SimplE
- class SimplE(*, embedding_dim: int = 200, clamp_score: tuple[float | None, float] | tuple[float, float | None] | float | None = None, entity_initializer: str | Callable[[Tensor], Tensor] | None = None, relation_initializer: str | Callable[[Tensor], Tensor] | None = None, regularizer: str | Regularizer | None = None, regularizer_kwargs: Mapping[str, Any] | None = None, **kwargs)[source]
Bases:
ERModel
[tuple
[Tensor
,Tensor
],tuple
[Tensor
,Tensor
],tuple
[Tensor
,Tensor
]]An implementation of SimplE [kazemi2018].
SimplE learns two \(d\)-dimensional vectors for each entity and each relation, stored in
Embedding
, and applies theSimplEInteraction
on top.Note
In the code in their repository, the score is clamped to \([-20, 20]\). That is not mentioned in the paper, so it is made optional.
See also
Official implementation: https://github.com/Mehran-k/SimplE
Improved implementation in pytorch: https://github.com/baharefatemi/SimplE
Initialize the model.
- Parameters:
embedding_dim (int) – the embedding dimension
clamp_score (Clamp | float | None) – whether to clamp scores, cf.
ClampedInteraction
entity_initializer (Hint[Initializer]) – the entity representation initializer
relation_initializer (Hint[Initializer]) – the relation representation initializer
regularizer (Hint[Regularizer]) – the regularizer, defaults to
SimplE.regularizer_default
regularizer_kwargs (OptionalKwargs) – additional keyword-based parameters passed to the regularizer, defaults to
SimplE.regularizer_default_kwargs
kwargs – additional keyword-based parameters passed to
ERModel.__init__()
Attributes Summary
The default strategy for optimizing the model's hyper-parameters
The default parameters for the default loss function class
The power sum settings used by [trouillon2016] for SimplE
Attributes Documentation
- hpo_default: ClassVar[Mapping[str, Any]] = {'embedding_dim': {'high': 256, 'low': 16, 'q': 16, 'type': <class 'int'>}}
The default strategy for optimizing the model’s hyper-parameters
- loss_default_kwargs: ClassVar[Mapping[str, Any]] = {}
The default parameters for the default loss function class
- regularizer_default_kwargs: ClassVar[Mapping[str, Any]] = {'normalize': True, 'p': 2.0, 'weight': 20}
The power sum settings used by [trouillon2016] for SimplE