Originally, I tried the prebuilt, had a few issues with networking, and gave up because I've had a history of having issues with how people package pi images sometimes.
So just following the linux install instructions:
It says
If you install tensorflow 1.8.0, that was compiled for python 3.4 and a pi has python 3.5.
You also need to install libatlas-base-dev libopenjp2-7-dev libtiff5
Also run
I found a few issues trying to get donkeycar running on a pi, following the instructions at https://docs.donkeycar.com/guide/install_software/#install-donkeycar-on-linux
Originally, I tried the prebuilt, had a few issues with networking, and gave up because I've had a history of having issues with how people package pi images sometimes.
So just following the linux install instructions: It says
If you install tensorflow 1.8.0, that was compiled for python 3.4 and a pi has python 3.5. You also need to install
libatlas-base-dev libopenjp2-7-dev libtiff5
Also runSo put together,
At this point, also run raspi-config and enable camera and i2c in the interfacing options, and reboot.
There is then a link to calibrate car, which references
donkey createcar
, but doesn't tell you what it is.If you skip around to the drive the car page, it says to run
Sadly,
donkey createcar
has never heard of this--path
flag, but if you run it without anything, it says it's missing the path.I think that's all the changes I had to make to get it to work.