RL_2023

Official Repository for the Reinforcement Learning Course at Artificial Intelligence and Robotics (Sapienza University of Rome).

Prof. Roberto Capobianco

TAs: Andrea Fanti and Michela Proietti

Here you will find the code used during practical lessons.

Please refer to the classroom page for further information.

Dependencies

requirements.txt contains all the Python libraries needed to run the notebooks, meant to be used with Python 3.8.*. To install all of them with pip in a single command (linux/macos) run:

pip install -r requirements.txt

Differences with previous years

The code of this year uses the gymnasium library instead of gym, which was used in previous years. A good overview of the differences can be found here.

Syncing the fork

In order to sync the fork and merge the new changes, add the remote by:

git remote add upstream https://github.com/KRLGroup/RL_2023.git

then, simply fetch from this upstream:

git fetch upstream
git checkout main

Now you either perform a merge or a pull:

Merge

git merge upstream/main

Pull

git pull upstream main

Then, finally push to your origin repo by:

git push origin HEAD:main