Skip to content

pass_through_token_embedder

allennlp.modules.token_embedders.pass_through_token_embedder

[SOURCE]

PassThroughTokenEmbedder

@TokenEmbedder.register("pass_through")
class PassThroughTokenEmbedder(TokenEmbedder):
 | def __init__(self, hidden_dim: int) -> None

Assumes that the input is already vectorized in some way, and just returns it.

Registered as a TokenEmbedder with name "pass_through".

Parameters

  • hidden_dim : int

get_output_dim

class PassThroughTokenEmbedder(TokenEmbedder):
 | ...
 | def get_output_dim(self)

forward

class PassThroughTokenEmbedder(TokenEmbedder):
 | ...
 | def forward(self, tokens: torch.Tensor) -> torch.Tensor