NotImplementedError: Saving the model to HDF5 format requires the model to be a Functional model or a Sequential model. It does not work for subclassed models, because such models are defined via the body of a Python method, which isn't safely serializable. Consider saving to the Tensorflow SavedModel format (by setting save_format="tf") or using save_weights.

[nicewinter]

When running the notebook titled "vae_faces.ipynb", following code threw a '' error. Note I have created a virtual env. with Python=3.10 and used 'pip install -r requirements.txt' to install all dependencies in this repo before running the notebook.

vae.fit(
train,
epochs=EPOCHS,
callbacks=[
model_checkpoint_callback,
tensorboard_callback,
ImageGenerator(num_img=10, latent_dim=Z_DIM),
],
)

Epoch 1/10
1583/1583 [==============================] - ETA: 0s - loss: 93.3229 - reconstruction_loss: 62.5427 - kl_loss: 16.0481

NotImplementedError Traceback (most recent call last)
Cell In[21], line 1
----> 1 vae.fit(
2 train,
3 epochs=EPOCHS,
4 callbacks=[
5 model_checkpoint_callback,
6 tensorboard_callback,
7 ImageGenerator(num_img=10, latent_dim=Z_DIM),
8 ],
9 )

File ~\AppData\Local\anaconda3\envs\genai\lib\site-packages\keras\utils\traceback_utils.py:70, in filter_traceback..error_handler(*args, **kwargs)
67 filtered_tb = _process_traceback_frames(e.traceback)
68 # To get the full stack trace, call:
69 # tf.debugging.disable_traceback_filtering()
---> 70 raise e.with_traceback(filtered_tb) from None
71 finally:
72 del filtered_tb

File ~\AppData\Local\anaconda3\envs\genai\lib\site-packages\keras\saving\save.py:153, in save_model(model, filepath, overwrite, include_optimizer, save_format, signatures, options, save_traces)
144 if (
145 save_format == "h5"
146 or (h5py is not None and isinstance(filepath, h5py.File))
147 or saving_utils.is_hdf5_filepath(filepath)
148 ):
149 # TODO(b/130258301): add utility method for detecting model type.
150 if not model._is_graph_network and not isinstance(
151 model, sequential.Sequential
152 ):
--> 153 raise NotImplementedError(
154 "Saving the model to HDF5 format requires the model to be a "
155 "Functional model or a Sequential model. It does not work for "
156 "subclassed models, because such models are defined via the "
157 "body of a Python method, which isn't safely serializable. "
158 "Consider saving to the Tensorflow SavedModel format (by "
159 'setting save_format="tf") or using save_weights.'
160 )
161 hdf5_format.save_model_to_hdf5(
162 model, filepath, overwrite, include_optimizer
163 )
164 else:

NotImplementedError: Saving the model to HDF5 format requires the model to be a Functional model or a Sequential model. It does not work for subclassed models, because such models are defined via the body of a Python method, which isn't safely serializable. Consider saving to the Tensorflow SavedModel format (by setting save_format="tf") or using save_weights.