Tilak

@bvbellomo, To expedite the trouble-shooting process, could you please provide a complete code and the TensorFlow version you are using. Batch normalization applies a transformation that maintains the mean output...

@bvbellomo, I was able to execute the code without any issues. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/b4debb257be9de80ae25bf3395fad1a4/untitled484.ipynb) and let us know if we are missing anything here.

@DaniPol, In order to expedite the trouble-shooting process, could you please provide a code in reproducible format or colab gist. Thank you!

@DaniPol, I was facing a different error while executing the provided code. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/35adc3c2ea43f19ced577c6485024b8e/untitled461.ipynb). Could you provide the dependencies and code in the indentation format....

@mortezamg63 , As mentioned in the above comment, contrib layers are deprecated and remove old `tf.contrib.layers` and replace them with [TF Slim](https://github.com/google-research/tf-slim) symbols. Also check [TF Addons](https://www.tensorflow.org/addons) for other `tf.contrib`...

@gadagashwini, I was able to reproduce the issue on tensorflow [v2.8](https://colab.research.google.com/gist/tilakrayal/5df9137af0838c148edb46e720bb07ba/2-8university_project.ipynb), [v2.9](https://colab.research.google.com/gist/tilakrayal/15494785b3fa1709b7a323e88701b180/2_9university_project.ipynb) and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/2bbd9c2c7032c117f65f216bc4337840/nightlyuniversity_project.ipynb).

@gadagashwini, I was able to reproduce the issue on tensorflow [v2.8](https://colab.research.google.com/gist/tilakrayal/f60312fa161df2b67b32f2f98950dc13/2-8tf_load_bug.ipynb), [v2.9](https://colab.research.google.com/gist/tilakrayal/77caabeae424334c57f6160f7477ab2c/2-9tf_load_bug.ipynb) and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/eff56bc8fa2938378036576007d2d1b1/nightlytf_load_bug.ipynb).

Hi @jessechancy , Could you please share a reproducible code that supports your statement so that the issue can be easily understood. Thank you!

@gadagashwini, I was able to reproduce the issue on tensorflow v2.8, v2.9 and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/070478047f33441a391d0114e83072c4/16732.ipynb).

@gadagashwini, I was able to reproduce the issue on tensorflow [v2.8](https://colab.research.google.com/gist/tilakrayal/3e1bcdb5658c0008b60f1f12b0cd39a7/2-8keras_mha_savedmodel_weight_bug.ipynb), [v2.9](https://colab.research.google.com/gist/tilakrayal/3739623a9addc0a3f67a5a6ef4b829d3/2-9keras_mha_savedmodel_weight_bug.ipynb) and nightly. Kindly find the gist of it [here](https://colab.research.google.com/gist/tilakrayal/2eb502183a26900857495d111a9f7e1d/nightly-keras_mha_savedmodel_weight_bug.ipynb).