Attention: This project is work-in-progress. The current soft-actor-critic implementation still requires a few adjustments. The next step would be to adjust this to the newest version of the algorithm that eliminates the need for a value network.

Project 2: Continuous Control

Introduction

For this project, you will 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 four numbers, corresponding to torque applicable to two joints. Every entry in the action vector should be a number between -1 and 1.

Distributed Training

For this project, we will provide you with 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.

The second version is useful for algorithms like PPO, A3C, and D4PG that use multiple (non-interacting, parallel) copies of the same agent to distribute the task of gathering experience.

Solving the Environment

Note that your project submission need only solve one of the two versions of the environment.

Option 1: Solve the First Version

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

Option 2: Solve the Second Version

The barrier for solving the second version of the environment is slightly different, to take into account the presence of many agents. In particular, your agents must get an average score of +30 (over 100 consecutive episodes, and over all 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 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.

Getting Started

Reacher

Unity's Reacher Environment is an environment in which an agent must control robotic hands with the goal of keeping the ends of the arms within a sphere. This repository uses the 20-agent environment in which a single agent must control 20 robotic arms.

The agent interacts with the environment via the following:

• It is fed 20 sets of observations each with a vector of 33 elements
• For each of the 20 arms, the agent must provide an action vector representing 4 continous actions

Prerequisites

• Anaconda

• Python 3.6

• A conda environment created as follows

  • Linux or Mac:

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

  • Windows

  conda create --name drlnd python=3.6 
  activate drlnd
  

• Required dependencies

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

Cloning the repository

1. Clone the repository

2. cd Reacher

3. Download Unity Reacher Environment

   • Linux: click here

   • Linux Headless: click here

   • Mac OSX: click here

   • Windows (32-bit): click here

   • Windows (64-bit): click here

     • 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.)

4. Unzip to git directory

5. jupyter notebook

6. You can train your own agent via main.ipynb or watch a single episode of the pre-trained network via Visualization.ipynb

Instructions

Follow the instructions in Continuous_Control.ipynb to get started with training your own agent!