---

license: apache-2.0 tags: - text-generation - cybersecurity - chatbot - transformers - edge-device-friendly - dex-ai datasets: - suryanshp1/kali-linux-pentesting-data - AlicanKiraz0/All-CVE-Records-Training-Dataset language: - en library_name: transformers pipeline_tag: text-generation

🛡️ Dex — Cybersecurity Chatbot AI (Made by Dex-Community)

Dex (Digital Exploit eXpert) is an AI model tailored for the cybersecurity and hacking community. It acts as a friendly chatbot that helps with:

• Cybersecurity topics
• Capture The Flag (CTF) guidance
• Basic exploit reasoning
• Code & payload understanding
• Custom AI assistant logic

---

📊 Model Information

Key Details	Value
Model Name	dexcommunity/dex
Base Architecture	Causal Language Model
Framework	🤗 Transformers
Optimized For	Edge devices (Raspberry Pi Zero etc)
Author	Gaurav Chouhan aka ghosthets
License	Apache-2.0

---

⚙️ Usage (Python Example)

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

tokenizer = AutoTokenizer.from_pretrained("dexcommunity/dex")
model = AutoModelForCausalLM.from_pretrained("dexcommunity/dex")

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

def ask_dex(prompt):
    inputs = tokenizer(f"User: {prompt}\nDex:", return_tensors="pt").to(device)
    output = model.generate(inputs.input_ids, max_length=256, pad_token_id=tokenizer.eos_token_id)
    return tokenizer.decode(output[0], skip_special_tokens=True).split("Dex:")[-1].strip()

print(ask_dex("Explain SQL Injection in simple words"))

✅ Features

✅ Lightweight & optimized for edge devices (e.g. Raspberry Pi Zero W/2W)
✅ Helpful in CTF, Bug Bounty, and Penetration Testing topics
✅ Trainable on your own dataset via LoRA or PEFT methods
✅ Fine-tuned on initial cyber security corpora

🔮 Future Plans

🧠 Add OWASP Top 10 understanding
💡 Enhance with Exploit DB & CVE logic
🤖 Hugging Face Space + GUI Gradio interface

🧠 Made With 💙 by
👨‍💻 Gaurav Chouhan

Aka ghosthets 🇮🇳
GitHub: ghosthets
LinkedIn: linkedin.com/in/ghosthets

📎 License
Licensed under Apache-2.0. Use freely with attribution.