{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "bpOEhPfXa9Nz" }, "source": [ "## HW2: Value Based Methods (DQN vs DDQN)" ] }, { "cell_type": "markdown", "metadata": { "id": "zSHRAecMbEKC" }, "source": [ "> - Full Name: **[Full Name]**\n", "> - Student ID: **[Stundet ID]**" ] }, { "cell_type": "markdown", "metadata": { "id": "NsmqMy9fZSgU" }, "source": [ "## **Overview**\n", "\n", "DQN, introduced in 2013, revolutionized deep reinforcement learning. \n", "In this notebook, you'll use PyTorch to train a Deep Q-Learning (DQN) agent on the [Cart-Pole](https://gymnasium.farama.org/environments/classic_control/cart_pole/) task from [Gymnasium](https://gymnasium.farama.org/). You'll also implement Double DQN (DDQN), an improved version with better stability, convergence, and test performance.\n", "\n", "### In this notebook:\n", "- Explore the Cart-Pole environment and observe an untrained agent.\n", "- Set up a Gymnasium environment.\n", "- Implement and train DQN and DDQN from scratch.\n", "- Compare their performance to understand strengths and weaknesses.\n", "- Render the trained agent’s behavior.\n", "\n", "Before starting, import the necessary packages. Helper functions are provided for visualization and rendering.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "O_4FcKWExplP" }, "source": [ "## **Setup** \n", "\n", "First, install the required packages. If you're using Colab, everything should work smoothly. However, on a local system, you may encounter some dependency installation challenges. \n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "H6T0r2sdxplQ", "outputId": "567fa4d6-187f-4179-ca79-ee5e4c8b0322" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Get:1 https://cloud.r-project.org/bin/linux/ubuntu jammy-cran40/ InRelease [3,632 B]\n", "Get:2 https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64 InRelease [1,581 B]\n", "Get:3 https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64 Packages [1,315 kB]\n", "Hit:4 http://archive.ubuntu.com/ubuntu jammy InRelease\n", "Get:5 http://security.ubuntu.com/ubuntu jammy-security InRelease [129 kB]\n", "Get:6 http://archive.ubuntu.com/ubuntu jammy-updates InRelease [128 kB]\n", "Hit:7 https://ppa.launchpadcontent.net/deadsnakes/ppa/ubuntu jammy InRelease\n", "Hit:8 https://ppa.launchpadcontent.net/graphics-drivers/ppa/ubuntu jammy InRelease\n", "Hit:9 https://ppa.launchpadcontent.net/ubuntugis/ppa/ubuntu jammy InRelease\n", "Get:10 https://r2u.stat.illinois.edu/ubuntu jammy InRelease [6,555 B]\n", "Get:11 http://archive.ubuntu.com/ubuntu jammy-backports InRelease [127 kB]\n", "Get:12 https://r2u.stat.illinois.edu/ubuntu jammy/main amd64 Packages [2,656 kB]\n", "Get:13 http://archive.ubuntu.com/ubuntu jammy-updates/main amd64 Packages [2,911 kB]\n", "Get:14 http://archive.ubuntu.com/ubuntu jammy-updates/universe amd64 Packages [1,526 kB]\n", "Get:15 https://r2u.stat.illinois.edu/ubuntu jammy/main all Packages [8,677 kB]\n", "Fetched 17.5 MB in 5s (3,184 kB/s)\n", "Reading package lists...\n", "W: Skipping acquire of configured file 'main/source/Sources' as repository 'https://r2u.stat.illinois.edu/ubuntu jammy InRelease' does not seem to provide it (sources.list entry misspelt?)\n", "Reading package lists...\n", "Building dependency tree...\n", "Reading state information...\n", "ffmpeg is already the newest version (7:4.4.2-0ubuntu0.22.04.1).\n", "The following additional packages will be installed:\n", " libfontenc1 libxfont2 libxkbfile1 x11-xkb-utils xfonts-base xfonts-encodings\n", " xfonts-utils xserver-common\n", "The following NEW packages will be installed:\n", " libfontenc1 libxfont2 libxkbfile1 x11-xkb-utils xfonts-base xfonts-encodings\n", " xfonts-utils xserver-common xvfb\n", "0 upgraded, 9 newly installed, 0 to remove and 22 not upgraded.\n", "Need to get 7,815 kB of archives.\n", "After this operation, 11.9 MB of additional disk space will be used.\n", "Get:1 http://archive.ubuntu.com/ubuntu jammy/main amd64 libfontenc1 amd64 1:1.1.4-1build3 [14.7 kB]\n", "Get:2 http://archive.ubuntu.com/ubuntu jammy/main amd64 libxfont2 amd64 1:2.0.5-1build1 [94.5 kB]\n", "Get:3 http://archive.ubuntu.com/ubuntu jammy/main amd64 libxkbfile1 amd64 1:1.1.0-1build3 [71.8 kB]\n", "Get:4 http://archive.ubuntu.com/ubuntu jammy/main amd64 x11-xkb-utils amd64 7.7+5build4 [172 kB]\n", "Get:5 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-encodings all 1:1.0.5-0ubuntu2 [578 kB]\n", "Get:6 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-utils amd64 1:7.7+6build2 [94.6 kB]\n", "Get:7 http://archive.ubuntu.com/ubuntu jammy/main amd64 xfonts-base all 1:1.0.5 [5,896 kB]\n", "Get:8 http://archive.ubuntu.com/ubuntu jammy-updates/main amd64 xserver-common all 2:21.1.4-2ubuntu1.7~22.04.12 [28.7 kB]\n", "Get:9 http://archive.ubuntu.com/ubuntu jammy-updates/universe amd64 xvfb amd64 2:21.1.4-2ubuntu1.7~22.04.12 [864 kB]\n", "Fetched 7,815 kB in 2s (3,620 kB/s)\n", "debconf: unable to initialize frontend: Dialog\n", "debconf: (No usable dialog-like program is installed, so the dialog based frontend cannot be used. at /usr/share/perl5/Debconf/FrontEnd/Dialog.pm line 78, <> line 9.)\n", "debconf: falling back to frontend: Readline\n", "debconf: unable to initialize frontend: Readline\n", "debconf: (This frontend requires a controlling tty.)\n", "debconf: falling back to frontend: Teletype\n", "dpkg-preconfigure: unable to re-open stdin: \n", "Selecting previously unselected package libfontenc1:amd64.\n", "(Reading database ... 124926 files and directories currently installed.)\n", "Preparing to unpack .../0-libfontenc1_1%3a1.1.4-1build3_amd64.deb ...\n", "Unpacking libfontenc1:amd64 (1:1.1.4-1build3) ...\n", "Selecting previously unselected package libxfont2:amd64.\n", "Preparing to unpack .../1-libxfont2_1%3a2.0.5-1build1_amd64.deb ...\n", "Unpacking libxfont2:amd64 (1:2.0.5-1build1) ...\n", "Selecting previously unselected package libxkbfile1:amd64.\n", "Preparing to unpack .../2-libxkbfile1_1%3a1.1.0-1build3_amd64.deb ...\n", "Unpacking libxkbfile1:amd64 (1:1.1.0-1build3) ...\n", "Selecting previously unselected package x11-xkb-utils.\n", "Preparing to unpack .../3-x11-xkb-utils_7.7+5build4_amd64.deb ...\n", "Unpacking x11-xkb-utils (7.7+5build4) ...\n", "Selecting previously unselected package xfonts-encodings.\n", "Preparing to unpack .../4-xfonts-encodings_1%3a1.0.5-0ubuntu2_all.deb ...\n", "Unpacking xfonts-encodings (1:1.0.5-0ubuntu2) ...\n", "Selecting previously unselected package xfonts-utils.\n", "Preparing to unpack .../5-xfonts-utils_1%3a7.7+6build2_amd64.deb ...\n", "Unpacking xfonts-utils (1:7.7+6build2) ...\n", "Selecting previously unselected package xfonts-base.\n", "Preparing to unpack .../6-xfonts-base_1%3a1.0.5_all.deb ...\n", "Unpacking xfonts-base (1:1.0.5) ...\n", "Selecting previously unselected package xserver-common.\n", "Preparing to unpack .../7-xserver-common_2%3a21.1.4-2ubuntu1.7~22.04.12_all.deb ...\n", "Unpacking xserver-common (2:21.1.4-2ubuntu1.7~22.04.12) ...\n", "Selecting previously unselected package xvfb.\n", "Preparing to unpack .../8-xvfb_2%3a21.1.4-2ubuntu1.7~22.04.12_amd64.deb ...\n", "Unpacking xvfb (2:21.1.4-2ubuntu1.7~22.04.12) ...\n", "Setting up libfontenc1:amd64 (1:1.1.4-1build3) ...\n", "Setting up xfonts-encodings (1:1.0.5-0ubuntu2) ...\n", "Setting up libxkbfile1:amd64 (1:1.1.0-1build3) ...\n", "Setting up libxfont2:amd64 (1:2.0.5-1build1) ...\n", "Setting up x11-xkb-utils (7.7+5build4) ...\n", "Setting up xfonts-utils (1:7.7+6build2) ...\n", "Setting up xfonts-base (1:1.0.5) ...\n", "Setting up xserver-common (2:21.1.4-2ubuntu1.7~22.04.12) ...\n", "Setting up xvfb (2:21.1.4-2ubuntu1.7~22.04.12) ...\n", "Processing triggers for man-db (2.10.2-1) ...\n", "Processing triggers for fontconfig (2.13.1-4.2ubuntu5) ...\n", "Processing triggers for libc-bin (2.35-0ubuntu3.8) ...\n", "/sbin/ldconfig.real: /usr/local/lib/libtbb.so.12 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_5.so.3 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtcm_debug.so.1 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtbbmalloc.so.2 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libur_adapter_level_zero.so.0 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtbbmalloc_proxy.so.2 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libhwloc.so.15 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtcm.so.1 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libumf.so.0 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtbbbind.so.3 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libur_loader.so.0 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libur_adapter_opencl.so.0 is not a symbolic link\n", "\n", "/sbin/ldconfig.real: /usr/local/lib/libtbbbind_2_0.so.3 is not a symbolic link\n", "\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.9/1.9 MB\u001b[0m \u001b[31m47.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[?25hRequirement already satisfied: gymnasium[box2d] in /usr/local/lib/python3.11/dist-packages (1.0.0)\n", "Requirement already satisfied: numpy>=1.21.0 in /usr/local/lib/python3.11/dist-packages (from gymnasium[box2d]) (1.26.4)\n", "Requirement already satisfied: cloudpickle>=1.2.0 in /usr/local/lib/python3.11/dist-packages (from gymnasium[box2d]) (3.1.1)\n", "Requirement already satisfied: typing-extensions>=4.3.0 in /usr/local/lib/python3.11/dist-packages (from gymnasium[box2d]) (4.12.2)\n", "Requirement already satisfied: farama-notifications>=0.0.1 in /usr/local/lib/python3.11/dist-packages (from gymnasium[box2d]) (0.0.4)\n", "Collecting box2d-py==2.3.5 (from gymnasium[box2d])\n", " Downloading box2d-py-2.3.5.tar.gz (374 kB)\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m374.4/374.4 kB\u001b[0m \u001b[31m26.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[?25h Building wheel for box2d-py (setup.py) ... \u001b[?25l\u001b[?25hdone\n", " Created wheel for box2d-py: filename=box2d_py-2.3.5-cp311-cp311-linux_x86_64.whl size=2379446 sha256=01dcaf5b40c9697dc280991e9dcbde5c77b5cf0fa8e1ad4b91f09185b95a1540\n", " Stored in directory: /root/.cache/pip/wheels/ab/f1/0c/d56f4a2bdd12bae0a0693ec33f2f0daadb5eb9753c78fa5308\n", "Successfully built box2d-py\n", "Installing collected packages: box2d-py\n", "Successfully installed box2d-py-2.3.5\n" ] } ], "source": [ "!sudo apt-get update --quiet\n", "!pip install imageio --quiet\n", "!sudo apt-get install -y xvfb ffmpeg --quiet\n", "!pip install swig --quiet\n", "!pip install gymnasium[box2d]" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "rQsT67I7xplS", "outputId": "386d398a-4fa1-4392-c362-d4383db7a673" }, "outputs": [ { "data": { "text/plain": [ "device(type='cuda')" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import gymnasium as gym\n", "import random\n", "import matplotlib\n", "from matplotlib import pyplot as plt\n", "import numpy as np\n", "from collections import namedtuple, deque\n", "\n", "import base64\n", "import json\n", "import imageio\n", "import IPython\n", "\n", "import torch\n", "import torch.nn as nn\n", "import torch.optim as optim\n", "import torch.nn.functional as F\n", "\n", "\n", "# if GPU is to be used\n", "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", "device" ] }, { "cell_type": "markdown", "metadata": { "id": "l0NlmaWvxplS" }, "source": [ "## **Helper Functions** \n", "This section contains functions for visualizing your results. \n" ] }, { "cell_type": "code", "execution_count": 46, "metadata": { "id": "2TSVEQFFxplS" }, "outputs": [], "source": [ "# @title helper functions\n", "\n", "# disable warnings\n", "import logging\n", "logging.getLogger().setLevel(logging.ERROR)\n", "\n", "# set up matplotlib\n", "is_ipython = 'inline' in matplotlib.get_backend()\n", "if is_ipython:\n", " from IPython import display\n", "plt.ion()\n", "plt.xkcd(scale=1, length=100, randomness=2)\n", "matplotlib.rcParams['figure.figsize'] = (12, 6)\n", "\n", "\n", "def plot_rewards(sum_of_rewards,i,show_result=False):\n", " plt.figure(1)\n", " rewards = torch.tensor(sum_of_rewards, dtype=torch.float)\n", " if show_result:\n", " plt.title('Result')\n", " else:\n", " plt.clf()\n", " plt.title(f'Training the Agent number {i}')\n", " plt.xlabel('Episode')\n", " plt.ylabel('Reward')\n", " plt.plot(rewards.numpy())\n", " # Take 50 episode averages and plot them too\n", " length = len(rewards)\n", " init_len = min(49, length)\n", " init_means = np.cumsum(rewards[:init_len]) / (1 + np.arange(init_len))\n", " if length > 50:\n", " means = rewards.unfold(0, 50, 1).mean(1).view(-1)\n", " means = torch.cat((init_means, means))\n", " else:\n", " means = init_means\n", " plt.plot(means.numpy())\n", "\n", " plt.pause(0.001)\n", " if is_ipython:\n", " if not show_result:\n", " display.display(plt.gcf())\n", " display.clear_output(wait=True)\n", " else:\n", " display.display(plt.gcf())\n", "\n", "def plot_smooth(DDQN_mean_rewards,DDQN_min_rewards,DDQN_max_rewards,DQN_mean_rewards,DQN_min_rewards,DQN_max_rewards):\n", " plt.figure(figsize=(12,7))\n", "\n", " # Plot DDQN\n", " DDQN, = plt.plot(range(len(DDQN_mean_rewards)), DDQN_mean_rewards, color='blue', label='DDQN')\n", " plt.fill_between(range(len(DDQN_min_rewards)), DDQN_min_rewards, DDQN_max_rewards, color='blue', alpha=0.2)\n", "\n", " # Plot DQN\n", " DQN, = plt.plot(range(len(DQN_mean_rewards)), DQN_mean_rewards, color='red', label='DQN')\n", " plt.fill_between(range(len(DQN_min_rewards)), DQN_min_rewards, DQN_max_rewards, color='red', alpha=0.2)\n", "\n", " # Fix legend\n", " plt.legend(handles=[DDQN, DQN])\n", " plt.show()\n", "\n", "\n", "\n", "def plot_values(values):\n", " plt.figure(figsize=(15, 9))\n", "\n", " # Iterate over each value set\n", " for i, value in enumerate(values):\n", " for n, Data in enumerate(value):\n", " plt.plot(range(len(Data)), Data, label=f\"Values of selected trained Agent Number {i+1}, Evaluation {n+1}\")\n", "\n", " plt.title('Test Episode Mean Q values')\n", " plt.xlabel(\"Episodes\")\n", " plt.ylabel(\"Value\")\n", " plt.grid(True)\n", " plt.legend()\n", " plt.show()\n", "\n", "\n", "\n", "def embed_mp4(filename):\n", " video = open(filename,'rb').read()\n", " b64 = base64.b64encode(video)\n", " tag = '''\n", " '''.format(b64.decode())\n", " return IPython.display.HTML(tag)\n", "\n", "\n", "def create_policy_eval_video(env, agent, filename, num_episodes=1, fps=30):\n", " filename = filename + \".mp4\"\n", " with imageio.get_writer(filename, fps=fps) as video:\n", " for _ in range(num_episodes):\n", " state, info = env.reset()\n", " video.append_data(env.render())\n", " while True:\n", " state = torch.from_numpy(state).unsqueeze(0).to(device)\n", " action,_ = agent.act(state, greedy=True)\n", " state, reward, terminated, truncated, info = env.step(action)\n", " video.append_data(env.render())\n", " if terminated or truncated:\n", " break\n", " return embed_mp4(filename)\n", "\n", "\n", "def save_progress(sum_of_rewards, PATH):\n", " # Convert the list to a JSON string\n", " json_data = json.dumps(sum_of_rewards)\n", " # Write the JSON data to a file\n", " with open(PATH + str('.json'), \"w\") as file:\n", " file.write(json_data)\n", "\n", "\n", "def load_progress(PATH):\n", " with open(PATH + str('.json'), \"r\") as file:\n", " json_data = file.read()\n", " # Load the JSON data back into a Python list\n", " return json.loads(json_data)" ] }, { "cell_type": "markdown", "metadata": { "id": "6QAc3RPgxplS" }, "source": [ "# **Explore the Environment** \n", "\n", "Let's explore the Gym Cart-Pole environment. \n", "First, we need to create the environment and set `rgb_array` as the render mode for visualization. \n" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "kkOb315uxplS", "outputId": "589eff99-8afc-417e-a369-a23ba7d6ff8a" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Observations: 4\n", "Actions: 2\n" ] } ], "source": [ "env = gym.make(\"CartPole-v1\", render_mode=\"rgb_array\")\n", "# TODO: Print the observation space and action space\n", "print('Observations:', env.observation_space.shape[0])\n", "print('Actions:', env.action_space.n)" ] }, { "cell_type": "markdown", "metadata": { "id": "ebrWxAFrxplS" }, "source": [ "Complete the following class to create an agent that selects actions randomly from the action space. \n" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "id": "MgHtlH8fxplT" }, "outputs": [], "source": [ "class RandomAgent():\n", "\n", " def __init__(self,env_name,mode='rgb_array'):\n", " self.env =gym.make(env_name, render_mode='rgb_array')\n", "\n", "\n", " def act(self,state = None,greedy = None):\n", " # TODO: Select and return a random action\n", " action = self.env.action_space.sample()\n", " return action,0\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "dcGBHs_2xplT" }, "source": [ "**Monitor the random Agent perfomance**" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 502 }, "id": "wgKiOogxxplT", "outputId": "280f468f-f227-4f5f-bb3a-cea4247bbd37" }, "outputs": [ { "data": { "text/html": [ "\n", " " ], "text/plain": [ "" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "random_agent = RandomAgent(\"CartPole-v1\")\n", "create_policy_eval_video(random_agent.env, random_agent, 'random_policy', num_episodes=5)" ] }, { "cell_type": "markdown", "metadata": { "id": "JRFrfnYpxplT" }, "source": [ "## **Main Components of DDQN and Its Variants** \n", "\n", "### **Deep Q Network (DQN)** \n", "\n", "DQN uses a neural network to estimate $Q(s,a)$ values. In theory, the network takes both state and action as input and outputs a single $Q(s,a)$ value. However, in practice, it takes only the state as input and outputs a vector of Q-values, where each value corresponds to an action in the action space. \n", "\n", " \n", "\n", "Now, let's define the Deep Q Network. \n" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "id": "I81sC3E5xplT" }, "outputs": [], "source": [ "\n", "class QNetwork(nn.Module):\n", "\n", " def __init__(self, state_size, action_size):\n", "\n", " super(QNetwork, self).__init__()\n", "\n", " self.relu = nn.ReLU()\n", " self.fc1 = nn.Linear(state_size, 512)\n", " self.fc2 = nn.Linear(512,125)\n", " self.fc3 = nn.Linear(125,action_size)\n", "\n", "\n", " def forward(self, state):\n", "\n", " x = self.relu(self.fc1(state))\n", " x = self.relu(self.fc2(x))\n", " x = self.relu(self.fc3(x))\n", "\n", " return x\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "pJ5N9yZjxplT" }, "source": [ "## **Experience Replay Buffer** \n", "\n", "To train the network, we need data. We store **transitions**—tuples of (state, action, reward, next state, termination)—in a replay buffer for later sampling. \n", "\n", "- **state (s):** The current situation the agent is experiencing. \n", "- **action (a):** The action taken by the agent. \n", "- **next state (s'):** The new state after taking the action. \n", "- **reward (r):** The feedback received for the action. \n", "- **termination (done):** A boolean indicating if the episode has ended. \n", "\n", "These transitions are stored in the **Experience Replay Buffer**, allowing us to sample and train the Q-network efficiently. \n", "\n", "A stack of transitions $(s, a, s', r, done)$ looks like this: \n", "\n", " \n", "\n", "**We shuffle the sampled data to break temporal correlations before training.** \n", "\n", "Now, let's define the Experience Replay Buffer class. \n" ] }, { "cell_type": "markdown", "metadata": { "id": "oqMrze5oxplT" }, "source": [ "## **Experience Replay Buffer** \n", "\n", "- `push`: Stores transitions from the environment. \n", "- `sample`: Shuffles and samples transitions. \n", "- `__len__`: Returns the number of stored transitions. \n", "- `get_size`: Returns the buffer size (same as `__len__`). \n" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "id": "3xn-7fyXxplT" }, "outputs": [], "source": [ "class ReplayMemory(object):\n", "\n", " def __init__(self, capacity, batch_size):\n", " self.memory = deque([],maxlen = capacity)\n", " self.batch_size = batch_size\n", " self.experience = namedtuple(\"Experience\", field_names = (\"state\", \"action\", \"reward\", \"next_state\", \"done\"))\n", "\n", " def push(self, state, action, reward, next_state, done):\n", " e = self.experience(state, action, reward, next_state, done)\n", " self.memory.append(e)\n", "\n", " def sample(self, batch_size = None):\n", "\n", " if batch_size is None:\n", " batch_size = self.batch_size\n", " experiences = random.sample(self.memory, k = batch_size)\n", " states = np.vstack([e.state.detach().cpu() for e in experiences if e is not None])\n", " actions = np.vstack([e.action for e in experiences if e is not None])\n", " rewards = np.vstack([e.reward for e in experiences if e is not None])\n", " next_states = np.vstack([e.next_state.detach().cpu() for e in experiences if e is not None])\n", " dones = np.vstack([e.done for e in experiences if e is not None]).astype(np.uint8)\n", " return states, actions, rewards, next_states, dones\n", "\n", " def __len__(self):\n", " return len(self.memory)\n", "\n", " def get_size(self):\n", " return self.__len__()" ] }, { "cell_type": "markdown", "metadata": { "id": "b0ctqz-zxplT" }, "source": [ "## **DQN Agent** \n", "\n", "DQN, the first Deep RL algorithm, uses TD learning similar to Q-learning, aiming to minimize the distance: \n", "\n", "$r_i + \\gamma \\cdot max_{a'} Q_\\theta'(s_i',a_i') - Q_\\theta(s_i,a_i)$ \n", "\n", "The more appropriate cost function is: \n", "\n", "$[r_i + \\gamma \\cdot max_{a'} Q_\\theta(s_i',a_i') - Q_\\theta(s_i,a_i)]^2$ $(1)$ \n", "\n", "Instead of a tabular method, DQNs use deep networks to estimate Q-values. \n", "\n", "In this class, we implement the original DQN with an experience replay buffer. The training process is as follows: \n", "\n", "**Network Updating** \n", "\n", "1. Gather data. Once the buffer reaches a certain size, start training the Q-networks. \n", "2. Sample transitions and feed the States (instead of State-action pairs) to the Q-network to estimate $Q_\\theta(s_i,a_i)$. \n", "3. Feed Next_States to the Q-network to estimate $Q_\\theta(s_i',a_i')$. \n", "4. Use the average of equation $(1)$ to update the network via backpropagation. \n" ] }, { "cell_type": "markdown", "metadata": { "id": "BVK_MrnFxplT" }, "source": [ "## **$\\epsilon-\\text{greedy policy}$** \n", "\n", "Exploration is crucial in RL algorithms. In DQN, exploration is achieved using the $\\epsilon-\\text{greedy policy}$.\n", "\n", "$$\n", "\\pi(a \\mid s) =\n", "\\begin{cases}\n", "\\arg\\max\\limits_{a} Q(s, a), & \\text{with probability } 1 - \\epsilon, \\\\\n", "\\text{random action}, & \\text{with probability } \\epsilon.\n", "\\end{cases}\n", "$$\n", "\n", "\n", "With probability $\\epsilon$, a random action is chosen for exploration, and with probability $1 - \\epsilon$, the best action is selected for exploitation. \n", "\n", "$\\epsilon$ starts high and gradually decreases over time. The decay follows this formula:\n", "\n", "$$\\varepsilon = \\varepsilon_{end} + \\left(\\varepsilon_{start} - \\varepsilon_{end}\\right)\\exp\\left(-\\frac{\\text{steps done}}{\\text{decay rate}}\\right)$$ \n", "\n", "**The $\\epsilon-\\text{greedy policy}$ is used in both DQN and DDQN.** \n", "\n", "## **Original DQN Pseudocode** \n", "\n", " \n", "\n", "### Properties of the Algorithm:\n", "\n", "- Uses a single network to estimate both $Q(s,a)$ and $Q(s',a')$, which can lead to instability—this is the problem DDQN aims to solve. **It's up to you to figure out why this is a problem.** \n", "- Utilizes the $\\epsilon-\\text{greedy}$ policy for exploration.\n", "\n", "**Hint:** \n", "- A single network leads to unstable learning, hence the introduction of a target network in modified versions. \n", "- The $\\epsilon-\\text{greedy}$ policy can sometimes hinder the learning process, despite addressing the exploration problem. \n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "2VJnYgyMxplT" }, "outputs": [], "source": [ "class DQNAgent(object):\n", "\n", " def __init__(self, q_network, memory, optimizer, criterion, params):\n", " # TODO: Create policy net based on the q-net\n", " self.policy_net = q_network.to(device)\n", "\n", " # TODO: Setup the agent's memory\n", " self.reply_buffer = memory(params['BUFFER_SIZE'],params['BATCH_SIZE'])\n", " # criterion, optimizer and params\n", " self.criterion = criterion()\n", " self.optimizer = optimizer(self.policy_net.parameters(), lr=params['LR'],amsgrad=True)\n", "\n", " self.gamma = params['GAMMA']\n", " self.eps = {'START': params['EPS_START'], 'END': params['EPS_END'], 'DECAY': params['EPS_DECAY']}\n", " self.steps_done = 0\n", " self.Loss = []\n", "\n", "\n", " def step(self, state, action, reward, next_state, done):\n", " # TODO: Save the experience in the memory of the agent\n", "\n", " self.reply_buffer.push(state, action, reward, next_state, done)\n", " # TODO: Increment the steps counter\n", " self.steps_done +=1\n", "\n", "\n", " if self.reply_buffer.get_size() > self.reply_buffer.batch_size:\n", " # TODO: Sample a batch from memory and learn from it\n", " states, actions, rewards, next_states,dones= self.reply_buffer.sample()\n", " self.learn(states, actions, rewards, next_states,dones)\n", "\n", "\n", " def act(self, state, greedy=False):\n", " self.eps_threshold = self.eps['END'] + (self.eps['START'] - self.eps['END']) * np.exp(- self.steps_done / self.eps['DECAY'])\n", "\n", " if greedy or random.random() > self.eps_threshold:\n", " with torch.no_grad(): # TODO: Select greedy action\n", " Q_values = self.policy_net(state)\n", " action = torch.argmax(Q_values).item()\n", " max_Q = torch.max(Q_values).item()\n", " else: # TODO: Select random action\n", " action = random.choices(range(2))[0]\n", " max_Q = 0\n", " return action,max_Q\n", "\n", " def learn(self, states, actions, rewards, next_states, dones):\n", "\n", "\n", "\n", " states = torch.from_numpy(states).float().to(device)\n", " actions = torch.from_numpy(actions).long().to(device)\n", " rewards = torch.from_numpy(rewards).float().to(device)\n", " next_states = torch.from_numpy(next_states).float().to(device)\n", " dones = torch.from_numpy(dones).float().to(device)\n", " # TODO: Compute the predicted Q-values using the policy network\n", " state_Q = self.policy_net(states)\n", " state_Q = state_Q.gather(dim = 1, index = actions)\n", "\n", "\n", "\n", " with torch.no_grad():\n", " next_state_Q = self.policy_net(next_states).detach().max(1)[0].unsqueeze(1)\n", " belman_backup = rewards + self.gamma*next_state_Q*(1-dones)\n", "\n", " # TODO: Compute the loss and do backpropagation\n", " loss = self.criterion(state_Q,belman_backup)\n", " self.optimizer.zero_grad()\n", " loss.backward()\n", " self.optimizer.step()\n", "\n", "\n", "\n", "\n", " def save(self, PATH):\n", " torch.save(self.policy_net, PATH + '_policy.pt')\n", "\n", "\n", " def load(self, PATH):\n", " self.policy_net = torch.load(PATH + '_policy.pt')\n" ] }, { "cell_type": "markdown", "metadata": { "id": "Pep4oRAjxplU" }, "source": [ "# **Training the DQN Agents** \n", "\n", "Now that you've implemented DQN and explored the Gym CartPole environment, it's time to train an actual agent. \n", "\n", "**First, we train the DQN agent.** \n" ] }, { "cell_type": "markdown", "metadata": { "id": "YbYIPpZaxplU" }, "source": [ "## **Setting Up Essentials for DQN** \n", "\n", "- **CartPole Environment Setup** \n", "- **Hyperparameter Initialization** \n", "- **Creating DQN Agents with Different Random Seeds** \n", " - Why? Training multiple agents with different seeds helps assess DQN's consistency and robustness. \n", " - We use **5 random seeds** for better evaluation. \n", "- **Defining Optimizer and Loss Function** \n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "bNCRCzX0xplU" }, "outputs": [], "source": [ "def create_dqn_agent(seed, QNetwork, ReplayMemory, optimizer, criterion, params):\n", " # TODO: set the different seeds\n", " torch.manual_seed(seed)\n", " random.seed(seed)\n", " np.random.seed(seed)\n", "\n", " # Instantiate the agent\n", " return DQNAgent(QNetwork(n_observations, n_actions), ReplayMemory, optimizer, criterion, params)\n", "\n", "\n", "\n", "\n", "params = {\n", "\n", " 'BUFFER_SIZE': int(100000), # size of the replay buffer\n", " 'BATCH_SIZE':128, # number of experiences sampled from memory\n", " 'GAMMA': 0.99 , # discount factor\n", " 'EPS_START':0.99 , # starting value of epsilon\n", " 'EPS_END': 0.01 , # final value of epsilon\n", " 'EPS_DECAY': 10000 , # rate of exponential decay of epsilon \n", " 'LR': 0.0001 # learning rate of the optimizer\n", "}\n", "\n", "# Srrting the seeds\n", "seeds = [1, 10, 15, 43, 63]\n", "\n", "\n", "env = gym.make(\"CartPole-v1\")\n", "state, _ = env.reset()\n", "# Get number of actions from gym action space\n", "n_actions = env.action_space.n\n", "# Get the number of state observations\n", "n_observations = env.observation_space.shape[0]\n", "# TODO: Choose optimizer and loss function\n", "optimizer = optim.Adam\n", "criterion = nn.MSELoss\n", "\n", "\n", "# Create multiple agents with different seeds\n", "DQN_agents = []\n", "dqn_sum_of_rewards = []\n", "\n", "for seed in seeds:\n", " # TODO: create the Agent instance\n", " dqn_Agent = create_dqn_agent(seed, QNetwork, ReplayMemory, optimizer, criterion, params)\n", " # TODO: Append the Agent to the agents list\n", " DQN_agents.append(dqn_Agent)\n", " dqn_sum_of_rewards.append([])" ] }, { "cell_type": "markdown", "metadata": { "id": "QXFBHgbhxplU" }, "source": [ "## **Training Loop for the DQN Agent** \n", "\n", "With the hyperparameters set, we can now train the agent. \n", "\n", "The code is designed for convenient retraining. simply re-run the segment to continue training if the agent hasn’t achieved satisfactory results. \n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 34 }, "id": "3_1rzGEXxplU", "outputId": "c5aacfa6-c1e6-4f6d-9096-01376eb16ea4" }, "outputs": [ { "data": { "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# TODO: Set the number of training episodes\n", "num_episodes = 400\n", "\n", "\n", "for i,dqn_Agent in enumerate(DQN_agents):\n", "\n", " for e in range(1, num_episodes + 1):\n", " # TODO: reset the environment\n", " state,_ = env.reset()\n", " state = torch.tensor(state, dtype=torch.float32, device=device).unsqueeze(0)\n", " episode_reward, done = 0, False\n", " # TODO: in the loop take actions,go to next step and get the rewards\n", " while True:\n", "\n", " action,_ = dqn_Agent.act(state)\n", " next_state, reward, done,truncated ,_,= env.step(action)\n", " next_state = torch.tensor(next_state, dtype=torch.float32, device=device).unsqueeze(0)\n", " # TODO: Save the Data to the Agent memory and train the Agent\n", " dqn_Agent.step(state, action, reward, next_state, done)\n", "\n", " episode_reward += reward\n", " # TODO: break the loop if the episode is terminated or truncated\n", " if done or truncated:\n", " break\n", " state = next_state\n", "\n", " dqn_sum_of_rewards[i].append(episode_reward)\n", " # Save model every 50 episodes and plot the returns (change the rate if needed) , you can turn the plot off.\n", " if e % 50 == 0:\n", "\n", " plot_rewards(dqn_sum_of_rewards[i],i)\n", " path = f'DQN_{i}_Network' + (str(len(dqn_sum_of_rewards[i]))).zfill(4)\n", " dqn_Agent.save(path)\n", " save_progress(dqn_sum_of_rewards[i], path)\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "Hq7ueyTGxplU" }, "source": [ "## **Why DDQN?** \n", "\n", "DQN and its variants use a replay buffer and Q-network, but DQN struggles with learning stability. **Can you identify why?** \n", "To address this, we introduce DDQN—a modified version with improved stability, capable of handling more complex environments. \n", "\n", "---\n", "\n", "## **DDQN Agent** \n", "\n", "### **Deep Q-Networks in DDQN** \n", "\n", "DDQN uses two networks: \n", "\n", "- **Online Network**: Estimates $Q_\\theta(s_i,a_i)$, representing the value of a state-action pair. its parameters is denoted with $\\theta$\n", "- **Target Network**: Estimates $Q_{\\theta'}(s_i',a_i')$, representing the value of the next state-action pair. its parameters is denoted with $\\theta'$\n", "\n", "Both networks have identical architectures but serve different roles. The target network helps stabilize training by reducing overestimation bias. \n", "\n", "### **Network Updating** \n", "\n", "In DQN, a single network estimates both $Q(s_i,a_i)$ and $Q(s_i',a_i')$, leading to instability. DDQN introduces a key change: \n", "\n", "- The **online network** selects the best action: \n", " $$ \\arg\\max_a Q_\\theta(s'_i,a) $$\n", "- The **target network** estimates its value: \n", " $$ Q_{\\theta'}(s_i, \\arg\\max_a Q_\\theta(s'_i,a)) $$ \n", "\n", "This modifies the loss function to: \n", "\n", "$$ [r_i + \\gamma \\cdot Q_{\\theta'}(s_i, \\arg\\max_a Q_\\theta(s'_i,a)) - Q_\\theta(s_i,a_i)]^2 $$ \n", "\n", "In this version DDQN, we also use $\\text{soft replacemet}$ which is not a part of original DDQN, that another modification that is applied.\n", "* `soft_update` : The target network is updated at every step with a soft update controlled by the hyperparameter $\\tau$, which was previously defined. The target is updated according to: $$\\theta' \\leftarrow \\tau \\theta + (1 - \\tau) \\theta'$$\n", "\n", "## DDQN Psudocode\n", "\n", " \n", "\n", "\n", "Now, let's define the **DDQN class**. \n" ] }, { "cell_type": "markdown", "metadata": { "id": "MFqyOXMIxplU" }, "source": [ "## **DDQN Agent Class**" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "vtv2Nf4HxplU" }, "outputs": [], "source": [ "class DDQNAgent(object):\n", "\n", " def __init__(self, q_network, memory, optimizer, criterion, params):\n", " # TODO: Create policy and target nets based on the q-net\n", " self.target_net = q_network.to(device)\n", " self.policy_net = q_network.to(device)\n", " self.Loss = []\n", " # TODO: Setup the agent's memory\n", " self.reply_buffer = memory(params['BUFFER_SIZE'],params['BATCH_SIZE'])\n", " # criterion, optimizer and params\n", " self.criterion = criterion()\n", " self.optimizer = optimizer(self.policy_net.parameters(), lr=params['LR'],amsgrad=True)\n", " self.tau = params['TAU']\n", " self.gamma = params['GAMMA']\n", " self.update_rate = params['UPDATE_RATE']\n", " self.eps = {'START': params['EPS_START'], 'END': params['EPS_END'], 'DECAY': params['EPS_DECAY']}\n", " self.steps_done = 0\n", "\n", "\n", " # TODO: Set the tagert network parameters equal to policy network\n", " self.soft_update(tau = 1)\n", "\n", " def step(self, state, action, reward, next_state, done):\n", " # TODO: Increment the steps_done 1 step each time\n", " self.steps_done += 1\n", " # TODO: Save the experience in the memory of the agent\n", " self.reply_buffer.push(state, action, reward, next_state, done)\n", " if self.reply_buffer.get_size() > self.reply_buffer.batch_size:\n", " # TODO: Sample a batch from memory and learn from it\n", " states, actions, rewards, next_states,dones= self.reply_buffer.sample()\n", " self.learn(states, actions, rewards, next_states,dones)\n", "\n", "\n", " def act(self, state, greedy=False,eps_threshold = None):\n", " self.eps_threshold = self.eps['END'] + (self.eps['START'] - self.eps['END']) * np.exp(- self.steps_done / self.eps['DECAY'])\n", "\n", " if greedy or random.random() > self.eps_threshold:\n", " with torch.no_grad(): # TODO: Select greedy action\n", " Q_values = self.target_net.forward(state)\n", " action = torch.argmax(Q_values).item()\n", " max_Q = torch.max(Q_values).item()\n", " else: # TODO: Select random action\n", " action = random.choices(range(2))[0]\n", " max_Q = 0\n", " return action,max_Q\n", "\n", " def learn(self, states, actions, rewards, next_states, dones):\n", "\n", "\n", "\n", " states = torch.from_numpy(states).float().to(device)\n", " actions = torch.from_numpy(actions).long().to(device)\n", " rewards = torch.from_numpy(rewards).float().to(device)\n", " next_states = torch.from_numpy(next_states).float().to(device)\n", " dones = torch.from_numpy(dones).float().to(device)\n", "\n", " # TODO: Compute the predicted Q-values using the policy network\n", " state_Q = self.policy_net(states)\n", " state_Q = state_Q.gather(dim = 1, index = actions).reshape(params['BATCH_SIZE'],1)\n", "\n", "\n", "\n", " with torch.no_grad():\n", " # TODO: Select the best next action using the policy network (online network)\n", " best_next_action = torch.argmax(self.policy_net(next_states),dim = 1).detach().unsqueeze(1)\n", " # TODO: Get the Q-values of the (next state,best next action) using the target network and selected best next action\n", " next_state_Q = self.target_net(next_states).detach().gather(1, best_next_action).detach()\n", " # TODO: Compute the Bellman backup target: r + γ * Q(next state,best next action) * (1 - done)\n", " belman_backup = rewards + self.gamma*next_state_Q*(1-dones)\n", " # TODO: Compute the loss and do backpropagation\n", " loss = self.criterion(state_Q,belman_backup)\n", " self.optimizer.zero_grad()\n", " loss.backward()\n", " self.optimizer.step()\n", " # TODO: do soft replacement if steps_done % update rate = 0\n", " if self.steps_done % self.update_rate == 0:\n", " self.soft_update(tau = self.tau)\n", "\n", "\n", " def soft_update(self,tau):\n", " # TODO: Soft update of all weights in the target network\n", "\n", " for target_param, pi_param in zip(self.target_net.parameters(), self.policy_net.parameters()):\n", " target_param.data.copy_(tau * pi_param.data + (1 - tau) * target_param.data)\n", "\n", "\n", " def save(self, PATH):\n", " torch.save(self.policy_net, PATH + '_policy.pt')\n", " torch.save(self.target_net, PATH + '_target.pt')\n", "\n", " def load(self, PATH):\n", " self.policy_net = torch.load(PATH + '_policy.pt')\n", " self.target_net = torch.load(PATH + '_target.pt')" ] }, { "cell_type": "markdown", "metadata": { "id": "Cd7tRn5vEtxr" }, "source": [ "## **Setting Up Essentials for DDQN** \n", "\n", "- **CartPole Environment Setup** \n", "- **Hyperparameter Initialization** \n", "- **Creating DDQN Agents with Different Random Seeds** \n", " - Why? Training multiple agents with different seeds helps assess DDQN's consistency and robustness. \n", " - We use **5 random seeds** for better evaluation. \n", "- **Defining Optimizer and Loss Function** \n" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "id": "LdkgUQl2RlBv" }, "outputs": [], "source": [ "def create_ddqn_agent(seed, QNetwork, ReplayMemory, optimizer, criterion, params):\n", " # TODO: set the different seeds\n", " torch.manual_seed(seed)\n", " random.seed(seed)\n", " np.random.seed(seed)\n", " # TODO: Instantiate the agent\n", " return DDQNAgent(QNetwork(n_observations, n_actions), ReplayMemory, optimizer, criterion, params)\n", "\n", "params = {\n", " 'UPDATE_RATE': 10 , # how often to update the network\n", " 'BUFFER_SIZE': int(100000), # size of the replay buffer\n", " 'BATCH_SIZE':128, # number of experiences sampled from memory\n", " 'GAMMA': 0.99 , # discount factor\n", " 'EPS_START':0.99 , # starting value of epsilon\n", " 'EPS_END': 0.01 , # final value of epsilon\n", " 'EPS_DECAY': 10000 , # rate of exponential decay of epsilon 35000\n", " 'TAU':0.009, # update rate of the target network 0.008\n", " 'LR': 0.0001 # learning rate of the optimizer\n", "}\n", "\n", "# setting different seed values\n", "seeds = [1, 10, 15, 43, 63]\n", "\n", "# TODO: Setup the Environment\n", "env = gym.make(\"CartPole-v1\")\n", "state, _ = env.reset()\n", "# Get number of actions from gym action space\n", "n_actions = env.action_space.n\n", "# Get the number of state observations\n", "n_observations = env.observation_space.shape[0]\n", "# TODO: Choose optimizer and loss function\n", "optimizer = optim.Adam\n", "criterion = nn.MSELoss\n", "\n", "# TODO: Create multiple agents with different seeds\n", "agents = []\n", "sum_of_rewards = []\n", "\n", "for seed in seeds:\n", " # TODO: create the Agent instance\n", " Agent = create_ddqn_agent(seed, QNetwork, ReplayMemory, optimizer, criterion, params)\n", " # TODO: Append the Agent to the agents list\n", " agents.append(Agent)\n", " sum_of_rewards.append([])" ] }, { "cell_type": "markdown", "metadata": { "id": "FKgBONebcSHp" }, "source": [ "## **Training Loop for DDQN Agent** \n", "\n", "With the hyperparameters set, we can now train the agent. The following code allows for easy re-training—if the agent's performance isn't satisfactory, simply re-run the segment to continue training for more episodes. \n" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 34 }, "id": "hbyia1ACd6Ju", "outputId": "75a11a2e-9d31-4f7f-dd19-204c6f5b3483" }, "outputs": [ { "data": { "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# TODO: Set the number of training episodes\n", "num_episodes = 400\n", "\n", "\n", "for i,DDQN_agent in enumerate(agents):\n", "\n", " for e in range(1, num_episodes + 1):\n", " # TODO: reset the environment\n", " state,_ = env.reset()\n", " state = torch.tensor(state, dtype=torch.float32, device=device).unsqueeze(0)\n", " episode_reward, done = 0, False\n", " # TODO: in the loop take actions,go to next step and get the rewards\n", " while True:\n", "\n", " action,_ = DDQN_agent.act(state)\n", " next_state, reward, done,truncated ,_,= env.step(action)\n", " next_state = torch.tensor(next_state, dtype=torch.float32, device=device).unsqueeze(0)\n", " # TODO: Save the Data to the Agent memory and train the Agent\n", " DDQN_agent.step(state, action, reward, next_state, done)\n", " episode_reward += reward\n", " # TODO: break the loop if the episode is terminated or truncated\n", " if done or truncated:\n", " break\n", " state = next_state\n", "\n", " sum_of_rewards[i].append(episode_reward)\n", " # Save model every 50 episodes and plot the returns (change the rate if needed) , you can turn the plot off.\n", " if e % 50 == 0:\n", "\n", " plot_rewards(sum_of_rewards[i],i)\n", " path = f'DDQN_{i}_Network' + (str(len(sum_of_rewards[i]))).zfill(4)\n", " DDQN_agent.save(path)\n", " save_progress(sum_of_rewards[i], path)\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "8-pmbcW1n-l9" }, "source": [ "## **Computing the Moving Average of Results (5 points)** \n", "\n", "We compute a moving average to smooth the results for both DDQN and DQN across all seeds. This includes the average of the minimum, maximum, and actual returns across all seeds and episodes. \n" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "id": "bzNZAA2o3Xyu" }, "outputs": [], "source": [ "def moving_average(data, window_size=25):\n", " return np.convolve(data, np.ones(window_size)/window_size, mode='valid')\n", "\n", "# Compute moving averages for DDQN\n", "DDQN_mean_rewards = moving_average(np.array(sum_of_rewards).mean(axis=0))\n", "DDQN_min_rewards = moving_average(np.array(sum_of_rewards).min(axis=0))\n", "DDQN_max_rewards = moving_average(np.array(sum_of_rewards).max(axis=0))\n", "\n", "# Compute moving averages for DQN\n", "DQN_mean_rewards = moving_average(np.array(dqn_sum_of_rewards).mean(axis=0))\n", "DQN_min_rewards = moving_average(np.array(dqn_sum_of_rewards).min(axis=0))\n", "DQN_max_rewards = moving_average(np.array(dqn_sum_of_rewards).max(axis=0))" ] }, { "cell_type": "markdown", "metadata": { "id": "CDY4CVrjochx" }, "source": [ "## **Visualizing the Outputs (5 points)** \n", "\n", "Plotting the smoothed results with: \n", "- Lower bound as the average of minimum returns \n", "- Upper bound as the average of maximum returns \n", "- Actual moving average of the returns \n", "\n", "for smoother plots, you can increase the **window_size** in `moving_average` function.\n" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 610 }, "id": "2P-JEtDL33xM", "outputId": "7d5101ac-5d09-4cc0-85dd-9571fdc3196d" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_smooth(DDQN_mean_rewards,DDQN_min_rewards,DDQN_max_rewards,DQN_mean_rewards,DQN_min_rewards,DQN_max_rewards)" ] }, { "cell_type": "markdown", "metadata": { "id": "t3EWv6crnryw" }, "source": [ "## **Takeaway Questions**\n", "- Which algorithm, DQN or DDQN, exhibits more stable learning?\n", " - Consider mean returns and the tightness of the upper and lower bounds in the training plot.\n", "- Which algorithm struggles less during learning?" ] }, { "cell_type": "markdown", "metadata": { "id": "PdK1XgRiIXBP" }, "source": [ "## **Model Evaluation** \n", "\n", "To evaluate the model, we measure the **average reward** and its **standard deviation** using the following function. \n" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "id": "h-ToISzSmBxl" }, "outputs": [], "source": [ "def evaluate_policy(env, agent, num_episodes=3):\n", " # TODO: Initialize sum of rewards\n", " total_reward = np.zeros((num_episodes,1))\n", " Episode_values = []\n", "\n", " for episode in range(num_episodes):\n", "\n", " # TODO: Initialize environment\n", " state,_ = env.reset()\n", " state = torch.tensor(state, dtype=torch.float32, device=device).unsqueeze(0)\n", " done,truncated = False,False\n", " values =[]\n", " mean_value =[]\n", " # TODO: In the loop use greedy policy to evaluate your Agent\n", " while True:\n", "\n", " action,Q = agent.act(state,True)\n", " next_state,reward,done,truncated,_ = env.step(action)\n", " total_reward[episode]+=reward\n", " # appending Q values for later use\n", " ###################################################\n", " values.append(Q)\n", " mean_value.append(np.array(values).mean())\n", " ###################################################\n", " if done or truncated:\n", " break\n", " next_state = torch.tensor(next_state, dtype=torch.float32, device=device)\n", " state = next_state\n", "\n", " Episode_values.append(mean_value)\n", "\n", " # TODO: Return mean and std of rewards\n", " mean = sum(total_reward)/num_episodes\n", " std = sum((total_reward - mean)**2)/num_episodes\n", "\n", "\n", " return mean,std,Episode_values\n" ] }, { "cell_type": "markdown", "metadata": { "id": "pOzO_aW_Yimj" }, "source": [ "**Run both trained policies for 3 episodes, observe the mean and std of returns, and plot the mean Q-values per episode. If the agent passes the evaluation bar, plot their mean values.**\n", "\n", "## **Plot DDQN values**" ] }, { "cell_type": "code", "execution_count": 47, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 989 }, "id": "1AgNnxeBl_V1", "outputId": "26de7fef-3db2-481d-d52c-7071c0e1a826" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "evaluating the 0th agent mean_reward = [316.33333333] +/- [366.88888889]\n", "\n", "evaluating the 1th agent mean_reward = [9.33333333] +/- [0.22222222]\n", "\n", "evaluating the 2th agent mean_reward = [477.33333333] +/- [1027.55555556]\n", "\n", "evaluating the 3th agent mean_reward = [113.] +/- [60.66666667]\n", "\n", "evaluating the 4th agent mean_reward = [370.33333333] +/- [11534.88888889]\n", "\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "DDQN_values = []\n", "# TODO: Run the each Trained Policy for 3 episodes a observe the mean and std of recived Returns and mean Q values over each episode\n", "for i,DDQN_agent in enumerate(agents):\n", " # TODO Evaluate the trained Agents\n", " mean_reward, std_reward,mean_values = evaluate_policy(env, DDQN_agent,num_episodes = 3)\n", " print(f\"evaluating the {i}th agent mean_reward = {mean_reward} +/- {std_reward}\\n\")\n", " if mean_reward >= 450:\n", " DDQN_values.append(mean_values)\n", "\n", "if len(DDQN_values) != 0:\n", " # TODO: Plot the mean values\n", " plot_values(DDQN_values)\n", "else:\n", " print('[Info] ... the Agent Did not pass the minimum requirement Please train it more.')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## **Plot DQN values**" ] }, { "cell_type": "code", "execution_count": 48, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "ztVRLsjgJJYK", "outputId": "4c220a1c-65b4-4acb-9d08-940d2805ccfb" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "evaluating the 0th agent mean_reward = [153.33333333] +/- [16.88888889]\n", "\n", "evaluating the 1th agent mean_reward = [9.] +/- [0.66666667]\n", "\n", "evaluating the 2th agent mean_reward = [286.33333333] +/- [7574.22222222]\n", "\n", "evaluating the 3th agent mean_reward = [374.] +/- [1418.66666667]\n", "\n", "evaluating the 4th agent mean_reward = [384.66666667] +/- [3146.88888889]\n", "\n", "[Info] ... the Agent Did not pass the minimum requirement Please train it more.\n" ] } ], "source": [ "DQN_values = []\n", "# TODO: Run the each Trained Policy for 3 episodes a observe the mean and std of recived Returns and mean Q values over each episode\n", "for i,DQN_agent in enumerate(DQN_agents):\n", " # TODO Evaluate the trained Agents\n", " mean_reward, std_reward,mean_values = evaluate_policy(env, DQN_agent,num_episodes = 3)\n", " print(f\"evaluating the {i}th agent mean_reward = {mean_reward} +/- {std_reward}\\n\")\n", " if mean_reward >= 450:\n", " DQN_values.append(mean_values)\n", "\n", "\n", "if len(DQN_values) != 0:\n", " # TODO: Plot the mean values\n", " plot_values(DQN_values)\n", "else:\n", " print('[Info] ... the Agent Did not pass the minimum requirement Please train it more.')" ] }, { "cell_type": "markdown", "metadata": { "id": "pyohSdzqEC3Z" }, "source": [ "## **Watch the best Agent's performance**\n", "**Select one of the best agents from the evaluation step and render its performance.**\n", "\n", "**Rendering DDQN Agent**" ] }, { "cell_type": "code", "execution_count": 41, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 519 }, "id": "QGQ3aE8fDLjb", "outputId": "310a9ac1-94ad-4350-dc56-4dc54528c3f7" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[info] ... rendering the DDQN Agent performance\n" ] }, { "data": { "text/html": [ "\n", " " ], "text/plain": [ "" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "env = gym.make(\"CartPole-v1\", render_mode='rgb_array')\n", "print('[info] ... rendering the DDQN Agent performance')\n", "# TODO: from the previous evaluation Select the best Agent to render the performance\n", "create_policy_eval_video(env, agents[2], 'greedy_policy', 3)" ] }, { "cell_type": "markdown", "metadata": { "id": "ZSPQW_n9Nix_" }, "source": [ "**Rendering DQN Agent**" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "RJdb9KpadSRp" }, "outputs": [], "source": [ "env = gym.make(\"CartPole-v1\", render_mode='rgb_array')\n", "print('[info] ... rendering the DQN Agent performance')\n", "# TODO: from the previous evaluation Select the best Agent to render the performance\n", "create_policy_eval_video(env, DQN_agents[...], 'greedy_policy', 3)" ] }, { "cell_type": "markdown", "metadata": { "id": "dDKTwX3explX" }, "source": [ "## **Takeaway Questions**\n", "- Which agent is trained better considering the evaluation and rendering the performance?\n", "- One of the DDQN's goal is to prevent Q values over estimation. Did you observe this phenomenon after ploting the Mean Q values?" ] } ], "metadata": { "accelerator": "GPU", "colab": { "gpuType": "T4", "provenance": [] }, "kaggle": { "accelerator": "nvidiaTeslaT4", "dataSources": [], "dockerImageVersionId": 30665, "isGpuEnabled": true, "isInternetEnabled": true, "language": "python", "sourceType": "notebook" }, "kernelspec": { "display_name": "Python 3", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.1" } }, "nbformat": 4, "nbformat_minor": 0 }