fix(launcher): use afterany dependency for allow_to_fail pipelines

[yeyu-nvidia]

Summary

• nemo-run's SlurmExecutor defaults to dependency_type="afterok", which cancels all downstream Slurm tasks when a predecessor times out (TIMEOUT) or fails
• For pipelines with allow_to_fail=True, this changes the dependency type to "afterany" so subsequent tasks run regardless of predecessor exit status
• This unblocks EAGLE3 multi-step pipelines where task_0 (data generation) may time out but task_1+ should still run on whatever data was produced

Test plan

• Verify existing launcher unit tests pass (uv run python3 -m pytest tests/ -v in tools/launcher/)
• Submit an EAGLE3 pipeline with allow_to_fail: true and confirm task_1 runs after task_0 times out
• Verify pipelines without allow_to_fail still use default afterok behavior

🤖 Generated with Claude Code

Summary by CodeRabbit

• New Features
  • Experiments can now continue executing downstream tasks even when upstream tasks fail or timeout, improving workflow resilience and enabling more robust experiment pipelines.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 26dba2c4-106d-43c4-aa06-b0c2bf60569a

📥 Commits

Reviewing files that changed from the base of the PR and between 0b42c14 and 4edc5a6.

📒 Files selected for processing (1)

• tools/launcher/core.py

---

📝 Walkthrough

Walkthrough

Added conditional logic in run_jobs function that checks job.allow_to_fail and whether the executor has a dependency_type attribute, then sets executor.dependency_type to "afterany" to enable downstream tasks to proceed independently of predecessor failures.

Changes

Cohort / File(s)	Summary
Task Dependency Configuration tools/launcher/core.py	Added conditional logic to set executor dependency_type to "afterany" when a job allows failure, enabling downstream tasks to continue regardless of predecessor timeout or failure states.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and specifically describes the main change: using afterany dependency type for pipelines with allow_to_fail enabled.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Security Anti-Patterns	✅ Passed	The pull request adds conditional logic to set executor.dependency_type without introducing security anti-patterns like unsafe deserialization, code execution, or dangerous configurations.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch yeyu/afterany-pipeline-fix

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

PR Preview Action v1.8.1
🚀 View preview at https://NVIDIA.github.io/Model-Optimizer/pr-preview/pr-1248/
Built to branch gh-pages at 2026-04-13 17:06 UTC. Preview will be ready when the GitHub Pages deployment is complete.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 76.91%. Comparing base (5ff1d7b) to head (4edc5a6).
⚠️ Report is 2 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1248   +/-   ##
=======================================
  Coverage   76.91%   76.91%           
=======================================
  Files         350      350           
  Lines       40481    40481           
=======================================
  Hits        31137    31137           
  Misses       9344     9344           

Flag	Coverage Δ
unit	55.53% <ø> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.