Reducecalculation#
Reducecalculation
Enumerations | |
enum | miopenReduceCalculationOp_t { MIOPEN_REDUCE_CALCULATION_PROD , MIOPEN_REDUCE_CALCULATION_SUM } |
Functions | |
miopenStatus_t | miopenGetReduceCalculationWorkspaceSize (miopenHandle_t handle, const miopenTensorDescriptor_t xDesc, const int32_t dim, const miopenReduceCalculationOp_t reduceCalculationOp, const miopenTensorDescriptor_t reduceDesc, size_t *sizeInBytes) |
Helper function to query the minimum workspace size required by the ReduceTensor call. More... | |
miopenStatus_t | miopenReduceCalculationForward (miopenHandle_t handle, miopenReduceCalculationNanPropagation_t nanPropagation, void *workspace, size_t workspaceSizeInBytes, const miopenTensorDescriptor_t xDesc, const void *x, const int32_t dim, const miopenReduceCalculationOp_t reduceCalculationOp, const miopenTensorDescriptor_t reduceDesc, void *y) |
Execute a reducecalculation forward layer. More... | |
Detailed Description
Enumeration Type Documentation
◆ miopenReduceCalculationOp_t
Function Documentation
◆ miopenGetReduceCalculationWorkspaceSize()
miopenStatus_t miopenGetReduceCalculationWorkspaceSize | ( | miopenHandle_t | handle, |
const miopenTensorDescriptor_t | xDesc, | ||
const int32_t | dim, | ||
const miopenReduceCalculationOp_t | reduceCalculationOp, | ||
const miopenTensorDescriptor_t | reduceDesc, | ||
size_t * | sizeInBytes | ||
) |
Helper function to query the minimum workspace size required by the ReduceTensor call.
- Parameters
-
[in] handle MIOpen Handle [in] xDesc Tensor descriptor for data input tensor x [in] dim Dimension to calculation. [in] yDesc Tensor descriptor for output data tensor y [out] sizeInBytes Pointer to data to return the minimum workspace size
- Returns
- miopenStatus_t
◆ miopenReduceCalculationForward()
miopenStatus_t miopenReduceCalculationForward | ( | miopenHandle_t | handle, |
miopenReduceCalculationNanPropagation_t | nanPropagation, | ||
void * | workspace, | ||
size_t | workspaceSizeInBytes, | ||
const miopenTensorDescriptor_t | xDesc, | ||
const void * | x, | ||
const int32_t | dim, | ||
const miopenReduceCalculationOp_t | reduceCalculationOp, | ||
const miopenTensorDescriptor_t | reduceDesc, | ||
void * | y | ||
) |
Execute a reducecalculation forward layer.
- Parameters
-
[in] handle MIOpen handle [in] nanPropagation Nan number propagation mode [in] workspace Address of the allocated workspace data [in] workspaceSizeInBytes Size in bytes of the allocated workspace data [in] xDesc Tensor descriptor for data input tensor x [in] x Data tensor x [in] dim Dimension to calculation. [in] yDesc Tensor descriptor for output data tensor y [out] y Data tensor y
- Returns
- miopenStatus_t