v0.1.3
PlanAI v0.1.3 Release Notes
We are pleased to announce the release of PlanAI v0.1.3, which includes important improvements and crucial bug fixes, focusing significantly on provenance handling and traceability enhancements in the dashboard.
New Features
- Dashboard Trace Improvements: Improved the expand/collapse functionality for trace prefixes in the dashboard by adding interactive elements and stylized displays, enhancing usability and clarity (#267186b).
- Tracing Mechanism: Introduced a tracing mechanism in the dashboard to improve the visibility and management of provenance, enhancing overall traceability (#f85a189).
Enhancements
- Cache Enhancements: Updated the system to store
input_task
with cache data while ensuring cache updates occur even on cache hits (#f8dca3e). - CSS and Styling Updates: Refactored dashboard style by separating CSS into distinct files, enabling easier maintenance and future enhancements (#2910cbd4d, #6add5e0).
- Expanded Configuration Options: Added
enable_trace
option for JoinedTaskWorker to allow more precise control over task traceability (#c97f1da).
Bug Fixes
- Task Provenance Copy Fix: Enhanced
Task.copy_input_provenance
to return model copies rather than references, mitigating corruption risks inherent in reference passing (#2eb2f8d). - Provenance Corruption Fix: Addressed a major issue where provenance data could become corrupted due to Python's reference handling by ensuring copies are used where necessary (#e3bd7bf).
- Notification Tasks Fix: Resolved an issue with notification tasks improperly removing provenance, which led to premature invocation of joiners (#f85a189).
- Testing Improvements: Ensured test reliability by copying test result arrays to avoid unexpected deletions when contexts exit (#a339573).
Maintenance
- Code Quality Improvements: Installed pre-commit hooks for Flake8 and Black to enforce code style and quality (#9d9ef9c).
This release notably improves the reliability of provenance management and enhances user interaction with traceable data paths in the dashboard, along with making significant strides towards a more stable caching system.
For further details and documentation, please visit our documentation page. We appreciate your feedback and continued support. If you encounter any issues or have suggestions, feel free to open an issue on our GitHub repository.
Thank you for using PlanAI!