In this project I have used the Reacher environment to train an agent.
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.
In this project, I've used the Unity environment that contains 20 identical agents, each with its own copy of the environment.
This environment 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.
An 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.
-
Download the environment from one of the links below that matches your operating system:
- Version 2: Twenty (20) Agents
- Linux: click here
- Mac OSX: click here
- Windows (32-bit): click here
- Windows (64-bit): click here
- Version 2: Twenty (20) Agents
-
Please make sure you have Python 3.6+ installed on your machine and to install the other dependencies that are quired to run this project. Please run the below command from p2_continuous-control folder.
conda install --yes --file requirements.txt
The file Report.md contains the strategies and the methodologies followed to train the agent.
The agent was able to solve the environment in 226 episodes Please refer to checkpoint_solved.pth file for finding the saved weights.