๐Ÿ›ก๏ธ Dex โ€” Your Personal Cybersecurity AI Sidekick

โ€œBuilt for the curious, optimized for the underground.โ€

Dex (Digital Exploit eXpert) is an intelligent cybersecurity-oriented conversational AI built for ethical hackers, CTF warriors, and knowledge-hungry learners. Whether you're crafting payloads, understanding CVEs, or just chatting โ€” Dex makes cyber learning fun, fast & fearless.


๐Ÿง  Model Intelligence

Feature Description
๐Ÿ”Ž Name ghosthets/Dex
๐Ÿง  Type Causal Language Model
๐ŸŽฏ Domain Cybersecurity, CTFs, Ethical Hacking
๐Ÿงช Usage Educational, research, AI assistant for cyber learners
๐Ÿงฐ Tools Text reasoning, payload discussion, exploit explanation

๐Ÿ’ฅ Dex Is Trained To:

  • ๐Ÿ’ฌ Talk like a pro on cybersecurity topics
  • ๐Ÿ” Explain CVEs, XSS, CSRF, SQLi, etc.
  • ๐Ÿงน Assist in CTF logic & recon mindset
  • ๐Ÿ’ก Generate payload hints
  • ๐Ÿ‘พ Help you build your own AI security bots
  • โš™๏ธ Run locally with just CPU or on edge boards

โš™๏ธ Run It in Python (Transformers)

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

tokenizer = AutoTokenizer.from_pretrained("ghosthets/Dex")
model = AutoModelForCausalLM.from_pretrained("ghosthets/Dex")
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model.to(device)

def ask_dex(prompt):
    input_text = f"User: {prompt}\nDex:"
    inputs = tokenizer(input_text, return_tensors="pt").to(device)
    outputs = model.generate(
        inputs.input_ids,
        attention_mask=inputs.attention_mask,
        max_length=256,
        do_sample=True,
        top_k=40,
        top_p=0.92,
        temperature=0.7,
        pad_token_id=tokenizer.eos_token_id
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True).split("Dex:")[-1].strip()

print(ask_dex("How to identify an XSS vulnerability?"))

โšก Edge-Device Ready

๐Ÿ’ป Lightweight. Efficient. Ready-to-Hack.

Dex is engineered to run smoothly on low-power devices like:

  • ๐Ÿ”น Raspberry Pi Zero W
  • ๐Ÿ”น Pi 2W & Pi 4
  • ๐Ÿ”น Tinker Boards / NanoPi / Pocket PCs
  • ๐Ÿ”น Light virtual containers and offline setups

Despite its small size, Dex delivers powerful outputs and will soon be trained on more curated data, making it sharper, smarter, and more secure.


๐Ÿ‘จโ€๐Ÿ’ป Perfect For

  • Ethical Hackers
  • Red Teamers & Pentesters
  • Bug Bounty Hunters
  • Cybersecurity Students
  • AI + Security Researchers
  • Solo Cyber Ninjas

๐Ÿง  Built With โค๏ธ by

Gaurav Chouhan aka ghosthets ๐Ÿ”— GitHub: ghosthets ๐Ÿ“ From India ๐Ÿ‡ฎ๐Ÿ‡ณ

Downloads last month
23
Safetensors
Model size
124M params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Datasets used to train ghosthets/Dex