Implement simple blue theme for universal visibility
Siraja704
789ec36
- Remove dark/light mode detection complexity
- Use blue color scheme that works in both themes:
- Main heading: Dark blue (#1565c0)
- Section headings: Medium blue (#1976d2)
- Regular text: Light blue (#2196f3)
- Icons and SVG: Light blue (#2196f3)
- Links: Dark blue (#1565c0)
- Add blue borders to sections for consistency
- Ensures text is visible in both light and dark backgrounds
- Simple, clean, and universally readable design
Simplify CSS for clear text visibility
Siraja704
d1daad9
Fix text and icon visibility with proper contrast
Siraja704
88ce90f
Enhance UI with modern dark/light mode CSS
Siraja704
b8cabae
- Add comprehensive CSS variables for dynamic theming
- Implement automatic dark/light mode detection using prefers-color-scheme
- Improve text visibility and contrast in both modes
- Add modern Inter font family and smooth transitions
- Style all components (medical disclaimers, info boxes, buttons, etc.)
- Create responsive grid layout for conditions section
- Add hover effects and better visual hierarchy
- Improve accessibility and readability across all themes
- Update Gradio theme configuration for better integration
Features:
- Automatic system theme detection
- Improved contrast ratios for accessibility
- Modern card-based design with shadows
- Consistent color scheme throughout the interface
- Better mobile responsiveness
Fix model loading compatibility issues
Siraja704
5c6141b
- Switch from JAX to TensorFlow backend for better model compatibility
- Use traditional HuggingFace Hub download approach instead of keras.saving
- Add huggingface_hub dependency for proper model downloading
- Fix 'weights_store' error by using compatible loading method
- Update documentation to reflect TensorFlow backend usage
Resolves: ValueError: Expected a model.weights.h5 or model.weights.npz file
Resolves: UnboundLocalError: local variable 'weights_store' referenced before assignment
Security: Remove hardcoded HF token, use environment variable only
Siraja704
0da07af
- Remove hardcoded token from code for security
- Token now only sourced from HF_TOKEN environment variable
- Maintains proper authentication while keeping token private
- Updated error handling to guide users on token setup
Upgrade to Keras 3.0 with JAX backend and direct HF model loading
Siraja704
c67397a
- Set KERAS_BACKEND to 'jax' for improved performance
- Use keras.saving.load_model('hf://Siraja704/DermaAI') for direct model loading
- Remove dependency on huggingface_hub package
- Add jax[cpu] to requirements for JAX backend support
- Update README with modern technical stack information
- Simplify model loading process with new Keras 3.0 features
Add authentication support for private model access
Siraja704
4835d52
- Add HF_TOKEN support for accessing gated/private model repository
- Update README.md with setup instructions for authentication
- Add detailed error handling for authentication issues
- Provide clear instructions for adding HF_TOKEN to Space secrets
Fix Gradio compatibility and update dependencies
Siraja704
a20be1c
- Remove deprecated @gr.utils.require_minimum_gradio_version decorator
- Update Gradio version to 5.0.0+ for better compatibility
- Update requirements.txt with compatible package versions
- Add .gitignore to exclude virtual environment and cache files
- Fix import compatibility issues with newer Gradio versions
