---
license: apache-2.0
task_categories:
- text-ranking
---

# TARQ Dataset

This dataset is associated with the paper [Equip Pre-ranking with Target Attention by Residual Quantization](https://huggingface.co/papers/2509.16931).

TARQ is a novel pre-ranking framework designed for industrial recommendation systems. It bridges the gap between efficiency and effectiveness by bringing the modeling power of Target Attention (TA) into the latency-critical pre-ranking stage through Residual Quantization.

## Resources

- **Paper:** [Equip Pre-ranking with Target Attention by Residual Quantization](https://huggingface.co/papers/2509.16931)
- **Code:** [https://github.com/zyody/tarq_sigir2026](https://github.com/zyody/tarq_sigir2026)