Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

"Only one DPU kernel" error #153

Closed
jimd6776 opened this issue Sep 20, 2020 · 1 comment
Closed

"Only one DPU kernel" error #153

jimd6776 opened this issue Sep 20, 2020 · 1 comment

Comments

@jimd6776
Copy link

Hello,

I ran the Vitis AI tutorial with an example DenseNetX model successfully for Alveo U50 but when i reach the step for running the application (app_mt.py) it instantly throws the following error:

assert len(subgraphs) == 1 # only one DPU kernel

It seems len(subgraphs) equals 2. Why is that and how I fix this?

Thanks in advance

@qianglin-xlnx
Copy link
Contributor

Hi @jimd6776
Different model will have different number of subgraphs. For DenseNetX model, it has only 1 subgraphs. Following are the logs of running the app_mt.py

/usr/bin/python3 app_mt.py -m model_dir/densenetx.xmodel -t 8

Running on Alveo U50

Command line options:
--image_dir : images
--threads : 8
--model : model_dir/densenetx.xmodel

FPS=1192.25, total frames = 10000 , time=8.3875 seconds
Correct: 9152 Wrong: 848 Accuracy: 0.9152

For the subgraphs more than 1, you can refer to https://github.com/Xilinx/Vitis-AI/blob/master/VART/samples/pose_detection/src/pose.cpp to modify your application.

@jimd6776 jimd6776 closed this as completed Oct 3, 2020
andyluo7 pushed a commit that referenced this issue Dec 17, 2020
* updating the zynq_setup script

*  resolving the compilation issue when targetting dpuv2

* removed patch file and change the zynq_setup.sh tvm installation

* resolving the board setup for zcu104

*  minor documentation changes

* update docker installation for TVM

* downgrade h5py installation to 2.8.0
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants