keras-core

Highlights

• Add ops keras.random.shuffle, keras.ops.image.map_coordinates
• Add support for tf.SparseTensor with TensorFlow backend (same level of support as legacy tf.keras)
• Make Discretization layer backend agnostic
• Bug fixes and performance improvements (in particular significantly reduced memory usage during training for JAX)

What's Changed

• Make jax2tf import conditional by @nkovela1 in https://github.com/keras-team/keras-core/pull/911
• Separate the metrics variables from non-trainable variables. by @qlzh727 in https://github.com/keras-team/keras-core/pull/910
• Update jax trainer function to save memory buffer. by @qlzh727 in https://github.com/keras-team/keras-core/pull/897
• [GHA] fix several codecov issues by @kiukchung in https://github.com/keras-team/keras-core/pull/912
• Add example to fine-tune TorchVision models using Keras by @soumik12345 in https://github.com/keras-team/keras-core/pull/909
• Refactor merging layers tests as parametrized tests. by @hertschuh in https://github.com/keras-team/keras-core/pull/913
• Add ops.random.shuffle by @james77777778 in https://github.com/keras-team/keras-core/pull/907
• Add ops.map_coordinates by @james77777778 in https://github.com/keras-team/keras-core/pull/906
• Makes ops.split in torch consistent with other backends by @james77777778 in https://github.com/keras-team/keras-core/pull/914
• Add: MultipleChoice with Transfer Learning example by @awsaf49 in https://github.com/keras-team/keras-core/pull/884
• Update test_case.py by @PatReis in https://github.com/keras-team/keras-core/pull/917
• Fix JAX RNN backend issue. by @qlzh727 in https://github.com/keras-team/keras-core/pull/924
• Adds JAX version checking for Export by @nkovela1 in https://github.com/keras-team/keras-core/pull/923
• make discretization backend agnostic by @divyashreepathihalli in https://github.com/keras-team/keras-core/pull/896
• Memory optimization for jax trainer. by @qlzh727 in https://github.com/keras-team/keras-core/pull/888
• [*.py] Standardise docstring usage of "Default to" by @SamuelMarks in https://github.com/keras-team/keras-core/pull/921
• Fix the inconsistency in affine_transform between numpy and jax by @james77777778 in https://github.com/keras-team/keras-core/pull/926
• Add support for Tensorflow SparseTensors: merging layers. by @hertschuh in https://github.com/keras-team/keras-core/pull/925
• fix torch error for bincount by @divyashreepathihalli in https://github.com/keras-team/keras-core/pull/927
• Refactor torch's affine_transform by @james77777778 in https://github.com/keras-team/keras-core/pull/929
• Increase test coverage + Fix save_model_to_hdf5 + Improve is_remote_path + Fix is_remote_path by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/900
• Slice serialization by @gleize in https://github.com/keras-team/keras-core/pull/932
• Some dtype fixes by @fchollet in https://github.com/keras-team/keras-core/pull/935
• Hacky fix for dictionary output with tf 2.14 by @mattdangerw in https://github.com/keras-team/keras-core/pull/933

New Contributors

• @awsaf49 made their first contribution in https://github.com/keras-team/keras-core/pull/884
• @PatReis made their first contribution in https://github.com/keras-team/keras-core/pull/917
• @SamuelMarks made their first contribution in https://github.com/keras-team/keras-core/pull/921

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.6...v0.1.7

Highlights

• Performance optimizations in normalization layers
• New op: ops.nn.moments
• Add TorchModuleWrapper to make it possible to use PyTorch Modules in Keras Core models with the PyTorch backend. Modules added as attributes of a layer are automatically wrapped (so that their weights are tracked).
• Add ignore_class argument in SparseCategoricalCrossentropy.
• Add new experimental data parallel and model parallel distribution API for JAX.

What's Changed

• Fix where op by @AakashKumarNain in https://github.com/keras-team/keras-core/pull/770
• Add skeleton for the common distribution backend. by @qlzh727 in https://github.com/keras-team/keras-core/pull/765
• Add more H5 backwards compatibility tests and lambda support by @nkovela1 in https://github.com/keras-team/keras-core/pull/775
• Add implementation for Distribution API by @qlzh727 in https://github.com/keras-team/keras-core/pull/776
• Removes tf.keras from transposed conv test by @sampathweb in https://github.com/keras-team/keras-core/pull/782
• Fix channels_first format bug for rescaling layer by @haifeng-jin in https://github.com/keras-team/keras-core/pull/792
• Fix argument order in stft and istft by @james77777778 in https://github.com/keras-team/keras-core/pull/790
• fixed squeeze axis inconsistency by @jackd in https://github.com/keras-team/keras-core/pull/788
• Replace tf.data with tf_data in preprocessing tests by @sampathweb in https://github.com/keras-team/keras-core/pull/787
• Reimplement the data parallel distribution. by @qlzh727 in https://github.com/keras-team/keras-core/pull/785
• Split fix by @jackd in https://github.com/keras-team/keras-core/pull/789
• Fix bug: Softmax function to support tuple axes + Add more tests by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/778
• Add LayoutMap which will be used for model parallelism by @qlzh727 in https://github.com/keras-team/keras-core/pull/793
• Fix bug: Log_Softmax with tuple axes + Add tests by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/794
• Port tab transformer example to keras core by @anas-rz in https://github.com/keras-team/keras-core/pull/621
• Update example losses on docstrings by @Frightera in https://github.com/keras-team/keras-core/pull/795
• Add model parallel distribution. by @qlzh727 in https://github.com/keras-team/keras-core/pull/797
• Increase-test-coverage-in-activations_test.py by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/799
• add ignore_class param to sparse CE loss by @AakashKumarNain in https://github.com/keras-team/keras-core/pull/599
• Convert pretraining_BERT.py example to keras core by @pranavvp16 in https://github.com/keras-team/keras-core/pull/801
• fixed apply_mask bug in losses by @jackd in https://github.com/keras-team/keras-core/pull/802
• Replace tf.keras in center-crop and resize tests by @sampathweb in https://github.com/keras-team/keras-core/pull/804
• Add a check for class_id argument by @Frightera in https://github.com/keras-team/keras-core/pull/796
• Add a dummy init for distribution, so it will be treat as a package. by @qlzh727 in https://github.com/keras-team/keras-core/pull/809
• Allow test using dynamic shapes to run on JAX. by @hertschuh in https://github.com/keras-team/keras-core/pull/810
• Keep "mse" as the metric name in the log by @haifeng-jin in https://github.com/keras-team/keras-core/pull/812
• Fixed misc mixed precision issues by @mattdangerw in https://github.com/keras-team/keras-core/pull/805
• Don't double cast layer norm weights in mixed precision by @mattdangerw in https://github.com/keras-team/keras-core/pull/815
• Add loss attribute for backward compatibility by @haifeng-jin in https://github.com/keras-team/keras-core/pull/813
• Add TorchModuleWrapper by @soumik12345 in https://github.com/keras-team/keras-core/pull/779
• Make sure the Test CI action fails on bad lint by @kiukchung in https://github.com/keras-team/keras-core/pull/818
• Format confusion_metrics_test by @mattdangerw in https://github.com/keras-team/keras-core/pull/821
• Fix the Jax distributed variable load issue. by @qlzh727 in https://github.com/keras-team/keras-core/pull/817
• Always save the build config when building a layer by @mattdangerw in https://github.com/keras-team/keras-core/pull/820
• Uses backend.shape in random preprocessing by @sampathweb in https://github.com/keras-team/keras-core/pull/822
• Don't lock functional models by @mattdangerw in https://github.com/keras-team/keras-core/pull/823
• Refactor conv_transpose padding by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/811
• Add rank>2 support for stft by @james77777778 in https://github.com/keras-team/keras-core/pull/825
• Fix reduce_shape + Add tests to operation_utils_test.py by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/826
• Fix weird broken test that never ran by @mattdangerw in https://github.com/keras-team/keras-core/pull/827
• Convert "English-to-Spanish translation with KerasNLP" to keras-core by @freedomtan in https://github.com/keras-team/keras-core/pull/777
• Replace assertTrue(isinstance(...)) with assertIsInstance for easier … by @hertschuh in https://github.com/keras-team/keras-core/pull/831
• Improve the flexibility of standardize_dtype and fix pad in torch backend by @james77777778 in https://github.com/keras-team/keras-core/pull/828
• Fix the distribution logic for the JAX trainer. by @qlzh727 in https://github.com/keras-team/keras-core/pull/832
• Add a docstring for compile by @mattdangerw in https://github.com/keras-team/keras-core/pull/838
• Enable codecov by @kiukchung in https://github.com/keras-team/keras-core/pull/840
• Add support for Tensorflow SparseTensors: core classes. by @hertschuh in https://github.com/keras-team/keras-core/pull/839
• bert-mlm port to keras-core with tf backend by @Mrutyunjay01 in https://github.com/keras-team/keras-core/pull/843
• Enable codecov flags and coverage carryover for keras_core and keras_… by @kiukchung in https://github.com/keras-team/keras-core/pull/841
• Update minimum python version to 3.9 to align with TF/JAX by @sampathweb in https://github.com/keras-team/keras-core/pull/848
• Update the eager_build logic for jax trainer. by @qlzh727 in https://github.com/keras-team/keras-core/pull/845
• Update return type for optimizer.apply by @mattdangerw in https://github.com/keras-team/keras-core/pull/850
• Add test,codecov,pypi status badges to README by @kiukchung in https://github.com/keras-team/keras-core/pull/852
• Fix autocast scope again by @mattdangerw in https://github.com/keras-team/keras-core/pull/849
• Add Export for JAX Backend by @nkovela1 in https://github.com/keras-team/keras-core/pull/819
• Increase test coverage + Fixing bug in static_call by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/847
• Add a loss scale optimizer by @mattdangerw in https://github.com/keras-team/keras-core/pull/851
• Increase test coverage in keras_core/utils by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/856
• Increase test coverage in keras_core/trainers by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/860
• Add a note in README.md regarding pre-import of backend by @SuryanarayanaY in https://github.com/keras-team/keras-core/pull/863
• Add optimizer weights to the model summary. by @qlzh727 in https://github.com/keras-team/keras-core/pull/859
• Add support for Tensorflow SparseTensors: core layers. by @hertschuh in https://github.com/keras-team/keras-core/pull/854
• Increase test coverage in keras_core/backend by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/864
• Add support for Tensorflow SparseTensors: reshaping layers. by @hertschuh in https://github.com/keras-team/keras-core/pull/865
• Respect the $KERAS_HOME environment variable for get_file by @mattdangerw in https://github.com/keras-team/keras-core/pull/868
• Don't convert int array input to floatx by @mattdangerw in https://github.com/keras-team/keras-core/pull/867
• run_layer_test also runs symbolic calls when input_data is provided. by @hertschuh in https://github.com/keras-team/keras-core/pull/870
• Increase test coverage in utils + Fix Bug in cast_to_common_dtype by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/875
• Fix stop_gradient in np by @james77777778 in https://github.com/keras-team/keras-core/pull/872
• Add sparseness inference for TensorFlow backend ops and base layer. by @hertschuh in https://github.com/keras-team/keras-core/pull/878
• Fix summary when optimizer is None by @mattdangerw in https://github.com/keras-team/keras-core/pull/879
• Add ops.nn.moments and speed-up normalization layers by @james77777778 in https://github.com/keras-team/keras-core/pull/866
• Don't show optimizer params if not optimizer or not built by @mattdangerw in https://github.com/keras-team/keras-core/pull/885
• Export dtype utils in keras_core.backend by @mattdangerw in https://github.com/keras-team/keras-core/pull/886
• Set actions.yml to run with read-only permissions by @pnacht in https://github.com/keras-team/keras-core/pull/882
• Fix model.save API for legacy H5 format by @nkovela1 in https://github.com/keras-team/keras-core/pull/891
• Add missing dtypes. by @gleize in https://github.com/keras-team/keras-core/pull/887
• Increase tests in utils by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/880
• syntax fix in cct example by @anas-rz in https://github.com/keras-team/keras-core/pull/898
• Use ops.rsqrt, improve normalization layers and enable ops fusion in tflite by @james77777778 in https://github.com/keras-team/keras-core/pull/892
• fixes #550: move torch tensors to cpu before to numpy array by @dkgaraujo in https://github.com/keras-team/keras-core/pull/853

New Contributors

• @kiukchung made their first contribution in https://github.com/keras-team/keras-core/pull/818
• @Mrutyunjay01 made their first contribution in https://github.com/keras-team/keras-core/pull/843
• @SuryanarayanaY made their first contribution in https://github.com/keras-team/keras-core/pull/863
• @pnacht made their first contribution in https://github.com/keras-team/keras-core/pull/882
• @gleize made their first contribution in https://github.com/keras-team/keras-core/pull/887
• @dkgaraujo made their first contribution in https://github.com/keras-team/keras-core/pull/853

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.5...v0.1.6

Highlights

• Add model.export() flow.
• Add ops rsqrt, extract_sequences, rfft, stft, irfft, istft.
• Bug fixes and performance improvements.
• ops.shape now returns a tuple with all backends (rather than a TF tensor with the TF backend). The tuple may contain scalar tensors for dynamic dimensions, and contains integers otherwise for static dimensions.

What's Changed

• Ensure backwards compatibility for legacy H5 saving format by @nkovela1 in https://github.com/keras-team/keras-core/pull/682
• Encapsulate build logic by @grasskin in https://github.com/keras-team/keras-core/pull/679
• Fix the --install arg doesn't work by @haifeng-jin in https://github.com/keras-team/keras-core/pull/686
• Convert to tensor for jax ops.numpy.transpose by @mattdangerw in https://github.com/keras-team/keras-core/pull/687
• Fix custom functional reload issue by @fchollet in https://github.com/keras-team/keras-core/pull/693
• Update the softmax default to axis=-1 by @mattdangerw in https://github.com/keras-team/keras-core/pull/694
• Random Flip for batched inputs by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/698
• Fix TextVectorization + Sequential bug by @fchollet in https://github.com/keras-team/keras-core/pull/696
• Add Export for TF backend by @nkovela1 in https://github.com/keras-team/keras-core/pull/692
• Fix typo in WhileLoop docstring in core.py by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/703
• Add activation docstrings // ops.nn.py by @Frightera in https://github.com/keras-team/keras-core/pull/705
• Remove tf dependency from math_test by @Frightera in https://github.com/keras-team/keras-core/pull/706
• Little cleanup in backend - config.py / export.py by @Frightera in https://github.com/keras-team/keras-core/pull/709
• Fix Issue in SegmentMax's Symbolic Call in keras_core/ops/math.py by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/713
• Replace tf.keras for expected result in Conv1D Test by @sampathweb in https://github.com/keras-team/keras-core/pull/725
• Add some additional H5 saving test coverage by @nkovela1 in https://github.com/keras-team/keras-core/pull/724
• Always return a tuple from ops.shape by @mattdangerw in https://github.com/keras-team/keras-core/pull/722
• Only compute mean once in layernorm by @mattdangerw in https://github.com/keras-team/keras-core/pull/721
• Add rsqrt to ops API by @shivance in https://github.com/keras-team/keras-core/pull/708
• Fix position encoder in examples/.../token_learner.py by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/727
• Support keras.Input in get_source_inputs by @haifeng-jin in https://github.com/keras-team/keras-core/pull/732
• Convert cyclegan to keras-core by @freedomtan in https://github.com/keras-team/keras-core/pull/728
• Add rms_scaling in LayerNormalization by @shivance in https://github.com/keras-team/keras-core/pull/726
• Docstrings for numpy ops by @FayazRahman in https://github.com/keras-team/keras-core/pull/714
• Add extract_sequences, rfft and stft to ops.math by @james77777778 in https://github.com/keras-team/keras-core/pull/717
• Apply scipy LazyModule by @james77777778 in https://github.com/keras-team/keras-core/pull/734
• Support torch symbolic call for normalization layer in preprocessing by @haifeng-jin in https://github.com/keras-team/keras-core/pull/739
• Replace tf.keras with numpy for conv2d and conv3d tests by @sampathweb in https://github.com/keras-team/keras-core/pull/740
• Enhanced Docstrings and Examples in /ops/function.py, /ops/math.py and /ops/nn.py. by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/736
• Fixed uniqueness bug and bits->bytes by @jackd in https://github.com/keras-team/keras-core/pull/747
• Add Examples to numpy.py functions by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/745
• Use convert_to_numpy in hashing and tests by @james77777778 in https://github.com/keras-team/keras-core/pull/748
• Convert ddim by @freedomtan in https://github.com/keras-team/keras-core/pull/743
• Add irfft and istft by @james77777778 in https://github.com/keras-team/keras-core/pull/742
• ported hashed-crossing by @asingh9530 in https://github.com/keras-team/keras-core/pull/685
• Add a high level API for distribution for JAX backend. by @qlzh727 in https://github.com/keras-team/keras-core/pull/741
• An attempt to add MPS device to torch backend by @M7Saad in https://github.com/keras-team/keras-core/pull/750
• Remove tf.keras from depthwise and separable conv tests by @sampathweb in https://github.com/keras-team/keras-core/pull/752
• Fix bug + Add Tests + Enhance docstrings (shape_equal) by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/751
• Fix standardize_shape by @james77777778 in https://github.com/keras-team/keras-core/pull/753
• Fix naming in mobilenet_v3 and densenet by @james77777778 in https://github.com/keras-team/keras-core/pull/759
• Enhance docstrings + easy to understand examples (nn.py) by @Faisal-Alsrheed in https://github.com/keras-team/keras-core/pull/758
• Remove backend.DYNAMIC_SHAPES_OK as dynamic shapes are supported in a… by @hertschuh in https://github.com/keras-team/keras-core/pull/763

New Contributors

• @Faisal-Alsrheed made their first contribution in https://github.com/keras-team/keras-core/pull/703
• @shivance made their first contribution in https://github.com/keras-team/keras-core/pull/708
• @jackd made their first contribution in https://github.com/keras-team/keras-core/pull/747
• @M7Saad made their first contribution in https://github.com/keras-team/keras-core/pull/750

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.4...v0.1.5

Highlights

• Add Lion optimizer
• Enable model.predict() and model.evaluate() with the NumPy backend
• Add ops cosh, sinh, arctanh, arcsinh, arccosh
• Add ops segment_max, digitize
• Bug fixes and performance improvements

What's Changed

• Set min TF Version by @sampathweb in https://github.com/keras-team/keras-core/pull/619
• Avoid some tree operations by @haifeng-jin in https://github.com/keras-team/keras-core/pull/620
• Add Lion optimizer by @james77777778 in https://github.com/keras-team/keras-core/pull/610
• Typo in layer tutorial by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/624
• fix load of models with Add and Concat layers by @freedomtan in https://github.com/keras-team/keras-core/pull/626
• Port/test learnable resizer to keras core backend agnostic by @anas-rz in https://github.com/keras-team/keras-core/pull/622
• Port fixres keras core by @anas-rz in https://github.com/keras-team/keras-core/pull/630
• Only ignore xx_mask arguments for the build shapes dict by @mattdangerw in https://github.com/keras-team/keras-core/pull/632
• Add numpy trainer by @james77777778 in https://github.com/keras-team/keras-core/pull/633
• [split_dataset] migrating from tf.keras to keras_core by @asingh9530 in https://github.com/keras-team/keras-core/pull/505
• Add hyperbolic ops by @FayazRahman in https://github.com/keras-team/keras-core/pull/634
• Port visualizing what convnets learn by @anas-rz in https://github.com/keras-team/keras-core/pull/640
• Port the ImageClassifier guide to keras_core by @tirthasheshpatel in https://github.com/keras-team/keras-core/pull/608
• Port simsiam to keras-core by @anas-rz in https://github.com/keras-team/keras-core/pull/644
• Fix the ImageClassifier demo and minimize the use of tensorflow by @tirthasheshpatel in https://github.com/keras-team/keras-core/pull/645
• Port neural decision forest by @anas-rz in https://github.com/keras-team/keras-core/pull/631
• Removes ref in Global Average Pool and Elu/Prelu Activations by @sampathweb in https://github.com/keras-team/keras-core/pull/646
• Small fixes on image.extract_patches by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/647
• Fix the missing weight name for Dense layer by @qlzh727 in https://github.com/keras-team/keras-core/pull/648
• Update learning_rate type to float in learning_rate_scheduler's logs by @siyavash in https://github.com/keras-team/keras-core/pull/649
• Performance improvement for MHA by @haifeng-jin in https://github.com/keras-team/keras-core/pull/654
• Add digitize op for Discretization layer by @abuelnasr0 in https://github.com/keras-team/keras-core/pull/641
• support passing jax.random.PRNGKey inputs in jax by @GallagherCommaJack in https://github.com/keras-team/keras-core/pull/651
• Removes tf.keras from Pooling tests - Max, Average and Global pooling by @sampathweb in https://github.com/keras-team/keras-core/pull/657
• Add numpy bincount docstrings by @dranaivo in https://github.com/keras-team/keras-core/pull/655
• Add a segment_max op by @ianstenbit in https://github.com/keras-team/keras-core/pull/660
• Port mlp_image_classification.py to all backends by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/663
• update requirements for all cpu install by @haifeng-jin in https://github.com/keras-team/keras-core/pull/666
• Random rotation port by @kamathis4 in https://github.com/keras-team/keras-core/pull/584
• Port TensorFlow implementation of the RandomCrop layer by @soumik12345 in https://github.com/keras-team/keras-core/pull/493
• Backend agnostic port of object detection using vit by @soumik12345 in https://github.com/keras-team/keras-core/pull/668
• Remove self.support_jit = False in RandomFlip by @james77777778 in https://github.com/keras-team/keras-core/pull/665
• Port Compact Convolutional Transformers to backend agnostic by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/669
• Move mlp_image_classification to backend agnostic folder by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/671
• Change saving API routing to deprecate SavedModel format by @nkovela1 in https://github.com/keras-team/keras-core/pull/674
• Convert neural machine translation with transformer to keras-core by @freedomtan in https://github.com/keras-team/keras-core/pull/623
• Make ner transformers work again by @freedomtan in https://github.com/keras-team/keras-core/pull/675
• Do not default to file root by @grasskin in https://github.com/keras-team/keras-core/pull/678
• Convert "text generation with miniature gpt" to Keras Core by @freedomtan in https://github.com/keras-team/keras-core/pull/676
• torch image added by @asingh9530 in https://github.com/keras-team/keras-core/pull/659
• Added some docstrings ops/nn.py by @Frightera in https://github.com/keras-team/keras-core/pull/677

New Contributors

• @asingh9530 made their first contribution in https://github.com/keras-team/keras-core/pull/505
• @FayazRahman made their first contribution in https://github.com/keras-team/keras-core/pull/634
• @siyavash made their first contribution in https://github.com/keras-team/keras-core/pull/649
• @abuelnasr0 made their first contribution in https://github.com/keras-team/keras-core/pull/641
• @dranaivo made their first contribution in https://github.com/keras-team/keras-core/pull/655

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.3...v0.1.4

Highlights

• Add compatibility with legacy whole-model h5 saving and loading
• Add keras_core.ops.fft and keras_core.ops.fft2
• Add keras_core.ops.image.affine_transform
• Add keras_core.ops.image.extract_patches
• Bug fixes and performance improvements
• Now compatible with TensorFlow versions 2.8 to 2.13

What's Changed

• Add Nadam for torch backend by @haifeng-jin in https://github.com/keras-team/keras-core/pull/551
• Fix multi-dimension sample weights passed to mean metrics by @mattdangerw in https://github.com/keras-team/keras-core/pull/552
• Update operation_utils.py - Added docstrings by @sqali in https://github.com/keras-team/keras-core/pull/514
• Node test by @kamathis4 in https://github.com/keras-team/keras-core/pull/557
• Add affine_transform op to all backends by @james77777778 in https://github.com/keras-team/keras-core/pull/477
• Add FFT Ops by @abheesht17 in https://github.com/keras-team/keras-core/pull/480
• Added CategoryEncoding to keras_core.ops.nn by @hazemessamm in https://github.com/keras-team/keras-core/pull/490
• Symbolic Args test by @kamathis4 in https://github.com/keras-team/keras-core/pull/536
• Converted to Keras Core: DeepLabV3Plus by @anas-rz in https://github.com/keras-team/keras-core/pull/545
• test and move backend agnostic example by @anas-rz in https://github.com/keras-team/keras-core/pull/567
• Port few shot learning with reptile by @anas-rz in https://github.com/keras-team/keras-core/pull/564
• Add some additional TF SavedModel tests by @nkovela1 in https://github.com/keras-team/keras-core/pull/569
• Fixes Torch-GPU Test failure on initializers/random_initializers_test.py by @sampathweb in https://github.com/keras-team/keras-core/pull/575
• Keras tensor functions by @kamathis4 in https://github.com/keras-team/keras-core/pull/576
• add fori_loop op to all 3 backends by @GallagherCommaJack in https://github.com/keras-team/keras-core/pull/462
• Fix Module Utils _available attribute by @sampathweb in https://github.com/keras-team/keras-core/pull/579
• Add some docstrings to keras_core/ops/numpy.py by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/558
• Add random seed support for torch dropout by @haifeng-jin in https://github.com/keras-team/keras-core/pull/568
• Port Keypoint detection to backend agnostic-keras-core by @anas-rz in https://github.com/keras-team/keras-core/pull/546
• Make random flip backend agnostic by @AmedeoBiolatti in https://github.com/keras-team/keras-core/pull/515
• Convert RandomTranslation to the backend-agnostic implementation by @james77777778 in https://github.com/keras-team/keras-core/pull/572
• doc: DocString for numpy.amax by @pranavvp16 in https://github.com/keras-team/keras-core/pull/582
• Add backend-agnostic video transformers example by @soumik12345 in https://github.com/keras-team/keras-core/pull/583
• Converted cutmix example to Keras Core by @cosmo3769 in https://github.com/keras-team/keras-core/pull/527
• docString for numpy.amin by @pranavvp16 in https://github.com/keras-team/keras-core/pull/586
• Update Tests for specific backend imports by @sampathweb in https://github.com/keras-team/keras-core/pull/591
• implement patch extraction in ops by @anas-rz in https://github.com/keras-team/keras-core/pull/581
• Converted to keras core external attention by @anas-rz in https://github.com/keras-team/keras-core/pull/529
• Backend consistency in handling of empty tensor in min and max by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/592
• Fix more spots torch != numpy for reductions by @mattdangerw in https://github.com/keras-team/keras-core/pull/596
• Backend-agnostic port of Image classification with Vision Transformer by @soumik12345 in https://github.com/keras-team/keras-core/pull/589
• More backend agnostic in cutmix and mixup file by @cosmo3769 in https://github.com/keras-team/keras-core/pull/593
• Always install the correct version with pip_build by @mattdangerw in https://github.com/keras-team/keras-core/pull/595
• Export ops.cond by @AmedeoBiolatti in https://github.com/keras-team/keras-core/pull/577
• Convert pretrained word embeddings to Keras Core by @freedomtan in https://github.com/keras-team/keras-core/pull/600
• port gcam backend agnostic by @anas-rz in https://github.com/keras-team/keras-core/pull/601
• Converted NER with transformers example to keras_core by @pranavvp16 in https://github.com/keras-team/keras-core/pull/594
• Add tf.keras backwards compat for nearly all non-experimental symbols by @fchollet in https://github.com/keras-team/keras-core/pull/603
• Cache self.call() signature by @haifeng-jin in https://github.com/keras-team/keras-core/pull/606
• Porting over changes to Torch and JAX keras-core distribution guides … by @hertschuh in https://github.com/keras-team/keras-core/pull/607
• Improve TensorFlowTrainer compatibility for TF<2.9 by @taehoonlee in https://github.com/keras-team/keras-core/pull/598
• Convert RandomZoom to backend-agnostic and improve affine_transform by @james77777778 in https://github.com/keras-team/keras-core/pull/574
• Add the unstack operation to keras core by @tirthasheshpatel in https://github.com/keras-team/keras-core/pull/597
• Port digit addition rnn example to keras-core by @anas-rz in https://github.com/keras-team/keras-core/pull/614
• Set Jit compile only if Model supports it by @sampathweb in https://github.com/keras-team/keras-core/pull/616
• Port big transfer to keras core by @anas-rz in https://github.com/keras-team/keras-core/pull/615
• Implements H5 legacy saving for Keras Core by @nkovela1 in https://github.com/keras-team/keras-core/pull/605
• Add docstrings in numpy by @guillaumebaquiast in https://github.com/keras-team/keras-core/pull/617

New Contributors

• @james77777778 made their first contribution in https://github.com/keras-team/keras-core/pull/477
• @hazemessamm made their first contribution in https://github.com/keras-team/keras-core/pull/490
• @GallagherCommaJack made their first contribution in https://github.com/keras-team/keras-core/pull/462
• @guillaumebaquiast made their first contribution in https://github.com/keras-team/keras-core/pull/558
• @AmedeoBiolatti made their first contribution in https://github.com/keras-team/keras-core/pull/515
• @pranavvp16 made their first contribution in https://github.com/keras-team/keras-core/pull/582
• @cosmo3769 made their first contribution in https://github.com/keras-team/keras-core/pull/527
• @freedomtan made their first contribution in https://github.com/keras-team/keras-core/pull/600
• @taehoonlee made their first contribution in https://github.com/keras-team/keras-core/pull/598

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.2...v0.1.3

Bug fixes and performance improvements.

What's Changed

• Add adagrad for torch by @haifeng-jin in https://github.com/keras-team/keras-core/pull/548
• Add adamax for torch by @haifeng-jin in https://github.com/keras-team/keras-core/pull/549
• Converted UK Ireland accent recognition to work with Keras Core by @fbadine in https://github.com/keras-team/keras-core/pull/537
• port actor_critic_cartpole to keras core by @anas-rz in https://github.com/keras-team/keras-core/pull/542
• Resolve shape via ops in broadcast_to operation by @sampathweb in https://github.com/keras-team/keras-core/pull/547

New Contributors

• @fbadine made their first contribution in https://github.com/keras-team/keras-core/pull/537

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.1...v0.1.2

Highlights

• TensorFlow dependency removed: it is now possible to use Keras Core with JAX or PyTorch without installing TensorFlow.
• Significant performance improvements for Keras optimizers with PyTorch eager execution. Keras optimizers with the PyTorch backend are now at performance parity with native PyTorch optimizers.
• Added the NumPy backend. Note that it does not support training. You can use it to create and call models / layers, write custom evaluation loops, etc.

What's Changed

• bug fix for torch gpu by @haifeng-jin in https://github.com/keras-team/keras-core/pull/438
• Fix typo in README by @abheesht17 in https://github.com/keras-team/keras-core/pull/441
• Use SeedGenerator backend when creating variable by @sampathweb in https://github.com/keras-team/keras-core/pull/439
• Add names to the input tensor, small typo by @Frightera in https://github.com/keras-team/keras-core/pull/449
• Remove unused variables // adam & adamw by @Frightera in https://github.com/keras-team/keras-core/pull/444
• fixed order of parameters in stateless_apply in JAX distributed example by @martin-gorner in https://github.com/keras-team/keras-core/pull/458
• Added JAX distributed training guide. by @hertschuh in https://github.com/keras-team/keras-core/pull/464
• optimize torch performance by @haifeng-jin in https://github.com/keras-team/keras-core/pull/465
• bug fix for seed generator by @haifeng-jin in https://github.com/keras-team/keras-core/pull/471
• Export DTypePolicy as Policy for backward compat by @abheesht17 in https://github.com/keras-team/keras-core/pull/470
• Make it possible to run Keras Core without TensorFlow. by @fchollet in https://github.com/keras-team/keras-core/pull/475
• Weighted metrics without metrics by @mihirparadkar in https://github.com/keras-team/keras-core/pull/474
• Constant Initializer Error fixed by @kamathis4 in https://github.com/keras-team/keras-core/pull/479
• Complete missing docstrings in keras-core/ops/core.py by @Frightera in https://github.com/keras-team/keras-core/pull/494
• Port zero-dce example to keras-core by @soumik12345 in https://github.com/keras-team/keras-core/pull/486
• Updated math.py & nn.py with docstrings. by @Sheiphan in https://github.com/keras-team/keras-core/pull/484
• Add a requirements file for multi-backend cuda by @mattdangerw in https://github.com/keras-team/keras-core/pull/472
• Stop tensorflow from eating all GPU memory by @mattdangerw in https://github.com/keras-team/keras-core/pull/473
• Adding docstrings to Keras Core ops in numpy.py file by @sukhdeepg in https://github.com/keras-team/keras-core/pull/507
• Fix Torch conv transpose by @chenmoneygithub in https://github.com/keras-team/keras-core/pull/520
• Port mirnet example to keras-core by @soumik12345 in https://github.com/keras-team/keras-core/pull/491
• Testing Timeseries anomaly detection Framework agnostic by @anas-rz in https://github.com/keras-team/keras-core/pull/501
• Added docstring ops/image/resize by @Frightera in https://github.com/keras-team/keras-core/pull/509
• Converted to Keras Core: A Vision Transformer without Attention by @anas-rz in https://github.com/keras-team/keras-core/pull/497
• Optimizer torch optimizer performance by @haifeng-jin in https://github.com/keras-team/keras-core/pull/482
• Clean up variable vs weight in base layer by @mattdangerw in https://github.com/keras-team/keras-core/pull/521
• Convert to Keras_Core: Compact Convolutional Transformer by @anas-rz in https://github.com/keras-team/keras-core/pull/523
• Framework agnostic autoencoder for image deniosing by @anas-rz in https://github.com/keras-team/keras-core/pull/524
• Input layer test by @kamathis4 in https://github.com/keras-team/keras-core/pull/511
• Update symbolic_arguments.py by @sqali in https://github.com/keras-team/keras-core/pull/513
• Orthogonal Inititializer test by @kamathis4 in https://github.com/keras-team/keras-core/pull/526
• Convert to Keras Core: Token Learner by @anas-rz in https://github.com/keras-team/keras-core/pull/528
• Add Adam optimizer parallel implementation for torch backend by @haifeng-jin in https://github.com/keras-team/keras-core/pull/531
• Adding: Numpy Backend by @ariG23498 in https://github.com/keras-team/keras-core/pull/483
• Add AdamW implementation with torch parallel ops by @haifeng-jin in https://github.com/keras-team/keras-core/pull/532
• Fix SavedModel integration and add associated tests by @nkovela1 in https://github.com/keras-team/keras-core/pull/522
• Add RMSprop for Torch by @haifeng-jin in https://github.com/keras-team/keras-core/pull/533
• Add Adadelta for torch by @haifeng-jin in https://github.com/keras-team/keras-core/pull/534

New Contributors

• @abheesht17 made their first contribution in https://github.com/keras-team/keras-core/pull/441
• @Frightera made their first contribution in https://github.com/keras-team/keras-core/pull/449
• @fchollet made their first contribution in https://github.com/keras-team/keras-core/pull/475
• @mihirparadkar made their first contribution in https://github.com/keras-team/keras-core/pull/474
• @kamathis4 made their first contribution in https://github.com/keras-team/keras-core/pull/479
• @soumik12345 made their first contribution in https://github.com/keras-team/keras-core/pull/486
• @Sheiphan made their first contribution in https://github.com/keras-team/keras-core/pull/484
• @sukhdeepg made their first contribution in https://github.com/keras-team/keras-core/pull/507
• @anas-rz made their first contribution in https://github.com/keras-team/keras-core/pull/501
• @sqali made their first contribution in https://github.com/keras-team/keras-core/pull/513

Full Changelog: https://github.com/keras-team/keras-core/compare/v0.1.0...v0.1.1

Initial preview release. See the announcement post for details on capabilities, credits, etc.