---
license: apache-2.0
tags:
- pruned
- math
- optimized
- wanda
- activation-pruning
base_model: Qwen/Qwen2.5-3B-Instruct
pipeline_tag: text-generation
---

# Qwen2.5-3B-Instruct-math-medium-light

> 🎯 **MATH-optimized** | 📦 **Medium Light** pruning | ⚡ **15% weights pruned**

This model is a **moderate-lightly pruned** version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct), specialized for **MATH** tasks using activation-aware weight pruning (Wanda-style).

## ✨ Key Features

- **Specialization**: Optimized for Math tasks
- **Pruning Method**: Wanda-style (|W| × |activation|) importance scoring
- **Size Reduction**: 15% weights pruned
- **Use Case**: Noticeable compression with good performance

## 📊 Performance Comparison

| Category | Original | Pruned | Change |
|----------|----------|--------|--------|
| Python | 100.0% | 100.0% | → |
| Html | 6.7% | 6.7% | → |
| Trivia | 100.0% | 100.0% | → |
| **Math** | 60.0% | 60.0% ⭐ | → |
| Reasoning | N/A | N/A |  |
| Medical | 100.0% | 100.0% | → |
| Linux | 100.0% | 100.0% | → |
| Writing | 73.3% | 80.0% | ↑ 6.7% |

**Average**: 77.1% → 78.1% (+1.0%)

**Math Retention**: 100.0% of original performance

![Comparison Graph](comparison_graph.png)

## 🚀 Quick Start

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("CompactAI/Qwen2.5-3B-Instruct-math-medium-light")
tokenizer = AutoTokenizer.from_pretrained("CompactAI/Qwen2.5-3B-Instruct-math-medium-light")

# Example usage
inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

## 📋 Technical Details

| Property | Value |
|----------|-------|
| Base Model | [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) |
| Specialization | Math |
| Prune Mode | Medium Light |
| Pruning Method | Activation-based weight pruning (Wanda) |
| Weight Reduction | 15% weights pruned |

## 🔗 Related Models

This model is part of the **Qwen2.5-3B-Instruct** pruned model collection. Other variants:
- Extra-light (minimal pruning)
- Light
- Medium-light  
- Medium
- Medium-heavy
- Heavy
- Extra-heavy (maximum compression)

## 📜 License

This model inherits the license from the base model [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct).

---
*Generated by ZANNPS [Zeto Automatic Neural Network Pruning System]*