Vitis AI – Getting Started - LIVE ONLINE
This course introduces the Vitis AI development Toolkit for the AI inference on Xilinx Hardware platforms in conjunction with DNN algorithms, model inference, associated frameworks for model development.
The basics of Machine Learning (ML) and challenges in Neural Network scenarios are revisited. Along with input of pre-trained models either from Tensorflow or Caffe the concepts of the Vitis AI development kit are shown with tools to prune and optimize the trained models. This provides a properly scaled base for mapping. From this the AI Compiler generates deployable code that can then be run on a FPGA fabric microarchitecture.
To efficiently ramp and properly evaluate such a project, Vitis AI development kit tools can be used for analyzing the model performance and debugging.
The theoretical content is supplemented by exercises carried out by the participant.
Duration: 2 mornings of 4 hours each
Times: 9.00 a.m. - 11.00 a.m. Lecture part 1
11.00 a.m. - 11.15 a.m. 15 minute break
11.15 a.m. - 1.15 p.m. Lecture part 2
Exercises: self paced by the participants. Estimated time for completion appr. 2-3 hours. At the end of a lecture, the exercises to be performed by participants are discussed. The results / sample solutions are presented by the trainer the next day.
After registration: the participant receives the presentation documents in electronic form (PDF) as well as the workbook for the exercises, the login data and a list of
requirements to be done in advanced.
- Illustrating the Vitis AI tool flow
- Optimizing a model using the AI quantizer and AI compiler
- Utilizing the architectural features of the Deep Learning Processor Unit (DPU)
- Utilizing the Vitis AI Library to optimize pre-processing and post-processing functions
- Creating a custom platform and application
- Deploying a design
- Introduction to the Vitis AI Development Environment
- Brief Overview of ML Concepts
- Frameworks Supported by the Vitis AI Development Environment
- AI Optimizer
- AI Quantizer and AI Compiler
- AI Profiler and AI Debugger
- Introduction to the Deep Learning Processor Unit (DPU)
- DPU Architectures
- Vitis AI Library for efficient AI Inference with the DPU
- Creating a Custom Hardware Platform
- Vitis AI Quantizer and Compiler
- Creating a VART Application
- Creating a Hardware Platform
- Vitis AI Development Environment
- Xilinx Alveo Accelerator cards, Xilinx Zynq MPSoCs and ACAPs
- Grundkenntnisse über "machine learning"-Konzepte
- Basiswissen in den Programmiersprachen C / C ++ / Python
- Software-Entwicklungsablauf in Vitis