Comment by sarmike31

I kind of want to blanket recommend but that's not very helpful

I would suggest starting with Andrej Karpathy's YouTube video: https://youtu.be/kCc8FmEb1nY?is=oiDsrBYJg_MUUmoD

This video is excellent. I'm a huge fan. Also the video is zero commitment and instantly available which makes it a good way to check you are interested.

The book by Sebastian Raschka is slightly less accessible but very reasonably priced and the experience of working through a book is a lot nicer than skipping back and forth in a video (for me). Sebastian's blog posts on recent architectures are absolutely great too.

I've read Raschka's "Build Your Own LLM from Scratch" book and really enjoyed it. I haven't tried the code yet, but the code from his previous Python ML book worked great.

actually just use chatgpt. There is a new mode of learning thats now avaiable that doesnt require you to read about things that are already discovered leaving you with a shallow knowledge.

you can now play the inventor and start with question "i want to build next token prediction software" and go as far as you can with your current knowledge while brainstroming with chatgpt as a rubber duck.

dont not start with a course on probabality , linear algebra or calculus . do not watch 3 hr videos or 3blue animations .

there is agood video on this way to learn.

https://www.youtube.com/watch?v=cbiyPOn-__M&t=380s