3.8 KiB
3.8 KiB
Changelog
All notable changes to ComfyUI BuimenLabo will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
[1.0.0] - 2024-01-01
Added
- 🚀 Initial release of ComfyUI BuimenLabo unified package
- 🌍 Multi-Language Locale Toggle with 20 language support
- Floating draggable button interface
- Left-click for quick toggle, right-click for full menu
- Native language names with flag icons
- Position and language memory between sessions
- Touch and mobile device support
- 🔍 Gemini Pose Analyzer with 22 language support
- Google Gemini 1.5 Flash API integration
- Detailed pose descriptions for OpenPose enhancement
- Single and multiple person detection
- Left/right limb distinction and gaze direction analysis
- Direct English output for Stable Diffusion prompts
- 🦙 Advanced LLM Text Generation with llama-cpp-python
- Interactive JavaScript UI with real-time generation
- All-In-One node for complete workflow
- Smart VRAM management with auto/manual cleanup
- GPU acceleration support (CUDA, Metal, ROCm)
- Multiple generation parameters and safety controls
- 🎛️ Smart ControlNet Loader
- Automatic SD vs SDXL model detection
- Intelligent model switching without manual selection
- Error prevention for model compatibility issues
- 📦 Unified Package Structure
- Combined 4 previously separate packages
- No naming conflicts or architectural issues
- Consolidated API endpoints and routes
- Unified documentation and installation
Features
- 20 Languages: English, Japanese, Chinese, Korean, Spanish, French, German, Russian, Italian, Portuguese, Dutch, Swedish, Norwegian, Danish, Finnish, Polish, Czech, Hungarian, Turkish, Thai
- API Endpoints: Direct HTTP API access for external integrations
- Memory Management: Advanced VRAM monitoring and cleanup
- Cross-Platform: Windows, Linux, macOS support
- GPU Acceleration: CUDA, Metal, ROCm compatibility
- Interactive UI: JavaScript-powered real-time interfaces
- ComfyUI Manager: Full registration metadata for easy installation
Documentation
- Comprehensive English README with usage examples
- Complete Japanese README (README_JP.md)
- Windows installation script (install.bat)
- Package metadata (pyproject.toml, package.json)
- ComfyUI Manager registration (comfyui_manager.json)
- API documentation and troubleshooting guides
Requirements
- Python 3.8+
- ComfyUI compatible
- Optional: NVIDIA/AMD GPU for acceleration
- API Keys: Google Gemini for pose analysis
Installation Methods
- ComfyUI Manager (recommended)
- Git clone with pip install
- Windows install.bat script
- Manual dependency installation
Blog Integration
- Prominent author blog link: https://note.com/hirodream44
- Documentation references throughout
- Community support and examples
Technical Details
- Node Classes: 7 total nodes across 4 categories
- JavaScript Extensions: 2 UI enhancement scripts
- API Routes: 4 HTTP endpoints for direct access
- Memory Footprint: Optimized for VRAM efficiency
- Error Handling: Comprehensive fallback systems
- Compatibility: No conflicts with existing nodes
[Planned Features]
Version 1.1.0
- Additional language support expansion
- Enhanced pose analysis accuracy
- More LLM model format support
- Advanced ControlNet automation
- Performance optimizations
Version 1.2.0
- Custom language pack system
- Pose analysis caching
- Distributed LLM inference
- ControlNet training integration
- Workflow templates
For detailed usage instructions, visit: https://note.com/hirodream44