Qwen3-4B-html-medium

🎯 HTML-optimized | 📦 Medium pruning | ⚡ 35% weights pruned

This model is a moderately pruned version of Qwen/Qwen3-4B, specialized for HTML tasks using activation-aware weight pruning (Wanda-style).

✨ Key Features

• Specialization: Optimized for Html tasks
• Pruning Method: Wanda-style (|W| × |activation|) importance scoring
• Size Reduction: 35% weights pruned
• Use Case: Balanced trade-off between size and accuracy

📊 Performance Comparison

Category	Original	Pruned	Change
Python	20.0%	60.0%	↑ 40.0%
Html	0.0%	0.0% ⭐	→
Trivia	100.0%	100.0%	→
Math	100.0%	100.0%	→
Reasoning	N/A	N/A
Medical	100.0%	100.0%	→
Linux	100.0%	100.0%	→
Writing	93.3%	80.0%	↓ 13.3%

Average: 73.3% → 77.1% (+3.8%)

🚀 Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("CompactAI/Qwen3-4B-html-medium")
tokenizer = AutoTokenizer.from_pretrained("CompactAI/Qwen3-4B-html-medium")

# 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/Qwen3-4B
Specialization	Html
Prune Mode	Medium
Pruning Method	Activation-based weight pruning (Wanda)
Weight Reduction	35% weights pruned

🔗 Related Models

This model is part of the Qwen3-4B 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/Qwen3-4B.

---

Generated by ZANNPS [Zeto Automatic Neural Network Pruning System]