New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
new PhaseII trainings for DeepCSV and DeepJet #42
Conversation
A new Pull Request was created by @mneukum (Max Neukum) for branch master. @perrotta, @smuzaffar, @mrodozov, @cmsbuild, @slava77, @jpata can you please review it and eventually sign? Thanks. |
Pull request #42 was updated. |
|
Pull request #42 was updated. |
@slava77 Let me know, if I should clean up the commit history before merging. |
Pull request #42 was updated. |
If I understood this correctly
Please check/confirm. |
it looks like moving of the files in multiple commits still keeps them in the git history and increases the repo size more than necessary. |
Pull request #42 was updated. |
I can confirm the 2nd bullet point. To my knowledge, there is no definite date for it yet. |
If the timeline for deploying in offline is unclear and there are no other (e.g. analysis) use cases for this file, it may be better not to push it in; partly to avoid a case of having a file that will never be used. BTW, is DeepFlavourV01_PhaseII/model.onnx going to be used in the HLT TDR? |
Yes, both
For HLT studies after the TDR we would make use of these files, but I think we don't need them in a central release. So I would also be fine with not pushing them. |
we will need to make a backport V01-03-00...V01-09-00 has many more changes already, not restricted to phase-2. @smuzaffar @mrodozov please make a CMSSW_11_1_X branch for this repository so that a backport to it (V01-03-00 + the diffs of this PR) can be made. |
Pull request #42 was updated. |
I updated the PR to only include the model files used in HLT TDR. |
https://github.com/cms-data/RecoBTag-Combined/tree/CMSSW_11_1_X is now available for 11.1.X. @mneukum , can you please submit a PR for CMSSW_11_1_X branch too |
I just created a new PR to CMSSW_11_1_X branch: |
@cmsbuild please test |
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-4f0049/13017/summary.html Comparison SummarySummary:
|
I see that the discriminant shape changed quite significantly for the phase-2 deepCSV (default offline). There are not enough events to conclude from the ROC curves in the DQM plots if this is OK, although it seems like it's fine (red is with this PR)
Please clarify which slide in https://indico.cern.ch/event/1008226/#73-status-of-phase2-training-o corresponds to this update (unfortunately I was not able to match). |
@slava77 |
+1 based on the preceding review; jenkins tests and #42 (comment) |
ping |
@smuzaffar @silviodonato |
+externals |
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. @silviodonato, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2) |
Adding updated versions of DeepCSV and DeepJet training for PhaseII.