VinF/deer

DEEp Reinforcement learning framework

What it solves

DeeR is a modular Python library designed to simplify the implementation of Deep Reinforcement Learning (DRL) algorithms, allowing developers to adapt the framework to various needs and environments.

How it works

It provides a set of pre-built, modular components that can be implemented across different environments, including those using OpenAI Gym. The library includes several core DRL techniques such as Double Q-learning, prioritized Experience Replay, and Deep Deterministic Policy Gradient (DDPG).

Who it’s for

It is intended for developers and researchers working with Deep Reinforcement Learning who need a flexible, modular framework to build and test agents.

Highlights

  • Modular design for easy adaptation
  • Support for Double Q-learning and DDPG
  • Prioritized Experience Replay
  • Combined Reinforcement via Abstract Representations (CRAR)
  • Compatible with OpenAI Gym environments

Related

  • Project
  • Project
  • Project
  • Project