DNN models for the ParticleNet tagger

[hqucms]

This PR adds the MXNet models of the ParticleNet tagger.

[cmsbuild]

A new Pull Request was created by @hqucms (Huilin Qu) for branch master.

@perrotta, @smuzaffar, @mrodozov, @cmsbuild, @slava77, @tulamor can you please review it and eventually sign? Thanks.
@emilbols, @smoortga, @riga, @HeinerTholen, @JyothsnaKomaragiri, @mverzett, @ferencek, @andrzejnovak, @pvmulder, @acaudron this is something you requested to watch as well.
cms-bot commands are listed here

[cmsbuild]

The tests are being triggered in jenkins.
Tested with other pull request(s) cms-sw/cmssw#28902
https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/4614/console Started: 2020/02/12 08:13

[cmsbuild]

+1
Tested at: 909c7c7
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-9fe682/4614/summary.html
CMSSW: CMSSW_11_1_X_2020-02-11-2300
SCRAM_ARCH: slc7_amd64_gcc820

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-9fe682/4614/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 1100 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 2694005
• DQMHistoTests: Total failures: 1
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 2693658
• DQMHistoTests: Total skipped: 346
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 33 files compared)
• Checked 147 log files, 16 edm output root files, 34 DQM output files

[slava77]

In DeepBoostedJet we have version first, and then full and decorrelated subdirs.
Please adjust the structure, especially if "MD" and non-MD are expected to appear simultaneously

It may be a good time to start populating README.md in the main directory and perhaps also in the subdirs to keep track of the files details.

[hqucms]

@slava77

In DeepBoostedJet we have version first, and then full and decorrelated subdirs.
Please adjust the structure, especially if "MD" and non-MD are expected to appear simultaneously

I would prefer to keep the current structure as in the ParticleNet case, the two versions are not as closely coupled as in the case for DeepAK8 -- The nominal (non-MD) ParticleNet targets Top/W/Z/H tagging, similar to DeepAK8, while ParticleNet-MD is for mass-independent 2-prong tagging (X->bb, X->cc, X->qq) -- This means that we actually might want to update one but not the other.

It may be a good time to start populating README.md in the main directory and perhaps also in the subdirs to keep track of the files details.

Good idea! I will add a readme for them.

[slava77]

I would prefer to keep the current structure as in the ParticleNet case, the two versions are not as closely coupled as in the case for DeepAK8

is ParticleNet name uniquely identifying the potential uses?
Should we at least add AK8 to the name of the files or dirs?
Perhaps ParticleNet/MD and ParticleNet/General (or alike) is still better; so that the training files are compartmentalized a bit better.

[hqucms]

@slava77
How about ParticleNetAK8/General/V00 and ParticleNetAK8/MD-2prong/V00?

[slava77]

How about ParticleNetAK8/General/V00 and ParticleNetAK8/MD-2prong/V00?

sounds good to me.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-9fe682/4984/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 1102 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 2680373
• DQMHistoTests: Total failures: 40
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 2680014
• DQMHistoTests: Total skipped: 319
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 33 files compared)
• Checked 147 log files, 16 edm output root files, 34 DQM output files

[cmsbuild]

The tests are being triggered in jenkins.
Tested with other pull request(s) cms-sw/cmssw#28902
https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/5001/console Started: 2020/03/05 04:32

[cmsbuild]

+1
Tested at: aca78b9
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-9fe682/5001/summary.html
CMSSW: CMSSW_11_1_X_2020-03-04-2300
SCRAM_ARCH: slc7_amd64_gcc820

[cmsbuild]

Comparison job queued.

[cmsbuild]

Comparison is ready
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-9fe682/5001/summary.html

Comparison Summary:

• No significant changes to the logs found
• Reco comparison results: 1102 differences found in the comparisons
• DQMHistoTests: Total files compared: 34
• DQMHistoTests: Total histograms compared: 2680577
• DQMHistoTests: Total failures: 40
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 2680218
• DQMHistoTests: Total skipped: 319
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 33 files compared)
• Checked 147 log files, 16 edm output root files, 34 DQM output files

[slava77]

+1

this appears to be ready to go based on what was seen so far in cms-sw/cmssw#28902

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @davidlange6, @silviodonato, @fabiocos (and backports should be raised in the release meeting by the corresponding L2)

[silviodonato]

+1