TESTEVAL
收藏资源简介:
TESTEVAL是一个专注于评估大型语言模型在测试用例生成能力的基准数据集,由阿尔伯塔大学等机构创建。该数据集包含210个从LeetCode平台收集的Python程序,分为三个不同的任务:整体覆盖、目标行/分支覆盖和目标路径覆盖。数据集的创建过程涉及从LeetCode收集程序,并根据程序的复杂性进行筛选。TESTEVAL主要用于解决软件测试中自动化测试用例生成的问题,特别是在理解程序逻辑和执行路径方面。
TESTEVAL is a benchmark dataset dedicated to evaluating the test case generation capabilities of large language models (LLMs), developed by the University of Alberta and other research institutions. This dataset contains 210 Python programs collected from the LeetCode platform, and is divided into three distinct tasks: overall coverage, target line/branch coverage, and target path coverage. The construction process of TESTEVAL involves collecting programs from LeetCode and screening them based on their complexity. TESTEVAL is primarily designed to address the challenges of automated test case generation in software testing, especially in scenarios requiring understanding of program logic and execution paths.

- 1TESTEVAL: Benchmarking Large Language Models for Test Case Generation阿尔伯塔大学 · 2024年



