-
-
Notifications
You must be signed in to change notification settings - Fork 9.9k
[CLI] Improve CLI arg parsing for -O
/--compilation-config
#20156
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
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 🚀 |
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.
Summary of Changes
Hello @ProExpertProg, 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 significantly enhances the command-line interface's argument parsing capabilities, particularly for compilation configuration flags. My changes introduce support for nested key-value pairs, streamline the handling of compilation levels, and add a valuable warning system for duplicate arguments, making the CLI more flexible, user-friendly, and robust.
Highlights
- Enhanced CLI Argument Parsing for
-O
: I've extended theFlexibleArgumentParsing
to support nested key-value pairs for the-O
/--compilation-config
flag, similar to how--hf-overrides
works. This allows for more granular control over compilation settings directly from the command line (e.g.,-O.use_inductor=true
). - Automatic
-O<level>
Conversion: The CLI now automatically converts shorthand compilation level flags like-O1
into-O.level=1
. This ensures that these level settings can be properly merged with other detailed-O.<key>=<value>
flags, providing a more consistent parsing experience. - Duplicate Argument Warning: I've implemented a new feature that detects and warns users when duplicate dictionary keys are provided via CLI arguments (e.g.,
--hf-overrides.key1 val1 --hf-overrides.key1 val2
). This improves user feedback and helps prevent unexpected behavior due to overwritten settings. - Improved
CompilationConfig
Handling: TheCompilationConfig.from_cli
method has been simplified by removing its internal logic for parsing-O<level>
flags, as this is now handled upstream by theFlexibleArgumentParser
. This centralizes the parsing logic and makes theCompilationConfig
class more focused on validation. - Comprehensive Unit Tests: New unit tests have been added to
tests/test_utils.py
to cover the new-O
parsing capabilities, including nested keys, level conversions, and the duplicate argument warning mechanism, ensuring the robustness of the CLI changes.
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. ↩
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.
Code Review
This pull request enhances the command-line argument parsing for compilation configurations (-O
/--compilation-config
). It introduces support for nested keys, shorthand for optimization levels (e.g., -O1
), and warnings for duplicate arguments. The changes are well-tested.
My review identified a potential issue with the handling of non-numeric -O
arguments and a minor inconsistency in a docstring. I've provided suggestions to address these points to improve the robustness and clarity of the code.
19e18b2
to
5ee2e5a
Compare
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.
This makes sense.
One thing that I'm not entirely sure about is the +,-
syntax. At this point, I felt like it might be better to just input as list?
There's no minus, but inputting as a list would require the parser to know that the arg is expecting a list. Right now the parser is agnostic to expected types, so + tells it to use a list. If we input as a JSON list, we need a bunch of quotes again, which is not ideal. |
Will add back the |
…gumentParser Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
d3c8f31
to
c5007fd
Compare
This reverts commit d3c8f31. Signed-off-by: luka <luka@neuralmagic.com>
Signed-off-by: luka <luka@neuralmagic.com>
c5007fd
to
ebcde56
Compare
Signed-off-by: luka <luka@neuralmagic.com>
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.
This looks solid to me, thanks for the work! Looking forward to using it
…roject#20156) Signed-off-by: luka <luka@neuralmagic.com>
…roject#20156) Signed-off-by: luka <luka@neuralmagic.com> Signed-off-by: avigny <47987522+avigny@users.noreply.github.com>
…roject#20156) Signed-off-by: luka <luka@neuralmagic.com>
Essential Elements of an Effective PR Description Checklist
supported_models.md
andexamples
for a new model.Purpose
While #18208 added really neat support for nested keys (
--hf-overrides.key1.key2=val1
instead of--hf-overrides='{"key1":{"key2":"val1"}}'
, they were not supported for-O
. This PR adds that support and makes the following additional improvements:-O1
to-O.level=1
in theFlexibleArgumentParsing
to support merging with other-O.<...>
flagsTest Plan
Added Python unit tests.
Test Result
All passing.
(Optional) Documentation Update
All CLI documentation is in the
vllm --help
message, which comes from docstrings, which I updated.