Simplest gym environment with discrete actions?

Web: https://www.reddit.com/r/reinforcementlearning/comments/vfiwnl/simplest_gym_environment_with_discrete_actions/

Hi there,

What is the simplest \`gym\` environment with a discrete action space?

I'm getting started with reinforcement learning and having fun doing some of my own implementations of standard algorithms (DQN, VPG, PPO, ...). It's been fun letting my agents loose in Super Mario Bros., but debugging my implementations has been a challenge. I'd like to find a simple environment to iterate rapidly on my models. Any recommendations?

(Ideally, I'd like inputs to be screen pixels too, but that's …

environment reinforcementlearning