CompGCN

class CompGCN(*, triples_factory: CoreTriplesFactory, embedding_dim: int = 64, encoder_kwargs: Mapping[str, Any] | None = None, interaction: str | Interaction[Tensor, RelationRepresentation, Tensor] | None = None, interaction_kwargs: Mapping[str, Any] | None = None, **kwargs)

Bases: ERModel[Tensor, RelationRepresentation, Tensor]

An implementation of CompGCN from [vashishth2020].

This model uses graph convolutions, and composition functions.

Initialize the model.

Parameters:

• triples_factory (CoreTriplesFactory) – The triples factory.

• embedding_dim (int) – The embedding dimension to be used if embedding_specification is not given explicitly in encoder_kwargs.

• encoder_kwargs (Mapping[str, Any] | None) – Additional keyword arguments for the encoder, cf. pykeen.nn.representation.CombinedCompGCNRepresentations.

• interaction (Interaction) – The interaction function to use as decoder.

• interaction_kwargs (Mapping[str, Any] | None) – Additional keyword based arguments for the interaction function.

• kwargs – Additional keyword based arguments passed to pykeen.models.ERModel.

Attributes Summary

hpo_default

The default strategy for optimizing the model's hyper-parameters

Attributes Documentation

hpo_default: ClassVar[Mapping[str, Any]] = {'embedding_dim': {'high': 512, 'low': 32, 'q': 32, 'type': <class 'int'>}}

The default strategy for optimizing the model’s hyper-parameters