Keras

Breaking changes

There are no known breaking changes in this release compared to 3.0.1.

API changes

- Add `keras.random.binomial` and `keras.random.beta` RNG functions.
- Add masking support to `BatchNormalization`.
- Add `keras.losses.CTC` (loss function for sequence-to-sequence tasks) as well as the lower-level operation `keras.ops.ctc_loss`.
- Add `ops.random.alpha_dropout` and `layers.AlphaDropout`.
- Add gradient accumulation support for all backends, and enable optimizer EMA for JAX and torch

**Full Changelog**: https://github.com/keras-team/keras/compare/v3.0.1...v3.0.2

This is a minor release focused on bug fixes and performance improvements.

What's Changed

* Bug fixes and performance improvements.
* Add `stop_evaluating` and `stop_predicting` model attributes for callbacks, similar to `stop_training`.
* Add `keras.device()` scope for managing device placement in a multi-backend way.
* Support dict items in `PyDataset`.
* Add `hard_swish` activation and op.
* Fix cuDNN LSTM performance on TensorFlow backend.
* Add a `force_download` arg to `get_file` to force cache invalidation.

**Full Changelog**: https://github.com/keras-team/keras/compare/v3.0.0...v3.0.1

Major updates

See the [release announcement](https://keras.io/keras_3/) for a detailed list of major changes. Main highlights compared to Keras 2 are:

- Keras can now be run on top of JAX, PyTorch, TensorFlow, and even NumPy (note that the NumPy backend is inference-only).
- New low-level `keras.ops` API for building cross-framework components.
- New large-scale model distribution `keras.distribution` based on JAX.
- New stateless API for layers, models, optimizers, and metrics.

Breaking changes

See [this thread](https://github.com/keras-team/keras/issues/18467) for a complete list of breaking changes, as well as the [Keras 3 migration guide](https://keras.io/guides/migrating_to_keras_3/).

What's Changed
* Typofixes for `StringLookup` documentation by cw118 in https://github.com/keras-team/keras/pull/18333
* Fix ModelCheckpoint trained-on batch counting when using steps_per_execution>1 by jasnyj in https://github.com/keras-team/keras/pull/17632
* Fix legacy optimizer handling in `compile_from_config()`. by nkovela1 in https://github.com/keras-team/keras/pull/18492
* Remove options arg from ModelCheckpoint callback for Keras V3 saving, streamline ModelCheckpoint saving flow. Parameterize associated tests. by nkovela1 in https://github.com/keras-team/keras/pull/18545
* Use TENSORFLOW_VERSION when available during pip_build script by sampathweb in https://github.com/keras-team/keras/pull/18739

New Contributors
* cw118 made their first contribution in https://github.com/keras-team/keras/pull/18333
* jasnyj made their first contribution in https://github.com/keras-team/keras/pull/17632

**Full Changelog**: https://github.com/keras-team/keras/compare/v2.14.0...v2.15.0

What's Changed
* [keras/layers/normalization] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17965
* Update Python ver to 3.9 in Dockerfile by sampathweb in https://github.com/keras-team/keras/pull/18076
* [keras/saving/legacy/saved_model] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17978
* [keras/metrics] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17972
* Update example losses to bce- metrics/confusion_metrics.py by Frightera in https://github.com/keras-team/keras/pull/18045
* [keras/layers/regularization] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17968
* [keras/applications/efficientnet.py] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17758
* [keras/models] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17974
* [keras/mixed_precision] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17973
* Update indentation level - losses.py by Frightera in https://github.com/keras-team/keras/pull/18043
* Remove github user sushreebarsa from assignees. by shmishra99 in https://github.com/keras-team/keras/pull/18058
* [keras/preprocessing] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17977
* [keras/testing_infra] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17979
* [keras/benchmarks/benchmark_util.py] Use var rather than string literal for `is None` checks on `measure_performance` by SamuelMarks in https://github.com/keras-team/keras/pull/17980
* Some cleanup // Optimizers by Frightera in https://github.com/keras-team/keras/pull/18124
* [keras/layers/preprocessing] fix comments in RandomWidth, change to 'horizontally' instead of 'vertically' by qibolee in https://github.com/keras-team/keras/pull/18113
* [keras/applications] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17954
* [keras/estimator] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17957
* Fix CategoricalFocalCE documentation by Frightera in https://github.com/keras-team/keras/pull/18144
* [keras/utils] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17953
* [keras/layers/pooling] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17966
* use io_utils.print_msg in datset_utils by pedrobrs in https://github.com/keras-team/keras/pull/18054
* Update indentation level - activations.py by Frightera in https://github.com/keras-team/keras/pull/18036
* [keras/legacy_tf_layers] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17971
* [keras/layers/convolutional] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17963
* [keras/engine] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17956
* [keras/feature_column] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17958
* Fix markdown rendering issue by sachinprasadhs in https://github.com/keras-team/keras/pull/18205
* Consistently use "pickleable" instead of "picklable" by NicoWeio in https://github.com/keras-team/keras/pull/18140
* Use prefetch() after batching // image_dataset.py by Frightera in https://github.com/keras-team/keras/pull/18160
* Update Argument padding for zero_padding1d.py by SuryanarayanaY in https://github.com/keras-team/keras/pull/18223
* RGB image data is not grayscale image data by misterrioes in https://github.com/keras-team/keras/pull/18133
* updated read.me by VaishnaviMudaliar in https://github.com/keras-team/keras/pull/18226
* Fixed the typo in policy.py by tilakrayal in https://github.com/keras-team/keras/pull/18233
* Update example losses probabilistic_metrics.py by Frightera in https://github.com/keras-team/keras/pull/18234
* Fixed "reset_state" of R2Score metric by pdyakov in https://github.com/keras-team/keras/pull/18251
* Ignore hidden folders for image_dataset_from_directory by sachinprasadhs in https://github.com/keras-team/keras/pull/18177
* Improve error message for input data to fit. by tomrtk in https://github.com/keras-team/keras/pull/18042
* refactor: _log_epoch_metrics() by arjun-234 in https://github.com/keras-team/keras/pull/18274
* Refactor test cases to improve unit test quality by freddiewanah in https://github.com/keras-team/keras/pull/18303
* typo in disable_interactie_logging by ganeshiva in https://github.com/keras-team/keras/pull/18314
* Removes Python 3.8 support by sampathweb in https://github.com/keras-team/keras/pull/18332
* Update requirements.txt by qlzh727 in https://github.com/keras-team/keras/pull/18382

New Contributors
* sampathweb made their first contribution in https://github.com/keras-team/keras/pull/18076
* shmishra99 made their first contribution in https://github.com/keras-team/keras/pull/18058
* qibolee made their first contribution in https://github.com/keras-team/keras/pull/18113
* pedrobrs made their first contribution in https://github.com/keras-team/keras/pull/18054
* NicoWeio made their first contribution in https://github.com/keras-team/keras/pull/18140
* SuryanarayanaY made their first contribution in https://github.com/keras-team/keras/pull/18223
* misterrioes made their first contribution in https://github.com/keras-team/keras/pull/18133
* VaishnaviMudaliar made their first contribution in https://github.com/keras-team/keras/pull/18226
* pdyakov made their first contribution in https://github.com/keras-team/keras/pull/18251
* tomrtk made their first contribution in https://github.com/keras-team/keras/pull/18042
* arjun-234 made their first contribution in https://github.com/keras-team/keras/pull/18274
* freddiewanah made their first contribution in https://github.com/keras-team/keras/pull/18303
* ganeshiva made their first contribution in https://github.com/keras-team/keras/pull/18314

**Full Changelog**: https://github.com/keras-team/keras/compare/v2.13.1...v2.14.0

What's Changed
* [keras/layers/normalization] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17965
* Update Python ver to 3.9 in Dockerfile by sampathweb in https://github.com/keras-team/keras/pull/18076
* [keras/saving/legacy/saved_model] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17978
* [keras/metrics] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17972
* Update example losses to bce- metrics/confusion_metrics.py by Frightera in https://github.com/keras-team/keras/pull/18045
* [keras/layers/regularization] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17968
* [keras/applications/efficientnet.py] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17758
* [keras/models] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17974
* [keras/mixed_precision] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17973
* Update indentation level - losses.py by Frightera in https://github.com/keras-team/keras/pull/18043
* Remove github user sushreebarsa from assignees. by shmishra99 in https://github.com/keras-team/keras/pull/18058
* [keras/preprocessing] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17977
* [keras/testing_infra] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17979
* [keras/benchmarks/benchmark_util.py] Use var rather than string literal for `is None` checks on `measure_performance` by SamuelMarks in https://github.com/keras-team/keras/pull/17980
* Some cleanup // Optimizers by Frightera in https://github.com/keras-team/keras/pull/18124
* [keras/layers/preprocessing] fix comments in RandomWidth, change to 'horizontally' instead of 'vertically' by qibolee in https://github.com/keras-team/keras/pull/18113
* [keras/applications] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17954
* [keras/estimator] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17957
* Fix CategoricalFocalCE documentation by Frightera in https://github.com/keras-team/keras/pull/18144
* [keras/utils] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17953
* [keras/layers/pooling] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17966
* use io_utils.print_msg in datset_utils by pedrobrs in https://github.com/keras-team/keras/pull/18054
* Update indentation level - activations.py by Frightera in https://github.com/keras-team/keras/pull/18036
* [keras/legacy_tf_layers] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17971
* [keras/layers/convolutional] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17963
* [keras/engine] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17956
* [keras/feature_column] Standardise docstring usage of "Default to" by SamuelMarks in https://github.com/keras-team/keras/pull/17958
* Fix markdown rendering issue by sachinprasadhs in https://github.com/keras-team/keras/pull/18205
* Consistently use "pickleable" instead of "picklable" by NicoWeio in https://github.com/keras-team/keras/pull/18140
* Use prefetch() after batching // image_dataset.py by Frightera in https://github.com/keras-team/keras/pull/18160
* Update Argument padding for zero_padding1d.py by SuryanarayanaY in https://github.com/keras-team/keras/pull/18223
* RGB image data is not grayscale image data by misterrioes in https://github.com/keras-team/keras/pull/18133
* updated read.me by VaishnaviMudaliar in https://github.com/keras-team/keras/pull/18226
* Fixed the typo in policy.py by tilakrayal in https://github.com/keras-team/keras/pull/18233
* Update example losses probabilistic_metrics.py by Frightera in https://github.com/keras-team/keras/pull/18234
* Fixed "reset_state" of R2Score metric by pdyakov in https://github.com/keras-team/keras/pull/18251
* Ignore hidden folders for image_dataset_from_directory by sachinprasadhs in https://github.com/keras-team/keras/pull/18177
* Improve error message for input data to fit. by tomrtk in https://github.com/keras-team/keras/pull/18042
* refactor: _log_epoch_metrics() by arjun-234 in https://github.com/keras-team/keras/pull/18274
* Refactor test cases to improve unit test quality by freddiewanah in https://github.com/keras-team/keras/pull/18303
* typo in disable_interactie_logging by ganeshiva in https://github.com/keras-team/keras/pull/18314
* Removes Python 3.8 support by sampathweb in https://github.com/keras-team/keras/pull/18332

New Contributors
* sampathweb made their first contribution in https://github.com/keras-team/keras/pull/18076
* shmishra99 made their first contribution in https://github.com/keras-team/keras/pull/18058
* qibolee made their first contribution in https://github.com/keras-team/keras/pull/18113
* pedrobrs made their first contribution in https://github.com/keras-team/keras/pull/18054
* NicoWeio made their first contribution in https://github.com/keras-team/keras/pull/18140
* SuryanarayanaY made their first contribution in https://github.com/keras-team/keras/pull/18223
* misterrioes made their first contribution in https://github.com/keras-team/keras/pull/18133
* VaishnaviMudaliar made their first contribution in https://github.com/keras-team/keras/pull/18226
* pdyakov made their first contribution in https://github.com/keras-team/keras/pull/18251
* tomrtk made their first contribution in https://github.com/keras-team/keras/pull/18042
* arjun-234 made their first contribution in https://github.com/keras-team/keras/pull/18274
* freddiewanah made their first contribution in https://github.com/keras-team/keras/pull/18303
* ganeshiva made their first contribution in https://github.com/keras-team/keras/pull/18314

**Full Changelog**: https://github.com/keras-team/keras/compare/v2.13.1...v2.14.0-rc0