CLIP_DELTA = 1.0
EPSILON_START = 1.0
EPSILON_MIN = .3
EPSILON_DECAY = 500000
UPDATE_FREQUENCY = 1
REPLAY_MEMORY_SIZE = 1000000
BATCH_SIZE = 32
FREEZE_INTERVAL = 1000
DETERMINISTIC = True
if __name__ == "__main__":
logging.basicConfig(level=logging.INFO)
# --- Parse parameters ---
parameters = process_args(sys.argv[1:], Defaults)
if parameters.deterministic:
rng = np.random.RandomState(123456)
else:
rng = np.random.RandomState()
# --- Instantiate environment ---
env = MG_two_storages_env(rng)
# --- Instantiate qnetwork ---
qnetwork = MyQNetwork(env, parameters.rms_decay, parameters.rms_epsilon,
parameters.momentum, parameters.clip_delta,
parameters.freeze_interval, parameters.batch_size,
parameters.update_rule, rng)
# --- Instantiate agent ---
CLIP_DELTA = 1.0
EPSILON_START = 1.0
EPSILON_MIN = 0.2
EPSILON_DECAY = 10000
UPDATE_FREQUENCY = 1
REPLAY_MEMORY_SIZE = 1000000
BATCH_SIZE = 32
NETWORK_TYPE = "General_DQN_0"
FREEZE_INTERVAL = 100
DETERMINISTIC = True
if __name__ == "__main__":
logging.basicConfig(level=logging.INFO)
# --- Parse parameters ---
parameters = process_args(sys.argv[1:], Defaults)
if parameters.deterministic:
rng = np.random.RandomState(12345)
else:
rng = np.random.RandomState()
# --- Instantiate environment ---
env = mountain_car_env(rng)
# --- Instantiate qnetwork ---
qnetwork = MyQNetwork(
env,
parameters.rms_decay,
parameters.rms_epsilon,
parameters.momentum,
parameters.clip_delta,