add Intel TBB #710

wds15 · 2019-06-16T09:50:46Z

Submisison Checklist

Run tests: ./runCmdStanTests.py src/test
Declare copyright holder and open-source license: see below

Summary

This adds the TBB targets to the makefiles. This PR needs to be merged with the Stan-math PR stan-dev/math#1180 and the stan PR stan-dev/stan#2769

Intended Effect

Make the Intel TBB available to stan.

How to Verify

Tests will link against the tbb, tbbmalloc & tbbmalloc_proxy libraries.

Side Effects:

Documentation:

Copyright and Licensing

Please list the copyright holder for the work you are submitting (this will be you or your assignee, such as a university or company): Sebastian Weber

By submitting this pull request, the copyright holder is agreeing to license the submitted work under the following licenses:

Code: BSD 3-clause (https://opensource.org/licenses/BSD-3-Clause)
Documentation: CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

stan-buildbot · 2019-06-16T10:06:48Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.99)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.99)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.01)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.0)
Result: 1.00018007699
Commit hash: 17a94d9

stan-buildbot · 2019-06-16T11:02:13Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.13)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.97)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.04)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.15)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.18)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.14)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.03)
(performance.compilation, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.15)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.02)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.02)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.03)
Result: 1.05870650445
Commit hash: 17a94d9

stan-buildbot · 2019-06-16T11:13:15Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 1.0)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.01)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 0.98)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.02)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 0.99)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 0.99)
Result: 1.00151595872
Commit hash: a508c59

wds15 · 2019-06-16T11:16:31Z

@serban-nicusor-toptal How do the performance tests work when I use the "build with parameters" facility to use non-default stan and stan-math refs to build this PR? Moreover, is it possible to instruct the performance suite to test on macOS specifically (and maybe Linux)?

stan-buildbot · 2019-06-16T11:23:56Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.14)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.95)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.04)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.13)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.18)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.15)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.02)
(performance.compilation, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.01)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.13)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.13)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.03)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.05)
Result: 1.06294752099
Commit hash: a508c59

stan-buildbot · 2019-06-16T13:43:48Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 1.0)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.03)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 0.95)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 0.99)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.99)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.0)
Result: 0.99783416764
Commit hash: de3cb06

stan-buildbot · 2019-06-16T14:18:35Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.1)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.97)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.03)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.21)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.14)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.16)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.02)
(performance.compilation, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.1)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.13)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.03)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.98)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.05)
Result: 1.06266792755
Commit hash: de3cb06

stan-buildbot · 2019-06-17T10:48:43Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.1)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.98)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.03)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.18)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.14)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.14)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.02)
(performance.compilation, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.08)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.12)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.04)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.03)
Result: 1.05885646212
Commit hash: de3cb06

serban-nicusor-toptal · 2019-06-17T20:09:37Z

@wds15 Hey,

It will get the base branch develop,
Checkout cmdstan and submodules from that branch/commit hash in build with parameters.
Then run tests on develop with latest stable submodules versus the branches/hashes in build with parameters.

https://github.com/stan-dev/performance-tests-cmdstan/blob/master/compare-git-hashes.sh

I am currently working on a custom branch in git and job for Jenkins to allow running the tests on mac/win/linux with more arguments for ease of customization.
I will add the possibility to run on a custom set of oses.

See:

Still needs a bit of refactoring to work on win.

mitzimorris · 2019-07-18T18:11:25Z

is this still status WIP?

wds15 · 2019-07-18T18:28:32Z

Yes...the tbb still needs to have settlement on the license

rok-cesnovar

This PR needs a merge with recent develop. I also have a question regarding the manually set CXX_TYPE for clang and g++.

.travis.yml

rok-cesnovar

Additionally the following changes need to be made:

cmdstan manual instructions need to be updated for the STAN_THREADS case, ming32-make, ...
cmdstan wiki Licensing section should have the same note on GPL-2 compliance similar to the TBB Stan Math PR.
update wiki with installation instructions for Windows (PATH and mingw32-make)

See more here: stan-dev/math#1376 (comment)

stan-buildbot · 2019-10-02T19:53:08Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.97)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.03)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.03)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 0.99)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.01)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.03)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 0.99)
Result: 1.00284648814
Commit hash: eb125a9

wds15 · 2019-10-03T07:53:57Z

Right now I am outputting on windows a message about adding the tbb library path to the PATH variable... but we could also just create a wrapper .bat file for each model on Windows. The wrapper will have the hard-coded TBB path in it, set the PATH variable and then start the stan program. This way we also get to control which TBB is being used on the system as we can control to put our TBB in the first place of the PATH. That wrapper should be easy to write.

wds15 · 2019-10-03T08:14:15Z

For reference, here is the windows search policy for dynamic libraries:

https://docs.microsoft.com/en-us/windows/win32/dlls/dynamic-link-library-search-order#standard-search-order-for-desktop-applications

rok-cesnovar · 2019-10-03T08:22:30Z

and then start the stan program.

By this you mean stan compiler? That sound like a nice convenience for Windows users and I very much approve.

But honestly, I dont feel this convenience .bat file is necessary in this initial TBB PR. If you feel that it will be easy to write then go ahead. If not, we can do this separately after this initial batch of TBB PRs get in.

rok-cesnovar · 2019-10-03T08:27:30Z

Or did you mean that a user would run their models through this .bat file?

wds15 · 2019-10-03T08:29:11Z

Yes, users would call the .bat file instead of the .exe file.

I would generate the .bat with the makefile.

So calling the Bernoulli example becomes

bernoulli.bat sample ...

that's it. The bernoulli.bat just wraps the exe call which it will precede with a reset PATH variable.

rok-cesnovar · 2019-10-03T08:31:02Z

This means that PATH is set each time a model is run? Or am I misunderstanding.

What about just setting PATH on make build of CmdStan?

rok-cesnovar

This is almost ready, the doc changes are great, user warnings also.

Please fix the build path issue and remove the changes to the stan submodule reference.

install-tbb.bat

stan-buildbot · 2019-10-06T11:51:28Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.97)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.99)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.01)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.04)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.0)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 0.99)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.01)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.99)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 0.99)
Result: 1.00059541574
Commit hash: 8c18f2b

stan-buildbot · 2019-10-06T12:01:16Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.97)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.98)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 0.98)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.01)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 0.99)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 0.99)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.95)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.0)
Result: 0.98936325739
Commit hash: 8c18f2b

wds15 · 2019-10-06T12:23:05Z

I just did

git checkout origin/develop stan

and it did not change anything as it looks to me. So I am not sure what I should update wrt to the submodule.

EDIT: or did u mean I update the stan submodule to the respective stan with the changes for the TBB? I don't think that is ok... we always need to point to some stan develop branch in cmdstan - so I was assuming that we update cmdstan / stan / math and then the submodule auto-update will get it straight.

rok-cesnovar · 2019-10-06T12:32:51Z

If you look at the changes in the PR it looks as we are changing the stan submodule.which we should not.

stan-buildbot · 2019-10-06T13:17:02Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.96)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.98)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 0.98)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 0.97)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.01)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.01)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 0.99)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.99)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 0.98)
Result: 0.99085982256
Commit hash: 8c18f2b

rok-cesnovar

Looks great! Merge once it passes.

wds15 · 2019-10-06T14:42:33Z

Cool. Let's also wait for the pipeline of stan-math PR-1180 to turn green before we merge this. Then we know that all will work out.

(I really hope that this time Jenkins is stable, the thing started to give again flaky behavior here and there)

rok-cesnovar · 2019-10-06T14:49:10Z

Did you setup 1180 to the TBB stan and cmdstan branches? Yeah, it would be great if we had that green light before.

I would then merge these three in quick succession and then deal with 1376. Then things will get easier.

Will handheld Jenkins if needed, no problem :)

rok-cesnovar · 2019-10-06T14:50:45Z

On the bright side we have come to the point where the only thing keeping us from having TBB in is having to wait for the tests to finish. I am psyched :)

wds15 · 2019-10-06T14:54:29Z

Me too ... looking forward to this.

So yes, I started the Jenkins for PR 1180 of stan-math using the "build with parameters" facility. That lets you test the 1180 with the respective upstream branches which match it.

I will update the PR test for 1180. If you don't mind I will take out the performance things there as these are documented in 1376.

BTW... the very last one to merge is the changes currently on the branch feature/intel-tbb-init in cmdstan. That's a super easy one which can do at the very end (to ensure proper interface initialization).

rok-cesnovar · 2019-10-06T14:59:22Z

I will update the PR test for 1180. If you don't mind I will take out the performance things there as these are documented in 1376.

You mean the PR text? Yes, please do that. Just reference to 1376 for performance stuff. 1180 doesnt really change anything outside of Mac's use of tbbmalloc in terms of performance.

BTW... the very last one to merge is the changes currently on the branch feature/intel-tbb-init in cmdstan.

Yes, I remember. That should be a simple thing.

makefile

stan-buildbot · 2019-10-06T15:06:02Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.01)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.98)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 0.98)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.05)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.06)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 0.99)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.0)
(performance.compilation, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 0.99)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.01)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.98)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.01)
Result: Regex did not match anything
Commit hash: 8b481cd

…mdstan into feature/intel-tbb-lib

wds15 · 2019-10-06T15:37:14Z

Now the final message looks nicer.

rok-cesnovar · 2019-10-06T15:59:56Z

Great, thanks.

Cmstan and Stan PRs are ready and waiting. When/if 1180 passes we can then merge all three. If all goes well and Jenkins plays nicely we should be there in around 3 hours (1h for distribution test and 1h30min for upstream). I will definitely still be at my computer then to merge. Should I just do it then? If all goes green I think we should.

And we then handle 1376 tomorrow once that passes? And the final Cmdstan PR to init the number of threads.

wds15 · 2019-10-06T16:14:31Z

Sure - go ahead once 1180 turns green. Awesome.

stan-buildbot · 2019-10-06T16:28:29Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.06)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.99)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.99)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 0.98)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.01)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.01)
(performance.compilation, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.01)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 0.99)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.01)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.97)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.0)
Result: 1.00140277691
Commit hash: e725788

stan-buildbot · 2019-10-06T16:38:15Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.99)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.0)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 0.98)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.0)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.01)
(performance.compilation, 1.02)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 1.0)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 0.99)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 0.98)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.0)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.0)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 0.99)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.03)
Result: 0.99966520225
Commit hash: e725788

stan-buildbot · 2019-10-06T19:37:41Z

(stat_comp_benchmarks/benchmarks/gp_pois_regr/gp_pois_regr.stan, 1.17)
(stat_comp_benchmarks/benchmarks/low_dim_corr_gauss/low_dim_corr_gauss.stan, 0.98)
(stat_comp_benchmarks/benchmarks/irt_2pl/irt_2pl.stan, 1.05)
(stat_comp_benchmarks/benchmarks/pkpd/one_comp_mm_elim_abs.stan, 1.17)
(stat_comp_benchmarks/benchmarks/eight_schools/eight_schools.stan, 1.2)
(stat_comp_benchmarks/benchmarks/gp_regr/gp_regr.stan, 1.15)
(stat_comp_benchmarks/benchmarks/arK/arK.stan, 1.02)
(performance.compilation, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan, 0.99)
(stat_comp_benchmarks/benchmarks/low_dim_gauss_mix/low_dim_gauss_mix.stan, 1.01)
(stat_comp_benchmarks/benchmarks/sir/sir.stan, 1.11)
(stat_comp_benchmarks/benchmarks/pkpd/sim_one_comp_mm_elim_abs.stan, 1.08)
(stat_comp_benchmarks/benchmarks/garch/garch.stan, 1.03)
(stat_comp_benchmarks/benchmarks/gp_regr/gen_gp_data.stan, 1.04)
(stat_comp_benchmarks/benchmarks/arma/arma.stan, 1.07)
Result: 1.07110834232
Commit hash: e725788

add TBB targets

17a94d9

add TBB build specifics

a508c59

wds15 changed the title ~~add TBB targets~~ WIP add TBB targets Jun 16, 2019

ensure correct CXX_TYPE on travis

de3cb06

serban-nicusor-toptal modified the milestones: 2.20.0, 2.19.1++ Jul 18, 2019

rok-cesnovar requested changes Sep 29, 2019

View reviewed changes

.travis.yml Outdated Show resolved Hide resolved

.travis.yml Outdated Show resolved Hide resolved

rok-cesnovar mentioned this pull request Oct 2, 2019

integrate Intel TBB stan-dev/math#1376

Merged

5 tasks

rok-cesnovar requested changes Oct 2, 2019

View reviewed changes

wds15 added 2 commits October 2, 2019 21:22

Merge remote-tracking branch 'origin/develop' into feature/intel-tbb-lib

9430713

add a note for windows builds

eb125a9

bump documentation

8c18f2b

rok-cesnovar requested changes Oct 6, 2019

View reviewed changes

install-tbb.bat Show resolved Hide resolved

align stan submodule

8b481cd

rok-cesnovar approved these changes Oct 6, 2019

View reviewed changes

rok-cesnovar reviewed Oct 6, 2019

View reviewed changes

makefile Outdated Show resolved Hide resolved

rok-cesnovar reviewed Oct 6, 2019

View reviewed changes

makefile Outdated Show resolved Hide resolved

wds15 added 2 commits October 6, 2019 17:15

Merge branch 'feature/intel-tbb-lib' of https://github.com/stan-dev/c…

e8b1b2a

…mdstan into feature/intel-tbb-lib

address review comments

dc03e9b

rok-cesnovar approved these changes Oct 6, 2019

View reviewed changes

make echo nicer at the end of the build target

e725788

rok-cesnovar merged commit a58ddcd into develop Oct 6, 2019

wds15 mentioned this pull request Oct 7, 2019

Add Intel TBB #709

Closed

mitzimorris deleted the feature/intel-tbb-lib branch January 21, 2020 16:26

add Intel TBB #710

add Intel TBB #710

Conversation

wds15 commented Jun 16, 2019 • edited by rok-cesnovar Loading

Submisison Checklist

Summary

Intended Effect

How to Verify

Side Effects:

Documentation:

Copyright and Licensing

stan-buildbot commented Jun 16, 2019

stan-buildbot commented Jun 16, 2019

stan-buildbot commented Jun 16, 2019

wds15 commented Jun 16, 2019

stan-buildbot commented Jun 16, 2019

stan-buildbot commented Jun 16, 2019

stan-buildbot commented Jun 16, 2019

stan-buildbot commented Jun 17, 2019

serban-nicusor-toptal commented Jun 17, 2019

mitzimorris commented Jul 18, 2019

wds15 commented Jul 18, 2019

rok-cesnovar left a comment

Choose a reason for hiding this comment

rok-cesnovar left a comment

Choose a reason for hiding this comment

stan-buildbot commented Oct 2, 2019

wds15 commented Oct 3, 2019

wds15 commented Oct 3, 2019

rok-cesnovar commented Oct 3, 2019

rok-cesnovar commented Oct 3, 2019

wds15 commented Oct 3, 2019

rok-cesnovar commented Oct 3, 2019

rok-cesnovar left a comment • edited Loading

Choose a reason for hiding this comment

stan-buildbot commented Oct 6, 2019

stan-buildbot commented Oct 6, 2019

wds15 commented Oct 6, 2019 • edited Loading

rok-cesnovar commented Oct 6, 2019

stan-buildbot commented Oct 6, 2019

rok-cesnovar left a comment

Choose a reason for hiding this comment

wds15 commented Oct 6, 2019

rok-cesnovar commented Oct 6, 2019

rok-cesnovar commented Oct 6, 2019

wds15 commented Oct 6, 2019

rok-cesnovar commented Oct 6, 2019 • edited Loading

stan-buildbot commented Oct 6, 2019

wds15 commented Oct 6, 2019

rok-cesnovar commented Oct 6, 2019

wds15 commented Oct 6, 2019

stan-buildbot commented Oct 6, 2019

stan-buildbot commented Oct 6, 2019

stan-buildbot commented Oct 6, 2019

wds15 commented Jun 16, 2019 •

edited by rok-cesnovar

Loading

rok-cesnovar left a comment •

edited

Loading

wds15 commented Oct 6, 2019 •

edited

Loading

rok-cesnovar commented Oct 6, 2019 •

edited

Loading