21.02
|
OpenCL kernel to fuse the batch normalization node to a preceding convolution node. More...
#include <CLFuseBatchNormalizationKernel.h>
Public Member Functions | |
CLFuseBatchNormalizationKernel () | |
Default constructor. More... | |
CLFuseBatchNormalizationKernel (const CLFuseBatchNormalizationKernel &)=delete | |
Prevent instances of this class from being copied (As this class contains pointers) More... | |
CLFuseBatchNormalizationKernel & | operator= (const CLFuseBatchNormalizationKernel &)=delete |
Prevent instances of this class from being copied (As this class contains pointers) More... | |
CLFuseBatchNormalizationKernel (CLFuseBatchNormalizationKernel &&)=default | |
Allow instances of this class to be moved. More... | |
CLFuseBatchNormalizationKernel & | operator= (CLFuseBatchNormalizationKernel &&)=default |
Allow instances of this class to be moved. More... | |
~CLFuseBatchNormalizationKernel ()=default | |
Default destructor. More... | |
void | configure (const ICLTensor *input_weights, const ICLTensor *bn_mean, const ICLTensor *bn_var, ICLTensor *fused_weights, ICLTensor *fused_bias, const ICLTensor *input_bias=nullptr, const ICLTensor *bn_beta=nullptr, const ICLTensor *bn_gamma=nullptr, float epsilon=0.001f, FuseBatchNormalizationType fbn_type=FuseBatchNormalizationType::CONVOLUTION) |
Set the source, destination of the kernel. More... | |
void | configure (const CLCompileContext &compile_context, const ICLTensor *input_weights, const ICLTensor *bn_mean, const ICLTensor *bn_var, ICLTensor *fused_weights, ICLTensor *fused_bias, const ICLTensor *input_bias=nullptr, const ICLTensor *bn_beta=nullptr, const ICLTensor *bn_gamma=nullptr, float epsilon=0.001f, FuseBatchNormalizationType fbn_type=FuseBatchNormalizationType::CONVOLUTION) |
Set the source, destination of the kernel. More... | |
void | run (const Window &window, cl::CommandQueue &queue) override |
Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. More... | |
Public Member Functions inherited from ICLKernel | |
ICLKernel () | |
Constructor. More... | |
cl::Kernel & | kernel () |
Returns a reference to the OpenCL kernel of this object. More... | |
template<typename T > | |
void | add_1D_array_argument (unsigned int &idx, const ICLArray< T > *array, const Strides &strides, unsigned int num_dimensions, const Window &window) |
Add the passed 1D array's parameters to the object's kernel's arguments starting from the index idx. More... | |
void | add_1D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 1D tensor's parameters to the object's kernel's arguments starting from the index idx. More... | |
void | add_1D_tensor_argument_if (bool cond, unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 1D tensor's parameters to the object's kernel's arguments starting from the index idx if the condition is true. More... | |
void | add_2D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx. More... | |
void | add_2D_tensor_argument_if (bool cond, unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx if the condition is true. More... | |
void | add_3D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx. More... | |
void | add_4D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window) |
Add the passed 4D tensor's parameters to the object's kernel's arguments starting from the index idx. More... | |
virtual void | run_op (ITensorPack &tensors, const Window &window, cl::CommandQueue &queue) |
Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. More... | |
template<typename T > | |
void | add_argument (unsigned int &idx, T value) |
Add the passed parameters to the object's kernel's arguments starting from the index idx. More... | |
void | set_lws_hint (const cl::NDRange &lws_hint) |
Set the Local-Workgroup-Size hint. More... | |
cl::NDRange | lws_hint () const |
Return the Local-Workgroup-Size hint. More... | |
void | set_wbsm_hint (const cl_int &wbsm_hint) |
Set the workgroup batch size modifier hint. More... | |
cl_int | wbsm_hint () const |
Return the workgroup batch size modifier hint. More... | |
const std::string & | config_id () const |
Get the configuration ID. More... | |
void | set_target (GPUTarget target) |
Set the targeted GPU architecture. More... | |
void | set_target (cl::Device &device) |
Set the targeted GPU architecture according to the CL device. More... | |
GPUTarget | get_target () const |
Get the targeted GPU architecture. More... | |
size_t | get_max_workgroup_size () |
Get the maximum workgroup size for the device the CLKernelLibrary uses. More... | |
template<unsigned int dimension_size> | |
void | add_tensor_argument (unsigned &idx, const ICLTensor *tensor, const Window &window) |
template<typename T , unsigned int dimension_size> | |
void | add_array_argument (unsigned &idx, const ICLArray< T > *array, const Strides &strides, unsigned int num_dimensions, const Window &window) |
Add the passed array's parameters to the object's kernel's arguments starting from the index idx. More... | |
Public Member Functions inherited from IKernel | |
IKernel () | |
Constructor. More... | |
virtual | ~IKernel ()=default |
Destructor. More... | |
virtual bool | is_parallelisable () const |
Indicates whether or not the kernel is parallelisable. More... | |
virtual BorderSize | border_size () const |
The size of the border for that kernel. More... | |
const Window & | window () const |
The maximum window the kernel can be executed on. More... | |
Static Public Member Functions | |
static Status | validate (const ITensorInfo *input_weights, const ITensorInfo *bn_mean, const ITensorInfo *bn_var, const ITensorInfo *fused_weights, const ITensorInfo *fused_bias, const ITensorInfo *input_bias=nullptr, const ITensorInfo *bn_beta=nullptr, const ITensorInfo *bn_gamma=nullptr, float epsilon=0.001f, FuseBatchNormalizationType fbn_type=FuseBatchNormalizationType::CONVOLUTION) |
Static function to check if given info will lead to a valid configuration of CLFuseBatchNormalizationKernel. More... | |
Static Public Member Functions inherited from ICLKernel | |
static constexpr unsigned int | num_arguments_per_1D_array () |
Returns the number of arguments enqueued per 1D array object. More... | |
static constexpr unsigned int | num_arguments_per_1D_tensor () |
Returns the number of arguments enqueued per 1D tensor object. More... | |
static constexpr unsigned int | num_arguments_per_2D_tensor () |
Returns the number of arguments enqueued per 2D tensor object. More... | |
static constexpr unsigned int | num_arguments_per_3D_tensor () |
Returns the number of arguments enqueued per 3D tensor object. More... | |
static constexpr unsigned int | num_arguments_per_4D_tensor () |
Returns the number of arguments enqueued per 4D tensor object. More... | |
static cl::NDRange | gws_from_window (const Window &window) |
Get the global work size given an execution window. More... | |
OpenCL kernel to fuse the batch normalization node to a preceding convolution node.
Definition at line 35 of file CLFuseBatchNormalizationKernel.h.
Default constructor.
Definition at line 102 of file CLFuseBatchNormalizationKernel.cpp.
|
delete |
Prevent instances of this class from being copied (As this class contains pointers)
|
default |
Allow instances of this class to be moved.
|
default |
Default destructor.
void configure | ( | const ICLTensor * | input_weights, |
const ICLTensor * | bn_mean, | ||
const ICLTensor * | bn_var, | ||
ICLTensor * | fused_weights, | ||
ICLTensor * | fused_bias, | ||
const ICLTensor * | input_bias = nullptr , |
||
const ICLTensor * | bn_beta = nullptr , |
||
const ICLTensor * | bn_gamma = nullptr , |
||
float | epsilon = 0.001f , |
||
FuseBatchNormalizationType | fbn_type = FuseBatchNormalizationType::CONVOLUTION |
||
) |
Set the source, destination of the kernel.
[in] | input_weights | Input weights tensor for convolution or depthwise convolution layer. Data type supported: F16/F32. Data layout supported: NCHW, NHWC |
[in] | bn_mean | Batch normalization layer mean tensor. Same as input_weights |
[in] | bn_var | Batch normalization layer variance tensor. Same as input_weights |
[out] | fused_weights | Output fused weights tensor. It can be a nullptr in case of in-place computation. Same as input_weights |
[out] | fused_bias | Output fused bias tensor. It can be a nullptr in case of in-place computation and input_bias != nullptr. Same as input_weights |
[in] | input_bias | (Optional) Input bias tensor for convolution or depthwise convolution layer. It can be a nullptr in case the bias tensor is not required. Same as input_weights |
[in] | bn_beta | (Optional) Batch normalization layer beta tensor. It can be a nullptr in case the beta tensor is not required. Same as input_weights |
[in] | bn_gamma | (Optional) Batch normalization layer gamma tensor. It can be a nullptr in case the gamma tensor is not required. Same as input_weights |
[in] | epsilon | (Optional) Batch normalization layer epsilon parameter. Defaults to 0.001f. |
[in] | fbn_type | (Optional) Fused batch normalization type. Defaults to CONVOLUTION. |
Definition at line 108 of file CLFuseBatchNormalizationKernel.cpp.
References CLKernelLibrary::get().
void configure | ( | const CLCompileContext & | compile_context, |
const ICLTensor * | input_weights, | ||
const ICLTensor * | bn_mean, | ||
const ICLTensor * | bn_var, | ||
ICLTensor * | fused_weights, | ||
ICLTensor * | fused_bias, | ||
const ICLTensor * | input_bias = nullptr , |
||
const ICLTensor * | bn_beta = nullptr , |
||
const ICLTensor * | bn_gamma = nullptr , |
||
float | epsilon = 0.001f , |
||
FuseBatchNormalizationType | fbn_type = FuseBatchNormalizationType::CONVOLUTION |
||
) |
Set the source, destination of the kernel.
[in] | compile_context | The compile context to be used. |
[in] | input_weights | Input weights tensor for convolution or depthwise convolution layer. Data type supported: F16/F32. Data layout supported: NCHW, NHWC |
[in] | bn_mean | Batch normalization layer mean tensor. Same as input_weights |
[in] | bn_var | Batch normalization layer variance tensor. Same as input_weights |
[out] | fused_weights | Output fused weights tensor. It can be a nullptr in case of in-place computation. Same as input_weights |
[out] | fused_bias | Output fused bias tensor. It can be a nullptr in case of in-place computation and input_bias != nullptr. Same as input_weights |
[in] | input_bias | (Optional) Input bias tensor for convolution or depthwise convolution layer. It can be a nullptr in case the bias tensor is not required. Same as input_weights |
[in] | bn_beta | (Optional) Batch normalization layer beta tensor. It can be a nullptr in case the beta tensor is not required. Same as input_weights |
[in] | bn_gamma | (Optional) Batch normalization layer gamma tensor. It can be a nullptr in case the gamma tensor is not required. Same as input_weights |
[in] | epsilon | (Optional) Batch normalization layer epsilon parameter. Defaults to 0.001f. |
[in] | fbn_type | (Optional) Fused batch normalization type. Defaults to CONVOLUTION. |
Definition at line 116 of file CLFuseBatchNormalizationKernel.cpp.
References CLBuildOptions::add_option(), CLBuildOptions::add_option_if(), ARM_COMPUTE_ERROR_ON, ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, arm_compute::auto_init_if_empty(), arm_compute::calculate_max_window(), ICloneable< T >::clone(), arm_compute::CONVOLUTION, arm_compute::create_kernel(), ITensorInfo::data_type(), ITensorInfo::dimension(), arm_compute::quantization::epsilon, arm_compute::float_to_string_with_full_precision(), arm_compute::get_cl_type_from_data_type(), arm_compute::get_padding_info(), arm_compute::has_padding_changed(), ITensor::info(), arm_compute::NHWC, CLBuildOptions::options(), arm_compute::support::cpp11::to_string(), and arm_compute::validate_arguments().
|
delete |
Prevent instances of this class from being copied (As this class contains pointers)
|
default |
Allow instances of this class to be moved.
|
overridevirtual |
Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.
[in] | window | Region on which to execute the kernel. (Must be a valid region of the window returned by window()). |
[in,out] | queue | Command queue on which to enqueue the kernel. |
Reimplemented from ICLKernel.
Definition at line 190 of file CLFuseBatchNormalizationKernel.cpp.
References ICLKernel::add_1D_tensor_argument(), ICLKernel::add_3D_tensor_argument(), ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, Window::collapse(), Window::DimZ, arm_compute::enqueue(), Window::first_slice_window_1D(), Window::first_slice_window_3D(), ICLKernel::lws_hint(), and IKernel::window().
|
static |
Static function to check if given info will lead to a valid configuration of CLFuseBatchNormalizationKernel.
[in] | input_weights | Input weights tensor info for convolution or depthwise convolution layer. Data type supported: F16/F32. Data layout supported: NCHW, NHWC |
[in] | bn_mean | Batch normalization layer mean tensor info. Same as input_weights |
[in] | bn_var | Batch normalization layer variance tensor info. Same as input_weights |
[in] | fused_weights | Output fused weights tensor info. It can be a nullptr in case of in-place computation. Same as input_weights |
[in] | fused_bias | Output fused bias tensor info. It can be a nullptr in case of in-place computation and input_bias != nullptr. Same as input_weights |
[in] | input_bias | (Optional) Input bias tensor info for convolution or depthwise convolution layer. It can be a nullptr in case the bias tensor is not required. Same as input_weights |
[in] | bn_beta | (Optional) Batch normalization layer beta tensor info. It can be a nullptr in case the beta tensor is not required. Same as input_weights |
[in] | bn_gamma | (Optional) Batch normalization layer gamma tensor info. It can be a nullptr in case the gamma tensor is not required. Same as input_weights |
[in] | epsilon | (Optional) Batch normalization layer epsilon parameter. Defaults to 0.001f. |
[in] | fbn_type | (Optional) Fused batch normalization type. Defaults to CONVOLUTION. |
Definition at line 181 of file CLFuseBatchNormalizationKernel.cpp.
References ARM_COMPUTE_RETURN_ON_ERROR, and arm_compute::validate_arguments().
Referenced by CLFuseBatchNormalization::validate().