Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[RLlib] Add gradient logging as default. #47451

Merged

Conversation

simonsays1980
Copy link
Collaborator

@simonsays1980 simonsays1980 commented Sep 2, 2024

Why are these changes needed?

Gradients logging is highly important for training neural networks. It enables users to monitor training performance and health. Right now, a user can only monitor gradients when she uses clipping and then also only, if the "global_norm" is used for clipping. This PR instead proposes gradients logging as a default option in RLlib.

Related issue number

Checks

  • I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
  • I've run scripts/format.sh to lint the changes in this PR.
  • I've included any doc changes needed for https://docs.ray.io/en/master/.
    • I've added any new APIs to the API Reference. For example, if I added a
      method in Tune, I've added it in doc/source/tune/api/ under the
      corresponding .rst file.
  • I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
  • Testing Strategy
    • Unit tests
    • Release tests
    • This PR is not tested :(

…om clipping to log the global norm to results. Added option to 'AlgorithmConfig.debugging' and set default to 'True'.

Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
@sven1977 sven1977 changed the title [RLlib] - Add gradient logging as default [RLlib] Add gradient logging as default. Sep 2, 2024
@sven1977
Copy link
Contributor

sven1977 commented Sep 2, 2024

This is great! Left some questions/nits. ...

Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
@@ -500,6 +500,7 @@ def __init__(self, algo_class: Optional[type] = None):
self.logger_config = None
self.log_level = "WARN"
self.log_sys_usage = True
self.log_gradients = True
Copy link
Contributor

Choose a reason for hiding this comment

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

nit nit: Can you move this to the default-value block, that's headered with reporting(). Keeps the c'tor structured.

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Great catch! Moved it to reporting defaults.

Copy link
Contributor

@sven1977 sven1977 left a comment

Choose a reason for hiding this comment

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

LGTM. Just please move the default value in the AlgorithmConfig constructor into the block commented as # .reporting().

@sven1977 sven1977 enabled auto-merge (squash) September 2, 2024 15:35
@github-actions github-actions bot added the go add ONLY when ready to merge, run all tests label Sep 2, 2024
Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
@sven1977 sven1977 enabled auto-merge (squash) September 9, 2024 16:46
…ming out.

Signed-off-by: simonsays1980 <simon.zehnder@gmail.com>
@sven1977 sven1977 merged commit 6a62b0f into ray-project:master Sep 17, 2024
5 checks passed
ujjawal-khare pushed a commit to ujjawal-khare-27/ray that referenced this pull request Oct 15, 2024
Signed-off-by: ujjawal-khare <ujjawal.khare@dream11.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
go add ONLY when ready to merge, run all tests
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants