PromptSet

PromptSet是由威斯康星大学麦迪逊分校创建的一个专为程序员设计的数据集，包含超过61,000个独特的开发者提示，这些提示用于开源Python程序中。数据集通过从GitHub上抓取符合特定标准的Python代码文件创建，主要关注使用如OpenAI、Anthropic等流行语言模型库的项目。PromptSet旨在通过提供大量实际使用的提示，帮助研究者和开发者理解和优化大型语言模型在实际应用中的表现，特别是在提示管理和静态分析方面。

PromptSet is a dataset developed by the University of Wisconsin-Madison specifically for programmers. It contains over 61,000 unique developer prompts designed for use in open-source Python projects. The dataset is built by scraping Python code files that meet predefined criteria from GitHub, with a primary focus on projects leveraging popular large language model (LLM) libraries such as OpenAI and Anthropic. PromptSet aims to help researchers and developers understand and optimize the performance of large language models (LLMs) in real-world applications, especially in the domains of prompt management and static analysis, by providing a large corpus of real-world, actively used prompts.