Tensorflow-decision-forests

Fix

- Regression with Mean Squared Error loss and Mean Average error loss
incorrectly clamped the gradients, leading to incorrect predictions.

Features

- Compatibility with TensorFlow 2.15.0.
- Inspector support for uplifting.

Fix

- Less warnings emitted when loading a model.

Features

- Add support for monotonic constraints.
- Add support for Mean average error (MAE) loss.
- Add support for Poisson loss.

Breaking Changes

- TF-DF no longer supports Python 3.8 since Tensorflow dropped its support.

Features

- Compatibility with Tensorflow 2.14.0
- Contrib: Training preprocessing jointly on the input features, labels and
weights

Fix

- Incorrect model predictions for models without features
- Data race for model resources

Features

- Compatibility with Tensorflow 2.13.0

Features

- Support for multi-task learning.
- New tutorial for TF-DF <--> TF.js
- Support for uplift modeling in the model inspector.
- New tutorial for Uplift modeling.
- Bump Bazel version to 6.1.0.

Fix

- Regex to generate Bazel workspace.
- Remove warning when converting Keras -> YDF.
- Fixed default hyperparameter issue Github 172.
- Various documentation issues fixed.