Implementation of MP Control + DDPG to manage a continous action space
Within a python enviroment run:
pip install .
- Check
notes/notes.pdffor project details
Instructions to run:
- Open main.py in VSCode
- create a log directory in the same directory as main.py
- Check for the #%% comment at the top of the screen.
- The jupyter extension should be available and a
run belowbutton should be available - Press said button and a window will open with running code
- Log of the process is stored in
log/events.log - a plot of different losses and rewards will be output to
misc/train_plot.jpg