twangodev/devpost-hacks

该数据集是一个从Devpost网站抓取的hackathon项目提交集合，并丰富了每个项目链接的GitHub仓库的README文件。数据集主要用于研究用途，如评估LLM评委、训练项目摘要模型、研究hackathon获胜因素等。数据集包含多个配置，每个配置对应不同的hackathon项目，并提供了详细的数据模式和加载方法。数据集的来源和许可信息也被提及，但没有统一的开源许可。This dataset is a collection of hackathon project submissions scraped from the Devpost website, with supplementary enrichments added to the README files of the GitHub repositories linked by each project. It is primarily intended for research applications, including evaluating LLM-based judges, training project summarization models, and investigating factors associated with hackathon winning outcomes, among other research directions. The dataset offers multiple configurations, each corresponding to different sets of hackathon projects, and provides detailed data schemas and loading procedures. The source and licensing information of the dataset are also included in the documentation, yet no unified open-source license is applicable to the entire dataset.