[Docs epic] Close post-v1.2.0 documentation gaps for 26.06

[lbliii]

Goal

Close every actionable documentation gap introduced after NeMo Curator v1.2.0, and make the next release's docs/release notes complete enough for users to adopt new features and migrate breaking behavior.

Audit boundary and method

• Primary boundary: commits reachable from main but absent from the v1.2.0 tag, reviewed through PR Add benchmark max timeout guard #2116 (merged 2026-06-29 19:34:32 UTC)
• Original inventory: 67 PRs merged after the v1.2.0 publication timestamp
• Supplemental code-freeze audit: 22 additional release-relevant PRs identified by Sara that merged after the 26.04 code freeze but before the release publication timestamp
• Evidence reviewed: PR titles/bodies, changed file paths, git v1.2.0..main, and the current fern/versions/main/pages/ documentation tree
• Decision rule: create focused docs work for user-facing features, public API/configuration changes, behavior changes, and migration hazards; bundle smaller release-note and reference gaps instead of creating noise

The publication timestamp alone was insufficient: it omitted release-relevant commits that missed v1.2.0 despite merging before the release was published. The supplemental disposition is tracked in #2144, #2145, and #2146.

Planned documentation workstreams

• [Docs] Publish 26.06 release notes and migration checklist #2120 — Publish 26.06 release notes and a migration checklist
• [Docs] Add a Nemotron-CLIMB data-curation recipe #2119 — Document the Nemotron-CLIMB curation recipe
• [Docs] Add an audio tagging, quality-metrics, normalization, and LLM PnC guide #2123 — Document audio tagging, quality metrics, normalization, and LLM PnC
• [Docs] Document long-form audio cutting for ALM pretraining #2122 — Document long-form audio cutting for ALM pretraining
• [Docs] Document the Nemotron OCR synthetic-data-generation pipeline #2121 — Document the Nemotron OCR synthetic-data pipeline
• [Docs] Document video caption-quality evaluation with Summarize-then-Align #2125 — Document the video caption-quality evaluation workflow
• [Docs] Replace legacy resumability guidance and document deterministic task identity #2124 — Replace legacy resumability guidance and document deterministic task identity
• [Docs] Complete the experimental translation configuration reference #2126 — Complete the experimental translation configuration reference
• [Docs] Document filetype-aware input-extension defaults in deduplication workflows #2127 — Document filetype-aware deduplication input defaults
• [Docs] Document Nemotron-Parse inference metrics, engine tuning, and Ray fanout #2128 — Document new Nemotron-Parse tuning and metrics
• [Docs] Document stage worker sizing and with_ backend overrides #2129 — Document stage scheduling/configuration overrides across Ray Data and Xenna
• [Docs] Rewrite InferenceServer docs for typed Ray Serve and Dynamo backends #2146 — Rewrite InferenceServer docs for typed Ray Serve and Dynamo backends
• [Docs] Update video caption model matrix and Qwen migration #2145 — Update the video caption model matrix and Qwen migration
• [Docs] Close late code-freeze docs and release-note gaps #2144 — Close late code-freeze docs and release-note gaps

Child issues are linked above for traceability.

Complete PR disposition

Dedicated docs workstreams required (18)

These PRs add or change user-facing workflows/APIs and need more than a release-note bullet:

#1727, #1835, #1863, #1898, #1899, #1980, #2033, #2036, #2038, #2045, #2054, #2058, #2062, #2070, #2072, #2077, #2086, #2110

Release-notes or migration coverage required (8)

These are smaller fixes, platform constraints, or already-partially-documented removals. They belong in the release/migration task rather than separate feature pages:

#2024, #2049, #2053, #2055, #2057, #2104, #2107, #2109

Documentation already landed; verify in release sweep (9)

No standalone follow-up is currently justified, but the release-notes owner should verify navigation, versioning, and consistency:

#1983, #1961, #1991, #2000, #2029, #2030, #2071, #2096, #2114

No public docs change required (32)

Dependency/security lockfile updates, CI/release plumbing, test-only changes, and internal benchmark-harness maintenance:

#1966, #2001, #2002, #2009, #2021, #2022, #2026, #2032, #2035, #2039, #2041, #2043, #2046, #2047, #2048, #2068, #2079, #2081, #2082, #2083, #2084, #2085, #2087, #2088, #2091, #2099, #2100, #2103, #2105, #2108, #2115, #2116

Supplemental code-freeze disposition

Sara's follow-up list added 22 release-relevant PRs that were outside the original publication-time query.

New dedicated or bundled workstreams

• Inference Server and Dynamo ([Docs] Rewrite InferenceServer docs for typed Ray Serve and Dynamo backends #2146): refactor: extract Ray Serve backend boundary - inference (1/5) #1813, feat!: Typed Serve Config + Dyamo config stub - inference (2/5) #1820, feat: Placement Group based subprocess manager - inference (3/5) #1833, feat: (experimental) Dynamo aggregated backend - inference (4/5) #1834, feat: (experimental) Dynamo Disaggregated serving + validators + router polish - inference (5/5) #1861, Bump Ray 2.55.1 / CUDA 12.9.1 / torch 2.10 / vllm 0.18; HAProxy ingress #1895, chore: Update dynamo to 1.1.0 #1957
• Video caption models ([Docs] Update video caption model matrix and Qwen migration #2145): Add qwen3 #1827, Add Nemotron Omni Model for Video caption pipeline #1883
• Semantic KMeans documentation ([Docs] Close late code-freeze docs and release-note gaps #2144): Add fit_data_fraction and cache_path to KMeansStage #1927

Added to existing workstreams

• Nemotron-CLIMB ([Docs] Add a Nemotron-CLIMB data-curation recipe #2119): [Tutorial] Merge multiple file prefixes generated by MegatronTokenizerWriter #1427
• Audio tagging ([Docs] Add an audio tagging, quality-metrics, normalization, and LLM PnC guide #2123): Add generic audio processing pipeline for ASR and TTS data preparation #1679
• Translation ([Docs] Complete the experimental translation configuration reference #2126): Add translation #1844
• Python migration/release notes ([Docs] Publish 26.06 release notes and migration checklist #2120): Bump python requirements to 3.11 min and 3.13 max #1855

Bundled release-note coverage (#2144 and #2120)

#1746, #1762, #1763, #1774, #1870, #1880, #1888, #1890

Epic-level acceptance criteria

• Every PR in the dedicated and release-only buckets is covered by a child issue or an explicit release-note entry
• fern/versions/main/pages/about/release-notes/index.mdx describes 26.06 rather than 26.04
• Breaking changes and removed/renamed options have migration instructions
• New pages are included in Fern navigation and pass docs validation
• Code samples match the merged APIs and avoid user-supplied task_id
• Links between concept, tutorial, API-reference, and release-note pages are complete
• Each child issue records the PRs it closes documentation coverage for

Out of scope

• Re-documenting internal benchmark infrastructure
• Adding public docs for dependency-only or CI-only changes
• Backporting new 26.06 feature docs into frozen v26.04 pages unless a correction is required