Welcome to the "Sustainability in the Digital Age" series
In an era where digital technologies are reshaping industries and daily life, the environmental impact of AI systems has become a growing concern. This course explores efficient AI methodologies to address these challenges. From deep learning model compression to low-bit quantization and collaborative inference, we delve into techniques that enhance computational efficiency and reduce energy consumption. In Week 2, we focus on low-bit quantization specifically for large language models (LLMs), showcasing cutting-edge open-source tools and models. Join us to learn how to build sustainable AI systems while pushing the boundaries of innovation.
This course is part of the Sustainability in the Digital Age series, a collaborative project between colleagues from Stanford University, SAP and the Hasso Plattner Institute.
Kurssprache: English
Advanced
Kursinformationen
Is this course for me?
Prerequisites
- Basic understanding of Machine Learning and Deep Learning principles
- Proficiency in Python programming
- Familiarity with neural networks is recommended
Knowledge
There two preparatory course options we found, which should be suitable in preparation for this course:
More comprehensive, in German: Praktische Einführung in Deep Learning für Computer Vision
Focused on the basics, in English: Week 5 of CS50’s Introduction to Artificial Intelligence with Python
Time required: The course runs for two weeks with a total workload of approximately 6-10 hours.
All learning materials (Video lectures and slides, additional reading materials and case studies, self-tests) are available from the start of the course. The final exam is activated at the end of the first week and remains open until the course ends, giving participants two weeks to complete the content and one week for the exam.
Was Teilnehmende lernen werden
- Efficient deep learning techniques with a focus on sustainability
- Principles of model compression and low-bit quantization
- Collaborative inference strategies to optimize resource usage
- Application of low-bit quantization for large language models (LLMs)
- Overview and practical use of open-source tools for efficient AI
Für wen dieser Kurs gedacht ist
- Students
- Professionals
- Lifelong learners
Für diesen Kurs einschreiben
Anforderungen für Leistungsnachweise
- Den Leistungsnachweis erhält, wer in der Summe aller benoteten Aufgaben mindestens 50% der Höchstpunktzahl erreicht hat.
- Die Teilnahmebestätigung erhält, wer auf mindestens 50% der Kursunterlagen zugegriffen hat.
Mehr Informationen finden Sie in den Richtlinien für Leistungsnachweise.
Dieser Kurs wird angeboten von
PD Dr. Haojin Yang is a senior researcher and multimedia and machine learning (MML) research group leader at Hasso-Plattner-Institute (HPI). Since 2019, he has been habilitated for a professorship. His research focuses on efficient deep learning, model acceleration and compression, and AI agentic systems.