MultiTrainLoss Class Reference

Inheritance diagram for MultiTrainLoss:

Collaboration diagram for MultiTrainLoss:

Public Member Functions
	__init__ (self, alpha_mass=0, ignore_index=-1, reduction="mean")
	forward (self, x_input, x_target, edge_input, edge_target, u_input, u_target)

Public Attributes
	alpha_mass = alpha_mass
	Parameter controlling the importance of mass term in loss.
	LCA_CE
	LCA cross-entropy.
	mass_CE
	Mass cross-entropy.

Detailed Description

Sum of cross-entropies for training against LCAS and mass hypotheses.

Args:
    alpha_mass (float): Weight of mass cross-entropy term in the loss.
    ignore_index (int): Index to ignore in the computation (e.g. padding).
    reduction (str): Type of reduction to be applied on the batch (``sum`` or ``mean``).

Definition at line 13 of file multiTrain.py.

Constructor & Destructor Documentation

__init__()

__init__	(		self,
			alpha_mass = 0,
			ignore_index = -1,
			reduction = "mean" )

Initialization

Definition at line 23 of file multiTrain.py.

    ):
        """
        Initialization
        """
        super().__init__()
 
        
        self.alpha_mass = alpha_mass
 
        
        self.LCA_CE = nn.CrossEntropyLoss(
            ignore_index=ignore_index, reduction=reduction
        )
        
        self.mass_CE = nn.CrossEntropyLoss(
            ignore_index=ignore_index, reduction=reduction
        )
 
        assert alpha_mass >= 0, "Alpha should be positive"
 

Member Function Documentation

forward()

forward	(		self,
			x_input,
			x_target,
			edge_input,
			edge_target,
			u_input,
			u_target )

Called internally by PyTorch to propagate the input.

Definition at line 48 of file multiTrain.py.

    def forward(self, x_input, x_target, edge_input, edge_target, u_input, u_target):
        """
        Called internally by PyTorch to propagate the input.
        """
 
        LCA_loss = self.LCA_CE(
            edge_input,
            edge_target,
        )
 
        mass_loss = (
            self.mass_CE(
                x_input,
                x_target,
            )
            if self.alpha_mass > 0
            else 0
        )
 
        return LCA_loss + self.alpha_mass * mass_loss

Member Data Documentation

alpha_mass

alpha_mass = alpha_mass

Parameter controlling the importance of mass term in loss.

Definition at line 35 of file multiTrain.py.

LCA_CE

LCA_CE

Initial value:

=  nn.CrossEntropyLoss(
            ignore_index=ignore_index, reduction=reduction
        )

LCA cross-entropy.

Definition at line 38 of file multiTrain.py.

mass_CE

mass_CE

Initial value:

=  nn.CrossEntropyLoss(
            ignore_index=ignore_index, reduction=reduction
        )

Mass cross-entropy.

Definition at line 42 of file multiTrain.py.

---

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

• analysis/scripts/grafei/model/multiTrain.py