Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Make reprojection layer optional #1676

Merged
merged 1 commit into from
Jun 8, 2020
+15 −12
Merged

Make reprojection layer optional #1676

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
27 changes: 15 additions & 12 deletions flair/models/sequence_tagger_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ def __init__(
dropout: float = 0.0,
word_dropout: float = 0.05,
locked_dropout: float = 0.5,
reproject_to: int = None,
reproject_embeddings: Union[bool,int] = True,
train_initial_hidden_state: bool = False,
rnn_type: str = "LSTM",
pickle_module: str = "pickle",
Expand All @@ -93,7 +93,8 @@ def __init__(
:param rnn_layers: number of RNN layers
:param dropout: dropout probability
:param word_dropout: word dropout probability
:param reproject_to: set this to control the dimensionality of the reprojection layer
:param reproject_embeddings: if True, adds trainable linear map on top of embedding layer. If False, no map.
If you set this to an integer, you can control the dimensionality of the reprojection layer
:param locked_dropout: locked dropout probability
:param train_initial_hidden_state: if True, trains initial hidden state of RNN
:param beta: Parameter for F-beta score for evaluation and training annealing
Expand Down Expand Up @@ -157,14 +158,14 @@ def __init__(
self.locked_dropout = flair.nn.LockedDropout(locked_dropout)

embedding_dim: int = self.embeddings.embedding_length
rnn_input_dim: int = embedding_dim

# if no dimensionality for reprojection layer is set, reproject to equal dimension
self.reproject_to = reproject_to
if self.reproject_to is None: self.reproject_to = embedding_dim
rnn_input_dim: int = self.reproject_to
# optional reprojection layer on top of word embeddings
self.reproject_embeddings = reproject_embeddings
if self.reproject_embeddings:
if type(self.reproject_embeddings) == int:
rnn_input_dim = self.reproject_embeddings

self.relearn_embeddings: bool = True
if self.relearn_embeddings:
self.embedding2nn = torch.nn.Linear(embedding_dim, rnn_input_dim)

self.train_initial_hidden_state = train_initial_hidden_state
Expand Down Expand Up @@ -244,7 +245,7 @@ def _get_state_dict(self):
"rnn_type": self.rnn_type,
"beta": self.beta,
"weight_dict": self.weight_dict,
"reproject_to": self.reproject_to,
"reproject_embeddings": self.reproject_embeddings,
}
return model_state

Expand All @@ -268,7 +269,9 @@ def _init_model_with_state_dict(state):
)
beta = 1.0 if "beta" not in state.keys() else state["beta"]
weights = None if "weight_dict" not in state.keys() else state["weight_dict"]
reproject_to = None if "reproject_to" not in state.keys() else state["reproject_to"]
reproject_embeddings = True if "reproject_embeddings" not in state.keys() else state["reproject_embeddings"]
if "reproject_to" in state.keys():
reproject_embeddings = state["reproject_to"]

model = SequenceTagger(
hidden_size=state["hidden_size"],
Expand All @@ -285,7 +288,7 @@ def _init_model_with_state_dict(state):
rnn_type=rnn_type,
beta=beta,
loss_weights=weights,
reproject_to=reproject_to,
reproject_embeddings=reproject_embeddings,
)
model.load_state_dict(state["state_dict"])
return model
Expand Down Expand Up @@ -642,7 +645,7 @@ def forward(self, sentences: List[Sentence]):
if self.use_locked_dropout > 0.0:
sentence_tensor = self.locked_dropout(sentence_tensor)

if self.relearn_embeddings:
if self.reproject_embeddings:
sentence_tensor = self.embedding2nn(sentence_tensor)

if self.use_rnn:
Expand Down