This project is built using poetry, but it doesn't have to be run with it as long as you set up your own virtual environment.
If not using poetry, create a virtual environment and install this directory as a package (pip install .) inside the virtual environment.
To train a simple classifier with 1 dimensional convolutions, run the paramnet.run.train_conv module.
$ poetry run python -m paramnet.run.train_conv path/to/bag/folder
OR
$ python -m paramnet.run.train_conv path/to/bag/folder
where path/to/bag/folder is a folder containing the bag files you want to use as training/validation data.
The bag files need to contain /scan and /joystick topics, and currently the classifier is only binary and assigns classes based on the buttons[0] value of the joystick messages.
It is relatively simple to change this in the code.