# epsilon=0.9BATCH_SIZE=32GAMMA=0.99EPS_START=1EPS_END=0.02EPS_DECAY=1000000EPS_RANDOM_COUNT=50000# 前50000步纯随机用于探索TARGET_UPDATE=1000# stepsRENDER=False lr=1e-4INITIAL_MEMORY=10000MEMORY_SIZE=10*INITIAL_MEMORYn_episode=100000#10000000MODEL_STORE_PATH='./models'#+'DQN_pytorch_pong'mod...
['tetris', 'lost_luggage', 'pitfall2', 'pong', 'koolaid', 'breakout', 'hero', 'jamesbond', 'alien', 'road_runner', 'tennis', 'beam_rider', 'entombed', 'freeway', 'double_dunk', 'seaquest', 'king_kong', 'backgammon', 'casino', 'tic_tac_toe_3d', 'mr_do', 'zaxxon', ...
global Memory % Fix size of the Pong Field fieldHeight = 64; fieldWidth = 32; % --- Step 1: Initializazion --- % % --- 1. Setup Learning Parameters --- % if Octave StartLearning = 1600 ; % # of Random Episodes before Learning starts NEpisodes = 20000; % Octave is much slower...
1. ['tetris', 'lost_luggage', 'pitfall2', 'pong', 'koolaid', 'breakout', 'hero', 'jamesbond', 'alien', 'road_runner', 'tennis', 'beam_rider', 'entombed', 'freeway', 'double_dunk', 'seaquest', 'king_kong', 'backgammon', 'casino', 'tic_tac_toe_3d', 'mr_do', 'zaxxon...