Use pip install tf-nightly within .travis.yml
#852
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
instead of pip installing the nightly whl URLs.
chihuahua
changed the title
pip install tf-nightly
|
Well it looks like that helped, but now some tests are failing on a missing GLIBCXX_3.4.20. It sounds like that might be because we're using trusty (14.04) for Travis but apparently TensorFlow builds are on 16.04 and hence only support GLIBCXX versions from what 16.04 ships with on up: ContinuumIO/anaconda-issues#483 (comment) And I guess Travis has no real plans to support 16.04 Ubuntu, or for that matter anything beyond 14.04, ugh: travis-ci/travis-ci#5821 So... maybe there's a way we can install a newer glibcxx on the machine prior to running our tests? |
|
The folks who reported tensorflow/tensorflow#15777 ended up dockerizing their tests: horovod/horovod#136 ... |
chihuahua
changed the title
pip install tf-nightlypip install tf-nightly within travis.yml
chihuahua
changed the title
pip install tf-nightly within travis.ymlpip install tf-nightly within .travis.yml
nfelt
approved these changes
Jan 5, 2018
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The security certificate of ci.tensorflow.org expired, causing TensorBoard's tests on travis to fail. See tensorflow/tensorflow#15827.
This change enables our travis environment to successfully install tf-nightly.
Unfortunately, the travis tests still fail after this PR goes in, but that is because of tensorflow/tensorflow#15777. We are still deciding what to do there - we might have to dockerize tests.
Anyways, fixing the issue caused by the expired certificate will let us focus on that different issue. Thanks to @nfelt for suggesting that we try using
pip install tf-nightly.