def main(): print('Cart Pole') env = gym.make('CartPole-v1') q_learn = QLearning(env, num_episodes=3000) q_learn.run()
# Implement Q-learning and use this to solve the cartpole-environment import gym # Source: https://github.com/JoeSnow7/Reinforcement-Learning/blob/master/Cartpole%20Q-learning.ipynb # We define a class to contain the learning algorithm from QLearning import QLearning env = gym.make("CartPole-v0") agent = QLearning(env) agent.train() agent.run()
ui_ = UI(MATRIX, ui2sig_, sig2ui_) # UI是主进程 app_ = ui_.run_() process_ = MoveUI(ui2sig_, sig2ui_) # MoveUI是子进程 process_.daemon = True process_.start() app_.exec_() process_.terminate() # %% Env + UI if args_.env_ui: env_ui(Env) # %% QLearning + Env if args_.q_env: q_learning_ = QLearning(MATRIX) # QLearning是主进程 q_learning_.run(5) # %% QLearning + UI if args_.q_ui: # q_learning_ = q_base_ui(QLearning, mute_train=args.mute_train) q_learning_ = q_base_ui( QLearning, train_rounds=10, e_schedule=(0.5, 0.95, 10), test_rounds=30, secs=0.03, mute_train=args_.mute_train, ) # %% Sarsa + Env if args_.s_env: