Project: Continous Control

This project is about training a deep reinforcment learning agent how to move a double jointed arm to target locations.

About the environment

In this environment, a double-jointed arm can move to target locations. A reward of +0.1 is provided for each step that the agent's hand is in the goal location. Thus, the goal of your agent is to maintain its position at the target location for as many time steps as possible.

The observation space consists of 33 variables corresponding to position, rotation, velocity, and angular velocities of the arm. Each action is a vector with four numbers, corresponding to torque applicable to two joints. Every entry in the action vector should be a number between -1 and 1.

The task is episodic, and in order to solve the environment, your agent must get an average score of +30 over 100 consecutive episodes.

Setting up the environment

Download environment

The Environment can be downloaded at

• Linux: click here
• Mac OSX: click here
• Windows (32-bit): click here
• Windows (64-bit): click here

Install dependencies

Dependencies needed to get the programm running are gathered in the requirements.txt to install those execute the command:

 pip install requirements.txt

I had problems to install torch==0.4.0 so if you execute

pip install torch==0.4.0 -f https://download.pytorch.org/whl/torch_stable.html 

before the command above it should work as expected

Run Agent

Run Continous_Control_train.ipynb to train the agent

Files

• Continuous_Control.ipynb - template given by udacity
• Continuous_Control_train.ipynb - juptyter notebook for training of the agent
• model.py - classes for the networks used by actor and critic
• agent_ddpg.py - class of the DDPG agent
• solved_735_actor_local.pth - weights of the first time the agent reached a average over 100 episodes of >30
• end_actor_local.pth - weights of the trained agent after 1000 episodes