v2.3
Merry Xmas and v2.3 updates:
- Refactored prompt templates and processing pipeline to support dynamic variable substitution, enabling flexible and reusable prompt construction.
- Enhanced error handling and streamlined model/API detection logic for improved stability and performance.
- Added native support for custom LLM APIs — users can now configure external models by specifying a custom API endpoint and API key for translation tasks.
- Conducted comprehensive code refactoring, resolved existing bugs, and implemented various performance and usability improvements.