tensorflow2での重みの保存でのエラー

Question

下のように，学習済みのモデルの重みの保存をしようとしているのですが，

model_unet.save_weights("saved_model.ckpt", save_format="tf")

以下のようなエラーが出ます．

Traceback (most recent call last):
  File "main_trained_model_saved.py", line 177, in <module>
    train(parser)
  File "main_trained_model_saved.py", line 142, in train
    model_unet.save_weights("saved_model.ckpt", save_format="tf")
  File "C:\Users\scoop\apps\anaconda3\2020.11\envs\py36_tfgpu\lib\site-packages\tensorflow\python\keras\engine\training.py", line 2046, in save_weights
    self._assert_weights_created()
  File "C:\Users\scoop\apps\anaconda3\2020.11\envs\py36_tfgpu\lib\site-packages\tensorflow\python\keras\engine\training.py", line 2439, in _assert_weights_created
    if self.dynamic:
  File "C:\Users\scoop\apps\anaconda3\2020.11\envs\py36_tfgpu\lib\site-packages\tensorflow\python\keras\engine\base_layer_v1.py", line 850, in dynamic
    return any(layer._dynamic for layer in self._flatten_layers())
  File "C:\Users\scoop\apps\anaconda3\2020.11\envs\py36_tfgpu\lib\site-packages\tensorflow\python\keras\engine\base_layer_v1.py", line 850, in <genexpr>
    return any(layer._dynamic for layer in self._flatten_layers())
AttributeError: 'Model' object has no attribute '_dynamic'

解決策をご教示頂けますと助かります．