Hugging Face's logo

tanhuajie2001
/
Reason-RFT-Spatial-Transformation-Qwen2-VL-2B

Safetensors
English
qwen2_vl
Model card Files
xet
Community
Reason-RFT-Spatial-Transformation-Qwen2-VL-2B
File size: 6,377 Bytes 
82bb1f2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
25bf90b
 
 
51064cd
82bb1f2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
 
---
license: apache-2.0
language:
- en
datasets:
- tanhuajie2001/Reason-RFT-CoT-Dataset
metrics:
- accuracy
base_model:
- Qwen/Qwen2-VL-2B-Instruct
---

<div align="center">
<img src="https://github.com/tanhuajie/Reason-RFT/raw/main/assets/logo.png" width="500"/>
</div>

# 🤗 Reason-RFT CoT Dateset
*The model checkpoints in our project "Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning"*.


<p align="center">
    </a>&nbsp&nbsp⭐️ <a href="https://tanhuajie.github.io/ReasonRFT/">Project</a></a>&nbsp&nbsp │ &nbsp&nbsp🌎 <a href="https://github.com/tanhuajie/Reason-RFT">Github</a>&nbsp&nbsp │ &nbsp&nbsp🔥 <a href="https://huggingface.co/datasets/tanhuajie2001/Reason-RFT-CoT-Dataset">Dataset</a>&nbsp&nbsp │ &nbsp&nbsp📑 <a href="https://arxiv.org/abs/2503.20752">ArXiv</a>&nbsp&nbsp │ &nbsp&nbsp💬 <a href="https://github.com/tanhuajie/Reason-RFT/raw/main/assets/wechat.png">WeChat</a>
</p>

<p align="center">
</a>&nbsp&nbsp🤖 <a href="https://github.com/FlagOpen/RoboBrain/">RoboBrain</a>: Aim to Explore ReasonRFT Paradigm to Enhance RoboBrain's Embodied Reasoning Capabilities.
</p>

## ♣️ Model List

| Tasks                  | Reason-RFT-Zero-2B        | Reason-RFT-Zero-7B       | Reason-RFT-2B        | Reason-RFT-7B             |
|------------------------|---------------------------|---------------------|---------------------------|---------------------------|
| Visual Counting        | [🤗VC-GRPO-Zero-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Visual-Counting-Qwen2-VL-2B) | [🤗VC-GRPO-Zero-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Visual-Counting-Qwen2-VL-7B) | [🤗VC-GRPO-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Visual-Counting-Qwen2-VL-2B) | [🤗VC-GRPO-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Visual-Counting-Qwen2-VL-7B) |
| Structure Perception   | [🤗SP-GRPO-Zero-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Structure-Perception-Qwen2-VL-2B) | [🤗SP-GRPO-Zero-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Structure-Perception-Qwen2-VL-7B) | [🤗SP-GRPO-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Structure-Perception-Qwen2-VL-2B) | [🤗SP-GRPO-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Structure-Perception-Qwen2-VL-7B) |
| Spatial Transformation | [🤗ST-GRPO-Zero-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Spatial-Transformation-Qwen2-VL-2B) | [🤗ST-GRPO-Zero-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Zero-Spatial-Transformation-Qwen2-VL-7B) | [🤗ST-GRPO-2B](https://huggingface.co/tanhuajie2001/Reason-RFT-Spatial-Transformation-Qwen2-VL-2B) | [🤗ST-GRPO-7B](https://huggingface.co/tanhuajie2001/Reason-RFT-Spatial-Transformation-Qwen2-VL-7B) | 
| ***Embodied Tasks***   | 🤖 *Stay Turned*   | 🤖 *Stay Turned*   | 🤖 *Stay Turned*   | 🤖 *Stay Turned*  | 


## 🔥 Overview
Visual reasoning abilities play a crucial role in understanding complex multimodal data, advancing both domain-specific applications and artificial general intelligence (AGI).
Existing methods improve VLM reasoning via Chain-of-Thought (CoT) supervised fine-tuning, using meticulously annotated training data to enhance visual reasoning capabilities.
However, this training paradigm may lead to overfitting and cognitive rigidity, restricting the model's ability to transfer visual reasoning skills across domains and limiting its real-world applicability.
To address these limitations, we propose **Reason-RFT**, a novel reinforcement fine-tuning framework that significantly enhances generalization capabilities in visual reasoning tasks.
**Reason-RFT** introduces a two-phase training framework for visual reasoning: (1) Supervised Fine-Tuning (SFT) with curated Chain-of-Thought (CoT) data activates the reasoning potential of Vision-Language Models (VLMs), followed by (2) Group Relative Policy Optimization (GRPO)-based reinforcement learning that generates multiple reasoning-response pairs, significantly enhancing generalization in visual reasoning tasks.
To evaluate **Reason-RFT**'s visual reasoning capabilities, we reconstructed a comprehensive dataset spanning visual counting, structure perception, and spatial transformation, serving as a benchmark to systematically assess visual cognition, geometric understanding, and spatial generalization.
Experimental results demonstrate Reasoning-RFT's three key advantages: **(1) Performance Enhancement**: achieving state-of-the-art results across multiple tasks, outperforming most mainstream open-source and proprietary models; 
**(2) Generalization Superiority**: consistently maintaining robust performance across diverse tasks and domains, outperforming alternative training paradigms; 
**(3) Data Efficiency**: excelling in few-shot learning scenarios while surpassing full-dataset SFT baselines; 
**Reason-RFT** introduces a novel paradigm in visual reasoning, significantly advancing multimodal research.

<div align="center">
<img src="https://github.com/tanhuajie/Reason-RFT/raw/main/assets/overview.png" />
</div>

## 🗞️ News

- **`2025-04-12`**: ⭐️ We released our [Models](https://huggingface.co/tanhuajie2001/Reason-RFT-Spatial-Transformation-Qwen2-VL-2B) to huggingface for [General Visual Reasoning Tasks](#GeneralVisualTasks).
- **`2025-04-04`**: 🤗 We released our [datasets](https://huggingface.co/datasets/tanhuajie2001/Reason-RFT-CoT-Dataset/) to huggingface for [General Visual Reasoning Tasks](#GeneralVisualTasks).
- **`2025-04-02`**: 🔥 We released codes and scripts for training/evaluation on [General Visual Reasoning Tasks](#GeneralVisualTasks).
- **`2025-03-29`**: 🌍 We released the [repository](https://github.com/tanhuajie/Reason-RFT/) and [roadmap](#RoadMap) for **Reason-RFT**.
- **`2025-03-26`**: 📑 We released our initial [ArXiv paper](https://arxiv.org/abs/2503.20752/) of **Reason-RFT**.


## ⭐️ Usage

*Please refer to [Reason-RFT](https://github.com/tanhuajie/Reason-RFT) for more details.*

## 📑 Citation
If you find this project useful, welcome to cite us.
```bib
@article{tan2025reason,
  title={Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning},
  author={Tan, Huajie and Ji, Yuheng and Hao, Xiaoshuai and Lin, Minglan and Wang, Pengwei and Wang, Zhongyuan and Zhang, Shanghang},
  journal={arXiv preprint arXiv:2503.20752},
  year={2025}
}
```