slupart/qrecc

QReCC Topics

This repository hosts the QReCC topics with passage relevance. This dataset complements the QReCC retrieval setup outlined in the Apple ML-QReCC GitHub repository.

Train split has 63501 examples, and test split 16451 examples. Relevant passages are in the field "Truth_passages".

from datasets import load_dataset

def main():
    # 1. Load the dataset
    dataset = load_dataset("slupart/qrecc")

    # 2. Show the available splits
    print("Available splits:", dataset.keys())

    # 3. Display the dataset structure / features
    print("\nDataset Features:\n", dataset["train"].features)

    # 4. Show how many examples per split
    for split in dataset:
        print(f"\nSplit '{split}' has {len(dataset[split])} examples")

    # 5. Print a few sample entries
    for i in range(3):
        example = dataset["train"][i]
        print(f"\nExample {i}:")
        print(" Context:", example.get("Context"))
        print(" Question:", example.get("Question"))
        print(" Human Rewrite:", example.get("Truth_rewrite"))
        print(" Answer:", example.get("Truth_answer"))
        print(" Relevant passage (qrel):", example.get("Truth_passages"))

if __name__ == "__main__":
    main()

The QReCC collection of passages, by url_id can be found here slupart/qrecc-passages

Contact & Citation

If you use this dataset in academic or applied work, you can cite the original QReCC dataset and our work:

• The original QReCC benchmark.
• Our work, DiSCo: LLM Knowledge Distillation for Efficient Sparse Retrieval in Conversational Search

@inproceedings{lupart2025disco,
  title={DiSCo: LLM Knowledge Distillation for Efficient Sparse Retrieval in Conversational Search},
  author={Lupart, Simon and Aliannejadi, Mohammad and Kanoulas, Evangelos},
  booktitle={Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval},
  pages={9--19},
  year={2025}
}