def define_agent(self, width, height, num_actions):
return NStepDQNAgent(config=Config(
num_actions=num_actions,
encoder=LayerEncoder(width, height, treasure_position=True),
optimizer=SharedAdamOptimizer(0.001),
network=CNN(hidden_units=[128]),
policy=EpsilonGreedyPolicy(1, 0.01, 25000),
discount=0.95,
n_step=16))
def define_agent(self, width, height, num_actions):
return NStepDQNAgent(
config=Config(num_actions=num_actions,
encoder=OneHotEncoder(width, height),
optimizer=AdamOptimizer(0.01),
network=MLP(),
policy=EpsilonGreedyPolicy(1, 0.01, 1000),
discount=0.95,
n_step=8))
def define_agent(self, width, height, num_actions):
return DQNAgent(config=Config(num_actions=num_actions,
encoder=OneHotEncoder(width, height),
optimizer=AdamOptimizer(0.01),
network=MLP(),
policy=EpsilonGreedyPolicy(1, 0.01, 500),
discount=0.95,
capacity=100,
batch_size=16))
def define_agent(self, width, height, num_actions):
return DQNAgent(
config=Config(
num_actions=num_actions,
encoder=LayerEncoder(width, height, treasure_position=True),
optimizer=AdamOptimizer(0.001),
network=CNN(hidden_units=[128]),
policy=EpsilonGreedyPolicy(1, 0.01, 50000),
discount=0.95,
capacity=10000,
batch_size=8,
target_sync=100,
double_q=True))