THe code is:

class generatorend(tf.keras.Model): 
def __init__(self): super().__init__(name="generator_ender") 
self.conv1 = tf.keras.layers.Conv2D(480, kernel_size = [3,3], strides =[2,2],activation=tfc.GDN(inverse = True,name="gdn1")) 
self.conv2 = tf.keras.layers.Conv2DTranspose(480, kernel_size = [3,3], strides =[2,2],activation=tfc.GDN(inverse = True, name="gdn1"))
self.acti = tf.keras.layers.Activation(tf.keras.activations.tanh)

def call(self, inputs):
x = self.conv1(inputs)
x = self.conv2(x)
x = self.acti(x)
return x

the code above is tested with:

model = generatorend() 
dummy_input = tf.random.normal([1, 16,24, 8]) 
model(dummy_input)  # This will build the model 
model.save('test_model')  # Now try to save it

The error is :

Traceback (most recent call last):
File "test.py", line 399, in
model.save('test_model') # Now try to save it
File "/usr/local/lib/python3.8/dist-packages/keras/utils/traceback_utils.py", line 70, in error_handler
raise e.with_traceback(filtered_tb) from None
File "/usr/lib/python3.8/contextlib.py", line 120, in exit
next(self.gen)
File "/usr/local/lib/python3.8/dist-packages/tensorflow_compression/python/layers/gdn.py", line 381, in call
if not callable(self.alpha_parameter) and self.alpha == 1 and self.rectify:
tensorflow.python.framework.errors_impl.OperatorNotAllowedInGraphError: Exception encountered when calling layer 'gdn1' (type GDN).

Using a symbolic tf.Tensor as a Python bool is not allowed: AutoGraph did convert this function. This might indicate you are trying to use an unsupported feature.

Call arguments received by layer 'gdn1' (type GDN):
• args=('tf.Tensor(shape=(None, None, None, 480), dtype=float32)',)
• kwargs=<class 'inspect._empty'>`
tested with simple code:
however, encountered with :

It is clear that all params stays the same and the only difference is conv2d and conv2dTranspose. Thus i think it may be a bug

I want to use tfc.GDN in my DL based image compression, however, i find it hard to work well with conv2dTranspose. I was wondering if there is anyone who encountered the same? it is an offical issue?
versions:
tensorboard 2.12.3
tensorboard-data-server 0.7.2
tensorboard-plugin-wit 1.8.1
tensorflow 2.12.1
tensorflow-addons 0.20.0
tensorflow-compression 2.12.0
tensorflow-datasets 4.9.2
tensorflow-estimator 2.12.0
tensorflow-io-gcs-filesystem 0.31.0
tensorflow-metadata 1.14.0
tensorflow-model-optimization 0.8.0
tensorflow-probability 0.20.0