-
Notifications
You must be signed in to change notification settings - Fork 40
Insights: keras-team/tf-keras
Overview
Could not load contribution data
Please try again later
2 Pull requests merged by 1 person
-
Disallow pickle loading in npz files.
#849 merged
Jun 3, 2025 -
Adds support for Call-Context Arguments
#843 merged
May 21, 2025
3 Pull requests opened by 3 people
-
Bump protobuf from 4.21.6 to 4.25.8
#850 opened
Jun 17, 2025 -
Backport implementation of precision-recall gain AUC (Keras PR #21370)
#851 opened
Jun 17, 2025
30 Issues closed by 13 people
-
Model stops training with variable-size dataset
#131 closed
Jun 19, 2025 -
MultiHeadAttention quietly ignores masking when annotated with tf.function
#56 closed
Jun 17, 2025 -
Optimizers no longer support trainable variables with partially known shapes
#161 closed
Jun 17, 2025 -
Keras generates non-unique names for stateful LSTM's states
#283 closed
Jun 14, 2025 -
layers.pop no longer removes the last layer
#75 closed
Jun 14, 2025 -
Keras bisect and editable install
#86 closed
Jun 13, 2025 -
Model Loading Fails
#106 closed
Jun 13, 2025 -
Feature: Flops calculation
#6 closed
Jun 12, 2025 -
Support gradient checkpointing for functional models via the standard fit method
#98 closed
Jun 12, 2025 -
Tensorflow/Keras 2.10 initializer randomness changes : How to get 100% reproducible results?
#432 closed
Jun 12, 2025 -
Toggle aggregated batch losses with tf.keras.callbacks on_train_batch_end
#393 closed
Jun 11, 2025 -
Spectral Normalization layer does not work
#816 closed
Jun 7, 2025 -
Neural Network generates zeros as predictions
#82 closed
Jun 6, 2025 -
Cannot create MultiHeadAttention layer from config with weights
#157 closed
Jun 6, 2025 -
Running mean and variance with `tf.keras.layers.Normalization`
#344 closed
May 31, 2025 -
keras metrics will not accept complex data
#85 closed
May 30, 2025 -
Keras functional API can not save right weights in h5 files
#40 closed
May 28, 2025 -
Importing `tf_keras` to use Keras 2 in TensorFlow 2.16 fails
#756 closed
May 28, 2025 -
tf.keras.losses.SparseCategoricalCrossentropy has logical/Doc bug
#846 closed
May 27, 2025 -
metrics.reset_state ignores shape of weights
#90 closed
May 24, 2025 -
Unable to restore a layer of class TextVectorization - Text Classification
#116 closed
May 22, 2025 -
model doesn't learn when using shuffle='batch'
#120 closed
May 22, 2025 -
Slow training for tf.keras.layers.Embedding with the new optimizers
#45 closed
May 22, 2025 -
Model.load_weights: NotFoundError if have brackets in name and no extension
#353 closed
May 20, 2025 -
unable to save function, WeightNormalization with Conv1DTranspose error
#327 closed
May 20, 2025 -
`Cannot iterate over a scalar tensor.` with split_dataset
#17 closed
May 20, 2025
2 Issues opened by 1 person
-
Inconsistent results when running tf.keras.layers.InputLayer
#848 opened
May 27, 2025 -
tf.keras.losses.SparseCategoricalCrossentropy has logical/Doc bug
#847 opened
May 27, 2025
55 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.
-
Embedding layer after a Dense layer raises UserWarning: Gradients do not exist
#826 commented on
May 21, 2025 • 0 new comments -
Reshape layer drops mask from previous layers
#134 commented on
May 28, 2025 • 0 new comments -
Tokenizer converts padding integers to OOV when oov_token is not None
#648 commented on
May 28, 2025 • 0 new comments -
keras.models.load_model resets the optimizer's state
#70 commented on
Jun 7, 2025 • 0 new comments -
tf.keras computes incorrect loss values with Masking
#130 commented on
Jun 7, 2025 • 0 new comments -
Extra GPU-CPU memory transfer when broadcasting operations between integer tensors
#132 commented on
Jun 10, 2025 • 0 new comments -
Memory leak in saving and loading a keras model containing CategoricalEncoding and Lookup layers
#121 commented on
Jun 10, 2025 • 0 new comments -
Documentation for GitHub Codespaces
#59 commented on
Jun 11, 2025 • 0 new comments -
Models with custom metrics cannot be loaded from trace
#642 commented on
Jun 11, 2025 • 0 new comments -
SparseKerasTensor looses shape after casting
#91 commented on
Jun 11, 2025 • 0 new comments -
Cannot Build Intermediate Model to Nested Layers
#643 commented on
Jun 11, 2025 • 0 new comments -
Addition of Reflect Padding Functionality for tf.keras Convolutional Layers
#766 commented on
Jun 11, 2025 • 0 new comments -
Support for StableHLO generation with JAX backend
#26 commented on
Jun 11, 2025 • 0 new comments -
Feature request: Add a "pandas field selection layer" to allow saving a specification of what inputs are needed in what order / what the output of the model corresponds to
#34 commented on
Jun 11, 2025 • 0 new comments -
Print warning when using `tf.keras.Model.fit()` with a `tf.data` dataset and `batch_size` argument is supplied
#96 commented on
Jun 11, 2025 • 0 new comments -
Training (or Start/End of epoch) signal to Loss function
#178 commented on
Jun 11, 2025 • 0 new comments -
Can only set jit compile WA for CPU device?
#97 commented on
Jun 11, 2025 • 0 new comments -
L-BFGS-B optimizer in Keras
#271 commented on
Jun 11, 2025 • 0 new comments -
How to create `tf.keras.layers.Input` without batch_size dimension?
#273 commented on
Jun 11, 2025 • 0 new comments -
Missing methods to easily access reset_state and states within keras.Model
#61 commented on
Jun 11, 2025 • 0 new comments -
Non adaptable FeatureSpace features
#302 commented on
Jun 11, 2025 • 0 new comments -
Serializing a `tf.keras.Model` in half-precision
#65 commented on
Jun 11, 2025 • 0 new comments -
image_dataset_from_directory supporting reshuffle_each_iteration
#354 commented on
Jun 11, 2025 • 0 new comments -
Persist best epoch per trial
#367 commented on
Jun 11, 2025 • 0 new comments -
BinaryCrossentropy should support a way to reduce over NO axes
#378 commented on
Jun 11, 2025 • 0 new comments -
request the adoption of the tf.io dependency to extend the list of supported images in this API.
#386 commented on
Jun 11, 2025 • 0 new comments -
Feature request: image_dataset_from_directory to provide 'test' option in subset
#402 commented on
Jun 11, 2025 • 0 new comments -
Using output names instead of node names for loss, exporting and serving
#528 commented on
Jun 11, 2025 • 0 new comments -
Class-Balanced Loss based on Effectivene number of examples
#538 commented on
Jun 11, 2025 • 0 new comments -
MobilenetV3 hard_sigmoid could be reimplemented using only keras.layers.
#105 commented on
Jun 11, 2025 • 0 new comments -
Support custom signatures= in tf.keras.callbacks.ModelCheckpoint
#80 commented on
Jun 11, 2025 • 0 new comments -
Gradient accumulation support?
#107 commented on
Jun 11, 2025 • 0 new comments -
steps_per_execution autotune
#593 commented on
Jun 11, 2025 • 0 new comments -
Add MLP-Mixer models `keras.applications`
#109 commented on
Jun 11, 2025 • 0 new comments -
Feature request: optimizer support with slot variables in different dtype to the variable.
#629 commented on
Jun 11, 2025 • 0 new comments -
Complex-valued variance scaling initializers
#111 commented on
Jun 11, 2025 • 0 new comments -
`convolution_op` method for ConvNDTranspose
#93 commented on
Jun 11, 2025 • 0 new comments -
Unroll factor for keras.layers.RNN or performance fix for TensorArray / while_loop.
#92 commented on
Jun 11, 2025 • 0 new comments -
Setting `is_adapted` to True when PreprocessingLayer is initialized with parameters
#78 commented on
Jun 11, 2025 • 0 new comments -
[Contributors Wanted] A real-world model benchmark
#637 commented on
Jun 11, 2025 • 0 new comments -
Scipy affine transform
#89 commented on
Jun 11, 2025 • 0 new comments -
[Contributors Wanted] Better `sample_weights` test coverage for Metrics
#117 commented on
Jun 11, 2025 • 0 new comments -
Support relative (in addition to absolute) `min_delta` parameters in `keras.callbacks.ReduceLROnPlateau`
#119 commented on
Jun 11, 2025 • 0 new comments -
Add max false positive rate option to tf.keras.metrics.AUC
#170 commented on
Jun 12, 2025 • 0 new comments -
Fitting with generators and sample weights crashes when batch size varies between steps with 3D output
#483 commented on
Jun 12, 2025 • 0 new comments -
Duplicate logging messages after calling `model.save()`
#104 commented on
Jun 13, 2025 • 0 new comments -
image_dataset_from_directory uses wrong directory when labels is list
#69 commented on
Jun 13, 2025 • 0 new comments -
TensorFlow 'NoneType' object is not subscriptable during fit()
#41 commented on
Jun 14, 2025 • 0 new comments -
Model fails to run on multi output model using keras.utils.Sequence dataset
#160 commented on
Jun 14, 2025 • 0 new comments -
[Contributors wanted!] A model debugging utility
#101 commented on
Jun 14, 2025 • 0 new comments -
InvalidArgumentError: Graph execution error:
#66 commented on
Jun 14, 2025 • 0 new comments -
Load Model using load_model throws error when there is a Lambda layer in the model
#345 commented on
Jun 15, 2025 • 0 new comments -
How to create a dynamic embedding layer in keras ?
#646 commented on
Jun 15, 2025 • 0 new comments -
Feature request: re-add activation histograms
#118 commented on
Jun 16, 2025 • 0 new comments -
ONEDNN - Layer Normalization error
#397 commented on
Jun 16, 2025 • 0 new comments