- Added `requires_contexts`, `requires_functions`, and `requires_expected_output` to evaluator list output
Removed
- Removed data loaders. Input variables to be used as replacements for data loaders.
1.5.2
Added
- Added evaluator execution by name
Changed
- Evaluator uses new endpoints instead of deprecated ones in all functions
1.5.1
Added
- Added 'include' parameter to listing execution logs to include optional fields
Changed
- Dropped support for running only skill validators
1.5.0
Added
- Added tags to evaluator and judge runs - Added tag filtering to execution logs
Changed
- Renamed the evaluator demonstration output and prompt parameters
1.4.0
Added - Introduced beta namespace for the client. - Judges functionality for improved AI evaluation. - Score field is now optional in `ExecutionLogList`. - Colab example links for better accessibility. - Refined example code formatting. - Enhanced async usage documentation. - Renamed custom evaluator for clarity.
1.3.3
Added
- Evaluators list and get methods - Improved docstrings for the SDK