Remove experimental SmoothedHingeLoss training task #455
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
Contributor
ashelkovnykov
commented
Feb 14, 2020
ashelkovnykov
commented
- Experimental results when comparing Smoothed Hinge Loss to Logistic Regression were substantially worse and also increased training time significantly due to slower convergence
- Removed GLMLossFunction and derived classes, since all training tasks in Photon ML are GLM tasks
- Experimental results when comparing Smoothed Hinge Loss to Logistic Regression were substantially worse and also reduced training time significantly due to slower convergence - Removed GLMLossFunction and derived classes, since all training tasks in Photon ML are GLM tasks
|
@cmjiang @yunboouyang @lguo |
yunboouyang
approved these changes
May 19, 2020
| @@ -52,6 +52,8 @@ object SparkSessionConfiguration { | |||
| classOf[GeneralizedLinearModel], | |||
| classOf[GeneralizedLinearOptimizationProblem[_]], | |||
| classOf[GLMOptimizationConfiguration], | |||
| classOf[HessianDiagonalAggregator], | |||
There was a problem hiding this comment.
Why do we need to add this class?
There was a problem hiding this comment.
I noticed that we had the ValueAndGradientAggregator and HessianVectorAggregator being registered with Kryo, but not the HessianDiagonalAggregator or the HessianMatrixAggregator. Since all four function the same way, it stands to reason that they should either all be registered or none of them be registered. Since objects of these four class types will for sure be distributed, therefore they should all be registered.
Apparently, Kryo was able to figure this out on its own, but I think it best to make it explicit.
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.