Question about ID mapping between QA and Contexts

#2
by RobFirth - opened

Hi There,

Thanks for providing this resource. I was using it to test some configurations on a rag pipeline. I might be missing something obvious but it seems that the passage id in the question-answer split does not align with passage ids in the text-corpus subset.

Any pointers would be massively appreciated! Cheers

RAG Datasets org

For this dataset there is no information about from which passages the answer stems. ID columns are nothing more than ids for the row. Feel free to update the dataset here and drop them.

https://huggingface.co/datasets/rag-datasets/rag-mini-bioasq provides gold passages though.

tillwenke changed discussion status to closed

Sign up or log in to comment