For this project, an agent is trained to work with the Reacher 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 4 numbers, corresponding to torque applicable to two joints. Every entry in the action vector should be a number between -1 and 1.
In this project, it's provide two separate versions of the Unity environment:
- The first version contains a single agent.
- The second version contains 20 identical agents, each with its own copy of the environment.
For solve this project, I used the second version (20 agents) and a DDPG algorithm for training.
The trained agent could be used in both environments.
To solve this environment with multiple agents, every agent must get an average score of +30 (over 100 consecutive episodes).
Specifically,
- After each episode, we add up the rewards that each agent received (without discounting), to get a score for each agent. This yields 20 (potentially different) scores. We then take the average of these 20 scores.
- This yields an average score for each episode (where the average is over all 20 agents).
The environment is considered solved, when the average (over 100 episodes) of those average scores is at least +30.
-
Download the environment from one of the links below. You need only select the environment that matches your operating system:
-
Version 1: One (1) Agent
- Linux: click here
- Mac OSX: click here
- Windows (32-bit): click here
- Windows (64-bit): click here
-
Version 2: Twenty (20) Agents
- 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 (version 1) or this link (version 2) 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.)
-
-
Place the file in this folder, and unzip (or decompress) the file.
To set up your python environment to run the code in this repository, follow the instructions below.
-
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
-
Install all dependencies from
requirements.txt:
pip install -r requirements.txt- Install Pytorch version 0.4.0 with your correct Cuda version (in my case, I'm using cuda 10.0).
conda install -n drlnd pytorch=0.4.0 cudatoolkit=10.0 -c pytorch4 - Create an IPython kernel for the drlnd environment.
python -m ipykernel install --user --name drlnd --display-name "drlnd"Follow the instructions in Report.ipynb to get started with training the agent!
In episode 182 (and after 46:42 minutes using Udacity's Workspace GPU), the agent achieved the expected result 👍 (moving avg > 30).
So let's see what happen to the agent:
| We started with a random agent | After 182 episodes the double-jointed arm was pretty good on folling the target |
|---|---|
 |
 |
The single-agent images above are just to make easy to understand the behavior of the environment and the trained agent. The agent was trained with the multi agent environment.
You can see the same trained agent executing a multi-agent simulation in the image below:
