---
title: TinyR1-32B-Preview-AWQ
canonical_url: "https://www.modelscope.cn/models/ShelterW/TinyR1-32B-Preview-AWQ"
md_url: "https://www.modelscope.cn/models/ShelterW/TinyR1-32B-Preview-AWQ.md"
repository: ShelterW/TinyR1-32B-Preview-AWQ
last_updated: 2025-02-25
license: apache-2.0
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - qwen2
architectures:
  - Qwen2ForCausalLM
base_model:
  - deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
base_model_relation: quantized
parameters: 5.7B
tensor_type:
  - I32
  - F16
  - BF16
library_name:
  - safetensors
inference_backends:
  - "deploy_task text/emb"
  - "lmdeploy 0.9.1"
  - "lmdeploy_turbomind 0.9.1"
  - "sglang 0.5.2"
  - "vllm 0.9.2"
downloads: 137
stars: 0
---

# TinyR1-32B-Preview-AWQ

> TinyR1-32B-Preview-AWQ - ShelterW 在 ModelScope 开源的模型。Model Name: Tiny-R1-32B-Preview Title: SuperDistillation Achieves Near-R1 Performance with Just 5% of Parameters.

ShelterW/TinyR1-32B-Preview-AWQ 是 ModelScope 魔搭社区上的 5.7B 参数text-generation模型，采用 apache-2.0 许可，基于 deepseek-ai/DeepSeek-R1-Distill-Qwen-32B 构建，可用 deploy_task text/emb、lmdeploy 0.9.1、lmdeploy_turbomind 0.9.1 部署。

- **Repository**: ShelterW/TinyR1-32B-Preview-AWQ
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 5.7B
- **Base model**: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
- **Inference backends**: deploy_task text/emb, lmdeploy 0.9.1, lmdeploy_turbomind 0.9.1, sglang 0.5.2, vllm 0.9.2
- **Downloads**: 137
- **Stars**: 0
- **Last updated**: 2025-02-25

Source: https://www.modelscope.cn/models/ShelterW/TinyR1-32B-Preview-AWQ

---

**Model Name**: Tiny-R1-32B-Preview   
**Title**: SuperDistillation Achieves Near-R1 Performance with Just 5% of Parameters.

# Introduction
We introduce our first-generation reasoning model, Tiny-R1-32B-Preview, which outperforms the 70B model Deepseek-R1-Distill-Llama-70B and nearly matches the full R1 model in math.


## Evaluation 
| Model                           | Math (AIME 2024)        | Coding (LiveCodeBench)  | Science (GPQA-Diamond) |
| ------------------------------- | ------------------- | ----------------------- | ---------------------- |
| Deepseek-R1-Distill-Qwen-32B    | 72.6        	| 57.2           	  | 62.1            |
| Deepseek-R1-Distill-Llama-70B   | 70.0                  |  57.5                   | 65.2                   |
| Deepseek-R1                     | 79.8                | 65.9                    | 71.5               |
| Tiny-R1-32B-Preview (Ours)       | 78.1                | 61.6                    | 65.0      

All scores are reported as pass@1.
For AIME 2024, we sample 16 responses, and for GPQA-Diamond, we sample 4 responses, both using average overall accuracy for stable evaluation.

## Approach
| Model                           | Math (AIME 2024)        | Coding (LiveCodeBench)  | Science (GPQA-Diamond) |
| ------------------------------- | ------------------- | ----------------------- | ---------------------- |
| Math-Model (Ours)                | 73.1                | -                       | -                      |
| Code-Model (Ours)              | -                   | 63.4                    | -                      |
| Science-Model (Ours)             | -                   | -                       | 64.5                   |
| Tiny-R1-32B-Preview (Ours)             | 78.1                | 61.6                    | 65.0 


We applied supervised fine-tuning (SFT) to Deepseek-R1-Distill-Qwen-32B across three target domains—Mathematics, Code, and Science — using the [360-LLaMA-Factory](https://github.com/Qihoo360/360-LLaMA-Factory/) training framework to produce three domain-specific models. We used questions from open-source data as seeds, and used DeepSeek-R1 to generate responses for mathematics, coding, and science tasks separately, creating specialized models for each domain. Building on this, we leveraged the Mergekit tool from the Arcee team to combine multiple models, creating Tiny-R1-32B-Preview, which demonstrates strong overall performance.



## Data

#### 1. Math
58.3k CoT trajectories from [open-r1/OpenR1-Math-220k](https://huggingface.co/datasets/open-r1/OpenR1-Math-220k), default subset

#### 2. Coding
19k CoT trajectories [open-thoughts/OpenThoughts-114k](https://huggingface.co/datasets/open-thoughts/OpenThoughts-114k), coding subset

#### 3. Science
We used R1 to generate 8 CoT trajectories on 7.6k seed examples, and got 60.8k CoT trajectories in total; the seed examples are as follows:  
- 2.7k seed examples from [simplescaling/data_ablation_full59K](https://huggingface.co/datasets/simplescaling/data_ablation_full59K), science and health science subset
- 4.9k seed examples from [open-thoughts/OpenThoughts-114k](https://huggingface.co/datasets/open-thoughts/OpenThoughts-114k), science subset



## Open Source Plan
We will publish a technical report as soon as possible and open-source our training and evaluation code, selected training data, and evaluation logs. Having benefited immensely from the open-source community, we are committed to giving back in every way we can.



## Contributors

*360 Team:*
Lin Sun,
Guangxiang Zhao,
Xiaoqi Jian,
Weihong Lin,
Yongfu Zhu,
Change Jia,
Linglin Zhang,
Jinzhu Wu,
Sai-er Hu,
Xiangzheng Zhang

*PKU Team:*
Yuhan Wu,
Zihan Jiang,
Wenrui Liu,
Junting Zhou,
Bin Cui,
Tong Yang


## Citation
```
@misc{tinyr1proj,
      title={SuperDistillation Achieves Near-R1 Performance with Just 5% of Parameters.}, 
      author={TinyR1 Team},
      year={2025},
      eprint={},
      archivePrefix={},
      primaryClass={},
      url={https://huggingface.co/qihoo360/TinyR1-32B-Preview}, 
}
```