Fix warnings

Implement angle function in keras.ops (#21200)

Fix Trainer.get_compile_config base case (empty dict) (#21212)

Recognize /tfhub as a remote location. (#21211)

Adds Support For Custom Call-Context Arguments (#21204)

[OpenVINO Backend] Support numpy min operation (#21168)

Update version number

Fix nightly releases (#21203)

mlx - new image ops implemented (#21202)

Add pre-commit to the common requirements file (#21199)

[Keras 3 OpenVINO Backend]: Support numpy.log1p operation #29487 (#21129

Add openvino to the basic requirements file (#21198)

Add back shell/format.sh, but it just runs pre-commit (#21197)

Redundant imports; no path hacking in package (#21187)

Don't scale gradients if overwriting variable with gradient. (#21193)

Return explicitly layout if already set on variable. (#21194)

Allow Embedding subclasses to only override compute_output_shape. (…

Fix Embedding.compute_output_spec with a non-KerasTensor input. (#…

Bump the python group with 5 updates (#21114)

mlx - test updates and signbit implementation (#21180)

Fix dtype detection for JAX types. (#21184)

Add sparse support to ops.ones_like and ops.zeros_like. (#21181)

[OpenVINO BACKEND] - feat: implement numpy.nonzero for openvino backe…

Add sparse_sigmoid activation (#21175)

Fix Embedding test with ragged tensors on GPU. (#21177)

[OpenVINO Backend]: support numpy.ndim (#21176)

Add tf.RaggedTensor support to Embedding layer. (#21171)

WIP: Add PyTorch backend support for LSTM with CuDNN optimization (#2…

Improve tf.RaggedTensor support in DataAdapters. (#21170)