# In[ ]:
for episode in range(EPISODES):
observation = env.reset()
QL.curr_episode = episode
while True:
if RENDER_ENV: env.render()
valid_move = False
action = None
while not valid_move:
# Choose an action based on observation
if action == None: action = QL.choose_action(observation)
observation_, reward, done, info = env.step(action)
valid_move = info['valid']
reward = QL.calculate_reward(valid_move, done, reward,
observation_)
QL.save_experience(observation=observation,
action=action,
reward=reward,
observation_=observation_,
is_game_over=done,
is_move_valid=valid_move)
action = (action + 1) % QL.n_y
from q_state import next_state, random_state, actions
from q_learning import QLearning
from q_table import QTable
if __name__ == "__main__":
episode = 100
model_save_interval = 10
table = QTable(actions)
learning = QLearning(table)
for step in range(episode):
init_state = random_state()
i = 0
reward = 0
while reward != 1:
state = init_state
while True:
i += 1
action = learning.choose_action(state)
state2, reward, done = next_state(state, action, table)
learning.learn(state, action, reward, state2, done)
if done:
break
state = state2
print(init_state, i, len(table.q_table))
if (step + 1) % model_save_interval == 0:
table.save()