Unity Reacher

ECE276C Final Project - Fall 2020 UCSD MS ECE

Team Members

• Alexander Behnke - [email protected]
• Anirudh Swaminathan - [email protected]
• Sai Akhil Suggu - [email protected]

Description

This repository houses code for the final project of ECE 276C - Robot Reinforcement Learning taken at UCSD in Fall 2020.

Base Code

The base implementation of DDPG and D4PG on Unity Reacher environment was provided in this repository. The Unity Reacher environment is provided by this repository. The specific environment that is in use in this project is the Unity Reacher environment.

Installation Instructions

• Install dependencies on top of python3.6.8 (the version on DSMLP server)

pip3 install -r requirements.txt

• Follow env installation instructions as given in the Getting Started section in problem 2 README. Ensure that when on server, follow the instruction marked (for AWS) to download the env without visualization in this directory