Add Phase-2 Level-1 trigger ML MET emulator [14_0_0_pre3] #1193
Conversation
jmduarte
changed the title
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. Attempts to compile this PR succeeded!
|
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. I found no issues with the code checks!
I found no issues with the headers!
|
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. I found 1 files that did not meet formatting requirements:
Please run
|
aloeliger
added
Phase-2
| switch (abs(pdgId)) { | ||
| case 211: | ||
| return 1; | ||
| case 130: | ||
| return 2; | ||
| case 22: | ||
| return 3; | ||
| case 13: | ||
| return 4; | ||
| case 11: | ||
| return 5; | ||
| default: | ||
| return 0; | ||
| } |
There was a problem hiding this comment.
As a slightly picky thing, would you mind commenting the actual particle name after their PDG codes? It may save a few people a trip to the handbook.
And just so I understand, is the return value here supposed to help with indexing a one hot encoded feature?
There was a problem hiding this comment.
Added the particle names.
Yes, this is so we can input these into a Keras embedding layer: https://keras.io/api/layers/core_layers/embedding/
It's sort of like a one-hot encoding, but potentially more efficient (smaller dimensional space).
| @@ -12,6 +12,8 @@ | |||
| <use name="tensorflow"/> | |||
| <use name="roottmva"/> | |||
| <use name="hls"/> | |||
| <use name="hls4mlEmulatorExtras"/> | |||
| <use name="L1METML"/> | |||
There was a problem hiding this comment.
Can I ask what the earliest version of CMSSW is that has L1METML? This isn't a problem for you, but we may need to make sure the IB is up to a version that supports your model in CMSSW.
There was a problem hiding this comment.
This was merged on December 20: cms-sw/cmsdist#8901 so I believe CMSSW_14_0_0_pre2 has it (built on December 22), but if not, definitely CMSSW_14_0_0_pre3 has it.
aloeliger
added
the
Physics Affecting
|
@jmduarte Just a ping on the review comments |
aloeliger
changed the base branch from
phase2-l1t-integration-13_3_0_pre3
to
phase2-l1t-integration-14_0_0_pre3
jmduarte
force-pushed
the
l1metml_1330pre3
branch
from
442eb89 to
61c2ae0
Compare
|
Rebased PR, and replied to questions, sorry for the delay! |
jmduarte
changed the title
jmduarte
force-pushed
the
l1metml_1330pre3
branch
from
5b61dd3 to
315058a
Compare
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. Attempts to compile this PR succeeded!
|
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. I found no issues with the code checks!
I found no issues with the headers!
|
|
Hello, I'm triggerDoctor. @aloeliger is testing this script for L1T offline software validation. I found no files with code format issues!
|
|
@jmduarte In order to start constructing a protoytype branch for the menu team to validate, this PR is being merged immediately. If there is no corresponding PR to CMSSW, please make it now. |
aloeliger
merged commit 180bc4a
into
cms-l1t-offline:phase2-l1t-integration-14_0_0_pre3
|
@aloeliger thanks! I see you already commented on the CMSSW PR cms-sw#43633. One question I had is should I open a backport to 14_0_X in CMSSW? Or should I ask this in the other PR? |
PR description:
PR validation: