The book 'Artificial Intelligence and Machine Learning for Programmers in PyTorch' by Laurence Moroney covers a broad spectrum of AI development, moving from basic neural network construction to modern generative models. The author employs a code-first approach, starting with simple PyTorch projects like image classification before progressing to data handling and natural language processing.
The curriculum includes traditional sequence modeling techniques such as RNNs and LSTMs, as well as time-series analysis using NASA weather data. This foundational work serves to illustrate the evolution of AI before introducing more complex architectures like Transformers and the Hugging Face ecosystem.
A significant portion of the book focuses on practical deployment, detailing how to serve models using TorchServe and Flask. Moroney also addresses real-world application scenarios by exploring fine-tuning, Retrieval-Augmented Generation (RAG), and running local LLMs via Ollama.
The final chapters shift focus to generative image models, covering diffusion techniques and the Diffusers library. The book concludes with a practical guide on using LoRA to adapt large pre-trained models to specific datasets, spanning over 400 pages of technical content.
