GPUStack UI v2.2.0-dev

A modern, production-ready AI chat interface for GPUStack with advanced authentication, performance monitoring, and intelligent file processing.

✨ Features

Core Functionality

• 💬 AI Chat Interface with streaming responses and stop functionality
• 📁 Enhanced File Processing - PDF, DOCX, TXT, images with intelligent parsing
• 🔍 Web Search Integration via Tavily API
• 🎨 Modern UI with dark/light themes and responsive design

Export/Import System

• 📄 Markdown Export - Human-readable format with proper headers
• 📝 Plain Text Export - Universal compatibility
• 🔧 JSON Export - Complete data fidelity for technical users
• 📥 Multi-format Import - Auto-detects and parses different file types

Performance & Monitoring

• ⚡ Response Metrics - Real-time timing and token count tracking
• 🏥 Health Monitoring - System status and resource usage
• 🔄 Async Backend - Improved concurrency with connection pooling
• 🛑 Stream Control - Stop long-running responses

🚀 Quick Start

Option 1: Automated Setup (Recommended)

git clone https://github.com/your-username/gpustack-ui.git
cd gpustack-ui
git checkout v2.0.0
./setup.sh

Option 2: Docker Compose

git clone https://github.com/your-username/gpustack-ui.git
cd gpustack-ui
git checkout v2.0.0
cp .env.example .env
# Edit .env with your GPUStack server URL
docker-compose up -d

📖 Complete Documentation

For detailed deployment instructions, troubleshooting, and advanced configuration:

📋 DEPLOYMENT.md - Complete deployment guide

🛠️ Prerequisites

• Docker & Docker Compose (recommended)
• GPUStack server running and accessible
• Tavily API key (optional, for web search)

⚙️ Configuration

Environment Configuration

Optional .env file configuration (API keys managed via settings system):

# Environment
ENV=development

# JWT Authentication (optional)
JWT_SECRET_KEY=your-secret-key

# Backend Configuration
BACKEND_PORT=8001
WORKERS=3

# Note: GPUStack and Tavily APIs are configured via the backend settings system

Frontend Configuration

Update frontend/public/config.js:

window.CONFIG = {
    BACKEND_URL: 'http://localhost:8001'  // or your server IP
};

🎯 Usage

Chat Interface

• Start conversations with AI models
• Stream responses with real-time metrics
• Stop long responses with the stop button
• Switch between available models

File Processing

• Upload PDF, DOCX, TXT, or image files
• Get intelligent content extraction and metadata
• Ask questions about uploaded documents

Web Search

• Search current information with Tavily integration
• Get AI-summarized search results with sources
• Combine search results with chat conversations

Export/Import

• Export conversations in Markdown, Text, or JSON
• Import previous conversations from any supported format
• Share conversations easily with colleagues

🔗 API Endpoints

Chat & Inference

• POST /api/inference/stream - Streaming chat responses
• POST /api/inference/infer - Direct inference
• GET /api/models - Available models

File & Tools

• POST /api/files/upload - Enhanced file processing
• POST /api/tools/search - AI-powered web search

Monitoring

• GET /health - Health check and system metrics
• GET / - API status

🚀 What's New in v2.0.0

Major Enhancements

• Multi-format Export/Import - Markdown, Text, JSON support
• Enhanced File Processing - Intelligent parsing with metadata
• Performance Metrics - Real-time response timing and token counts
• Stop Chat Functionality - Interrupt long-running responses
• Async Backend - Improved concurrency and connection pooling

UI/UX Improvements

• Dropdown Export Menu - Easy format selection
• Response Metrics Display - Performance monitoring in chat
• Better File Status - Enhanced upload feedback
• Improved Error Handling - Better user feedback

🆙 Upgrading from v1.x

v2.0.0 is backward compatible - just pull the latest code and rebuild:

git pull
git checkout v2.0.0
docker-compose build --no-cache
docker-compose up -d

🤝 Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Make your changes and test thoroughly
4. Submit a pull request to the dev-improvements branch

📞 Support

• Issues: GitHub Issues
• Docs: Check DEPLOYMENT.md for detailed setup instructions
• Logs: Use docker-compose logs for troubleshooting

📄 License

[Add your license information here]

---

GPUStack UI v2.0.0 - Built for enhanced AI interactions with intelligent file processing and flexible conversation management.

Tailwind CSS Workflow

Building CSS for Production

Run:

npm run build:css

This will generate a minified CSS file at frontend/public/styles/output.css.

Live Development (Watch Mode)

Run:

npm run build:css:watch

This will watch for changes in your HTML/JS/CSS and automatically rebuild output.css.

Usage in HTML

Reference the built CSS in your HTML:

<link rel="stylesheet" href="./styles/output.css">

Notes

• All CSS assets are now served locally. No CDN or external dependencies are required for Tailwind.
• Remove any old references to tailwind.min.css, tailwind.prod.css, or similar files.