---
license: mit
language:
  - en
base_model: Qwen/Qwen3-0.6B
tags:
  - lora
  - merged
  - qwen3
  - chatbot
  - university
pipeline_tag: text-generation
---

# UTN-Qwen3-0.6B-LoRA-merged

Qwen3-0.6B finetuned with LoRA (r=64, alpha=128) on UTN domain data, then merged into a standalone model. Ready for direct inference without PEFT.

## Usage

```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "saeedbenadeeb/UTN-Qwen3-0.6B-LoRA-merged"

tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
    trust_remote_code=True,
)

messages = [
    {"role": "system", "content": "You are a helpful assistant for the University of Technology Nuremberg (UTN)."},
    {"role": "user", "content": "What are the admission requirements for AI & Robotics?"},
]

prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True, enable_thinking=False)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

with torch.no_grad():
    output = model.generate(**inputs, max_new_tokens=512, temperature=0.3, top_p=0.9, do_sample=True)

print(tokenizer.decode(output[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
```

## Training

- **Base**: [Qwen/Qwen3-0.6B](https://huggingface.co/Qwen/Qwen3-0.6B)
- **Method**: LoRA (r=64, alpha=128, dropout=0.05, all linear layers)
- **Data**: 1,289 UTN Q&A pairs, 5 epochs, lr=3e-4
- **Hardware**: NVIDIA A40

## Evaluation (Validation Set, 17 examples)

| Metric | Score |
|--------|-------|
| ROUGE-1 | 0.5924 |
| ROUGE-2 | 0.4967 |
| ROUGE-L | 0.5687 |