Add PartcileNet ONNX models for Run 3 PFScouting

[alintulu]

This PR adds the necessary files for running 3 types of inference for Run 3 PFScouting. The 3 models include

• AK4 flavour-tagging
• AK8 b-tagging
• AK8 mass regression

You can find plots over the performance of each model here.

We plan to redo the training for each model as soon as the improved Run 3 sample is ready (hopefully within a month). However as this PR is in combination with another to CMSSW adding a custom Run 3 PFScouting NanoAOD (cms-sw/cmssw#40438), we would like to include these first versions of the models already now.

[cmsbuild]

A new Pull Request was created by @alintulu (Adelina Lintuluoto) for branch master.

@smuzaffar, @aandvalenzuela, @iarspider, @clacaputo, @cmsbuild, @mandrenguyen can you please review it and eventually sign? Thanks.
@AlexDeMoor, @emilbols, @JyothsnaKomaragiri, @AnnikaStein, @missirol, @andrzejnovak, @demuller this is something you requested to watch as well.
@perrotta, @dpiparo, @rappoccio you are the release manager for this.
cms-bot commands are listed here

[alintulu]

Hi, what is the status on this PR? We are ready to move on with the CMSSW PR (cms-sw/cmssw#40438), however it requires this PR to be merged first.

[smuzaffar]

@alintulu , these are all new files so should be OK to merge in to data repo. But lets first test with with the cmssw PR. I have started the tests for cms-sw/cmssw#40438

[clacaputo]

+reconstruction

• tested in Custom Run 3 PFScouting NanoAOD cms-sw/cmssw#40438

[rappoccio]

+1

[mmusich]

as cms-sw/cmssw#40438 was merged, shouldn't this one be merged as well?
The IB now complains about

----- Begin Fatal Exception 12-Oct-2023 07:01:19 CEST-----------------------
An exception of category 'ConfigFileReadError' occurred while
   [0] Processing the python configuration file named step2_NANO.py
Exception Message:
 unknown python problem occurred.
RuntimeError: An exception of category 'FileInPathError' occurred.
Exception Message:
edm::FileInPath unable to find file RecoBTag/Combined/data/Run3Scouting/ParticleNetAK4/V00/particle-net.onnx anywhere in the search path.
The search path is defined by: CMSSW_SEARCH_PATH
${CMSSW_SEARCH_PATH} is: /data/cmsbld/jenkins/workspace/ib-run-relvals/CMSSW_13_3_X_2023-10-11-2300/poison:/data/cmsbld/jenkins/workspace/ib-run-relvals/CMSSW_13_3_X_2023-10-11-2300/src:/data/cmsbld/jenkins/workspace/ib-run-relvals/CMSSW_13_3_X_2023-10-11-2300/external/slc7_amd64_gcc11/data:/cvmfs/cms-ib.cern.ch/sw/x86_64/nweek-02806/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/poison:/cvmfs/cms-ib.cern.ch/sw/x86_64/nweek-02806/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/src:/cvmfs/cms-ib.cern.ch/sw/x86_64/nweek-02806/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/external/slc7_amd64_gcc11/data
Current directory is: /data/cmsbld/jenkins/workspace/ib-run-relvals/CMSSW_13_3_X_2023-10-11-2300/pyRelval/2500.5_ScoutingNanodata13X

see log.

Looks like externals signature is missing here @smuzaffar

[smuzaffar]

+externals

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next master IBs after it passes the integration tests.

[smuzaffar]

cms-sw/cmsdist#8750 is merged and 11h IB should have the new data files

[alintulu]

Hi @smuzaffar! I tried running a module that access a file part of this PR, but it returned the following error message

Exception Message:
edm::FileInPath unable to find file RecoBTag/Combined/data/Run3Scouting/ParticleNetAK4/V00/particle-net.onnx anywhere in the search path.
The search path is defined by: CMSSW_SEARCH_PATH
${CMSSW_SEARCH_PATH} is: /tmp/adlintul/CMSSW_13_3_X_2023-10-11-2300/poison:/tmp/adlintul/CMSSW_13_3_X_2023-10-11-2300/src:/tmp/adlintul/CMSSW_13_3_X_2023-10-11-2300/external/slc7_amd64_gcc11/data:/cvmfs/cms-ib.cern.ch/sw/x86_64/week0/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/poison:/cvmfs/cms-ib.cern.ch/sw/x86_64/week0/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/src:/cvmfs/cms-ib.cern.ch/sw/x86_64/week0/slc7_amd64_gcc11/cms/cmssw-patch/CMSSW_13_3_X_2023-10-11-2300/external/slc7_amd64_gcc11/data
Current directory is: /tmp/adlintul/CMSSW_13_3_X_2023-10-11-2300/src/2500.5_ScoutingNanodata13X

I ran it in the IB release CMSSW_13_3_X_2023-10-11-2300. Do you know if I need to specify something for it to access the correct path or if I need to wait for a proper release before it is accessible? Thank you!

[smuzaffar]

@alintulu , you need to wait for today's 11h IB i.e. CMSSW_13_3_X_2023-10-12-1100 (hopefully it will be available in 4 hours)

[alintulu]

@alintulu , you need to wait for today's 11h IB i.e. CMSSW_13_3_X_2023-10-12-1100 (hopefully it will be available in 4 hours)

Aaah sorry! I didn't make the connection with 11h, thank you!