Hands-On Reinforcement Learning with Python
上QQ阅读APP看书,第一时间看更新

Getting Started with OpenAI and TensorFlow

OpenAI is a non-profit, open source artificial intelligence (AI) research company founded by Elon Musk and Sam Altman that aims to build a general AI. They are sponsored by top industry leaders and top-notch companies. OpenAI comes in two flavors, Gym and Universe, using which we can simulate realistic environments, build reinforcement learning (RL) algorithms, and test our agents in those environments. TensorFlow is an open source machine learning library by Google that is extensively used for numerical computation. We will use OpenAI and TensorFlow for building and evaluating powerful RL algorithms in the upcoming chapters.

In this chapter, you will learn about the following:

  • Setting up your machine by installing Anaconda, Docker, OpenAI Gym, and Universe and TensorFlow
  • Simulating an environment using OpenAI Gym and Universe
  • Training a robot to walk
  • Building a video game bot
  • Fundamentals of TensorFlow
  • Using TensorBoard