[RLlib] [fix] [metrics] avoid biasing the throughput #57215
+39
−26
Conversation
This file contains hidden or 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
Signed-off-by: simonsays1980 <[email protected]>
Signed-off-by: simonsays1980 <[email protected]>
simonsays1980
added
rllib
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
kamil-kaczmarek
changed the title
kamil-kaczmarek
requested changes
Oct 7, 2025
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
… there. Signed-off-by: simonsays1980 <[email protected]>
| self.metrics.log_value( | ||
| key=(mid, NUM_MODULE_STEPS_TRAINED_LIFETIME), | ||
| value=module_batch_size, | ||
| reduce="sum", |
There was a problem hiding this comment.
In the learner we report this with_throughput=True
Let's add with_throughput=True here to standardize this and make sure we have consistent metrics.
There was a problem hiding this comment.
We'd need to add with_throughput=True here.
kamil-kaczmarek
requested changes
Oct 8, 2025
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
…tialLearner' and fixed some linting. Signed-off-by: simonsays1980 <[email protected]>
simonsays1980
force-pushed
the
fix-module-steps-throughputs
branch
from
7c42aaa to
ddcaeab
Compare
simonsays1980
requested review from
kamil-kaczmarek
and removed request for
a team,
edoakes and
jjyao
kamil-kaczmarek
requested changes
Oct 9, 2025
…bleLearner'. Signed-off-by: simonsays1980 <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Why are these changes needed?
Computing the
num_module_steps_trained_(lifetime)_throughputmetrics are biased due to the way how we record throughput times in a loop over module batches. This PR offers a fix to this bias.Related issue number
Checks
git commit -s) in this PR.method in Tune, I've added it in
doc/source/tune/api/under thecorresponding
.rstfile.