def test_sl1338hc_vs_rl(sl1338_hc_model_path, rl_models_directory): rl_model_path = rl_models_directory + 'rl1_model.h5' players = [ SLPlayer1338hc(name='SLPlayer1', game_model_path=sl1338_hc_model_path), RLPlayer(name='RLPlayer1', model_path=rl_model_path), SLPlayer1338hc(name='SLPlayer2', game_model_path=sl1338_hc_model_path), RLPlayer(name='RLPlayer2', model_path=rl_model_path) ] get_function_name() run_game(players)
def test_sl_3lp_222_vs_rl(sl3lp222_model_path, rl_models_directory): rl_model_path = rl_models_directory + 'rl1_model.h5' players = [ SLPlayer3LP222(name='SLPlayer1', game_model_path=sl3lp222_model_path), RLPlayer(name='RLPlayer1', model_path=rl_model_path), SLPlayer3LP222(name='SLPlayer2', game_model_path=sl3lp222_model_path), RLPlayer(name='RLPlayer2', model_path=rl_model_path) ] get_function_name() run_game(players)
def test_rl_vs_challenge(weights_directory): model_path = weights_directory + 'rl1_model.h5' players = [ RLPlayer(name='RLPlayer1', model_path=model_path), ChallengePlayer(name='ChallengePlayer1'), RLPlayer(name='RLPlayer2', model_path=model_path), ChallengePlayer(name='ChallengePlayer2') ] get_function_name() run_game(players)
def test_rl_vs_greedy(weights_directory): model_path = weights_directory + 'rl1_model.h5' players = [ RLPlayer(name='RLPlayer1', model_path=model_path), GreedyPlayer(name='Greedy1'), RLPlayer(name='RLPlayer2', model_path=model_path), GreedyPlayer(name='Greedy2') ] get_function_name() run_game(players)
def test_rl_vs_random(weights_directory): model_path = weights_directory + 'rl1_model.h5' players = [ RLPlayer(name='RLPlayer1', model_path=model_path), RandomPlayer(name='Track'), RLPlayer(name='RLPlayer2', model_path=model_path), RandomPlayer(name='Trick') ] get_function_name() run_game(players)
def test_rl_greedy_vs_rl_challenge(weights_directory): model_path = weights_directory + 'rl1_model_challenge.h5' model_path_greedy = weights_directory + 'rl1_model.h5' players = [ RLPlayer(name='RLGreedyPlayer1', model_path=model_path_greedy), RLPlayer(name='RLChallengePlayer1', model_path=model_path), RLPlayer(name='RLGreedyPlayer2', model_path=model_path_greedy), RLPlayer(name='RLChallengePlayer2', model_path=model_path) ] get_function_name() run_game(players)
def run(log_dir, episodes, rounds): model_path = log_dir + '/rl1_model.h5' rl_player = RLPlayer(name='RL1', model_path=model_path, rounds=rounds) players = [ rl_player, ChallengePlayer(name='Tick'), ChallengePlayer(name='Trick'), ChallengePlayer(name='Track') ] players = [ rl_player, GreedyPlayer(name='Tick'), GreedyPlayer(name='Trick'), GreedyPlayer(name='Track') ] sum_won = 0 for e in range(episodes): tournament = Tournament() [tournament.register_player(player) for player in players] tournament.play(rounds=rounds, use_counting_factor=False) rl_player.replay() sum_won += rl_player.won[0] print_stats_winning(rl_player.won_stich, rl_player.won, e, sum_won) rl_player.reset_stats() rl_player.model.save(model_path)