Softmax module for neural networks. More...

Inheritance diagram for Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >:

Collaboration diagram for Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >:

[legend]

Public Types
using	ModuleBase = Module< TDeviceType, TInput, TOutput >
	Alias for base module type.

using	MR = std::conditional_t< TDeviceType==DeviceType::Cuda, CudaMemoryResource, CpuMemoryResource >
	Memory resource type used for tensors, selected based on device type.

Public Types inherited from Mila::Dnn::Module< TDeviceType, TInput, TOutput >
using	MR = std::conditional_t< TDeviceType==DeviceType::Cuda, CudaMemoryResource, CpuMemoryResource >

Public Member Functions
	Softmax (const std::string &device_name, const SoftmaxConfig &config)
	Constructs a new Softmax module with a device name.

	Softmax (std::shared_ptr< DeviceContext > device_context, const SoftmaxConfig &config)
	Constructs a new Softmax module with a provided device context.

void	backward (const Tensor< TInput, MR > &input, const Tensor< TOutput, MR > &output_grad, Tensor< TInput, MR > &input_grad)
	Performs the backward pass of the Softmax operation.

void	forward (const Tensor< TInput, MR > &input, Tensor< TOutput, MR > &output)
	Performs the forward pass of the softmax operation.

int64_t	getAxis () const
	Gets the axis used for softmax computation.

void	load (ModelArchive &archive) override
	Deserializes the module state from a ZIP archive.

size_t	parameterCount () const override
	Gets the number of trainable parameters in this module.

void	save (ModelArchive &zip) const override
	Serializes the module state to a ZIP archive.

std::string	toString () const override
	Generates a string representation of this module's configuration.

Public Member Functions inherited from Mila::Dnn::Module< TDeviceType, TInput, TOutput >
	Module (const std::string &device_name, const ComponentConfig &config)
	Constructor with device name.

	Module (std::shared_ptr< DeviceContext > context, const ComponentConfig &config)
	Constructor with a specific device context.

virtual	~Module ()=default
	Virtual destructor for proper cleanup in derived classes.

std::shared_ptr< Compute::DeviceContext >	getDeviceContext () const
	Get the device context for this module.

Compute::DeviceType	getDeviceType () const
	Get the device type of the current device context.

std::string	getName () const
	Get the name of the module.

const auto &	getParameterTensors () const
	Get the parameter tensors of this module.

const ComputePrecision::Policy &	getPrecision () const

const auto &	getStateTensors () const
	Get the state tensors of this module.

bool	isTraining () const
	Check if the module is in training mode.

virtual void	setTraining (bool is_training)
	Set the training mode of this module.

Private Member Functions
void	createOperation ()
	Creates the appropriate softmax operation for the current device.

Private Attributes
OperationAttributes	attributes_
	Operation attributes and configuration.

SoftmaxConfig	config_
	Configuration for the Softmax module.

std::shared_ptr< UnaryOperation< TDeviceType, TInput, TOutput > >	operation_ { nullptr }
	The operation that implements the softmax calculation.

std::vector< std::shared_ptr< Tensor< TOutput, MR > > >	output_state_
	Collection of output state tensors for caching.

std::vector< std::shared_ptr< Tensor< TInput, MR > > >	parameters_
	Collection of parameters for this module (empty for Softmax).

Additional Inherited Members
Protected Member Functions inherited from Mila::Dnn::Module< TDeviceType, TInput, TOutput >
const std::string	parametersToString () const
	Helper method to convert parameters to string representation.

const std::string	stateToString () const
	Helper method to convert state tensors to string representation.

Protected Attributes inherited from Mila::Dnn::Module< TDeviceType, TInput, TOutput >
std::unordered_map< std::string, std::shared_ptr< Tensor< TOutput, MR > > >	parameter_map_ = {}
	Map of parameter names to parameter tensors.

std::unordered_map< std::string, std::shared_ptr< Tensor< TOutput, MR > > >	state_map_ = {}
	Map of state names to state tensors.

Detailed Description

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>
requires ValidFloatTensorTypes<TInput, TOutput>
class Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >

Softmax module for neural networks.

This class implements the softmax function, which is often used in the final layer of a neural network to convert raw scores into probabilities. The softmax operation normalizes the input values by applying:

softmax(x_i) = exp(x_i) / sum(exp(x_j)) for all j

where the sum is computed over the specified axis. This normalization ensures all values sum to 1, allowing them to be interpreted as probabilities for classification tasks.

Template Parameters

TDeviceType	The device type (CPU or CUDA) on which to perform computations.
TInput	The data type of the input tensor elements.
TOutput	The data type of the output tensor elements, defaults to TInput.

Member Typedef Documentation

◆ ModuleBase

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

using Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::ModuleBase = Module<TDeviceType, TInput, TOutput>

export

Alias for base module type.

◆ MR

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

using Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::MR = std::conditional_t<TDeviceType == DeviceType::Cuda, CudaMemoryResource, CpuMemoryResource>

export

Memory resource type used for tensors, selected based on device type.

Constructor & Destructor Documentation

◆ Softmax() [1/2]

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::Softmax	(	const std::string &	device_name,
		const SoftmaxConfig &	config
	)

inlineexplicitexport

Constructs a new Softmax module with a device name.

Creates a new DeviceContext internally using the provided device name. This constructor is useful for creating standalone modules without pre-existing device contexts.

Parameters

device_name	The name of the device to use (e.g., "CPU", "CUDA:0").
config	Configuration parameters for the Softmax module.

Exceptions

std::invalid_argument	If the device name is invalid or the configuration is invalid
std::runtime_error	If device type doesn't match template parameter TDeviceType

Here is the call graph for this function:

◆ Softmax() [2/2]

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::Softmax	(	std::shared_ptr< DeviceContext >	device_context,
		const SoftmaxConfig &	config
	)

inlineexplicitexport

Constructs a new Softmax module with a provided device context.

Uses a pre-existing DeviceContext instance. This constructor is useful when integrating the module into a larger network that shares device contexts across modules.

Parameters

device_context	The device context to use for this module.
config	Configuration parameters for the Softmax module.

Exceptions

std::invalid_argument	If device_context is null or configuration is invalid
std::runtime_error	If device context type doesn't match template parameter TDeviceType

Here is the call graph for this function:

Member Function Documentation

◆ backward()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

void Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::backward	(	const Tensor< TInput, MR > &	input,
		const Tensor< TOutput, MR > &	output_grad,
		Tensor< TInput, MR > &	input_grad
	)

inlineexport

Performs the backward pass of the Softmax operation.

Computes the gradient of the softmax function with respect to its inputs. The gradient of softmax is more complex than most activations because each output depends on all inputs in the same dimension.

Parameters

input	The input tensor from the forward pass.
output_grad	The gradient of loss with respect to the output.
input_grad	The tensor to store gradients with respect to input.

◆ createOperation()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

void Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::createOperation ( )

inlineexportprivate

Creates the appropriate softmax operation for the current device.

Instantiates either a CPU or CUDA softmax operation based on the device type. Sets the axis attribute needed by the operation to properly apply softmax along the specified dimension.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ forward()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

void Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::forward	(	const Tensor< TInput, MR > &	input,
		Tensor< TOutput, MR > &	output
	)

inlineexport

Performs the forward pass of the softmax operation.

Computes the softmax of the input tensor along the specified axis and writes the result to the output tensor. The operation exponentiates each element and then normalizes by the sum of all exponentiated values along the specified axis.

Parameters

input	The input tensor to apply softmax to.
output	The tensor where softmax results will be stored.

◆ getAxis()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

int64_t Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::getAxis ( ) const

inlineexport

Gets the axis used for softmax computation.

Returns: int64_t The axis along which softmax is applied.

Here is the call graph for this function:

◆ load()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

void Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::load ( ModelArchive & archive )

inlineoverrideexportvirtual

Deserializes the module state from a ZIP archive.

Implementation of the Module interface for deserialization. Since Softmax has no learnable parameters, this is a no-op implementation.

Parameters

zip	ZIP archive for deserialization

Implements Mila::Dnn::Module< TDeviceType, TInput, TOutput >.

◆ parameterCount()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

size_t Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::parameterCount ( ) const

inlineoverrideexportvirtual

Gets the number of trainable parameters in this module.

The Softmax module has no trainable parameters as it's a fixed mathematical operation.

Returns: size_t Always returns 0 as Softmax has no parameters.

Implements Mila::Dnn::Module< TDeviceType, TInput, TOutput >.

◆ save()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

void Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::save ( ModelArchive & zip ) const

inlineoverrideexportvirtual

Serializes the module state to a ZIP archive.

Implementation of the Module interface for serialization. Since Softmax has no learnable parameters, this is a no-op implementation.

Parameters

zip	ZIP archive for serialization

Implements Mila::Dnn::Module< TDeviceType, TInput, TOutput >.

◆ toString()

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

std::string Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::toString ( ) const

inlineoverrideexportvirtual

Generates a string representation of this module's configuration.

Returns: std::string A formatted string with module name, axis, device, and precision info

Implements Mila::Dnn::Module< TDeviceType, TInput, TOutput >.

Here is the call graph for this function:

Member Data Documentation

◆ attributes_

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

OperationAttributes Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::attributes_

exportprivate

Operation attributes and configuration.

◆ config_

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

SoftmaxConfig Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::config_

exportprivate

Configuration for the Softmax module.

◆ operation_

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

std::shared_ptr<UnaryOperation<TDeviceType, TInput, TOutput> > Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::operation_ { nullptr }

exportprivate

The operation that implements the softmax calculation.

◆ output_state_

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

std::vector<std::shared_ptr<Tensor<TOutput, MR> > > Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::output_state_

exportprivate

Collection of output state tensors for caching.

◆ parameters_

template<DeviceType TDeviceType = DeviceType::Cuda, typename TInput = float, typename TOutput = TInput>

std::vector<std::shared_ptr<Tensor<TInput, MR> > > Mila::Dnn::Softmax< TDeviceType, TInput, TOutput >::parameters_

exportprivate

Collection of parameters for this module (empty for Softmax).

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

/home/runner/work/Mila/Mila/Mila/Src/Dnn/Modules/Layers/Softmax.ixx

Public Types

Public Member Functions

Private Member Functions

Private Attributes

Additional Inherited Members

Detailed Description

Member Typedef Documentation

◆ ModuleBase

◆ MR

Constructor & Destructor Documentation

◆ Softmax() [1/2]

◆ Softmax() [2/2]

Member Function Documentation

◆ backward()

◆ createOperation()

◆ forward()

◆ getAxis()

◆ load()

◆ parameterCount()

◆ save()

◆ toString()

Member Data Documentation

◆ attributes_

◆ config_

◆ operation_

◆ output_state_

◆ parameters_