Submission of Local Linearity Regularizer + Adversarially Trained model

[emmyqin]

This is a submission for the unrestricted adversarial challenge: Phase I. The entry is uses a pre-trained ImageNet model with Local Linearity Regularizer, then adversarially trained using birds-or-bicycles dataset (train and extras) as provided by the challenge.

[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).

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

Once you've signed (or fixed any issues), please reply here with @googlebot 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.

ℹ️ Googlers: Go here for more info.

[googlebot]

CLAs look good, thanks!

ℹ️ Googlers: Go here for more info.

[nottombrown]

This looks good to me. Thanks for the link to a github repo that's clear and easy to execute https://github.com/deepmind/deepmind-research/tree/master/unrestricted_advx

I'd like to merge this in, but it looks like the python 3.5 tests are bitrotten, and since I've left Google, I no longer have admin access to disable 3.5 tests so that this could be merged.

@carlini, would you be up to do the following?

1) Disable the bitrotten python 3.5 travis tests (but leave 2.7 intact) 2) Either merge this in, or give me write-access so that I can merge things in the future (if that is easy)

Thanks!

[carlini]

Sorry for the delay on this. Tom you should have had access--not sure why you didn't, you do now.

Do we want to remove the 3.5 tests or should we fix them?

[juesato]

Commenting so that Github know to notify me about new comments.

[juesato]

Hi @nottombrown , are you able to merge this now?

[nottombrown]

Merged! Thanks for you patience @juesato and @emmyqin

[nottombrown]

@carlini - I'll remove the bitrotten 3.5 test for now and will accept PRs to fix them