kiyam/ddro-nq-pq

DDRO-Generative-Document-Retrieval

This collection contains four generative retrieval models trained using Direct Document Relevance Optimization (DDRO), a lightweight alternative to reinforcement learning for aligning docid generation with document-level relevance through pairwise ranking.

The models are trained on two benchmark datasets (MS MARCO (MS300K) and Natural Questions (NQ320K)) with two types of document identifiers:

• PQ (Product Quantization): captures deep semantic features for complex queries.
• TU (Title + URL): leverages surface-level lexical signals for keyword-driven retrieval.

📌 Models

Dataset	Docid Type	Model Name	MRR@10	R@10
MS MARCO (MS300K)	PQ	ddro-msmarco-pq	45.76	73.02
MS MARCO (MS300K)	TU	ddro-msmarco-tu	50.07	74.01
📍Natural Questions (NQ320K)	PQ	ddro-nq-pq	55.51	67.31
Natural Questions (NQ320K)	TU	ddro-nq-tu	45.99	55.98

---

🚀 Intended Uses

• Generative document retrieval and ranking
• Open-domain question answering
• Semantic and keyword-oriented search
• Research and benchmarking in Information Retrieval (IR)

🏗️ Model Architecture

• Base: T5-base
• Training: Supervised Fine-tuning (SFT) + Pairwise Ranking (Direct L2R)

---

📖 Citation

If you use these models, please cite:

@inproceedings{anonymous2025ddro,
  title={Lightweight and Direct Document Relevance Optimization for Generative Information Retrieval},
  author={Anonymous},
  booktitle={Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR ’25)},
  year={2025},
}

---

🌟 Highlights

• No reinforcement learning or reward modeling
• Lightweight and efficient optimization
• Public checkpoints for reproducibility

---