Dnn.Components.Swiglu Module Reference

Exported Modules
module	Compute.DeviceTypeTraits
module	Dnn.TensorTypes
module	Dnn.Components.Gelu
module	Compute.OperationTraits
module	Serialization.Metadata
module	Compute.DeviceType
module	Dnn.ITensor
module	Serialization.Mode
module	Dnn.ComponentType
module	Compute.IExecutionContext
module	Compute.UnaryOperation
module	Dnn.TensorDataType
module	Dnn.Tensor
module	Compute.DeviceId
module	Compute.Device
module	Dnn.Components.SwigluConfig
module	Dnn.TensorDataTypeTraits
module	Serialization.ModelArchive
module	Compute.CpuMemoryResource
module	Dnn.Component
module	Compute.ExecutionContextFactory
module	Serialization.Tensor

Classes
class	Mila::Dnn::Swiglu< TDeviceType, TPrecision >
	SwiGLU activation component. More...

Typedefs
using	ComponentBase = Component<TDeviceType, TPrecision>
using	MR = typename DeviceTypeTraits<TDeviceType>::memory_resource
using	OpType = typename OperationTraits<OperationType::SwigluOp, TDeviceType, TPrecision>::type
using	TensorType = Tensor<TPrecision, MR>

Functions
	Swiglu (const std::string &name, const SwigluConfig &config, std::optional< DeviceId > device_id=std::nullopt)
	~Swiglu () override=default
TensorType &	backward (const TensorType &input, const TensorType &output_grad)
void	createOperation ()
TensorType &	forward (const TensorType &input)
static std::unique_ptr< Swiglu >	fromArchive_ (ModelArchive &archive, const std::string &component_name, IExecutionContext *exec_context)
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.
const ComponentType	getType () const override
	Get the component type identifier.
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.
size_t	parameterCount () const override
	Return number of trainable parameters.
void	save_ (ModelArchive &archive, SerializationMode mode) const override
void	synchronize () override
	Wait for outstanding device work submitted by this component.
std::string	toString () const override
	Produce a short, human-readable description of the component.
void	validateBuildContext (const BuildContext &build_context) const
static void	validateMetadata_ (const SerializationMetadata &meta, const std::string &component_name)

Variables
SwigluConfig	config_
std::unique_ptr< TensorType >	input_grad_ { nullptr }
shape_t	input_shape_
std::shared_ptr< OpType >	operation_ { nullptr }
std::unique_ptr< TensorType >	output_ { nullptr }
std::optional< TensorType >	output_view_
std::unique_ptr< IExecutionContext >	owned_exec_context_ { nullptr }

Files
file	/__w/Mila/Mila/Mila/Src/Dnn/Components/Activations/Swiglu/Swiglu.ixx
	SwiGLU activation component implementation.