Skip to content

Conversation

@alok
Copy link
Contributor

@alok alok commented Jun 17, 2018

Fixes #2233

@alok alok changed the title ## What do these changes do? Fix #2233 Jun 17, 2018
@AmplabJenkins
Copy link

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/Ray-PRB/6084/
Test PASSed.

@alok
Copy link
Contributor Author

alok commented Jun 17, 2018

@richardliaw

@richardliaw
Copy link
Contributor

Thanks for the patch; I'd like to hold on this for a bit (basically waiting on #2170 and #1646, which I'm trying to patch) in order to reduce introducing more variability in regressions

Copy link
Contributor

@richardliaw richardliaw left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good, but holding until #1646

@richardliaw richardliaw dismissed their stale review June 19, 2018 17:50

Both PRs have landed

@richardliaw
Copy link
Contributor

Hey @alok can you verify there's no performance regression for tuned_examples/pendulum-ppo with this change?

@robertnishihara robertnishihara changed the title Fix #2233 [rllib] Replace tf.minimum with tf.maximum in PPO loss. Jun 19, 2018
Copy link
Contributor

@ericl ericl left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good. Any difference on benchmarks?

@richardliaw
Copy link
Contributor

Addressed in #2366

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants