[FIX] Prevent inverse normalization of quantile validation loss#1432
Merged
[FIX] Prevent inverse normalization of quantile validation loss#1432
Conversation
nasaul
approved these changes
Jan 13, 2026
Collaborator
nasaul
left a comment
nasaul
left a comment
There was a problem hiding this comment.
The solution is correct. However I think that we need to add a test to check consistency across the different losses so this won't happen again.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
When using DistributionLoss for training and sCRPS/MQLoss/HuberMQLoss for validation,
_inv_normalizationwas applied to quantiles that were already in the original scale (via scale_decouple), causing validation loss to be computed on incorrectly scaled values.This fix tracks when the output comes from a scaled distribution and skips
_inv_normalizationin those cases.