API References

Similarity Model

SimilarityModel

Main class which contains encoder models with the head layer.

Encoders

Encoder	Base class for encoder abstraction
SwitchEncoder	Allows use alternative embeddings based on input data.
FasttextEncoder	Creates a fasttext encoder, which generates vector for a list of tokens based in given fasttext model

Head Layers

EmptyHead	Returns input embeddings without any modification
GatedHead	Disables or amplifies some components of input embedding.
SequentialHead	A torch.nn.Sequential-like head layer that you can freely add any layers.
SkipConnectionHead	Unites the idea of gated head and residual connections.
SoftmaxEmbeddingsHead	Provides a concatenation of the independent softmax embeddings groups as a head layer
StackedProjectionHead	Stacks any number of projection layers with specified output sizes.
WideningHead	Implements narrow-wide-narrow architecture.