Render model output using Showdown.js v2.1.0 instead of <pre> tags.
RecapAnon
3884647
Remove unused statusContainer id.
RecapAnon
05ff161
Add deferred model loading with 'Load model' button.
RecapAnon
f210955
Revert back to plain Miku.sh prompt.
RecapAnon
89480b6
Update system prompt from character card and remove soft CoT.
RecapAnon
64582fd
Fix indexeddb-export-import import again.
RecapAnon
45ec7ad
Wrap message text in pre tags for proper markdown line break display
RecapAnon
e7d7941
- Updated addMessage function to wrap content in pre elements
- Updated createStreamingMessageContainer for streaming pre tag support
- Added CSS styling for message-content pre elements
- Ensures proper line break preservation in chat messages
Replace emoji avatars with image avatars for chat participants
RecapAnon
589d9e7
- Update addMessage() function to use image elements instead of emojis
- Update createStreamingMessageContainer() function for assistant avatar
- Replace default assistant message emoji with image in index.html
- Assistant avatar: https://files.catbox.moe/cbclyf.png
- User avatar: https://files.catbox.moe/m5qftn.png
Update styles.
RecapAnon
a8dadb9
Add web-based database management UI with file-based import/export
RecapAnon
855c2d3
- Created manage_database.html with Import/Export/Clear database buttons
- Added file-based data storage with data/dummy.json for seed data
- Modified index.js to auto-load database.json if present, fallback to dummy.json
- Removed hardcoded DB_EXPORT_DATA constant
- Export downloads JSON file to user's Downloads folder
- Import uses HTML5 File API for file selection
- Added 'Manage Database' link in main chat header
- Maintains no-backend architecture with browser-only operations
- Faster initialization with pre-calculated embeddings from database.json
Fix browser errors: model loading, favicon, and source map warnings
RecapAnon
6de6dbf
Add browser-based Qwen3-0.6B chatbot with Transformers.js
RecapAnon
939eb15
- Create index.html with modern chat UI
- Implement real-time streaming responses
- Add status indicators for model loading
- Include responsive design for mobile
- Use Transformers.js 3.1.2 from CDN
- Configure Qwen3-0.6B-ONNX with q4f16 quantization
