Representation

class Representation(max_id: int, shape: int | Sequence[int] = 64, normalizer: str | Callable[[Tensor], Tensor] | type[Callable[[Tensor], Tensor]] | None = None, normalizer_kwargs: Mapping[str, Any] | None = None, regularizer: str | Regularizer | type[Regularizer] | None = None, regularizer_kwargs: Mapping[str, Any] | None = None, dropout: float | None = None, unique: bool | None = None)

Bases: Module, ExtraReprMixin, ABC

A base class for obtaining representations for entities/relations.

A representation module maps integer IDs to representations, which are tensors of floats.

max_id defines the upper bound of indices we are allowed to request (exclusively). For simple embeddings this is equivalent to num_embeddings, but more a more appropriate word for general non-embedding representations, where the representations could come from somewhere else, e.g. a GNN encoder.

shape describes the shape of a single representation. In case of a vector embedding, this is just a single dimension. For others, e.g. pykeen.models.RESCAL, we have 2-d representations, and in general it can be any fixed shape.

We can look at all representations as a tensor of shape (max_id, *shape), and this is exactly the result of passing indices=None to the forward method.

We can also pass multi-dimensional indices to the forward method, in which case the indices’ shape becomes the prefix of the result shape: (*indices.shape, *self.shape).

Initialize the representation module.

Parameters:

• max_id (int) – The maximum ID (exclusively). Valid Ids reach from 0, …, max_id-1

• shape (tuple[int, ...]) – The shape of an individual representation.

• normalizer (Callable[[Tensor], Tensor] | None) – A normalization function, which is applied to the selected representations in every forward pass.

• normalizer_kwargs (OptionalKwargs) – Additional keyword arguments passed to the normalizer

• regularizer (Regularizer | None) – An output regularizer, which is applied to the selected representations in forward pass

• regularizer_kwargs (OptionalKwargs) – Additional keyword arguments passed to the regularizer

• dropout (Dropout | None) – The optional dropout probability

• unique (bool | None) – whether to optimize for calculating representations for same indices only once. This is only useful if the calculation of representations is either significantly more expensive than an index-based lookup and duplicate indices are expected, e.g., when using negative sampling and large batch sizes

Attributes Summary

device

Return the device.

Methods Summary

forward([indices])	Get representations for indices.
iter_extra_repr()	Iterate over components for extra_repr().
post_parameter_update()	Apply constraints which should not be included in gradients.
reset_parameters()	Reset the module's parameters.

Attributes Documentation

device

Return the device.

Methods Documentation

forward(indices: Tensor | None = None) → Tensor

Get representations for indices.

Note

depending on Representation.unique, this implementation will use an optimization for duplicate indices. It is generally only recommended if computing individual representation is expensive, e.g., since it involves message passing, or a large encoder networks, but discouraged for cheap lookups, e.g., a plain embedding lookup.

Parameters:

indices (Tensor | None) – shape: s The indices, or None. If None, this is interpreted as torch.arange(self.max_id) (although implemented more efficiently).

Returns:

shape: (*s, *self.shape) The representations.

Return type:

Tensor

iter_extra_repr() → Iterable[str]

Iterate over components for extra_repr().

Return type:

Iterable[str]

post_parameter_update()

Apply constraints which should not be included in gradients.

reset_parameters() → None

Reset the module’s parameters.

Return type:

None