[rllib] Decreases log quantity for learning tests by pseudo-rnd-thoughts · Pull Request #59005 · ray-project/ray

pseudo-rnd-thoughts · 2025-11-26T16:23:32Z

Description

Reviewing our testing logs, they can often be incredibly long. This PR aims to reduce them by changing three things

By default, the CLIReporter in run_rllib_example_script_experiment will report an algorithms training results at least every 5 seconds. This PR adds a tune-max-report-freq argument that we keep at 5 for end-users while in tests we change it to 30 seconds
Change the verbosity of the tune results from 2 to 1 when testing
Removed WARNING impala_learner.py:576 -- No old learner state to remove from the queue. warnings

Signed-off-by: Mark Towers <mark@anyscale.com>

pseudo-rnd-thoughts · 2025-11-27T12:16:53Z

Reviewing the buildkite logs, the problem isn't the Tune CLIReporter frequency as I originally believed but the print of the results after each train I believe.

github-actions · 2025-12-11T12:25:57Z

This pull request has been automatically marked as stale because it has not had
any activity for 14 days. It will be closed in another 14 days if no further activity occurs.
Thank you for your contributions.

You can always ask for help on our discussion forum or Ray's public slack channel.

If you'd like to keep this open, just leave any comment, and the stale label will be removed.

github-actions · 2025-12-26T00:41:19Z

This pull request has been automatically closed because there has been no more activity in the 14 days
since being marked stale.

Please feel free to reopen or open a new pull request if you'd still like this to be addressed.

Again, you can always ask for help on our discussion forum or Ray's public slack channel.

Thanks again for your contribution!

# Conflicts: # rllib/utils/test_utils.py

Signed-off-by: Mark Towers <mark@anyscale.com>

cursor · 2026-01-07T19:45:21Z

rllib/BUILD.bazel

    ],
    # Include the offline data files.
    data = [
+        "--tune-max-report-freq=30",


Argument incorrectly placed in Bazel data instead of args

High Severity

The --tune-max-report-freq=30 argument is added to the data attribute instead of the args attribute in multiple py_test rules. In Bazel, data specifies data files needed at runtime while args specifies command-line arguments. This affects 10 test definitions: learning_tests_cartpole_bc, learning_tests_cartpole_bc_gpu, learning_tests_cartpole_bc_with_offline_evaluation, learning_tests_cartpole_bc_with_offline_evaluation_gpu, learning_tests_pendulum_cql, learning_tests_pendulum_cql_gpu, learning_tests_pendulum_iql, learning_tests_pendulum_iql_gpu, learning_tests_cartpole_marwil, and learning_tests_cartpole_marwil_gpu. The argument won't be passed to the scripts and Bazel may fail trying to find a file with that name.

Additional Locations (2)

rllib/BUILD.bazel#L433-L434

rllib/BUILD.bazel#L514-L515

Signed-off-by: Mark Towers <mark@anyscale.com>

rllib/examples/utils.py

kamil-kaczmarek · 2026-01-08T08:02:39Z

rllib/examples/utils.py

nit: we can create CLIReporter regardless the value of the args.num_agents (L658).

kamil-kaczmarek

LGTM! Just left smal nit to improve

Signed-off-by: Mark Towers <mark@anyscale.com>

## Description Reviewing our testing logs, they can often be incredibly long. This PR aims to reduce them by changing three things 1. By default, the CLIReporter in `run_rllib_example_script_experiment` will report an algorithms training results at least every 5 seconds. This PR adds a `tune-max-report-freq` argument that we keep at 5 for end-users while in tests we change it to 30 seconds 2. Change the verbosity of the tune results from 2 to 1 when testing 3. Removed ` WARNING impala_learner.py:576 -- No old learner state to remove from the queue.` warnings --------- Signed-off-by: Mark Towers <mark@anyscale.com> Co-authored-by: Mark Towers <mark@anyscale.com> Signed-off-by: jasonwrwang <jasonwrwang@tencent.com>

## Description Reviewing our testing logs, they can often be incredibly long. This PR aims to reduce them by changing three things 1. By default, the CLIReporter in `run_rllib_example_script_experiment` will report an algorithms training results at least every 5 seconds. This PR adds a `tune-max-report-freq` argument that we keep at 5 for end-users while in tests we change it to 30 seconds 2. Change the verbosity of the tune results from 2 to 1 when testing 3. Removed ` WARNING impala_learner.py:576 -- No old learner state to remove from the queue.` warnings --------- Signed-off-by: Mark Towers <mark@anyscale.com> Co-authored-by: Mark Towers <mark@anyscale.com>

## Description Reviewing our testing logs, they can often be incredibly long. This PR aims to reduce them by changing three things 1. By default, the CLIReporter in `run_rllib_example_script_experiment` will report an algorithms training results at least every 5 seconds. This PR adds a `tune-max-report-freq` argument that we keep at 5 for end-users while in tests we change it to 30 seconds 2. Change the verbosity of the tune results from 2 to 1 when testing 3. Removed ` WARNING impala_learner.py:576 -- No old learner state to remove from the queue.` warnings --------- Signed-off-by: Mark Towers <mark@anyscale.com> Co-authored-by: Mark Towers <mark@anyscale.com> Signed-off-by: peterxcli <peterxcli@gmail.com>

[rllib] Reduce log frequency for learning tests

941977a

Signed-off-by: Mark Towers <mark@anyscale.com>

pseudo-rnd-thoughts added the rllib RLlib related issues label Nov 26, 2025

github-actions bot added the stale The issue is stale. It will be closed within 7 days unless there are further conversation label Dec 11, 2025

github-actions bot closed this Dec 26, 2025

pseudo-rnd-thoughts reopened this Dec 29, 2025

pseudo-rnd-thoughts added unstale A PR that has been marked unstale. It will not get marked stale again if this label is on it. and removed stale The issue is stale. It will be closed within 7 days unless there are further conversation labels Dec 29, 2025

Mark Towers added 2 commits January 7, 2026 19:14

Merge branch 'master' into reduce-learning-log-freq

b30b37b

# Conflicts: # rllib/utils/test_utils.py

Change verbose=1 when as_test=True

3088e6c

Signed-off-by: Mark Towers <mark@anyscale.com>

pseudo-rnd-thoughts marked this pull request as ready for review January 7, 2026 19:39

pseudo-rnd-thoughts requested a review from a team as a code owner January 7, 2026 19:39

Update BUILD.bazel to use tune-max-report-freq

8ac9635

Signed-off-by: Mark Towers <mark@anyscale.com>

pseudo-rnd-thoughts changed the title ~~[rllib] Increase log frequency for learning tests~~ [rllib] Decreases log quantity for learning tests Jan 7, 2026

cursor bot reviewed Jan 7, 2026

View reviewed changes

Improvements

5749b7b

Signed-off-by: Mark Towers <mark@anyscale.com>

cursor bot reviewed Jan 7, 2026

View reviewed changes

rllib/examples/utils.py Outdated Show resolved Hide resolved

kamil-kaczmarek assigned pseudo-rnd-thoughts and kamil-kaczmarek Jan 8, 2026

kamil-kaczmarek reviewed Jan 8, 2026

View reviewed changes

kamil-kaczmarek approved these changes Jan 8, 2026

View reviewed changes

Mark Towers added 2 commits January 8, 2026 09:44

Remove args.num_agents > 0 condition

d655236

Signed-off-by: Mark Towers <mark@anyscale.com>

Add progress_reporter for num_agents == 0 and > 0

59a4fb9

Signed-off-by: Mark Towers <mark@anyscale.com>

pseudo-rnd-thoughts added the go add ONLY when ready to merge, run all tests label Jan 8, 2026

ArturNiederfahrenhorst merged commit bbb55ac into ray-project:master Jan 9, 2026
7 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[rllib] Decreases log quantity for learning tests#59005

[rllib] Decreases log quantity for learning tests#59005
ArturNiederfahrenhorst merged 7 commits intoray-project:masterfrom
pseudo-rnd-thoughts:reduce-learning-log-freq

pseudo-rnd-thoughts commented Nov 26, 2025 •

edited

Loading

Uh oh!

pseudo-rnd-thoughts commented Nov 27, 2025

Uh oh!

github-actions bot commented Dec 11, 2025

Uh oh!

github-actions bot commented Dec 26, 2025

Uh oh!

cursor bot Jan 7, 2026

Uh oh!

Uh oh!

kamil-kaczmarek Jan 8, 2026

Uh oh!

pseudo-rnd-thoughts Jan 8, 2026

Uh oh!

kamil-kaczmarek left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

pseudo-rnd-thoughts commented Nov 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Uh oh!

pseudo-rnd-thoughts commented Nov 27, 2025

Uh oh!

github-actions bot commented Dec 11, 2025

Uh oh!

github-actions bot commented Dec 26, 2025

Uh oh!

cursor bot Jan 7, 2026

Choose a reason for hiding this comment

Argument incorrectly placed in Bazel data instead of args

Uh oh!

Uh oh!

kamil-kaczmarek Jan 8, 2026

Choose a reason for hiding this comment

Uh oh!

pseudo-rnd-thoughts Jan 8, 2026

Choose a reason for hiding this comment

Uh oh!

kamil-kaczmarek left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

pseudo-rnd-thoughts commented Nov 26, 2025 •

edited

Loading