Skip to content

ameyac11/LiteNeTX_Frontend

Repository files navigation

🌟 LiteNeTX Frontend

React · Vite · Tailwind CSS · Three.js

React Vite TypeScript Tailwind CSS Three.js


🌐 Live Demo · 🎥 Demo · 📝 Kaggle Writeup · 🔗 DOI


📸 Preview

LiteNeTX Home Page
🏠 Home Page · Watch Demo
LiteNeTX Model Playground
🧪 Model Playground · Watch Demo

The client app for LiteNeTX — a lightweight CNN family built entirely from scratch (no pretrained models or transfer learning). Explore architectures, run live inference, and visualize model performance through an interactive web UI.

Related repository: LiteNeTX Backend — FastAPI inference API that powers this app.


📖 About

LiteNeTX is a research-driven image classification project with three custom models for FashionMNIST, CIFAR-10, and CIFAR-100. This repo is the React frontend deployed at litenetx.in; it connects to the backend for uploads, predictions, and example galleries.

Model Dataset Highlights
LiteNeTX-FMNIST FashionMNIST Compact conv blocks · grayscale 28×28
LiteNeTX-C10 CIFAR-10 Residual stages · 32×32 RGB
LiteNeTX-C100 CIFAR-100 Wide SE-ResNet · 100-class head

Training hardware: NVIDIA Tesla T4 ×2 (30 GB VRAM)


✨ Features

  • 🛠️ Scratch-Built CNNs — Showcases LiteNeTX models developed from scratch — no pretrained weights
  • 🧪 Model Laboratory — Interactive playground with upload & real-time inference
  • 🏗️ Architecture Explorer — Layer-by-layer breakdown of all three LiteNeTX models
  • 🌐 3D Neural Network — Immersive Three.js visualization of network topology
  • 📈 Performance Insights — Accuracy & parameter reduction charts vs. baselines
  • 📊 Model Comparison — Side-by-side specs for FMNIST · CIFAR-10 · CIFAR-100
  • 🌓 Dark / Light Theme — Polished UI with smooth page transitions
  • 📱 Responsive — Desktop & mobile ready
  • 📡 Keep-Alive — Backend warm-up for faster cold-start inference

🛠️ Tech Stack

⚛️ React 18 · TypeScript · Vite
🎨 Tailwind CSS · Radix UI · Framer Motion
🌐 React Three Fiber · Drei · Three.js
🔄 React Query · React Hook Form · Zod
📊 Recharts — performance visualizations
🗺️ React Router · Vercel Analytics

🚀 Quick Start

npm install
npm run dev

Create a .env file and point the app at your local or deployed API:

VITE_API_BASE_URL=http://localhost:8000

🌐 App → http://localhost:8080
⚡ Backend → LiteNeTX_Backend


🌟 Support

If you find this project useful or interesting, please consider giving it a ⭐ on GitHub! Your support helps make the project more visible and encourages further development.


📜 License

License: AGPL-3.0

Licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).
Copyright © 2026 Ameya Sanjay Chopade · See LICENSE for details.

About

LiteNeTX — lightweight React+TypeScript UI to visualize and demo trained CNN models.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages