Gated-linear-unit (GLU-family) activation component. More...

Inheritance diagram for Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >:

Public Types
using	ComponentBase = Component<TDeviceType, TPrecision>
using	MR = typename DeviceTypeTraits<TDeviceType>::memory_resource
using	TensorType = Tensor<TPrecision, MR>

Public Member Functions
	Swiglu (const std::string &name, const SwigluConfig &config, std::optional< DeviceId > device_id=std::nullopt)
TensorType &	backward (const TensorType &input, const TensorType &output_grad)
TensorType &	forward (const TensorType &input)
DeviceId	getDeviceId () const override
	Get the compute device id associated with this component.
std::vector< ITensor * >	getGradients () const override
	Return non-owning pointers to parameter gradient tensors.
MemoryStats	getMemoryStats () const override
	Return memory allocation breakdown.
std::vector< ITensor * >	getParameters () const override
	Return non-owning pointers to parameter tensors.
MemoryStats	getRequiredMemory (const BuildContext &context) const override
	What onBuilding() would allocate for this context, without allocating.
const ComponentType	getType () const override
	Get the component type identifier.
void	installSharedOutput (std::shared_ptr< TensorType > output)
	Install a shared output slot (activation pooling).
dim_t	parameterCount () const override
	Return number of trainable parameters.
void	save_ (ModelArchive &archive, SerializationMode mode) const override
void	synchronize () override
std::string	toString () const override
	Produce a short, human-readable description of the component.
Public Member Functions inherited from Mila::Dnn::Component< TDeviceType, TPrecision >
	Component (const std::string &name)
	Construct component with required name identifier.
virtual void	build (const BuildContext &context) final
	Build the component with the provided BuildContext (canonical overload).
const std::string	getName () const
	Get the component's name identifier.
virtual std::vector< std::string >	getParameterNames () const
	List all available parameter names for this component.
TrainingMode	getTrainingMode () const noexcept
	The current runtime behavioral mode of this Component.
virtual bool	isBuilt () const final
	Returns true if build() has completed successfully.
virtual void	load_ (ModelArchive &archive, SerializationMode mode)
	Restore this component's parameters from its archive scope.
virtual void	loadParameter (const std::string &, const Serialization::ITensorBlob &)
	Load a parameter from serialized tensor data.
virtual void	requireSerializableParameters () const
	Verify this component can serialize whatever parameters it owns.
virtual void	saveFlatTensors (Serialization::SafeTensorsWriter &writer, const std::string &prefix, Serialization::TensorSavePass pass) const
	Drive this component's tensors through one pass of a flat safetensors save.
void	setTrainingMode (TrainingMode mode)
	Set the runtime behavioral mode for this Component.
virtual void	zeroGradients ()
	Clear all model-owned gradients for this component.

Static Public Member Functions
static std::unique_ptr< Swiglu >	fromArchive_ (ModelArchive &archive, const std::string &component_name, IExecutionContext *exec_context)
Static Public Member Functions inherited from Mila::Dnn::Component< TDeviceType, TPrecision >
static constexpr DeviceType	getDeviceType ()
	Compile-time device type for this component instance.
static constexpr TensorDataType	getPrecision () noexcept
	Compile-time tensor precision for this component instance.

Protected Member Functions
void	onBuilding (const BuildContext &build_context) override
	Hook invoked by build() to allocate component buffers.
void	onExecutionContextSet () override
	Lifecycle hook: Called immediately after ExecutionContext is set.
void	onTrainingModeChanging (TrainingMode training_mode) override
	Hook called before TrainingMode transitions.
Protected Member Functions inherited from Mila::Dnn::Component< TDeviceType, TPrecision >
IExecutionContext *	getExecutionContext () const
	Get the shared execution context.
bool	hasExecutionContext () const noexcept
	Check if execution context has been set.
template<TensorDataType TParameterPrecision, typename TMemoryResource>
void	loadParameterFromBlob (const std::string &param_name, const Serialization::ITensorBlob &blob, Tensor< TParameterPrecision, TMemoryResource > &target, const shape_t &expected_shape)
	Load a tensor blob into a parameter tensor with validation.
template<TensorDataType TParameterPrecision, typename TMemoryResource>
void	saveParameterToArchive (ModelArchive &archive, const std::string &parameter_name, const Tensor< TParameterPrecision, TMemoryResource > &parameter) const
	Write one parameter tensor into the archive under "tensors/<name>".
template<TensorDataType TParameterPrecision, typename TMemoryResource>
void	saveParameterToWriter (Serialization::SafeTensorsWriter &writer, const std::string &flat_name, const Tensor< TParameterPrecision, TMemoryResource > &parameter, Serialization::TensorSavePass pass) const
	Drive one parameter through one pass of a flat safetensors save.
void	setExecutionContext (IExecutionContext *context)
	Set the execution context for this component.

Additional Inherited Members
Protected Types inherited from Mila::Dnn::Component< TDeviceType, TPrecision >
using	HostStagingMemoryResource
	Host memory a device-resident parameter stages through.
Protected Attributes inherited from Mila::Dnn::Component< TDeviceType, TPrecision >
BuildContext	build_context_ { shape_t{ 1 }, RuntimeMode::Training }
	The BuildContext stored at build time.

Detailed Description

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>
requires PrecisionSupportedOnDevice<TPrecision, TDeviceType>
class Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >

Gated-linear-unit (GLU-family) activation component.

Splits the input along the feature axis into gate|value halves and computes out = TGate(gate) * value The gate function is a compile-time parameter: TGate = Silu realizes SwiGLU (the existing optimized SwigluOp), TGate = Gelu realizes GeGLU (GegluOp, the Gemma FFN). The default keeps the historical SwiGLU behavior unchanged.

Delegates work to the device operation selected by the gate (SwigluOp / GegluOp).

Member Function Documentation

◆ getDeviceId()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

DeviceId Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getDeviceId ( ) const

inlineoverridevirtual

Get the compute device id associated with this component.

Must return the device on which parameters and operations execute.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ getGradients()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

std::vector< ITensor * > Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getGradients ( ) const

inlineoverridevirtual

Return non-owning pointers to parameter gradient tensors.

Gradient buffers are allocated only when the component is built in training mode, so a component built for inference returns an empty vector. Stateless components return empty in either mode. This is the accessor counterpart to getParameters() and does not throw on mode.

Returns: Vector of gradient pointers; empty when built for inference.

Exceptions

std::runtime_error if called before the component has been built.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ getMemoryStats()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

MemoryStats Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getMemoryStats ( ) const

inlineoverridevirtual

Return memory allocation breakdown.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ getParameters()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

std::vector< ITensor * > Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getParameters ( ) const

inlineoverridevirtual

Return non-owning pointers to parameter tensors.

The returned tensor pointers remain valid for the lifetime of the component. Order should be canonical (weights before biases).

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ getRequiredMemory()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

MemoryStats Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getRequiredMemory ( const BuildContext & context ) const

inlineoverridevirtual

What onBuilding() would allocate for this context, without allocating.

The output feature dimension is halved: the input carries gate and value concatenated, and the gated product emits one of them. See Specifications/MemoryFootprint.md.

Reimplemented from Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ getType()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

const ComponentType Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::getType ( ) const

inlineoverridevirtual

Get the component type identifier.

Used for serialization and runtime type identification.

Returns: Component type enum value.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ installSharedOutput()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::installSharedOutput ( std::shared_ptr< TensorType > output )

inline

Install a shared output slot (activation pooling).

Must be called before build(): onBuilding then skips output self-allocation after validating the slot's storage covers the build shape. forward() already always returns a shape-adjusted view, so a wider slot never leaks its geometry to callers. Mirrors Linear::installSharedWeight; self-allocation remains the default. The slot is owned and memory-accounted by the installer.

◆ onBuilding()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::onBuilding ( const BuildContext & build_context )

inlineoverrideprotectedvirtual

Hook invoked by build() to allocate component buffers.

Receives the stored BuildContext. Implementations must use config.allocationSeqLen() when sizing output buffers – this is the single call that makes Inference and Training allocate the correct buffer sizes automatically without per-component logic.

// Example -- Linear component:
shape_t out_shape =
{
    config.batchSize(),
    config.allocationSeqLen(),   // 1 for Inference, T for Training
    config_.getOutputFeatures()
};
output_ = std::make_unique<TensorType>( device, out_shape,
    this->getName() + ".output" );

The default implementation forwards to the legacy onBuilding( const shape_t& ) overload for backwards compatibility. New components should override this overload directly.

Note: Do not call build() or onBuilding() from within this hook.; Implementations should either succeed fully or leave no partial state, as a failed build() may be retried.

Takes the build-time configuration; use its allocationSeqLen() to obtain the correct output buffer sequence dimension.

Reimplemented from Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ onExecutionContextSet()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::onExecutionContextSet ( )

inlineoverrideprotectedvirtual

Lifecycle hook: Called immediately after ExecutionContext is set.

Override this to perform initialization that requires a valid ExecutionContext. At the time this is called, getExecutionContext() is guaranteed to return a valid context.

Common uses:

Composite components: Create and configure child components.
Device resource allocation: Query device capabilities.

Default implementation does nothing.

Exceptions

Any	exception thrown will cause setExecutionContext() to fail and restore the component to a "context not set" state.

Reimplemented from Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ onTrainingModeChanging()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::onTrainingModeChanging ( TrainingMode training_mode )

inlineoverrideprotectedvirtual

Hook called before TrainingMode transitions.

Called by setTrainingMode() after validation and lock acquisition, before the internal state is updated. Derived classes override to respond to the transition – e.g. zeroing gradient buffers on transition to Eval, or re-enabling dropout on transition to Training.

The default implementation is a no-op.

Takes the incoming TrainingMode.

Reimplemented from Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ parameterCount()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

dim_t Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::parameterCount ( ) const

inlineoverridevirtual

Return number of trainable parameters.

For leaf components this is the element count of owned parameter tensors. CompositeComponent and Network implementations should return the recursive aggregate across all children.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ save_()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::save_	(	ModelArchive &	archive,
		SerializationMode	mode ) const

inlineoverridevirtual

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ synchronize()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

void Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::synchronize ( )

inlineoverridevirtual

         @brief Convenience accessor -- true if currently in Eval mode.

         Equivalent to getTrainingMode() == TrainingMode::Eval.
         Valid for both RuntimeMode::Inference and RuntimeMode::Training
         built components.

         @return true if in Eval mode.
        &zwj;/

bool isEvalMode() const noexcept { return getTrainingMode() == TrainingMode::Eval; }

    RuntimeMode getRuntimeMode() const noexcept
    {
        return build_context_.getRuntimeMode();
    }

    bool isInferenceMode() const noexcept
    {
        return build_context_.isInferenceMode();
    }

    bool isTrainingMode() const noexcept
    {
        return build_context_.isTrainingMode();
    }

====================================================================

Synchronization

    /**
       @brief Wait for outstanding device work submitted by this component.

       On CPU this may be a no-op. Use to ensure results are visible to
       the host or to measure synchronous timings.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

◆ toString()

template<DeviceType TDeviceType, TensorDataType TPrecision, ActivationType TGate = ActivationType::Silu>

std::string Mila::Dnn::Swiglu< TDeviceType, TPrecision, TGate >::toString ( ) const

inlineoverridevirtual

Produce a short, human-readable description of the component.

Implementations should keep output concise and avoid throwing.

Implements Mila::Dnn::Component< TDeviceType, TPrecision >.

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

Mila/Src/Dnn/Components/FFN/Swiglu/Swiglu.ixx

Public Types

Public Member Functions

Static Public Member Functions

Protected Member Functions

Additional Inherited Members

Detailed Description

Member Function Documentation

◆ getDeviceId()

◆ getGradients()

◆ getMemoryStats()

◆ getParameters()

◆ getRequiredMemory()

◆ getType()

◆ installSharedOutput()

◆ onBuilding()

◆ onExecutionContextSet()

◆ onTrainingModeChanging()

◆ parameterCount()

◆ save_()

◆ synchronize()

Synchronization

◆ toString()