[v1] add cli sampler

[hiyouga]

What does this PR do?

Fixes # (issue)

Before submitting

• Did you read the contributor guideline?
• Did you write any new necessary tests?

[gemini-code-assist]

Summary of Changes

Hello @hiyouga, 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 introduces a new command-line interface (CLI) sampler, enabling users to interact with models in a chat-like manner or perform batch inference. It involves significant refactoring of core components, including renaming ModelLoader to ModelEngine and introducing a Renderer class to handle chat template application and message parsing. The plugin registration system has been updated for better flexibility, and various internal utilities and type definitions have been refined to support these new features and improve code clarity.

Highlights

• New CLI Sampler: Introduced a new command-line interface (CLI) sampler that allows for interactive chat with models and batch inference capabilities. This includes a SyncSampler to bridge asynchronous generation with synchronous CLI interaction.
• Core Component Refactoring: The ModelLoader class has been refactored and renamed to ModelEngine, centralizing model-related functionalities. A new Renderer class was introduced to handle chat template application and message parsing, supporting different templates via a plugin system.
• Enhanced Plugin System: The plugin registration mechanism has been updated to support registering multiple methods under a single plugin name, offering greater flexibility and organization for extending functionality.
• Accelerator Device Handling Improvements: Changes were made to how accelerator devices are set and retrieved, with set_device renamed to set_device_index and a new get_device_index function, along with updates to DistributedInterface to use current_device.
• Data Handling and Type System Updates: The data engine's index adjustment and sample selection logic were streamlined by converting plugin classes into direct functions. Additionally, the type definitions for Content, Message, SFTSample, DPOSample, ToolCall, and ModelInput were significantly expanded and clarified.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request 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 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. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

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

1. 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. ↩

[gemini-code-assist]

Code Review

This pull request introduces a command-line interface (CLI) for interactive sampling/chat, which is a great addition. The implementation involves significant and thoughtful refactoring across the board, including the model loading, data processing, and plugin systems, which improves modularity and maintainability. I've identified a couple of issues that need attention: a bug in the ChatML prompt rendering for tool calls, a bug in an updated test case due to the refactoring, and two new files that contain only commented-out code and should probably be removed.