Data Science Code Translation (DSCT) is the process of converting source code from one data science library to target code in another library while maintaining functional equivalence, enabling interoperability across data science ecosystems. While Large Language Models (LLMs) have demonstrated considerable progress in Data Science Code Generation (DSCG), their performance in DSCT remains insufficiently studied. To address this gap, we introduce a… See the full description on the dataset page:
https://huggingface.co/datasets/birdsql/DSCT.
