Deployment on Data Center cards - 3.0 English

Vitis AI User Guide (UG1414)

Document ID
Release Date
3.0 English
  • Copy the compiled model to the work place
    [Docker]$ cp compiled/resnet50_tf.xmodel ~
  • Download the vitis_ai_runtime_r3.0.0_image_video.tar.gz package and unzip it in the docker container.
    [Docker]# cd /workspace/examples
    [Docker]# wget -O vitis_ai_runtime_r3.0.0_image_video.tar.gz
    [Docker]# tar -xzvf vitis_ai_runtime_r3.0.0_image_video.tar.gz -C vai_runtime
  • Environment variable setup in docker container.
    source /workspace/board_setup/vck5000/ DPUCVDX8H_8pe_normal
  • Build and Run the resnet50 example.
    [Docker] cd /workspace/examples/vai_runtime/resnet50
    [Docker] bash -x
    [Docker] ./resnet50 ~/resnet50_tf.xmodel

The result is shown below.

Note: To improve the user experience, the pre-built models are already available in Model Zoo. For more information, see Quick Start For Cloud.