

Parameter	Input/Output	Description
workspace	Input	Workspace memory address applied for on the Device side.
workspaceSize	Input	Specifies the workspace size applied for on the Device side, obtained from the first-phase API aclnnDropoutGetWorkspaceSize.
executor	Input	Operator executor that contains the operator computation flow.
stream	Input	Specifies the stream on which the task is executed.

Return value

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

Constraints

Deterministic computation:
- aclnnDropout 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_dropout.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: initializes resources.
  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 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 side 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 strides for 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 aclCreateTensor 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 manual.
  // Specify 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. Customize the construction based on the API.
  std::vector<int64_t> selfShape = {4, 2};
  std::vector<int64_t> outShape = {4, 2};
  std::vector<int64_t> maskOutShape = {16};
  void* selfDeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  void* maskOutDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* out = nullptr;
  aclTensor* maskOut = nullptr;
  std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
  std::vector<float> outHostData(8, 0);
  std::vector<uint8_t> maskOutHostData(16, 0);
  double p = 0.5;
  bool train = true;
  int64_t seed = 1234;
  int64_t offset = 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 maskOut aclTensor.
  ret = CreateAclTensor(maskOutHostData, maskOutShape, &maskOutDeviceAddr, aclDataType::ACL_UINT8, &maskOut);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 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 aclnnDropout.
  ret = aclnnDropoutGetWorkspaceSize(self, p, train, seed, offset, out, maskOut, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnDropoutGetWorkspaceSize 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 aclnnDropout second-phase API.
  ret = aclnnDropout(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnDropout 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 side memory to the host side. Modify based on the specific API definition.
  auto size = GetShapeSize(outShape);
  std::vector<float> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), outDeviceAddr,
                    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 ret);
  for (int64_t i = 0; i < size; i++) {
    LOG_PRINT("result[%ld] is: %f\n", i, resultData[i]);
  }

  // 6. Release aclTensor and aclScalar. Modify based on the specific API definition.
  aclDestroyTensor(self);
  aclDestroyTensor(out);
  aclDestroyTensor(maskOut);

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