for rknn-toolkit2 V1.6.0
[https://github.com/airockchip/rknn-toolkit2/tree/master]
- git clone https://github.com/airockchip/rknn-toolkit2/tree/master
- cd rknn-toolkit2/rknn-toolkit2/packages/
- pip3 install rknn_toolkit2-1.6.0+81f21f4d-cp38-cp38-linux_x86_64.whl (python 3.8 for this example)
- Run in python terminal (library test, successful when no errors are returned)
- from rknnlite.api import RKNNLite
Orange pi 5 (ubuntu 22.04 by Joshua-Riek) [https://github.com/Joshua-Riek/ubuntu-rockchip/wiki/Orange-Pi-5-Plus]
- git clone https://github.com/airockchip/rknn-toolkit2/tree/master
- cd rknn-toolkit2/rknn_toolkit_lite2/packages/
- pip3 install rknn_toolkit_lite2-1.6.0-cp310-cp310-linux_aarch64.whl (python 3.10 for this example)
- Run in python terminal (library test, successful when no errors are returned)
- from rknnlite.api import RKNNLite
- cd dmonitoring_python
- sudo python convert_dmonitoring.py
- cd dmonitoring_python
- sudo python run_dmonitoring.py
When the library test returns an error, it means that the library version is mismatch
- suggest to make python virtualenv using "python[version] -m venv [virtual-environment-name]" https://www.freecodecamp.org/news/how-to-setup-virtual-environments-in-python/
In the openpilot folde, there is a folder called openpilot.py which is the modified version of the openpilot model runner you can transfer over to the openpilot version, and add in support for RKNN (this is already done in the development fork of openpilot for Kommu)