The python side customizable class for measure candidate generation. More...

#include <search_strategy.h>

Inheritance diagram for tvm::meta_schedule::PySearchStrategyNode:

Collaboration diagram for tvm::meta_schedule::PySearchStrategyNode:

Public Types
using	FInitializeWithTuneContext = SearchStrategy::FInitializeWithTuneContext

using	FPreTuning = SearchStrategy::FPreTuning

using	FPostTuning = SearchStrategy::FPostTuning

using	FGenerateMeasureCandidates = SearchStrategy::FGenerateMeasureCandidates

using	FNotifyRunnerResults = SearchStrategy::FNotifyRunnerResults

using	FClone = SearchStrategy::FClone

Public Types inherited from tvm::runtime::Object
typedef void(*	FDeleter) (Object *self)
	Object deleter. More...

using	RefCounterType = std::atomic< int32_t >

Public Member Functions
void	VisitAttrs (tvm::AttrVisitor *v)

void	InitializeWithTuneContext (const TuneContext &context) final
	Initialize the search strategy with tuning context. More...

void	PreTuning (int max_trials, int num_trials_per_iter, const Array< tir::Schedule > &design_spaces, const Optional< Database > &database, const Optional< CostModel > &cost_model) final
	Pre-tuning for the search strategy. More...

void	PostTuning () final
	Post-tuning for the search strategy. More...

Optional< Array< MeasureCandidate > >	GenerateMeasureCandidates () final
	Generate measure candidates from design spaces for measurement. More...

void	NotifyRunnerResults (const Array< MeasureCandidate > &measure_candidates, const Array< RunnerResult > &results)
	Update the search strategy with measurement results. More...

SearchStrategy	Clone () const final
	Clone the search strategy. More...

	TVM_DECLARE_FINAL_OBJECT_INFO (PySearchStrategyNode, SearchStrategyNode)

Public Member Functions inherited from tvm::meta_schedule::SearchStrategyNode
virtual	~SearchStrategyNode ()=default
	Virtual destructor. More...

	TVM_DECLARE_BASE_OBJECT_INFO (SearchStrategyNode, Object)

Public Member Functions inherited from tvm::runtime::Object
uint32_t	type_index () const

std::string	GetTypeKey () const

size_t	GetTypeKeyHash () const

template<typename TargetType >
bool	IsInstance () const

bool	unique () const

	Object ()

	Object (const Object &other)

	Object (Object &&other)

Object &	operator= (const Object &other)

Object &	operator= (Object &&other)

Public Attributes
FInitializeWithTuneContext	f_initialize_with_tune_context
	The packed function to the `InitializeWithTuneContext` method. More...

FPreTuning	f_pre_tuning
	The packed function to the `PreTuning` method. More...

FPostTuning	f_post_tuning
	The packed function to the `PostTuning` method. More...

FGenerateMeasureCandidates	f_generate_measure_candidates
	The packed function to the `GenerateMeasureCandidates` method. More...

FNotifyRunnerResults	f_notify_runner_results
	The packed function to the `NotifyRunnerResults` method. More...

FClone	f_clone
	The packed function to the `Clone` method. More...

Static Public Attributes
static constexpr const char *	_type_key = "meta_schedule.PySearchStrategy"

Static Public Attributes inherited from tvm::meta_schedule::SearchStrategyNode
static constexpr const char *	_type_key = "meta_schedule.SearchStrategy"

Static Public Attributes inherited from tvm::runtime::Object
static constexpr const char *	_type_key = "runtime.Object"

static constexpr bool	_type_final = false

static constexpr uint32_t	_type_child_slots = 0

static constexpr bool	_type_child_slots_can_overflow = true

static constexpr bool	_type_has_method_visit_attrs = true

static constexpr bool	_type_has_method_sequal_reduce = false

static constexpr bool	_type_has_method_shash_reduce = false

static constexpr uint32_t	_type_index = TypeIndex::kDynamic

Additional Inherited Members
Static Public Member Functions inherited from tvm::runtime::Object
static std::string	TypeIndex2Key (uint32_t tindex)
	Get the type key of the corresponding index from runtime. More...

static size_t	TypeIndex2KeyHash (uint32_t tindex)
	Get the type key hash of the corresponding index from runtime. More...

static uint32_t	TypeKey2Index (const std::string &key)
	Get the type index of the corresponding key from runtime. More...

static uint32_t	_GetOrAllocRuntimeTypeIndex ()

static uint32_t	RuntimeTypeIndex ()

Protected Member Functions inherited from tvm::runtime::Object
void	IncRef ()
	developer function, increases reference counter. More...

void	DecRef ()
	developer function, decrease reference counter. More...

Static Protected Member Functions inherited from tvm::runtime::Object
static uint32_t	GetOrAllocRuntimeTypeIndex (const std::string &key, uint32_t static_tindex, uint32_t parent_tindex, uint32_t type_child_slots, bool type_child_slots_can_overflow)
	Get the type index using type key. More...

Protected Attributes inherited from tvm::runtime::Object
uint32_t	type_index_ {0}
	Type index(tag) that indicates the type of the object. More...

RefCounterType	ref_counter_ {0}
	The internal reference counter. More...

FDeleter	deleter_ = nullptr
	deleter of this object to enable customized allocation. If the deleter is nullptr, no deletion will be performed. The creator of the object must always set the deleter field properly. More...

Detailed Description

The python side customizable class for measure candidate generation.

Member Typedef Documentation

◆ FClone

using tvm::meta_schedule::PySearchStrategyNode::FClone = SearchStrategy::FClone

◆ FGenerateMeasureCandidates

using tvm::meta_schedule::PySearchStrategyNode::FGenerateMeasureCandidates = SearchStrategy::FGenerateMeasureCandidates

◆ FInitializeWithTuneContext

using tvm::meta_schedule::PySearchStrategyNode::FInitializeWithTuneContext = SearchStrategy::FInitializeWithTuneContext

◆ FNotifyRunnerResults

using tvm::meta_schedule::PySearchStrategyNode::FNotifyRunnerResults = SearchStrategy::FNotifyRunnerResults

◆ FPostTuning

using tvm::meta_schedule::PySearchStrategyNode::FPostTuning = SearchStrategy::FPostTuning

◆ FPreTuning

using tvm::meta_schedule::PySearchStrategyNode::FPreTuning = SearchStrategy::FPreTuning

Member Function Documentation

◆ Clone()

SearchStrategy tvm::meta_schedule::PySearchStrategyNode::Clone ( ) const

finalvirtual

Clone the search strategy.

Returns: The cloned search strategy.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ GenerateMeasureCandidates()

Optional<Array<MeasureCandidate> > tvm::meta_schedule::PySearchStrategyNode::GenerateMeasureCandidates ( )

finalvirtual

Generate measure candidates from design spaces for measurement.

Returns: The measure candidates generated, nullptr if finished.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ InitializeWithTuneContext()

void tvm::meta_schedule::PySearchStrategyNode::InitializeWithTuneContext ( const TuneContext & context )

finalvirtual

Initialize the search strategy with tuning context.

Parameters

context The tuning context for initialization.

Note: This method is supposed to be called only once before every other method.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ NotifyRunnerResults()

void tvm::meta_schedule::PySearchStrategyNode::NotifyRunnerResults	(	const Array< MeasureCandidate > &	measure_candidates,
		const Array< RunnerResult > &	results
	)

virtual

Update the search strategy with measurement results.

Parameters

measure_candidates	The candidates to be measured.
results	The measurement results from the runner.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ PostTuning()

void tvm::meta_schedule::PySearchStrategyNode::PostTuning ( )

finalvirtual

Post-tuning for the search strategy.

Note: Post-tuning is supposed to be called after the tuning process and before we reset the search strategy with another pre-tuning. Post-tuning can be empty.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ PreTuning()

void tvm::meta_schedule::PySearchStrategyNode::PreTuning	(	int	max_trials,
		int	num_trials_per_iter,
		const Array< tir::Schedule > &	design_spaces,
		const Optional< Database > &	database,
		const Optional< CostModel > &	cost_model
	)

finalvirtual

Pre-tuning for the search strategy.

Parameters

max_trials	The maximum number of trials.
num_trials_per_iter	The number of trials per iteration.
design_spaces	The design spaces used during tuning process.
database	The database used during tuning process.
cost_model	The cost model used during tuning process.

Note: Pre-tuning is supposed to be called before the tuning process and after the initialization. Because the search strategy is stateful, we can always call pretuning and reset the search strategy.

Implements tvm::meta_schedule::SearchStrategyNode.

◆ TVM_DECLARE_FINAL_OBJECT_INFO()

tvm::meta_schedule::PySearchStrategyNode::TVM_DECLARE_FINAL_OBJECT_INFO	(	PySearchStrategyNode	,
		SearchStrategyNode
	)

◆ VisitAttrs()

void tvm::meta_schedule::PySearchStrategyNode::VisitAttrs ( tvm::AttrVisitor * v )

inline

Member Data Documentation

◆ _type_key

constexpr const char* tvm::meta_schedule::PySearchStrategyNode::_type_key = "meta_schedule.PySearchStrategy"

staticconstexpr

◆ f_clone

FClone tvm::meta_schedule::PySearchStrategyNode::f_clone

The packed function to the Clone method.

◆ f_generate_measure_candidates

FGenerateMeasureCandidates tvm::meta_schedule::PySearchStrategyNode::f_generate_measure_candidates

The packed function to the GenerateMeasureCandidates method.

◆ f_initialize_with_tune_context

FInitializeWithTuneContext tvm::meta_schedule::PySearchStrategyNode::f_initialize_with_tune_context

The packed function to the InitializeWithTuneContext method.

◆ f_notify_runner_results

FNotifyRunnerResults tvm::meta_schedule::PySearchStrategyNode::f_notify_runner_results

The packed function to the NotifyRunnerResults method.

◆ f_post_tuning

FPostTuning tvm::meta_schedule::PySearchStrategyNode::f_post_tuning

The packed function to the PostTuning method.

◆ f_pre_tuning

FPreTuning tvm::meta_schedule::PySearchStrategyNode::f_pre_tuning

The packed function to the PreTuning method.

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

include/tvm/meta_schedule/search_strategy.h

Public Types

Public Member Functions

Public Attributes

Static Public Attributes

Additional Inherited Members

Detailed Description

Member Typedef Documentation

◆ FClone

◆ FGenerateMeasureCandidates

◆ FInitializeWithTuneContext

◆ FNotifyRunnerResults

◆ FPostTuning

◆ FPreTuning

Member Function Documentation

◆ Clone()

◆ GenerateMeasureCandidates()

◆ InitializeWithTuneContext()

◆ NotifyRunnerResults()

◆ PostTuning()

◆ PreTuning()

◆ TVM_DECLARE_FINAL_OBJECT_INFO()

◆ VisitAttrs()

Member Data Documentation

◆ _type_key

◆ f_clone

◆ f_generate_measure_candidates

◆ f_initialize_with_tune_context

◆ f_notify_runner_results

◆ f_post_tuning

◆ f_pre_tuning