PyRagix v0.1.0 - Web Interface

PyRagix v0.1.0 - Web Interface Release

This is the first official release of PyRagix, featuring a complete web interface alongside the existing console-based RAG system.

🚀 New Features

Web Interface

• Modern Web UI: Beautiful, responsive dark theme interface accessible at http://localhost:8000/web/
• Real-time Search: Interactive document search with live status indicators
• TypeScript Frontend: ES2024-powered frontend with professional UI/UX
• REST API: Complete FastAPI backend with automatic documentation at /docs
• Configurable Options: Search results count, source display, and debug mode controls

Enhanced RAG Pipeline

• FAISS IVF Indexing: Advanced indexing with configurable clustering for better performance
• GPU Acceleration: Intelligent GPU detection with graceful CPU fallback
• Memory Optimizations: Adaptive memory settings, tiled OCR, and automatic cleanup
• Advanced Configuration: Extensive tuning options via settings.json

📦 What's Included

• Console Interface (query_rag.py): Interactive command-line RAG queries
• Web Interface (web_server.py): Modern web application with REST API
• Document Ingestion (ingest_folder.py): Process PDF, HTML, and images with OCR
• Modular Architecture: Clean, typed Python with proper error handling
• Cross-Platform: Works on Windows, Linux, and macOS

🛠 Installation & Usage

1. Clone the repository
2. Install dependencies: pip install -r requirements.txt
3. Install Ollama and pull a model: ollama pull llama3.1:8b-instruct-q4_0
4. Process documents: python ingest_folder.py ./your_documents
5. Start web interface: python web_server.py or start_web.bat

📋 Requirements

• Python 3.8+
• 16GB RAM recommended (8GB minimum)
• CUDA GPU optional (automatic detection)
• Ollama for answer generation

See the README for detailed setup instructions and configuration options.

---

Full Changelog:
https://github.com/psarno/PyRagix/commits/v0.1.0