Use of Deep Q-Learning to build a RL Agent. Case of the Gym Acrobot-v1 environment.
Install the project environment
conda env create -f env.yml
conda activate acrobot
Train the agent with
python train.py
The agent pytorch model will be saved in a file called state.pt.
You can change this name and other parameter in the file train.py.
Simulate the environment and the agent with
python test.py