- It is recommended to use python version >= 3.5
- In the command line, after moving to the required directory where run.py file is present, type
python3 run.py
- For each question, the learning methods are implemented in algorithms.py file which is imported in run.py file and used for demonstrations.
- Code/Q1: Gridworld problem solved using SARSA, Q-learning
- Code/Q2: Cart-pole balancing using DQN (developed the code on template that was given)
- Question file Q1 contains the two files run.py and algorithms.py while Q2 contains only run.py file.
- run.py contains the code that performs the experiment
- algorithms.py contains the implementation of the algorithms (Q-learning and SARSA for Q1, Policy gradient method for Q2)
- log files are erased for memory purposes
- https://github.com/xinleipan/gym-gridworld : Gridworld implementation
- https://github.com/openai/gym