Soft Actor-Critic

Implementation of Soft Actor-Critic Algorithm in TF2

Notes: https://shivamshakti.dev/posts/sac

Usage

• Create a virtual environment for Python (I use this setup)

• Install the dependencies

  pip install -r requirements.txt
  

• Run the training script

  cd src
  python main.py # Uses `MountainCarContinuous-v0` by default
  

• Run the evaluation script

  python play.py --model_name <PATH_TO_SAVED_MODEL>
  

References

• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• Soft Actor-Critic Algorithms and Applications
• Learning to Walk via Deep Reinforcement Learning
• Open AI SAC