recbole.model.loss

Common Loss in recommender system

class recbole.model.loss.BPRLoss(gamma=1e-10)

Bases: torch.nn.modules.module.Module

BPRLoss, based on Bayesian Personalized Ranking

Parameters

gamma (-) – Small value to avoid division by zero

Shape:

• Pos_score: (N)

• Neg_score: (N), same shape as the Pos_score

• Output: scalar.

Examples:

>>> loss = BPRLoss()
>>> pos_score = torch.randn(3, requires_grad=True)
>>> neg_score = torch.randn(3, requires_grad=True)
>>> output = loss(pos_score, neg_score)
>>> output.backward()

forward(pos_score, neg_score)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class recbole.model.loss.EmbLoss(norm=2)

Bases: torch.nn.modules.module.Module

EmbLoss, regularization on embeddings

forward(*embeddings, require_pow=False)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class recbole.model.loss.EmbMarginLoss(power=2)

Bases: torch.nn.modules.module.Module

EmbMarginLoss, regularization on embeddings

forward(*embeddings)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool

class recbole.model.loss.RegLoss

Bases: torch.nn.modules.module.Module

RegLoss, L2 regularization on model parameters

forward(parameters)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

training: bool