search

Xilinx / Vitis-AI

Vitis AI is Xilinx’s development stack for AI inference on Xilinx hardware platforms, including both edge devices and Alveo cards.
https://www.xilinx.com/ai
Apache License 2.0
1.41k stars 621 forks source link

Module without vitis_ai_library #737

Open RookieCaoChao opened 2 years ago

RookieCaoChao commented 2 years ago

Regarding the resnet.py in the demo that failed to find the vitis_ai_library module, how can I find this module? Code address: https://github.com/Xilinx/Vitis-AI/blob/v1.4.1/demo/Vitis-AI-Library/samples/graph_runner/resnet50_graph_runner_py/resnet50.py or Regarding the python API described in the ug1414.pdf document, there is a program in creat_graph_runner as shown in the figure: image The vitis_ai_library API cannot be found, what is the reason? (The prerequisite for this is that the board has been installed, and it can quantify the module, compile the module, and get the xmodel I want) Thank you in advance

qianglin-xlnx commented 2 years ago

Hi @China-CC-lon Graph_runner API is based on Vitis-AI-Library. So you need to install it.
For VAI2.0, it's libvitis_ai_library-2.0.0-r64.aarch64.rpm You can find the library in: https://github.com/Xilinx/Vitis-AI/tree/master/setup/mpsoc/VART/2021.2 Note that it's based on 2021.2.

RookieCaoChao commented 2 years ago

嗨@China-CC-lon Graphrunner API 基于 Vitis-AI-Library。所以你需要安装它。 对于 VAI2.0，它是 libvitis​​ai_library-2.0.0-r64.aarch64.rpm 您可以在以下位置找到该库：https ://github.com/Xilinx/Vitis-AI/tree/master/setup/mpsoc/VART/2021.2 注意它基于 2021.2。

Hi，qianglin-xlnx OK, thanks for your help! I'll try it right away.

RookieCaoChao commented 2 years ago

Hi @qianglin-xlnx, I am using VCK5000 in Ubuntu20.04 environment, according to the information you provided, it is an RPM file, do you have any deb file suitable for Ubuntu system?

qianglin-xlnx commented 2 years ago

@wanghy-xlnx Could you please help to provide the deb files for VCK5000?

@China-CC-lon You can also try to build the VART source files to generate the deb packages. Refer to: https://github.com/Xilinx/Vitis-AI/blob/master/tools/Vitis-AI-Runtime/VART/quick_start_for_edge.md