Skip to content

Fix kl div #1

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
jelkhouryy wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix kl div #1

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
5 changes: 4 additions & 1 deletion src/probabilistic_dag_model/probabilistic_dag_autoencoder.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,7 +93,10 @@ def ELBO_loss(self, X_pred, X):

if self.regr > 0:
kl_loss = torch.nn.KLDivLoss(reduction='mean')
regularizer = kl_loss(self.probabilistic_dag.edge_log_params, self.prior_p * torch.ones_like(self.probabilistic_dag.edge_log_params))
logsigmoid = torch.nn.LogSigmoid()
ones = torch.ones_like(self.probabilistic_dag.edge_log_params)
regularizer = kl_loss(logsigmoid(self.probabilistic_dag.edge_log_params, self.prior_p * ones)) #KLDivLoss expects logits as first argument
regularizer += kl_loss(torch.log(ones - torch.sigmoid(self.probabilistic_dag.edge_log_params)), (1 - self.prior_p) * ones)
ELBO_loss = ELBO_loss + self.regr * regularizer

return ELBO_loss
Expand Down