walk-bot

PPO implementation of Humanoid-v2 from Open-AI gym

• Project completed as part of Prof. Capobianco's Deep Reinforcement Learning class
• Sapienza University of Rome

Demo

Installation

• Python 3.6.4
• Please see requirements.txt for necessary modules

Repo code structure

• Report (PDF) detailing the implemenation and PPO

• PyTorch implementation: taken from Move37 and altered to make project

• Tensorflow implementation: my personal implementation of the above (P.S. very slow, to be improved)

Sample Results:

•