[Python][Relax] Fix YaRN correction dim calculation

[akaashrp]

Precompute inv_theta_log_scale

[gemini-code-assist]

Summary of Changes

Hello @akaashrp, 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 addresses a potential issue in the YaRN (Yet another RoPE Normalization) scaling mechanism by refactoring how a key scaling factor, inv_theta_log_scale, is handled. By precomputing this value and passing it explicitly, the changes aim to improve the accuracy and consistency of the correction dimension calculation, ensuring that the YaRN implementation functions as intended across various contexts within the Relax framework.

Highlights

• YaRN Scaling Factor Precomputation: Introduced a new helper function _prepare_yarn_rope_scaling to precompute the inv_theta_log_scale factor for YaRN (Yet another RoPE Normalization) rope scaling configurations, ensuring this value is readily available.
• Correction Dimension Calculation Update: Modified the yarn_find_correction_dim formula to directly utilize the precomputed inv_theta_log_scale, replacing the previous calculation of 1.0 / (2 * math.log(theta)) within the function.
• Parameter Propagation: Updated yarn_find_correction_range, rope_freq_yarn, and switch_rope_freq_func to accept and pass the new inv_theta_log_scale parameter, ensuring its consistent use across YaRN-related functions.
• Initialization Integration: Integrated the _prepare_yarn_rope_scaling function into the __init__ methods of FlashInferPagedKVCache and PagedKVCache to ensure that YaRN scaling parameters are correctly prepared at the cache initialization stage.
• Type Hint Enhancement: Adjusted type hints for the theta parameter in YaRN-related functions from float to Union[float, tir.PrimExpr] and added inv_theta_log_scale as an optional parameter with the same type hint to support more flexible expressions.

🧠 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 refactors the YaRN RoPE scaling implementation to pre-compute inv_theta_log_scale. This is a good optimization that avoids repeated calculations. The changes are logically correct. I have a couple of suggestions to further simplify the code by removing an unused parameter that was left over from the refactoring.

[akaashrp]

cc @tqchen

[guan404ming]

LGTM, thanks!

[akaashrp]

Thanks for approving. I don't have permissions to merge, so it would be great if you could merge this!

[MasterJH5574]

@akaashrp Thanks for the fix! I just merged it.

[guan404ming]

Thanks @MasterJH5574 for helping merge!