Fix HSDP Registering Device Mesh

[tomlifu]

What does this PR do ?

This PR adds a fix that allows HSDP device mesh to be registered in EP submeshes.

:warning: For major changes (either in lines of code or in its impact), please make sure to first share discuss a design-doc with the team.

Contribution process

flowchart LR
    A[Pre-checks] --> B[PR Tests]
    subgraph Code Review/Approval
        C1[Expert Review] --> C2[Final Review]
    end
    B --> C1
    C2 --> D[Merge]

Pre-checks

• [ ] I want this PR in a versioned release and have added the appropriate Milestone (e.g., Core 0.8)
• [ ] I have added relevant unit tests
• [ ] I have added relevant functional tests
• [ ] I have added proper typing to my code Typing guidelines
• [ ] I have added relevant documentation
• [ ] I have run the autoformatter.sh on my PR

Code review

The following process is enforced via the CODEOWNERS file for changes into megatron/core. For changes outside of megatron/core, it is up to the PR author whether or not to tag the Final Reviewer team.

For MRs into `main` branch

(Step 1): Add PR label Expert Review

(Step 2): Collect the expert reviewers reviews

1. Attach the Expert Review label when your PR is ready for review.
2. GitHub auto-assigns expert reviewers based on your changes. They will get notified and pick up your PR soon.

:warning: Only proceed to the next step once all reviewers have approved, merge-conflict are resolved and the CI is passing.
Final Review might get declined if these requirements are not fulfilled.

(Step 3): Final Review

1. Add Final Review label
2. GitHub auto-assigns final reviewers based on your changes. They will get notified and pick up your PR soon.

(Optional Step 4): Cherry-pick into release branch

If this PR also needs to be merged into core_r* release branches, after this PR has been merged, select Cherry-pick to open a new PR into the release branch.

For MRs into `dev` branch

The proposed review process for `dev` branch is under active discussion.

MRs are mergable after one approval by either [email protected] or [email protected].

Merging your PR

Any member of core-adlr and core-nemo will be able to merge your PR.

[copy-pr-bot[bot]]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[cspades]

May need @xuwchen @shjwudp to chime in, is the DP-Outer in the device mesh compatible with expert device mesh, or do we need a DP-Outer dimension in the expert mesh?

[ko3n1g]

/ok to test af8628c

[shjwudp]

May need @xuwchen @shjwudp to chime in, is the DP-Outer in the device mesh compatible with expert device mesh, or do we need a DP-Outer dimension in the expert mesh?

Yes, this PR fix the case where HSDP could not take effect on DP when EDP is enabled. For EDP-HSDP, we still need an EDP-specific HSDP mesh.

[cspades]

May need @xuwchen @shjwudp to chime in, is the DP-Outer in the device mesh compatible with expert device mesh, or do we need a DP-Outer dimension in the expert mesh?

Yes, this PR fix the case where HSDP could not take effect on DP when EDP is enabled. For EDP-HSDP, we still need an EDP-specific HSDP mesh.

Copy that, so this is a WAR or solution for non-EP HSDP, but we don't support EP EDP-HSDP yet. No problem for now.

[yanring]

/ok to test 1c3e5ce

[yanring]

Hey @tomlifu, could you open a mirror PR to main?

[tomlifu]

Hey @tomlifu, could you open a mirror PR to main?

https://github.com/NVIDIA/Megatron-LM/pull/2467