[infer] Support infer cache impl

[Jintao-Huang]

No description provided.

[gemini-code-assist]

Summary of Changes

Hello @Jintao-Huang, 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 inference capabilities by introducing explicit support for different cache implementations within the PyTorch inference engine. It provides a new configuration option for users to select their preferred caching strategy, which is then seamlessly integrated into the model's generation process, aiming to improve performance or memory usage during inference.

Highlights

• Cache Implementation Argument: A new cache_impl argument has been introduced in InferArguments to allow users to specify the desired cache implementation for inference.
• PtEngine Integration: The PtEngine class and its from_model_template method have been updated to accept and store the cache_impl argument.
• Dynamic Cache Application: The specified cache_impl is now dynamically passed as cache_implementation to the generate_kwargs during model generation within PtEngine, enabling flexible caching strategies.

🧠 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 support for specifying a cache implementation for the PyTorch inference backend by adding the cache_impl argument. The changes are well-integrated, passing the new argument from InferArguments down to the model's generate method. The implementation is straightforward and correct. I have a couple of suggestions to improve code clarity and maintainability.

[gemini-code-assist]

For better type clarity and consistency, the type hint for cache_impl should be Optional[str] since its default value is None. This aligns with its usage and its definition in other parts of the codebase, such as in PtEngine.

Suggested change

	cache_impl: str = None
	cache_impl: Optional[str] = None

[gemini-code-assist]

This block of code, which prepares generate_kwargs by adding the cache_implementation, is duplicated in the _infer_full method at lines 408-410. To improve maintainability and adhere to the DRY (Don't Repeat Yourself) principle, I recommend extracting this logic into a private helper method within the PtEngine class. This would centralize the logic for preparing these arguments.