Kaggle versus Gradient / other advice to the "younger" you?

Just stepping in to the course.

I’ve watched the first video, read the first course pages and feel a bit overwhelmed by the number of links as well as the choices.

Appreciate a couple of pointers:

(1) We’re told:

Instead of Kaggle, you can also use Paperspace Gradient.

Which one should I use? I’d rather use a system that is an industry standard; on the other hand, I’d prefer the system that minimizes “sysadmin” complexity.

(2) Same with the choice of Google Colab versus nbviewer.

(3) A more general question: those who have recently finished the first one or two or three lessons, what specific bits of knowledge or advice do you have for “the younger you” - starting out a few weeks previously?


Hey there,

There are lots of options to you to start, that is true, but they vary in the complexity of the setup.
You have to make you own choice depending on how comfortable you feel with python (dependency management there can be a pain), jupyter notebook and ssh.

For new beginners I would suggest just using Google Colab.

For people who already know their way around, I would say, get a VM on Paperspace and set it up by following the steps in Lesson 0 or one of the forum threads.

:exclamation: But the most important is just to get you experimenting with notebooks and training models as quickly as possible. You can always upgrade your tools later :wink: