[Doc][Neuron] Updates to reflect Neuron SDK 2.24 #20145
Conversation
|
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @elaineyz, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request primarily focuses on updating the documentation related to AWS Neuron integration within vLLM. The changes aim to provide up-to-date information regarding Neuron SDK versions, installation procedures, and supported features, ensuring users have accurate guidance for deploying vLLM on AWS Neuron hardware.
Highlights
- Documentation Update for AWS Neuron SDK: Updated the
aws_neuron.mddocumentation to reflect the latest AWS Neuron SDK 2.24, clarifying its support on a dedicated AWS Neuron fork of vLLM. - Installation Instructions and Features: Revised installation instructions for Neuron 2.23 libraries in the main vLLM branch and introduced new installation steps for the Neuron 2.24 fork, highlighting its support for prefix caching and disaggregated inference.
- Documentation Structure and Clarity: Reordered navigation and platform listings to prioritize AWS Neuron documentation and removed outdated limitations regarding multi-modal and multi-node support, improving the accuracy of the documentation.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
Signed-off-by: Elaine Zhao <elaineyz@amazon.com>
There was a problem hiding this comment.
Code Review
This pull request updates the Neuron documentation to reflect changes in Neuron SDK 2.24, including distinguishing between features supported in vLLM main (SDK 2.23) and those in the AWS Neuron fork (SDK 2.24). The changes also reorder some navigation and list items for consistency. One potential issue was identified regarding an updated file path in the installation instructions for the AWS Neuron fork, which could lead to installation failures if the corresponding file structure in the external repository is not as implied by the change.
| @@ -14,7 +14,7 @@ This describes how to set up your environment to run vLLM on Neuron. | |||
| - Python: 3.9 or newer | |||
| - Pytorch 2.5/2.6 | |||
| - Accelerator: NeuronCore-v2 (in trn1/inf2 chips) or NeuronCore-v3 (in trn2 chips) | |||
| - AWS Neuron SDK 2.23 | |||
| - AWS Neuron SDK 2.23 (the latest Neuron SDK 2.24 is only supported on the AWS Neuron forked repo, see details below) | |||
There was a problem hiding this comment.
nit/ replace with:
- AWS Neuron SDK 2.23.0 or greater
Note: The latest AWS Neuron SDK 2.24.0 may only be installed via the AWS Neuron Fork. See details below.
You should also hyperlink the "below" in see details below with that section's permalink.
| AWS Neuron maintains a [Github fork of vLLM](https://github.com/aws-neuron/upstreaming-to-vllm/tree/neuron-2.23-vllm-v0.7.2) at | ||
| <https://github.com/aws-neuron/upstreaming-to-vllm/tree/neuron-2.23-vllm-v0.7.2>, which contains several features in addition to what's | ||
| available on vLLM V0. Please utilize the AWS Fork for the following features: | ||
| The Neuron features that are upstreamed to vLLM main reflect Neuron SDK 2.23. Please run the following to |
There was a problem hiding this comment.
Improve readability with a more straight forward sentence structure with action/directive.
Example:
"All features up to Neuron 2.23.0 have been upstreamed to vLLM. To install Neuron 2.23.0, please refer to ..."
| <https://github.com/aws-neuron/upstreaming-to-vllm/tree/neuron-2.24-vllm-v0.7.2>. Please utilize the AWS fork for the following features: | ||
|
|
||
| - Prefix caching | ||
| - Disaggregated inference |
There was a problem hiding this comment.
Hyperlink the tutorials for APC and DI. They should be available here: https://awsdocs-neuron.readthedocs-hosted.com/en/latest/libraries/nxd-inference/tutorials/index.html
There was a problem hiding this comment.
DI is not yet published; will update once it's available. Added a link for APC.
Signed-off-by: Elaine Zhao <elaineyz@amazon.com>
Signed-off-by: Elaine Zhao <elaineyz@amazon.com>
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.Purpose
Update Neuron documentation to reflect changes in the latest Neuron 2.24 release
Test Plan
Locally verified the doc updates by running
mkdocs serveTest Result
Changes rendered as expected
(Optional) Documentation Update
Yes, sole purpose of this PR.