There was a problem while building and running in standalone mode. The logs can be found here.

/run standalone

/run cmssw
/run checks

Looks like cuts are still too loose at high pT and in some regions, I will take a look.

Here is a timing comparison: Reset: 188.2 -> 428.9

If it's from this PR, is the timer out of place now?

/run all

Here is a timing comparison: Reset: 188.2 -> 428.9

If it's from this PR, is the timer out of place now?

Maybe because it is multi-stream? Sometimes the time increases for each category can be unreliable.

The PR was built and ran successfully in standalone mode. Here are some of the comparison plots.

The full set of validation and comparison plots can be found here.

Here is a timing comparison:

   Evt    Hits       MD       LS      T3       T5       pLS       pT5      pT3      TC       Reset    Event     Short             Rate
   avg     30.9    378.1    272.8    125.9     52.2    697.0     11.1    127.1    132.7    188.5      1.7    2018.0    1290.0+/- 307.3     619.9   explicit[s=4] (target branch)
   avg     28.6    381.5    275.9    119.6     49.3    687.5     12.1    126.8    134.7    428.5      1.9    2246.3    1530.1+/- 370.2     682.6   explicit[s=4] (this PR)

Here is a timing comparison: Reset: 188.2 -> 428.9

If it's from this PR, is the timer out of place now?

Maybe because it is multi-stream? Sometimes the time increases for each category can be unreliable.

the last test came up with a similar result.
These are CPU-based, btw.

Current improvement plot, seems a decent chunk of the pT3's get extended to pT4's.

The PR was built and ran successfully with CMSSW. Here are some plots.

The full set of validation and comparison plots can be found here.

No noticeable improvement in momentum resolution.

Plot showing tracks are increasing length outside barrel.