README.md

Structure

This repo includes

• Pytorch notes and its implementations
• Classic RL algorithms
• notes I made going through the resources.
• training and testing a RL agent in multiple environment in metadrive simulator(open source autonomous driving simulator)

Execution

You can run the test codes normally on google collab or jupyter notebook

• For RL training and scripts
• install metadrive
• move the "set_1 folder under metadrive/examples"
• cd under the metadrive and run

python -m metadrive.examples.file_name

(Remember to comment out some part of codes to customise the settings)

Resources

• A tutorial on MADDPG(not imp right now) - https://medium.com/machine-intelligence-and-deep-learning-lab/a-tutorial-on-maddpg-53241ae8aac

• Davild silver playlist - https://www.davidsilver.uk/teaching/

• Policy based, on/off policy, model based/free - https://stats.stackexchange.com/questions/407230/what-is-the-difference-between-policy-based-on-policy-value-based-off-policy

• Q learning - https://www.avenga.com/magazine/q-learning-applications/#:~:text=The%20optimal%20value%20function%20

• Deep RL bootcamp - https://www.youtube.com/watch?v=qaMdN6LS9rA

• Policy and value iteration - https://medium.com/@m.alzantot/deep-reinforcement-learning-demysitifed-episode-2-policy-iteration-value-iteration-and-q-978f9e89ddaa

• Policy iteration : https://towardsdatascience.com/policy-iteration-in-rl-an-illustration-6d58bdcb87a7

• PPO - https://huggingface.co/blog/deep-rl-ppo

• Theory + implementation reference - https://huggingface.co/learn/deep-rl-course/unit8/introduction