Release Notes for Version 0.1.5 of llm-duckduckgo-chat
Overview
Version 0.1.5 of the `llm-duckduckgo-chat` plugin introduces several enhancements and bug fixes aimed at improving the functionality and reliability of interactions with the DuckDuckGo AI Chat service.
Key Changes
Model Name Correction
- **Updated Model Name**: The model name for "Mixtral Small 3" has been corrected to "Mistral Small 3" in the README documentation to ensure consistency and accuracy.
New Features
- **VQD Hash Handling**:
- Introduced support for handling `vqdhash`, a new header that enhances the request's integrity and security.
- The `DuckChatModel` class now fetches and utilizes both `vqd` and `vqdhash` for improved request handling.
Code Improvements
- **llm chat**:
- In the previous version of the plugin ``llm chat`` did't work. With release llm chat with duckchat models works as expected.
- **Error Handling**:
- Enhanced error handling when fetching `vqd` and `vqdhash`, ensuring that appropriate exceptions are raised in case of failures.
- Improved response handling in the `fetch_response` method to manage rate limits more effectively.
Testing Enhancements
- **Unit Tests**:
- Updated unit tests to cover the new `vqdhash` functionality, ensuring that both `vqd` and `vqdhash` are correctly processed and validated in various scenarios.
- Added assertions in tests to verify the correct handling of the new headers.
Version Bump
- **Version Update**: The version number has been incremented from `0.1.4` to `0.1.5` to reflect the changes made in this release.
Conclusion
This release focuses on improving the robustness of the `llm-duckduckgo-chat` plugin by adding support for new features and enhancing existing functionalities. I encourage everyone to update to version 0.1.5 to take advantage of these improvements.
