Skip to content
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

[pull] main from huggingface:main #10

Merged
merged 2 commits into from
Feb 21, 2025
Merged

[pull] main from huggingface:main #10

merged 2 commits into from
Feb 21, 2025

Conversation

pull[bot]
Copy link

@pull pull bot commented Feb 21, 2025

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.1)

Can you help keep this open source service alive? 💖 Please sponsor : )

Summary by Sourcery

This pull request introduces a new code_format reward function to encourage models to generate code in a specific format. It also updates the installation instructions and evaluation commands in the README, and adds a new test suite for the code_format reward function.

New Features:

  • Introduces a new reward function, code_format, to incentivize the model to generate code responses in a specific format, including <think> and <answer> tags with a language identifier.

Enhancements:

  • Updates the installation instructions in the README to use uv pip install without the --link-mode=copy flag, and adds a tip for Hugging Face cluster users to suppress cache warnings.
  • Updates the evaluation commands in the README to include generation_parameters={max_new_tokens:32768,temperature:0.0} in the MODEL_ARGS to control generation parameters.
  • Adds langdetect as a dependency for LightEval's extended tasks.

Tests:

  • Adds a new test suite, TestCodeFormat, to verify the code_format reward function with various correct and incorrect code formats, including different languages and multiple code blocks.

lewtun and others added 2 commits February 21, 2025 14:52
* Fix lighteval cmd

* Fix typo

* Pin lighteval

* Hacks to the max

* Fix slurm

* Fix

* Pin lighteval

* Pin l

---------

Co-authored-by: [email protected] <[email protected]>
@pull pull bot added the ⤵️ pull label Feb 21, 2025
@pull pull bot merged commit 8322b31 into Stars1233:main Feb 21, 2025
Copy link

gitstream-cm bot commented Feb 21, 2025

🚨 gitStream Monthly Automation Limit Reached 🚨

Your organization has exceeded the number of pull requests allowed for automation with gitStream.
Monthly PRs automated: 1913/250

To continue automating your PR workflows and unlock additional features, please contact LinearB.

Copy link

sourcery-ai bot commented Feb 21, 2025

Reviewer's Guide by Sourcery

This pull request introduces a new code format reward function, updates installation instructions in the README, modifies dependencies in setup.py, and corrects a typo in the Makefile and slurm script. The code format reward function checks if the generated code follows a specific format with think/answer tags and code blocks. The installation instructions are updated to use uv pip install without --link-mode=copy. The dependencies in setup.py are updated, and flash_attn is removed from extras['train']. Finally, gpu_memory_utilisation is corrected to gpu_memory_utilization in the Makefile and slurm script.

Sequence diagram for the code format reward

sequenceDiagram
    participant User
    participant GRPO
    participant Reward Function

    User->>GRPO: Provides completion
    GRPO->>Reward Function: Calls get_code_format_reward(language)
    Reward Function->>Reward Function: Defines code_format_reward(completions, **kwargs)
    Reward Function->>Reward Function: Applies regex pattern to completion content
    Reward Function-->>GRPO: Returns reward (1.0 if match, 0.0 if not)
    GRPO-->>User: Returns reward
Loading

File-Level Changes

Change Details Files
Added a new reward function to check if the generated code follows a specific format.
  • Implemented get_code_format_reward function to check for code format.
  • Added tests for correct and incorrect code formats.
  • Added tests for multiple code blocks and different languages.
  • Added tests for multiline code blocks.
tests/test_rewards.py
src/open_r1/rewards.py
src/open_r1/grpo.py
Updated the installation instructions in the README to use uv pip install without --link-mode=copy and added a tip for Hugging Face cluster users.
  • Removed --link-mode=copy from uv pip install commands.
  • Added a tip for Hugging Face cluster users to set UV_LINK_MODE=copy.
  • Added installation of setuptools and flash-attn.
README.md
Added generation_parameters to MODEL_ARGS in the README examples.
  • Added generation_parameters={max_new_tokens:32768,temperature:0.0} to MODEL_ARGS for AIME 2024.
  • Added generation_parameters={max_new_tokens:32768,temperature:0.6,top_p:0.95} to MODEL_ARGS for LiveCodeBench.
README.md
Updated dependencies in setup.py and removed flash_attn from extras['train'].
  • Updated lighteval to a new commit hash.
  • Updated vllm to version 0.7.2.
  • Added langdetect as a dependency.
  • Removed flash_attn from extras['train'].
setup.py
Replaced gpu_memory_utilisation with gpu_memory_utilization in the Makefile and slurm script.
  • Replaced gpu_memory_utilisation with gpu_memory_utilization in the Makefile.
  • Replaced gpu_memory_utilisation with gpu_memory_utilization in the slurm script.
Makefile
slurm/evaluate.slurm

Tips and commands

Interacting with Sourcery

  • Trigger a new review: Comment @sourcery-ai review on the pull request.
  • Continue discussions: Reply directly to Sourcery's review comments.
  • Generate a GitHub issue from a review comment: Ask Sourcery to create an
    issue from a review comment by replying to it. You can also reply to a
    review comment with @sourcery-ai issue to create an issue from it.
  • Generate a pull request title: Write @sourcery-ai anywhere in the pull
    request title to generate a title at any time. You can also comment
    @sourcery-ai title on the pull request to (re-)generate the title at any time.
  • Generate a pull request summary: Write @sourcery-ai summary anywhere in
    the pull request body to generate a PR summary at any time exactly where you
    want it. You can also comment @sourcery-ai summary on the pull request to
    (re-)generate the summary at any time.
  • Generate reviewer's guide: Comment @sourcery-ai guide on the pull
    request to (re-)generate the reviewer's guide at any time.
  • Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
    pull request to resolve all Sourcery comments. Useful if you've already
    addressed all the comments and don't want to see them anymore.
  • Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
    request to dismiss all existing Sourcery reviews. Especially useful if you
    want to start fresh with a new review - don't forget to comment
    @sourcery-ai review to trigger a new review!
  • Generate a plan of action for an issue: Comment @sourcery-ai plan on
    an issue to generate a plan of action for it.

Customizing Your Experience

Access your dashboard to:

  • Enable or disable review features such as the Sourcery-generated pull request
    summary, the reviewer's guide, and others.
  • Change the review language.
  • Add, remove or edit custom review instructions.
  • Adjust other review settings.

Getting Help

Copy link

coderabbitai bot commented Feb 21, 2025

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.


Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

❤️ Share
🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

  • Review comments: Directly reply to a review comment made by CodeRabbit. Example:
    • I pushed a fix in commit <commit_id>, please review it.
    • Generate unit testing code for this file.
    • Open a follow-up GitHub issue for this discussion.
  • Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
    • @coderabbitai generate unit testing code for this file.
    • @coderabbitai modularize this function.
  • PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
    • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
    • @coderabbitai read src/utils.ts and generate unit testing code.
    • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
    • @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

  • @coderabbitai pause to pause the reviews on a PR.
  • @coderabbitai resume to resume the paused reviews.
  • @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
  • @coderabbitai full review to do a full review from scratch and review all the files again.
  • @coderabbitai summary to regenerate the summary of the PR.
  • @coderabbitai generate docstrings to generate docstrings for this PR. (Beta)
  • @coderabbitai resolve resolve all the CodeRabbit review comments.
  • @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
  • @coderabbitai help to get help.

Other keywords and placeholders

  • Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
  • Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
  • Add @coderabbitai anywhere in the PR title to generate the title automatically.

CodeRabbit Configuration File (.coderabbit.yaml)

  • You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
  • Please see the configuration documentation for more information.
  • If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

  • Visit our Documentation for detailed information on how to use CodeRabbit.
  • Join our Discord Community to get help, request features, and share feedback.
  • Follow us on X/Twitter for updates and announcements.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants