qwen x86 flex backend

[malay-nagda]

What does this PR do ?

Valid flex dispatcher backend for H100, B200, B300 for Qwen3 30B_a3B and 235B_a22B

• Before refactor- default was deepep
• After refactor- default is None

This PR changes back the default for specific systems to deepep in perf scripts.

Changelog

moe_flex_dispatcher_backend="deepep",

GitHub Actions CI

See the CI sectionin the Contributing doc for how to trigger the CI. A Nvidia developer will need to approve and trigger the CI for external contributors.

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation?
• Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
  • Reviewer: Does the PR have correct import guards for all optional libraries?

If you haven't finished some of the above items you can still open "Draft" PR.

Additional Information

• Related to # (issue)

Summary by CodeRabbit

• New Features
  • Added new dispatcher backend configuration for Qwen3 large language models, enabling enhanced inference optimization and improved model dispatch flexibility.

[coderabbitai]

📝 Walkthrough

Walkthrough

Added moe_flex_dispatcher_backend="deepep" configuration parameter to two Qwen3 model base configs (235B A22B and 30B A3B). These changes only modify configuration values without altering control flow or error handling logic.

Changes

Cohort / File(s)	Summary
Qwen3 Configuration scripts/performance/configs/qwen/qwen3_workload_base_configs.py	Added moe_flex_dispatcher_backend="deepep" field to BASE_QWEN3_235B_A22B_CONFIG and BASE_QWEN3_30B_A3B_CONFIG definitions.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 2 | ❌ 2

❌ Failed checks (2 inconclusive)

Check name	Status	Explanation	Resolution
Title check	❓ Inconclusive	The title 'qwen x86 flex backend' is vague and doesn't clearly convey the specific change: adding moe_flex_dispatcher_backend configuration to Qwen model configs.	Consider a more descriptive title like 'Add moe_flex_dispatcher_backend=deepep to Qwen3 configs' that clearly indicates what configuration is being added and why.
Test Results For Major Changes	❓ Inconclusive	Unable to verify PR changes without access to actual file contents and repository structure. Shell commands provided cannot be executed in this environment.	Please provide the actual file changes, documentation, and test results associated with this PR for proper assessment.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

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

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch malay/dispatcher_qwen_2602

Tip

Issue Planner is now in beta. Read the docs and try it out! Share your feedback on Discord.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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