[BugFix] Better account of composite distributions in PPO #2622

vmoens · 2024-12-02T16:24:12Z

Stack from ghstack (oldest at bottom):

-> [BugFix] Better account of composite distributions in PPO #2622

Happy to massage this a little bit if the API isn't optimal!
Do you think we're covering all (edge) cases?

cc @louisfaury @albertbou92

[ghstack-poisoned]

ghstack-source-id: 3d86f99bc5b20a53e4092d786e96a5f7e83405ac Pull Request resolved: #2622

pytorch-bot · 2024-12-02T16:24:17Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2622

📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 11 New Failures, 2 Unrelated Failures

As of commit aa29ad9 with merge base d537dcb ():

NEW FAILURES - The following jobs have failed:

Continuous Benchmark (PR) / CPU Pytest benchmark (gh)
FAILED ../../../../../../tmp/test_objectives_benchmarks.py::test_sac_speed[True-None] - torch._dynamo.exc.Unsupported: Graph break under GenericContextWrappingVariable
Continuous Benchmark (PR) / GPU Pytest benchmark (gh)
FAILED ../../../../tmp/test_objectives_benchmarks.py::test_sac_speed[True-None] - torch._dynamo.exc.Unsupported: Graph break under GenericContextWrappingVariable
SOTA Tests on Linux / tests (3.9, 12.1) / linux-job (gh)
RuntimeError: Command docker exec -t 8f1346060336615f5ba76c8d44ae80404052f41d4f1876148f3108164fa158f6 /exec failed with exit code 1
Unit-tests on Linux / tests-cpu (3.10) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-cpu (3.11) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-cpu (3.12) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-cpu (3.9) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-cpu-oldget (3.12) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-gpu (3.11, 12.1) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]
Unit-tests on Linux / tests-olddeps (3.8, 11.6) / linux-job (gh)
RuntimeError: Command docker exec -t 8797fcb2caee0c84ffa7237c2466ec7246962b1a345b080a2e135055ac48d39c /exec failed with exit code 1
Unit-tests on Linux / tests-stable-gpu (3.10, 11.8) / linux-job (gh)
test/test_loggers.py::TestMLFlowLogger::test_log_video[steps1]

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

Habitat Tests on Linux / tests (3.9, 12.1) / linux-job (gh) (trunk failure)
AttributeError: _ARRAY_API not found
Unit-tests on Windows / unittests-cpu / windows-job (gh) (trunk failure)
CommandNotFoundError: Your shell has not been properly configured to use 'conda deactivate'.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

ghstack-source-id: 3d86f99bc5b20a53e4092d786e96a5f7e83405ac Pull Request resolved: #2622

louisfaury · 2024-12-02T21:47:36Z

lgtm -- thanks @vmoens for looking into this!

Update

aa29ad9

[ghstack-poisoned]

vmoens added a commit that referenced this pull request Dec 2, 2024

[BugFix] Better account of composite distributions in PPO

237b101

ghstack-source-id: 3d86f99bc5b20a53e4092d786e96a5f7e83405ac Pull Request resolved: #2622

facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Dec 2, 2024

vmoens added the bug Something isn't working label Dec 2, 2024

vmoens merged commit aa29ad9 into gh/vmoens/46/base Dec 2, 2024
60 of 73 checks passed

vmoens added a commit that referenced this pull request Dec 2, 2024

[BugFix] Better account of composite distributions in PPO

90c8e40

ghstack-source-id: 3d86f99bc5b20a53e4092d786e96a5f7e83405ac Pull Request resolved: #2622

vmoens deleted the gh/vmoens/46/head branch December 2, 2024 17:44

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[BugFix] Better account of composite distributions in PPO #2622

[BugFix] Better account of composite distributions in PPO #2622

vmoens commented Dec 2, 2024 •

edited

Loading

pytorch-bot bot commented Dec 2, 2024 •

edited

Loading

louisfaury commented Dec 2, 2024

[BugFix] Better account of composite distributions in PPO #2622

[BugFix] Better account of composite distributions in PPO #2622

Conversation

vmoens commented Dec 2, 2024 • edited Loading

pytorch-bot bot commented Dec 2, 2024 • edited Loading

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2622

❌ 11 New Failures, 2 Unrelated Failures

louisfaury commented Dec 2, 2024

vmoens commented Dec 2, 2024 •

edited

Loading

pytorch-bot bot commented Dec 2, 2024 •

edited

Loading