The objective of this repository is to research some AI and deep learning techniques to train an agent to play a game. The main focus is to personally learn this technology.
A game engine will be developed along side the agent, adding features to make the game more complex, and observe the agent behaviour on different scenarios.
This script is used to start a Magic game.
Usage:
Optional arguments:
- --game_type Path to the game configurations file (.json). Defaults to 'config/game_normal.json'.
- --deck_1 Deck file for player 1 (.txt). Defaults to 'config/deck_basic.txt'.
- --deck_2 Deck file for player 2 (.txt). Defaults to 'config/deck_basic.txt'.
Basic usage:
python test_game.py