-
Notifications
You must be signed in to change notification settings - Fork 208
docs: Add doc for nano-v3 #1694
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
Conversation
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
📝 WalkthroughWalkthroughNew documentation file introducing a guide for post-training the Nemotron 3 Nano model using NeMo RL. The guide covers data preparation, repository setup, launch script configuration with resource specifications, and execution instructions. Changes
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~5 minutes Pre-merge checks and finishing touches❌ Failed checks (1 inconclusive)
✅ Passed checks (3 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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. Comment |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actionable comments posted: 1
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
docs/guides/nemotron-3-nano.md
🧰 Additional context used
📓 Path-based instructions (2)
docs/**/*.md
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Update docs/index.md when a new markdown doc is added under docs/**/*.md or a markdown file is renamed, ensuring the document appears in the most appropriate section
Files:
docs/guides/nemotron-3-nano.md
!(**/tests/**|**/test_*.py|**/test_*.sh)
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Add the NVIDIA copyright header to all Python files and shell scripts (excluding tests). The header should include the current year
Files:
docs/guides/nemotron-3-nano.md
⏰ 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). (4)
- GitHub Check: build-container / main
- GitHub Check: Lint check
- GitHub Check: Post automodel integration comment / Comment on PR
- GitHub Check: Post submodule check comment / Comment on PR
🔇 Additional comments (3)
docs/guides/nemotron-3-nano.md (3)
1-71: Verify that docs/index.md has been updated to include this new guide.Per the coding guidelines, when a new markdown documentation file is added under
docs/**/*.md, thedocs/index.mdfile must be updated to ensure the document appears in the most appropriate section. Since this file is not included in the review context, please confirm that it has been updated to referencedocs/guides/nemotron-3-nano.md.🔎 How to verify
Check that
docs/index.mdcontains an entry linking to the new guide. It should appear in a section like "Guides" or similar. For example:- [Nemotron 3 Nano Post-Training Guide](guides/nemotron-3-nano.md)
64-64: Clarify theray.subscript reference.Line 64 references
ray.subwithout a full path or explanation. This file is not created or described in the documentation. Please clarify whetherray.subis:
- A file that users should have in their repository root
- A file provided elsewhere in the NeMo RL repository
- Something that needs to be created by the user
Adding a brief explanation or a link to the relevant documentation would improve clarity.
3-71: Documentation content looks solid.The guide provides clear step-by-step instructions for data preparation, code checkout, launch script configuration, and training execution. The bash commands are syntactically correct, the data splitting logic is sound (using
head -n -1000for training andtail -n 1000for validation), and the SLURM configuration is well-documented with helpful placeholders.
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
docs/index.md
Outdated
| guides/deepseek.md | ||
| model-quirks.md | ||
| guides/async-grpo.md | ||
| guides/nemotron-3-nano.md |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
could we maybe have this at the top of this guides section so it's easy to see. also might be good to add it somewhere even higher up as well so it's easy to spot
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Addressed in 242d9bb
terrykong
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nice. can we also have the front page readme news bullet for nano v3 point to this doc?
docs/guides/nemotron-3-nano.md
Outdated
| # Install Hugging Face CLI | ||
| pip install huggingface-hub | ||
|
|
||
| # Download RL data blend | ||
| hf download nvidia/Nemotron-3-Nano-RL-Training-Blend --repo-type dataset --local-dir=data |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
prob can one-line this (also since we don't tell users about pip in the main flow, so they may not even have that):
uvx --from huggingface-hub huggingface-cli download nvidia/Nemotron-3-Nano-RL-Training-Blend --repo-type dataset --local-dir=data
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Addressed in a040a8f
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
Signed-off-by: Yi-Fu Wu <yifu.wu@gmail.com>
Updated the existing bullet point to point to these instructions. |
What does this PR do ?
Adds instructions for training nano-v3 with nemo rl.
Issues
List issues that this PR closes (syntax):
Usage
# Add a code snippet demonstrating how to use thisBefore your PR is "Ready for review"
Pre checks:
Additional Information
Summary by CodeRabbit
Documentation
✏️ Tip: You can customize this high-level summary in your review settings.