Berkeley pacmanCTF using Reinforcement Learning

Training

Training together

Train for 5 times and play for 2 times, the weights will store in *.train

python capture.py -r superTeam -x 5 -n 7

Testing

If need to use pretrain weights, need to uncomment some lines in __init__, otherwise, best weight is use

Play against baselineTeam

python capture.py -r superTeam -b baselineTeam