Deep RL Project: Collaboration and Competition

Introduction

For this project, I trained an agent to solve the Tennis environment.

In this environment, two agents control rackets to bounce a ball over a net. If an agent hits the ball over the net, it receives a reward of +0.1. If an agent lets a ball hit the ground or hits the ball out of bounds, it receives a reward of -0.01. Thus, the goal of each agent is to keep the ball in play.

The observation space consists of 8 variables corresponding to the position and velocity of the ball and racket. Each agent receives its own, local observation. Two continuous actions are available, corresponding to movement toward (or away from) the net, and jumping.

The task is episodic, and in order to solve the environment, our agents must get an average score of +0.5 (over 100 consecutive episodes, after taking the maximum over both agents). Specifically,

• After each episode, we add up the rewards that each agent received (without discounting), to get a score for each agent. This yields 2 (potentially different) scores. We then take the maximum of these 2 scores.
• This yields a single score for each episode.

The environment is considered solved, when the average (over 100 episodes) of those scores is at least +0.5.

Getting Started

To set up your python environment to run the code in this repository, follow the instructions below.

1. Create (and activate) a new environment with Python 3.6.

   • Linux or Mac:

   conda create --name drlnd python=3.6
   source activate drlnd

   • Windows:

   conda create --name drlnd python=3.6 
   activate drlnd

2. Clone the repository (if you haven't already!), and navigate to the python/ folder. Then, install several dependencies.

   git clone https://github.com/udacity/deep-reinforcement-learning.git
   cd deep-reinforcement-learning/python
   pip install .

3. Create an IPython kernel for the drlnd environment.

   python -m ipykernel install --user --name drlnd --display-name "drlnd"

4. Before running code in a notebook, change the kernel to match the drlnd environment by using the drop-down Kernel menu.

5. Download the environment from one of the links below. You need only select the environment that matches your operating system:

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

   (For Windows users) Check out this link if you need help with determining if your computer is running a 32-bit version or 64-bit version of the Windows operating system.

   (For AWS) If you'd like to train the agent on AWS (and have not enabled a virtual screen), then please use this link to obtain the "headless" version of the environment. You will not be able to watch the agent without enabling a virtual screen, but you will be able to train the agent. (To watch the agent, you should follow the instructions to enable a virtual screen, and then download the environment for the Linux operating system above.)

6. Place the file in the root folder of this GitHub repository, and unzip (or decompress) the file.

7. If not working in AWS, change file_name in file main.py (row 18) to the correct path to the above env file.

The code provided here was tested on an Ubuntu AWS headless (no visual) operating system, but should work for all above provided you follow the correct instructions.

Instructions

2 options:

1. Follow the instructions in Tennis.ipynb to see the training and performance of the trained agent. Best pre-trained weights are saved in checkpoints folder.

2. To train the agent via the command line, run ./main.py. You'll see statistics outputed to the command, which we'll log to the 'runs' directory for visualizing via tensorboard. To start tensorboard run tensorboard --logdir runs.

To load a saved model in evaluation mode run ./main.py --eval --load=<path to files>. This will load saved weights from checkpoint files. A pre-trained model is included, to load it run ./main.py --eval --load=checkpoints/solved_best. Full file names are named 'solved_best.0.actor.pth', 'solved_best.0.critic.pth', you should only pass the common part, in this case 'best'.