My Google ML Bootcamp and Tensorflow Certified Developer Journey
I passed my Tensorflow Certification Exam on 16th Nov 2022. While it costs $100 USD, I had coupon code from Google’s ML Bootcamp so I didn’t had to pay for the exam. My experience in hackathons, internships and following courses helped me clear this exam. The exam really test your model building skills within time constraint, with 5 questions and increasing difficulty.
Google ML Bootcamp 2022
The ML bootcamp was organised by Google this year in July for students through a test. Almost 200 students were selected, including me. The program included free access to Deep Learning Specialization course, TensorFlow Developer Professional Certificate course, Challenges, Tech talks, Office hours, company 1–on-1s and Tensorflow Fecilitation classes. The program lasts for 4 months and I can say it was really helpful for me in developing skills.
Tensorflow Certification Exam
The exam consists of 5 questions testing your skills on building models for regression, classification, forecasting and generation. The datasets can be images, texts or series that also require you to format your inputs and outputs, but the exam will give you instructions for organising data.
I faced difficulty in solving the last or the hardest problem, other than that the verification and installations took a lot more time for me. I used my Ideapad 3 and the GPUs were working fine, however still to manage time I used Google Colab for training another model simultaneoulsy.
and I passed with the following number of test cases:
Question 1: 4/5
Question 2: 5/5
Question 3: 5/5
Question 4: 4/5
Question 5: 5/5
- Practise more problems, and focus on optimisation.
- Focus on reducing loss.
- Keep colab with GPU accelaration handy.
- Use callbacks, early stopping and learning rate scheduler will be helpful.
- Start solving with templates (prepare beforehand).
- You’ve got 5 hours, be calm and utilise every second.
Link to setting up GPU on gaming laptop: https://atufashireen.medium.com/installing-tensorflow-with-gpu-support-on-windows-10-with-nvidia-graphics-card-in-5-steps-4bde9f9dac10