mteb/SWEbenchLiteRR

SWEbenchLiteRR是一个用于软件问题定位的Massive Text Embedding Benchmark（MTEB）数据集。它包含四个配置：corpus（文本数据）、qrels（相关性评分）、queries（查询语句）和top_ranked（排名靠前的文本ID）。数据集支持多元语言，并采用MIT许可证。SWEbenchLiteRR is a Massive Text Embedding Benchmark (MTEB) dataset for software issue localization. It includes four configurations: corpus (text data), qrels (relevance scores), queries (query statements), and top_ranked (top-ranked text IDs). The dataset is multilingual and is licensed under MIT.