def test_deepq():
"""
test DeepQ on atari
"""
logger.configure()
set_global_seeds(SEED)
env = make_atari(ENV_ID)
env = bench.Monitor(env, logger.get_dir())
env = wrap_atari_dqn(env)
model = DQN(env=env,
policy=CnnPolicy,
learning_rate=1e-4,
buffer_size=10000,
exploration_fraction=0.1,
exploration_final_eps=0.01,
train_freq=4,
learning_starts=10000,
target_network_update_freq=1000,
gamma=0.99,
prioritized_replay=True,
prioritized_replay_alpha=0.6,
checkpoint_freq=10000)
model.learn(total_timesteps=NUM_TIMESTEPS)
env.close()
del model, env
def test_deepq():
"""
test DeepQ on atari
"""
logger.configure()
set_global_seeds(SEED)
env = make_atari(ENV_ID)
env = bench.Monitor(env, logger.get_dir())
env = wrap_atari_dqn(env)
q_func = deepq_models.cnn_to_mlp(convs=[(32, 8, 4), (64, 4, 2),
(64, 3, 1)],
hiddens=[256],
dueling=True)
model = DeepQ(env=env,
policy=q_func,
learning_rate=1e-4,
buffer_size=10000,
exploration_fraction=0.1,
exploration_final_eps=0.01,
train_freq=4,
learning_starts=10000,
target_network_update_freq=1000,
gamma=0.99,
prioritized_replay=True,
prioritized_replay_alpha=0.6,
checkpoint_freq=10000)
model.learn(total_timesteps=NUM_TIMESTEPS)
env.close()
del model, env
def __init__(self, env=None):
"""Return only every `skip`-th frame"""
super(FireOtherSkipEnv, self).__init__(env)
# most recent raw observations (for max pooling across time steps)
self._obs_buffer = deque(maxlen=2)
self.env = wrap_atari_dqn(self.env)
model_output='/home/jingjia16/stable-baselines/scripts/deepq_pong.zip'
if os.path.exists(model_output):
self.model = DQN.load(model_output)
else:
print("failed to load the model")
def main():
"""
run a trained model for the pong problem
"""
env = gym.make("PongNoFrameskip-v4")
env = deepq.wrap_atari_dqn(env)
model = DeepQ.load("pong_model.pkl", env)
while True:
obs, done = env.reset(), False
episode_rew = 0
while not done:
env.render()
action, _ = model.predict(obs)
obs, rew, done, _ = env.step(action)
episode_rew += rew
print("Episode reward", episode_rew)
def main():
"""
run the atari test
"""
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--env', help='environment ID', default='BreakoutNoFrameskip-v4')
parser.add_argument('--seed', help='RNG seed', type=int, default=0)
parser.add_argument('--prioritized', type=int, default=1)
parser.add_argument('--prioritized-replay-alpha', type=float, default=0.6)
parser.add_argument('--dueling', type=int, default=1)
parser.add_argument('--num-timesteps', type=int, default=int(10e6))
parser.add_argument('--checkpoint-freq', type=int, default=10000)
parser.add_argument('--checkpoint-path', type=str, default=None)
args = parser.parse_args()
logger.configure()
set_global_seeds(args.seed)
env = make_atari(args.env)
env = bench.Monitor(env, logger.get_dir())
env = wrap_atari_dqn(env)
q_func = deepq_models.cnn_to_mlp(
convs=[(32, 8, 4), (64, 4, 2), (64, 3, 1)],
hiddens=[256],
dueling=bool(args.dueling),
)
model = DeepQ(
env=env,
policy=q_func,
learning_rate=1e-4,
buffer_size=10000,
exploration_fraction=0.1,
exploration_final_eps=0.01,
train_freq=4,
learning_starts=10000,
target_network_update_freq=1000,
gamma=0.99,
prioritized_replay=bool(args.prioritized),
prioritized_replay_alpha=args.prioritized_replay_alpha,
checkpoint_freq=args.checkpoint_freq,
checkpoint_path=args.checkpoint_path,
)
model.learn(total_timesteps=args.num_timesteps)
env.close()
def main():
"""
Run the atari test
"""
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--env',
help='environment ID',
default='BreakoutNoFrameskip-v4')
parser.add_argument('--seed', help='RNG seed', type=int, default=0)
parser.add_argument('--prioritized', type=int, default=1)
parser.add_argument('--dueling', type=int, default=1)
parser.add_argument('--prioritized-replay-alpha', type=float, default=0.6)
parser.add_argument('--num-timesteps', type=int, default=int(10e6))
args = parser.parse_args()
logger.configure()
set_global_seeds(args.seed)
env = make_atari(args.env)
env = bench.Monitor(env, logger.get_dir())
env = wrap_atari_dqn(env)
policy = partial(CnnPolicy, dueling=args.dueling == 1)
model = DQN(
env=env,
policy=policy,
learning_rate=1e-4,
buffer_size=10000,
exploration_fraction=0.1,
exploration_final_eps=0.01,
train_freq=4,
learning_starts=10000,
target_network_update_freq=1000,
gamma=0.99,
prioritized_replay=bool(args.prioritized),
prioritized_replay_alpha=args.prioritized_replay_alpha,
)
model.learn(total_timesteps=args.num_timesteps)
env.close()