Reinforcement Learning Algorithms in GridWorld

This repository contains lab session materials (code) for the SNU Reinforcement Learning course, 2026 Spring.

Download slides from here: Google Drive Link

# clone the repository
git clone https://github.com/SNU-IntelligentMotionLab/2026Spring_RL_Lab.git

Open PowerShell in the repository folder and run:

Set-ExecutionPolicy -Scope Process -ExecutionPolicy RemoteSigned
.\setup.ps1

# create and activate a virtual environment
python3 -m venv lab_env # or
python -m venv lab_env
# Windows
.\lab_env\Scripts\activate
# Linux/Mac
source lab_env/bin/activate

# install required packages
pip install -r requirements.txt

For each lab session, do cd Lab[num]_[theme] (e.g., cd Lab1_DP) and refer to README for instructions.

Current labs:

• Lab1_DP: Dynamic Programming
• Lab2_ModelFree: Monte Carlo, SARSA, and Q-Learning
• Lab3_DeepValueRL: Deep SARSA
• Lab4_PG: REINFORCE
• Lab5_DQN: DQN

TensorBoard is used in the deep RL labs for reward/loss visualization and is included in requirements.txt.