dmm: (Default)
Dataflow matrix machines (by Anhinga anhinga) ([personal profile] dmm) wrote in [personal profile] anhinga_anhinga 2020-02-16 07:20 pm (UTC)

I also did not quite get what's the point until recently. Julia? Julia... whatever... I did not care much. (It does look like MATLAB, actually; its creators all were power users of MATLAB.)

But I have a good friend who is using Julia (because that's what his co-employees are using at work), and recently I got some extra motivation to make an extra effort to look at what he is doing in general, including Julia ecosystem, to see if some of it might be holding the keys for my problems.

And, obviously, machine learning frameworks is one place where Python seems to have an overwhelming advantage, nothing else just comes close to modern Python machine learning frameworks. So when I've seen Julia Flux a few weeks ago, which seems to be so much better than any of Python mainstream frameworks, my whole thinking about Julia changed, and that was because of the thinking in this thread: https://anhinga-anhinga.livejournal.com/84046.html?thread=724558#t724558

And when I looked closer at Julia I've seen the quality of thinking in that community. These people are as strong as Haskell folks in terms of programming language theory and practice, it's just amazing.


Post a comment in response:

This account has disabled anonymous posting.
If you don't have an account you can create one now.
HTML doesn't work in the subject.
More info about formatting