1. Introduction

Attention: This is the cuDNN 8.0.1 Preview release. This Preview release is for early testing and feedback, therefore, for production use of cuDNN, continue to use cuDNN 7.6.5. This release is subject to change based on ongoing performance tuning and functional testing. For feedback on the new backend API and deprecations, email cudnn@nvidia.com.

NVIDIA® CUDA® Deep Neural Network library™ (cuDNN) offers a context-based API that allows for easy multithreading and (optional) interoperability with NVIDIA® CUDA® streams. The cuDNN Datatypes Reference API describes all the types and enums of the cuDNN library API. The cuDNN API Reference describes the API of all the routines in the cuDNN library.

The cuDNN library as well as this API document has been split into the following libraries:
  • cudnn_ops_infer - This entity contains the routines related to cuDNN context creation and destruction, tensor descriptor management, tensor utility routines, and the inference portion of common ML algorithms such as batch normalization, softmax, dropout, etc.
  • cudnn_ops_train - This entity contains common training routines and algorithms, such as batch normalization, softmax, dropout, etc. The cudnn_ops_train library depends on cudnn_ops_infer.
  • cudnn_cnn_infer - This entity contains all routines related to convolutional neural networks needed at inference time. The cudnn_cnn_infer library depends on cudnn_ops_infer.
  • cudnn_cnn_train - This entity contains all routines related to convolutional neural networks needed during training time. The cudnn_cnn_train library depends on cudnn_ops_infer, cudnn_ops_train, and cudnn_cnn_infer.
  • cudnn_adv_infer - This entity contains all other features and algorithms. This includes RNNs, CTC loss, and Multihead Attention. The cudnn_adv_infer library depends on cudnn_ops_infer.
  • cudnn_adv_train - This entity contains all the training counterparts of cudnn_adv_infer. The cudnn_adv_train library depends on cudnn_ops_infer, cudnn_ops_train, and cudnn_adv_infer.
  • cudnn - This is an optional shim layer between the application layer and the cuDNN code. This layer opportunistically opens the correct library for the API at runtime.

2. Added, Deprecated, And Removed API Functions

2.1. API Changes For cuDNN 8.0.0 Preview

The following tables show which API functions were added, deprecated, and removed for the cuDNN 8.0.0 Preview Release.
Table 1. API functions and data types that were added
New functions and data types
cudnnAdvInferVersionCheck()
cudnnAdvTrainVersionCheck()
cudnnBackendAttributeName_t
cudnnBackendAttributeType_t
cudnnBackendCreateDescriptor()
cudnnBackendDescriptor_t
cudnnBackendDescriptorType_t
cudnnBackendDestroyDescriptor()
cudnnBackendExecute()
cudnnBackendFinalize()
cudnnBackendGetAttribute()
cudnnBackendHeurMode_t
cudnnBackendInitialize()
cudnnBackendKnobType_t
cudnnBackendLayoutType_t
cudnnBackendNumericalNote_t
cudnnBackendSetAttribute()
cudnnBuildRNNDynamic()
cudnnCTCLoss_v8()
cudnnDeriveNormTensorDescriptor()
cudnnForwardMode_t
cudnnGenStatsMode_t
cudnnGetCTCLossDescriptor_v8()
cudnnGetCTCLossDescriptorEx()
cudnnGetCTCLossWorkspaceSize_v8
cudnnGetFilterSizeInBytes()
cudnnGetFoldedConvBackwardDataDescriptors()
cudnnGetNormalizationBackwardWorkspaceSize()
cudnnGetNormalizationForwardTrainingWorkspaceSize()
cudnnGetNormalizationTrainingReserveSpaceSize()
cudnnGetRNNDescriptor_v8()
cudnnGetRNNMatrixMathType()
cudnnGetRNNTempSpaceSizes()
cudnnGetRNNWeightParams()
cudnnGetRNNWeightSpaceSize()
cudnnLRNDescriptor_t
cudnnNormAlgo_t
cudnnNormalizationBackward()
cudnnNormalizationForwardInference()
cudnnNormalizationForwardTraining()
cudnnNormMode_t
cudnnNormOps_t
cudnnOpsInferVersionCheck()
cudnnOpsTrainVersionCheck()
cudnnPointwiseMode_t
cudnnRNNBackwardData_v8()
cudnnRNNBackwardWeights_v8()
cudnnRNNForward()
cudnnRNNGetClip_v8()
cudnnRNNSetClip_v8()
cudnnSetCTCLossDescriptor_v8()
cudnnSetRNNDescriptor_v8()
cudnnSeverity_t
For our deprecation policy, refer to the Backward Compatibility And Deprecation Policy section in the cuDNN Developer Guide.
Table 2. API functions and data types that were deprecated
Deprecated functions and data types Replaced with
cudnnCopyAlgorithmDescriptor()  
cudnnCreateAlgorithmDescriptor()  
cudnnCreatePersistentRNNPlan() cudnnBuildRNNDynamic()
cudnnDestroyAlgorithmDescriptor()  
cudnnDestroyPersistentRNNPlan()  
cudnnFindRNNBackwardDataAlgorithmEx()  
cudnnFindRNNBackwardWeightsAlgorithmEx()  
cudnnFindRNNForwardInferenceAlgorithmEx()  
cudnnFindRNNForwardTrainingAlgorithmEx()  
cudnnGetAlgorithmDescriptor()  
cudnnGetAlgorithmPerformance()  
cudnnGetAlgorithmSpaceSize()  
cudnnGetRNNBackwardDataAlgorithmMaxCount()  
cudnnGetRNNBackwardWeightsAlgorithmMaxCount()  
  • cudnnGetRNNDescriptor_v6()
  • cudnnGetRNNMatrixMathType()
  • cudnnGetRNNBiasMode()
  • cudnnGetRNNPaddingMode()
  • cudnnGetRNNProjectionLayers()
cudnnGetRNNDescriptor_v8()
cudnnGetRNNForwardInferenceAlgorithmMaxCount()  
cudnnGetRNNForwardTrainingAlgorithmMaxCount()  
  • cudnnGetRNNLinLayerBiasParams()
  • cudnnGetRNNLinLayerMatrixParams()
cudnnGetRNNWeightParams()
cudnnGetRNNParamsSize() cudnnGetRNNWeightSpaceSize()
  • cudnnGetRNNWorkspaceSize()
  • cudnnGetRNNTrainingReserveSize()
cudnnGetRNNTempSpaceSizes()
cudnnPersistentRNNPlan_t  
cudnnRestoreAlgorithm()  
  • cudnnRNNBackwardData()
  • cudnnRNNBackwardDataEx()
cudnnRNNBackwardData_v8()
  • cudnnRNNBackwardWeights()
  • cudnnRNNBackwardWeightsEx()
cudnnRNNBackwardWeights_v8()
  • cudnnRNNForwardInference()
  • cudnnRNNForwardInferenceEx()
  • cudnnRNNForwardTraining()
  • cudnnRNNForwardTrainingEx()
cudnnRNNForward()
cudnnRNNGetClip() cudnnRNNGetClip_v8()
cudnnRNNSetClip() cudnnRNNSetClip_v8()
cudnnSaveAlgorithm()  
cudnnSetAlgorithmDescriptor()  
cudnnSetAlgorithmPerformance()  
cudnnSetPersistentRNNPlan()  
cudnnSetRNNAlgorithmDescriptor()  
  • cudnnSetRNNBiasMode()
  • cudnnSetRNNDescriptor_v6()
  • cudnnSetRNNMatrixMathType()
  • cudnnSetRNNPaddingMode()
  • cudnnSetRNNProjectionLayers()
cudnnSetRNNDescriptor_v8()
Table 3. API functions and data types that were removed
Removed functions and data types
cudnnConvolutionBwdDataPreference_t
cudnnConvolutionBwdFilterPreference_t
cudnnConvolutionFwdPreference_t
cudnnGetConvolutionBackwardDataAlgorithm()
cudnnGetConvolutionBackwardFilterAlgorithm()
cudnnGetConvolutionForwardAlgorithm()
cudnnGetRNNDescriptor()
cudnnSetRNNDescriptor()

3. cudnn_ops_infer.so Library

3.1. Data Type References

3.1.1. Pointer To Opaque Struct Types

3.1.1.1. cudnnActivationDescriptor_t

cudnnActivationDescriptor_t is a pointer to an opaque structure holding the description of an activation operation. cudnnCreateActivationDescriptor() is used to create one instance, and cudnnSetActivationDescriptor() must be used to initialize this instance.

3.1.1.2. cudnnCTCLossDescriptor_t

cudnnCTCLossDescriptor_t is a pointer to an opaque structure holding the description of a CTC loss operation. cudnnCreateCTCLossDescriptor() is used to create one instance, cudnnSetCTCLossDescriptor() is used to initialize this instance, and cudnnDestroyCTCLossDescriptor() is used to destroy this instance.

3.1.1.3. cudnnDropoutDescriptor_t

cudnnDropoutDescriptor_t is a pointer to an opaque structure holding the description of a dropout operation. cudnnCreateDropoutDescriptor() is used to create one instance, cudnnSetDropoutDescriptor() is used to initialize this instance, cudnnDestroyDropoutDescriptor() is used to destroy this instance, cudnnGetDropoutDescriptor() is used to query fields of a previously initialized instance, cudnnRestoreDropoutDescriptor() is used to restore an instance to a previously saved off state.

3.1.1.4. cudnnFilterDescriptor_t

cudnnFilterDescriptor_t is a pointer to an opaque structure holding the description of a filter dataset. cudnnCreateFilterDescriptor() is used to create one instance, and cudnnSetFilter4dDescriptor() or cudnnSetFilterNdDescriptor() must be used to initialize this instance.

3.1.1.5. cudnnHandle_t

cudnnHandle_t is a pointer to an opaque structure holding the cuDNN library context. The cuDNN library context must be created using cudnnCreate() and the returned handle must be passed to all subsequent library function calls. The context should be destroyed at the end using cudnnDestroy(). The context is associated with only one GPU device, the current device at the time of the call to cudnnCreate(). However, multiple contexts can be created on the same GPU device.

3.1.1.6. cudnnLRNDescriptor_t

cudnnLRNDescriptor_t is a pointer to an opaque structure holding the parameters of a local response normalization. cudnnCreateLRNDescriptor() is used to create one instance, and the routine cudnnSetLRNDescriptor() must be used to initialize this instance.

3.1.1.7. cudnnOpTensorDescriptor_t

cudnnOpTensorDescriptor_t is a pointer to an opaque structure holding the description of a Tensor Core operation, used as a parameter to cudnnOpTensor(). cudnnCreateOpTensorDescriptor() is used to create one instance, and cudnnSetOpTensorDescriptor() must be used to initialize this instance.

3.1.1.8. cudnnPoolingDescriptor_t

cudnnPoolingDescriptor_t is a pointer to an opaque structure holding the description of a pooling operation. cudnnCreatePoolingDescriptor() is used to create one instance, and cudnnSetPoolingNdDescriptor() or cudnnSetPooling2dDescriptor() must be used to initialize this instance.

3.1.1.9. cudnnReduceTensorDescriptor_t

cudnnReduceTensorDescriptor_t is a pointer to an opaque structure holding the description of a tensor reduction operation, used as a parameter to cudnnReduceTensor(). cudnnCreateReduceTensorDescriptor() is used to create one instance, and cudnnSetReduceTensorDescriptor() must be used to initialize this instance.

3.1.1.10. cudnnSpatialTransformerDescriptor_t

cudnnSpatialTransformerDescriptor_t is a pointer to an opaque structure holding the description of a spatial transformation operation. cudnnCreateSpatialTransformerDescriptor() is used to create one instance, cudnnSetSpatialTransformerNdDescriptor() is used to initialize this instance, and cudnnDestroySpatialTransformerDescriptor() is used to destroy this instance.

3.1.1.11. cudnnTensorDescriptor_t

cudnnCreateTensorDescriptor_t is a pointer to an opaque structure holding the description of a generic n-D dataset. cudnnCreateTensorDescriptor() is used to create one instance, and one of the routines cudnnSetTensorNdDescriptor(), cudnnSetTensor4dDescriptor() or cudnnSetTensor4dDescriptorEx() must be used to initialize this instance.

3.1.1.12. cudnnTensorTransformDescriptor_t

cudnnTensorTransformDescriptor_t is an opaque structure containing the description of the tensor transform. Use the cudnnCreateTensorTransformDescriptor() function to create an instance of this descriptor, and cudnnDestroyTensorTransformDescriptor() function to destroy a previously created instance.

3.1.2. Enumeration Types

3.1.2.1. cudnnActivationMode_t

cudnnActivationMode_t is an enumerated type used to select the neuron activation function used in cudnnActivationForward(), cudnnActivationBackward(), and cudnnConvolutionBiasActivationForward().

Values
CUDNN_ACTIVATION_SIGMOID

Selects the sigmoid function.

CUDNN_ACTIVATION_RELU

Selects the rectified linear function.

CUDNN_ACTIVATION_TANH

Selects the hyperbolic tangent function.

CUDNN_ACTIVATION_CLIPPED_RELU

Selects the clipped rectified linear function.

CUDNN_ACTIVATION_ELU

Selects the exponential linear function.

CUDNN_ACTIVATION_IDENTITY

Selects the identity function, intended for bypassing the activation step in cudnnConvolutionBiasActivationForward(). (The cudnnConvolutionBiasActivationForward() function must use CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_​PRECOMP_GEMM.) Does not work with cudnnActivationForward() or cudnnActivationBackward().

3.1.2.2. cudnnAlgorithm_t

This function has been deprecated in cuDNN 8.0.

3.1.2.3. cudnnBatchNormMode_t

cudnnBatchNormMode_t is an enumerated type used to specify the mode of operation in cudnnBatchNormalizationForwardInference(), cudnnBatchNormalizationForwardTraining(), cudnnBatchNormalizationBackward() and cudnnDeriveBNTensorDescriptor() routines.

Values
CUDNN_BATCHNORM_PER_ACTIVATION

Normalization is performed per-activation. This mode is intended to be used after the non-convolutional network layers. In this mode, the tensor dimensions of bnBias and bnScale and the parameters used in the cudnnBatchNormalization* functions, are 1xCxHxW.

CUDNN_BATCHNORM_SPATIAL

Normalization is performed over N+spatial dimensions. This mode is intended for use after convolutional layers (where spatial invariance is desired). In this mode the bnBias and bnScale tensor dimensions are 1xCx1x1.

CUDNN_BATCHNORM_SPATIAL_PERSISTENT

This mode is similar to CUDNN_BATCHNORM_SPATIAL but it can be faster for some tasks.

An optimized path may be selected for CUDNN_DATA_FLOAT and CUDNN_DATA_HALF types, compute capability 6.0 or higher for the following two batch normalization API calls: cudnnBatchNormalizationForwardTraining(), and cudnnBatchNormalizationBackward(). In the case of cudnnBatchNormalizationBackward(), the savedMean and savedInvVariance arguments should not be NULL.

The rest of this section applies to NCHW mode only:

This mode may use a scaled atomic integer reduction that is deterministic but imposes more restrictions on the input data range. When a numerical overflow occurs, the algorithm may produce NaN-s or Inf-s (infinity) in output buffers.

When Inf-s/NaN-s are present in the input data, the output in this mode is the same as from a pure floating-point implementation.

For finite but very large input values, the algorithm may encounter overflows more frequently due to a lower dynamic range and emit Inf-s/NaN-s while CUDNN_BATCHNORM_SPATIAL will produce finite results. The user can invoke cudnnQueryRuntimeError() to check if a numerical overflow occurred in this mode.

3.1.2.4. cudnnBatchNormOps_t

cudnnBatchNormOps_t is an enumerated type used to specify the mode of operation in cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize(), cudnnBatchNormalizationForwardTrainingEx(), cudnnGetBatchNormalizationBackwardExWorkspaceSize(), cudnnBatchNormalizationBackwardEx(), and cudnnGetBatchNormalizationTrainingExReserveSpaceSize() functions.

Values
CUDNN_BATCHNORM_OPS_BN

Only batch normalization is performed, per-activation.

CUDNN_BATCHNORM_OPS_BN_ACTIVATION

First, the batch normalization is performed, and then the activation is performed.

CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION

Performs the batch normalization, then element-wise addition, followed by the activation operation.

3.1.2.5. cudnnCTCLossAlgo_t

cudnnCTCLossAlgo_t is an enumerated type that exposes the different algorithms available to execute the CTC loss operation.

Values
CUDNN_CTC_LOSS_ALGO_DETERMINISTIC

Results are guaranteed to be reproducible.

CUDNN_CTC_LOSS_ALGO_NON_DETERMINISTIC

Results are not guaranteed to be reproducible.

3.1.2.6. cudnnDataType_t

cudnnDataType_t is an enumerated type indicating the data type to which a tensor descriptor or filter descriptor refers.

Values
CUDNN_DATA_FLOAT

The data is a 32-bit single-precision floating-point (float).

CUDNN_DATA_DOUBLE

The data is a 64-bit double-precision floating-point (double).

CUDNN_DATA_HALF

The data is a 16-bit floating-point.

CUDNN_DATA_INT8

The data is an 8-bit signed integer.

CUDNN_DATA_UINT8

The data is an 8-bit unsigned integer.

CUDNN_DATA_INT32

The data is a 32-bit signed integer.

CUDNN_DATA_INT8x4

The data is 32-bit elements each composed of 4 8-bit signed integers. This data type is only supported with tensor format CUDNN_TENSOR_NCHW_VECT_C.

CUDNN_DATA_INT8x32

The data is 32-element vectors, each element being an 8-bit signed integer. This data type is only supported with the tensor format CUDNN_TENSOR_NCHW_VECT_C. Moreover, this data type can only be used with algo 1, meaning, CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_​PRECOMP_GEMM. For more information, see cudnnConvolutionFwdAlgo_t.

CUDNN_DATA_UINT8x4

The data is 32-bit elements each composed of 4 8-bit unsigned integers. This data type is only supported with tensor format CUDNN_TENSOR_NCHW_VECT_C.

3.1.2.7. cudnnDeterminism_t

cudnnDeterminism_t is an enumerated type used to indicate if the computed results are deterministic (reproducible). For more information, see Reproducibility (determinism).

Values
CUDNN_NON_DETERMINISTIC

Results are not guaranteed to be reproducible.

CUDNN_DETERMINISTIC

Results are guaranteed to be reproducible.

3.1.2.8. cudnnDivNormMode_t

cudnnDivNormMode_t is an enumerated type used to specify the mode of operation in cudnnDivisiveNormalizationForward() and cudnnDivisiveNormalizationBackward().

Values
CUDNN_DIVNORM_PRECOMPUTED_MEANS
The means tensor data pointer is expected to contain means or other kernel convolution values precomputed by the user. The means pointer can also be NULL, in that case, it's considered to be filled with zeroes. This is equivalent to spatial LRN.
Note: In the backward pass, the means are treated as independent inputs and the gradient over means is computed independently. In this mode, to yield a net gradient over the entire LCN computational graph, the destDiffMeans result should be backpropagated through the user's means layer (which can be implemented using average pooling) and added to the destDiffData tensor produced by cudnnDivisiveNormalizationBackward().

3.1.2.9. cudnnErrQueryMode_t

cudnnErrQueryMode_t is an enumerated type passed to cudnnQueryRuntimeError() to select the remote kernel error query mode.

Values
CUDNN_ERRQUERY_RAWCODE

Read the error storage location regardless of the kernel completion status.

CUDNN_ERRQUERY_NONBLOCKING

Report if all tasks in the user stream of the cuDNN handle were completed. If that is the case, report the remote kernel error code.

CUDNN_ERRQUERY_BLOCKING

Wait for all tasks to complete in the user stream before reporting the remote kernel error code.

3.1.2.10. cudnnFoldingDirection_t

cudnnFoldingDirection_t is an enumerated type used to select the folding direction. For more information, see cudnnTensorTransformDescriptor_t.

Data Member
CUDNN_TRANSFORM_FOLD = 0U

Selects folding.

CUDNN_TRANSFORM_UNFOLD = 1U

Selects unfolding.

3.1.2.11. cudnnIndicesType_t

cudnnIndicesType_t is an enumerated type used to indicate the data type for the indices to be computed by the cudnnReduceTensor() routine. This enumerated type is used as a field for the cudnnReduceTensorDescriptor_t descriptor.

Values
CUDNN_32BIT_INDICES

Compute unsigned int indices.

CUDNN_64BIT_INDICES

Compute unsigned long indices.

CUDNN_16BIT_INDICES

Compute unsigned short indices.

CUDNN_8BIT_INDICES

Compute unsigned char indices.

3.1.2.12. cudnnLRNMode_t

cudnnLRNMode_t is an enumerated type used to specify the mode of operation in cudnnLRNCrossChannelForward() and cudnnLRNCrossChannelBackward().

Values
CUDNN_LRN_CROSS_CHANNEL_DIM1

LRN computation is performed across tensor's dimension dimA[1].

3.1.2.13. cudnnMathType_t

cudnnMathType_t is an enumerated type used to indicate if the use of Tensor Core operations is permitted in a given library routine.

Values
CUDNN_DEFAULT_MATH

Tensor Core operations are not used on pre-NVIDIA A100 GPU devices. On A100 GPU devices, Tensor Core TF32 operation is permitted.

CUDNN_TENSOR_OP_MATH

The use of Tensor Core operations is permitted but will not actively perform datatype down conversion on tensors in order to utilize Tensor Cores.

CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION

The use of Tensor Core operations is permitted and will actively perform datatype down conversion on tensors in order to utilize Tensor Cores.

CUDNN_FMA_MATH

Restricted to only kernels that use FMA instructions.

On pre-NVIDIA A100 GPU devices, CUDNN_DEFAULT_MATH and CUDNN_FMA_MATH have the same behavior: Tensor Core kernels will not be selected. With NVIDIA Ampere GPU architecture and CUDA Toolkit 11, CUDNN_DEFAULT_MATH permits TF32 Tensor Core operation and CUDNN_FMA_MATH does not. The TF32 behavior for CUDNN_DEFAULT_MATH can be explicitly disabled by the environment variable NVIDIA_TF32_OVERRIDE=0.

3.1.2.14. cudnnNanPropagation_t

cudnnNanPropagation_t is an enumerated type used to indicate if a given routine should propagate Nan numbers. This enumerated type is used as a field for the cudnnActivationDescriptor_t descriptor and cudnnPoolingDescriptor_t descriptor.

Values
CUDNN_NOT_PROPAGATE_NAN

Nan numbers are not propagated.

CUDNN_PROPAGATE_NAN

Nan numbers are propagated.

3.1.2.15. cudnnNormAlgo_t

cudnnNormAlgo_t is an enumerated type used to specify the algorithm to execute the normalization operation.

Values
CUDNN_NORM_ALGO_STANDARD

Standard normalization is performed.

CUDNN_NORM_ALGO_PERSIST

This mode is similar to CUDNN_NORM_ALGO_STANDARD, however it only supports CUDNN_NORM_PER_CHANNEL and can be faster for some tasks.

An optimized path may be selected for CUDNN_DATA_FLOAT and CUDNN_DATA_HALF types, compute capability 6.0 or higher for the following two normalization API calls: cudnnNormalizationForwardTraining() and cudnnNormalizationBackward(). In the case of cudnnNormalizationBackward(), the savedMean and savedInvVariance arguments should not be NULL.

The rest of this section applies to NCHW mode only: This mode may use a scaled atomic integer reduction that is deterministic but imposes more restrictions on the input data range. When a numerical overflow occurs, the algorithm may produce NaN-s or Inf-s (infinity) in output buffers.

When Inf-s/NaN-s are present in the input data, the output in this mode is the same as from a pure floating-point implementation.

For finite but very large input values, the algorithm may encounter overflows more frequently due to a lower dynamic range and emit Inf-s/NaN-s while CUDNN_NORM_ALGO_STANDARD will produce finite results. The user can invoke cudnnQueryRuntimeError() to check if a numerical overflow occurred in this mode.

3.1.2.16. cudnnNormMode_t

cudnnNormMode_t is an enumerated type used to specify the mode of operation in cudnnNormalizationForwardInference(), cudnnNormalizationForwardTraining(), cudnnBatchNormalizationBackward(), cudnnGetNormalizationForwardTrainingWorkspaceSize(), cudnnGetNormalizationBackwardWorkspaceSize(), and cudnnGetNormalizationTrainingReserveSpaceSize() routines.

Values
CUDNN_NORM_PER_ACTIVATION

Normalization is performed per-activation. This mode is intended to be used after the non-convolutional network layers. In this mode, the tensor dimensions of normBias and normScale and the parameters used in the cudnnNormalization* functions, are 1xCxHxW.

CUDNN_NORM_PER_CHANNEL

Normalization is performed per-channel over N+spatial dimensions. This mode is intended for use after convolutional layers (where spatial invariance is desired). In this mode, the normBias and normScale tensor dimensions are 1xCx1x1.

3.1.2.17. cudnnNormOps_t

cudnnNormOps_t is an enumerated type used to specify the mode of operation in cudnnGetNormalizationForwardTrainingWorkspaceSize(), cudnnNormalizationForwardTraining(), cudnnGetNormalizationBackwardWorkspaceSize(), cudnnNormalizationBackward(), and cudnnGetNormalizationTrainingReserveSpaceSize() functions.

Values
CUDNN_NORM_OPS_NORM

Only normalization is performed.

CUDNN_NORM_OPS_NORM_ACTIVATION

First, the normalization is performed, then the activation is performed.

CUDNN_NORM_OPS_NORM_ADD_ACTIVATION

Performs the normalization, then element-wise addition, followed by the activation operation.

3.1.2.18. cudnnOpTensorOp_t

cudnnOpTensorOp_t is an enumerated type used to indicate the Tensor Core operation to be used by the cudnnOpTensor() routine. This enumerated type is used as a field for the cudnnOpTensorDescriptor_t descriptor.

Values
CUDNN_OP_TENSOR_ADD

The operation to be performed is addition.

CUDNN_OP_TENSOR_MUL

The operation to be performed is multiplication.

CUDNN_OP_TENSOR_MIN

The operation to be performed is a minimum comparison.

CUDNN_OP_TENSOR_MAX

The operation to be performed is a maximum comparison.

CUDNN_OP_TENSOR_SQRT

The operation to be performed is square root, performed on only the A tensor.

CUDNN_OP_TENSOR_NOT

The operation to be performed is negation, performed on only the A tensor.

3.1.2.19. cudnnPoolingMode_t

cudnnPoolingMode_t is an enumerated type passed to cudnnSetPooling2dDescriptor() to select the pooling method to be used by cudnnPoolingForward() and cudnnPoolingBackward().

Values
CUDNN_POOLING_MAX

The maximum value inside the pooling window is used.

CUDNN_POOLING_AVERAGE_COUNT_INCLUDE_PADDING

Values inside the pooling window are averaged. The number of elements used to calculate the average includes spatial locations falling in the padding region.

CUDNN_POOLING_AVERAGE_COUNT_EXCLUDE_PADDING

Values inside the pooling window are averaged. The number of elements used to calculate the average excludes spatial locations falling in the padding region.

CUDNN_POOLING_MAX_DETERMINISTIC

The maximum value inside the pooling window is used. The algorithm used is deterministic.

3.1.2.20. cudnnReduceTensorIndices_t

cudnnReduceTensorIndices_t is an enumerated type used to indicate whether indices are to be computed by the cudnnReduceTensor() routine. This enumerated type is used as a field for the cudnnReduceTensorDescriptor_t descriptor.

Values
CUDNN_REDUCE_TENSOR_NO_INDICES

Do not compute indices.

CUDNN_REDUCE_TENSOR_FLATTENED_INDICES

Compute indices. The resulting indices are relative, and flattened.

3.1.2.21. cudnnReduceTensorOp_t

cudnnReduceTensorOp_t is an enumerated type used to indicate the Tensor Core operation to be used by the cudnnReduceTensor() routine. This enumerated type is used as a field for the cudnnReduceTensorDescriptor_t descriptor.

Values
CUDNN_REDUCE_TENSOR_ADD

The operation to be performed is addition.

CUDNN_REDUCE_TENSOR_MUL

The operation to be performed is multiplication.

CUDNN_REDUCE_TENSOR_MIN

The operation to be performed is a minimum comparison.

CUDNN_REDUCE_TENSOR_MAX

The operation to be performed is a maximum comparison.

CUDNN_REDUCE_TENSOR_AMAX

The operation to be performed is a maximum comparison of absolute values.

CUDNN_REDUCE_TENSOR_AVG

The operation to be performed is averaging.

CUDNN_REDUCE_TENSOR_NORM1

The operation to be performed is addition of absolute values.

CUDNN_REDUCE_TENSOR_NORM2

The operation to be performed is a square root of the sum of squares.

CUDNN_REDUCE_TENSOR_MUL_NO_ZEROS

The operation to be performed is multiplication, not including elements of value zero.

3.1.2.22. cudnnRNNAlgo_t

cudnnRNNAlgo_t is an enumerated type used to specify the algorithm used in the cudnnRNNForwardInference(), cudnnRNNForwardTraining(), cudnnRNNBackwardData() and cudnnRNNBackwardWeights() routines.

Values
CUDNN_RNN_ALGO_STANDARD
Each RNN layer is executed as a sequence of operations. This algorithm is expected to have robust performance across a wide range of network parameters.
CUDNN_RNN_ALGO_PERSIST_STATIC

The recurrent parts of the network are executed using a persistent kernel approach. This method is expected to be fast when the first dimension of the input tensor is small (meaning, a small minibatch).

CUDNN_RNN_ALGO_PERSIST_STATIC is only supported on devices with compute capability >= 6.0.

CUDNN_RNN_ALGO_PERSIST_DYNAMIC

The recurrent parts of the network are executed using a persistent kernel approach. This method is expected to be fast when the first dimension of the input tensor is small (meaning, a small minibatch). When using CUDNN_RNN_ALGO_PERSIST_DYNAMIC persistent kernels are prepared at runtime and are able to optimize using the specific parameters of the network and active GPU. As such, when using CUDNN_RNN_ALGO_PERSIST_DYNAMIC a one-time plan preparation stage must be executed. These plans can then be reused in repeated calls with the same model parameters.

The limits on the maximum number of hidden units supported when using CUDNN_RNN_ALGO_PERSIST_DYNAMIC are significantly higher than the limits when using CUDNN_RNN_ALGO_PERSIST_STATIC, however throughput is likely to significantly reduce when exceeding the maximums supported by CUDNN_RNN_ALGO_PERSIST_STATIC. In this regime, this method will still outperform CUDNN_RNN_ALGO_STANDARD for some cases.

CUDNN_RNN_ALGO_PERSIST_DYNAMIC is only supported on devices with compute capability >= 6.0 on Linux machines.

3.1.2.23. cudnnSamplerType_t

cudnnSamplerType_t is an enumerated type passed to cudnnSetSpatialTransformerNdDescriptor() to select the sampler type to be used by cudnnSpatialTfSamplerForward() and cudnnSpatialTfSamplerBackward().

Values
CUDNN_SAMPLER_BILINEAR
Selects the bilinear sampler.

3.1.2.24. cudnnSeverity_t

cudnnSeverity_t is an enumerated type passed to the customized callback function for logging that users may set. This enumerate describes the severity level of the item, so the customized logging call back may react differently.

Values
CUDNN_SEV_FATAL

This value indicates a fatal error emitted by cuDNN.

CUDNN_SEV_ERROR

This value indicates a normal error emitted by cuDNN.

CUDNN_SEV_WARNING

This value indicates a warning emitted by cuDNN.

CUDNN_SEV_INFO

This value indicates a piece of information (for example, API log) emitted by cuDNN.

3.1.2.25. cudnnSoftmaxAlgorithm_t

cudnnSoftmaxAlgorithm_t is used to select an implementation of the softmax function used in cudnnSoftmaxForward() and cudnnSoftmaxBackward().

Values
CUDNN_SOFTMAX_FAST

This implementation applies the straightforward softmax operation.

CUDNN_SOFTMAX_ACCURATE

This implementation scales each point of the softmax input domain by its maximum value to avoid potential floating point overflows in the softmax evaluation.

CUDNN_SOFTMAX_LOG

This entry performs the log softmax operation, avoiding overflows by scaling each point in the input domain as in CUDNN_SOFTMAX_ACCURATE.

3.1.2.26. cudnnSoftmaxMode_t

cudnnSoftmaxMode_t is used to select over which data the cudnnSoftmaxForward() and cudnnSoftmaxBackward() are computing their results.

Values
CUDNN_SOFTMAX_MODE_INSTANCE

The softmax operation is computed per image (N) across the dimensions C,H,W.

CUDNN_SOFTMAX_MODE_CHANNEL

The softmax operation is computed per spatial location (H,W) per image (N) across the dimension C.

3.1.2.27. cudnnStatus_t

cudnnStatus_t is an enumerated type used for function status returns. All cuDNN library functions return their status, which can be one of the following values:

Values
CUDNN_STATUS_SUCCESS

The operation was completed successfully.

CUDNN_STATUS_NOT_INITIALIZED

The cuDNN library was not initialized properly. This error is usually returned when a call to cudnnCreate() fails or when cudnnCreate() has not been called prior to calling another cuDNN routine. In the former case, it is usually due to an error in the CUDA Runtime API called by cudnnCreate() or by an error in the hardware setup.

CUDNN_STATUS_ALLOC_FAILED

Resource allocation failed inside the cuDNN library. This is usually caused by an internal cudaMalloc() failure.

To correct, prior to the function call, deallocate previously allocated memory as much as possible.

CUDNN_STATUS_BAD_PARAM

An incorrect value or parameter was passed to the function.

To correct, ensure that all the parameters being passed have valid values.

CUDNN_STATUS_ARCH_MISMATCH

The function requires a feature absent from the current GPU device. Note that cuDNN only supports devices with compute capabilities greater than or equal to 3.0.

To correct, compile and run the application on a device with appropriate compute capability.

CUDNN_STATUS_MAPPING_ERROR

An access to GPU memory space failed, which is usually caused by a failure to bind a texture.

To correct, prior to the function call, unbind any previously bound textures.

Otherwise, this may indicate an internal error/bug in the library.

CUDNN_STATUS_EXECUTION_FAILED

The GPU program failed to execute. This is usually caused by a failure to launch some cuDNN kernel on the GPU, which can occur for multiple reasons.

To correct, check that the hardware, an appropriate version of the driver, and the cuDNN library are correctly installed.

Otherwise, this may indicate an internal error/bug in the library.

CUDNN_STATUS_INTERNAL_ERROR

An internal cuDNN operation failed.

CUDNN_STATUS_NOT_SUPPORTED

The functionality requested is not presently supported by cuDNN.

CUDNN_STATUS_LICENSE_ERROR

The functionality requested requires some license and an error was detected when trying to check the current licensing. This error can happen if the license is not present or is expired or if the environment variable NVIDIA_LICENSE_FILE is not set properly.

CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING

The runtime library that is required by RNN calls (libcuda.so or nvcuda.dll) cannot be found in predefined search paths.

CUDNN_STATUS_RUNTIME_IN_PROGRESS

Some tasks in the user stream are not completed.

CUDNN_STATUS_RUNTIME_FP_OVERFLOW

Numerical overflow occurred during the GPU kernel execution.

3.1.2.28. cudnnTensorFormat_t

cudnnTensorFormat_t is an enumerated type used by cudnnSetTensor4dDescriptor() to create a tensor with a pre-defined layout. For a detailed explanation of how these tensors are arranged in memory, see the Data Layout Formats section in the cuDNN Developer Guide.

Values
CUDNN_TENSOR_NCHW

This tensor format specifies that the data is laid out in the following order: batch size, feature maps, rows, columns. The strides are implicitly defined in such a way that the data are contiguous in memory with no padding between images, feature maps, rows, and columns; the columns are the inner dimension and the images are the outermost dimension.

CUDNN_TENSOR_NHWC

This tensor format specifies that the data is laid out in the following order: batch size, rows, columns, feature maps. The strides are implicitly defined in such a way that the data are contiguous in memory with no padding between images, rows, columns, and feature maps; the feature maps are the inner dimension and the images are the outermost dimension.

CUDNN_TENSOR_NCHW_VECT_C

This tensor format specifies that the data is laid out in the following order: batch size, feature maps, rows, columns. However, each element of the tensor is a vector of multiple feature maps. The length of the vector is carried by the data type of the tensor. The strides are implicitly defined in such a way that the data are contiguous in memory with no padding between images, feature maps, rows, and columns; the columns are the inner dimension and the images are the outermost dimension. This format is only supported with tensor data types CUDNN_DATA_INT8x4, CUDNN_DATA_INT8x32, and CUDNN_DATA_UINT8x4.

The CUDNN_TENSOR_NCHW_VECT_C can also be interpreted in the following way: The NCHW INT8x32 format is really N x (C/32) x H x W x 32 (32 Cs for every W), just as the NCHW INT8x4 format is N x (C/4) x H x W x 4 (4 Cs for every W). Hence, the VECT_C name - each W is a vector (4 or 32) of Cs.

3.2. API Functions

3.2.1. cudnnActivationForward()

cudnnStatus_t cudnnActivationForward(
    cudnnHandle_t handle,
    cudnnActivationDescriptor_t     activationDesc,
    const void                     *alpha,
    const cudnnTensorDescriptor_t   xDesc,
    const void                     *x,
    const void                     *beta,
    const cudnnTensorDescriptor_t   yDesc,
    void                           *y)
This routine applies a specified neuron activation function element-wise over each input value.
Note:
  • In-place operation is allowed for this routine; meaning, xData and yData pointers may be equal. However, this requires xDesc and yDesc descriptors to be identical (particularly, the strides of the input and output must match for an in-place operation to be allowed).
  • All tensor formats are supported for 4 and 5 dimensions, however, the best performance is obtained when the strides of xDesc and yDesc are equal and HW-packed. For more than 5 dimensions the tensors must have their spatial dimensions packed.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

activationDesc

Input. Activation descriptor. For more information, see cudnnActivationDescriptor_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc

Input. Handle to the previously initialized input tensor descriptor. For more information, see cudnnTensorDescriptor_t.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

yDesc

Input. Handle to the previously initialized output tensor descriptor.

y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The parameter mode has an invalid enumerant value.
  • The dimensions n, c, h, w of the input tensor and output tensor differ.
  • The datatype of the input tensor and output tensor differs.
  • The strides nStride, cStride, hStride, wStride of the input tensor and output tensor differ and in-place operation is used (meaning, x and y pointers are equal).
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.2. cudnnAddTensor()

cudnnStatus_t cudnnAddTensor(
    cudnnHandle_t                     handle,
    const void                       *alpha,
    const cudnnTensorDescriptor_t     aDesc,
    const void                       *A,
    const void                       *beta,
    const cudnnTensorDescriptor_t     cDesc,
    void                             *C)

This function adds the scaled values of a bias tensor to another tensor. Each dimension of the bias tensor A must match the corresponding dimension of the destination tensor C or must be equal to 1. In the latter case, the same value from the bias tensor for those dimensions will be used to blend into the C tensor.

Note: Up to dimension 5, all tensor formats are supported. Beyond those dimensions, this routine is not supported

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with the prior value in the destination tensor as follows:
dstValue = alpha[0]*srcValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

aDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

A

Input. Pointer to data of the tensor described by the aDesc descriptor.

cDesc

Input. Handle to a previously initialized tensor descriptor.

C

Input/Output. Pointer to data of the tensor described by the cDesc descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function executed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

The dimensions of the bias tensor refer to an amount of data that is incompatible with the output tensor dimensions or the dataType of the two tensor descriptors are different.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.3. cudnnBatchNormalizationForwardInference()

 cudnnStatus_t cudnnBatchNormalizationForwardInference(
      cudnnHandle_t                    handle,
      cudnnBatchNormMode_t             mode,
      const void                      *alpha,
      const void                      *beta,
      const cudnnTensorDescriptor_t    xDesc,
      const void                      *x,
      const cudnnTensorDescriptor_t    yDesc,
      void                            *y,
      const cudnnTensorDescriptor_t    bnScaleBiasMeanVarDesc,
      const void                      *bnScale,
      const void                      *bnBias,
      const void                      *estimatedMean,
      const void                      *estimatedVariance,
      double                           epsilon)
This function performs the forward batch normalization layer computation for the inference phase. This layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015.
Note:
  • Only 4D and 5D tensors are supported.
  • The input transformation performed by this function is defined as:
    y = beta*y + alpha *[bnBias + (bnScale * (x-estimatedMean)/sqrt(epsilon + estimatedVariance)]
  • The epsilon value has to be the same during training, backpropagation and inference.
  • For the training phase, use cudnnBatchNormalizationForwardTraining().
  • Higher performance can be obtained when HW-packed tensors are used for all of x and dx.
For more information, see cudnnDeriveBNTensorDescriptor() for the secondary tensor descriptor generation for the parameters used in this function.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

alpha, beta
Inputs. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Input. Handles to the previously initialized tensor descriptors.

*x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc, for the layer’s x input data.

*y

Input. Data pointer to GPU memory associated with the tensor descriptor yDesc, for the youtput of the batch normalization layer.

bnScaleBiasMeanVarDesc, bnScale, bnBias

Inputs. Tensor descriptors and pointers in device memory for the batch normalization scale and bias parameters (in the original paper bias is referred to as beta and scale as gamma).

estimatedMean, estimatedVariance

Inputs. Mean and variance tensors (these have the same descriptor as the bias and scale). The resultRunningMean and resultRunningVariance, accumulated during the training phase from the cudnnBatchNormalizationForwardTraining() call, should be passed as inputs here.

epsilon

Input. Epsilon value used in the batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h.

Supported configurations

This function supports the following combinations of data types for various descriptors.

Table 4. Supported configurations
Data Type Configurations xDesc bnScaleBiasMeanVarDesc alpha, beta yDesc
INT8_CONFIG CUDNN_DATA_INT8 CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_INT8
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the pointers alpha, beta, x, y, bnScale, bnBias, estimatedMean, estimatedInvVariance is NULL.
  • The number of xDesc or yDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported.)
  • bnScaleBiasMeanVarDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • epsilon value is less than CUDNN_BN_MIN_EPSILON.
  • Dimensions or data types mismatch for xDesc, yDesc.

3.2.4. cudnnCopyAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

3.2.5. cudnnCreate()

cudnnStatus_t cudnnCreate(cudnnHandle_t *handle)

This function initializes the cuDNN library and creates a handle to an opaque structure holding the cuDNN library context. It allocates hardware resources on the host and device and must be called prior to making any other cuDNN library calls.

The cuDNN library handle is tied to the current CUDA device (context). To use the library on multiple devices, one cuDNN handle needs to be created for each device.

For a given device, multiple cuDNN handles with different configurations (for example, different current CUDA streams) may be created. Because cudnnCreate() allocates some internal resources, the release of those resources by calling cudnnDestroy() will implicitly call cudaDeviceSynchronize; therefore, the recommended best practice is to call cudnnCreate/cudnnDestroy outside of performance-critical code paths.

For multithreaded applications that use the same device from different threads, the recommended programming model is to create one (or a few, as is convenient) cuDNN handle(s) per thread and use that cuDNN handle for the entire life of the thread.

Parameters

handle

Output. Pointer to pointer where to store the address to the allocated cuDNN handle. For more information, see cudnnHandle_t.

Returns

CUDNN_STATUS_BAD_PARAM

Invalid (NULL) input pointer supplied.

CUDNN_STATUS_NOT_INITIALIZED

No compatible GPU found, CUDA driver not installed or disabled, CUDA runtime API initialization failed.

CUDNN_STATUS_ARCH_MISMATCH

NVIDIA GPU architecture is too old.

CUDNN_STATUS_ALLOC_FAILED

Host memory allocation failed.

CUDNN_STATUS_INTERNAL_ERROR

CUDA resource allocation failed.

CUDNN_STATUS_LICENSE_ERROR

cuDNN license validation failed (only when the feature is enabled).

CUDNN_STATUS_SUCCESS

cuDNN handle was created successfully.

3.2.6. cudnnCreateActivationDescriptor()

cudnnStatus_t cudnnCreateActivationDescriptor(
        cudnnActivationDescriptor_t   *activationDesc)

This function creates an activation descriptor object by allocating the memory needed to hold its opaque structure. For more information, see cudnnActivationDescriptor_t.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.7. cudnnCreateAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnCreateAlgorithmDescriptor(
    cudnnAlgorithmDescriptor_t *algoDesc)

This function creates an algorithm descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.8. cudnnCreateAlgorithmPerformance()

cudnnStatus_t cudnnCreateAlgorithmPerformance(
    cudnnAlgorithmPerformance_t *algoPerf,
    int                         numberToCreate)

This function creates multiple algorithm performance objects by allocating the memory needed to hold their opaque structures.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.9. cudnnCreateDropoutDescriptor()

cudnnStatus_t cudnnCreateDropoutDescriptor(
    cudnnDropoutDescriptor_t    *dropoutDesc)

This function creates a generic dropout descriptor object by allocating the memory needed to hold its opaque structure. For more information, see cudnnDropoutDescriptor_t.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.10. cudnnCreateFilterDescriptor()

cudnnStatus_t cudnnCreateFilterDescriptor(
    cudnnFilterDescriptor_t *filterDesc)

This function creates a filter descriptor object by allocating the memory needed to hold its opaque structure. For more information, see cudnnFilterDescriptor_t.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.11. cudnnCreateLRNDescriptor()

cudnnStatus_t cudnnCreateLRNDescriptor(
            cudnnLRNDescriptor_t    *poolingDesc)

This function allocates the memory needed to hold the data needed for LRN and DivisiveNormalization layers operation and returns a descriptor used with subsequent layer forward and backward calls.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

cudnnCreateOpTensorDescriptor()

cudnnStatus_t cudnnCreateOpTensorDescriptor(
    cudnnOpTensorDescriptor_t* 	opTensorDesc)

This function creates a tensor pointwise math descriptor. For more information, see cudnnOpTensorDescriptor_t.

Parameters

opTensorDesc

Output. Pointer to the structure holding the description of the tensor pointwise math such as add, multiply, and more.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

Tensor pointwise math descriptor passed to the function is invalid.

CUDNN_STATUS_ALLOC_FAILED

Memory allocation for this tensor pointwise math descriptor failed.

3.2.13. cudnnCreatePoolingDescriptor()

cudnnStatus_t cudnnCreatePoolingDescriptor(
    cudnnPoolingDescriptor_t    *poolingDesc)

This function creates a pooling descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.14. cudnnCreateReduceTensorDescriptor()

cudnnStatus_t cudnnCreateReduceTensorDescriptor(
	cudnnReduceTensorDescriptor_t*	reduceTensorDesc)

This function creates a reduced tensor descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_BAD_PARAM

reduceTensorDesc is a NULL pointer.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.15. cudnnCreateSpatialTransformerDescriptor()

cudnnStatus_t cudnnCreateSpatialTransformerDescriptor(
    cudnnSpatialTransformerDescriptor_t *stDesc)

This function creates a generic spatial transformer descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

3.2.16. cudnnCreateTensorDescriptor()

cudnnStatus_t cudnnCreateTensorDescriptor(
    cudnnTensorDescriptor_t *tensorDesc)

This function creates a generic tensor descriptor object by allocating the memory needed to hold its opaque structure. The data is initialized to all zeros.

Parameters

tensorDesc

Input. Pointer to pointer where the address to the allocated tensor descriptor object should be stored.

Returns

CUDNN_STATUS_BAD_PARAM

Invalid input argument.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

CUDNN_STATUS_SUCCESS

The object was created successfully.

3.2.17. cudnnCreateTensorTransformDescriptor()

cudnnStatus_t cudnnCreateTensorTransformDescriptor(
	cudnnTensorTransformDescriptor_t *transformDesc);

This function creates a tensor transform descriptor object by allocating the memory needed to hold its opaque structure. The tensor data is initialized to be all zero. Use the cudnnSetTensorTransformDescriptor() function to initialize the descriptor created by this function.

Parameters

transformDesc
Output. A pointer to an uninitialized tensor transform descriptor.

Returns

CUDNN_STATUS_SUCCESS
The descriptor object was created successfully.
CUDNN_STATUS_BAD_PARAM
The transformDesc is NULL.
CUDNN_STATUS_ALLOC_FAILED
The memory allocation failed.

3.2.18. cudnnDeriveBNTensorDescriptor()

cudnnStatus_t cudnnDeriveBNTensorDescriptor(
    cudnnTensorDescriptor_t         derivedBnDesc,
    const cudnnTensorDescriptor_t   xDesc,
    cudnnBatchNormMode_t            mode)

This function derives a secondary tensor descriptor for the batch normalization scale, invVariance, bnBias, and bnScale subtensors from the layer's x data descriptor.

Use the tensor descriptor produced by this function as the bnScaleBiasMeanVarDesc parameter for the cudnnBatchNormalizationForwardInference() and cudnnBatchNormalizationForwardTraining() functions, and as the bnScaleBiasDiffDesc parameter in the cudnnBatchNormalizationBackward() function.

The resulting dimensions will be:
  • 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for BATCHNORM_MODE_SPATIAL
  • 1xCxHxW for 4D and 1xCxDxHxW for 5D for BATCHNORM_MODE_PER_ACTIVATION mode
For HALF input data type the resulting tensor descriptor will have a FLOAT type. For other data types, it will have the same type as the input data.
Note:
  • Only 4D and 5D tensors are supported.
  • The derivedBnDesc should be first created using cudnnCreateTensorDescriptor().
  • xDesc is the descriptor for the layer's x data and has to be set up with proper dimensions prior to calling this function.

Parameters

derivedBnDesc

Output. Handle to a previously created tensor descriptor.

xDesc

Input. Handle to a previously created and initialized layer's x data descriptor.

mode

Input. Batch normalization layer mode of operation.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

Invalid Batch Normalization mode.

3.2.19. cudnnDeriveNormTensorDescriptor()

cudnnStatus_t CUDNNWINAPI
cudnnDeriveNormTensorDescriptor(cudnnTensorDescriptor_t derivedNormScaleBiasDesc,
                                cudnnTensorDescriptor_t derivedNormMeanVarDesc,
                                const cudnnTensorDescriptor_t xDesc,
   	                         cudnnNormMode_t mode,
                                int groupCnt)

This function derives tensor descriptors for the normalization mean, invariance, normBias, and normScale subtensors from the layer's x data descriptor and norm mode. normalization, mean, and invariance share the same descriptor while bias and scale share the same descriptor.

Use the tensor descriptor produced by this function as the normScaleBiasDesc or normMeanVarDesc parameter for the cudnnNormalizationForwardInference() and cudnnNormalizationForwardTraining() functions, and as the dNormScaleBiasDesc and normMeanVarDesc parameter in the cudnnNormalizationBackward() function.

The resulting dimensions will be:
  • 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for CUDNN_NORM_PER_ACTIVATION
  • 1xCxHxW for 4D and 1xCxDxHxW for 5D for CUDNN_NORM_PER_CHANNEL mode
For HALF input data type the resulting tensor descriptor will have a FLOAT type. For other data types, it will have the same type as the input data.
  • Only 4D and 5D tensors are supported.
  • The derivedNormScaleBiasDesc and derivedNormMeanVarDesc should be created first using cudnnCreateTensorDescriptor().
  • xDesc is the descriptor for the layer's x data and has to be set up with proper dimensions prior to calling this function.

Parameters

derivedNormScaleBiasDesc

Output. Handle to a previously created tensor descriptor.

derivedNormMeanVarDesc

Output. Handle to a previously created tensor descriptor.

xDesc

Input. Handle to a previously created and initialized layer's x data descriptor.

mode

Input. The normalization layer mode of operation.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

Invalid Batch Normalization mode.

3.2.20. cudnnDestroy()

cudnnStatus_t cudnnDestroy(cudnnHandle_t handle)

This function releases the resources used by the cuDNN handle. This function is usually the last call with a particular handle to the cuDNN handle. Because cudnnCreate() allocates some internal resources, the release of those resources by calling cudnnDestroy() will implicitly call cudaDeviceSynchronize; therefore, the recommended best practice is to call cudnnCreate/cudnnDestroy outside of performance-critical code paths.

Parameters

handle

Input. Pointer to the cuDNN handle to be destroyed.

Returns

CUDNN_STATUS_SUCCESS

The cuDNN context destruction was successful.

CUDNN_STATUS_BAD_PARAM

Invalid (NULL) pointer supplied.

3.2.21. cudnnDestroyActivationDescriptor()

cudnnStatus_t cudnnDestroyActivationDescriptor(
        cudnnActivationDescriptor_t activationDesc)

This function destroys a previously created activation descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.22. cudnnDestroyAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnDestroyAlgorithmDescriptor(
        cudnnActivationDescriptor_t algorithmDesc)

This function destroys a previously created algorithm descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.23. cudnnDestroyAlgorithmPerformance()

cudnnStatus_t cudnnDestroyAlgorithmPerformance(
        cudnnAlgorithmPerformance_t     algoPerf)

This function destroys a previously created algorithm descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.24. cudnnDestroyDropoutDescriptor()

cudnnStatus_t cudnnDestroyDropoutDescriptor(
    cudnnDropoutDescriptor_t dropoutDesc)

This function destroys a previously created dropout descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.25. cudnnDestroyFilterDescriptor()

cudnnStatus_t cudnnDestroyFilterDescriptor(
    cudnnFilterDescriptor_t filterDesc)

This function destroys a previously created tensor 4D descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.26. cudnnDestroyLRNDescriptor()

cudnnStatus_t cudnnDestroyLRNDescriptor(
    cudnnLRNDescriptor_t lrnDesc)

This function destroys a previously created LRN descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.27. cudnnDestroyOpTensorDescriptor()

cudnnStatus_t cudnnDestroyOpTensorDescriptor(
    cudnnOpTensorDescriptor_t   opTensorDesc)

This function deletes a tensor pointwise math descriptor object.

Parameters

opTensorDesc

Input. Pointer to the structure holding the description of the tensor pointwise math to be deleted.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

3.2.28. cudnnDestroyPoolingDescriptor()

cudnnStatus_t cudnnDestroyPoolingDescriptor(
    cudnnPoolingDescriptor_t poolingDesc)

This function destroys a previously created pooling descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.29. cudnnDestroyReduceTensorDescriptor()

cudnnStatus_t cudnnDestroyReduceTensorDescriptor(
    cudnnReduceTensorDescriptor_t   tensorDesc)

This function destroys a previously created reduce tensor descriptor object. When the input pointer is NULL, this function performs no destroy operation.

Parameters

tensorDesc

Input. Pointer to the reduce tensor descriptor object to be destroyed.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.30. cudnnDestroySpatialTransformerDescriptor()

cudnnStatus_t cudnnDestroySpatialTransformerDescriptor(
    cudnnSpatialTransformerDescriptor_t stDesc)

This function destroys a previously created spatial transformer descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.31. cudnnDestroyTensorDescriptor()

cudnnStatus_t cudnnDestroyTensorDescriptor(cudnnTensorDescriptor_t tensorDesc)

This function destroys a previously created tensor descriptor object. When the input pointer is NULL, this function performs no destroy operation.

Parameters

tensorDesc

Input. Pointer to the tensor descriptor object to be destroyed.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

3.2.32. cudnnDestroyTensorTransformDescriptor()

cudnnStatus_t cudnnDestroyTensorTransformDescriptor(
	cudnnTensorTransformDescriptor_t transformDesc);

Destroys a previously created tensor transform descriptor.

Parameters

transformDesc
Input. The tensor transform descriptor to be destroyed.

Returns

CUDNN_STATUS_SUCCESS
The descriptor was destroyed successfully.

3.2.33. cudnnDivisiveNormalizationForward()

cudnnStatus_t cudnnDivisiveNormalizationForward(
    cudnnHandle_t                    handle,
    cudnnLRNDescriptor_t             normDesc,
    cudnnDivNormMode_t               mode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *means,
    void                            *temp,
    void                            *temp2,
    const void                      *beta,
    const cudnnTensorDescriptor_t    yDesc,
    void                            *y)
This function performs the forward spatial DivisiveNormalization layer computation. It divides every value in a layer by the standard deviation of its spatial neighbors as described in What is the Best Multi-Stage Architecture for Object Recognition, Jarrett 2009, Local Contrast Normalization Layer section. Note that DivisiveNormalization only implements the x/max(c, sigma_x) portion of the computation, where sigma_x is the variance over the spatial neighborhood of x. The full LCN (Local Contrastive Normalization) computation can be implemented as a two-step process:
x_m = x-mean(x);
y = x_m/max(c, sigma(x_m));

The x-mean(x) which is often referred to as "subtractive normalization" portion of the computation can be implemented using cuDNN average pooling layer followed by a call to addTensor.

Note: Supported tensor formats are NCHW for 4D and NCDHW for 5D with any non-overlapping non-negative strides. Only 4D and 5D tensors are supported.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

normDesc

Input. Handle to a previously initialized LRN parameter descriptor. This descriptor is used for both LRN and DivisiveNormalization layers.

divNormMode

Input. DivisiveNormalization layer mode of operation. Currently only CUDNN_DIVNORM_PRECOMPUTED_MEANS is implemented. Normalization is performed using the means input tensor that is expected to be precomputed by the user.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Input. Tensor descriptor objects for the input and output tensors. Note that xDesc is shared between x, means, temp, and temp2 tensors.

x

Input. Input tensor data pointer in device memory.

means

Input. Input means tensor data pointer in device memory. Note that this tensor can be NULL (in that case its values are assumed to be zero during the computation). This tensor also doesn't have to contain means, these can be any values, a frequently used variation is a result of convolution with a normalized positive kernel (such as Gaussian).

temp, temp2

Workspace. Temporary tensors in device memory. These are used for computing intermediate values during the forward pass. These tensors do not have to be preserved as inputs from forward to the backward pass. Both use xDesc as their descriptor.

y

Output. Pointer in device memory to a tensor for the result of the forward DivisiveNormalization computation.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the tensor pointers x, y, temp, temp2 is NULL.
  • Number of input tensor or output tensor dimensions is outside of [4,5] range.
  • A mismatch in dimensions between any two of the input or output tensors.
  • For in-place computation when pointers x == y, a mismatch in strides between the input data and output data tensors.
  • Alpha or beta pointer is NULL.
  • LRN descriptor parameters are outside of their valid ranges.
  • Any of the tensor strides are negative.
CUDNN_STATUS_UNSUPPORTED

The function does not support the provided configuration, for example, any of the input and output tensor strides mismatch (for the same dimension) is a non-supported configuration.

3.2.34. cudnnDropoutForward()

cudnnStatus_t cudnnDropoutForward(
    cudnnHandle_t                       handle,
    const cudnnDropoutDescriptor_t      dropoutDesc,
    const cudnnTensorDescriptor_t       xdesc,
    const void                         *x,
    const cudnnTensorDescriptor_t       ydesc,
    void                               *y,
    void                               *reserveSpace,
    size_t                              reserveSpaceSizeInBytes)

This function performs forward dropout operation over x returning results in y. If dropout was used as a parameter to cudnnSetDropoutDescriptor(), the approximately dropout fraction of x values will be replaced by a 0, and the rest will be scaled by 1/(1-dropout). This function should not be running concurrently with another cudnnDropoutForward() function using the same states.

Note:
  • Better performance is obtained for fully packed tensors.
  • This function should not be called during inference.

Parameters

handle

Input. Handle to a previously created cuDNN context.

dropoutDesc

Input. Previously created dropout descriptor object.

xDesc

Input. Handle to a previously initialized tensor descriptor.

x

Input. Pointer to data of the tensor described by the xDesc descriptor.

yDesc

Input. Handle to a previously initialized tensor descriptor.

y

Output. Pointer to data of the tensor described by the yDesc descriptor.

reserveSpace

Output. Pointer to user-allocated GPU memory used by this function. It is expected that the contents of reserveSpace does not change between cudnnDropoutForward() and cudnnDropoutBackward() calls.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided memory for the reserve space.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The number of elements of input tensor and output tensors differ.
  • The datatype of the input tensor and output tensors differs.
  • The strides of the input tensor and output tensors differ and in-place operation is used (meaning, x and y pointers are equal).
  • The provided reserveSpaceSizeInBytes is less than the value returned by cudnnDropoutGetReserveSpaceSize().
  • cudnnSetDropoutDescriptor() has not been called on dropoutDesc with the non-NULLstates argument.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.35. cudnnDropoutGetReserveSpaceSize()

cudnnStatus_t cudnnDropoutGetReserveSpaceSize(
    cudnnTensorDescriptor_t     xDesc,
    size_t                     *sizeInBytes)

This function is used to query the amount of reserve needed to run dropout with the input dimensions given by xDesc. The same reserve space is expected to be passed to cudnnDropoutForward() and cudnnDropoutBackward(), and its contents is expected to remain unchanged between cudnnDropoutForward() and cudnnDropoutBackward() calls.

Parameters

xDesc

Input. Handle to a previously initialized tensor descriptor, describing input to a dropout operation.

sizeInBytes

Output. Amount of GPU memory needed as reserve space to be able to run dropout with an input tensor descriptor specified by xDesc.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

3.2.36. cudnnDropoutGetStatesSize()

cudnnStatus_t cudnnDropoutGetStatesSize(
    cudnnHandle_t       handle,
    size_t             *sizeInBytes)

This function is used to query the amount of space required to store the states of the random number generators used by cudnnDropoutForward() function.

Parameters

handle

Input. Handle to a previously created cuDNN context.

sizeInBytes

Output. Amount of GPU memory needed to store random generator states.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

3.2.37. cudnnGetActivationDescriptor()

cudnnStatus_t cudnnGetActivationDescriptor(
        const cudnnActivationDescriptor_t   activationDesc,
        cudnnActivationMode_t              *mode,
        cudnnNanPropagation_t              *reluNanOpt,
        double                             *coef)

This function queries a previously initialized generic activation descriptor object.

Parameters

activationDesc

Input. Handle to a previously created activation descriptor.

mode

Output. Enumerant to specify the activation mode.

reluNanOpt

Output. Enumerant to specify the Nan propagation mode.

coef

Output. Floating point number to specify the clipping threshold when the activation mode is set to CUDNN_ACTIVATION_CLIPPED_RELU or to specify the alpha coefficient when the activation mode is set to CUDNN_ACTIVATION_ELU.

Returns

CUDNN_STATUS_SUCCESS

The object was queried successfully.

3.2.38. cudnnGetAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnGetAlgorithmDescriptor(        
        const cudnnAlgorithmDescriptor_t    algoDesc,
        cudnnAlgorithm_t                    *algorithm)

This function queries a previously initialized generic algorithm descriptor object.

Parameters

algorithmDesc

Input. Handle to a previously created algorithm descriptor.

algorithm

Input. Struct to specify the algorithm.

Returns

CUDNN_STATUS_SUCCESS

The object was queried successfully.

3.2.39. cudnnGetAlgorithmPerformance()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnGetAlgorithmPerformance(
        const cudnnAlgorithmPerformance_t   algoPerf,
        cudnnAlgorithmDescriptor_t*         algoDesc,
        cudnnStatus_t*                      status,
        float*                              time,
        size_t*                             memory)

This function queries a previously initialized generic algorithm performance object.

Parameters

algoPerf

Input/Output. Handle to a previously created algorithm performance object.

algoDesc

Output. The algorithm descriptor which the performance results describe.

status

Output. The cuDNN status returned from running the algoDesc algorithm.

timecoef

Output. The GPU time spent running the algoDesc algorithm.

memory

Output. The GPU memory needed to run the algoDesc algorithm.

Returns

CUDNN_STATUS_SUCCESS

The object was queried successfully.

3.2.40. cudnnGetAlgorithmSpaceSize()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnGetAlgorithmSpaceSize(    
    cudnnHandle_t               handle,
    cudnnAlgorithmDescriptor_t  algoDesc,
    size_t*                     algoSpaceSizeInBytes)

This function queries for the amount of host memory needed to call cudnnSaveAlgorithm(), much like the “get workspace size” function query for the amount of device memory needed.

Parameters

handle

Input. Handle to a previously created cuDNN context.

algoDesc

Input. A previously created algorithm descriptor.

algoSpaceSizeInBytes

Output. Amount of host memory needed as a workspace to be able to save the metadata from the specified algoDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the arguments is NULL.

3.2.41. cudnnGetCallback()

cudnnStatus_t cudnnGetCallback(
        unsigned            mask,
        void                **udata,
        cudnnCallback_t     fptr)

This function queries the internal states of cuDNN error reporting functionality.

Parameters

mask

Output. Pointer to the address where the current internal error reporting message bit mask will be outputted.

udata

Output. Pointer to the address where the current internally stored udata address will be stored.

fptr

Output. Pointer to the address where the current internally stored callback function pointer will be stored. When the built-in default callback function is used, NULL will be outputted.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

If any of the input parameters are NULL.

3.2.42. cudnnGetCudartVersion()

size_t cudnnGetCudartVersion()

The same version of a given cuDNN library can be compiled against different NVIDIA® CUDA® Toolkit™ versions. This routine returns the CUDA Toolkit version that the currently used cuDNN library has been compiled against.

3.2.43. cudnnGetDropoutDescriptor()

cudnnStatus_t cudnnGetDropoutDescriptor(
    cudnnDropoutDescriptor_t    dropoutDesc,
    cudnnHandle_t               handle,
    float                      *dropout,
    void                       **states,
    unsigned long long         *seed)

This function queries the fields of a previously initialized dropout descriptor.

Parameters

dropoutDesc

Input. Previously initialized dropout descriptor.

handle

Input. Handle to a previously created cuDNN context.

dropout

Output. The probability with which the value from input is set to 0 during the dropout layer.

states

Output. Pointer to user-allocated GPU memory that holds random number generator states.

seed

Output. Seed used to initialize random number generator states.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

One or more of the arguments was an invalid pointer.

3.2.44. cudnnGetErrorString()

const char * cudnnGetErrorString(cudnnStatus_t status)

This function converts the cuDNN status code to a NULL terminated (ASCIIZ) static string. For example, when the input argument is CUDNN_STATUS_SUCCESS, the returned string is CUDNN_STATUS_SUCCESS. When an invalid status value is passed to the function, the returned string is CUDNN_UNKNOWN_STATUS.

Parameters

status

Input. cuDNN enumerant status code.

Returns

Pointer to a static, NULL terminated string with the status name.

3.2.45. cudnnGetFilter4dDescriptor()

cudnnStatus_t cudnnGetFilter4dDescriptor(
    const cudnnFilterDescriptor_t     filterDesc,
    cudnnDataType_t            *dataType,
    cudnnTensorFormat_t        *format,
    int                        *k,
    int                        *c,
    int                        *h,
    int                        *w)

This function queries the parameters of the previously initialized filter descriptor object.

Parameters

filterDesc

Input. Handle to a previously created filter descriptor.

datatype

Output. Data type.

format

Output. Type of format.

k

Output. Number of output feature maps.

c

Output. Number of input feature maps.

h

Output. Height of each filter.

w

Output. Width of each filter.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

3.2.46. cudnnGetFilterNdDescriptor()

cudnnStatus_t cudnnGetFilterNdDescriptor(
    const cudnnFilterDescriptor_t   wDesc,
    int                             nbDimsRequested,
    cudnnDataType_t                *dataType,
    cudnnTensorFormat_t            *format,
    int                            *nbDims,
    int                             filterDimA[])

This function queries a previously initialized filter descriptor object.

Parameters

wDesc

Input. Handle to a previously initialized filter descriptor.

nbDimsRequested

Input. Dimension of the expected filter descriptor. It is also the minimum size of the arrays filterDimA in order to be able to hold the results

datatype

Output. Data type.

format

Output. Type of format.

nbDims

Output. Actual dimension of the filter.

filterDimA

Output. Array of dimension of at least nbDimsRequested that will be filled with the filter parameters from the provided filter descriptor.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

The parameter nbDimsRequested is negative.

3.2.47. cudnnGetFilterSizeInBytes()

cudnnStatus_t
cudnnGetFilterSizeInBytes(const cudnnFilterDescriptor_t filterDesc, size_t *size) ;

This function returns the size of the filter tensor in memory with respect to the given descriptor. It can be used to know the amount of GPU memory to be allocated to hold that filter tensor.

Parameters

filterDesc

Input. handle to a previously initialized filter descriptor.

size

Output. size in bytes needed to hold the tensor in GPU memory.

Returns

CUDNN_STATUS_SUCCESS

filterDesc is valid.

CUDNN_STATUS_BAD_PARAM

filerDesc is invald.

3.2.48. cudnnGetLRNDescriptor()

cudnnStatus_t cudnnGetLRNDescriptor(
    cudnnLRNDescriptor_t    normDesc,
    unsigned               *lrnN,
    double                 *lrnAlpha,
    double                 *lrnBeta,
    double                 *lrnK)

This function retrieves values stored in the previously initialized LRN descriptor object.

Parameters

normDesc

Output. Handle to a previously created LRN descriptor.

lrnN, lrnAlpha, lrnBeta, lrnK

Output. Pointers to receive values of parameters stored in the descriptor object. See cudnnSetLRNDescriptor() for more details. Any of these pointers can be NULL (no value is returned for the corresponding parameter).

Returns

CUDNN_STATUS_SUCCESS

Function completed successfully.

3.2.49. cudnnGetOpTensorDescriptor()

cudnnStatus_t cudnnGetOpTensorDescriptor(
    const cudnnOpTensorDescriptor_t opTensorDesc,
    cudnnOpTensorOp_t               *opTensorOp,
    cudnnDataType_t                 *opTensorCompType,
    cudnnNanPropagation_t           *opTensorNanOpt)

This function returns the configuration of the passed tensor pointwise math descriptor.

Parameters

opTensorDesc

Input. Tensor pointwise math descriptor passed to get the configuration from.

opTensorOp

Output. Pointer to the tensor pointwise math operation type, associated with this tensor pointwise math descriptor.

opTensorCompType

Output. Pointer to the cuDNN data-type associated with this tensor pointwise math descriptor.

opTensorNanOpt

Output. Pointer to the NAN propagation option associated with this tensor pointwise math descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

Input tensor pointwise math descriptor passed is invalid.

3.2.50. cudnnGetPooling2dDescriptor()

cudnnStatus_t cudnnGetPooling2dDescriptor(
    const cudnnPoolingDescriptor_t      poolingDesc,
    cudnnPoolingMode_t                 *mode,
    cudnnNanPropagation_t              *maxpoolingNanOpt,
    int                                *windowHeight,
    int                                *windowWidth,
    int                                *verticalPadding,
    int                                *horizontalPadding,
    int                                *verticalStride,
    int                                *horizontalStride)

This function queries a previously created 2D pooling descriptor object.

Parameters

poolingDesc

Input. Handle to a previously created pooling descriptor.

mode

Output. Enumerant to specify the pooling mode.

maxpoolingNanOpt

Output. Enumerant to specify the Nan propagation mode.

windowHeight

Output. Height of the pooling window.

windowWidth

Output. Width of the pooling window.

verticalPadding

Output. Size of vertical padding.

horizontalPadding

Output. Size of horizontal padding.

verticalStride

Output. Pooling vertical stride.

horizontalStride

Output. Pooling horizontal stride.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

3.2.51. cudnnGetPooling2dForwardOutputDim()

cudnnStatus_t cudnnGetPooling2dForwardOutputDim(
    const cudnnPoolingDescriptor_t      poolingDesc,
    const cudnnTensorDescriptor_t       inputDesc,
    int                                *outN,
    int                                *outC,
    int                                *outH,
    int                                *outW)

This function provides the output dimensions of a tensor after 2d pooling has been applied.

Each dimension h and w of the output images is computed as follows:
    outputDim = 1 + (inputDim + 2*padding - windowDim)/poolingStride;
    

Parameters

poolingDesc

Input. Handle to a previously initialized pooling descriptor.

inputDesc

Input. Handle to the previously initialized input tensor descriptor.

N

Output. Number of images in the output.

C

Output. Number of channels in the output.

H

Output. Height of images in the output.

W

Output. Width of images in the output.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • poolingDesc has not been initialized.
  • poolingDesc or inputDesc has an invalid number of dimensions (2 and 4 respectively are required).

3.2.52. cudnnGetPoolingNdDescriptor()

cudnnStatus_t cudnnGetPoolingNdDescriptor(
const cudnnPoolingDescriptor_t      poolingDesc,
int                                 nbDimsRequested,
cudnnPoolingMode_t                 *mode,
cudnnNanPropagation_t              *maxpoolingNanOpt,
int                                *nbDims,
int                                 windowDimA[],
int                                 paddingA[],
int                                 strideA[])

This function queries a previously initialized generic pooling descriptor object.

Parameters

poolingDesc

Input. Handle to a previously created pooling descriptor.

nbDimsRequested

Input. Dimension of the expected pooling descriptor. It is also the minimum size of the arrays windowDimA, paddingA, and strideA in order to be able to hold the results.

mode

Output. Enumerant to specify the pooling mode.

maxpoolingNanOpt

Input. Enumerant to specify the Nan propagation mode.

nbDims

Output. Actual dimension of the pooling descriptor.

windowDimA

Output. Array of dimension of at least nbDimsRequested that will be filled with the window parameters from the provided pooling descriptor.

paddingA

Output. Array of dimension of at least nbDimsRequested that will be filled with the padding parameters from the provided pooling descriptor.

strideA

Output. Array of dimension at least nbDimsRequested that will be filled with the stride parameters from the provided pooling descriptor.

Returns

CUDNN_STATUS_SUCCESS

The object was queried successfully.

CUDNN_STATUS_NOT_SUPPORTED

The parameter nbDimsRequested is greater than CUDNN_DIM_MAX.

3.2.53. cudnnGetPoolingNdForwardOutputDim()

cudnnStatus_t cudnnGetPoolingNdForwardOutputDim(
    const cudnnPoolingDescriptor_t  poolingDesc,
    const cudnnTensorDescriptor_t   inputDesc,
    int                             nbDims,
    int                             outDimA[])

This function provides the output dimensions of a tensor after Nd pooling has been applied.

Each dimension of the (nbDims-2)-D images of the output tensor is computed as follows:
    outputDim = 1 + (inputDim + 2*padding - windowDim)/poolingStride;
    

Parameters

poolingDesc

Input. Handle to a previously initialized pooling descriptor.

inputDesc

Input. Handle to the previously initialized input tensor descriptor.

nbDims

Input. Number of dimensions in which pooling is to be applied.

outDimA

Output. Array of nbDims output dimensions.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • poolingDesc has not been initialized.
  • The value of nbDims is inconsistent with the dimensionality of poolingDesc and inputDesc.

3.2.54. cudnnGetProperty()

cudnnStatus_t cudnnGetProperty(
    libraryPropertyType     type,
    int                    *value)

This function writes a specific part of the cuDNN library version number into the provided host storage.

Parameters

type

Input. Enumerant type that instructs the function to report the numerical value of the cuDNN major version, minor version, or the patch level.

value

Output. Host pointer where the version information should be written.

Returns

CUDNN_STATUS_INVALID_VALUE

Invalid value of the type argument.

CUDNN_STATUS_SUCCESS

Version information was stored successfully at the provided address.

3.2.55. cudnnGetReduceTensorDescriptor()

cudnnStatus_t cudnnGetReduceTensorDescriptor(
    const cudnnReduceTensorDescriptor_t reduceTensorDesc,
    cudnnReduceTensorOp_t               *reduceTensorOp,
    cudnnDataType_t                     *reduceTensorCompType,
    cudnnNanPropagation_t               *reduceTensorNanOpt,
    cudnnReduceTensorIndices_t          *reduceTensorIndices,
    cudnnIndicesType_t                  *reduceTensorIndicesType)

This function queries a previously initialized reduce tensor descriptor object.

Parameters

reduceTensorDesc

Input. Pointer to a previously initialized reduce tensor descriptor object.

reduceTensorOp

Output. Enumerant to specify the reduce tensor operation.

reduceTensorCompType

Output. Enumerant to specify the computation datatype of the reduction.

reduceTensorNanOpt

Input. Enumerant to specify the Nan propagation mode.

reduceTensorIndices

Output. Enumerant to specify the reduced tensor indices.

reduceTensorIndicesType

Output. Enumerant to specify the reduce tensor indices type.

Returns

CUDNN_STATUS_SUCCESS

The object was queried successfully.

CUDNN_STATUS_BAD_PARAM

reduceTensorDesc is NULL.

3.2.56. cudnnGetReductionIndicesSize()

cudnnStatus_t cudnnGetReductionIndicesSize(
    cudnnHandle_t                       handle,
    const cudnnReduceTensorDescriptor_t reduceDesc,
    const cudnnTensorDescriptor_t       aDesc,
    const cudnnTensorDescriptor_t       cDesc,
    size_t                              *sizeInBytes)

This is a helper function to return the minimum size of the index space to be passed to the reduction given the input and output tensors.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

reduceDesc

Input. Pointer to a previously initialized reduce tensor descriptor object.

aDesc

Input. Pointer to the input tensor descriptor.

cDesc

Input. Pointer to the output tensor descriptor.

sizeInBytes

Output. Minimum size of the index space to be passed to the reduction.

Returns

CUDNN_STATUS_SUCCESS

The index space size is returned successfully.

3.2.57. cudnnGetReductionWorkspaceSize()

cudnnStatus_t cudnnGetReductionWorkspaceSize(
    cudnnHandle_t                       handle,
    const cudnnReduceTensorDescriptor_t reduceDesc,
    const cudnnTensorDescriptor_t       aDesc,
    const cudnnTensorDescriptor_t       cDesc,
    size_t                              *sizeInBytes)

This is a helper function to return the minimum size of the workspace to be passed to the reduction given the input and output tensors.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

reduceDesc

Input. Pointer to a previously initialized reduce tensor descriptor object.

aDesc

Input. Pointer to the input tensor descriptor.

cDesc

Input. Pointer to the output tensor descriptor.

sizeInBytes

Output. Minimum size of the index space to be passed to the reduction.

Returns

CUDNN_STATUS_SUCCESS

The workspace size is returned successfully.

3.2.58. cudnnGetStream()

cudnnStatus_t cudnnGetStream(
    cudnnHandle_t   handle,
    cudaStream_t   *streamId)

This function retrieves the user CUDA stream programmed in the cuDNN handle. When the user's CUDA stream is not set in the cuDNN handle, this function reports the null-stream.

Parameters

handle

Input. Pointer to the cuDNN handle.

streamID

Output. Pointer where the current CUDA stream from the cuDNN handle should be stored.

Returns

CUDNN_STATUS_BAD_PARAM

Invalid (NULL) handle.

CUDNN_STATUS_SUCCESS

The stream identifier was retrieved successfully.

3.2.59. cudnnGetTensor4dDescriptor()

cudnnStatus_t cudnnGetTensor4dDescriptor(
    const cudnnTensorDescriptor_t  tensorDesc,
    cudnnDataType_t         *dataType,
    int                     *n,
    int                     *c,
    int                     *h,
    int                     *w,
    int                     *nStride,
    int                     *cStride,
    int                     *hStride,
    int                     *wStride)

This function queries the parameters of the previously initialized tensor4D descriptor object.

Parameters

tensorDesc

Input. Handle to a previously initialized tensor descriptor.

datatype

Output. Data type.

n

Output. Number of images.

c

Output. Number of feature maps per image.

h

Output. Height of each feature map.

w

Output. Width of each feature map.

nStride

Output. Stride between two consecutive images.

cStride

Output. Stride between two consecutive feature maps.

hStride

Output. Stride between two consecutive rows.

wStride

Output. Stride between two consecutive columns.

Returns

CUDNN_STATUS_SUCCESS

The operation succeeded.

3.2.60. cudnnGetTensorNdDescriptor()

cudnnStatus_t cudnnGetTensorNdDescriptor(
    const cudnnTensorDescriptor_t   tensorDesc,
    int                             nbDimsRequested,
    cudnnDataType_t                *dataType,
    int                            *nbDims,
    int                             dimA[],
    int                             strideA[])

This function retrieves values stored in a previously initialized tensor descriptor object.

Parameters

tensorDesc

Input. Handle to a previously initialized tensor descriptor.

nbDimsRequested

Input. Number of dimensions to extract from a given tensor descriptor. It is also the minimum size of the arrays dimA and strideA. If this number is greater than the resulting nbDims[0], only nbDims[0] dimensions will be returned.

datatype

Output. Data type.

nbDims

Output. Actual number of dimensions of the tensor will be returned in nbDims[0].

dimA

Output. Array of dimensions of at least nbDimsRequested that will be filled with the dimensions from the provided tensor descriptor.

strideA

Input. Array of dimension of at least nbDimsRequested that will be filled with the strides from the provided tensor descriptor.

Returns

CUDNN_STATUS_SUCCESS

The results were returned successfully.

CUDNN_STATUS_BAD_PARAM

Either tensorDesc or nbDims pointer is NULL.

3.2.61. cudnnGetTensorSizeInBytes()

cudnnStatus_t cudnnGetTensorSizeInBytes(
    const cudnnTensorDescriptor_t   tensorDesc,
    size_t                         *size)

This function returns the size of the tensor in memory in respect to the given descriptor. This function can be used to know the amount of GPU memory to be allocated to hold that tensor.

Parameters

tensorDesc

Input. Handle to a previously initialized tensor descriptor.

size

Output. Size in bytes needed to hold the tensor in GPU memory.

Returns

CUDNN_STATUS_SUCCESS

The results were returned successfully.

3.2.62. cudnnGetTensorTransformDescriptor()

cudnnStatus_t cudnnGetTensorTransformDescriptor(
	cudnnTensorTransformDescriptor_t transformDesc,
	uint32_t nbDimsRequested,
	cudnnTensorFormat_t *destFormat,
	int32_t padBeforeA[],
	int32_t padAfterA[],
	uint32_t foldA[],
	cudnnFoldingDirection_t *direction);

This function returns the values stored in a previously initialized tensor transform descriptor.

Parameters

transformDesc
Input. A previously initialized tensor transform descriptor.
nbDimsRequested
Input. The number of dimensions to consider. For more information, see the Tensor Descriptor section in the cuDNN Developer Guide.
destFormat
Output. The transform format that will be returned.
padBeforeA[]
Output. An array filled with the amount of padding to add before each dimension. The dimension of this padBeforeA[] parameter is equal to nbDimsRequested.
padAfterA[]
Output. An array filled with the amount of padding to add after each dimension. The dimension of this padBeforeA[] parameter is equal to nbDimsRequested.
foldA[]
Output. An array that was filled with the folding parameters for each spatial dimension. The dimension of this foldA[] array is nbDimsRequested-2.
direction
Output. The setting that selects folding or unfolding. For more information, see cudnnFoldingDirection_t.

Returns

CUDNN_STATUS_SUCCESS
The results were obtained successfully.
CUDNN_STATUS_BAD_PARAM
If transformDesc is NULL or if nbDimsRequested is less than 3 or greater than CUDNN_DIM_MAX.

3.2.63. cudnnGetVersion()

size_t cudnnGetVersion()

This function returns the version number of the cuDNN library. It returns the CUDNN_VERSION defined present in the cudnn.h header file. Starting with release R2, the routine can be used to identify dynamically the current cuDNN library used by the application. The defined CUDNN_VERSION can be used to have the same application linked against different cuDNN versions using conditional compilation statements.

3.2.64. cudnnInitTransformDest()

cudnnStatus_t cudnnInitTransformDest(
	const cudnnTensorTransformDescriptor_t transformDesc,
	const cudnnTensorDescriptor_t srcDesc,
	cudnnTensorDescriptor_t destDesc,
	size_t *destSizeInBytes);
This function initializes and returns a destination tensor descriptor destDesc for tensor transform operations. The initialization is done with the desired parameters described in the transform descriptor cudnnTensorDescriptor_t.
Note: The returned tensor descriptor will be packed.

Parameters

transformDesc
Input. Handle to a previously initialized tensor transform descriptor.
srcDesc
Input. Handle to a previously initialized tensor descriptor.
destDesc
Output. Handle of the tensor descriptor that will be initialized and returned.
destSizeInBytes
Output. A pointer to hold the size, in bytes, of the new tensor.

Returns

CUDNN_STATUS_SUCCESS
The tensor descriptor was initialized successfully.
CUDNN_STATUS_BAD_PARAM
If either srcDesc or destDesc is NULL, or if the tensor descriptor’s nbDims is incorrect. For more information, see the Tensor Descriptor section in the cuDNN Developer Guide.
CUDNN_STATUS_NOT_SUPPORTED
If the provided configuration is not 4D.
CUDNN_STATUS_EXECUTION_FAILED
Function failed to launch on the GPU.

3.2.65. cudnnLRNCrossChannelForward()

cudnnStatus_t cudnnLRNCrossChannelForward(
    cudnnHandle_t                    handle,
    cudnnLRNDescriptor_t             normDesc,
    cudnnLRNMode_t                   lrnMode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *beta,
    const cudnnTensorDescriptor_t    yDesc,
    void                            *y)

This function performs the forward LRN layer computation.

Note: Supported formats are: positive-strided, NCHW and NHWC for 4D x and y, and only NCDHW DHW-packed for 5D (for both x and y). Only non-overlapping 4D and 5D tensors are supported. NCHW layout is preferred for performance.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

normDesc

Input. Handle to a previously initialized LRN parameter descriptor.

lrnMode

Input. LRN layer mode of operation. Currently only CUDNN_LRN_CROSS_CHANNEL_DIM1 is implemented. Normalization is performed along the tensor's dimA[1].

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Input. Tensor descriptor objects for the input and output tensors.

x

Input. Input tensor data pointer in device memory.

y

Output. Output tensor data pointer in device memory.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the tensor pointers x, y is NULL.
  • Number of input tensor dimensions is 2 or less.
  • LRN descriptor parameters are outside of their valid ranges.
  • One of the tensor parameters is 5D but is not in NCDHW DHW-packed format.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • Any of the input tensor datatypes is not the same as any of the output tensor datatype.
  • x and y tensor dimensions mismatch.
  • Any tensor parameters strides are negative.

3.2.66. cudnnNormalizationForwardInference()

cudnnStatus_t
cudnnNormalizationForwardInference(cudnnHandle_t handle,
                                   cudnnNormMode_t mode,
                   	            cudnnNormOps_t normOps,
                                   cudnnNormAlgo_t algo,
                         	      const void *alpha,
                               	const void *beta, 
                               	const cudnnTensorDescriptor_t xDesc,
                        	       const void *x,
                               	const cudnnTensorDescriptor_t normScaleBiasDesc,
                               	const void *normScale,
                               	const void *normBias,
                 	              const cudnnTensorDescriptor_t normMeanVarDesc,
                               	const void *estimatedMean,
                               	const void *estimatedVariance,
                             	  const cudnnTensorDescriptor_t zDesc,
                               	const void *z,
                                   cudnnActivationDescriptor_t activationDesc,
                               	const cudnnTensorDescriptor_t yDesc,
            	                   void *y,
                               	double epsilon,
                               	int groupCnt);
This function performs the forward normalization layer computation for the inference phase. Per-channel normalization layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015.
Note:
  • Only 4D and 5D tensors are supported.
  • The input transformation performed by this function is defined as:
    y = beta*y + alpha *[normBias + (normScale * (x-estimatedMean)/sqrt(epsilon + estimatedVariance)]
  • The epsilon value has to be the same during training, backpropagation, and inference.
  • For the training phase, use cudnnNormalizationForwardTraining().
  • Higher performance can be obtained when HW-packed tensors are used for all of x and y.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.

normOps

Input. Mode of post-operative. Currently, CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are not supported.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

alpha, beta
Inputs. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Input. Handles to the previously initialized tensor descriptors.

*x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc, for the layer’s x input data.

*y

Output. Data pointer to GPU memory associated with the tensor descriptor yDesc, for the y output of the normalization layer.

zDesc, *z

Input. Tensor descriptors and pointers in device memory for residual addition to the result of the normalization operation, prior to the activation. zDesc and *z are optional and are only used when normOps is CUDNN_NORM_OPS_NORM_ADD_ACTIVATION, otherwise users may pass NULL. When in use, z should have exactly the same dimension as x and the final output y. For more information, see cudnnTensorDescriptor_t.

Since normOps is only supported for CUDNN_NORM_OPS_NORM, we can set these to NULL for now.

normScaleBiasDesc, normScale, normBias

Inputs. Tensor descriptors and pointers in device memory for the normalization scale and bias parameters (in the original paper bias is referred to as beta and scale as gamma).

normMeanVarDesc, estimatedMean, estimatedVariance

Inputs. Mean and variance tensors and their tensor descriptors. The estimatedMean and estimatedVariance inputs, accumulated during the training phase from the cudnnNormalizationForwardTraining() call, should be passed as inputs here.

activationDesc

Input. Descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION then this activation is used, otherwise the user may pass NULL. Since normOps is only supported for CUDNN_NORM_OPS_NORM, we can set these to NULL for now.

epsilon

Input. Epsilon value used in the normalization formula. Its value should be equal to or greater than zero.

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the pointers alpha, beta, x, y, normScale, normBias, estimatedMean, and estimatedInvVariance is NULL.
  • The number of xDesc or yDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • normScaleBiasDesc and normMeanVarDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for per-channel, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • epsilon value is less than zero.
  • Dimensions or data types mismatch for xDesc and yDesc.
CUDNN_STATUS_NOT_SUPPORTED

A compute or data type other than FLOAT was chosen, or an unknown algorithm type was chosen.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.67. cudnnOpsInferVersionCheck()

cudnnStatus_t cudnnOpsInferVersionCheck(void)

This function is the first of a series of corresponding functions that check for consistent library versions among DLL files for different modules.

Returns

CUDNN_STATUS_SUCCESS

The version of this DLL file is consistent with cuDNN DLLs on which it depends.

CUDNN_STATUS_VERSION_MISMATCH

The version of this DLL file does not match that of a cuDNN DLLs on which it depends.

3.2.68. cudnnOpTensor()

cudnnStatus_t cudnnOpTensor(
    cudnnHandle_t                     handle,
    const cudnnOpTensorDescriptor_t   opTensorDesc,
    const void                       *alpha1,
    const cudnnTensorDescriptor_t     aDesc,
    const void                       *A,
    const void                       *alpha2,
    const cudnnTensorDescriptor_t     bDesc,
    const void                       *B,
    const void                       *beta,
    const cudnnTensorDescriptor_t     cDesc,
    void                             *C)

This function implements the equation C = op(alpha1[0] * A, alpha2[0] * B) + beta[0] * C, given the tensors A, B, and C and the scaling factors alpha1, alpha2, and beta. The op to use is indicated by the descriptor cudnnOpTensorDescriptor_t, meaning, the type of opTensorDesc. Currently-supported ops are listed by the cudnnOpTensorOp_t enum.

The following restrictions on the input and destination tensors apply:
  • Each dimension of the input tensor A must match the corresponding dimension of the destination tensor C, and each dimension of the input tensor B must match the corresponding dimension of the destination tensor C or must be equal to 1. In the latter case, the same value from the input tensor B for those dimensions will be used to blend into the C tensor.
  • The data types of the input tensors A and B, and the destination tensor C, must satisfy Table 5.
Table 5. Supported Datatypes
opTensorCompType in opTensorDesc A B C (destination)
FLOAT FLOAT FLOAT FLOAT
FLOAT INT8 INT8 FLOAT
FLOAT HALF HALF FLOAT
DOUBLE DOUBLE DOUBLE DOUBLE
FLOAT FLOAT FLOAT HALF
FLOAT HALF HALF HALF
FLOAT INT8 INT8 INT8
FLOAT FLOAT FLOAT INT8
Note: All tensor formats up to dimension five (5) are supported. This routine does not support tensor formats beyond these dimensions.

Parameters

handle

Input. Handle to a previously created cuDNN context.

opTensorDesc

Input. Handle to a previously initialized op tensor descriptor.

alpha1, alpha2, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

aDesc, bDesc, cDesc

Input. Handle to a previously initialized tensor descriptor.

A, B

Input. Pointer to data of the tensors described by the aDesc and bDesc descriptors, respectively.

C

Input/Output. Pointer to data of the tensor described by the cDesc descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function executed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimensions of the bias tensor and the output tensor dimensions are above 5.
  • opTensorCompType is not set as stated above.
CUDNN_STATUS_BAD_PARAM

The data type of the destination tensor C is unrecognized, or the restrictions on the input and destination tensors, stated above, are not met.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.69. cudnnPoolingForward()

cudnnStatus_t cudnnPoolingForward(
    cudnnHandle_t                    handle,
    const cudnnPoolingDescriptor_t   poolingDesc,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *beta,
    const cudnnTensorDescriptor_t    yDesc,
    void                            *y)

This function computes pooling of input values (meaning, the maximum or average of several adjacent values) to produce an output with smaller height and/or width.

Note:

Parameters

handle

Input. Handle to a previously created cuDNN context.

poolingDesc

Input. Handle to a previously initialized pooling descriptor.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc

Input. Handle to the previously initialized input tensor descriptor. Must be of type FLOAT, DOUBLE, HALF or INT8. For more information, see cudnnDataType_t.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

yDesc

Input. Handle to the previously initialized output tensor descriptor. Must be of type FLOAT, DOUBLE, HALF or INT8. For more information, see cudnnDataType_t.

y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions n, c of the input tensor and output tensors differ.
  • The datatype of the input tensor and output tensors differs.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.70. cudnnQueryRuntimeError()

cudnnStatus_t cudnnQueryRuntimeError(
    cudnnHandle_t            handle,
    cudnnStatus_t           *rstatus,
    cudnnErrQueryMode_t      mode,
    cudnnRuntimeTag_t       *tag)

cuDNN library functions perform extensive input argument checking before launching GPU kernels. The last step is to verify that the GPU kernel actually started. When a kernel fails to start, CUDNN_STATUS_EXECUTION_FAILED is returned by the corresponding API call. Typically, after a GPU kernel starts, no runtime checks are performed by the kernel itself - numerical results are simply written to output buffers.

When the CUDNN_BATCHNORM_SPATIAL_PERSISTENT mode is selected in cudnnBatchNormalizationForwardTraining() or cudnnBatchNormalizationBackward(), the algorithm may encounter numerical overflows where CUDNN_BATCHNORM_SPATIAL performs just fine albeit at a slower speed. The user can invoke cudnnQueryRuntimeError() to make sure numerical overflows did not occur during the kernel execution. Those issues are reported by the kernel that performs computations.

cudnnQueryRuntimeError() can be used in polling and blocking software control flows. There are two polling modes (CUDNN_ERRQUERY_RAWCODE and CUDNN_ERRQUERY_NONBLOCKING) and one blocking mode CUDNN_ERRQUERY_BLOCKING.

CUDNN_ERRQUERY_RAWCODE reads the error storage location regardless of the kernel completion status. The kernel might not even start and the error storage (allocated per cuDNN handle) might be used by an earlier call.

CUDNN_ERRQUERY_NONBLOCKING checks if all tasks in the user stream are completed. The cudnnQueryRuntimeError() function will return immediately and report CUDNN_STATUS_RUNTIME_IN_PROGRESS in rstatus if some tasks in the user stream are pending. Otherwise, the function will copy the remote kernel error code to rstatus.

In the blocking mode (CUDNN_ERRQUERY_BLOCKING), the function waits for all tasks to drain in the user stream before reporting the remote kernel error code. The blocking flavor can be further adjusted by calling cudaSetDeviceFlags with the cudaDeviceScheduleSpin, cudaDeviceScheduleYield, or cudaDeviceScheduleBlockingSync flag.

CUDNN_ERRQUERY_NONBLOCKING and CUDNN_ERRQUERY_BLOCKING modes should not be used when the user stream is changed in the cuDNN handle, meaning, cudnnSetStream() is invoked between functions that report runtime kernel errors and the cudnnQueryRuntimeError() function.

The remote error status reported in rstatus can be set to: CUDNN_STATUS_SUCCESS, CUDNN_STATUS_RUNTIME_IN_PROGRESS, or CUDNN_STATUS_RUNTIME_FP_OVERFLOW. The remote kernel error is automatically cleared by cudnnQueryRuntimeError().

Note: The cudnnQueryRuntimeError() function should be used in conjunction with cudnnBatchNormalizationForwardTraining() and cudnnBatchNormalizationBackward() when the cudnnBatchNormMode_t argument is CUDNN_BATCHNORM_SPATIAL_PERSISTENT.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rstatus

Output. Pointer to the user's error code storage.

mode

Input. Remote error query mode.

tag

Input/Output. Currently, this argument should be NULL.

Returns

CUDNN_STATUS_SUCCESS

No errors detected (rstatus holds a valid value).

CUDNN_STATUS_BAD_PARAM

Invalid input argument.

CUDNN_STATUS_INTERNAL_ERROR

A stream blocking synchronization or a non-blocking stream query failed.

CUDNN_STATUS_MAPPING_ERROR

The device cannot access zero-copy memory to report kernel errors.

3.2.71. cudnnReduceTensor()

cudnnStatus_t cudnnReduceTensor(
    cudnnHandle_t                           handle,
    const cudnnReduceTensorDescriptor_t     reduceTensorDesc,
    void                                   *indices,
    size_t                                  indicesSizeInBytes,
    void                                   *workspace,
    size_t                                  workspaceSizeInBytes,
    const void                             *alpha,
    const cudnnTensorDescriptor_t           aDesc,
    const void                             *A,
    const void                             *beta,
    const cudnnTensorDescriptor_t           cDesc,
    void                                   *C)

This function reduces tensor A by implementing the equation C = alpha * reduce op ( A ) + beta * C, given tensors A and C and scaling factors alpha and beta. The reduction op to use is indicated by the descriptor reduceTensorDesc. Currently-supported ops are listed by the cudnnReduceTensorOp_t enum.

Each dimension of the output tensor C must match the corresponding dimension of the input tensor A or must be equal to 1. The dimensions equal to 1 indicate the dimensions of A to be reduced.

The implementation will generate indices for the min and max ops only, as indicated by the cudnnReduceTensorIndices_t enum of the reduceTensorDesc. Requesting indices for the other reduction ops results in an error. The data type of the indices is indicated by the cudnnIndicesType_t enum; currently only the 32-bit (unsigned int) type is supported.

The indices returned by the implementation are not absolute indices but relative to the dimensions being reduced. The indices are also flattened, meaning, not coordinate tuples.

The data types of the tensors A and C must match if of type double. In this case, alpha and beta and the computation enum of reduceTensorDesc are all assumed to be of type double.

The HALF and INT8 data types may be mixed with the FLOAT data types. In these cases, the computation enum of reduceTensorDesc is required to be of type FLOAT.
Note:

Up to dimension 8, all tensor formats are supported. Beyond those dimensions, this routine is not supported.

Parameters

handle

Input. Handle to a previously created cuDNN context.

reduceTensorDesc

Input. Handle to a previously initialized reduce tensor descriptor.

indices

Output. Handle to a previously allocated space for writing indices.

indicesSizeInBytes

Input. Size of the above previously allocated space.

workspace

Input. Handle to a previously allocated space for the reduction implementation.

workspaceSizeInBytes

Input. Size of the above previously allocated space.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

aDesc, cDesc

Input. Handle to a previously initialized tensor descriptor.

A

Input. Pointer to data of the tensor described by the aDesc descriptor.

C

Input/Output. Pointer to data of the tensor described by the cDesc descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function executed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimensions of the input tensor and the output tensor are above 8.
  • reduceTensorCompType is not set as stated above.
CUDNN_STATUS_BAD_PARAM

The corresponding dimensions of the input and output tensors all match, or the conditions in the above paragraphs are unmet.

CUDNN_INVALID_VALUE

The allocations for the indices or workspace are insufficient.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.72. cudnnRestoreAlgorithm()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnRestoreAlgorithm(
    cudnnHandle_t               handle,
    void*                       algoSpace,
    size_t                      algoSpaceSizeInBytes,
    cudnnAlgorithmDescriptor_t  algoDesc)

This function reads algorithm metadata from the host memory space provided by the user in algoSpace, allowing the user to use the results of RNN finds from previous cuDNN sessions.

Parameters

handle

Input. Handle to a previously created cuDNN context.

algoDesc

Input. A previously created algorithm descriptor.

algoSpace

Input. Pointer to the host memory to be read.

algoSpaceSizeInBytes

Input. Amount of host memory needed as a workspace to be able to hold the metadata from the specified algoDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The metadata is from a different cuDNN version.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions is met:

  • One of the arguments is NULL.
  • The metadata is corrupted.

3.2.73. cudnnRestoreDropoutDescriptor()

cudnnStatus_t cudnnRestoreDropoutDescriptor(
    cudnnDropoutDescriptor_t dropoutDesc,
    cudnnHandle_t            handle,
    float                    dropout,
    void                    *states,
    size_t                   stateSizeInBytes,
    unsigned long long       seed)

This function restores a dropout descriptor to a previously saved-off state.

Parameters

dropoutDesc

Input/Output. Previously created dropout descriptor.

handle

Input. Handle to a previously created cuDNN context.

dropout

Input. Probability with which the value from an input tensor is set to 0 when performing dropout.

states

Input. Pointer to GPU memory that holds random number generator states initialized by a prior call to cudnnSetDropoutDescriptor().

stateSizeInBytes

Input. Size in bytes of buffer holding random number generator states.

seed

Input. Seed used in prior call to cudnnSetDropoutDescriptor() that initialized states buffer. Using a different seed from this has no effect. A change of seed, and subsequent update to random number generator states can be achieved by calling cudnnSetDropoutDescriptor().

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_INVALID_VALUE

States buffer size (as indicated in stateSizeInBytes) is too small.

3.2.74. cudnnSaveAlgorithm()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnSaveAlgorithm(
    cudnnHandle_t          		handle,
    cudnnAlgorithmDescriptor_t 	algoDesc,
    void*                  		algoSpace
    size_t                 		algoSpaceSizeInBytes)

This function writes algorithm metadata into the host memory space provided by the user in algoSpace, allowing the user to preserve the results of RNN finds after cuDNN exits.

Parameters

handle

Input. Handle to a previously created cuDNN context.

algoDesc

Input. A previously created algorithm descriptor.

algoSpace

Input. Pointer to the host memory to be written.

algoSpaceSizeInBytes

Input. Amount of host memory needed as a workspace to be able to save the metadata from the specified algoDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions is met:

  • One of the arguments is NULL.
  • algoSpaceSizeInBytes is too small.

3.2.75. cudnnScaleTensor()

cudnnStatus_t cudnnScaleTensor(
    cudnnHandle_t                   handle,
    const cudnnTensorDescriptor_t   yDesc,
    void                           *y,
    const void                     *alpha)

This function scales all the elements of a tensor by a given factor.

Parameters

handle

Input. Handle to a previously created cuDNN context.

yDesc

Input. Handle to a previously initialized tensor descriptor.

y

Input/Output. Pointer to data of the tensor described by the yDesc descriptor.

alpha

Input. Pointer in the host memory to a single value that all elements of the tensor will be scaled with. For more information, see Scaling Parameters in the cuDNN Developer Guide.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

One of the provided pointers is nil.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.76. cudnnSetActivationDescriptor()

cudnnStatus_t cudnnSetActivationDescriptor(
    cudnnActivationDescriptor_t         activationDesc,
    cudnnActivationMode_t               mode,
    cudnnNanPropagation_t               reluNanOpt,
    double                              coef)

This function initializes a previously created generic activation descriptor object.

Parameters

activationDesc

Input/Output. Handle to a previously created pooling descriptor.

mode

Input. Enumerant to specify the activation mode.

reluNanOpt

Input. Enumerant to specify the Nan propagation mode.

coef

Input. Floating point number. When the activation mode (see cudnnActivationMode_t) is set to CUDNN_ACTIVATION_CLIPPED_RELU, this input specifies the clipping threshold; and when the activation mode is set to CUDNN_ACTIVATION_RELU, this input specifies the upper bound.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

mode or reluNanOpt has an invalid enumerant value.

3.2.77. cudnnSetAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnSetAlgorithmDescriptor(
    cudnnAlgorithmDescriptor_t      algorithmDesc,
    cudnnAlgorithm_t                algorithm)

This function initializes a previously created generic algorithm descriptor object.

Parameters

algorithmDesc

Input/Output. Handle to a previously created algorithm descriptor.

algorithm

Input. Struct to specify the algorithm.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

3.2.78. cudnnSetAlgorithmPerformance()

This function has been deprecated in cuDNN 8.0.

cudnnStatus_t cudnnSetAlgorithmPerformance(
    cudnnAlgorithmPerformance_t	    algoPerf,
    cudnnAlgorithmDescriptor_t      algoDesc,
    cudnnStatus_t                   status,
    float                           time,
    size_t                          memory)

This function initializes a previously created generic algorithm performance object.

Parameters

algoPerf

Input/Output. Handle to a previously created algorithm performance object.

algoDesc

Input. The algorithm descriptor which the performance results describe.

status

Input. The cuDNN status returned from running the algoDesc algorithm.

time

Input. The GPU time spent running the algoDesc algorithm.

memory

Input. The GPU memory needed to run the algoDesc algorithm.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

mode or reluNanOpt has an invalid enumerate value.

3.2.79. cudnnSetCallback()

cudnnStatus_t cudnnSetCallback(
        unsigned            mask,
        void                *udata,
        cudnnCallback_t     fptr)

This function sets the internal states of cuDNN error reporting functionality.

Parameters

mask

Input. An unsigned integer. The four least significant bits (LSBs) of this unsigned integer are used for switching on and off the different levels of error reporting messages. This applies for both the default callbacks, and for the customized callbacks. The bit position is in correspondence with the enum of cudnnSeverity_t. The user may utilize the predefined macros CUDNN_SEV_ERROR_EN, CUDNN_SEV_WARNING_EN, and CUDNN_SEV_INFO_EN to form the bit mask. When a bit is set to 1, the corresponding message channel is enabled.

For example, when bit 3 is set to 1, the API logging is enabled. Currently, only the log output of level CUDNN_SEV_INFO is functional; the others are not yet implemented. When used for turning on and off the logging with the default callback, the user may pass NULL to udata and fptr. In addition, the environment variable CUDNN_LOGDEST_DBG must be set. For more information, see the Backward compatibility and deprecation policy section in the cuDNN Developer Guide.
  • CUDNN_SEV_INFO_EN= 0b1000 (functional).
  • CUDNN_SEV_ERROR_EN= 0b0010 (not yet functional).
  • CUDNN_SEV_WARNING_EN= 0b0100 (not yet functional).

The output of CUDNN_SEV_FATAL is always enabled and cannot be disabled.

udata

Input. A pointer provided by the user. This pointer will be passed to the user’s custom logging callback function. The data it points to will not be read, nor be changed by cuDNN. This pointer may be used in many ways, such as in a mutex or in a communication socket for the user’s callback function for logging. If the user is utilizing the default callback function, or doesn’t want to use this input in the customized callback function, they may pass in NULL.

fptr
Input. A pointer to a user-supplied callback function. When NULL is passed to this pointer, then cuDNN switches back to the built-in default callback function. The user-supplied callback function prototype must be similar to the following (also defined in the header file):
void customizedLoggingCallback (cudnnSeverity_t sev, void *udata, const cudnnDebug_t *dbg, const char *msg);
  • The structure cudnnDebug_t is defined in the header file. It provides the metadata, such as time, time since start, stream ID, process and thread ID, that the user may choose to print or store in their customized callback.
  • The variable msg is the logging message generated by cuDNN. Each line of this message is terminated by \0, and the end of the message is terminated by \0\0. Users may select what is necessary to show in the log, and may reformat the string.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

3.2.80. cudnnSetDropoutDescriptor()

cudnnStatus_t cudnnSetDropoutDescriptor(
    cudnnDropoutDescriptor_t    dropoutDesc,
    cudnnHandle_t               handle,
    float                       dropout,
    void                       *states,
    size_t                      stateSizeInBytes,
    unsigned long long          seed)

This function initializes a previously created dropout descriptor object. If the states argument is equal to NULL, then the random number generator states won't be initialized, and only the dropout value will be set. No other function should be writing to the memory pointed at by the states argument while this function is running. The user is expected not to change the memory pointed at by states for the duration of the computation.

Parameters

dropoutDesc

Input/Output. Previously created dropout descriptor object.

handle

Input. Handle to a previously created cuDNN context.

dropout

Input. The probability with which the value from input is set to zero during the dropout layer.

states

Output. Pointer to user-allocated GPU memory that will hold random number generator states.

stateSizeInBytes

Input. Specifies the size in bytes of the provided memory for the states.

seed

Input. Seed used to initialize random number generator states.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_INVALID_VALUE

sizeInBytes is less than the value returned by cudnnDropoutGetStatesSize().

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.81. cudnnSetFilter4dDescriptor()

cudnnStatus_t cudnnSetFilter4dDescriptor(
    cudnnFilterDescriptor_t    filterDesc,
    cudnnDataType_t            dataType,
    cudnnTensorFormat_t        format,
    int                        k,
    int                        c,
    int                        h,
    int                        w)

This function initializes a previously created filter descriptor object into a 4D filter. The layout of the filters must be contiguous in memory.

Tensor format CUDNN_TENSOR_NHWC has limited support in cudnnConvolutionForward(), cudnnConvolutionBackwardData(), and cudnnConvolutionBackwardFilter().

Parameters

filterDesc

Input/Output. Handle to a previously created filter descriptor.

datatype

Input. Data type.

format
Input.Type of the filter layout format. If this input is set to CUDNN_TENSOR_NCHW, which is one of the enumerant values allowed by cudnnTensorFormat_t descriptor, then the layout of the filter is in the form of KCRS, where:
  • K represents the number of output feature maps
  • C is the number of input feature maps
  • R is the number of rows per filter
  • S is the number of columns per filter

If this input is set to CUDNN_TENSOR_NHWC, then the layout of the filter is in the form of KRSC. For more information, see cudnnTensorFormat_t.

k

Input. Number of output feature maps.

c

Input. Number of input feature maps.

h

Input. Height of each filter.

w

Input. Width of each filter.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the parameters k, c, h, w is negative or dataType or format has an invalid enumerant value.

3.2.82. cudnnSetFilterNdDescriptor()

cudnnStatus_t cudnnSetFilterNdDescriptor(
    cudnnFilterDescriptor_t filterDesc,
    cudnnDataType_t         dataType,
    cudnnTensorFormat_t     format,
    int                     nbDims,
    const int               filterDimA[])

This function initializes a previously created filter descriptor object. The layout of the filters must be contiguous in memory.

The tensor format CUDNN_TENSOR_NHWC has limited support in cudnnConvolutionForward(), cudnnConvolutionBackwardData(), and cudnnConvolutionBackwardFilter().

Parameters

filterDesc

Input/Output. Handle to a previously created filter descriptor.

datatype

Input. Data type.

format

Input.Type of the filter layout format. If this input is set to CUDNN_TENSOR_NCHW, which is one of the enumerant values allowed by cudnnTensorFormat_t descriptor, then the layout of the filter is as follows:

  • For N=4, a 4D filter descriptor, the filter layout is in the form of KCRS:
    • K represents the number of output feature maps
    • C is the number of input feature maps
    • R is the number of rows per filter
    • S is the number of columns per filter
  • For N=3, a 3D filter descriptor, the number S (number of columns per filter) is omitted.
  • For N=5 and greater, the layout of the higher dimensions immediately follows RS.

On the other hand, if this input is set to CUDNN_TENSOR_NHWC, then the layout of the filter is as follows:

  • For N=4, a 4D filter descriptor, the filter layout is in the form of KRSC.
  • For N=3, a 3D filter descriptor, the number S (number of columns per filter) is omitted and the layout of C immediately follows R.
  • For N=5 and greater, the layout of the higher dimensions are inserted between S and C. For more information, see cudnnTensorFormat_t.
nbDims

Input. Dimension of the filter.

filterDimA

Input. Array of dimension nbDims containing the size of the filter for each dimension.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the elements of the array filterDimA is negative or dataType or format has an invalid enumerant value.

CUDNN_STATUS_NOT_SUPPORTED

The parameter nbDims exceeds CUDNN_DIM_MAX.

3.2.83. cudnnSetLRNDescriptor()

cudnnStatus_t cudnnSetLRNDescriptor(
    cudnnLRNDescriptor_t   normDesc,
    unsigned               lrnN,
    double                 lrnAlpha,
    double                 lrnBeta,
    double                 lrnK)

This function initializes a previously created LRN descriptor object.

Note:
  • Macros CUDNN_LRN_MIN_N, CUDNN_LRN_MAX_N, CUDNN_LRN_MIN_K, CUDNN_LRN_MIN_BETA defined in cudnn.h specify valid ranges for parameters.
  • Values of double parameters will be cast down to the tensor datatype during computation.

Parameters

normDesc

Output. Handle to a previously created LRN descriptor.

lrnN

Input. Normalization window width in elements. The LRN layer uses a window [center-lookBehind, center+lookAhead], where lookBehind = floor( (lrnN-1)/2 ), lookAhead = lrnN-lookBehind-1. So for n=10, the window is [k-4...k...k+5] with a total of 10 samples. For the DivisiveNormalization layer, the window has the same extents as above in all spatial dimensions (dimA[2], dimA[3], dimA[4]). By default, lrnN is set to 5 in cudnnCreateLRNDescriptor().

lrnAlpha

Input. Value of the alpha variance scaling parameter in the normalization formula. Inside the library code, this value is divided by the window width for LRN and by (window width)^#spatialDimensions for DivisiveNormalization. By default, this value is set to 1e-4 in cudnnCreateLRNDescriptor().

lrnBeta

Input. Value of the beta power parameter in the normalization formula. By default, this value is set to 0.75 in cudnnCreateLRNDescriptor().

lrnK

Input. Value of the k parameter in the normalization formula. By default, this value is set to 2.0.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

One of the input parameters was out of valid range as described above.

3.2.84. cudnnSetOpTensorDescriptor()

cudnnStatus_t cudnnSetOpTensorDescriptor(
    cudnnOpTensorDescriptor_t   opTensorDesc,
    cudnnOpTensorOp_t           opTensorOp,
    cudnnDataType_t             opTensorCompType,
    cudnnNanPropagation_t       opTensorNanOpt)

This function initializes a tensor pointwise math descriptor.

Parameters

opTensorDesc

Output. Pointer to the structure holding the description of the tensor pointwise math descriptor.

opTensorOp

Input. Tensor pointwise math operation for this tensor pointwise math descriptor.

opTensorCompType

Input. Computation datatype for this tensor pointwise math descriptor.

opTensorNanOpt

Input. NAN propagation policy.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the input parameters passed is invalid.

3.2.85. cudnnSetPooling2dDescriptor()

cudnnStatus_t cudnnSetPooling2dDescriptor(
    cudnnPoolingDescriptor_t    poolingDesc,
    cudnnPoolingMode_t          mode,
    cudnnNanPropagation_t       maxpoolingNanOpt,
    int                         windowHeight,
    int                         windowWidth,
    int                         verticalPadding,
    int                         horizontalPadding,
    int                         verticalStride,
    int                         horizontalStride)

This function initializes a previously created generic pooling descriptor object into a 2D description.

Parameters

poolingDesc

Input/Output. Handle to a previously created pooling descriptor.

mode

Input. Enumerant to specify the pooling mode.

maxpoolingNanOpt

Input. Enumerant to specify the Nan propagation mode.

windowHeight

Input. Height of the pooling window.

windowWidth

Input. Width of the pooling window.

verticalPadding

Input. Size of vertical padding.

horizontalPadding

Input. Size of horizontal padding

verticalStride

Input. Pooling vertical stride.

horizontalStride

Input. Pooling horizontal stride.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the parameters windowHeight, windowWidth, verticalStride, horizontalStride is negative or mode or maxpoolingNanOpt has an invalid enumerate value.

3.2.86. cudnnSetPoolingNdDescriptor()

cudnnStatus_t cudnnSetPoolingNdDescriptor(
    cudnnPoolingDescriptor_t     poolingDesc,
    const cudnnPoolingMode_t     mode,
    const cudnnNanPropagation_t  maxpoolingNanOpt,
    int                          nbDims,
    const int                    windowDimA[],
    const int                    paddingA[],
    const int                    strideA[])

This function initializes a previously created generic pooling descriptor object.

Parameters

poolingDesc

Input/Output. Handle to a previously created pooling descriptor.

mode

Input. Enumerant to specify the pooling mode.

maxpoolingNanOpt

Input. Enumerant to specify the Nan propagation mode.

nbDims

Input. Dimension of the pooling operation. Must be greater than zero.

windowDimA

Input. Array of dimension nbDims containing the window size for each dimension. The value of array elements must be greater than zero.

paddingA

Input. Array of dimension nbDims containing the padding size for each dimension. Negative padding is allowed.

strideA

Input. Array of dimension nbDims containing the striding size for each dimension. The value of array elements must be greater than zero (meaning, negative striding size is not allowed).

Returns

CUDNN_STATUS_SUCCESS

The object was initialized successfully.

CUDNN_STATUS_NOT_SUPPORTED

If (nbDims > CUDNN_DIM_MAX-2).

CUDNN_STATUS_BAD_PARAM

Either nbDims, or at least one of the elements of the arrays windowDimA or strideA is negative, or mode or maxpoolingNanOpt has an invalid enumerate value.

3.2.87. cudnnSetReduceTensorDescriptor()

cudnnStatus_t cudnnSetReduceTensorDescriptor(
    cudnnReduceTensorDescriptor_t   reduceTensorDesc,
    cudnnReduceTensorOp_t           reduceTensorOp,
    cudnnDataType_t                 reduceTensorCompType,
    cudnnNanPropagation_t           reduceTensorNanOpt,
    cudnnReduceTensorIndices_t      reduceTensorIndices,
    cudnnIndicesType_t              reduceTensorIndicesType)

This function initializes a previously created reduce tensor descriptor object.

Parameters

reduceTensorDesc

Input/Output. Handle to a previously created reduce tensor descriptor.

reduceTensorOp

Input. Enumerant to specify the reduce tensor operation.

reduceTensorCompType

Input. Enumerant to specify the computation datatype of the reduction.

reduceTensorNanOpt

Input. Enumerant to specify the Nan propagation mode.

reduceTensorIndices

Input. Enumerant to specify the reduced tensor indices.

reduceTensorIndicesType

Input. Enumerant to specify the reduce tensor indices type.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

reduceTensorDesc is NULL (reduceTensorOp, reduceTensorCompType, reduceTensorNanOpt, reduceTensorIndices or reduceTensorIndicesType has an invalid enumerant value).

3.2.88. cudnnSetSpatialTransformerNdDescriptor()

cudnnStatus_t cudnnSetSpatialTransformerNdDescriptor(
    cudnnSpatialTransformerDescriptor_t     stDesc,
    cudnnSamplerType_t                      samplerType,
    cudnnDataType_t                         dataType,
    const int                               nbDims,
    const int                               dimA[])

This function initializes a previously created generic spatial transformer descriptor object.

Parameters

stDesc

Input/Output. Previously created spatial transformer descriptor object.

samplerType

Input. Enumerant to specify the sampler type.

dataType

Input. Data type.

nbDims

Input. Dimension of the transformed tensor.

dimA

Input. Array of dimension nbDims containing the size of the transformed tensor for every dimension.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • Either stDesc or dimA is NULL.
  • Either dataType or samplerType has an invalid enumerant value

3.2.89. cudnnSetStream()

cudnnStatus_t cudnnSetStream(
    cudnnHandle_t   handle,
    cudaStream_t    streamId)

This function sets the user's CUDA stream in the cuDNN handle. The new stream will be used to launch cuDNN GPU kernels or to synchronize to this stream when cuDNN kernels are launched in the internal streams. If the cuDNN library stream is not set, all kernels use the default (NULL) stream. Setting the user stream in the cuDNN handle guarantees the issue-order execution of cuDNN calls and other GPU kernels launched in the same stream.

Parameters

handle

Input. Pointer to the cuDNN handle.

streamID

Input. New CUDA stream to be written to the cuDNN handle.

Returns

CUDNN_STATUS_BAD_PARAM

Invalid (NULL) handle.

CUDNN_STATUS_MAPPING_ERROR

Mismatch between the user stream and the cuDNN handle context.

CUDNN_STATUS_SUCCESS

The new stream was set successfully.

3.2.90. cudnnSetTensor()

cudnnStatus_t cudnnSetTensor(
    cudnnHandle_t                   handle,
    const cudnnTensorDescriptor_t   yDesc,
    void                           *y,
    const void                     *valuePtr)

This function sets all the elements of a tensor to a given value.

Parameters

handle

Input. Handle to a previously created cuDNN context.

yDesc

Input. Handle to a previously initialized tensor descriptor.

y

Input/Output. Pointer to data of the tensor described by the yDesc descriptor.

valuePtr

Input. Pointer in host memory to a single value. All elements of the y tensor will be set to value[0]. The data type of the element in value[0] has to match the data type of tensor y.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

One of the provided pointers is nil.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.91. cudnnSetTensor4dDescriptor()

cudnnStatus_t cudnnSetTensor4dDescriptor(
    cudnnTensorDescriptor_t tensorDesc,
    cudnnTensorFormat_t     format,
    cudnnDataType_t         dataType,
    int                     n,
    int                     c,
    int                     h,
    int                     w)

This function initializes a previously created generic tensor descriptor object into a 4D tensor. The strides of the four dimensions are inferred from the format parameter and set in such a way that the data is contiguous in memory with no padding between dimensions.

Note: The total size of a tensor including the potential padding between dimensions is limited to 2 Giga-elements of type datatype.

Parameters

tensorDesc

Input/Output. Handle to a previously created tensor descriptor.

format

Input. Type of format.

datatype

Input. Data type.

n

Input. Number of images.

c

Input. Number of feature maps per image.

h

Input. Height of each feature map.

w

Input. Width of each feature map.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the parameters n, c, h, w was negative or format has an invalid enumerant value or dataType has an invalid enumerant value.

CUDNN_STATUS_NOT_SUPPORTED

The total size of the tensor descriptor exceeds the maximum limit of 2 Giga-elements.

3.2.92. cudnnSetTensor4dDescriptorEx()

cudnnStatus_t cudnnSetTensor4dDescriptorEx(
    cudnnTensorDescriptor_t     tensorDesc,
    cudnnDataType_t             dataType,
    int                         n,
    int                         c,
    int                         h,
    int                         w,
    int                         nStride,
    int                         cStride,
    int                         hStride,
    int                         wStride)

This function initializes a previously created generic tensor descriptor object into a 4D tensor, similarly to cudnnSetTensor4dDescriptor() but with the strides explicitly passed as parameters. This can be used to lay out the 4D tensor in any order or simply to define gaps between dimensions.

Note:
  • At present, some cuDNN routines have limited support for strides. Those routines will return CUDNN_STATUS_NOT_SUPPORTED if a 4D tensor object with an unsupported stride is used. cudnnTransformTensor() can be used to convert the data to a supported layout.
  • The total size of a tensor including the potential padding between dimensions is limited to 2 Giga-elements of type datatype.

Parameters

tensorDesc

Input/Output. Handle to a previously created tensor descriptor.

datatype

Input. Data type.

n

Input. Number of images.

c

Input. Number of feature maps per image.

h

Input. Height of each feature map.

w

Input. Width of each feature map.

nStride

Input. Stride between two consecutive images.

cStride

Input. Stride between two consecutive feature maps.

hStride

Input. Stride between two consecutive rows.

wStride

Input. Stride between two consecutive columns.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the parameters n, c, h, w or nStride, cStride, hStride, wStride is negative or dataType has an invalid enumerant value.

CUDNN_STATUS_NOT_SUPPORTED

The total size of the tensor descriptor exceeds the maximum limit of 2 Giga-elements.

3.2.93. cudnnSetTensorNdDescriptor()

cudnnStatus_t cudnnSetTensorNdDescriptor(
    cudnnTensorDescriptor_t tensorDesc,
    cudnnDataType_t         dataType,
    int                     nbDims,
    const int               dimA[],
    const int               strideA[])

This function initializes a previously created generic tensor descriptor object.

Note: The total size of a tensor including the potential padding between dimensions is limited to 2 Giga-elements of type datatype. Tensors are restricted to having at least 4 dimensions, and at most CUDNN_DIM_MAX dimensions (defined in cudnn.h). When working with lower dimensional data, it is recommended that the user create a 4D tensor, and set the size along unused dimensions to 1.

Parameters

tensorDesc

Input/Output. Handle to a previously created tensor descriptor.

datatype

Input. Data type.

nbDims
Input. Dimension of the tensor.
Note: Do not use 2 dimensions. Due to historical reasons, the minimum number of dimensions in the filter descriptor is three. For more information, see cudnnGetRNNLinLayerBiasParams().
dimA

Input. Array of dimension nbDims that contain the size of the tensor for every dimension. The size along unused dimensions should be set to 1. By convention, the ordering of dimensions in the array follows the format - [N, C, D, H, W], with W occupying the smallest index in the array.

strideA

Input. Array of dimension nbDims that contain the stride of the tensor for every dimension. By convention, the ordering of the strides in the array follows the format - [Nstride, Cstride, Dstride, Hstride, Wstride], with Wstride occupying the smallest index in the array.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the elements of the array dimA was negative or zero, or dataType has an invalid enumerant value.

CUDNN_STATUS_NOT_SUPPORTED

The parameter nbDims is outside the range [4, CUDNN_DIM_MAX], or the total size of the tensor descriptor exceeds the maximum limit of 2 Giga-elements.

3.2.94. cudnnSetTensorNdDescriptorEx()

cudnnStatus_t cudnnSetTensorNdDescriptorEx(
    cudnnTensorDescriptor_t tensorDesc,
    cudnnTensorFormat_t     format,
    cudnnDataType_t         dataType,
    int                     nbDims,
    const int               dimA[])

This function initializes an n-D tensor descriptor.

Parameters

tensorDesc

Output. Pointer to the tensor descriptor struct to be initialized.

format

Input. Tensor format.

dataType

Input. Tensor data type.

nbDims
Input. Dimension of the tensor.
Note: Do not use 2 dimensions. Due to historical reasons, the minimum number of dimensions in the filter descriptor is three. For more information, see cudnnGetRNNLinLayerBiasParams().
dimA

Input. Array containing the size of each dimension.

Returns

CUDNN_STATUS_SUCCESS

The function was successful.

CUDNN_STATUS_BAD_PARAM

Tensor descriptor was not allocated properly; or input parameters are not set correctly.

CUDNN_STATUS_NOT_SUPPORTED

Dimension size requested is larger than maximum dimension size supported.

3.2.95. cudnnSetTensorTransformDescriptor()

cudnnStatus_t cudnnSetTensorTransformDescriptor(
	cudnnTensorTransformDescriptor_t transformDesc,
	const uint32_t nbDims,
	const cudnnTensorFormat_t destFormat,
	const int32_t padBeforeA[],
	const int32_t padAfterA[],
	const uint32_t foldA[],
	const cudnnFoldingDirection_t direction);

This function initializes a tensor transform descriptor that was previously created using the cudnnCreateTensorTransformDescriptor() function.

Parameters

transformDesc

Output. The tensor transform descriptor to be initialized.

nbDims

Input. The dimensionality of the transform operands. Must be greater than 2. For more information, see the Tensor Descriptor section from the cuDNN Developer Guide.

destFormat

Input. The desired destination format.

padBeforeA[]

Input. An array that contains the amount of padding that should be added before each dimension. Set to NULL for no padding.

padAfterA[]

Input. An array that contains the amount of padding that should be added after each dimension. Set to NULL for no padding.

foldA[]

Input. An array that contains the folding parameters for each spatial dimension (dimensions 2 and up). Set to NULL for no folding.

direction

Input. Selects folding or unfolding. This input has no effect when folding parameters are all <= 1. For more information, see cudnnFoldingDirection_t.

Returns

CUDNN_STATUS_SUCCESS

The function was launched successfully.

CUDNN_STATUS_BAD_PARAM

The parameter transformDesc is NULL, or if direction is invalid, or nbDims is <= 2.

CUDNN_STATUS_NOT_SUPPORTED

If the dimension size requested is larger than maximum dimension size supported (meaning, one of the nbDims is larger than CUDNN_DIM_MAX), or if destFromat is something other than NCHW or NHWC.

3.2.96. cudnnSoftmaxForward()

cudnnStatus_t cudnnSoftmaxForward(
    cudnnHandle_t                    handle,
    cudnnSoftmaxAlgorithm_t          algorithm,
    cudnnSoftmaxMode_t               mode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *beta,
    const cudnnTensorDescriptor_t    yDesc,
    void                            *y)

This routine computes the softmax function.

Note: All tensor formats are supported for all modes and algorithms with 4 and 5D tensors. Performance is expected to be highest with NCHW fully-packed tensors. For more than 5 dimensions tensors must be packed in their spatial dimensions

Parameters

handle

Input. Handle to a previously created cuDNN context.

algorithm

Input. Enumerant to specify the softmax algorithm.

mode

Input. Enumerant to specify the softmax mode.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

yDesc

Input. Handle to the previously initialized output tensor descriptor.

y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions n, c, h, w of the input tensor and output tensors differ.
  • The datatype of the input tensor and output tensors differ.
  • The parameters algorithm or mode have an invalid enumerant value.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.97. cudnnSpatialTfGridGeneratorForward()

cudnnStatus_t cudnnSpatialTfGridGeneratorForward(
    cudnnHandle_t                               handle,
    const cudnnSpatialTransformerDescriptor_t   stDesc,
    const void                                 *theta,
    void                                       *grid)

This function generates a grid of coordinates in the input tensor corresponding to each pixel from the output tensor.

Note: Only 2d transformation is supported.

Parameters

handle

Input. Handle to a previously created cuDNN context.

stDesc

Input. Previously created spatial transformer descriptor object.

theta

Input. Affine transformation matrix. It should be of size n*2*3 for a 2d transformation, where n is the number of images specified in stDesc.

grid

Output. A grid of coordinates. It is of size n*h*w*2 for a 2d transformation, where n, h, w is specified in stDesc. In the 4th dimension, the first coordinate is x, and the second coordinate is y.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is NULL.
  • One of the parameters grid or theta is NULL.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimension of the transformed tensor specified in stDesc > 4.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.98. cudnnSpatialTfSamplerForward()

cudnnStatus_t cudnnSpatialTfSamplerForward(
    cudnnHandle_t                              handle,
    const cudnnSpatialTransformerDescriptor_t  stDesc,
    const void                                *alpha,
    const cudnnTensorDescriptor_t              xDesc,
    const void                                *x,
    const void                                *grid,
    const void                                *beta,
    cudnnTensorDescriptor_t                    yDesc,
    void                                      *y)

This function performs a sampler operation and generates the output tensor using the grid given by the grid generator.

Note: Only 2d transformation is supported.

Parameters

handle

Input. Handle to a previously created cuDNN context.

stDesc

Input. Previously created spatial transformer descriptor object.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with prior value in the destination tensor as follows:
dstValue = alpha[0]*srcValue + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

grid

Input. A grid of coordinates generated by cudnnSpatialTfGridGeneratorForward().

yDesc

Input. Handle to the previously initialized output tensor descriptor.

y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is NULL.
  • One of the parameters x, y or grid is NULL.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimension of transformed tensor > 4.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.99. cudnnTransformFilter()

cudnnStatus_t cudnnTransformTensorEx(
	cudnnHandle_t handle,
	const cudnnTensorTransformDescriptor_t transDesc,                                    
	const void *alpha,                                           
	const cudnnFilterDescriptor_t srcDesc,                        
	const void *srcData,   
	const void *beta, 
	const cudnnFilterDescriptor_t destDesc, 
	void *destData);

This function converts the filter between different formats, data types, or dimensions based on the described transformation. It can be used to convert a filter with an unsupported layout format to a filter with a supported layout format.

This function copies the scaled data from the input filter srcDesc to the output tensor destDesc with a different layout. If the filter descriptors of srcDesc and destDesc have different dimensions, they must be consistent with folding and padding amount and order specified in transDesc.

The srcDesc and destDesc tensors must not overlap in any way (imeaning, tensors cannot be transformed in place).
Note: When performing a folding transform or a zero-padding transform, the scaling factors (alpha, beta) should be set to (1, 0). However, unfolding transforms support any (alpha, beta) values. This function is thread safe.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

transDesc

Input. A descriptor containing the details of the requested filter transformation. For more information, see cudnnTensorTransformDescriptor_t.

alpha, beta

Input. Pointers, in the host memory, to the scaling factors used to scale the data in the input tensor srcDesc. beta is used to scale the destination tensor, while alpha is used to scale the source tensor. For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

The beta scaling value is not honored in the folding and zero-padding cases. Unfolding supports any (alpha, beta).

srcDesc, destDesc

Input. Handles to the previously initiated filter descriptors. srcDesc and destDesc must not overlap. For more information, see cudnnTensorDescriptor_t.

srcData, destData

Input. Pointers, in the host memory, to the data of the tensor described by srcDesc and destDesc respectively.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

A parameter is uninitialized or initialized incorrectly, or the number of dimensions is different between srcDesc and destDesc.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. Also, in the folding and padding paths, any value other than A=1 and B=0 will result in a CUDNN_STATUS_NOT_SUPPORTED.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.100. cudnnTransformTensor()

cudnnStatus_t cudnnTransformTensor(
    cudnnHandle_t                  handle,
    const void                    *alpha,
    const cudnnTensorDescriptor_t  xDesc,
    const void                    *x,
    const void                    *beta,
    const cudnnTensorDescriptor_t  yDesc,
    void                          *y)

This function copies the scaled data from one tensor to another tensor with a different layout. Those descriptors need to have the same dimensions but not necessarily the same strides. The input and output tensors must not overlap in any way (meaning, tensors cannot be transformed in place). This function can be used to convert a tensor with an unsupported format to a supported one.

Parameters

handle

Input. Handle to a previously created cuDNN context.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with prior value in the destination tensor as follows:
dstValue = alpha[0]*srcValue + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

xDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

x

Input. Pointer to data of the tensor described by the xDesc descriptor.

yDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

y

Output. Pointer to data of the tensor described by the yDesc descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

The dimensions n, c, h, w or the dataType of the two tensor descriptors are different.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

3.2.101. cudnnTransformTensorEx()

cudnnStatus_t cudnnTransformTensorEx(
	cudnnHandle_t handle,
	const cudnnTensorTransformDescriptor_t transDesc,                                    
	const void *alpha,                                           
	const cudnnTensorDescriptor_t srcDesc,                        
	const void *srcData,   
	const void *beta, 
	const cudnnTensorDescriptor_t destDesc, 
	void *destData);

This function converts the tensor layouts between different formats. It can be used to convert a tensor with an unsupported layout format to a tensor with a supported layout format.

This function copies the scaled data from the input tensor srcDesc to the output tensor destDesc with a different layout. The tensor descriptors of srcDesc and destDesc should have the same dimensions but need not have the same strides.

The srcDesc and destDesc tensors must not overlap in any way (meaning, tensors cannot be transformed in place).
Note: When performing a folding transform or a zero-padding transform, the scaling factors (alpha,beta) should be set to (1, 0). However, unfolding transforms support any (alpha,beta) values. This function is thread safe.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

transDesc

Input. A descriptor containing the details of the requested tensor transformation. For more information, see cudnnTensorTransformDescriptor_t.

alpha, beta

Input. Pointers, in the host memory, to the scaling factors used to scale the data in the input tensor srcDesc. beta is used to scale the destination tensor, while alpha is used to scale the source tensor. For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

The beta scaling value is not honored in the folding and zero-padding cases. Unfolding supports any (alpha, beta).

srcDesc, destDesc

Input. Handles to the previously initiated tensor descriptors. srcDesc and destDesc must not overlap. For more information, see cudnnTensorDescriptor_t.

srcData, destData

Input. Pointers, in the host memory, to the data of the tensor described by srcDesc and destDesc respectively.

Returns

CUDNN_STATUS_SUCCESS

The function was launched successfully.

CUDNN_STATUS_BAD_PARAM

A parameter is uninitialized or initialized incorrectly, or the number of dimensions is different between srcDesc and destDesc.

CUDNN_STATUS_NOT_SUPPORTED

Function does not support the provided configuration. Also, in the folding and padding paths, any value other than A=1 and B=0 will result in a CUDNN_STATUS_NOT_SUPPORTED.

CUDNN_STATUS_EXECUTION_FAILED

Function failed to launch on the GPU.

4. cudnn_ops_train.so Library

4.1. API Functions

4.1.1. cudnnActivationBackward()

cudnnStatus_t cudnnActivationBackward(
    cudnnHandle_t                    handle,
    cudnnActivationDescriptor_t      activationDesc,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    yDesc,
    const void                      *y,
    const cudnnTensorDescriptor_t    dyDesc,
    const void                      *dy,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *beta,
    const cudnnTensorDescriptor_t    dxDesc,
    void                            *dx)
This routine computes the gradient of a neuron activation function.
Note:
  • In-place operation is allowed for this routine; meaning dy and dx pointers may be equal. However, this requires the corresponding tensor descriptors to be identical (particularly, the strides of the input and output must match for an in-place operation to be allowed).
  • All tensor formats are supported for 4 and 5 dimensions, however, the best performance is obtained when the strides of yDesc and xDesc are equal and HW-packed. For more than 5 dimensions the tensors must have their spatial dimensions packed.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

activationDesc

Input. Activation descriptor. See cudnnActivationDescriptor_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

yDesc

Input. Handle to the previously initialized input tensor descriptor. For more information, see cudnnTensorDescriptor_t.

y

Input. Data pointer to GPU memory associated with the tensor descriptor yDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyDesc.

xDesc

Input. Handle to the previously initialized output tensor descriptor.

x

Input. Data pointer to GPU memory associated with the output tensor descriptor xDesc.

dxDesc

Input. Handle to the previously initialized output differential tensor descriptor.

dx

Output. Data pointer to GPU memory associated with the output tensor descriptor dxDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The strides nStride, cStride, hStride, wStride of the input differential tensor and output differential tensor differ and in-place operation is used.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimensions n, c, h, w of the input tensor and output tensor differ.
  • The datatype of the input tensor and output tensor differs.
  • The strides nStride, cStride, hStride, wStride of the input tensor and the input differential tensor differ.
  • The strides nStride, cStride, hStride, wStride of the output tensor and the output differential tensor differ.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

4.1.2. cudnnBatchNormalizationBackward()

cudnnStatus_t cudnnBatchNormalizationBackward(
      cudnnHandle_t                    handle,
      cudnnBatchNormMode_t             mode,
      const void                      *alphaDataDiff,
      const void                      *betaDataDiff,
      const void                      *alphaParamDiff,
      const void                      *betaParamDiff,
      const cudnnTensorDescriptor_t    xDesc,
      const void                      *x,
      const cudnnTensorDescriptor_t    dyDesc,
      const void                      *dy,
      const cudnnTensorDescriptor_t    dxDesc,
      void                            *dx,
      const cudnnTensorDescriptor_t    bnScaleBiasDiffDesc,
      const void                      *bnScale,
      void                            *resultBnScaleDiff,
      void                            *resultBnBiasDiff,
      double                           epsilon,
      const void                      *savedMean,
      const void                      *savedInvVariance)
This function performs the backward batch normalization layer computation. This layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015. .
Note:
  • Only 4D and 5D tensors are supported.
  • The epsilon value has to be the same during training, backpropagation, and inference.
  • Higher performance can be obtained when HW-packed tensors are used for all of x, dy, dx.

For more information, see cudnnDeriveBNTensorDescriptor() for the secondary tensor descriptor generation for the parameters used in this function.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

*alphaDataDiff, *betaDataDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient output dx with a prior value in the destination tensor as follows:
dstValue = alphaDataDiff[0]*resultValue + betaDataDiff[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

*alphaParamDiff, *betaParamDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient outputs resultBnScaleDiff and resultBnBiasDiff with prior values in the destination tensor as follows:
dstValue = alphaParamDiff[0]*resultValue + betaParamDiff[0]*priorDstValue

For more information, see Scaling Parameters.

xDesc, dxDesc, dyDesc

Inputs. Handles to the previously initialized tensor descriptors.

*x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc, for the layer’s x data.

*dy

Inputs. Data pointer to GPU memory associated with the tensor descriptor dyDesc, for the backpropagated differential dy input.

*dx

Inputs. Data pointer to GPU memory associated with the tensor descriptor dxDesc, for the resulting differential output with respect to x.

bnScaleBiasDiffDesc
Input. Shared tensor descriptor for the following five tensors: bnScale, resultBnScaleDiff, resultBnBiasDiff, savedMean, savedInvVariance. The dimensions for this tensor descriptor are dependent on normalization mode. For more information, see cudnnDeriveBNTensorDescriptor().
Note: The data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.
*bnScale
Input. Pointer in the device memory for the batch normalization scale parameter (in the original paper the quantity scale is referred to as gamma).
Note: The bnBias parameter is not needed for this layer's computation.
resultBnScaleDiff, resultBnBiasDiff
Outputs. Pointers in device memory for the resulting scale and bias differentials computed by this routine. Note that these scale and bias gradients are weight gradients specific to this batch normalization operation, and by definition are not backpropagated.
epsilon

Input. Epsilon value used in batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h. The same epsilon value should be used in forward and backward functions.

*savedMean, *savedInvVariance
Inputs. Optional cache parameters containing saved intermediate results that were computed during the forward pass. For this to work correctly, the layer's x and bnScale data have to remain unchanged until this backward function is called.
Note: Both these parameters can be NULL but only at the same time. It is recommended to use this cache since the memory overhead is relatively small.

Supported configurations

This function supports the following combinations of data types for various descriptors.
Table 6. Supported configurations
Data Type Configurations xDesc bnScaleBiasMeanVarDesc alphaDataDiff, betaDataDiff,alphaParamDiff, betaParamDiff dyDesc dxDesc
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_HALF CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • Any of the pointers alpha, beta, x, dy, dx, bnScale, resultBnScaleDiff, resultBnBiasDiff is NULL.
  • The number of xDesc or yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • bnScaleBiasDiffDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of savedMean, savedInvVariance pointers is NULL.
  • epsilon value is less than CUDNN_BN_MIN_EPSILON.
  • Dimensions or data types mismatch for any pair of xDesc, dyDesc, dxDesc.

4.1.3. cudnnBatchNormalizationBackwardEx()

cudnnStatus_t cudnnBatchNormalizationBackwardEx (
    cudnnHandle_t                       handle,
    cudnnBatchNormMode_t                mode,
    cudnnBatchNormOps_t                 bnOps,
    const void                          *alphaDataDiff,
    const void                          *betaDataDiff,
    const void                          *alphaParamDiff,
    const void                          *betaParamDiff,
    const cudnnTensorDescriptor_t       xDesc,
    const void                          *xData,
    const cudnnTensorDescriptor_t       yDesc,
    const void                          *yData,
    const cudnnTensorDescriptor_t       dyDesc,
    const void                          *dyData,
    const cudnnTensorDescriptor_t       dzDesc,
    void                                *dzData,
    const cudnnTensorDescriptor_t       dxDesc,
    void                                *dxData,
    const cudnnTensorDescriptor_t       dBnScaleBiasDesc,
    const void                          *bnScaleData,
    const void                          *bnBiasData,
    void                                *dBnScaleData,
    void                                *dBnBiasData,
    double                              epsilon,
    const void                          *savedMean,
    const void                          *savedInvVariance,
    const cudnnActivationDescriptor_t   activationDesc,
    void                                *workspace,
    size_t                              workSpaceSizeInBytes
    void                                *reserveSpace
    size_t                              reserveSpaceSizeInBytes);
This function is an extension of the cudnnBatchNormalizationBackward() for performing the backward batch normalization layer computation with a fast NHWC semi-persistent kernel. This API will trigger the new semi-persistent NHWC kernel when the following conditions are true:

If workspace is NULL and workSpaceSizeInBytes of zero is passed in, this API will function exactly like the non-extended function cudnnBatchNormalizationBackward.

This workspace is not required to be clean. Moreover, the workspace does not have to remain unchanged between the forward and backward pass, as it is not used for passing any information.

This extended function can accept a *workspace pointer to the GPU workspace, and workSpaceSizeInBytes, the size of the workspace, from the user.

The bnOps input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.

Only 4D and 5D tensors are supported. The epsilon value has to be the same during the training, the backpropagation, and the inference.

When the tensor layout is NCHW, higher performance can be obtained when HW-packed tensors are used for x, dy, dx.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

bnOps
Input. Mode of operation. Currently, CUDNN_BATCHNORM_OPS_BN_ACTIVATION and CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnBatchNormOps_t. This input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.
*alphaDataDiff, *betaDataDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient output dx with a prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

*alphaParamDiff, *betaParamDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient outputs dBnScaleData and dBnBiasData with prior values in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, *x, yDesc, *yData, dyDesc, *dyData

Inputs. Tensor descriptors and pointers in the device memory for the layer's x data, backpropagated gradient input dy, the original forward output y data. yDesc and yData are not needed if bnOps is set to CUDNN_BATCHNORM_OPS_BN, users may pass NULL. For more information, see cudnnTensorDescriptor_t.

dzDesc, *dzData, dxDesc, *dxData
Outputs. Tensor descriptors and pointers in the device memory for the computed gradient output dz, and dx. dzDesc and *dzData are not needed when bnOps is CUDNN_BATCHNORM_OPS_BN or CUDNN_BATCHNORM_OPS_BN_ACTIVATION, users may pass NULL. For more information, see cudnnTensorDescriptor_t.
dBnScaleBiasDesc

Input. Shared tensor descriptor for the following six tensors: bnScaleData, bnBiasData, dBnScaleData, dBnBiasData, savedMean, and savedInvVariance. For more information, see cudnnDeriveBNTensorDescriptor().

The dimensions for this tensor descriptor are dependent on normalization mode.
Note: The data type of this tensor descriptor must be float for FP16 and FP32 input tensors and double for FP64 input tensors.
For more information, see cudnnTensorDescriptor_t.
*bnScaleData

Input. Pointer in the device memory for the batch normalization scale parameter (in the original paper the quantity scale is referred to as gamma).

*bnBiasData
Input. Pointers in the device memory for the batch normalization bias parameter (in the original paper bias is referred to as beta). This parameter is used only when activation should be performed.
*dBnScaleData, dBnBiasData
Inputs. Pointers in the device memory for the gradients of bnScaleData and bnBiasData, respectively.
epsilon

Input. Epsilon value used in batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h. The same epsilon value should be used in forward and backward functions.

*savedMean, *savedInvVariance
Inputs. Optional cache parameters containing saved intermediate results computed during the forward pass. For this to work correctly, the layer's x and bnScaleData, bnBiasData data has to remain unchanged until this backward function is called. Note that both these parameters can be NULL but only at the same time. It is recommended to use this cache since the memory overhead is relatively small.
activationDesc
Input. Descriptor for the activation operation. When the bnOps input is set to either CUDNN_BATCHNORM_OPS_BN_ACTIVATION or CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION then this activation is used, otherwise user may pass NULL.
workspace
Input. Pointer to the GPU workspace. If workspace is NULL and workSpaceSizeInBytes of zero is passed in, then this API will function exactly like the non-extended function cudnnBatchNormalizationBackward().
workSpaceSizeInBytes
Input. The size of the workspace. It must be large enough to trigger the fast NHWC semi-persistent kernel by this function.
*reserveSpace
Input. Pointer to the GPU workspace for the reserveSpace.
reserveSpaceSizeInBytes
Input. The size of the reserveSpace. It must be equal or larger than the amount required by cudnnGetBatchNormalizationTrainingExReserveSpaceSize().

Supported configurations

This function supports the following combinations of data types for various descriptors.
Table 7. Supported configurations
Data Type Configurations xDesc, yDesc dBnScaleBiasDesc alphaDataDiff, betaDataDiff,alphaParamDiff, betaParamDiff dyDesc, dzDesc, dxDesc
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • Any of the pointers alphaDataDiff, betaDataDiff, alphaParamDiff, betaParamDiff, x, dy, dx, bnScale, resultBnScaleDiff, resultBnBiasDiff is NULL.
  • The number of xDesc or yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • dBnScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of savedMean, savedInvVariance pointers is NULL.
  • epsilon value is less than CUDNN_BN_MIN_EPSILON.
  • Dimensions or data types mismatch for any pair of xDesc, dyDesc, dxDesc.

4.1.4. cudnnBatchNormalizationForwardTraining()

 cudnnStatus_t cudnnBatchNormalizationForwardTraining(
      cudnnHandle_t                    handle,
      cudnnBatchNormMode_t             mode,
      const void                      *alpha,
      const void                      *beta,
      const cudnnTensorDescriptor_t    xDesc,
      const void                      *x,
      const cudnnTensorDescriptor_t    yDesc,
      void                            *y,
      const cudnnTensorDescriptor_t    bnScaleBiasMeanVarDesc,
      const void                      *bnScale,
      const void                      *bnBias,
      double                           exponentialAverageFactor,
      void                            *resultRunningMean,
      void                            *resultRunningVariance,
      double                           epsilon,
      void                            *resultSaveMean,
      void                            *resultSaveInvVariance)
This function performs the forward batch normalization layer computation for the training phase. This layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015.
Note:
  • Only 4D and 5D tensors are supported.
  • The epsilon value has to be the same during training, backpropagation, and inference.
  • For the inference phase, use cudnnBatchNormalizationForwardInference.
  • Higher performance can be obtained when HW-packed tensors are used for both x and y.
See cudnnDeriveBNTensorDescriptor() for the secondary tensor descriptor generation for the parameters used in this function.

Parameters

handle

Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

alpha, beta
Inputs. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Tensor descriptors and pointers in device memory for the layer's x and y data. For more information, see cudnnTensorDescriptor_t.

*x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc, for the layer’s x input data.

*y

Input. Data pointer to GPU memory associated with the tensor descriptor yDesc, for the youtput of the batch normalization layer.

bnScaleBiasMeanVarDesc

Shared tensor descriptor desc for the secondary tensor that was derived by cudnnDeriveBNTensorDescriptor(). The dimensions for this tensor descriptor are dependent on the normalization mode.

bnScale, bnBias
Inputs. Pointers in device memory for the batch normalization scale and bias parameters (in the original paper bias is referred to as beta and scale as gamma). Note that bnBias parameter can replace the previous layer's bias parameter for improved efficiency.
exponentialAverageFactor
Input. Factor used in the moving average computation as follows:
runningMean = runningMean*(1-factor) + newMean*factor
Use a factor=1/(1+n) at N-th call to the function to get Cumulative Moving Average (CMA) behavior such that:
CMA[n] = (x[1]+...+x[n])/n

This is proved below:

CMA[n+1] = (n*CMA[n]+x[n+1])/(n+1)
= ((n+1)*CMA[n]-CMA[n])/(n+1) + x[n+1]/(n+1)
= CMA[n]*(1-1/(n+1))+x[n+1]*1/(n+1)
= CMA[n]*(1-factor) + x(n+1)*factor
resultRunningMean, resultRunningVariance

Inputs/Outputs. Running mean and variance tensors (these have the same descriptor as the bias and scale). Both of these pointers can be NULL but only at the same time. The value stored in resultRunningVariance (or passed as an input in inference mode) is the sample variance and is the moving average of variance[x] where the variance is computed either over batch or spatial+batch dimensions depending on the mode. If these pointers are not NULL, the tensors should be initialized to some reasonable values or to 0.

epsilon

Input. Epsilon value used in the batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h. The same epsilon value should be used in forward and backward functions.

resultSaveMean, resultSaveInvVariance
Outputs. Optional cache to save intermediate results computed during the forward pass. These buffers can be used to speed up the backward pass when supplied to the cudnnBatchNormalizationBackward() function. The intermediate results stored in resultSaveMean and resultSaveInvVariance buffers should not be used directly by the user. Depending on the batch normalization mode, the results stored in resultSaveInvVariance may vary. For the cache to work correctly, the input layer data must remain unchanged until the backward function is called. Note that both parameters can be NULL but only at the same time. In such a case, intermediate statistics will not be saved, and cudnnBatchNormalizationBackward() will have to re-compute them. It is recommended to use this cache as the memory overhead is relatively small because these tensors have a much lower product of dimensions than the data tensors.

Supported configurations

This function supports the following combinations of data types for various descriptors.
Table 8. Supported configurations
Data Type Configurations xDesc bnScaleBiasMeanVarDesc alpha, beta yDesc
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the pointers alpha, beta, x, y, bnScale, bnBias is NULL.
  • The number of xDesc or yDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • bnScaleBiasMeanVarDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of resultSaveMean, resultSaveInvVariance pointers are NULL.
  • Exactly one of resultRunningMean, resultRunningInvVariance pointers are NULL.
  • epsilon value is less than CUDNN_BN_MIN_EPSILON.
  • Dimensions or data types mismatch for xDesc, yDesc.

4.1.5. cudnnBatchNormalizationForwardTrainingEx()

 cudnnStatus_t cudnnBatchNormalizationForwardTrainingEx(
    cudnnHandle_t                       handle,
    cudnnBatchNormMode_t                mode,
    cudnnBatchNormOps_t                 bnOps,
    const void                          *alpha,
    const void                          *beta,
    const cudnnTensorDescriptor_t       xDesc,
    const void                          *xData,
    const cudnnTensorDescriptor_t       zDesc, 
    const void                          *zData, 
    const cudnnTensorDescriptor_t       yDesc,
    void                                *yData,
    const cudnnTensorDescriptor_t       bnScaleBiasMeanVarDesc,
    const void                          *bnScaleData,
    const void                          *bnBiasData, 
    double                              exponentialAverageFactor,
    void                                *resultRunningMeanData,
    void                                *resultRunningVarianceData,
    double                              epsilon,
    void                                *saveMean,
    void                                *saveInvVariance,
    const cudnnActivationDescriptor_t   activationDesc, 
    void                                *workspace,
    size_t                              workSpaceSizeInBytes
    void                                *reserveSpace
    size_t                              reserveSpaceSizeInBytes);

This function is an extension of the cudnnBatchNormalizationForwardTraining() for performing the forward batch normalization layer computation.

This API will trigger the new semi-persistent NHWC kernel when the following conditions are true:

If workspace is NULL and workSpaceSizeInBytes of zero is passed in, this API will function exactly like the non-extended function cudnnBatchNormalizationForwardTraining().

This workspace is not required to be clean. Moreover, the workspace does not have to remain unchanged between the forward and backward pass, as it is not used for passing any information.

This extended function can accept a *workspace pointer to the GPU workspace, and workSpaceSizeInBytes, the size of the workspace, from the user.

The bnOps input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.

Only 4D and 5D tensors are supported. The epsilon value has to be the same during the training, the backpropagation, and the inference.

When the tensor layout is NCHW, higher performance can be obtained when HW-packed tensors are used for x, dy, dx.

Parameters

handle

Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

bnOps
Input. Mode of operation for the fast NHWC kernel. See cudnnBatchNormOps_t. This input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.
*alpha, *beta
Inputs. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, *xData, zDesc, *zData, yDesc, *yData

Tensor descriptors and pointers in device memory for the layer's input x and output y, and for the optional z tensor input for residual addition to the result of the batch normalization operation, prior to the activation. The optional zDes and *zData descriptors are only used when bnOps is CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION, otherwise users may pass NULL. When in use, z should have exactly the same dimension as x and the final output y. For more information, see cudnnTensorDescriptor_t.

bnScaleBiasMeanVarDesc

Shared tensor descriptor desc for the secondary tensor that was derived by cudnnDeriveBNTensorDescriptor(). The dimensions for this tensor descriptor are dependent on the normalization mode.

*bnScaleData, *bnBiasData
Inputs. Pointers in device memory for the batch normalization scale and bias parameters (in the original paper, bias is referred to as beta and scale as gamma). Note that bnBiasData parameter can replace the previous layer's bias parameter for improved efficiency.
exponentialAverageFactor
Input. Factor used in the moving average computation as follows:
runningMean = runningMean*(1-factor) + newMean*factor
Use a factor=1/(1+n) at N-th call to the function to get Cumulative Moving Average (CMA) behavior such that:
CMA[n] = (x[1]+...+x[n])/n

This is proved below:

Writing
CMA[n+1] = (n*CMA[n]+x[n+1])/(n+1)
= ((n+1)*CMA[n]-CMA[n])/(n+1) + x[n+1]/(n+1)
= CMA[n]*(1-1/(n+1))+x[n+1]*1/(n+1)
= CMA[n]*(1-factor) + x(n+1)*factor
*resultRunningMeanData, *resultRunningVarianceData

Inputs/Outputs. Pointers to the running mean and running variance data. Both these pointers can be NULL but only at the same time. The value stored in resultRunningVarianceData (or passed as an input in inference mode) is the sample variance and is the moving average of variance[x] where the variance is computed either over batch or spatial+batch dimensions depending on the mode. If these pointers are not NULL, the tensors should be initialized to some reasonable values or to 0.

epsilon

Input. Epsilon value used in the batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h. The same epsilon value should be used in forward and backward functions.

*saveMean, *saveInvVariance
Outputs. Optional cache parameters containing saved intermediate results computed during the forward pass. For this to work correctly, the layer's x and bnScaleData, bnBiasData data has to remain unchanged until this backward function is called. Note that both these parameters can be NULL but only at the same time. It is recommended to use this cache since the memory overhead is relatively small.
activationDesc
Input. The tensor descriptor for the activation operation. When the bnOps input is set to either CUDNN_BATCHNORM_OPS_BN_ACTIVATION or CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION then this activation is used, otherwise user may pass NULL.
*workspace, workSpaceSizeInBytes
Inputs. *workspace is a pointer to the GPU workspace, and workSpaceSizeInBytes is the size of the workspace. When *workspace is not NULL and *workSpaceSizeInBytes is large enough, and the tensor layout is NHWC and the data type configuration is supported, then this function will trigger a new semi-persistent NHWC kernel for batch normalization. The workspace is not required to be clean. Also, the workspace does not need to remain unchanged between the forward and backward passes.
*reserveSpace
Input. Pointer to the GPU workspace for the reserveSpace.
reserveSpaceSizeInBytes
Input. The size of the reserveSpace. Must be equal or larger than the amount required by cudnnGetBatchNormalizationTrainingExReserveSpaceSize().

Supported configurations

This function supports the following combinations of data types for various descriptors.
Table 9. Supported configurations
Data Type Configurations xDesc bnScaleBiasMeanVarDesc alpha, beta zDesc yDesc
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_HALF CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT Not supported CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE Not supported CUDNN_DATA_DOUBLE

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the pointers alpha, beta, x, y, bnScaleData, bnBiasData is NULL.
  • The number of xDesc or yDesc tensor descriptor dimensions is not within the [4,5] range (only 4D and 5D tensors are supported).
  • bnScaleBiasMeanVarDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of saveMean, saveInvVariance pointers are NULL.
  • Exactly one of resultRunningMeanData, resultRunningInvVarianceData pointers are NULL.
  • epsilon value is less than CUDNN_BN_MIN_EPSILON.
  • Dimensions or data types mismatch for xDesc, yDesc.

4.1.6. cudnnDivisiveNormalizationBackward()

 cudnnStatus_t cudnnDivisiveNormalizationBackward(
    cudnnHandle_t                    handle,
    cudnnLRNDescriptor_t             normDesc,
    cudnnDivNormMode_t               mode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *means,
    const void                      *dy,
    void                            *temp,
    void                            *temp2,
    const void                      *beta,
    const cudnnTensorDescriptor_t    dxDesc,
    void                            *dx,
    void                            *dMeans)

This function performs the backward DivisiveNormalization layer computation.

Note: Supported tensor formats are NCHW for 4D and NCDHW for 5D with any non-overlapping non-negative strides. Only 4D and 5D tensors are supported.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

normDesc

Input. Handle to a previously initialized LRN parameter descriptor (this descriptor is used for both LRN and DivisiveNormalization layers).

mode

Input. DivisiveNormalization layer mode of operation. Currently only CUDNN_DIVNORM_PRECOMPUTED_MEANS is implemented. Normalization is performed using the means input tensor that is expected to be precomputed by the user.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, x, means

Input. Tensor descriptor and pointers in device memory for the layer's x and means data. Note that the means tensor is expected to be precomputed by the user. It can also contain any valid values (not required to be actual means, and can be for instance a result of a convolution with a Gaussian kernel).

dy

Input. Tensor pointer in device memory for the layer's dy cumulative loss differential data (error backpropagation).

temp, temp2

Workspace. Temporary tensors in device memory. These are used for computing intermediate values during the backward pass. These tensors do not have to be preserved from forward to backward pass. Both use xDesc as a descriptor.

dxDesc

Input. Tensor descriptor for dx and dMeans.

dx, dMeans

Output. Tensor pointers (in device memory) for the layers resulting cumulative gradients dx and dMeans (dLoss/dx and dLoss/dMeans). Both share the same descriptor.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the tensor pointers x, dx, temp, tmep2, dy is NULL.
  • Number of any of the input or output tensor dimensions is not within the [4,5] range.
  • Either alpha or beta pointer is NULL.
  • A mismatch in dimensions between xDesc and dxDesc.
  • LRN descriptor parameters are outside of their valid ranges.
  • Any of the tensor strides is negative.
CUDNN_STATUS_UNSUPPORTED

The function does not support the provided configuration, for example, any of the input and output tensor strides mismatch (for the same dimension) is a non-supported configuration.

4.1.7. cudnnDropoutBackward()

cudnnStatus_t cudnnDropoutBackward(
    cudnnHandle_t                   handle,
    const cudnnDropoutDescriptor_t  dropoutDesc,
    const cudnnTensorDescriptor_t   dydesc,
    const void                     *dy,
    const cudnnTensorDescriptor_t   dxdesc,
    void                           *dx,
    void                           *reserveSpace,
    size_t                          reserveSpaceSizeInBytes)

This function performs backward dropout operation over dy returning results in dx. If during forward dropout operation value from x was propagated to y then during backward operation value from dy will be propagated to dx, otherwise, dx value will be set to 0.

Note: Better performance is obtained for fully packed tensors.

Parameters

handle

Input. Handle to a previously created cuDNN context.

dropoutDesc

Input. Previously created dropout descriptor object.

dyDesc

Input. Handle to a previously initialized tensor descriptor.

dy

Input. Pointer to data of the tensor described by the dyDesc descriptor.

dxDesc

Input. Handle to a previously initialized tensor descriptor.

dx

Output. Pointer to data of the tensor described by the dxDesc descriptor.

reserveSpace

Input. Pointer to user-allocated GPU memory used by this function. It is expected that reserveSpace was populated during a call to cudnnDropoutForward and has not been changed.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided memory for the reserve space

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The number of elements of input tensor and output tensors differ.
  • The datatype of the input tensor and output tensors differs.
  • The strides of the input tensor and output tensors differ and in-place operation is used (i.e., x and y pointers are equal).
  • The provided reserveSpaceSizeInBytes is less then the value returned by cudnnDropoutGetReserveSpaceSize.
  • cudnnSetDropoutDescriptor has not been called on dropoutDesc with the non-NULLstates argument.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

cudnnGetBatchNormalizationBackwardExWorkspaceSize()

cudnnStatus_t cudnnGetBatchNormalizationBackwardExWorkspaceSize(
    cudnnHandle_t                       handle,
    cudnnBatchNormMode_t                mode,
    cudnnBatchNormOps_t                 bnOps,
    const cudnnTensorDescriptor_t       xDesc,
    const cudnnTensorDescriptor_t       yDesc, 
    const cudnnTensorDescriptor_t       dyDesc,
    const cudnnTensorDescriptor_t       dzDesc, 
    const cudnnTensorDescriptor_t       dxDesc,
    const cudnnTensorDescriptor_t       dBnScaleBiasDesc,
    const cudnnActivationDescriptor_t   activationDesc, 
    size_t                              *sizeInBytes);

This function returns the amount of GPU memory workspace the user should allocate to be able to call cudnnGetBatchNormalizationBackwardExWorkspaceSize() function for the specified bnOps input setting. The workspace allocated will then be passed to the function cudnnGetBatchNormalizationBackwardExWorkspaceSize().

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

bnOps

Input. Mode of operation for the fast NHWC kernel. For more information, see cudnnBatchNormOps_t. This input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.

xDesc, yDesc, dyDesc, dzDesc, dxDesc

Tensor descriptors and pointers in the device memory for the layer's x data, back propagated differential dy (inputs), the optional y input data, the optional dz output, and the dx output, which is the resulting differential with respect to x. For more information, see cudnnTensorDescriptor_t.

dBnScaleBiasDesc

Input. Shared tensor descriptor for the following six tensors: bnScaleData, bnBiasData, dBnScaleData, dBnBiasData, savedMean, and savedInvVariance. This is the shared tensor descriptor desc for the secondary tensor that was derived by cudnnDeriveBNTensorDescriptor(). The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

activationDesc
Input. Descriptor for the activation operation. When the bnOps input is set to either CUDNN_BATCHNORM_OPS_BN_ACTIVATION or CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION, then this activation is used, otherwise user may pass NULL.
*sizeInBytes

Output. Amount of GPU memory required for the workspace, as determined by this function, to be able to execute the cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize() function with the specified bnOps input setting.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • Number of xDesc, yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • dBnScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Dimensions or data types mismatch for any pair of xDesc, dyDesc, dxDesc.

cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize()

 cudnnStatus_t cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize(
    cudnnHandle_t                           handle,
    cudnnBatchNormMode_t                    mode,
    cudnnBatchNormOps_t                     bnOps,
    const cudnnTensorDescriptor_t           xDesc,
    const cudnnTensorDescriptor_t           zDesc,
    const cudnnTensorDescriptor_t           yDesc, 
    const cudnnTensorDescriptor_t           bnScaleBiasMeanVarDesc,
    const cudnnActivationDescriptor_t       activationDesc, 
    size_t                                  *sizeInBytes);

This function returns the amount of GPU memory workspace the user should allocate to be able to call cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize() function for the specified bnOps input setting. The workspace allocated should then be passed by the user to the function cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize().

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

bnOps
Input. Mode of operation for the fast NHWC kernel. For more information, see cudnnBatchNormOps_t. This input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.
xDesc, zDesc, yDesc

Tensor descriptors and pointers in the device memory for the layer's x data, the optional z input data, and the y output. zDesc is only needed when bnOps is CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION, otherwise the user may pass NULL. For more information, see cudnnTensorDescriptor_t.

bnScaleBiasMeanVarDesc

Input. Shared tensor descriptor for the following six tensors: bnScaleData, bnBiasData, dBnScaleData, dBnBiasData, savedMean, and savedInvVariance. This is the shared tensor descriptor desc for the secondary tensor that was derived by cudnnDeriveBNTensorDescriptor(). The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

activationDesc
Input. Descriptor for the activation operation. When the bnOps input is set to either CUDNN_BATCHNORM_OPS_BN_ACTIVATION or CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION then this activation is used, otherwise the user may pass NULL.
*sizeInBytes
Output. Amount of GPU memory required for the workspace, as determined by this function, to be able to execute the cudnnGetBatchNormalizationForwardTrainingExWorkspaceSize() function with the specified bnOps input setting.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • Number of xDesc, yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • dBnScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for spatial, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Dimensions or data types mismatch for xDesc, yDesc.

4.1.10. cudnnGetBatchNormalizationTrainingExReserveSpaceSize()

cudnnStatus_t cudnnGetBatchNormalizationTrainingExReserveSpaceSize(
    cudnnHandle_t                       handle,
    cudnnBatchNormMode_t                mode,
    cudnnBatchNormOps_t                 bnOps,
    const cudnnActivationDescriptor_t   activationDesc, 
    const cudnnTensorDescriptor_t       xDesc, 
    size_t                              *sizeInBytes);

This function returns the amount of reserve GPU memory workspace the user should allocate for the batch normalization operation, for the specified bnOps input setting. In contrast to the workspace, the reserved space should be preserved between the forward and backward calls, and the data should not be altered.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode

Input. Mode of operation (spatial or per-activation). For more information, see cudnnBatchNormMode_t.

bnOps
Input. Mode of operation for the fast NHWC kernel. For more information, see cudnnBatchNormOps_t. This input can be used to set this function to perform either only the batch normalization, or batch normalization followed by activation, or batch normalization followed by element-wise addition and then activation.
xDesc

Tensor descriptors for the layer's x data. For more information, see cudnnTensorDescriptor_t.

activationDesc
Input. Descriptor for the activation operation. When the bnOps input is set to either CUDNN_BATCHNORM_OPS_BN_ACTIVATION or CUDNN_BATCHNORM_OPS_BN_ADD_ACTIVATION then this activation is used, otherwise user may pass NULL.
*sizeInBytes
Output. Amount of GPU memory reserved.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The xDesc tensor descriptor dimension is not within the [4,5] range (only 4D and 5D tensors are supported).

4.1.11. cudnnGetNormalizationBackwardWorkspaceSize()

cudnnStatus_t
cudnnGetNormalizationBackwardWorkspaceSize(cudnnHandle_t handle,
                                           cudnnNormMode_t mode,
                                           cudnnNormOps_t normOps,
                                           cudnnNormAlgo_t algo,
                                           const cudnnTensorDescriptor_t xDesc,
                                           const cudnnTensorDescriptor_t yDesc,
                                           const cudnnTensorDescriptor_t dyDesc,
     	                                  const cudnnTensorDescriptor_t dzDesc,
                                           const cudnnTensorDescriptor_t dxDesc,
                                           const cudnnTensorDescriptor_t dNormScaleBiasDesc,
      	                                 const cudnnActivationDescriptor_t activationDesc,
                                           const cudnnTensorDescriptor_t normMeanVarDesc,
                                           size_t *sizeInBytes,
                                           int groupCnt);

This function returns the amount of GPU memory workspace the user should allocate to be able to call cudnnNormalizationBackward() function for the specified normOps and algo input setting. The workspace allocated will then be passed to the function cudnnNormalizationBackward().

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode
Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.
normOps

Input. Mode of post-operative. Currently CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnNormOps_t. This input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

xDesc, yDesc, dyDesc, dzDesc, dxDesc

Tensor descriptors and pointers in the device memory for the layer's x data, back propagated differential dy (inputs), the optional y input data, the optional dz output, and the dx output, which is the resulting differential with respect to x. For more information, see cudnnTensorDescriptor_t.

dNormScaleBiasDesc

Input. Shared tensor descriptor for the following four tensors: normScaleData, normBiasData, dNormScaleData, dNormBiasData. The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

activationDesc

Input. Descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION, then this activation is used, otherwise the user may pass NULL.

normMeanVarDesc

Input. Shared tensor descriptor for the following tensors: savedMean and savedInvVariance. The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

*sizeInBytes

Output. Amount of GPU memory required for the workspace, as determined by this function, to be able to execute the cudnnGetNormalizationForwardTrainingWorkspaceSize() function with the specified normOps input setting.

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • Number of xDesc, yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • dNormScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for per-channel, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Dimensions or data types mismatch for any pair of xDesc, dyDesc, dxDesc.

4.1.12. cudnnGetNormalizationForwardTrainingWorkspaceSize()

cudnnStatus_t
cudnnGetNormalizationForwardTrainingWorkspaceSize(cudnnHandle_t handle,
                                                  cudnnNormMode_t mode,
                                                  cudnnNormOps_t normOps,
                                                  cudnnNormAlgo_t algo,
                                  	            const cudnnTensorDescriptor_t xDesc,
                                                  const cudnnTensorDescriptor_t zDesc,
                                                  const cudnnTensorDescriptor_t yDesc,
                             	                 const cudnnTensorDescriptor_t normScaleBiasDesc,
                                                  const cudnnActivationDescriptor_t activationDesc,
                                                  const cudnnTensorDescriptor_t normMeanVarDesc,
                                                  size_t *sizeInBytes,
                                                  int groupCnt);

This function returns the amount of GPU memory workspace the user should allocate to be able to call cudnnNormalizationForwardTraining() function for the specified normOps and algo input setting. The workspace allocated should then be passed by the user to the function cudnnNormalizationForwardTraining().

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode
Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.
normOps

Input. Mode of post-operative. Currently CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnNormOps_t. This input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

xDesc, zDesc, yDesc

Tensor descriptors and pointers in the device memory for the layer's x data, the optional z input data, and the y output. zDesc is only needed when normOps is CUDNN_NORM_OPS_NORM_ADD_ACTIVATION, otherwise the user may pass NULL. For more information, see cudnnTensorDescriptor_t.

normScaleBiasDesc

Input. Shared tensor descriptor for the following tensors: normScaleData and normBiasData. The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

activationDesc

Input. Descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION, then this activation is used, otherwise the user may pass NULL.

normMeanVarDesc

Input. Shared tensor descriptor for the following tensors: savedMean and savedInvVariance. The dimensions for this tensor descriptor are dependent on normalization mode. Note that the data type of this tensor descriptor must be float for FP16 and FP32 input tensors, and double for FP64 input tensors.

*sizeInBytes

Output. Amount of GPU memory required for the workspace, as determined by this function, to be able to execute the cudnnGetNormalizationForwardTrainingWorkspaceSize() function with the specified normOps input setting.

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • Number of xDesc, yDesc or zDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • normScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for per-channel, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Dimensions or data types mismatch for xDesc, yDesc.

4.1.13. cudnnGetNormalizationTrainingReserveSpaceSize()

cudnnStatus_t
cudnnGetNormalizationTrainingReserveSpaceSize(cudnnHandle_t handle,
                                              cudnnNormMode_t mode,
                     	                     cudnnNormOps_t normOps,
                                              cudnnNormAlgo_t algo,
                                              const cudnnActivationDescriptor_t activationDesc,
                                          	const cudnnTensorDescriptor_t xDesc,
                                              size_t *sizeInBytes,
                                              int groupCnt);

This function returns the amount of reserve GPU memory workspace the user should allocate for the normalization operation, for the specified normOps input setting. In contrast to the workspace, the reserved space should be preserved between the forward and backward calls, and the data should not be altered.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode
Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.
normOps

Input. Mode of post-operative. Currently CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnNormOps_t . This input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

xDesc

Tensor descriptors for the layer's x data. For more information, see cudnnTensorDescriptor_t.

activationDesc

Input. Descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION then this activation is used, otherwise the user may pass NULL.

*sizeInBytes

Output. Amount of GPU memory reserved.

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • The xDesc tensor descriptor dimension is not within the [4,5] range (only 4D and 5D tensors are supported).

4.1.14. cudnnLRNCrossChannelBackward()

cudnnStatus_t cudnnLRNCrossChannelBackward(
    cudnnHandle_t                    handle,
    cudnnLRNDescriptor_t             normDesc,
    cudnnLRNMode_t                   lrnMode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    yDesc,
    const void                      *y,
    const cudnnTensorDescriptor_t    dyDesc,
    const void                      *dy,
    const cudnnTensorDescriptor_t    xDesc,
    const void                      *x,
    const void                      *beta,
    const cudnnTensorDescriptor_t    dxDesc,
    void                            *dx)

This function performs the backward LRN layer computation.

Note: Supported formats are: positive-strided, NCHW and NHWC for 4D x and y, and only NCDHW DHW-packed for 5D (for both x and y). Only non-overlapping 4D and 5D tensors are supported. NCHW layout is preferred for performance.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

normDesc

Input. Handle to a previously initialized LRN parameter descriptor.

lrnMode

Input. LRN layer mode of operation. Currently, only CUDNN_LRN_CROSS_CHANNEL_DIM1 is implemented. Normalization is performed along the tensor's dimA[1].

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

yDesc, y

Input. Tensor descriptor and pointer in device memory for the layer's y data.

dyDesc, dy

Input. Tensor descriptor and pointer in device memory for the layer's input cumulative loss differential data dy (including error backpropagation).

xDesc, x

Input. Tensor descriptor and pointer in device memory for the layer's x data. Note that these values are not modified during backpropagation.

dxDesc, dx

Output. Tensor descriptor and pointer in device memory for the layer's resulting cumulative loss differential data dx (including error backpropagation).

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the tensor pointers x, y is NULL.
  • Number of input tensor dimensions is 2 or less.
  • LRN descriptor parameters are outside of their valid ranges.
  • One of the tensor parameters is 5D but is not in NCDHW DHW-packed format.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • Any of the input tensor datatypes is not the same as any of the output tensor datatype.
  • Any pairwise tensor dimensions mismatch for x, y, dx, dy.
  • Any tensor parameters strides are negative.

4.1.15. cudnnNormalizationBackward()

cudnnStatus_t
cudnnNormalizationBackward(cudnnHandle_t handle,
                           cudnnNormMode_t mode,
                           cudnnNormOps_t normOps,
                           cudnnNormAlgo_t algo,
                           const void *alphaDataDiff,
          	             const void *betaDataDiff,
                           const void *alphaParamDiff,
                           const void *betaParamDiff,
                           const cudnnTensorDescriptor_t xDesc,
                           const void *xData,
                           const cudnnTensorDescriptor_t yDesc,
                           const void *yData,
                           const cudnnTensorDescriptor_t dyDesc,
                           const void *dyData,
               	        const cudnnTensorDescriptor_t dzDesc,
                           void *dzData,
                           const cudnnTensorDescriptor_t dxDesc,
                           void *dxData,
                           const cudnnTensorDescriptor_t dNormScaleBiasDesc,
                           const void *normScaleData,
                           const void *normBiasData,
                           void *dNormScaleData,
                           void *dNormBiasData,
                           double epsilon,
                           const cudnnTensorDescriptor_t normMeanVarDesc,
                           const void *savedMean,
                           const void *savedInvVariance,
                           cudnnActivationDescriptor_t activationDesc,
                           void *workSpace,
                           size_t workSpaceSizeInBytes,
                           void *reserveSpace,
                           size_t reserveSpaceSizeInBytes,
           	            int groupCnt)
This function performs backward normalization layer computation that is specified by mode. Per-channel normalization layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015.
Note: Only 4D and 5D tensors are supported.

The epsilon value has to be the same during training, backpropagation, and inference. This workspace is not required to be clean. Moreover, the workspace does not have to remain unchanged between the forward and backward pass, as it is not used for passing any information.

This function can accept a *workspace pointer to the GPU workspace, and workSpaceSizeInBytes, the size of the workspace, from the user.

The normOps input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

When the tensor layout is NCHW, higher performance can be obtained when HW-packed tensors are used for x, dy, dx.

Higher performance for CUDNN_NORM_PER_CHANNEL mode can be obtained when the following conditions are true:
  • All tensors, namely, x, y, dz, dy, and dx must be NHWC-fully packed, and must be of the type CUDNN_DATA_HALF.
  • The tensor C dimension should be a multiple of 4.
  • The input parameter mode must be set to CUDNN_NORM_PER_CHANNEL.
  • The input parameter algo must be set to CUDNN_NORM_ALGO_PERSIST.
  • Workspace is not NULL.
  • workSpaceSizeInBytes is equal or larger than the amount required by cudnnGetNormalizationBackwardWorkspaceSize().
  • reserveSpaceSizeInBytes is equal or larger than the amount required by cudnnGetNormalizationTrainingReserveSpaceSize().
  • The content in reserveSpace stored by cudnnNormalizationForwardTraining() must be preserved.

Parameters

handle
Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.
mode
Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.
normOps

Input. Mode of post-operative. Currently CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnNormOps_t. This input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

*alphaDataDiff, *betaDataDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient output dx with a prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

*alphaParamDiff, *betaParamDiff
Inputs. Pointers to scaling factors (in host memory) used to blend the gradient outputs dNormScaleData and dNormBiasData with prior values in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, *xData, yDesc, *yData, dyDesc, *dyData

Inputs. Tensor descriptors and pointers in the device memory for the layer's x data, backpropagated gradient input dy, the original forward output y data. yDesc and yData are not needed if normOps is set to CUDNN_NORM_OPS_NORM, users may pass NULL. For more information, see cudnnTensorDescriptor_t.

dzDesc, *dzData, dxDesc, *dxData

Outputs. Tensor descriptors and pointers in the device memory for the computed gradient output dz and dx. dzDesc and *dzData is not needed when normOps is CUDNN_NORM_OPS_NORM or CUDNN_NORM_OPS_NORM_ACTIVATION, users may pass NULL. For more information, see cudnnTensorDescriptor_t.

dNormScaleBiasDesc
Input. Shared tensor descriptor for the following six tensors: normScaleData, normBiasData, dNormScaleData, and dNormBiasData. The dimensions for this tensor descriptor are dependent on normalization mode.
Note: The data type of this tensor descriptor must be float for FP16 and FP32 input tensors and double for FP64 input tensors.

For more information, see cudnnTensorDescriptor_t.

*normScaleData

Input. Pointer in the device memory for the normalization scale parameter (in the original paper the quantity scale is referred to as gamma).

*normBiasData

Input. Pointers in the device memory for the normalization bias parameter (in the original paper bias is referred to as beta). This parameter is used only when activation should be performed.

*dNormScaleData, dNormBiasData

Inputs. Pointers in the device memory for the gradients of normScaleData and normBiasData, respectively.

epsilon

Input. Epsilon value used in normalization formula. Its value should be equal to or greater than zero. The same epsilon value should be used in forward and backward functions.

normMeanVarDesc
Input. Shared tensor descriptor for the following tensors: savedMean and savedInvVariance. The dimensions for this tensor descriptor are dependent on normalization mode.
Note: The data type of this tensor descriptor must be float for FP16 and FP32 input tensors and double for FP64 input tensors.

For more information, see cudnnTensorDescriptor_t.

*savedMean, *savedInvVariance

Inputs. Optional cache parameters containing saved intermediate results computed during the forward pass. For this to work correctly, the layer's x and normScaleData, normBiasData data has to remain unchanged until this backward function is called. Note that both these parameters can be NULL but only at the same time. It is recommended to use this cache since the memory overhead is relatively small.

activationDesc

Input. Descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION then this activation is used, otherwise the user may pass NULL.

workspace

Input. Pointer to the GPU workspace.

workSpaceSizeInBytes

Input. The size of the workspace. It must be large enough to trigger the fast NHWC semi-persistent kernel by this function.

*reserveSpace

Input. Pointer to the GPU workspace for the reserveSpace.

reserveSpaceSizeInBytes

Input. The size of the reserveSpace. It must be equal or larger than the amount required by cudnnGetNormalizationTrainingReserveSpaceSize().

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • Any of the pointers alphaDataDiff, betaDataDiff, alphaParamDiff, betaParamDiff, xData, dyData, dxData, normScaleData, dNormScaleData, and dNormBiasData is NULL.
  • The number of xDesc or yDesc or dxDesc tensor descriptor dimensions is not within the range of [4,5] (only 4D and 5D tensors are supported).
  • dNormScaleBiasDesc dimensions not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for per-channel, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of savedMean, savedInvVariance pointers is NULL.
  • epsilon value is less than zero.
  • Dimensions or data types mismatch for any pair of xDesc, dyDesc, dxDesc, dNormScaleBiasDesc, and normMeanVarDesc.

4.1.16. cudnnNormalizationForwardTraining()

cudnnStatus_t
cudnnNormalizationForwardTraining(cudnnHandle_t handle,
                                  cudnnNormMode_t mode,
                                  cudnnNormOps_t normOps,
                                  cudnnNormAlgo_t algo,
                              	const void *alpha,
                              	const void *beta,
                              	const cudnnTensorDescriptor_t xDesc,
                     	         const void *xData,
                              	const cudnnTensorDescriptor_t normScaleBiasDesc,
                              	const void *normScale,
                              	const void *normBias,
                            	  double exponentialAverageFactor,
                              	const cudnnTensorDescriptor_t normMeanVarDesc,
                              	void *resultRunningMean,
                           	   void *resultRunningVariance,
                                  double epsilon,
                              	void *resultSaveMean,
                              	void *resultSaveInvVariance,
                                  cudnnActivationDescriptor_t activationDesc,
                       	       const cudnnTensorDescriptor_t zDesc,
                              	const void *zData,
                              	const cudnnTensorDescriptor_t yDesc,
                              	void *yData,
                              	void *workspace,
                              	size_t workSpaceSizeInBytes,
                              	void *reserveSpace,
                              	size_t reserveSpaceSizeInBytes,
                              	int groupCnt);
This function performs the forward normalization layer computation for the training phase. Depending on mode, different normalization operations will be performed. Per-channel layer is based on the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift, S. Ioffe, C. Szegedy, 2015.
Note:
  • Only 4D and 5D tensors are supported.
  • The epsilon value has to be the same during training, back propagation, and inference.
  • For the inference phase, use cudnnNormalizationForwardInference().
  • Higher performance can be obtained when HW-packed tensors are used for both x and y.
This API will trigger the new semi-persistent NHWC kernel when the following conditions are true:

This workspace is not required to be clean. Moreover, the workspace does not have to remain unchanged between the forward and backward pass, as it is not used for passing any information. This extended function can accept a *workspace pointer to the GPU workspace, and workSpaceSizeInBytes, the size of the workspace, from the user.

The normOps input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

Only 4D and 5D tensors are supported. The epsilon value has to be the same during the training, the backpropagation, and the inference.

When the tensor layout is NCHW, higher performance can be obtained when HW-packed tensors are used for xData, yData.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor. For more information, see cudnnHandle_t.

mode
Input. Mode of operation (per-channel or per-activation). For more information, see cudnnNormMode_t.
normOps

Input. Mode of post-operative. Currently CUDNN_NORM_OPS_NORM_ACTIVATION and CUDNN_NORM_OPS_NORM_ADD_ACTIVATION are only supported in the NHWC layout. For more information, see cudnnNormOps_t. This input can be used to set this function to perform either only the normalization, or normalization followed by activation, or normalization followed by element-wise addition and then activation.

algo

Input. Algorithm to be performed. For more information, see cudnnNormAlgo_t.

*alpha, *beta
Inputs. Pointers to scaling factors (in host memory) used to blend the layer output value with prior value in the destination tensor as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc, yDesc

Input. Handles to the previously initialized tensor descriptors.

*xData

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc, for the layer’s x input data.

*yData

Output. Data pointer to GPU memory associated with the tensor descriptor yDesc, for the y output of the normalization layer.

zDesc, *zData

Input. Tensor descriptors and pointers in device memory for residual addition to the result of the normalization operation, prior to the activation. zDesc and *zData are optional and are only used when normOps is CUDNN_NORM_OPS_NORM_ADD_ACTIVATION, otherwise the user may pass NULL. When in use, z should have exactly the same dimension as xData and the final output yData. For more information, see cudnnTensorDescriptor_t.

normScaleBiasDesc, normScale, normBias

Inputs. Tensor descriptors and pointers in device memory for the normalization scale and bias parameters (in the original paper bias is referred to as beta and scale as gamma). The dimensions for the tensor descriptor are dependent on the normalization mode.

exponentialAverageFactor
Input. Factor used in the moving average computation as follows:
runningMean = runningMean*(1-factor) + newMean*factor
Use a factor=1/(1+n) at N-th call to the function to get Cumulative Moving Average (CMA) behavior such that:
CMA[n] = (x[1]+...+x[n])/n
This is proved below:
Writing
CMA[n+1] = (n*CMA[n]+x[n+1])/(n+1)
= ((n+1)*CMA[n]-CMA[n])/(n+1) + x[n+1]/(n+1)
= CMA[n]*(1-1/(n+1))+x[n+1]*1/(n+1)
= CMA[n]*(1-factor) + x(n+1)*factor
normMeanVarDesc

Inputs. Tensor descriptor used for following tensors: resultRunningMean, resultRunningVariance, resultSaveMean, resultSaveInvVariance.

*resultRunningMean, *resultRunningVariance

Inputs/Outputs. Pointers to the running mean and running variance data. Both these pointers can be NULL but only at the same time. The value stored in resultRunningVariance (or passed as an input in inference mode) is the sample variance and is the moving average of variance[x] where the variance is computed either over batch or spatial+batch dimensions depending on the mode. If these pointers are not NULL, the tensors should be initialized to some reasonable values or to 0.

epsilon

Input. Epsilon value used in the normalization formula. Its value should be equal to or greater than zero.

*resultSaveMean, *resultSaveInvVariance

Outputs. Optional cache parameters containing saved intermediate results computed during the forward pass. For this to work correctly, the layer's x and normScale, normBias data has to remain unchanged until this backward function is called. Note that both these parameters can be NULL but only at the same time. It is recommended to use this cache since the memory overhead is relatively small.

activationDesc

Input. The tensor descriptor for the activation operation. When the normOps input is set to either CUDNN_NORM_OPS_NORM_ACTIVATION or CUDNN_NORM_OPS_NORM_ADD_ACTIVATION then this activation is used, otherwise the user may pass NULL.

*workspace, workSpaceSizeInBytes

Inputs. *workspace is a pointer to the GPU workspace, and workSpaceSizeInBytes is the size of the workspace. When *workspace is not NULL and *workSpaceSizeInBytes is large enough, and the tensor layout is NHWC and the data type configuration is supported, then this function will trigger a semi-persistent NHWC kernel for normalization. The workspace is not required to be clean. Also, the workspace does not need to remain unchanged between the forward and backward passes.

*reserveSpace

Input. Pointer to the GPU workspace for the reserveSpace.

reserveSpaceSizeInBytes

Input. The size of the reserveSpace. Must be equal or larger than the amount required by cudnnGetNormalizationTrainingReserveSpaceSize().

groutCnt

Input. Only support 1 for now.

Returns

CUDNN_STATUS_SUCCESS

The computation was performed successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • One of the pointers alpha, beta, xData, yData, normScale, and normBias is NULL.
  • The number of xDesc or yDesc tensor descriptor dimensions is not within the [4,5] range (only 4D and 5D tensors are supported).
  • normScaleBiasDesc dimensions are not 1xCx1x1 for 4D and 1xCx1x1x1 for 5D for per-channel mode, and are not 1xCxHxW for 4D and 1xCxDxHxW for 5D for per-activation mode.
  • Exactly one of resultSaveMean, resultSaveInvVariance pointers are NULL.
  • Exactly one of resultRunningMean, resultRunningInvVariance pointers are NULL.
  • epsilon value is less than zero.
  • Dimensions or data types mismatch for xDesc, yDesc.

4.1.17. cudnnOpsTrainVersionCheck()

cudnnStatus_t cudnnOpsTrainVersionCheck(void)

This function checks whether the version of the OpsTrain subset of the library is consistent with the other sub-libraries.

Returns

CUDNN_STATUS_SUCCESS

The version is consistent with other sub-libraries.

CUDNN_STATUS_VERSION_MISMATCH

The version of OpsTrain is not consistent with other sub-libraries. Users should check the installation and make sure all sub-component versions are consistent.

4.1.18. cudnnPoolingBackward()

cudnnStatus_t cudnnPoolingBackward(
    cudnnHandle_t                       handle,
    const cudnnPoolingDescriptor_t      poolingDesc,
    const void                         *alpha,
    const cudnnTensorDescriptor_t       yDesc,
    const void                         *y,
    const cudnnTensorDescriptor_t       dyDesc,
    const void                         *dy,
    const cudnnTensorDescriptor_t       xDesc,
    const void                         *xData,
    const void                         *beta,
    const cudnnTensorDescriptor_t       dxDesc,
    void                               *dx)

This function computes the gradient of a pooling operation.

As of cuDNN version 6.0, a deterministic algorithm is implemented for max backwards pooling. This algorithm can be chosen via the pooling mode enum of poolingDesc. The deterministic algorithm has been measured to be up to 50% slower than the legacy max backwards pooling algorithm, or up to 20% faster, depending upon the use case.
Note: All tensor formats are supported, best performance is expected when using HW-packed tensors. Only 2 and 3 spatial dimensions are allowed

Parameters

handle

Input. Handle to a previously created cuDNN context.

poolingDesc

Input. Handle to the previously initialized pooling descriptor.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

yDesc

Input. Handle to the previously initialized input tensor descriptor.

y

Input. Data pointer to GPU memory associated with the tensor descriptor yDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyData.

xDesc

Input. Handle to the previously initialized output tensor descriptor.

x

Input. Data pointer to GPU memory associated with the output tensor descriptor xDesc.

dxDesc

Input. Handle to the previously initialized output differential tensor descriptor.

dx

Output. Data pointer to GPU memory associated with the output tensor descriptor dxDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions n, c, h, w of the yDesc and dyDesc tensors differ.
  • The strides nStride, cStride, hStride, wStride of the yDesc and dyDesc tensors differ.
  • The dimensions n, c, h, w of the dxDesc and dxDesc tensors differ.
  • The strides nStride, cStride, hStride, wStride of the xDesc and dxDesc tensors differ.
  • The datatype of the four tensors differ.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The wStride of input tensor or output tensor is not 1.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

4.1.19. cudnnSoftmaxBackward()

cudnnStatus_t cudnnSoftmaxBackward(
    cudnnHandle_t                    handle,
    cudnnSoftmaxAlgorithm_t          algorithm,
    cudnnSoftmaxMode_t               mode,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    yDesc,
    const void                      *yData,
    const cudnnTensorDescriptor_t    dyDesc,
    const void                      *dy,
    const void                      *beta,
    const cudnnTensorDescriptor_t    dxDesc,
    void                            *dx)

This routine computes the gradient of the softmax function.

Note:
  • In-place operation is allowed for this routine; meaning, dy and dx pointers may be equal. However, this requires dyDesc and dxDesc descriptors to be identical (particularly, the strides of the input and output must match for in-place operation to be allowed).
  • All tensor formats are supported for all modes and algorithms with 4 and 5D tensors. Performance is expected to be highest with NCHW fully-packed tensors. For more than 5 dimensions tensors must be packed in their spatial dimensions.

Parameters

handle

Input. Handle to a previously created cuDNN context.

algorithm

Input. Enumerant to specify the softmax algorithm.

mode

Input. Enumerant to specify the softmax mode.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

yDesc

Input. Handle to the previously initialized input tensor descriptor.

y

Input. Data pointer to GPU memory associated with the tensor descriptor yDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyData.

dxDesc

Input. Handle to the previously initialized output differential tensor descriptor.

dx

Output. Data pointer to GPU memory associated with the output tensor descriptor dxDesc.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions n, c, h, w of the yDesc, dyDesc and dxDesc tensors differ.
  • The strides nStride, cStride, hStride, wStride of the yDesc and dyDesc tensors differ.
  • The datatype of the three tensors differs.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

4.1.20. cudnnSpatialTfGridGeneratorBackward()

cudnnStatus_t cudnnSpatialTfGridGeneratorBackward(
    cudnnHandle_t                               handle,
    const cudnnSpatialTransformerDescriptor_t   stDesc,
    const void                                 *dgrid,
    void                                       *dtheta)

This function computes the gradient of a grid generation operation.

Note: Only 2d transformation is supported.

Parameters

handle

Input. Handle to a previously created cuDNN context.

stDesc

Input. Previously created spatial transformer descriptor object.

dgrid

Input. Data pointer to GPU memory contains the input differential data.

dtheta

Output. Data pointer to GPU memory contains the output differential data.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is NULL.
  • One of the parameters dgrid or dtheta is NULL.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimension of the transformed tensor specified in stDesc > 4.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

4.1.21. cudnnSpatialTfSamplerBackward()

cudnnStatus_t cudnnSpatialTfSamplerBackward(
    cudnnHandle_t                              handle,
    const cudnnSpatialTransformerDescriptor_t  stDesc,
    const void                                *alpha,
    const cudnnTensorDescriptor_t              xDesc,
    const void                                *x,
    const void                                *beta,
    const cudnnTensorDescriptor_t              dxDesc,
    void                                      *dx,
    const void                                *alphaDgrid,
    const cudnnTensorDescriptor_t              dyDesc,
    const void                                *dy,
    const void                                *grid,
    const void                                *betaDgrid,
    void                                      *dgrid)

This function computes the gradient of a sampling operation.

Note: Only 2d transformation is supported.

Parameters

handle

Input. Handle to a previously created cuDNN context.

stDesc

Input. Previously created spatial transformer descriptor object.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the source value with prior value in the destination tensor as follows:
dstValue = alpha[0]*srcValue + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

dxDesc

Input. Handle to the previously initialized output differential tensor descriptor.

dx

Output. Data pointer to GPU memory associated with the output tensor descriptor dxDesc.

alphaDgrid, betaDgrid
Input. Pointers to scaling factors (in host memory) used to blend the gradient outputs dgrid with prior value in the destination pointer as follows:
dstValue = alpha[0]*srcValue + beta[0]*priorDstValue

For more information, see the Scaling Parameters section in the cuDNN Developer Guide.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyDesc.

grid

Input. A grid of coordinates generated by cudnnSpatialTfGridGeneratorForward().

dgrid

Output. Data pointer to GPU memory contains the output differential data.

Returns

CUDNN_STATUS_SUCCESS

The call was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is NULL.
  • One of the parameters x, dx, y, dy, grid, dgrid is NULL.
  • The dimension of dy differs from those specified in stDesc.
CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration. See the following for some examples of non-supported configurations:

  • The dimension of transformed tensor > 4.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

5. cudnn_cnn_infer.so Library

5.1. Data Type References

5.1.1. Pointer To Opaque Struct Types

5.1.1.1. cudnnConvolutionDescriptor_t

cudnnConvolutionDescriptor_t is a pointer to an opaque structure holding the description of a convolution operation. cudnnCreateConvolutionDescriptor() is used to create one instance, and cudnnSetConvolutionNdDescriptor() or cudnnSetConvolution2dDescriptor() must be used to initialize this instance.

5.1.2. Struct Types

5.1.2.1. cudnnConvolutionBwdDataAlgoPerf_t

cudnnConvolutionBwdDataAlgoPerf_t is a structure containing performance results returned by cudnnFindConvolutionBackwardDataAlgorithm() or heuristic results returned by cudnnGetConvolutionBackwardDataAlgorithm_v7().

Data Members
cudnnConvolutionBwdDataAlgo_t algo

The algorithm runs to obtain the associated performance metrics.

cudnnStatus_t status

If any error occurs during the workspace allocation or timing of cudnnConvolutionBackwardData(), this status will represent that error. Otherwise, this status will be the return status of cudnnConvolutionBackwardData().

  • CUDNN_STATUS_ALLOC_FAILED if any error occurred during workspace allocation or if the provided workspace is insufficient.
  • CUDNN_STATUS_INTERNAL_ERROR if any error occurred during timing calculations or workspace deallocation.
  • Otherwise, this will be the return status of cudnnConvolutionBackwardData().
float time

The execution time of cudnnConvolutionBackwardData() (in milliseconds).

size_t memory

The workspace size (in bytes).

cudnnDeterminism_t determinism

The determinism of the algorithm.

cudnnMathType_t mathType

The math type provided to the algorithm.

int reserved[3]

Reserved space for future properties.

5.1.2.2. cudnnConvolutionFwdAlgoPerf_t

cudnnConvolutionFwdAlgoPerf_t is a structure containing performance results returned by cudnnFindConvolutionForwardAlgorithm() or heuristic results returned by cudnnGetConvolutionForwardAlgorithm_v7().

Data Members
cudnnConvolutionFwdAlgo_t algo

The algorithm runs to obtain the associated performance metrics.

cudnnStatus_t status

If any error occurs during the workspace allocation or timing of cudnnConvolutionForward(), this status will represent that error. Otherwise, this status will be the return status of cudnnConvolutionForward().

  • CUDNN_STATUS_ALLOC_FAILED if any error occurred during workspace allocation or if the provided workspace is insufficient.
  • CUDNN_STATUS_INTERNAL_ERROR if any error occurred during timing calculations or workspace deallocation.
  • Otherwise, this will be the return status of cudnnConvolutionForward().
float time

The execution time of cudnnConvolutionForward() (in milliseconds).

size_t memory

The workspace size (in bytes).

cudnnDeterminism_t determinism

The determinism of the algorithm.

cudnnMathType_t mathType

The math type provided to the algorithm.

int reserved[3]

Reserved space for future properties.

5.1.3. Enumeration Types

5.1.3.1. cudnnBackendAttributeName_t

cudnnBackendAttributeName_t is an enumerated type that indicates the backend descriptor attributes that can be set or get via cudnnBackendSetAttribute() and cudnnBackendGetAttribute() functions. The backend descriptor to which an attribute belongs is identified by the prefix of the attribute name.
typedef enum {
    CUDNN_ATTR_ACTIVATION_COEF            = 0,
    CUDNN_ATTR_ACTIVATION_MODE            = 1,
    CUDNN_ATTR_ACTIVATION_NAN_PROPAGATION = 2,

    CUDNN_ATTR_POINTWISE_MODE            = 10,
    CUDNN_ATTR_POINTWISE_MATH_PREC       = 11,
    CUDNN_ATTR_POINTWISE_NAN_PROPAGATION = 12,
    CUDNN_ATTR_POINTWISE_RELU_LOWER_CLIP = 13,
    CUDNN_ATTR_POINTWISE_RELU_UPPER_CLIP = 14,

    CUDNN_ATTR_CONVOLUTION_COMP_TYPE        = 100,
    CUDNN_ATTR_CONVOLUTION_CONVOLUTION_MODE = 101,
    CUDNN_ATTR_CONVOLUTION_DILATIONS        = 102,
    CUDNN_ATTR_CONVOLUTION_FILTER_STRIDES   = 103,
    CUDNN_ATTR_CONVOLUTION_MODE             = 104,
    CUDNN_ATTR_CONVOLUTION_POST_PADDINGS    = 105,
    CUDNN_ATTR_CONVOLUTION_PRE_PADDINGS     = 106,
    CUDNN_ATTR_CONVOLUTION_SPATIAL_DIMS     = 107,

    CUDNN_ATTR_DATA_PTR_OPERATION_ACTIVATION_FORWARD_X = 200,
    CUDNN_ATTR_DATA_PTR_OPERATION_ACTIVATION_FORWARD_Y = 201,
    CUDNN_ATTR_DATA_PTR_OPERATION_ADD_TENSOR_A         = 202,
    CUDNN_ATTR_DATA_PTR_OPERATION_ADD_TENSOR_C         = 203,
    CUDNN_ATTR_DATA_PTR_OPERATION_CONVOLUTION_W        = 204,
    CUDNN_ATTR_DATA_PTR_OPERATION_CONVOLUTION_X        = 205,
    CUDNN_ATTR_DATA_PTR_OPERATION_CONVOLUTION_Y        = 206,

    CUDNN_ATTR_ENGINECFG_ENGINE            = 300,
    CUDNN_ATTR_ENGINECFG_INTERMEDIATE_INFO = 301,
    CUDNN_ATTR_ENGINECFG_KNOB_CHOICES      = 302,

    CUDNN_ATTR_EXECUTION_PLAN_HANDLE                      = 400,
    CUDNN_ATTR_EXECUTION_PLAN_ENGINE_CONFIG               = 401,
    CUDNN_ATTR_EXECUTION_PLAN_WORKSPACE_SIZE              = 402,
    CUDNN_ATTR_EXECUTION_PLAN_INTERMEDIATES_USER_PROVIDED = 403,
    CUDNN_ATTR_EXECUTION_PLAN_RUN_ONLY_INTERMEDIATE       = 404,

    CUDNN_ATTR_INTERMEDIATE_DEPENDENCY_ATTRIB_NAME = 500,
    CUDNN_ATTR_INTERMEDIATE_INFO_DEPENDENCIES      = 501,
    CUDNN_ATTR_INTERMEDIATE_INFO_SIZE              = 502,

    CUDNN_ATTR_KNOB_CHOICE_ENGINE     = 600,
    CUDNN_ATTR_KNOB_CHOICE_KNOB_TYPE  = 601,
    CUDNN_ATTR_KNOB_CHOICE_KNOB_VALUE = 602,

    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_ALPHA        = 700,
    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_BETA         = 701,
    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_CONV_DESC    = 702,
    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_w            = 703,
    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_x            = 704,
    CUDNN_ATTR_OPERATION_CONVOLUTION_FORWARD_y            = 705,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_ALPHA       = 706,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_BETA        = 707,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_CONV_DESC   = 708,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_w           = 709,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_dx          = 710,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_DATA_dy          = 711,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_ALPHA     = 712,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_BETA      = 713,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_CONV_DESC = 714,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_dw        = 715,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_x         = 716,
    CUDNN_ATTR_OPERATION_CONVOLUTION_BWD_FILTER_dy        = 717,

    CUDNN_ATTR_OPERATION_POINTWISE_PW_DESCRIPTOR = 750,
    CUDNN_ATTR_OPERATION_POINTWISE_XDESC         = 751,
    CUDNN_ATTR_OPERATION_POINTWISE_BDESC         = 752,
    CUDNN_ATTR_OPERATION_POINTWISE_YDESC         = 753,
    CUDNN_ATTR_OPERATION_POINTWISE_ALPHA1        = 754,
    CUDNN_ATTR_OPERATION_POINTWISE_ALPHA2        = 755,

    CUDNN_ATTR_OPERATION_GENSTATS_MODE      = 770,
    CUDNN_ATTR_OPERATION_GENSTATS_MATH_PREC = 771,
    CUDNN_ATTR_OPERATION_GENSTATS_XDESC     = 772,
    CUDNN_ATTR_OPERATION_GENSTATS_SUMDESC   = 773,
    CUDNN_ATTR_OPERATION_GENSTATS_SQSUMDESC = 774,

    CUDNN_ATTR_OPERATIONGRAPH_HANDLE                 = 800,
    CUDNN_ATTR_OPERATIONGRAPH_OPS                    = 801,
    CUDNN_ATTR_OPERATIONGRAPH_ENGINE_GLOBAL_COUNT    = 802,
    CUDNN_ATTR_OPERATIONGRAPH_ENGINE_SUPPORTED_COUNT = 803,

    CUDNN_ATTR_TENSOR_BYTE_ALIGNMENT = 900,
    CUDNN_ATTR_TENSOR_DATA_TYPE      = 901,
    CUDNN_ATTR_TENSOR_DIMENSIONS     = 902,
    CUDNN_ATTR_TENSOR_STRIDES        = 903,
    CUDNN_ATTR_TENSOR_VECTOR_COUNT   = 904,
    CUDNN_ATTR_TENSOR_UNIQUE_ID      = 905,
    CUDNN_ATTR_TENSOR_IS_VIRTUAL     = 906,

    CUDNN_ATTR_VARIANT_PACK_UNIQUE_IDS    = 1000,
    CUDNN_ATTR_VARIANT_PACK_DATA_POINTERS = 1001,
    CUDNN_ATTR_VARIANT_PACK_INTERMEDIATES = 1002,
    CUDNN_ATTR_VARIANT_PACK_WORKSPACE     = 1003,

    CUDNN_ATTR_LAYOUT_QUERY_INFO                   = 1100,
    CUDNN_ATTR_LAYOUT_QUERY_OPERATION_GRAPH        = 1101,
    CUDNN_ATTR_LAYOUT_QUERY_ENGINE_GLOBAL_INDEX    = 1102,
    CUDNN_ATTR_LAYOUT_QUERY_ENGINE_SUPPORTED_INDEX = 1103,

    CUDNN_ATTR_KNOB_INFO_TYPE          = 1200,
    CUDNN_ATTR_KNOB_INFO_MAXIMUM_VALUE = 1201,
    CUDNN_ATTR_KNOB_INFO_MINIMUM_VALUE = 1202,
    CUDNN_ATTR_KNOB_INFO_STRIDE        = 1203,

    CUDNN_ATTR_ENGINE_OPERATION_GRAPH = 1300,
    CUDNN_ATTR_ENGINE_GLOBAL_INDEX    = 1301,
    CUDNN_ATTR_ENGINE_SUPPORTED_INDEX = 1302,
    CUDNN_ATTR_ENGINE_KNOB_INFOS      = 1303,
    CUDNN_ATTR_ENGINE_NUMERICAL_NOTE  = 1304,

    CUDNN_ATTR_ENGINEHEUR_MODE            = 1400,
    CUDNN_ATTR_ENGINEHEUR_OPERATION_GRAPH = 1401,
    CUDNN_ATTR_ENGINEHEUR_RESULTS         = 1402,

} cudnnBackendAttributeName_t;

5.1.3.2. cudnnBackendAttributeType_t

cudnnBackendAttributeType_t is an enumerated type that indicates type of backend descriptor attributes that can be set and get via cudnnBackendSetAttribute() and cudnnBackendGetAttribute() functions.
typedef enum {
    CUDNN_TYPE_HANDLE = 0,
    CUDNN_TYPE_DATA_TYPE,
    CUDNN_TYPE_BOOLEAN,
    CUDNN_TYPE_INT64,
    CUDNN_TYPE_FLOAT,
    CUDNN_TYPE_DOUBLE,
    CUDNN_TYPE_VOID_PTR,
    CUDNN_TYPE_CONVOLUTION_MODE,
    CUDNN_TYPE_ACTIVATION_MODE,
    CUDNN_TYPE_HEUR_MODE,
    CUDNN_TYPE_KNOB_TYPE,
    CUDNN_TYPE_NAN_PROPOGATION,
    CUDNN_TYPE_NUMERICAL_NOTE,
    CUDNN_TYPE_ATTRIB_NAME,
    CUDNN_TYPE_POINTWISE_MODE,
    CUDNN_TYPE_BACKEND_DESCRIPTOR,
    CUDNN_TYPE_GENSTATS_MODE
} cudnnBackendAttributeType_t;

5.1.3.3. cudnnBackendDescriptorType_t

cudnnBackendDescriptor_t is an enumerated type that indicates type of backend descriptors. Users create a backend descriptor of a particular type by passing a value from this enumerate to cudnnBackendCreateDescriptor() function.
typedef enum {
    CUDNN_BACKEND_ACTIVATION_DESCRIPTOR = 0,
    CUDNN_BACKEND_POINTWISE_DESCRIPTOR,
    CUDNN_BACKEND_CONVOLUTION_DESCRIPTOR,
    CUDNN_BACKEND_ENGINE_DESCRIPTOR,
    CUDNN_BACKEND_ENGINECFG_DESCRIPTOR,
    CUDNN_BACKEND_ENGINEHEUR_DESCRIPTOR,
    CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR,
    CUDNN_BACKEND_INTERMEDIATE_DEPENDENCY_DESCRIPTOR,
    CUDNN_BACKEND_INTERMEDIATE_INFO_DESCRIPTOR,
    CUDNN_BACKEND_KNOB_CHOICE_DESCRIPTOR,
    CUDNN_BACKEND_KNOB_INFO_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_ACTIVATION_FORWARD_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_CONVOLUTION_FORWARD_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_CONVOLUTION_BACKWARD_FILTER_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_CONVOLUTION_BACKWARD_DATA_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_POINTWISE_DESCRIPTOR,
    CUDNN_BACKEND_OPERATION_GEN_STATS_DESCRIPTOR,
    CUDNN_BACKEND_OPERATIONGRAPH_DESCRIPTOR,,
    CUDNN_BACKEND_VARIANT_PACK_DESCRIPTOR,
    CUDNN_BACKEND_TENSOR_DESCRIPTOR,
    CUDNN_BACKEND_LAYOUT_INFO_DESCRIPTOR,
} cudnnBackendDescriptorType_t;

5.1.3.4. cudnnBackendHeurMode_t

cudnnBackendHeurMode_t is an enumerated type that indicates the operation mode of a CUDNN_BACKEND_ENGINE_HEUR_DESCRIPTOR.
typedef enum {
    CUDNN_HEUR_MODE_INSTANT = 0
}

5.1.3.5. cudnnBackendKnobType_t

cudnnBackendKnobType_t is an enumerated type that indicates the type of performance knobs. Performance knobs are runtime settings to an engine that will affect its performance. Users can query for an array of performance knobs and their valid value range from a CUDNN_BACKEND_ENGINE_DESCRIPTOR using cudnnBackendGetAttribute() function. Users can set the choice for each knob using the cudnnBackendSetAttribute() function with a CUDNN_BACKEND_KNOB_CHOICE_DESCRIPTOR descriptor.
typedef enum {
    CUDNN_KNOB_TYPE_SPLIT_K          = 0,
    CUDNN_KNOB_TYPE_SWIZZLE          = 1,
    CUDNN_KNOB_TYPE_TILE_SIZE        = 2,
    CUDNN_KNOB_TYPE_USE_TEX          = 3,
    CUDNN_KNOB_TYPE_EDGE             = 4,
    CUDNN_KNOB_TYPE_KBLOCK           = 5,
    CUDNN_KNOB_TYPE_LDGA             = 6,
    CUDNN_KNOB_TYPE_LDGB             = 7,
    CUDNN_KNOB_TYPE_CHUNK_K          = 8,
    CUDNN_KNOB_TYPE_SPLIT_H          = 9,
    CUDNN_KNOB_TYPE_WINO_TILE        = 10,
    CUDNN_KNOB_TYPE_MULTIPLY         = 11,
    CUDNN_KNOB_TYPE_SPLIT_K_BUF      = 12,
    CUDNN_KNOB_TYPE_TILEK            = 13,
    CUDNN_KNOB_TYPE_STAGES           = 14,
    CUDNN_KNOB_TYPE_REDUCTION_MODE   = 15,
    CUDNN_KNOB_TYPE_CTA_SPLIT_K_MODE = 16,
    CUDNN_KNOB_TYPE_SPLIT_K_SLC      = 17,
    CUDNN_KNOB_TYPE_IDX_MODE         = 18,
    CUDNN_KNOB_TYPE_SLICED           = 19,
    CUDNN_KNOB_TYPE_SPLIT_RS         = 20,
    CUDNN_KNOB_TYPE_SINGLEBUFFER     = 21,

    CUDNN_KNOB_TYPE_COUNTS = 22,
} cudnnBackendKnobType_t;

5.1.3.6. cudnnBackendLayoutType_t

cudnnBackendLayoutType_t is an enumerated type that indicates queryable layout requirements of an engine. Users can query for layout requirements from a CUDNN_BACKEND_ENGINE_DESC descriptor using cudnnBackendGetAttribute() function.
typedef enum {
    CUDNN_LAYOUT_TYPE_REQUIRED_NCHW,
    CUDNN_LAYOUT_TYPE_PREFERRED_NCHW,
    CUDNN_LAYOUT_TYPE_REQUIRED_PADC8_FOLDED_NHWC,
} cudnnBackendLayoutType_t;

5.1.3.7. cudnnBackendNumericalNote_t

cudnnBackendNumericalNot_t is an enumerated type that indicates queryable numerical properties of an engine. Users can query for an array of numerical notes from an CUDNN_BACKEND_ENGINE_DESC using cudnnBackendGetAttribute() function.
typedef enum {
    CUDNN_NUMERICAL_NOTE_TENSOR_CORE = 0,
    CUDNN_NUMERICAL_NOTE_DOWN_CONVERT_INPUTS,
    CUDNN_NUMERICAL_NOTE_REDUCED_PRECISION_REDUCTION,
    CUDNN_NUMERICAL_NOTE_FFT,
    CUDNN_NUMERICAL_NOTE_NONDETERMINISTIC,
    CUDNN_NUMERICAL_NOTE_WINOGRAD,
    CUDNN_NUMERICAL_NOTE_TYPE_COUNT
} cudnnBackendNumericalNote_t;

5.1.3.8. cudnnConvolutionBwdDataAlgo_t

cudnnConvolutionBwdDataAlgo_t is an enumerated type that exposes the different algorithms available to execute the backward data convolution operation.

Values
CUDNN_CONVOLUTION_BWD_DATA_ALGO_0

This algorithm expresses the convolution as a sum of matrix products without actually explicitly forming the matrix that holds the input tensor data. The sum is done using the atomic add operation, thus the results are non-deterministic.

CUDNN_CONVOLUTION_BWD_DATA_ALGO_1

This algorithm expresses the convolution as a matrix product without actually explicitly forming the matrix that holds the input tensor data. The results are deterministic.

CUDNN_CONVOLUTION_BWD_DATA_ALGO_FFT

This algorithm uses a Fast-Fourier Transform approach to compute the convolution. A significant memory workspace is needed to store intermediate results. The results are deterministic.

CUDNN_CONVOLUTION_BWD_DATA_ALGO_FFT_TILING

This algorithm uses the Fast-Fourier Transform approach but splits the inputs into tiles. A significant memory workspace is needed to store intermediate results but less than CUDNN_CONVOLUTION_BWD_DATA_ALGO_FFT for large size images. The results are deterministic.

CUDNN_CONVOLUTION_BWD_DATA_ALGO_WINOGRAD

This algorithm uses the Winograd Transform approach to compute the convolution. A reasonably sized workspace is needed to store intermediate results. The results are deterministic.

CUDNN_CONVOLUTION_BWD_DATA_ALGO_WINOGRAD_NONFUSED

This algorithm uses the Winograd Transform approach to compute the convolution. A significant workspace may be needed to store intermediate results. The results are deterministic.

5.1.3.9. cudnnConvolutionBwdFilterAlgo_t

cudnnConvolutionBwdFilterAlgo_t is an enumerated type that exposes the different algorithms available to execute the backward filter convolution operation.

Values
CUDNN_CONVOLUTION_BWD_FILTER_ALGO_0

This algorithm expresses the convolution as a sum of matrix products without actually explicitly forming the matrix that holds the input tensor data. The sum is done using the atomic add operation, thus the results are non-deterministic.

CUDNN_CONVOLUTION_BWD_FILTER_ALGO_1

This algorithm expresses the convolution as a matrix product without actually explicitly forming the matrix that holds the input tensor data. The results are deterministic.

CUDNN_CONVOLUTION_BWD_FILTER_ALGO_FFT

This algorithm uses the Fast-Fourier Transform approach to compute the convolution. A significant workspace is needed to store intermediate results. The results are deterministic.

CUDNN_CONVOLUTION_BWD_FILTER_ALGO_3

This algorithm is similar to CUDNN_CONVOLUTION_BWD_FILTER_ALGO_0 but uses some small workspace to precompute some indices. The results are also non-deterministic.

CUDNN_CONVOLUTION_BWD_FILTER_WINOGRAD_NONFUSED

This algorithm uses the Winograd Transform approach to compute the convolution. A significant workspace may be needed to store intermediate results. The results are deterministic.

CUDNN_CONVOLUTION_BWD_FILTER_ALGO_FFT_TILING

This algorithm uses the Fast-Fourier Transform approach to compute the convolution but splits the input tensor into tiles. A significant workspace may be needed to store intermediate results. The results are deterministic.

5.1.3.10. cudnnConvolutionFwdAlgo_t

cudnnConvolutionFwdAlgo_t is an enumerated type that exposes the different algorithms available to execute the forward convolution operation.

Values
CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_GEMM

This algorithm expresses the convolution as a matrix product without actually explicitly forming the matrix that holds the input tensor data.

CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_PRECOMP_GEMM

This algorithm expresses convolution as a matrix product without actually explicitly forming the matrix that holds the input tensor data, but still needs some memory workspace to precompute some indices in order to facilitate the implicit construction of the matrix that holds the input tensor data.

CUDNN_CONVOLUTION_FWD_ALGO_GEMM

This algorithm expresses the convolution as an explicit matrix product. A significant memory workspace is needed to store the matrix that holds the input tensor data.

CUDNN_CONVOLUTION_FWD_ALGO_DIRECT

This algorithm expresses the convolution as a direct convolution (for example, without implicitly or explicitly doing a matrix multiplication).

CUDNN_CONVOLUTION_FWD_ALGO_FFT

This algorithm uses the Fast-Fourier Transform approach to compute the convolution. A significant memory workspace is needed to store intermediate results.

CUDNN_CONVOLUTION_FWD_ALGO_FFT_TILING

This algorithm uses the Fast-Fourier Transform approach but splits the inputs into tiles. A significant memory workspace is needed to store intermediate results but less than CUDNN_CONVOLUTION_FWD_ALGO_FFT for large size images.

CUDNN_CONVOLUTION_FWD_ALGO_WINOGRAD

This algorithm uses the Winograd Transform approach to compute the convolution. A reasonably sized workspace is needed to store intermediate results.

CUDNN_CONVOLUTION_FWD_ALGO_WINOGRAD_NONFUSED

This algorithm uses the Winograd Transform approach to compute the convolution. A significant workspace may be needed to store intermediate results.

5.1.3.11. cudnnConvolutionMode_t

cudnnConvolutionMode_t is an enumerated type used by cudnnSetConvolution2dDescriptor() to configure a convolution descriptor. The filter used for the convolution can be applied in two different ways, corresponding mathematically to a convolution or to a cross-correlation. (A cross-correlation is equivalent to a convolution with its filter rotated by 180 degrees.)

Values
CUDNN_CONVOLUTION

In this mode, a convolution operation will be done when applying the filter to the images.

CUDNN_CROSS_CORRELATION

In this mode, a cross-correlation operation will be done when applying the filter to the images.

5.1.3.12. cudnnGenStatsMode_t

cudnnGenStatsMode_t is an enumerated type to indicate the statistics mode in the backend statistics generation operation.

Values
CUDNN_GENSTATS_SUM_SQSUM

In this mode, the sum and sum of squares of the input tensor along the specified dimensions are computed and written out. The reduction dimensions currently supported are limited per channel, however additional support may be added upon request.

5.1.3.13. cudnnPointwiseMode_t

cudnnPointwiseMode_t is an enumerated type to indicate the intended pointwise math operation in the backend pointwise operation descriptor.

Values
CUDNN_POINTWISE_ADD

In this mode, a pointwise add between two tensors is computed.

CUDNN_POINTWISE_MUL

In this mode, a pointwise multiplication between two tensors is computed.

CUDNN_POINTWISE_MIN

In this mode, a pointwise minimum is taken between two tensors.

CUDNN_POINTWISE_MAX

In this mode, a pointwise max is taken between two tensors.

CUDNN_POINTWISE_SQRT

In this mode, a pointwise square root of the input tensor is computed.

CUDNN_POINTWISE_IDENTITY_FWD

In this mode, an identity activation (pointwise value copy) is performed.

CUDNN_POINTWISE_RELU_FWD

In this mode, a pointwise ReLu activation is computed.

CUDNN_POINTWISE_TANH_FWD

In this mode, a pointwise tanh activation is computed.

CUDNN_POINTWISE_SIGMOID_FWD

In this mode, a pointwise sigmoid activation is computed.

CUDNN_POINTWISE_ELU_FWD

In this mode, a pointwise Exponential Linear Unit activation is computed.

5.1.3.14. cudnnReorderType_t

typedef enum {
	CUDNN_DEFAULT_REORDER = 0,
	CUDNN_NO_REORDER      = 1,
	} cudnnReorderType_t;		

cudnnReorderType_t is an enumerated type to set the convolution reordering type. The reordering type can be set by cudnnSetConvolutionReorderType() and its status can be read by cudnnGetConvolutionReorderType().

5.1.4. Data Types Found In cudnn_backend.h

5.1.4.1. cudnnBackendDescriptor_t

cudnnBackendDescriptor_t is a typedef void pointer to one of many opaque descriptor structures. The type of structure that it points to is determined by the argument when allocating the memory for the opaque structure using cudnnBackendCreateDescriptor().

Attributes of a descriptor can be set using cudnnBackendSetAttribute(). After all required attributes of a descriptor are set, the descriptor can be finalized by cudnnBackendFinalizeDescriptor(). From a finalized descriptor, one can query its queryable attributes using cudnnBackendGetAttribute(). Finally, the memory allocated for a descriptor can be freed using cudnnBackendDestroyDescriptor().

5.2. API Functions

5.2.1. cudnnBackendCreateDescriptor()

cudnnStatus_t cudnnBackendCreateDescriptor(cudnnBackendDescriptorType_t descriptorType, cudnnBackendDescriptor_t *descriptor)
This function allocates memory:
  • in the descriptor for a given descriptor type
  • at the location pointed by the descriptor
Note: The void * is a pointer to the cudnnBackend_t.

Parameters

descriptorType

Input. One among the enumerated cudnnBackendDescriptorType_t.

descriptor

Input. Pointer to an instance of cudnnBackendDescriptor_t to be created.

Returns

CUDNN_STATUS_SUCCESS

The creation was successful.

CUDNN_STATUS_ALLOC_FAILED

The memory allocation failed.

5.2.2. cudnnBackendDestroyDescriptor()

cudnnStatus_t cudnnBackendDestroyDescriptor(cudnnBackendDescriptor_tdescriptor)

This function destroys instances of cudnnBackendDescriptor_t that were previously created using cudnnBackendCreateDescriptor().

Parameters

descriptor

Input. Instance of cudnnBackendDescriptor_t previously created by cudnnBackendCreateDescriptor().

Returns

CUDNN_STATUS_SUCCESS

The memory was destroyed successfully.

CUDNN_STATUS_ALLOC_FAILED

The destruction of memory failed.

Undefined Behavior

The descriptor was altered between the Create and Destroy Descriptor.

Undefined

The value pointed by the descriptor will be Undefined after the memory is free and done.

5.2.3. cudnnBackendExecute()

cudnnStatus_ cudnnBackendExecute(cudnnHandle_t handle, cudnnBackendDescriptor_t executionPlan, cudnnBackendDescriptor_t varianPack)
This function executes:
  • the given Engine Configuration Plan on the VariantPack
  • the finalized ExecutionPlan on the data

The data and the working space are encapsulated in the VariantPack.

Parameters

executionPlan

Input. Pointer to the cuDNN handle to be destroyed.

variantPack
Input. Pointer to the finalized VariantPack consisting of:
  • Data pointer for each non-virtual pointer of the operation set in the execution plan.
  • Pointer to user-allocated workspace in global memory at least as large as the size queried from CUDNN_BACKEND_.

Returns

CUDNN_STATUS_SUCCESS

The ExecutionPlan was executed successfully.

5.2.4. cudnnBackendFinalize()

cudnnStatus_t cudnnbBackendFinalize(cudnnBackendDescriptor descriptor)

This function finalizes the memory pointed to by the descriptor. The type of finalization is done depending on the descriptorType argument with which the descriptor was created using cudnnBackendCreate() or initialized using cudnnBackendInitialize().

cudnnBackendFinalize() also checks all the attributes set between the create/initialization and finalize phase. If successful, cudnnBackendFinalize() returns CUDNN_STATUS_SUCCESS and the finalized state of the descriptor is set to true. In this state, setting attributes using cudnnBackendSetAttribute() is not allowed. Getting attributes using cudnnBackendGetAttribute() is only allowed when the finalized state of the descriptor is true.

Parameters

descriptor

Input. Instance of cudnnBackendDescriptor_t to finalize.

Returns

CUDNN_STATUS_SUCCESS

The descriptor was finalized successfully.

CUDNN_STATUS_BAD_PARAM

Invalid descriptor attribute values or combination thereof is encountered.

CUDNN_STATUS_NOT_SUPPORTED

Descriptor attribute values or combinations therefore not supported by the current version of cuDNN is encountered.

CUDNN_STATUS_INTERNAL_ERROR

Some internal errors are encountered.

5.2.5. cudnnBackendGetAttribute()

cudnnStatus_t cudnnBackendGetAttribute(
    cudnnBackendDescriptor_t descriptor,
    cudnnBackendAttributeName_t attributeName,
    cudnnBackendAttributeType_t attributeType,
    int64_t requestedElementCount,
    int64_t *elementCount,
    void *arrayOfElements);

This function retrieves the value(s) of an attribute of a descriptor. attributeName is the name of the attribute whose value is requested. The attributeType is the type of attribute. requestsedElementCount is the number of elements to be potentially retrieved. The number of elements for the requested attribute is stored in elementCount. The retrieved values are stored in arrayOfElements. When the attribute is expected to have a single value, arrayOfElements can be pointer to the output value. This function will return CUDNN_STATUS_NOT_INTIALIZED if the descriptor was already successfully finalized.

Parameters

descriptor

Input. Instance of cudnnBackendDescriptor_t whose attribute the user wants to retrieve.

attributeName

Input. The name of the attribute being get from the on the descriptor.

attributeType

Input. The type of attribute.

requestedElementCount

Input. Number of elements to output to arrayOfElements.

elementCount

Input. Output pointer for the number of elements the descriptor attribute has. Note that cudnnBackendGetAttribute() will only write the least of this and requestedElementCount elements to arrayOfElements.

arrayOfElements

Input. Output pointer to element of attribute type.

Returns

CUDNN_STATUS_SUCCESS

The attributeName was given to the descriptor successfully.

CUDNN_STATUS_NOT_INITIALIZED

The descriptor has not been successfully finalized using cudnnBackendFinalizeDescriptor().

5.2.6. cudnnBackendInitialize()

cudnnStatus_t cudnnBackendInitialize(cudnnBackendDescriptor_t descriptor, cudnnBackendDescriptorType_t descriptorType, size_t sizeInBytes)

This function repurposes a pre-allocated memory pointed to by a descriptor of size sizeInByte to a backend descriptor of type descriptorType. The necessary size for a descriptor type can be acquired by calling the function cudnnBackendGetSizeOf(). The finalized state of the descriptor is set to false.

Parameters

descriptor

Input. Instance of cudnnBackendDescriptor_t to be initialized.

descriptorType

Input. Enumerated value for the type of cuDNN backend descriptor.

sizeInBytes

Input. Size of memory pointed to by descriptor.

Returns

CUDNN_STATUS_SUCCESS

The memory was initialized successfully.

CUDNN_STATUS_BAD_PARAM
An invalid or inconsistent argument value is encountered. For example:
  • descriptor is a nullptr
  • sizeInBytes is less than the size required by the descriptor type

5.2.7. cudnnBackendSetAttribute()

cudnnStatus_t cudnnBackendSetAttribute(
    cudnnBackendDescriptor_t descriptor,
    cudnnBackendAttributeName_t attributeName,
    cudnnBackendAttributeType_t attributeType,
    int64_t elementCount,
    void *arrayOfElements);

This function sets an attribute of a descriptor to value(s) provided as a pointer. descriptor is the descriptor to be set. attributeName is the name of the attribute to be set. attributeType is the type of attribute. The value to which the attribute is set, is pointed by the arrayOfElements. The number of elements is given by elementCount. This function will return CUDNN_STATUS_NOT_INTIALIZED if the descriptor is already successfully finalized using cudnnBackendFinalize().

Parameters

descriptor

Input. Instance of cudnnBackendDescriptor_t whose attribute is being set.

attributeName

Input. The name of the attribute being set on the descriptor.

attributeType

Input. The type of attribute.

elementCount

Input. Number of elements being set.

arrayOfElements

Input. The starting location for an array from where to read the values from.

Returns

CUDNN_STATUS_SUCCESS

The attributeName was set to the descriptor.

CUDNN_STATUS_NOT_INITIALIZED

The backend descriptor pointed to by descriptor is already in finalized state.

CUDNN_STATUS_BAD_PARAM
The function is called with arguments that correspond to invalid values. Some possible causes are:
  • attributeName is not a settable attribute of descriptor
  • attributeType is incorrect for this attributeName.
  • elemCount value is unexpected.
  • arrayOfElements contains values invalid for the attributeType.
CUDNN_STATUS_NOT_SUPPORTED

The value(s) to which the attributes are being set is not supported by the current version of cuDNN.

5.2.8. cudnnConvolutionBackwardData()

cudnnStatus_t cudnnConvolutionBackwardData(
    cudnnHandle_t                       handle,
    const void                         *alpha,
    const cudnnFilterDescriptor_t       wDesc,
    const void                         *w,
    const cudnnTensorDescriptor_t       dyDesc,
    const void                         *dy,
    const cudnnConvolutionDescriptor_t  convDesc,
    cudnnConvolutionBwdDataAlgo_t       algo,
    void                               *workSpace,
    size_t                              workSpaceSizeInBytes,
    const void                         *beta,
    const cudnnTensorDescriptor_t       dxDesc,
    void                               *dx)

This function computes the convolution data gradient of the tensor dy, where y is the output of the forward convolution in cudnnConvolutionForward(). It uses the specified algo, and returns the results in the output tensor dx. Scaling factors alpha and beta can be used to scale the computed result or accumulate with the current dx.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

wDesc

Input. Handle to a previously initialized filter descriptor. For more information, see cudnnFilterDescriptor_t.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor. For more information, see cudnnTensorDescriptor_t.

dy

Input. Data pointer to GPU memory associated with the input differential tensor descriptor dyDesc.

convDesc

Input. Previously initialized convolution descriptor. For more information, see cudnnConvolutionDescriptor_t.

algo

Input. Enumerant that specifies which backward data convolution algorithm should be used to compute the results. For more information, see cudnnConvolutionBwdDataAlgo_t.

workSpace

Input. Data pointer to GPU memory to a workspace needed to be able to execute the specified algorithm. If no workspace is needed for a particular algorithm, that pointer can be nil.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

dxDesc

Input. Handle to the previously initialized output tensor descriptor.

dx

Input/Output. Data pointer to GPU memory associated with the output tensor descriptor dxDesc that carries the result.

Supported configurations

This function supports the following combinations of data types for wDesc, dyDesc, convDesc, and dxDesc.
Data Type Configurations wDesc, dyDesc and dxDesc Data Type convDesc Data Type
TRUE_HALF_CONFIG (only supported on architectures with true FP16 support, meaning, compute capability 5.3 and later) CUDNN_DATA_HALF CUDNN_DATA_HALF
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Supported algorithms

Note: Specifying a separate algorithm can cause changes in performance, support and computation determinism. See the following for a list of algorithm options, and their respective supported parameters and deterministic behavior.

The table below shows the list of the supported 2D and 3D convolutions. The 2D convolutions are described first, followed by the 3D convolutions.

For the following terms, the short-form versions shown in the parentheses are used in the table below, for brevity:
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_0 (_ALGO_0)
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_1 (_ALGO_1)
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_FFT (_FFT)
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_FFT_TILING (_FFT_TILING)
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_WINOGRAD (_WINOGRAD)
  • CUDNN_CONVOLUTION_BWD_DATA_ALGO_WINOGRAD_NONFUSED (_WINOGRAD_NONFUSED)
  • CUDNN_TENSOR_NCHW (_NCHW)
  • CUDNN_TENSOR_NHWC (_NHWC)
  • CUDNN_TENSOR_NCHW_VECT_C (_NCHW_VECT_C)
Table 10. For 2D convolutions: wDesc: _NHWC
Filter descriptor wDesc: _NHWC (see cudnnTensorFormat_t)
Algo Name Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important

_ALGO_0

_ALGO_1

  NHWC HWC-packed NHWC HWC-packed TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

 
Table 11. For 2D convolutions: wDesc: _NCHW
Filter descriptor wDesc: _NCHW.
Algo Name Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important

_ALGO_0

No NCHW CHW-packed All except _NCHW_VECT_C. TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_ALGO_1 Yes NCHW CHW-packed All except _NCHW_VECT_C. TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_FFT Yes NCHW CHW-packed NCHW HW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

dxDesc feature map height + 2 * convDesc zero-padding height must equal 256 or less

dxDesc feature map width + 2 * convDesc zero-padding width must equal 256 or less

convDesc vertical and horizontal filter stride must equal 1

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

_FFT_TILING Yes NCHW CHW-packed NCHW HW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG is also supported when the task can be handled by 1D FFT, meaning, one of the filter dimensions, width or height is 1.

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

When neither of wDesc filter dimension is 1, the filter width and height must not be larger than 32

When either of wDesc filter dimension is 1, the largest filter dimension should not exceed 256

convDesc vertical and horizontal filter stride must equal 1 when either the filter width or filter height is 1, otherwise, the stride can be 1 or 2

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

_WINOGRAD Yes NCHW CHW-packed All except _NCHW_VECT_C. PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

convDesc vertical and horizontal filter stride must equal 1

wDesc filter height must be 3

wDesc filter width must be 3

_WINOGRAD_NONFUSED Yes NCHW CHW-packed All except _NCHW_VECT_C. TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

convDesc vertical and horizontal filter stride must equal 1

wDesc filter (height, width) must be (3,3) or (5,5)

If wDesc filter (height, width) is (5,5) then the data type config TRUE_HALF_CONFIG is not supported

Table 12. For 3D convolutions: wDesc: _NCHW
Filter descriptor wDesc: _NCHW.
Algo Name Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important
_ALGO_0 Yes NCDHW CDHW-packed All except _NCDHW_VECT_C. PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_ALGO_1 Yes NCDHW

CDHW-packed

NCDHW

CDHW-packed

TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIGDOUBLE_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

_FFT_TILING Yes NCDHW CDHW-packed NCDHW DHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

wDesc filter height must equal 16 or less

wDesc filter width must equal 16 or less

wDesc filter depth must equal 16 or less

convDesc must have all filter strides equal to 1

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

wDesc filter depth must be greater than convDesc zero-padding width

Table 13. For 3D convolutions: wDesc: _NHWC
Filter descriptor wDesc: _NHWC
Algo Name (3D Convolutions) Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important
_ALGO_1 Yes NDHWC

DHWC-packed

NDHWC

DHWC-packed

TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: Greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

Returns

CUDNN_STATUS_SUCCESS

The operation was launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • At least one of the following is NULL: handle, dyDesc, wDesc, convDesc, dxDesc, dy, w, dx, alpha, beta
  • wDesc and dyDesc have a non-matching number of dimensions
  • wDesc and dxDesc have a non-matching number of dimensions
  • wDesc has fewer than three number of dimensions
  • wDesc, dxDesc, and dyDesc have a non-matching data type.
  • wDesc and dxDesc have a non-matching number of input feature maps per image (or group in case of grouped convolutions).
  • dyDesc spatial sizes do not match with the expected size as determined by cudnnGetConvolutionNdForwardOutputDim
CUDNN_STATUS_NOT_SUPPORTED
At least one of the following conditions are met:
  • dyDesc or dxDesc have a negative tensor striding
  • dyDesc, wDesc or dxDesc has a number of dimensions that is not 4 or 5
  • The chosen algo does not support the parameters provided; see above for an exhaustive list of parameters that support each algo
  • dyDesc or wDesc indicate an output channel count that isn't a multiple of group count (if group count has been set in convDesc).
CUDNN_STATUS_MAPPING_ERROR

An error occurs during the texture binding of the filter data or the input differential tensor data

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

5.2.9. cudnnConvolutionBiasActivationForward()

cudnnStatus_t cudnnConvolutionBiasActivationForward(
    cudnnHandle_t                       handle,
    const void                         *alpha1,
    const cudnnTensorDescriptor_t       xDesc,
    const void                         *x,
    const cudnnFilterDescriptor_t       wDesc,
    const void                         *w,
    const cudnnConvolutionDescriptor_t  convDesc,
    cudnnConvolutionFwdAlgo_t           algo,
    void                               *workSpace,
    size_t                              workSpaceSizeInBytes,
    const void                         *alpha2,
    const cudnnTensorDescriptor_t       zDesc,
    const void                         *z,
    const cudnnTensorDescriptor_t       biasDesc,
    const void                         *bias,
    const cudnnActivationDescriptor_t   activationDesc,
    const cudnnTensorDescriptor_t       yDesc,
    void                               *y)
This function applies a bias and then an activation to the convolutions or cross-correlations of cudnnConvolutionForward(), returning results in y. The full computation follows the equation y = act ( alpha1 * conv(x) + alpha2 * z + bias ).
Note:

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha1, alpha2
Input. Pointers to scaling factors (in host memory) used to blend the computation result of convolution with z and bias as follows:
y = act ( alpha1 * conv(x) + alpha2 * z + bias )

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

wDesc

Input. Handle to a previously initialized filter descriptor. For more information, see cudnnFilterDescriptor_t.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

convDesc

Input. Previously initialized convolution descriptor. For more information, see cudnnConvolutionDescriptor_t.

algo

Input. Enumerant that specifies which convolution algorithm should be used to compute the results. For more information, see cudnnConvolutionFwdAlgo_t.

workSpace

Input. Data pointer to GPU memory to a workspace needed to be able to execute the specified algorithm. If no workspace is needed for a particular algorithm, that pointer can be nil.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

zDesc

Input. Handle to a previously initialized tensor descriptor.

z

Input. Data pointer to GPU memory associated with the tensor descriptor zDesc.

biasDesc

Input. Handle to a previously initialized tensor descriptor.

bias

Input. Data pointer to GPU memory associated with the tensor descriptor biasDesc.

activationDesc

Input. Handle to a previously initialized activation descriptor. For more information, see cudnnActivationDescriptor_t.

yDesc

Input. Handle to a previously initialized tensor descriptor.

y

Input/Output. Data pointer to GPU memory associated with the tensor descriptor yDesc that carries the result of the convolution.

For the convolution step, this function supports the specific combinations of data types for xDesc, wDesc, convDesc, and yDesc as listed in the documentation of cudnnConvolutionForward(). The following table specifies the supported combinations of data types for x, y, z, bias, and alpha1/alpha2.

Table 14. Supported combinations of data types (X = CUDNN_DATA)
x w convDesc y and z bias alpha1/alpha2
X_DOUBLE X_DOUBLE X_DOUBLE X_DOUBLE X_DOUBLE X_DOUBLE
X_FLOAT X_FLOAT X_FLOAT X_FLOAT X_FLOAT X_FLOAT
X_HALF X_HALF X_FLOAT X_HALF X_HALF X_FLOAT
X_INT8 X_INT8 X_INT32 X_INT8 X_FLOAT X_FLOAT
X_INT8 X_INT8 X_INT32 X_FLOAT X_FLOAT X_FLOAT
X_INT8x4 X_INT8x4 X_INT32 X_INT8x4 X_FLOAT X_FLOAT
X_INT8x4 X_INT8x4 X_INT32 X_FLOAT X_FLOAT X_FLOAT
X_UINT8 X_INT8 X_INT32 X_INT8 X_FLOAT X_FLOAT
X_UINT8 X_INT8 X_INT32 X_FLOAT X_FLOAT X_FLOAT
X_UINT8x4 X_INT8x4 X_INT32 X_INT8x4 X_FLOAT X_FLOAT
X_UINT8x4 X_INT8x4 X_INT32 X_FLOAT X_FLOAT X_FLOAT
X_INT8x32 X_INT8x32 X_INT32 X_INT8x32 X_FLOAT X_FLOAT

Returns

In addition to the error values listed by the documentation of cudnnConvolutionForward(), the possible error values returned by this function and their meanings are listed below.
CUDNN_STATUS_SUCCESS

The operation was launched successfully.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • At least one of the following is NULL: handle, xDesc, wDesc, convDesc, yDesc, zDesc, biasDesc, activationDesc, xData, wData, yData, zData, bias, alpha1, alpha2.
  • The number of dimensions of xDesc, wDesc, yDesc, zDesc is not equal to the array length of convDesc + 2.
CUDNN_STATUS_NOT_SUPPORTED
The function does not support the provided configuration. Some examples of non-supported configurations are as follows:
  • The mode of activationDesc is neither CUDNN_ACTIVATION_RELU or CUDNN_ACTIVATION_IDENTITY.
  • The reluNanOpt of activationDesc is not CUDNN_NOT_PROPAGATE_NAN.
  • The second stride of biasDesc is not equal to one.
  • The first dimension of biasDesc is not equal to one.
  • The second dimension of biasDesc and the first dimension of filterDesc are not equal.
  • The data type of biasDesc does not correspond to the data type of yDesc as listed in the above data types table.
  • zDesc and destDesc do not match.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

5.2.10. cudnnConvolutionForward()

cudnnStatus_t cudnnConvolutionForward(
    cudnnHandle_t                       handle,
    const void                         *alpha,
    const cudnnTensorDescriptor_t       xDesc,
    const void                         *x,
    const cudnnFilterDescriptor_t       wDesc,
    const void                         *w,
    const cudnnConvolutionDescriptor_t  convDesc,
    cudnnConvolutionFwdAlgo_t           algo,
    void                               *workSpace,
    size_t                              workSpaceSizeInBytes,
    const void                         *beta,
    const cudnnTensorDescriptor_t       yDesc,
    void                               *y)

This function executes convolutions or cross-correlations over x using filters specified with w, returning results in y. Scaling factors alpha and beta can be used to scale the input tensor and the output tensor respectively.

Note: The routine cudnnGetConvolution2dForwardOutputDim() or cudnnGetConvolutionNdForwardOutputDim() can be used to determine the proper dimensions of the output tensor descriptor yDesc with respect to xDesc, convDesc, and wDesc.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

wDesc

Input. Handle to a previously initialized filter descriptor. For more information, see cudnnFilterDescriptor_t.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

convDesc

Input. Previously initialized convolution descriptor. For more information, see cudnnConvolutionDescriptor_t.

algo

Input. Enumerant that specifies which convolution algorithm should be used to compute the results. For more information, see cudnnConvolutionFwdAlgo_t.

workSpace

Input. Data pointer to GPU memory to a workspace needed to be able to execute the specified algorithm. If no workspace is needed for a particular algorithm, that pointer can be nil.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

yDesc

Input. Handle to a previously initialized tensor descriptor.

y

Input/Output. Data pointer to GPU memory associated with the tensor descriptor yDesc that carries the result of the convolution.

Supported configurations

This function supports the following combinations of data types for xDesc, wDesc, convDesc, and yDesc.

Table 15. Supported configurations
Data Type Configurations xDesc and wDesc convDesc yDesc
TRUE_HALF_CONFIG (only supported on architectures with true FP16 support, meaning, compute capability 5.3 and later) CUDNN_DATA_HALF CUDNN_DATA_HALF CUDNN_DATA_HALF
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT CUDNN_DATA_HALF
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE
INT8_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_INT8 CUDNN_DATA_INT32 CUDNN_DATA_INT8
INT8_EXT_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_INT8 CUDNN_DATA_INT32 CUDNN_DATA_FLOAT
INT8x4_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_INT8x4 CUDNN_DATA_INT32 CUDNN_DATA_INT8x4
INT8x4_EXT_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_INT8x4 CUDNN_DATA_INT32 CUDNN_DATA_FLOAT
UINT8x4_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_UINT8x4 CUDNN_DATA_INT32 CUDNN_DATA_UINT8x4
UINT8x4_EXT_CONFIG (only supported on architectures with DP4A support, meaning, compute capability 6.1 and later) CUDNN_DATA_UINT8x4 CUDNN_DATA_INT32 CUDNN_DATA_FLOAT
INT8x32_CONFIG (only supported on architectures with IMMA support, meaning compute capability 7.5 and later) CUDNN_DATA_INT8x32 CUDNN_DATA_INT32 CUDNN_DATA_INT8x32

Supported algorithms

Note: For this function, all algorithms perform deterministic computations. Specifying a separate algorithm can cause changes in performance and support.

The table below shows the list of the supported 2D and 3D convolutions. The 2D convolutions are described first, followed by the 3D convolutions.

For the following terms, the short-form versions shown in the parenthesis are used in the table below, for brevity:
  • CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_GEMM (_IMPLICIT_GEMM)
  • CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_PRECOMP_GEMM (_IMPLICIT_PRECOMP_GEMM)
  • CUDNN_CONVOLUTION_FWD_ALGO_GEMM (_GEMM)
  • CUDNN_CONVOLUTION_FWD_ALGO_DIRECT (_DIRECT)
  • CUDNN_CONVOLUTION_FWD_ALGO_FFT (_FFT)
  • CUDNN_CONVOLUTION_FWD_ALGO_FFT_TILING (_FFT_TILING)
  • CUDNN_CONVOLUTION_FWD_ALGO_WINOGRAD (_WINOGRAD)
  • CUDNN_CONVOLUTION_FWD_ALGO_WINOGRAD_NONFUSED (_WINOGRAD_NONFUSED)
  • CUDNN_TENSOR_NCHW (_NCHW)
  • CUDNN_TENSOR_NHWC (_NHWC)
  • CUDNN_TENSOR_NCHW_VECT_C (_NCHW_VECT_C)
Table 16. For 2D convolutions: wDesc: _NCHW
Filter descriptor wDesc: _NCHW (see cudnnTensorFormat_t)

convDesc Group count support: Greater than 0, for all algos.

Algo Name Tensor Formats Supported for xDesc Tensor Formats Supported for yDesc Data Type Configurations Supported Important
_IMPLICIT_GEMM All except _NCHW_VECT_C. All except _NCHW_VECT_C. TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: Greater than 0 for all dimensions
_IMPLICIT_PRECOMP_GEMM All except _NCHW_VECT_C. All except _NCHW_VECT_C. TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: 1 for all dimensions
_GEMM All except _NCHW_VECT_C. All except _NCHW_VECT_C. PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: 1 for all dimensions
_FFT NCHW HW-packed NCHW HW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

xDesc feature map height + 2 * convDesc zero-padding height must equal 256 or less

xDesc feature map width + 2 * convDesc zero-padding width must equal 256 or less

convDesc vertical and horizontal filter stride must equal 1

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

_FFT_TILING PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG is also supported when the task can be handled by 1D FFT, meaning, one of the filter dimensions, width or height is 1.

Dilation: 1 for all dimensions

When neither of wDesc filter dimension is 1, the filter width and height must not be larger than 32

When either of wDesc filter dimension is 1, the largest filter dimension should not exceed 256

convDesc vertical and horizontal filter stride must equal 1 when either the filter width or filter height is 1, otherwise the stride can be a 1 or 2

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

_WINOGRAD All except_NCHW_VECT_C. All except_NCHW_VECT_C. PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc vertical and horizontal filter stride must equal 1

wDesc filter height must be 3

wDesc filter width must be 3

_WINOGRAD_NONFUSED TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc vertical and horizontal filter stride must equal 1

wDesc filter (height, width) must be (3,3) or (5,5)

If wDesc filter (height, width) is (5,5), then data type config TRUE_HALF_CONFIG is not supported.

_DIRECT Currently not implemented in cuDNN.
Table 17. For 2D convolutions: wDesc: _NCHWC
Filter descriptor wDesc: _NCHWC

convDesc Group count support: Greater than 0.

Algo Name xDesc yDesc Data Type Configurations Supported Important
_IMPLICIT_GEMM

_IMPLICIT_PRECOMP_GEMM

_NCHW_VECT_C _NCHW_VECT_C INT8x4_CONFIG

UINT8x4_CONFIG

INT8x4_EXT_CONFIG

UINT8x4_EXT_CONFIG

Dilation: 1 for all dimensions
_IMPLICIT_PRECOMP_GEMM _NCHW_VECT_C _NCHW_VECT_C INT8x32_CONFIG Dilation: 1 for all dimensions

Requires compute capability 7.2 or above.

Table 18. For 2D convolutions: wDesc: _NHWC
Filter descriptor wDesc: _NHWC

convDesc Group count support: Greater than 0.

Algo Name xDesc yDesc Data Type Configurations Supported Important
_IMPLICIT_GEMM

_IMPLICIT_PRECOMP_GEMM

NHWC fully-packed NHWC fully-packed INT8_CONFIG

INT8_EXT_CONFIG

UINT8_CONFIG

UINT8_EXT_CONFIG

Dilation: 1 for all dimensions

Input and output feature maps must be a multiple of 4. Output features maps can be non-multiple in the case of INT8_EXT_CONFIG or UINT8_EXT_CONFIG.

_IMPLICIT_GEMM

_IMPLICIT_PRECOMP_GEMM

NHWC HWC-packed.

NHWC HWC-packed.

NCHW CHW-packed

TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

 
Table 19. For 3D convolutions: wDesc: _NCHW
Filter descriptor wDesc: _NCHW

convDesc Group count support: Greater than 0, for all algos.

Algo Name xDesc yDesc Data Type Configurations Supported Important
_IMPLICIT_GEMM All except _NCHW_VECT_C. All except _NCHW_VECT_C. PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: Greater than 0 for all dimensions
_IMPLICIT_PRECOMP_GEMM Dilation: Greater than 0 for all dimensions
_FFT_TILING NCDHW DHW-packed NCDHW DHW-packed Dilation: 1 for all dimensions

wDesc filter height must equal 16 or less

wDesc filter width must equal 16 or less

wDesc filter depth must equal 16 or less

convDesc must have all filter strides equal to 1

wDesc filter height must be greater than convDesc zero-padding height

wDesc filter width must be greater than convDesc zero-padding width

wDesc filter depth must be greater than convDesc zero-padding width

Table 20. For 3D convolutions: wDesc: _NHWC
Filter descriptor wDesc: _NHWC

convDesc Group count support: Greater than 0, for all algos.

Algo Name xDesc yDesc Data Type Configurations Supported Important
_IMPLICIT_PRECOMP_GEMM NDHWC

DHWC-packed

NDHWC

DHWC-packed

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: Greater than 0 for all dimensions
Note: Tensors can be converted to and from CUDNN_TENSOR_NCHW_VECT_C with cudnnTransformTensor().

Returns

CUDNN_STATUS_SUCCESS

The operation was launched successfully.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • At least one of the following is NULL: handle, xDesc, wDesc, convDesc, yDesc, xData, w, yData, alpha, beta
  • xDesc and yDesc have a non-matching number of dimensions
  • xDesc and wDesc have a non-matching number of dimensions
  • xDesc has fewer than three number of dimensions
  • xDesc's number of dimensions is not equal to convDesc array length + 2
  • xDesc and wDesc have a non-matching number of input feature maps per image (or group in case of grouped convolutions)
  • yDesc or wDesc indicate an output channel count that isn't a multiple of group count (if group count has been set in convDesc).
  • xDesc, wDesc, and yDesc have a non-matching data type
  • For some spatial dimension, wDesc has a spatial size that is larger than the input spatial size (including zero-padding size)
CUDNN_STATUS_NOT_SUPPORTED
At least one of the following conditions are met:
  • xDesc or yDesc have negative tensor striding
  • xDesc, wDesc, or yDesc has a number of dimensions that is not 4 or 5
  • yDesc spatial sizes do not match with the expected size as determined by cudnnGetConvolutionNdForwardOutputDim()
  • The chosen algo does not support the parameters provided; see above for an exhaustive list of parameters supported for each algo
CUDNN_STATUS_MAPPING_ERROR

An error occurred during the texture binding of the filter data.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

5.2.11. cudnnCreateConvolutionDescriptor()

cudnnStatus_t cudnnCreateConvolutionDescriptor(
    cudnnConvolutionDescriptor_t *convDesc)

This function creates a convolution descriptor object by allocating the memory needed to hold its opaque structure. For more information, see cudnnConvolutionDescriptor_t.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

5.2.12. cudnnDestroyConvolutionDescriptor()

cudnnStatus_t cudnnDestroyConvolutionDescriptor(
    cudnnConvolutionDescriptor_t convDesc)

This function destroys a previously created convolution descriptor object.

Returns

CUDNN_STATUS_SUCCESS
The descriptor was destroyed successfully.

5.2.13. cudnnFindConvolutionBackwardDataAlgorithm()

cudnnStatus_t cudnnFindConvolutionBackwardDataAlgorithm(
    cudnnHandle_t                          handle,
    const cudnnFilterDescriptor_t          wDesc,
    const cudnnTensorDescriptor_t          dyDesc,
    const cudnnConvolutionDescriptor_t     convDesc,
    const cudnnTensorDescriptor_t          dxDesc,
    const int                              requestedAlgoCount,
    int                                   *returnedAlgoCount,
    cudnnConvolutionBwdDataAlgoPerf_t     *perfResults)

This function attempts all algorithms available for cudnnConvolutionBackwardData(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionBwdDataAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionBackwardDataAlgorithmMaxCount().

Note:
  • This function is host blocking.
  • It is recommended to run this function prior to allocating layer data; doing otherwise may needlessly inhibit some algorithm options due to resource usage.

Parameters

handle

Input. Handle to a previously created cuDNN context.

wDesc

Input. Handle to a previously initialized filter descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dxDesc

Input. Handle to the previously initialized output tensor descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • wDesc, dyDesc, or dxDesc is not allocated properly.
  • wDesc, dyDesc, or dxDesc has fewer than 1 dimension.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_ALLOC_FAILED

This function was unable to allocate memory to store sample input, filters and output.

CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

5.2.14. cudnnFindConvolutionBackwardDataAlgorithmEx()

cudnnStatus_t cudnnFindConvolutionBackwardDataAlgorithmEx(
    cudnnHandle_t                          handle,
    const cudnnFilterDescriptor_t          wDesc,
    const void                            *w,
    const cudnnTensorDescriptor_t          dyDesc,
    const void                            *dy,
    const cudnnConvolutionDescriptor_t     convDesc,
    const cudnnTensorDescriptor_t          dxDesc,
    void                                  *dx,
    const int                              requestedAlgoCount,
    int                                   *returnedAlgoCount,
    cudnnConvolutionBwdDataAlgoPerf_t     *perfResults,
    void                                  *workSpace,
    size_t                                 workSpaceSizeInBytes)

This function attempts all algorithms available for cudnnConvolutionBackwardData(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionBwdDataAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionBackwardDataAlgorithmMaxCount().

Note: This function is host blocking.

Parameters

handle

Input. Handle to a previously created cuDNN context.

wDesc

Input. Handle to a previously initialized filter descriptor.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the filter descriptor dyDesc.

convDesc

Input. Previously initialized convolution descriptor.

dxDesc

Input. Handle to the previously initialized output tensor descriptor.

dxDesc

Input/Output. Data pointer to GPU memory associated with the tensor descriptor dxDesc. The content of this tensor will be overwritten with arbitrary values.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workSpace

Input. Data pointer to GPU memory is a necessary workspace for some algorithms. The size of this workspace will determine the availability of algorithms. A nil pointer is considered a workSpace of 0 bytes.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • wDesc, dyDesc, or dxDesc is not allocated properly.
  • wDesc, dyDesc, or dxDesc has fewer than 1 dimension.
  • w, dy, or dx is nil.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

5.2.15. cudnnFindConvolutionForwardAlgorithm()

cudnnStatus_t cudnnFindConvolutionForwardAlgorithm(
    cudnnHandle_t                      handle,
    const cudnnTensorDescriptor_t      xDesc,
    const cudnnFilterDescriptor_t      wDesc,
    const cudnnConvolutionDescriptor_t convDesc,
    const cudnnTensorDescriptor_t      yDesc,
    const int                          requestedAlgoCount,
    int                               *returnedAlgoCount,
    cudnnConvolutionFwdAlgoPerf_t     *perfResults)

This function attempts all algorithms available for cudnnConvolutionForward(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionFwdAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionForwardAlgorithmMaxCount().

Note:
  • This function is host blocking.
  • It is recommended to run this function prior to allocating layer data; doing otherwise may needlessly inhibit some algorithm options due to resource usage.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

wDesc

Input. Handle to a previously initialized filter descriptor.

convDesc

Input. Previously initialized convolution descriptor.

yDesc

Input. Handle to the previously initialized output tensor descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • xDesc, wDesc, or yDesc are not allocated properly.
  • xDesc, wDesc, or yDesc has fewer than 1 dimension.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_ALLOC_FAILED

This function was unable to allocate memory to store sample input, filters and output.

CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

5.2.16. cudnnFindConvolutionForwardAlgorithmEx()

cudnnStatus_t cudnnFindConvolutionForwardAlgorithmEx(
    cudnnHandle_t                      handle,
    const cudnnTensorDescriptor_t      xDesc,
    const void                        *x,
    const cudnnFilterDescriptor_t      wDesc,
    const void                        *w,
    const cudnnConvolutionDescriptor_t convDesc,
    const cudnnTensorDescriptor_t      yDesc,
    void                              *y,
    const int                          requestedAlgoCount,
    int                               *returnedAlgoCount,
    cudnnConvolutionFwdAlgoPerf_t     *perfResults,
    void                              *workSpace,
    size_t                             workSpaceSizeInBytes)

This function attempts all algorithms available for cudnnConvolutionForward(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionFwdAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionForwardAlgorithmMaxCount().

Note: This function is host blocking.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

wDesc

Input. Handle to a previously initialized filter descriptor.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

convDesc

Input. Previously initialized convolution descriptor.

yDesc

Input. Handle to the previously initialized output tensor descriptor.

y

Input/Output. Data pointer to GPU memory associated with the tensor descriptor yDesc. The content of this tensor will be overwritten with arbitrary values.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workSpace

Input. Data pointer to GPU memory is a necessary workspace for some algorithms. The size of this workspace will determine the availability of algorithms. A nil pointer is considered a workSpace of 0 bytes.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • xDesc, wDesc, or yDesc are not allocated properly.
  • xDesc, wDesc, or yDesc has fewer than 1 dimension.
  • x, w, or y is nil.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

5.2.17. cudnnGetConvolution2dDescriptor()

cudnnStatus_t cudnnGetConvolution2dDescriptor(
    const cudnnConvolutionDescriptor_t  convDesc,
    int                                *pad_h,
    int                                *pad_w,
    int                                *u,
    int                                *v,
    int                                *dilation_h,
    int                                *dilation_w,
    cudnnConvolutionMode_t             *mode,
    cudnnDataType_t                    *computeType)

This function queries a previously initialized 2D convolution descriptor object.

Parameters

convDesc

Input/Output. Handle to a previously created convolution descriptor.

pad_h

Output. Zero-padding height: number of rows of zeros implicitly concatenated onto the top and onto the bottom of input images.

pad_w

Output. Zero-padding width: number of columns of zeros implicitly concatenated onto the left and onto the right of input images.

u

Output. Vertical filter stride.

v

Output. Horizontal filter stride.

dilation_h

Output. Filter height dilation.

dilation_w

Output. Filter width dilation.

mode

Output. Convolution mode.

computeType

Output. Compute precision.

Returns

CUDNN_STATUS_SUCCESS

The operation was successful.

CUDNN_STATUS_BAD_PARAM

The parameter convDesc is nil.

5.2.18. cudnnGetConvolution2dForwardOutputDim()

cudnnStatus_t cudnnGetConvolution2dForwardOutputDim(
    const cudnnConvolutionDescriptor_t  convDesc,
    const cudnnTensorDescriptor_t       inputTensorDesc,
    const cudnnFilterDescriptor_t       filterDesc,
    int                                *n,
    int                                *c,
    int                                *h,
    int                                *w)

This function returns the dimensions of the resulting 4D tensor of a 2D convolution, given the convolution descriptor, the input tensor descriptor and the filter descriptor This function can help to setup the output tensor and allocate the proper amount of memory prior to launch the actual convolution.

Each dimension h and w of the output images is computed as follows:

    outputDim = 1 + ( inputDim + 2*pad - (((filterDim-1)*dilation)+1) )/convolutionStride;
    
Note: The dimensions provided by this routine must be strictly respected when calling cudnnConvolutionForward() or cudnnConvolutionBackwardBias(). Providing a smaller or larger output tensor is not supported by the convolution routines.

Parameters

convDesc

Input. Handle to a previously created convolution descriptor.

inputTensorDesc

Input. Handle to a previously initialized tensor descriptor.

filterDesc

Input. Handle to a previously initialized filter descriptor.

n

Output. Number of output images.

c

Output. Number of output feature maps per image.

h

Output. Height of each output feature map.

w

Output. Width of each output feature map.

Returns

CUDNN_STATUS_BAD_PARAM

One or more of the descriptors has not been created correctly or there is a mismatch between the feature maps of inputTensorDesc and filterDesc.

CUDNN_STATUS_SUCCESS

The object was set successfully.

5.2.19. cudnnGetConvolutionBackwardDataAlgorithmMaxCount()

cudnnStatus_t cudnnGetConvolutionBackwardDataAlgorithmMaxCount(
    cudnnHandle_t       handle,
    int                 *count)

This function returns the maximum number of algorithms which can be returned from cudnnFindConvolutionBackwardDataAlgorithm() and cudnnGetConvolutionForwardAlgorithm_v7(). This is the sum of all algorithms plus the sum of all algorithms with Tensor Core operations supported for the current device.

Parameters

handle

Input. Handle to a previously created cuDNN context.

count

Output. The resulting maximum number of algorithms.

Returns

CUDNN_STATUS_SUCCESS

The function was successful.

CUDNN_STATUS_BAD_PARAM

The provided handle is not allocated properly.

5.2.20. cudnnGetConvolutionBackwardDataAlgorithm_v7()

cudnnStatus_t cudnnGetConvolutionBackwardDataAlgorithm_v7(
    cudnnHandle_t                          handle,
    const cudnnFilterDescriptor_t          wDesc,
    const cudnnTensorDescriptor_t          dyDesc,
    const cudnnConvolutionDescriptor_t     convDesc,
    const cudnnTensorDescriptor_t          dxDesc,
    const int                              requestedAlgoCount,
    int                                   *returnedAlgoCount,
    cudnnConvolutionBwdDataAlgoPerf_t     *perfResults)

This function serves as a heuristic for obtaining the best suited algorithm for cudnnConvolutionBackwardData() for the given layer specifications. This function will return all algorithms (including CUDNN_TENSOR_OP_MATH and CUDNN_DEFAULT_MATH versions of algorithms where CUDNN_TENSOR_OP_MATH may be available) sorted by expected (based on internal heuristic) relative performance with the fastest being index 0 of perfResults. For an exhaustive search for the fastest algorithm, use cudnnFindConvolutionBackwardDataAlgorithm(). The total number of resulting algorithms can be queried through the returnedAlgoCount variable.

Parameters

handle

Input. Handle to a previously created cuDNN context.

wDesc

Input. Handle to a previously initialized filter descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dxDesc

Input. Handle to the previously initialized output tensor descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters handle, wDesc, dyDesc, convDesc, dxDesc, perfResults, returnedAlgoCount is NULL.
  • The numbers of feature maps of the input tensor and output tensor differ.
  • The dataType of the two tensor descriptors or the filters are different.
  • requestedAlgoCount is less than or equal to 0.

5.2.21. cudnnGetConvolutionBackwardDataWorkspaceSize()

cudnnStatus_t cudnnGetConvolutionBackwardDataWorkspaceSize(
    cudnnHandle_t                       handle,
    const cudnnFilterDescriptor_t       wDesc,
    const cudnnTensorDescriptor_t       dyDesc,
    const cudnnConvolutionDescriptor_t  convDesc,
    const cudnnTensorDescriptor_t       dxDesc,
    cudnnConvolutionBwdDataAlgo_t       algo,
    size_t                             *sizeInBytes)

This function returns the amount of GPU memory workspace the user needs to allocate to be able to call cudnnConvolutionBackwardData() with the specified algorithm. The workspace allocated will then be passed to the routine cudnnConvolutionBackwardData(). The specified algorithm can be the result of the call to cudnnGetConvolutionBackwardDataAlgorithm_v7() or can be chosen arbitrarily by the user. Note that not every algorithm is available for every configuration of the input tensor and/or every configuration of the convolution descriptor.

Parameters

handle

Input. Handle to a previously created cuDNN context.

wDesc

Input. Handle to a previously initialized filter descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dxDesc

Input. Handle to the previously initialized output tensor descriptor.

algo

Input. Enumerant that specifies the chosen convolution algorithm.

sizeInBytes

Output. Amount of GPU memory needed as workspace to be able to execute a forward convolution with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The numbers of feature maps of the input tensor and output tensor differ.
  • The dataType of the two tensor descriptors or the filter are different.
CUDNN_STATUS_NOT_SUPPORTED

The combination of the tensor descriptors, filter descriptor and convolution descriptor is not supported for the specified algorithm.

5.2.22. cudnnGetConvolutionForwardAlgorithmMaxCount()

cudnnStatus_t cudnnGetConvolutionForwardAlgorithmMaxCount(
    cudnnHandle_t   handle,
    int             *count)

This function returns the maximum number of algorithms which can be returned from cudnnFindConvolutionForwardAlgorithm() and cudnnGetConvolutionForwardAlgorithm_v7(). This is the sum of all algorithms plus the sum of all algorithms with Tensor Core operations supported for the current device.

Parameters

handle

Input. Handle to a previously created cuDNN context.

count

Output. The resulting maximum number of algorithms.

Returns

CUDNN_STATUS_SUCCESS

The function was successful.

CUDNN_STATUS_BAD_PARAM

The provided handle is not allocated properly.

5.2.23. cudnnGetConvolutionForwardAlgorithm_v7()

cudnnStatus_t cudnnGetConvolutionForwardAlgorithm_v7(
    cudnnHandle_t                       handle,
    const cudnnTensorDescriptor_t       xDesc,
    const cudnnFilterDescriptor_t       wDesc,
    const cudnnConvolutionDescriptor_t  convDesc,
    const cudnnTensorDescriptor_t       yDesc,
    const int                           requestedAlgoCount,
    int                                *returnedAlgoCount,
    cudnnConvolutionFwdAlgoPerf_t      *perfResults)

This function serves as a heuristic for obtaining the best suited algorithm for cudnnConvolutionForward() for the given layer specifications. This function will return all algorithms (including CUDNN_TENSOR_OP_MATH and CUDNN_DEFAULT_MATH versions of algorithms where CUDNN_TENSOR_OP_MATH may be available) sorted by expected (based on internal heuristic) relative performance with the fastest being index 0 of perfResults. For an exhaustive search for the fastest algorithm, use cudnnFindConvolutionForwardAlgorithm(). The total number of resulting algorithms can be queried through the returnedAlgoCount variable.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

wDesc

Input. Handle to a previously initialized convolution filter descriptor.

convDesc

Input. Previously initialized convolution descriptor.

yDesc

Input. Handle to the previously initialized output tensor descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters handle, xDesc, wDesc, convDesc, yDesc, perfResults, returnedAlgoCount is NULL.
  • Either yDesc or wDesc have different dimensions from xDesc.
  • The data types of tensors xDesc, yDesc or wDesc are not all the same.
  • The number of feature maps in xDesc and wDesc differs.
  • The tensor xDesc has a dimension smaller than 3.
  • requestedAlgoCount is less than or equal to 0.

5.2.24. cudnnGetConvolutionForwardWorkspaceSize()

cudnnStatus_t cudnnGetConvolutionForwardWorkspaceSize(
    cudnnHandle_t   handle,
    const   cudnnTensorDescriptor_t         xDesc,
    const   cudnnFilterDescriptor_t         wDesc,
    const   cudnnConvolutionDescriptor_t    convDesc,
    const   cudnnTensorDescriptor_t         yDesc,
    cudnnConvolutionFwdAlgo_t               algo,
    size_t                                 *sizeInBytes)

This function returns the amount of GPU memory workspace the user needs to allocate to be able to call cudnnConvolutionForward() with the specified algorithm. The workspace allocated will then be passed to the routine cudnnConvolutionForward(). The specified algorithm can be the result of the call to cudnnGetConvolutionForwardAlgorithm_v7() or can be chosen arbitrarily by the user. Note that not every algorithm is available for every configuration of the input tensor and/or every configuration of the convolution descriptor.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized x tensor descriptor.

wDesc

Input. Handle to a previously initialized filter descriptor.

convDesc

Input. Previously initialized convolution descriptor.

yDesc

Input. Handle to the previously initialized y tensor descriptor.

algo

Input. Enumerant that specifies the chosen convolution algorithm.

sizeInBytes

Output. Amount of GPU memory needed as workspace to be able to execute a forward convolution with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters handle, xDesc, wDesc, convDesc, yDesc is NULL.
  • The tensor yDesc or wDesc are not of the same dimension as xDesc.
  • The tensor xDesc, yDesc or wDesc are not of the same data type.
  • The numbers of feature maps of the tensor xDesc and wDesc differ.
  • The tensor xDesc has a dimension smaller than 3.
CUDNN_STATUS_NOT_SUPPORTED

The combination of the tensor descriptors, filter descriptor and convolution descriptor is not supported for the specified algorithm.

5.2.25. cudnnGetConvolutionGroupCount()

cudnnStatus_t cudnnGetConvolutionGroupCount(
    cudnnConvolutionDescriptor_t    convDesc,
    int                            *groupCount)

This function returns the group count specified in the given convolution descriptor.

Returns

CUDNN_STATUS_SUCCESS

The group count was returned successfully.

CUDNN_STATUS_BAD_PARAM

An invalid convolution descriptor was provided.

5.2.26. cudnnGetConvolutionMathType()

cudnnStatus_t cudnnGetConvolutionMathType(
    cudnnConvolutionDescriptor_t    convDesc,
    cudnnMathType_t                *mathType)

This function returns the math type specified in a given convolution descriptor.

Returns

CUDNN_STATUS_SUCCESS

The math type was returned successfully.

CUDNN_STATUS_BAD_PARAM

An invalid convolution descriptor was provided.

5.2.27. cudnnGetConvolutionNdDescriptor()

cudnnStatus_t cudnnGetConvolutionNdDescriptor(
    const cudnnConvolutionDescriptor_t  convDesc,
    int                                 arrayLengthRequested,
    int                                *arrayLength,
    int                                 padA[],
    int                                 filterStrideA[],
    int                                 dilationA[],
    cudnnConvolutionMode_t             *mode,
    cudnnDataType_t                    *dataType)

This function queries a previously initialized convolution descriptor object.

Parameters

convDesc

Input/Output. Handle to a previously created convolution descriptor.

arrayLengthRequested

Input. Dimension of the expected convolution descriptor. It is also the minimum size of the arrays padA, filterStrideA, and dilationA in order to be able to hold the results

arrayLength

Output. Actual dimension of the convolution descriptor.

padA

Output. Array of dimension of at least arrayLengthRequested that will be filled with the padding parameters from the provided convolution descriptor.

filterStrideA

Output. Array of dimension of at least arrayLengthRequested that will be filled with the filter stride from the provided convolution descriptor.

dilationA

Output. Array of dimension of at least arrayLengthRequested that will be filled with the dilation parameters from the provided convolution descriptor.

mode

Output. Convolution mode of the provided descriptor.

datatype

Output. Datatype of the provided descriptor.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor convDesc is nil.
  • The arrayLengthRequest is negative.
CUDNN_STATUS_NOT_SUPPORTED

The arrayLengthRequested is greater than CUDNN_DIM_MAX-2.

5.2.28. cudnnGetConvolutionNdForwardOutputDim()

cudnnStatus_t cudnnGetConvolutionNdForwardOutputDim(
    const cudnnConvolutionDescriptor_t  convDesc,
    const cudnnTensorDescriptor_t       inputTensorDesc,
    const cudnnFilterDescriptor_t       filterDesc,
    int                                 nbDims,
    int                                 tensorOuputDimA[])

This function returns the dimensions of the resulting n-D tensor of a nbDims-2-D convolution, given the convolution descriptor, the input tensor descriptor and the filter descriptor This function can help to setup the output tensor and allocate the proper amount of memory prior to launch the actual convolution.

Each dimension of the (nbDims-2)-D images of the output tensor is computed as follows:
    outputDim = 1 + ( inputDim + 2*pad - (((filterDim-1)*dilation)+1) )/convolutionStride;
    
Note: The dimensions provided by this routine must be strictly respected when calling cudnnConvolutionForward() or cudnnConvolutionBackwardBias(). Providing a smaller or larger output tensor is not supported by the convolution routines.

Parameters

convDesc

Input. Handle to a previously created convolution descriptor.

inputTensorDesc

Input. Handle to a previously initialized tensor descriptor.

filterDesc

Input. Handle to a previously initialized filter descriptor.

nbDims

Input. Dimension of the output tensor.

tensorOuputDimA

Output. Array of dimensions nbDims that contains on exit of this routine the sizes of the output tensor.

Returns

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters convDesc, inputTensorDesc, and filterDesc is nil.
  • The dimension of the filter descriptor filterDesc is different from the dimension of input tensor descriptor inputTensorDesc.
  • The dimension of the convolution descriptor is different from the dimension of input tensor descriptor inputTensorDesc-2.
  • The features map of the filter descriptor filterDesc is different from the one of input tensor descriptor inputTensorDesc.
  • The size of the dilated filter filterDesc is larger than the padded sizes of the input tensor.
  • The dimension nbDims of the output array is negative or greater than the dimension of input tensor descriptor inputTensorDesc.
CUDNN_STATUS_SUCCESS

The routine exited successfully.

5.2.29. cudnnGetConvolutionReorderType()

cudnnStatus_t cudnnGetConvolutionReorderType(
	cudnnConvolutionDescriptor_t convDesc,
	cudnnReorderType_t *reorderType);		

This function retrieves the convolution reorder type from the given convolution descriptor.

Parameters

convDesc
Input. The convolution descriptor from which the reorder type should be retrieved.
reorderType
Output. The retrieved reorder type. For more information, see cudnnReorderType_t.

Returns

CUDNN_STATUS_BAD_PARAM
One of the inputs to this function is not valid.
CUDNN_STATUS_SUCCESS
The reorder type is retrieved successfully.

5.2.30. cudnnGetFoldedConvBackwardDataDescriptors()

cudnnStatus_t
cudnnGetFoldedConvBackwardDataDescriptors(const cudnnHandle_t handle,
                                          const cudnnFilterDescriptor_t filterDesc,
                                          const cudnnTensorDescriptor_t diffDesc,
                                          const cudnnConvolutionDescriptor_t convDesc,
                                          const cudnnTensorDescriptor_t gradDesc,
                                          const cudnnTensorFormat_t transformFormat,
                                          cudnnFilterDescriptor_t foldedFilterDesc,
                                          cudnnTensorDescriptor_t paddedDiffDesc,
                                          cudnnConvolutionDescriptor_t foldedConvDesc,
                                          cudnnTensorDescriptor_t foldedGradDesc,
                                          cudnnTensorTransformDescriptor_t filterFoldTransDesc,
                                          cudnnTensorTransformDescriptor_t diffPadTransDesc,
                                          cudnnTensorTransformDescriptor_t gradFoldTransDesc,
                                          cudnnTensorTransformDescriptor_t gradUnfoldTransDesc) ;

This function calculates folding descriptors for backward data gradient. It takes as input the data descriptors along with convolution descriptor and computes the folded data descriptors and the folding transform descriptors. These can then be used to do the actual folding transform.

Parameters

handle

Input. Handle to a previously created cuDNN context.

filterDesc
Input. Filter descriptor before folding.
diffDesc

Input. Diff descriptor before folding.

convDesc

Input. Convolution descriptor before folding.

gradDesc

Input. Gradient descriptor before folding.

transformFormat

Input. Transform format for folding.

foldedFilterDesc

Output. Folded filter descriptor.

paddedDiffDesc

Output. Padded Diff descriptor.

foldedConvDesc

Output. Folded convolution descriptor.

foldedGradDesc

Output. Folded gradient descriptor.

filterFoldTransDesc

Output. Folding transform descriptor for filter.

diffPadTransDesc

Output. Folding transform descriptor for Desc.

gradFoldTransDesc

Output. Folding transform descriptor for gradient.

gradUnfoldTransDesc

Output. Unfolding transform descriptor for folded gradient.

Returns

CUDNN_STATUS_SUCCESS

Folded descriptors were computed successfully.

CUDNN_STATUS_BAD_PARAM

If any of the input parameters is NULL or if the input tensor has more than 4 dimensions.

CUDNN_STATUS_EXECUTION_FAILED

Computing the folded descriptors failed.

5.2.31. cudnnIm2Col()

cudnnStatus_t cudnnIm2Col(
    cudnnHandle_t                   handle,
    cudnnTensorDescriptor_t         srcDesc,
    const void                      *srcData,
    cudnnFilterDescriptor_t         filterDesc,   
    cudnnConvolutionDescriptor_t    convDesc,
    void                            *colBuffer)
This function constructs the A matrix necessary to perform a forward pass of GEMM convolution. This A matrix has a height of batch_size*y_height*y_width and width of input_channels*filter_height*filter_width, where:
  • batch_size is xDesc first dimension
  • y_height/y_width are computed from cudnnGetConvolutionNdForwardOutputDim()
  • input_channels is xDesc second dimension
  • filter_height/filter_width are wDesc third and fourth dimension
The A matrix is stored in format HW fully-packed in GPU memory.

Parameters

handle

Input. Handle to a previously created cuDNN context.

srcDesc

Input. Handle to a previously initialized tensor descriptor.

srcData

Input. Data pointer to GPU memory associated with the input tensor descriptor.

filterDesc

Input. Handle to a previously initialized filter descriptor.

convDesc

Input. Handle to a previously initialized convolution descriptor.

colBuffer

Output. Data pointer to GPU memory storing the output matrix.

Returns

CUDNN_STATUS_BAD_PARAM

srcData or colBuffer is NULL.

CUDNN_STATUS_NOT_SUPPORTED

Any of srcDesc, filterDesc, convDesc has dataType of CUDNN_DATA_INT8, CUDNN_DATA_INT8x4, CUDNN_DATA_INT8 or CUDNN_DATA_INT8x4convDesc has groupCount larger than 1.

CUDNN_STATUS_EXECUTION_FAILED

The CUDA kernel execution was unsuccessful.

CUDNN_STATUS_SUCCESS

The output data array is successfully generated.

5.2.32. cudnnReorderFilterAndBias()

cudnnStatus_t cudnnReorderFilterAndBias(
	cudnnHandle_t handle,
	const cudnnFilterDescriptor_t filterDesc,
	cudnnReorderType_t reorderType,
	const void *filterData,
	void *reorderedFilterData,
	int reorderBias,
	const void *biasData,
	void *reorderedBiasData);		

This function cudnnReorderFilterAndBias() reorders the filter and bias values. It can be used to enhance the inference time by separating the reordering operation from convolution.

For example, convolutions in a neural network of multiple layers can require reordering of kernels at every layer, which can take up a significant fraction of the total inference time. Using this function, the reordering can be done one time on the filter and bias data followed by the convolution operations at the multiple layers, thereby enhancing the inference time.

Parameters

filterDesc
Input. Descriptor for the kernel dataset.
reorderType
Input. Setting to either perform reordering or not. For more information, see cudnnReorderType_t.
filterData
Input. Pointer to the filter (kernel) data location in the device memory.
reorderedFilterData
Input. Pointer to the location in the device memory where the reordered filter data will be written to, by this function.
reorderBias
Input. If > 0, then reorders the bias data also. If <= 0 then does not perform reordering operations on the bias data.
biasData
Input. Pointer to the bias data location in the device memory.
reorderedBiasData
Input. Pointer to the location in the device memory where the reordered bias data will be written to, by this function.

Returns

CUDNN_STATUS_SUCCESS
Reordering was successful.
CUDNN_STATUS_EXECUTION_FAILED
Either the reordering of the filter data or of the bias data failed.

5.2.33. cudnnSetConvolution2dDescriptor()

cudnnStatus_t cudnnSetConvolution2dDescriptor(
    cudnnConvolutionDescriptor_t    convDesc,
    int                             pad_h,
    int                             pad_w,
    int                             u,
    int                             v,
    int                             dilation_h,
    int                             dilation_w,
    cudnnConvolutionMode_t          mode,
    cudnnDataType_t                 computeType)

This function initializes a previously created convolution descriptor object into a 2D correlation. This function assumes that the tensor and filter descriptors correspond to the forward convolution path and checks if their settings are valid. That same convolution descriptor can be reused in the backward path provided it corresponds to the same layer.

Parameters

convDesc

Input/Output. Handle to a previously created convolution descriptor.

pad_h

Input. Zero-padding height: number of rows of zeros implicitly concatenated onto the top and onto the bottom of input images.

pad_w

Input. Zero-padding width: number of columns of zeros implicitly concatenated onto the left and onto the right of input images.

u

Input. Vertical filter stride.

v

Input. Horizontal filter stride.

dilation_h

Input. Filter height dilation.

dilation_w

Input. Filter width dilation.

mode

Input. Selects between CUDNN_CONVOLUTION and CUDNN_CROSS_CORRELATION.

computeType

Input. Compute precision.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor convDesc is nil.
  • One of the parameters pad_h, pad_w is strictly negative.
  • One of the parameters u, v is negative or zero.
  • One of the parameters dilation_h, dilation_w is negative or zero.
  • The parameter mode has an invalid enumerant value.

5.2.34. cudnnSetConvolutionGroupCount()

cudnnStatus_t cudnnSetConvolutionGroupCount(
    cudnnConvolutionDescriptor_t    convDesc,
    int                             groupCount)

This function allows the user to specify the number of groups to be used in the associated convolution.

Returns

CUDNN_STATUS_SUCCESS

The group count was set successfully.

CUDNN_STATUS_BAD_PARAM

An invalid convolution descriptor was provided

5.2.35. cudnnSetConvolutionMathType()

cudnnStatus_t cudnnSetConvolutionMathType(
    cudnnConvolutionDescriptor_t    convDesc,
    cudnnMathType_t                 mathType)

This function allows the user to specify whether or not the use of tensor op is permitted in the library routines associated with a given convolution descriptor.

Returns

CUDNN_STATUS_SUCCESS

The math type was set successfully.

CUDNN_STATUS_BAD_PARAM

Either an invalid convolution descriptor was provided or an invalid math type was specified.

5.2.36. cudnnSetConvolutionNdDescriptor()

cudnnStatus_t cudnnSetConvolutionNdDescriptor(
    cudnnConvolutionDescriptor_t    convDesc,
    int                             arrayLength,
    const int                       padA[],
    const int                       filterStrideA[],
    const int                       dilationA[],
    cudnnConvolutionMode_t          mode,
    cudnnDataType_t                 dataType)

This function initializes a previously created generic convolution descriptor object into a n-D correlation. That same convolution descriptor can be reused in the backward path provided it corresponds to the same layer. The convolution computation will be done in the specified dataType, which can be potentially different from the input/output tensors.

Parameters

convDesc

Input/Output. Handle to a previously created convolution descriptor.

arrayLength

Input. Dimension of the convolution.

padA

Input. Array of dimension arrayLength containing the zero-padding size for each dimension. For every dimension, the padding represents the number of extra zeros implicitly concatenated at the start and at the end of every element of that dimension.

filterStrideA

Input. Array of dimension arrayLength containing the filter stride for each dimension. For every dimension, the filter stride represents the number of elements to slide to reach the next start of the filtering window of the next point.

dilationA

Input. Array of dimension arrayLength containing the dilation factor for each dimension.

mode

Input. Selects between CUDNN_CONVOLUTION and CUDNN_CROSS_CORRELATION.

datatype
Input. Selects the data type in which the computation will be done.
Note:CUDNN_DATA_HALF in cudnnSetConvolutionNdDescriptor() with HALF_CONVOLUTION_BWD_FILTER is not recommended as it is known to not be useful for any practical use case for training and will be considered to be blocked in a future cuDNN release. The use of CUDNN_DATA_HALF for input tensors in cudnnSetTensorNdDescriptor() and CUDNN_DATA_FLOAT in cudnnSetConvolutionNdDescriptor() with HALF_CONVOLUTION_BWD_FILTER is recommended and is used with the automatic mixed precision (AMP) training in many well known deep learning frameworks.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor convDesc is nil.
  • The arrayLengthRequest is negative.
  • The enumerant mode has an invalid value.
  • The enumerant datatype has an invalid value.
  • One of the elements of padA is strictly negative.
  • One of the elements of strideA is negative or zero.
  • One of the elements of dilationA is negative or zero.
CUDNN_STATUS_NOT_SUPPORTED

At least one of the following conditions are met:

  • The arrayLengthRequest is greater than CUDNN_DIM_MAX.

5.2.37. cudnnSetConvolutionReorderType()

cudnnStatus_t cudnnSetConvolutionReorderType(
	cudnnConvolutionDescriptor_t convDesc, 
	cudnnReorderType_t reorderType);		

This function sets the convolution reorder type for the given convolution descriptor.

Parameters

convDesc
Input. The convolution descriptor for which the reorder type should be set.
reorderType
Input. Set the reorder type to this value. For more information, see cudnnReorderType_t.

Returns

CUDNN_STATUS_BAD_PARAM
The reorder type supplied is not supported.
CUDNN_STATUS_SUCCESS
Reorder type is set successfully.

6. cudnn_cnn_train.so Library

6.1. Data Type References

6.1.1. Pointer To Opaque Struct Types

6.1.1.1. cudnnFusedOpsConstParamPack_t

cudnnFusedOpsConstParamPack_t is a pointer to an opaque structure holding the description of the cudnnFusedOps constant parameters. Use the function cudnnCreateFusedOpsConstParamPack() ​to create one instance of this structure, and the function cudnnDestroyFusedOpsConstParamPack() to destroy a previously-created descriptor.

6.1.1.2. cudnnFusedOpsPlan_t

cudnnFusedOpsPlan_t is a pointer to an opaque structure holding the description of the cudnnFusedOpsPlan. This descriptor contains the plan information, including the problem type and size, which kernels should be run, and the internal workspace partition. Use the function cudnnCreateFusedOpsPlan() to create one instance of this structure, and the function cudnnDestroyFusedOpsPlan() to destroy a previously-created descriptor.

6.1.1.3. cudnnFusedOpsVariantParamPack_t

cudnnFusedOpsVariantParamPack_t is a pointer to an opaque structure holding the description of the cudnnFusedOps variant parameters. Use the function cudnnCreateFusedOpsVariantParamPack() to create one instance of this structure, and the function cudnnDestroyFusedOpsVariantParamPack() to destroy a previously-created descriptor.

6.1.2. Struct Types

6.1.2.1. cudnnConvolutionBwdFilterAlgoPerf_t

cudnnConvolutionBwdFilterAlgoPerf_t is a structure containing performance results returned by cudnnFindConvolutionBackwardFilterAlgorithm() or heuristic results returned by cudnnGetConvolutionBackwardFilterAlgorithm_v7().

Data Members
cudnnConvolutionBwdFilterAlgo_t algo

The algorithm runs to obtain the associated performance metrics.

cudnnStatus_t status

If any error occurs during the workspace allocation or timing of cudnnConvolutionBackwardFilter(), this status will represent that error. Otherwise, this status will be the return status of cudnnConvolutionBackwardFilter().

  • CUDNN_STATUS_ALLOC_FAILED if any error occurred during workspace allocation or if the provided workspace is insufficient.
  • CUDNN_STATUS_INTERNAL_ERROR if any error occurred during timing calculations or workspace deallocation.
  • Otherwise, this will be the return status of cudnnConvolutionBackwardFilter().
float time

The execution time of cudnnConvolutionBackwardFilter() (in milliseconds).

size_t memory

The workspace size (in bytes).

cudnnDeterminism_t determinism

The determinism of the algorithm.

cudnnMathType_t mathType

The math type provided to the algorithm.

int reserved[3]

Reserved space for future properties.

6.1.3. Enumeration Types

6.1.3.1. cudnnFusedOps_t

The cudnnFusedOps_t type is an enumerated type to select a specific sequence of computations to perform in the fused operations.

Member Description
CUDNN_FUSED_SCALE_BIAS_ACTIVATION_CONV_BNSTATS = 0 On a per-channel basis, performs these operations in this order: scale, add bias, activation, convolution, and generate batchnorm statistics.
CUDNN_FUSED_SCALE_BIAS_ACTIVATION_WGRAD = 1 On a per-channel basis, performs these operations in this order: scale, add bias, activation, convolution backward weights, and generate batchnorm statistics.
CUDNN_FUSED_BN_FINALIZE_STATISTICS_TRAINING = 2 Computes the equivalent scale and bias from ySum, ySqSum and learned scale, bias.

Optionally update running statistics and generate saved stats

CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE = 3 Computes the equivalent scale and bias from the learned running statistics and the learned scale, bias.
CUDNN_FUSED_CONV_SCALE_BIAS_ADD_ACTIVATION = 4 On a per-channel basis, performs these operations in this order: convolution, scale, add bias, element-wise addition with another tensor, and activation.
CUDNN_FUSED_SCALE_BIAS_ADD_ACTIVATION_GEN_BITMASK = 5 On a per-channel basis, performs these operations in this order: scale and bias on one tensor, scale, and bias on a second tensor, element-wise addition of these two tensors, and on the resulting tensor perform activation, and generate activation bit mask.
CUDNN_FUSED_DACTIVATION_FORK_DBATCHNORM = 6 On a per-channel basis, performs these operations in this order: backward activation, fork (meaning, write out gradient for the residual branch), and backward batch norm.

6.1.3.2. cudnnFusedOpsConstParamLabel_t

The cudnnFusedOpsConstParamLabel_t is an enumerated type for the selection of the type of the cudnnFusedOps descriptor. For more information, see cudnnSetFusedOpsConstParamPackAttribute().

typedef enum {
	CUDNN_PARAM_XDESC                          = 0,
	CUDNN_PARAM_XDATA_PLACEHOLDER              = 1,
	CUDNN_PARAM_BN_MODE                        = 2,
	CUDNN_PARAM_BN_EQSCALEBIAS_DESC            = 3,
	CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER         = 4,
	CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER          = 5,
	CUDNN_PARAM_ACTIVATION_DESC                = 6,
	CUDNN_PARAM_CONV_DESC                      = 7,
	CUDNN_PARAM_WDESC                          = 8,
	CUDNN_PARAM_WDATA_PLACEHOLDER              = 9,
	CUDNN_PARAM_DWDESC                         = 10,
	CUDNN_PARAM_DWDATA_PLACEHOLDER             = 11,
	CUDNN_PARAM_YDESC                          = 12,
	CUDNN_PARAM_YDATA_PLACEHOLDER              = 13,
	CUDNN_PARAM_DYDESC                         = 14,
	CUDNN_PARAM_DYDATA_PLACEHOLDER             = 15,
	CUDNN_PARAM_YSTATS_DESC                    = 16,
	CUDNN_PARAM_YSUM_PLACEHOLDER               = 17,
	CUDNN_PARAM_YSQSUM_PLACEHOLDER             = 18,
	CUDNN_PARAM_BN_SCALEBIAS_MEANVAR_DESC      = 19,
	CUDNN_PARAM_BN_SCALE_PLACEHOLDER           = 20,
	CUDNN_PARAM_BN_BIAS_PLACEHOLDER            = 21,
	CUDNN_PARAM_BN_SAVED_MEAN_PLACEHOLDER      = 22,
	CUDNN_PARAM_BN_SAVED_INVSTD_PLACEHOLDER    = 23,
	CUDNN_PARAM_BN_RUNNING_MEAN_PLACEHOLDER    = 24,
	CUDNN_PARAM_BN_RUNNING_VAR_PLACEHOLDER     = 25,
	CUDNN_PARAM_ZDESC                          = 26,
	CUDNN_PARAM_ZDATA_PLACEHOLDER              = 27,
	CUDNN_PARAM_BN_Z_EQSCALEBIAS_DESC          = 28,
	CUDNN_PARAM_BN_Z_EQSCALE_PLACEHOLDER       = 29,
	CUDNN_PARAM_BN_Z_EQBIAS_PLACEHOLDER        = 30,
	CUDNN_PARAM_ACTIVATION_BITMASK_DESC        = 31,
	CUDNN_PARAM_ACTIVATION_BITMASK_PLACEHOLDER = 32,
	CUDNN_PARAM_DXDESC                         = 33,
	CUDNN_PARAM_DXDATA_PLACEHOLDER             = 34,
	CUDNN_PARAM_DZDESC                         = 35,
	CUDNN_PARAM_DZDATA_PLACEHOLDER             = 36,
	CUDNN_PARAM_BN_DSCALE_PLACEHOLDER          = 37,
	CUDNN_PARAM_BN_DBIAS_PLACEHOLDER           = 38,
	} cudnnFusedOpsConstParamLabel_t;
Short-form used Stands for
Setter cudnnSetFusedOpsConstParamPackAttribute()
Getter cudnnGetFusedOpsConstParamPackAttribute()
X_PointerPlaceHolder_t cudnnFusedOpsPointerPlaceHolder_t
X_ prefix in the Attribute column Stands for CUDNN_PARAM_ in the enumerator name
Table 21. CUDNN_FUSED_SCALE_BIAS_ACTIVATION_CONV_BNSTATS
For the attribute CUDNN_FUSED_SCALE_BIAS_ACTIVATION_CONV_BNSTATS in cudnnFusedOp_t
Attribute Expected Descriptor Type Passed in, in the Setter Description Default Value After Creation
X_XDESC In the setter, the *param should be xDesc, a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout, and datatype of the x (input) tensor. NULL
X_XDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether xData pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_MODE In the setter, the *param should be a pointer to a previously initialized cudnnBatchNormMode_t*. Describes the mode of operation for the scale, bias and the statistics.

As of cuDNN 7.6.0, only CUDNN_BATCHNORM_SPATIAL and CUDNN_BATCHNORM_SPATIAL_PERSISTENT are supported, meaning, scale, bias, and statistics are all per-channel.

CUDNN_BATCHNORM_PER_ACTIVATION
X_BN_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout, and datatype of the batchNorm equivalent scale and bias tensors. The shapes must match the mode specified in CUDNN_PARAM_BN_MODE. If set to NULL, both scale and bias operation will become a NOP. NULL
X_BN_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the scale operation becomes a NOP.

CUDNN_PTR_NULL
X_BN_EQBIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the bias operation becomes a NOP.

CUDNN_PTR_NULL
X_ACTIVATION_DESC In the setter, the *param should be a pointer to a previously initialized cudnnActivationDescriptor_t*. Describes the activation operation.

As of cuDNN 7.6.0, only activation modes of CUDNN_ACTIVATION_RELU and CUDNN_ACTIVATION_IDENTITY are supported. If set to NULL or if the activation mode is set to CUDNN_ACTIVATION_IDENTITY, then the activation in the op sequence becomes a NOP.

NULL
X_CONV_DESC In the setter, the *param should be a pointer to a previously initialized cudnnConvolutionDescriptor_t*. Describes the convolution operation. NULL
X_WDESC In the setter, the *param should be a pointer to a previously initialized cudnnFilterDescriptor_t*. Filter descriptor describing the size, layout and datatype of the w (filter) tensor. NULL
X_WDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether w (filter) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_YDESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the y (output) tensor. NULL
X_YDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether y (output) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_YSTATS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the sum of y and sum of y square tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE.

If set to NULL, the y statistics generation operation will become a NOP.

NULL
X_YSUM_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether sum of y pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, the y statistics generation operation will become a NOP.

CUDNN_PTR_NULL
X_YSQSUM_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether sum of y square pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, the y statistics generation operation will become a NOP.

CUDNN_PTR_NULL
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and need to be at least element-aligned or 16 bytes-aligned, respectively.
As of cuDNN 7.6.0, if the conditions in Table 22 are met, then the fully fused fast path will be triggered. Otherwise, a slower partially fused path will be triggered.
Table 22. Conditions for Fully Fused Fast Path (Forward)
Parameter Condition
Device compute capability Need to be one of 7.0, 7.2 or 7.5.
CUDNN_PARAM_XDESC

CUDNN_PARAM_XDATA_PLACEHOLDER

Tensor is 4 dimensional

Datatype is CUDNN_DATA_HALF

Layout is NHWC fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

Tensor’s C dimension is a multiple of 8.

CUDNN_PARAM_BN_EQSCALEBIAS_DESC

CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER

CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER

If either one of scale and bias operation is not turned into a NOP:

Tensor is 4 dimensional with shape 1xCx1x1

Datatype is CUDNN_DATA_HALF

Layout is fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

CUDNN_PARAM_CONV_DESC

CUDNN_PARAM_WDESC

CUDNN_PARAM_WDATA_PLACEHOLDER

Convolution descriptor’s mode needs to be CUDNN_CROSS_CORRELATION.

Convolution descriptor’s dataType needs to be CUDNN_DATA_FLOAT.

Convolution descriptor’s dilationA is (1,1).

Convolution descriptor’s group count needs to be 1.

Convolution descriptor’s mathType needs to be CUDNN_TENSOR_OP_MATH or CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION.

Filter is in NHWC layout

Filter’s data type is CUDNN_DATA_HALF

Filter’s K dimension is a multiple of 32

Filter size RxS is either 1x1 or 3x3

If filter size RxS is 1x1, convolution descriptor’s padA needs to be (0,0) and filterStrideA needs to be (1,1).

Filter’s alignment is CUDNN_PTR_16B_ALIGNED

CUDNN_PARAM_YDESC

CUDNN_PARAM_YDATA_PLACEHOLDER

Tensor is 4 dimensional

Datatype is CUDNN_DATA_HALF

Layout is NHWC fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

CUDNN_PARAM_YSTATS_DESC

CUDNN_PARAM_YSUM_PLACEHOLDER

CUDNN_PARAM_YSQSUM_PLACEHOLDER

If the generate statistics operation is not turned into a NOP:

Tensor is 4 dimensional with shape 1xKx1x1

Datatype is CUDNN_DATA_FLOAT

Layout is fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

Table 23. CUDNN_FUSED_SCALE_BIAS_ACTIVATION_WGRAD
For the attribute CUDNN_FUSED_SCALE_BIAS_ACTIVATION_WGRAD in cudnnFusedOp_t
Attribute Expected Descriptor Type Passed in, in the Setter Description Default Value After Creation
X_XDESC In the setter, the *param should be xDesc, a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout and datatype of the x (input) tensor NULL
X_XDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether xData pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_MODE In the setter, the *param should be a pointer to a previously initialized cudnnBatchNormMode_t*. Describes the mode of operation for the scale, bias and the statistics.

As of cuDNN 7.6.0, only CUDNN_BATCHNORM_SPATIAL and CUDNN_BATCHNORM_SPATIAL_PERSISTENT are supported, meaning, scale, bias, and statistics are all per-channel.

CUDNN_BATCHNORM_PER_ACTIVATION
X_BN_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout and datatype of the batchNorm equivalent scale and bias tensors. The shapes must match the mode specified in CUDNN_PARAM_BN_MODE. If set to NULL, both scale and bias operation will become a NOP. NULL
X_BN_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the scale operation becomes a NOP.

CUDNN_PTR_NULL
X_BN_EQBIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the bias operation becomes a NOP.

CUDNN_PTR_NULL
X_ACTIVATION_DESC In the setter, the *param should be a pointer to a previously initialized cudnnActivationDescriptor_t*. Describes the activation operation.

As of cuDNN 7.6.0, only activation mode of CUDNN_ACTIVATION_RELU and CUDNN_ACTIVATION_IDENTITY is supported. If set to NULL or if the activation mode is set to CUDNN_ACTIVATION_IDENTITY, then the activation in the op sequence becomes a NOP.

NULL
X_CONV_DESC In the setter, the *param should be a pointer to a previously initialized cudnnConvolutionDescriptor_t*. Describes the convolution operation. NULL
X_DWDESC In the setter, the *param should be a pointer to a previously initialized cudnnFilterDescriptor_t*. Filter descriptor describing the size, layout and datatype of the dw (filter gradient output) tensor. NULL
X_DWDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether dw (filter gradient output) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_DYDESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the dy (gradient input) tensor. NULL
X_DYDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether dy (gradient input) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.
As of cuDNN 7.6.0, if the conditions in Table 24 are met, then the fully fused fast path will be triggered. Otherwise a slower partially fused path will be triggered.
Table 24. Conditions for Fully Fused Fast Path (Backward)
Parameter Condition
Device compute capability Needs to be one of 7.0, 7.2 or 7.5.
CUDNN_PARAM_XDESC

CUDNN_PARAM_XDATA_PLACEHOLDER

Tensor is 4 dimensional

Datatype is CUDNN_DATA_HALF

Layout is NHWC fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

Tensor’s C dimension is a multiple of 8.

CUDNN_PARAM_BN_EQSCALEBIAS_DESC

CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER

CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER

If either one of scale and bias operation is not turned into a NOP:

Tensor is 4 dimensional with shape 1xCx1x1

Datatype is CUDNN_DATA_HALF

Layout is fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

CUDNN_PARAM_CONV_DESC

CUDNN_PARAM_DWDESC

CUDNN_PARAM_DWDATA_PLACEHOLDER

Convolution descriptor’s mode needs to be CUDNN_CROSS_CORRELATION.

Convolution descriptor’s dataType needs to be CUDNN_DATA_FLOAT.

Convolution descriptor’s dilationA is (1,1)

Convolution descriptor’s group count needs to be 1.

Convolution descriptor’s mathType needs to be CUDNN_TENSOR_OP_MATH or CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION.

Filter gradient is in NHWC layout

Filter gradient’s data type is CUDNN_DATA_HALF

Filter gradient’s K dimension is a multiple of 32.

Filter gradient size RxS is either 1x1 or 3x3

If filter gradient size RxS is 1x1, convolution descriptor’s padA needs to be (0,0) and filterStrideA needs to be (1,1).

Filter gradient’s alignment is CUDNN_PTR_16B_ALIGNED

CUDNN_PARAM_DYDESC

CUDNN_PARAM_DYDATA_PLACEHOLDER

Tensor is 4 dimensional

Datatype is CUDNN_DATA_HALF

Layout is NHWC fully packed

Alignment is CUDNN_PTR_16B_ALIGNED

Table 25. CUDNN_FUSED_BN_FINALIZE_STATISTICS_TRAINING
For the attribute CUDNN_FUSED_BN_FINALIZE_STATISTICS_TRAINING in cudnnFusedOp_t
Attribute Expected Descriptor Type Passed in, in the Setter Description Default Value After Creation
X_BN_MODE In the setter, the *param should be a pointer to a previously initialized cudnnBatchNormMode_t*. Describes the mode of operation for the scale, bias and the statistics.

As of cuDNN 7.6.0, only CUDNN_BATCHNORM_SPATIAL and CUDNN_BATCHNORM_SPATIAL_PERSISTENT are supported, meaning, scale, bias and statistics are all per-channel.

CUDNN_BATCHNORM_PER_ACTIVATION
X_YSTATS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the sum of y and sum of y square tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE. NULL
X_YSUM_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether sum of y pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_YSQSUM_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether sum of y square pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_SCALEBIAS_MEANVAR_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t. A common tensor descriptor describing the size, layout and datatype of the batchNorm trained scale, bias and statistics tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE (similar to the bnScaleBiasMeanVarDesc field in the cudnnBatchNormalization* API). NULL
X_BN_SCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm trained scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If the output of BN_EQSCALE is not needed, then this is not needed and may be NULL.

CUDNN_PTR_NULL
X_BN_BIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm trained bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If neither output of BN_EQSCALE or BN_EQBIAS is needed, then this is not needed and may be NULL.

CUDNN_PTR_NULL
X_BN_SAVED_MEAN_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm saved mean pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_SAVED_INVSTD_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm saved inverse standard deviation pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_RUNNING_MEAN_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm running mean pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_RUNNING_VAR_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm running variance pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the batchNorm equivalent scale and bias tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE.

If neither output of BN_EQSCALE or BN_EQBIAS is needed, then this is not needed and may be NULL.

NULL
X_BN_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_EQBIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
Table 26. CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE
For the attribute CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE in cudnnFusedOp_t
Attribute Expected Descriptor Type Passed in, in the Setter Description Default Value After Creation
X_BN_MODE In the setter, the *param should be a pointer to a previously initialized cudnnBatchNormMode_t*. Describes the mode of operation for the scale, bias and the statistics.

As of cuDNN 7.6.0, only CUDNN_BATCHNORM_SPATIAL and CUDNN_BATCHNORM_SPATIAL_PERSISTENT are supported, meaning, scale, bias and statistics are all per-channel.

CUDNN_BATCHNORM_PER_ACTIVATION
X_BN_SCALEBIAS_MEANVAR_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t. A common tensor descriptor describing the size, layout and datatype of the batchNorm trained scale, bias and statistics tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE (similar to the bnScaleBiasMeanVarDesc field in the cudnnBatchNormalization* API). NULL
X_BN_SCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm trained scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_BIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm trained bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_RUNNING_MEAN_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm running mean pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_RUNNING_VAR_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether the batchNorm running variance pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the batchNorm equivalent scale and bias tensors. The shapes need to match the mode specified in CUDNN_PARAM_BN_MODE. NULL
X_BN_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent scale pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
X_BN_EQBIAS_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent bias pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the computation for this output becomes a NOP.

CUDNN_PTR_NULL
Table 27. CUDNN_FUSED_CONVOLUTION_SCALE_BIAS_ADD_RELU
For the attribute CUDNN_FUSED_CONVOLUTION_SCALE_BIAS_ADD_RELU in cudnnFusedOp_t

This operation performs the following computation, where * denotes convolution operator: y=1(w*x)+2 z+b

Attribute Expected Descriptor Type Passed in, in the Setter Description Default Value After Creation
X_XDESC In the setter, the *param should be xDesc, a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout and datatype of the x (input) tensor. NULL
X_XDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether xData pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_CONV_DESC In the setter, the *param should be a pointer to a previously initialized cudnnConvolutionDescriptor_t*. Describes the convolution operation. NULL
X_WDESC In the setter, the *param should be a pointer to a previously initialized cudnnFilterDescriptor_t*. Filter descriptor describing the size, layout and datatype of the w (filter) tensor. NULL
X_WDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether w (filter) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL
X_BN_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the α1 scale and bias tensors. The tensor should have shape (1,K,1,1), K is the number of output features. NULL
X_BN_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm equivalent scale or α1 tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then α1 scaling becomes an NOP.

CUDNN_PTR_NULL
X_ZDESC In the setter, the *param should be xDesc, a pointer to a previously initialized cudnnTensorDescriptor_t. Tensor descriptor describing the size, layout and datatype of the z tensor.

If unset, then z scale-add term becomes a NOP.

NULL
CUDNN_PARAM_ZDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether z tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then z scale-add term becomes a NOP.

CUDNN_PTR_NULL
CUDNN_PARAM_BN_Z_EQSCALEBIAS_DESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the α2 tensor.

If set to NULL then scaling for input z becomes a NOP.

NULLPTR
CUDNN_PARAM_BN_Z_EQSCALE_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether batchnorm z-equivalent scaling pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *.

If set to CUDNN_PTR_NULL, then the scaling for input z becomes a NOP.

CUDNN_PTR_NULL
X_ACTIVATION_DESC In the setter, the *param should be a pointer to a previously initialized cudnnActivationDescriptor_t*. Describes the activation operation.

As of 7.6.0, only activation modes of CUDNN_ACTIVATION_RELU and CUDNN_ACTIVATION_IDENTITY are supported. If set to NULL or if the activation mode is set to CUDNN_ACTIVATION_IDENTITY, then the activation in the op sequence becomes a NOP.

NULL
X_YDESC In the setter, the *param should be a pointer to a previously initialized cudnnTensorDescriptor_t*. Tensor descriptor describing the size, layout and datatype of the y (output) tensor. NULL
X_YDATA_PLACEHOLDER In the setter, the *param should be a pointer to a previously initialized X_PointerPlaceHolder_t*. Describes whether y (output) tensor pointer in the VariantParamPack will be NULL, or if not, user promised pointer alignment *. CUDNN_PTR_NULL

6.1.3.3. cudnnFusedOpsPointerPlaceHolder_t

cudnnFusedOpsPointerPlaceHolder_t is an enumerated type used to select the alignment type of the cudnnFusedOps descriptor pointer.
Member Description
CUDNN_PTR_NULL = 0 Indicates that the pointer to the tensor in the variantPack will be NULL.
CUDNN_PTR_ELEM_ALIGNED = 1 Indicates that the pointer to the tensor in the variantPack will not be NULL, and will have element alignment.
CUDNN_PTR_16B_ALIGNED = 2 Indicates that the pointer to the tensor in the variantPack will not be NULL, and will have 16 byte alignment.

6.1.3.4. cudnnFusedOpsVariantParamLabel_t

The cudnnFusedOpsVariantParamLabel_t is an enumerated type that is used to set the buffer pointers. These buffer pointers can be changed in each iteration.

typedef enum {
	CUDNN_PTR_XDATA                              = 0,
	CUDNN_PTR_BN_EQSCALE                         = 1,
	CUDNN_PTR_BN_EQBIAS                          = 2,
	CUDNN_PTR_WDATA                              = 3,
	CUDNN_PTR_DWDATA                             = 4,
	CUDNN_PTR_YDATA                              = 5,
	CUDNN_PTR_DYDATA                             = 6,
	CUDNN_PTR_YSUM                               = 7,
	CUDNN_PTR_YSQSUM                             = 8,
	CUDNN_PTR_WORKSPACE                          = 9,
	CUDNN_PTR_BN_SCALE                           = 10,
	CUDNN_PTR_BN_BIAS                            = 11,
	CUDNN_PTR_BN_SAVED_MEAN                      = 12,
	CUDNN_PTR_BN_SAVED_INVSTD                    = 13,
	CUDNN_PTR_BN_RUNNING_MEAN                    = 14,
	CUDNN_PTR_BN_RUNNING_VAR                     = 15,
	CUDNN_PTR_ZDATA                              = 16,
	CUDNN_PTR_BN_Z_EQSCALE                       = 17,
	CUDNN_PTR_BN_Z_EQBIAS                        = 18,
	CUDNN_PTR_ACTIVATION_BITMASK                 = 19,
	CUDNN_PTR_DXDATA                             = 20,
	CUDNN_PTR_DZDATA                             = 21,
	CUDNN_PTR_BN_DSCALE                          = 22,
	CUDNN_PTR_BN_DBIAS                           = 23,
	CUDNN_SCALAR_SIZE_T_WORKSPACE_SIZE_IN_BYTES  = 100,
	CUDNN_SCALAR_INT64_T_BN_ACCUMULATION_COUNT   = 101,
	CUDNN_SCALAR_DOUBLE_BN_EXP_AVG_FACTOR        = 102,
	CUDNN_SCALAR_DOUBLE_BN_EPSILON               = 103,
	} cudnnFusedOpsVariantParamLabel_t;
Table 28. Legend For Tables in This Section
Short-form used Stands for
Setter cudnnSetFusedOpsVariantParamPackAttribute()
Getter cudnnGetFusedOpsVariantParamPackAttribute()
X_ prefix in the Attribute key column Stands for CUDNN_PTR_ or CUDNN_SCALAR_ in the enumerator name.
Table 29. CUDNN_FUSED_SCALE_BIAS_ACTIVATION_CONV_BNSTATS
For the attribute CUDNN_FUSED_SCALE_BIAS_ACTIVATION_CONV_BNSTATS in cudnnFusedOp_t
Attribute key Expected Descriptor Type Passed in, in the Setter I/O Type Description Default Value
X_XDATA void * input Pointer to x (input) tensor on device, need to agree with previously set CUDNN_PARAM_XDATA_PLACEHOLDER attribute *. NULL
X_BN_EQSCALE void * input Pointer to batchnorm equivalent scale tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER attribute *. NULL
X_BN_EQBIAS void * input Pointer to batchnorm equivalent bias tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER attribute *. NULL
X_WDATA void * input Pointer to w (filter) tensor on device, need to agree with previously set CUDNN_PARAM_WDATA_PLACEHOLDER attribute *. NULL
X_YDATA void * output Pointer to y (output) tensor on device, need to agree with previously set CUDNN_PARAM_YDATA_PLACEHOLDER attribute *. NULL
X_YSUM void * output Pointer to sum of y tensor on device, need to agree with previously set CUDNN_PARAM_YSUM_PLACEHOLDER attribute *. NULL
X_YSQSUM void * output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_YSQSUM_PLACEHOLDER attribute *. NULL
X_WORKSPACE void * input Pointer to user allocated workspace on device. Can be NULL if the workspace size requested is 0. NULL
X_SIZE_T_WORKSPACE_SIZE_IN_BYTES size_t * input Pointer to a size_t value in host memory describing the user allocated workspace size in bytes. The amount needs to be equal or larger than the amount requested in cudnnMakeFusedOpsPlan. 0
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.
Table 30. CUDNN_FUSED_SCALE_BIAS_ACTIVATION_WGRAD
For the attribute CUDNN_FUSED_SCALE_BIAS_ACTIVATION_WGRAD in cudnnFusedOp_t
Attribute key Expected Descriptor Type Passed in, in the Setter I/O Type Description Default Value
X_XDATA void * input Pointer to x (input) tensor on device, need to agree with previously set CUDNN_PARAM_XDATA_PLACEHOLDER attribute *. NULL
X_BN_EQSCALE void * input Pointer to batchnorm equivalent scale tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER attribute *. NULL
X_BN_EQBIAS void * input Pointer to batchnorm equivalent bias tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER attribute *. NULL
X_DWDATA void * output Pointer to dw (filter gradient output) tensor on device, need to agree with previously set CUDNN_PARAM_WDATA_PLACEHOLDER attribute *. NULL
X_DYDATA void * input Pointer to dy (gradient input) tensor on device, need to agree with previously set CUDNN_PARAM_YDATA_PLACEHOLDER attribute *. NULL
X_WORKSPACE void * input Pointer to user allocated workspace on device. Can be NULL if the workspace size requested is 0. NULL
X_SIZE_T_WORKSPACE_SIZE_IN_BYTES size_t * input Pointer to a size_t value in host memory describing the user allocated workspace size in bytes. The amount needs to be equal or larger than the amount requested in cudnnMakeFusedOpsPlan. 0
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.
Table 31. CUDNN_FUSED_BN_FINALIZE_STATISTICS_TRAINING
For the attribute CUDNN_FUSED_BN_FINALIZE_STATISTICS_TRAINING in cudnnFusedOp_t
Attribute key Expected Descriptor Type Passed in, in the Setter I/O Type Description Default Value
X_YSUM void * input Pointer to sum of y tensor on device, need to agree with previously set CUDNN_PARAM_YSUM_PLACEHOLDER attribute *. NULL
X_YSQSUM void * input Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_YSQSUM_PLACEHOLDER attribute *. NULL
X_BN_SCALE void * input Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_SCALE_PLACEHOLDER attribute *. NULL
X_BN_BIAS void * input Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_BIAS_PLACEHOLDER attribute *. NULL
X_BN_SAVED_MEAN void * output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_SAVED_MEAN_PLACEHOLDER attribute *. NULL
X_BN_SAVED_INVSTD void * output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_SAVED_INVSTD_PLACEHOLDER attribute *. NULL
X_BN_RUNNING_MEAN void * input/output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_RUNNING_MEAN_PLACEHOLDER attribute *. NULL
X_BN_RUNNING_VAR void * input/output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_RUNNING_VAR_PLACEHOLDER attribute *. NULL
X_BN_EQSCALE void * output Pointer to batchnorm equivalent scale tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER attribute *. NULL
X_BN_EQBIAS void * output Pointer to batchnorm equivalent bias tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER attribute *. NULL
X_INT64_T_BN_ACCUMULATION_COUNT int64_t * input Pointer to a scalar value in int64_t on host memory.

This value should describe the number of tensor elements accumulated in the sum of y and sum of y square tensors.

For example, in the single GPU use case, if the mode is CUDNN_BATCHNORM_SPATIAL or CUDNN_BATCHNORM_SPATIAL_PERSISTENT, the value should be equal to N*H*W of the tensor from which the statistics are calculated.

In multi-GPU use case, if all-reduce has been performed on the sum of y and sum of y square tensors, this value should be the sum of the single GPU accumulation count on each of the GPUs.

0
X_DOUBLE_BN_EXP_AVG_FACTOR double * input Pointer to a scalar value in double on host memory.

Factor used in the moving average computation. See exponentialAverageFactor in cudnnBatchNormalization* APIs.

0.0
X_DOUBLE_BN_EPSILON double * input Pointer to a scalar value in double on host memory.

A conditioning constant used in the batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h.

See exponentialAverageFactor in cudnnBatchNormalization* APIs.

0.0
X_WORKSPACE void * input Pointer to user allocated workspace on device. Can be NULL if the workspace size requested is 0. NULL
X_SIZE_T_WORKSPACE_SIZE_IN_BYTES size_t * input Pointer to a size_t value in host memory describing the user allocated workspace size in bytes. The amount needs to be equal or larger than the amount requested in cudnnMakeFusedOpsPlan. 0
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.
Table 32. CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE
For the attribute CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE in cudnnFusedOp_t
Attribute key Expected Descriptor Type Passed in, in the Setter I/O Type Description Default Value
X_BN_SCALE void * input Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_SCALE_PLACEHOLDER attribute *. NULL
X_BN_BIAS void * input Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_BIAS_PLACEHOLDER attribute *. NULL
X_BN_RUNNING_MEAN void * input/output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_RUNNING_MEAN_PLACEHOLDER attribute *. NULL
X_BN_RUNNING_VAR void * input/output Pointer to sum of y square tensor on device, need to agree with previously set CUDNN_PARAM_BN_RUNNING_VAR_PLACEHOLDER attribute *. NULL
X_BN_EQSCALE void * output Pointer to batchnorm equivalent scale tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER attribute *. NULL
X_BN_EQBIAS void * output Pointer to batchnorm equivalent bias tensor on device, need to agree with previously set CUDNN_PARAM_BN_EQBIAS_PLACEHOLDER attribute *. NULL
X_DOUBLE_BN_EPSILON double * input Pointer to a scalar value in double on host memory.

A conditioning constant used in the batch normalization formula. Its value should be equal to or greater than the value defined for CUDNN_BN_MIN_EPSILON in cudnn.h.

See exponentialAverageFactor in cudnnBatchNormalization* APIs.

0.0
X_WORKSPACE void * input Pointer to user allocated workspace on device. Can be NULL if the workspace size requested is 0. NULL
X_SIZE_T_WORKSPACE_SIZE_IN_BYTES size_t * input Pointer to a size_t value in host memory describing the user allocated workspace size in bytes. The amount needs to be equal or larger than the amount requested in cudnnMakeFusedOpsPlan. 0
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.
Table 33. CUDNN_FUSED_SCALE_BIAS_ADD_RELU
For the attribute CUDNN_FUSED_BN_FINALIZE_STATISTICS_INFERENCE in cudnnFusedOp_t
Attribute key Expected Descriptor Type Passed in, in the Setter I/O Type Description Default Value
X_XDATA void * input Pointer to x (image) tensor on device, need to agree with previously set CUDNN_PARAM_XDATA_PLACEHOLDER attribute *. NULL
X_WDATA void * input Pointer to w (filter) tensor on device, need to agree with previously set CUDNN_PARAM_WDATA_PLACEHOLDER attribute *. NULL
X_BN_EQSCALE void * input Pointer to alpha1 or batchnorm equivalent scale tensor on device; need to agree with previously set CUDNN_PARAM_BN_EQSCALE_PLACEHOLDER attribute *. NULL
X_ZDATA void * input Pointer to z ( tensor on device; Need to agree with previously set CUDNN_PARAM_YDATA_PLACEHOLDER attribute *. NULL
X_BN_Z_EQSCALE void * input Pointer to alpha2, equivalent scale tensor for z; Need to agree with previously set CUDNN_PARAM_BN_Z_EQSCALE_PLACEHOLDER attribute *. NULL
X_BN_Z_EQBIAS void * input Pointer to batchnorm equivalent bias tensor on device, need to agree with previously set CUDNN_PARAM_BN_Z_EQBIAS_PLACEHOLDER attribute *. NULL
X_YDATA void * output Pointer to y (output) tensor on device, need to agree with previously set CUDNN_PARAM_YDATA_PLACEHOLDER attribute *. NULL
X_WORKSPACE void * input Pointer to user allocated workspace on device. Can be NULL if the workspace size requested is 0. NULL
X_SIZE_T_WORKSPACE_SIZE_IN_BYTES size_t * input Pointer to a size_t value in host memory describing the user allocated workspace size in bytes. The amount needs to be equal or larger than the amount requested in cudnnMakeFusedOpsPlan. 0
Note:
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_NULL, then the device pointer in the VariantParamPack needs to be NULL as well.
  • If the corresponding pointer placeholder in ConstParamPack is set to CUDNN_PTR_ELEM_ALIGNED or CUDNN_PTR_16B_ALIGNED, then the device pointer in the VariantParamPack may not be NULL and needs to be at least element-aligned or 16 bytes-aligned, respectively.

6.2. API Functions

6.2.1. cudnnConvolutionBackwardBias()

cudnnStatus_t cudnnConvolutionBackwardBias(
    cudnnHandle_t                    handle,
    const void                      *alpha,
    const cudnnTensorDescriptor_t    dyDesc,
    const void                      *dy,
    const void                      *beta,
    const cudnnTensorDescriptor_t    dbDesc,
    void                            *db)

This function computes the convolution function gradient with respect to the bias, which is the sum of every element belonging to the same feature map across all of the images of the input tensor. Therefore, the number of elements produced is equal to the number of features maps of the input tensor.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*resultValue + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

dyDesc

Input. Handle to the previously initialized input tensor descriptor. For more information, see cudnnTensorDescriptor_t.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyDesc.

dbDesc

Input. Handle to the previously initialized output tensor descriptor.

db

Output. Data pointer to GPU memory associated with the output tensor descriptor dbDesc.

Returns

CUDNN_STATUS_SUCCESS

The operation was launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters n, height, width of the output tensor is not 1.
  • The numbers of feature maps of the input tensor and output tensor differ.
  • The dataType of the two tensor descriptors is different.

6.2.2. cudnnConvolutionBackwardFilter()

cudnnStatus_t cudnnConvolutionBackwardFilter(
    cudnnHandle_t                       handle,
    const void                         *alpha,
    const cudnnTensorDescriptor_t       xDesc,
    const void                         *x,
    const cudnnTensorDescriptor_t       dyDesc,
    const void                         *dy,
    const cudnnConvolutionDescriptor_t  convDesc,
    cudnnConvolutionBwdFilterAlgo_t     algo,
    void                               *workSpace,
    size_t                              workSpaceSizeInBytes,
    const void                         *beta,
    const cudnnFilterDescriptor_t       dwDesc,
    void                               *dw)

This function computes the convolution weight (filter) gradient of the tensor dy, where y is the output of the forward convolution in cudnnConvolutionForward(). It uses the specified algo, and returns the results in the output tensor dw. Scaling factors alpha and beta can be used to scale the computed result or accumulate with the current dw.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

alpha, beta
Input. Pointers to scaling factors (in host memory) used to blend the computation result with prior value in the output layer as follows:
dstValue = alpha[0]*result + beta[0]*priorDstValue

For more information, see Scaling Parameters in the cuDNN Developer Guide.

xDesc

Input. Handle to a previously initialized tensor descriptor. For more information, see cudnnTensorDescriptor_t.

x

Input. Data pointer to GPU memory associated with the tensor descriptor xDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the backpropagation gradient tensor descriptor dyDesc.

convDesc

Input. Previously initialized convolution descriptor. For more information, see cudnnConvolutionDescriptor_t.

algo

Input. Enumerant that specifies which convolution algorithm should be used to compute the results. For more information, see cudnnConvolutionBwdFilterAlgo_t.

workSpace

Input. Data pointer to GPU memory to a workspace needed to be able to execute the specified algorithm. If no workspace is needed for a particular algorithm, that pointer can be nil.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

dwDesc

Input. Handle to a previously initialized filter gradient descriptor. For more information, see cudnnFilterDescriptor_t.

dw

Input/Output. Data pointer to GPU memory associated with the filter gradient descriptor dwDesc that carries the result.

Supported configurations

This function supports the following combinations of data types for xDesc, dyDesc, convDesc, and dwDesc.

Data Type Configurations xDesc, dyDesc, and dwDesc Data Type convDesc Data Type
TRUE_HALF_CONFIG (only supported on architectures with true FP16 support, meaning, compute capability 5.3 and later) CUDNN_DATA_HALF CUDNN_DATA_HALF
PSEUDO_HALF_CONFIG CUDNN_DATA_HALF CUDNN_DATA_FLOAT
FLOAT_CONFIG CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT
DOUBLE_CONFIG CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE

Supported algorithms

Note: Specifying a separate algorithm can cause changes in performance, support and computation determinism. See the following table for an exhaustive list of algorithm options and their respective supported parameters and deterministic behavior.

The table below shows the list of the supported 2D and 3D convolutions. The 2D convolutions are described first, followed by the 3D convolutions.

For the following terms, the short-form versions shown in the parentheses are used in the table below, for brevity:
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_0 (_ALGO_0)
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_1 (_ALGO_1)
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_3 (_ALGO_3)
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_FFT (_FFT)
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_FFT_TILING (_FFT_TILING)
  • CUDNN_CONVOLUTION_BWD_FILTER_ALGO_WINOGRAD_NONFUSED (_WINOGRAD_NONFUSED)
  • CUDNN_TENSOR_NCHW (_NCHW)
  • CUDNN_TENSOR_NHWC (_NHWC)
  • CUDNN_TENSOR_NCHW_VECT_C (_NCHW_VECT_C)
Table 34. For 2D convolutions: dwDesc: _NHWC
Filter descriptor dwDesc: _NHWC (see cudnnTensorFormat_t)
Algo Name Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important
_ALGO_0 and _ALGO_1   NHWC HWC-packed NHWC HWC-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

 
Table 35. For 2D convolutions: Desc: _NCHW
Filter descriptor Desc: _NCHW
Algo Name Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important
_ALGO_0 No All except _NCHW_VECT_C NCHW CHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_ALGO_1 Yes All except _NCHW_VECT_C NCHW CHW-packed TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: Greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_FFT Yes NCHW CHW-packed NCHW CHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

xDesc feature map height + 2 * convDesc zero-padding height must equal 256 or less

xDesc feature map width + 2 * convDesc zero-padding width must equal 256 or less

convDesc vertical and horizontal filter stride must equal 1

dwDesc filter height must be greater than convDesc zero-padding height

dwDesc filter width must be greater than convDesc zero-padding width

_ALGO_3 Yes All except _NCHW_VECT_C NCHW CHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

_WINOGRAD_NONFUSED Yes All except _NCHW_VECT_C NCHW CHW-packed TRUE_HALF_CONFIG

PSEUDO_HALF_CONFIG

FLOAT_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

convDesc vertical and horizontal filter stride must equal 1

wDesc filter (height, width) must be (3,3) or (5,5)

If wDesc filter (height, width) is (5,5), then the data type config TRUE_HALF_CONFIG is not supported.

_FFT_TILING Yes NCHW CHW-packed NCHW CHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: 1 for all dimensions

convDesc Group Count Support: Greater than 0

xDesc width or height must equal 1

dyDesc width or height must equal 1 (the same dimension as in xDesc). The other dimension must be less than or equal to 256, meaning, the largest 1D tile size currently supported.

convDesc vertical and horizontal filter stride must equal 1

dwDesc filter height must be greater than convDesc zero-padding height

dwDesc filter width must be greater than convDesc zero-padding width

Table 36. For 3D convolutions: wDesc: _NCHW
Filter descriptor wDesc: _NCHW.
Algo Name (3D Convolutions) Deterministic (Yes or No) Tensor Formats Supported for dyDesc Tensor Formats Supported for dxDesc Data Type Configurations Supported Important
_ALGO_0 No All except _NCDHW_VECT_C. NCDHW CDHW-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

_ALGO_3 No NCDHW fully-packed NCDHW fully-packed PSEUDO_HALF_CONFIG

FLOAT_CONFIG

DOUBLE_CONFIG

Dilation: greater than 0 for all dimensions

convDesc Group Count Support: Greater than 0

Returns

CUDNN_STATUS_SUCCESS

The operation was launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • At least one of the following is NULL: handle, xDesc, dyDesc, convDesc, dwDesc, xData, dyData, dwData, alpha, beta
  • xDesc and dyDesc have a non-matching number of dimensions
  • xDesc and dwDesc have a non-matching number of dimensions
  • xDesc has fewer than three number of dimensions
  • xDesc, dyDesc, and dwDesc have a non-matching data type.
  • xDesc and dwDesc have a non-matching number of input feature maps per image (or group in case of grouped convolutions).
  • yDesc or wDesc indicate an output channel count that isn't a multiple of group count (if group count has been set in convDesc).
CUDNN_STATUS_NOT_SUPPORTED
At least one of the following conditions are met:
  • xDesc or dyDesc have negative tensor striding
  • xDesc, dyDesc or dwDesc has a number of dimensions that is not 4 or 5
  • The chosen algo does not support the parameters provided; see above for exhaustive list of parameter support for each algo
CUDNN_STATUS_MAPPING_ERROR

An error occurs during the texture binding of the filter data.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

6.2.3. cudnnCreateFusedOpsConstParamPack()

cudnnStatus_t cudnnCreateFusedOpsConstParamPack(
	cudnnFusedOpsConstParamPack_t *constPack, 
	cudnnFusedOps_t ops);		

This function creates an opaque structure to store the various problem size information, such as the shape, layout and the type of tensors, and the descriptors for convolution and activation, for the selected sequence of cudnnFusedOps computations.

Parameters

constPack
Input. The opaque structure that is created by this function. For more information, see cudnnFusedOpsConstParamPack_t.
ops
Input. The specific sequence of computations to perform in the cudnnFusedOps computations, as defined in the enumerant type cudnnFusedOps_t.

Returns

CUDNN_STATUS_BAD_PARAM
If either constPack or ops is NULL.
CUDNN_STATUS_SUCCESS
If the descriptor is created successfully.
CUDNN_STATUS_NOT_SUPPORTED
If the ops enum value is not supported or reserved for future use.

6.2.4. cudnnCreateFusedOpsPlan()

cudnnStatus_t cudnnCreateFusedOpsPlan(
	cudnnFusedOpsPlan_t *plan, 
	cudnnFusedOps_t ops);		

This function creates the plan descriptor for the cudnnFusedOps computation. This descriptor contains the plan information, including the problem type and size, which kernels should be run, and the internal workspace partition.

Parameters

plan
Input. A pointer to the instance of the descriptor created by this function.
ops
Input. The specific sequence of fused operations computations for which this plan descriptor should be created. For more information, see cudnnFusedOps_t.

Returns

CUDNN_STATUS_BAD_PARAM
If either the input *plan is NULL or the ops input is not a valid cudnnFusedOp enum.
CUDNN_STATUS_NOT_SUPPORTED
The ops input provided is not supported.
CUDNN_STATUS_SUCCESS
The plan descriptor is created successfully.

6.2.5. cudnnCreateFusedOpsVariantParamPack()

cudnnStatus_t cudnnCreateFusedOpsVariantParamPack(
	cudnnFusedOpsVariantParamPack_t *varPack, 
	cudnnFusedOps_t ops);		

This function creates a descriptor for cudnnFusedOps constant parameters.

Parameters

varPack
Input. Pointer to the descriptor created by this function. For more information, see cudnnFusedOpsVariantParamPack_t.
ops
Input. The specific sequence of fused operations computations for which this descriptor should be created.

Returns

CUDNN_STATUS_SUCCESS
The descriptor is successfully created.
CUDNN_STATUS_BAD_PARAM
If any input is invalid.

6.2.6. cudnnDestroyFusedOpsConstParamPack()

cudnnStatus_t cudnnDestroyFusedOpsConstParamPack(
	cudnnFusedOpsConstParamPack_t constPack);		

This function destroys a previously-created cudnnFusedOpsConstParamPack_t structure.

Parameters

constPack
Input. The cudnnFusedOpsConstParamPack_t structure that should be destroyed.

Returns

CUDNN_STATUS_SUCCESS
If the descriptor is destroyed successfully.
CUDNN_STATUS_INTERNAL_ERROR
If the ops enum value is not supported or invalid.

6.2.7. cudnnDestroyFusedOpsPlan()

cudnnStatus_t cudnnDestroyFusedOpsPlan(
	cudnnFusedOpsPlan_t plan);		

This function destroys the plan descriptor provided.

Parameters

plan
Input. The descriptor that should be destroyed by this function.

Returns

CUDNN_STATUS_SUCCESS
If either the plan descriptor is NULL or the descriptor is successfully destroyed.

6.2.8. cudnnDestroyFusedOpsVariantParamPack()

cudnnStatus_t cudnnDestroyFusedOpsVariantParamPack(
	cudnnFusedOpsVariantParamPack_t varPack);		

This function destroys a previously-created descriptor for cudnnFusedOps constant parameters.

Parameters

varPack
Input. The descriptor that should be destroyed.

Returns

CUDNN_STATUS_SUCCESS
The descriptor is successfully destroyed.

6.2.9. cudnnFindConvolutionBackwardFilterAlgorithm()

cudnnStatus_t cudnnFindConvolutionBackwardFilterAlgorithm(
cudnnHandle_t                          handle,
const cudnnTensorDescriptor_t          xDesc,
const cudnnTensorDescriptor_t          dyDesc,
const cudnnConvolutionDescriptor_t     convDesc,
const cudnnFilterDescriptor_t          dwDesc,
const int                              requestedAlgoCount,
int                                   *returnedAlgoCount,
cudnnConvolutionBwdFilterAlgoPerf_t   *perfResults)

This function attempts all algorithms available for cudnnConvolutionBackwardFilter(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionBwdFilterAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionBackwardFilterAlgorithmMaxCount().

Note:
  • This function is host blocking.
  • It is recommended to run this function prior to allocating layer data; doing otherwise may needlessly inhibit some algorithm options due to resource usage.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dwDesc

Input. Handle to a previously initialized filter descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • xDesc, dyDesc, or dwDesc are not allocated properly.
  • xDesc, dyDesc, or dwDesc has fewer than 1 dimension.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_ALLOC_FAILED

This function was unable to allocate memory to store sample input, filters and output.

CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

6.2.10. cudnnFindConvolutionBackwardFilterAlgorithmEx()

cudnnStatus_t cudnnFindConvolutionBackwardFilterAlgorithmEx(
    cudnnHandle_t                          handle,
    const cudnnTensorDescriptor_t          xDesc,
    const void                            *x,
    const cudnnTensorDescriptor_t          dyDesc,
    const void                            *dy,
    const cudnnConvolutionDescriptor_t     convDesc,
    const cudnnFilterDescriptor_t          dwDesc,
    void                                  *dw,
    const int                              requestedAlgoCount,
    int                                   *returnedAlgoCount,
    cudnnConvolutionBwdFilterAlgoPerf_t   *perfResults,
    void                                  *workSpace,
    size_t                                 workSpaceSizeInBytes)

This function attempts all algorithms available for cudnnConvolutionBackwardFilter(). It will attempt both the provided convDescmathType and CUDNN_DEFAULT_MATH (assuming the two differ).

Note: Algorithms without the CUDNN_TENSOR_OP_MATH availability will only be tried with CUDNN_DEFAULT_MATH, and returned as such.

Memory is allocated via cudaMalloc(). The performance metrics are returned in the user-allocated array of cudnnConvolutionBwdFilterAlgoPerf_t. These metrics are written in a sorted fashion where the first element has the lowest compute time. The total number of resulting algorithms can be queried through the API cudnnGetConvolutionBackwardFilterAlgorithmMaxCount().

Note: This function is host blocking.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

x

Input. Data pointer to GPU memory associated with the filter descriptor xDesc.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

dy

Input. Data pointer to GPU memory associated with the tensor descriptor dyDesc.

convDesc

Input. Previously initialized convolution descriptor.

dwDesc

Input. Handle to a previously initialized filter descriptor.

dw

Input/Output. Data pointer to GPU memory associated with the filter descriptor dwDesc. The content of this tensor will be overwritten with arbitrary values.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workSpace

Input. Data pointer to GPU memory is a necessary workspace for some algorithms. The size of this workspace will determine the availability of algorithms. A nil pointer is considered a workSpace of 0 bytes.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workSpace.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • handle is not allocated properly.
  • xDesc, dyDesc, or dwDesc are not allocated properly.
  • xDesc, dyDesc, or dwDesc has fewer than 1 dimension.
  • x, dy, or dw is nil.
  • Either returnedCount or perfResults is nil.
  • requestedCount is less than 1.
CUDNN_STATUS_INTERNAL_ERROR

At least one of the following conditions are met:

  • The function was unable to allocate necessary timing objects.
  • The function was unable to deallocate necessary timing objects.
  • The function was unable to deallocate sample input, filters and output.

6.2.11. cudnnFusedOpsExecute()

cudnnStatus_t cudnnFusedOpsExecute(
	cudnnHandle_t handle, 
	const cudnnFusedOpsPlan_t plan, 
	cudnnFusedOpsVariantParamPack_t varPack);		

This function executes the sequence of cudnnFusedOps operations.

Parameters

handle
Input. Pointer to the cuDNN library context.
plan
Input. Pointer to a previously-created and initialized plan descriptor.
varPack
Input. Pointer to the descriptor to the variant parameters pack.

Returns

CUDNN_STATUS_BAD_PARAM
If the type of cudnnFusedOps_t in the plan descriptor is unsupported.

6.2.12. cudnnGetConvolutionBackwardFilterAlgorithmMaxCount()

cudnnStatus_t cudnnGetConvolutionBackwardFilterAlgorithmMaxCount(
    cudnnHandle_t       handle,
    int                 *count)

This function returns the maximum number of algorithms which can be returned from cudnnFindConvolutionBackwardFilterAlgorithm() and cudnnGetConvolutionForwardAlgorithm_v7(). This is the sum of all algorithms plus the sum of all algorithms with Tensor Core operations supported for the current device.

Parameters

handle

Input. Handle to a previously created cuDNN context.

count

Output. The resulting maximum count of algorithms.

Returns

CUDNN_STATUS_SUCCESS

The function was successful.

CUDNN_STATUS_BAD_PARAM

The provided handle is not allocated properly.

6.2.13. cudnnGetConvolutionBackwardFilterAlgorithm_v7()

cudnnStatus_t cudnnGetConvolutionBackwardFilterAlgorithm_v7(
    cudnnHandle_t                          handle,
    const cudnnTensorDescriptor_t          xDesc,
    const cudnnTensorDescriptor_t          dyDesc,
    const cudnnConvolutionDescriptor_t     convDesc,
    const cudnnFilterDescriptor_t          dwDesc,
    const int                              requestedAlgoCount,
    int                                   *returnedAlgoCount,
    cudnnConvolutionBwdFilterAlgoPerf_t   *perfResults)

This function serves as a heuristic for obtaining the best suited algorithm for cudnnConvolutionBackwardFilter() for the given layer specifications. This function will return all algorithms (including CUDNN_TENSOR_OP_MATH and CUDNN_DEFAULT_MATH versions of algorithms where CUDNN_TENSOR_OP_MATH may be available) sorted by expected (based on internal heuristic) relative performance with fastest being index 0 of perfResults. For an exhaustive search for the fastest algorithm, use cudnnFindConvolutionBackwardFilterAlgorithm(). The total number of resulting algorithms can be queried through the returnedAlgoCount variable.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dwDesc

Input. Handle to a previously initialized filter descriptor.

requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • One of the parameters handle, xDesc, dyDesc, convDesc, dwDesc, perfResults, returnedAlgoCount is NULL.
  • The numbers of feature maps of the input tensor and output tensor differ.
  • The dataType of the two tensor descriptors or the filter are different.
  • requestedAlgoCount is less than or equal to 0.

6.2.14. cudnnGetConvolutionBackwardFilterWorkspaceSize()

cudnnStatus_t cudnnGetConvolutionBackwardFilterWorkspaceSize(
    cudnnHandle_t                       handle,
    const cudnnTensorDescriptor_t       xDesc,
    const cudnnTensorDescriptor_t       dyDesc,
    const cudnnConvolutionDescriptor_t  convDesc,
    const cudnnFilterDescriptor_t       dwDesc,
    cudnnConvolutionBwdFilterAlgo_t     algo,
    size_t                             *sizeInBytes)

This function returns the amount of GPU memory workspace the user needs to allocate to be able to call cudnnConvolutionBackwardFilter() with the specified algorithm. The workspace allocated will then be passed to the routine cudnnConvolutionBackwardFilter(). The specified algorithm can be the result of the call to cudnnGetConvolutionBackwardFilterAlgorithm_v7() or can be chosen arbitrarily by the user. Note that not every algorithm is available for every configuration of the input tensor and/or every configuration of the convolution descriptor.

Parameters

handle

Input. Handle to a previously created cuDNN context.

xDesc

Input. Handle to the previously initialized input tensor descriptor.

dyDesc

Input. Handle to the previously initialized input differential tensor descriptor.

convDesc

Input. Previously initialized convolution descriptor.

dwDesc

Input. Handle to a previously initialized filter descriptor.

algo

Input. Enumerant that specifies the chosen convolution algorithm.

sizeInBytes

Output. Amount of GPU memory needed as workspace to be able to execute a forward convolution with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The numbers of feature maps of the input tensor and output tensor differ.
  • The dataType of the two tensor descriptors or the filter are different.
CUDNN_STATUS_NOT_SUPPORTED

The combination of the tensor descriptors, filter descriptor and convolution descriptor is not supported for the specified algorithm.

6.2.15. cudnnGetFusedOpsConstParamPackAttribute()

cudnnStatus_t cudnnGetFusedOpsConstParamPackAttribute(
	const cudnnFusedOpsConstParamPack_t constPack,
	cudnnFusedOpsConstParamLabel_t paramLabel,
	void *param,
	int *isNULL);		

This function retrieves the values of the descriptor pointed to by the param pointer input. The type of the descriptor is indicated by the enum value of paramLabel input.

Parameters

constPack
Input. The opaque cudnnFusedOpsConstParamPack_t structure that contains the various problem size information, such as the shape, layout and the type of tensors, and the descriptors for convolution and activation, for the selected sequence of cudnnFusedOps_t computations.
paramLabel
Input. Several types of descriptors can be retrieved by this getter function. The param input points to the descriptor itself, and this input indicates the type of the descriptor pointed to by the param input. The cudnnFusedOpsConstParamLabel_t enumerant type enables the selection of the type of the descriptor. Refer to the param description below.
param
Input. Data pointer to the host memory associated with the descriptor that should be retrieved. The type of this descriptor depends on the value of paramLabel. For the given paramLabel, if the associated value inside the constPack is set to NULL or by default NULL, then cuDNN will copy the value or the opaque structure in the constPack to the host memory buffer pointed to by param. For more information, see the table in cudnnFusedOpsConstParamLabel_t.
isNULL
Input/Output. Users must pass a pointer to an integer in the host memory in this field. If the value in the constPack associated with the given paramLabel is by default NULL or previously set by the user to NULL, then cuDNN will write a non-zero value to the location pointed by isNULL.

Returns

CUDNN_STATUS_SUCCESS
The descriptor values are retrieved successfully.
CUDNN_STATUS_BAD_PARAM
If either constPack, param or isNULL is NULL; or if paramLabel is invalid.

6.2.16. cudnnGetFusedOpsVariantParamPackAttribute()

cudnnStatus_t cudnnGetFusedOpsVariantParamPackAttribute(
	const cudnnFusedOpsVariantParamPack_t varPack,
	cudnnFusedOpsVariantParamLabel_t paramLabel,
	void *ptr);		

This function retrieves the settings of the variable parameter pack descriptor.

Parameters

varPack
Input. Pointer to the cudnnFusedOps variant parameter pack (varPack) descriptor.
paramLabel
Input. Type of the buffer pointer parameter (in the varPack descriptor). For more information, see cudnnFusedOpsConstParamLabel_t. The retrieved descriptor values vary according to this type.
ptr
Output. Pointer to the host or device memory where the retrieved value is written by this function. The data type of the pointer, and the host/device memory location, depend on the paramLabel input selection. For more information, see cudnnFusedOpsVariantParamLabel_t.

Returns

CUDNN_STATUS_SUCCESS
The descriptor values are retrieved successfully.
CUDNN_STATUS_BAD_PARAM
If either varPack or ptr is NULL, or if paramLabel is set to invalid value.

6.2.17. cudnnMakeFusedOpsPlan()

cudnnStatus_t cudnnMakeFusedOpsPlan(
	cudnnHandle_t handle,
	cudnnFusedOpsPlan_t plan,
	const cudnnFusedOpsConstParamPack_t constPack,
	size_t *workspaceSizeInBytes); 		

This function determines the optimum kernel to execute, and the workspace size the user should allocate, prior to the actual execution of the fused operations by cudnnFusedOpsExecute().

Parameters

handle
Input. Pointer to the cuDNN library context.
plan
Input. Pointer to a previously-created and initialized plan descriptor.
constPack
Input. Pointer to the descriptor to the const parameters pack.
workspaceSizeInBytes
Output. The amount of workspace size the user should allocate for the execution of this plan.

Returns

CUDNN_STATUS_BAD_PARAM
If any of the inputs is NULL, or if the type of cudnnFusedOps_t in the constPack descriptor is unsupported.
CUDNN_STATUS_SUCCESS
The function executed successfully.

6.2.18. cudnnSetFusedOpsConstParamPackAttribute()

cudnnStatus_t cudnnSetFusedOpsConstParamPackAttribute(
	cudnnFusedOpsConstParamPack_t constPack,
	cudnnFusedOpsConstParamLabel_t paramLabel,
	const void *param);		

This function sets the descriptor pointed to by the param pointer input. The type of the descriptor to be set is indicated by the enum value of the paramLabel input.

Parameters

constPack
Input. The opaque cudnnFusedOpsConstParamPack_t structure that contains the various problem size information, such as the shape, layout and the type of tensors, the descriptors for convolution and activation, and settings for operations such as convolution and activation.
paramLabel
Input. Several types of descriptors can be set by this setter function. The param input points to the descriptor itself, and this input indicates the type of the descriptor pointed to by the param input. The cudnnFusedOpsConstParamLabel_t enumerant type enables the selection of the type of the descriptor.
param
Input. Data pointer to the host memory, associated with the specific descriptor. The type of the descriptor depends on the value of paramLabel. For more information, see the table in cudnnFusedOpsConstParamLabel_t.

If this pointer is set to NULL, then the cuDNN library will record as such. If not, then the values pointed to by this pointer (meaning, the value or the opaque structure underneath) will be copied into the constPack during ​cudnnSetFusedOpsConstParamPackAttribute() operation.

Returns

CUDNN_STATUS_SUCCESS
The descriptor is set successfully.
CUDNN_STATUS_BAD_PARAM
If constPack is NULL, or if paramLabel or the ops setting for constPack is invalid.

6.2.19. cudnnSetFusedOpsVariantParamPackAttribute()

cudnnStatus_t cudnnSetFusedOpsVariantParamPackAttribute(
	cudnnFusedOpsVariantParamPack_t varPack,
	cudnnFusedOpsVariantParamLabel_t paramLabel,
	void *ptr);		

This function sets the variable parameter pack descriptor.

Parameters

varPack
Input. Pointer to the cudnnFusedOps variant parameter pack (varPack) descriptor.
paramLabel
Input. Type to which the buffer pointer parameter (in the varPack descriptor) is set by this function. For more information, see cudnnFusedOpsConstParamLabel_t.
ptr
Input. Pointer, to the host or device memory, to the value to which the descriptor parameter is set. The data type of the pointer, and the host/device memory location, depend on the paramLabel input selection. For more information, see cudnnFusedOpsVariantParamLabel_t.

Returns

CUDNN_STATUS_BAD_PARAM
If varPack is NULL or if paramLabel is set to an unsupported value.
CUDNN_STATUS_SUCCESS
The descriptor was set successfully.

7. cudnn_adv_infer.so Library

7.1. Data Type References

7.1.1. Pointer To Opaque Struct Types

7.1.1.1. cudnnAttnDescriptor_t

cudnnAttnDescriptor_t is a pointer to an opaque structure holding parameters of the multi-head attention layer such as:
  • weight and bias tensor shapes (vector lengths before and after linear projections)
  • parameters that can be set in advance and do not change when invoking functions to evaluate forward responses and gradients (number of attention heads, softmax smoothing/sharpening coefficient)
  • other settings that are necessary to compute temporary buffer sizes.

Use the cudnnCreateAttnDescriptor() function to create an instance of the attention descriptor object and cudnnDestroyAttnDescriptor() to delete the previously created descriptor. Use the cudnnSetAttnDescriptor() function to configure the descriptor.

7.1.1.2. cudnnPersistentRNNPlan_t

This function is deprecated starting in cuDNN 8.0.0.

cudnnPersistentRNNPlan_t is a pointer to an opaque structure holding a plan to execute a dynamic persistent RNN. cudnnCreatePersistentRNNPlan() is used to create and initialize one instance.

7.1.1.3. cudnnRNNDataDescriptor_t

cudnnRNNDataDescriptor_t is a pointer to an opaque structure holding the description of an RNN data set. The function cudnnCreateRNNDataDescriptor() is used to create one instance, and cudnnSetRNNDataDescriptor() must be used to initialize this instance.

7.1.1.4. cudnnRNNDescriptor_t

cudnnRNNDescriptor_t is a pointer to an opaque structure holding the description of an RNN operation. cudnnCreateRNNDescriptor() is used to create one instance.

7.1.1.5. cudnnSeqDataDescriptor_t

cudnnSeqDataDescriptor_t is a pointer to an opaque structure holding parameters of the sequence data container or buffer. The sequence data container is used to store fixed size vectors defined by the VECT dimension. Vectors are arranged in additional three dimensions: TIME, BATCH and BEAM.

The TIME dimension is used to bundle vectors into sequences of vectors. The actual sequences can be shorter than the TIME dimension, therefore, additional information is needed about each sequence length and how unused (padding) vectors should be saved.

It is assumed that the sequence data container is fully packed. The TIME, BATCH and BEAM dimensions can be in any order when vectors are traversed in the ascending order of addresses. Six data layouts (permutation of TIME, BATCH and BEAM) are possible.

The cudnnSeqDataDescriptor_t object holds the following parameters:
  • data type used by vectors
  • TIME, BATCH, BEAM and VECT dimensions
  • data layout
  • the length of each sequence along the TIME dimension
  • an optional value to be copied to output padding vectors

Use the cudnnCreateSeqDataDescriptor() function to create one instance of the sequence data descriptor object and cudnnDestroySeqDataDescriptor() to delete a previously created descriptor. Use the cudnnSetSeqDataDescriptor() function to configure the descriptor.

This descriptor is used by multi-head attention API functions.

7.1.2. Enumeration Types

7.1.2.1. cudnnDirectionMode_t

cudnnDirectionMode_t is an enumerated type used to specify the recurrence pattern in the cudnnRNNForwardInference(), cudnnRNNForwardTraining(), cudnnRNNBackwardData() and cudnnRNNBackwardWeights() routines.

Values
CUDNN_UNIDIRECTIONAL
The network iterates recurrently from the first input to the last.
CUDNN_BIDIRECTIONAL
Each layer of the network iterates recurrently from the first input to the last and separately from the last input to the first. The outputs of the two are concatenated at each iteration giving the output of the layer.

7.1.2.2. cudnnForwardMode_t

cudnnForwardMode_t is an enumerated type to specify inference or training mode in RNN API. This parameter allows the cuDNN library to tune more precisely the size of the workspace buffer that could be different in inference and training regimens.

Values
CUDNN_FWD_MODE_INFERENCE

Selects the inference mode.

CUDNN_FWD_MODE_TRAINING

Selects the training mode.

7.1.2.3. cudnnMultiHeadAttnWeightKind_t

cudnnMultiHeadAttnWeightKind_t is an enumerated type that specifies a group of weights or biases in the cudnnGetMultiHeadAttnWeights() function.

Values
CUDNN_MH_ATTN_Q_WEIGHTS

Selects the input projection weights for queries.

CUDNN_MH_ATTN_K_WEIGHTS

Selects the input projection weights for keys.

CUDNN_MH_ATTN_V_WEIGHTS

Selects the input projection weights for values.

CUDNN_MH_ATTN_O_WEIGHTS

Selects the output projection weights.

CUDNN_MH_ATTN_Q_BIASES

Selects the input projection biases for queries.

CUDNN_MH_ATTN_K_BIASES

Selects the input projection biases for keys.

CUDNN_MH_ATTN_V_BIASES

Selects the input projection biases for values.

CUDNN_MH_ATTN_O_BIASES

Selects the output projection biases.

7.1.2.4. cudnnRNNBiasMode_t

cudnnRNNBiasMode_t is an enumerated type used to specify the number of bias vectors for RNN functions. See the description of the cudnnRNNMode_t enumerated type for the equations for each cell type based on the bias mode.

Values
CUDNN_RNN_NO_BIAS

Applies RNN cell formulas that do not use biases.

CUDNN_RNN_SINGLE_INP_BIAS

Applies RNN cell formulas that use one input bias vector in the input GEMM.

CUDNN_RNN_DOUBLE_BIAS

Applies RNN cell formulas that use two bias vectors.

CUDNN_RNN_SINGLE_REC_BIAS

Applies RNN cell formulas that use one recurrent bias vector in the recurrent GEMM.

7.1.2.5. cudnnRNNClipMode_t

cudnnRNNClipMode_t is an enumerated type used to select the LSTM cell clipping mode. It is used with cudnnRNNSetClip(), cudnnRNNGetClip() functions, and internally within LSTM cells.

Values
CUDNN_RNN_CLIP_NONE

Disables LSTM cell clipping.

CUDNN_RNN_CLIP_MINMAX

Enables LSTM cell clipping.

7.1.2.6. cudnnRNNDataLayout_t

cudnnRNNDataLayout_t is an enumerated type used to select the RNN data layout. It is used in the API calls cudnnGetRNNDataDescriptor() and cudnnSetRNNDataDescriptor().

Values
CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_UNPACKED

Data layout is padded, with outer stride from one time-step to the next.

CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_PACKED

The sequence length is sorted and packed as in the basic RNN API.

CUDNN_RNN_DATA_LAYOUT_BATCH_MAJOR_UNPACKED

Data layout is padded, with outer stride from one batch to the next.

7.1.2.7. cudnnRNNInputMode_t

cudnnRNNInputMode_t is an enumerated type used to specify the behavior of the first layer in the cudnnRNNForwardInference(), cudnnRNNForwardTraining(), cudnnRNNBackwardData() and cudnnRNNBackwardWeights() routines.

Values
CUDNN_LINEAR_INPUT

A biased matrix multiplication is performed at the input of the first recurrent layer.

CUDNN_SKIP_INPUT

No operation is performed at the input of the first recurrent layer. If CUDNN_SKIP_INPUT is used the leading dimension of the input tensor must be equal to the hidden state size of the network.

7.1.2.8. cudnnRNNMode_t

cudnnRNNMode_t is an enumerated type used to specify the type of network used in the cudnnRNNForwardInference, cudnnRNNForwardTraining, cudnnRNNBackwardData and cudnnRNNBackwardWeights routines.

Values
CUDNN_RNN_RELU

A single-gate recurrent neural network with a ReLU activation function.

In the forward pass, the output ht for a given iteration can be computed from the recurrent input ht-1 and the previous layer input xt, given the matrices W, R and the bias vectors, where ReLU(x) = max(x, 0).

If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS (default mode), then the following equation with biases bW and bR applies:
ht = ReLU(Wixt + Riht-1 + bWi + bRi)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_SINGLE_REC_BIAS, then the following equation with bias b applies:
ht = ReLU(Wixt + Riht-1 + bi)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_NO_BIAS, then the following equation applies:
ht = ReLU(Wixt + Riht-1)
CUDNN_RNN_TANH

A single-gate recurrent neural network with a tanh activation function.

In the forward pass, the output ht for a given iteration can be computed from the recurrent input ht-1 and the previous layer input xt, given the matrices W, R and the bias vectors, and where tanh is the hyperbolic tangent function.

If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS (default mode), then the following equation with biasesbW and bR applies:
ht = tanh(Wixt + Riht-1 + bWi + bRi)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_SINGLE_REC_BIAS, then the following equation with bias b applies:
ht = tanh(Wixt + Riht-1 + bi)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_NO_BIAS, then the following equation applies:
ht = tanh(Wixt + Riht-1)
CUDNN_LSTM

A four-gate Long Short-Term Memory (LSTM) network with no peephole connections.

In the forward pass, the output ht and cell output ct for a given iteration can be computed from the recurrent input ht-1, the cell input ct-1 and the previous layer input xt, given the matrices W, R and the bias vectors.

In addition, the following applies:
  • σ is the sigmoid operator such that: σ(x) = 1 / (1 + e-x),
  • represents a point-wise multiplication,
  • tanh is the hyperbolic tangent function, and
  • it, ft, ot, c't represent the input, forget, output and new gates respectively.
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS (default mode), then the following equations with biases bW and bR apply:
it = σ(Wixt + Riht-1 + bWi + bRi)
ft = σ(Wfxt + Rfht-1 + bWf + bRf)
ot = σ(Woxt + Roht-1 + bWo + bRo)
c't = tanh(Wcxt + Rcht-1 + bWc + bRc)
ct = ft ◦ ct-1 + it ◦ c't
ht = ot ◦ tanh(ct)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_SINGLE_REC_BIAS, then the following equations with bias b apply:
it = σ(Wixt + Riht-1 + bi)
ft = σ(Wfxt + Rfht-1 + bf)
ot = σ(Woxt + Roht-1 + bo)
c't = tanh(Wcxt + Rcht-1 + bc)
ct = ft ◦ ct-1 + it ◦ c't
ht = ot ◦ tanh(ct)
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_NO_BIAS, then the following equations apply:
it = σ(Wixt + Riht-1)
ft = σ(Wfxt + Rfht-1)
ot = σ(Woxt + Roht-1)
c't = tanh(Wcxt + Rcht-1)
ct = ft ◦ ct-1 + it ◦ c't
ht = ot◦tanh(ct)
CUDNN_GRU

A three-gate network consisting of Gated Recurrent Units.

In the forward pass, the output ht for a given iteration can be computed from the recurrent input ht-1 and the previous layer input xt given matrices W, R and the bias vectors.

In addition, the following applies:
  • σ is the sigmoid operator such that: σ(x) = 1 / (1 + e-x),
  • represents a point-wise multiplication,
  • tanh is the hyperbolic tangent function, and
  • it, rt, h't represent the input, reset, new gates respectively.
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS (default mode), then the following equations with biases bW and bR apply:
it = σ(Wixt + Riht-1 + bWi + bRu)
rt = σ(Wrxt + Rrht-1 + bWr + bRr)
h't = tanh(Whxt + rt◦(Rhht-1 + bRh) + bWh)
ht = (1 - it) ◦ h't + it ◦ ht-1
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS, then the following equations with bias b apply:
it = σ(Wixt + Riht-1 + bi)
rt = σ(Wrxt + Rrht-1 + br)
h't = tanh(Whxt + rt ◦ (Rhht-1) + bWh)
ht = (1 - it) ◦ h't + it ◦ ht-1
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_SINGLE_REC_BIAS, then the following equations with bias b apply:
it = σ(Wixt + Riht-1 + bi)
rt = σ(Wrxt + Rrht-1 + br)
h't = tanh(Whxt + rt ◦ (Rhht-1 + bRh))
ht = (1 - it) ◦ h't + it ◦ ht-1
If cudnnRNNBiasMode_t biasMode in rnnDesc is CUDNN_RNN_NO_BIAS, then the following equations apply:
it = σ(Wixt + Riht-1)
rt = σ(Wrxt + Rrht-1)
h't = tanh(Whxt + rt ◦ (Rhht-1))
ht = (1 - it) ◦ h't + it ◦ ht-1

7.1.2.9. cudnnRNNPaddingMode_t

cudnnRNNPaddingMode_t is an enumerated type used to enable or disable the padded input/output.

Values
CUDNN_RNN_PADDED_IO_DISABLED

Disables the padded input/output.

CUDNN_RNN_PADDED_IO_ENABLED

Enables the padded input/output.

7.1.2.10. cudnnSeqDataAxis_t

cudnnSeqDataAxis_t is an enumerated type that indexes active dimensions in the dimA[] argument that is passed to the cudnnSetSeqDataDescriptor() function to configure the sequence data descriptor of type cudnnSeqDataDescriptor_t.

cudnnSeqDataAxis_t constants are also used in the axis[] argument of the cudnnSetSeqDataDescriptor() call to define the layout of the sequence data buffer in memory.

See cudnnSetSeqDataDescriptor() for a detailed description on how to use the cudnnSeqDataAxis_t enumerated type.

The CUDNN_SEQDATA_DIM_COUNT macro defines the number of constants in the cudnnSeqDataAxis_t enumerated type. This value is currently set to 4.

Values
CUDNN_SEQDATA_TIME_DIM

Identifies the TIME (sequence length) dimension or specifies the TIME in the data layout.

CUDNN_SEQDATA_BATCH_DIM

Identifies the BATCH dimension or specifies the BATCH in the data layout.

CUDNN_SEQDATA_BEAM_DIM

Identifies the BEAM dimension or specifies the BEAM in the data layout.

CUDNN_SEQDATA_VECT_DIM

Identifies the VECT (vector) dimension or specifies the VECT in the data layout.

7.2. API Functions

7.2.1. cudnnAdvInferVersionCheck()

cudnnStatus_t cudnnAdvInferVersionCheck(void)

This function checks to see whether the version of the AdvInfer subset of the library is consistent with the other sub-libraries.

Returns

CUDNN_STATUS_SUCCESS

The version is consistent with other sub-libraries.

CUDNN_STATUS_VERSION_MISMATCH

The version of AdvInfer is not consistent with other sub-libraries. Users should check the installation and make sure all sub-component versions are consistent.

7.2.2. cudnnBuildRNNDynamic()

cudnnStatus_t cudnnBuildRNNDynamic(
	cudnnHandle_t handle,
	cudnnRNNDescriptor_t rnnDesc,
	int32_t miniBatch);

This function compiles the RNN persistent code using CUDA runtime compilation library (NVRTC) when the CUDNN_RNN_ALGO_PERSIST_DYNAMIC algo is selected. The code is tailored to the current GPU and specific hyperparameters (miniBatch). This call is expected to be expensive in terms of runtime and should be invoked infrequently. Note that the CUDNN_RNN_ALGO_PERSIST_DYNAMIC algo does not support variable length sequences within the batch.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

miniBatch

Input. The exact number of sequences in a batch.

Returns

CUDNN_STATUS_SUCCESS

The code was built and linked successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING

The prerequisite runtime library could not be found.

CUDNN_STATUS_NOT_SUPPORTED

The current hyperparameters are invalid.

7.2.3. cudnnCreateAttnDescriptor()

cudnnStatus_t cudnnCreateAttnDescriptor(cudnnAttnDescriptor_t *attnDesc);

This function creates one instance of an opaque attention descriptor object by allocating the host memory for it and initializing all descriptor fields. The function writes NULL to attnDesc when the attention descriptor object cannot be allocated.

Use the cudnnSetAttnDescriptor() function to configure the attention descriptor and cudnnDestroyAttnDescriptor() to destroy it and release the allocated memory.

Parameters

attnDesc
Output. Pointer where the address to the newly created attention descriptor should be written.

Returns

CUDNN_STATUS_SUCCESS
The descriptor object was created successfully.
CUDNN_STATUS_BAD_PARAM
An invalid input argument was encountered (attnDesc=NULL).
CUDNN_STATUS_ALLOC_FAILED
The memory allocation failed.

7.2.4. cudnnCreatePersistentRNNPlan()

This function has been deprecated in cuDNN 8.0. Use cudnnBuildRNNDynamic() instead of cudnnCreatePersistentRNNPlan().

cudnnStatus_t cudnnCreatePersistentRNNPlan(
    cudnnRNNDescriptor_t        rnnDesc,
    const int                   minibatch,
    const cudnnDataType_t       dataType,
    cudnnPersistentRNNPlan_t   *plan)

This function creates a plan to execute persistent RNNs when using the CUDNN_RNN_ALGO_PERSIST_DYNAMIC algo. This plan is tailored to the current GPU and problem hyperparameters. This function call is expected to be expensive in terms of runtime and should be used infrequently. For more information, see cudnnRNNDescriptor_t, cudnnDataType_t, and cudnnPersistentRNNPlan_t.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

CUDNN_STATUS_RUNTIME_PREREQUISITE_MISSING

A prerequisite runtime library cannot be found.

CUDNN_STATUS_NOT_SUPPORTED

The current hyperparameters are invalid.

7.2.5. cudnnCreateRNNDataDescriptor()

cudnnStatus_t cudnnCreateRNNDataDescriptor(
    cudnnRNNDataDescriptor_t *RNNDataDesc)

This function creates a RNN data descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The RNN data descriptor object was created successfully.

CUDNN_STATUS_BAD_PARAM

RNNDataDesc is NULL.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

7.2.6. cudnnCreateRNNDescriptor()

cudnnStatus_t cudnnCreateRNNDescriptor(
    cudnnRNNDescriptor_t    *rnnDesc)

This function creates a generic RNN descriptor object by allocating the memory needed to hold its opaque structure.

Returns

CUDNN_STATUS_SUCCESS

The object was created successfully.

CUDNN_STATUS_ALLOC_FAILED

The resources could not be allocated.

7.2.7. cudnnCreateSeqDataDescriptor()

cudnnStatus_t cudnnCreateSeqDataDescriptor(cudnnSeqDataDescriptor_t *seqDataDesc);	

This function creates one instance of an opaque sequence data descriptor object by allocating the host memory for it and initializing all descriptor fields. The function writes NULL to seqDataDesc when the sequence data descriptor object cannot be allocated.

Use the cudnnSetSeqDataDescriptor() function to configure the sequence data descriptor and cudnnDestroySeqDataDescriptor() to destroy it and release the allocated memory.

Parameters

seqDataDesc
Output. Pointer where the address to the newly created sequence data descriptor should be written.

Returns

CUDNN_STATUS_SUCCESS
The descriptor object was created successfully.
CUDNN_STATUS_BAD_PARAM
An invalid input argument was encountered (seqDataDesc=NULL).
CUDNN_STATUS_ALLOC_FAILED
The memory allocation failed.

7.2.8. cudnnDestroyAttnDescriptor()

cudnnStatus_t cudnnDestroyAttnDescriptor(cudnnAttnDescriptor_t attnDesc);

This function destroys the attention descriptor object and releases its memory. The attnDesc argument can be NULL. Invoking cudnnDestroyAttnDescriptor() with a NULL argument is a no operation (NOP).

The cudnnDestroyAttnDescriptor() function is not able to detect if the attnDesc argument holds a valid address. Undefined behavior will occur in case of passing an invalid pointer, not returned by the cudnnCreateAttnDescriptor() function, or in the double deletion scenario of a valid address.

Parameters

attnDesc
Input. Pointer to the attention descriptor object to be destroyed.

Returns

CUDNN_STATUS_SUCCESS
The descriptor was destroyed successfully.

7.2.9. cudnnDestroyPersistentRNNPlan()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnDestroyPersistentRNNPlan(
    cudnnPersistentRNNPlan_t plan)

This function destroys a previously created persistent RNN plan object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

7.2.10. cudnnDestroyRNNDataDescriptor()

cudnnStatus_t cudnnDestroyRNNDataDescriptor(
    cudnnRNNDataDescriptor_t RNNDataDesc)

This function destroys a previously created RNN data descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The RNN data descriptor object was destroyed successfully.

7.2.11. cudnnDestroyRNNDescriptor()

cudnnStatus_t cudnnDestroyRNNDescriptor(
    cudnnRNNDescriptor_t rnnDesc)

This function destroys a previously created RNN descriptor object.

Returns

CUDNN_STATUS_SUCCESS

The object was destroyed successfully.

7.2.12. cudnnDestroySeqDataDescriptor()

cudnnStatus_t cudnnDestroySeqDataDescriptor(cudnnSeqDataDescriptor_t seqDataDesc);

This function destroys the sequence data descriptor object and releases its memory. The seqDataDesc argument can be NULL. Invoking cudnnDestroySeqDataDescriptor() with a NULL argument is a no operation (NOP).

The cudnnDestroySeqDataDescriptor() function is not able to detect if the seqDataDesc argument holds a valid address. Undefined behavior will occur in case of passing an invalid pointer, not returned by the cudnnCreateSeqDataDescriptor() function, or in the double deletion scenario of a valid address.

Parameters

seqDataDesc
Input. Pointer to the sequence data descriptor object to be destroyed.

Returns

CUDNN_STATUS_SUCCESS
The descriptor was destroyed successfully.

7.2.13. cudnnFindRNNForwardInferenceAlgorithmEx()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnFindRNNForwardInferenceAlgorithmEx(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    const void                     *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                     *cx,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const cudnnTensorDescriptor_t   *yDesc,
    void                           *y,
    const cudnnTensorDescriptor_t   hyDesc,
    void                           *hy,
    const cudnnTensorDescriptor_t   cyDesc,
    void                           *cy,
    const float                    findIntensity,
    const int                      requestedAlgoCount,
    int                            *returnedAlgoCount,
    cudnnAlgorithmPerformance_t    *perfResults,
    void                           *workspace,
    size_t                          workSpaceSizeInBytes)

This function attempts all available cuDNN algorithms for cudnnRNNForwardInference(), using user-allocated GPU memory. It outputs the parameters that influence the performance of the algorithm to a user-allocated array of cudnnAlgorithmPerformance_t. These parameter metrics are written in sorted fashion where the first element has the lowest compute time.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc

Input. An array of fully packed tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

x

Input. Data pointer to GPU memory associated with the tensor descriptors in the array xDesc. The data are expected to be packed contiguously with the first element of iteration n+1 following directly from the last element of iteration n.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in xDesc.
y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc. The data are expected to be packed contiguously with the first element of iteration n+1 following directly from the last element of iteration n.

hyDesc
Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc
Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

findIntensity

Input.This input was previously unused in versions prior to 7.2.0. It is used in cuDNN 7.2.0 and later versions to control the overall runtime of the RNN find algorithms, by selecting the percentage of a large Cartesian product space to be searched.

  • Setting findIntensity within the range (0,1.] will set a percentage of the entire RNN search space to search. When findIntensity is set to 1.0, a full search is performed over all RNN parameters.
  • When findIntensity is set to 0.0f, a quick, minimal search is performed. This setting has the best runtime. However, in this case the parameters returned by this function will not correspond to the best performance of the algorithm; a longer search might discover better parameters. This option will execute up to three instances of the configured RNN problem. Runtime will vary proportionally to RNN problem size, as it will in the other cases, hence no guarantee of an explicit time bound can be given.
  • Setting findIntensity within the range [-1.,0) sets a percentage of a reduced Cartesian product space to be searched. This reduced search space has been heuristically selected to have good performance. The setting of -1.0 represents a full search over this reduced search space.
  • Values outside the range [-1,1] are truncated to the range [-1,1], and then interpreted as per the above.
  • Setting findIntensity to 1.0 in cuDNN 7.2 and later versions is equivalent to the behavior of this function in versions prior to cuDNN 7.2.0.
  • This function times the single RNN executions over large parameter spaces - one execution per parameter combination. The times returned by this function are latencies.
requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, cxDesc, wDesc, hyDesc, cyDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, cxDesc, wDesc, yDesc, hyDesc, cyDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

7.2.14. cudnnGetAttnDescriptor()

cudnnStatus_t cudnnGetAttnDescriptor(
    cudnnAttnDescriptor_t attnDesc,
    unsigned *attnMode,
    int *nHeads,
    double *smScaler,
    cudnnDataType_t *dataType,
    cudnnDataType_t *computePrec,
    cudnnMathType_t *mathType,
    cudnnDropoutDescriptor_t *attnDropoutDesc,
    cudnnDropoutDescriptor_t *postDropoutDesc,
    int *qSize,
    int *kSize,
    int *vSize,
    int *qProjSize,
    int *kProjSize,
    int *vProjSize,
    int *oProjSize,
    int *qoMaxSeqLength,
    int *kvMaxSeqLength,
    int *maxBatchSize,
    int *maxBeamSize);

This function retrieves settings from the previously created attention descriptor. The user can assign NULL to any pointer except attnDesc when the retrieved value is not needed.

Parameters

attnDesc
Input. Attention descriptor.
attnMode
Output. Pointer to the storage for binary attention flags.
nHeads
Output. Pointer to the storage for the number of attention heads.
smScaler
Output. Pointer to the storage for the softmax smoothing/sharpening coefficient.
dataType
Output. Data type for attention weights, sequence data inputs, and outputs.
computePrec
Output. Pointer to the storage for the compute precision.
mathType
Output. NVIDIA Tensor Core settings.
attnDropoutDesc
Output. Descriptor of the dropout operation applied to the softmax output.
postDropoutDesc
Output. Descriptor of the dropout operation applied to the multi-head attention output.
qSize, kSize, vSize
Output. Q, K, and V embedding vector lengths.
qProjSize, kProjSize, vProjSize
Output. Q, K, and V embedding vector lengths after input projections.
oProjSize
Output. Pointer to store the output vector length after projection.
qoMaxSeqLength
Output. Largest sequence length expected in sequence data descriptors related to Q, O, dQ, dO inputs and outputs.
kvMaxSeqLength
Output. Largest sequence length expected in sequence data descriptors related to K, V, dK, dV inputs and outputs.
maxBatchSize
Output. Largest batch size expected in the cudnnSeqDataDescriptor_t container.
maxBeamSize
Output. Largest beam size expected in the cudnnSeqDataDescriptor_t container.

Returns

CUDNN_STATUS_SUCCESS
Requested attention descriptor fields were retrieved successfully.
CUDNN_STATUS_BAD_PARAM
An invalid input argument was found.

7.2.15. cudnnGetMultiHeadAttnBuffers()

cudnnStatus_t cudnnGetMultiHeadAttnBuffers(
	cudnnHandle_t handle,
	const cudnnAttnDescriptor_t attnDesc,
	size_t *weightSizeInBytes,
	size_t *workSpaceSizeInBytes,
	size_t *reserveSpaceSizeInBytes);
This function computes weight, work, and reserve space buffer sizes used by the following functions:
  • cudnnMultiHeadAttnForward()
  • cudnnMultiHeadAttnBackwardData()
  • cudnnMultiHeadAttnBackwardWeights()
Assigning NULL to the reserveSpaceSizeInBytes argument indicates that the user does not plan to invoke multi-head attention gradient functions: cudnnMultiHeadAttnBackwardData() and cudnnMultiHeadAttnBackwardWeights(). This situation occurs in the inference mode.
Note:NULL cannot be assigned to weightSizeInBytes and workSpaceSizeInBytes pointers.

The user must allocate weight, work, and reserve space buffer sizes in the GPU memory using cudaMalloc() with the reported buffer sizes. The buffers can be also carved out from a larger chunk of allocated memory but the buffer addresses must be at least 16B aligned.

The work-space buffer is used for temporary storage. Its content can be discarded or modified after all GPU kernels launched by the corresponding API complete. The reserve-space buffer is used to transfer intermediate results from cudnnMultiHeadAttnForward() to cudnnMultiHeadAttnBackwardData(), and from cudnnMultiHeadAttnBackwardData() to cudnnMultiHeadAttnBackwardWeights(). The content of the reserve-space buffer cannot be modified until all GPU kernels launched by the above three multi-head attention API functions finish.

All multi-head attention weight and bias tensors are stored in a single weight buffer. For speed optimizations, the cuDNN API may change tensor layouts and their relative locations in the weight buffer based on the provided attention parameters. Use the cudnnGetMultiHeadAttnWeights() function to obtain the start address and the shape of each weight or bias tensor.

Parameters

handle

Input. The current cuDNN context handle.

attnDesc

Input. Pointer to a previously initialized attention descriptor.

weightSizeInBytes

Output. Minimum buffer size required to store all multi-head attention trainable parameters.

workSpaceSizeInBytes

Output. Minimum buffer size required to hold all temporary surfaces used by the forward and gradient multi-head attention API calls.

reserveSpaceSizeInBytes

Output. Minimum buffer size required to store all intermediate data exchanged between forward and backward (gradient) multi-head attention functions. Set this parameter to NULL in the inference mode indicating that gradient API calls will not be invoked.

Returns

CUDNN_STATUS_SUCCESS

The requested buffer sizes were computed successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was found.

7.2.16. cudnnGetMultiHeadAttnWeights()

cudnnStatus_t cudnnGetMultiHeadAttnWeights(
    cudnnHandle_t handle,
    const cudnnAttnDescriptor_t attnDesc,
    cudnnMultiHeadAttnWeightKind_t wKind,
    size_t weightSizeInBytes,
    const void *weights,
    cudnnTensorDescriptor_t wDesc,
    void **wAddr);

This function obtains the shape of the weight or bias tensor. It also retrieves the start address of tensor data located in the weight buffer. Use the wKind argument to select a particular tensor. For more information, see cudnnMultiHeadAttnWeightKind_t for the description of the enumerant type.

Biases are used in the input and output projections when the CUDNN_ATTN_ENABLE_PROJ_BIASES flag is set in the attention descriptor. See cudnnSetAttnDescriptor() for the description of flags to control projection biases.

When the corresponding weight or bias tensor does not exist, the function writes NULL to the storage location pointed by wAddr and returns zeros in the wDesc tensor descriptor. The return status of the cudnnGetMultiHeadAttnWeights() function is CUDNN_STATUS_SUCCESS in this case.

The cuDNNmultiHeadAttention sample code demonstrates how to access multi-head attention weights. Although the buffer with weights and biases should be allocated in the GPU memory, the user can copy it to the host memory and invoke the cudnnGetMultiHeadAttnWeights() function with the host weights address to obtain tensor pointers in the host memory. This scheme allows the user to inspect trainable parameters directly in the CPU memory.

Parameters

handle

Input. The current cuDNN context handle.

attnDesc

Input. A previously configured attention descriptor.

wKind

Input. Enumerant type to specify which weight or bias tensor should be retrieved.

weightSizeInBytes

Input. Buffer size that stores all multi-head attention weights and biases.

weights

Input. Pointer to the weight buffer in the host or device memory.

wDesc

Output. The descriptor specifying weight or bias tensor shape. For weights, the wDesc.dimA[] array has three elements: [nHeads, projected size, original size]. For biases, the wDesc.dimA[] array also has three elements: [nHeads, projected size, 1]. The wDesc.strideA[] array describes how tensor elements are arranged in memory.

wAddr

Output. Pointer to a location where the start address of the requested tensor should be written. When the corresponding projection is disabled, the address written to wAddr is NULL.

Returns

CUDNN_STATUS_SUCCESS

The weight tensor descriptor and the address of data in the device memory were successfully retrieved.

CUDNN_STATUS_BAD_PARAM

An invalid or incompatible input argument was encountered. For example, wKind did not have a valid value or weightSizeInBytes was too small.

7.2.17. cudnnGetRNNBackwardDataAlgorithmMaxCount()

This function has been deprecated in cuDNN 8.0.

7.2.18. cudnnGetRNNBiasMode()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNDescriptor_v8() instead of cudnnGetRNNBiasMode()

cudnnStatus_t cudnnGetRNNBiasMode(
   cudnnRNNDescriptor_t   rnnDesc,
   cudnnRNNBiasMode_t     *biasMode)

This function retrieves the RNN bias mode that was configured by cudnnSetRNNBiasMode(). The default value of biasMode in rnnDesc after cudnnCreateRNNDescriptor() is CUDNN_RNN_DOUBLE_BIAS.

Parameters

rnnDesc

Input. A previously created RNN descriptor.

*biasMode
Output. Pointer to where RNN bias mode should be saved.

Returns

CUDNN_STATUS_BAD_PARAM

Either the rnnDesc or *biasMode is NULL.

CUDNN_STATUS_SUCCESS

The biasMode parameter was retrieved successfully.

7.2.19. cudnnGetRNNDataDescriptor()

cudnnStatus_t cudnnGetRNNDataDescriptor(
    cudnnRNNDataDescriptor_t       RNNDataDesc,
    cudnnDataType_t                *dataType,
    cudnnRNNDataLayout_t           *layout,
    int                            *maxSeqLength,
    int                            *batchSize,
    int                            *vectorSize,
    int                            arrayLengthRequested,
    int                            seqLengthArray[],
    void                           *paddingFill);

This function retrieves a previously created RNN data descriptor object.

Parameters

RNNDataDesc

Input. A previously created and initialized RNN descriptor.

dataType

Output. Pointer to the host memory location to store the datatype of the RNN data tensor.

layout

Output. Pointer to the host memory location to store the memory layout of the RNN data tensor.

maxSeqLength

Output. The maximum sequence length within this RNN data tensor, including the padding vectors.

batchSize

Output. The number of sequences within the mini-batch.

vectorSize

Output. The vector length (meaning, embedding size) of the input or output tensor at each time-step.

arrayLengthRequested

Input. The number of elements that the user requested for seqLengthArray.

seqLengthArray

Output. Pointer to the host memory location to store the integer array describing the length (meaning, number of time-steps) of each sequence. This is allowed to be a NULL pointer if arrayLengthRequested is 0.

paddingFill

Output. Pointer to the host memory location to store the user defined symbol. The symbol should be interpreted as the same data type as the RNN data tensor.

Returns

CUDNN_STATUS_SUCCESS

The parameters are fetched successfully.

CUDNN_STATUS_BAD_PARAM

Any one of these have occurred:

  • Any of RNNDataDesc, dataType, layout, maxSeqLength, batchSize, vectorSize, paddingFill is NULL.
  • seqLengthArray is NULL while arrayLengthRequested is greater than zero.
  • arrayLengthRequested is less than zero.

7.2.20. cudnnGetRNNDescriptor_v6()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNDescriptor_v8() instead of cudnnGetRNNDescriptor_v6().

cudnnStatus_t cudnnGetRNNDescriptor_v6(
	cudnnHandle_t handle,
	cudnnRNNDescriptor_t rnnDesc,
	int *hiddenSize,
	int *numLayers,
    cudnnDropoutDescriptor_t *dropoutDesc,
	cudnnRNNInputMode_t *inputMode,
	cudnnDirectionMode_t *direction,
	cudnnRNNMode_t *cellMode,
	cudnnRNNAlgo_t *algo,
	cudnnDataType_t *mathPrec) {

This function retrieves RNN network parameters that were configured by cudnnSetRNNDescriptor_v6(). All pointers passed to the function should be not-NULL or CUDNN_STATUS_BAD_PARAM is reported. The function does not check the validity of retrieved parameters.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously created and initialized RNN descriptor.

hiddenSize

Output. Pointer to where the size of the hidden state should be stored (the same value is used in every RNN layer).

numLayers

Output. Pointer to where the number of RNN layers should be stored.

dropoutDesc

Output. Pointer to where the handle to a previously configured dropout descriptor should be stored.

inputMode

Output. Pointer to where the mode of the first RNN layer should be saved.

direction

Output. Pointer to where RNN uni-directional/bi-directional mode should be saved.

mode

Output. Pointer to where RNN cell type should be saved.

algo

Output. Pointer to where RNN algorithm type should be stored.

mathPrec

Output. Pointer to where the math precision type should be stored.

Returns

CUDNN_STATUS_SUCCESS

RNN parameters were successfully retrieved from the RNN descriptor.

CUDNN_STATUS_BAD_PARAM

At least one pointer passed to the function is NULL.

7.2.21. cudnnGetRNNDescriptor_v8()

cudnnStatus_t cudnnGetRNNDescriptor_v8(
	cudnnRNNDescriptor_t rnnDesc,
	cudnnRNNAlgo_t *algo,
	cudnnRNNMode_t *cellMode,
	cudnnRNNBiasMode_t *biasMode,
	cudnnDirectionMode_t *dirMode,
	cudnnRNNInputMode_t *inputMode,
	cudnnDataType_t *dataType,
	cudnnDataType_t *mathPrec,
	cudnnMathType_t *mathType,
	int32_t *inputSize,
	int32_t *hiddenSize,
	int32_t *projSize,
	int32_t *numLayers,
    cudnnDropoutDescriptor_t *dropoutDesc,
	uint32_t *auxFlags);

This function retrieves RNN network parameters that were configured by cudnnSetRNNDescriptor_v8(). The user can assign NULL to any pointer except rnnDesc when the retrieved value is not needed. The function does not check the validity of retrieved parameters.

Parameters

rnnDesc

Input. A previously created and initialized RNN descriptor.

algo

Output. Pointer to where RNN algorithm type should be stored.

cellMode
Output. Pointer to where RNN cell type should be saved.
biasMode

Output. Pointer to where RNN bias mode cudnnRNNBiasMode_t should be saved.

dirMode

Output. Pointer to where RNN uni-directional/bi-directional mode should be saved.

inputMode

Output. Pointer to where the mode of the first RNN layer should be saved.

dataType

Output. Pointer to where the data type of RNN weights/biases should be stored.

mathPrec

Output. Pointer to where the math precision type should be stored.

mathType

Output. Pointer to where the preferred option for Tensor Cores are saved.

inputSize

Output. Pointer to where the RNN input vector size is stored.

hiddenSize

Output. Pointer to where the size of the hidden state should be stored (the same value is used in every RNN layer).

projSize

Output. Pointer to where the LSTM cell output size after the recurrent projection is stored.

numLayers

Output. Pointer to where the number of RNN layers should be stored.

dropoutDesc

Output. Pointer to where the handle to a previously configured dropout descriptor should be stored.

auxFlags

Output. Pointer to miscellaneous RNN options (flags) that do not require passing additional numerical values to configure.

Returns

CUDNN_STATUS_SUCCESS

RNN parameters were successfully retrieved from the RNN descriptor.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was found (rnnDesc was NULL).

CUDNN_STATUS_NOT_INITIALIZED

The RNN descriptor was configured with the legacy cudnnSetRNNDescriptor_v6() call.

7.2.22. cudnnGetRNNForwardInferenceAlgorithmMaxCount()

This function has been deprecated in cuDNN 8.0.

7.2.23. cudnnGetRNNForwardTrainingAlgorithmMaxCount()

This function has been deprecated in cuDNN 8.0.

7.2.24. cudnnGetRNNLinLayerBiasParams()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNWeightParams() instead of cudnnGetRNNLinLayerBiasParams().

cudnnStatus_t cudnnGetRNNLinLayerBiasParams(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       pseudoLayer,
    const cudnnTensorDescriptor_t   xDesc,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const int                       linLayerID,
    cudnnFilterDescriptor_t         linLayerBiasDesc,
    void                           **linLayerBias)

This function is used to obtain a pointer and a descriptor of every RNN bias column vector in each pseudo-layer within the recurrent network defined by rnnDesc and its input width specified in xDesc.

Note: The cudnnGetRNNLinLayerBiasParams() function was changed in cuDNN version 7.1.1 to match the behavior of cudnnGetRNNLinLayerMatrixParams().

The cudnnGetRNNLinLayerBiasParams() function returns the RNN bias vector size in two dimensions: rows and columns.

Due to historical reasons, the minimum number of dimensions in the filter descriptor is three. In previous versions of the cuDNN library, the function returns the total number of vector elements in linLayerBiasDesc as follows:
filterDimA[0]=total_size, 
filterDimA[1]=1, 
filterDimA[2]=1
For more information, see the description of the cudnnGetFilterNdDescriptor() function.
In cuDNN 7.1.1, the format was changed to:
filterDimA[0]=1, 
filterDimA[1]=rows, 
filterDimA[2]=1 (number of columns)
In both cases, the format field of the filter descriptor should be ignored when retrieved by cudnnGetFilterNdDescriptor().
The RNN implementation in cuDNN uses two bias vectors before the cell non-linear function. Note that the RNN implementation in cuDNN depends on the number of bias vectors before the cell non-linear function. Refer to the equations in the cudnnRNNMode_t description, for the enumerant type based on the value of cudnnRNNBiasMode_tbiasMode in rnnDesc. If nonexistent biases are referenced by linLayerID, then this function sets linLayerBiasDesc to a zeroed filter descriptor where:
filterDimA[0]=0, 
filterDimA[1]=0, and 
filterDimA[2]=2
and sets linLayerBias to NULL. Refer to the details for the function parameter linLayerID to determine the relevant values of linLayerID based on biasMode.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously initialized RNN descriptor.

pseudoLayer
Input. The pseudo-layer to query. In uni-directional RNNs, a pseudo-layer is the same as a physical layer (pseudoLayer=0 is the RNN input layer, pseudoLayer=1 is the first hidden layer). In bi-directional RNNs, there are twice as many pseudo-layers in comparison to physical layers.
  • pseudoLayer=0 refers to the forward part of the physical input layer
  • pseudoLayer=1 refers to the backward part of the physical input layer
  • pseudoLayer=2 is the forward part of the first hidden layer, and so on
xDesc

Input. A fully packed tensor descriptor describing the input to one recurrent iteration (to retrieve the RNN input width).

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

linLayerID

Input. Linear ID index of the weight matrix.

If cellMode in rnnDesc was set to CUDNN_RNN_RELU or CUDNN_RNN_TANH:
  • Value 0 references the weight matrix used in conjunction with the input from the previous layer or input to the RNN model.
  • Value 1 references the weight matrix used in conjunction with the hidden state from the previous time step or the initial hidden state.
If cellMode in rnnDesc was set to CUDNN_LSTM:
  • Values 0, 1, 2, and 3 reference weight matrices used in conjunction with the input from the previous layer or input to the RNN model.
  • Values 4, 5, 6, and 7 reference weight matrices used in conjunction with the hidden state from the previous time step or the initial hidden state.
  • Value 8 corresponds to the projection matrix, if enabled.
Values and their LSTM gates:
  • linLayerID0 and 4 correspond to the input gate.
  • linLayerID1 and 5 correspond to the forget gate.
  • linLayerID2 and 6 correspond to the new cell state calculations with a hyperbolic tangent.
  • linLayerID3 and 7 correspond to the output gate.
If cellMode in rnnDesc was set to CUDNN_GRU:
  • Values 0, 1, and 2 reference weight matrices used in conjunction with the input from the previous layer or input to the RNN model.
  • Values 3, 4, and 5 reference weight matrices used in conjunction with the hidden state from the previous time step or the initial hidden state.
Values and their GRU gates:
  • linLayerID0 and 3 correspond to the reset gate.
  • linLayerID1 and 4 references to the update gate.
  • linLayerID2 and 5 correspond to the new hidden state calculations with a hyperbolic tangent.
linLayerBiasDesc

Output. Handle to a previously created filter descriptor.

linLayerBias

Output. Data pointer to GPU memory associated with the filter descriptor linLayerBiasDesc.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • One of the following arguments is NULL: handle, rnnDesc, xDesc, wDesc, linLayerBiasDesc, linLayerBias.
  • A data type mismatch was detected between rnnDesc and other descriptors.
  • Minimum requirement for the w pointer alignment is not satisfied.
  • The value of pseudoLayer or linLayerID is out of range.
CUDNN_STATUS_INVALID_VALUE

Some elements of the linLayerBias vector are outside the w buffer boundaries as specified by the wDesc descriptor.

7.2.25. cudnnGetRNNLinLayerMatrixParams()

This function has been deprecated in cuDNN 8.0 . Use cudnnGetRNNWeightParams() instead of cudnnGetRNNLinLayerMatrixParams().

cudnnStatus_t cudnnGetRNNLinLayerMatrixParams(
cudnnHandle_t                   handle,
const cudnnRNNDescriptor_t      rnnDesc,
const int                       pseudoLayer,
const cudnnTensorDescriptor_t   xDesc,
const cudnnFilterDescriptor_t   wDesc,
const void                     *w,
const int                       linLayerID,
cudnnFilterDescriptor_t         linLayerMatDesc,
void                           **linLayerMat)

This function is used to obtain a pointer and a descriptor of every RNN weight matrix in each pseudo-layer within the recurrent network defined by rnnDesc and its input width specified in xDesc.

Note: The cudnnGetRNNLinLayerMatrixParams() function was enhanced in cuDNN version 7.1.1 without changing its prototype. Instead of reporting the total number of elements in each weight matrix in the linLayerMatDesc filter descriptor, the function returns the matrix size as two dimensions: rows and columns. Moreover, when a weight matrix does not exist, for example, due to CUDNN_SKIP_INPUT mode, the function returns NULL in linLayerMat and all fields of linLayerMatDesc are zero.

The cudnnGetRNNLinLayerMatrixParams() function returns the RNN matrix size in two dimensions: rows and columns. This allows the user to easily print and initialize RNN weight matrices. Elements in each weight matrix are arranged in the row-major order. Due to historical reasons, the minimum number of dimensions in the filter descriptor is three. In previous versions of the cuDNN library, the function returned the total number of weights in linLayerMatDesc as follows: filterDimA[0]=total_size, filterDimA[1]=1, filterDimA[2]=1 (see the description of the cudnnGetFilterNdDescriptor() function). In cuDNN 7.1.1, the format was changed to: filterDimA[0]=1, filterDimA[1]=rows, filterDimA[2]=columns. In both cases, the "format" field of the filter descriptor should be ignored when retrieved by cudnnGetFilterNdDescriptor().

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously initialized RNN descriptor.

pseudoLayer
Input. The pseudo-layer to query. In uni-directional RNNs, a pseudo-layer is the same as a physical layer (pseudoLayer=0 is the RNN input layer, pseudoLayer=1 is the first hidden layer). In bi-directional RNNs, there are twice as many pseudo-layers in comparison to physical layers.
  • pseudoLayer=0 refers to the forward part of the physical input layer
  • pseudoLayer=1 refers to the backward part of the physical input layer
  • pseudoLayer=2 is the forward part of the first hidden layer, and so on
xDesc

Input. A fully packed tensor descriptor describing the input to one recurrent iteration (to retrieve the RNN input width).

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

linLayerID
Input. The linear layer to obtain information about:
    • If mode in rnnDesc was set to CUDNN_RNN_RELU or CUDNN_RNN_TANH:
      • Value 0 references the bias applied to the input from the previous layer (relevant if biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_DOUBLE_BIAS).
      • Value 1 references the bias applied to the recurrent input (relevant if biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS or CUDNN_RNN_SINGLE_REC_BIAS).
    • If mode in rnnDesc was set to CUDNN_LSTM:
      • Values of 0, 1, 2 and 3 reference bias applied to the input from the previous layer (relevant if biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_DOUBLE_BIAS).
      • Values of 4, 5, 6 and 7 reference bias applied to the recurrent input (relevant if biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS or CUDNN_RNN_SINGLE_REC_BIAS).
      • Values and their associated gates:
        • Values 0 and 4 reference the input gate.
        • Values 1 and 5 reference the forget gate.
        • Values 2 and 6 reference the new memory gate.
        • Values 3 and 7 reference the output gate.
    • If mode in rnnDesc was set to CUDNN_GRU:
      • Values of 0, 1 and 2 reference bias applied to the input from the previous layer (relevant if biasMode in rnnDesc is CUDNN_RNN_SINGLE_INP_BIAS or CUDNN_RNN_DOUBLE_BIAS).
      • Values of 3, 4 and 5 reference bias applied to the recurrent input (relevant if biasMode in rnnDesc is CUDNN_RNN_DOUBLE_BIAS or CUDNN_RNN_SINGLE_REC_BIAS).
      • Values and their associated gates:
        • Values 0 and 3 reference the reset gate.
        • Values 1 and 4 reference the update gate.
        • Values 2 and 5 reference the new memory gate.
For more information on modes and bias modes, see cudnnRNNMode_t.
linLayerMatDesc

Output. Handle to a previously created filter descriptor. When the weight matrix does not exist, the returned filer descriptor has all fields set to zero.

linLayerMat

Output. Data pointer to GPU memory associated with the filter descriptor linLayerMatDesc. When the weight matrix does not exist, the returned pointer is NULL.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • One of the following arguments is NULL: handle, rnnDesc, xDesc, wDesc, linLayerMatDesc, linLayerMat.
  • A data type mismatch was detected between rnnDesc and other descriptors.
  • Minimum requirement for the w pointer alignment is not satisfied.
  • The value of pseudoLayer or linLayerID is out of range.
CUDNN_STATUS_INVALID_VALUE

Some elements of the linLayerMat vector are outside the w buffer boundaries as specified by the wDesc descriptor.

7.2.26. cudnnGetRNNMatrixMathType()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNDescriptor_v8() instead of cudnnGetRNNMatrixMathType().

cudnnStatus_t cudnnGetRNNMatrixMathType(
	cudnnRNNDescriptor_t rnnDesc,
	cudnnMathType_t *mType);

This function retrieves the preferred settings for NVIDIA Tensor Cores on Volta (SM 7.0) or higher GPUs. See the cudnnMathType_t description for more details.

Parameters

rnnDesc

Input. A previously created and initialized RNN descriptor.

mType

Output. Address where the preferred Tensor Core settings should be stored.

Returns

CUDNN_STATUS_SUCCESS

The requested RNN descriptor field was retrieved successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was found (rnnDesc or mType was NULL).

7.2.27. cudnnGetRNNPaddingMode()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNDescriptor_v8() instead of cudnnGetRNNPaddingMode().

cudnnStatus_t cudnnGetRNNPaddingMode(
    cudnnRNNDescriptor_t        rnnDesc,
    cudnnRNNPaddingMode_t       *paddingMode)

This function retrieves the RNN padding mode from the RNN descriptor.

Parameters

rnnDesc

Input/Output. A previously created RNN descriptor.

*paddingMode

Input. Pointer to the host memory where the RNN padding mode is saved.

Returns

CUDNN_STATUS_SUCCESS

The RNN padding mode parameter was retrieved successfully.

CUDNN_STATUS_BAD_PARAM

Either the rnnDesc or *paddingMode is NULL.

7.2.28. cudnnGetRNNParamsSize()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNWeightSpaceSize() instead of cudnnGetRNNParamsSize().

cudnnStatus_t cudnnGetRNNParamsSize(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const cudnnTensorDescriptor_t   xDesc,
    size_t                         *sizeInBytes,
    cudnnDataType_t                 dataType)

This function is used to query the amount of parameter space required to execute the RNN described by rnnDesc with input dimensions defined by xDesc.

Parameters

s
handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously initialized RNN descriptor.

xDesc

Input. A fully packed tensor descriptor describing the input to one recurrent iteration.

sizeInBytes

Output. Minimum amount of GPU memory needed as parameter space to be able to execute an RNN with the specified descriptor and input tensors.

dataType

Input. The data type of the parameters.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • The descriptor rnnDesc is invalid.
  • The descriptor xDesc is invalid.
  • The descriptor xDesc is not fully packed.
  • The combination of dataType and tensor descriptor data type is invalid.
CUDNN_STATUS_NOT_SUPPORTED

The combination of the RNN descriptor and tensor descriptors is not supported.

7.2.29. cudnnGetRNNProjectionLayers()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNDescriptor_v8() instead of cudnnGetRNNProjectionLayers().

cudnnStatus_t cudnnGetRNNProjectionLayers(
    cudnnHandle_t           handle,
    cudnnRNNDescriptor_t    rnnDesc,
    int                     *recProjSize,
    int                     *outProjSize)

This function retrieves the current RNN projection parameters. By default, the projection feature is disabled so invoking this function will yield recProjSize equal to hiddenSize and outProjSize set to zero. The cudnnSetRNNProjectionLayers() method enables the RNN projection.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously created and initialized RNN descriptor.

recProjSize

Output. Pointer where the recurrent projection size should be stored.

outProjSize

Output. Pointer where the output projection size should be stored.

Returns

CUDNN_STATUS_SUCCESS

RNN projection parameters were retrieved successfully.

CUDNN_STATUS_BAD_PARAM

A NULL pointer was passed to the function.

7.2.30. cudnnGetRNNTempSpaceSizes()

cudnnStatus_t cudnnGetRNNTempSpaceSizes(
	cudnnHandle_t handle,
	cudnnRNNDescriptor_t rnnDesc,
	cudnnForwardMode_t fMode,
    cudnnRNNDataDescriptor_t xDesc,
	size_t *workSpaceSize,
	size_t *reserveSpaceSize);

This function computes the work and reserve space buffer sizes based on the RNN network geometry stored in rnnDesc, designated usage (inference or training) defined by the fMode argument, and the current RNN data dimensions (maxSeqLength, batchSize) retrieved from xDesc. When RNN data dimensions change, the cudnnGetRNNTempSpaceSizes() must be called again because RNN temporary buffer sizes are not monotonic.

The user can assign NULL to workSpaceSize or reserveSpaceSize pointers when the corresponding value is not needed.

Parameters

handle

Input. The current cuDNN context handle.

rnnDesc

Input. A previously initialized RNN descriptor.

fMode
Input. Specifies whether temporary buffers are used in inference or training modes. The reserve-space buffer is not used during inference. Therefore, the returned size of the reserve space buffer will be zero when the fMode argument is CUDNN_FWD_MODE_INFERENCE.
xDesc

Input. A single RNN data descriptor that specifies current RNN data dimensions: maxSeqLength and batchSize.

workSpaceSize

Output. Minimum amount of GPU memory in bytes needed as a workspace buffer. The workspace buffer is not used to pass intermediate results between APIs but as a temporary read/write buffer.

reserveSpaceSize

Output. Minimum amount of GPU memory in bytes needed as the reserve-space buffer. The reserve space buffer is used to pass intermediate results from cudnnRNNForward() to RNN BackwardData and BackwardWeights routines that compute first order derivatives with respect to RNN inputs or trainable weight and biases.

Returns

CUDNN_STATUS_SUCCESS

RNN temporary buffer sizes were computed successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was detected.

CUDNN_STATUS_NOT_SUPPORTED

An incompatible or unsupported combination of input arguments was detected.

7.2.31. cudnnGetRNNWeightParams()

cudnnStatus_t cudnnGetRNNWeightParams(
	cudnnHandle_t handle,
	cudnnRNNDescriptor_t rnnDesc,
	int32_t pseudoLayer,
	size_t weightSpaceSize,
	const void *weightSpace,
	int32_t linLayerID,
    cudnnTensorDescriptor_t mDesc,
	void **mAddr,
    cudnnTensorDescriptor_t bDesc,
	void **bAddr);

This function is used to obtain the start address and shape of every RNN weight matrix and bias vector in each pseudo-layer within the recurrent network.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc
Input. A previously initialized RNN descriptor.
pseudoLayer
Input. The pseudo-layer to query. In uni-directional RNNs, a pseudo-layer is the same as a physical layer (pseudoLayer=0 is the RNN input layer, pseudoLayer=1 is the first hidden layer). In bi-directional RNNs, there are twice as many pseudo-layers in comparison to physical layers:
  • pseudoLayer=0 refers to the forward direction sub-layer of the physical input layer
  • pseudoLayer=1 refers to the backward direction sub-layer of the physical input layer
  • pseudoLayer=2 is the forward direction sub-layer of the first hidden layer, and so on
weightSpaceSize

Input. Size of the weight space buffer in bytes.

weightSpace

Input. Pointer to the weight space buffer.

linLayerID

Input. Weight matrix or bias vector linear ID index.

If cellMode in rnnDesc was set to CUDNN_RNN_RELU or CUDNN_RNN_TANH:
  • Value 0 references the weight matrix or bias vector used in conjunction with the input from the previous layer or input to the RNN model.
  • Value 1 references the weight matrix or bias vector used in conjunction with the hidden state from the previous time step or the initial hidden state.
If cellMode in rnnDesc was set to CUDNN_LSTM:
  • Values 0, 1, 2 and 3 reference weight matrices or bias vectors used in conjunction with the input from the previous layer or input to the RNN model.
  • Values 4, 5, 6 and 7 reference weight matrices or bias vectors used in conjunction with the hidden state from the previous time step or the initial hidden state.
  • Value 8 corresponds to the projection matrix, if enabled (there is no bias in this operation).
If cellMode in rnnDesc was set to CUDNN_GRU:
  • Values 0, 1 and 2 reference weight matrices or bias vectors used in conjunction with the input from the previous layer or input to the RNN model.
  • Values 3, 4 and 5 reference weight matrices or bias vectors used in conjunction with the hidden state from the previous time step or the initial hidden state.
Values and their GRU gates:
  • linLayerID0 and 3 correspond to the reset gate.
  • linLayerID1 and 4 reference to the update gate.
  • linLayerID2 and 5 correspond to the new hidden state calculations with hyperbolic tangent.

For more information on modes and bias modes, see cudnnRNNMode_t.

mDesc

Output. Handle to a previously created tensor descriptor. The shape of the corresponding weight matrix is returned in this descriptor in the following format: dimA[3] = {1, rows, cols}. The reported number of tensor dimensions is zero when the weight matrix does not exist. This situation occurs for input GEMM matrices of the first layer when CUDNN_SKIP_INPUT is selected or for the LSTM projection matrix when the feature is disabled.

mAddr

Output. Pointer to the beginning of the weight matrix within the weight space buffer. When the weight matrix does not exist, the returned address is NULL.

bDesc

Output. Handle to a previously created tensor descriptor. The shape of the corresponding bias vector is returned in this descriptor in the following format: dimA[3] = {1, rows, 1}. The reported number of tensor dimensions is zero when the bias vector does not exist.

bAddr

Output. Pointer to the beginning of the bias vector within the weight space buffer. When the bias vector does not exist, the returned address is NULL.

Returns

CUDNN_STATUS_SUCCESS

The query was completed successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was encountered. For example, the value of pseudoLayer is out of range or linLayerID is negative or larger than 8.

CUDNN_STATUS_INVALID_VALUE

Some weight/bias elements are outside the weight space buffer boundaries.

CUDNN_STATUS_NOT_INITIALIZED

The RNN descriptor was configured with the legacy cudnnSetRNNDescriptor_v6() call.

7.2.32. cudnnGetRNNWeightSpaceSize()

cudnnStatus_t cudnnGetRNNWeightSpaceSize(
	cudnnHandle_t handle,
	cudnnRNNDescriptor_t rnnDesc,
	size_t *weightSpaceSize);

This function reports the required size of the weight space buffer in bytes. The weight space buffer holds all RNN weight matrices and bias vectors.

Parameters

handle

Input. The current cuDNN context handle.

rnnDesc
Input. A previously initialized RNN descriptor.
weightSpaceSize

Output. Minimum size in bytes of GPU memory needed for all RNN trainable parameters.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was encountered. For example, any input argument was NULL.

CUDNN_STATUS_NOT_INITIALIZED

The RNN descriptor was configured with the legacy cudnnSetRNNDescriptor_v6() call.

7.2.33. cudnnGetRNNWorkspaceSize()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNTempSpaceSizes() instead of cudnnGetRNNWorkspaceSize().

cudnnStatus_t cudnnGetRNNWorkspaceSize(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    size_t                         *sizeInBytes)

This function is used to query the amount of work space required to execute the RNN described by rnnDesc with input dimensions defined by xDesc.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. Workspace that is allocated, based on the size that this function provides, cannot be used for sequences longer than seqLength.

xDesc

Input. An array of tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. For example, if you have multiple time series in a batch, they can be different lengths. This dimension is the batch size for the particular iteration of the sequence, and so it should decrease when a sequence in the batch has been terminated.

Each tensor descriptor must have the same second dimension (vector length).

sizeInBytes

Output. Minimum amount of GPU memory needed as workspace to be able to execute an RNN with the specified descriptor and input tensors.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors in xDesc is invalid.
  • The descriptors in xDesc have inconsistent second dimensions, strides or data types.
  • The descriptors in xDesc have increasing first dimensions.
  • The descriptors in xDesc are not fully packed.
CUDNN_STATUS_NOT_SUPPORTED

The data types in tensors described by xDesc are not supported.

7.2.34. cudnnGetSeqDataDescriptor()

cudnnStatus_t cudnnGetSeqDataDescriptor(
	const cudnnSeqDataDescriptor_t seqDataDesc,
	cudnnDataType_t *dataType,
	int *nbDims,
	int nbDimsRequested,
	int dimA[],
	cudnnSeqDataAxis_t axes[],
	size_t *seqLengthArraySize,
	size_t seqLengthSizeRequested,
	int seqLengthArray[],
	void *paddingFill);

This function retrieves settings from a previously created sequence data descriptor. The user can assign NULL to any pointer except seqDataDesc when the retrieved value is not needed. The nbDimsRequested argument applies to both dimA[] and axes[] arrays. A positive value of nbDimsRequested or seqLengthSizeRequested is ignored when the corresponding array, dimA[], axes[], or seqLengthArray[] is NULL.

The cudnnGetSeqDataDescriptor() function does not report the actual strides in the sequence data buffer. Those strides can be handy in computing the offset to any sequence data element. The user must precompute strides based on the axes[] and dimA[] arrays reported by the cudnnGetSeqDataDescriptor() function. Below is sample code that performs this task:
// Array holding sequence data strides.
size_t strA[CUDNN_SEQDATA_DIM_COUNT] = {0};
 
// Compute strides from dimension and order arrays.
size_t stride = 1;
for (int i = nbDims - 1; i >= 0; i--) {
	int j = int(axes[i]);
	if (unsigned(j) < CUDNN_SEQDATA_DIM_COUNT-1 && strA[j] == 0) {
   	strA[j] = stride;
   	stride *= dimA[j];
	} else {
	    fprintf(stderr, "ERROR: invalid axes[%d]=%d\n\n", i, j);
    	abort();
	}
}
Now, the strA[] array can be used to compute the index to any sequence data element, for example:
// Using four indices (batch, beam, time, vect) with ranges already checked.
size_t base = strA[CUDNN_SEQDATA_BATCH_DIM] * batch
	        + strA[CUDNN_SEQDATA_BEAM_DIM]  * beam
       	 + strA[CUDNN_SEQDATA_TIME_DIM]  * time;
val = seqDataPtr[base + vect];

The above code assumes that all four indices (batch, beam, time, vect) are less than the corresponding value in the dimA[] array. The sample code also omits the strA[CUDNN_SEQDATA_VECT_DIM] stride because its value is always 1, meaning, elements of one vector occupy a contiguous block of memory.

Parameters

seqDataDesc
Input. Sequence data descriptor.
dataType
Output. Data type used in the sequence data buffer.
nbDims
Output. The number of active dimensions in the dimA[] and axes[] arrays.
nbDimsRequested
Input. The maximum number of consecutive elements that can be written to dimA[] and axes[] arrays starting from index zero. The recommended value for this argument is CUDNN_SEQDATA_DIM_COUNT.
dimA[]
Output. Integer array holding sequence data dimensions.
axes[]
Output. Array of cudnnSeqDataAxis_t that defines the layout of sequence data in memory.
seqLengthArraySize
Output. The number of required elements in seqLengthArray[] to save all sequence lengths.
seqLengthSizeRequested
Input. The maximum number of consecutive elements that can be written to the seqLengthArray[] array starting from index zero.
seqLengthArray[]
Output. Integer array holding sequence lengths.
paddingFill
Output. Pointer to a storage location of dataType with the fill value that should be written to all padding vectors. Use NULL when an explicit initialization of output padding vectors was not requested.

Returns

CUDNN_STATUS_SUCCESS
Requested sequence data descriptor fields were retrieved successfully.
CUDNN_STATUS_BAD_PARAM
An invalid input argument was found.
CUDNN_STATUS_INTERNAL_ERROR
An inconsistent internal state was encountered.

7.2.35. cudnnMultiHeadAttnForward()

cudnnStatus_t cudnnMultiHeadAttnForward(
	cudnnHandle_t handle,
	const cudnnAttnDescriptor_t attnDesc,
	int currIdx,
	const int loWinIdx[],
	const int hiWinIdx[],
	const int devSeqLengthsQO[],
	const int devSeqLengthsKV[],
	const cudnnSeqDataDescriptor_t qDesc,
	const void *queries,
	const void *residuals,
	const cudnnSeqDataDescriptor_t kDesc,
	const void *keys,
	const cudnnSeqDataDescriptor_t vDesc,
	const void *values,
	const cudnnSeqDataDescriptor_t oDesc,
       void *out,
	size_t weightSizeInBytes,
	const void *weights,
	size_t workSpaceSizeInBytes,
	void *workSpace,
	size_t reserveSpaceSizeInBytes,
	void *reserveSpace);

The cudnnMultiHeadAttnForward() function computes the forward responses of the multi-head attention layer. When reserveSpaceSizeInBytes=0 and reserveSpace=NULL, the function operates in the inference mode in which backward (gradient) functions are not invoked, otherwise, the training mode is assumed. In the training mode, the reserve space is used to pass intermediate results from cudnnMultiHeadAttnForward() to cudnnMultiHeadAttnBackwardData() and from cudnnMultiHeadAttnBackwardData() to cudnnMultiHeadAttnBackwardWeights().

In the inference mode, the currIdx specifies the time-step or sequence index of the embedding vectors to be processed. In this mode, the user can perform one iteration for time-step zero (currIdx=0), then update Q, K, V vectors and the attention window, and execute the next step (currIdx=1). The iterative process can be repeated for all time-steps.

When all Q time-steps are available (for example, in the training mode or in the inference mode on the encoder side in self-attention),the user can assign a negative value to currIdx and the cudnnMultiHeadAttnForward() API will automatically sweep through all Q time-steps.

The loWinIdx[] and hiWinIdx[] host arrays specify the attention window size for each Q time-step. In a typical self-attention case, the user must include all previously visited embedding vectors but not the current or future vectors. In this situation, the user should set:
currIdx=0: loWinIdx[0]=0; hiWinIdx[0]=0;  // initial time-step, no attention window
currIdx=1: loWinIdx[1]=0; hiWinIdx[1]=1;  // attention window spans one vector
currIdx=2: loWinIdx[2]=0; hiWinIdx[2]=2;  // attention window spans two vectors
(...)

When currIdx is negative in cudnnMultiHeadAttnForward(), the loWinIdx[] and hiWinIdx[] arrays must be fully initialized for all time-steps. When cudnnMultiHeadAttnForward() is invoked with currIdx=0, currIdx=1, currIdx=2, etc., then the user can update loWinIdx[currIdx] and hiWinIdx[currIdx] elements only before invoking the forward response function. All other elements in the loWinIdx[] and hiWinIdx[] arrays will not be accessed. Any adaptive attention window scheme can be implemented that way.

Use the following settings when the attention window should be the maximum size, for example, in cross-attention:
currIdx=0: loWinIdx[0]=0; hiWinIdx[0]=maxSeqLenK;
currIdx=1: loWinIdx[1]=0; hiWinIdx[1]=maxSeqLenK;
currIdx=2: loWinIdx[2]=0; hiWinIdx[2]=maxSeqLenK;
(...)
The maxSeqLenK value above should be equal to or larger than dimA[CUDNN_SEQDATA_TIME_DIM] in the kDesc descriptor. A good choice is to use maxSeqLenK=INT_MAX from limits.h.
Note: The actual length of any K sequence defined in seqLengthArray[] in cudnnSetSeqDataDescriptor() can be shorter than maxSeqLenK. The effective attention window span is computed based on seqLengthArray[] stored in the K sequence descriptor and indices held in loWinIdx[] and hiWinIdx[] arrays.

devSeqLengthsQO[] and devSeqLengthsKV[] are pointers to device (not host) arrays with Q, O, and K, V sequence lengths. Note that the same information is also passed in the corresponding descriptors of type cudnnSeqDataDescriptor_t on the host side. The need for extra device arrays comes from the asynchronous nature of cuDNN calls and limited size of the constant memory dedicated to GPU kernel arguments. When the cudnnMultiHeadAttnForward() API returns, the sequence length arrays stored in the descriptors can be immediately modified for the next iteration. However, the GPU kernels launched by the forward call may not have started at this point. For this reason, copies of sequence arrays are needed on the device side to be accessed directly by GPU kernels. Those copies cannot be created inside the cudnnMultiHeadAttnForward() function for very large K, V inputs without the device memory allocation and CUDA stream synchronization.

To reduce the cudnnMultiHeadAttnForward() API overhead, devSeqLengthsQO[] and devSeqLengthsKV[] device arrays are not validated to contain the same settings as seqLengthArray[] in the sequence data descriptors.

Sequence lengths in the kDesc and vDesc descriptors should be the same. Similarly, sequence lengths in the qDesc and oDesc descriptors should match. The user can define six different data layouts in the qDesc, kDesc, vDesc and oDesc descriptors. See the cudnnSetSeqDataDescriptor() function for the discussion of those layouts. All multi-head attention API calls require that the same layout is used in all sequence data descriptors.

In the transformer model, the multi-head attention block is tightly coupled with the layer normalization and residual connections. cudnnMultiHeadAttnForward() does not encompass the layer normalization but it can be used to handle residual connections as depicted in the following figure.

Queries and residuals share the same qDesc descriptor in cudnnMultiHeadAttnForward(). When residual connections are disabled, the residuals pointer should be NULL. When residual connections are enabled, the vector length in qDesc should match the vector length specified in the oDesc descriptor, so that a vector addition is feasible.

The queries, keys, and values pointers are not allowed to be NULL, even when K and V are the same inputs or Q, K, V are the same inputs.

Parameters

handle

Input. The current cuDNN context handle.

attnDesc

Input. A previously initialized attention descriptor.

currIdx

Input. Time-step in queries to process. When the currIdx argument is negative, all Q time-steps are processed. When currIdx is zero or positive, the forward response is computed for the selected time-step only. The latter input can be used in inference mode only, to process one time-step while updating the next attention window and Q, R, K, V inputs in-between calls.

loWinIdx[], hiWinIdx[]

Input. Two host integer arrays specifying the start and end indices of the attention window for each Q time-step. The start index in K, V sets is inclusive, and the end index is exclusive.

devSeqLengthsQO[]

Input. Device array specifying sequence lengths of query, residual, and output sequence data.

devSeqLengthsKV[]

Input. Device array specifying sequence lengths of key and value input data.

qDesc

Input. Descriptor for the query and residual sequence data.

queries

Input. Pointer to queries data in the device memory.

residuals

Input. Pointer to residual data in device memory. Set this argument to NULL if no residual connections are required.

kDesc

Input. Descriptor for the keys sequence data.

keys

Input. Pointer to keys data in device memory.

vDesc

Input. Descriptor for the values sequence data.

values

Input. Pointer to values data in device memory.

oDesc

Input. Descriptor for the multi-head attention output sequence data.

out

Output. Pointer to device memory where the output response should be written.

weightSizeInBytes

Input. Size of the weight buffer in bytes where all multi-head attention trainable parameters are stored.

weights

Input. Pointer to the weight buffer in device memory.

workSpaceSizeInBytes

Input. Size of the work-space buffer in bytes used for temporary API storage.

workSpace

Input/Output. Pointer to the work-space buffer in device memory.

reserveSpaceSizeInBytes

Input. Size of the reserve-space buffer in bytes used for data exchange between forward and backward (gradient) API calls. This parameter should be zero in the inference mode and non-zero in the training mode.

reserveSpace

Input/Output. Pointer to the reserve-space buffer in device memory. This argument should be NULL in inference mode and non-NULL in the training mode.

Returns

CUDNN_STATUS_SUCCESS

No errors were detected while processing API input arguments and launching GPU kernels.

CUDNN_STATUS_BAD_PARAM
An invalid or incompatible input argument was encountered. Some examples include:
  • a required input pointer was NULL
  • currIdx was out of bound
  • the descriptor value for attention, query, key, value, and output were incompatible with one another
CUDNN_STATUS_EXECUTION_FAILED

The process of launching a GPU kernel returned an error, or an earlier kernel did not complete successfully.

CUDNN_STATUS_INTERNAL_ERROR

An inconsistent internal state was encountered.

CUDNN_STATUS_NOT_SUPPORTED

A requested option or a combination of input arguments is not supported.

CUDNN_STATUS_ALLOC_FAILED

Insufficient amount of shared memory to launch a GPU kernel.

7.2.36. cudnnRNNForward()

cudnnStatus_t cudnnRNNForward(
    cudnnHandle_t handle,
    cudnnRNNDescriptor_t rnnDesc,
    cudnnForwardMode_t fwdMode,
    const int32_t devSeqLengths[],
    cudnnRNNDataDescriptor_t xDesc,
    const void *x,
    cudnnRNNDataDescriptor_t yDesc,
    void *y,
    cudnnTensorDescriptor_t hDesc,
    const void *hx,
    void *hy,
    cudnnTensorDescriptor_t cDesc,
    const void *cx,
    void *cy,
    size_t weightSpaceSize,
    const void *weightSpace,
    size_t workSpaceSize,
    void *workSpace,
    size_t reserveSpaceSize,
    void *reserveSpace);

This routine computes the forward response of the recurrent neural network described by rnnDesc with inputs in x, hx, cx, and weights/biases in the weightSpace buffer. The outputs are written to y, hy, and cy. This function is able to store intermediate data required for training when the fwdMode parameter is set to CUDNN_FWD_MODE_TRAINING. Note that work and reserve space buffer sizes should be computed by the cudnnGetRNNTempSpaceSizes() function with the same fwdMode settings.

Parameters

handle

Input. The current cuDNN context handle.

rnnDesc
Input. A previously initialized RNN descriptor.
fwdMode

Input. Specifies inference or training mode (CUDNN_FWD_MODE_INFERENCE and CUDNN_FWD_MODE_TRAINING). In the training mode, additional data is stored in the reserve space buffer. This information is used in the backward pass to compute derivatives.

devSeqLengths

Input. A copy of seqLengthArray from xDesc or yDesc RNN data descriptor. The devSeqLengths array must be stored in GPU memory as it is read directly by GPU kernels.

xDesc

Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength, batchSize, and seqLengthArray need to match that of yDesc.

x

Input. Data pointer to the GPU memory associated with the RNN data descriptor xDesc. The vectors are expected to be laid out in memory according to the layout specified by xDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

yDesc
Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength, batchSize, and seqLengthArray must match that of xDesc. The parameter vectorSize depends on whether LSTM projection is enabled and whether the network is bidirectional. Specifically:
  • For unidirectional models, the parameter vectorSize must match the hiddenSize argument passed to cudnnSetRNNDescriptor_v8(). If the LSTM projection is enabled, the vectorSize must be the same as the projSize argument passed to cudnnSetRNNDescriptor_v8().
  • For bidirectional models, if the RNN cellMode is CUDNN_LSTM and the projection feature is enabled, the parameter vectorSize must be 2x the projSize argument passed to cudnnSetRNNDescriptor_v8(). Otherwise, it should be 2x the hiddenSize value.
y

Output. Data pointer to the GPU memory associated with the RNN data descriptor yDesc. The vectors are expected to be laid out in memory according to the layout specified by yDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

hDesc
Input. A tensor descriptor describing the initial or final hidden state of RNN. Hidden state data are fully packed. The first dimension of the tensor depends on the dirMode argument passed to the cudnnSetRNNDescriptor_v8() function.
  • If dirMode is CUDNN_UNIDIRECTIONAL, then the first dimension should match the numLayers argument passed to cudnnSetRNNDescriptor_v8().
  • If dirMode is CUDNN_BIDIRECTIONAL, then the first dimension should be double the numLayers argument passed to cudnnSetRNNDescriptor_v8().
The second dimension must match the batchSize parameter described in xDesc. The third dimension depends on whether RNN mode is CUDNN_LSTM and whether the LSTM projection is enabled. Specifically:
  • If RNN mode is CUDNN_LSTM and LSTM projection is enabled, the third dimension must match the projSize argument passed to the cudnnSetRNNProjectionLayers() call.
  • Otherwise, the third dimension must match the hiddenSize argument passed to the cudnnSetRNNDescriptor_v8() call used to initialize rnnDesc.
hx

Input. Pointer to the GPU buffer with the RNN initial hidden state. Data dimensions are described by the hDesc tensor descriptor. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

hy

Output. Pointer to the GPU buffer where the final RNN hidden state should be stored. Data dimensions are described by the hDesc tensor descriptor. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cDesc
Input. A tensor descriptor describing the initial or final cell state for LSTM networks only. Cell state data are fully packed. The first dimension of the tensor depends on the dirMode argument passed to the cudnnSetRNNDescriptor_v8() call.
  • If dirMode is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument passed to cudnnSetRNNDescriptor_v8().
  • If dirMode is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument passed to cudnnSetRNNDescriptor_v8().

The second tensor dimension must match the batchSize parameter in xDesc. The third dimension must match the hiddenSize argument passed to the cudnnSetRNNDescriptor_v8() call.

cx

Input. Pointer to the GPU buffer with the initial LSTM state data. Data dimensions are described by the cDesc tensor descriptor. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

cy

Output. Pointer to the GPU buffer where final LSTM state data should be stored. Data dimensions are described by the cDesc tensor descriptor. If a NULL pointer is passed, the final LSTM cell state will not be saved.

weightSpaceSize

Input. Specifies the size in bytes of the provided weight-space buffer.

weightSpace

Input. Address of the weight space buffer in GPU memory.

workSpaceSize

Input. Specifies the size in bytes of the provided workspace buffer.

workSpace

Input/Output. Address of the workspace buffer in GPU memory to store temporary data.

reserveSpaceSize

Input. Specifies the size in bytes of the reserve-space buffer.

reserveSpace

Output. Address of the reserve-space buffer in GPU memory.

Returns

CUDNN_STATUS_SUCCESS

No errors were detected while processing API input arguments and launching GPU kernels.

CUDNN_STATUS_NOT_SUPPORTED
At least one of the following conditions are met:
  • variable sequence length input is passed while CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is specified
  • CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is requested on pre-Pascal devices
  • the 'double' floating point type is used for input/output and the CUDNN_RNN_ALGO_PERSIST_STATIC algo
CUDNN_STATUS_BAD_PARAM
An invalid or incompatible input argument was encountered. For example:
  • some input descriptors are NULL
  • at least one of the settings in rnnDesc, xDesc, yDesc, hDesc, or cDesc descriptors is invalid
  • weightSpaceSize, workSpaceSize, or reserveSpaceSize is too small
CUDNN_STATUS_EXECUTION_FAILED

The process of launching a GPU kernel returned an error, or an earlier kernel did not complete successfully.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate CPU memory.

7.2.37. cudnnRNNForwardInference()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNForward() instead of cudnnRNNForwardInference().

cudnnStatus_t cudnnRNNForwardInference(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    const void                     *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                     *cx,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const cudnnTensorDescriptor_t   *yDesc,
    void                           *y,
    const cudnnTensorDescriptor_t   hyDesc,
    void                           *hy,
    const cudnnTensorDescriptor_t   cyDesc,
    void                           *cy,
    void                           *workspace,
    size_t                          workSpaceSizeInBytes)

This routine executes the recurrent neural network described by rnnDesc with inputs x, hx, and cx, weights w and outputs y, hy, and cy. workspace is required for intermediate storage. This function does not store intermediate data required for training; cudnnRNNForwardTraining() should be used for that purpose.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc
Input. An array of seqLength fully packed tensor descriptors. Each descriptor in the array should have three dimensions that describe the input data format to one recurrent iteration (one descriptor per RNN time-step). The first dimension (batch size) of the tensors may decrease from iteration n to iteration n+1 but may not increase. Each tensor descriptor must have the same second dimension (RNN input vector length, inputSize). The third dimension of each tensor should be 1. Input data are expected to be arranged in the column-major order so strides in xDesc should be set as follows:
strideA[0]=inputSize, strideA[1]=1, strideA[2]=1
x

Input. Data pointer to GPU memory associated with the array of tensor descriptors xDesc. The input vectors are expected to be packed contiguously with the first vector of iteration (time-step) n+1 following directly from the last vector of iteration n. In other words, input vectors for all RNN time-steps should be packed in the contiguous block of GPU memory with no gaps between the vectors.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in xDesc.
y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc. The data are expected to be packed contiguously with the first element of iteration n+1 following directly from the last element of iteration n.

hyDesc
Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc
Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, cxDesc, wDesc, hyDesc, cyDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, cxDesc, wDesc, yDesc, hyDesc, cyDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

7.2.38. cudnnRNNForwardInferenceEx()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNForward() instead of cudnnRNNForwardInference().

cudnnStatus_t cudnnRNNForwardInferenceEx(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const cudnnRNNDataDescriptor_t  xDesc,
    const void                      *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                      *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                      *cx,
    const cudnnFilterDescriptor_t   wDesc,
    const void                      *w,
    const cudnnRNNDataDescriptor_t  yDesc,
    void                            *y,
    const cudnnTensorDescriptor_t   hyDesc,
    void                            *hy,
    const cudnnTensorDescriptor_t   cyDesc,
    void                            *cy,
    const cudnnRNNDataDescriptor_t  kDesc,
    const void                      *keys,
    const cudnnRNNDataDescriptor_t  cDesc,
    void                            *cAttn,
    const cudnnRNNDataDescriptor_t  iDesc,
    void                            *iAttn,
    const cudnnRNNDataDescriptor_t  qDesc,
    void                            *queries,
    void                            *workSpace,
    size_t                          workSpaceSizeInBytes)

This routine is the extended version of the cudnnRNNForwardInference() function. The cudnnRNNForwardTrainingEx() function allows the user to use an unpacked (padded) layout for input x and output y. In the unpacked layout, each sequence in the mini-batch is considered to be of fixed length, specified by maxSeqLength in its corresponding RNNDataDescriptor. Each fixed-length sequence, for example, the nth sequence in the mini-batch, is composed of a valid segment, specified by the seqLengthArray[n] in its corresponding RNNDataDescriptor, and a padding segment to make the combined sequence length equal to maxSeqLength.

With unpacked layout, both sequence major (meaning, time major) and batch major are supported. For backward compatibility, the packed sequence major layout is supported. However, similar to the non-extended function cudnnRNNForwardInference(), the sequences in the mini-batch need to be sorted in descending order according to length.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

xDesc

Input. A previously initialized RNN Data descriptor. The dataType, layout, maxSeqLength, batchSize, and seqLengthArray need to match that of yDesc.

x

Input. Data pointer to the GPU memory associated with the RNN data descriptor xDesc. The vectors are expected to be laid out in memory according to the layout specified by xDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.

The second dimension must match the batchSize parameter described in xDesc. The third dimension depends on whether RNN mode is CUDNN_LSTM and whether LSTM projection is enabled. Specifically:

  • If RNN mode is CUDNN_LSTM and LSTM projection is enabled, the third dimension must match the recProjSize argument passed to cudnnSetRNNProjectionLayers() call used to set rnnDesc.
  • Otherwise, the third dimension must match the hiddenSize argument used to initialize rnnDesc.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.

The second dimension must match the batchSize parameter in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc.

cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc

Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength , batchSize, and seqLengthArray must match that of dyDesc and dxDesc. The parameter vectorSize depends on whether RNN mode is CUDNN_LSTM and whether LSTM projection is enabled and whether the network is bidirectional. Specifically:

  • For unidirectional network, if the RNN mode is CUDNN_LSTM and LSTM projection is enabled, the parameter vectorSize must match the recProjSize argument passed to cudnnSetRNNProjectionLayers() call used to set rnnDesc. If the network is bidirectional, then multiply the value by 2.
  • Otherwise, for a unidirectional network, the parameter vectorSize must match the hiddenSize argument used to initialize rnnDesc. If the network is bidirectional, then multiply the value by 2.
y

Output. Data pointer to the GPU memory associated with the RNN data descriptor yDesc. The vectors are expected to be laid out in memory according to the layout specified by yDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

hyDesc

Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The descriptor must be set exactly the same way as hxDesc.

hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc

Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The descriptor must be set exactly the same way as cxDesc.

cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

kDesc

Reserved. User may pass in NULL.

keys

Reserved. Users may pass in NULL.

cDesc

Reserved. Users may pass in NULL.

cAttn

Reserved. Users may pass in NULL.

iDesc

Reserved. Users may pass in NULL.

iAttn

Reserved. Users may pass in NULL.

qDesc

Reserved. Users may pass in NULL.

queries

Reserved. Users may pass in NULL.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

At least one of the following conditions are met:

  • Variable sequence length input is passed in while CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used.
  • CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used on pre-Pascal devices.
  • Double input/output is used for CUDNN_RNN_ALGO_PERSIST_STATIC.
CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors in xDesc, yDesc, hxDesc, cxDesc, wDesc, hyDesc, cyDesc is invalid, or have incorrect strides or dimensions.
  • reserveSpaceSizeInBytes is too small.
  • workSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

7.2.39. cudnnRNNGetClip()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNGetClip_v8() instead of cudnnRNNGetClip().

cudnnStatus_t cudnnRNNGetClip(
    cudnnHandle_t               handle,
    cudnnRNNDescriptor_t        rnnDesc,
    cudnnRNNClipMode_t          *clipMode,
    cudnnNanPropagation_t       *clipNanOpt,
    double                      *lclip,
    double                      *rclip);

Retrieves the current LSTM cell clipping parameters, and stores them in the arguments provided.

Parameters

*clipMode

Output. Pointer to the location where the retrieved clipMode is stored. The clipMode can be CUDNN_RNN_CLIP_NONE in which case no LSTM cell state clipping is being performed; or CUDNN_RNN_CLIP_MINMAX, in which case the cell state activation to other units are being clipped.

*lclip, *rclip

Output. Pointers to the location where the retrieved LSTM cell clipping range [lclip, rclip] is stored.

*clipNanOpt

Output. Pointer to the location where the retrieved clipNanOpt is stored.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

If any of the pointer arguments provided are NULL.

7.2.40. cudnnRNNGetClip_v8()

cudnnStatus_t cudnnRNNGetClip_v8(
	cudnnRNNDescriptor_t rnnDesc,
	cudnnRNNClipMode_t *clipMode,
	cudnnNanPropagation_t *clipNanOpt,
	double *lclip,
	double *rclip);

Retrieves the current LSTM cell clipping parameters, and stores them in the arguments provided. The user can assign NULL to any pointer except rnnDesc when the retrieved value is not needed. The function does not check the validity of retrieved parameters.

Parameters

rnnDesc

Input. A previously initialized RNN descriptor.

clipMode
Output. Pointer to the location where the retrieved cudnnRNNClipMode_t value is stored. The clipMode can be CUDNN_RNN_CLIP_NONE in which case no LSTM cell state clipping is being performed; or CUDNN_RNN_CLIP_MINMAX, in which case the cell state activation to other units are being clipped.
clipNanOpt

Output. Pointer to the location where the retrieved cudnnNanPropagation_t value is stored.

lclip, rclip

Output. Pointers to the location where the retrieved LSTM cell clipping range [lclip, rclip] is stored.

Returns

CUDNN_STATUS_SUCCESS

LSTM clipping parameters were successfully retrieved from the RNN descriptor.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was found (rnnDesc was NULL).

7.2.41. cudnnRNNSetClip()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNSetClip_v8() instead of cudnnRNNSetClip().

cudnnStatus_t cudnnRNNSetClip(
    cudnnHandle_t               handle,
    cudnnRNNDescriptor_t        rnnDesc,
    cudnnRNNClipMode_t          clipMode,
    cudnnNanPropagation_t       clipNanOpt,
    double                      lclip,
    double                      rclip);

Sets the LSTM cell clipping mode. The LSTM clipping is disabled by default. When enabled, clipping is applied to all layers. This cudnnRNNSetClip() function may be called multiple times.

Parameters

clipMode

Input. Enables or disables the LSTM cell clipping. When clipMode is set to CUDNN_RNN_CLIP_NONE no LSTM cell state clipping is performed. When clipMode is CUDNN_RNN_CLIP_MINMAX the cell state activation to other units is clipped.

lclip, rclip

Input. The range [lclip, rclip] to which the LSTM cell clipping should be set.

clipNanOpt

Input. When set to CUDNN_PROPAGATE_NAN (see the description for cudnnNanPropagation_t), NaN is propagated from the LSTM cell, or it can be set to one of the clipping range boundary values, instead of propagating.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

Returns this value if lclip > rclip; or if either lclip or rclip is NaN.

7.2.42. cudnnRNNSetClip_v8()

cudnnStatus_t cudnnRNNSetClip_v8(
	cudnnRNNDescriptor_t rnnDesc,
	cudnnRNNClipMode_t clipMode,
	cudnnNanPropagation_t clipNanOpt,
	double lclip,
	double rclip);

Sets the LSTM cell clipping mode. The LSTM clipping is disabled by default. When enabled, clipping is applied to all layers. This cudnnRNNSetClip() function does not affect the work, reserve, and weight-space buffer sizes and may be called multiple times.

Parameters

rnnDesc

Input. A previously initialized RNN descriptor.

clipMode
Input. Enables or disables the LSTM cell clipping. When clipMode is set to CUDNN_RNN_CLIP_NONE no LSTM cell state clipping is performed. When clipMode is CUDNN_RNN_CLIP_MINMAX the cell state activation to other units is clipped.
clipNanOpt

Input. When set to CUDNN_PROPAGATE_NAN (see the description for cudnnNanPropagation_t), NaN is propagated from the LSTM cell, or it can be set to one of the clipping range boundary values, instead of propagating.

lclip, rclip

Input. The range [lclip, rclip] to which the LSTM cell clipping should be set.

Returns

CUDNN_STATUS_SUCCESS

The function completed successfully.

CUDNN_STATUS_BAD_PARAM
An invalid input argument was found, for example:
  • rnnDesc was NULL
  • lclip > rclip
  • either lclip or rclip is NaN
CUDNN_STATUS_BAD_PARAM

The dimensions of the bias tensor refer to an amount of data that is incompatible with the output tensor dimensions or the dataType of the two tensor descriptors are different.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

7.2.43. cudnnSetAttnDescriptor()

cudnnStatus_t cudnnSetAttnDescriptor(
	cudnnAttnDescriptor_t attnDesc,
	unsigned attnMode,
	int nHeads,
	double smScaler,
	cudnnDataType_t dataType,
	cudnnDataType_t computePrec,
	cudnnMathType_t mathType,
	cudnnDropoutDescriptor_t attnDropoutDesc,
	cudnnDropoutDescriptor_t postDropoutDesc,
	int qSize,
	int kSize,
	int vSize,
	int qProjSize,
	int kProjSize,
	int vProjSize,
	int oProjSize,
	int qoMaxSeqLength,
	int kvMaxSeqLength,
	int maxBatchSize,
	int maxBeamSize);

This function configures a multi-head attention descriptor that was previously created using the cudnnCreateAttnDescriptor() function. The function sets attention parameters that are necessary to compute internal buffer sizes, dimensions of weight and bias tensors, or to select optimized code paths.

Input sequence data descriptors in cudnnMultiHeadAttnForward(), cudnnMultiHeadAttnBackwardData() and cudnnMultiHeadAttnBackwardWeights() functions are checked against the configuration parameters stored in the attention descriptor. Some parameters must match exactly while max arguments such as maxBatchSize or qoMaxSeqLength establish upper limits for the corresponding dimensions.

The multi-head attention model can be described by the following equations:

h i = W V , i V softmax smScaler K T W K , i T W Q , i q ,   for   i   = 0   ...   nHeads - 1

MultiHeadAttn q , K , V , W Q , W K , W V , W O = i = 0 nHeads - 1 W O , i h i

Where:
  • nHeads is the number of independent attention heads that evaluate h i vectors.
  • q is a primary input, a single query column vector.
  • K , V are two matrices of key and value column vectors.

For simplicity, the above equations are presented using a single embedding vector q but the API can handle multiple q candidates in the beam search scheme, process q vectors from multiple sequences bundled into a batch, or automatically iterate through all embedding vectors (time-steps) of a sequence. Thus, in general, q , K , V inputs are tensors with additional pieces of information such as the active length of each sequence or how unused padding vectors should be saved.

In some publications, W O , i matrices are combined into one output projection matrix and h i vectors are merged explicitly into a single vector. This is an equivalent notation. In the library, W O , i matrices are conceptually treated the same way as W Q , i , W K , i or W V , i input projection weights. See the description of the cudnnGetMultiHeadAttnWeights() function for more details.

Weight matrices W Q , i , W K , i , W V , i and W O , i play similar roles, adjusting vector lengths in q , K , V inputs and in the multi-head attention final output. The user can disable any or all projections by setting qProjSize , kProjSize , vProjSize or oProjSize arguments to zero.

Embedding vector sizes in q , K , V and the vector lengths after projections need to be selected in such a way that matrix multiplications described above are feasible. Otherwise, CUDNN_STATUS_BAD_PARAM is returned by the cudnnSetAttnDescriptor() function. All four weight matrices are used when it is desirable to maintain rank deficiency of W K Q , i = W K , i T W Q, i or W OV , i = W O , i W V , i matrices to eliminate one or more dimensions during linear transformations in each head. This is a form of feature extraction. In such cases, the projected sizes are smaller than the original vector lengths.

For each attention head, weight matrix sizes are defined as follows:
  • W Q , i - size qProjSize   x   qSize ,   i = O   ..   nHeads - 1
  • W K , i - size kProjSize   x   kSize ,   i = 0   ..   nHeads - 1 ,   kProjSize = qProjSize
  • W V , i - size vProjSize   x   vSize ,   i = 0   ..   nHeads - 1
  • W O , i - size oProjSize   x   vProjSize   >   0   ?   vProjSize   :   vSize ,   i = 0   ..   nHeads - 1

When the output projection is disabled oProjSize = 0 , the output vector length is nHead *   vProjSize   >   0   ?   vProjSize   :   vSize , meaning, the output is a concatenation of all h i vectors. In the alternative interpretation, a concatenated matrix W O = W O , 0 ,   W O , 1 ,   W O , 2 ,   ... forms the identity matrix.

Softmax is a normalized, exponential vector function that takes and outputs vectors of the same size. The multi-head attention API utilizes softmax of the CUDNN_SOFTMAX_ACCURATE type to reduce the likelihood of the floating-point overflow.

The smScaler parameter is the softmax sharpening/smoothing coefficient. When smScaler=1.0, softmax uses the natural exponential function exp(x) or 2.7183*. When smScaler<1.0, for example smScaler=0.2, the function used by the softmax block will not grow as fast because exp(0.2*x) ≈ 1.2214*.

The smScaler parameter can be adjusted to process larger ranges of values fed to softmax. When the range is too large (or smScaler is not sufficiently small for the given range), the output vector of the softmax block becomes categorical, meaning, one vector element is close to 1.0 and other outputs are zero or very close to zero. When this occurs, the Jacobian matrix of the softmax block is also close to zero so deltas are not back-propagated during training from output to input except through residual connections, if these connections are enabled. The user can set smScaler to any positive floating-point value or even zero. The smScaler parameter is not trainable.

The qoMaxSeqLength, kvMaxSeqLength, maxBatchSize, and maxBeamSize arguments declare the maximum sequence lengths, maximum batch size, and maximum beam size respectively, in the cudnnSeqDataDescriptor_t containers. The actual dimensions supplied to forward and backward (gradient) API functions should not exceed the max limits. The max arguments should be set carefully because too large values will result in excessive memory usage due to oversized work and reserve space buffers.

The attnMode argument is treated as a binary mask where various on/off options are set. These options can affect the internal buffer sizes, enforce certain argument checks, select optimized code execution paths, or enable attention variants that do not require additional numerical arguments. An example of such options is the inclusion of biases in input and output projections.

The attnDropoutDesc and postDropoutDesc arguments are descriptors that define two dropout layers active in the training mode. The first dropout operation defined by attnDropoutDesc, is applied directly to the softmax output. The second dropout operation, specified by postDropoutDesc, alters the multi-head attention output, just before the point where residual connections are added.
Note: The cudnnSetAttnDescriptor() function performs a shallow copy of attnDropoutDesc and postDropoutDesc, meaning, the addresses of both dropout descriptors are stored in the attention descriptor and not the entire structure. Therefore, the user should keep dropout descriptors during the entire life of the attention descriptor.

Parameters

attnDesc

Output. Attention descriptor to be configured.

attnMode

Input. Enables various attention options that do not require additional numerical values. See the table below for the list of supported flags. The user should assign a preferred set of bitwise OR-ed flags to this argument.

nHeads

Input. Number of attention heads.

smScaler

Input. Softmax smoothing (1.0 >= smScaler >= 0.0) or sharpening (smScaler > 1.0) coefficient. Negative values are not accepted.

dataType

Input. Data type used to represent attention inputs, attention weights and attention outputs.

computePrec

Input. Compute precision.

mathType

Input. NVIDIA Tensor Core settings.

attnDropoutDesc

Input. Descriptor of the dropout operation applied to the softmax output. See the table below for a list of unsupported features.

postDropoutDesc

Input. Descriptor of the dropout operation applied to the multi-head attention output, just before the point where residual connections are added. See the table below for a list of unsupported features.

qSize, kSize, vSize

Input. Q , K , V embedding vector lengths.

qProjSize, kProjSize, vProjSize

Input. Q , K , V embedding vector lengths after input projections. Use zero to disable the corresponding projection.

oProjSize

Input. The h i vector length after the output projection. Use zero to disable this projection.

qoMaxSeqLength

Input. Largest sequence length expected in sequence data descriptors related to Q , O , dQ and dO inputs and outputs.

kvMaxSeqLength

Input. Largest sequence length expected in sequence data descriptors related to K , V , dK and dV inputs and outputs.

maxBatchSize

Input. Largest batch size expected in any cudnnSeqDataDescriptor_t container.

maxBeamSize

Input. Largest beam size expected in any cudnnSeqDataDescriptor_t container.

Supported attnMode flags

CUDNN_ATTN_QUERYMAP_ALL_TO_ONE

Forward declaration of mapping between Q and K , V vectors when the beam size is greater than one in the Q input. Multiple Q vectors from the same beam bundle map to the same K , V vectors. This means that beam sizes in the K , V sets are equal to one.

CUDNN_ATTN_QUERYMAP_ONE_TO_ONE

Forward declaration of mapping between Q and K , V vectors when the beam size is greater than one in the Q input. Multiple Q vectors from the same beam bundle map to different K , V vectors. This requires beam sizes in K , V sets to be the same as in the Q input.

CUDNN_ATTN_DISABLE_PROJ_BIASES

Use no biases in the attention input and output projections.

CUDNN_ATTN_ENABLE_PROJ_BIASES

Use extra biases in the attention input and output projections. In this case the projected K ¯ vectors are computed as K i ¯ = W K , i K + b * 1 , 1 , ..., 1 1 × n , where n is the number of columns in the K matrix. In other words, the same column vector b is added to all columns of K after the weight matrix multiplication.

Supported combinations of dataType, computePrec, and mathType

Table 37. Supported combinations
dataType computePrec mathType
CUDNN_DATA_DOUBLE CUDNN_DATA_DOUBLE CUDNN_DEFAULT_MATH
CUDNN_DATA_FLOAT CUDNN_DATA_FLOAT CUDNN_DEFAULT_MATH, CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION
CUDNN_DATA_HALF CUDNN_DATA_HALF CUDNN_DEFAULT_MATH, CUDNN_TENSOR_OP_MATH, CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION

Unsupported features

  1. The dropout option is currently not supported by the multi-head attention API. Assign NULL to attnDropoutDesc and postDropoutDesc arguments when configuring the attention descriptor.
  2. The CUDNN_ATTN_ENABLE_PROJ_BIASES option is not supported in the multi-head attention gradient functions.
  3. The paddingFill argument in cudnnSeqDataDescriptor_t is currently ignored by all multi-head attention functions.

Returns

CUDNN_STATUS_SUCCESS

The attention descriptor was configured successfully.

CUDNN_STATUS_BAD_PARAM
An invalid input argument was encountered. Some examples include:
  • post projection Q and K sizes were not equal
  • dataType, computePrec, or mathType were invalid
  • one or more of the following arguments were either negative or zero: nHeads, qSize, kSize, vSize, qoMaxSeqLength, kvMaxSeqLength, maxBatchSize, maxBeamSize
  • one or more of the following arguments were negative: qProjSize, kProjSize, vProjSize, smScaler
CUDNN_STATUS_NOT_SUPPORTED

A requested option or a combination of input arguments is not supported.

7.2.44. cudnnSetPersistentRNNPlan()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnSetPersistentRNNPlan(
    cudnnRNNDescriptor_t        rnnDesc,
    cudnnPersistentRNNPlan_t    plan)

This function sets the persistent RNN plan to be executed when using rnnDesc and CUDNN_RNN_ALGO_PERSIST_DYNAMIC algo.

Returns

CUDNN_STATUS_SUCCESS

The plan was set successfully.

CUDNN_STATUS_BAD_PARAM

The algo selected in rnnDesc is not CUDNN_RNN_ALGO_PERSIST_DYNAMIC.

7.2.45. cudnnSetRNNAlgorithmDescriptor()

This function has been deprecated in cuDNN 8.0.

7.2.46. cudnnSetRNNBiasMode()

This function has been deprecated in cuDNN 8.0. Use cudnnSetRNNDescriptor_v8() instead of cudnnSetRNNBiasMode().

cudnnStatus_t cudnnSetRNNBiasMode(
   cudnnRNNDescriptor_t   rnnDesc, 
   cudnnRNNBiasMode_t     biasMode)

The cudnnSetRNNBiasMode() function sets the number of bias vectors for a previously created and initialized RNN descriptor. This function should be called to enable the specified bias mode in an RNN. The default value of biasMode in rnnDesc after cudnnCreateRNNDescriptor() is CUDNN_RNN_DOUBLE_BIAS.

Parameters

rnnDesc

Input/Output. A previously created RNN descriptor.

biasMode

Input. Sets the number of bias vectors. For more information, see cudnnRNNBiasMode_t.

Returns

CUDNN_STATUS_BAD_PARAM

Either the rnnDesc is NULL or biasMode has an invalid enumerant value.

CUDNN_STATUS_SUCCESS

The biasMode was set successfully.

CUDNN_STATUS_NOT_SUPPORTED

Non-default bias mode (an enumerated type besides CUDNN_RNN_DOUBLE_BIAS) applied to an RNN algo other than CUDNN_RNN_ALGO_STANDARD.

7.2.47. cudnnSetRNNDataDescriptor()

cudnnStatus_t cudnnSetRNNDataDescriptor(
    cudnnRNNDataDescriptor_t       RNNDataDesc,
    cudnnDataType_t                dataType,
    cudnnRNNDataLayout_t           layout,
    int                            maxSeqLength,
    int                            batchSize,
    int                            vectorSize,
    const int                      seqLengthArray[],
    void                           *paddingFill);

This function initializes a previously created RNN data descriptor object. This data structure is intended to support the unpacked (padded) layout for input and output of extended RNN inference and training functions. A packed (unpadded) layout is also supported for backward compatibility.

Parameters

RNNDataDesc

Input/Output. A previously created RNN descriptor. For more information, see cudnnRNNDataDescriptor_t.

dataType

Input. The datatype of the RNN data tensor. For more information, see cudnnDataType_t.

layout

Input. The memory layout of the RNN data tensor.

maxSeqLength

Input. The maximum sequence length within this RNN data tensor. In the unpacked (padded) layout, this should include the padding vectors in each sequence. In the packed (unpadded) layout, this should be equal to the greatest element in seqLengthArray.

batchSize

Input. The number of sequences within the mini-batch.

vectorSize

Input. The vector length (embedding size) of the input or output tensor at each time-step.

seqLengthArray

Input. An integer array with batchSize number of elements. Describes the length (number of time-steps) of each sequence. Each element in seqLengthArray must be greater than 0 but less than or equal to maxSeqLength. In the packed layout, the elements should be sorted in descending order, similar to the layout required by the non-extended RNN compute functions.

paddingFill

Input. A user-defined symbol for filling the padding position in RNN output. This is only effective when the descriptor is describing the RNN output, and the unpacked layout is specified. The symbol should be in the host memory, and is interpreted as the same data type as that of the RNN data tensor. If a NULL pointer is passed in, then the padding position in the output will be undefined.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_NOT_SUPPORTED

dataType is not one of CUDNN_DATA_HALF, CUDNN_DATA_FLOAT or CUDNN_DATA_DOUBLE.

CUDNN_STATUS_BAD_PARAM

Any one of these have occurred:

  • RNNDataDesc is NULL.
  • Any one of maxSeqLength, batchSize or vectorSize is less than or equal to zero.
  • An element of seqLengthArray is less than or equal to zero or greater than maxSeqLength.
  • Layout is not one of CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_UNPACKED, CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_PACKED or CUDNN_RNN_DATA_LAYOUT_BATCH_MAJOR_UNPACKED.
CUDNN_STATUS_ALLOC_FAILED
The allocation of internal array storage has failed.

7.2.48. cudnnSetRNNDescriptor_v6()

This function has been deprecated in cuDNN 8.0. Use cudnnSetRNNDescriptor_v8() instead of cudnnSetRNNDescriptor_v6().

cudnnStatus_t cudnnSetRNNDescriptor_v6(
	cudnnHandle_t                    handle,
	cudnnRNNDescriptor_t             rnnDesc,
	const int                        hiddenSize,
	const int                        numLayers,
	cudnnDropoutDescriptor_t         dropoutDesc,
	cudnnRNNInputMode_t              inputMode,
	cudnnDirectionMode_t             direction,
	cudnnRNNMode_t                   mode,
	cudnnRNNAlgo_t                   algo,
	cudnnDataType_t                  mathPrec)

This function initializes a previously created RNN descriptor object.

Note: Larger networks, for example, longer sequences or more layers, are expected to be more efficient than smaller networks.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input/Output. A previously created RNN descriptor.

hiddenSize

Input. Size of the internal hidden state for each layer.

numLayers

Input. Number of stacked layers.

dropoutDesc

Input. Handle to a previously created and initialized dropout descriptor. Dropout will be applied between layers, for example, a single layer network will have no dropout applied.

inputMode

Input. Specifies the behavior at the input to the first layer

direction

Input. Specifies the recurrence pattern, for example, bidirectional.

mode

Input. Specifies the type of RNN to compute.

algo

Input. Specifies which RNN algorithm should be used to compute the results.

mathPrec
Input. Math precision. This parameter is used for controlling the math precision in RNN. The following applies:
  • For the input/output in FP16, the parameter mathPrec can be CUDNN_DATA_HALF or CUDNN_DATA_FLOAT.
  • For the input/output in FP32, the parameter mathPrec can only be CUDNN_DATA_FLOAT.
  • For the input/output in FP64, double type, the parameter mathPrec can only be CUDNN_DATA_DOUBLE.

Returns

CUDNN_STATUS_SUCCESS

The object was set successfully.

CUDNN_STATUS_BAD_PARAM

Either at least one of the parameters hiddenSize or numLayers was zero or negative, one of inputMode, direction, mode, algo or dataType has an invalid enumerant value, dropoutDesc is an invalid dropout descriptor or rnnDesc has not been created correctly.

7.2.49. cudnnSetRNNDescriptor_v8()

cudnnStatus_t cudnnSetRNNDescriptor_v8(
	cudnnRNNDescriptor_t rnnDesc,
	cudnnRNNAlgo_t algo,
	cudnnRNNMode_t cellMode,
	cudnnRNNBiasMode_t biasMode,
	cudnnDirectionMode_t dirMode,
	cudnnRNNInputMode_t inputMode,
	cudnnDataType_t dataType,
	cudnnDataType_t mathPrec,
	cudnnMathType_t mathType,
	int32_t inputSize,
	int32_t hiddenSize,
	int32_t projSize,
	int32_t numLayers,
    cudnnDropoutDescriptor_t dropoutDesc,
    uint32_t auxFlags);

This function initializes a previously created RNN descriptor object. The RNN descriptor configured by cudnnSetRNNDescriptor_v8() was enhanced to store all information needed to compute the total number of adjustable weights/biases in the RNN model.

Parameters

rnnDesc

Input. A previously initialized RNN descriptor.

algo
Input. RNN algo (CUDNN_RNN_ALGO_STANDARD, CUDNN_RNN_ALGO_PERSIST_STATIC, or CUDNN_RNN_ALGO_PERSIST_DYNAMIC).
cellMode

Input. Specifies the RNN cell type in the entire model (RELU, TANH, LSTM, GRU).

biasMode

Input. Sets the number of bias vectors (CUDNN_RNN_NO_BIAS, CUDNN_RNN_SINGLE_INP_BIAS, CUDNN_RNN_SINGLE_REC_BIAS, CUDNN_RNN_DOUBLE_BIAS). The two single bias settings are functionally the same for RELU, TANH and LSTM cell types. For differences in GRU cells, see the description of CUDNN_GRU in the cudnnRNNMode_t enumerated type.

dirMode

Input. Specifies the recurrence pattern: CUDNN_UNIDIRECTIONAL or CUDNN_BIDIRECTIONAL. In bidirectional RNNs, the hidden states passed between physical layers are concatenations of forward and backward hidden states.

inputMode

Input. Specifies how the input to the RNN model is processed by the first layer. When inputMode is CUDNN_LINEAR_INPUT, original input vectors of size inputSize are multiplied by the weight matrix to obtain vectors of hiddenSize. When inputMode is CUDNN_SKIP_INPUT, the original input vectors to the first layer are used as is without multiplying them by the weight matrix.

dataType

Input. Specifies data type for RNN weights/biases and input and output data.

mathPrec
Input. This parameter is used to control the compute math precision in the RNN model. The following applies:
  • For the input/output in FP16, the parameter mathPrec can be CUDNN_DATA_HALF or CUDNN_DATA_FLOAT.
  • For the input/output in FP32, the parameter mathPrec can only be CUDNN_DATA_FLOAT.
  • For the input/output in FP64, double type, the parameter mathPrec can only be CUDNN_DATA_DOUBLE.
mathType
Input. Sets the preferred option to use NVIDIA Tensor Cores accelerators on Volta (SM 7.0) or higher GPU-s).
  • When dataType is CUDNN_DATA_HALF, the mathType parameter can be CUDNN_DEFAULT_MATH or CUDNN_TENSOR_OP_MATH. The ALLOW_CONVERSION setting is treated the same CUDNN_TENSOR_OP_MATH for this data type.
  • When dataType is CUDNN_DATA_FLOAT, the mathType parameter can be CUDNN_DEFAULT_MATH or CUDNN_TENSOR_OP_MATH_ALLOW_CONVERSION. When the latter settings are used, original weights and intermediate results will be down-converted to CUDNN_DATA_HALF before they are used in another recursive iteration.
  • When dataType is CUDNN_DATA_DOUBLE, the mathType parameter can be CUDNN_DEFAULT_MATH.

This option has an advisory status meaning Tensor Cores may not be always utilized, for example, due to specific GEMM dimensions restrictions.

inputSize

Input. Size of the input vector in the RNN model. When the inputMode=CUDNN_SKIP_INPUT, the inputSize should match the hiddenSize value.

hiddenSize

Input. Size of the hidden state vector in the RNN model. The same hidden size is used in all RNN layers.

projSize

Input. The size of the LSTM cell output after the recurrent projection. This value should not be larger than hiddenSize. It is legal to set projSize equal to hiddenSize, however, in this case, the recurrent projection feature is disabled. The recurrent projection is an additional matrix multiplication in the LSTM cell to project hidden state vectors ht into smaller vectors rt = Wrht, where Wr is a rectangular matrix with projSize rows and hiddenSize columns. When the recurrent projection is enabled, the output of the LSTM cell (both to the next layer and unrolled in-time) is rt instead of ht. The recurrent projection can be enabled for LSTM cells and CUDNN_RNN_ALGO_STANDARD only.

numLayers

Input. Number of stacked, physical layers in the deep RNN model. When dirMode= CUDNN_BIDIRECTIONAL, the physical layer consists of two pseudo-layers corresponding to forward and backward directions.

dropoutDesc

Input. Handle to a previously created and initialized dropout descriptor. Dropout operation will be applied between physical layers. A single layer network will have no dropout applied. Dropout is used in the training mode only.

auxFlags

Input. This argument is used to pass miscellaneous switches that do not require additional numerical values to configure the corresponding feature. In future cuDNN releases, this parameter will be used to extend the RNN functionality without adding new API functions (applicable options should be bitwise OR-ed). Currently, this parameter is used to enable or disable padded input/output (CUDNN_RNN_PADDED_IO_DISABLED, CUDNN_RNN_PADDED_IO_ENABLED). When the padded I/O is enabled, layouts CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_UNPACKED and CUDNN_RNN_DATA_LAYOUT_BATCH_MAJOR_UNPACKED are permitted in RNN data descriptors.

Returns

CUDNN_STATUS_SUCCESS

The RNN descriptor was configured successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was detected.

CUDNN_STATUS_NOT_SUPPORTED

The dimensions of the bias tensor refer to an amount of data that is incompatible with the output tensor dimensions or the dataType of the two tensor descriptors are different.

CUDNN_STATUS_EXECUTION_FAILED

An incompatible or unsupported combination of input arguments was detected.

7.2.50. cudnnSetRNNMatrixMathType()

This function has been deprecated in cuDNN 8.0. Use cudnnSetRNNDescriptor_v8() instead of cudnnSetRNNMatrixMathType().

cudnnStatus_t cudnnSetRNNMatrixMathType(
    cudnnRNNDescriptor_t    rnnDesc,
    cudnnMathType_t         mType)

This function sets the preferred option to use NVIDIA Tensor Cores accelerators on Volta GPUs (SM 7.0 or higher). When the mType parameter is CUDNN_TENSOR_OP_MATH, inference and training RNN APIs will attempt use Tensor Cores when weights/biases are of type CUDNN_DATA_HALF or CUDNN_DATA_FLOAT. When RNN weights/biases are stored in the CUDNN_DATA_FLOAT format, the original weights and intermediate results will be down-converted to CUDNN_DATA_HALF before they are used in another recursive iteration.

Parameters

rnnDesc

Input. A previously created and initialized RNN descriptor.

mType

Input. A preferred compute option when performing RNN GEMMs (general matrix-matrix multiplications). This option has an advisory status meaning that Tensor Cores may not be utilized, for example, due to specific GEMM dimensions.

Returns

CUDNN_STATUS_SUCCESS

The preferred compute option for the RNN network was set successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input parameter was detected.

7.2.51. cudnnSetRNNPaddingMode()

This function has been deprecated in cuDNN 8.0. Use cudnnSetRNNDescriptor_v8() instead of cudnnSetRNNPaddingMode().

cudnnStatus_t cudnnSetRNNPaddingMode(
    cudnnRNNDescriptor_t        rnnDesc,
    cudnnRNNPaddingMode_t       paddingMode)

This function enables or disables the padded RNN input/output for a previously created and initialized RNN descriptor. This information is required before calling the cudnnGetRNNWorkspaceSize() and cudnnGetRNNTrainingReserveSize() functions, to determine whether additional workspace and training reserve space is needed. By default, the padded RNN input/output is not enabled.

Parameters

rnnDesc

Input/Output. A previously created RNN descriptor.

paddingMode

Input. Enables or disables the padded input/output. For more information, see cudnnRNNPaddingMode_t.

Returns

CUDNN_STATUS_SUCCESS

The paddingMode was set successfully.

CUDNN_STATUS_BAD_PARAM

Either the rnnDesc is NULL or paddingMode has an invalid enumerant value.

7.2.52. cudnnSetRNNProjectionLayers()

This function has been deprecated in cuDNN 8.0. Use cudnnSetRNNDescriptor_v8() instead of cudnnSetRNNProjectionLayers().

cudnnStatus_t cudnnSetRNNProjectionLayers(
    cudnnHandle_t           handle,
    cudnnRNNDescriptor_t    rnnDesc,
    int                     recProjSize,
    int                     outProjSize)

The cudnnSetRNNProjectionLayers() function should be called to enable the recurrent and/or output projection in a recursive neural network. The recurrent projection is an additional matrix multiplication in the LSTM cell to project hidden state vectors ht into smaller vectors rt = Wrht, where Wr is a rectangular matrix with recProjSize rows and hiddenSize columns. When the recurrent projection is enabled, the output of the LSTM cell (both to the next layer and unrolled in-time) is rt instead of ht. The dimensionality of it, ft, ot, and ct vectors used in conjunction with non-linear functions remains the same as in the canonical LSTM cell. To make this possible, the shapes of matrices in the LSTM formulas (see cudnnRNNMode_t type), such as Wi in hidden RNN layers or Ri in the entire network, become rectangular versus square in the canonical LSTM mode. Obviously, the result of Ri* Wr is a square matrix but it is rank deficient, reflecting the compression of LSTM output. The recurrent projection is typically employed when the number of independent (adjustable) weights in the RNN network with projection is smaller in comparison to canonical LSTM for the same hiddenSize value.

The recurrent projection can be enabled for LSTM cells and CUDNN_RNN_ALGO_STANDARD only. The recProjSize parameter should be smaller than the hiddenSize value. It is legal to set recProjSize equal to hiddenSize but in that case the recurrent projection feature is disabled.

The output projection is currently not implemented.

For more information on the recurrent and output RNN projections, see the paper by Hasim Sak, et al.: Long Short-Term Memory Based Recurrent Neural Network Architectures For Large Vocabulary Speech Recognition.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously created and initialized RNN descriptor.

recProjSize

Input. The size of the LSTM cell output after the recurrent projection. This value should not be larger than hiddenSize.

outProjSize

Input. This parameter should be zero.

Returns

CUDNN_STATUS_SUCCESS

RNN projection parameters were set successfully.

CUDNN_STATUS_BAD_PARAM

An invalid input argument was detected (for example, NULL handles, negative values for projection parameters).

CUDNN_STATUS_NOT_SUPPORTED

Projection applied to RNN algo other than CUDNN_RNN_ALGO_STANDARD, cell type other than CUDNN_LSTM, recProjSize larger than hiddenSize.

7.2.53. cudnnSetSeqDataDescriptor()

cudnnStatus_t cudnnSetSeqDataDescriptor(
	cudnnSeqDataDescriptor_t seqDataDesc,
    cudnnDataType_t dataType,
	int nbDims,
	const int dimA[],
	const cudnnSeqDataAxis_t axes[],
	size_t seqLengthArraySize,
	const int seqLengthArray[],
	void *paddingFill);
This function initializes a previously created sequence data descriptor object. In the most simplified view, this descriptor defines dimensions (dimA) and the data layout (axes) of a four-dimensional tensor. All four dimensions of the sequence data descriptor have unique identifiers that can be used to index the dimA[] array:
CUDNN_SEQDATA_TIME_DIM
CUDNN_SEQDATA_BATCH_DIM
CUDNN_SEQDATA_BEAM_DIM
CUDNN_SEQDATA_VECT_DIM

For example, to express information that vectors in our sequence data buffer are five elements long, we need to assign dimA[CUDNN_SEQDATA_VECT_DIM]=5 in the dimA[] array.

The number of active dimensions in the dimA[] and axes[] arrays is defined by the nbDims argument. Currently, the value of this argument should be four. The actual size of the dimA[] and axes[] arrays should be declared using the CUDNN_SEQDATA_DIM_COUNT macro.

The cudnnSeqDataDescriptor_t container is treated as a collection of fixed length vectors that form sequences, similarly to words (vectors of characters) constructing sentences. The TIME dimension spans the sequence length. Different sequences are bundled together in a batch. A BATCH may be a group of individual sequences or beams. A BEAM is a cluster of alternative sequences or candidates. When thinking about the beam, consider a translation task from one language to another. You may want to keep around and experiment with several translated versions of the original sentence before selecting the best one. The number of candidates kept around is the BEAM size.

Every sequence can have a different length, even within the same beam, so vectors toward the end of the sequence can be just padding. The paddingFill argument specifies how the padding vectors should be written in output sequence data buffers. The paddingFill argument points to one value of type dataType that should be copied to all elements in padding vectors. Currently, the only supported value for paddingFill is NULL which means this option should be ignored. In this case, elements of the padding vectors in output buffers will have undefined values.

It is assumed that a non-empty sequence always starts from the time index zero. The seqLengthArray[] must specify all sequence lengths in the container so the total size of this array should be dimA[CUDNN_SEQDATA_BATCH_DIM] * dimA[CUDNN_SEQDATA_BEAM_DIM]. Each element of the seqLengthArray[] array should have a non-negative value, less than or equal to dimA[CUDNN_SEQDATA_TIME_DIM; the maximum sequence length. Elements in seqLengthArray[] are always arranged in the same batch-major order, meaning, when considering BEAM and BATCH dimensions, BATCH is the outer or the slower changing index when we traverse the array in ascending order of the addresses. Using a simple example, the seqLengthArray[] array should hold sequence lengths in the following order:
{batch_idx=0, beam_idx=0}
{batch_idx=0, beam_idx=1}
{batch_idx=1, beam_idx=0}
{batch_idx=1, beam_idx=1}
{batch_idx=2, beam_idx=0}
{batch_idx=2, beam_idx=1}
when dimA[CUDNN_SEQDATA_BATCH_DIM]=3 and dimA[CUDNN_SEQDATA_BEAM_DIM]=2.
Data stored in the cudnnSeqDataDescriptor_t container must comply with the following constraints:
  • All data is fully packed. There are no unused spaces or gaps between individual vector elements or consecutive vectors.
  • The most inner dimension of the container is the vector. In other words, the first contiguous group of dimA[CUDNN_SEQDATA_VECT_DIM] elements belongs to the first vector, followed by elements of the second vector, and so on.
The axes argument in the cudnnSetSeqDataDescriptor() function is a bit more complicated. This array should have the same capacity as dimA[]. The axes[] array specifies the actual data layout in the GPU memory. In this function, the layout is described in the following way: as we move from one element of a vector to another in memory by incrementing the element pointer, what is the order of VECT, TIME, BATCH, and BEAM dimensions that we encounter. Let us assume that we want to define the following data layout:
that corresponds to tensor dimensions:
int dimA[CUDNN_SEQDATA_DIM_COUNT];
dimA[CUDNN_SEQDATA_TIME_DIM]  = 4;
dimA[CUDNN_SEQDATA_BATCH_DIM] = 3;
dimA[CUDNN_SEQDATA_BEAM_DIM]  = 2;
dimA[CUDNN_SEQDATA_VECT_DIM]  = 5;
Now, let’s initialize the axes[] array. Note that the most inner dimension is described by the last active element of axes[]. There is only one valid configuration here as we always traverse a full vector first. Thus, we need to write CUDNN_SEQDATA_VECT_DIM in the last active element of axes[].
cudnnSeqDataAxis_t axes[CUDNN_SEQDATA_DIM_COUNT];
axes[3] = CUDNN_SEQDATA_VECT_DIM;   // 3 = nbDims-1
Now, let’s work on the remaining three elements of axes[]. When we reach the end of the first vector, we jump to the next beam, therefore:
axes[2] = CUDNN_SEQDATA_BEAM_DIM;
When we approach the end of the second vector, we move to the next batch, therefore:
axes[1] = CUDNN_SEQDATA_BATCH_DIM;
The last (outermost) dimension is TIME:
axes[0] = CUDNN_SEQDATA_TIME_DIM;

The four values of the axes[] array fully describe the data layout depicted in the figure.

The sequence data descriptor allows the user to select 3! = 6 different data layouts or permutations of BEAM, BATCH and TIME dimensions. The multi-head attention API supports all six layouts.

Parameters

seqDataDesc
Output. Pointer to a previously created sequence data descriptor.
dataType
Input. Data type of the sequence data buffer (CUDNN_DATA_HALF, CUDNN_DATA_FLOAT or CUDNN_DATA_DOUBLE).
nbDims
Input. Must be 4. The number of active dimensions in dimA[] and axes[] arrays. Both arrays should be declared to contain at least CUDNN_SEQDATA_DIM_COUNT elements.
dimA[]
Input. Integer array specifying sequence data dimensions. Use the cudnnSeqDataAxis_t enumerated type to index all active dimA[] elements.
axes[]
Input. Array of cudnnSeqDataAxis_t that defines the layout of sequence data in memory. The first nbDims elements of axes[] should be initialized with the outermost dimension in axes[0] and the innermost dimension in axes[nbDims-1].
seqLengthArraySize
Input. Number of elements in the sequence length array, seqLengthArray[].
seqLengthArray[]
Input. An integer array that defines all sequence lengths of the container.
paddingFill
Input. Must be NULL. Pointer to a value of dataType that is used to fill up output vectors beyond the valid length of each sequence or NULL to ignore this setting.

Returns

CUDNN_STATUS_SUCCESS
All input arguments were validated and the sequence data descriptor was successfully updated.
CUDNN_STATUS_BAD_PARAM
An invalid input argument was found. Some examples include:
  • seqDataDesc=NULL
  • dateType was not a valid type of cudnnDataType_t
  • nbDims was negative or zero
  • seqLengthArraySize did not match the expected length
  • some elements of seqLengthArray[] were invalid
CUDNN_STATUS_NOT_SUPPORTED
An unsupported input argument was encountered. Some examples include:
  • nbDims is not equal to 4
  • paddingFill is not NULL
CUDNN_STATUS_ALLOC_FAILED
Failed to allocate storage for the sequence data descriptor object.

8. cudnn_adv_train.so Library

8.1. Data Type References

8.1.1. Enumeration Types

8.1.1.1. cudnnLossNormalizationMode_t

cudnnLossNormalizationMode_t is an enumerated type that controls the input normalization mode for a loss function. This type can be used with cudnnSetCTCLossDescriptorEx().

Values
CUDNN_LOSS_NORMALIZATION_NONE

The input probs of cudnnCTCLoss() function is expected to be the normalized probability, and the output gradients is the gradient of loss with respect to the unnormalized probability.

CUDNN_LOSS_NORMALIZATION_SOFTMAX

The input probs of cudnnCTCLoss() function is expected to be the unnormalized activation from the previous layer, and the output gradients is the gradient with respect to the activation. Internally the probability is computed by softmax normalization.

8.1.1.2. cudnnWgradMode_t

cudnnWgradMode_t is an enumerated type that selects how buffers holding gradients of the loss function, computed with respect to trainable parameters, are updated. Currently, this type is used by the cudnnGetMultiHeadAttnWeights() function only.

Values
CUDNN_WGRAD_MODE_ADD

A weight gradient component corresponding to a new batch of inputs is added to previously evaluated weight gradients. Before using this mode, the buffer holding weight gradients should be initialized to zero. Alternatively, the first API call outputting to an uninitialized buffer should use the CUDNN_WGRAD_MODE_SET option.

CUDNN_WGRAD_MODE_SET
A weight gradient component, corresponding to a new batch of inputs, overwrites previously stored weight gradients in the output buffer.

8.2. API Functions

8.2.1. cudnnAdvTrainVersionCheck()

cudnnStatus_t cudnnAdvTrainVersionCheck(void)

This function checks whether the version of the AdvTrain subset of the library is consistent with the other sub-libraries.

Returns

CUDNN_STATUS_SUCCESS

The version is consistent with other sub-libraries.

CUDNN_STATUS_VERSION_MISMATCH

The version of AdvTrain is not consistent with other sub-libraries. Users should check the installation and make sure all sub-component versions are consistent.

8.2.2. cudnnCreateCTCLossDescriptor()

cudnnStatus_t cudnnCreateCTCLossDescriptor(
    cudnnCTCLossDescriptor_t* ctcLossDesc)

This function creates a CTC loss function descriptor.

Parameters

ctcLossDesc

Output. CTC loss descriptor to be set. For more information, see cudnnCTCLossDescriptor_t.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

CTC loss descriptor passed to the function is invalid.

CUDNN_STATUS_ALLOC_FAILED

Memory allocation for this CTC loss descriptor failed.

8.2.3. cudnnCTCLoss()

cudnnStatus_t cudnnCTCLoss(
    cudnnHandle_t                        handle,
    const   cudnnTensorDescriptor_t      probsDesc,
    const   void                        *probs,
    const   int                          hostLabels[],
    const   int                          hostLabelLengths[],
    const   int                          hostInputLengths[],
    void                                *costs,
    const   cudnnTensorDescriptor_t      gradientsDesc,
    const   void                        *gradients,
    cudnnCTCLossAlgo_t                   algo,
    const   cudnnCTCLossDescriptor_t     ctcLossDesc,
    void                                *workspace,
    size_t                              *workSpaceSizeInBytes)
This function returns the CTC costs and gradients, given the probabilities and labels.
Note: This function can have an inconsistent interface depending on the cudnnLossNormalizationMode_t chosen (bound to the cudnnCTCLossDescriptor_t with cudnnSetCTCLossDescriptorEx()). For the CUDNN_LOSS_NORMALIZATION_NONE, this function has an inconsistent interface, for example, the probs input is probability normalized by softmax, but the gradients output is with respect to the unnormalized activation. However, for CUDNN_LOSS_NORMALIZATION_SOFTMAX, the function has a consistent interface; all values are normalized by softmax.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

probsDesc

Input. Handle to the previously initialized probabilities tensor descriptor. For more information, see cudnnTensorDescriptor_t.

probs

Input. Pointer to a previously initialized probabilities tensor. These input probabilities are normalized by softmax.

hostLabels

Input. Pointer to a previously initialized labels list, in CPU memory.

hostLabelLengths

Input. Pointer to a previously initialized lengths list in CPU memory, to walk the above labels list.

hostInputLengths

Input. Pointer to a previously initialized list of the lengths of the timing steps in each batch, in CPU memory.

costs

Output. Pointer to the computed costs of CTC.

gradientsDesc

Input. Handle to a previously initialized gradient tensor descriptor.

gradients

Output. Pointer to the computed gradients of CTC. These computed gradient outputs are with respect to the unnormalized activation.

algo

Input. Enumerant that specifies the chosen CTC loss algorithm. For more information, see cudnnCTCLossAlgo_t.

ctcLossDesc

Input. Handle to the previously initialized CTC loss descriptor. For more information, see cudnnCTCLossDescriptor_t.

workspace

Input. Pointer to GPU memory of a workspace needed to be able to execute the specified algorithm.

sizeInBytes

Input. Amount of GPU memory needed as workspace to be able to execute the CTC loss computation with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions of probsDesc do not match the dimensions of gradientsDesc.
  • The inputLengths do not agree with the first dimension of probsDesc.
  • The workSpaceSizeInBytes is not sufficient.
  • The labelLengths is greater than 256.
CUDNN_STATUS_NOT_SUPPORTED

A compute or data type other than FLOAT was chosen, or an unknown algorithm type was chosen.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

8.2.4. cudnnCTCLoss_v8()

cudnnStatus_t cudnnCTCLoss_v8(
    cudnnHandle_t                        handle,
    cudnnCTCLossAlgo_t                   algo,
    const   cudnnCTCLossDescriptor_t     ctcLossDesc,
    const   cudnnTensorDescriptor_t      probsDesc,
    const   void                        *probs,
    const   int                          labels[],
    const   int                          labelLengths[],
    const   int                          inputLengths[],
    void                                *costs,
    const   cudnnTensorDescriptor_t      gradientsDesc,
    const   void                        *gradients,
    size_t                              *workSpaceSizeInBytes,
    void                                *workspace)
This function returns the CTC costs and gradients, given the probabilities and labels. Many CTC API functions were updated in v8 with the _v8 suffix to support CUDA graphs. Label and input data is now passed in GPU memory, and cudnnCTCLossDescriptor_t should be set using cudnnSetCTCLossDescriptor_v8().
Note: This function can have an inconsistent interface depending on the cudnnLossNormalizationMode_t chosen (bound to the cudnnCTCLossDescriptor_t with cudnnSetCTCLossDescriptorEx()). For the CUDNN_LOSS_NORMALIZATION_NONE, this function has an inconsistent interface, for example, the probs input is probability normalized by softmax, but the gradients output is with respect to the unnormalized activation. However, for CUDNN_LOSS_NORMALIZATION_SOFTMAX, the function has a consistent interface; all values are normalized by softmax.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information, see cudnnHandle_t.

algo

Input. Enumerant that specifies the chosen CTC loss algorithm. For more information, see cudnnCTCLossAlgo_t.

ctcLossDesc

Input. Handle to the previously initialized CTC loss descriptor. To use this _v8 function, this descriptor must be set using cudnnSetCTCLossDescriptor_v8(). For more information, see cudnnCTCLossDescriptor_t.

probsDesc

Input. Handle to the previously initialized probabilities tensor descriptor. For more information, see cudnnTensorDescriptor_t.

probs

Input. Pointer to a previously initialized probabilities tensor. These input probabilities are normalized by softmax.

labels

Input. Pointer to a previously initialized labels list, in GPU memory.

labelLengths

Input. Pointer to a previously initialized lengths list in GPU memory, to walk the above labels list.

inputLengths

Input. Pointer to a previously initialized list of the lengths of the timing steps in each batch, in GPU memory.

costs

Output. Pointer to the computed costs of CTC.

gradientsDesc

Input. Handle to a previously initialized gradient tensor descriptor.

gradients

Output. Pointer to the computed gradients of CTC. These computed gradient outputs are with respect to the unnormalized activation.

workspace

Input. Pointer to GPU memory of a workspace needed to be able to execute the specified algorithm.

sizeInBytes

Input. Amount of GPU memory needed as a workspace to be able to execute the CTC loss computation with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions of probsDesc do not match the dimensions of gradientsDesc.
  • The inputLengths do not agree with the first dimension of probsDesc.
  • The workSpaceSizeInBytes is not sufficient.
  • The labelLengths is greater than 256.
CUDNN_STATUS_NOT_SUPPORTED

A compute or data type other than FLOAT was chosen, or an unknown algorithm type was chosen.

CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

8.2.5. cudnnDestroyCTCLossDescriptor()

cudnnStatus_t cudnnDestroyCTCLossDescriptor(
    cudnnCTCLossDescriptor_t 	ctcLossDesc)

This function destroys a CTC loss function descriptor object.

Parameters

ctcLossDesc

Input. CTC loss function descriptor to be destroyed.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

8.2.6. cudnnFindRNNBackwardDataAlgorithmEx()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnFindRNNBackwardDataAlgorithmEx(
    cudnnHandle_t                    handle,
    const cudnnRNNDescriptor_t       rnnDesc,
    const int                        seqLength,
    const cudnnTensorDescriptor_t    *yDesc,
    const void                       *y,
    const cudnnTensorDescriptor_t    *dyDesc,
    const void                       *dy,
    const cudnnTensorDescriptor_t    dhyDesc,
    const void                       *dhy,
    const cudnnTensorDescriptor_t    dcyDesc,
    const void                       *dcy,
    const cudnnFilterDescriptor_t    wDesc,
    const void                       *w,
    const cudnnTensorDescriptor_t    hxDesc,
    const void                       *hx,
    const cudnnTensorDescriptor_t    cxDesc,
    const void                       *cx,
    const cudnnTensorDescriptor_t    *dxDesc,
    void                             *dx,
    const cudnnTensorDescriptor_t    dhxDesc,
    void                             *dhx,
    const cudnnTensorDescriptor_t    dcxDesc,
    void                             *dcx,
    const float                      findIntensity,
    const int                        requestedAlgoCount,
    int                              *returnedAlgoCount,
    cudnnAlgorithmPerformance_t      *perfResults,
    void                             *workspace,
    size_t                           workSpaceSizeInBytes,
    const void                       *reserveSpace,
    size_t                           reserveSpaceSizeInBytes)

This function attempts all available cuDNN algorithms for cudnnRNNBackwardData(), using user-allocated GPU memory. It outputs the parameters that influence the performance of the algorithm to a user-allocated array of cudnnAlgorithmPerformance_t. These parameter metrics are written in sorted fashion where the first element has the lowest compute time.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in dyDesc.
y

Input. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

dyDesc
Input. An array of fully packed tensor descriptors describing the gradient at the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the second dimension of the tensor n in dxDesc.
dy

Input. Data pointer to GPU memory associated with the tensor descriptors in the array dyDesc.

dhyDesc
Input. A fully packed tensor descriptor describing the gradients at the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dhy

Input. Data pointer to GPU memory associated with the tensor descriptor dhyDesc. If a NULL pointer is passed, the gradients at the final hidden state of the network will be initialized to zero.

dcyDesc
Input. A fully packed tensor descriptor describing the gradients at the final cell state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dcy

Input. Data pointer to GPU memory associated with the tensor descriptor dcyDesc. If a NULL pointer is passed, the gradients at the final cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

dxDesc

Input. An array of fully packed tensor descriptors describing the gradient at the input of each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

dx

Output. Data pointer to GPU memory associated with the tensor descriptors in the array dxDesc.

dhxDesc
Input. A fully packed tensor descriptor describing the gradient at the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dhx

Output. Data pointer to GPU memory associated with the tensor descriptor dhxDesc. If a NULL pointer is passed, the gradient at the hidden input of the network will not be set.

dcxDesc
Input. A fully packed tensor descriptor describing the gradient at the initial cell state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in dxDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dcx

Output. Data pointer to GPU memory associated with the tensor descriptor dcxDesc. If a NULL pointer is passed, the gradient at the cell input of the network will not be set.

findIntensity

Input.This input was previously unused in versions prior to cuDNN 7.2.0. It is used in cuDNN 7.2.0 and later versions to control the overall runtime of the RNN find algorithms, by selecting the percentage of a large Cartesian product space to be searched.

  • Setting findIntensity within the range (0,1.] will set a percentage of the entire RNN search space to search. When findIntensity is set to 1.0, a full search is performed over all RNN parameters.
  • When findIntensity is set to 0.0f, a quick, minimal search is performed. This setting has the best runtime. However, in this case the parameters returned by this function will not correspond to the best performance of the algorithm; a longer search might discover better parameters. This option will execute up to three instances of the configured RNN problem. Runtime will vary proportionally to RNN problem size, as it will in the other cases, hence no guarantee of an explicit time bound can be given.
  • Setting findIntensity within the range [-1.,0) sets a percentage of a reduced Cartesian product space to be searched. This reduced search space has been heuristically selected to have good performance. The setting of -1.0 represents a full search over this reduced search space.
  • Values outside the range [-1,1] are truncated to the range [-1,1], and then interpreted as per the above.
  • Setting findIntensity to 1.0 in cuDNN 7.2 and later versions is equivalent to the behavior of this function in versions prior to cuDNN 7.2.0.
  • This function times the single RNN executions over large parameter spaces - one execution per parameter combination. The times returned by this function are latencies.
requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors dhxDesc, wDesc, hxDesc, cxDesc, dcxDesc, dhyDesc, dcyDesc or one of the descriptors in yDesc, dxdesc, dydesc is invalid.
  • The descriptors in one of yDesc, dxDesc, dyDesc, dhxDesc, wDesc, hxDesc, cxDesc, dcxDesc, dhyDesc, dcyDesc has incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.7. cudnnFindRNNBackwardWeightsAlgorithmEx()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnFindRNNBackwardWeightsAlgorithmEx(
    cudnnHandle_t                    handle,
    const cudnnRNNDescriptor_t       rnnDesc,
    const int                        seqLength,
    const cudnnTensorDescriptor_t    *xDesc,
    const void                       *x,
    const cudnnTensorDescriptor_t    hxDesc,
    const void                       *hx,
    const cudnnTensorDescriptor_t    *yDesc,
    const void                       *y,
    const float                      findIntensity,
    const int                        requestedAlgoCount,
    int                              *returnedAlgoCount,
    cudnnAlgorithmPerformance_t      *perfResults,
    const void                       *workspace,
    size_t                           workSpaceSizeInBytes,
    const cudnnFilterDescriptor_t    dwDesc,
    void                             *dw,
    const void                       *reserveSpace,
    size_t                           reserveSpaceSizeInBytes)

This function attempts all available cuDNN algorithms for cudnnRNNBackwardWeights(), using user-allocated GPU memory. It outputs the parameters that influence the performance of the algorithm to a user-allocated array of cudnnAlgorithmPerformance_t. These parameter metrics are written in sorted fashion where the first element has the lowest compute time.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc

Input. An array of fully packed tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

x

Input. Data pointer to GPU memory associated with the tensor descriptors in the array xDesc.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in dyDesc.
y

Input. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

findIntensity

Input.This input was previously unused in versions prior to cuDNN 7.2.0. It is used in cuDNN 7.2.0 and later versions to control the overall runtime of the RNN find algorithms, by selecting the percentage of a large Cartesian product space to be searched.

  • Setting findIntensity within the range (0,1.] will set a percentage of the entire RNN search space to search. When findIntensity is set to 1.0, a full search is performed over all RNN parameters.
  • When findIntensity is set to 0.0f, a quick, minimal search is performed. This setting has the best runtime. However, in this case the parameters returned by this function will not correspond to the best performance of the algorithm; a longer search might discover better parameters. This option will execute up to three instances of the configured RNN problem. Runtime will vary proportionally to RNN problem size, as it will in the other cases, hence no guarantee of an explicit time bound can be given.
  • Setting findIntensity within the range [-1.,0) sets a percentage of a reduced Cartesian product space to be searched. This reduced search space has been heuristically selected to have good performance. The setting of -1.0 represents a full search over this reduced search space.
  • Values outside the range [-1,1] are truncated to the range [-1,1], and then interpreted as per the above.
  • Setting findIntensity to 1.0 in cuDNN 7.2 and later versions is equivalent to the behavior of this function in versions prior to cuDNN 7.2.0.
  • This function times the single RNN executions over large parameter spaces - one execution per parameter combination. The times returned by this function are latencies.
requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

dwDesc

Input. Handle to a previously initialized filter descriptor describing the gradients of the weights for the RNN.

dw

Input/Output. Data pointer to GPU memory associated with the filter descriptor dwDesc.

reserveSpace

Input. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, dwDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, yDesc, dwDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.8. cudnnFindRNNForwardTrainingAlgorithmEx()

This function has been deprecated in cuDNN 8.0.
cudnnStatus_t cudnnFindRNNForwardTrainingAlgorithmEx(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    const void                     *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                     *cx,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const cudnnTensorDescriptor_t  *yDesc,
    void                           *y,
    const cudnnTensorDescriptor_t   hyDesc,
    void                           *hy,
    const cudnnTensorDescriptor_t   cyDesc,
    void                           *cy,
    const float                    findIntensity,
    const int                      requestedAlgoCount,
    int                            *returnedAlgoCount,
    cudnnAlgorithmPerformance_t    *perfResults,
    void                           *workspace,
    size_t                          workSpaceSizeInBytes,
    void                           *reserveSpace,
    size_t                          reserveSpaceSizeInBytes)

This function attempts all available cuDNN algorithms for cudnnRNNForwardTraining(), using user-allocated GPU memory. It outputs the parameters that influence the performance of the algorithm to a user-allocated array of cudnnAlgorithmPerformance_t. These parameter metrics are written in sorted fashion where the first element has the lowest compute time.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

xDesc

Input. An array of fully packed tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

x

Input. Data pointer to GPU memory associated with the tensor descriptors in the array xDesc.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in xDesc.
y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

hyDesc
Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc
Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

findIntensity

Input.This input was previously unused in versions prior to cuDNN 7.2.0. It is used in cuDNN 7.2.0 and later versions to control the overall runtime of the RNN find algorithms, by selecting the percentage of a large Cartesian product space to be searched.

  • Setting findIntensity within the range (0,1.] will set a percentage of the entire RNN search space to search. When findIntensity is set to 1.0, a full search is performed over all RNN parameters.
  • When findIntensity is set to 0.0f, a quick, minimal search is performed. This setting has the best runtime. However, in this case the parameters returned by this function will not correspond to the best performance of the algorithm; a longer search might discover better parameters. This option will execute up to three instances of the configured RNN problem. Runtime will vary proportionally to RNN problem size, as it will in the other cases, hence no guarantee of an explicit time bound can be given.
  • Setting findIntensity within the range [-1.,0) sets a percentage of a reduced Cartesian product space to be searched. This reduced search space has been heuristically selected to have good performance. The setting of -1.0 represents a full search over this reduced search space.
  • Values outside the range [-1,1] are truncated to the range [-1,1], and then interpreted as per the above.
  • Setting findIntensity to 1.0 in cuDNN 7.2 and later versions is equivalent to the behavior of this function in versions prior to cuDNN 7.2.0.
  • This function times the single RNN executions over large parameter spaces - one execution per parameter combination. The times returned by this function are latencies.
requestedAlgoCount

Input. The maximum number of elements to be stored in perfResults.

returnedAlgoCount

Output. The number of output elements stored in perfResults.

perfResults

Output. A user-allocated array to store performance metrics sorted ascending by compute time.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, cxDesc, wDesc, hyDesc, cyDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, cxDesc, wDesc, yDesc, hyDesc, cyDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.9. cudnnGetCTCLossDescriptor()

cudnnStatus_t cudnnGetCTCLossDescriptor(
    cudnnCTCLossDescriptor_t         ctcLossDesc,
    cudnnDataType_t*                 compType)

This function returns the configuration of the passed CTC loss function descriptor.

Parameters

ctcLossDesc

Input. CTC loss function descriptor passed, from which to retrieve the configuration.

compType

Output. Compute type associated with this CTC loss function descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

Input OpTensor descriptor passed is invalid.

8.2.10. cudnnGetCTCLossDescriptorEx()

cudnnStatus_t cudnnGetCTCLossDescriptorEx(
    cudnnCTCLossDescriptor_t         ctcLossDesc,
    cudnnDataType_t                 *compType,
    cudnnLossNormalizationMode_t    *normMode,
    cudnnNanPropagation_t           *gradMode)

This function returns the configuration of the passed CTC loss function descriptor.

Parameters

ctcLossDesc

Input. CTC loss function descriptor passed, from which to retrieve the configuration.

compType

Output. Compute type associated with this CTC loss function descriptor.

normMode

Output. Input normalization type for this CTC loss function descriptor. For more information, see cudnnLossNormalizationMode_t.

gradMode

Output. NaN propagation type for this CTC loss function descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

Input OpTensor descriptor passed is invalid.

8.2.11. cudnnGetCTCLossDescriptor_v8()

cudnnStatus_t cudnnGetCTCLossDescriptor_v8(
    cudnnCTCLossDescriptor_t         ctcLossDesc,
    cudnnDataType_t                 *compType,
    cudnnLossNormalizationMode_t    *normMode,
    cudnnNanPropagation_t           *gradMode,
    int                             *maxLabelLength)

This function returns the configuration of the passed CTC loss function descriptor.

Parameters

ctcLossDesc

Input. CTC loss function descriptor passed, from which to retrieve the configuration.

compType

Output. Compute type associated with this CTC loss function descriptor.

normMode

Output. Input normalization type for this CTC loss function descriptor. For more information, see cudnnLossNormalizationMode_t.

gradMode

Output. NaN propagation type for this CTC loss function descriptor.

maxLabelLength

Output. The max label length for this CTC loss function descriptor.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

Input OpTensor descriptor passed is invalid.

8.2.12. cudnnGetCTCLossWorkspaceSize()

cudnnStatus_t cudnnGetCTCLossWorkspaceSize(
    cudnnHandle_t                        handle,
    const   cudnnTensorDescriptor_t      probsDesc,
    const   cudnnTensorDescriptor_t      gradientsDesc,
    const   int                         *labels,
    const   int                         *labelLengths,
    const   int                         *inputLengths,
    cudnnCTCLossAlgo_t                   algo,
    const   cudnnCTCLossDescriptor_t     ctcLossDesc,
    size_t                              *sizeInBytes)

This function returns the amount of GPU memory workspace the user needs to allocate to be able to call cudnnCTCLoss() with the specified algorithm. The workspace allocated will then be passed to the routine cudnnCTCLoss().

Parameters

handle

Input. Handle to a previously created cuDNN context.

probsDesc

Input. Handle to the previously initialized probabilities tensor descriptor.

gradientsDesc

Input. Handle to a previously initialized gradient tensor descriptor.

labels

Input. Pointer to a previously initialized labels list.

labelLengths

Input. Pointer to a previously initialized lengths list, to walk the above labels list.

inputLengths

Input. Pointer to a previously initialized list of the lengths of the timing steps in each batch.

algo

Input. Enumerant that specifies the chosen CTC loss algorithm

ctcLossDesc

Input. Handle to the previously initialized CTC loss descriptor.

sizeInBytes

Output. Amount of GPU memory needed as workspace to be able to execute the CTC loss computation with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions of probsDesc do not match the dimensions of gradientsDesc.
  • The inputLengths do not agree with the first dimension of probsDesc.
  • The workSpaceSizeInBytes is not sufficient.
  • The labelLengths is greater than 256.
CUDNN_STATUS_NOT_SUPPORTED

A compute or data type other than FLOAT was chosen, or an unknown algorithm type was chosen.

8.2.13. cudnnGetCTCLossWorkspaceSize_v8()

cudnnStatus_t cudnnGetCTCLossWorkspaceSize_v8(
    cudnnHandle_t                        handle,
    cudnnCTCLossAlgo_t                   algo,
    const   cudnnCTCLossDescriptor_t     ctcLossDesc,
    const   cudnnTensorDescriptor_t      probsDesc,
    const   cudnnTensorDescriptor_t      gradientsDesc,
    size_t                              *sizeInBytes

This function returns the amount of GPU memory workspace the user needs to allocate to be able to call cudnnCTCLoss_v8() with the specified algorithm. The workspace allocated will then be passed to the routine cudnnCTCLoss_v8().

Parameters

handle

Input. Handle to a previously created cuDNN context.

algo

Input. Enumerant that specifies the chosen CTC loss algorithm.

ctcLossDesc

Input. Handle to the previously initialized CTC loss descriptor.

probsDesc

Input. Handle to the previously initialized probabilities tensor descriptor.

gradientsDesc

Input. Handle to a previously initialized gradient tensor descriptor.

sizeInBytes

Output. Amount of GPU memory needed as workspace to be able to execute the CTC loss computation with the specified algo.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The dimensions of probsDesc do not match the dimensions of gradientsDesc.
CUDNN_STATUS_NOT_SUPPORTED

A compute or data type other than FLOAT was chosen, or an unknown algorithm type was chosen.

8.2.14. cudnnGetRNNTrainingReserveSize()

This function has been deprecated in cuDNN 8.0. Use cudnnGetRNNTempSpaceSizes() instead of cudnnGetRNNWorkspaceSize().

cudnnStatus_t cudnnGetRNNTrainingReserveSize(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    size_t                         *sizeInBytes)

This function is used to query the amount of reserved space required for training the RNN described by rnnDesc with input dimensions defined by xDesc. The same reserved space buffer must be passed to cudnnRNNForwardTraining(), cudnnRNNBackwardData(), and cudnnRNNBackwardWeights(). Each of these calls overwrites the contents of the reserved space, however it can safely be backed up and restored between calls if reuse of the memory is desired.

Parameters

handle

Input. Handle to a previously created cuDNN library descriptor.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc

Input. An array of tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

sizeInBytes

Output. Minimum amount of GPU memory needed as reserve space to be able to train an RNN with the specified descriptor and input tensors.

Returns

CUDNN_STATUS_SUCCESS

The query was successful.

CUDNN_STATUS_BAD_PARAM
At least one of the following conditions are met:
  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors in xDesc is invalid.
  • The descriptors in xDesc have inconsistent second dimensions, strides or data types.
  • The descriptors in xDesc have increasing first dimensions.
  • The descriptors in xDesc are not fully packed.
CUDNN_STATUS_NOT_SUPPORTED

The data types in tensors described by xDesc are not supported.

8.2.15. cudnnMultiHeadAttnBackwardData()

cudnnStatus_t cudnnMultiHeadAttnBackwardData(
	cudnnHandle_t handle,
	const cudnnAttnDescriptor_t attnDesc,
	const int loWinIdx[],
	const int hiWinIdx[],
	const int devSeqLengthsDQDO[],
	const int devSeqLengthsDKDV[],
	const cudnnSeqDataDescriptor_t doDesc,
	const void *dout,
	const cudnnSeqDataDescriptor_t dqDesc,
	void *dqueries,
	const void *queries,
	const cudnnSeqDataDescriptor_t dkDesc,
	void *dkeys,
	const void *keys,
	const cudnnSeqDataDescriptor_t dvDesc,
	void *dvalues,
	const void *values,
	size_t weightSizeInBytes,
	const void *weights,
	size_t workSpaceSizeInBytes,
	void *workSpace,
	size_t reserveSpaceSizeInBytes,
	void *reserveSpace);

This function computes exact, first-order derivatives of the multi-head attention block with respect to its inputs: Q, K, V. If y=F(w) is a vector-valued function that represents the multi-head attention layer and it takes some vector x ϵ n as an input (with all other parameters and inputs constant), and outputs vector y ϵ m , then cudnnMultiHeadAttnBackwardData() computes the result of y i / x j T δ out where δ out is the m × 1 gradient of the loss function with respect to multi-head attention outputs. The δ out gradient is back propagated through prior layers of the deep learning model. y i / x j is the m × n Jacobian matrix of F(x). The input is supplied via the dout argument and gradient results for Q, K, V are written to the dqueries, dkeys, and dvalues buffers.

The cudnnMultiHeadAttnBackwardData() function does not output partial derivatives for residual connections because this result is equal to δ out . If the multi-head attention model enables residual connections sourced directly from Q, then the dout tensor needs to be added to dqueries to obtain the correct result of the latter. This operation is demonstrated in the cuDNN multiHeadAttention sample code.

The cudnnMultiHeadAttnBackwardData() function must be invoked after cudnnMultiHeadAttnForward(). The loWinIdx[], hiWinIdx[], queries, keys, values, weights, and reserveSpace arguments should be the same as in the cudnnMultiHeadAttnForward() call. devSeqLengthsDQDO[] and devSeqLengthsDKDV[] device arrays should contain the same start and end attention window indices as devSeqLengthsQO[] and devSeqLengthsKV[] arrays in the forward function invocation.
Note:cudnnMultiHeadAttnBackwardData() does not verify that sequence lengths stored in devSeqLengthsDQDO[] and devSeqLengthsDKDV[] contain the same settings as seqLengthArray[] in the corresponding sequence data descriptor.

Parameters

handle

Input. The current context handle.

attnDesc

Input. A previously initialized attention descriptor.

loWinIdx[], hiWinIdx[]

Input. Two host integer arrays specifying the start and end indices of the attention window for each Q time-step. The start index in K, V sets is inclusive, and the end index is exclusive.

devSeqLengthsDQDO[]

Input. Device array containing a copy of the sequence length array from the dqDesc or doDesc sequence data descriptor.

devSeqLengthsDKDV[]

Input. Device array containing a copy of the sequence length array from the dkDesc or dvDesc sequence data descriptor.

doDesc

Input. Descriptor for the δ out gradients (vectors of partial derivatives of the loss function with respect to the multi-head attention outputs).

dout

Pointer to δ out gradient data in the device memory.

dqDesc

Input. Descriptor for queries and dqueries sequence data.

dqueries

Output. Device pointer to gradients of the loss function computed with respect to queries vectors.

queries

Input. Pointer to queries data in the device memory. This is the same input as in cudnnMultiHeadAttnForward().

dkDesc

Input. Descriptor for keys and dkeys sequence data.

dkeys

Output. Device pointer to gradients of the loss function computed with respect to keys vectors.

keys

Input. Pointer to keys data in the device memory. This is the same input as in cudnnMultiHeadAttnForward().

dvDesc

Input. Descriptor for values and dvalues sequence data.

dvalues

Output. Device pointer to gradients of the loss function computed with respect to values vectors.

values

Input. Pointer to values data in the device memory. This is the same input as in cudnnMultiHeadAttnForward().

weightSizeInBytes

Input. Size of the weight buffer in bytes where all multi-head attention trainable parameters are stored.

weights

Input. Address of the weight buffer in the device memory.

workSpaceSizeInBytes

Input. Size of the work-space buffer in bytes used for temporary API storage.

workSpace

Input/Output. Address of the work-space buffer in the device memory.

reserveSpaceSizeInBytes

Input. Size of the reserve-space buffer in bytes used for data exchange between forward and backward (gradient) API calls.

reserveSpace

Input/Output. Address to the reserve-space buffer in the device memory.

Returns

CUDNN_STATUS_SUCCESS

No errors were detected while processing API input arguments and launching GPU kernels.

CUDNN_STATUS_BAD_PARAM

An invalid or incompatible input argument was encountered.

CUDNN_STATUS_EXECUTION_FAILED

The process of launching a GPU kernel returned an error, or an earlier kernel did not complete successfully.

CUDNN_STATUS_INTERNAL_ERROR

An inconsistent internal state was encountered.

CUDNN_STATUS_NOT_SUPPORTED

A requested option or a combination of input arguments is not supported.

CUDNN_STATUS_ALLOC_FAILED

Insufficient amount of shared memory to launch a GPU kernel.

8.2.16. cudnnMultiHeadAttnBackwardWeights()

cudnnStatus_t cudnnMultiHeadAttnBackwardWeights(
	cudnnHandle_t handle,
	const cudnnAttnDescriptor_t attnDesc,
	cudnnWgradMode_t addGrad,
	const cudnnSeqDataDescriptor_t qDesc,
	const void *queries,
	const cudnnSeqDataDescriptor_t kDesc,
	const void *keys,
	const cudnnSeqDataDescriptor_t vDesc,
	const void *values,
	const cudnnSeqDataDescriptor_t doDesc,
	const void *dout,
	size_t weightSizeInBytes,
	const void *weights,
	void *dweights,
	size_t workSpaceSizeInBytes,
	void *workSpace,
	size_t reserveSpaceSizeInBytes,
	void *reserveSpace);

This function computes exact, first-order derivatives of the multi-head attention block with respect to its trainable parameters: projection weights and projection biases. If y=F(w) is a vector-valued function that represents the multi-head attention layer and it takes some vector x ϵ n of flatten weights or biases as an input (with all other parameters and inputs fixed), and outputs vector y ϵ m , then cudnnMultiHeadAttnBackwardWeights() computes the result of y i / x j T δ out where δ out is the m × 1 gradient of the loss function with respect to multi-head attention outputs. The δ out gradient is back propagated through prior layers of the deep learning model. y i / x j is the m × n Jacobian matrix of F(w). The δ out input is supplied via the dout argument.

All gradient results with respect to weights and biases are written to the dweights buffer. The size and the organization of the dweights buffer is the same as the weights buffer that holds multi-head attention weights and biases. The cuDNN multiHeadAttention sample code demonstrates how to access those weights.

Gradient of the loss function with respect to weights or biases is typically computed over multiple batches. In such a case, partial results computed for each batch should be summed together. The addGrad argument specifies if the gradients from the current batch should be added to previously computed results or the dweights buffer should be overwritten with the new results.

The cudnnMultiHeadAttnBackwardWeights() function should be invoked after cudnnMultiHeadAttnBackwardData(). The queries, keys, values, weights, and reserveSpace arguments should be the same as in cudnnMultiHeadAttnForward() and cudnnMultiHeadAttnBackwardData() calls. The dout argument should be the same as in cudnnMultiHeadAttnBackwardData().

Parameters

handle

Input. The current context handle.

attnDesc

Input. A previously initialized attention descriptor.

addGrad

Input. Weight gradient output mode.

qDesc

Input. Descriptor for the query sequence data.

queries

Input. Pointer to queries sequence data in the device memory.

kDesc

Input. Descriptor for the keys sequence data.

keys

Input. Pointer to keys sequence data in the device memory.

vDesc

Input. Descriptor for the values sequence data.

values

Input. Pointer to values sequence data in the device memory.

doDesc

Input. Descriptor for the δ out gradients (vectors of partial derivatives of the loss function with respect to the multi-head attention outputs).

dout

Input. Pointer to δ out gradient data in the device memory.

weightSizeInBytes

Input. Size of the weights and dweights buffers in bytes.

weights

Input. Address of the weight buffer in the device memory.

dweights

Output. Address of the weight gradient buffer in the device memory.

workSpaceSizeInBytes

Input. Size of the work-space buffer in bytes used for temporary API storage.

workSpace

Input/Output. Address of the work-space buffer in the device memory.

reserveSpaceSizeInBytes

Input. Size of the reserve-space buffer in bytes used for data exchange between forward and backward (gradient) API calls.

reserveSpace

Input/Output. Address to the reserve-space buffer in the device memory.

Returns

CUDNN_STATUS_SUCCESS

No errors were detected while processing API input arguments and launching GPU kernels.

CUDNN_STATUS_BAD_PARAM

An invalid or incompatible input argument was encountered.

CUDNN_STATUS_EXECUTION_FAILED

The process of launching a GPU kernel returned an error, or an earlier kernel did not complete successfully.

CUDNN_STATUS_INTERNAL_ERROR

An inconsistent internal state was encountered.

CUDNN_STATUS_NOT_SUPPORTED

A requested option or a combination of input arguments is not supported.

8.2.17. cudnnRNNBackwardData()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNBackwardData_v8() instead of cudnnRNNBackwardData().

cudnnStatus_t cudnnRNNBackwardData(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *yDesc,
    const void                     *y,
    const cudnnTensorDescriptor_t  *dyDesc,
    const void                     *dy,
    const cudnnTensorDescriptor_t   dhyDesc,
    const void                     *dhy,
    const cudnnTensorDescriptor_t   dcyDesc,
    const void                     *dcy,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                     *cx,
    const cudnnTensorDescriptor_t  *dxDesc,
    void                           *dx,
    const cudnnTensorDescriptor_t   dhxDesc,
    void                           *dhx,
    const cudnnTensorDescriptor_t   dcxDesc,
    void                           *dcx,
    void                           *workspace,
    size_t                          workSpaceSizeInBytes,
    const void                     *reserveSpace,
    size_t                          reserveSpaceSizeInBytes)

This routine executes the recurrent neural network described by rnnDesc with output gradients dy, dhy, and dhc, weights w and input gradients dx, dhx, and dcx. workspace is required for intermediate storage. The data in reserveSpace must have previously been generated by cudnnRNNForwardTraining(). The same reserveSpace data must be used for future calls to cudnnRNNBackwardWeights() if they execute on the same input data.

Parameters

handle

Input. Handle to a previously created cuDNN context. For more information , see cudnnHandle_t.

rnnDesc

Input. A previously initialized RNN descriptor. For more information, see cudnnRNNDescriptor_t.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in the cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). For more information, see cudnnTensorDescriptor_t. The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in dyDesc.
y

Input. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

dyDesc
Input. An array of fully packed tensor descriptors describing the gradient at the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in dxDesc.
dy

Input. Data pointer to GPU memory associated with the tensor descriptors in the array dyDesc.

dhyDesc
Input. A fully packed tensor descriptor describing the gradients at the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dhy

Input. Data pointer to GPU memory associated with the tensor descriptor dhyDesc. If a NULL pointer is passed, the gradients at the final hidden state of the network will be initialized to zero.

dcyDesc
Input. A fully packed tensor descriptor describing the gradients at the final cell state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dcy

Input. Data pointer to GPU memory associated with the tensor descriptor dcyDesc. If a NULL pointer is passed, the gradients at the final cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN. For more information, see cudnnFilterDescriptor_t.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the second dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the second dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

dxDesc

Input. An array of fully packed tensor descriptors describing the gradient at the input of each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

dx

Output. Data pointer to GPU memory associated with the tensor descriptors in the array dxDesc.

dhxDesc
Input. A fully packed tensor descriptor describing the gradient at the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dhx

Output. Data pointer to GPU memory associated with the tensor descriptor dhxDesc. If a NULL pointer is passed, the gradient at the hidden input of the network will not be set.

dcxDesc
Input. A fully packed tensor descriptor describing the gradient at the initial cell state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
dcx

Output. Data pointer to GPU memory associated with the tensor descriptor dcxDesc. If a NULL pointer is passed, the gradient at the cell input of the network will not be set.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors dhxDesc, wDesc, hxDesc, cxDesc, dcxDesc, dhyDesc, dcyDesc or one of the descriptors in yDesc, dxdesc, dydesc is invalid.
  • The descriptors in one of yDesc, dxDesc, dyDesc, dhxDesc, wDesc, hxDesc, cxDesc, dcxDesc, dhyDesc, dcyDesc has incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.19. cudnnRNNBackwardDataEx()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNBackwardData_v8 instead of cudnnRNNBackwardDataEx().

cudnnStatus_t cudnnRNNBackwardDataEx(
    cudnnHandle_t                     handle,
    const cudnnRNNDescriptor_t        rnnDesc,
    const cudnnRNNDataDescriptor_t    yDesc,
    const void                        *y,
    const cudnnRNNDataDescriptor_t    dyDesc,
    const void                        *dy,
    const cudnnRNNDataDescriptor_t    dcDesc,
    const void                        *dcAttn,
    const cudnnTensorDescriptor_t     dhyDesc,
    const void                        *dhy,
    const cudnnTensorDescriptor_t     dcyDesc,
    const void                        *dcy,
    const cudnnFilterDescriptor_t     wDesc,
    const void                        *w,
    const cudnnTensorDescriptor_t     hxDesc,
    const void                        *hx,
    const cudnnTensorDescriptor_t     cxDesc,
    const void                        *cx,
    const cudnnRNNDataDescriptor_t    dxDesc,
    void                              *dx,
    const cudnnTensorDescriptor_t     dhxDesc,
    void                              *dhx,
    const cudnnTensorDescriptor_t     dcxDesc,
    void                              *dcx,
    const cudnnRNNDataDescriptor_t    dkDesc,
    void                              *dkeys,
    void                              *workSpace,
    size_t                            workSpaceSizeInBytes,
    void                              *reserveSpace,
    size_t                            reserveSpaceSizeInBytes)

This routine is the extended version of the function cudnnRNNBackwardData(). This function cudnnRNNBackwardDataEx() allows the user to use unpacked (padded) layout for input y and output dx.

In the unpacked layout, each sequence in the mini-batch is considered to be of fixed length, specified by maxSeqLength in its corresponding RNNDataDescriptor. Each fixed-length sequence, for example, the nth sequence in the mini-batch, is composed of a valid segment specified by the seqLengthArray[n] in its corresponding RNNDataDescriptor; and a padding segment to make the combined sequence length equal to maxSeqLength.

With the unpacked layout, both sequence major (meaning, time major) and batch major are supported. For backward compatibility, the packed sequence major layout is supported. However, similar to the non-extended function cudnnRNNBackwardData(), the sequences in the mini-batch need to be sorted in descending order according to length.

Parameters

handle

Input. Handle to a previously created This function is deprecated starting in cuDNN 8.0.0. context.

rnnDesc

Input. A previously initialized RNN descriptor.

yDesc

Input. A previously initialized RNN data descriptor. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

y

Input. Data pointer to the GPU memory associated with the RNN data descriptor yDesc. The vectors are expected to be laid out in memory according to the layout specified by yDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported. Must contain the exact same data previously produced by cudnnRNNForwardTrainingEx().

dyDesc

Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength, batchSize, vectorSize, and seqLengthArray need to match the yDesc previously passed to cudnnRNNForwardTrainingEx().

dy

Input. Data pointer to the GPU memory associated with the RNN data descriptor dyDesc. The vectors are expected to be laid out in memory according to the layout specified by dyDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

dhyDesc
Input. A fully packed tensor descriptor describing the gradients at the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc. Additionally:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.

The second dimension must match the batchSize parameter in xDesc. The third dimension depends on whether the RNN mode is CUDNN_LSTM and whether LSTM projection is enabled. Additionally:

  • If the RNN mode is CUDNN_LSTM and LSTM projection is enabled, the third dimension must match the recProjSize argument passed to cudnnSetRNNProjectionLayers() call used to set rnnDesc.
  • Otherwise, the third dimension must match the hiddenSize argument used to initialize rnnDesc.
dhy

Input. Data pointer to GPU memory associated with the tensor descriptor dhyDesc. If a NULL pointer is passed, the gradients at the final hidden state of the network will be initialized to zero.

dcyDesc
Input. A fully packed tensor descriptor describing the gradients at the final cell state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc. Additionally:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.

The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.

dcy

Input. Data pointer to GPU memory associated with the tensor descriptor dcyDesc. If a NULL pointer is passed, the gradients at the final cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

hxDesc

Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero. Must contain the exact same data previously passed into cudnnRNNForwardTrainingEx(), or be NULL if NULL was previously passed to cudnnRNNForwardTrainingEx().

cxDesc

Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero. Must contain the exact same data previously passed into cudnnRNNForwardTrainingEx(), or be NULL if NULL was previously passed to cudnnRNNForwardTrainingEx().

dxDesc

Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength, batchSize, vectorSize and seqLengthArray need to match that of xDesc previously passed to cudnnRNNForwardTrainingEx().

dx

Output. Data pointer to the GPU memory associated with the RNN data descriptor dxDesc. The vectors are expected to be laid out in memory according to the layout specified by dxDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

dhxDesc

Input. A fully packed tensor descriptor describing the gradient at the initial hidden state of the RNN. The descriptor must be set exactly the same way as dhyDesc.

dhx

Output. Data pointer to GPU memory associated with the tensor descriptor dhxDesc. If a NULL pointer is passed, the gradient at the hidden input of the network will not be set.

dcxDesc

Input. A fully packed tensor descriptor describing the gradient at the initial cell state of the RNN. The descriptor must be set exactly the same way as dcyDesc.

dcx

Output. Data pointer to GPU memory associated with the tensor descriptor dcxDesc. If a NULL pointer is passed, the gradient at the cell input of the network will not be set.

dkDesc

Reserved. User may pass in NULL.

dkeys

Reserved. User may pass in NULL.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

At least one of the following conditions are met:

  • Variable sequence length input is passed in while CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used.
  • CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used on pre-Pascal devices.
  • Double input/output is used for CUDNN_RNN_ALGO_PERSIST_STATIC.
CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors yDesc, dxdesc, dydesc, dhxDesc, wDesc, hxDesc, cxDesc, dcxDesc, dhyDesc, dcyDesc is invalid or has incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.20. cudnnRNNBackwardWeights()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNBackwardWeights_v8() instead of cudnnRNNBackwardWeights().

cudnnStatus_t cudnnRNNBackwardWeights(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    const void                     *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t  *yDesc,
    const void                     *y,
    const void                     *workspace,
    size_t                          workSpaceSizeInBytes,
    const cudnnFilterDescriptor_t   dwDesc,
    void                           *dw,
    const void                     *reserveSpace,
    size_t                          reserveSpaceSizeInBytes)

This routine accumulates weight gradients dw from the recurrent neural network described by rnnDesc with inputs x, hx and outputs y. The mode of operation in this case is additive, the weight gradients calculated will be added to those already existing in dw. workspace is required for intermediate storage. The data in reserveSpace must have previously been generated by cudnnRNNBackwardData().

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc

Input. An array of fully packed tensor descriptors describing the input to each recurrent iteration (one descriptor per iteration). The first dimension (batch size) of the tensors may decrease from element n to element n+1 but may not increase. Each tensor descriptor must have the same second dimension (vector length).

x

Input. Data pointer to GPU memory associated with the tensor descriptors in the array xDesc.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in dyDesc.
y

Input. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

dwDesc

Input. Handle to a previously initialized filter descriptor describing the gradients of the weights for the RNN.

dw

Input/Output. Data pointer to GPU memory associated with the filter descriptor dwDesc.

reserveSpace

Input. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, dwDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, yDesc, dwDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.22. cudnnRNNBackwardWeightsEx()

This function has been deprecated in cuDNN 8.0. Use cudnnRNNBackwardWeights_v8() instead of cudnnRNNBackwardWeightsEX().

cudnnStatus_t cudnnRNNBackwardWeightsEx(
    cudnnHandle_t                    handle,
    const cudnnRNNDescriptor_t       rnnDesc,
    const cudnnRNNDataDescriptor_t   xDesc,
    const void                       *x,
    const cudnnTensorDescriptor_t    hxDesc,
    const void                       *hx,
    const cudnnRNNDataDescriptor_t   yDesc,
    const void                       *y,
    void                             *workSpace,
    size_t                           workSpaceSizeInBytes,
    const cudnnFilterDescriptor_t    dwDesc,
    void                             *dw,
    void                             *reserveSpace,
    size_t                           reserveSpaceSizeInBytes)

This routine is the extended version of the function cudnnRNNBackwardWeights(). This function cudnnRNNBackwardWeightsEx() allows the user to use an unpacked (padded) layout for input x and output dw.

In the unpacked layout, each sequence in the mini-batch is considered to be of fixed length, specified by maxSeqLength in its corresponding RNNDataDescriptor. Each fixed-length sequence, for example, the nth sequence in the mini-batch, is composed of a valid segment specified by the seqLengthArray[n] in its corresponding RNNDataDescriptor; and a padding segment to make the combined sequence length equal to maxSeqLength.

With the unpacked layout, both sequence major (meaning, time major) and batch major are supported. For backward compatibility, the packed sequence major layout is supported. However, similar to the non-extended function cudnnRNNBackwardWeights(), the sequences in the mini-batch need to be sorted in descending order according to length.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

xDesc

Input. A previously initialized RNN data descriptor. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

x

Input. Data pointer to GPU memory associated with the tensor descriptors in the array xDesc. Must contain the exact same data previously passed into cudnnRNNForwardTrainingEx().

hxDesc

Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero. Must contain the exact same data previously passed into cudnnRNNForwardTrainingEx(), or be NULL if NULL was previously passed to cudnnRNNForwardTrainingEx().

yDesc

Input. A previously initialized RNN data descriptor. Must match or be the exact same descriptor previously passed into cudnnRNNForwardTrainingEx().

y

Input. Data pointer to GPU memory associated with the output tensor descriptor yDesc. Must contain the exact same data previously produced by cudnnRNNForwardTrainingEx().

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

dwDesc

Input. Handle to a previously initialized filter descriptor describing the gradients of the weights for the RNN.

dw

Input/Output. Data pointer to GPU memory associated with the filter descriptor dwDesc.

reserveSpace

Input. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

The function does not support the provided configuration.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors xDesc, yDesc, hxDesc, dwDesc is invalid, or has incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.23. cudnnRNNForwardTraining()

This function is deprecated starting in cuDNN 8.0.0.

Use cudnnRNNForward() instead of cudnnRNNForwardTraining().

cudnnStatus_t cudnnRNNForwardTraining(
    cudnnHandle_t                   handle,
    const cudnnRNNDescriptor_t      rnnDesc,
    const int                       seqLength,
    const cudnnTensorDescriptor_t  *xDesc,
    const void                     *x,
    const cudnnTensorDescriptor_t   hxDesc,
    const void                     *hx,
    const cudnnTensorDescriptor_t   cxDesc,
    const void                     *cx,
    const cudnnFilterDescriptor_t   wDesc,
    const void                     *w,
    const cudnnTensorDescriptor_t  *yDesc,
    void                           *y,
    const cudnnTensorDescriptor_t   hyDesc,
    void                           *hy,
    const cudnnTensorDescriptor_t   cyDesc,
    void                           *cy,
    void                           *workspace,
    size_t                          workSpaceSizeInBytes,
    void                           *reserveSpace,
    size_t                          reserveSpaceSizeInBytes)

This routine executes the recurrent neural network described by rnnDesc with inputs x, hx, and cx, weights w and outputs y, hy, and cy. workspace is required for intermediate storage. reserveSpace stores data required for training. The same reserveSpace data must be used for future calls to cudnnRNNBackwardData() and cudnnRNNBackwardWeights() if these execute on the same input data.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

seqLength

Input. Number of iterations to unroll over. The value of this seqLength must not exceed the value that was used in cudnnGetRNNWorkspaceSize() function for querying the workspace size required to execute the RNN.

xDesc
Input. An array of seqLength fully packed tensor descriptors. Each descriptor in the array should have three dimensions that describe the input data format to one recurrent iteration (one descriptor per RNN time-step). The first dimension (batch size) of the tensors may decrease from iteration element n to iteration element n+1 but may not increase. Each tensor descriptor must have the same second dimension (RNN input vector length, inputSize). The third dimension of each tensor should be 1. Input vectors are expected to be arranged in the column-major order so strides in xDesc should be set as follows:
strideA[0]=inputSize, strideA[1]=1, strideA[2]=1
x

Input. Data pointer to GPU memory associated with the array of tensor descriptors xDesc. The input vectors are expected to be packed contiguously with the first vector of iterations (time-step) n+1 following directly the last vector of iteration n. In other words, input vectors for all RNN time-steps should be packed in the contiguous block of GPU memory with no gaps between the vectors.

hxDesc
Input. A fully packed tensor descriptor describing the initial hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc
Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc
Input. An array of fully packed tensor descriptors describing the output from each recurrent iteration (one descriptor per iteration). The second dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the second dimension should match the hiddenSize argument.
  • If direction is CUDNN_BIDIRECTIONAL the second dimension should match double the hiddenSize argument.
The first dimension of the tensor n must match the first dimension of the tensor n in xDesc.
y

Output. Data pointer to GPU memory associated with the output tensor descriptor yDesc.

hyDesc
Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc
Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The first dimension of the tensor depends on the direction argument used to initialize rnnDesc:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.
The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.
cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors hxDesc, cxDesc, wDesc, hyDesc, cyDesc or one of the descriptors in xDesc, yDesc is invalid.
  • The descriptors in one of xDesc, hxDesc, cxDesc, wDesc, yDesc, hyDesc, cyDesc have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.24. cudnnRNNForwardTrainingEx()

This function has been deprecated starting in cuDNN 8.0. Use cudnnRNNForward() instead of cudnnRNNForwardTrainingEx().

cudnnStatus_t cudnnRNNForwardTrainingEx(
    cudnnHandle_t                        handle,
    const cudnnRNNDescriptor_t           rnnDesc,
    const cudnnRNNDataDescriptor_t       xDesc,
    const void                           *x,
    const cudnnTensorDescriptor_t        hxDesc,
    const void                           *hx,
    const cudnnTensorDescriptor_t        cxDesc,
    const void                           *cx,
    const cudnnFilterDescriptor_t        wDesc,
    const void                           *w,
    const cudnnRNNDataDescriptor_t       yDesc,
    void                                 *y,
    const cudnnTensorDescriptor_t        hyDesc,
    void                                 *hy,
    const cudnnTensorDescriptor_t        cyDesc,
    void                                 *cy,
    const cudnnRNNDataDescriptor_t       kDesc,
    const void                           *keys,
    const cudnnRNNDataDescriptor_t       cDesc,
    void                                 *cAttn,
    const cudnnRNNDataDescriptor_t       iDesc,
    void                                 *iAttn,
    const cudnnRNNDataDescriptor_t       qDesc,
    void                                 *queries,
    void                                 *workSpace,
    size_t                               workSpaceSizeInBytes,
    void                                 *reserveSpace,
    size_t                               reserveSpaceSizeInBytes);

This routine is the extended version of the cudnnRNNForwardTraining() function. The cudnnRNNForwardTrainingEx() allows the user to use unpacked (padded) layout for input x and output y.

In the unpacked layout, each sequence in the mini-batch is considered to be of fixed length, specified by maxSeqLength in its corresponding RNNDataDescriptor. Each fixed-length sequence, for example, the nth sequence in the mini-batch, is composed of a valid segment specified by the seqLengthArray[n] in its corresponding RNNDataDescriptor; and a padding segment to make the combined sequence length equal to maxSeqLength.

With the unpacked layout, both sequence major (meaning, time major) and batch major are supported. For backward compatibility, the packed sequence major layout is supported. However, similar to the non-extended function cudnnRNNForwardTraining(), the sequences in the mini-batch need to be sorted in descending order according to length.

Parameters

handle

Input. Handle to a previously created cuDNN context.

rnnDesc

Input. A previously initialized RNN descriptor.

xDesc

Input. A previously initialized RNN Data descriptor. The dataType, layout, maxSeqLength, batchSize, and seqLengthArray need to match that of yDesc.

x

Input. Data pointer to the GPU memory associated with the RNN data descriptor xDesc. The input vectors are expected to be laid out in memory according to the layout specified by xDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

hxDesc

Input. A fully packed tensor descriptor describing the initial hidden state of the RNN.

The first dimension of the tensor depends on the direction argument used to initialize rnnDesc. Moreover:
  • If direction is CUDNN_UNIDIRECTIONAL then the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL then the first dimension should match double the numLayers argument.

The second dimension must match the batchSize parameter in xDesc. The third dimension depends on whether RNN mode is CUDNN_LSTM and whether LSTM projection is enabled. Additionally:

  • If RNN mode is CUDNN_LSTM and LSTM projection is enabled, the third dimension must match the recProjSize argument passed to cudnnSetRNNProjectionLayers() call used to set rnnDesc.
  • Otherwise, the third dimension must match the hiddenSize argument used to initialize rnnDesc.
hx

Input. Data pointer to GPU memory associated with the tensor descriptor hxDesc. If a NULL pointer is passed, the initial hidden state of the network will be initialized to zero.

cxDesc

Input. A fully packed tensor descriptor describing the initial cell state for LSTM networks.

The first dimension of the tensor depends on the direction argument used to initialize rnnDesc. Additionally:
  • If direction is CUDNN_UNIDIRECTIONAL the first dimension should match the numLayers argument.
  • If direction is CUDNN_BIDIRECTIONAL the first dimension should match double the numLayers argument.

The second dimension must match the first dimension of the tensors described in xDesc. The third dimension must match the hiddenSize argument used to initialize rnnDesc. The tensor must be fully packed.

cx

Input. Data pointer to GPU memory associated with the tensor descriptor cxDesc. If a NULL pointer is passed, the initial cell state of the network will be initialized to zero.

wDesc

Input. Handle to a previously initialized filter descriptor describing the weights for the RNN.

w

Input. Data pointer to GPU memory associated with the filter descriptor wDesc.

yDesc

Input. A previously initialized RNN data descriptor. The dataType, layout, maxSeqLength, batchSize, and seqLengthArray need to match that of dyDesc and dxDesc. The parameter vectorSize depends on whether the RNN mode is CUDNN_LSTM and whether LSTM projection is enabled and whether the network is bidirectional. Specifically:

  • For unidirectional network, if the RNN mode is CUDNN_LSTM and LSTM projection is enabled, the parameter vectorSize must match the recProjSize argument passed to cudnnSetRNNProjectionLayers() call used to set rnnDesc. If the network is bidirectional, then multiply the value by 2.
  • Otherwise, for unidirectional network, the parameter vectorSize must match the hiddenSize argument used to initialize rnnDesc. If the network is bidirectional, then multiply the value by 2.
y

Output. Data pointer to GPU memory associated with the RNN data descriptor yDesc. The input vectors are expected to be laid out in memory according to the layout specified by yDesc. The elements in the tensor (including elements in the padding vector) must be densely packed, and no strides are supported.

hyDesc

Input. A fully packed tensor descriptor describing the final hidden state of the RNN. The descriptor must be set exactly the same as hxDesc.

hy

Output. Data pointer to GPU memory associated with the tensor descriptor hyDesc. If a NULL pointer is passed, the final hidden state of the network will not be saved.

cyDesc

Input. A fully packed tensor descriptor describing the final cell state for LSTM networks. The descriptor must be set exactly the same as cxDesc.

cy

Output. Data pointer to GPU memory associated with the tensor descriptor cyDesc. If a NULL pointer is passed, the final cell state of the network will not be saved.

kDesc

Reserved. Users may pass in NULL.

keys

Reserved. Users may pass in NULL.

cDesc

Reserved. Users may pass in NULL.

cAttn

Reserved. Users may pass in NULL.

iDesc

Reserved. Users may pass in NULL.

iAttn

Reserved. Users may pass in NULL.

qDesc

Reserved. Users may pass in NULL.

queries

Reserved. Users may pass in NULL.

workspace

Input. Data pointer to GPU memory to be used as a workspace for this call.

workSpaceSizeInBytes

Input. Specifies the size in bytes of the provided workspace.

reserveSpace

Input/Output. Data pointer to GPU memory to be used as a reserve space for this call.

reserveSpaceSizeInBytes

Input. Specifies the size in bytes of the provided reserveSpace.

Returns

CUDNN_STATUS_SUCCESS

The function launched successfully.

CUDNN_STATUS_NOT_SUPPORTED

At least one of the following conditions are met:

  • Variable sequence length input is passed in while CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used.
  • CUDNN_RNN_ALGO_PERSIST_STATIC or CUDNN_RNN_ALGO_PERSIST_DYNAMIC is used on pre-Pascal devices.
  • Double input/output is used for CUDNN_RNN_ALGO_PERSIST_STATIC.
CUDNN_STATUS_BAD_PARAM

At least one of the following conditions are met:

  • The descriptor rnnDesc is invalid.
  • At least one of the descriptors xDesc, yDesc, hxDesc, cxDesc, wDesc, hyDesc, and cyDesc is invalid, or have incorrect strides or dimensions.
  • workSpaceSizeInBytes is too small.
  • reserveSpaceSizeInBytes is too small.
CUDNN_STATUS_EXECUTION_FAILED

The function failed to launch on the GPU.

CUDNN_STATUS_ALLOC_FAILED

The function was unable to allocate memory.

8.2.25. cudnnSetCTCLossDescriptor()

cudnnStatus_t cudnnSetCTCLossDescriptor(
    cudnnCTCLossDescriptor_t        ctcLossDesc,
    cudnnDataType_t                 compType)

This function sets a CTC loss function descriptor. See also the extended version cudnnSetCTCLossDescriptorEx() to set the input normalization mode.

When the extended version cudnnSetCTCLossDescriptorEx() is used with normMode set to CUDNN_LOSS_NORMALIZATION_NONE and the gradMode set to CUDNN_NOT_PROPAGATE_NAN, then it is the same as the current function cudnnSetCTCLossDescriptor(), meaning:
cudnnSetCtcLossDescriptor(*) = cudnnSetCtcLossDescriptorEx(*, normMode=CUDNN_LOSS_NORMALIZATION_NONE, gradMode=CUDNN_NOT_PROPAGATE_NAN)

Parameters

ctcLossDesc

Output. CTC loss descriptor to be set.

compType

Input. Compute type for this CTC loss function.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

At least one of input parameters passed is invalid.

8.2.26. cudnnSetCTCLossDescriptorEx()

cudnnStatus_t cudnnSetCTCLossDescriptorEx(
    cudnnCTCLossDescriptor_t        ctcLossDesc,
    cudnnDataType_t                 compType,
    cudnnLossNormalizationMode_t    normMode,
    cudnnNanPropagation_t           gradMode)

This function is an extension of cudnnSetCTCLossDescriptor(). This function provides an additional interface normMode to set the input normalization mode for the CTC loss function, and gradMode to control the NaN propagation type.

When this function cudnnSetCTCLossDescriptorEx() is used with normMode set to CUDNN_LOSS_NORMALIZATION_NONE and the gradMode set to CUDNN_NOT_PROPAGATE_NAN, then it is the same as cudnnSetCTCLossDescriptor(), meaning:
cudnnSetCtcLossDescriptor(*) = cudnnSetCtcLossDescriptorEx(*, normMode=CUDNN_LOSS_NORMALIZATION_NONE, gradMode=CUDNN_NOT_PROPAGATE_NAN)

Parameters

ctcLossDesc

Output. CTC loss descriptor to be set.

compType

Input. Compute type for this CTC loss function.

normMode

Input. Input normalization type for this CTC loss function. For more information, see cudnnLossNormalizationMode_t.

gradMode

Input. NaN propagation type for this CTC loss function. For L the sequence length, R the number of repeated letters in the sequence, and T the length of sequential data, the following applies: when a sample with L+R > T is encountered during the gradient calculation, if gradMode is set to CUDNN_PROPAGATE_NAN (see cudnnNanPropagation_t), then the CTC loss function does not write to the gradient buffer for that sample. Instead, the current values, even not finite, are retained. If gradMode is set to CUDNN_NOT_PROPAGATE_NAN, then the gradient for that sample is set to zero. This guarantees finite gradient.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

At least one of input parameters passed is invalid.

8.2.27. cudnnSetCTCLossDescriptor_v8()

cudnnStatus_t cudnnSetCTCLossDescriptorEx(
    cudnnCTCLossDescriptor_t        ctcLossDesc,
    cudnnDataType_t                 compType,
    cudnnLossNormalizationMode_t    normMode,
    cudnnNanPropagation_t           gradMode,
    int                             maxLabelLength)

Many CTC API functions are updated in cuDNN version 8.0.0 to support CUDA graphs. In order to do so, a new parameter is needed, maxLabelLength. Now that label and input data are assumed to be in GPU memory, this information is not otherwise readily available.

Parameters

ctcLossDesc

Output. CTC loss descriptor to be set.

compType

Input. Compute type for this CTC loss function.

normMode

Input. Input normalization type for this CTC loss function. For more information, see cudnnLossNormalizationMode_t.

gradMode

Input. NaN propagation type for this CTC loss function. For L the sequence length, R the number of repeated letters in the sequence, and T the length of sequential data, the following applies: when a sample with L+R > T is encountered during the gradient calculation, if gradMode is set to CUDNN_PROPAGATE_NAN (see cudnnNanPropagation_t), then the CTC loss function does not write to the gradient buffer for that sample. Instead, the current values, even not finite, are retained. If gradMode is set to CUDNN_NOT_PROPAGATE_NAN, then the gradient for that sample is set to zero. This guarantees finite gradient.

maxLabelLength

Input. The maximum label length from the labels data.

Returns

CUDNN_STATUS_SUCCESS

The function returned successfully.

CUDNN_STATUS_BAD_PARAM

At least one of input parameters passed is invalid.

Notice

Notice

This document is provided for information purposes only and shall not be regarded as a warranty of a certain functionality, condition, or quality of a product. NVIDIA Corporation (“NVIDIA”) makes no representations or warranties, expressed or implied, as to the accuracy or completeness of the information contained in this document and assumes no responsibility for any errors contained herein. NVIDIA shall have no liability for the consequences or use of such information or for any infringement of patents or other rights of third parties that may result from its use. This document is not a commitment to develop, release, or deliver any Material (defined below), code, or functionality.

NVIDIA reserves the right to make corrections, modifications, enhancements, improvements, and any other changes to this document, at any time without notice.

Customer should obtain the latest relevant information before placing orders and should verify that such information is current and complete.

NVIDIA products are sold subject to the NVIDIA standard terms and conditions of sale supplied at the time of order acknowledgement, unless otherwise agreed in an individual sales agreement signed by authorized representatives of NVIDIA and customer (“Terms of Sale”). NVIDIA hereby expressly objects to applying any customer general terms and conditions with regards to the purchase of the NVIDIA product referenced in this document. No contractual obligations are formed either directly or indirectly by this document.

NVIDIA products are not designed, authorized, or warranted to be suitable for use in medical, military, aircraft, space, or life support equipment, nor in applications where failure or malfunction of the NVIDIA product can reasonably be expected to result in personal injury, death, or property or environmental damage. NVIDIA accepts no liability for inclusion and/or use of NVIDIA products in such equipment or applications and therefore such inclusion and/or use is at customer’s own risk.

NVIDIA makes no representation or warranty that products based on this document will be suitable for any specified use. Testing of all parameters of each product is not necessarily performed by NVIDIA. It is customer’s sole responsibility to evaluate and determine the applicability of any information contained in this document, ensure the product is suitable and fit for the application planned by customer, and perform the necessary testing for the application in order to avoid a default of the application or the product. Weaknesses in customer’s product designs may affect the quality and reliability of the NVIDIA product and may result in additional or different conditions and/or requirements beyond those contained in this document. NVIDIA accepts no liability related to any default, damage, costs, or problem which may be based on or attributable to: (i) the use of the NVIDIA product in any manner that is contrary to this document or (ii) customer product designs.

No license, either expressed or implied, is granted under any NVIDIA patent right, copyright, or other NVIDIA intellectual property right under this document. Information published by NVIDIA regarding third-party products or services does not constitute a license from NVIDIA to use such products or services or a warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual property rights of the third party, or a license from NVIDIA under the patents or other intellectual property rights of NVIDIA.

Reproduction of information in this document is permissible only if approved in advance by NVIDIA in writing, reproduced without alteration and in full compliance with all applicable export laws and regulations, and accompanied by all associated conditions, limitations, and notices.

THIS DOCUMENT AND ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND SEPARATELY, “MATERIALS”) ARE BEING PROVIDED “AS IS.” NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE. TO THE EXTENT NOT PROHIBITED BY LAW, IN NO EVENT WILL NVIDIA BE LIABLE FOR ANY DAMAGES, INCLUDING WITHOUT LIMITATION ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL, PUNITIVE, OR CONSEQUENTIAL DAMAGES, HOWEVER CAUSED AND REGARDLESS OF THE THEORY OF LIABILITY, ARISING OUT OF ANY USE OF THIS DOCUMENT, EVEN IF NVIDIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Notwithstanding any damages that customer might incur for any reason whatsoever, NVIDIA’s aggregate and cumulative liability towards customer for the products described herein shall be limited in accordance with the Terms of Sale for the product.

VESA DisplayPort

DisplayPort and DisplayPort Compliance Logo, DisplayPort Compliance Logo for Dual-mode Sources, and DisplayPort Compliance Logo for Active Cables are trademarks owned by the Video Electronics Standards Association in the United States and other countries.

HDMI

HDMI, the HDMI logo, and High-Definition Multimedia Interface are trademarks or registered trademarks of HDMI Licensing LLC.

OpenCL

OpenCL is a trademark of Apple Inc. used under license to the Khronos Group Inc.

Trademarks

NVIDIA, the NVIDIA logo, and cuBLAS, CUDA, CUDA Toolkit, cuDNN, DALI, DIGITS, DGX, DGX-1, DGX-2, DGX Station, DLProf, GPU, JetPack, Jetson, Kepler, Maxwell, NCCL, Nsight Compute, Nsight Systems, NVCaffe, NVIDIA Ampere GPU architecture, NVIDIA Deep Learning SDK, NVIDIA Developer Program, NVIDIA GPU Cloud, NVLink, NVSHMEM, PerfWorks, Pascal, SDK Manager, T4, Tegra, TensorRT, TensorRT Inference Server, Tesla, TF-TRT, Triton Inference Server, Turing, and Volta are trademarks and/or registered trademarks of NVIDIA Corporation in the United States and other countries. Other company and product names may be trademarks of the respective companies with which they are associated.