Tensorflow.compat.v2.__internal__.tracking' has no attribute 'TrackableSaver' Error

Question:

I got this error after installing Tensorflow.js. Previously this program was working. Could it be a problem with the versions? I’m really curious as to what’s causing it.
Thanks in advance.

File ~OneDriveMasaüstüBitirme Projeneural_network(sinir_ağları).py:61
    model = build_model()

  File ~OneDriveMasaüstüBitirme Projeneural_network(sinir_ağları).py:29 in build_model
    model = keras.Sequential([

  File C:ProgramDataAnaconda3libsite-packagestensorflowpythontrackablebase.py:205 in _method_wrapper
    result = method(self, *args, **kwargs)

  File C:ProgramDataAnaconda3libsite-packageskerasutilstraceback_utils.py:67 in error_handler
    raise e.with_traceback(filtered_tb) from None

  File C:ProgramDataAnaconda3libsite-packageskerasenginetraining.py:3331 in saver_with_op_caching
    return tf.__internal__.tracking.TrackableSaver(

AttributeError: module 'tensorflow.compat.v2.__internal__.tracking' has no attribute 'TrackableSaver'

I was planning to convert my model with Tensorflow.js and run it over the web. But when I installed Tensorflow.js I got this error in the program.

Asked By: faik mermer

||
Source

Answers:

Update keras with pip install 'keras>=2.9.0' for keras-team/keras@af70910.

-    self._trackable_saver = saver_with_op_caching(self)
+    self._checkpoint = tf.train.Checkpoint(root=weakref.ref(self))
Answered By: aaron

I had to create another environment on Jupyter and re-install all the libraries from scratch since it seemed all the libraries started having issues, not just TensorFlow. It is now working with no errors.

Answered By: hotspice34