Python bench.Monitor方法代码示例

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_mujoco_env(env_id, seed, reward_scale=1.0):
    """
    Create a wrapped, monitored gym.Env for MuJoCo.
    """
    rank = MPI.COMM_WORLD.Get_rank()
    myseed = seed  + 1000 * rank if seed is not None else None
    set_global_seeds(myseed)
    env = gym.make(env_id)
    logger_path = None if logger.get_dir() is None else os.path.join(logger.get_dir(), str(rank))
    env = Monitor(env, logger_path, allow_early_resets=True)
    env.seed(seed)

    if reward_scale != 1.0:
        from baselines.common.retro_wrappers import RewardScaler
        env = RewardScaler(env, reward_scale)

    return env 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def train(env_id, num_timesteps, seed):
    whoami  = mpi_fork(num_cpu)
    if whoami == "parent":
        return
    import baselines.common.tf_util as U
    logger.session().__enter__()
    sess = U.single_threaded_session()
    sess.__enter__()

    rank = MPI.COMM_WORLD.Get_rank()
    if rank != 0:
        logger.set_level(logger.DISABLED)
    workerseed = seed + 10000 * MPI.COMM_WORLD.Get_rank()
    set_global_seeds(workerseed)
    env = gym.make(env_id)
    def policy_fn(name, ob_space, ac_space):
        return MlpPolicy(name=name, ob_space=env.observation_space, ac_space=env.action_space,
            hid_size=32, num_hid_layers=2)
    env = bench.Monitor(env, osp.join(logger.get_dir(), "%i.monitor.json" % rank))
    env.seed(workerseed)
    gym.logger.setLevel(logging.WARN)

    trpo_mpi.learn(env, policy_fn, timesteps_per_batch=1024, max_kl=0.01, cg_iters=10, cg_damping=0.1,
        max_timesteps=num_timesteps, gamma=0.99, lam=0.98, vf_iters=5, vf_stepsize=1e-3)
    env.close() 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def train(env_id, num_timesteps, seed):
    from baselines.pposgd import mlp_policy, pposgd_simple
    U.make_session(num_cpu=1).__enter__()
    logger.session().__enter__()
    set_global_seeds(seed)
    env = gym.make(env_id)
    def policy_fn(name, ob_space, ac_space):
        return mlp_policy.MlpPolicy(name=name, ob_space=ob_space, ac_space=ac_space,
            hid_size=64, num_hid_layers=2)
    env = bench.Monitor(env, osp.join(logger.get_dir(), "monitor.json"))
    env.seed(seed)
    gym.logger.setLevel(logging.WARN)
    pposgd_simple.learn(env, policy_fn, 
            max_timesteps=num_timesteps,
            timesteps_per_batch=2048,
            clip_param=0.2, entcoeff=0.0,
            optim_epochs=10, optim_stepsize=3e-4, optim_batchsize=64,
            gamma=0.99, lam=0.95,
        )
    env.close() 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def main():
    logger.configure()
    env = make_atari('PongNoFrameskip-v4')
    env = bench.Monitor(env, logger.get_dir())
    env = deepq.wrap_atari_dqn(env)

    model = deepq.learn(
        env,
        "conv_only",
        convs=[(32, 8, 4), (64, 4, 2), (64, 3, 1)],
        hiddens=[256],
        dueling=True,
        lr=1e-4,
        total_timesteps=int(1e7),
        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,
    )

    model.save('pong_model.pkl')
    env.close() 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_env(env_id, env_type, subrank=0, seed=None, reward_scale=1.0, gamestate=None, wrapper_kwargs={}):
    mpi_rank = MPI.COMM_WORLD.Get_rank() if MPI else 0
    if env_type == 'atari':
        env = make_atari(env_id)
    elif env_type == 'retro':
        import retro
        gamestate = gamestate or retro.State.DEFAULT
        env = retro_wrappers.make_retro(game=env_id, max_episode_steps=10000, use_restricted_actions=retro.Actions.DISCRETE, state=gamestate)
    else:
        env = gym.make(env_id)

    env.seed(seed + subrank if seed is not None else None)
    env = Monitor(env,
                  logger.get_dir() and os.path.join(logger.get_dir(), str(mpi_rank) + '.' + str(subrank)),
                  allow_early_resets=True)

    if env_type == 'atari':
        env = wrap_deepmind(env, **wrapper_kwargs)
    elif env_type == 'retro':
        env = retro_wrappers.wrap_deepmind_retro(env, **wrapper_kwargs)

    if reward_scale != 1:
        env = retro_wrappers.RewardScaler(env, reward_scale)

    return env 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_atari_env(env_id, num_env, seed, hparams=None, wrapper_kwargs=None, start_index=0, nsteps=5, **kwargs):
    """
    Create a wrapped, monitored SubprocVecEnv for Atari.
    """
    if wrapper_kwargs is None: wrapper_kwargs = {}
    def make_env(rank): # pylint: disable=C0111
        def _thunk():
            env = make_atari(env_id)
            env.seed(seed + rank)
            env = Monitor(env, logger.get_dir() and os.path.join(logger.get_dir(), str(rank)))

            if rank == start_index and 'video_log_dir' in kwargs:
                env = VideoLogMonitor(env, kwargs['video_log_dir'] + '_rgb', write_attention_video=kwargs['write_attention_video'], hparams=hparams, nsteps=nsteps)

            return wrap_deepmind(env, **wrapper_kwargs)
        return _thunk
    set_global_seeds(seed)

    env_fns = [make_env(i + start_index) for i in range(num_env)]

    global my_subproc_vec_env
    assert my_subproc_vec_env == None
    my_subproc_vec_env = SubprocVecEnv(env_fns)

    return my_subproc_vec_env 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def _make_atari_env(env_id, num_env, seed, wrapper_kwargs=None, start_index=0):
    """
    Create a wrapped, monitored SubprocVecEnv for Atari. This is the same as the one used
     in OpenAI Baselines.
    """
    from baselines.common.vec_env.subproc_vec_env import SubprocVecEnv

    wrapper_kwargs = {} if wrapper_kwargs is None else wrapper_kwargs

    def make_env(rank):  # pylint: disable=C0111
        def _thunk():
            env = make_atari(env_id)
            env.seed(seed + rank)
            env = Monitor(
                env,
                logger.get_dir() and os.path.join(logger.get_dir(), str(rank)),
                allow_early_resets=True,
            )
            return wrap_deepmind(env, **wrapper_kwargs)

        return _thunk

    set_global_seeds(seed)
    return SubprocVecEnv([make_env(i + start_index) for i in range(num_env)]) 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def main(args):
    U.make_session(num_cpu=1).__enter__()
    set_global_seeds(args.seed)
    env = gym.make(args.env_id)

    def policy_fn(name, ob_space, ac_space, reuse=False):
        return mlp_policy.MlpPolicy(name=name, ob_space=ob_space, ac_space=ac_space,
                                    reuse=reuse, hid_size=args.policy_hidden_size, num_hid_layers=2)
    env = bench.Monitor(env, logger.get_dir() and
                        osp.join(logger.get_dir(), "monitor.json"))
    env.seed(args.seed)
    gym.logger.setLevel(logging.WARN)
    task_name = get_task_name(args)
    args.checkpoint_dir = osp.join(args.checkpoint_dir, task_name)
    args.log_dir = osp.join(args.log_dir, task_name)
    dataset = Mujoco_Dset(expert_path=args.expert_path, traj_limitation=args.traj_limitation)
    savedir_fname = learn(env,
                          policy_fn,
                          dataset,
                          max_iters=args.BC_max_iter,
                          ckpt_dir=args.checkpoint_dir,
                          log_dir=args.log_dir,
                          task_name=task_name,
                          verbose=True)
    avg_len, avg_ret = runner(env,
                              policy_fn,
                              savedir_fname,
                              timesteps_per_batch=1024,
                              number_trajs=10,
                              stochastic_policy=args.stochastic_policy,
                              save=args.save_sample,
                              reuse=True) 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def main():
    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('--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 = deepq.wrap_atari_dqn(env)
    model = deepq.models.cnn_to_mlp(
        convs=[(32, 8, 4), (64, 4, 2), (64, 3, 1)],
        hiddens=[256],
        dueling=bool(args.dueling),
    )
    act = deepq.learn(
        env,
        q_func=model,
        lr=1e-4,
        max_timesteps=args.num_timesteps,
        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)
    )
    # act.save("pong_model.pkl") XXX
    env.close() 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def train(env_id, num_timesteps, seed):
    from baselines.trpo_mpi.nosharing_cnn_policy import CnnPolicy
    from baselines.trpo_mpi import trpo_mpi
    import baselines.common.tf_util as U
    rank = MPI.COMM_WORLD.Get_rank()
    sess = U.single_threaded_session()
    sess.__enter__()
    if rank == 0:
        logger.configure()
    else:
        logger.configure(format_strs=[])

    workerseed = seed + 10000 * MPI.COMM_WORLD.Get_rank()
    set_global_seeds(workerseed)
    env = make_atari(env_id)
    def policy_fn(name, ob_space, ac_space): #pylint: disable=W0613
        return CnnPolicy(name=name, ob_space=env.observation_space, ac_space=env.action_space)
    env = bench.Monitor(env, logger.get_dir() and osp.join(logger.get_dir(), str(rank)))
    env.seed(workerseed)

    env = wrap_deepmind(env)
    env.seed(workerseed)

    trpo_mpi.learn(env, policy_fn, timesteps_per_batch=512, max_kl=0.001, cg_iters=10, cg_damping=1e-3,
        max_timesteps=int(num_timesteps * 1.1), gamma=0.98, lam=1.0, vf_iters=3, vf_stepsize=1e-4, entcoeff=0.00)
    env.close() 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def train(env_id, num_timesteps, seed, policy, r_ex_coef, r_in_coef, lr_alpha, lr_beta, reward_freq):
    from baselines.common import set_global_seeds
    from baselines.common.vec_env.vec_normalize import VecNormalize
    from baselines.ppo2 import ppo2
    from baselines.ppo2.policies import MlpPolicy, MlpPolicyIntrinsicReward
    import gym
    import tensorflow as tf
    from baselines.common.vec_env.dummy_vec_env import DummyVecEnv
    ncpu = 1
    config = tf.ConfigProto(allow_soft_placement=True,
                            intra_op_parallelism_threads=ncpu,
                            inter_op_parallelism_threads=ncpu)
    config.gpu_options.allow_growth = True
    tf.Session(config=config).__enter__()
    def make_env():
        env = gym.make(env_id)
        env = bench.Monitor(env, logger.get_dir())
        return env
    env = DummyVecEnv([make_env])
    env = VecNormalize(env)

    set_global_seeds(seed)
    if policy == 'mlp':
        policy = MlpPolicy
    elif policy == 'mlp_int':
        policy = MlpPolicyIntrinsicReward
    else:
        raise NotImplementedError
    ppo2.learn(policy=policy, env=env, nsteps=2048, nminibatches=32,
        lam=0.95, gamma=0.99, noptepochs=10, log_interval=1,
        ent_coef=0.0,
        lr_alpha=lr_alpha,
        cliprange=0.2,
        total_timesteps=num_timesteps,
        r_ex_coef=r_ex_coef,
        r_in_coef=r_in_coef,
        lr_beta=lr_beta,
        reward_freq=reward_freq) 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_atari_env(env_id, num_env, seed, wrapper_kwargs=None, start_index=0):
    """
    Create a wrapped, monitored SubprocVecEnv for Atari.
    """
    if wrapper_kwargs is None: wrapper_kwargs = {}
    def make_env(rank): # pylint: disable=C0111
        def _thunk():
            env = make_atari(env_id)
            env.seed(seed + rank)
            env = Monitor(env, logger.get_dir() and os.path.join(logger.get_dir(), str(rank)))
            return wrap_deepmind(env, **wrapper_kwargs)
        return _thunk
    set_global_seeds(seed)
    return SubprocVecEnv([make_env(i + start_index) for i in range(num_env)]) 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_mujoco_env(env_id, seed):
    """
    Create a wrapped, monitored gym.Env for MuJoCo.
    """
    set_global_seeds(seed)
    env = gym.make(env_id)
    env = Monitor(env, logger.get_dir())
    env.seed(seed)
    return env 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def make_robotics_env(env_id, seed, rank=0):
    """
    Create a wrapped, monitored gym.Env for MuJoCo.
    """
    set_global_seeds(seed)
    env = gym.make(env_id)
    env = FlattenDictWrapper(env, ['observation', 'desired_goal'])
    env = Monitor(
        env, logger.get_dir() and os.path.join(logger.get_dir(), str(rank)),
        info_keywords=('is_success',))
    env.seed(seed)
    return env 

# 需要导入模块: from baselines import bench [as 别名]
# 或者: from baselines.bench import Monitor [as 别名]
def main():
    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 = deepq.wrap_atari_dqn(env)

    deepq.learn(
        env,
        "conv_only",
        convs=[(32, 8, 4), (64, 4, 2), (64, 3, 1)],
        hiddens=[256],
        dueling=bool(args.dueling),
        lr=1e-4,
        total_timesteps=args.num_timesteps,
        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,
    )

    env.close()