What is your preferred ML development environment?

[kristopher-wood]

Hi all!

I put this in Show and Tell not because I have anything to show, but because I hope you might!

What is your preferred ML Development environment? Which OS, language(s), IDE, libraries, etc make your life easier as an ML developer/researcher?

Currently I'm running Python 3.9 on Windows 10 using Spyder IDE in an Anaconda environment. I'm finding it very difficult to get some things working because I'm constantly fighting with Windows to get the various components to place nice together.

I hope some of you might have found a better way so that I can imitate your configurations!

[MaxHalford]

What exactly are you struggling with? A long time ago I used to develop with a Windows machine, but it wasn't fun. Now I'm on MacOS and it's great. The instructions in the CONTRIBUTING.md file should get do the trick. But I've never tried them on Windows.

[kristopher-wood]

I got chantilly run to work in Windows by running it inside of an Ubuntu for Windows terminal! Now to see if I can get the rest of my dev tools to work in that environment....

[kristopher-wood]

Any tips on getting chantilly working with river instead of creme?

[MaxHalford]

Yes you'd have to change the Chantilly codebase to use learn_one instead of fit_one. But we're not maintaining because we're working on something new with @raphaelsty which will be more delightful and powerful :)

[kristopher-wood]

Do you need more developers for this more delightful something? 😉

I'm new to python but have over two decades professional development experience in general and learn new languages quickly. I’d be happy to contribute!

[MaxHalford]

For sure you'd be welcome to contribute! We first have to work on some core stuff, but once we release an alpha version there'll be loads of stuff to do.