bayesian optimization draft

[libbyandhelen]

This is a draft for bayesian optimization (work in progress) #15 @gaocegege @YujiOshima

[googlebot]

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed (or fixed any issues), please reply here (e.g. I signed it!) and we'll verify it.

---

What to do if you already signed the CLA

Individual signers

• It's possible we don't have your GitHub username or you're using a different email address on your commit. Check your existing CLA data and verify that your email is set on your git commits.

Corporate signers

• Your company has a Point of Contact who decides which employees are authorized to participate. Ask your POC to be added to the group of authorized contributors. If you don't know who your Point of Contact is, direct the Google project maintainer to go/cla#troubleshoot (Public version).
• The email used to register you as an authorized contributor must be the email used for the Git commit. Check your existing CLA data and verify that your email is set on your git commits.
• The email used to register you as an authorized contributor must also be attached to your GitHub account.

[libbyandhelen]

/assign @gaocegege @YujiOshima

[googlebot]

CLAs look good, thanks!

[gaocegege]

/hold

[gaocegege]

Thanks for your awesome contribution! I will take a look soon.

[gaocegege]

When the PR is ready to merge, please use /hold cancel to cancel the hold.

[gaocegege]

Generally LGTM. @YujiOshima I think we should support python suggestion by default thus we could refer to the implementation in other systems, e.g. https://github.com/tobegit3hub/advisor

[gaocegege]

Please remove the new line here.

[gaocegege]

Inconsistent blanks here. Please fix it.

[gaocegege]

/ok-to-test

Please rebase the master and I think the test should pass.

[YujiOshima]

You should use /api/api.proto.
We should generate suggestion API from the same file not to make inconsistency between go and python.

[YujiOshima]

@libbyandhelen Please move suggestion/python/api/ directory to api/python and store your bayesian-optimization source codes under suggestion/bayesianoptimization directory in conformity to other services.

[YujiOshima]

@libbyandhelen And please dockerize your suggestion service and add k8s manifest.
I'm sorry there is no nice documents. Please refer other services.
https://github.com/kubeflow/hp-tuning/blob/master/suggestion/random/Dockerfile
https://github.com/kubeflow/hp-tuning/tree/master/manifests/vizier/suggestion/random

[gaocegege]

@libbyandhelen If you have any question about dockerize your suggestion service and add k8s manifest please at me and I can help you.

[libbyandhelen]

/hold cancel

[gaocegege]

I think you may have discussions with @YujiOshima about api proto. Thus I only reviewed other code and LGTM.

Thanks

[libbyandhelen]

@gaocegege OK, got it.
@YujiOshima
Do you think the api proto is OK now? I did not change the protocol in api.proto.
I used SetSuggestionParameters to set all the parameters needed for the suggestion service, and StopSuggestion for deleting the relevant information of the specific study stored in the service.
completed_trials in GenerateTrials is used to pass information of trials for now, since the interface for DB is not ready.

[YujiOshima]

@libbyandhelen OK, we will add DB interface for python later.
Finally, please add build and deploy test for this suggestion to test script.
Since this is in python, you don't need go build test. Only gcloud container builds submit and deploy.
https://github.com/kubeflow/hp-tuning/blob/master/test/scripts/build.sh
https://github.com/kubeflow/hp-tuning/blob/master/test/scripts/run-tests.sh

[gaocegege]

I think it is the last step! 🎉

Thanks for your contribution!

[YujiOshima]

@libbyandhelen Thank you for your great contribution!
/lgtm

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: YujiOshima

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [YujiOshima]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment