DPU-PYNQ v1.3.0 release

Announcements
1

Dear PYNQ users,

We have DPU-PYNQ officially supporting Vitis AI 1.3.0 now:

This release is compatible with:

  • Vitis AI 1.3.0
  • Vitis / Vivado 2020.2
  • PYNQ v2.6.0 image
  • ZCU104, ZCU111, and Ultra96

Some notes:

  1. From this version going forward, we will deprecate DNNDK support; this will be consistent with Vitis AI.
  2. From Vitis AI, *.xmodel files will be used as DPU models instead of *.elf files. So previous models before v1.3.0 need to be re-trained.
  3. Pytorch models are supported. We will add examples later.
  4. Zynq 7000 is still not supported since there are some issues with VART on armv7l.

Please use the PYNQ discussion forum to report any issue.

Thanks,

PYNQ team

2 Likes
2

Would it be possible to have a YOLO or other object detection example notebook?

1 Like
3

Do you guys plan to integrate vitis AI 1.4 soon?