Summary of Changes

Hello @e06084, 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 significant enhancement to the evaluation framework by enabling the collection and summarization of detailed metric scores, particularly beneficial for RAG (Retrieval Augmented Generation) evaluations. This feature provides a more granular understanding of performance across different metrics, offering insights into average, min, max, and standard deviation. The accompanying documentation has been thoroughly revised to reflect these changes and provide clearer guidance, while older, less relevant examples have been removed to streamline the project.

Highlights

• Metric Score Collection: Introduced functionality to collect and store individual metric scores (e.g., for RAG evaluation scenarios) within the SummaryModel during the execution process.
• Comprehensive Score Statistics: Added methods to SummaryModel for calculating detailed statistics for collected metric scores, including average, minimum, maximum, and standard deviation. It also provides overall and per-metric average scores.
• Enhanced RAG Evaluation Documentation: Significantly updated the Chinese documentation for RAG evaluation metrics, clarifying field requirements, calculation methods, scoring standards, configuration options, and best practices for combining metrics.
• Codebase Streamlining: Removed several outdated RAG evaluation example scripts, simplifying the examples/rag directory and improving code maintainability.
• Robust Testing: Added new unit tests for LocalExecutor and a dedicated test file for SummaryModel to ensure the correct collection, calculation, and reporting of metric scores in various scenarios.

spark没有支持

spark没有支持

done