basilmh25/qwen_Cybersecucity

Qwen3-4B-Instruct-2507 Cybersecurity LoRA

Model Description

This repository contains a LoRA fine-tuned version of Qwen3-4B-Instruct-2507, specialized for cybersecurity-related instruction following and question answering.

The model was trained using LoRA (Low-Rank Adaptation) with the Unsloth framework on the Cybersecurity Dataset Fenrir v2.1.

The goal of this fine-tuning is to improve the model's ability to answer cybersecurity questions, explain security concepts, assist with defensive security tasks, and provide technically accurate responses while maintaining the general conversational capabilities of the base model.

---

Base Model

• Model: unsloth/Qwen3-4B-Instruct-2507
• Architecture: Qwen3
• Parameters: 4 Billion
• License: Apache-2.0

The model builds upon the official Unsloth implementation of Qwen3-4B-Instruct-2507, which is optimized for efficient fine-tuning and inference. :contentReference[oaicite:0]{index=0}

---

Training Dataset

Dataset

• AlicanKiraz0/Cybersecurity-Dataset-Fenrir-v2.1

The dataset contains cybersecurity-focused instruction-response pairs covering topics such as:

• Network Security
• Malware Analysis
• Web Security
• Cryptography
• Incident Response
• Digital Forensics
• Penetration Testing
• Vulnerability Assessment
• Threat Intelligence
• Secure Programming
• General Cybersecurity Knowledge

---

Fine-Tuning Method

• Method: Supervised Fine-Tuning (SFT)
• Adapter: LoRA
• Framework: Unsloth
• Base Model: Qwen3-4B-Instruct-2507

Only LoRA adapter weights were trained while the original model weights remained frozen.

---

Intended Use

This model is intended for:

• Cybersecurity education
• Security awareness
• Technical question answering
• Security documentation assistance
• Capture The Flag (CTF) learning
• Blue Team support
• SOC analyst assistance
• Research

---

Limitations

This model:

• is not a replacement for professional security expertise;
• may generate inaccurate or outdated information;
• should not be relied upon as the sole source for security-critical decisions;
• should always be validated against official documentation and current security guidance.

---

Example

User

Explain the difference between a vulnerability, an exploit, and a payload.

Assistant

A vulnerability is a weakness in software or a system that can be abused.

An exploit is the technique or code used to take advantage of that vulnerability.

A payload is the code executed after successful exploitation, such as opening a reverse shell, installing malware, or escalating privileges.

---

Loading the LoRA Adapter

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

base_model = AutoModelForCausalLM.from_pretrained(
    "unsloth/Qwen3-4B-Instruct-2507"
)

tokenizer = AutoTokenizer.from_pretrained(
    "unsloth/Qwen3-4B-Instruct-2507"
)

model = PeftModel.from_pretrained(
    base_model,
    "YOUR_USERNAME/YOUR_MODEL_NAME"
)

---

Citation

If you use this model in your work, please cite both:

• Qwen3
• Unsloth
• Cybersecurity Dataset Fenrir v2.1

---

Acknowledgements

Thanks to:

• Alibaba Qwen Team
• Unsloth
• AlicanKiraz for the Cybersecurity Dataset Fenrir v2.1

---

Disclaimer

This model is intended for educational, research, and defensive cybersecurity purposes only.

Users are responsible for complying with all applicable laws and regulations. The authors assume no responsibility for misuse of this model.