cannot import name 'convert_all_kernels_in_model' from 'tensorflow.python.keras.utils'

[aliavni]

Hi,

I am trying to import ModelTraining with from imageai.Prediction.Custom import ModelTraining and I get the following error:

Traceback (most recent call last):
  File "<input>", line 1, in <module>
  File "/Applications/PyCharm.app/Contents/plugins/python/helpers/pydev/_pydev_bundle/pydev_import_hook.py", line 21, in do_import
    module = self._system_import(name, *args, **kwargs)
  File ".../python3.7/site-packages/imageai/Prediction/Custom/__init__.py", line 4, in <module>
    from ..DenseNet.densenet import DenseNetImageNet121
  File "/Applications/PyCharm.app/Contents/plugins/python/helpers/pydev/_pydev_bundle/pydev_import_hook.py", line 21, in do_import
    module = self._system_import(name, *args, **kwargs)
  File ".../python3.7/site-packages/imageai/Prediction/DenseNet/densenet.py", line 21, in <module>
    from tensorflow.python.keras.utils import convert_all_kernels_in_model
ImportError: cannot import name 'convert_all_kernels_in_model' from 'tensorflow.python.keras.utils' (.../python3.7/site-packages/tensorflow_core/python/keras/utils/__init__.py)

This issue gets resolved when I update all tensorflow.python.keras imports with tensorflow.keras in ImageAI library. For example, I change from tensorflow.python.keras.utils import convert_all_kernels_in_model to from tensorflow.keras.utils import convert_all_kernels_in_model.

Is there a specific tensorflow version ImageAI requires?

Here is some info about my environment:

Python 3.6.10
imageai==2.1.5
tensorflow==2.1.0
Keras==2.3.1

[harshithere]

Currently, ImageAI works on tensorflow version 1.14 (and higher) but is not compactible with 2.0 or higher. Thats the reason you face this issue. Try downgrading your tensorflow version.
I faced similar issues and the downgrade helped

[aliavni]

Thanks @harshithere. Closing this issue now.

[400lbhacker]

on colab... at the following magic line at the begining of your code cell
%tensorflow_version 1.x

[ekesdf]

Update tf to 2.4.0 and it will work and just comment out the line this should help