Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.
6026 Discussions

Cannot create ShapeOf layer pool10/input_rank/shape_of id:145 when running ./demo_squeezenet_downloa

Zitong-Xu
Beginner
212 Views

I'm working on Ubuntu 18.04 with NCS

 

I followed instructions in Install Intel® Distribution of OpenVINO™ toolkit for Linux* 

with version: 2020.3

 

Here is my dmesg and lsusb results:

Screen Shot 2022-01-14 at 1.53.12 AM.png

Screen Shot 2022-01-14 at 1.53.43 AM.png

which I think the

Myriad VPU [Movidius Neural Compute Stick] 

is detected in my system.

 

However, I came across this problem while running the demo example:

both cmds are not working

sudo ./demo_squeezenet_download_convert_run.sh

sudo ./demo_squeezenet_download_convert_run.sh -d MYRIAD

Screen Shot 2022-01-14 at 1.58.32 AM.png

 

[ INFO ] Loading network files
[ ERROR ] Cannot create ShapeOf layer pool10/input_rank/shape_of id:145
Error on or near line 217; exiting with status 1

 

I saw similar problems in the Forums, I thought perhaps because I do not have IE Plugin or IE core in my ubuntu 18.04 system, but I have to use 2020.3 instead of the latest version since I am using NCS.

 

Could you help me solve this problem? Thanks a lot.

0 Kudos
3 Replies
Peh_Intel
Moderator
187 Views

Hi Zitong,


Thanks for reaching out to us.


The encountered error is due to the mismatch of Intermediate Representation (IR) version and OpenVINO™ version.


I believe that you’ve run the validation demo (demo_squeezenet_download_convert_run.sh) in the later OpenVINO™ version before. Hence, the later model version had been downloaded and converted into IR, and then saved into the /home/<username>/openvino_models directory.


After that, when you run the validation demo in OpenVINO™ 2020.3, the steps of downloading and converting model will be replaced by loading the IR from cache.


Remove the openvino_models folder and re-run the validation demo in OpenVINO™ 2020.3.



Regards,

Peh


Zitong-Xu
Beginner
179 Views

Hello Peh,

 

Thanks for the quick reply!

 

I successfully ran the validation demo in OpenVINO 2020.3 after removing the openvino_models directory (yes I had run the 2021.3 version before).

 

Thanks for your assistance~

 

Regards,

Zitong

Peh_Intel
Moderator
175 Views

Hi Zitong,


I am glad that I was able to help.


This thread will no longer be monitored since this issue has been resolved. If you need any additional information from Intel, please submit a new question.



Regards,

Peh


Reply