docs(extraction): H200 support matrix + section anchors and API crosslinks (NVBugs 6261554)

[kheiss-uwzoo]

Summary

Docs-only extraction update (11 files). Combines the #2240 H200 NVL support-matrix fix and #2245 anchor/crosslink hygiene into one reviewable batch. Rebased on current main (no Helm README or runtime changes).

NVBugs 6261554 — H200 NVL Parakeet unsupported

• Document self-hosted Parakeet ASR (nimOperator.audio) as not supported on H200 NVL, in addition to B200 and other Blackwell GPUs, in prerequisites-support-matrix.md
• Update the model hardware table so the H200 NVL column shows Not supported⁴ for Parakeet audio/video rows
• Expand the intro callout and footnote ⁴; core PDF/multimodal extraction on those GPUs remains unchanged

Navigation — section anchors and on-page TOC

• Add stable { #anchor } IDs on heavily cross-linked headings in prerequisites-support-matrix.md, deployment-options.md, workflow-document-ingestion.md, nemo-retriever-api-reference.md, and concepts.md
• Add an On this page TOC to prerequisites-support-matrix.md (anchors aligned with current #image-captioning, not legacy 26.05 labels)

Crosslinks and handoffs

• Place Python API guide links adjacent to Python snippets in embedding.md, nimclient.md, and audio-video.md
• Remove circular docs.nvidia.com/.../overview/ self-links; route prerequisites/deployment handoffs through getting-started-about.md and deployment-options.md (overview.md, deployment-options.md, troubleshoot.md, getting-started-about.md)
• Standardize link CTAs to refer to (for example nemo-retriever-api-reference.md, concepts.md)

Supersedes: #2240, #2245

Out of scope (already on main or separate PRs): Helm chart README edits, OCR v2 defaults, custom-metadata.md consolidation → vdbs.md, starter-kits.md rename

Files changed

File	Change
prerequisites-support-matrix.md	H200 NVL Parakeet unsupported; TOC + section anchors
deployment-options.md	Section anchors; internal getting-started handoffs
concepts.md	Section anchors; API guide crosslink wording
embedding.md, nimclient.md, audio-video.md	API guide crosslinks by Python snippets
nemo-retriever-api-reference.md	PDF pre-split anchor; refer-to CTA
workflow-document-ingestion.md	Section anchors
getting-started-about.md, overview.md, troubleshoot.md	Remove circular overview links

Test plan

• MkDocs build passes
• check-nrl-doc-leakage.ps1 -Base main passes on the branch diff
• Spot-check anchor links on prerequisites-support-matrix.md (#on-this-page, #model-hardware-requirements, #image-captioning)
• Confirm H200 NVL column shows Not supported⁴ for Parakeet audio in the hardware table
• After docs publish: verify latest/extraction/prerequisites-support-matrix/ renders the H200 footnote and on-page TOC

[greptile-apps]

Greptile Summary

Documentation-only update to the NeMo Retriever extraction docs, combining a support-matrix fix (NVBugs 6261554) with anchor/crosslink hygiene across 11 files. No runtime, Helm, or Python code changes are included.

• Marks H200 NVL as Not supported⁴ for self-hosted Parakeet ASR across the callout, table, and footnote in prerequisites-support-matrix.md, consistent with B200/Blackwell coverage; footnote ⁴ and the intro paragraph are expanded accordingly.
• Adds stable { #anchor } IDs to key headings in six files and places "For parameter details, refer to the Python API guide" crosslinks directly above Python code snippets in embedding.md, nimclient.md, and audio-video.md.
• Replaces circular docs.nvidia.com/.../extraction/overview/ self-links in overview.md, deployment-options.md, troubleshoot.md, and getting-started-about.md with local relative links to getting-started-about.md and deployment-options.md.

Confidence Score: 5/5

Documentation-only change with no runtime, Helm, or Python code modifications; safe to merge.

All changes are Markdown edits: adding MkDocs anchor IDs, updating the H200 NVL Parakeet support status consistently across the callout text, table cells, and footnote, and replacing outbound circular links with local relative references. The H200 NVL "Not supported⁴" change is applied in all three places it needs to appear (intro paragraph, table rows, footnote), and the new anchors resolve pre-existing broken in-page links.

No files require special attention.

Important Files Changed

Filename	Overview
docs/docs/extraction/prerequisites-support-matrix.md	Core change: H200 NVL Parakeet rows flipped from 1¹ to Not supported⁴; intro callout and footnote ⁴ updated consistently; three new section anchors added.
docs/docs/extraction/deployment-options.md	Two section anchors added; circular docs.nvidia.com link replaced with local getting-started-about.md reference in two places.
docs/docs/extraction/concepts.md	Five section anchors added to top-level headings; API guide crosslink phrasing standardized to "For parameter details, refer to".
docs/docs/extraction/embedding.md	API guide crosslink moved from page intro and duplicated before each of the three Python code snippet blocks for closer proximity.
docs/docs/extraction/nimclient.md	Three API guide crosslinks added before Custom NIMs, Real-World Examples, and UDFs sections; no structural changes.
docs/docs/extraction/audio-video.md	API guide crosslink inserted before hosted-inference Python snippet; no other changes.
docs/docs/extraction/nemo-retriever-api-reference.md	Stable anchor added to PDF pre-split heading (referenced from concepts.md); "See" changed to "Refer to" per style standardization.
docs/docs/extraction/workflow-document-ingestion.md	Two section anchors added to Ingest and extract and Choose how you call the library headings.
docs/docs/extraction/getting-started-about.md	Circular docs.nvidia.com link replaced with cleaner description; deploy step description tightened.
docs/docs/extraction/overview.md	Circular docs.nvidia.com self-link removed from Related links section.
docs/docs/extraction/troubleshoot.md	Circular docs.nvidia.com link replaced with local getting-started-about.md reference in the Related links footer.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["prerequisites-support-matrix.md\n(H200 NVL: Not supported⁴ for Parakeet ASR)"] -->|"crosslinks via #model-hardware-requirements\n#core-and-advanced-pipeline-features\n#related-topics"| A
    A -->|"audio-video.md anchor"| B["audio-video.md\n(#parakeet-hosted-inference-build-nvidia)"]
    B -->|"API guide crosslink"| C["nemo-retriever-api-reference.md\n(#pdf-pre-splitting-for-parallel-ingest)"]
    D["embedding.md\nnimclient.md"] -->|"API guide crosslinks\n(near Python snippets)"| C
    E["deployment-options.md\noverview.md\ntroubleshoot.md\ngetting-started-about.md"] -->|"circular docs.nvidia.com links removed\nreplaced with local relative links"| F["getting-started-about.md\ndeployment-options.md"]
    G["concepts.md\nworkflow-document-ingestion.md"] -->|"section anchors added"| G

Loading

%%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%%
flowchart TD
    A["prerequisites-support-matrix.md\n(H200 NVL: Not supported⁴ for Parakeet ASR)"] -->|"crosslinks via #model-hardware-requirements\n#core-and-advanced-pipeline-features\n#related-topics"| A
    A -->|"audio-video.md anchor"| B["audio-video.md\n(#parakeet-hosted-inference-build-nvidia)"]
    B -->|"API guide crosslink"| C["nemo-retriever-api-reference.md\n(#pdf-pre-splitting-for-parallel-ingest)"]
    D["embedding.md\nnimclient.md"] -->|"API guide crosslinks\n(near Python snippets)"| C
    E["deployment-options.md\noverview.md\ntroubleshoot.md\ngetting-started-about.md"] -->|"circular docs.nvidia.com links removed\nreplaced with local relative links"| F["getting-started-about.md\ndeployment-options.md"]
    G["concepts.md\nworkflow-document-ingestion.md"] -->|"section anchors added"| G

Loading

_{Reviews (3): Last reviewed commit: "Merge branch 'main' into docs/extraction..." | Re-trigger Greptile}

[randerzander]

one nit, can be followup PR