Does anyone have any guides, tips or primers for Nx (or ML in general)?
I guess specifically things that would be helpful before reading Genetic Algorithms in Elixir. Or does the book contain everything you might need? (If anyone’s interested we will probably run a book club for it soon - PM me if you’re interested, I should have a few copies to give away!)
I’m also in the same boat: I’m really interested in everything ML related, but I’m “just an average trained programmer”, who has forgotten all about matrixes in highschool/university.
Although i really really like the content you’re creating @seanmor5, I can’t seem to grok most of it, given my modest background. Anything targetted to mere mortals would be great. Even if it’s a one year course, I’d probably take it, on my own pace.
Industry standard advice for getting up to speed as quick as possible on ML is the free fastai course: https://course.fast.ai. I’ve heard it’s required for new OpenAI engineers. I found it to be very helpful in explaining most key concepts.
The Hugging Face transformers course is also a great follow-up: Introduction - Hugging Face Course. Given that Bumblebee allows you to use pre-trained Hugging Face models, it’s probably a particularly helpful follow-up to the fastai course.
Of course you’re stuck with using Python if you follow these tutorials but it’s probably the best way to start to get a conceptual understanding of all this stuff before using Nx/Axon/Bumblebee.
I got some familiarity with ML via two books, one by The Pragmatic Programmer and one by no starch press. The first uses Python and the second is a visual approach.
Of course, they have nothing to do with Nx… Maybe @seanmor5 will come with a new book (wishful thinking).
I’m definitely interested to join the book club.
Then, I can only second the book suggested by @bdarla , “Programming Machine Learning”. I’m almost finished it (16 out 20 chapters) and I rewritten all the examples with Nx/Axon so far (althought, I suspect not the best ML code ).
We’ll probably start the Tailwind book club first (let me know if you want to join that as well!) and so hope to get this one started some time after that depending on when we reach at least 10 people interested (currently at 6 for this book)
I have read the tw book already and tbh not a big fan of tw (for my own personal taste, nothing wrong with tw, and the book is great! Remember that DevTalk thread I started?), but yes am totally up for Genetic Algorithm one!