Vitis AI 1.2 release uses container technology to distribute the AI software. The release consists of the following components.
- Tools container
- Runtime package for MPSoC
- Public Github for examples (https://github.com/Xilinx/Vitis-AI)
- Vitis AI Model Zoo (https://github.com/Xilinx/AI-Model-Zoo/)
The tools container consists of the following:
- Containers distributed via dockerhub: https://hub.docker.com/r/xilinx/vitis-ai/tags
- Unified Compiler flow includes
- Compiler flow for DPUCAHX8H (cloud)
- Compiler flow for DPUCZDX8G (embedded)
- Compiler flow for DPUCADX8G (cloud)
- Pre-built conda environment to run frameworks:
- conda activate vitis-ai-caffe for Caffe-based flows
- conda activate vitis-ai-tensorflow for TensorFlow-based flows
- PyTorch-based DPU flows
- Alveo Runtime tools
Runtime Package for MPSoC Devices
- Container path URL: https://www.xilinx.com/bin/public/openDownload?filename=vitis_ai_2020.1-r1.2.0.tar.gz
- PetaLinux SDK and Cross compiler tool chain
- VAI board packages based on 2020.1 release, including Vitis AI new generation runtime VART.
- Models and overlaybins at https://github.com/Xilinx/Vitis-AI:
- All public pre-trained models
- All Zynq UltraScale+ MPSoCs and Alveo overlays
- Scripts are included to automate download and install Models and Overlays.