Build an A2C agent that learns to play Sonic with Tensorflow (tutorial)

⚠️ We launched a new free Deep Reinforcement Learning Course with Hugging Face 🤗 from beginner to expert, here : huggingface.co/deep-rl-course/unit0/introduction
---

Today we'll learn to implement an advantage actor critic agent that learns to play Sonic the Hedgehog on Genesis (MegaDrive)

There is some problems of incrustations at the end of the video sorry for that.

This video is part of the free Deep Reinforcement Learning Course, a series of articles and videos where you'll master the skills and architectures you need, to become a deep reinforcement learning expert. Start learning now here: bit.ly/2MX2mNE

The A2C article is here: medium.freecodecamp.org/an-intro-to-advantage-acto…

The implementation repository: github.com/simoninithomas/Deep_reinforcement_learn…

The excellent Aborghi PPO implementation: github.com/aborghi/retro_contest_agent

Generalized advantage estimation paper: arxiv.org/pdf/1506.02438.pdf