Generative AI is now at the center of Artificial Intelligence in both academia and industry, with both significant potential and complex challenges. Such a transformative moment is brought by the remarkable advances in parameterizing generative models with deep neural networks. In this graduate course, students will systematically go through deep generative models in detail, equipping themselves with comprehensive knowledge and skills to address various tasks with deep generative models or understand in principle the issues of them. At the core of the curriculum is various representative deep generative models that are commonly used in research and practice, such as diffusion models, generative adversarial networks, variational autoencoders and autoregressive models. In this course we will cover their theoretical principles, implementation details as well as relative pros and cons. In this way students are empowered with a clear understanding of these models and a practical guidance of using them to develop task-specific innovative solutions. On top of a principled journey into the secrets behind generative ai, this course will also demonstrate the frontier of practical application of deep generative models, with case studies in language modeling, art and embodied ai. This enhances their ability to catch up with current circumstances and opportunities ahead. Finally, this course will also place a strong emphasis on the responsible use of deep generative models, by discussing how to evaluate these models, detect their output, and handle their ethical issues. To improve their learning efficiency, students will conduct example hands-on training of various deep generative models utilizing open-source datasets, and at the same time participate in literature survey and presentation of latest academic and industrial advances. In this way their problem-solving abilities will be strengthened, and their vision will be broadened. Upon completion of this course, students are expected to build a scientific and structured understanding of generative ai, and master necessary skills to make their own contributions for relevant academia and industry.