This 10-day full-time instructor-led deep-dive course for coders consists of three parts. The first part, an overview over deep learning and deep neural networks, which problems they are applicable to, how they work and how they are implemented on a very high level (using pytorch building blocks to be precise) on day 1. The second part teaches how to code deep learning using deep neural networks efficiently for various problem settings such as image classification, multi-class classification, tabular data, audio, image segmentation, superresolution, neural style transfer, GAN and NLP on days 2-6. The third part re-creates large parts of fast.ai and pytorch as an optional module for those who want to dive deep into the inner workings of deep learning during days 7-10. The days of this course are structured such that the mornings consist of recorded lectures presenting the jupyter notebooks with the course contents and the afternoons consist of paper reading and presentation groups (reading several of the original seminal and brand new publication that drive the field), code presentation groups and guided coding. Participants are encouraged to apply the learned content on their own datasets or rehearse or prepare materials during the evenings.