Skip to content

[codex] publish 26.06 release notes and migration checklist#2143

Open
lbliii wants to merge 7 commits into
NVIDIA-NeMo:mainfrom
lbliii:codex/docs-26-06-release-notes
Open

[codex] publish 26.06 release notes and migration checklist#2143
lbliii wants to merge 7 commits into
NVIDIA-NeMo:mainfrom
lbliii:codex/docs-26-06-release-notes

Conversation

@lbliii

@lbliii lbliii commented Jun 30, 2026

Copy link
Copy Markdown
Contributor

Summary

Publishes the NeMo Curator 26.06 (v1.3.0) release-level documentation rollup tracked by #2120 and the supplemental code-freeze audits in #2118.

Release-boundary update

PR #2063 is now on main, so the branch no longer says resumability is unreleased. The release notes and migration checklist document released Pipeline.run(checkpoint_path=...) behavior, its pre-existing-Ray requirement, local checkpoint storage, resumability-safe stage validation, and NoneTask / FailedTask semantics.

PR #2059 is also included as a new capability: deterministic source sharding across SLURM arrays, shard completion manifests, and incomplete-shard retry discovery.

Supplemental coverage

The update explicitly captures #1427, #1679, #1746, #1762, #1763, #1774, #1844, #1855, #1870, #1880, #1888, #1890, #1895, #1927, #1957, #2059, and #2063. Detailed #1927 guidance is in #2149, native resumability is in #2134, and SLURM arrays are in #2154.

Merge order / documentation dependencies

Merge this PR after the detailed feature workstreams so its canonical internal links resolve to their final pages:

Validation

  • fern check: passed with 0 errors
  • git diff --check: passed
  • all Python examples in the release notes and migration page parse with ast
  • coverage audit: all 17 supplemental/code-freeze PR numbers are linked
  • dependency audit: versions verified against current pyproject.toml, uv.lock, Dockerfile, and HAProxy installer
  • fern docs broken-links: 10 planned sibling-page links above plus the 22 pre-existing repository errors

Closes #2120
Part of #2144
Parent tracker: #2118

@copy-pr-bot

copy-pr-bot Bot commented Jun 30, 2026

Copy link
Copy Markdown

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.

lbliii added 2 commits July 1, 2026 14:00
Signed-off-by: Lawrence Lane <llane@nvidia.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>
Signed-off-by: Lawrence Lane <llane@nvidia.com>
@lbliii lbliii marked this pull request as ready for review July 1, 2026 21:00
@lbliii lbliii requested a review from a team as a code owner July 1, 2026 21:00
@lbliii lbliii requested review from meatybobby and removed request for a team July 1, 2026 21:00
@greptile-apps

greptile-apps Bot commented Jul 1, 2026

Copy link
Copy Markdown
Contributor

Greptile Summary

This PR publishes the NeMo Curator 26.06 (v1.3.0) documentation rollup: it replaces the stale 26.04 release notes page with the 26.06 release notes and adds a new migration checklist page wired into the Fern navigation.

  • fern/versions/main.yml — adds the 26.06 Migration Checklist entry between Overview and Migration FAQ with a matching slug.
  • fern/versions/main/pages/about/release-notes/index.mdx — replaces 26.04 content with the 26.06 feature, improvements, fixes, breaking-change, and dependency-change sections; all cross-page anchor fragments include the numbered prefix and match their headings in the migration file.
  • fern/versions/main/pages/about/release-notes/26-06-migration.mdx — new 10-step checklist (Python upgrade → task-ID ownership → sentinel construction → video preprocessing removal → caption variant rename → typed inference server → ASR worker rename → stage scheduling → architecture-specific dependencies → resumability), with before/after code examples throughout.

Confidence Score: 5/5

Documentation-only change with no executable code; safe to merge once sibling feature PRs listed in the merge-order section are in place.

All three files are Fern MDX documentation. The navigation entry in main.yml is correct and consistent with the slug pattern used elsewhere. Cross-page anchor fragments in index.mdx include the numbered prefix that matches the numbered headings in 26-06-migration.mdx, resolving the concern from the prior review round. Version numbers across files are mutually consistent and match the PR description. No code logic, schema, or runtime behavior is changed.

No files require special attention. The 10 sibling documentation PRs listed in the merge-order section should be merged first so the internal cross-links resolve to their final pages.

Important Files Changed

Filename Overview
fern/versions/main.yml Adds the 26.06 Migration Checklist page to the Release Notes navigation section, correctly placed between Overview and Migration FAQ with a consistent slug pattern.
fern/versions/main/pages/about/release-notes/26-06-migration.mdx New 10-step migration checklist for upgrading to 26.06 (v1.3.0); covers Python upgrade, task-ID ownership, sentinel construction, video preprocessing removal, caption variant rename, typed inference server, ASR worker rename, stage scheduling, architecture-specific dependencies, and resumability setup.
fern/versions/main/pages/about/release-notes/index.mdx Replaces the stale 26.04 release notes with the 26.06 (v1.3.0) rollup; all cross-page anchor fragments include the numbered prefix and match their target headings in the migration file.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[Release Notes Navigation] --> B[Overview index.mdx 26.06 Release Notes]
    A --> C[26.06 Migration Checklist 26-06-migration.mdx]
    A --> D[Migration FAQ]
    B -->|links to| C
    C --> S1[1. Upgrade Python]
    C --> S2[2. Stop supplying task_id]
    C --> S3[3. Instantiate sentinel tasks]
    C --> S4[4. Remove model_does_preprocess]
    C --> S5[5. Replace qwen variant]
    C --> S6[6. Replace InferenceModelConfig]
    C --> S7[7. Rename ASR num_workers]
    C --> S8[8. Normalize stage scheduling]
    C --> S9[9. Architecture-specific deps]
    C --> S10[10. Configure resumability]
    C --> S11[11. Post-upgrade checks]
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[Release Notes Navigation] --> B[Overview index.mdx 26.06 Release Notes]
    A --> C[26.06 Migration Checklist 26-06-migration.mdx]
    A --> D[Migration FAQ]
    B -->|links to| C
    C --> S1[1. Upgrade Python]
    C --> S2[2. Stop supplying task_id]
    C --> S3[3. Instantiate sentinel tasks]
    C --> S4[4. Remove model_does_preprocess]
    C --> S5[5. Replace qwen variant]
    C --> S6[6. Replace InferenceModelConfig]
    C --> S7[7. Rename ASR num_workers]
    C --> S8[8. Normalize stage scheduling]
    C --> S9[9. Architecture-specific deps]
    C --> S10[10. Configure resumability]
    C --> S11[11. Post-upgrade checks]
Loading

Reviews (6): Last reviewed commit: "Merge branch 'main' into codex/docs-26-0..." | Re-trigger Greptile

Signed-off-by: Lawrence Lane <llane@nvidia.com>
@lbliii lbliii force-pushed the codex/docs-26-06-release-notes branch from 303c2a2 to 28712ec Compare July 2, 2026 02:10
Signed-off-by: Lawrence Lane <llane@nvidia.com>
@lbliii

lbliii commented Jul 2, 2026

Copy link
Copy Markdown
Contributor Author

Adversarial release-doc review follow-up is now included in commit 86e286845:

  • Added a What's New entry and breaking-change notice for the expanded Qwen/Nemotron caption model matrix.
  • Added the required migration from the removed qwen caption variant to qwen2.5 or qwen3.
  • Added a What's New entry and breaking-change notice for typed Ray Serve/Dynamo inference-server configuration.
  • Added before/after migration guidance for the removed InferenceModelConfig API.
  • Updated the migration summary, section anchors, and post-upgrade checks.

This makes #2147 and #2148 explicit content dependencies of the release rollup. Please merge the detailed feature PRs first and keep #2143 last so the consolidated 26.06 page can be rebased once and retain all release-note material.

Validation after rebasing on the current remote branch:

  • fern check: 0 errors
  • git diff --check: passed
  • all 15 Python code blocks in the migration checklist parse with ast

Cross-PR integration and publication staging are tracked in #2160.

@github-actions

github-actions Bot commented Jul 6, 2026

Copy link
Copy Markdown
Contributor

@lbliii

lbliii commented Jul 7, 2026

Copy link
Copy Markdown
Contributor Author

@meatybobby friendly ping on this docs PR for the #2118 release-docs workstream — when you have a moment, could you take a look at the 26.06 release notes and migration checklist?

Docs preview: https://nvidia-preview-codex-docs-26-06-release-notes.docs.buildwithfern.com/nemo/curator

Thanks!

@lbliii lbliii added the docs-release-26.06 Documentation work tracked by the 26.06 release epic (#2118) label Jul 13, 2026 — with ChatGPT Codex Connector
Comment on lines +163 to +166
- **Final CUDA and serving stack:** The release container uses CUDA 12.9.1. Core Ray dependencies require `ray[default,data]>=2.55.1`; the inference-server extra adds `ray[serve,llm]>=2.55.1`, `vllm>=0.14.1,<0.19`, and `ai-dynamo==1.1.0` on supported x86_64 non-macOS systems. The 26.06 lock resolves Ray 2.55.1, vLLM 0.18.1, and AI-Dynamo 1.1.0. The container builds HAProxy 2.8.20 for Ray Serve HAProxy ingress ([#1895](https://github.com/NVIDIA-NeMo/Curator/pull/1895), [#1957](https://github.com/NVIDIA-NeMo/Curator/pull/1957)).
- **PyTorch and CUDA wheels:** Dependency resolution pins PyTorch and torchaudio 2.10.0 and torchvision 0.25.0. Linux x86_64 and aarch64 installations resolve the CUDA 12.9 builds (`torch==2.10.0+cu129`, `torchaudio==2.10.0+cu129`, and `torchvision==0.25.0+cu129`).
- **`nemo_text_processing` on Arm:** The audio extra installs `nemo_text_processing` on x86_64 Linux, but no longer installs it on aarch64 or macOS because its `pynini` dependency is not supported there. Audio stages that import it, including inverse text normalization and WER processing, must be omitted or supplied through a separately supported environment on those platforms ([#2049](https://github.com/NVIDIA-NeMo/Curator/pull/2049)).
- **`flash-attn` is no longer a direct video dependency:** The `video_cuda12` extra no longer pins or installs `flash-attn`. Built-in video pipelines use their declared dependencies; custom code that imports FlashAttention must install a compatible build explicitly ([#2107](https://github.com/NVIDIA-NeMo/Curator/pull/2107)). This applies to both x86_64 and aarch64.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looking at the pyproject this seems accurate: https://github.com/NVIDIA-NeMo/Curator/blob/r1.3.0/pyproject.toml

Maybe we should have @ayushdg / @praateekmahajan / @suiyoubi confirm if extra context is needed for any of this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

docs-release-26.06 Documentation work tracked by the 26.06 release epic (#2118)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Docs] Publish 26.06 release notes and migration checklist

2 participants