TinyML: Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers 1st Edition

This is a fantastic, well-written, highly-entertaining resource for devs of all levels curious about running machine learning models on resource-limited devices and looking to play with edge computing. It goes beyond Google's online documentation and gives practical demos and explanations that make sense.Basically, you can follow along the book by running pre-built notebooks in Google Colab to train ML models, then compiling the code to binaries, which you then flashing to the microcontroller - Arduino, SparkFun Edge, and Stm32f7 Discovery Kit are supported with great instructions for all three platforms.For future versions of the book, I'd like to see:- instructions for those working in Windows. All the makefiles and build scripts are MacOS/Linux, but providing a facility for those working in Windows environments would be nice, too (Windows Subsystem for Linux, Visual Studio's nmake, cygwin, cloud environment, virtual machine, etc.).- notebook locations on Github and Google Colab have moved out of 'experimental' status and so the URLs have changed, so some poking around is required to find the code (not hard - the dedicated notebook for the "Hello, world!: example now lives in the /train directory in the repos).- the book doesn't mention having a serial breakout programmer, just the microcontroller and a USB cable. I had to order the serial adapter separately.Overall, the book is really well laid out with a friendly voice and demos that are truly fun to work through. The approach to running the examples, then explaining the concepts for running ML on embedded environments and underlying C++ constructs is a great way to present the material. I prefer this than the traditional 6 chapters of iterative building and only at the end you arrive at the finished product. This gives you something to play with right away.

The authors have tried to explain a difficult subject in an easy to understand manner. What I like about the book is that after a nice introductory chapters, in chapter 4 they show overview of all the required steps of ML (building the model in Colab , training, testing, converting the model in TensorFlow) in one chapter to help you envision the whole process. They leave the detail of each step for the following chapters. Another important point is for modeling they choose a simple sine function with analytical solution to explain the actual vs predicted values to show how the model prediction can be improved by additional layers and optimization.

Like: the authors provide detailed explanations of code and its purpose (and going beyond to explain the C++ syntax). The detailed explanations are valuable in moving beyond the 'blackbox' approach of 'just do this'. The colab pages were good even though some were out of date using an obsolete tensorflow version.Moderate dislike, but reality: this field is changing so fast and there are so many updates, some git URLs are out of date. Some point to other git URLs. Be prepared to cobble together what's in the book with what's been changed on-line. I suspect this is completely out of the author's controlDislike: Even though this is a project oriented book, it is not possible to complete on a Windows 11 machine even though the opening pages suggest so. A few pages later the authors make the statement that "Git and Make are often preinstalled on modern operating systems". Well yes, on Mac and Linux, but not Windows 11. So yes, on Windows 11 you can manually install git, make, g++, python3, wget, numpy, Pillow, and a unix-type Find (making sure it is invoked before a windows find.exe). And if you do so, you then run into the Makefile making path names incompatible with g++So presumable the way forward on a Windows 10 or 11 machine is to use wsl to load Ubuntu and run under Linux. The authors should have been much clearer on this requirement.

I was curious about this topic in the overall ML domain and I stumbled on this book.The book does a great job to go deep in this topic and I love the approach it took to cover the relevant topics without going to lengthy rudimentary details that might be academic. I like the way chapters are divided into certain deep dive topics on say, how to deploy a model for a board or the code it generates, that you can revisit again later when you feel like. So, as a result, you can skip ahead, learn about a project, and again come back to review the topic to understand the details. I like this hands-on approach of this ML on a small device topic without sacrificing the quick productivity gain out of this book.Excellent jobs from the authors and I am looking forward to their next book!

I am using the TinyML book to develop usable, hands-on competence with Tensorflow and machine learning. The book is a great starting point for learning this technology. You don't need a supercomputer, you can run the programs in this book from your PC connected to very low cost devices from Arduino, SparkFun and other vendors. You don't have to be working in the tiny ML space ( < 1mW, small memory, etc. ) to benefit from this book as the concepts are applicable to whatever the opposite of TinyML is. The contagious enthusiasm of the authors is very impressive. Look for them on youtube, they know this material very well and are very generous with help and support. The book is biased toward how to get things done, and is light on deep theory.