metadata
library_name: transformers
license: mit
metrics:
- accuracy
- bleu
- code_eval
pipeline_tag: text-generation
tags:
- cybersecurity
- web-development
- multilingual
- hindi
- hinglish
- code-generation
- security
- ddos-protection
- sql-injection
- xss-prevention
🔒 Hinglish Cybersecurity & Web Development Expert
Model Overview
HinglishCyberSec is a fine-tuned CodeLlama-7B model specialized in multilingual cybersecurity and full-stack web development. It understands and generates code in Hindi, English, and Hinglish (Hindi+English mix).
🎯 What This Model Does
| Capability | Description |
|---|---|
| 🌐 Multilingual | Responds in Hindi, English, or Hinglish |
| 🛡️ Cybersecurity | SQL injection, XSS, DDoS, JWT, Encryption, CSRF protection |
| 💻 Web Development | HTML, CSS, JavaScript, React, Flask, Express, PHP |
| 🔒 Security Headers | CSP, HSTS, X-Frame-Options implementation |
| 📊 Rate Limiting | DDoS protection, API rate limiting |
| 🔐 Authentication | JWT, bcrypt, session management |
| 📝 Code Explanation | Explains code in simple Hinglish |
Model Details
Basic Information
- Developed by: Harsh Verma
- Model type: Causal Language Model (Fine-tuned CodeLlama-7B)
- Language(s): Hindi, English, Hinglish (Hindi + English mix)
- License: MIT
- Base Model: CodeLlama-7B
- Fine-tuned on: Custom multilingual cybersecurity + web development dataset
Model Sources
- Repository: [Your ]
- Demo: [https://huggingface.co/spaces/Harsh2026verma/Rudra-chatbot]
- Dataset: [Link to your dataset]
Uses
Direct Use
This model can be used for:
# 1. Generate secure code in Hinglish
response = model.generate("SQL injection se bachne ka tarika batao")
# 2. Create web applications
response = model.generate("Ek responsive navbar banao with logo and 3 links")
# 3. Security audit
response = model.generate("Is code mein SQL injection vulnerability hai?")
# 4. DDoS protection
response = model.generate("Flask mein rate limiting kaise lagayein?")