README.md

---
license: apache-2.0
datasets:
- hotpotqa/hotpot_qa
base_model:
- Qwen/Qwen2.5-7B-Instruct
---


## Model Card for RAG-R1

### Model Details

* **Model Name:** RAG-R1-sq-7b
* **Version:** 1.0
* **Model Type:** RAG
* **Developers:** Zhiwen Tan, Jiaming Huang, Qintong Wu, Hongxuan Zhang, Chenyi Zhuang, Jinjie Gu

[![Paper](https://img.shields.io/badge/arXiv-2507.02962-b5212f.svg)](https://arxiv.org/abs/2507.02962) [![Code](https://img.shields.io/badge/GitHub-Repository-blue.svg?logo=github)](https://github.com/inclusionAI/AWorld-RL/tree/main/RAG-R1)

### Overview

RAG-R1 is a deepsearch training framework designed to enable LLMs to adaptively leverage internal and external knowledge during the reasoning process. 
We further expand the generation and retrieval processes within the framework from single-query mode to multi-query parallelism, aimed at reducing inference time and enhancing the model's capabilities. 
Extensive experiments on seven question-answering benchmarks demonstrate that our method outperforms the strongest baseline by up to 13.2% and decreases inference time by 11.1%.

### Framework

<img src="RAG-R1.png" style="width:100%;">
<h5 align="center"> Overall framework of RAG-R1.</h5>

### Performance

<img src="RAG-R1-result.png" style="width:100%;">
<h5 align="left">Performance comparisons on QA benchmarks under the EM metric. The best and second
best results are bold and underlined, respectively.</h5>

### Acknowledgements
RAG-R1 is inspired by [Deepseek-R1](https://github.com/deepseek-ai/DeepSeek-R1) with its implementation based on [veRL](https://github.com/volcengine/verl) and [Search-r1](https://github.com/PeterGriffinJin/Search-R1). We deeply appreciate the contributions of these teams to open-source research and development.

### Citation
Please cite our repo if our works are helpful for your research.
```
@article{RAG-R1,
  title={RAG-R1 : Incentivize the Search and Reasoning Capabilities of LLMs through Multi-query Parallelism}, 
  author={Zhiwen Tan and Jiaming Huang and Qintong Wu and Hongxuan Zhang and Chenyi Zhuang and Jinjie Gu},
  journal={arXiv preprint arXiv:2507.02962},
  year={2025}
}
```