keras-team/keras

Issues

• `keras.ops.repeat` cannot return an exptected shape when `x` is a `KerasTensor` and the `axis` is `None`

  #19821 opened 10 hours ago
  3

• activity_regularizer is omitted from get_config/serialization

  #19820 opened 10 hours ago
  2

• Error in masked BatchNormalization with > 3 dimensions

  #19818 opened 2 days ago
  9

• Merge layers do not pass masks to the `compute_mask` function

  #19817 opened 3 days ago
  1

• `inputs` argument cannot be empty

  #19816 opened 3 days ago
  1

• model.compile(jit_compile=True) for PyTorch has no intended effect (Keras 3.3)

  #19815 opened 3 days ago
  1

• While retraining a model I get worse values for MSE from the last epoch of last trainment to the first of the next

  #19813 opened 3 days ago
  1

• Model attribute stop_evaluating does not work

  #19812 opened 3 days ago
  2

• Unexpected depreciation warning in tensorflow 2.16.1

  #19811 opened 4 days ago
  0

• [Bug] keras Add layer failing to build

  #19810 opened 4 days ago
  0

• Enhancement Request: XLA Compatibility for `keras.layers.Embedding

  #19809 opened 4 days ago
  2

• Unable to load finetuned keras model.

  #19808 opened 5 days ago
  3

• Can't load LayerNormalization with RMS scaling

  #19806 opened 4 days ago
  0

• Clearer errors for `add_metric`

  #19804 opened 6 days ago
  0

• ValueError when trying to load a .keras model created using Functional API

  #19803 opened 6 days ago
  2

• `split_dataset` cannot split a dataset containing nested tuples

  #19797 opened 7 days ago
  0

• Dropout variables are inconsistent across save/load

  #19796 opened 7 days ago
  4

• Unknown reference at progbar.py

  #19795 opened 7 days ago
  1

• TokenAndPositionEmbedding layer warning---Respecting padding mask or destroying it?

  #19794 opened 7 days ago
  0

• model.keras format much slower to load

  #19793 opened 7 days ago
  8

• TimeDistributed layer with nested model no longer working in TensorFlow 2.16.1

  #19792 opened 5 days ago
  0

• [Question] When is it absolutely necessary to use a `Lambda` layer?

  #19791 opened 7 days ago
  1

• Most of the outbound links are broken in Tensorflow documentation

  #19784 opened 2 days ago
  3

• [BUG] Conflicting loss_weights implementation in Keras3 for single output case.

  #19780 opened 10 days ago
  2

• Keras `__init__.py` structure isn't readable by a static type checker

  #19779 opened 10 days ago
  5

• model.predict time stdout

  #19778 opened 10 days ago
  4

• Loading model fails: can only concatenate tuple

  #19776 opened 12 days ago
  1

• ops.linspace broken in Tensorflow when num is a tf.Tensor

  #19772 opened 11 days ago
  4

• model.export() fails to save output layer name in the SavedModel

  #19771 opened 7 days ago
  9

• [BUG] keras.layers.StringLookup and Vocabulary of Tensors

  #19770 opened 10 days ago
  3

• Inconsistent assertion in keras.layers.MultiHeadAttention

  #19769 opened 12 days ago
  0

• Question: Best practice for direct variable initialization

  #19767 opened 13 days ago
  5

• Specifying shape of normal sampling with partially empty keras tensor

  #19766 opened 12 days ago
  3

• compatibility with pytorch 2.3

  #19765 opened 13 days ago
  1

• Getting Nan values in prediction

  #19762 opened 15 days ago
  2

• Please explain label_mode='int' in keras.utils.image_dataset_from_directory

  #19761 opened 15 days ago
  2

• optimizers.Adam no longer accepts tf.Variable

  #19760 opened 15 days ago
  1

• Recommendation for writing tests for ModelParallel distribution

  #19757 opened 16 days ago
  2

• keras.ops.select doesn't accept tuples as input

  #19755 opened 16 days ago
  1

• RNN layer: len is not well defined for a symbolic Tensor. Please call `x.shape` rather than `len(x)` for shape information.

  #19754 opened 17 days ago
  3

• Conv3D generates output on Tensorflow backend while crashes on pytorch backend

  #19753 opened 17 days ago
  3

• Add `on_epoch_begin` to `utils.Sequence`

  #19749 opened 18 days ago
  2

• Custom Generator crash for variable input dimensions

  #19748 opened 14 days ago
  4

• "Keras.wrappers" <module not found error> in 'python IDLE 3.12.0'

  #19747 opened 17 days ago
  4

• I wrote a custom layer like this and i am not confident of serializing this layer finally, how to write this in best way?

  #19745 opened 18 days ago
  2

• ModuleNotFoundError: No module named 'keras.utils.generic_utils'

  #19744 opened 18 days ago
  2

• Unexpected Losses with Sample Weights

  #19740 opened 19 days ago
  1

• Has any plan for int4?

  #19738 opened 20 days ago
  1

• Bidirectional layer rejects custom activations

  #19737 opened 21 days ago
  4

• The .fit() method passes unsized batches to a custom loss function.

  #19732 opened 21 days ago
  3