How confident can you pass the TensorFlow Developer Certificate exam?

I just finished the specialization. Honestly, I think this specialization is quite easy. Some reviewers said some topic like wavenet is not covered. Does this specialization cover enough material for me to take the certificate exam?

Here’s the exam handbook

1 Like

Thanks, I know there is a handbook. Just curious about personal exam experience.

and so @lumuse ?

Hi @Jose_Antonio_Garcia,
Thanks for the great question,
I have been trying to write about my experience taking the specialization then I came across your question so now I have a reason to get some time doing that.

Let me start by the TLDR; This specialization is enough to pass the exam if you have a good grasp on the concept covered in this specialization. If you have done this work before and in various occasions you had to tweak things for the model to get a better results or a at least a result, the answer is Yes. On the other hand, if this is your first time going over TensorFlow and its applications, I would advice you to read couple of books on the topic and try to practice the concepts to pass the test.

For me, I used Tensorflow in some projects and I have tried the JS library of Tensorflow but nothing advanced in Tensorflow. I am not a machine learning scientitst or engineer but I work with data and I use other technologies more often.
I finished the tensorflow developer specialization in couple of weeks given that I was studying for about 3 hours a day. I read chapters on the some of the concepts but nothing major. I also watched the MIT intro to Deep Learning which I really recommend (link).

I took the exam afterwords but I had a difficulty passing the exam since it starts easy but it progresses rather difficultly. As of now, and you can check the exam handbook, there are five parts of the exam but based on my knwoeldge and experience, I had trouble during the fourth and fifth parts since it was not easy or natural to me to converge or get a satisfying error rate. After some thinking and checking resources online, I modifed my code and passed the exam.
I took me almost three hours (of five allowed) to pass the exam.

My advise:

  1. Practice using PyCharm and set it up the way Tensorflow recommends. (Most time consuming part).
    2.A. Prepare yourself to navigate through Tensorflow Documentation and its datasets.
    2.B. Be familiar with the datasets on Tensorflow documentation.
  2. Try to review examples or tutorials on the exam topics and be sure of using the tensorflow docs datasets.
  3. Have a good night of sleep and good breakfast if you like that.
  4. Set up your environment for a long test. (3-5 hours )

Please feel free to ask about the topics and if you needed help setting up PyCharm the recommended way,

Good luck,