Rocky start in learning Tensorflow 2.5

Started learning Python and Tensorflow about a month ago, reading from the web. Like Python as it is intuitive. Tensorflow is another animal! Too much syntax to get it right.

Using Pycharm 21.2, Python 3.9 and Tensor Flow 2.5.0.

Reading “Learn TensorFlow 2.0 Implement Machine Learning And Deep Learning Models With Python”. I tried the programs in the book to learn. But get errors in running them. Ditto with programs from the web. Used compat.v1 for the different versions but still get errors here and there. Try this one from the book and see if you can it running without any error.

A. Maybe it is IDE dependant? Hope not.

B. Easy to understand book on Tensorflow?

C. New and better Tensor Flow 2.6 coming soon?

D. Best alternative to Tensor Flow? Sci-kit?


The bast place to start for modern examples is and, we work to keep those examples tested and up to date with best-practices.

That example runs fine in Colab, what error are you getting?

  • compat.v1
  • tf.estimator

If you’re learning, or writing new code: don’t use these.

If you need tree models see: TensorFlow Decision Forests, which implements them in Keras.


Noted and many thanks.

I will try the tutorial and guides you mentioned.