Inheritance diagram for Digikam::SearchLlamaBackend:

Public Slots
void	slotCancel () override
void	slotRunInference (const QString &prompt) override
Public Slots inherited from Digikam::SearchLanguageBackend
virtual void	slotCancel ()
virtual void	slotRunInference (const QString &prompt)=0
	Run inference asynchronously.

Signals
void	signalRequestInference (const QString &prompt, int maxTokens, float temperature)
void	signalRequestLoad (const QString &path)
Signals inherited from Digikam::SearchLanguageBackend
void	signalInferenceCancelled ()
void	signalInferenceError (const QString &errorMessage)
void	signalInferenceProgress (int step)
void	signalModelLoaded (bool success)
void	signalModelLoadProgress (int percent)
void	signalRawOutputReady (const QString &output)

Public Member Functions
	SearchLlamaBackend (QObject *const parent=nullptr)
QString	backendName () const override
	Human-readable backend identifier ("llama.cpp", "onnx", "mock").
bool	isModelLoaded () const override
bool	loadModel (const QString &modelPath) override
	Load the model from a local path (resolved by SearchNlModelManager).
QString	modelPath () const override
void	setMaxTokens (int maxTokens)
void	setTemperature (float temperature)
void	unloadModel () override
Public Member Functions inherited from Digikam::SearchLanguageBackend
	SearchLanguageBackend (QObject *const parent=nullptr)

Member Function Documentation

◆ backendName()

QString Digikam::SearchLlamaBackend::backendName ( ) const

overridevirtual

Implements Digikam::SearchLanguageBackend.

◆ isModelLoaded()

bool Digikam::SearchLlamaBackend::isModelLoaded ( ) const

overridevirtual

Implements Digikam::SearchLanguageBackend.

◆ loadModel()

bool Digikam::SearchLlamaBackend::loadModel ( const QString & modelPath )

overridevirtual

Implementations should perform heavy loading off the GUI thread and report completion through signalModelLoaded(). Returns false immediately if the path is invalid.

Implements Digikam::SearchLanguageBackend.