cp: docs: Add news items for FP8 Quantization, MoE optimization, and NeMo-RL V0.3 (1301) into r0.4.0

[chtruong814]

beep boop [🤖]: Hi @snowmanwwg 👋,

we've cherry picked #1301 into  for you! 🚀

Please review and approve this cherry pick by your convenience!

Summary by CodeRabbit

• Documentation
  • Updated the News section with four new entries:
    • 9/30/2025: Accelerated RL on GCP with NeMo RL.
    • 9/27/2025: FP8 Quantization in NeMo RL.
    • 8/15/2025: Optimizing weight transfer in large MoE models by 10x.
    • 7/31/2025: NeMo-RL V0.3—scalable, performant post-training via Megatron-Core.
  • Improves visibility into recent milestones and capabilities.

[coderabbitai]

📝 Walkthrough

Walkthrough

Adds four new bullet entries to the README’s News section dated 9/30/2025, 9/27/2025, 8/15/2025, and 7/31/2025. No other files or public entities are modified.

Changes

Cohort / File(s)	Summary
Docs: README News updates README.md	Appends four News items: Accelerated RL on GCP with NeMo RL (9/30/2025), FP8 Quantization in NeMo RL (9/27/2025), MoE weight transfer optimization (8/15/2025), and NeMo‑RL V0.3 announcement (7/31/2025).

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Possibly related PRs

• docs: Restructure README with backend-specific quick start and setup guides #1091 — Also restructures README, including News/Previous News sections; overlaps with this change area.
• docs: Add news items for FP8 Quantization, MoE optimization, and NeMo-RL V0.3 #1301 — Adds the same FP8, MoE optimization, and NeMo‑RL V0.3 News items to README.

Suggested labels

r0.4.0, CI:docs

Suggested reviewers

• snowmanwwg
• terrykong

Pre-merge checks and finishing touches

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Title Check	⚠️ Warning	The title embeds internal cherry-pick metadata, quoted commit messages, and branch references that obscure the actual change and exceed the recommended brevity; it fails to succinctly convey that the PR simply adds new items to the News section.	Please simplify the title to focus on the core update, for example: “docs: Add news items for FP8 Quantization, MoE optimization, and NeMo-RL V0.3,” removing the “cp:” prefix, PR number, and branch details.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Docstring Coverage	✅ Passed	No functions found in the changes. Docstring coverage check skipped.
Test Results For Major Changes	✅ Passed	Based on the provided summary and PR objectives, this change only updates README.md to add four News entries and does not introduce code changes, features, refactors, or behavior that could affect numerics or performance. There are no alterations to public entities and the only review feedback concerns a Markdown formatting fix. Since the changes are purely documentation and minor in nature, test results or performance benchmarks are not required for this PR to pass this check.

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch cherry-pick-1301-r0.4.0

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

README.md (1)

8-11: Conform nested list indentation to markdownlint (MD007)

Indent nested list items by 2 spaces (currently 4), to satisfy tooling and keep consistency.

-    * Student generates on-policy sequences and aligns logits to a larger teacher via KL, achieving near-larger-model quality at lower cost than RL. See [On-policy Distillation](#on-policy-distillation).
+  * Student generates on-policy sequences and aligns logits to a larger teacher via KL, achieving near-larger-model quality at lower cost than RL. See [On-policy Distillation](#on-policy-distillation).

As flagged by markdownlint (MD007). [Based on static analysis hints]

📜 Review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 9689bca and f90a890.

📒 Files selected for processing (1)

• README.md (1 hunks)

🧰 Additional context used

🪛 markdownlint-cli2 (0.18.1)

README.md

9-9: Unordered list indentation
Expected: 2; Actual: 4

(MD007, ul-indent)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (5)

• GitHub Check: Lint check
• GitHub Check: Lint check
• GitHub Check: Lint check
• GitHub Check: Post submodule check comment / Comment on PR
• GitHub Check: Post automodel integration comment / Comment on PR

🔇 Additional comments (2)

README.md (2)

7-7: LGTM on added News items

The FP8 Quantization, MoE optimization, and NeMo‑RL V0.3 entries look correct and well‑linked.

Also applies to: 10-11

---

6-11: Confirm cherry-pick scope
The PR objective lists three entries (FP8 Quantization 9/27, MoE optimization 8/15, NeMo-RL V0.3 7/31), but README also includes the GCP RL item (9/30). Was this extra entry intended for r0.4.0?