Update filter script

[hamishivi]

I made some updates while running offline filtering, pushing these upstream.

[gemini-code-assist]

Summary of Changes

Hello @hamishivi, 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 significantly improves the filter_existing_dataset_correctness.py script by adding more control over the evaluation process and enriching the output data. It allows for a unified judging mechanism across various samples and provides a way to annotate the original dataset with detailed correctness metrics, which is invaluable for understanding and analyzing model performance.

Highlights

• Judge Override Capability: Introduced a new --judge_override command-line argument, allowing users to specify a single judge or verifier to be used for all samples, overriding the dataset's default judge. This enhances flexibility in evaluating correctness across diverse datasets.
• Dataset Annotation Feature: Added an --annotate_original_dataset argument that, when provided with a path, saves the original dataset with additional fields: total_rollouts, total_correct_rollouts, and passrate, enabling more detailed analysis of sample performance.
• Enhanced Correctness Calculation: The _avg_correctness function was updated to handle different input keys for outputs (either output or outputs), extract query information from messages if available, and now returns both the average score and the total number of rollouts for each sample.

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 several enhancements to the dataset filtering script. Key changes include adding a --judge_override argument for more flexible verification, a --annotate_original_dataset option to enrich the source data with correctness metrics, and improved handling of sample data structures. The script now also passes the full query to verifiers, enabling more context-aware evaluation. Overall, these are valuable updates. I've provided a couple of suggestions to improve data handling robustness and ensure the correctness of the annotated metrics.