Deep Learning with R, Third Edition - François Chollet

Deep learning from the ground up using R and the powerful Keras library! Deep Learning with R, Third Edition introduces deep learning from scratch with examples that use the R language and the Keras library. Each chapter offers practical code examples that build your understanding of deep learning layer by layer. You'll appreciate the intuitive explanations, crisp illustrations, and clear examples. In this expanded third edition you'll find fresh chapters on the transformers architecture, building your own GPT-like large language model, and image generation with diffusion models. Plus, even DL veterans will benefit from the insightful explanations on the nature of deep learning. In Deep Learning with R, Third Edition you will learn: - Deep learning from first principles \n - The latest features of Keras \n - Image classification and image segmentation \n - Time series forecasting \n - Text classification and machine translation \n - Text and image generation--build your own LLMs and diffusion models! \n - Scaling and tuning models For R programmers, the R interface to the Keras deep learning library is a powerful head start on building deep learning models without switching to Python. It provides a simple, consistent API that makes deep learning accessible and simplifies the process of building neural networks, even if you have no prior experience in advanced machine learning. About the book Deep Learning with R, Third Edition introduces R programmers to the latest advances in deep learning. In it, you'll explore how to use Keras 3 and R to build and train deep learning models, all without advanced math or low-level programming. You'll get started on core DL tasks like computer vision and natural language processing, and you'll take your first steps into the world of transformers, LLMs, and the foundations of modern AI. You'll learn to fine-tune and evaluate your models for peak performance, and dive into advanced methods like transfer learning and model interpretability. This expanded third edition brings cutting-edge coverage of transformers, building your own GPT-style language model, and creating images with diffusion models--all in R. About the reader For readers with intermediate R skills. No previous experience with Keras, TensorFlow, Jax, Torch, or deep learning is required. About the author François Chollet is the creator of Keras. Tomasz Kalinowski is a software engineer at Posit Software, PBC (formerly RStudio) and maintainer of the Keras