Build a Large Language Model (From Scratch)

I'd never quite understood how an LLM functions under the hood and what the hell a tensor was... Now I understand it to a pretty deep level. While I haven't built my own LLM (yet), it's made me understand enough about what happens under the hood to massively improve my use of LLM's for coding tasks (in terms of token consumption, but also in terms of right first time code).

Very readable and finally, a book which takes you from zero! Been trying all the code from the repo and all works first time. I looked up this book after watching the author on a Lex Fridman podcast and Lex mentions his book so I knew it must be good. Not disappointed. If you are like me, an engineer, and need to "build it to get it"! This book is for you. No deep maths, you just need to know a bit of python.

Yes, this is an extraordinarily well written book about how to build an LLM. The language is friendly and easy to understand. Each step is profoundly explained. On top, it is absolutely clear that the author loves his work and provides a positive attitude in tackling such a model. I highly recommend this book, it opens all doors into the profound knowledge about LLMs.

An excellent text and resources for understanding LLMs and their NLP applications. Build your own. Learn to load weights from pre-trained models. Fine-tune for Classification or Chatbot.

The most helpful and inspirational NLP book I've read. (I'm currently just over 1/2 way through.) It's lucid and easy to follow for anyone with intermedia Python skills and a basic understanding of NLP. Raschka anticipates what you need to know at any given stage. I did an MSc in NLP/SP a couple of years ago and we went through a lot of these topics and similar code sequences then. But reading this has bedded it all down and helped me to understand how train my own models in the future. Looking forward to reading his upcoming book on reasoning models after I finish this.

Awesome book . Not for someone who wants to go too deep but great to begin with and get good idea of basics .

Cards on the table, I only read the first three chapters. But actually I think these chapters are well written with clear explanations. I really wanted to understand how the attention mechanism works. I've read so many books and blogs but couldn't see it in my mind's eye. This book fixed that for me.

From the blurb: "you'll develop an in-depth understanding of how LLMs work". You won't. He introduces stages of the model build process without explaining what they are for; there is no high level conceptual discussion or insight. There are diversions to work through unnecessary and irrelevant simplified versions of ideas; you then do not know what is actually going on under the hood in the real world (or your own project). There are graphics which require effort to understand in themselves rather than add understanding. There are paragraphs alluding to concepts and ideas not explained or covered. There are simplified examples which don't stand up to scrutiny. There are pages of the obvious while explanation is absent. If you want a book that goes through the steps of calling python code without even API level discussion of those calls, and where the libraries used are not even the mainstream (and available) libraries then this may be for you. I am having to go and research each stage and idea independently to try and get an understanding. Definitely not a "keeper".

A great book to start and understand AI.

One of the best books on the subject of AI and LLMs that I’ve come across so far. The author explains the components and inner workings of an LLM perfectly. By the end of the book, you’ll have a clear understanding of how an LLM and a chatbot work internally. The only prerequisites for this book are a basic understanding of how neural networks are structured and some knowledge of vectors and matrices; with these, all the steps and code examples are easy to follow. I’m already looking forward to the author’s next book on reasoning models.

Contains useful stuff, but the author is very sloppy with numbers, which is very annoying comparing one figure with the next. He often rounds off numbers in a wrong way, especially when they are small ('1') or ('2'). The author prefers printing Python that a LLM can also give me. The structure of the book is not logical. Introduces a loss function without mathematical representation. You will need chatgpt to make shots and let you explain. Even chapter titles are misleading, as chapter 7 'pretraining' - no, it is the first introduction of a specific loss function. Book needs a lot of re-work There is not even one flow diagram representing the total flow through a LLM

Poor printing quality: paper is so thin so one can see letters from back side while reading front side. Also for some reason main cover is not alligned with the rest of the book. Overall impression like it was printed at home

İçerik çok güzel ama ben basım kağıdını beğenmedim