33 states for each agent, 20 agents in multi agent environment. Each agent has 4 continous actions. The environment is considered solved when an average score of 30+ for all agents is maintained for 100 episodes.
First clone the udacity deep reinforcement learning repo and navigate to it's directory then
cd python
pip install .
this installs the required dependencies.
Download the Reacher Unity-ML environment from one of the following links:
Windows: click here
Mac: click here
Put it in the root of the cloned project folder and unzip to Reacher_20_agents folder.
You should now be able to build and run the project.
Run the following command to train the agent
python runner.py