訓練一個玩瑪利歐的 RL 代理程式

建立於：2020 年 12 月 17 日 | 最後更新：2024 年 2 月 05 日 | 最後驗證：未驗證

作者： Yuansong Feng, Suraj Subramanian, Howard Wang, Steven Guo.

本教學課程將引導您了解深度強化學習的基礎知識。最後，您將實作一個由 AI 驅動的瑪利歐（使用 Double Deep Q-Networks），它可以自己玩遊戲。

雖然本教學課程不需要事先了解 RL，但您可以熟悉這些 RL 概念，並將此方便的 速查表 作為您的伴侶。完整的程式碼可在此處找到 這裡。

%%bash
pip install gym-super-mario-bros==7.4.0
pip install tensordict==0.3.0
pip install torchrl==0.3.0

import torch
from torch import nn
from torchvision import transforms as T
from PIL import Image
import numpy as np
from pathlib import Path
from collections import deque
import random, datetime, os

# Gym is an OpenAI toolkit for RL
import gym
from gym.spaces import Box
from gym.wrappers import FrameStack

# NES Emulator for OpenAI Gym
from nes_py.wrappers import JoypadSpace

# Super Mario environment for OpenAI Gym
import gym_super_mario_bros

from tensordict import TensorDict
from torchrl.data import TensorDictReplayBuffer, LazyMemmapStorage

RL 定義

環境 代理程式與之互動並從中學習的世界。

動作 \(a\)：代理程式如何回應環境。所有可能動作的集合稱為動作空間。

狀態 \(s\)：環境的目前特徵。環境可以處於的所有可能狀態的集合稱為狀態空間。

獎勵 \(r\)：獎勵是環境對代理程式的關鍵回饋。它是驅使代理程式學習並改變其未來動作的原因。多個時間步長的獎勵聚合稱為 回報。

最佳動作值函數 \(Q^*(s,a)\)：如果您從狀態 \(s\) 開始，採取任意動作 \(a\)，然後對於每個未來的時間步長採取最大化回報的動作，則給出預期的回報。\(Q\) 可以說是代表狀態中動作的「品質」。我們嘗試近似此函數。

環境

初始化環境

在瑪利歐中，環境由水管、蘑菇和其他元件組成。

當瑪利歐採取動作時，環境會以變更的（下一個）狀態、獎勵和其他資訊來回應。

# Initialize Super Mario environment (in v0.26 change render mode to 'human' to see results on the screen)
if gym.__version__ < '0.26':
    env = gym_super_mario_bros.make("SuperMarioBros-1-1-v0", new_step_api=True)
else:
    env = gym_super_mario_bros.make("SuperMarioBros-1-1-v0", render_mode='rgb', apply_api_compatibility=True)

# Limit the action-space to
#   0. walk right
#   1. jump right
env = JoypadSpace(env, [["right"], ["right", "A"]])

env.reset()
next_state, reward, done, trunc, info = env.step(action=0)
print(f"{next_state.shape},\n {reward},\n {done},\n {info}")

/usr/local/lib/python3.10/dist-packages/gym/envs/registration.py:555: UserWarning:

WARN: The environment SuperMarioBros-1-1-v0 is out of date. You should consider upgrading to version `v3`.

/usr/local/lib/python3.10/dist-packages/gym/envs/registration.py:627: UserWarning:

WARN: The environment creator metadata doesn't include `render_modes`, contains: ['render.modes', 'video.frames_per_second']

/usr/local/lib/python3.10/dist-packages/gym/utils/passive_env_checker.py:233: DeprecationWarning:

`np.bool8` is a deprecated alias for `np.bool_`.  (Deprecated NumPy 1.24)

(240, 256, 3),
 0.0,
 False,
 {'coins': 0, 'flag_get': False, 'life': 2, 'score': 0, 'stage': 1, 'status': 'small', 'time': 400, 'world': 1, 'x_pos': 40, 'y_pos': 79}

預處理環境

環境資料以 next_state 的形式返回給代理程式。如您在上面看到的，每個狀態都由一個 [3, 240, 256] 大小的陣列表示。通常，這比我們的代理程式需要的資訊更多；例如，瑪利歐的動作不取決於水管或天空的顏色！

我們使用 Wrappers 在將環境資料傳送給代理程式之前對其進行預處理。

GrayScaleObservation 是一個常見的 Wrapper，用於將 RGB 影像轉換為灰階；這樣做可以減少狀態表示的大小，而不會遺失有用的資訊。現在每個狀態的大小：[1, 240, 256]

ResizeObservation 將每個觀察值下取樣為方形影像。新大小：[1, 84, 84]

SkipFrame 是一個自訂 Wrapper，它繼承自 gym.Wrapper 並實作 step() 函數。因為連續影格變化不大，所以我們可以跳過 n 個中間影格，而不會遺失太多資訊。第 n 個影格會聚合每個跳過影格累積的獎勵。

FrameStack 是一個封裝器，能讓我們將環境中連續的影格壓縮成單一的觀察點，以便餵給我們的學習模型。這樣一來，我們就能根據 Mario 在前幾個影格中的移動方向來判斷他是正在著陸還是跳躍。

class SkipFrame(gym.Wrapper):
    def __init__(self, env, skip):
        """Return only every `skip`-th frame"""
        super().__init__(env)
        self._skip = skip

    def step(self, action):
        """Repeat action, and sum reward"""
        total_reward = 0.0
        for i in range(self._skip):
            # Accumulate reward and repeat the same action
            obs, reward, done, trunk, info = self.env.step(action)
            total_reward += reward
            if done:
                break
        return obs, total_reward, done, trunk, info


class GrayScaleObservation(gym.ObservationWrapper):
    def __init__(self, env):
        super().__init__(env)
        obs_shape = self.observation_space.shape[:2]
        self.observation_space = Box(low=0, high=255, shape=obs_shape, dtype=np.uint8)

    def permute_orientation(self, observation):
        # permute [H, W, C] array to [C, H, W] tensor
        observation = np.transpose(observation, (2, 0, 1))
        observation = torch.tensor(observation.copy(), dtype=torch.float)
        return observation

    def observation(self, observation):
        observation = self.permute_orientation(observation)
        transform = T.Grayscale()
        observation = transform(observation)
        return observation


class ResizeObservation(gym.ObservationWrapper):
    def __init__(self, env, shape):
        super().__init__(env)
        if isinstance(shape, int):
            self.shape = (shape, shape)
        else:
            self.shape = tuple(shape)

        obs_shape = self.shape + self.observation_space.shape[2:]
        self.observation_space = Box(low=0, high=255, shape=obs_shape, dtype=np.uint8)

    def observation(self, observation):
        transforms = T.Compose(
            [T.Resize(self.shape, antialias=True), T.Normalize(0, 255)]
        )
        observation = transforms(observation).squeeze(0)
        return observation


# Apply Wrappers to environment
env = SkipFrame(env, skip=4)
env = GrayScaleObservation(env)
env = ResizeObservation(env, shape=84)
if gym.__version__ < '0.26':
    env = FrameStack(env, num_stack=4, new_step_api=True)
else:
    env = FrameStack(env, num_stack=4)

將上述封裝器應用到環境後，最終封裝的狀態包含 4 個堆疊在一起的灰階連續影格，如左側圖片所示。每次 Mario 採取一個動作，環境就會回應該結構的狀態。該結構由大小為 [4, 84, 84] 的 3D 陣列表示。

代理人 (Agent)

我們建立一個類別 Mario 來代表我們在遊戲中的代理人。Mario 應該能夠：

• 根據當前狀態（環境）按照最佳動作策略來行動 (Act)。

• 記住 (Remember) 經驗。經驗 = (當前狀態, 當前動作, 獎勵, 下一個狀態)。Mario 會快取 (caches) 並在之後回憶 (recalls) 他的經驗來更新他的動作策略。

• 學習 (Learn) 隨著時間推移，學習更好的動作策略。

class Mario:
    def __init__():
        pass

    def act(self, state):
        """Given a state, choose an epsilon-greedy action"""
        pass

    def cache(self, experience):
        """Add the experience to memory"""
        pass

    def recall(self):
        """Sample experiences from memory"""
        pass

    def learn(self):
        """Update online action value (Q) function with a batch of experiences"""
        pass

在接下來的章節中，我們將填寫 Mario 的參數並定義他的函數。

行動 (Act)

對於任何給定的狀態，代理人可以選擇執行最佳動作（利用 (exploit)）或隨機動作（探索 (explore)）。

Mario 以 self.exploration_rate 的機率隨機探索；當他選擇利用時，他依靠 MarioNet（在 Learn 區段中實作）來提供最佳動作。

class Mario:
    def __init__(self, state_dim, action_dim, save_dir):
        self.state_dim = state_dim
        self.action_dim = action_dim
        self.save_dir = save_dir

        self.device = "cuda" if torch.cuda.is_available() else "cpu"

        # Mario's DNN to predict the most optimal action - we implement this in the Learn section
        self.net = MarioNet(self.state_dim, self.action_dim).float()
        self.net = self.net.to(device=self.device)

        self.exploration_rate = 1
        self.exploration_rate_decay = 0.99999975
        self.exploration_rate_min = 0.1
        self.curr_step = 0

        self.save_every = 5e5  # no. of experiences between saving Mario Net

    def act(self, state):
        """
    Given a state, choose an epsilon-greedy action and update value of step.

    Inputs:
    state(``LazyFrame``): A single observation of the current state, dimension is (state_dim)
    Outputs:
    ``action_idx`` (``int``): An integer representing which action Mario will perform
    """
        # EXPLORE
        if np.random.rand() < self.exploration_rate:
            action_idx = np.random.randint(self.action_dim)

        # EXPLOIT
        else:
            state = state[0].__array__() if isinstance(state, tuple) else state.__array__()
            state = torch.tensor(state, device=self.device).unsqueeze(0)
            action_values = self.net(state, model="online")
            action_idx = torch.argmax(action_values, axis=1).item()

        # decrease exploration_rate
        self.exploration_rate *= self.exploration_rate_decay
        self.exploration_rate = max(self.exploration_rate_min, self.exploration_rate)

        # increment step
        self.curr_step += 1
        return action_idx

快取與回憶 (Cache and Recall)

這兩個函數充當 Mario 的「記憶」過程。

cache()：每次 Mario 執行一個動作，他都會將 experience 儲存到他的記憶中。他的經驗包括目前的狀態 (state)、執行的動作 (action)、來自該動作的獎勵 (reward)、下一個狀態 (next state) 以及遊戲是否結束 (done)。

recall()：Mario 從他的記憶中隨機抽取一批經驗，並用它來學習遊戲。

class Mario(Mario):  # subclassing for continuity
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.memory = TensorDictReplayBuffer(storage=LazyMemmapStorage(100000, device=torch.device("cpu")))
        self.batch_size = 32

    def cache(self, state, next_state, action, reward, done):
        """
        Store the experience to self.memory (replay buffer)

        Inputs:
        state (``LazyFrame``),
        next_state (``LazyFrame``),
        action (``int``),
        reward (``float``),
        done(``bool``))
        """
        def first_if_tuple(x):
            return x[0] if isinstance(x, tuple) else x
        state = first_if_tuple(state).__array__()
        next_state = first_if_tuple(next_state).__array__()

        state = torch.tensor(state)
        next_state = torch.tensor(next_state)
        action = torch.tensor([action])
        reward = torch.tensor([reward])
        done = torch.tensor([done])

        # self.memory.append((state, next_state, action, reward, done,))
        self.memory.add(TensorDict({"state": state, "next_state": next_state, "action": action, "reward": reward, "done": done}, batch_size=[]))

    def recall(self):
        """
        Retrieve a batch of experiences from memory
        """
        batch = self.memory.sample(self.batch_size).to(self.device)
        state, next_state, action, reward, done = (batch.get(key) for key in ("state", "next_state", "action", "reward", "done"))
        return state, next_state, action.squeeze(), reward.squeeze(), done.squeeze()

學習 (Learn)

Mario 在底層使用 DDQN 演算法。DDQN 使用兩個 ConvNets - \(Q_{online}\) 和 \(Q_{target}\) - 來獨立地近似最佳動作價值函數。

在我們的實作中，我們跨 \(Q_{online}\) 和 \(Q_{target}\) 分享特徵生成器 features，但為每個網路維護單獨的 FC 分類器。\(\theta_{target}\)（\(Q_{target}\) 的參數）被凍結以防止通過反向傳播進行更新。相反，它會定期與 \(\theta_{online}\) 同步（稍後會詳細介紹）。

神經網路 (Neural Network)

class MarioNet(nn.Module):
    """mini CNN structure
  input -> (conv2d + relu) x 3 -> flatten -> (dense + relu) x 2 -> output
  """

    def __init__(self, input_dim, output_dim):
        super().__init__()
        c, h, w = input_dim

        if h != 84:
            raise ValueError(f"Expecting input height: 84, got: {h}")
        if w != 84:
            raise ValueError(f"Expecting input width: 84, got: {w}")

        self.online = self.__build_cnn(c, output_dim)

        self.target = self.__build_cnn(c, output_dim)
        self.target.load_state_dict(self.online.state_dict())

        # Q_target parameters are frozen.
        for p in self.target.parameters():
            p.requires_grad = False

    def forward(self, input, model):
        if model == "online":
            return self.online(input)
        elif model == "target":
            return self.target(input)

    def __build_cnn(self, c, output_dim):
        return nn.Sequential(
            nn.Conv2d(in_channels=c, out_channels=32, kernel_size=8, stride=4),
            nn.ReLU(),
            nn.Conv2d(in_channels=32, out_channels=64, kernel_size=4, stride=2),
            nn.ReLU(),
            nn.Conv2d(in_channels=64, out_channels=64, kernel_size=3, stride=1),
            nn.ReLU(),
            nn.Flatten(),
            nn.Linear(3136, 512),
            nn.ReLU(),
            nn.Linear(512, output_dim),
        )

TD 估計 & TD 目標 (TD Estimate & TD Target)

學習涉及兩個值：

TD 估計 (TD Estimate) - 針對給定狀態 \(s\) 預測的最佳 \(Q^*\)。

\[{TD}_e = Q_{online}^*(s,a)\]

TD 目標 (TD Target) - 目前獎勵和下一個狀態 \(s'\) 中估計的 \(Q^*\) 的聚合。

\[a' = argmax_{a} Q_{online}(s', a)\]

\[{TD}_t = r + \gamma Q_{target}^*(s',a')\]

因為我們不知道下一個動作 \(a'\) 會是什麼，所以我們使用最大化下一個狀態 \(s'\) 中 \(Q_{online}\) 的動作 \(a'\)。

請注意，我們在 td_target() 上使用了 @torch.no_grad() 裝飾器來停用此處的梯度計算（因為我們不需要在 \(\theta_{target}\) 上進行反向傳播）。

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.gamma = 0.9

    def td_estimate(self, state, action):
        current_Q = self.net(state, model="online")[
            np.arange(0, self.batch_size), action
        ]  # Q_online(s,a)
        return current_Q

    @torch.no_grad()
    def td_target(self, reward, next_state, done):
        next_state_Q = self.net(next_state, model="online")
        best_action = torch.argmax(next_state_Q, axis=1)
        next_Q = self.net(next_state, model="target")[
            np.arange(0, self.batch_size), best_action
        ]
        return (reward + (1 - done.float()) * self.gamma * next_Q).float()

更新模型 (Updating the model)

當 Mario 從他的重播緩衝區中採樣輸入時，我們計算 \(TD_t\) 和 \(TD_e\) 並將此損失反向傳播到 \(Q_{online}\) 以更新其參數 \(\theta_{online}\)（\(\alpha\) 是傳遞給 optimizer 的學習率 lr）。

\[\theta_{online} \leftarrow \theta_{online} + \alpha \nabla(TD_e - TD_t)\]

\(\theta_{target}\) 不會通過反向傳播進行更新。相反，我們會定期將 \(\theta_{online}\) 複製到 \(\theta_{target}\)。

\[\theta_{target} \leftarrow \theta_{online}\]

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.optimizer = torch.optim.Adam(self.net.parameters(), lr=0.00025)
        self.loss_fn = torch.nn.SmoothL1Loss()

    def update_Q_online(self, td_estimate, td_target):
        loss = self.loss_fn(td_estimate, td_target)
        self.optimizer.zero_grad()
        loss.backward()
        self.optimizer.step()
        return loss.item()

    def sync_Q_target(self):
        self.net.target.load_state_dict(self.net.online.state_dict())

儲存檢查點 (Save checkpoint)

class Mario(Mario):
    def save(self):
        save_path = (
            self.save_dir / f"mario_net_{int(self.curr_step // self.save_every)}.chkpt"
        )
        torch.save(
            dict(model=self.net.state_dict(), exploration_rate=self.exploration_rate),
            save_path,
        )
        print(f"MarioNet saved to {save_path} at step {self.curr_step}")

將它們放在一起 (Putting it all together)

class Mario(Mario):
    def __init__(self, state_dim, action_dim, save_dir):
        super().__init__(state_dim, action_dim, save_dir)
        self.burnin = 1e4  # min. experiences before training
        self.learn_every = 3  # no. of experiences between updates to Q_online
        self.sync_every = 1e4  # no. of experiences between Q_target & Q_online sync

    def learn(self):
        if self.curr_step % self.sync_every == 0:
            self.sync_Q_target()

        if self.curr_step % self.save_every == 0:
            self.save()

        if self.curr_step < self.burnin:
            return None, None

        if self.curr_step % self.learn_every != 0:
            return None, None

        # Sample from memory
        state, next_state, action, reward, done = self.recall()

        # Get TD Estimate
        td_est = self.td_estimate(state, action)

        # Get TD Target
        td_tgt = self.td_target(reward, next_state, done)

        # Backpropagate loss through Q_online
        loss = self.update_Q_online(td_est, td_tgt)

        return (td_est.mean().item(), loss)

日誌 (Logging)

import numpy as np
import time, datetime
import matplotlib.pyplot as plt


class MetricLogger:
    def __init__(self, save_dir):
        self.save_log = save_dir / "log"
        with open(self.save_log, "w") as f:
            f.write(
                f"{'Episode':>8}{'Step':>8}{'Epsilon':>10}{'MeanReward':>15}"
                f"{'MeanLength':>15}{'MeanLoss':>15}{'MeanQValue':>15}"
                f"{'TimeDelta':>15}{'Time':>20}\n"
            )
        self.ep_rewards_plot = save_dir / "reward_plot.jpg"
        self.ep_lengths_plot = save_dir / "length_plot.jpg"
        self.ep_avg_losses_plot = save_dir / "loss_plot.jpg"
        self.ep_avg_qs_plot = save_dir / "q_plot.jpg"

        # History metrics
        self.ep_rewards = []
        self.ep_lengths = []
        self.ep_avg_losses = []
        self.ep_avg_qs = []

        # Moving averages, added for every call to record()
        self.moving_avg_ep_rewards = []
        self.moving_avg_ep_lengths = []
        self.moving_avg_ep_avg_losses = []
        self.moving_avg_ep_avg_qs = []

        # Current episode metric
        self.init_episode()

        # Timing
        self.record_time = time.time()

    def log_step(self, reward, loss, q):
        self.curr_ep_reward += reward
        self.curr_ep_length += 1
        if loss:
            self.curr_ep_loss += loss
            self.curr_ep_q += q
            self.curr_ep_loss_length += 1

    def log_episode(self):
        "Mark end of episode"
        self.ep_rewards.append(self.curr_ep_reward)
        self.ep_lengths.append(self.curr_ep_length)
        if self.curr_ep_loss_length == 0:
            ep_avg_loss = 0
            ep_avg_q = 0
        else:
            ep_avg_loss = np.round(self.curr_ep_loss / self.curr_ep_loss_length, 5)
            ep_avg_q = np.round(self.curr_ep_q / self.curr_ep_loss_length, 5)
        self.ep_avg_losses.append(ep_avg_loss)
        self.ep_avg_qs.append(ep_avg_q)

        self.init_episode()

    def init_episode(self):
        self.curr_ep_reward = 0.0
        self.curr_ep_length = 0
        self.curr_ep_loss = 0.0
        self.curr_ep_q = 0.0
        self.curr_ep_loss_length = 0

    def record(self, episode, epsilon, step):
        mean_ep_reward = np.round(np.mean(self.ep_rewards[-100:]), 3)
        mean_ep_length = np.round(np.mean(self.ep_lengths[-100:]), 3)
        mean_ep_loss = np.round(np.mean(self.ep_avg_losses[-100:]), 3)
        mean_ep_q = np.round(np.mean(self.ep_avg_qs[-100:]), 3)
        self.moving_avg_ep_rewards.append(mean_ep_reward)
        self.moving_avg_ep_lengths.append(mean_ep_length)
        self.moving_avg_ep_avg_losses.append(mean_ep_loss)
        self.moving_avg_ep_avg_qs.append(mean_ep_q)

        last_record_time = self.record_time
        self.record_time = time.time()
        time_since_last_record = np.round(self.record_time - last_record_time, 3)

        print(
            f"Episode {episode} - "
            f"Step {step} - "
            f"Epsilon {epsilon} - "
            f"Mean Reward {mean_ep_reward} - "
            f"Mean Length {mean_ep_length} - "
            f"Mean Loss {mean_ep_loss} - "
            f"Mean Q Value {mean_ep_q} - "
            f"Time Delta {time_since_last_record} - "
            f"Time {datetime.datetime.now().strftime('%Y-%m-%dT%H:%M:%S')}"
        )

        with open(self.save_log, "a") as f:
            f.write(
                f"{episode:8d}{step:8d}{epsilon:10.3f}"
                f"{mean_ep_reward:15.3f}{mean_ep_length:15.3f}{mean_ep_loss:15.3f}{mean_ep_q:15.3f}"
                f"{time_since_last_record:15.3f}"
                f"{datetime.datetime.now().strftime('%Y-%m-%dT%H:%M:%S'):>20}\n"
            )

        for metric in ["ep_lengths", "ep_avg_losses", "ep_avg_qs", "ep_rewards"]:
            plt.clf()
            plt.plot(getattr(self, f"moving_avg_{metric}"), label=f"moving_avg_{metric}")
            plt.legend()
            plt.savefig(getattr(self, f"{metric}_plot"))

讓我們開始玩吧！ (Let's play!)

在本範例中，我們將訓練迴圈執行 40 個回合，但為了讓 Mario 真正學習他世界的方式，我們建議將迴圈執行至少 40,000 個回合！

use_cuda = torch.cuda.is_available()
print(f"Using CUDA: {use_cuda}")
print()

save_dir = Path("checkpoints") / datetime.datetime.now().strftime("%Y-%m-%dT%H-%M-%S")
save_dir.mkdir(parents=True)

mario = Mario(state_dim=(4, 84, 84), action_dim=env.action_space.n, save_dir=save_dir)

logger = MetricLogger(save_dir)

episodes = 40
for e in range(episodes):

    state = env.reset()

    # Play the game!
    while True:

        # Run agent on the state
        action = mario.act(state)

        # Agent performs action
        next_state, reward, done, trunc, info = env.step(action)

        # Remember
        mario.cache(state, next_state, action, reward, done)

        # Learn
        q, loss = mario.learn()

        # Logging
        logger.log_step(reward, loss, q)

        # Update state
        state = next_state

        # Check if end of game
        if done or info["flag_get"]:
            break

    logger.log_episode()

    if (e % 20 == 0) or (e == episodes - 1):
        logger.record(episode=e, epsilon=mario.exploration_rate, step=mario.curr_step)

Using CUDA: True

Episode 0 - Step 163 - Epsilon 0.9999592508251706 - Mean Reward 635.0 - Mean Length 163.0 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 1.881 - Time 2025-02-03T17:21:52
Episode 20 - Step 5007 - Epsilon 0.9987490329557962 - Mean Reward 667.429 - Mean Length 238.429 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 57.036 - Time 2025-02-03T17:22:49
Episode 39 - Step 8854 - Epsilon 0.9977889477081997 - Mean Reward 656.6 - Mean Length 221.35 - Mean Loss 0.0 - Mean Q Value 0.0 - Time Delta 46.039 - Time 2025-02-03T17:23:35

結論 (Conclusion)

在本教學課程中，我們看到了如何使用 PyTorch 來訓練一個玩遊戲的 AI。您可以使用相同的方法來訓練 AI 玩 OpenAI gym 中的任何遊戲。希望您喜歡本教學課程，請隨時透過 我們的 github 與我們聯繫！