zsychina / Shared-Autonomy-DRL

Implementation of the paper Shared Autonomy via Deep Reinforcement Learning

Implementation of Shared Autonomy via Deep Reinforcement Learning

*正在施工 Unfinished

本仓库是对这篇论文的复现

This repo is an unofficial implementation of this paper.

This implementation differs from previous work that previous work used tenserflow where this implementation uses PyTorch

Paper info:
REDDY S, DRAGAN A, LEVINE S. Shared Autonomy via Deep Reinforcement Learning[C/OL]//Robotics: Science and Systems XIV. Robotics: Science and Systems Foundation, 2018. DOI:10.15607/RSS.2018.XIV.005

安装 Install

conda install -c conda-forge gym
conda install pytorch

使用 Usage

1. run pretrain.py to get a pretrained policy network and a target network without human in the loop
2. run train.py to fine-tuning the networks with human in the loop
3. run run.py to test the result of the trained networks

背景 Background

最近在浅浅地学习一些RL相关内容，建立本repo仅做学习记录。本人工程渣，如有错误，请不吝赐教！
小弟拜谢。

This repo is for my learning records. My engineering sucks, if there are mistakes, please let me know!
Thank you bro.

参考 References

• Paper
• Deep Q-Learning
• Double Q-Learning