CLQLSTMLayerNormalizationKernel Class Reference

Interface for the kernel to do layer normalization. More...

#include <CLQLSTMLayerNormalizationKernel.h>

Collaboration diagram for CLQLSTMLayerNormalizationKernel:

Public Member Functions
	CLQLSTMLayerNormalizationKernel ()
	Default constructor. More...
	CLQLSTMLayerNormalizationKernel (const CLQLSTMLayerNormalizationKernel &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
CLQLSTMLayerNormalizationKernel &	operator= (const CLQLSTMLayerNormalizationKernel &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
	CLQLSTMLayerNormalizationKernel (CLQLSTMLayerNormalizationKernel &&)=default
	Allow instances of this class to be moved. More...
CLQLSTMLayerNormalizationKernel &	operator= (CLQLSTMLayerNormalizationKernel &&)=default
	Allow instances of this class to be moved. More...
	~CLQLSTMLayerNormalizationKernel ()=default
	Default destructor. More...
void	configure (const ICLTensor *input, ICLTensor *output, const ICLTensor *weight, const ICLTensor *bias)
	Initialise the kernel's input and outputs. More...
void	configure (const CLCompileContext &compile_context, const ICLTensor *input, ICLTensor *output, const ICLTensor *weight, const ICLTensor *bias)
	Initialise the kernel's input and outputs. 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...
CLKernelType	type () const
	Returns the CL kernel type. 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...
void	add_5D_tensor_argument (unsigned int &idx, const ICLTensor *tensor, const Window &window)
	Add the passed 5D tensor's parameters to the object's kernel's arguments starting from the index idx. More...
void	add_3d_tensor_nhw_argument (unsigned int &idx, const ICLTensor *tensor)
	Add the passed NHW 3D tensor's parameters to the object's kernel's arguments by passing strides, dimensions and the offset to the first valid element in bytes. More...
void	add_4d_tensor_nhwc_argument (unsigned int &idx, const ICLTensor *tensor)
	Add the passed NHWC 4D tensor's parameters to the object's kernel's arguments by passing strides, dimensions and the offset to the first valid element in bytes. 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...
cl::NDRange	get_cached_gws () const
	Get the cached gws used to enqueue this kernel. More...
void	cache_gws (const cl::NDRange &gws)
	Cache the latest gws used to enqueue this kernel. 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...
bool	is_window_configured () const
	Function to check if the embedded window of this kernel has been configured. More...

Static Public Member Functions
static Status	validate (const ITensorInfo *input, const ITensorInfo *output, const ITensorInfo *weight, const ITensorInfo *bias)
	Static function to check if given info will lead to a valid configuration of CLQLSTMLayerNormalizationKernel. More...
Static Public Member Functions inherited from ICLKernel
constexpr static unsigned int	num_arguments_per_3d_tensor_nhw ()
	Returns the number of arguments enqueued per NHW 3D Tensor object. More...
constexpr static unsigned int	num_arguments_per_4d_tensor_nhwc ()
	Returns the number of arguments enqueued per NHWC 4D Tensor object. More...
constexpr static unsigned int	num_arguments_per_1D_array ()
	Returns the number of arguments enqueued per 1D array object. More...
constexpr static unsigned int	num_arguments_per_1D_tensor ()
	Returns the number of arguments enqueued per 1D tensor object. More...
constexpr static unsigned int	num_arguments_per_2D_tensor ()
	Returns the number of arguments enqueued per 2D tensor object. More...
constexpr static unsigned int	num_arguments_per_3D_tensor ()
	Returns the number of arguments enqueued per 3D tensor object. More...
constexpr static 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, bool use_dummy_work_items)
	Get the global work size given an execution window. More...

Detailed Description

Interface for the kernel to do layer normalization.

Definition at line 34 of file CLQLSTMLayerNormalizationKernel.h.

Constructor & Destructor Documentation

CLQLSTMLayerNormalizationKernel() [1/3]

CLQLSTMLayerNormalizationKernel

(

)

Default constructor.

Definition at line 91 of file CLQLSTMLayerNormalizationKernel.cpp.

    : _input(nullptr), _weight(nullptr), _bias(nullptr), _output(nullptr)
{
    _type = CLKernelType::ELEMENTWISE;
}

References arm_compute::ELEMENTWISE.

CLQLSTMLayerNormalizationKernel() [2/3]

CLQLSTMLayerNormalizationKernel ( const CLQLSTMLayerNormalizationKernel &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

CLQLSTMLayerNormalizationKernel() [3/3]

CLQLSTMLayerNormalizationKernel ( CLQLSTMLayerNormalizationKernel &&  )

default

Allow instances of this class to be moved.

~CLQLSTMLayerNormalizationKernel()

~CLQLSTMLayerNormalizationKernel ( )

default

Default destructor.

Member Function Documentation

configure() [1/2]

void configure	(	const CLCompileContext &	compile_context,
		const ICLTensor *	input,
		ICLTensor *	output,
		const ICLTensor *	weight,
		const ICLTensor *	bias
	)

Initialise the kernel's input and outputs.

Parameters

[in]	compile_context	The compile context to be used.
[in]	input	Source tensor with 2 dimensions. Data types supported: QSYMM16.
[out]	output	Destination tensor. Data type supported: same as input
[in]	weight	Weight tensor. Data types supported: Same as input.
[in]	bias	Bias tensor. Data types supported: S32.

Definition at line 97 of file CLQLSTMLayerNormalizationKernel.cpp.

{
    ARM_COMPUTE_ERROR_ON_NULLPTR(input, weight, bias, output);
    auto padding_info = get_padding_info({input, weight, bias, output});
 
    ARM_COMPUTE_ERROR_THROW_ON(validate_arguments(input->info(), output->info(), weight->info(), bias->info()));
 
    _input  = input;
    _weight = weight;
    _bias   = bias;
    _output = output;
 
    const uint32_t num_elems_processed_per_iteration = max_cl_vector_width / input->info()->element_size();
 
    int32_t                       output_multiplier{};
    int32_t                       output_shift{};
    const UniformQuantizationInfo quan_info = _weight->info()->quantization_info().uniform();
    const Status                  status =
        quantization::calculate_quantized_multiplier(quan_info.scale, &output_multiplier, &output_shift);
    output_shift *= -1;
 
    // Set build options
    CLBuildOptions build_opts;
    build_opts.add_option("-DDATA_TYPE=" + get_cl_type_from_data_type(input->info()->data_type()));
    build_opts.add_option("-DVEC_SIZE=" + support::cpp11::to_string(num_elems_processed_per_iteration));
    build_opts.add_option("-DWIDTH=" + support::cpp11::to_string(input->info()->dimension(0)));
    build_opts.add_option("-DOUTPUT_MULTIPLIER=" + support::cpp11::to_string(output_multiplier));
    build_opts.add_option("-DOUTPUT_SHIFT=" + support::cpp11::to_string(output_shift));
    build_opts.add_option("-DMIN_BOUND=" +
                          support::cpp11::to_string(std::get<0>(
                              quantization::get_min_max_values_from_quantized_data_type(input->info()->data_type()))));
    build_opts.add_option("-DMAX_BOUND=" +
                          support::cpp11::to_string(std::get<1>(
                              quantization::get_min_max_values_from_quantized_data_type(input->info()->data_type()))));
 
    // Create kernel
    _kernel = create_kernel(compile_context, "qlstm_layer_normalization", build_opts.options());
 
    // Configure kernel window
    auto win_config = validate_and_configure_window(input->info(), output->info());
    ARM_COMPUTE_ERROR_THROW_ON(win_config.first);
    ICLKernel::configure_internal(win_config.second);
 
    // Set config_id for enabling LWS tuning
    _config_id = "qlstm_layer_normalization_";
    _config_id += lower_string(string_from_data_type(input->info()->data_type()));
    _config_id += "_";
    _config_id += support::cpp11::to_string(input->info()->dimension(0));
    _config_id += "_";
    _config_id += support::cpp11::to_string(input->info()->dimension(1));
    ARM_COMPUTE_ERROR_ON(has_padding_changed(padding_info));
}

References CLBuildOptions::add_option(), ARM_COMPUTE_ERROR_ON, ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, bias, arm_compute::quantization::calculate_quantized_multiplier(), arm_compute::create_kernel(), arm_compute::get_cl_type_from_data_type(), arm_compute::quantization::get_min_max_values_from_quantized_data_type(), arm_compute::get_padding_info(), arm_compute::has_padding_changed(), ITensor::info(), arm_compute::test::validation::input, arm_compute::lower_string(), num_elems_processed_per_iteration, CLBuildOptions::options(), ITensorInfo::quantization_info(), UniformQuantizationInfo::scale, arm_compute::string_from_data_type(), arm_compute::support::cpp11::to_string(), QuantizationInfo::uniform(), arm_compute::cpu::kernels::validate_and_configure_window(), and arm_compute::cpu::kernels::validate_arguments().

configure() [2/2]

void configure	(	const ICLTensor *	input,
		ICLTensor *	output,
		const ICLTensor *	weight,
		const ICLTensor *	bias
	)

Initialise the kernel's input and outputs.

Parameters

[in]	input	Source tensor with 2 dimensions. Data types supported: QSYMM16.
[out]	output	Destination tensor. Data type supported: same as input
[in]	weight	Weight tensor. Data types supported: Same as input.
[in]	bias	Bias tensor. Data types supported: S32.

Definition at line 154 of file CLQLSTMLayerNormalizationKernel.cpp.

{
    configure(CLKernelLibrary::get().get_compile_context(), input, output, weight, bias);
}

References bias, CLKernelLibrary::get(), and arm_compute::test::validation::input.

operator=() [1/2]

CLQLSTMLayerNormalizationKernel& operator= ( CLQLSTMLayerNormalizationKernel &&  )

default

Allow instances of this class to be moved.

operator=() [2/2]

CLQLSTMLayerNormalizationKernel& operator= ( const CLQLSTMLayerNormalizationKernel &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

run()

void run ( const Window &  window, cl::CommandQueue &  queue  )

overridevirtual

Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.

Note

The queue is not flushed by this method, and therefore the kernel will not have been executed by the time this method returns.

Parameters

[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 172 of file CLQLSTMLayerNormalizationKernel.cpp.

{
    ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(this);
    ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(ICLKernel::window(), window);
 
    Window slice = window.first_slice_window_2D();
    // Set slice step equal to width to force gws[0] to 1, as each thread normalizes across all rows
    slice.set_dimension_step(Window::DimX, _input->info()->dimension(0));
 
    Window weight_window;
    Window weight_slice;
 
    weight_window.use_tensor_dimensions(_weight->info()->tensor_shape());
    weight_slice = weight_window.first_slice_window_1D();
 
    do
    {
        unsigned int idx = 0;
        add_2D_tensor_argument(idx, _input, slice);
        add_1D_tensor_argument(idx, _weight, weight_slice);
        add_1D_tensor_argument(idx, _bias, weight_slice);
        add_2D_tensor_argument(idx, _output, slice);
 
        enqueue(queue, *this, slice, lws_hint());
    } while (window.slide_window_slice_2D(slice));
}

References ICLKernel::add_1D_tensor_argument(), ICLKernel::add_2D_tensor_argument(), ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW, ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL, ITensorInfo::dimension(), Window::DimX, arm_compute::enqueue(), Window::first_slice_window_1D(), Window::first_slice_window_2D(), ITensor::info(), ICLKernel::lws_hint(), arm_compute::test::validation::reference::slice(), Window::slide_window_slice_2D(), ITensorInfo::tensor_shape(), Window::use_tensor_dimensions(), and IKernel::window().

validate()

Status validate ( const ITensorInfo *  input, const ITensorInfo *  output, const ITensorInfo *  weight, const ITensorInfo *  bias  )

static

Static function to check if given info will lead to a valid configuration of CLQLSTMLayerNormalizationKernel.

Parameters

[in]	input	Source tensor info with 2 dimensions. Data types supported: QSYMM16.
[in]	output	Destination info tensor. Data type supported: same as input
[in]	weight	Weight info tensor. Data types supported: Same as input.
[in]	bias	Bias tensor info. Data types supported: S32.

Returns

a status

Definition at line 162 of file CLQLSTMLayerNormalizationKernel.cpp.

{
    ARM_COMPUTE_RETURN_ON_ERROR(validate_arguments(input, output, weight, bias));
    ARM_COMPUTE_RETURN_ON_ERROR(validate_and_configure_window(input->clone().get(), output->clone().get()).first);
    return Status{};
}

References ARM_COMPUTE_RETURN_ON_ERROR, bias, ICloneable< T >::clone(), arm_compute::test::validation::input, arm_compute::cpu::kernels::validate_and_configure_window(), and arm_compute::cpu::kernels::validate_arguments().

---

The documentation for this class was generated from the following files:

• src/core/CL/kernels/CLQLSTMLayerNormalizationKernel.h
• src/core/CL/kernels/CLQLSTMLayerNormalizationKernel.cpp