You can deploy this model and start earning money today!

Rauhan/
Qwen2.5-3B-GRPO-GSM325

Name: Qwen2.5-3B-GRPO-GSM325
Availability: InStock

Views

No views yet

Qwen2.5-3B-GRPO-325

🚀 Qwen2.5-3B-GRPO-325 is a fine-tuned version of Qwen2.5 3B, trained using GRPO (Guided Reinforcement Policy Optimization) and DeepSeek R1’s open-source recipe. This model significantly enhances the base Qwen2.5-3B in mathematical problem-solving, logical reasoning, and structured response generation, pushing it towards GPT-4o1-style advanced reasoning.

🔥 Trained entirely on a Free Google Colab Tesla T4 GPU: Training Notebook

🚀 Try inference using this Google Colab notebook: Inference Notebook

🚀 With more resources and extended training, this model could be pushed even further!

Model Details

Base Model: unsloth/Qwen2.5-3B-instruct-unsloth-bnb-4bit
Fine-tuning Method: GRPO with structured reinforcement
Dataset: 325 curated questions from GSM8K (math reasoning)
Format Adherence: XML-based structured reasoning
Notable Improvements:
- Mathematical accuracy ✔
- Logical consistency ✔
- Structured XML-format responses ✔
- GPT-4o1-like step-by-step reasoning ✔

Usage

Example Input & Output

Input (User Query)

If 2x+5=10. Solve for x.

Output (Model Response)

xml

<reasoning>
To solve for x, we need to isolate x on one side of the equation. This can be done by subtracting 5 from both sides of the equation.
</reasoning>
<answer>
2x + 5 - 5 = 10 - 5,
2x = 5,
2x / 2 = 5 / 2,
x = 2.5
</answer>

Installation & Inference

Hugging Face Transformers

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Rauhan/Qwen2.5-3B-GRPO-325"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Limitations & Future Work

🚧 Limitations:

Limited by small dataset size (325 questions)
Training done on a single Free Google Colab Tesla T4 GPU
Some long-form reasoning may need further fine-tuning

🚀 Future Improvements:

Training on a larger dataset (more GSM8K questions + other logical reasoning datasets)
Extending fine-tuning using DeepSeek R1’s full training pipeline
Further quantization for faster and memory-efficient inference

License & Citation

This model is released under Apache 2.0 License. If you use this model in your research, please cite:

@misc{Qwen2.5-3B-GRPO-325,
  author = {Rauhan},
  title = {Qwen2.5-3B-GRPO-325},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/Rauhan/Qwen2.5-3B-GRPO-325}
}

🚀 This model demonstrates how even small models can achieve great results with the right fine-tuning techniques! 🚀

About the Author

🔗 Portfolio & Contact Information:

🌍 Website: rauhanahmed.org
🏢 GitHub: github.com/rauhanAhmed
💼 LinkedIn: linkedin.com/in/rauhan-ahmed
🐦 Twitter (X): x.com/ahmed_rauh46040
📧 Email: rauhaan.siddiqui@gmail.com

Feel free to reach out for collaborations, AI research, or any inquiries! 🚀

Rauhan/Qwen2.5-3B-GRPO-GSM325