Questions Regarding the vww_patch Inference Tutorial in the mcunet Project

[winffke]

Description
While following the vww_patch inference tutorial in the mcunet project, I noticed the following issues:

The tutorial includes the following command:

export PYTHONPATH=${PYTHONPATH}:$(pwd)
python examples/vww.py # To apply patch-based inference, please use example/vww_patchbased.py
However, the comment seems to contain an error. Specifically, the path to vww_patchbased.py should be examples/vww_patchbased.py instead of example/vww_patchbased.py.

After downloading the vww_patch code and running it, I noticed that the code does not seem to perform patch-based inference as expected. Instead, it skips patch-based inference and directly enters layer-based inference. Could this be a bug?

While following the build process, I encountered errors. After referencing Issue #102, the results produced were still incorrect.

Questions and Suggestions
Could you please verify and correct the tutorial instructions if there are errors?
Can you confirm whether the skipping of patch-based inference is intended behavior or a bug in the code?
Is it possible to provide a complete STM32CubeIDE project file for this example to help users better understand and replicate the results?
Lastly, I would appreciate it if you could confirm the current status of the tutorial and investigate the issues mentioned.
If further discussion or clarification is needed, feel free to contact me via email: 2827583596@qq.com.

Thank you for your attention and support!