Update title, subtitle, and tags

README.md

@@ -1,190 +1,105 @@
 ---
-title: Z Image Turbo
-emoji: 🚀
+title: Z-Image Turbo + GLM-4.6V / DeepSeek-3.2 Thinking
+emoji: 🎨
 colorFrom: purple
 colorTo: pink
 sdk: gradio
-sdk_version: "4.20.0"
+sdk_version: 5.31.0
 app_file: app.py
-pinned: false
+pinned: true
+license: apache-2.0
+tags:
+  - text-to-image
+  - image-to-image
+  - diffusion
+  - transformer
+  - z-image
+  - z-image-turbo
+  - ai-art
+  - image-generation
+  - image-editing
+  - glm-4
+  - glm-4.6v
+  - glm-4v
+  - vision-language
+  - multimodal
+  - prompt-generation
+  - ai-assistant
+  - zhipu
+  - zai
+  - deepseek
+  - deepseek-reasoner
+  - deepseek-3.2
+  - thinking
+  - reasoning
+  - multilingual
+  - rtl
+  - arabic
+  - hindi
+  - spanish
+  - portuguese
+short_description: Image Gen & Edit with GLM-4.6V + DeepSeek-3.2
 ---
 
-# Z Image Turbo - Production Edition
-
-A production-ready, optimized image generation and transformation application built with Gradio and Diffusers. This application addresses common production challenges including error handling, resource management, and performance optimization.
-
-## 🚀 Features
+# ⚡ Z-Image Generation & Transformation Demo
 
-### Core Functionality
-- **Image Generation**: Create high-quality images from text prompts
-- **Image Transformation**: Modify existing images with AI guidance
-- **Multiple Styles**: Apply artistic styles to generations
-- **Aspect Ratios**: Support for various image dimensions
-
-### Production Optimizations
-- ✅ **PyTorch 2.0+ Compilation** with graceful fallback for older versions
-- ✅ **xformers Memory Optimization** with CPU fallback when not available
-- ✅ **Dynamic Resource Management** with CPU offloading for memory-constrained environments
-- ✅ **Intelligent Caching System** with LRU eviction and TTL
-- ✅ **Comprehensive Error Handling** with user-friendly messages
-- ✅ **Real-time System Monitoring** with performance metrics
-- ✅ **Modular Architecture** with clear separation of concerns
-
-## 📋 System Requirements
-
-### Minimum Requirements
-- Python 3.9+
-- GPU with 6GB+ VRAM (recommended)
-- 8GB+ RAM
-- 10GB+ disk space
-
-### Recommended
-- GPU with 12GB+ VRAM
-- PyTorch 2.0+ for optimal performance
-- xformers installed for memory efficiency
-
-## 🔧 Configuration
-
-### Environment Variables
-```bash
-# Optional: Override default settings
-HF_TOKEN=your_huggingface_token
-CUDA_VISIBLE_DEVICES=0
-PYTORCH_CUDA_ALLOC_CONF=max_split_size_mb:128
-```
-
-### Performance Tuning
-The application automatically optimizes based on available hardware:
-- Detects optimal data type (bfloat16/float16/float32)
-- Enables CPU offloading for limited GPU memory
-- Adjusts batch sizes and concurrency limits
-
-## 📊 Monitoring
-
-### System Metrics
-- Memory usage (CPU and GPU)
-- CPU utilization
-- Generation count and success rate
-- Cache hit rate
-- Uptime and error tracking
-
-### Health Checks
-The application provides comprehensive health monitoring with real-time metrics tracking.
-
-## 🐛 Troubleshooting
-
-### Common Issues
-
-#### 1. "Out of Memory" Error
-**Solution**: The app automatically handles OOM scenarios and suggests:
-- Using smaller image sizes
-- Enabling CPU offloading
-- Restarting the space
-
-#### 2. "Could not compile UNet" Warning
-**Expected Behavior**: This is normal for PyTorch < 2.0. The app continues without compilation.
-
-#### 3. "Could not enable xformers" Warning
-**Expected Behavior**: Falls back to standard attention mechanism.
-
-#### 4. Slow Generation
-**Check**:
-- GPU memory availability
-- Model loading status
-- System resource usage
-
-### Logs
-The application logs to:
-- Console output for real-time monitoring
-- `z_image_turbo.log` file for persistent logs
-
-## 🎯 Best Practices
-
-### For Users
-1. **Start with simple prompts** to test performance
-2. **Use appropriate image sizes** for your hardware
-3. **Leverage caching** by reusing similar prompts
-4. **Monitor system resources** in the System Monitor tab
-
-### For Developers
-1. **Check error codes** for debugging
-2. **Monitor performance metrics** regularly
-3. **Use the health check endpoint** for monitoring
-4. **Review logs** for troubleshooting
-
-## 🔍 Error Codes
-
-| Code | Meaning | Resolution |
-|------|---------|------------|
-| 1001 | Model Load Error | Check model access and permissions |
-| 1002 | Generation Error | Check prompt and resources |
-| 1003 | Transform Error | Verify input image and prompt |
-| 2001 | Invalid Input | Check input parameters |
-| 3001 | Resource Error | Free up memory or use smaller images |
-| 4001 | Network Error | Check internet connection |
-| 5001 | Cache Error | Clear cache or restart |
-| 9999 | Unknown Error | Check logs for details |
-
-## 📈 Performance Optimization
-
-### Automatic Optimizations
-1. **Model Compilation**: PyTorch 2.0+ compilation when available
-2. **Memory Efficiency**: xformers attention and VAE slicing
-3. **Resource Management**: CPU offloading for memory constraints
-4. **Smart Caching**: LRU cache with intelligent eviction
-
-### Manual Tuning
-- Adjust `max_size` in CacheManager for memory/concurrency balance
-- Modify inference steps for quality/speed trade-off
-- Use aspect ratios that match your hardware capabilities
-
-## 🔐 Security
-
-### Best Practices
-- No user data is persisted beyond cache TTL
-- Images are not stored permanently
-- API keys are environment variables only
-- Safe tensor format for model weights
-
-## 📝 Changelog
-
-### v2.0.0 Production
-- **Added**: Comprehensive error handling with error codes
-- **Added**: PyTorch 2.0+ compilation with fallback
-- **Added**: xformers optimization with CPU fallback
-- **Added**: Intelligent caching system with LRU and TTL
-- **Added**: Real-time system monitoring
-- **Added**: Production-ready logging
-- **Added**: Health check endpoints
-- **Improved**: Resource management and cleanup
-- **Improved**: User experience with better error messages
-- **Improved**: Performance with automatic optimizations
-
-### v1.0.0 Initial
-- Basic image generation
-- Simple UI
-- Model integration
-
-## 🤝 Contributing
-
-When contributing:
-1. Follow PEP 8 style guidelines
-2. Add comprehensive error handling
-3. Include logging for debugging
-4. Update documentation
-5. Test with various hardware configurations
-
-## 📄 License
-
-This project is licensed under the MIT License.
-
-## 🙏 Acknowledgments
-
-- [Tongyi-MAI/Z-Image-Turbo](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) for the base model
-- [Diffusers](https://github.com/huggingface/diffusers) library
-- [Gradio](https://github.com/gradio-app/gradio) for the UI framework
-- Hugging Face Spaces for hosting
+Generate stunning AI images and transform existing photos using **Alibaba's Z-Image-Turbo** - a next-generation diffusion transformer model.
+
+## ✨ Features
+
+🎨 **Text-to-Image Generation**
+- 10 style presets (Photorealistic, Cinematic, Anime, Digital Art, and more)
+- 18 aspect ratios from 1024px to 2048px
+- AI-powered prompt polishing with Qwen2.5-72B
+
+🖼️ **Image-to-Image Transformation**
+- Transform photos into different art styles
+- Adjustable strength control (subtle to complete change)
+- Style-focused AI prompt enhancement
+
+⚡ **Optimized Performance**
+- Runs on ZeroGPU (NVIDIA H200)
+- FlashAttention-2 via SDPA
+- ~20-25 seconds per generation
+
+## 🚀 How to Use
+
+### Generate Tab
+1. Enter a description of your image
+2. (Optional) Enable "Polish Prompt" for AI-enhanced prompts
+3. Select a style and aspect ratio
+4. Click "Generate"
+
+### Transform Tab
+1. Upload an image
+2. Describe the transformation you want
+3. Adjust strength (lower = subtle, higher = dramatic)
+4. Click "Transform"
+
+## 🎯 Example Prompts
+
+- *"Ancient dragon perched on a Gothic cathedral at dusk, stormy purple sky"*
+- *"Cyberpunk samurai in a neon-lit rainy alley, glowing armor"*
+- *"Cozy witch cottage interior, cauldrons bubbling, magical atmosphere"*
+
+## 📊 Technical Details
+
+| Spec | Value |
+|------|-------|
+| Model | Tongyi-MAI/Z-Image-Turbo |
+| Steps | 9 (optimized) |
+| Precision | BFloat16 |
+| Max Resolution | 2048x2048 |
+
+## 🔗 Links
+
+- [Z-Image-Turbo Model](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo)
+- [Z-Image GitHub](https://github.com/Tongyi-MAI/Z-Image)
+
+## 📜 License
+
+Model: Apache 2.0 License
 
 ---
 
-**Created with ❤️ by AI Agent Framework Specialist**
+**Created by [@lulavc](https://huggingface.co/lulavc)**