v0.6.3 (2022-01-18)
Added
- New feature A callback can now be passed to
IntegratedGradientsvia thetarget_fnargument, in order to calculate the scalar target dimension from the model output. This is to bypass the requirement of passingtargetdirectly toexplainwhen thetargetof interest may depend on the prediction output. See the example in the docs. (#523). - A new comprehensive Introduction to explainability added to the documentation (#510).
Changed
- Python 3.6 has been deprecated from the supported versions as it has reached end-of-life.
Fixed
- Fix a bug with passing background images to
AnchorImageleading to an error (#542). - Fix a bug with rounding errors being introduced in
CounterfactualRLTabular(#550).
Development
- Docstrings have been updated and consolidated (#548). For developers, docstring conventions have been documented in CONTRIBUTING.md.
numpytyping has been updated to be compatible withnumpy 1.22(#543). This is a prerequisite for upgrading totensorflow 2.7.- To further improve reliability, strict
Optionaltype-checking withmypyhas been reinstated (#541). - The Alibi CI tests now include Windows and MacOS platforms (#575).
- The maximum
tensorflowversion has been bumped from 2.6 to 2.7 (#377).