add repetition_penalty for openai chat api #932
Conversation
![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 @shihaobai, 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 enhances the OpenAI chat completion API integration by introducing a new parameter. It allows users to control the repetition penalty during text generation, providing more fine-grained control over the output and improving the diversity of generated text.
Highlights
- API Enhancement: Added support for the
repetition_penaltyparameter in the OpenAI chat completion API endpoint.
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.
Code Review
This pull request introduces the repetition_penalty parameter to the OpenAI chat API. A potential issue exists where sending null for repetition_penalty can lead to a TypeError during parameter validation. A code suggestion has been provided to coalesce None to the default value (1.0).
| "do_sample": True, | ||
| "presence_penalty": request.presence_penalty, | ||
| "frequency_penalty": request.frequency_penalty, | ||
| "repetition_penalty": request.repetition_penalty, |
There was a problem hiding this comment.
If request.repetition_penalty is None, it can cause a TypeError during parameter validation in SamplingParams. Coalesce None to the default value (1.0).
"repetition_penalty": request.repetition_penalty if request.repetition_penalty is not None else 1.0This reverts commit 4ca8b78.
No description provided.