5 Pull requests merged by 2 people

• Fix symbolic call of logsumexp with int axis.

  #21428 merged Jun 27, 2025

• Progress bar now handles steps_per_execution.

  #21422 merged Jun 26, 2025

• Fix divide by zero error in math_test.py

  #21419 merged Jun 25, 2025

• Fix JAX GPU tests.

  #21402 merged Jun 24, 2025

• Disable loading functions within deserialization.

  #21412 merged Jun 24, 2025

14 Pull requests opened by 10 people

• update a small feature in __getitem__ for openvino backend

  #21407 opened Jun 21, 2025

• [OpenVINO backend] support tri, triu, and tril

  #21408 opened Jun 21, 2025

• [Keras 3 OpenVINO Backend]: Support numpy.flip operation

  #21410 opened Jun 22, 2025

• [OpenVINO backend] supporting mul with bools

  #21411 opened Jun 22, 2025

• Improve error message for class_weight with PyTorch DataLoaders

  #21414 opened Jun 24, 2025

• `mlx` - more changes to tests to pass CI

  #21415 opened Jun 24, 2025

• Enable ROCm devices to correctly use keras.layers.CuDNNGRU and keras.…

  #21417 opened Jun 24, 2025

• [OpenVINO backend] Remove usage of deprecated `openvino.runtime`

  #21418 opened Jun 24, 2025

• [OpenVINO backend] support bool for convert_to_tensor

  #21423 opened Jun 26, 2025

• Running TPU tests on linux-x86-ct6e-44-1tpu

  #21425 opened Jun 26, 2025

• Fix: Update SparseCategoricalCrossentropy example

  #21427 opened Jun 26, 2025

• Only allow deserialization of `KerasSaveable`s by module and name.

  #21429 opened Jun 27, 2025

• Support non-tf.data.Dataset instances in multi-host setting.

  #21430 opened Jun 27, 2025

• Worth not to hide exceptions

  #21432 opened Jun 28, 2025

10 Issues closed by 4 people

• as_list() is not defined on an unknown TensorShape.

  #20573 closed Jun 28, 2025

• Keras does not save weights properly

  #19693 closed Jun 27, 2025

• ValueError: Passed in object <KerasVariable shape=(1, 64), dtype=float64, path=dense_15/kernel> of type 'Variable', not tf.Tensor or tf.Variable or ExtensionType.

  #19874 closed Jun 27, 2025

• Progress bar showing wrong step number

  #20861 closed Jun 26, 2025

• Inconsistent execution results by the PyTorch backend

  #20235 closed Jun 26, 2025

• ValueError: File not found: filepath=Backend\C4__256g_000040000.keras. Please ensure the file is an accessible `.keras` zip file.

  #20259 closed Jun 26, 2025

• RFC consider pre-commit for linters

  #20604 closed Jun 25, 2025

• MultiHeadAttention's use_causal_mask is broken

  #21284 closed Jun 25, 2025

• loss_weights depending on epoch number

  #20294 closed Jun 25, 2025

• Is this Bug?

  #20984 closed Jun 25, 2025

3 Issues opened by 3 people

• Wrong example for SparseCategoricalCrossentropy

  #21426 opened Jun 26, 2025

• [Feature Suggestion] Improving Built-In Visualization System With Layered-Style Architecture Generation

  #21421 opened Jun 25, 2025

• `keras.ops.image.map_coordinates` breaks with TF backend

  #21420 opened Jun 25, 2025

18 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [OpenVINO Backend] support ops.slice_update

  #21362 commented on Jun 24, 2025 • 1 new comment

• IGNORE THIS PR

  #21389 commented on Jun 24, 2025 • 0 new comments

• Modified compute_output_shape function to handle broadcasting behavior in layers.Rescaling

  #21351 commented on Jun 23, 2025 • 0 new comments

• [OpenVINO backend] Update `ndim` behavior

  #21317 commented on Jun 24, 2025 • 0 new comments

• Implement numpy.prod method for OpenVino backend

  #21282 commented on Jun 24, 2025 • 0 new comments

• Add memory usage monitor callback

  #21245 commented on Jun 23, 2025 • 0 new comments

• DRAFT: Add custom variable updater.

  #21225 commented on Jun 26, 2025 • 0 new comments

• Support numpy.prod operation

  #21188 commented on Jun 22, 2025 • 0 new comments

• LossScaleOptimizer crash under jit_compile=True with MirroredStrategy and mixed_precision

  #21366 commented on Jun 26, 2025 • 0 new comments

• pytorch backend lstm very +10x slow (maybe batch size with pytorch backend has different semantics than the traditional Keras semantics?)

  #20902 commented on Jun 26, 2025 • 0 new comments

• Failed to convert LSTM layers to ONNX on Torch backend

  #21390 commented on Jun 26, 2025 • 0 new comments

• High memory consumption with model.fit in TF 2.15 + Keras 3.0.2

  #19071 commented on Jun 26, 2025 • 0 new comments

• Distributed training with tensorflow doesn't work with Keras 3.7 and above

  #21172 commented on Jun 25, 2025 • 0 new comments

• Unable to see any speedup from multi-GPU training using JAX backend for LSTM / GRU

  #20998 commented on Jun 25, 2025 • 0 new comments

• Weight loading: `skip_mismatch` is not respected for H5 format

  #21404 commented on Jun 24, 2025 • 0 new comments

• Incorrect error message given for named inputs when unnamed input provided to fit()

  #19127 commented on Jun 23, 2025 • 0 new comments

• Documentation: better info about class_weight and/or DataAdapters

  #21355 commented on Jun 23, 2025 • 0 new comments

• Incorrect values returned by the evaluate method in Model API

  #20788 commented on Jun 23, 2025 • 0 new comments