DevOps Summary Benchmark

[lightislost]

en

Feature Request

Develop and open-source a summary benchmark.

Motivation

Dataset Investigation: Conduct research on summary datasets to understand their composition, quality, and applicable scenarios.

Corpus Collection: Focus on the collection of summary corpora to ensure a sufficient and diverse data source to support the construction of the benchmark.

Benchmark Construction: Based on the preliminary research and corpus collection achievements, build a summary benchmark to ensure it comprehensively assesses the performance of summary technologies.

Open-source Benchmark: Make the constructed summary benchmark open-source, allowing the entire community to benefit from it, and improve the transparency and reliability of summary technologies.

zh

功能请求

构建并开源一个summary benchmark。

动机

数据集调研：进行summary数据集的调研，了解现有的summary数据集的构成、质量以及适用场景。
语料收集：专注于summary语料的收集，以保证有足够的、多样化的数据来源来支撑benchmark的构建。
Benchmark构建：依据前期的调研和语料收集成果，构建summary benchmark，确保它能全面评估summary技术的性能。
开源Benchmark：将构建好的summary benchmark开源，让整个社区都能从中受益，提高summary技术的透明度和可靠性。