How to improve after specialization?


I have completed this specialization as well as the GAN specialization that DeepLearning.AI offers. While I enjoyed both courses and felt like they gave a great overview of the field, it is a leap to go from these specializations to being able to translate a research paper to code, or build a new model from scratch. I’ve tried without success for weeks to find a private tutor who is advanced enough to help me bridge the gap in ability. It seems that the pool of people capable of coding models from papers and doing original research is so small that those people tend to be in far too high demand to have time to tutor people, regardless of price.

Does anyone here have recommendations for where I could go to learn advanced PyTorch/Tensorflow, how to build models from scratch, and how to learn how to read a paper like the DeepAR paper or Google’s TFT paper and code the models based on the papers? I will take a look at more advanced courses if anyone has some ideas, or I will pay a 1-on-1 tutor very well to help me through this.

Thank you

1 Like

Hi @Alexander_Valarus,

recently a similar question came up. Feel free to check out this thread. Maybe the TF specialisation or MLOps specialisation might be worth a look: After completing DLS, what's next - #4 by Christian_Simonis

When it comes to understanding and use methods from papers, I can definitely recommend to get your hands dirty and get practitioners experience which is usually something where you will encounter new challenges and lern a lot, e.g. if you apply the method to a new data set or play around with parameters etc.

On this note, this page might be interesting for you: as well as following authors, thought leaders or institutions of your interests on Github for example.

Personally, I guess there is no shortcut but if you can keep the motivation up and practice regularly, I am optimistic you will succeed.
What do you think?

Best regards

1 Like