v0.2.0 - Production Features

TinyForgeAI v0.2.0 - Production Features

This release adds production-ready features including real-time updates, Docker orchestration, authentication, and comprehensive testing.

New Features

WebSocket Support

• Real-time training progress updates via WebSocket connections
• Multiple channels: jobs, logs, stats
• Automatic reconnection with exponential backoff
• Client-side integration in dashboard UI

Docker Compose

• Full-stack orchestration with docker-compose.yml
• Dashboard API (Dockerfile.dashboard)
• Training worker (Dockerfile.training)
• Nginx reverse proxy with WebSocket support
• HuggingFace cache volume for faster model downloads

Authentication

• API key authentication for programmatic access
• HTTP Basic authentication
• Session token management
• Login/logout/verify endpoints
• Configurable via environment variables

Dashboard API Tests

• 38 comprehensive unit tests
• Coverage for all endpoints (jobs, services, inference, stats, logs)
• Integration tests for complete workflows

Improvements

• Dashboard UI now shows real-time job progress updates
• Progress percentage displayed in jobs table
• Cancel button for pending/running jobs
• Updated API version to 0.2.0
• Enhanced .env.example with auth settings

Quick Start

# Clone and install
git clone https://github.com/foremsoft/TinyForgeAI.git
cd TinyForgeAI
pip install -e ".[all]"

# Run with Docker Compose
docker-compose up --build

# Or run locally
make dashboard  # Start dashboard API
cd dashboard && python -m http.server 3000  # Start UI

Documentation

• Getting Started
• Training Basics
• RAG Tutorial
• Architecture Guide

Contributors

Thanks to all contributors who made this release possible!

---

Full Changelog: v0.1.0...v0.2.0