tvm
Public Types | Public Member Functions | Static Public Member Functions | Public Attributes | Static Public Attributes | List of all members
tvm::meta_schedule::PySearchStrategyNode Class Reference

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 Member Functions

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)
 

Static Public Member Functions

static void RegisterReflection ()
 

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"
 

Detailed Description

The python side customizable class for measure candidate generation.

Member Typedef Documentation

◆ FClone

◆ FGenerateMeasureCandidates

◆ FInitializeWithTuneContext

◆ FNotifyRunnerResults

◆ FPostTuning

◆ 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
contextThe 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_candidatesThe candidates to be measured.
resultsThe 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_trialsThe maximum number of trials.
num_trials_per_iterThe number of trials per iteration.
design_spacesThe design spaces used during tuning process.
databaseThe database used during tuning process.
cost_modelThe 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.

◆ RegisterReflection()

static void tvm::meta_schedule::PySearchStrategyNode::RegisterReflection ( )
inlinestatic

◆ TVM_DECLARE_FINAL_OBJECT_INFO()

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

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: