This project was made for the final project for the DRL-IASD 2021-2022 Class. The only included sumbissions are the best_submission.csv file.
There are multiple ways to run the main (best) algorithm -DDPG. The easiest way is to use the jupyter notebook file drl_project_main.ipnyb. An alternative is available in the main_ddpg.py file.
Some of our most notable deprecated models and tryouts are put for good measure in the deprecated folder.
ddpg.py : Containts the main class for the ddpg agent. drl_project_main.ipynb : The main notebook to run to execute the code. env.py : File containing a gym implementation of the environment for ease of use. main_ddpg.py : main file to execute to run the code. models.py : contains the actor-critic models for the DDPG. utils.py : some usefuls functions.