AttributeError:'tensorflow.python.framework.fast_tensor_util' has no attribute 'appendfloat32arraytotensorproto'

Viewed 822

TensorFlow 2.2.0 + Keras 2.3.1 on Python 3.7

How to fix this AttributeError? I reinstall the TensorFlow but it does not work.

enter image description here

enter image description here

1 Answers

I had an exact same issue with Python 3.5.6 :: Anaconda, Inc. but I don't think it's Python's or TF version fault.

Frankly speaking, decided to create a new py-env like this (given that I use conda distro):

Created this env.yml file:

name: tf2
channels:
  - defaults
  - anaconda
dependencies:
  - jupyter
  - matplotlib
  - numpy
  - python=3.7
  - scikit-learn
  - tensorflow

then ran

conda env create -f env.yml 
conda activate tf2
python -m ipykernel install --user --name=tf2
jupyter notebook

and it all works:

enter image description here

Related