Add HLT_DoubleMediumPFPuppiParTTauh30_eta2p1 path by agrubercms · Pull Request #51705 · cms-sw/cmssw

agrubercms · 2026-08-14T10:22:45Z

PR description:

This PR introduces a new Phase-2 tau HLT path (HLT_DoubleMediumPFPuppiParTTauh30_eta2p1) based on a ParticleTransformer model. The model (+documentation) can be found in this PR to the cms-data repository. It was developed to also be able to be used for b-tagging, but was not yet tested for this purpose - to illustrate this, the PR includes b- and c- discriminators, which are currently commented out. Most implementation choices were made to maintain compatibility with the existing tools used by the BTV group (most notably the hltupler framework to produce the training ntuples).
Development progress and measurements were presented regularly in Tau internal meetings, e.g. here and here.

The new model significantly outperforms the existing DeepTau model across all evaluated metrics - these are tau tagging efficiency, trigger rate and computational runtime.

PR validation:

The standard tests have been run:

scram b runtests use-ibeos
runTheMatrix.py -l limited -i all --ibeos

Model performance has been evaluated using the same sample composition as was used in the training, but with a different subgroup of the samples to ensure evaluation is done on events not used in the model training.

To test this PR, the PR to cms-data (mentioned above) containing the onnx model would need to be included.

cmsbuild · 2026-08-14T10:23:09Z

cms-bot internal usage

cmsbuild · 2026-08-14T10:25:05Z

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50596

Code check has found code style and quality issues which could be resolved by applying following patch(s)

code-format:
https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50596/code-format.patch
e.g. curl -k https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50596/code-format.patch | patch -p1
You can also run scram build code-format to apply code format directly

cmsbuild · 2026-08-14T10:33:17Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50597

cmsbuild · 2026-08-14T10:33:44Z

Pull request #51705 was updated.

cmsbuild · 2026-08-14T11:56:51Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50599

cmsbuild · 2026-08-14T11:57:14Z

Pull request #51705 was updated.

mmusich

can the new path be added also to the timing menu?
Also can the UParT sequence be added to the NGT scouting path?

mmusich · 2026-08-14T13:53:36Z

test parameters:

pull_request = Add hltParticleTransformer model for new Phase-2 path HLT_DoubleMediumPFPuppiParTTauh30_eta2p1 cms-data/RecoBTag-Combined#73
workflows = ph2_hlt
enable = hlt_p2_integration, hlt_p2_timing

cmsbuild · 2026-08-17T19:16:59Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50643

There are other open Pull requests which might conflict with changes you have proposed:
- File HLTrigger/Configuration/python/HLT_75e33_cff.py modified in PR(s): Rename HLT_LooseDeepTauPFTauHPS150_L2NN_eta2p1 to L1NN #51718
- File HLTrigger/Configuration/python/HLT_75e33_timing_cff.py modified in PR(s): Rename HLT_LooseDeepTauPFTauHPS150_L2NN_eta2p1 to L1NN #51718

cmsbuild · 2026-08-17T19:17:25Z

Pull request #51705 was updated.

mmusich · 2026-08-17T19:31:19Z

@cmsbuild, please test

cmsbuild · 2026-08-18T01:55:35Z

-1

Failed Tests: RelVals
Size: This PR adds an extra 64KB to repository
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55413/summary.html
COMMIT: e16e97b
CMSSW: CMSSW_20_1_X_2026-08-17-1100/el9_amd64_gcc14
Additional Tests: HLT_P2_INTEGRATION,HLT_P2_TIMING
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/51705/55413/install.sh to create a dev area with all the needed externals and cmssw changes.

The following merge commits were also included on top of IB + this PR after doing git cms-merge-topic:

You can see more details here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55413/git-recent-commits.json
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55413/git-merge-result

HLT P2 Timing: chart
DAS Queries: The DAS query tests failed, see the summary page for details.

Failed RelVals

37634.758

----- Begin Fatal Exception 18-Aug-2026 03:31:59 CEST-----------------------
An exception of category 'ScheduleExecutionFailure' occurred while
   [0] Calling beginJob
Exception Message:
Unrunnable schedule
Dependent module later on Path
  module 'hltParticleFlowRecHitHBHE' depends on 'hltHbhereco' which is later on path HLT_DoubleMediumPFPuppiParTTauh30_eta2p1
----- End Fatal Exception -------------------------------------------------

cmsbuild · 2026-08-19T08:10:57Z

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-51705/50662

cmsbuild · 2026-08-19T08:11:24Z

Pull request #51705 was updated.

mmusich · 2026-08-19T08:12:13Z

please test

cmsbuild · 2026-08-19T23:31:01Z

-1

Size: This PR adds an extra 24KB to repository
Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55462/summary.html
COMMIT: fa8a1b3
CMSSW: CMSSW_20_1_X_2026-08-18-2300/el9_amd64_gcc14
Additional Tests: HLT_P2_INTEGRATION,HLT_P2_TIMING
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmssw/51705/55462/install.sh to create a dev area with all the needed externals and cmssw changes.

The following merge commits were also included on top of IB + this PR after doing git cms-merge-topic:

@mmusich ScoutingTrackMonitor: fix for the axes' labels according to the pT cut #51721

You can see more details here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55462/git-recent-commits.json
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2d4510/55462/git-merge-result

HLT P2 Timing: chart
DAS Queries: The DAS query tests failed, see the summary page for details.

Comparison Summary

Summary:

You potentially added 9 lines to the logs
ROOTFileChecks: Some differences in event products or their sizes found
Reco comparison results: 17 differences found in the comparisons
DQMHistoTests: Total files compared: 59
DQMHistoTests: Total histograms compared: 4089582
DQMHistoTests: Total failures: 2605
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 4086959
DQMHistoTests: Total skipped: 18
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 58 files compared)
Checked 249 log files, 203 edm output root files, 59 DQM output files
TriggerResults: no differences found

NVIDIA_H100 Comparison Summary

Summary:

No significant changes to the logs found
ROOTFileChecks: Some differences in event products or their sizes found
Reco comparison results: 95 differences found in the comparisons
DQMHistoTests: Total files compared: 7
DQMHistoTests: Total histograms compared: 173739
DQMHistoTests: Total failures: 8798
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 164941
DQMHistoTests: Total skipped: 0
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 6 files compared)
Checked 25 log files, 20 edm output root files, 7 DQM output files
TriggerResults: no differences found

NVIDIA_L4 Comparison Summary

Summary:

You potentially added 1 lines to the logs
ROOTFileChecks: Some differences in event products or their sizes found
Reco comparison results: 102 differences found in the comparisons
DQMHistoTests: Total files compared: 7
DQMHistoTests: Total histograms compared: 173739
DQMHistoTests: Total failures: 8897
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 164842
DQMHistoTests: Total skipped: 0
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 6 files compared)
Checked 25 log files, 20 edm output root files, 7 DQM output files
TriggerResults: no differences found

NVIDIA_L40S Comparison Summary

Summary:

No significant changes to the logs found
ROOTFileChecks: Some differences in event products or their sizes found
Reco comparison results: 90 differences found in the comparisons
DQMHistoTests: Total files compared: 7
DQMHistoTests: Total histograms compared: 173739
DQMHistoTests: Total failures: 8527
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 165212
DQMHistoTests: Total skipped: 0
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 6 files compared)
Checked 25 log files, 20 edm output root files, 7 DQM output files
TriggerResults: no differences found

NVIDIA_T4 Comparison Summary

Summary:

No significant changes to the logs found
ROOTFileChecks: Some differences in event products or their sizes found
Reco comparison results: 108 differences found in the comparisons
DQMHistoTests: Total files compared: 7
DQMHistoTests: Total histograms compared: 173739
DQMHistoTests: Total failures: 9554
DQMHistoTests: Total nulls: 0
DQMHistoTests: Total successes: 164185
DQMHistoTests: Total skipped: 0
DQMHistoTests: Total Missing objects: 0
DQMHistoSizes: Histogram memory added: 0.0 KiB( 6 files compared)
Checked 25 log files, 20 edm output root files, 7 DQM output files
TriggerResults: no differences found

Max Memory Comparisons exceeding threshold NVIDIA_H100

@cms-sw/core-l2 , I found 1 workflow step(s) with memory usage exceeding the error threshold:

Expand to see workflows ...

Error: Workflow 37634.7503_TTbar_14TeV+Run4D127_HLTHeterogeneousValid step2 max memory diff 227.1 exceeds +/- 30.0 MiB

mmusich · 2026-08-20T18:19:56Z

-1

Failure is unrelated.

@agrubercms just a couple of extra requests. Can you add your new path:

here:

cmssw/Validation/HLTrigger/python/HLTGenValidation_cff.py

Lines 562 to 565 in 03fbc40

    
           hltPathsToCheck = cms.vstring( 
        
               'HLT_DoubleMediumChargedIsoPFTauHPS40_eta2p1', 
        
               'HLT_DoubleMediumDeepTauPFTauHPS35_eta2p1' 
        
           ),

and

here:

cmssw/HLTrigger/NGTScouting/python/hltTriggerObjects_cff.py

Line 24 in 15a5d43

pathNames=cms.vstring(

?

cmsbuild added this to the CMSSW_20_1_X milestone Aug 14, 2026

cmsbuild added reconstruction-pending hlt-pending pending-signatures tests-pending orp-pending code-checks-pending changes-dataformats labels Aug 14, 2026

cmsbuild added code-checks-rejected and removed code-checks-pending labels Aug 14, 2026

agrubercms force-pushed the GloParT_clean_PR branch from 11b9638 to 2858f55 Compare August 14, 2026 10:29

cmsbuild added code-checks-pending and removed code-checks-rejected labels Aug 14, 2026

cmsbuild added code-checks-approved and removed code-checks-pending labels Aug 14, 2026

agrubercms force-pushed the GloParT_clean_PR branch from 2858f55 to 73a79fc Compare August 14, 2026 11:55

cmsbuild added code-checks-pending and removed code-checks-approved labels Aug 14, 2026

cmsbuild added code-checks-approved and removed code-checks-pending labels Aug 14, 2026

mmusich reviewed Aug 14, 2026

View reviewed changes

cmsbuild added the requires-external label Aug 14, 2026

cmsbuild mentioned this pull request Aug 17, 2026

Rename HLT_LooseDeepTauPFTauHPS150_L2NN_eta2p1 to L1NN #51718

Merged

cmsbuild added code-checks-pending and removed code-checks-approved labels Aug 17, 2026

cmsbuild added code-checks-approved and removed code-checks-pending labels Aug 17, 2026

cmsbuild added tests-started and removed tests-pending labels Aug 17, 2026

cmsbuild added tests-rejected and removed tests-started labels Aug 18, 2026

Phase-2 ParticleTransformer tau path

fa8a1b3

agrubercms force-pushed the GloParT_clean_PR branch from e16e97b to fa8a1b3 Compare August 19, 2026 08:09

cmsbuild added tests-pending code-checks-pending and removed tests-rejected code-checks-approved labels Aug 19, 2026

cmsbuild added code-checks-approved and removed code-checks-pending labels Aug 19, 2026

cmsbuild added tests-started and removed tests-pending labels Aug 19, 2026

cmsbuild mentioned this pull request Aug 19, 2026

Added ROOT dictionaries in DataFormats #51739

Merged

cmsbuild added tests-rejected and removed tests-started labels Aug 19, 2026

Uh oh!

Conversation

agrubercms commented Aug 14, 2026

PR description:

PR validation:

Uh oh!

cmsbuild commented Aug 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cmsbuild commented Aug 14, 2026

Uh oh!

cmsbuild commented Aug 14, 2026

Uh oh!

cmsbuild commented Aug 14, 2026

Uh oh!

cmsbuild commented Aug 14, 2026

Uh oh!

cmsbuild commented Aug 14, 2026

Uh oh!

mmusich left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

mmusich commented Aug 14, 2026

Uh oh!

cmsbuild commented Aug 17, 2026

Uh oh!

cmsbuild commented Aug 17, 2026

Uh oh!

mmusich commented Aug 17, 2026

Uh oh!

cmsbuild commented Aug 18, 2026

Failed RelVals

Uh oh!

cmsbuild commented Aug 19, 2026

Uh oh!

cmsbuild commented Aug 19, 2026

Uh oh!

mmusich commented Aug 19, 2026

Uh oh!

cmsbuild commented Aug 19, 2026

Comparison Summary

NVIDIA_H100 Comparison Summary

NVIDIA_L4 Comparison Summary

NVIDIA_L40S Comparison Summary

NVIDIA_T4 Comparison Summary

Max Memory Comparisons exceeding threshold NVIDIA_H100

Uh oh!

mmusich commented Aug 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

cmsbuild commented Aug 14, 2026 •

edited

Loading