Goal: plays multiplayer games with alts to efficiently clear enemies Current status: working on making it play Diep.io Architecture:
- DDQN_Model.py is a model that creates 2 from the class ConvNet in Convolutional_Neural_Network.py
- Get_environment gets the environemnt by taking a screenshot into a numpy array.
- Gather_training_points takes a screenshot of the reward and reads it using OCR.
- Main_Bot.py intergrates everything together
Note: Everything complete, just need to test. May add an Adversial Inverse Reinforcement Learning algorithm
https://colab.research.google.com/drive/1HShBe_qfRYvRIBe9ff8NrtXhzsCuI-Eo?usp=sharing