Repository for the course IFT-7014: Directed reading on

Deep Reinforcement Learning

by Luc Coupal, Université Laval, Montréal, QC, Canada, Luc.Coupal.1@uLaval.ca

Under the supervision of:

Professor Brahim Chaib-draa, Directeur du programme de baccalauréat en génie logiciel de l'Université Laval, Québec, QC, Canada, Brahim.Chaib-draa@ift.ulaval.ca

---

Essay on:

• Maximum Entropy RL:
  • Soft Actor-Critic
  • See my blog post Soft Actor-Critic part 1: intuition and theoretical aspect for more details on SAC and MaxEnt-RL
• Classical RL:
  • Actor-Critic
  • Basic policy gradient
• A reflexion on design, architecture and implementation details

---

Watch mp4 video - Soft Actor-Critic Post training - Test run on 2X harder LunarLanderContinuous-v2 environment

---

Install instruction:

1. Create & activate a new virtual environment (I recommand using conda, ... it's a walk in the park)

   conda create --name myNewVirtualEnvironmentName python=3.7
   conda activate myNewVirtualEnvironmentName

2. Clone the GitHub repository & install dependencies:

   git clone https://github.com/RedLeader962/LectureDirigeDRLimplementation.git
   cd LectureDirigeDRLimplementation
   pip install -e .

   This will automaticaly install those dependencies in myNewVirtualEnvironmentName :

    'gym>=0.14.0'
    'tensorflow>=1.14.0,<2.0',
    'matplotlib>=3.1.0',
    'numpy>=1.16.4',
    'seaborn>=0.9.0',
    'pytest',
   

3. Enjoy DRL script

---