[dev] [Deepseek-V4] Use the full CSA denominator for fused indexer loss

[hxbai]

• I, the PR author, have personally reviewed every line of this PR.

What does this PR do?

This PR extends #5960 to the fused CSA indexer-loss paths.

The fused implementation previously used FlashMLA's partial lse_indexer as the teacher denominator. That value excludes the sliding-window contribution and attention sink, and it does not represent all causally visible compressed KV tokens required by the dense loss. As a result, fused indexer-loss values and gradients differed from the unfused/native implementation.

This PR:

• Uses FlashMLA's full sparse-attention LSE plus the attention sink for sparse indexer loss.
• Recomputes the dense teacher LSE over all causally visible compressed KV tokens and combines it with the sliding-window and sink mass.
• Supports both SBHD and packed THD inputs, including CP causal offsets.
• Applies the fix to both FusedCSAIndexerSparseAttnFunc and FusedCSAIndexerSparseAttnFromTopkFunc.
• Chunks teacher-LSE recomputation to bound temporary memory usage.
• Updates the native fused reference to use the same full-denominator oracle as the unfused path.
• Adds helper, fused-kernel, SBHD/THD, and CP parity coverage.

Issue tracking

Linked issue: #5776

Related PR: #5960

⚠️ For major changes (either in lines of code or in its impact), please make sure to first share a design doc with the team. If you're unsure what's the best way to do so, contact @NVIDIA/mcore-oncall.

Issue tracking

For PRs from open-source community contributors:

• New features: a linked issue is required. Please open a feature request and reference it here before submitting the PR.
• Small updates (bug fixes, minor improvements): a linked issue is recommended and will accelerate the PR review process.

Linked issue:

Contribution process

Pre-checks

• 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

Feel free to message or comment @NVIDIA/mcore-oncall to help accelerate your merge into main. The less complex your PR is, the faster it will be approved and merged!

All PRs start as draft. If you open a non-draft PR, it will be automatically converted to draft.

Step 1: Mark PR as "Ready for Review"

1. When your PR is ready, click Ready for Review.
2. An oncall reviewer is auto-assigned and expert reviewers are notified based on your changes.
   • Some PRs may jump straight to step 2. This is determined by .github/CODEOWNERS.

⚠️ Only mark as ready once merge-conflicts are resolved and the CI is passing.
Final Review might get declined if these requirements are not fulfilled.

Step 2: Final Review

For PRs that change megatron/core, once all expert reviewers have approved, the Final Review label is applied automatically and final reviewers are assigned.

For PRs outside megatron/core, this step is skipped.

Step 3: Approved

Once all required reviewers have approved, the Approved label is applied automatically.

Merge

Any member of mcore-engineers will be able to merge your PR.

[copy-pr-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.