[Enhancement] Log mean/max KL across PPO update phase

[pjain03]

Description

In ppo.py (and ppo_continuous_action.py), losses/approx_kl and losses/old_approx_kl reflect the last minibatch of the final epoch in each update. This proposes logging simple aggregates over the PPO update phase:

• losses/approx_kl_mean (mean over all minibatches/epochs in the update)
• losses/approx_kl_max (max over all minibatches/epochs)
• losses/old_approx_kl_mean
• losses/old_approx_kl_max
  This is non-performance-impacting, small quality-of-life improvement: these aggregates can make KL trends easier to read in noisier settings. No behavior changes, logging only.

Types of changes

• Bug fix
• New feature
• New algorithm
• Documentation

Checklist:

• I've read the CONTRIBUTION guide (required).
• I have ensured pre-commit run --all-files passes (required).
• I have updated the tests accordingly (if applicable).
• I have updated the documentation and previewed the changes via mkdocs serve.
  • I have explained note-worthy implementation details.
  • I have explained the logged metrics.
  • I have added links to the original paper and related papers.

If you need to run benchmark experiments for a performance-impacting changes:

• I have contacted @vwxyzjn to obtain access to the openrlbenchmark W&B team.
• I have used the benchmark utility to submit the tracked experiments to the openrlbenchmark/cleanrl W&B project, optionally with --capture_video.
• I have performed RLops with python -m openrlbenchmark.rlops.
  • For new feature or bug fix:
    • I have used the RLops utility to understand the performance impact of the changes and confirmed there is no regression.
  • For new algorithm:
    • I have created a table comparing my results against those from reputable sources (i.e., the original paper or other reference implementation).
  • I have added the learning curves generated by the python -m openrlbenchmark.rlops utility to the documentation.
  • I have added links to the tracked experiments in W&B, generated by python -m openrlbenchmark.rlops ....your_args... --report, to the documentation.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Preview	Comments	Updated (UTC)
cleanrl	Ready	Preview	Comment	Aug 27, 2025 1:28am