Cheat sheets based on each chapter of Chip Huyen's book "Designing Machine Learning Systems"
As highlighted by the author this book is not meant for people who wish to learn machine learning. It's a timely description of the Machine Learning in production landscape. The presented description of machine learning system design is modular, where each module roughly corresponds to a chapter while insisting on mutual influence between different modules and on the fact that putting machine learning systems to production is an iterative process.
The book is written in a light and often humourous tone, while providing quality content and extensive descriptions with numerous relevant examples spanning all of a machine learning system's lifecycle.
The author, Chip Huyen, is included among Linkedin Top Voices in Software Development (2019) and Top Voices in Data Science & AI (2020). She also teaches CS 329S: Machine Learning Systems Design at Stanford University. The book summarized in these cheat sheets is also an Amazon #1 bestseller in Artificial Intelligence. Given the author's credentials and the book's popularity, future MLOps practices should be greatly influenced by the content of this book.
I made these cheat sheets in order to better navigate and memorize all the knowledge in DMLS. They contain the main information presented in every Chapter without the full arguments, examples and rationales. Here's a few things to note about their content :
- They represent my own subjective views on what should be remembered from the book.
- They are cheat sheets, not summaries: You need to read the book to get the full argument, then use these to navigate what you learned.
- These cheat sheets were not proofread by the author, nor do they live up to their writing/formatting standards.
Each PDF hereabove summarize one chapter from the book as a Mind Map. In short, Mind Maps are knowledge management tools used for brainstorming ideas, transcribing thought processes, and just generally putting down thoughts in an intuitively browseable manner. To make them, I used FreePlane, an open-source MindMapping tool which is pretty intuitive and fast to get a hang of. The .mm files are native FreePlane mindmaps which can be modified through it. Here's an example view of one of the Mind Maps:
I may correct some typos or add some details I missed throughout the Chapters. However, if anyone feels like they would like to append their own knowledge to the diagrams, I will add a new shape to the diagram legends (next to examples and tools) as a marker for knowledge that's not part of the book.