Dnn.Components.Gelu Module Reference

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

Classes
class	Mila::Dnn::Gelu< TDeviceType, TPrecision >
	Gaussian Error Linear Unit (GELU) activation component. More...

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

Functions
	Gelu (const std::string &name, const GeluConfig &config, std::optional< DeviceId > device_id=std::nullopt)
	~Gelu () override=default
TensorType &	backward (const TensorType &input, const TensorType &output_grad)
	Compute gradients with respect to the component input.
void	createOperation ()
	Create backend UnaryOperation from OperationRegistry.
TensorType &	forward (const TensorType &input)
	Run the forward computation for this GELU component.
static std::unique_ptr< Gelu >	fromArchive_ (ModelArchive &archive, const std::string &component_name, IExecutionContext *exec_context)
ApproximationMethod	getApproximationMethod () const
	Return the configured GELU approximation method.
DeviceId	getDeviceId () const override
	Get the device identifier for this component.
std::vector< ITensor * >	getGradients () const override
	Get parameter gradient tensors.
MemoryStats	getMemoryStats () const override
	Return memory allocation breakdown.
std::vector< ITensor * >	getParameters () const override
	Get trainable parameter tensors.
const ComponentType	getType () const override
	Get the component type identifier.
void	onBuilding (const BuildContext &build_context) override
	Hook invoked during build() to initialize backend operation.
void	onExecutionContextSet () override
	Hook invoked after ExecutionContext is set.
void	onTrainingModeChanging (TrainingMode training_mode) override
	Hook invoked when training mode changes.
size_t	parameterCount () const override
	Number of trainable parameters.
void	save_ (ModelArchive &archive, SerializationMode mode) const override
	Persist component state to archive.
void	synchronize () override
	Wait for all asynchronous work submitted by this component to complete.
std::string	toString () const override
	Generate human-readable description of the component.
static void	validateMetadata_ (const SerializationMetadata &meta, const std::string &component_name)
	Validate metadata from archive during deserialization.

Variables
GeluConfig	config_
std::unique_ptr< TensorType >	input_grad_ { nullptr }
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/Gelu/Gelu.ixx
	GELU activation component implementation.