This is an experiment to build a simple game (Yahtzee) and train an AI Agent using Deep Q-Learning.
To play an interactive game yourself, run:
python3 cli.py
The AI model has 2 agents:
- Deciding which dice to keep/re-roll
- Deciding where to score the dice
To train the model, run:
python3 train.py --roll-weights=path/to/roll.hdf5 \
--slot-weights=path/to/slot.hdf5