In artificial intelligence, technical performance plays an increasingly important role. This involves optimising the code, data and models for performance. Among other things, the use of different hardware resources is taken into account.
When using machine learning technologies, it is often an individual solution that has its own characteristics, which has to be configured for a specific use case. Since in some cases a large number of configurable parameters are involved, automated tuning procedures are often used here.
Performed by: SSE
- Introduction to AI and hardware structures
- Memory requirements
- ML Tuning
The aim of the two-day course is to build an understanding of AI performance and give ideas on how to implement the above techniques.