README.md

---
license: mit
datasets:
- zwhe99/DeepMath-103K
language:
- en
metrics:
- accuracy
base_model:
- zwhe99/Qwen2.5-7B-orz
tags:
- math
- reasoning
- rl
- qwen
- qwen2
model-index:
- name: DeepMath-Zero-7B
  results:
  - task:
      type: text-generation
    dataset:
      name: MATH500
      type: MATH500
    metrics:
    - type: pass@1
      value: 0.855
      name: pass@1
      verified: false
  - task:
      type: text-generation
    dataset:
      name: AMC23
      type: AMC23
    metrics:
    - type: pass@1
      value: 0.647
      name: pass@1
      verified: false
  - task:
      type: text-generation
    dataset:
      name: OlympiadBench
      type: OlympiadBench
    metrics:
    - type: pass@1
      value: 0.51
      name: pass@1
      verified: false
  - task:
      type: text-generation
    dataset:
      name: MinervaMath
      type: MinervaMath
    metrics:
    - type: pass@1
      value: 0.453
      name: pass@1
      verified: false
  - task:
      type: text-generation
    dataset:
      name: AIME24
      type: AIME24
    metrics:
    - type: pass@1
      value: 0.204
      name: pass@1
      verified: false
    - type: pass@1
      value: 0.175
      name: pass@1
      verified: false
---

# DeepMath-Zero-7B

<table>
  <tr>
    <td style="padding: 0;">
      <a href="https://huggingface.co/datasets/zwhe99/DeepMath-103K">
        <img src="https://img.shields.io/badge/Data-4d5eff?style=for-the-badge&logo=huggingface&logoColor=ffffff&labelColor" alt="Data">
      </a>
    </td>
    <td style="padding: 0;">
      <a href="https://huggingface.co/collections/zwhe99/deepmath-6816e139b7f467f21a459a9a">
        <img src="https://img.shields.io/badge/Model-4d5eff?style=for-the-badge&logo=huggingface&logoColor=ffffff&labelColor" alt="Data">
      </a>
    </td>
    <td style="padding: 0;">
      <a href="https://github.com/zwhe99/DeepMath">
        <img src="https://img.shields.io/badge/Code-000000?style=for-the-badge&logo=github&logoColor=white" alt="Code">
      </a>
    </td>
    <td style="padding: 0;">
      <a href="https://arxiv.org/abs/2504.11456">
        <img src="https://img.shields.io/badge/arXiv-2504.11456-b31b1b.svg?style=for-the-badge" alt="arXiv">
      </a>
    </td>
  </tr>
</table>

<!-- Provide a quick summary of what the model is/does. -->
DeepMath-Zero-7B is created by finetuning Qwen/Qwen2.5-7B on DeepMath-103K dataset via RL (using ORZ chat template).

## 📖 Overview

**`DeepMath-103K`** is meticulously curated to push the boundaries of mathematical reasoning in language models. Key features include:

**1. Challenging Problems**: DeepMath-103K has a strong focus on difficult mathematical problems (primarily Levels 5-9), significantly raising the complexity bar compared to many existing open datasets.

<div align="center"> <img src="./assets/github-difficulty.png" width="90%"/>

<sub>Difficulty distribution comparison.</sub> </div>

**2. Broad Topical Diversity**: The dataset spans a wide spectrum of mathematical subjects, including Algebra, Calculus, Number Theory, Geometry, Probability, and Discrete Mathematics.

<div align="center"> <img src="./assets/github-domain.png" width="50%"/>

<sub>Hierarchical breakdown of mathematical topics covered in DeepMath-103K.</sub></div>

**4. Rigorous Decontamination**: Built from diverse sources, the dataset underwent meticulous decontamination against common benchmarks using semantic matching. This minimizes test set leakage and promotes fair model evaluation.

<div align="center"> <img src="./assets/github-contamination-case.png" width="80%"/>

<sub>Detected contamination examples. Subtle conceptual overlaps can also be identified.</sub> </div>

**5. Rich Data Format**: Each sample in `DeepMath-103K` is structured with rich information to support various research applications:

<div align="center"> <img src="./assets/github-data-sample.png" width="90%"/>

<sub>A data sample from DeepMath-103K.</sub> </div>

- **Question**: The mathematical problem statement.
- **Final Answer**: A reliably verifiable final answer, enabling robust rule-based reward functions for RL.
- **Difficulty**: A numerical score for difficulty-aware training or analysis.
- **Topic**: Hierarchical classification for topic-specific applications.
- **R1 Solutions**: Three distinct reasoning paths from DeepSeek-R1, valuable for supervised fine-tuning (SFT) or knowledge distillation.

## 📊Main Results

`DeepMath-Zero-7B` and `DeepMath-1.5B` are trained on the `DeepMath-103K` dataset via RL. These models are initialized from `Qwen2.5-7B-Base` and `R1-Distill-Qwen-1.5B`, respectively.


|          Model           | MATH 500 |  AMC23   | Olympiad Bench | Minerva Math |  AIME24  |  AIME25  |
| :----------------------: | :------: | :------: | :------------: | :----------: | :------: | :------: |
|     Qwen2.5-7B-Base      |   54.8   |   35.3   |      27.8      |     16.2     |   7.7    |   5.4    |
|  Open-Reasoner-Zero-7B   |   81.8   |   58.9   |      47.9      |     38.4     |   15.6   |   14.4   |
| Qwen-2.5-7B-SimpleRL-Zoo |   77.0   |   55.8   |      41.0      |     41.2     |   15.6   |   8.7    |
|     [DeepMath-Zero-7B](https://huggingface.co/zwhe99/DeepMath-Zero-7B)     | **85.5** | **64.7** |    **51.0**    |   **45.3**   | **20.4** | **17.5** |

|          Model          | MATH 500 |  AMC23   | Olympiad Bench | Minerva Math |  AIME24  |  AIME25  |
| :---------------------: | :------: | :------: | :------------: | :----------: | :------: | :------: |
|  R1-Distill-Qwen-1.5B   |   84.7   |   72.0   |      53.1      |     36.6     |   29.4   |   24.8   |
| DeepScaleR-1.5B-Preview |   89.4   |   80.3   |      60.9      |     42.2     | **42.3** |   29.6   |
|  Still-3-1.5B-Preview   |   86.6   |   75.8   |      55.7      |     38.7     |   30.8   |   24.6   |
|   [DeepMath-1.5B](https://huggingface.co/zwhe99/DeepMath-1.5B)         | **89.9** | **82.3** |    **61.8**    |   **42.5**   |   37.3   | **30.8** |

## 🙏 Acknowledgements

This work can not be done without the help of the following works:

- **[verl](https://github.com/volcengine/verl)**: A very fast reinforcement learning framework.
- **[Vivacem/MMIQC](https://huggingface.co/datasets/Vivacem/MMIQC)**: A mixture of question-response pairs extracted from Mathematics Stack Exchange pages.
- **[TIGER-Lab/WebInstructSub](https://huggingface.co/datasets/TIGER-Lab/WebInstructSub)**: Instruction data from MathStackExchange and ScienceStackExchange.
- **[AI-MO/NuminaMath-CoT](https://huggingface.co/datasets/AI-MO/NuminaMath-CoT)**: Approximately 860k math problems.

## 📚 Citation
```bibtex
@article{deepmath,
  title={DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and  Verifiable Mathematical Dataset for Advancing Reasoning},
  author={He, Zhiwei and Liang, Tian and Xu, Jiahao and Liu, Qiuzhi and Chen, Xingyu and Wang, Yue and Song, Linfeng and Yu, Dian and Liang, Zhenwen and Wang, Wenxuan and Zhang, Zhuosheng and Wang, Rui and Tu, Zhaopeng and Mi, Haitao and Yu, Dong},
  year={2025},
  eprint={2504.11456},
  archivePrefix={arXiv},
  primaryClass={cs.CL},
  url={https://arxiv.org/abs/2504.11456}, 
}
```