Psiz

Release r0.12.4

* Update as many tests as possible to be Keras backend agnostic.
* Update doc generation.

Release r0.12.2

Major Changes
* Bump python version support (now supports 3.9, 3.10, 3.11)
* Update models and layers to use Keras 3 API.
* The majority of Keras objects are now backend agnostic (they do not invoke TF objects) allowing users to use TensorFlow, Pytorch, or Jax.
* Exceptions are objects which rely on TensorFlow Probability (tensorflow-probability), which includes all concrete classes of `StochasticEmbedding`. Migrating these objects to backend agnostic versions is still a work in progress.
* Update gate objects.
* Streamline `Gate` abstract class.
* Require using mixing weights. If you were previously using indices, you must convert your indices to one-hot float weights ahead of time.
* Remove redundant `BranchGate` layer. This layer does not have a clear use case given that it is easy to specify a functional or subclassed model with multiple branches.
* Remove deprecated classes and associated tests:
* `RankSimilarity`: Use `SoftRank` layer instead.
* `RateSimilarity`: Use `Logistic` layer instead.
* `ProgressBarRe`: users should use a package like `tqdm` instead (24 )
* `DistanceBased`: Use a concrete class of the `Proximity` layer instead.
* Remove (temporarily) RNN based functionality since the Keras 3 API changed the assumptions associated with RNN inputs, which requires a deep refactor of the existing PsiZ RNN cell objects.

**Full Changelog**: https://github.com/psiz-org/psiz/compare/v0.12.1...v0.12.2

This release has no feature additions. It merely bumps the TensorFlow and TensorFlow Probability requirements to the latest versions and updates the PsiZ codebase to be compatible with the latest versions.

This release adds some new features, but primarily focuses on reorganization of the core functionality of PsiZ into easy-to-use percept, proximity, and behavior modules.

Major Changes

* Source code related to architecting a "similarity function" has been reorganized to reflect the more general notion of a "proximity function" (which subsumes the notion of similarity, dissimilarity, kernel, and distance).
* Added new abstract `Proximity` layer which includes optional `activation` argument.
* Updated `Minkowski` and `MinkowskiStochastic` layer to inherit from `Proximity`
* Recommended approach is to use `activation` argument to modify output of Minkowski distance.
* Added experimental `CosineSimilarity` layer (which inherits from `Proximity`)
* Added experimental `InnerProduct` layer (which inherits from `Proximity`)
* Added experimental `GeneralizedInnerProduct` layer (which inherits from `Proximity`)
* `DistanceBased` layer is now deprecated in favor of using "activation" argument of `Proximity`
* Some functions have been re-homed as "activation" layers.
* Moved to `psiz.keras.layers.activations.ExponentialSimilarity`
* Moved to `psiz.keras.layers.activations.HeavyTailedSimilarity`
* Moved to `psiz.keras.layers.activations.InverseSimilarity`
* Moved to `psiz.keras.layers.activations.StudentsTSimilarity`
* `RankSimilarity` and `RateSimilarity` have been deprecated in favor of `SoftRank` and `Logistic`. In practice, this change means that users have increased responsibility for wiring up the layers, but model readability is now better. Examples, tutorials, and tests have all been updated to reflect these changes. To update code please see the examples.
* Cell-based layers have been moved to `experimental` until a stable API can be determined.

Minor Changes
* New `SoftRank` layer includes optional temperature parameter for controlling determinism.

* Add missing `name_scope` attribute in serialization of behavior layers.

A minor, mostly cosmetic release.

Breaking changes

* Removed deprecated objects.

Other changes

* Fixed missing autodoc generated API on readthedocs.