FAQs
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

modelzoo deployment error

yessine
Associate III

‎2024-06-25 02:48 PM - last edited on ‎2024-06-25 04:26 PM by

i downloaded stmzoo file from github. followed the steps in how to install stm modelzoo. created a virtual machine envirement. installed all the requirements (manually because i kept having errors when installing them all automatically using the command install -r requirement.txt) . however . when i want to deploy it using the command python deploy.py after changing my place to the deployement folder . i keep get this error that say no module found with the name common-deploy although it exist in the modelzoo folder under common folder . why??(windows)

Labels:
0 Kudos
3 REPLIES 3
MSHAH.3
ST Employee

‎2024-06-27 05:47 AM - edited ‎2024-06-27 05:48 AM

@yessine , starting from stm32ai-modelzoo v2.0.0 onwards, you have only a single entry point to the model zoo code for each case. That is the stm32ai_main.py file located in stm32ai-modelzoo/<usecase>/src directory.


To launch any of the available operation_mode (train, evaluate, deploy etc.) you have to be in the stm32ai-modelzoo/<usecase>/src directory. If not there already you can do it from terminal by cd command to that folder. Then you can either configure the operation_mode by modifying the user_config.yaml file in this folder or can use a pre-filled default deployment file with modifying only few parameters. Finally, to launch the scripts you have to use the command:
> python stm32ai_main.py

 

You can find more details on how to deploy a model on STM32 platforms using model zoo in the readme files in each deployment folder. Here is a guide for deploying an Image Classification model on STM32 MCU. Let us know if you still have some issues.

0 Kudos
yessine
Associate III

‎2024-07-03 07:14 AM

first of all . thanks for the guide you provided . it really helped me doing some progress. Now , after giving the model path and switching the mode to deployment in the .yaml file in src ,i had some other issues ( 3 type of errors i think ) : first of all i had 'login faield' problem' when he asked for my ST account credentials even though i'am 100% they are correct . the second type of errors are related to the hydra   files and third type of errors are related to some files . honestly im really lost in these errors so can u provide me some help , guide, or even some explanations Screenshot 2024-07-03 161018.png

0 Kudos
MSHAH.3
ST Employee
In response to yessine

‎2024-07-03 07:31 AM

It might be a problem with proxy. Are you running the scripts behind some proxy server? If so, we do provide the guide on how to use STM32ai dev-cloud in settings where you are running behind the proxy servers Wiki article. There are also some guidelines on how to install the local version on your machine (if the online version is not working due to proxy or some firewall).

Furthermore, to avoiding any misalignments due to some file paths or typo issues, may be try to use the deployment_config.yaml file in the "object_detection/src/config_file_examples/".


You can call this python scripts with this file simply from src directory by issuing following command:

python stm32ai_main.py --config-path ./config_file_examples/ --config-name deployment_config.yaml

Edit the deployment_config.yaml file  to point to correct paths of your CubeIDE.exe, stm32ai.exe file and that is all that you need to change.

Once you can reproduce these basic steps, then you can experiment more, for example you can edit the model path to use a different model.

If this does not solve the issue, share the configuration file and model file for us to be able to reproduce the issue.

0 Kudos
Top