[ allennlp.modules.token_embedders.token_embedder ]
class TokenEmbedder(torch.nn.Module, Registrable)
TokenEmbedder is a
Module that takes as input a tensor with integer ids that have
been output from a
TokenIndexer and outputs
a vector per token in the input. The input typically has shape
(batch_size, num_tokens, num_characters), and the output is of shape
output_dim). The simplest
TokenEmbedder is just an embedding layer, but for
character-level input, it could also be some kind of character encoder.
We add a single method to the basic
get_output_dim(). This lets us
more easily compute output dimensions for the
which we might need when defining model parameters such as LSTMs or linear layers, which need
to know their input dimension before the layers are called.
default_implementation = "embedding"
| def get_output_dim(self) -> int
Returns the final output dimension that this
TokenEmbedder uses to represent each
token. This is
not the shape of the returned tensor, but the last element of that shape.