As someone who has been working in ML for years, I can only recommend to stay away from anything recent. Grab an old bayesian statistics textbook and learn the fundamentals, then progress to learning the major frameworks like Pytorch. Try to write every part of a CNN, RNN and Transformer architecture and training pipeline yourself the first time (including data loaders, but maybe leave out CUDA matrix kernels). Stay the hell away from wrappers for other people's wrappers like Langchain. Their documentation is often not just outdated, but flat out wrong regarding the fundamentals. Huggingface is great if you know the basics and thus how to fix things if their standard wrappers break.
You can try Theodoridis if you can find a first or second edition. It is old enough to not be diluted by the recent craze but still recent enough to cover all the necessary fundamentals. There is also a new edition coming out soon, but that seems to have been heavily tainted by the ChatGPT hype.
