/train-procgen-pytorch

Pytorch implementation on OpenAI's Procgen ppo-baseline, built from scratch.

Primary LanguageJupyter Notebook

Goal Misgeneralization in Deep RL

Code for the paper Goal Misgeneralization in Deep Reinforcement Learning, based on the repository https://github.com/joonleesky/train-procgen-pytorch

Requirements

First, install the modified version of the procgen library available here.

  • python>=3.6
  • torch 1.3
  • procgen
  • pyyaml