11-03-2020 12:23 AM
I am trying to build a custom tensorflow/caffe model for zynq-7000. As there are two paths for this IE. DNNDK or VITIS AI - I went through the tutorials and guides.
As per these documentations they have mentioned it for zcu102/ 104 platform, Is it possible to run an external model on zynq-7000 with DPU ?
if yes what procedure i can follow? and what should be the required file types ? IF We take new dataset then can we train it using VITIS AI for Zynq Platforms?
11-10-2020 07:27 PM
Hi, I also want to get a definite answer to this question, and whether xilinx has successfully deployed Vitis AI on the zynq7000 board.
我也非常想知道这个问题的答案，同时Xilinx 官方人员是否在zynq7000系列的开发板上成功部署过Vitis AI.
我按照https://github.com/gewuek/vitis_ai_custom_platform_flow / https://www.xilinx.com/html_docs/vitis_ai/1_2/gjp1592545687662.html 以及其他流程操作时，总会碰到无法安装packagegroup-petalinux-vitisai 的问题。
11-25-2020 01:21 AM
DNNDK toolset is recommended for Zynq-7000 platform. Here's an AR which may give you some guidelines.
We don't deliver Vitis AI runtime package for the zynq-7000 target, as discussed in a previous post.