OutrageousGiraffe8 so basically replacing to:
self.d1 = ReLU()

I'm sorry for late response. You could probably replicate it by instead of using activation="relu" you used new ReLU layer after Dense. Or if this does not do, extract part of the model in separate Sequential model, for example
Sequential([ Dense(128), BatchNormalization(), ReLU(), ])

Hi OutrageousGiraffe8

when I save model using tf.keras.save_model

This should create a new Model in the system (not artifact), models have their own entity and UID.
Are you creating the Task with output_uri=" gs://bucket/folder " ?

self.d1 = Dense(128, dtype=tf.float32) self.d1_a = ReLU()

I found out that if I work with one model than that model is saved but if I work with different model that model is not saved, everything else being the same. Both models use Functional API

Edit: I have further pinpointed a problem to a ReLU layer. If i use
layers.ReLU()it does not work, but if I use
layers.Activation("relu")it works.

OutrageousGiraffe8 this sounds like a bug, how can we reproduce it?
Maybe a add another layer here?
https://github.com/allegroai/clearml/blob/a47f127679ebf5912690f7c3e60791a2daa5c984/examples/frameworks/tensorflow/tensorflow_mnist.py#L40

Hi OutrageousGiraffe8
I was not able to reproduce 😞
Python 3.8 Ubuntu + TF 2.8
I get both metrics and model stored and uploaded
Any idea?