env = gym.make('LunarLanderContinuous-v2')
agent = Agent(alpha=0.000025, beta = 0.00025, input_dims=[8], tau=0.001, env=env, batch_size=64, layer1_size=400, layer2_size=300, n_actions=2)
episodes = 1000
np.random.seed(42)
tau_hist = []
score_hist = []
for i in range(episodes):
done = False
score = 0
state = env.reset()
while not done:
act = agent.choose_action(state)
next_state, reward, done, _ = env.step(act)
agent.store(state, act, reward, next_state, int(done))
agent.learn()
score += reward
state = next_state
agent.save_models()
score_hist.append(score)
tau_hist.append(agent.tau)
avg_score = np.mean(score_hist[-100:])
print('episode ' + str(i + 1) + 'score %.2f' % score +
'average score %.2f' % avg_score)
episodes = np.arange(1, episodes + 1)
plot_curve(episodes, score_hist, tau_hist)
batch_size=64,
gamma=0.99)
agent.load_models()
np.random.seed(0)
score_history = []
for i in range(200):
obs = env.reset()
done = False
score = 0
step = 0
while not done:
step += 1
# print(obs)
act = agent.choose_action(obs)
# print(act)
new_state, reward, done, info = env.step(act)
agent.remember(obs, act, reward, new_state, int(done))
agent.learn()
score += reward
obs = new_state
env.render()
score_history.append(score)
# if i % 25 == 0:
# agent.save_models()
print('episode ', i, 'score %.2f' % score,
'trailing 128 games avg %.3f' % np.mean(score_history[-128:]),
'finished after ', step, ' episode')
