LowPtElectrons: final ID model weights file #16
Merged
Conversation
This file contains 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
|
A new Pull Request was created by @bainbrid for branch master. @perrotta, @smuzaffar, @mrodozov, @cmsbuild, @slava77, @jpata can you please review it and eventually sign? Thanks. |
|
We (i.e. @crovelli!) have investigated various hyperparameter and training configurations for our ID BDT model, with the aim of reducing the disk and memory footprint of the GBRForest. Attached is a plot that shows the performance of various models. Below is a table that summarises the disk footprint of the .root file that would be stored in cms-data. The "B-Parking" model in the table below is the "2020Sept15" model that is currently used by the CMSSW_11_3_X, merged from the PR #15. It is planned to be integrated within CMSSW_10_6_X as part of the back port cms-sw/cmssw#32372 for use with the UL reminiAOD compaign. It was trained based on BParking studies and it is our benchmark for all the studies below. It is indicated by the black curve in the attached plot. The other models, labeled DepthXX, are indicated in the plot as coloured dashed curves. They are trained on AOD inputs taken from B->J/psi(->ee)K samples produced with 10_6_X. The general improvement w.r.t. the baseline "B-Parking" model is understood to be due to the improved reconstruction from the UltraLegacy AOD samples w.r.t. the Prompt conditions used by the BParking samples. Our recommendation is to use the Depth10 model, so we plan to update this PR to use the Depth10 model. Please let us know if you have any objections.
|
…n 2; rm Autumn18 file
bainbrid
force-pushed
the
from-CMSSW_11_3_X-2020Nov28
branch
from
87555f6
to
01c2f68
Compare
|
Pull request #16 was updated. |
|
The last commit implements the changes discussed here #16 (comment) |
|
+1 based on tests in cms-sw/cmssw#32391 |
cmsbuild
added
reconstruction-approved
tests-started
and removed
reconstruction-pending
tests-pending
labels
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-697892/12048/summary.html Comparison SummarySummary:
|
|
I'm a bit surprised why the test here passed; @smuzaffar do we have a mechanism to hide the base release data area ? |
|
@slava77 , not currently for data externals but I can update https://github.com/cms-sw/cmssw-config/blob/scramv3/SCRAM/hooks/runtime/50-remove-release-external-lib to drop the data path from release too. let me update build rules and re-run tests for this PR |
|
please test with cms-sw/cmsdist#6538 |
|
@smuzaffar , if you agree, I will merge this PR with cms-sw/cmssw#32391 once the tests with cms-sw/cmsdist#6538 are finished |
|
please test with cms-sw/cmsdist#6538 |
|
@silviodonato , yes please wait for the results. |
|
-1 Failed Tests: UnitTests RelVals AddOn Unit TestsI found errors in the following unit tests: ---> test runtestPhysicsToolsPatAlgos had ERRORS ---> test testRecoMETMETProducers had ERRORS ---> test testPhase2PixelNtuple had ERRORS ---> test testTauEmbeddingProducers had ERRORS RelVals |
|
merge |
|
this looks good now, I mean bot now properly remove data path of release. |
silviodonato
added a commit
to cms-sw/cmsdist
that referenced
this pull request
Jan 11, 2021
cmsbuild
added a commit
to cms-sw/cmsdist
that referenced
this pull request
Jan 11, 2021
|
@smuzaffar I didn't see the PR from @cmsbuild , so I manually created the new branch (V01-03-00) on cms-data and the PR for cmsdist (cms-sw/cmsdist#6543). I hope everything is ok. |
|
you mean new tag V01-04-00 ... right |
|
the generated release already existed when it tried to create it. |
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.
This PR contains a .root file containing weights for a BDT model used by the LowPtGsfElectronIDProducer module.
cms-sw/cmssw#32391 depends in this PR.