Skip to content

resolving padding issue #18

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

Open
rainyrainyguo wants to merge 2 commits into
base: master
Choose a base branch
from
Open

resolving padding issue #18

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2237347
resolve pack_padded_sequence issue of models.py
rainyrainyguo Aug 24, 2018
f12499d
resolving pack_padded_sequence issue of models.py
rainyrainyguo Aug 24, 2018
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
6 changes: 3 additions & 3 deletions lang/models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -186,7 +186,7 @@ def forward(self, indices, lengths, noise, encode_only=False):

def encode(self, indices, lengths, noise):
embeddings = self.embedding(indices)
packed_embeddings = pack_padded_sequence(input=embeddings,
packed_embeddings = pack_padded_sequence(embeddings,
lengths=lengths,
batch_first=True)

Expand All @@ -195,7 +195,7 @@ def encode(self, indices, lengths, noise):
hidden = hidden / torch.norm(hidden, p=2, dim=1, keepdim=True)

if noise and self.noise_r > 0:
gauss_noise = torch.normal(means=torch.zeros(hidden.size()),
gauss_noise = torch.normal(mean=torch.zeros(hidden.size()),
std=self.noise_r)
hidden = hidden + Variable(gauss_noise.cuda())

Expand All @@ -213,7 +213,7 @@ def decode(self, hidden, batch_size, maxlen, indices=None, lengths=None):

embeddings = self.embedding_decoder(indices)
augmented_embeddings = torch.cat([embeddings, all_hidden], 2)
packed_embeddings = pack_padded_sequence(input=augmented_embeddings,
packed_embeddings = pack_padded_sequence(augmented_embeddings,
lengths=lengths,
batch_first=True)

Expand Down
12 changes: 6 additions & 6 deletions yelp/models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -155,7 +155,7 @@ def forward(self, whichdecoder, indices, lengths, noise=False, encode_only=False

def encode(self, indices, lengths, noise):
embeddings = self.embedding(indices)
packed_embeddings = pack_padded_sequence(input=embeddings,
packed_embeddings = pack_padded_sequence(embeddings,
lengths=lengths,
batch_first=True)

Expand All @@ -175,7 +175,7 @@ def encode(self, indices, lengths, noise):
# hidden = torch.div(hidden, norms.unsqueeze(1).expand_as(hidden))

if noise and self.noise_r > 0:
gauss_noise = torch.normal(means=torch.zeros(hidden.size()),
gauss_noise = torch.normal(mean=torch.zeros(hidden.size()),
std=self.noise_r)
hidden = hidden + to_gpu(self.gpu, Variable(gauss_noise))

Expand All @@ -197,7 +197,7 @@ def decode(self, whichdecoder, hidden, batch_size, maxlen, indices=None, lengths
embeddings = self.embedding_decoder2(indices)

augmented_embeddings = torch.cat([embeddings, all_hidden], 2)
packed_embeddings = pack_padded_sequence(input=augmented_embeddings,
packed_embeddings = pack_padded_sequence(augmented_embeddings,
lengths=lengths,
batch_first=True)

Expand Down Expand Up @@ -443,7 +443,7 @@ def forward(self, indices, lengths, noise, encode_only=False):

def encode(self, indices, lengths, noise):
embeddings = self.embedding(indices)
packed_embeddings = pack_padded_sequence(input=embeddings,
packed_embeddings = pack_padded_sequence(embeddings,
lengths=lengths,
batch_first=True)

Expand All @@ -463,7 +463,7 @@ def encode(self, indices, lengths, noise):
# hidden = torch.div(hidden, norms.unsqueeze(1).expand_as(hidden))

if noise and self.noise_r > 0:
gauss_noise = torch.normal(means=torch.zeros(hidden.size()),
gauss_noise = torch.normal(mean=torch.zeros(hidden.size()),
std=self.noise_r)
hidden = hidden + to_gpu(self.gpu, Variable(gauss_noise))

Expand All @@ -481,7 +481,7 @@ def decode(self, hidden, batch_size, maxlen, indices=None, lengths=None):

embeddings = self.embedding_decoder(indices)
augmented_embeddings = torch.cat([embeddings, all_hidden], 2)
packed_embeddings = pack_padded_sequence(input=augmented_embeddings,
packed_embeddings = pack_padded_sequence(augmented_embeddings,
lengths=lengths,
batch_first=True)

Expand Down