Fix Loss Function to Improve Model Convergence for AutoEncoder
#1460
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
/ok to test |
hsin-c
force-pushed
the
hsinc-bugfix-1455
branch
from
af11a73
to
01517bb
Compare
mdemoret-nv
approved these changes
Jan 17, 2024
There was a problem hiding this comment.
Reviewed the changes and have made a few commits to improve the readability and remove redundant calculations. The loss now equally weights all types of features. Once CI is passing and @hsin-c is good with the changes, we can merge.
mdemoret-nv
approved these changes
Jan 22, 2024
mdemoret-nv
changed the title
AutoEncoder
|
/merge |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR addresses an issue in the dfencoder model related to its convergence behavior. Previously, the model exhibited difficulty in converging when trained exclusively with numerical features.
This PR fixes the way different loss types are combined in the model's loss function to ensure that backpropagation works correctly.
Note: This may alter the exact values resulting from calling
fit()on the model. Before, categorical features were weighted much higher than binary or numerical categories (all numerical features shared a combined weight of 1, all binaries features shared a combined weight of 1, and each categorical feature had a weight of 1). Now all features are weighted equally which may impact the trained weights.Closes #1455