Repository navigation
[PWGJE] Add Particle Transformer b-jet tagger to jetTaggerHF - #18409
Open
ahmadtabikh wants to merge 1 commit into
Open
ahmadtabikh wants to merge 1 commit into
ahmadtabikh wants to merge 1 commit into
Conversation
ahmadtabikh
requested review from
alibuild,
fjonasALICE,
jaimenorman,
mhemmer-cern and
nzardosh
as code owners
October 22, 2026 12:20
|
O2 linter results: ❌ 0 errors, |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds a Particle Transformer (ParT) b-jet tagger to jetTaggerHF, alongside the existing GNN tagger. It writes its D_b score to the same scoreML / tagging tables.
jetTaggerHF.cxx: new `processAlgorithmParT (data and MCD, off by default) and a part ConfigurableGroup (nConstit, nFeat, nVec, padWrap, minTracks). The model is loaded from CCDB or from a local file.JetTaggingUtilities.h:analyzeJetTrackInfo4ParT()builds the 10 per-constituent features and the 4-vector.getDbFromProb()computes D_b directly from the probabilities, with no softmax since the exported model already applies it.MlResponseHfTagging.h:ParTBjetAllocatorreproduces the input preparation of weaver (the framework used to train the ParT model):pf_features (1, nFeat, nConstit),pf_vectors (1, nVec, nConstit)andpf_mask (1, 1, nConstit), with truncation atnConstit, wrap or zero padding (mask 0 on padded slots), and non-finite values set to 0.getModelOutputMultiInput()evaluates models with several input tensors.ParT_O2_pipeline_colored.pdf