Applied AI & AI in business /

The next step after taking an online deep learning course

March 24/5 min read
  • Reynaldo Boulogne
    Reynaldo Boulogne

So you finished your deep learning online course and realized that you don’t have the computing power or the necessary coding skills needed to practice deep learning on your own. 

Does this sound familiar? Then this might be the article for you.

So, you've just finished your online deep learning course (maybe it was Andrew Ng's famous Deep Learning Specialization?)... You’ve watched videos and lectures, you’ve read the books and perhaps even some papers. You understand how AI and deep learning actually work under the hood now and you’re ready to sink your teeth into some real-life problems to really hone your skills and go from theory to practice.

But then it hits you:

  1. You don’t have the computing power to train most models on those large image or text datasets.
  2. If you can train a model, your model tuning process is slow because you can only run one experiment at a time, instead of being able to run multiple experiments in parallel. It can take you days to get a refined model even on the simplest of deep learning datasets.
  3. Each time you try to tackle a new deep learning idea you spend countless hours on stuff that is not related to deep learning. Things such as: 
    trying to learn a number of  data science libraries to do something very specific for the problem you’re trying to solve
    time spent debugging your code
    looking for implementations of the deep learning models that you want to try (or maybe even spending time on implementing them yourself)

"Once I faced reality, I hit the wall"

If this rings a bell then you’re not alone. When I started my deep learning journey, these were exactly the kind of problems I ran into and I’m guessing many others too. Once I faced the reality outside of the curated examples of the online course I hit the wall, so to speak. There was a steep learning curve if I really wanted to work on deep learning problems using open source libraries and I needed to find access to a GPU that wasn’t expensive and didn’t require a lot of setup or maintenance work from me.

This is why I believe  the Peltarion Platform is a really valuable tool to use, and a tool I’d wish I’d had myself a few years ago. 

Let me tell you why.

Practice makes perfect

A crucial part of learning any new skill is practice. Whether you want to increase your experience, learn by doing or try your hand at practical deep learning, I believe the platform is a great place for that because of 3 reasons. 

  1. It makes your work-life easier - You can skip the hassle of learning and managing software libraries and the same goes for the hardware. The platform gives you access to the storage, computation, modelling and deployment resources you need – but without the tedious and oh so boring setup work.
  2. It encourages experimentation - You can explore, tweak and run several models at the same time. In that way, you’ll learn faster what works and what doesn’t for the data, problem or application at hand, rather than waiting for each experiment to finish. 
  3. It solves the data problem - Start right away by either importing your own data, or, if you don’t have this yourself - choose one of our 30 datasets from our data library and get going.

Ready to start? 

Click here to find out more about the platform and improve your deep learning skills. And if you need some inspiration to get started, check out these deep learning tutorials.

(By the way, the platform is free to try so go ahead and play with the platform before you commit to anything).

    • Reynaldo Boulogne

      Reynaldo Boulogne

      With over 15 years of experience, Reynaldo has worked within the intersection of business and technology across multiple sectors, most recently at Klarna and Spotify. He is passionate about innovation, leadership, and building things from scratch. Reynaldo is also a former Vice-chairman of the Stockholm based AI forum, Stockholm AI.

    02/ More on Business & AI