I've been using a pretty random mix of things including the PyTorch tutorial, some of the tutorials on how transformers work that got posted here months ago, reading papers, and (of course) asking GPT4. It probably isn't the most efficient way to learn.
I would say that learning how to actually build NNs is likely not that important. What's far more important is to know how to use LLMs as an API or library, which is of course 1% coding because the API is so easy and 99% figuring out what their limits are, how best to integrate them into workflows, how to design textual "protocols" to communicate with the AI, how to test non-deterministic systems and so on. Learning how to train a model from scratch is fun but to get competitive results is too expensive, so pragmatism requires focus on being a user for now.
Use perplexity.ai . Why not use AI to learn AI! The good thing I like about this tool is that it gives citations, so that you can learn further beyond summarization it does.
Any pointers / references / books that you’ve found particularly helpful in your learning journey?
I know about Karpathy’s video series (and accompanying repos). Anything else come to mind? Thanks!