Repository navigation
[TimesFM 2.5] Fix quantile loss channel indexing - #50704
Open
Roshanraj2580 wants to merge 1 commit into
Open
Roshanraj2580 wants to merge 1 commit into
Roshanraj2580 wants to merge 1 commit into
Conversation
Contributor
|
Thank you for your contribution 🤗! CI Security Gate — waiting for maintainer approvalCI is approved automatically only for authors who already have a contribution merged in this repository. A maintainer will review this PR and approve CI to run on it. |
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: timesfm2_5 |
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Fixes #50612
In
TimesFm2_5ModelForPrediction.forward, quantile loss indexing filtered channels via[i for i in range(normalized_preds.shape[-1]) if i != decode_index].Because channel 0 is the point-forecast head (
output_projection_point) rather than a quantile head, this index expression erroneously included channel 0 inquantile_tensorand evaluated it againstself.config.quantiles[0]. Consequently, channels 1..4 were shifted by one position, and channeldecode_index(the median quantile) received no quantile loss supervision.This PR:
list(range(1, normalized_preds.shape[-1]))in bothmodeling_timesfm2_5.pyandmodular_timesfm2_5.py, aligning channels 1..9 1:1 withconfig.quantiles.test_loss_quantile_channel_alignmentintests/models/timesfm2_5/test_modeling_timesfm2_5.py.Who can review?
@Cyrilvallez @Raushan-T