{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "xRQE_P5OYSTr"
      },
      "source": [
        " *by Gerard Caravaca Ibáñez*"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "2d-NLtNzYSqb"
      },
      "source": [
        "## **TD3 implementation**\n",
        "\n",
        "This notebook is an implementation of the TD3 algorithm for reinforcement learning, proposed in [1]. \n",
        "\n",
        "**TD3 (Twin Delayed Deep Deterministic Policy Gradient)** is a state-of-the-art reinforcement learning algorithm that is designed to learn continuous control policies in environments with high-dimensional state and action spaces. TD3 is an extension of the original DDPG algorithm, which was limited by its susceptibility to overestimation of the Q-function and sensitivity to hyperparameters.\n",
        "\n",
        "TD3 improves upon DDPG by introducing several key modifications, including the use of twin Q-networks to reduce overestimation bias, delayed policy updates to improve stability, and target policy smoothing to reduce variance.\n",
        "\n",
        "\n",
        "*[1] Fujimoto, S., Hoof, H., & Meger, D. (2018, July). Addressing function approximation error in actor-critic methods. In International conference on machine learning (pp. 1587-1596). PMLR.*"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "tE636EOu985l"
      },
      "source": [
        "# **Imports**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "70sYGfWnZIeY",
        "outputId": "0241a45a-c55d-4a68-a06a-10a3e90e1f9d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting box2d-py\n",
            "  Downloading box2d-py-2.3.8.tar.gz (374 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m374.5/374.5 kB\u001b[0m \u001b[31m9.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Building wheels for collected packages: box2d-py\n",
            "  \u001b[1;31merror\u001b[0m: \u001b[1msubprocess-exited-with-error\u001b[0m\n",
            "  \n",
            "  \u001b[31m×\u001b[0m \u001b[32mpython setup.py bdist_wheel\u001b[0m did not run successfully.\n",
            "  \u001b[31m│\u001b[0m exit code: \u001b[1;36m1\u001b[0m\n",
            "  \u001b[31m╰─>\u001b[0m See above for output.\n",
            "  \n",
            "  \u001b[1;35mnote\u001b[0m: This error originates from a subprocess, and is likely not a problem with pip.\n",
            "  Building wheel for box2d-py (setup.py) ... \u001b[?25lerror\n",
            "\u001b[31m  ERROR: Failed building wheel for box2d-py\u001b[0m\u001b[31m\n",
            "\u001b[0m\u001b[?25h  Running setup.py clean for box2d-py\n",
            "Failed to build box2d-py\n",
            "Installing collected packages: box2d-py\n",
            "  \u001b[1;31merror\u001b[0m: \u001b[1msubprocess-exited-with-error\u001b[0m\n",
            "  \n",
            "  \u001b[31m×\u001b[0m \u001b[32mRunning setup.py install for box2d-py\u001b[0m did not run successfully.\n",
            "  \u001b[31m│\u001b[0m exit code: \u001b[1;36m1\u001b[0m\n",
            "  \u001b[31m╰─>\u001b[0m See above for output.\n",
            "  \n",
            "  \u001b[1;35mnote\u001b[0m: This error originates from a subprocess, and is likely not a problem with pip.\n",
            "  Running setup.py install for box2d-py ... \u001b[?25l\u001b[?25herror\n",
            "\u001b[1;31merror\u001b[0m: \u001b[1mlegacy-install-failure\u001b[0m\n",
            "\n",
            "\u001b[31m×\u001b[0m Encountered error while trying to install package.\n",
            "\u001b[31m╰─>\u001b[0m box2d-py\n",
            "\n",
            "\u001b[1;35mnote\u001b[0m: This is an issue with the package mentioned above, not pip.\n",
            "\u001b[1;36mhint\u001b[0m: See above for output from the failure.\n",
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: gym[box2d] in /usr/local/lib/python3.9/dist-packages (0.25.2)\n",
            "Requirement already satisfied: gym-notices>=0.0.4 in /usr/local/lib/python3.9/dist-packages (from gym[box2d]) (0.0.8)\n",
            "Requirement already satisfied: numpy>=1.18.0 in /usr/local/lib/python3.9/dist-packages (from gym[box2d]) (1.22.4)\n",
            "Requirement already satisfied: cloudpickle>=1.2.0 in /usr/local/lib/python3.9/dist-packages (from gym[box2d]) (2.2.1)\n",
            "Requirement already satisfied: importlib-metadata>=4.8.0 in /usr/local/lib/python3.9/dist-packages (from gym[box2d]) (6.4.1)\n",
            "Collecting swig==4.*\n",
            "  Downloading swig-4.1.1-py2.py3-none-manylinux_2_5_x86_64.manylinux1_x86_64.whl (1.8 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.8/1.8 MB\u001b[0m \u001b[31m27.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting pygame==2.1.0\n",
            "  Downloading pygame-2.1.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (18.3 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m18.3/18.3 MB\u001b[0m \u001b[31m56.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hCollecting box2d-py==2.3.5\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[31m33.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.9/dist-packages (from importlib-metadata>=4.8.0->gym[box2d]) (3.15.0)\n",
            "Building wheels for collected packages: box2d-py\n",
            "  \u001b[1;31merror\u001b[0m: \u001b[1msubprocess-exited-with-error\u001b[0m\n",
            "  \n",
            "  \u001b[31m×\u001b[0m \u001b[32mpython setup.py bdist_wheel\u001b[0m did not run successfully.\n",
            "  \u001b[31m│\u001b[0m exit code: \u001b[1;36m1\u001b[0m\n",
            "  \u001b[31m╰─>\u001b[0m See above for output.\n",
            "  \n",
            "  \u001b[1;35mnote\u001b[0m: This error originates from a subprocess, and is likely not a problem with pip.\n",
            "  Building wheel for box2d-py (setup.py) ... \u001b[?25lerror\n",
            "\u001b[31m  ERROR: Failed building wheel for box2d-py\u001b[0m\u001b[31m\n",
            "\u001b[0m\u001b[?25h  Running setup.py clean for box2d-py\n",
            "Failed to build box2d-py\n",
            "Installing collected packages: swig, box2d-py, pygame\n",
            "  Running setup.py install for box2d-py ... \u001b[?25l\u001b[?25hdone\n",
            "\u001b[33m  DEPRECATION: box2d-py was installed using the legacy 'setup.py install' method, because a wheel could not be built for it. pip 23.1 will enforce this behaviour change. A possible replacement is to fix the wheel build issue reported above. Discussion can be found at https://github.com/pypa/pip/issues/8368\u001b[0m\u001b[33m\n",
            "\u001b[0m  Attempting uninstall: pygame\n",
            "    Found existing installation: pygame 2.3.0\n",
            "    Uninstalling pygame-2.3.0:\n",
            "      Successfully uninstalled pygame-2.3.0\n",
            "Successfully installed box2d-py-2.3.5 pygame-2.1.0 swig-4.1.1\n"
          ]
        }
      ],
      "source": [
        "!pip install box2d-py\n",
        "!pip install gym[box2d]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OOgTdhE796G8"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "import tensorflow.keras as keras\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.optimizers.legacy import Adam\n",
        "import gym\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "from gym.wrappers import RecordVideo\n",
        "import glob\n",
        "import io\n",
        "import base64\n",
        "from IPython.display import HTML\n",
        "from IPython import display"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CVWtIfGZQE2p",
        "outputId": "87687587-23c6-4d9f-e265-26b644b2f3bc"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "2.12.0\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/ipykernel/ipkernel.py:283: DeprecationWarning: `should_run_async` will not call `transform_cell` automatically in the future. Please pass the result to `transformed_cell` argument and any exception that happen during thetransform in `preprocessing_exc_tuple` in IPython 7.17 and above.\n",
            "  and should_run_async(code)\n"
          ]
        }
      ],
      "source": [
        "print(tf.__version__)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4jclbREX5Ps3",
        "outputId": "5495a9af-5e84-4f1e-fd50-0d093fe12fc0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "QR-_j-ZQ-yDh"
      },
      "source": [
        "# **Replay Buffer**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "LJUAwgV0YcZx"
      },
      "source": [
        "The following class implements a **replay buffer** based on numpy memory arrays. This buffer is used to store experiences during the traininig phase. It allows the agent to learn from past experiences by randomly sampling a batch of transitions from the buffer, instead of relying solely on the most recent experience. In this case I have decided to set a memory limit to avoid overloading the device's RAM.\n",
        "\n",
        "In particular this class implements two functions:\n",
        "\n",
        "*   Store: is used to store an experience in the replay buffer. It takes as input a state, an action, a reward, a new state, and a boolean flag indicating whether the episode is done (done). It updates the buffers by storing the provided values at the current storage index (id) and increments the storage index.\n",
        "*   Sample: is used to randomly sample a batch of experiences from the replay buffer. It takes as input the desired batch size. It first determines the maximum index to sample from, which is the minimum of the current storage index (self.curr_storage) and the maximum storage size (self.max_storage). It then randomly selects batch_size indices from the range of valid indices and uses them to retrieve corresponding samples from the state, action, reward, new state, and terminal buffers. Finally, it returns these samples as separate arrays.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4QFzMS7r-2YZ"
      },
      "outputs": [],
      "source": [
        "class ReplayBuffer:\n",
        "  def __init__(self, max_size, input_shape, n_actions):\n",
        "    self.max_storage = max_size\n",
        "    self.curr_storage = 0\n",
        "    self.curr_buffer = np.zeros((self.max_storage, *input_shape))\n",
        "    self.new_buffer = np.zeros((self.max_storage, *input_shape))\n",
        "    self.action_buffer = np.zeros((self.max_storage, n_actions))\n",
        "    self.reward_buffer = np.zeros(self.max_storage)\n",
        "    self.term_buffer = np.zeros(self.max_storage, dtype=bool)\n",
        "  \n",
        "  def store(self, state, action, reward, new_state, done):\n",
        "    id = self.curr_storage % self.max_storage\n",
        "\n",
        "    self.curr_buffer[id] = state\n",
        "    self.new_buffer[id] = new_state\n",
        "    self.action_buffer[id] = action\n",
        "    self.reward_buffer[id] = reward\n",
        "    self.term_buffer[id] = done\n",
        "\n",
        "    self.curr_storage += 1\n",
        "  \n",
        "  def sample(self, batch_size):\n",
        "    max = min(self.curr_storage, self.max_storage)\n",
        "    batch = np.random.choice(max, batch_size)\n",
        "\n",
        "    return self.curr_buffer[batch], self.action_buffer[batch], self.reward_buffer[batch], self.new_buffer[batch], self.term_buffer[batch]"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "nKuGmLOnCIrP"
      },
      "source": [
        "# **Critic**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "WqN2Sl5WYf9G"
      },
      "source": [
        "This class implements the critic neural network which approximates the Q-function taking into account the current state and action. This class overwrite the keras.Model class, taking this into account the init function initialize the model architecture and the call function do the forward pass throug the network. I decided to use a sequential network with 2 dense layers applying RELU function as activation.\n",
        "\n",
        "**Important:** if you want to run this notebook, modify the *chkpt_dir* parameter. This refers to the address of the directory that will be used to store the models during training. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "L_H9pPa7Bgeq"
      },
      "outputs": [],
      "source": [
        "class Critic(keras.Model):\n",
        "    def __init__(self, n1, n2, name='critic', chkpt_dir='/content/drive/MyDrive/MAI/ATCI/implementation/models'):\n",
        "        super(Critic, self).__init__()\n",
        "        self.l1_size = n1\n",
        "        self.l2_size = n2\n",
        "        self.model_name = name\n",
        "        self.checkpoint_dir = chkpt_dir\n",
        "        self.checkpoint_file = os.path.join(self.checkpoint_dir, name)\n",
        "\n",
        "        # Net architecture\n",
        "        self.l1 = Dense(self.l1_size, activation='relu')\n",
        "        self.l2 = Dense(self.l2_size, activation='relu')\n",
        "        self.q = Dense(1, activation=None)\n",
        "\n",
        "    def call(self, state, action):\n",
        "        q = self.l1(tf.concat([state, action], axis=1))\n",
        "        q = self.l2(q)\n",
        "\n",
        "        q = self.q(q)\n",
        "\n",
        "        return q"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "DMq0jeTCF8H5"
      },
      "source": [
        "# **Actor**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "99dkreCzYkJs"
      },
      "source": [
        "This class implements the actor neural networks which takes in the current state as input and outputs an action based on the current policy. All the concepts explained in the previous case also apply to this class.\n",
        "\n",
        "**Important:** if you want to run this notebook, modify the *chkpt* parameter. This refers to the address of the directory that will be used to store the models during training. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "E0EsO-UOGDB3"
      },
      "outputs": [],
      "source": [
        "class Actor(keras.Model):\n",
        "    def __init__(self, n1, n2, n_actions, name='actor',\n",
        "                 chkpt='/content/drive/MyDrive/MAI/ATCI/implementation/models'):\n",
        "        super(Actor, self).__init__()\n",
        "        self.l1_size = n1\n",
        "        self.l2_size = n2\n",
        "        self.n_actions = n_actions\n",
        "        self.model_name = name\n",
        "        self.checkpoint_dir = chkpt\n",
        "        self.checkpoint_file = os.path.join(self.checkpoint_dir, name)\n",
        "\n",
        "        # Net architecture\n",
        "        self.l1 = Dense(self.l1_size, activation='relu')\n",
        "        self.l2 = Dense(self.l2_size, activation='relu')\n",
        "        self.mu = Dense(self.n_actions, activation='tanh')\n",
        "\n",
        "    def call(self, state):\n",
        "        prob = self.l1(state)\n",
        "        prob = self.l2(prob)\n",
        "\n",
        "        mu = self.mu(prob)\n",
        "\n",
        "        return mu"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "uAurJoSTHNaF"
      },
      "source": [
        "# **Agent**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "VFYFEs0pYmfG"
      },
      "source": [
        "This class implements the complete agent,which learns to interact with an environment in order to achieve a specific goal. \n",
        "\n",
        "All the logic of the algorithm is implemented in this class. In this way, the three differential concepts of TD3 can be detected:\n",
        "\n",
        "*   Twin Q-networks: It can be seen from the initialisation of the class that 2 Critic models are used. Moreover, in the training function the minimum Q of the two models is chosen (pessimistic q-learning).\n",
        "*   Clipped action exploration: in the next action function it can be seen that some noise is introduced to the policy and that clipped exploration is implemented.\n",
        "*   Delayed policy update: in the training function you can see that the actor is only updated when update_actor_interval indicates.\n",
        "\n",
        "Following, each of the functions implemented in this class are explained:\n",
        "\n",
        "* The __init__ method is the constructor of the Agent class. It initializes the agent with various parameters, such as learning rates (lr_a and lr_c), input shape (input_shape), the environment (env), update interval for the actor network (update_actor_interval), number of actions (n_actions), maximum size of the replay buffer (max_size), layer sizes for the actor and critic models (l1_size and l2_size), a parameter for soft target network updates (tau), and a name for the agent (name).\n",
        "\n",
        "* The __next_action__ method is used to determine the next action to take based on the current observation. It takes as input an observation and optional parameters for noise (noise) and a discount factor (gamma). It first converts the observation to a tensor and passes it through the actor network (self.actor) to obtain the mean action (mu). Some noise is added to the mean action, and the resulting action is clipped to the valid action range. The method also increments the time step count.\n",
        "\n",
        "* The __save_mem method__ is used to store an experience in the agent's replay buffer (self.replay_buffer). It takes as input a state, action, reward, new state, and done flag, and calls the store method of the replay buffer to store the experience.\n",
        "\n",
        "* The __training__ method is used to train the agent's critic and actor networks. It takes an optional batch size (batch_size), soft target network update parameter (tau), and discount factor (gamma). If the replay buffer does not have enough experiences to form a batch, the method returns early. Otherwise, it samples a batch of experiences from the replay buffer. The states, actions, rewards, and new states are converted to tensors. Within a tf.GradientTape context, the target actions are computed by passing the new states through the target actor network (self.target_actor). Some noise is added to the target actions, which are then clipped to the valid action range. The target Q-values are computed by passing the new states and target actions through the target critic networks (self.target_critic1 and self.target_critic2). The current Q-values are obtained by passing the states and actions through the critic networks (self.critic1 and self.critic2). The minimum of the target Q-values is taken as the Q-value estimate. The critic loss is computed by comparing the target Q-values with the current Q-values. The loss is then used to compute the gradients for the critic networks (self.critic1 and self.critic2), and the optimizer is applied to update the critic networks' weights. The learn_step_count is incremented, and if it is not a multiple of update_actor_interval, the method returns early. Otherwise, within another tf.GradientTape context, the actor loss is computed by passing the states through the actor network (self.actor). The gradients for the actor network are computed using the actor loss, and the optimizer is applied to update the actor network's weights.\n",
        "\n",
        "* Finally, the __update_network__ method is called to update the weights of the target networks (self.target_actor, self.target_critic1, and self.target_critic2) based on the current network weights.\n",
        "\n",
        "\n",
        "One thing to note is that, although it can be clearly seen that for each model, the model itself uses an additional model called target, in practice only the main model is trained and the target model is a separate copy of the original network that is periodically updated to match the weights. This is done to improve the stability of the models in the training phase. \n",
        "\n",
        "Another tricky point is the use of the tf.squeeze function. This is used to size tensors from (batch_size,1) shape to (batch_size) shape.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "qxXxMe6iHHay"
      },
      "outputs": [],
      "source": [
        "class Agent:\n",
        "  def __init__(self, lr_a, lr_c, input_shape, env,\n",
        "               update_actor_interval = 2,\n",
        "               n_actions = 2, max_size = 1000000,\n",
        "               l1_size=400, l2_size=300, tau=0.005, name=''):\n",
        "\n",
        "    self.n_actions=n_actions\n",
        "    \n",
        "    self.replay_buffer = ReplayBuffer(max_size, input_shape, n_actions)\n",
        "    \n",
        "    self.update_actor_interval = update_actor_interval\n",
        "    self.time_step = 0\n",
        "    self.learn_step_count = 0\n",
        "    self.max_action = env.action_space.high[0]\n",
        "    self.min_action = env.action_space.low[0]\n",
        "\n",
        "    # Needed networks\n",
        "    self.actor = Actor(l1_size, l2_size, n_actions=n_actions, name='actor_'+name)\n",
        "    self.critic1 = Critic(l1_size, l2_size, name='critic1_'+name)\n",
        "    self.critic2 = Critic(l1_size, l2_size, name='critic2_'+name)\n",
        "    self.target_actor = Actor(l1_size, l2_size, n_actions=n_actions, name='target_actor_'+name)\n",
        "    self.target_critic1 = Critic(l1_size, l2_size, name='target_critic1_'+name)\n",
        "    self.target_critic2 = Critic(l1_size, l2_size, name='target_critic2_'+name)\n",
        "\n",
        "    # Optimizers\n",
        "    opt_actor = Adam(learning_rate=lr_a)\n",
        "    opt_critics = Adam(learning_rate=lr_c)\n",
        "\n",
        "    self.actor.compile(optimizer=opt_actor, loss='mean')\n",
        "    self.target_actor.compile(optimizer=opt_actor, loss='mean')\n",
        "    self.critic1.compile(optimizer=opt_critics, loss='mean_squared_error')\n",
        "    self.target_critic1.compile(optimizer=opt_critics, loss='mean_squared_error')\n",
        "    self.critic2.compile(optimizer=opt_critics, loss='mean_squared_error')\n",
        "    self.target_critic2.compile(optimizer=opt_critics, loss='mean_squared_error')\n",
        "\n",
        "    self.update_network(tau=tau)\n",
        "  \n",
        "  def next_action(self, observation, noise=0.1, gamma=0.99):\n",
        "    # Get policy with some noise\n",
        "    state = tf.convert_to_tensor([observation], dtype=tf.float32)\n",
        "    mu = self.actor(state)[0]\n",
        "    mu_ = mu + np.random.normal(scale=noise)\n",
        "\n",
        "    # Clipped action exploration\n",
        "    mu_ = tf.clip_by_value(mu_, self.min_action, self.max_action)\n",
        "    self.time_step += 1\n",
        "\n",
        "    return mu_\n",
        "\n",
        "  def save_mem(self, state, action, reward, new_state, done):\n",
        "        self.replay_buffer.store(state, action, reward, new_state, done)\n",
        "  \n",
        "  def training(self, batch_size=100, tau=0.005, gamma=0.99):\n",
        "    if self.replay_buffer.curr_storage < batch_size:\n",
        "      return\n",
        "\n",
        "    states, actions, rewards, new_states, dones = self.replay_buffer.sample(batch_size)\n",
        "\n",
        "    # convert to tensor for training\n",
        "    states = tf.convert_to_tensor(states, dtype=tf.float32)\n",
        "    actions = tf.convert_to_tensor(actions, dtype=tf.float32)\n",
        "    rewards = tf.convert_to_tensor(rewards, dtype=tf.float32)\n",
        "    new_states = tf.convert_to_tensor(new_states, dtype=tf.float32)\n",
        "    \n",
        "    with tf.GradientTape(persistent=True) as tape:\n",
        "      # Select action according to policy and add clipped noise \n",
        "      target_actions = self.target_actor(new_states)\n",
        "      target_actions = target_actions + tf.clip_by_value(np.random.normal(scale=0.2), -0.5, 0.5)\n",
        "      target_actions = tf.clip_by_value(target_actions, self.min_action,\n",
        "                                        self.max_action)\n",
        "      # Compute the target Q value\n",
        "      q1_ = self.target_critic1(new_states, target_actions)\n",
        "      q2_ = self.target_critic2(new_states, target_actions)\n",
        "      q1 = tf.squeeze(self.critic1(states, actions), 1)\n",
        "      q2 = tf.squeeze(self.critic2(states, actions), 1)\n",
        "      # shape is [batch_size, 1], want to collapse to [batch_size],\n",
        "      # squeeze removes dimensions of size 1 from the shape of a tensor.\n",
        "      q1_ = tf.squeeze(q1_, 1)\n",
        "      q2_ = tf.squeeze(q2_, 1)\n",
        "      # pessimistic double-Q learning\n",
        "      q = tf.math.minimum(q1_, q2_) \n",
        "\n",
        "      # Compute critic loss\n",
        "      target = rewards + gamma*q*(1-dones)\n",
        "      critic1_loss = keras.losses.MSE(target, q1)\n",
        "      critic2_loss = keras.losses.MSE(target, q2)\n",
        "\n",
        "    # Optimize the critic\n",
        "    critic1_gradient = tape.gradient(critic1_loss,\n",
        "                                self.critic1.trainable_variables)\n",
        "    critic2_gradient = tape.gradient(critic2_loss,\n",
        "                                self.critic2.trainable_variables)\n",
        "    self.critic1.optimizer.apply_gradients(\n",
        "          zip(critic1_gradient, self.critic1.trainable_variables))\n",
        "    self.critic2.optimizer.apply_gradients(\n",
        "          zip(critic2_gradient, self.critic2.trainable_variables))\n",
        "\n",
        "    self.learn_step_count += 1\n",
        "\n",
        "    # delayed policy update\n",
        "    if self.learn_step_count % self.update_actor_interval != 0:\n",
        "      return\n",
        "\n",
        "    with tf.GradientTape() as tape:\n",
        "      # Compute actor loss\n",
        "      new_actions = self.actor(states)\n",
        "      critic1_value = self.critic1(states, new_actions)\n",
        "      actor_loss = -tf.math.reduce_mean(critic1_value)\n",
        "\n",
        "    # Optimize the actor\n",
        "    actor_gradient = tape.gradient(actor_loss, self.actor.trainable_variables)\n",
        "    self.actor.optimizer.apply_gradients(zip(actor_gradient, self.actor.trainable_variables))\n",
        "\n",
        "    self.update_network(tau=tau)\n",
        "\n",
        "  def update_network(self, tau):\n",
        "    # Update weights of a network\n",
        "      weights = []\n",
        "      targets = self.target_actor.weights\n",
        "      for i, weight in enumerate(self.actor.weights):\n",
        "          weights.append(weight * tau + targets[i]*(1-tau))\n",
        "\n",
        "      self.target_actor.set_weights(weights)\n",
        "\n",
        "      weights = []\n",
        "      targets = self.target_critic1.weights\n",
        "      for i, weight in enumerate(self.critic1.weights):\n",
        "          weights.append(weight * tau + targets[i]*(1-tau))\n",
        "\n",
        "      self.target_critic1.set_weights(weights)\n",
        "\n",
        "      weights = []\n",
        "      targets = self.target_critic2.weights\n",
        "      for i, weight in enumerate(self.critic2.weights):\n",
        "          weights.append(weight * tau + targets[i]*(1-tau))\n",
        "\n",
        "      self.target_critic2.set_weights(weights)\n",
        "\n",
        "  def save_models(self):\n",
        "      print('Saving models ...')\n",
        "      self.actor.save_weights(self.actor.checkpoint_file)\n",
        "      self.critic1.save_weights(self.critic1.checkpoint_file)\n",
        "      self.critic2.save_weights(self.critic2.checkpoint_file)\n",
        "      self.target_actor.save_weights(self.target_actor.checkpoint_file)\n",
        "      self.target_critic1.save_weights(self.target_critic1.checkpoint_file)\n",
        "      self.target_critic2.save_weights(self.target_critic2.checkpoint_file)\n",
        "\n",
        "  def load_models(self):\n",
        "      print('Loading models ...')\n",
        "      self.actor.load_weights(self.actor.checkpoint_file)\n",
        "      self.critic1.load_weights(self.critic1.checkpoint_file)\n",
        "      self.critic2.load_weights(self.critic2.checkpoint_file)\n",
        "      self.target_actor.load_weights(self.target_actor.checkpoint_file)\n",
        "      self.target_critic1.load_weights(self.target_critic1.checkpoint_file)\n",
        "      self.target_critic2.load_weights(self.target_critic2.checkpoint_file)\n",
        "    \n",
        "  \n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "ie-4HhccFElW"
      },
      "source": [
        "# **Extra functions**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "3TY10d6zYrVX"
      },
      "source": [
        "In this section you can see the functions that have been used to run the experiments."
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "ekOJrYz9FOr3"
      },
      "source": [
        "Parameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4-fXBFsrFOXi"
      },
      "outputs": [],
      "source": [
        "ENV='Pendulum-v1'\n",
        "LR_ACTOR=0.001\n",
        "LR_CRITIC=0.002\n",
        "GAMMA=0.99\n",
        "TAU=0.005\n",
        "NOISE=0.1\n",
        "BATCH_SIZE=128\n",
        "L1_SIZE=512\n",
        "L2_SIZE=512\n",
        "UPDATE_ACTOR_INTERVAL=2\n",
        "N_ITERATIONS=300\n",
        "MAX_SIZE=100000\n",
        "EXP_NAME='pendulum'\n",
        "SAVE_POINTS=[100, 200, 250, 300, 350, 400, 450]"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "nt66Fcv9F1S3"
      },
      "source": [
        "Training function\n",
        "\n",
        "The train function trains an agent in a gym environment using the Agent class. It performs a specified number of iterations. In each iteration, the agent interacts with the environment, collects experiences, and updates its critic and actor networks. The function keeps track of the scores achieved in each episode and the average scores over time. At the end of training, the trained agent, episode scores, and average scores are returned."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WXE6M3g6FDAx"
      },
      "outputs": [],
      "source": [
        "def train():\n",
        "  env = gym.make(ENV)\n",
        "\n",
        "  agent = Agent(lr_a=LR_ACTOR, lr_c=LR_CRITIC, input_shape=env.observation_space.shape, env=env, update_actor_interval=UPDATE_ACTOR_INTERVAL,\n",
        "                n_actions=env.action_space.shape[0], l1_size=L1_SIZE, l2_size=L2_SIZE, max_size=MAX_SIZE, name=EXP_NAME)\n",
        "\n",
        "  best_score = env.reward_range[0]\n",
        "  scores=[]\n",
        "  avg_history=[]\n",
        "\n",
        "  with tf.device('GPU:0'):\n",
        "    tf.random.set_seed(123)\n",
        "    for i in range(N_ITERATIONS):\n",
        "      obs = env.reset()\n",
        "      done = False\n",
        "      score = 0\n",
        "      while not done:\n",
        "        action = agent.next_action(observation=obs, noise=NOISE, gamma=GAMMA)\n",
        "        new_obs, reward, done, info = env.step(action)\n",
        "        agent.save_mem(state=obs, action=action, reward=reward, new_state=new_obs, done=done)\n",
        "        agent.training(batch_size=BATCH_SIZE, tau=TAU)\n",
        "        score += reward\n",
        "        obs = new_obs\n",
        "\n",
        "      scores.append(score)\n",
        "      # Mean of last 50 scores\n",
        "      mean_score = np.mean(scores[-100:])\n",
        "\n",
        "      if mean_score > best_score:\n",
        "        best_score = mean_score\n",
        "\n",
        "      if i in SAVE_POINTS:\n",
        "        agent.save_models()\n",
        "\n",
        "      avg_history.append(mean_score)\n",
        "      print(f\"# Episode: {i}, Reward: {score}, Mean reward: {mean_score}.\")\n",
        "    \n",
        "    agent.save_models()\n",
        "    env.close()\n",
        "\n",
        "  return agent, scores, avg_history\n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "dg305ALJ8A60"
      },
      "source": [
        "Plot Training Curve"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "The show_curve function is used to visualize the training progress of an agent by plotting the average test reward over episodes. It takes two arguments: avg, which is a list representing the average test rewards at each episode, and score, which is a list representing the test scores achieved at each episode. The function creates a line plot with the x-axis representing the episode numbers and the y-axis representing the average test reward. The average test rewards are plotted in red, while the test scores are plotted in blue with reduced opacity. The function then displays the plot."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "HyAaDH5Ig_c3"
      },
      "outputs": [],
      "source": [
        "def show_curve(avg, score):\n",
        "  ep = [i  for i in range(len(avg))]\n",
        "  plt.plot( range(len(avg)),avg,'r')\n",
        "  plt.plot(score, color = 'b', alpha=0.5)\n",
        "  plt.xlabel(\"Episode\")\n",
        "  plt.ylabel(\"Average Test Reward\")\n",
        "  plt.show()"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "n0QgMVqw8Lni"
      },
      "source": [
        "Show Experiment Video"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n",
        "The show_video function is used to display a video file in the Jupyter Notebook environment. It looks for video files with the .mp4 extension in the \"video\" directory. If a video file is found, it reads the file, encodes it using base64, and displays it as an HTML5 video element. The video is set to autoplay, loop, and has controls for playback. The function uses the display and HTML functions from IPython to show the video. If no video file is found, it prints a message indicating that the video could not be found."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "pLd_ZAaXxeLb"
      },
      "outputs": [],
      "source": [
        "def show_video():\n",
        "  mp4list = glob.glob('video/*.mp4')\n",
        "  if len(mp4list) > 0:\n",
        "    mp4 = mp4list[0]\n",
        "    video = io.open(mp4, 'r+b').read()\n",
        "    encoded = base64.b64encode(video)\n",
        "    display.display(HTML(data='''<video alt=\"test\" autoplay \n",
        "                loop controls style=\"height: 400px;\">\n",
        "                <source src=\"data:video/mp4;base64,{0}\" type=\"video/mp4\" />\n",
        "             </video>'''.format(encoded.decode('ascii'))))\n",
        "  else: \n",
        "    print(\"Could not find video\")"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "KD4h8Qqy8yvA"
      },
      "source": [
        "Test Learned Behavior"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n",
        "The test_behavior function is used to test the behavior of an agent in a gym environment and record a video of the agent's interactions."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dH8KkUmA8vwM"
      },
      "outputs": [],
      "source": [
        "def test_behavior(agent):\n",
        "  env = RecordVideo(gym.make(ENV,render_mode='rgb_array',new_step_api=True),'video',new_step_api=True)\n",
        "    \n",
        "  obs = env.reset()\n",
        "  agent.load_models()\n",
        "  while True:\n",
        "      env.render()\n",
        "      action = agent.next_action(observation=obs, noise=NOISE, gamma=GAMMA)\n",
        "      new_obs, reward, done, truncated, info = env.step(action)\n",
        "      if done or truncated: break\n",
        "      obs = new_obs\n",
        "  env.close()\n",
        "  show_video()"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "JZEk1Yg6-eOV"
      },
      "source": [
        "Test Random Behavior"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n",
        "The random_behavior function is used to observe the behavior of an agent that takes random actions in a gym environment and record a video of its interactions. It does not rely on any pre-trained agent or specific algorithm."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "AdCgcsyw-gLj"
      },
      "outputs": [],
      "source": [
        "def random_behavior():\n",
        "  env = RecordVideo(gym.make(ENV,render_mode='rgb_array',new_step_api=True),'video',new_step_api=True)\n",
        "\n",
        "  observation = env.reset()\n",
        "\n",
        "  for _ in range(300):\n",
        "    env.render()\n",
        "    action = env.action_space.sample() # this takes random actions\n",
        "    observation, reward, terminated , truncated, info = env.step(action)\n",
        "    if terminated or truncated:\n",
        "      break\n",
        "  env.close()\n",
        "  show_video()"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "ESj1kJ4t9XRH"
      },
      "source": [
        "# **Experiments**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "dvCTvuOlYucK"
      },
      "source": [
        "In order to test the algorithm I used two Gym environments: \n",
        "\n",
        "*   Pendulum\n",
        "*   Lunar Lander in continious version"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "A0ELfvjv9cZj"
      },
      "source": [
        "# **1-Pendulum**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "s_TDcqqy9ZbX"
      },
      "outputs": [],
      "source": [
        "ENV='Pendulum-v1'\n",
        "LR_ACTOR=0.001\n",
        "LR_CRITIC=0.002\n",
        "GAMMA=0.99\n",
        "TAU=0.005\n",
        "NOISE=0.1\n",
        "BATCH_SIZE=128\n",
        "L1_SIZE=512\n",
        "L2_SIZE=512\n",
        "UPDATE_ACTOR_INTERVAL=2\n",
        "N_ITERATIONS=300\n",
        "MAX_SIZE=100000\n",
        "EXP_NAME='pendulum'"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "AUifM-hU-Boq"
      },
      "source": [
        "Behaviour before learning:"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "b7AnDMzBZAvG"
      },
      "source": [
        "As expected, the initial behaviour of the untrained agent is totally random."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nsFl8nv2-BAV",
        "outputId": "e4035bfc-9970-4f0e-f41f-bfbd0a236c7c"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/record_video.py:78: UserWarning: \u001b[33mWARN: Overwriting existing videos at /content/video folder (try specifying a different `video_folder` for the `RecordVideo` wrapper if this is not desired)\u001b[0m\n",
            "  logger.warn(\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<video alt=\"test\" autoplay \n",
              "                loop controls style=\"height: 400px;\">\n",
              "                <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\" />\n",
              "             </video>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "random_behavior()"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "220rrsOfA6BI"
      },
      "source": [
        "Training:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q9pDgPJwA4fI",
        "outputId": "a7463993-ef2c-4054-a861-8dd9902a3bc2"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/core.py:317: DeprecationWarning: \u001b[33mWARN: Initializing wrapper in old step API which returns one bool instead of two. It is recommended to set `new_step_api=True` to use new step API. This will be the default behaviour in future.\u001b[0m\n",
            "  deprecation(\n",
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/step_api_compatibility.py:39: DeprecationWarning: \u001b[33mWARN: Initializing environment in old step API which returns one bool instead of two. It is recommended to set `new_step_api=True` to use new step API. This will be the default behaviour in future.\u001b[0m\n",
            "  deprecation(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "# Episode: 0, Reward: -1562.7877723822164, Mean reward: -1562.7877723822164.\n",
            "# Episode: 1, Reward: -1361.4354621836085, Mean reward: -1462.1116172829124.\n",
            "# Episode: 2, Reward: -1343.92031217996, Mean reward: -1422.7145155819283.\n",
            "# Episode: 3, Reward: -1356.8107492610375, Mean reward: -1406.2385740017057.\n",
            "# Episode: 4, Reward: -1402.3624143374261, Mean reward: -1405.46334206885.\n",
            "# Episode: 5, Reward: -1610.3995975035982, Mean reward: -1439.619384641308.\n",
            "# Episode: 6, Reward: -1642.1109866652127, Mean reward: -1468.5467563590087.\n",
            "# Episode: 7, Reward: -1426.276993509213, Mean reward: -1463.2630360027842.\n",
            "# Episode: 8, Reward: -1526.5718965928436, Mean reward: -1470.297353846124.\n",
            "# Episode: 9, Reward: -1601.0449361066994, Mean reward: -1483.3721120721816.\n",
            "# Episode: 10, Reward: -1632.4195736371332, Mean reward: -1496.9218813053592.\n",
            "# Episode: 11, Reward: -1618.3070587284117, Mean reward: -1507.0373127572802.\n",
            "# Episode: 12, Reward: -1500.0601440474577, Mean reward: -1506.5006074719092.\n",
            "# Episode: 13, Reward: -1443.7748427851704, Mean reward: -1502.0201957085706.\n",
            "# Episode: 14, Reward: -1605.856108746786, Mean reward: -1508.9425899111184.\n",
            "# Episode: 15, Reward: -1407.7304553550618, Mean reward: -1502.6168315013647.\n",
            "# Episode: 16, Reward: -1431.5150723891027, Mean reward: -1498.4343750829962.\n",
            "# Episode: 17, Reward: -1542.1406742952418, Mean reward: -1500.86250281701.\n",
            "# Episode: 18, Reward: -1333.9310054769703, Mean reward: -1492.0766345359552.\n",
            "# Episode: 19, Reward: -1596.194111067101, Mean reward: -1497.2825083625125.\n",
            "# Episode: 20, Reward: -1544.1416115454483, Mean reward: -1499.5138942283666.\n",
            "# Episode: 21, Reward: -1334.8386853561851, Mean reward: -1492.028657461449.\n",
            "# Episode: 22, Reward: -1222.440017689133, Mean reward: -1480.3074122539572.\n",
            "# Episode: 23, Reward: -1244.7076291702547, Mean reward: -1470.4907546254697.\n",
            "# Episode: 24, Reward: -1849.8587307451246, Mean reward: -1485.6654736702558.\n",
            "# Episode: 25, Reward: -1139.5673601294225, Mean reward: -1472.3540077648393.\n",
            "# Episode: 26, Reward: -1209.5439464295478, Mean reward: -1462.6203017894582.\n",
            "# Episode: 27, Reward: -1283.6819527180066, Mean reward: -1456.2296464654776.\n",
            "# Episode: 28, Reward: -1311.7590436580167, Mean reward: -1451.2479015410822.\n",
            "# Episode: 29, Reward: -1727.5391400533667, Mean reward: -1460.457609491492.\n",
            "# Episode: 30, Reward: -1344.8025328583071, Mean reward: -1456.7268005678409.\n",
            "# Episode: 31, Reward: -1263.0352807211386, Mean reward: -1450.6739405726314.\n",
            "# Episode: 32, Reward: -1090.0322702825226, Mean reward: -1439.7454051092948.\n",
            "# Episode: 33, Reward: -964.6306197058785, Mean reward: -1425.7714408327238.\n",
            "# Episode: 34, Reward: -999.1826075827698, Mean reward: -1413.5831884541537.\n",
            "# Episode: 35, Reward: -820.7368043974894, Mean reward: -1397.1152333414686.\n",
            "# Episode: 36, Reward: -1001.6962232742645, Mean reward: -1386.428233069382.\n",
            "# Episode: 37, Reward: -616.4755407853949, Mean reward: -1366.1663201145402.\n",
            "# Episode: 38, Reward: -878.4856897301255, Mean reward: -1353.661688566222.\n",
            "# Episode: 39, Reward: -873.1413496213816, Mean reward: -1341.6486800926007.\n",
            "# Episode: 40, Reward: -997.7200058398942, Mean reward: -1333.2601758425346.\n",
            "# Episode: 41, Reward: -591.3227006294347, Mean reward: -1315.5949978612703.\n",
            "# Episode: 42, Reward: -1726.895061254992, Mean reward: -1325.1601156146126.\n",
            "# Episode: 43, Reward: -758.8838761360727, Mean reward: -1312.2902010810094.\n",
            "# Episode: 44, Reward: -987.714786203871, Mean reward: -1305.0774140837398.\n",
            "# Episode: 45, Reward: -1767.1234013579933, Mean reward: -1315.1218920679626.\n",
            "# Episode: 46, Reward: -761.3189814508729, Mean reward: -1303.3388514165351.\n",
            "# Episode: 47, Reward: -714.8536243176349, Mean reward: -1291.0787425186415.\n",
            "# Episode: 48, Reward: -1820.2651821640177, Mean reward: -1301.8784657767105.\n",
            "# Episode: 49, Reward: -1081.2301419013795, Mean reward: -1297.4654992992039.\n",
            "# Episode: 50, Reward: -1152.3863088325393, Mean reward: -1289.2574700282103.\n",
            "# Episode: 51, Reward: -1013.215303901061, Mean reward: -1282.2930668625593.\n",
            "# Episode: 52, Reward: -1278.043395431276, Mean reward: -1280.9755285275858.\n",
            "# Episode: 53, Reward: -1181.6151216893502, Mean reward: -1277.471615976152.\n",
            "# Episode: 54, Reward: -899.5065282463488, Mean reward: -1267.4144982543303.\n",
            "# Episode: 55, Reward: -890.3884405564133, Mean reward: -1253.0142751153867.\n",
            "# Episode: 56, Reward: -1012.4450936221324, Mean reward: -1240.420957254525.\n",
            "# Episode: 57, Reward: -1015.956456487028, Mean reward: -1232.2145465140813.\n",
            "# Episode: 58, Reward: -961.4900692646497, Mean reward: -1220.9129099675174.\n",
            "# Episode: 59, Reward: -503.2769183732298, Mean reward: -1198.9575496128482.\n",
            "# Episode: 60, Reward: -647.2510231762222, Mean reward: -1179.25417860363.\n",
            "# Episode: 61, Reward: -569.9195764795688, Mean reward: -1158.2864289586528.\n",
            "# Episode: 62, Reward: -1768.2528971160455, Mean reward: -1163.6502840200246.\n",
            "# Episode: 63, Reward: -1849.4999410253265, Mean reward: -1171.764785984828.\n",
            "# Episode: 64, Reward: -247.69127711997874, Mean reward: -1144.6014893522918.\n",
            "# Episode: 65, Reward: -124.520686421582, Mean reward: -1118.9372939736222.\n",
            "# Episode: 66, Reward: -250.96258953738746, Mean reward: -1095.326244316588.\n",
            "# Episode: 67, Reward: -123.12736136432878, Mean reward: -1066.9459780579696.\n",
            "# Episode: 68, Reward: -247.47727449729308, Mean reward: -1045.2169034383762.\n",
            "# Episode: 69, Reward: -486.35152685808197, Mean reward: -1023.0200517541957.\n",
            "# Episode: 70, Reward: -121.06801176253148, Mean reward: -994.5585797585372.\n",
            "# Episode: 71, Reward: -126.12178950966162, Mean reward: -970.3842418416068.\n",
            "# Episode: 72, Reward: -124.68299345969903, Mean reward: -948.4291013570182.\n",
            "# Episode: 73, Reward: -1768.81771946083, Mean reward: -958.9113031628295.\n",
            "# Episode: 74, Reward: -1317.8592103785847, Mean reward: -948.2713127554989.\n",
            "# Episode: 75, Reward: -125.96508157099453, Mean reward: -927.9992671843304.\n",
            "# Episode: 76, Reward: -124.86154925421897, Mean reward: -906.3056192408238.\n",
            "# Episode: 77, Reward: -125.10810520554872, Mean reward: -883.1341422905746.\n",
            "# Episode: 78, Reward: -1817.0037797866976, Mean reward: -893.239037013148.\n",
            "# Episode: 79, Reward: -485.14270681520463, Mean reward: -868.391108348385.\n",
            "# Episode: 80, Reward: -242.25248770571577, Mean reward: -846.3401074453332.\n",
            "# Episode: 81, Reward: -0.16088524511623073, Mean reward: -821.0826195358127.\n",
            "# Episode: 82, Reward: -0.37655954144025167, Mean reward: -799.289505320991.\n",
            "# Episode: 83, Reward: -246.84852421148202, Mean reward: -784.9338634111032.\n",
            "# Episode: 84, Reward: -238.85350418128604, Mean reward: -769.7272813430735.\n",
            "# Episode: 85, Reward: -126.51073678903167, Mean reward: -755.8427599909043.\n",
            "# Episode: 86, Reward: -0.07694916273697687, Mean reward: -735.8103745086737.\n",
            "# Episode: 87, Reward: -0.3540789145446045, Mean reward: -723.4879452712568.\n",
            "# Episode: 88, Reward: -1684.4994792339974, Mean reward: -739.6082210613343.\n",
            "# Episode: 89, Reward: -369.9571952701445, Mean reward: -729.5445379743095.\n",
            "# Episode: 90, Reward: -1714.0712311386797, Mean reward: -743.8715624802852.\n",
            "# Episode: 91, Reward: -125.9597728941272, Mean reward: -734.5643039255791.\n",
            "# Episode: 92, Reward: -1767.3403761160225, Mean reward: -735.3732102227997.\n",
            "# Episode: 93, Reward: -244.61978599329467, Mean reward: -725.0879284199441.\n",
            "# Episode: 94, Reward: -241.08306082269254, Mean reward: -710.1552939123205.\n",
            "# Episode: 95, Reward: -505.494147641591, Mean reward: -684.9227088379924.\n",
            "# Episode: 96, Reward: -123.51443717383053, Mean reward: -672.1666179524516.\n",
            "# Episode: 97, Reward: -490.983110234996, Mean reward: -667.6892076707987.\n",
            "# Episode: 98, Reward: -125.18571577681207, Mean reward: -633.7876183430548.\n",
            "# Episode: 99, Reward: -1.1043235683937573, Mean reward: -612.1851019763951.\n",
            "# Episode: 100, Reward: -127.29385311973925, Mean reward: -591.683252862139.\n",
            "# Episode: 101, Reward: -502.8341819494419, Mean reward: -581.4756304231066.\n",
            "# Episode: 102, Reward: -239.67339602983338, Mean reward: -560.7082304350777.\n",
            "# Episode: 103, Reward: -496.1652082366493, Mean reward: -546.9992321660237.\n",
            "# Episode: 104, Reward: -482.17880518355645, Mean reward: -538.6526777047679.\n",
            "# Episode: 105, Reward: -126.33762090037119, Mean reward: -523.3716613116471.\n",
            "# Episode: 106, Reward: -241.69381511593505, Mean reward: -507.95663574152314.\n",
            "# Episode: 107, Reward: -246.71170044522913, Mean reward: -492.57174062068714.\n",
            "# Episode: 108, Reward: -120.10206967396866, Mean reward: -475.74398062887354.\n",
            "# Episode: 109, Reward: -242.8586391075948, Mean reward: -470.5356150435608.\n",
            "# Episode: 110, Reward: -125.58743670460788, Mean reward: -460.1023433141286.\n",
            "# Episode: 111, Reward: -124.27458571661924, Mean reward: -451.18944349886954.\n",
            "# Episode: 112, Reward: -241.83636866399033, Mean reward: -420.6611129298284.\n",
            "# Episode: 113, Reward: -1.3274321462674412, Mean reward: -383.6976627522472.\n",
            "# Episode: 114, Reward: -123.94303897522518, Mean reward: -381.22269798935224.\n",
            "# Episode: 115, Reward: -373.94044732125946, Mean reward: -386.21109320734575.\n",
            "# Episode: 116, Reward: -118.73771596185165, Mean reward: -383.566595735835.\n",
            "# Episode: 117, Reward: -124.26929108545362, Mean reward: -383.5894343302576.\n",
            "# Episode: 118, Reward: -633.7380780478151, Mean reward: -391.314650401268.\n",
            "# Episode: 119, Reward: -1.541880794669653, Mean reward: -381.61845747999973.\n",
            "# Episode: 120, Reward: -121.50155744527255, Mean reward: -381.62712839365463.\n",
            "# Episode: 121, Reward: -124.52600256123664, Mean reward: -381.5952126546861.\n",
            "# Episode: 122, Reward: -124.85262059229919, Mean reward: -381.5986051973381.\n",
            "# Episode: 123, Reward: -376.208367292793, Mean reward: -353.7464181539774.\n",
            "# Episode: 124, Reward: -122.00326222363539, Mean reward: -329.8292991908783.\n",
            "# Episode: 125, Reward: -367.1106707356174, Mean reward: -334.6522109741708.\n",
            "# Episode: 126, Reward: -356.23341307469735, Mean reward: -339.2796482505803.\n",
            "# Episode: 127, Reward: -120.0949415648, Mean reward: -339.17938497776527.\n",
            "# Episode: 128, Reward: -826.2249272372575, Mean reward: -319.36380792677653.\n",
            "# Episode: 129, Reward: -500.2865687294518, Mean reward: -319.6666851650615.\n",
            "# Episode: 130, Reward: -378.71330013889093, Mean reward: -322.39590141372497.\n",
            "# Episode: 131, Reward: -126.12191029085925, Mean reward: -324.9151219146398.\n",
            "# Episode: 132, Reward: -237.24766276264194, Mean reward: -329.65254397906386.\n",
            "# Episode: 133, Reward: -126.32992069071126, Mean reward: -327.24217190864846.\n",
            "# Episode: 134, Reward: -373.22689524867855, Mean reward: -329.92963972999627.\n",
            "# Episode: 135, Reward: -122.79656442383008, Mean reward: -329.8553562826923.\n",
            "# Episode: 136, Reward: -477.8493830780354, Mean reward: -339.41080496099823.\n",
            "# Episode: 137, Reward: -237.9090219379713, Mean reward: -344.16190382146686.\n",
            "# Episode: 138, Reward: -245.17722494788754, Mean reward: -315.37545873574453.\n",
            "# Episode: 139, Reward: -121.40257875138226, Mean reward: -310.40436640536933.\n",
            "# Episode: 140, Reward: -494.4256033648532, Mean reward: -286.0114538498928.\n",
            "# Episode: 141, Reward: -125.55905050942413, Mean reward: -286.00343940219875.\n",
            "# Episode: 142, Reward: -1698.0615380999548, Mean reward: -284.6178626418774.\n",
            "# Episode: 143, Reward: -629.3311811993128, Mean reward: -292.3120905459978.\n",
            "# Episode: 144, Reward: -240.82374586366487, Mean reward: -292.3069042468172.\n",
            "# Episode: 145, Reward: -122.47784356970755, Mean reward: -284.64657816537954.\n",
            "# Episode: 146, Reward: -507.80681776733803, Mean reward: -292.3324257772497.\n",
            "# Episode: 147, Reward: -126.58288442134442, Mean reward: -285.04442126097666.\n",
            "# Episode: 148, Reward: -1760.6736737476926, Mean reward: -317.7541804203943.\n",
            "# Episode: 149, Reward: -364.72354768719316, Mean reward: -325.0265649027703.\n",
            "# Episode: 150, Reward: -123.27602769175391, Mean reward: -324.94620839421054.\n",
            "# Episode: 151, Reward: -1771.3319720752559, Mean reward: -350.31616419672685.\n",
            "# Episode: 152, Reward: -124.23363037757943, Mean reward: -348.00736888368175.\n",
            "# Episode: 153, Reward: -1730.8088551189248, Mean reward: -372.70024182132727.\n",
            "# Episode: 154, Reward: -123.48948701029381, Mean reward: -365.52645545786197.\n",
            "# Episode: 155, Reward: -0.6877930997149249, Mean reward: -363.0134589018489.\n",
            "# Episode: 156, Reward: -479.24491774042053, Mean reward: -367.7644809543385.\n",
            "# Episode: 157, Reward: -1749.4411637718683, Mean reward: -397.81907022087137.\n",
            "# Episode: 158, Reward: -234.7894012979782, Mean reward: -400.11281685335155.\n",
            "# Episode: 159, Reward: -124.0752807777356, Mean reward: -397.7371496867543.\n",
            "# Episode: 160, Reward: -1853.5012822111185, Mean reward: -432.2954265968846.\n",
            "# Episode: 161, Reward: -249.27536163254666, Mean reward: -434.7954421152031.\n",
            "# Episode: 162, Reward: -0.8212808710090668, Mean reward: -429.97514035934347.\n",
            "# Episode: 163, Reward: -125.80852591884518, Mean reward: -432.46476223479505.\n",
            "# Episode: 164, Reward: -475.3331263756432, Mean reward: -439.49256398280346.\n",
            "# Episode: 165, Reward: -121.21107206631153, Mean reward: -434.4379764777044.\n",
            "# Episode: 166, Reward: -124.66968982389449, Mean reward: -434.55661595494536.\n",
            "# Episode: 167, Reward: -240.72541497635777, Mean reward: -436.8857384327634.\n",
            "# Episode: 168, Reward: -124.8731233752167, Mean reward: -426.70843933931144.\n",
            "# Episode: 169, Reward: -485.66480025168914, Mean reward: -436.39089772845176.\n",
            "# Episode: 170, Reward: -126.66536474871323, Mean reward: -436.4941738745207.\n",
            "# Episode: 171, Reward: -118.65891999581214, Mean reward: -436.3768322232122.\n",
            "# Episode: 172, Reward: -246.17153340974937, Mean reward: -438.8032104795612.\n",
            "# Episode: 173, Reward: -363.2510662905924, Mean reward: -438.5440644595172.\n",
            "# Episode: 174, Reward: -484.09859319637314, Mean reward: -445.7859710789719.\n",
            "# Episode: 175, Reward: -236.38326880285177, Mean reward: -443.1714230403166.\n",
            "# Episode: 176, Reward: -126.36848232827298, Mean reward: -438.5741244253882.\n",
            "# Episode: 177, Reward: -238.04197712698976, Mean reward: -440.9330651366319.\n",
            "# Episode: 178, Reward: -238.67079798957732, Mean reward: -429.1819825516783.\n",
            "# Episode: 179, Reward: -242.8256705719036, Mean reward: -424.0327645885273.\n",
            "# Episode: 180, Reward: -485.0264553384061, Mean reward: -426.1590276925177.\n",
            "# Episode: 181, Reward: -1602.4835093430172, Mean reward: -455.68625967356076.\n",
            "# Episode: 182, Reward: -353.76008847672705, Mean reward: -458.0165081878425.\n",
            "# Episode: 183, Reward: -235.0935304679208, Mean reward: -460.19178038338674.\n",
            "# Episode: 184, Reward: -678.1178857306672, Mean reward: -466.2896001930265.\n",
            "# Episode: 185, Reward: -236.76264038541217, Mean reward: -468.5689217122581.\n",
            "# Episode: 186, Reward: -487.23733918412796, Mean reward: -468.75668083437995.\n",
            "# Episode: 187, Reward: -122.65071484673048, Mean reward: -466.45151469255524.\n",
            "# Episode: 188, Reward: -124.02019596121741, Mean reward: -464.0283741128218.\n",
            "# Episode: 189, Reward: -122.53127719499612, Mean reward: -464.05094808169406.\n",
            "# Episode: 190, Reward: -125.25444918012978, Mean reward: -456.66752499799964.\n",
            "# Episode: 191, Reward: -118.88667126332855, Mean reward: -456.5340774130777.\n",
            "# Episode: 192, Reward: -124.67504150867458, Mean reward: -425.066347481252.\n",
            "# Episode: 193, Reward: -501.1280225383371, Mean reward: -422.5022843080326.\n",
            "# Episode: 194, Reward: -370.4501802593436, Mean reward: -425.09481299594614.\n",
            "# Episode: 195, Reward: -241.75560982760368, Mean reward: -427.48036832110404.\n",
            "# Episode: 196, Reward: -363.8158122931417, Mean reward: -424.60054821162004.\n",
            "# Episode: 197, Reward: -121.0296843038079, Mean reward: -424.48948420926934.\n",
            "# Episode: 198, Reward: -0.5112335062875162, Mean reward: -389.28623540444124.\n",
            "# Episode: 199, Reward: -126.31769086994498, Mean reward: -384.5181182680963.\n",
            "# Episode: 200, Reward: -485.20503530877636, Mean reward: -391.75669842043675.\n",
            "# Episode: 201, Reward: -371.9951140300931, Mean reward: -363.76996125953355.\n",
            "# Episode: 202, Reward: -241.9497502315216, Mean reward: -366.1242836566123.\n",
            "# Episode: 203, Reward: -120.98942221760763, Mean reward: -333.92789499858605.\n",
            "# Episode: 204, Reward: -494.26052090390465, Mean reward: -341.34331567645813.\n",
            "# Episode: 205, Reward: -616.4256339666218, Mean reward: -353.65807249379634.\n",
            "# Episode: 206, Reward: -367.5930059462569, Mean reward: -351.425034257913.\n",
            "# Episode: 207, Reward: -237.27439865577907, Mean reward: -321.1816989555913.\n",
            "# Episode: 208, Reward: -123.68938582351767, Mean reward: -318.9596986461021.\n",
            "# Episode: 209, Reward: -125.2938839856628, Mean reward: -318.98407071026065.\n",
            "# Episode: 210, Reward: -728.4823474171959, Mean reward: -296.4836920143822.\n",
            "# Episode: 211, Reward: -124.97789897631857, Mean reward: -293.9977427612576.\n",
            "# Episode: 212, Reward: -119.85795048787519, Mean reward: -296.37847615359493.\n",
            "# Episode: 213, Reward: -0.971242988957709, Mean reward: -293.88173049499716.\n",
            "# Episode: 214, Reward: -376.00070795778646, Mean reward: -291.89508212664003.\n",
            "# Episode: 215, Reward: -124.73794321574529, Mean reward: -291.96561954962874.\n",
            "# Episode: 216, Reward: -122.47885047906468, Mean reward: -291.9218027627321.\n",
            "# Episode: 217, Reward: -587.688620484883, Mean reward: -298.86106687290265.\n",
            "# Episode: 218, Reward: -124.47928281090125, Mean reward: -298.8531900616163.\n",
            "# Episode: 219, Reward: -121.64504522908108, Mean reward: -291.57279496116416.\n",
            "# Episode: 220, Reward: -123.41753496219765, Mean reward: -291.5078383654338.\n",
            "# Episode: 221, Reward: -478.49870728914794, Mean reward: -298.7046341113006.\n",
            "# Episode: 222, Reward: -120.91837788140563, Mean reward: -296.1995710007336.\n",
            "# Episode: 223, Reward: -127.2760616707372, Mean reward: -291.4800709083366.\n",
            "# Episode: 224, Reward: -366.2734563044792, Mean reward: -289.12356817049874.\n",
            "# Episode: 225, Reward: -503.16430298528144, Mean reward: -294.4591888541473.\n",
            "# Episode: 226, Reward: -241.9717213622038, Mean reward: -296.77125363482594.\n",
            "# Episode: 227, Reward: -123.05558841254964, Mean reward: -294.4715258605371.\n",
            "# Episode: 228, Reward: -121.80718490347297, Mean reward: -292.13425359881506.\n",
            "# Episode: 229, Reward: -123.79532577609437, Mean reward: -289.75364670289883.\n",
            "# Episode: 230, Reward: -0.6850578776810112, Mean reward: -280.06681875368434.\n",
            "# Episode: 231, Reward: -471.10478355079016, Mean reward: -257.43924423783983.\n",
            "# Episode: 232, Reward: -124.29877757456948, Mean reward: -252.8500180197967.\n",
            "# Episode: 233, Reward: -367.56708852951965, Mean reward: -255.49948918102865.\n",
            "# Episode: 234, Reward: -357.09788583844744, Mean reward: -249.07908918318427.\n",
            "# Episode: 235, Reward: -473.53633139052744, Mean reward: -253.81456300328654.\n",
            "# Episode: 236, Reward: -474.2203954168461, Mean reward: -253.5542241279409.\n",
            "# Episode: 237, Reward: -124.15082681618107, Mean reward: -253.5842263673299.\n",
            "# Episode: 238, Reward: -235.75908427855242, Mean reward: -255.8190041336766.\n",
            "# Episode: 239, Reward: -125.9571484448382, Mean reward: -255.88752155867343.\n",
            "# Episode: 240, Reward: -368.78659108437006, Mean reward: -260.7581643967583.\n",
            "# Episode: 241, Reward: -0.8092003950439985, Mean reward: -258.39661497939255.\n",
            "# Episode: 242, Reward: -541.8134661029964, Mean reward: -266.73938347127904.\n",
            "# Episode: 243, Reward: -122.09277078577028, Mean reward: -259.15867843622766.\n",
            "# Episode: 244, Reward: -123.80453062400085, Mean reward: -254.2257654435208.\n",
            "# Episode: 245, Reward: -244.80746129818309, Mean reward: -254.2868024729324.\n",
            "# Episode: 246, Reward: -362.04810334426253, Mean reward: -254.2514482939548.\n",
            "# Episode: 247, Reward: -122.05596876655771, Mean reward: -254.2719739832098.\n",
            "# Episode: 248, Reward: -125.37210840676055, Mean reward: -256.76919148121925.\n",
            "# Episode: 249, Reward: -474.1867420017607, Mean reward: -263.72657250385555.\n",
            "# Episode: 250, Reward: -239.80757243639619, Mean reward: -258.81862324640804.\n",
            "# Episode: 251, Reward: -242.61737417057233, Mean reward: -256.23106844921756.\n",
            "# Episode: 252, Reward: -485.84023114353573, Mean reward: -261.1088780674579.\n",
            "# Episode: 253, Reward: -121.19956134226047, Mean reward: -261.1130808499509.\n",
            "# Episode: 254, Reward: -122.1725311107124, Mean reward: -253.6713210540871.\n",
            "# Episode: 255, Reward: -122.01004446138667, Mean reward: -243.7830092639824.\n",
            "# Episode: 256, Reward: -122.23518076405887, Mean reward: -238.87585276033843.\n",
            "# Episode: 257, Reward: -363.13613684553275, Mean reward: -241.39308752413348.\n",
            "# Episode: 258, Reward: -123.96380651125054, Mean reward: -241.39857593788813.\n",
            "# Episode: 259, Reward: -590.1960660907439, Mean reward: -250.69661957998977.\n",
            "# Episode: 260, Reward: -361.6273079238908, Mean reward: -243.35951879012364.\n",
            "# Episode: 261, Reward: -124.78511563576797, Mean reward: -243.35566312331267.\n",
            "# Episode: 262, Reward: -120.851830872494, Mean reward: -243.37554073100506.\n",
            "# Episode: 263, Reward: -365.70910960846214, Mean reward: -250.67029806339514.\n",
            "# Episode: 264, Reward: -598.043352046622, Mean reward: -255.1111509451718.\n",
            "# Episode: 265, Reward: -240.62694313127963, Mean reward: -257.4289309434825.\n",
            "# Episode: 266, Reward: -124.14126245732913, Mean reward: -257.4621791830478.\n",
            "# Episode: 267, Reward: -366.56377000308623, Mean reward: -253.03968217341185.\n",
            "# Episode: 268, Reward: -125.4914809119177, Mean reward: -253.0599261354322.\n",
            "# Episode: 269, Reward: -479.2894882132265, Mean reward: -260.2128149951151.\n",
            "# Episode: 270, Reward: -120.6822045282293, Mean reward: -260.1581083864357.\n",
            "# Episode: 271, Reward: -122.24722647550198, Mean reward: -253.0330787701628.\n",
            "# Episode: 272, Reward: -123.5513554825655, Mean reward: -253.08573832218596.\n",
            "# Episode: 273, Reward: -359.8891182636622, Mean reward: -257.7379994540445.\n",
            "# Episode: 274, Reward: -246.86296773367806, Mean reward: -255.34978968262848.\n",
            "# Episode: 275, Reward: -124.48266052282159, Mean reward: -247.77615683337928.\n",
            "# Episode: 276, Reward: -244.32212199841155, Mean reward: -247.82316484610342.\n",
            "# Episode: 277, Reward: -245.43571584904677, Mean reward: -250.27076739483343.\n",
            "# Episode: 278, Reward: -361.4745719035413, Mean reward: -255.0641151348348.\n",
            "# Episode: 279, Reward: -354.90591119778327, Mean reward: -259.6863268432686.\n",
            "# Episode: 280, Reward: -117.78860747055151, Mean reward: -262.02839783512593.\n",
            "# Episode: 281, Reward: -126.38756989652042, Mean reward: -255.13405356204055.\n",
            "# Episode: 282, Reward: -125.1538419620272, Mean reward: -255.15115484978966.\n",
            "# Episode: 283, Reward: -479.7216582908638, Mean reward: -257.3942462450166.\n",
            "# Episode: 284, Reward: -240.10265221730293, Mean reward: -255.0543415725937.\n",
            "# Episode: 285, Reward: -1.6677042294417748, Mean reward: -245.61696902937197.\n",
            "# Episode: 286, Reward: -123.1305317936327, Mean reward: -238.59517175690772.\n",
            "# Episode: 287, Reward: -123.73483675826444, Mean reward: -238.58685195574938.\n",
            "# Episode: 288, Reward: -120.7551181431495, Mean reward: -236.28677263304132.\n",
            "# Episode: 289, Reward: -238.8284109740542, Mean reward: -238.54419788362569.\n",
            "# Episode: 290, Reward: -246.42512770716246, Mean reward: -236.0969686160815.\n",
            "# Episode: 291, Reward: -364.9531862373078, Mean reward: -243.37984833292674.\n",
            "# Episode: 292, Reward: -125.50550185474584, Mean reward: -235.05368904796177.\n",
            "# Episode: 293, Reward: -238.08844118645425, Mean reward: -237.37360245597546.\n",
            "# Episode: 294, Reward: -369.49941348024845, Mean reward: -242.2875001131004.\n",
            "# Episode: 295, Reward: -506.84745463890573, Mean reward: -247.5282999799149.\n",
            "# Episode: 296, Reward: -502.7416480484216, Mean reward: -250.34217087399801.\n",
            "# Episode: 297, Reward: -518.3245393486237, Mean reward: -258.2675422856393.\n",
            "# Episode: 298, Reward: -126.61400713309797, Mean reward: -258.2923802601661.\n",
            "# Episode: 299, Reward: -238.47390221545294, Mean reward: -253.57812346443993.\n",
            "Saving models ...\n"
          ]
        }
      ],
      "source": [
        "agent, scores, avg_history = train()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "qzM6DahtEVvT",
        "outputId": "f083766c-a95c-48bb-9ad8-6e8ba2d52f4c"
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlAAAAGwCAYAAABmTltaAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAC74ElEQVR4nOydd7gU1fnHP7t7O+XSi4gUUQQRRRAEC6goWEOMaGxBjVh+mkQxGk0ixkrsmmgkRrEXbDEaK2IXFAWJqChYEAQu0i/l9ru/Pw6ze3Z2ZnZmd7be9/M899ndKWfOzsyd893v+55zAuFwOIwgCIIgCILgmmC2KyAIgiAIgpBviIASBEEQBEHwiAgoQRAEQRAEj4iAEgRBEARB8IgIKEEQBEEQBI+IgBIEQRAEQfCICChBEARBEASPFGW7AoVIc3Mzq1atok2bNgQCgWxXRxAEQRAEF4TDYbZs2cJOO+1EMOjsMYmASgOrVq2iZ8+e2a6GIAiCIAhJsGLFCnbeeWfHbURApYE2bdoA6gK0bds2y7URBEEQBMEN1dXV9OzZM9KOOyECKg0YYbu2bduKgBIEQRCEPMNN+o0kkQuCIAiCIHhEBJQgCIIgCIJHREAJgiAIgiB4RASUIAiCIAiCR0RACYIgCIIgeEQElCAIgiAIgkdEQAmCIAiCIHhEBJQgCIIgCIJHREAJgiAIgiB4RASUIAiCIAiCR0RA2XD33XfTu3dvysrKGDFiBPPmzct2lQRBEARByBFEQFkwc+ZMpkyZwlVXXcWCBQvYe++9GTduHD/99FO2qyYIgiAIQg4gAsqC2267jcmTJ3PmmWcycOBApk+fTkVFBTNmzMh21YQWQn09bN0KjY3x68Jh2LIFtm/PfL0Ed4TD2a5B9sn3c5Dv9RfSjwgoE/X19cyfP5+xY8dGlgWDQcaOHcvcuXMt96mrq6O6ujrmTxCSZf16uPlmuOUWuPVWMN9Or7yilt90k3qfCo2NMGMGvPZaauXohMPw3HPwwgvO2736qtqm0BqquXPhttvUdXSisRGmT4cXX8xMvTLJCy/AHXdAba2/5dbWwj33wNtv+1uumUWL1P/YihXJ7R8Ow+OPwyOPFN797ZWlS+Hhh2Hz5mzXxH9EQJlYt24dTU1NdO3aNWZ5165dqaqqstxn2rRpVFZWRv569uyZiaoKBcrq1dDQoN7X1ID5tvvhh+j7ZctSO9ZPP8Hy5bBgQWrl6NTWwmefqTLr6qy3aW6GDz9U22zY4N+xc4HXXlMO4ezZztutW6eu7aJFmalXJvn6a9Vg2jwyk2buXFizJv0C6tlnlQP87LPJ7d/YCEuWwLff2v8PtBQWLIDvvoPFi7NdE/8RAeUDV1xxBZs3b478rUj2Z4sgEB+2M3+uqYm+T/XXrVFWXZ0SNX5giD+wdyD0Y+nfp5AIJni6Gte1vh6amtJfn0wRDkevqd8OVKbD1oFAcvvp19Ov/6t8pb5evRbi/3lRtiuQa3Tq1IlQKMSaNWtilq9Zs4Zu3bpZ7lNaWkppaWkmqie0AHQBAvECys9GSS+rrg7Ky1MvU69/TQ1UVsZvozcqfjeyuUJJifN6/brW1kKrVumtT6aor49eX7+vrdEYZ4pkH+sioKIY93kh/p+LA2WipKSEoUOHMlvz35ubm5k9ezYjR47MYs2EloJZMOmCpKkpthHxy4Eyv08Fvf52Zer1LtQQR6LG1yw0C4V03FMGmRZQiUSwHSKgoujpCIWGOFAWTJkyhUmTJjFs2DCGDx/OHXfcwbZt2zjzzDOzXTWhBeAUwjP/ivNTQPn1C9GNMNAblUISUPq18upAFQr6d8l3B0oEVOoY97kIqBbCSSedxNq1a5k6dSpVVVXss88+vPrqq3GJ5YKQDpxCeOno1WSQDgfKTQ5UIQko/RwWFztv68apy0fy3YHSf5RICC91REC1QC688EIuvPDCbFdDaIE4OVDmh1AuOlBeQ3iF9GDVz2Gia1OoDlQ67imDTAgoXdCLA5U6xg/CQrrHDSQHShByDLMD5aZXW7Kkw4HyGsIrJAGlf5dEDafkQHnHz/w/O/T/iUQ9Ke0QARWlkB0oEVCCkGMYDxyjC7WVoxMKqddcd6DchPAK6ZepFwElDpR3dAGVLmHi5RraIQIqip5EXmiDioqAEoQcw2hYy8piP0O0QaqoUK+5KKDcOCstIYSXaGwnyYHyTqYFVLLjc+l1a8kCKhyO3ufNzZnvBJBuREAJQo5hCBBjTCarEJ4f4zXp5UFmhzGQEF56Q7PZJF298MLh+CE90oFeZ3GgUiOdY9jlAiKgBCHHcHKgjAbaEFC57kC1tJHIkw3hFeo58DNsY3Yv0iWgJITnH06zKBQCIqAEIccwHjqGSLIK4fkhoJqbY3sciQOVOl7ci5aQA9XUFN+I+lGuUXY6EAHlH+YOMYX0vw4ioAQh53AK4ZkdqFQwN9rZGsagtrZwkku95M+0BAfK6rNf5WYihJfsMURAKcSBEgQho2TKgTILpnQMY1BXZ11HvVEJhwvHgZEcqPj7yK/vJg5U/mF2oArpPgcRUIKQcxgPHadeeH44UOlq6PT62okjc6NSKL9Mkw3hFcr3h+h3McZQSpcDJcMY5D7iQAmCkFHMDpRTCC8VB8ooq3Vr9epXKM1N3oP5OIXyyzTZJPL6+vQ5KplEn+y6XTv1mm8OlITw/EMElCAIGSXTITyjoQuH/ZmXzk3X5ZbgQCVqfAsxvGF8h0Agel/lWw6UOFD+IUnkgiBkFHMSuT4QXTpCeG3aQNGOWTH9aMTdPDQLUUCFw8k7UFAYAsr4/qWl0cFe89mBSlb8yECaikK8x3VEQAlCDqGP3GvkQBmCRHeH/HSgysqi5fkhZNzY9uZ6F4KAamjw5jwUYnhDDzEb9684UKnVJ58RB0oQBE80NsL27cnt29QUFRdmB8p4+BQX+zMXnlVj5/YXYkMDfPeddSNmnsuvthaqq2H2bFizRi1z40CZv9umTbB6tbv6OWEe0dovzOcuGQcqHIZt25z3W7tWncutW73XMRW2bIG334ZVq+y30e8p4/61uqe2bvUuLLwIqKam+J6emzc7l2/8eMm1HKjmZlV3L3UJh2H9em/7bNyo9vFa3+ZmeP55mDs3/n/WuMfNHQo2b/YnXSDbFGW7AoKQb9TVwSefwJ57qs8vvQQjR0Lfvurzo4/C8uUwfLjKA2nVCvbaS60Lh2HFCqiqUg+TkhJVTtu2ar3eqOoCSu/NVl4eFSdWrF2rwnKGKNJZtgw+/RTGjYs+zNw4UNu2qTBfaan6/N578O67cMwxsPfe8MUX6rg9ekTFSevWqtGtqVHn67331N/hh0PnzrHlL1igtquogH33VQ/ymTNh/PjoubvjDvU6ZUr0fIE6L7W16lw3NKj3bdrA+++r73vyyVHBCfDkk/DNN3Dxxeq8btwIu+wSW585c2DRIjj9dHV+amujIanGRvjPf2C33VTd3n8fevaMD6uaG6KqKvj8czjwQFWmcZ5KS9U9VVOjypo9G7p3h0MPVcf4+mtV9i67wMqV8K9/qf1KSuCgg+KvVU2Nuj+srr8b1q6F556LCvUNG1R9fvxRCZ9334UxY9Sxzfehfo/aOVCrV8M//wlDh8Kxx8Yf/5tv4LPPoGNHdW9VVqr75ptvYrdraoJ58+DLL9X5O/VU9cPlzTdh6VJV9112gQkTVOP+wQcweLA6ZnFxtJxwGJ59Fr791rqn36ZNsHCh+n827gGdBQvUsfbeOyrSzAJq5kz1TNhpJzjiiOj9//jj6kfF3nvDqFHW1+zhh9V9DFEhUlQERx0Fe+wB77yj7sOddlLrVqyAV15RQtd8jhsb1fXbfXf1v/r999C1qzpvd9+ttikuVvU78EAYOFAtW7NGnZ8RI2L/l0Bdz4UL1fvNm6FLF3XeO3WKfRZUV8O6deo4a9fCzjvD2WfHlrVqlfp/6Ngx/jwY1Ner+06/htlCBJQgeOSzz2DWLNXId+6sHtbFxVEBZTzsPvwwuk+PHtChg3pgPfxwbHk//ggTJ6r3untjiBVjuR5yMzD/4tu4UT2g+vSBSZPi6/7hh/DVV9C7d2x5Tg5UbS3cfLN6cF55pVq2fr163bBBNfrvvKM+FxVFxY0hoGprYx2fWbPgF7+Ifi4qUo3U3Lnqc02N+u7btqlG0xBQBuvXxwqohx9WD/gpU1RDtWqVEkfz5qmHdlWVOv8GX3+tXhctgtdeU+/PPBN69Ypu8/rr6nXePNXwffcdnHuuEhLz5ql9Fy1SD3pD8Bx1VGw9zb/+p09Xr42NShga17pNGyUAamujDt3q1UrEnHMOPPGEWvbnP6sG16C6mjiamuAf/1D3z0UXRRtcL3z9dbzTZwiDVq3UdXnzTXVd2reP3c7ogVdSYu9AVVWp17VrrY//5ptRl+uHH5RYe/NN9blfPyXiqqrU/83770f3+/ZbdW9/9ZX63NSk7p+lS5XwBPW/29gIJ54Y3W/dOiVsrWhuVvflRx+p7zRqVOz6mhp44QX1vn9/JQxBCSKDxkZYvFi9X7pUCf2jj1b/u0uWqOXvvqv+R8aNi6+D7vgZory+Plrm3Lnq+hjf6YUXoud23brYspYuVcf68Uc45BD1v7PHHrDPPtFtGhrUMd97LyqgXn9dnd+OHdX31NGn2DGeed26wXnnRe/xrl3VfdPQEK3bjz+qc6A71fffr+6xKVPizwOoa3rTTWqfP/7R+YdkJhABJQgeMRquLVuiQwDov1wNR6FdOyUMINqIGPu2aaN+oX3/fWy4zxAaRUXRxG6IDQvqDpRZQBnlWzWuevmNjdEGvrhYNQ76eh3jIWyERYLB6PepqVHnQa+n8Vl30MzhKuOhu8sucMopqjH43/+U+Ny2LforV0+gNzD/AjZCFRs2qIa/sVGdd+O72IUx9F+wP/wQK6B0vvtOvc6frxw3/dwax2hoiL8WdqEQQ3xa9bbU61pTExVUoM6LHt6zmtl+27bo+a+pUY2RV4x6lZbCkUcqB2jZMvX9DjoI7rxTiRir4+v3r3F+7XK97K6L/h23b4+Gerp0US6T8QPEHMLUpyYaP15dtyVL1PH1e+b7763r064dnHWWuocaGuCZZ2LLtHJn9e+wbl30f1Tf1vw/ZWxj/v52YX/jPjr3XPW8+fJL5TDpP6r0a6Ef23zujXNWWxsVxdXV0brssgvsvz889VRsmcY9pf+v230//TjGurZt4YIL1P1cXw///ne03sYPxW3bVD2qq1W9iyzUybZt0e9UW+vfpOrJIjlQguAR46FVWxt9WFlN4XHmmarx0TEa2e7dlb1uxng4FBWph77hIDQ2Ro9RUWH/y8so3y43yqhbU1P0fSgUbWCsGjVdaBh1sDoHBla9CM0PcuM4RqhpyBAYMCB+e12gGJgFlLFtdXXs9sZy/Tvp9TBEI8Q6Nfq507cxizp923DYvYAyD5BqNCD6NTHQ83bc9GjSG+Fk8/CMeu+zj/rr00e5FYceGpt/55T/5rSdUW87AaXXOxyO1qe4WN0vRrnmHJrm5ui2bdrYn1e7xOZWrVRD37+/+nFj3tdqTj/9GhvCWP+OVvsZxzPXw+p8hMPR5W3aRP+M/a1+YOjHszt2Y2PsDwz9x5Qx/ITVQK9WItLYrlcvuPDC2O+iP886dFD/43vvHX2m2AlNN/duLiSki4ASBI/oDxNdSBjowsTsFBmvgYC1i2Q8RIwHjPErrKEh1oEyMDfaiZLKkxFQepnmB2lNjf2DTB8I1PwgNyeXQqxjYaw3vxr1NWhujtZZb8DMDYS5/vrxzPXQH+RWIitVAVVeHpvkbAiqRALK7PhYnXd9WSodGSBeqBq4EVDGDwCr7ZwcKP26gTpHxnk1zr+TgDLKDAZjj28WGFb3tP5/ZRxLL9PKadG/g37/6XWzE2x2Pyp09PvK+D7GM0E/V/r3049nd2zzDwz9muvPHPN+Vvec7jqar7n5eWZglXPpdRLyZO9vPxEBJQge0V0Y84PF6oEPzsJGX6c3QPqrWwcq0fGMB5AXAWUlQPQQnt3DTnda3Ago/cFtFk76w1z/bnYNmF1PKrsxfnSxoH8ffbmVgHLCridWeXlsGU4OlBEChngBlciBSvYXeioCyqkxNTDqbSUwrZK4je2Me95OQOlCySygzHWwaqx1AWXs29xsLeKtyvEqoNw4UPoyo06GGLFyWXXBZ1VnXbzZCShz6FXf1uqec3Idzc8zg0QCyk4c2f0YyhYioATBI7rrZLyvq4t/eOkOlIGVA6WjP4z0V7c5UOkI4enLtm+P7RGoi0hzuNIphGce6gCsf1nrOVsGdr+23Qgoq1ArxAo5u3GAjG30/dw6UHp9vAgorw6UHyG8RALK6jwYuAnhOTlQ5jrrAsqNA2Vsa76nvQooKwfKSkC5caCcRIxdWVbL3DhQ5jKdxJtVCM/sQJkHh/XiQOnDhbhxoKwcLzPmZ1G2EQElCB7RH4B6Imttrb2AsgrhGViF8MwOVENDrANlta/VZzNWAioYjG0w7PYBVYf6+uh227dHxUGHDrH76QLK/CB3cqCsQnh2DpSdA2AnlOyElZ2AstommRCe/rAPBu0bHXMjumFD9L0hoPReS2YSNXZuyJQDZbW/kwOVSggvWQGl7+slhKdfG/N+dXVqP7v/CbvyjTpZhbqNc5SoTC8OlJF/luieshLNRp3EgRIEIQb9H1d3CGpq7MNCbnOg3ITwEo0DZS5Tx8rq9xrCM/9qNI5lFlD6SOp2v7aTyYGyS5jV6+WXA5VIQDmh11MvUx9t3k5AWfXuNASUsc488jlkxoHSw1tm/HKg9EbcbQ6UlxCe0wTd+jG8OFB6mVYOlD78SG2tNwdK/zFmFep2cqCs8r3sksh1B8rYX/9fcutAGeWaHXUD41xb9UC2Ow5IDpQg5DV6+Mr4bKA7UMGgdZguUQjPbHlbhfDcOFDpCuHpifM6JSXRht0o00jATiaE59QLz86B0rETQXa/eNMdwjP3LLMTGkbZRk8rHUNA6dff3NBkOwfKTRK5GwfKGH7BKgfKuA7mkGY6Qnj6tUrkQNmVrwsMfXBRrwLKwCqEZ3bJ9P8ru7CXfh3MzwJjf/2Hm1FvM3YOlO6yiQMlCAJ1dfbipKYm/oHnNYTn1YHyGsLzI4ncSkDpU3cY9bYSRAZek8jdOFA6bpLI9QbYjQPldhgDK6Flftjr19n43lYOlI5R3+LiaN6U+Vpk0oFKJoSn59R4FVBmB8qMedtUHCirXpmJHCg79HtdFw5eksjtnFq7HCj9uxjb2OUzmZ8FgYB16oC+j45+zY0fjka5iRwoyYEShBaE01xxugNlfsg7hfB0zA8c40FWXx89diq98HQHyu7Xut0+oB5aVg83s4AqLnYWUFYNg9Uv60Q5UHZz2rlxoHQBZdcLL5kQnlWIy+xAWQmNxkZ3AioUsh/lO9s5UIlCeA0NsSLejHGekhVQfudAGXU272PgRkAZ+4dCsSP+m39EeHWgjHIgPgeqtDTWRYLY3EWI7TlsbGPu6dfYGJ/PZfdjyNhHP+/JOlBuQnjiQAlCnuH0T+vkQBm4HQfKeOAYD6UtW+InGTbvq3/OdAivosLegXJ66Ornxy4HSg+jgH8OlC6+7Bwoq20SOVBuBJTbHCgdXUDZzTOX7V54iRwocy6Y3aTShoByyoGyqnciB0rPyzMf0yoHCqLn3UsITyeRA6WPA2ZXvl4fq4FtzSG84uL44QjM94rVfW7u6Wd2oKzKMT+z9POerAPlJoQnDpQg5BlOAsrKgfIrhGdMH1JWFmuTm0l2IE0vvfDcOlBWoQY9LwoS98Iz7w/+5UBZTUXitE0qAsouhGeXA+VWQDmF8FJ1oOzm0UslB8pcXzsBZXx/p3GgDIz7zE0OlBH61JOv9QmQDdLlQOnCwShPH8bCrnzzeGRG/XQXSa+j7gAbxzffD1bhbCsHKpGAyqQDJTlQgpDHOIXwrBwoA7Ow8RrCM88vZ5cDZXc8g2zkQOkOlBsBFQ7H92Lyy4GyE0dWPZWMuhs4OS86VufSTQjPi4CyCuE1NcWeN2PMLq+4DeG57YWnO03mRs98LpMJ4RkCxC6Ep4dGzQ6Ufr70XnL6/6dVHp5d/a1IxYHSBaGOeSoUcwjPai5CJyFkFlBODpT5/9+NA+VGQIkDJQgFTrIOlIFVg+ZmHCjDgTJPnuklhKc3RskKKH3cJx2nHCh9f7OAsuqFB/GhNq8OlNckcjsBlaoDZSwzP+wThfCceuHZhfCsxlCyc9mc8CuJXBfHxrbme8dchpskcrMzZtxTdiE8/d4xO1DG8UpL47+v+TjJhvD8dqAg9scGJOdA6d/HyYFyyrMzH1MvQ3egnEJ4xncwO1B2zzC9/m6HFEkXIqAEwQNuc6CMh6/XufDMDyPj1RBQRhd2uxCgk+OgP3zMCdRuBVR9vZoR3djPwMmB0tEbO4htpPTu0zpODpRdErld/e1+8eokCvNZPbStcnWM5eYy7YYxaGyMfjcnB0p3MvTGTR+p3rhvkvmVbud6GHhNIte39epA6e6VGwfKKoSnX0NjW7OoMP8wsTqOkY/nVH8r7Bwo8/hQyQgo83ZeHCgd/d7Sy/crB8rOgdJFlvlHktUPNbv7JVuIgBIEDxj/1Poks8ZD2U0Iz+04UGYHyhAt5hCeHVZCytwbysCLgILo6Njt20eXlZerh69xLvQHuEEgEP9Q18WG3n1aJ5kcKKv6mx/KXkN4xjbmbvlmzKMxg7skcr0+Vj0tE+VAGfXWE/qTaWD8cqCsBJSTA6V3szcElL6NXQ5UIgdKP6/6/IzgLKCscsDcjN1kxs6B0nvMGWXZCTS7EJ6BOYRn5UA53Qt2DpQuoNq2jdZdxyySdOfPqL+5vrrA1gf31LESfHaOZbYQASUIHjD+YXXxYLz3EsLzmkRuoA+iaLW/UwjP7CQZ6EnkqQgo/bW4ODbZFeK77EN8I2UloMxWvZscKKv6m50sKwFlFmuJQmBOITyjrubxd8BaaOjHKiqKb9StusPr5eoDrRr3STINTLK98MzCMBCI39bJgdK72esCynyv2DlQ5oFRzec1EIgPIWdCQOnOpJMDZTUYq1sHyrjP9POfjANlzoHSQ3jGTAN297I5hKfnl5nrGwjE50GZz62V4DPn3YkDJQh5hJOA8sOBsgvhGSRyoJIJ4QUC3hwo48HoJKCsBKCeF2OVA2Xe3sAvB8r84LcSR055Ik7n1k5ANTWp45idFqtQlznfytyoJ0oi10N42XCg9Jwvq3wYc32NfQyMuhYVxTq85nCvnQOl3wtWIbxQyD4vKJ0CSq+TkwNlVZ5bBwpiOye4yYHSceNA6c85HTsHShdQVvU1Cyi7iY91xIEShDzGaAD8cKDcjAPl1YFyWmblQBl5R049q+waiUQOFCQWUG4cKL9yoOx6D4F1rhLYh/mclpkdKLOISRTCM5w7JwGVKITnxoGqrYV586Lh4epqVbdke+Hp59MqH8aqPlb5aRUVsfeFWwfK3GPSSkDZuTJucqDMxzDXPxGJksitynPrQBnbWiWRp+JAbd8evc5eHSjj3ozLbWxshHnzKN+wEpqaqHl/Pnz3HY0NYfjsM5g/HzZtshT/uZYDZXEZ8pNly5Zx7bXX8uabb1JVVcVOO+3Eaaedxp/+9CdKdvxEWbZsGX369Inbd+7cuey///6Rz08//TRXXnkly5YtY7fdduPGG2/kqKOOyth3EXIXJweqri5+QLpkx4GyEiBg7UAlCuH99BOsXQtdu0aXmR+YekNXVaWeYWPGqFBKMgLKPBCoscwc0jELKKtfqm564RUXq21KSuKdpaYmeOUV+Oij2OVW4sjJgbLCKoRnhK6MpGarZFinASeNz2axnKgXnh7CM7Z9801YuhQOPBB6944t74UX4MsvYdEi2H9/ePppOPxw7w5UOAyrV8fem1b3FTjnQOlixqoHXyIHyk5A6d8n3Q5U69awdWv8fkZ5iRyorVvhnXeUWBk40OJahMPwzDMU3fwNFA+Enj3Vzb/77jQ3l9JQ1wzNUBQIUxwCCMV916Ii+4FtzQ6U0XklFIr2DPXqQBUXA99/D+ecA0uWwMaNsGUL5fwSAgOoDr8MFYtoGHAnzP+RcmqooZyarwPw4HkxD65cc6AKRkB99dVXNDc3889//pN+/frx+eefM3nyZLZt28Ytt9wSs+0bb7zBnnvuGfncsWPHyPs5c+Zw8sknM23aNI455hgef/xxJkyYwIIFCxg0aFDGvo+Qmxj/sMavMYB27aLvjUbMryRys6Do0iW6vxVmcTFzpmpAAQ47LLrO3NNKb2zmzoX//Q86d4bhw70JKKPRNxo1XQAaeVFG3ay+h1cHyljevz+sWgW77RYvlJqa4pcZy82YG3izyHIad0tfpwsoq6EmrBwoA+OzuSeefs2sQnhGXUtKoHt3JYLr6uCbb9Tfz34GQ4ZEt//yS/W6YgX066fe//STdwG1dCk8/rg698Z3Mid8e3Gg9LkeIf5eMQsbqxwoXcTp9bZzoPQxoAy8CKg991T1btcO3ngjfj/j+Hq+kyG0jDBuUxN8+qn6/wMlfkeNiu5LYyP88pfw7LMUMxGog4Wfqg06daZ553IaL38TtvajmDcoCnSCfc6i8aAjgeLId23TRmkYuzqCxfhzpU1UfLcYVrZnZbgDr7xSzmGHRX+4GN9DL8O4N4tWLYcRI9SvOIP27em6bRNf1zfzRtF4OmzfQOP8hUAn2vRsT82KOmpefRv2uAOuvx5+/WsoLc25HKiCEVDjx49n/Pjxkc99+/bl66+/5p577okTUB07dqRbt26W5dx5552MHz+eSy+9FIBrr72WWbNmcddddzF9+nTLferq6qjTAr7VhmwXCg7joVBZqR7S4bASDaWlqrEyHop2ITwDt0nk+q/To46KdZGs9teXVVVFxRNEf03qWAko41Y2Xo1Gon9/+PZbVcdQSAms3r2VS2XUd9991XH22iv2exjvzSOeuw3hJXKguneHE05QAsBKQOkPeiusxqLR62lsozf4xvU3Y06ettrGrrs/RD8fdJC6z6qqYPHi2PV6CK+5WR1Pv3f22gv69oXNm1WD/t13yiXSBZSOnmjvVUCtW6deq6qixzdva5xHJwfKWFdWFv2BoZ9ztw6UOSyt18XsQJlHyNeJ3JsNDerG6tiRhml3wdO3wGmnwTnn0LS8FawrpXNZKWOOacPXX8eXo5dX3FxHcRgaGgJsWR+G+gDFoSJCoSBNTbHuVU1NVOgEg8CUKfDss1BaStFRx8GqTrB+vfrlsG4tTRfeQgOqDSyikeJwHXy6gIYpb8NT59LU1B9QQm/jqhp1wQIB6NEjon5C9TXw9kcUf14JtQPYsqVMuUX33UWnlTcT4DfUBkJ81KsXu1w1kP6n7Repr6UDtW4dxQ/cA9vXqpvvb39TKnPAAA7a3sCa+zbx9fpOvDh7EMH5H8PYo2j9y+H89O6P1L29Fr59Gy64AK64Qp3vsX8FQgwcEObgba/Qdr9DAQv1myEKRkBZsXnzZjroVsEOjjvuOGpra9l999257LLLOO644yLr5s6dy5QpU2K2HzduHM8//7ztcaZNm8bVV1/tW72F3EXvltu2rRILbduqB3BdXWzcH+xDeFbrID6E16ePcoF691aWvnlfM/pxzL/WrJKmjXrqDb555GXjO/ftCxMmqF/I7dqpfc44I7a87t3h5JOjn80Cytyo+dULzy7kadTfOF9HHaWO+d//xm5jNUSBFeY8H7sQni4crBwoXRTYCaj27WH0aHjppfj1esiprk59NovvVq3UX58+SkA5JdzrAzF67YVn3PO6o2L+LmYHynDo7ASUsY1+P7rthWe3nZUDZTdOUWT/L7+EV1+FLdVAgEYeBtbArbfCrbfSxLHAvoTufhP22UJR8R7A6XDwwSq5rKREXcjt2wlNfwAm/o7ypt/SQFu20gwEKfr7i4R2vQRGjaWutjMsXwFffQWhEHU9hkK4ktBTT8DMv6uKPf44xcXHw/wdFV2zBh54gOa6IA3jj4PBEykeP4WiOV/BNbNp/OwL2HNPmna/B7r1p7R0Fbz5PdQbvUE6wKmnQnMzofE/hx8/oIgDoOxYqgfvAYu/pXzLUjq0quc3bZ7hP1XD+WFZM1vP/AcN/+4Dvf4B7dtTfMtf4YH7CG0+AtrsT+0HDTB3E0Xbq9VDbPbsGFu1uE2In53bjZtugi3jTqDs2J9DfUhtsvPO1N50J6zsD3/9qxKJt9xC09y20P142j/zL7p9cSes/Sv84Q/xFy9DFKyA+uabb/j73/8e4z61bt2aW2+9lQMOOIBgMMizzz7LhAkTeP755yMiqqqqiq6mn/ldu3alyvh5ZcEVV1wRI7qqq6vp2bOnz99IyAX0xvCUU1Sj0aZN9AFsFlDm/ZIZB8oq/S6ZgTT1XjEGVg6UnYAyGu5DD7U/hhm7HCgD8zmwyoFycqAShTzN+w8YoJ7FdujzrjmF3iB+wExzCM8oz6ocK7Fl4OZzKBTN96qpiRVQ5nNgdl3MGOEjiL1H3DpQxj1vfCcrB8pw4oxtKyrU/44uhs1z0hkCyqsD5SSgzOfCSUAF165RiWGEIVQETY00lFfCFdfAc8/B0qU0hdtCczmh2kZYuJBi1gP94eOPo/sdfji89x7BbW8DTZRTQzVtad7Rh6t46wZC/1sAXy6jdlYYvoleqNpPZkLrvQmte0UtuPlmOP54il7RKtq1K5x3Hs1jD6Rx44HwLRR1KKf4gOEwuS+NHzfAvNdpWvw1LF5JKV8Be0DbSnUiNm6Au+5S54hl0LEjxcH2sLaG+nkqRFjepS3M+R8ddt2VzvdV8cP0T6hZ8C6NL7wE/J1A+3YEN/5xRxlrYP331C37DuhLUZcO6heLxeiweui0tl5dMCPXqrY+BL/5jXKgnnoKJk2i+YM5QBMhPlM3kd1NmiFyXkBdfvnl3HjjjY7bLF68mD322CPyeeXKlYwfP56JEycyefLkyPJOnTrFCJ399tuPVatWcfPNN8e4UF4pLS2lVI+1CAVPIKCeW4bWNhotowHyaxwoN9iF8MzL3Qio5mZnAeUVuxCeQaq98BKNmwWqjdAbd6vvYXagDEfRvI15GpdEITyrXo3mbZMRUKCEhiGgwH7iVrPrYqakJFpP/Tu7nUzYnNdkJ6B0d8tItrbKgdIdKL3edgNpmnvhme9pvS7mnmmOAurDOUAYevdRDs3GjTQefz4c1hWuvFJ9r2eAzyG036+h8QOKljfCtPWwehUQgKZGeFWpnVDPHvDoO5QvHgE/RB8ARcOPJXTpR/DNT9R+UwWhnZXdvG4ddat/hNq1hAJhmP5PlYiNxQ+Fdu1oHnUgDc8TWV9UBHTqRMMVU6Hf8TRdtg5Wb6asYydoN0Ylbm3bphIlf1yh6jhiP3jxXoq+aw83LFK/Nnr2pOLE3WBXpWzL+3SDY46h5uy9aHj69/BmmOKNawgUF8MddxDaNhbe2kRt/TLoMIjiQ3aFztbtYzCorrce2jV0VmRZMKhyvzp3pun3b8O6noT2OQrueRh23tmy3EyR8wLqkksu4QxznMBE3759I+9XrVrFIYccwqhRo7j33nsTlj9ixAhmzZoV+dytWzfWrFkTs82aNWtsc6aEloWVAAJ7B8rNVC562XZzR5lxk0SejIDSHSijLtkWUG564TkJKP3hbJVzpJeph2itzpe5Hk698MAfB8p8jnQBtXlz/ECE5nNgFg1mSkutBZTbYQzMeU12ITw9fGc1fYlVCE/fxspZ0kevT+RU6SE8cw5U3P9bczOhDz8AWsOwYeognTvT2DF2s8j/RtdOMOxnFK8F1oVV8lKrVsqpWvI17NKL4O2XwsGtKFsFaDlXxfsPJfR/+8Bzb1Pb8BPsf6SKkYfD1G05DNY0Exo7Gc7sEdnHzTAGMcJ50CCaDgTqoXQEYOQJtm4NZ50VyRYP/aENdA5QvBqVs7Qjaa5cy4SJDJHRpReNTzwNN2yiaNU3cMVEGDKE0H+BbVDXcTishyJtUFQrystj76GIA2XuvXrYYTT95TD4FIKHAdnVTkAeCKjOnTvTuXNnV9uuXLmSQw45hKFDh/LAAw8QtPsJpbFw4UK6d+8e+Txy5Ehmz57NRRddFFk2a9YsRo4c6bnuQuFhFyIzO1DmW89JQBnL9MY2kQPlZRgDA7cOlDm84ZeA0nvhGfjVC88pB8r43oZYcfoeugNlxuxAmddZhfDsksj1Zak4UBAvoNyE8PTjl5RYh/DcOlDmhs7OgdIFktVE02YBZdwbTrlNVoLYTQgvoQP1zjsEN6yF0k6q98QO7MaBium9FghEu+medJLK3u/WjVBbtZF5yISiIgiVhOCww6grBwxHLxCgbuddoS2Edonfx4z+v6vPRWn+IRTX4zAQiMzTEnIx/pw+SGtDA9CuHcW9hsGQ2HNhXM9EPwbLy2N7BcY5UBqpPIvSQc4LKLesXLmSMWPG0KtXL2655RbWal0mDffooYceoqSkhCE7VPVzzz3HjBkzuO+++yLb/u53v2P06NHceuutHH300Tz55JN88sknrtwsoeVg1/CnEsLTG+dEDx2rMs3LzMudksj1B5LxHfwQUIlyoNyMA+WmF55x/g1HQj+XxvcxTy/CqlUqhFFbS3j9J1A8jOZm1X2wePUP8NlPqlvhpk2wejXhI/akoSF64ZxCeGaXRsfckcDOYUr02W4kZzchPF0oWTlQRk82K5IVUFbjPCXjQJmnB7I7f24cKEsBVVsLf/oTQXZVYS5tZaLxk+KETTCoerlp9bYUUHrPNVNVrL6L1f+JHn43O1B6xwWnbBPjOHYzIOjv9cFAra65/n/nhPl8OAmoRBNdZ5qCEVCzZs3im2++4ZtvvmFnU1w0rD2xrr32Wn744QeKiorYY489mDlzJieccEJk/ahRo3j88cf585//zB//+Ed22203nn/+eRkDSgDsQ3jGAyeVEJ55gl8n/ArhWTVKfgoov0J4iUYiNx/HnKukbxMKoQZGeuzR6Db8F+7/kqbLX4ZFHSn5z13Q1F2NjLxmDTTUw8hqGieMiinXTQjPSkSb7wM9kTtVB8ouhKffX3rekpGsrZfhdK3teuEZ2IXwdIFkFmF6Oa5yoGprYd06ilo1EvrfBtjYJzIwWTAILFlC6JvvYOVesNNOEAgoB+qL/8Gb62msWQuj+tDYODx6jurrYc4c1cNu7lyC5XtGB2LaQSIHyumHj/m66efLTnCbB7w1sAvh2TlQerluBJQbB0ofDNTqmts5omb081FUFL3+dXXq/0T//xEHKk2cccYZCXOlJk2axKRJkxKWNXHiRCZOnOhTzYRCwi6EZzxw7H4huQnhmSdiTaZOTiE8tw5UukJ4Vg99NyE8cwOdKOm+uNj612vMQH8LF6oPbStVvknrPeHL+TRddTUwmmIjjrIjuRYg/NFHNBwd26DqdbLrhWcVCjNfn3QKKCsHShdQTU3xDbfTtU42iVx3oKwElHlU8GAQ2L6dxjnvwVcrCb73NQwPEnpzHnw8GsLNFLGB4N+eAs6DUQfAYYcRfPNN+O3hBIEAUwnvPQT23ZfQy89Q/PH1wG9opBZGnkLj4MegSy+K5t0D7z8fHTmyqIjQlN9BUWzSk91AmrYOlIYbB8rAPG6ZWwdKFzT6ddfPs9Wgoebj+OFAGXhxoIqLYwcbrauLra/dDAbZomAElCCkG6sxnAzMD4lUQnhuwnduBJbTkAYGRj2Dwfiu++kQUOaG2smBMsYKMjfQTr3wzO+tyg7VbiMy4uGJJ0KPHoTHnQ6PtKL54ZXQVE7JIWOh+yHw4ovqJ/uPKwjP/9RzL7ymJvt8OGNbiD23mciB0gWmeTwmq2NarTOSls2iws6BMrbTHSjjWupDHMTkQM2cSdPyxUAFwRWz4M1PCVEGHASt21Bc2kSwYmdYAcz5AD76iGDTSlXAnnsS+jJM4/8Wwv8WEuILigLNsMdAGoJh+KKZxoWLgK8o4gVgixrqf9w4mDyZ4OpB8GXsd0skoIzQp9V94eRAWU3Z5CSg7HKgrObCa2iIvb7JOFBWAqq2NvqjzOqaG3h1oIy/xkZ1DF1ASQhPEAqAROMX2YXwnMqwy2FJdHy7sF2iLvR6PY33egPhdw6U1WjgTjlQZWUqUdUsoJwGHgUXAmr269DYAO3aq/AOEA4VwT/+QdMEYA4U7wUsAv7v/1Se1C03w3ff0vjTBiDaJcnqfLsZSNMspP0QUHb3j1UvPD8cKLt5/uwcKCM0XF4evaeMVz3MFAnh/VQFy3+gMVgCRx1LsEc3qJ9HaMRB8NXxUFlJUQ8I/vxaOHeRmtyvsYFQELj5Vrj4YkJnLqHxiedUCG/4KIpv/Qu8tCdNQPP0njT+YSusW0/RL/8EPzsM9tknclMGn4n/bolCeEYOntWYW1YCypg42ny+y8tjZw5wI6B0p0kP4enLAwHrDhLm45ifZ1YhvHA4Oniqnw4UqHtg61b7ketFQAlCnuHk6Ng5UOZ9/XKg7OrldlBNq3qaBVQ6hjEwOx1OTp6dgNJdC7sQnhXFxcCaNYSm3w0cCIMGxVXAshdeq1bQuQvhtQEa5n0KHBY5vo7bgTSt9k1VQNmdC0gcwjPm7HM6ptU6Pa9Jx00vPOO9OY8qGIzWN/jxh2qb3QbA0KEEfj4U9j6PUBNwbfRYoRBq7pr+/WHbNoJ9S+B3qu98aGB/+P3vIRQiNDRI0d7AjpHd64aOInz4jnIuJ25GEC9z4Znv84aG2LCsXp6VYLASUDpuXB09RK+H8HQHSk+kt8Koo5MDFQpFp64yRJ6TA+VFQBnbioAShALGqwNl4JQD5cWBMspwEknJCCiddITwzA2QUwjP7LAAEA4T/vZ7aLeepq49CId3sjyOZV3Wr4EBAwhtrIXgaBg8WC8WiB0HKoY+fQivDdA4bwEMjgooK8HqZhwoA+MecPoF70ZA6cdxCuEZSblmAeUlhKeHJ51yzfRyzDlQZnEeEVelYQJLl8JXXxGc9xFQStPgITHHte2FV1ICJSUE25iOb8zzFoo9z3rdre4Zq3OQKITHjsPV1KhX/bxaOVAxHRs0zALKTWcLvZOInQMVDNr/fxhumPEdDEpKrAWVLqCc7l+vITyITSTXybUcqByphiDkPsnkQJn3deqF53UUcqvwYCohPPNDKdu98IyHaEw45LPPaD7//2D4cBp79YV774V336XouyWWZegUz3kHNm4k2H93OPNMNRvyDvSxuMAizLFjsN6GF15R03T8+CM0NNiG8BKNRO53CM88R5+O0YCFw9H6+JVEbnYHzce3c6Ase/Jt2UL5rdcpJ+lnPyOweSOUlcPuuwPRffTzazW2mN25DIViw2V63a2+r1UjnSiEB9Hvb74OVg6UnYAyJ3q7ESXm4TqsksidHCir72Cur3mZkXPvlwOlh/AgXqBLDpQg5ClOjk4yDpR5WS6E8HSyOg5UOEz5D19DXZ9o1ms4DHPmECYAnTrRuK4WVq8isHoloT2vhdNPg5oaipcMhe5j1JDGnTpF5tsp+p+afTV0xWXwvXmok9jvGncNdtuN8CGH0fjWu/DyjhhQ310JH3t6XBn697IaSNMq+TwZAaU3MnpXf6cGzAgtpeJAJRvCc+qFV1MDfP45ZWuXK/U6eDDBDX1g8PhIgfq9Egqpelv1YDNvZ35vhJKNutv1etXLMTo0uHGgjO9v90woKYmWl2wIz8mBMgs4PdHf7MLZHUOvu5OASrcDJSE8QShAvPbC8zIOlFcHyg79WHZiKtMhPNcjkX/8MWWvXAMVo+GQQ9SAlitWwE9raC4phyVLaFzXCJesoOirRQSWhuGRR1QZAJ/VEqKJJkIw4efQpw/Fy79RycRHHgH/sK6vrYAKBmHqVBp7PAYvbVQTsH73LeE1PwFdAOvpWdyG8FJxoJqaVK6Ycf7M51TvGWZc00QOlFOIxEpA6fdYwnGgAnWEHnkCgofSNEwNsV1bC3z7LWXUwl//ChdfTPA+4EfrOoVC6v8lkSi3Oq/GND3GOXAKaRmUlsY7fcb3Mh/H+P7me0h30MrLVd8EtyE8Lw6U1fH1AVLN4Tmrsab0scn0BHJz/fzOgUrkQOVaCE8ElCC4xCmEZ/dr07yvm3Gg/HKgjEbRrleQuZ5WAkpPME5VQFmFW2xDeAsWUE4NbN8GL/0X5s2LbBMecwi0b09jE7BvZ4oO2BcO2g3uuw969qRo+VB4r5qS7eupW72e5v/8BwYNoohGGD6cYJdOcaLS7ECFQvFJwOFAkIZfng67AjNmwIrlhN9+GypPjCkjUyE8YwDGpiZrJ8BA7xlmJaCSTSLXh5iorFQDtpvrYOlAvfpvQg/dD4HlNLc7GI4ZQ+3mOvjhByWgjjgCcBbbupvkJYSn1093oJy+J0QFlB8OFEQFVDocKKtpjXQBZQ6T2w3WaeRwOTlQ5kFq7cpxwosDJSE8QchTvPTC039tWpVhFcJL1oFKFMKzGj7AwKnhDodje/ZkLAdq7VpYU0V5qAHGjIUPPoC1P0W2aT7qGMB0vkaNiowaXfQi0AuKWjfT+PBTNC/6ChZ9RjENcNRRke9iNVq5/oCOE1Cag8PAgbBiOc1vvgMTJkIgECNC3E4mbJCMgDKcjK1bY8Z/tMQQUMY58yOEB1Hnq337qIBK6EB9MJsQTRBupun2O+GIWmq/6ASNDZR1aKXOLfH/O1bOUrIOFETPgV0Dr5djNOhuk8itytXLM0SDlQMVCMSP1ZRMCM/I99LPvZEHZvyA0I9jdwwrB8q8LBUHKhSKOmFGOUa9JIQnCAVIsuNApSOJ3A6jQXfTbRmsH0p6L5h05EBZhvA+/xyAspFD4MADYehQtayxETp3Jty7D2Dv2EV+gRcHCZ14Ag2t34SPPqKIMPz855HvYjUxsP6AdkwiHjAAXnuV8GeL4MsbYPBgwoceG/lOxveym0zY/P2TEVDgTUCB+xCe07XWz8u2beq1fXv4/vv4Ohjl1NXtOHY4TNl7swjRFXrsTNPKz+Hss6nd43cAlA3bM3JS3Agjp8mEzd/DqwNlDuFB7A8RXRy7SSI3Xzej/uZ1+hAEVvvq+xnr9HG2zP9z+jojnFtcrASLnqxudwwnB0o/jpu62lFeruojOVCCUKD40QvPqdx0JZE7CTK9nlaCQRdQyeQdeHWgiovCEQFVfuxY2I56uu63X2Qbo7FPNPdbcTHUlxapkND++1O052mwlxJfdtdHz7Gw2iYiuioroecuhFegBuX89FPCGw8C2rmaTDiVEJ6Vk2HVG0rHPImuXw6ULqAMrBpTY8DFwPp1lK75gVBJDzj1VJoe/hxWrqRm5QfAPpSPHBLZ1+le0ccrMgSrlZixur/NDpQbAWU06DU1sGSJ6pSp/6+5CeG5daD0IQisyjfKMhLRS0uVE2g1ea+R72Weo7OoSAmWZB0oq9HUE5XjRHk5bN6cfzlQOVINQch9khlIM93jQDnV0W5cIKt6mt8bGA9lfYwYL3gO4X23BDash1ARZUceYlmmWXDaCaiYHlpt21K8ex/b49rlQJm3iQmFHn88zX/4I3TpCuFmmt95Ty1fs4bgkq+guTmtITxw7g2loztQMUKQ5MeBgqgwqqy0HkPIKGfbNrVx2XdfEgCCe+8F5eU0nX8htG5NbfudYNh+lB1xsOVxwD4HyrytXw6UOQcKlIB6/HFYsMB6jCewD+FZXTcrAWU1tYvV9TCHu6ymVTHK0R0ofbmTgPLiQHXqlLgcJ8znQ3KgBKGASWcIz60DpZdtNWmtmxCeWwGV7AMrURJ5nJP32o4hAnr3pqxT65h1Q4fC/PmJHSg9idbKGQD772POgTITE/Zr147w6PEw/wN4YxbhF16EmioCq14mSD10O47mnocS7rSb9cHw34GyHQNLGxPI6oeAOZzpdL31HlqGA1VeHh053tKB+vIHmP4g5axXy0cMA6CpTz/YvJnahwLwQ4Cyttbf0/zZLAT0fDW3AioZB8pg1SrYc0/rbd04UK133NrG9fMawgOVN6RPtmsVwjPemx0os9NjdYyuXWH1aujWLf7YlZXR9+PHwy672Jfj1oGyqpeE8AShQPB7IE3zOq8OlMG2baoD2p57xj7Y9BBeoglOreoMqQsor+NAFb3yIjAa+venuFjNorJtGxxzTOwULpB47jezgHIKM1g5UFYulTkkFw6jkp7fmEX422+BnQgEAwSLS6BqNU1TryY86xEgELNPsiE8I4fFwG0IzzgnxmjkZrwIKGO9PnxCWZmKsv7wQ2TYLSB6Dus++RwIq152wSChMQfBoh3nOxikti5ajoGbJPLItC8eHChjkFQvAsocxgqHY+8Vva5GncyDsep12XdfdS/tvXf8OjchPIDRo2HlSvV+9WrrEJ6dA9W2LaxfHxt6NR/j2GNh7Fj1P2ima1c44QTo2BG6d3euq5vnWe/ean7vHVNTWgoo/ZznSghPBJQguMTvgTT9SiJfs0Y1ZMuWxcxOEnMsu6EMMimgEiX8sn49RXPeBUbDbrtRVAS//KX6bv37qwHAId6BMp974+FbVhZ1SMDZgXKbA2U5plP79tBjZ5pXBqHHznDNbwm2CsKpd9FctQa+/RboF3csHS8CSscQUFaTuuqYQ3hmzPeGGwEF0bLKytRwXZbbhcMqcQgoP30i/P5CQsX9YZHztfTiQCUSTfp7I2yV6Jzp+7ZuDccfr9LzlixR4svODRk8GNatU+Loww+tc7MqKuDgaLQyqRDesGHq77//VZ/tksgh3oH6+c9VZ1enTiTBoLV4AvVMGTTIel0yIbzhw5WotArhGdMP6T9exIEShDzGz4E0k00ityvbQH/gZFxAPfAA3HsvJYsXc9joP8OQIRRf/G+C6yth2y/U0/e55wjc/EcI7BiioKmJQLiZnXoE2dKzHa1bQ7t20SKNh30iB2rAANi4URlDzz8fXe7kQBk45UDpxzaInOOf/5wwXaHteAIdiwm1Anr1ovm7IOEP5kBFP8tyknGgdJzG49ExJ5EnItH1Nosbc4gLgJoaQq++Cd/sDtWboaiY8vEHw+BSQjtm3jHOt9W1dAr37ryzcl+M8JIXB8qoqyGu3ThQoZASRiUlSkBt324voLp1g1NOiZbhxjUxO1BuQnjmdVYTGxvlWDlQbdvC8uXujuGFZBwo83bGNTKGUiktFQElCHmN3wNpmtclG8LTQ1tWSeSGA6XX3epXsVMvPE8PrKoq+PWvIwc56IVL4YUdx6A3UAxvvw1bqgmyBdgWs/uvL25L82/jz4Nxzozvq8+tplNSAmPGxNfbjQPllANl60ABdOxIeOSxMFcbSLNfv6iAOvxXluW4EVBOLoFTbyidRA6UGbcOlEGcgHr7bZg0idDyEuBUtaxvXyralcbsbzT6Vu6rkwN1+OHqGlsNA+BWQOkDzVphVY4+/6CbfBxdQLkRQWDtQDmJL9vpkLAP4VkdN9sCyryPESauq1MCyi5pP5uIgBIEl6SrF56B35MJ2wkoI/lUr6f5vUFSAmruXHXw3XeH226DG25QltC4cQQ3t4MHgC3VECoi+NgTMHzX6L5lZYS6d8fqcObRvfW51ezwKqAS9cJzmpYlbiDNXXelmSB88gmMaXC0FlN1oAySTSJ3qk+i9XGOSTgMp50GK1cSCvYD47zsvntc0rRZQOnlOOVAmbf1EsIzi710Cyir92ZScaCcJuS2SyK3KjcdAsrIv/RKIBANwdfWKrdMF1CSAyUIeYrVA8FIJDU7O04hPINUx4EyJ1ebl0PyAsruwevInDnq9ZBD4Oij1d8OgiuArV+q0cUPOojA4XtAB3fFms+lUTe3Asopv8ZAD7e4eUjr5zhuKpfOnWnq0IXwhhVqnr6aGti6lfANJxIu7xPznRI5UMa9layAsksiN8YSMuNFQMW5T998o+JrJSWE3v8AznxXJeoNGBBJxtYnWwbvDpQZLyE88yjfbnKgjG10AaVP0OumXk7bmUWPmxwou3XJOlB+iRK7/zmv6AIKovep8f+QC4iAEgSXWIkfA3Oitt04Q+kYB0ov240DZZcL5JsDZQioHVOr6ASDqOSkHdN1eHlo2zlQlvk3O8ikA6Wf71BIvWk+7HDCTy+AFdFkk/Db7xIe715AGcsaG1MXUGYHKhSyFlCJroujgHrvPfU6fDih7l1Udy1TfXUHSh+Hym0OlFN93IbwDJJxoMLhaA/EdDhQ+lARifZ1cubscqCsjpsuBypZzD3xcm0IA5CBNAXBNYlCH1bCxOmh79c4UF5zoPTu1b4LqLo6NVgTwMiRcau9uApmzILRawjPzTAGSedAYRHCA5p/eQrhf/8Hjj4Ghu0YTX3xYsd62gkoq3Vuc6D0JHK9zm7cl0Tr4wTU+++r14MOsq2vLqD0IRT8cKDSGcLTw2tGLz4/HCir+1TP73J6jjiF8BI5WekWUKk6UBAvoHIlfAcioATBM3YPM71LtV3uUzqSyBP1wnMSUIl+IXsWUJ9+qnbq1An69Ytb7cVVsNs3HTlQBl574dkJ1pi69u6t+psfeKDa7tvvCO8Y+MiLA2W1rrQ09hwm40BZkeh6Ow0yGXGgDjwwrhwjhGcsb262F1CJcqDs1nl1oNxMJqyXYx46wulc6d/BiwOlv3q5FmDtQFkdx/zZLwGl3/+pOFDmCYXFgRKEPMYphAfWPYJyIYncXGZaQ3hG+G7kSMsT5acDlUoOlJ275JQDZeVAOuZAYZpMuLIS2lYSbm4m/P0y23p6EVCBQOz3t7x3Ghsp/uEbqK11PWCm4/VuaiL0zdfw7rvw5ZeUh3fEsrZvV9f/m29UxUaNcuVAGT8cQiFnQZ8uAeXGhdPfGyLQGLw0XQ6UWwFlXp9tB0ovy88QXq5N4wKSAyUIrkkUwrN64GViHKh0hPCKilR9PAuot99WrwccYLk6FQGluzrhsP85UGZnxpcQnnkuvF69YNEHSmQM709g6xbY3EwoVGldt7o6uPRSQkuPhuFHEArFi9Ly8mg+Tty988kncM45FH0aAo6hYXhrwodfjPHbOSkBNX48oTe6A30BKHvuU/jHT/DZZ9GK7LUXtGtHaFPsrmYHSg/hJeq671cOlNskcj8cKL0Mt/W3CuE5kWsOlFFWg3PH04TkQw6UCChB8EiiEJ5T6MfOxQqHsxPCs3uAlperX9ieBFR9Pbz1lnp/xBGWm/jlQDU2Rhtev3rhmceZSSWEp4eoYujVi/CiAOFP5sO6AIHrfgP8QGjomTD0emjdmtC/X4byGujbF2bMgBkzCBGG7SWE+hwE9TsmPtxxUW0dqMWL1XDXNTUUFe0HjWEa5y0gvPB/wJAYp8yM7fVeuRLeeIMQp8KAgbBmDWUbtqoht0G5bG3awJQpceUEg1HxYiWgnOaOs/psty5RDlRJSWyP2UwIqER5TH6G8ArVgcrFHCgRUILgknSG8Jqb3U3+q+PWgQLvIbyysiQE1Icfqlalc+foJF8mrEJQbtFdHcN9Cgbj5xyzOp5ZLFi5S4kElJUDlSiEF+dA9e0LBAivXQtr5wGNQBOheXNg4f0QDBKqvQGoju4TCBAKN8G77xB6914480l1gJEjYcAAyn84EPofAx07UlS3DW6dDps2wYsvqhM1ejTFVz8L586i4etlhGe9AcEhMULP7rzF8eaban3fXeDEEwEo6/8zqHlDjS6/334xF1UvR8/X0keVr69X7+0GTjXwK4RnjDHkZS48KwHlJYTnpVejVwfKvN6LA+U2xOgVoyw/HSgJ4QlCHpOuEB7ETrGRigPlVwjPHBZz9dCaNUu9Hn64bYvhlwOlJ5C7+WVvHtAvkYByOw6U5xBe+/aEr78BZn0GVQ0ErngIDuhJ6OSr4WOlVkO77Qq9OsKCBbBhA9x+O6E3doWXFighZRT8wQfwwQeUsx7eq4LBgym+6xpY/2H0eB07whNPULS5I/TrR+PXRUpAjbs0OQfqjTfU+sF7RhaV7b4LDD0rYTl66ExfbjUJLni7V7yE8MCdgHJyZyHxXHp6XbyIID+TyBM5UPpwCfniQImAEoQ8xosDZZAohJeMgDI7UHr5+nK3AsqpZ5Wrh9brr6vXww+33cSvHCi7aVzM2P0Stvo+5oH6vOZA2SWRmwnv2o9w636wAQKHAD0h9MB98NuPoKKC0ON/hjZFauf166FLFzW33uBxhPrXw8/+pZbPmgXr1lH+0E+wtAbmfUQRVbDrrmp21s8/h1tuge7dKdqGElAUEZ73MYyuhVZl3hyocBhmz1bnaPBekcVuQ6h2950fAspLCA9i75tkHSg3jkgqDlQmQnjGsnQJqEIfxkAElCC4JFEIz8mBsirDyoHyMvVBosmEUwnhmRvFhA/Xl16Cjz9W7z0IKC8hPDsHygm7X8JODpSxLhUBpedAWTmX5mWh8hI46CD13nBqQiHo0iVal+JiQm2LoRKVa3TuuQCUD62FK2dDXR1F5/8dfn1EXFyzuBho146GLjsT/qkZPvmEwMj9CAZNGdXaoWP44QcVEly5EkpLCQ3sDzuGs3KbxJ/IgUqUA+V2LCQ3YkqvS7ICymqdXRnJOFDG+UgkGFJJIteXiQPlHRFQguAStwNpOoXw7Eg2gRzS0wvPkwP1/vtqtOlwGM48E3r0sN3U3Csp1RwotwLKzWTP5l+4bpLIdffPVQjPVIbx/RONUeXUyJW3L4tMl1M0AbDICYuMAzVgEPwEzH6DwFuvEexXDiN+uSM3y6IOTU3w6qsq38noYXfAAYQqogrESUDp11e/7wyXrrnZnQOV6F5JFMIzhytTcaCMnoRW6+zK8CKgMjmMgb4sVx0o4/6QHChBKAD86oVn5UB5eeAkmsollV545kbR9lfwihVw/PHqZ+Ixx8A//+lYZ7dTW1iRaQfKzThQbgbSdNrPaiBNq/Pi1Mjp1yrhSOQHjyXcbzI814HAxipCSxbDkkdg2DCK9hpAY1lrWL+e0BUPwHtPwI8/RgvZc08ltP74R0IbrI/vhHn4gFDIWUC5HYTSvN7qnjafNzcCyg93NpkQntmBSiUHqlAcKKv/z1xABJQguMRtLzynsV/chPDcYpUDZXcsKwFlroNnB2rJEjjpJFi7VvW6mzkzoQJ0Oy6O075+5EBZNWjmX7hpy4GyKEc/ptV5cXSgEg2kqS1vLGtN+KabYWcIbNtAcMHz8OZy+ORjSj95m0ZaARDkUWB19Av9+tfwj39ETmTotWjZbgWUubekcY7cOFBeBJSVmEpGQNmVmY4Qnl5+NnKg3BzHC346UMZQF5IDJQh5jN+98HRSCeElcqDM5epCL2kB9Z//KPFUV6d6ev373/GxDQu8uAp2+2YrB8rYTsevEF6iRsw3AaVN5RLo2IHQ5LOg21JYtIiSlYvYVlsB5eWEfn4WnHAQDB2qrKM2bWLKc+pwYIdZQBnfxRDDTiI3kdhOh4AqL1frWrWKPX4yIbxE97reGy6VXnjmHMp8dqCKi6Nh3tpaCeEJQkHgVy+8dIfwrBwo/b25nq574dXXw+9+p8TT2LEqbNenj+s6Gw9FrwIqmRwo43skcg3c5EClI4ncuIaVldC6dSRnPA63AipRCK+5Ofo9I07ZbrvBbrtR2u14qNpxnAuBTtZl6fUoLXV/Ha1CeOCPA2UXAjWPqWRVF7tGvrQUJk+OF36tW8OIEfDtt+o69u9vXy+3DhSo6RI3bID27WPrnEz+lIEbB8os2PzADwFljNe1fbsSULkYwsshMyx1evfuTSAQiPn761//GrPNZ599xkEHHURZWRk9e/bkpptuiivn6aefZo899qCsrIy99tqLl19+OVNfQchhMtULzy1WA2naHUuvm11j7JTn0aqV9uGRR1SvrK5d4YUX4hKQE2E0Kl5DeMk4ULvvrjTeoYfGLrd6CPuZA+XUecBKiJWUwEUXwemnW34N2zpB1BExh2p19OXGwJXm8pzmCrSrj1v3CUz3EIkFlB85UF26qDmczQPju8kbA3WLG4JG58gj4cIL4Te/gZ13TlwvNyLzyCPh1FPjOxZ4daB03DhQw4crDd2zZ+I6umXwYOjWTY2okQr6hMISwssA11xzDZMnT458bqNZz9XV1RxxxBGMHTuW6dOns2jRIs466yzatWvHOeecA8CcOXM4+eSTmTZtGscccwyPP/44EyZMYMGCBQwaNCjj30fIHRKF8HbaST2g9Aeq2xBeKg8HLyE8twJKb2AqK9XUZoBSejfcoN5fdlliBWOBl0bFaj8vDlRxsWpAzVi5S4lyoIzt7D5bheYSuVb6feAknvv1U9PnWWnVigolEouKEocAIdaBMidKd+igBog0ix278twIqEMPVXUfOtS6DDfDGCSbAxUIqHNjRq93Oh0NLw6Umd691bVNJELsRDC4c6CGDFF/fjJokPpLFT2RPBcdqIITUG3atKFbt26W6x577DHq6+uZMWMGJSUl7LnnnixcuJDbbrstIqDuvPNOxo8fz6WXXgrAtddey6xZs7jrrruYPn16xr6HkLvYOSe77AKXX279izZRCC+Ru+VUD7uBNO0cqC5d1EgDvXrFlqcLBz1sMW6c9p3uuw+++04Vct557itrcZxUcqDcJpHbYd4v2XGgrHKgzA6UmxBeIvbcU/3ZYSUSdezuNbNzc/bZSiM7TY9jbAvu9PPBB6s/uzKMa5loGAM39THvZ4dx/UOh9Doayf5YAPU/euml3r67+Ry6EVC5jC6gcjEHKofMMH/461//SseOHRkyZAg333wzjcZMlcDcuXM5+OCDKdGeDuPGjePrr79m48aNkW3Gmn6yjBs3jrlz59oes66ujurq6pg/IT9ZvRruvVdpBDNuGj2zeHIK4enoja9XEg2kqQuoYFC9nzw5Pqyh5y106AADBqhcjwEDdmywdStcfbV6P3Wqq6RxK5IN4Vk5UMkKqF13Vc6E/ivZrxwosA7hGeXaiapMod9rZuFRUaEcx0QY91Oy51/f15hTzq9eeG4aWOPYfub9WJGKAwXu/kecHChzWDeXxIcbjOv07LORaRhz6jsUlID67W9/y5NPPslbb73Fueeeyw033MBll10WWV9VVUXXrl1j9jE+V1VVOW5jrLdi2rRpVFZWRv56+hlMFjLKzJmwahU8/HD8ulQaPat9/XagnEJ4hlPg1OC1a6c6W/Xsqco+6SSVlxGp0623wpo1Kp60w7FNhlRDeOGw6k0GyXeTLipSro3xr+62F54ZO8dPX+Z0bZMRzMmgH0cXUF6Fh0H//upv+PDk62Tck37nQLm5rzp0UP8LNsEK30jFgfJ6DLD+f/CS25ZrWDmckgPlgcsvv5wbb7zRcZvFixezxx57MGXKlMiywYMHU1JSwrnnnsu0adMoNXcB8ZErrrgi5tjV1dUiovIUw9lwIhmRY2AXwkvGgfLSC69zZyWGTL8NYjASmS0fUGvWqHnVAK6/PqUBXpJtVHTBaHy3VBsE/RyaQwQ77aTcmPp6dV8kE8IzyjYfKxnBnCqBQOz3NAsoL9ejbVs4+eTU6mNuHNORA2VHaSlcfHHuO1BejgHW3yefHai991bTPpaXw1dfqWXZdG/N5LyAuuSSSzjjjDMct+lr0wtoxIgRNDY2smzZMvr370+3bt1Ys2ZNzDbGZyNvym4bu7wqgNLS0rQKNCFzlJREfxGbScUlSpREnspDwU0vvEBAheMSYfuAvfZaFcLbbz+YODHpuoI/ITzju6XaAOp1MDtQbdooQTlvHrzySvIhPGN5tkN4hoCyGm5Br1+mMLuhfuVAuRUJmXhkZ8KBchrGAPJbQPXqpWaHamyE665Ty9z8yM0UOS+gOnfuTOfOnZPad+HChQSDQbrsGFxl5MiR/OlPf6KhoYHiHXfarFmz6N+/P+139FUdOXIks2fP5qKLLoqUM2vWLEaOHJnaFxHyAidjxY8QXjqSyO0acbsRyj2zdGl0ipabbkrZNknVgdK/o18NgtU4UPox9e3sPicaxiCbITz9WFYDfkLmG1ezA5WOkcizTS44UPkcwjMoKoLTToPZs2H//bNdmyg5L6DcMnfuXD766CMOOeQQ2rRpw9y5c7n44os57bTTIuLolFNO4eqrr+bXv/41f/jDH/j888+58847uf322yPl/O53v2P06NHceuutHH300Tz55JN88skn3Hvvvdn6akIGcROZ8iuEZ7yawypey7YTUL6Fif70J/UT8MgjYcyYFAtLPQdKx88Qnl03abdDEiTqhZcLITxzPbPpQCUK4aUzBypTZFpAFZoDpdOvn/rLJQpGQJWWlvLkk0/yl7/8hbq6Ovr06cPFF18ck5tUWVnJ66+/zgUXXMDQoUPp1KkTU6dOjQxhADBq1Cgef/xx/vznP/PHP/6R3Xbbjeeff17GgGohuHGg0hnCS6ZBtZtOxJdGet48ePppVYhpUNpkSdWB0vFLQIF9N2mnQTHthpDI1RAe+JNE7gdeQnjJjkSebTIdwitUBypXKRgBte+++/Lhhx8m3G7w4MG89957jttMnDiRiSnmeQj5SSZDeMZruhyoVIZGAOD776P5TqefroYX9oFUc6AMQiH/HBwnB8q8nd1nu154BrkWwkslidwP0hXCyyWRkEsOlDltQEgdVwJqyJAhBFye+QULFqRUIUHIJvoDqKnJ+sHnZwjPvNwLVjktdo6IZzZtgkMOgeXL1XwoFlMeJYtfDpQfjZJVCM9cr1RCePo2Vg5UtgVUMsnXfuFFQCU6T7kawjN6vTr1fk0Vt8MY+PmDQ1C4ElATJkyIvK+treUf//gHAwcOjCRWf/jhh3zxxRf83//9X1oqKQiZQn8A1dXFjhWZjhCeVaPmlbTkQP35z2q+u7594a23fG0B/MqB8lNAQeIQHsSeU9051Nfp+9iJpVwM4WXbgfJjGINcc1n22ksN2prkmLOucDuMQS45c4WCKwF11VVXRd6fffbZ/Pa3v+Xaa6+N22bFihX+1k4QsoidgEqGRALKj4E07d57blDmz4d//EO9/9e/1IBIPmI8yL3Wy7y9H2P4uEkiNzALJKdxtxIlkZuPnwkS9WLMtWEMvCSRexn8NNOkUzyBNwdK8BfP/zJPP/00v/rVr+KWn3baaTz77LO+VEoQsoXRiEJ0ji4zqYTw7EjFLUrkQHlm6lS188knq5lgfSaXHCgDr73w9HCcgV0vPKdysh3CK7RhDHIpfJcp9OsoDlRm8Xy7lZeX88EHH8Qt/+CDDyhLZWIkQcgBdAdn5Up49FGVBgS5F8KzcqBSDuEtX65GjYTovHc+k4s5UJA4Bwqc3SS3vfDM22aKXAvhBYOxg1k6hfDc5kC1RAEF0e8tDlRm8WyCX3TRRZx//vksWLCA4TsmQvroo4+YMWMGV155pe8VFIRMoouR//5XvX73XdSUSRa3ITwvWO2bcgjvvvtUgYceCrvt5r1SLki2F56xr5+zsuvnMFG5VmIo0bARuRzCy3YSOSgXym4uvGSGMWipIiEUUkO1OfXCa6nnJp14FlCXX345ffv25c477+TRRx8FYMCAATzwwAOceOKJvldQEDKJHsIzMI/onY5eeH4lkafUC6+uDu6/X71PYbLgRKTiFujnx895zNyG8JyW2fXCM++j75dLIbxsuDd60MKPyYRbugPlNA6UCCj/8fQIamxs5IYbbuCss84SsSQUJFbTn/iR6J2pJPKkx4EKh+HCC2HVKujSBbSet36TSmMXDLobr8ktViE8NwLKKgfKbb5TLg2kme0BKI08KKsu9pID5R6nEJ44UOnD0+1WVFTETTfdRGNjY7rqIwhxLF4Md98Npjme04KVgGrTRr2mM4SX9alcZsxQ4btgEB56KK0zraYSwtP3ydQ4UAb6veEkXu164VmJrlxyoLIVwgPrht9LDpTR0y3dPd5yFePaOSWRt1RxmU48n9LDDjuMd955Jx11EQRLZs6EtWshE508rUJ4hoAySOdAmsk0qClP5dLcDNdfr95fey2MH++9Eh5I1YEy8LsXnpepXLyE8JzGgcqmgILcCeFZNfxe6ta5M5xwQlqN05xm8GDo0cN6uDYJ4aUPz1kERx55JJdffjmLFi1i6NChtGrVKmb9cccd51vlBEGnvj79x3DjQKUzhOeFRA6U1XexZPZsNW1LZSVcdJH3ingkl3Kg3IwDlUoID6LXwWqfTJIohJdNB8rqWnrJgQJoydOVjh1rv66yUr22bZuZurQkPD+CjNHGb7vttrh1gUCAJquf8ILgA5n4hWwlOoxfcOkI4ZmPmxUH6l//Uq+nnpqRGEiqvfAMMpUDZWAVwnMTMs21Xni5lETuJKCyXbdCoXdvOOOM9E4n01LxLKCaXf+sFQRv1NaqsZf69LF+YGbiIWql/83Cyc+BNP1MULd771jm2rXw/PPqfRp73un45UClayBNp7nwDKyGMbAL4ZnFsZUQywSJhjHIpoBKNQdKsCcQUCJK8B/R9ULOMHs2PPIIfPWV9fpMPEStfh8kco/ckKmBNO3e2/LQQ9DQAPvtB3vv7f7gKWA02rmQA+VmHCi3OVB2zpJTEnkmycUkcsPwLCmJXycOlJDrJJVFsG3bNt555x2WL19OvSkx5be//a0vFRNaHtXV6nXzZuv12QrhmcWPFzKRRJ60AxUOR8N3kyd7P3CS+NULz88cKPAWwnPKgUrkQGUrhGdVn2yLlH79lHYfMCB+ndccKEHINJ4fQZ9++ilHHXUU27dvZ9u2bXTo0IF169ZRUVFBly5dREAJSWM0YHajZORzCM/ozp7JkcgTCqh334UlS6BVK/jlL71XIElytReelxCelfunr3PqcZftEF4uJZEXF8PRR1uvy7a4E4REeL4tL774Yo499lg2btxIeXk5H374IT/88ANDhw7llltuSUcdhRaC8WDPpoBKdwjPIF0hPE+98B56SL2efHL8WA1pJJVpN9I5DpS5t5x5m0RJ5Hb1dHKgMolTLpf5fS6Qy3UTBEhCQC1cuJBLLrmEYDBIKBSirq6Onj17ctNNN/HHP/4xHXUUWgiJHKhs5UAZpBrCcwrdpHMqF8syw2F44w31PsOzCgwcCH37JtftXG9I/Q7hJRLIdsLDqUx9v1wcSDPbDpQTkkQu5DqeBVRxcTHBHXd2ly5dWL5jqvrKykpWrFjhb+2EFkVLCOGZl/vRC8/OgXIsc9kyWLFCqZBRo9wf1Ae6dYNf/Qp23tn7vunshWd1DP2z29BbruZA5eIwBk7o5zHX6iYIkEQO1JAhQ/j444/ZbbfdGD16NFOnTmXdunU88sgjDGrJI5kJKZNLDlQgEB+6SzWE59Rw+jWMgV23+jiM2QT220/lQOUJ6eyF55cDZS7DKQcqk+SbgILo/2Eu1k0QPN+WN9xwA927dwfg+uuvp3379px//vmsXbuWe++91/cKCi2HXHCgjMbl3HNhzBj13q9eeE4OVCqNaiIHyhJDQB18cPIHzgLpyoECewHl1Tmyc6AkhOedlj5JsJDbeHaghg0bFnnfpUsXXn31VV8rJLRcsp1E3twcbdzatoX27dV7v0J4btZnbCBNQ0CNHu3+gDmA3zlQBk4OVDIhPB0J4SVPKkNeCEK68fwvM2PGDL7//vt01EVo4WQ7hGce58fc4KUzhJeOJHLLMrdtgyuvVHPfBYNwwAHeD5hF0tkLz20Iz8010l0op6lcMolVfcSBEoTk8XxbTps2jX79+rHLLrtw+umnc9999/HNN9+ko25CCyPbITxdfIRC9gLKC25DeKk4UDqOIbxwGH7xC7juOvX5jDPybobRdM6F59aBMofo7Mo072c1/Us2HSi9Tub3uYIkkQu5jOfbcunSpSxfvpxp06ZRUVHBLbfcQv/+/dl555057bTT0lFHoYWQbQGl98CzcqAM0tELL5VxoHQchzF44AF47TUoK4Onn4b77nN/sBwhnb3w/MyBMpedTDl+YycEjeXiQAmCN5K6LXv06MGpp57K7bffzp133snpp5/OmjVrePLJJ/2un9CCyIUcKKtjZbIXXqpYlRMIAOvXw5QpasE118AJJ+RlYkm6xoFyE8LzKnLtxHG2Q3hmATVihBqTq3Xr7NTLCcmBEnIZz4+g119/nbfffpu3336bTz/9lAEDBjB69GieeeYZDs6zHj1CbmHlQGXy17oearEK0+RDCM+2N98zz6hJBvfcEy6+2P1Bcoxc6IXnNYSXawNpmr/n+PGZq4NXxIESchnPAmr8+PF07tyZSy65hJdffpl27dqloVpCSySRgMpUCM/8qzcdITwzfoXwbB2oZ55RH04/3d/uaxkmnXPhpUtAma9tKh0GUiHbx08GyYESchnPt+Vtt93GAQccwE033cSee+7JKaecwr333suSJUvSUT+hhaBP5qoLKHNeUjoxJ/umqxee3gCn4kBZYTm57eZN8NZb6sMJJ6R2gCyTzV54XsfqytVeePkkoMSBEnIZz7flRRddxHPPPce6det49dVXGTVqFK+++iqDBg1i52TmZhAEYhuVhoboe11AZSqEZzTM6eqFpy/zO4nc0oF65211IocMgV13dX+AHCQbc+H55UCZQ3jiQCVGcqCEXCYpXR8Oh1mwYAGzZs3itdde46233qK5uZnOnTv7XT+hhaALJd2BskvsNrNlC8ydCzU1qdfBzxCeQaIQXtqSyMNh+Pe/1ftf/MKfg2SRXOiF51ZA6WU7lZsprIYxyHXEgRJyGc+/4Y499lg++OADqqur2XvvvRkzZgyTJ0/m4IMPlnwoIWnsBJTV5L5WzJgBGzfCunVw7LHJ1SEdITyrbdPpQMXxxRcEvvxczXd31lnuC89RsjkXnl+98LJFrtXHDZIDJeQyngXUHnvswbnnnstBBx1EZWVlOuoktEB0odTcrP6CwdjlTi7Nxo3qdcWK5OuQzhCeXQOddA7UN98QuPB26PUzOOII620aGuCNNwgQhssugx1zWOYz+vnJtxBerjhQ2QohJoM4UEIu4/kRdPPNN0fe19bWUlZW5muFhJaJOazQ2AglJfYT5drt26VL8nXIVAjPKXfJVdlNTTBpEqzqDKvmQJ8+8MEH0LdvdHLgcBhefBE2b4JOneGSS7xXOgfJhgOVyEW0W2c3jIGbMtJBPjpQkgMl5DKedX1zczPXXnstPXr0oHXr1nz33XcAXHnlldx///2+V9Atb7/9NoFAwPLv448/BmDZsmWW6z/88MOYsp5++mn22GMPysrK2GuvvXj55Zez8ZVaFOZQnRHGc+NAGe4TRCcATgZzCM983FRDeL4lkd95J8yZo5wlgMcfhx+WqZ52n30Gf/sb3HwzLPoMAkEC11+nQngFQDZyoJzq4Ga7XBEsuVYfN+y+u5ptqEePbNdEEOLxLKCuu+46HnzwQW666SZKSkoiywcNGsR9WZwaYtSoUaxevTrm7+yzz6ZPnz4MGzYsZts33ngjZruhQ4dG1s2ZM4eTTz6ZX//613z66adMmDCBCRMm8Pnnn2f6K7Uo7ASUGwdqzRp/6mAXwkt0fCe8hvASUlsL06apfSdMAAJgCCnC8O/nYOMGqNmuFo0bR2D4ft4rnqPkggPlNYSXKw6UQT4JqNGj1biveTZlo9BC8CygHn74Ye69915OPfVUQtoTbO+99+arr77ytXJeKCkpoVu3bpG/jh078p///IczzzyTgOlJ0bFjx5hti4uLI+vuvPNOxo8fz6WXXsqAAQO49tpr2Xfffbnrrrsy/ZVaFKk4ULqASqU3WzZDeK4btaefVpnyPXvCr34FI0dCeQX84gQo2nEfd+gIZ/0azj1PzdNRQGQzB0r/7KUXXq7lQOVTLzzID6EntEw8P4JWrlxJv3794pY3NzfToA/gk2VeeOEF1q9fz5lnnhm37rjjjqO2tpbdd9+dyy67jOOOOy6ybu7cuUwx5gzbwbhx43j++edtj1VXV0ddXV3kc3V1depfoIVhlQMFmRVQ6e6F50sS+T/+oV7PPZdAUUglkB9+uNoxHIZFi+DII2NimYXUAGVjJPJkybWk7VyrjyDkO54dqIEDB/Lee+/FLX/mmWcYMmSIL5Xyg/vvv59x48bFDO7ZunVrbr31Vp5++mleeuklDjzwQCZMmMALL7wQ2aaqqoquXbvGlNW1a1eqqqpsjzVt2jQqKysjfz179vT/CxU4bkN4mzerfOna2uhyvwVUJnvhGbhq1D76CD78EIqL4eyzo9sab/baC045JS4RrJAaSr1bux/fKxMhvFx1oArpvhCEbODZgZo6dSqTJk1i5cqVNDc389xzz/H111/z8MMP89///tf3Cl5++eXceOONjtssXryYPfbYI/L5xx9/5LXXXuOpp56K2a5Tp04x7tJ+++3HqlWruPnmm2NcKK9cccUVMeVWV1eLiPKI2xDejBlKRK1erWYlqa+PTSJPpwNl4GcIz3WjFg6DcY+deiqYRL4ThdRQGtfGL/cp2RCemzJzLQdKBJQg+ItnAfWzn/2MF198kWuuuYZWrVoxdepU9t13X1588UUOP/xw3yt4ySWXcMYZZzhu07dv35jPDzzwAB07dnQlikaMGMGsWbMin7t168YaU1bymjVr6Natm20ZpaWllJaWJjyWYI8XBwrAmHpRd6KMbVKtQ7oH0rRanzAv5cknYc4cqKiA667zXI9CwfjOfs+HrJ9/N9fMzbnPNcGSa/URhHwnqcfQQQcdFCM6DD755JO4Hm+p0rlzZ09TxITDYR544AF+9atfxSSH27Fw4UK6awMMjhw5ktmzZ3PRRRdFls2aNYuRI0d6qrfgDTc5UDrGcrNgKsgQ3vbtaiBMgCuu8Nynu5AaynQ5UE4Cyq9xoHIlhCc5UILgD54F1NatWwmFQpSXl0eWLVy4kCuvvJKXX36ZJrdzb6SJN998k++//56zzz47bt1DDz1ESUlJJFfrueeeY8aMGTHDL/zud79j9OjR3HrrrRx99NE8+eSTfPLJJ9x7770Z+w4tEa+98NIpoLIRwnNs1G6+GX78EXr1ihkQ0+t4RIWA8V38FlD6NU51HCi7XngSwhOEwsJ1EvmKFSsYOXJkJFF6ypQpbN++nV/96leMGDGCVq1aMWfOnHTW1RX3338/o0aNismJ0rn22msZOnQoI0aM4D//+Q8zZ86M6ak3atQoHn/8ce6991723ntvnnnmGZ5//nkGDRqUqa/QIkl2HCg/BZQ5hGcuMysDaa5YAUYO4E03gfbDpSU2gMa1yWQIz2sSuXm/XHF88nUYA0HIVVw/hi699FJqa2u58847ee6557jzzjt57733GDFiBN9++21Mb7ds8vjjj9uumzRpEpMmTUpYxsSJE5k4caKf1RISkOw4UJkI4aVSdsohvMsvh5oaOOggSPKezHbD7SfpcqD8FFDSC08QWgauBdS7777Lc889x/7778+JJ55It27dOPXUU2NyhQQhWcy/io0hxbIhoLIRwrNk7lw1TUsgAHfcEbdxSwzhpSsHyksIz22ZudoLL1vHF4RCw3UIb82aNfTp0weALl26UFFRwZFHHpm2igkti1wM4fndC8/N/jHrLr9cvZ55Juy7r/O2LQS/HSiDlhTCs/ssCII3PA2kGdR+QgWDwZi58AQhFfwK4aVCOnrhGbh1oCKfv/4a3n1Xqbmrr07+wBbHyGf8zoHKRAjPbbnpJtvHF4RCw/VjKBwOs/vuu2PMK7d161aGDBkSI6oANmzY4G8NhRZBLjhQ6QjhuU0ij/v84IPq9cgjwSa/sCWG8PKxF56BhPAEobBwLaAeeOCBdNZDaKHU1UFVlXsHqqgous5YppOrITy39QoEUF/woYfUgrPOSli22zoUAn7nQBmkM4Rn9znT5Fp9BCHfcS2g3PReEwSvzJoFn3wC5plv7ARUKJQ+AZULITxAnZTVq6FTJzjmmOQPWoBkw4GyukaFEMITBCE1PE8mLAh+Ysxjt3597HK7EJ4596UgQ3hPP60+nHQSOOQZtkQHyrj+fqVfJiNq3IbwzNtJCE8QCgufh6MTBG/U16vXmprY5U4OlE6+DKSZaByoyPaNDfD88+qDT2ORFVJDOXAgrFoFw4enp3w3bpOE8ARBABFQQpYxBJQhLoqL1RhQTjlQOrk+kKa+rysH6uN5ypbr0gUOPNCxzJbYALZpAz//uX/lJROuk154giCAhPCELGMIKIPSUvVqF8LTHajm5vwP4cVtP+t19eb44xMm+rTEEJ7fpENU2JWRbQGT7eMLQqHhWUBdc801bN++PW55TU0N11xzjS+VEloOiQSUkwPV2Bg/gnku98JL6EA1NqoEcvAtfBd3DMGRdIbwzKHhTCMCShD8xfO/9NVXX83WrVvjlm/fvp2rUxzwT2h5mAWUkRzsxoFqbMysA5X2EN5XXxHYsll1SRw9OmGZ0gCmTrIhPDdl5loIz0y2jy8I+Y5nARUOhyODaer873//o0OHDr5USmgZNDdH57wzSORA6TQ0ZHYYA4O0hfAWLiRAGCZNctVPX0J4qZNsL7xkcqByrReeIAip4TqJvH379gQCAQKBQMyI5ABNTU1s3bqV8847Ly2VFAoTs3gCZwEFsWLGbwGVqRCe1Xqqq+Hbb6MCykek4XSP2zGfpBeeIAiuBdQdd9xBOBzmrLPO4uqrr6aysjKyrqSkhN69ezNy5Mi0VFIoTMzhO4gKqLVr1WwmtbXRdWYRkm8hPAPLEN7nnwNhAvvuC/36uSpTHKjUaUkhvGwfXxAKDc8jkffp04cDDjiAIr9m8xRaLE4CCmDZsth1ZgFVECG8cBgIwJdfqs9HHeW+cCFlMhnCy7aAyfbxBaHQ8JwD1aZNGxYvXhz5/J///IcJEybwxz/+kXqrFlEQbLC6XTp0UGNBWZFuAZXxgTQ//5zA5LPh2Wdh5Y9AgMC4IzyX7dd2gr+98MxILzxBKCw8/0ufe+65LFmyBIDvvvuOk046iYqKCp5++mkuu+wy3ysoFC5WAqpVK7joIrCKBvsZwluzRg34vW5ddJldCM/q2F6xFFB3301g2xb4fJH63KsXgS6dXZcpAip10jkOlDhQglDYeBZQS5YsYZ999gHg6aefZvTo0Tz++OM8+OCDPPvss37XTyhgrARUKKREVNu28ev8cqDWr4d77oGFC+Hjj6PLE41EbuBLCK++HmbOjN14zz2lUcswyeZAFUIITxCE1PCcyBQOh2ne0dK88cYbHLNjtviePXuyTv85LwgJsBJQhvtTXh6/LhyOHReqoSHx1C5WZTz5ZPTzpk3R93a98Iz9kgnh2e7/yiuwcSOBdu1g0EHw44+w115JizM/tmuJpENA2e0nITxBKCw8C6hhw4Zx3XXXMXbsWN555x3uueceAL7//nu6du3qewWFwsXOgQJ7AaWTTAivvl718DPQBVk6Qni2DtQjj6gPBx4IQw/zXjAioNJBMsLIbn2uO1ByXwhCanj+TXTHHXewYMECLrzwQv70pz/Rb0eX62eeeYZRo0b5XkGhcHESUGVl8ev8COGZx5XSh0lwCuHZDUOQCMsk8i3V8MIL6v3BB1luL2SGZBwou2XmdSKgBKGw8exADR48mEWLFsUtv/nmmwm5GD1ZEAxSdaCSEVDGAJ0GNTXR9xkL4X3wvqr8sGEE+vYB7d9JQniZJZ0hPDMSwhOEwiKpf+lNmzZx3333ccUVV7BhwwYAvvzyS3766SdfKycUNk45UG4cqGRCeE4CKt0hvEhBb72l3k+enHh7L2WnuJ0gITxBENzj2YH67LPPOOyww2jXrh3Lli1j8uTJdOjQgeeee47ly5fz8MMPp6OeQgGSKw5UOKwak4yE8JYvJ1C1SnU1PPlkArOst/cTaSjtyWQSubEsFSczFURACYK/eHagpkyZwplnnsnSpUsp02yCo446infffdfXygmFjZOAKiqKn0/XjxwoQ0AZAk2f0NhuIE3zsVMK4c2fr+a7O+UUaNMmpUZMGsDUyeYwBpkm28cXhELDs4D6+OOPOffcc+OW9+jRg6qqKl8qJbQMnARUIBDvQvkRwjNEUnl59FhGGM8phOembCtiXKxNm+HLL5WAsgjfWR3TD6ThdE+6B9LUl4kDJQj5jWcBVVpaSnV1ddzyJUuW0Lmz+1GUBcEpBwri86D8dKCKiqICzSygfA3h1dbA9u2q7i+/DE2NBHr1gmHDLMuSHKjMkukQnpf1fiMCShD8xbWAWr58Oc3NzRx33HFcc801NOyIewQCAZYvX84f/vAHfvGLX6StokLhYQgoPVSnv0/kQDU0xLtGqQgo33vhhcMwYQLcfjvhr74m/LgawTNw2KG2haRDQAn2JCugvJSpL8/mNRMBJQj+4lpA9enTh3Xr1nHrrbeydetWunTpQk1NDaNHj6Zfv360adOG66+/Pp11FQoMQ0C1bh1dpgsoKwdKH/hSD+F5FVChULR8Yywo33vhffwxgU/mQWMDPPwQ4VWroG0lgYMOjGySiUZNGkr3uHWWks2BkhCeIBQOrnvhhXe0HpWVlcyaNYv333+fzz77jK1bt7LvvvsyduzYtFVSKCyammDDhlgBtXmzeu/kQIF9CM+tgDJcpqIiKC5W7+1CeBDtNZVUCO+BB1S+ExAmQLioBE46iUCrCu9lWSAhvNRJx0CabrcRASUI+Y2nYQwC2n/cgQceyIEHHuiwtSBY88or8Mkn0c+6A6XnQJWUxO7nlAOVagivtja6r35cXUB5CuHV1MATT6j3xx5H+PulhCeNh607uXIv3CANYOqkM4Tn1wCcfiECShD8xZOAuvLKK6moqHDc5rbbbkupQkLho4snsA/hJRJQViG8ROgCSg/hGfNgt2kTL6Csjp2Q//wHNm8m0HMQDBkCw/clfCDwauxmEsLLLQo5hCcIgr94ElCLFi2ixNyqaQTkiSAkQbICKhUHKhSKdaDWr1fvO3aM3V4XUOZljuxwnwInnQiBgO3+0gsvu2SyF162HaBsH18QCg1PAurf//43Xbp0SVddhBZCUVHsiOCGgAoEvIXwUkkiN4fwDAeqUyfr/TyF8DZuVHFKgBNOgFfc7y8OVGZJh6jI9PV1iwgoQfAX173wsu0uXX/99YwaNYqKigratWtnuc3y5cs5+uijqaiooEuXLlx66aU0mubuePvtt9l3330pLS2lX79+PPjgg3Hl3H333fTu3ZuysjJGjBjBvHnz0vCNWi7mcFurVurVPPJ4OhwoPYnciwPlpuwIzz+vKjdoEIGBA+L29yuMIw2g//jhQDmF8BIdK51k+/iCUGi4FlBhTwkg/lNfX8/EiRM5//zzLdc3NTVx9NFHU19fz5w5c3jooYd48MEHmTp1amSb77//nqOPPppDDjmEhQsXctFFF3H22Wfz2muvRbaZOXMmU6ZM4aqrrmLBggXsvffejBs3TiZK9ommpvgBNEtL1atZWCUSUE1NqTlQVjlQZgcqqRDek2q8J375S8u6ZzqJXBpKe9yKCrPoTTaEJ8MYCELh4FpAPfDAA1RWVqazLo5cffXVXHzxxey1116W619//XW+/PJLHn30UfbZZx+OPPJIrr32Wu6++27qd7TY06dPp0+fPtx6660MGDCACy+8kBNOOIHbb789Us5tt93G5MmTOfPMMxk4cCDTp0+noqKCGTNm2Natrq6O6urqmD/BGmPMJR1DKJkdKMOZMjALKP1zKiG87dvVsArgLKBchfA2b4Y331TvTzzR0sHyKwfKLdJQ2pOsgHJTZrLDH6QLEVCC4C+uBdSkSZMoNayCHGTu3LnstddedO3aNbJs3LhxVFdX88UXX0S2MY9XNW7cOObOnQsol2v+/Pkx2wSDQcaOHRvZxopp06ZRWVkZ+evZs6efX62gMMZcAiWQBg6MCiXzuE99+8Lee8P++6vPZgHV3Bz9bIivZJLIN21Sy4uKwPwbwXMvvFmzVGH9+8NuuyUUYJlIIhfc44erJyE8QWgZeJ4LL1epqqqKEU9A5LMxybHdNtXV1dTU1LBu3Tqampost3GaKPmKK65g8+bNkb8VK1b48ZUKEsOBatcOpkyBiROhQwc148mECbHbBgLw85/DQQepz1YCJtFULrW18NlnUFenPus5UOaRzjt0iA8jeg7hvfSSej366LhVEsLLTdy4S4UYwhMEITWyKqAuv/xyAoGA499XX32VzSq6orS0lLZt28b8CdYYDlR5uXKBjIf6PvvALrtY7+P04DcElJXQAfjwQ3juuejYU+YQXpHWD9WcQG4uN6ED1dwML7+s3u8QUHrdzXU1v08X0nA64/V6uA3h5RriQAmCv3gaxsBvLrnkEs444wzHbfr27euqrG7dusX1lluzZk1knfFqLNO3adu2LeXl5YRCIUKhkOU2RhlCahgOlNn9ccLpQW+eANgscowpYrZuVa+6gAqF4MQT4d134ccfVdTN7tiucqDmz4efflKjce4YpT9RLz4J4eUWfjhQTiG8XHKg5P4RhNRISkBt2rSJZ555hm+//ZZLL72UDh06sGDBArp27UqPHj1cl9O5c2c6d+6cTBXiGDlyJNdffz0//fRTZKyqWbNm0bZtWwYOHBjZ5mXDIdjBrFmzGDlyJAAlJSUMHTqU2bNnM2FHPKm5uZnZs2dz4YUX+lLPlo7uQPmBmxAeRIWTngMFsPvu6q+pKT6JHTyG8J5+Wr0ecUR8F0KsHSi747lBQnj+4DWE57Rdon0kiVwQCgfPAuqzzz5j7NixVFZWsmzZMiZPnkyHDh147rnnWL58OQ8//HA66sny5cvZsGEDy5cvp6mpiYULFwLQr18/WrduzRFHHMHAgQM5/fTTuemmm6iqquLPf/4zF1xwQST5/bzzzuOuu+7isssu46yzzuLNN9/kqaee4iUjbwWYMmUKkyZNYtiwYQwfPpw77riDbdu2ceaZZ6ble7U0Mu1AGblPZgFVZLrzrcSTTsIQXmMjPPKIen/aaZHF0gsv90lXCE+SyAWhsPEsoKZMmcIZZ5zBTTfdRJs2bSLLjzrqKE455RRfK6czdepUHnroocjnIUOGAPDWW28xZswYQqEQ//3vfzn//PMZOXIkrVq1YtKkSVxzzTWRffr06cNLL73ExRdfzJ133snOO+/Mfffdx7hx4yLbnHTSSaxdu5apU6dSVVXFPvvsw6uvvhqXWC4kRzIOlJscKLcOlJ5E7vXYjiG8V1+Fqiro3DkmgdxrDpQ4UNklnSE8q3IyiQgoQfAXzwLq448/5p///Gfc8h49ejj2VEuVBx980HLUcJ1evXrFhejMjBkzhk8//dRxmwsvvFBCdmnCbwcq2RCeVwGVMIT3wAPq9bTToLjYsa7ScOUWyfTC81qu8VlCeIJQOHjuhVdaWmo5UOSSJUt8y2cSChe/HSi/QniJju0Ywlu3Dl58Ub03dYpIZwhPHCh/SCaE5zZXymldth0oQRBSw7OAOu6447jmmmtoaGgAIBAIsHz5cv7whz/wi1/8wvcKCoVFunKgrAbSDIcTJ5G7PbZjL7zHHlNz3+27LwwebFt3GQcq90k1iTyRGJMQniAUDp4F1K233srWrVvp0qULNTU1jB49mn79+tGmTRuuv/76dNRRKCD8cqAMx8kphNfYGBVYqeZAOYbwjPBdgo4GfudACf6QyRBeNsm1+ghCvuM5B6qyspJZs2bx/vvv89lnn7F161b23XffuClSBMEKPxwoPYTiJKCM8B2kMYS3cCH8739q2AKLThSSRJ77pCuElygHShwoQchvkh5I88ADD+TAHYMFCoJb/HCggkH119TknAOlT1zsp4CKqc9TT6nX445Tc8E41N3VXHoeEAHlP6n2wpMQniC0HDwLqL/97W+WywOBAGVlZfTr14+DDz6YkNskE6HF0NQE9fXqvRcHyowbB6qhwVlAJZMDZV4GwH//q17NE/lZICG83MRrCC+ZcpMtw09yrT6CkO94FlC33347a9euZfv27bRv3x6AjRs3UlFRQevWrfnpp5/o27cvb731Fj179vS9wkL+ogsaryG8QCDWAXISUJ9+Cv/5j5pfz6CxUa3z6kAZWIbwli+HRYvUwcePt627XoZ5mYTwsk9LDeEJgpAanpPIb7jhBvbbbz+WLl3K+vXrWb9+PUuWLGHEiBHceeedLF++nG7dunHxxReno75CHmMIqNLSqOBxi7nhMfa3CuH95z/q/Y7B6gElnAyxBckNpBm3zBhzbP/9rWciNuH3VC5ukYbTmWSSyPNxGAMz2T6+IOQ7nh2oP//5zzz77LPsuuuukWX9+vXjlltu4Re/+AXfffcdN910kwxpIMSRTAK5gV0DZnag7GhsjLpP4FMOlDEFkDbyuFO9s+VACe5JZy88L/umg1wTcIKQ73h2oFavXk2j3hLtoLGxMTIS+U477cSWLVtSr51QUBi94nZMTegJOwFldqAAtBmGIpgFVCrjQAGwZQu88YZ671JA+T0SuYTw/MEvcZSorFwL4cl9IQip4VlAHXLIIZx77rkx06F8+umnnH/++Rx66KEALFq0iD59+vhXS6EgMBLIS0q872sXwrNyoFq3jt8/HI4KuFDIe0MZl0T+/PPKUtt997jBM+2QyYRzk3SF8BJdWxFQgpDfeBZQ999/Px06dGDo0KGUlpZSWlrKsGHD6NChA/fffz8ArVu35tZbb/W9skJ+46eA8upAQVRAeUkgtw3hPf64+nDKKQlbInO4UZLIcxcJ4QmC4BbPOVDdunVj1qxZfPXVVyxZsgSA/v37079//8g2hxxyiH81FAoGQ0AlE8LTSeRA2YXnjBysVAUU69fDrFnqvcXgmVZlhMP+J5FLA+gPLbUXntw/gpAaSQ+kuccee7DHHnv4WRehwMmUA6X3ttNJVUBFlr3+mjrwfvvBbru5LktCeLmJ1xCel+2SKSNdyH0gCP6SlID68ccfeeGFF1i+fDn1Rqu4g9tuu82XigmFh18CKhh07oVniCozfjlQgTd2uE8TJ3oqw0rYpdKouXVLBPdICE8QBLd4FlCzZ8/muOOOo2/fvnz11VcMGjSIZcuWEQ6H2XfffdNRR6FAyFQSeSIHKplB8iMCqqYG5s1TC3/+c1f7ml0sCeHlFpkM4Tl9TjfZPr4gFBqek8ivuOIKfv/737No0SLKysp49tlnWbFiBaNHj2aiy1/kQmHx00/w0Uf2wsXASOLOVggvlSTyCEuXEmhqgEGDoF8/9wWRnalcpJFMTDK98LyWZbyXHChBKBw8C6jFixfzq1/9CoCioiJqampo3bo111xzDTfeeKPvFRRyn9deg1degW++cd4uHTlQbhwoY1tfQniLFxMg7Np9Mpehf3ba1ku5qW4jREnVgbIry0pAZRoRUILgL54FVKtWrSJ5T927d+fbb7+NrFu3bp1/NRPyBsPZ2bDBebt0hPCsHChzDlSrVuo1ZQFVVx9ViS4mDzaX4caBEjJPupLIE11bcaAEIb/xnAO1//778/777zNgwACOOuooLrnkEhYtWsRzzz3H/vvvn446CjmOIQw2b3beLpVhDFJxoFq1gq1bk8uBCtTVQLiMcDgAixdDYwOBnXaCIUM8f4dsTOUijWRi0hXCs1ouITxBKBw8C6jbbruNrVu3AnD11VezdetWZs6cyW677SY98FoohjBwK6DSkQNlLAuHo8tCIRg6VOVoQRIO1DPPEDj/v9BrDOERhxL+dKE61uFjPbU+Tg6U1XZey011GyFKukN42SQX6iAIhYQnAdXU1MSPP/7I4B1TV7Rq1Yrp06enpWJC/uDVgfIzhKe7OoaAMurz85+rXO9HH1WfPQmocBiuu44Ag+GHZYRPmEg4fADQFg4/PKm6J3Kg0tHASaOZmGQcqGSdKnGgBKFw8JQDFQqFOOKII9i4cWO66iPkIYYwqK523i4dvfCslumuFEQFk6deeB9+CP/7H4HiIui5C+GaGqitgfIKAvsN8/4FSJwDlQ4HSkhMMjlQXsuSEJ4gFB6ek8gHDRrEd999l466CHmKIQy2brUfxBLS40Dpy+zyogzB5MmBuuceVe7++8OvfkX4gIMIE4CBA5WoSqLu+mjmTtt5LdfPMls6LSmElwt1EoR8xnMO1HXXXcfvf/97rr32WoYOHUoro4vTDtq2betb5YT8wBAG4bByodq3t94mHSOR6+vtco3MAiphEvn69fDUU+r9oYcCRYRvv4PwA0ug0tvYTzrZ6IUnjWRiMhHCEwdKEAoPzwLqqKOOAuC4444joP0HhsNhAoEATU4WhFCQ6L3eNm+2FlBNTdHt/OyFZ7XMPLSBIaAMoZdQwD3wgIr3DRlCoN+u8A1QXAx77glrk3eK/M6BkgbQHzLRC89rOelABJQg+ItnAfXWW2+lox5CHqOHpuwSyfUpE4uLvR8jmRCe2YEyaN3a4UDNzfDPf6r3559PIKgKiZkLL0mhI73wcp9kcp3cbJML10EElCD4i2cBNXr06HTUQ8hjdAfKLpHcEFDFxfHixw1+JJEbmKLOscyerQbLbNMGTj6ZwItqsS6gksXvJHI3SCOZmJYawhMEITWSaMrgvffe47TTTmPUqFGsXLkSgEceeYT333/f18oJ+YEXByqZ/CcziQSUXRK5gaOAuu8+9Xr66dC6tWUCuJ8hPKvtvJYrpEa6QnhewnnZINvHF4R8x7OAevbZZxk3bhzl5eUsWLCAuh19wzdv3swNN9zgewWF3MecA2VFKkMYgLcQXtIO1IYN8Pzz6v2vfx1TRrpCeJJEnlukuxdeNq+HhPAEwV88C6jrrruO6dOn869//YtiLZnlgAMOYMGCBb5WTsgPMuFAeQnhGXh2oJ54QlV08ODIVC1WAipZspFELo1kYrw6UG63y/UQntwbgpAangXU119/zcEHHxy3vLKykk2bNvlRJyGHCYdh4ULV09/ASw5UJhwo8z66gAoEoLzc4gDNzXD//er9mWdGdvYzhCdJ5LlJJnvhJbutH4iAEgR/8SygunXrxjfGjPQa77//Pn379vWlUkLu8t13Ksr197/Hjv9kUFsbHW9JJ5WJhME/B6pVK5uG44474NNPoawMTj017rh+hPCsHKxc66nV0pEQniAIbvEsoCZPnszvfvc7PvroIwKBAKtWreKxxx7j97//Peeff3466ijkELrz9OOP6lV3oMDahcpGCM/KgbIM382ZA5dfrt7ffjt07hy3SS72whMHyh8y2QvPbttMIAJKEPzFs4C6/PLLOeWUUzjssMPYunUrBx98MGeffTbnnnsuv/nNb9JRRwCuv/56Ro0aRUVFBe3atYtb/7///Y+TTz6Znj17Ul5ezoABA7jzzjtjtnn77bcJBAJxf1VVVTHb3X333fTu3ZuysjJGjBjBvHnz0va98g19nNSFC9WrISyM0JhVHpSfAioYTD6EFyeg5s6F8eOhoQF+8Qs491zb4zotc1P3ZB2sROWmuk1LJ5kcKK9l5WIOlCAIqeF5HKhAIMCf/vQnLr30Ur755hu2bt3KwIEDae04OmHq1NfXM3HiREaOHMn9Rq6Kxvz58+nSpQuPPvooPXv2ZM6cOZxzzjmEQiEuvPDCmG2//vrrmClnunTpEnk/c+ZMpkyZwvTp0xkxYgR33HEH48aN4+uvv47ZrqWyfXv0/eefK+1hOCvt2kFNjbWA8rsXnl24zmqZrYBauRKOOgq2bIExY+Chh+IK9jOEl41xoARvpCuE56X8dCEOlCD4i2cB9eijj3L88cdTUVHBwIED01EnS66++moAHnzwQcv1Z511Vsznvn37MnfuXJ577rk4AdWlSxdLFwvgtttuY/LkyZx55pkATJ8+nZdeeokZM2ZwuRHmacHU1ETf19XB8uVRYdG+PaxenSchvHBYuU2bNsGwYfDSS1BRYXvcdPXCszqWW5zK8dvtKmTSlUQuITxBKGw8h/AuvvhiunTpwimnnMLLL7+c03Pfbd68mQ4dOsQt32effejevTuHH344H3zwQWR5fX098+fPZ+zYsZFlwWCQsWPHMnfuXNvj1NXVUV1dHfNXqOgOFChhZDTWhiZNdwgvEHAXwnN0oB5/XImmkhLlPFmIJ/24fvTCs/qcDgdKXC1veD1fqeRAZfvaZPv4glBIeBZQq1ev5sknnyQQCHDiiSfSvXt3LrjgAubMmZOO+iXNnDlzmDlzJuecc05kWffu3Zk+fTrPPvsszz77LD179mTMmDGR8avWrVtHU1MTXbt2jSmra9eucXlSOtOmTaOysjLy17Nnz/R8qRxAd6AgNicqkwIqJQdq61a49FK14MorwcFJ9TOEZ/U5HUnk0jB6I9Ukci9TE2X72lgJOUEQksOzgCoqKuKYY47hscce46effuL2229n2bJlHHLIIey6666eyrr88sstk7r1v6+++sprFfn888/52c9+xlVXXcURRxwRWd6/f3/OPfdchg4dyqhRo5gxYwajRo3i9ttv93wMnSuuuILNmzdH/lasWJFSebmM2YHKZQFl60BNm6ZijbvuGhVSCY7rRwjPXGa6ytEbdGkovZGMKHUjiHMhhKcfU+4LQUgdzzlQOhUVFYwbN46NGzfyww8/sHjxYk/7X3LJJZxxxhmO23gdW+rLL7/ksMMO45xzzuHPf/5zwu2HDx8emcOvU6dOhEIh1qxZE7PNmjVr6Natm20ZpaWllCY7wFGeYThQZWVqvCddQLVvr16rq5XY0B/S27apV8d56BxIJoRn6UB9+THcfLP6cOutCQemkhBe4ZNqCM/L+2xfG7kfBME/khJQ27dv59///jePPfYYs2fPpmfPnpx88sk888wznsrp3LkznS3G3EmWL774gkMPPZRJkyZx/fXXu9pn4cKFdO/eHYCSkhKGDh3K7NmzmTBhAgDNzc3Mnj07LhG9pWIIqFatlIBqbIyua9tWPaCbmpRg0jtmbtmiXtu0Se64KTtQ4TD8+CMVd52qhiw44QQ47jjXx00lhGdXptN7r+W4KV+wJl0hvFy8DuJACYJ/eBZQv/zlL/nvf/9LRUUFJ554IldeeSUjR45MR91iWL58ORs2bGD58uU0NTWxcMcgRP369aN169Z8/vnnHHrooYwbN44pU6ZEcpZCoVBEpN1xxx306dOHPffck9raWu677z7efPNNXn/99chxpkyZwqRJkxg2bBjDhw/njjvuYNu2bZFeeS2Zhgb1B0ocrV8f60AVFSmBVF2twniGgGpuVmlH4J+Acu1AzZ9P0bU3w7ujYOMGKlgO++wDDz7oqRVJJYTn9jCSA5V9khGlyYbtJIQnCPmNZwEVCoV46qmnGDduHKFQKGbd559/zqBBg3yrnM7UqVN56KGHIp+H7Jjs9a233mLMmDE888wzrF27lkcffZRHH300sl2vXr1YtmwZoHrZXXLJJaxcuZKKigoGDx7MG2+8wSGHHBLZ/qSTTmLt2rVMnTqVqqoq9tlnH1599dW4xPKWiOE+BYMqhAexAioQgMpKJaDWr4cePdTy7duViAoE/AvhWTpQzU1A9J4M/Ps5OO9UymprOZalBIuLKDv5JLjxRtcV8SN3JdMhPMmB8kYy1yDfQ3hyXwhC6ngWUI899ljM5y1btvDEE09w3333MX/+/LQNa/Dggw/ajgEF8Je//IW//OUvjmVcdtllXHbZZQmPdeGFF0rIzgIjgbyiItpImwXULrvAihXwzTcweLBaboTvWrf21mPJjmDQQpTc8w8C138Ox5wN++4LCxYQvPpXQC0cdRRDL75YLbcY1sKJdIfw0lFOthvpfCPVEJ7bvKlcQASUIPhH0s3Zu+++y6RJk+jevTu33HILhx56KB9++KGfdRNyDMOBKi+PCiEjB8poVHbfXX1eujQ68naq+U9G+fr7GCG2ciWBG64nEG5SYzu98AK8+AIBmuH//k99HjvWs3jSj+tnCC8dOVAimvwhGVFq5/hJCE8QChtPDlRVVRUPPvgg999/P9XV1Zx44onU1dXx/PPPZ3RUciE7ODlQxueePVV4r6ZGTTa8yy5pEFCECcyeDSv6qAM89xyB5kYCpaVQ1wSfqnG9ghdeAH+7KaXWIl964QWD8dci1TJbCpnshWd33EwhAkoQ/MO1A3XsscfSv39/PvvsM+644w5WrVrF3//+93TWTcgxdAfKeAAbjbbxORiE3XZT75cuVa++CqjGRgJ33E5g2vXw6KNqRPEN6wl07UrgqqmwSy/YuSecciqBm1MTT/pxMxHCS0WYeXFBhFi8hvCS2SdXhEu2jy8IhYRrB+qVV17ht7/9Leeffz67GS2k0KLQHSgjPGfleuy2GyxaBN9+C4cd5oOA+uILAo99Cv1/Bm++SWDhWwRpBsKw7HsAAn+7k0B1J9B6SwZDNuV5IJdDeG7KFLwhITxBENzi2oF6//332bJlC0OHDmXEiBHcddddrFu3Lp11E3KMRDlQBsaI5LW16tWYGjApARUOwxlnEHjpRbj/flj4KQEgcP310Ku32ma/4QQOPcRRqCRLOkJ4btd5KVdvxCWE5w0J4QmCkAyuBdT+++/Pv/71L1avXs25557Lk08+yU477URzczOzZs1ii2EzCAWL4UDpAsrKgTJGt/AliXz2bPjkEwKEYfMmAALHHkPgoAPhtNPgjDPhyCPjGjVznZIlHSE8NwMvuiHZhluIJZO98HLleuRKPQQhn/HcxLRq1YqzzjqL999/n0WLFnHJJZfw17/+lS5dunCci5GdhfwgHFbTxdXVRZcZDlRFhX0OFMSLq5QE1F//qsrfZ28oLYPOXQhM+pU6RlER9OoFgYClgPKzkfAzhKcPn+ZnErnfZbZEkslP8yJks31tck3ICUI+k9Jv9P79+3PTTTfx448/8sQTT/hVJyEHWLkS/vlPePHF6DKrEJ6TA9XUFJ3WBZIQUDfeqByooiIC55wDU6bAOecQqCi3dJvSGcJLpdxUcmbclithu+RJNYSX7PUUASUI+Y0PQQ41OvmECRN44YUX/ChOcEFDA8ybF50ixW82bFCvmzZFlxluVFmZcw6UHsIz6hcKKefKNffcA5dfrt5fdx2BLp2hpESJqUDiqVz8CN8Z5UJ+hfBETHkj2R51yeyT7euRK/UQhELAp2ZGyDRvvQUvv6x68Ru5Rn5izHmnTxZsLCsudnag9HWGgGrVysNDe/FiuPhi9f4vf4E//CFOIFg1YukIj6SjF55dCC8VJITnD6nmQOVLErkgCKkjAioP2b4dPvlEvV+1CtIxAHwiAeWUA6U7UMY+JSUuD9zYCJMmKbtr/HiYOjWufLt8p3Q7UOZlXsswyGQSuZCYZM5dsiG8bF8ncaAEwT9EQOUh8+ZBfT2UlqrPb70VzU/yi/p69ZqqA2WsD7kdk+nGG+Hjj9VYCPfdF3nSmxueRCG8dDpQqZad7iRyCeF5I10hPKttsn09cqUeglAIiIDKQwz36dhjoX17JWxWrfL3GIZYMl7DYWsB5ZQDpa93JaAWLoSrr1bv//536NEjsiqXHKhkyzDIZBK5NJTeaCkhPLkvBCF1REDlGeFwNK+od2/o3l29X7PG3+OYQ3jNzdFcK7cOFESdLFcC6rLL1IF//nM49dSYVV4FlN8OlJ5nlq4QnlckhOcPXs+dW7EuITxBKGxEQOUZekituBi6dlXv0y2gjM/Gcd3kQOn7JRRQS5fCrFmqoFtvjXvCew3h+SVSDFIRUGYkhJdb+BnCy3VRKwJKEPxDBFSeYTg6oIRMt27qfVWVv8cxhE9Tk3K9jOMGg0oAeHWgEgqa6dPV65FHQp8+jptmw4HK9RBerjfc+UIySeRe3mf72oiAEgT/EAGVZxjCpqhINcSGA7VuXVTM+HkcUC6Unv8EzjlQ+jADrhyorVvhwQfV+/PPt9wkF5LI/SrbjcvhtRzJgUqeZByoZF2rbF8buR8EwT9EQOUZZiFTWakGtmxqUiLK7+NArIAyhiNwcqAgKpgSCqhwGM4+W43c2aePcqAscJMDpeN3npFfOVChUHrcIjsHUEhMqiG8fAqZigMlCP4hj9o8wyxkAoGoC+VnGM+tA2WVA6WvTyig7rkHZs5Ultqjj9pumEtJ5MmWAf6KGwnh+U8ySeQSwhOElokIqDzDyCkyhAzV1XT9aREsWcKaZf4NBpVIQJmTyO0cKMccqOpquPJK9f6mm2DUKNv6mEVINqZysaqLV6zqnWy5EsLzB68OVDL7iIAShMKjKNsVELwRI2Tq62HYMDoubQ8cyeb1PeBnk5Mq9/334Ztv4KST1GTBerK6kwNlkJQD9fe/q9Bd//7wm9841i+XHKhUQ3h261Kpc7pcrpZAukJ4uShSREAJgn/IozbPiBEyM2bA0qUUB1Xr3rDg86TLfeMNWLZMjSSgHwfcCahEDlScgNq4UQ1XAGq6liJnLe9VQOXqQJpO9RIHKvtICE8QBLeIgMozDEFSEmiA668HoPiCcyAQpGHdJli+PKXyFy1Sr14FlGcH6je/USJq4EBleyXA3PBkuheeX+NApcuBEgGVPMk4UKm6Vk77pRMRUILgHyKg8oyIkPl4Dvz4I+y0EyVn/wq6d6eBYnj3XQC+/x7mz0+u/I0b3edAGXjqhffss/DYY2qnGTNcDVOerRCegV8hvETCM9lyJYSXPMnkQOl46YWXbeGS7eMLQiEhj9o8IyJkvtphFZ1xBsWtS6FXrxgB9fzz8OKLsH594jLD4dgH66JFqTtQZgEV2X7Llmi+0+WXw4gRiSuIs4Ayym5pITx9H3Gg/EFCeIIguEUEVJ4RGcbgmy/Vm/33V6LGEFDvvANAzY4OeZs3Jy6zsTFWICxZEr/eaw6U8TkuB+qaa2D1ath112gPPBc4hfDS2ThZCahcDOG5acQFa1INx0kITxBaJiKg8oz6eqCujuIV36kF++2nRM0uu9BACSxZQnjV6ojg2b7dvqxNm1SoT3ebjOU6jY3xwyd4daBCIVRM8Y471IK//U2NAOoSNzko6XSg0hXCs9vOa9kyF17ypJoD5cb9yzXhkiv1EIR8RgRUntHQAKxaRTH1sMsu0K2bEjXl5TTs1AuApldnRRyTbdvsy7rjDnjoofi8861bYz8nkwMVDIQhHI4KqO1bYOJEVdjxx8NRR7n5uhGy7UClayDNVOucKIQnJCaZa+A1hJeojEyRa0JOEPIZEVB5RkMDsHIlJdRH8ocMUVPfb6Da5qXXI9vbOVD6vHk//uh8TKepXAwiD+SGBrj9dkIX/QZmzqS+phE2biR0yUXK7urdG+67z/mAFmTbgcpECC8VJITnD+kO4WX72oiAEgT/kIE084z6emDlSoppgOHDgaiAau63O03vBmmY/S7s2QzBoK0DtXFj9L3xMK2osBZcDQ0uc6CWL4cTT4SPPiLI6fD1VzTdOwPWrSPY8LmauO+ZZ6B9e8/f2yyOMtU4mR2oVFwikBBeLpJqCC/ZJPJskO3jC0IhIQIqzzAcKCsBRY8eNLTvSsPGbWqYAGBbp/FA57hy9N55hsiqqFACrbExdltXvfBWLIdz91UFt2tH6MCj4KUlsHoVAKHBg+CFp6BXr6S+t9cQXr70wvMrhJdtZyOf8XrurARUIBDfm9XLcTNFrgg5QSgEREDlGQ01jbClWgmoPfYAVFgoGIRmgjQcNp6GZ16G774FYPsfr4ORv4KhQ2PK2bAh+t7IeSouViKqujr2mJY5UIs+g7nblIirqSH4j+uUeBoyBJ57juDc3tBmCXzxBQweTOjivtDLn6e2mxBeLjlQOjKQZm6TbI86t/tk+9qIgBIE/xABlWc0bFJ2UQkN0LEjoB6GxcVQVwcNJ/+KhmdnQ9fuEA6zbc06NdL3okVqkrsd6A7Uli3qtaRECYWEAmrRIoInnwTbTlI/u5csIbB2Dey2m5oLpmNHQvOA3XdXf0AoxTvNaw6U3w2EhPAKF68hPKt9rByoXA7hZbseglAISBJ5nlG/QdlFxe1bx9gZhjPUcMAYGj5dBOecA2ecwfa23eHbbyPTvhjoAsrsQJmJEVDV6+GYYwhu26Gy3ngDflhGsKQY/vvfiKgzOy2phtTyOYSn4+dI5OkopyXiVwjPqSyrcsWBEoT8Jm8E1PXXX8+oUaOoqKigXbt2ltsEAoG4vyeffDJmm7fffpt9992X0tJS+vXrx4MPPhhXzt13303v3r0pKytjxIgRzJs3Lw3fKDkMB6q4Y9uY5REB1QANZW3VE7KsjO3jj6eZANx0kwqn7cAqB6qkJIGACocp/uOlsHw5wT69oXMXCCtrJnDqKRG3CeyndkmWbDlQZgGVqgPldwjPyoGya9CFxKQrhJdKOX4i94Ug+EfeCKj6+nomTpzI+eef77jdAw88wOrVqyN/EyZMiKz7/vvvOfrooznkkENYuHAhF110EWeffTavvfZaZJuZM2cyZcoUrrrqKhYsWMDee+/NuHHj+Omnn9L11TzRsHlHCK9zZczyGAGlD4y5xx5sP2qiWnjeedDcTH19bJjOCE8VF2tRvtpa5S59+SWNDTvGc/r4Y4rfeBlKSgjc9y8YNw4IQLfuBCccF1Mfs1DItIBK10CauRrC8+J8CLF4DeH55UBlAxFQguAfeZMDdfXVVwNYOkY67dq1o1u3bpbrpk+fTp8+fbj11lsBGDBgAO+//z63334748aNA+C2225j8uTJnHnmmZF9XnrpJWbMmMHll1/u07dJnvrNtQAUd24XszwyFlS9SUAFAmy/9lZav/MSvP8+3HsvGyacZ1l2ce0WKt5/F1buCh98AD+uAKDxyyD1bYfA3PkqeX3aNIKDBsKuwP/9H7RtS6A49lZKtwOVTwNp6mRiKhfjs9deYS0VPxLEEwmTXEsiFwQhdfLGgXLLBRdcQKdOnRg+fDgzZswgrCWvzJ07l7Fjx8ZsP27cOObOnQsol2v+/Pkx2wSDQcaOHRvZxoq6ujqqq6tj/tJBOAwN1WqSOzsB1dAQPwzBtvY7w7XXqg8XXcT6tz5Tk+S9+CI8+qjqkjdnDiW/Pp2K266FmU/CjysoKQ1CMETjF1/RMPdjdZzzzoaLLooKmM6dobQ0oWDKdwfKrxBeOh0oq/eCNySEJwiCW/LGgXLDNddcw6GHHkpFRQWvv/46//d//8fWrVv57W9/C0BVVRVdu3aN2adr165UV1dTU1PDxo0baWpqstzmq6++sj3utGnTIg5ZOmlqgvA2NdJlcdcOMetsQ3jsyHH63e/URMP/+Q/bTpkMHEmQZpoJwl13QbiZYjZT3ncnWN8O6uspP+ck6pvbULf8PZo/2wz9+1N863AIBhImQyeabNgrueJA+S2gUq2nWTQ59QQTrMlWEnk2EAElCP6RVQF1+eWXc+ONNzpus3jxYvbYMd5RIq688srI+yFDhrBt2zZuvvnmiIBKF1dccQVTpkyJfK6urqZnz56+H6ehgchQ4WYBZUyxYiWgtm9HtdyPPAIHHUTN/9Qkvu127cSG7WVqsMtQEcVTLqbi18fAE6qVL+8Gm6ugZo+fwT6qrGKbqVwy6UC5GYk8V3vhpTOEl+icCNYkkwNlvBqC1e58m0VttkWV3BeC4B9ZFVCXXHIJZ5xxhuM2ffv2Tbr8ESNGcO2111JXV0dpaSndunVjzZo1MdusWbOGtm3bUl5eTigUIhQKWW5jl1cFUFpaSmlpadL1dEt9PbBtGyGaCHXtFLMuoQMF0KYNfPQRNU9shMWVVA4sZ8OSBvjf/6BXL0pO6ExF6+h+RkJ5jYoaEgxGBYBVQ6FTaL3w8iGJXEJ4qeNGQHktz058i4AShPwmqwKqc+fOdO4cP82IXyxcuJD27dtHxM3IkSN5+eWXY7aZNWsWI0eOBKCkpIShQ4cye/bsSO+95uZmZs+ezYUXXpi2errFcKCKaVC5RxpOAipmfrvSUmoqu0E5tGu3Y8dhwyJlaGNtRoY0MARUcXH0wZtNB8ocwjPep/PXfbpyoNIp+vwss5BJJoRnvLpxoKxes0Wu1EMQCoG8yYFavnw5GzZsYPny5TQ1NbFw4UIA+vXrR+vWrXnxxRdZs2YN+++/P2VlZcyaNYsbbriB3//+95EyzjvvPO666y4uu+wyzjrrLN58802eeuopXnrppcg2U6ZMYdKkSQwbNozhw4dzxx13sG3btkivvGzS0ABs20YJ9a4EVJs2apRxY6BMA0MQmYfTMo8DZYgpIyk9MuceiXN50j2QZqZDeLnaC8/OdZKG0j1eRJPd9onOt4TwBKHwyBsBNXXqVB566KHI5yFDhgDw1ltvMWbMGIqLi7n77ru5+OKLCYfD9OvXLzIkgUGfPn146aWXuPjii7nzzjvZeeedue+++yJDGACcdNJJrF27lqlTp1JVVcU+++zDq6++GpdYng3qa5qgpkY5UF26xKyzElAdOyoBZUzVYmAIqMrYoaQoLobS0h3z6jXHulH6MSCxA1VoIbxM9MJLBgnh+YfTeXO635zOtdM22RRQgiCkTt4IqAcffNBxDKjx48czfvz4hOWMGTOGTz/91HGbCy+8MCdCdmYa1m4CwkpA7ZgyxcBqHKiOHWHZsvi57ZwEVCCghNO2bc4CKlEOVCZDeFaNQr4MpJmOJPJsuxz5RrKujJ14tdrG7nOmEQdKEPyj4MaBKmQa1mwAoLhVCRTFal87BwqUA6WHoAwBVVER7b0H0feGsLISWAYt1YFKFafzICG87ODFSbLbLpkcKAnhCUJ+kzcOlGA4UFDSpixundUwBu3aqQa7qUmJqMpKJQTMAqq+Xn02BNJxx8Hq1bDLLrHHaNMm+t5rDlS+D6Rp99lrGX47UIlCQ9JQusergMrnEJ7cF4KQOuJA5RH1azcDUFwZP+OvlQNVUhIVPUYYr74+6kaVlamcJ3MZ3brBkCFxJldM3rpXB8rP3B+7EF46HahUytX3ycQ4UJID5Q2vosJKQOVLCM8gV+ohCPmMCKg8wnCg3Aqo4mJo21a936y0V8yQBMXF1iE8A7OA6qQNPZXtHKhMh/D8IhNJ5OblgjNeQ3jJlCEOlCAUHiKg8oiGDao7XUn7VnHrrARUUVE0j8lwoAwBZSSIWzlQBk4OlLnBdsp5MjtGyWA+loTwYveRJPLUyUQIL9vXRgSUIPiH5EDlEXtfegQ9x/5Aqx7t4tYlcqCSEVDmh2yn2MHPCQZVfpXVtrpQSNV9MpdviIVAIPPTZKRaroTwco+WFMLL9vEFoZAQAZVHVA7sQeXAHpbrkhVQRthOH2XcDnOITxdQTg5UOgSU8WonoNLlQKVaRrqmcrFbLg1mYryG8JzCcXaCSUJ4glB4SAivQNDHgdJHDjdCeOYcKLMDZXafzOg98AycGhWrqVZSwepY5ilc8iGJXMaByl285kBJCE8QWjbiQBUIhgCqq4td5taBMrtLZszTvoCzSMqUA2X1alUfP45r9dkrfpwLHbswkjSQ7slECC9Xrkeu1UcQ8hlxoAoEKwGkC6itW1W4LVkHKpGAynQOlNOrVX38OK4fZWQiidy8XHAmGScpmTLM68WBEoT8RgRUgWAWQKGQaqxbtVLvw2E1mKZTDpQVxpR7Q4fGr8u2A+UUwsslB0pCePmBVwElITxBaNlICK9ACAajkwBDVBAFAtChA6xdC1VV9g6UXQjvzDNh0ybo3j1+XbZzoLLhQKUzhOdXErlTSEmIx8u5srsP8yWEV7FjCDnzPJeCIHhHBFSBEAgo0WTkQOmOUp8+SkB99128gNp5ZyWi+vSxLre83P5hm8sOlITwvJfVUvHiJHktw+lzNq7R0KHKld5998wfWxAKDQnhFRDG5MEQK6D69lWvVgKqSxf4wx9g9Gjvx3PKgcpkErlZSJnr5tdxrT57LcNpHKhUkBBe8vjhQLktO9vXpqQEBg9W0zgJgpAaIqAKCN1F0gVU797qYb1unXKiIDq8ASQvNpwcqHQmkds5T5lonFItN13jQJkFVK6FjvKBdCSRy3UQhMJFBFQBYThNEDsNS1kZ9Ngx/mY4DIMGqbyoVMm2A5WNJPJUy8hUErk02O5JxkkyL8uXEJ4gCP4hAqqA2GWX6PstW2LXGeKquBiOOMKf4zkJA0kit97H76lc3Dojgj3JukTJhv6clgmCkD9IEnkBoYftjIEzDfbdV+VADR8eHRsqVbLtQDkJqFzKgdJJ51QuEsJLjUyE8OR6CELhIAKqwNh9d1iyBNq3j13erh2cfba/x8pkDpSOmxBeuhyoVMtwCuGlegwJ4SVHJkJ4uZJELgiCf4iAKjCOPx7eeQf22Sf9xxIHynsZ6RoHyk40SSOdmGRDcV6ElyAIhYcIqAKjrAzGjcvMsdzmQBXSVC7pCOEFAiq53y8BJSG85EhFCIkDJQgtD0kiF5Im0Vx45lBbKlg1PPkSwtOxE1CpHsvuu0sjnZhMhvCs1gmCkJ+IgBKSJlFPO8N5aukhPGN6HUjvVC4inJIjnSE8p/0FQchvREAJSePkQOnr83kqF6d6uEUXUE4OVCr1kRBe6nh1oNyslxCeIBQuIqCEpEnk+KTLgXIzEnkuDaTpVkClKqSkkU6OZJPBJYQnCC0bEVBC0rh1oFr6QJrhcPR9OqdySbWslko6Q3hO96ggCPmNCCghabKdA2U1ibBTfVI9rt2xEqE7UImcDK8Yg6K2bSshvFRJZwjP6rNcG0HIb2QYAyFpsp0DlQ0HKhl0AeX3sY49FkaOhJ12khBeskgvPEEQkkEElJA02cqBMi/L9RBeIgGVSl1LS6MTRYtwSg4voeZke+GJIygIhYcIKCFpsuFA6b/2zQ1fJpLIU82BcjpGqo2rhPCSo08f2HNP2Gsv+20SncdkctHk2ghCfiMCSkgatzlQfiaRuwnlmd/7cdxUSGcIL1E50kgnpqwMJk503savEJ5cG0EoHCSJXEiaRA5UOkJ4bgVULjlQbkN4qTaoRTt+DoVC0jinC+mFJwiCgQgoIWkSCZZ0hfDM5Wfy130uh/AOPhhGjIBdd5UG22/86oXndp0gCLmPhPCEpGmJDlQyuA3hpUqvXupPRxppf0jUiUFCeILQ8hAHSkiaRDlQPXoo8dS1a+rHatsWunWDPfaIP2ah9MLzs0EVByo92J3PRD8W5HoIQuGRNwLq+uuvZ9SoUVRUVNCuXbu49Q8++CCBQMDy76effgLg7bfftlxfVVUVU9bdd99N7969KSsrY8SIEcybNy8TXzHvSORAjR0Lf/iDPwIqFIJzz4Vf/CL+mPmeA5VK2ZkoS0jsgB5wAOy7rwqfJtrPvE4QhPwkb0J49fX1TJw4kZEjR3L//ffHrT/ppJMYP358zLIzzjiD2tpaunTpErP866+/pq0xhDPErJ85cyZTpkxh+vTpjBgxgjvuuINx48bx9ddfx5XT0knkQAUCUFLi3/GyER7xo5xM5UA5lS34g52A6tNH/Tltb7e/IAj5Sd4IqKuvvhpQTpMV5eXllJeXRz6vXbuWN99801JsdenSxdLFArjtttuYPHkyZ555JgDTp0/npZdeYsaMGVx++eWW+9TV1VFXVxf5XF1d7eYr5T3ZbgwkhJfZMlsyyZ5HuQ6CULjkTQjPKw8//DAVFRWccMIJcev22WcfunfvzuGHH84HH3wQWV5fX8/8+fMZO3ZsZFkwGGTs2LHMnTvX9ljTpk2jsrIy8tezZ09/v0yOksiBSjeZCOGZy01nCE/IXdx0YrCiZ08oLlb5e3b7C4KQnxSsgLr//vs55ZRTYlyp7t27M336dJ599lmeffZZevbsyZgxY1iwYAEA69ato6mpia6mpJ2uXbvG5UnpXHHFFWzevDnyt2LFivR8qRwjUQ5UuslUD6dUy8pGCE+cj/TgVUAdfLDKA+ze3f0+giDkB1kN4V1++eXceOONjtssXryYPfSuVy6YO3cuixcv5pFHHolZ3r9/f/r37x/5PGrUKL799ltuv/32uG29UFpaSmlpadL75yt6CC3XQnj55ECJgMp9OndWTpIhhLxQlDeJEoIgeCGr/9qXXHIJZ5xxhuM2ffv29Vzufffdxz777MPQoUMTbjt8+HDef/99ADp16kQoFGLNmjUx26xZs4ZuhgcvRMh2I11ZGfuaLgcqGISmpuTL7d4dli1LvJ2fdW7TJvZVSI127eDSS5WIMkjm/s/2/4wgCP6RVQHVuXNnOnfu7GuZW7du5amnnmLatGmutl+4cCHdd/ysLCkpYejQocyePZsJEyYA0NzczOzZs7nwwgt9rWch4GUW+3Rw0EGw++7x4RG/69SqFWzalPz+hxwCpaUwYID1+nQ0qkcdBcOGwc47+1dmS8fco7R3b1izJprf5AURUIKQ/+SNubx8+XI2bNjA8uXLaWpqYuHChQD069eP1q1bR7abOXMmjY2NnHbaaXFl3HHHHfTp04c999yT2tpa7rvvPt58801ef/31yDZTpkxh0qRJDBs2jOHDh3PHHXewbdu2SK88IYo5hJZpQiHYaafo53Q5UG3aRAVUMuWWlMCYMfbr03H+SktVArOQPg45BEaP9ibWRTgJQuGQNwJq6tSpPPTQQ5HPQ4YMAeCtt95ijNY63X///Rx//PGWwxTU19dzySWXsHLlSioqKhg8eDBvvPEGhxxySGSbk046ibVr1zJ16lSqqqrYZ599ePXVV+MSy4XsO1Bm0uVA6WGwdDaA0rjmH17vMwnhCULhkDcC6sEHH7QdA0pnzpw5tusuu+wyLrvssoRlXHjhhRKyc0GuNQbpdKDSSa6dRyH9yLUWhPwnR7wDIR/JZQcqXQJKHCghFeQaC0LhkCNNn5CPZDsHyky6QnjarD9pnW4lV86jkD7kWgtC4SACSkialuhApQNpTFsecs0FIf/JkaZPyEdy7de0JJELuY5cY0EoHERACUnTEh2ohgb/yjWXL4NethxESAlC/pM3vfCE3CPXcqAqKtTAmhUV/tZJH0Bx61b/yjWYOFGNM9Wxo/9lC7lFrrm2giAkjwgoIWly0YE65ZT0lGuwZYv/5bdqpf6EwkeEkyAUDjnS9An5SEv8NZ0OB0poebSk/xlBKFREQAlJk2sOVCZobs52DYR8piX+6BCEQqUFNX2C3+RaDpQgCIIgZAoRUELSVFbGvhYyRx2lXkeOzG49hPxGHChBKBwkiVxImi5d4PzzwWLe5oJjv/2gXz9o3z7bNRHymU6dYKedYJddsl0TQRBSRQSUkBJdu2a7BpkhEIAOHbJdCyHfKSqCc87Jdi0EQfADCeEJgiAIgiB4RASUIAiCIAiCR0RACYIgCIIgeEQElCAIgiAIgkdEQAmCIAiCIHhEBJQgCIIgCIJHREAJgiAIgiB4RASUIAiCIAiCR0RACYIgCIIgeEQElCAIgiAIgkdEQAmCIAiCIHhEBJQgCIIgCIJHREAJgiAIgiB4RASUIAiCIAiCR4qyXYFCJBwOA1BdXZ3lmgiCIAiC4Baj3TbacSdEQKWBLVu2ANCzZ88s10QQBEEQBK9s2bKFyspKx20CYTcyS/BEc3Mzq1atok2bNgQCAV/Lrq6upmfPnqxYsYK2bdv6WnahIefKPXKuvCHnyz1yrrwh58s96ThX4XCYLVu2sNNOOxEMOmc5iQOVBoLBIDvvvHNaj9G2bVv553KJnCv3yLnyhpwv98i58oacL/f4fa4SOU8GkkQuCIIgCILgERFQgiAIgiAIHhEBlWeUlpZy1VVXUVpamu2q5Dxyrtwj58obcr7cI+fKG3K+3JPtcyVJ5IIgCIIgCB4RB0oQBEEQBMEjIqAEQRAEQRA8IgJKEARBEATBIyKgBEEQBEEQPCICKo+4++676d27N2VlZYwYMYJ58+Zlu0pZ5y9/+QuBQCDmb4899oisr62t5YILLqBjx460bt2aX/ziF6xZsyaLNc4s7777Lsceeyw77bQTgUCA559/PmZ9OBxm6tSpdO/enfLycsaOHcvSpUtjttmwYQOnnnoqbdu2pV27dvz6179m69atGfwWmSHRuTrjjDPi7rXx48fHbNNSztW0adPYb7/9aNOmDV26dGHChAl8/fXXMdu4+d9bvnw5Rx99NBUVFXTp0oVLL72UxsbGTH6VtOPmXI0ZMybu3jrvvPNitmkJ5wrgnnvuYfDgwZHBMUeOHMkrr7wSWZ9L95UIqDxh5syZTJkyhauuuooFCxaw9957M27cOH766adsVy3r7LnnnqxevTry9/7770fWXXzxxbz44os8/fTTvPPOO6xatYrjjz8+i7XNLNu2bWPvvffm7rvvtlx/00038be//Y3p06fz0Ucf0apVK8aNG0dtbW1km1NPPZUvvviCWbNm8d///pd3332Xc845J1NfIWMkOlcA48ePj7nXnnjiiZj1LeVcvfPOO1xwwQV8+OGHzJo1i4aGBo444gi2bdsW2SbR/15TUxNHH3009fX1zJkzh4ceeogHH3yQqVOnZuMrpQ035wpg8uTJMffWTTfdFFnXUs4VwM4778xf//pX5s+fzyeffMKhhx7Kz372M7744gsgx+6rsJAXDB8+PHzBBRdEPjc1NYV32mmn8LRp07JYq+xz1VVXhffee2/LdZs2bQoXFxeHn3766ciyxYsXh4Hw3LlzM1TD3AEI//vf/458bm5uDnfr1i188803R5Zt2rQpXFpaGn7iiSfC4XA4/OWXX4aB8McffxzZ5pVXXgkHAoHwypUrM1b3TGM+V+FwODxp0qTwz372M9t9Wuq5CofD4Z9++ikMhN95551wOOzuf+/ll18OB4PBcFVVVWSbe+65J9y2bdtwXV1dZr9ABjGfq3A4HB49enT4d7/7ne0+LfVcGbRv3z5833335dx9JQ5UHlBfX8/8+fMZO3ZsZFkwGGTs2LHMnTs3izXLDZYuXcpOO+1E3759OfXUU1m+fDkA8+fPp6GhIea87bHHHuyyyy5y3oDvv/+eqqqqmPNTWVnJiBEjIudn7ty5tGvXjmHDhkW2GTt2LMFgkI8++ijjdc42b7/9Nl26dKF///6cf/75rF+/PrKuJZ+rzZs3A9ChQwfA3f/e3Llz2WuvvejatWtkm3HjxlFdXR1xGwoR87kyeOyxx+jUqRODBg3iiiuuYPv27ZF1LfVcNTU18eSTT7Jt2zZGjhyZc/eVTCacB6xbt46mpqaYGwKga9eufPXVV1mqVW4wYsQIHnzwQfr378/q1au5+uqrOeigg/j888+pqqqipKSEdu3axezTtWtXqqqqslPhHMI4B1b3lbGuqqqKLl26xKwvKiqiQ4cOLe4cjh8/nuOPP54+ffrw7bff8sc//pEjjzySuXPnEgqFWuy5am5u5qKLLuKAAw5g0KBBAK7+96qqqizvPWNdIWJ1rgD+v707jWkqa+MA/i9ga8HBom1s3QgIorgQRSV1S2aqDHyYqOGDIpq6RCOKMYp74h6DmShxMh+YmIwao5GMZoiJ+4KQiIobCCqSVIvEpC5RQRBBsc/7YV5v3iu49B1Kkf5/yU3ae849PefJueTh3nPbmTNnIjw8HL1790ZZWRnWrFmDyspK/P333wD8L1bl5eWwWq1obGxEt27dkJeXh9jYWJSWlnaoecUEir5rycnJyuvhw4cjISEB4eHh+Ouvv6DX633YM+psZsyYobweNmwYhg8fjgEDBqCgoAA2m82HPfOtJUuW4M6dO6q1h9S6z8Xqf9fJDRs2DBaLBTabDQ8ePMCAAQPau5s+FxMTg9LSUtTW1uLo0aOw2+0oLCz0dbda4C2874DRaERgYGCLJw2ePn0Ks9nso151TAaDAQMHDoTD4YDZbMa7d+9QU1OjqsO4/eNjDL40r8xmc4sHFZqbm/Hy5Uu/j2FkZCSMRiMcDgcA/4xVRkYGjh8/josXL6Jv377K/m8598xmc6tz72NZZ/O5WLUmISEBAFRzy59ipdVqERUVhfj4eGRlZSEuLg6//fZbh5tXTKC+A1qtFvHx8bhw4YKyz+1248KFC7BarT7sWcdTX1+PBw8ewGKxID4+Hl26dFHFrbKyEtXV1YwbgIiICJjNZlV8Xr9+jeLiYiU+VqsVNTU1uHnzplInPz8fbrdb+SPvrx4/fowXL17AYrEA8K9YiQgyMjKQl5eH/Px8REREqMq/5dyzWq0oLy9XJZ3nzp1DaGgoYmNj22cg7eBrsWpNaWkpAKjmlj/E6nPcbjeampo63rxq0yXp5DW5ubmi0+lk//79cu/ePVm4cKEYDAbVkwb+KDMzUwoKCsTpdEpRUZFMmjRJjEajPHv2TEREFi1aJP3795f8/Hy5ceOGWK1WsVqtPu51+6mrq5OSkhIpKSkRAJKdnS0lJSXy6NEjERHZsWOHGAwGOXbsmJSVlcmUKVMkIiJC3r59q7SRlJQkI0aMkOLiYrl06ZJER0dLamqqr4bkNV+KVV1dnaxcuVKuXLkiTqdTzp8/LyNHjpTo6GhpbGxU2vCXWKWnp0v37t2loKBAXC6XsjU0NCh1vnbuNTc3y9ChQyUxMVFKS0vl9OnTYjKZZN26db4Yktd8LVYOh0O2bt0qN27cEKfTKceOHZPIyEiZOHGi0oa/xEpEZO3atVJYWChOp1PKyspk7dq1otFo5OzZsyLSseYVE6jvyO+//y79+/cXrVYrY8aMkatXr/q6Sz43ffp0sVgsotVqpU+fPjJ9+nRxOBxK+du3b2Xx4sUSFhYmwcHBMm3aNHG5XD7scfu6ePGiAGix2e12Efnnqww2bNggvXr1Ep1OJzabTSorK1VtvHjxQlJTU6Vbt24SGhoqc+fOlbq6Oh+Mxru+FKuGhgZJTEwUk8kkXbp0kfDwcFmwYEGLf2D8JVatxQmA7Nu3T6nzLedeVVWVJCcni16vF6PRKJmZmfL+/ft2Ho13fS1W1dXVMnHiROnRo4fodDqJioqSVatWSW1traodf4iViMi8efMkPDxctFqtmEwmsdlsSvIk0rHmlUZEpG2vaRERERF1blwDRUREROQhJlBEREREHmICRUREROQhJlBEREREHmICRUREROQhJlBEREREHmICRUREROQhJlBEREREHmICRUT0X1VVVdBoNMpvkXnDnDlzMHXqVK+1T0TtgwkUEXUac+bMgUajabElJSV90/H9+vWDy+XC0KFDvdxTIvreBfm6A0REbSkpKQn79u1T7dPpdN90bGBgIMxmsze6RUSdDK9AEVGnotPpYDabVVtYWBgAQKPRICcnB8nJydDr9YiMjMTRo0eVYz+9hffq1SukpaXBZDJBr9cjOjpalZyVl5fjp59+gl6vR8+ePbFw4ULU19cr5R8+fMCKFStgMBjQs2dPrF69Gp/+/Kjb7UZWVhYiIiKg1+sRFxen6hMRdUxMoIjIr2zYsAEpKSm4ffs20tLSMGPGDFRUVHy27r1793Dq1ClUVFQgJycHRqMRAPDmzRv8/PPPCAsLw/Xr13HkyBGcP38eGRkZyvG7du3C/v37sXfvXly6dAkvX75EXl6e6jOysrJw4MAB/PHHH7h79y6WL1+OWbNmobCw0HtBIKJ/T4iIOgm73S6BgYESEhKi2rZv3y4iIgBk0aJFqmMSEhIkPT1dREScTqcAkJKSEhER+eWXX2Tu3LmtftaePXskLCxM6uvrlX0nTpyQgIAAefLkiYiIWCwW+fXXX5Xy9+/fS9++fWXKlCkiItLY2CjBwcFy+fJlVdvz58+X1NTU/z8QROR1XANFRJ3Kjz/+iJycHNW+Hj16KK+tVquqzGq1fvapu/T0dKSkpODWrVtITEzE1KlTMXbsWABARUUF4uLiEBISotQfN24c3G43Kisr0bVrV7hcLiQkJCjlQUFBGDVqlHIbz+FwoKGhAZMnT1Z97rt37zBixAjPB09E7YYJFBF1KiEhIYiKimqTtpKTk/Ho0SOcPHkS586dg81mw5IlS7Bz5842af/jeqkTJ06gT58+qrJvXfhORL7BNVBE5FeuXr3a4v3gwYM/W99kMsFut+PgwYPYvXs39uzZAwAYPHgwbt++jTdv3ih1i4qKEBAQgJiYGHTv3h0WiwXFxcVKeXNzM27evKm8j42NhU6nQ3V1NaKiolRbv3792mrIROQFvAJFRJ1KU1MTnjx5otoXFBSkLP4+cuQIRo0ahfHjx+PQoUO4du0a/vzzz1bb2rhxI+Lj4zFkyBA0NTXh+PHjSrKVlpaGTZs2wW63Y/PmzXj+/DmWLl2K2bNno1evXgCAZcuWYceOHYiOjsagQYOQnZ2Nmpoapf0ffvgBK1euxPLly+F2uzF+/HjU1taiqKgIoaGhsNvtXogQEbUFJlBE1KmcPn0aFotFtS8mJgb3798HAGzZsgW5ublYvHgxLBYLDh8+jNjY2Fbb0mq1WLduHaqqqqDX6zFhwgTk5uYCAIKDg3HmzBksW7YMo0ePRnBwMFJSUpCdna0cn5mZCZfLBbvdjoCAAMybNw/Tpk1DbW2tUmfbtm0wmUzIysrCw4cPYTAYMHLkSKxfv76tQ0NEbUgj8smXkhARdVIajQZ5eXn8KRUi+te4BoqIiIjIQ0ygiIiIiDzENVBE5De4YoGI2gqvQBERERF5iAkUERERkYeYQBERERF5iAkUERERkYeYQBERERF5iAkUERERkYeYQBERERF5iAkUERERkYf+A5sCA6iZh7MhAAAAAElFTkSuQmCC",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "show_curve(avg_history, scores)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "UOTTN7wKZDiL"
      },
      "source": [
        "This graph shows that the problem could be solved using 300 episodes. However, from episode 210 onwards, the results are very similar and the training stabilises. Although the agent has managed to learn to solve the problem in a reduced number of episodes, the learning line that the reward shows is quite unstable and has peaks. This may be due to the fact that the parameters used are not the most optimal."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "id": "Gi54iOb_Dkp3",
        "outputId": "3dfb0725-6e77-4271-dd19-72e6a2efedb2"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/record_video.py:78: UserWarning: \u001b[33mWARN: Overwriting existing videos at /content/video folder (try specifying a different `video_folder` for the `RecordVideo` wrapper if this is not desired)\u001b[0m\n",
            "  logger.warn(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Loading models ...\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<video alt=\"test\" autoplay \n",
              "                loop controls style=\"height: 400px;\">\n",
              "                <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\" />\n",
              "             </video>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "test_behavior(agent)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "rsxvJ5U3ZGZA"
      },
      "source": [
        "This video shows how the agent has learned to use inertia effectively to achieve the fastest possible vertical positioning. Moreover, he manages to stay in that position, which is the objective of the problem."
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "p_03BMxoTWgl"
      },
      "source": [
        "# **2-LunarLander**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JgbOh9h5TVkI"
      },
      "outputs": [],
      "source": [
        "ENV='LunarLanderContinuous-v2'\n",
        "LR_ACTOR=0.001\n",
        "LR_CRITIC=0.002\n",
        "GAMMA=0.99\n",
        "TAU=0.005\n",
        "NOISE=0.1\n",
        "BATCH_SIZE=128\n",
        "L1_SIZE=512\n",
        "L2_SIZE=512\n",
        "UPDATE_ACTOR_INTERVAL=2\n",
        "N_ITERATIONS=400\n",
        "MAX_SIZE=100000\n",
        "EXP_NAME='lunar'\n",
        "SAVE_POINTS=[100, 200, 250, 300, 350, 400]"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "BzmyEcLqZKu5"
      },
      "source": [
        "This video shows how the agent has learned to use inertia effectively to achieve the fastest possible vertical positioning. Moreover, he manages to stay in that position, which is the objective of the problem."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 476
        },
        "id": "UQ4BuyZZ5ZOY",
        "outputId": "92b617a7-3c67-41f1-b863-e047b03bb4d1"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/record_video.py:78: UserWarning: \u001b[33mWARN: Overwriting existing videos at /content/video folder (try specifying a different `video_folder` for the `RecordVideo` wrapper if this is not desired)\u001b[0m\n",
            "  logger.warn(\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<video alt=\"test\" autoplay \n",
              "                loop controls style=\"height: 400px;\">\n",
              "                <source src=\"data:video/mp4;base64,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\" type=\"video/mp4\" />\n",
              "             </video>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "random_behavior()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UzKT0Rut5c7L",
        "outputId": "8828e0f7-e15f-4232-99d5-6745a8e0ca63"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/core.py:317: DeprecationWarning: \u001b[33mWARN: Initializing wrapper in old step API which returns one bool instead of two. It is recommended to set `new_step_api=True` to use new step API. This will be the default behaviour in future.\u001b[0m\n",
            "  deprecation(\n",
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/step_api_compatibility.py:39: DeprecationWarning: \u001b[33mWARN: Initializing environment in old step API which returns one bool instead of two. It is recommended to set `new_step_api=True` to use new step API. This will be the default behaviour in future.\u001b[0m\n",
            "  deprecation(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "# Episode: 0, Reward: -317.00588299169215, Mean reward: -317.00588299169215.\n",
            "# Episode: 1, Reward: -421.27300686331415, Mean reward: -369.1394449275032.\n",
            "# Episode: 2, Reward: -604.968276132241, Mean reward: -447.74905532908247.\n",
            "# Episode: 3, Reward: -1057.7104739971176, Mean reward: -600.2394099960912.\n",
            "# Episode: 4, Reward: -486.55860873379964, Mean reward: -577.503249743633.\n",
            "# Episode: 5, Reward: -449.35420010037814, Mean reward: -556.1450748030904.\n",
            "# Episode: 6, Reward: -713.9958220688386, Mean reward: -578.6951815553401.\n",
            "# Episode: 7, Reward: -710.742066617938, Mean reward: -595.2010421881649.\n",
            "# Episode: 8, Reward: -623.9746857697377, Mean reward: -598.3981136972285.\n",
            "# Episode: 9, Reward: -671.0811449407813, Mean reward: -605.6664168215838.\n",
            "# Episode: 10, Reward: -210.4783305659402, Mean reward: -569.7402271619799.\n",
            "# Episode: 11, Reward: -115.0221911868141, Mean reward: -531.8470574973827.\n",
            "# Episode: 12, Reward: -133.72554171374398, Mean reward: -501.22232551402584.\n",
            "# Episode: 13, Reward: -81.19858847003853, Mean reward: -471.2206300108839.\n",
            "# Episode: 14, Reward: -89.88899170601356, Mean reward: -445.7985207905592.\n",
            "# Episode: 15, Reward: -75.629684862501, Mean reward: -422.6629685450556.\n",
            "# Episode: 16, Reward: -47.59158938874976, Mean reward: -400.5999462417435.\n",
            "# Episode: 17, Reward: -362.57820036492086, Mean reward: -398.48762702636446.\n",
            "# Episode: 18, Reward: -232.28965787373116, Mean reward: -389.7403654920153.\n",
            "# Episode: 19, Reward: -301.12135991218855, Mean reward: -385.30941521302395.\n",
            "# Episode: 20, Reward: -139.61145658613535, Mean reward: -373.60951242126737.\n",
            "# Episode: 21, Reward: -244.97229041626318, Mean reward: -367.7623659664945.\n",
            "# Episode: 22, Reward: -166.72829269120786, Mean reward: -359.0217540849603.\n",
            "# Episode: 23, Reward: -224.17194085115608, Mean reward: -353.40301186688515.\n",
            "# Episode: 24, Reward: -289.86182566063394, Mean reward: -350.86136441863505.\n",
            "# Episode: 25, Reward: -235.20180660104913, Mean reward: -346.41291988718945.\n",
            "# Episode: 26, Reward: -367.93251797241226, Mean reward: -347.20994203849403.\n",
            "# Episode: 27, Reward: -379.4019851711283, Mean reward: -348.3596578646596.\n",
            "# Episode: 28, Reward: -169.25890259442176, Mean reward: -342.1837697518927.\n",
            "# Episode: 29, Reward: -286.11224387221887, Mean reward: -340.314718889237.\n",
            "# Episode: 30, Reward: -276.0166249463987, Mean reward: -338.24058682656477.\n",
            "# Episode: 31, Reward: -127.26372768306308, Mean reward: -331.64755997833026.\n",
            "# Episode: 32, Reward: -218.82943600506536, Mean reward: -328.2288289488374.\n",
            "# Episode: 33, Reward: -157.1201143117069, Mean reward: -323.19621969480414.\n",
            "# Episode: 34, Reward: -165.93905716458718, Mean reward: -318.7031579082265.\n",
            "# Episode: 35, Reward: -115.16716258118797, Mean reward: -313.04938026025326.\n",
            "# Episode: 36, Reward: -314.57341126593565, Mean reward: -313.0905702874339.\n",
            "# Episode: 37, Reward: -175.87203125733163, Mean reward: -309.47955610243116.\n",
            "# Episode: 38, Reward: -32.73904339271533, Mean reward: -302.38364552013076.\n",
            "# Episode: 39, Reward: -314.37523625977497, Mean reward: -302.6834352886218.\n",
            "# Episode: 40, Reward: -238.72495068979748, Mean reward: -301.1234722496261.\n",
            "# Episode: 41, Reward: -246.71338178788068, Mean reward: -299.8279939052988.\n",
            "# Episode: 42, Reward: -142.71648974386844, Mean reward: -296.1742379945679.\n",
            "# Episode: 43, Reward: -228.96177657828173, Mean reward: -294.64668205328866.\n",
            "# Episode: 44, Reward: -275.84604190453894, Mean reward: -294.22889004998314.\n",
            "# Episode: 45, Reward: -105.5486470413312, Mean reward: -290.12714563675155.\n",
            "# Episode: 46, Reward: -155.06039973229124, Mean reward: -287.2533850855928.\n",
            "# Episode: 47, Reward: -448.40794088626, Mean reward: -290.6107716647734.\n",
            "# Episode: 48, Reward: -201.38060398542066, Mean reward: -288.78974783458256.\n",
            "# Episode: 49, Reward: -183.42643902222156, Mean reward: -286.6824816583353.\n",
            "# Episode: 50, Reward: -546.7020498051294, Mean reward: -291.78090456317443.\n",
            "# Episode: 51, Reward: -236.66532645517492, Mean reward: -290.72098959955906.\n",
            "# Episode: 52, Reward: -246.8673739733887, Mean reward: -289.8935628896313.\n",
            "# Episode: 53, Reward: -98.65727527518754, Mean reward: -286.3521501560305.\n",
            "# Episode: 54, Reward: -56.777400273285444, Mean reward: -282.17806379452605.\n",
            "# Episode: 55, Reward: 152.86192746380277, Mean reward: -274.40949252205587.\n",
            "# Episode: 56, Reward: 170.5804837854585, Mean reward: -266.60265083245037.\n",
            "# Episode: 57, Reward: -127.70245529352314, Mean reward: -264.20781987488266.\n",
            "# Episode: 58, Reward: -184.81219086149733, Mean reward: -262.86213124753715.\n",
            "# Episode: 59, Reward: -123.86765957380311, Mean reward: -260.5455567196416.\n",
            "# Episode: 60, Reward: -336.4711329338211, Mean reward: -261.7902382969232.\n",
            "# Episode: 61, Reward: -355.6750866516587, Mean reward: -263.30451004458024.\n",
            "# Episode: 62, Reward: -54.75907280799288, Mean reward: -259.99426500907884.\n",
            "# Episode: 63, Reward: -353.783701506248, Mean reward: -261.4597249543471.\n",
            "# Episode: 64, Reward: -228.76561799117354, Mean reward: -260.9567386933752.\n",
            "# Episode: 65, Reward: -167.87566756776886, Mean reward: -259.5464194338963.\n",
            "# Episode: 66, Reward: -419.38995359459733, Mean reward: -261.9321438243545.\n",
            "# Episode: 67, Reward: -325.0313670566028, Mean reward: -262.8600735777699.\n",
            "# Episode: 68, Reward: -486.50601610806166, Mean reward: -266.1013191216872.\n",
            "# Episode: 69, Reward: -291.2265270949895, Mean reward: -266.46025066416297.\n",
            "# Episode: 70, Reward: -289.680549562538, Mean reward: -266.78729712752033.\n",
            "# Episode: 71, Reward: 175.71411730779232, Mean reward: -260.6414441492521.\n",
            "# Episode: 72, Reward: 11.931092203256064, Mean reward: -256.90757378825884.\n",
            "# Episode: 73, Reward: 242.90891015494174, Mean reward: -250.15329697821554.\n",
            "# Episode: 74, Reward: 232.98387219362706, Mean reward: -243.71146805592429.\n",
            "# Episode: 75, Reward: 211.7046910453925, Mean reward: -237.71915017301222.\n",
            "# Episode: 76, Reward: 36.5852169085756, Mean reward: -234.15675579532927.\n",
            "# Episode: 77, Reward: -333.35058732152856, Mean reward: -235.42847158412673.\n",
            "# Episode: 78, Reward: 217.5386096076324, Mean reward: -229.69471106271203.\n",
            "# Episode: 79, Reward: 240.7859618303707, Mean reward: -223.8137026515486.\n",
            "# Episode: 80, Reward: 202.99102732384532, Mean reward: -218.54450845432152.\n",
            "# Episode: 81, Reward: 248.19823012440432, Mean reward: -212.85252383750776.\n",
            "# Episode: 82, Reward: -33.01594734646966, Mean reward: -210.68581809665187.\n",
            "# Episode: 83, Reward: 244.24550967468204, Mean reward: -205.26996895651695.\n",
            "# Episode: 84, Reward: 255.76195238327594, Mean reward: -199.8460639995782.\n",
            "# Episode: 85, Reward: 237.10123724620445, Mean reward: -194.76528142695287.\n",
            "# Episode: 86, Reward: 242.92959767116082, Mean reward: -189.73430580513545.\n",
            "# Episode: 87, Reward: -37.28279937294455, Mean reward: -188.00190232295145.\n",
            "# Episode: 88, Reward: 220.2876610100787, Mean reward: -183.41437913943426.\n",
            "# Episode: 89, Reward: 232.89982340643212, Mean reward: -178.78866577781352.\n",
            "# Episode: 90, Reward: 230.41586728563445, Mean reward: -174.29191266722617.\n",
            "# Episode: 91, Reward: 271.70257713762635, Mean reward: -169.44414647369516.\n",
            "# Episode: 92, Reward: 159.09021989072363, Mean reward: -165.91151887837884.\n",
            "# Episode: 93, Reward: -587.2285801806544, Mean reward: -170.39361527521154.\n",
            "# Episode: 94, Reward: -434.8888466022839, Mean reward: -173.1777756049702.\n",
            "# Episode: 95, Reward: -208.30058648677038, Mean reward: -173.5436382183223.\n",
            "# Episode: 96, Reward: -81.27143005613809, Mean reward: -172.59237834036165.\n",
            "# Episode: 97, Reward: 208.83737692981504, Mean reward: -168.70023798046188.\n",
            "# Episode: 98, Reward: 207.90768267362694, Mean reward: -164.8961175698145.\n",
            "# Episode: 99, Reward: 206.94447625368548, Mean reward: -161.17771163157948.\n",
            "Saving models ...\n",
            "# Episode: 100, Reward: -294.8118310070248, Mean reward: -160.95577111173282.\n",
            "# Episode: 101, Reward: 206.1072681600317, Mean reward: -154.68196836149937.\n",
            "# Episode: 102, Reward: 281.76666741276813, Mean reward: -145.81461892604926.\n",
            "# Episode: 103, Reward: 171.83950022456332, Mean reward: -133.5191191838325.\n",
            "# Episode: 104, Reward: -447.82758234583775, Mean reward: -133.13180891995285.\n",
            "# Episode: 105, Reward: 173.61773928922923, Mean reward: -126.90208952605677.\n",
            "# Episode: 106, Reward: 199.91420035415672, Mean reward: -117.76298930182682.\n",
            "# Episode: 107, Reward: -455.9848803947525, Mean reward: -115.21541743959497.\n",
            "# Episode: 108, Reward: -159.26472612356457, Mean reward: -110.56831784313323.\n",
            "# Episode: 109, Reward: -187.845261921528, Mean reward: -105.73595901294073.\n",
            "# Episode: 110, Reward: -242.79442341921637, Mean reward: -106.0591199414735.\n",
            "# Episode: 111, Reward: -314.4964205387612, Mean reward: -108.05386223499296.\n",
            "# Episode: 112, Reward: -60.8710872578847, Mean reward: -107.32531769043435.\n",
            "# Episode: 113, Reward: -7.259810048931579, Mean reward: -106.5859299062233.\n",
            "# Episode: 114, Reward: 184.95886691311728, Mean reward: -103.83745132003197.\n",
            "# Episode: 115, Reward: -90.83522507636904, Mean reward: -103.98950672217066.\n",
            "# Episode: 116, Reward: 286.72671356718456, Mean reward: -100.64632369261132.\n",
            "# Episode: 117, Reward: 219.7476168848529, Mean reward: -94.82306552011357.\n",
            "# Episode: 118, Reward: -84.70205682639795, Mean reward: -93.34718950964023.\n",
            "# Episode: 119, Reward: -91.77373584181716, Mean reward: -91.25371326893654.\n",
            "# Episode: 120, Reward: 13.176220112053258, Mean reward: -89.72583650195466.\n",
            "# Episode: 121, Reward: -66.04697513686762, Mean reward: -87.9365833491607.\n",
            "# Episode: 122, Reward: 3.813005172847312, Mean reward: -86.23117037052015.\n",
            "# Episode: 123, Reward: 30.814469329432995, Mean reward: -83.68130626871424.\n",
            "# Episode: 124, Reward: 286.0864878579562, Mean reward: -77.92182313352833.\n",
            "# Episode: 125, Reward: 224.3291649836283, Mean reward: -73.32651341768157.\n",
            "# Episode: 126, Reward: 188.05448458554355, Mean reward: -67.76664339210201.\n",
            "# Episode: 127, Reward: 159.90984700957924, Mean reward: -62.373525070294924.\n",
            "# Episode: 128, Reward: -431.3276905836456, Mean reward: -64.99421295018716.\n",
            "# Episode: 129, Reward: -99.7353391426771, Mean reward: -63.130443902891756.\n",
            "# Episode: 130, Reward: 284.7895911781392, Mean reward: -57.52238174164638.\n",
            "# Episode: 131, Reward: 287.22475144301734, Mean reward: -53.377496950385584.\n",
            "# Episode: 132, Reward: 38.62912747847153, Mean reward: -50.80291131555022.\n",
            "# Episode: 133, Reward: 207.1597110592462, Mean reward: -47.16011306184068.\n",
            "# Episode: 134, Reward: -131.39780175913504, Mean reward: -46.814700507786156.\n",
            "# Episode: 135, Reward: -18.536628226210482, Mean reward: -45.84839516423638.\n",
            "# Episode: 136, Reward: -153.70561202913686, Mean reward: -44.2397171718684.\n",
            "# Episode: 137, Reward: -42.18554434002289, Mean reward: -42.9028523026953.\n",
            "# Episode: 138, Reward: 230.62850385102794, Mean reward: -40.26917683025788.\n",
            "# Episode: 139, Reward: 106.98186390246234, Mean reward: -36.0556058286355.\n",
            "# Episode: 140, Reward: -490.31592767018685, Mean reward: -38.57151559843939.\n",
            "# Episode: 141, Reward: 174.3208201394026, Mean reward: -34.361173579166554.\n",
            "# Episode: 142, Reward: -992.5234913051908, Mean reward: -42.859243594779784.\n",
            "# Episode: 143, Reward: 137.29481921707668, Mean reward: -39.196677636826195.\n",
            "# Episode: 144, Reward: 256.8196031259344, Mean reward: -33.87002118652146.\n",
            "# Episode: 145, Reward: 253.78617669968978, Mean reward: -30.27667294911126.\n",
            "# Episode: 146, Reward: 258.96686613824136, Mean reward: -26.136400290405927.\n",
            "# Episode: 147, Reward: 245.45434909002148, Mean reward: -19.197777390643115.\n",
            "# Episode: 148, Reward: 185.10192373551217, Mean reward: -15.332952113433787.\n",
            "# Episode: 149, Reward: 216.08815764701916, Mean reward: -11.337806146741382.\n",
            "# Episode: 150, Reward: 262.7082925212612, Mean reward: -3.2437027234774747.\n",
            "# Episode: 151, Reward: -17.071248757718493, Mean reward: -1.047761946502916.\n",
            "# Episode: 152, Reward: 269.66544668569554, Mean reward: 4.11756626008793.\n",
            "# Episode: 153, Reward: 255.0585054663092, Mean reward: 7.654724067502901.\n",
            "# Episode: 154, Reward: -56.99143148946241, Mean reward: 7.652583755341132.\n",
            "# Episode: 155, Reward: 254.8883091065191, Mean reward: 8.67284757176829.\n",
            "# Episode: 156, Reward: 8.277436558748377, Mean reward: 7.049817099501188.\n",
            "# Episode: 157, Reward: 212.31946879089338, Mean reward: 10.450036340345354.\n",
            "# Episode: 158, Reward: 309.9273263881561, Mean reward: 15.39743151284189.\n",
            "# Episode: 159, Reward: 266.65468869406413, Mean reward: 19.302654995520562.\n",
            "# Episode: 160, Reward: 212.68603602941812, Mean reward: 24.794226685152953.\n",
            "# Episode: 161, Reward: -48.5525542343689, Mean reward: 27.86545200932585.\n",
            "# Episode: 162, Reward: 290.05430598167027, Mean reward: 31.313585797222476.\n",
            "# Episode: 163, Reward: 272.7546915875227, Mean reward: 37.57896972816019.\n",
            "# Episode: 164, Reward: 265.50561682465616, Mean reward: 42.52168207631849.\n",
            "# Episode: 165, Reward: -243.9581031707284, Mean reward: 41.7608577202889.\n",
            "# Episode: 166, Reward: 252.3366751188105, Mean reward: 48.47812400742298.\n",
            "# Episode: 167, Reward: -32.773974207651406, Mean reward: 51.4006979359125.\n",
            "# Episode: 168, Reward: 278.82211115923815, Mean reward: 59.053979208585496.\n",
            "# Episode: 169, Reward: 269.0617072553255, Mean reward: 64.65686155208864.\n",
            "# Episode: 170, Reward: -9.475556258604783, Mean reward: 67.45891148512797.\n",
            "# Episode: 171, Reward: 245.5797865843922, Mean reward: 68.15756817789398.\n",
            "# Episode: 172, Reward: 236.8983192798549, Mean reward: 70.40724044865996.\n",
            "# Episode: 173, Reward: -41.606044159923385, Mean reward: 67.5620909055113.\n",
            "# Episode: 174, Reward: 37.06755447036201, Mean reward: 65.60292772827866.\n",
            "# Episode: 175, Reward: -419.7883785847724, Mean reward: 59.28799703197699.\n",
            "# Episode: 176, Reward: 16.84711656387806, Mean reward: 59.09061602853002.\n",
            "# Episode: 177, Reward: 17.30878560858966, Mean reward: 62.59720975783121.\n",
            "# Episode: 178, Reward: -4.2117337575149065, Mean reward: 60.37970632417974.\n",
            "# Episode: 179, Reward: -47.694908703117946, Mean reward: 57.49489761884485.\n",
            "# Episode: 180, Reward: 261.5974883265053, Mean reward: 58.08096222887145.\n",
            "# Episode: 181, Reward: 284.753105782645, Mean reward: 58.44651098545386.\n",
            "# Episode: 182, Reward: 225.53281924544427, Mean reward: 61.03199865137301.\n",
            "# Episode: 183, Reward: 10.852422710704431, Mean reward: 58.698067781733215.\n",
            "# Episode: 184, Reward: -289.74253797593053, Mean reward: 53.24302287814116.\n",
            "# Episode: 185, Reward: 251.4986638303434, Mean reward: 53.38699714398254.\n",
            "# Episode: 186, Reward: 46.73856589171402, Mean reward: 51.42508682618807.\n",
            "# Episode: 187, Reward: 239.93110469505845, Mean reward: 54.1972258668681.\n",
            "# Episode: 188, Reward: 276.7142481677652, Mean reward: 54.761491738444974.\n",
            "# Episode: 189, Reward: -127.60809553775071, Mean reward: 51.15641254900314.\n",
            "# Episode: 190, Reward: 19.424365252973985, Mean reward: 49.04649752867654.\n",
            "# Episode: 191, Reward: 298.13554679850034, Mean reward: 49.31082722528527.\n",
            "# Episode: 192, Reward: -22.244239143173985, Mean reward: 47.497482634946294.\n",
            "# Episode: 193, Reward: 220.25239778932735, Mean reward: 55.57229241464611.\n",
            "# Episode: 194, Reward: 272.49360803494585, Mean reward: 62.646116961018414.\n",
            "# Episode: 195, Reward: 242.75664417451173, Mean reward: 67.15668926763124.\n",
            "# Episode: 196, Reward: 191.20155568710453, Mean reward: 69.88141912506367.\n",
            "# Episode: 197, Reward: -19.99854410992033, Mean reward: 67.59305991466631.\n",
            "# Episode: 198, Reward: 228.62552011930006, Mean reward: 67.80023828912304.\n",
            "# Episode: 199, Reward: 15.154113754629293, Mean reward: 65.88233466413247.\n",
            "Saving models ...\n",
            "# Episode: 200, Reward: 248.0983853946604, Mean reward: 71.31143682814931.\n",
            "# Episode: 201, Reward: 283.8547146422926, Mean reward: 72.08891129297193.\n",
            "# Episode: 202, Reward: -40.497258867174565, Mean reward: 68.86627203017251.\n",
            "# Episode: 203, Reward: 301.298402149618, Mean reward: 70.16086104942306.\n",
            "# Episode: 204, Reward: 288.767886772147, Mean reward: 77.5268157406029.\n",
            "# Episode: 205, Reward: -374.5865011634521, Mean reward: 72.04477333607609.\n",
            "# Episode: 206, Reward: -210.56895167313263, Mean reward: 67.93994181580318.\n",
            "# Episode: 207, Reward: 60.86013871968626, Mean reward: 73.10839200694758.\n",
            "# Episode: 208, Reward: 16.86631954274536, Mean reward: 74.86970246361068.\n",
            "# Episode: 209, Reward: 217.14643621053972, Mean reward: 78.91961944493136.\n",
            "# Episode: 210, Reward: 262.94303224839894, Mean reward: 83.97699400160751.\n",
            "# Episode: 211, Reward: 159.34787131348594, Mean reward: 88.71543692012997.\n",
            "# Episode: 212, Reward: 252.1848460837185, Mean reward: 91.84599625354602.\n",
            "# Episode: 213, Reward: 266.0463694989302, Mean reward: 94.57905804902464.\n",
            "# Episode: 214, Reward: -337.72232503844054, Mean reward: 89.35224612950906.\n",
            "# Episode: 215, Reward: 208.49865120268183, Mean reward: 92.34558489229957.\n",
            "# Episode: 216, Reward: 255.33977501556953, Mean reward: 92.0317155067834.\n",
            "# Episode: 217, Reward: 267.98292905910074, Mean reward: 92.51406862852589.\n",
            "# Episode: 218, Reward: 282.2261691521652, Mean reward: 96.18335088831152.\n",
            "# Episode: 219, Reward: 298.67825971239563, Mean reward: 100.08787084385366.\n",
            "# Episode: 220, Reward: 120.99800535242758, Mean reward: 101.16608869625739.\n",
            "# Episode: 221, Reward: 186.24945073476576, Mean reward: 103.6890529549737.\n",
            "# Episode: 222, Reward: 257.84792002154, Mean reward: 106.22940210346067.\n",
            "# Episode: 223, Reward: -25.412064554474185, Mean reward: 105.66713676462159.\n",
            "# Episode: 224, Reward: 25.541615280610557, Mean reward: 103.06168803884813.\n",
            "# Episode: 225, Reward: 232.26201525206372, Mean reward: 103.14101654153248.\n",
            "# Episode: 226, Reward: 288.1843845779491, Mean reward: 104.14231554145654.\n",
            "# Episode: 227, Reward: -235.92407743401327, Mean reward: 100.18397629702059.\n",
            "# Episode: 228, Reward: 312.6065648005639, Mean reward: 107.6233188508627.\n",
            "# Episode: 229, Reward: 276.3778768539682, Mean reward: 111.38445101082917.\n",
            "# Episode: 230, Reward: 273.39941391689206, Mean reward: 111.2705492382167.\n",
            "# Episode: 231, Reward: -77.33745780618037, Mean reward: 107.62492714572473.\n",
            "# Episode: 232, Reward: 253.73959726413597, Mean reward: 109.77603184358135.\n",
            "# Episode: 233, Reward: 224.52035925298782, Mean reward: 109.94963832551875.\n",
            "# Episode: 234, Reward: 252.08740288348682, Mean reward: 113.78449037194497.\n",
            "# Episode: 235, Reward: 258.583356394405, Mean reward: 116.55569021815114.\n",
            "# Episode: 236, Reward: 301.9362927658928, Mean reward: 121.11210926610143.\n",
            "# Episode: 237, Reward: 251.055000602341, Mean reward: 124.04451471552507.\n",
            "# Episode: 238, Reward: 266.14158781936874, Mean reward: 124.3996455552085.\n",
            "# Episode: 239, Reward: 258.0417206347706, Mean reward: 125.91024412253157.\n",
            "# Episode: 240, Reward: -86.04525107795116, Mean reward: 129.95295088845393.\n",
            "# Episode: 241, Reward: 278.0675990611987, Mean reward: 130.9904186776719.\n",
            "# Episode: 242, Reward: 267.6654467282247, Mean reward: 143.59230805800604.\n",
            "# Episode: 243, Reward: 280.1131603273913, Mean reward: 145.02049146910917.\n",
            "# Episode: 244, Reward: 279.5532810959419, Mean reward: 145.24782824880924.\n",
            "# Episode: 245, Reward: 232.8112122586534, Mean reward: 145.0380786043989.\n",
            "# Episode: 246, Reward: -107.03250209245756, Mean reward: 141.37808492209192.\n",
            "# Episode: 247, Reward: 307.1097266546795, Mean reward: 141.9946386977385.\n",
            "# Episode: 248, Reward: -57.45671937144893, Mean reward: 139.5690522666689.\n",
            "# Episode: 249, Reward: 246.837409666567, Mean reward: 139.87654478686437.\n",
            "Saving models ...\n",
            "# Episode: 250, Reward: 264.166044489662, Mean reward: 139.8911223065484.\n",
            "# Episode: 251, Reward: -0.6443060496861364, Mean reward: 140.0553917336287.\n",
            "# Episode: 252, Reward: 255.35792949291599, Mean reward: 139.9123165617009.\n",
            "# Episode: 253, Reward: -161.20973215462345, Mean reward: 135.7496341854916.\n",
            "# Episode: 254, Reward: 257.2849813152287, Mean reward: 138.8923983135385.\n",
            "# Episode: 255, Reward: 260.2114400702295, Mean reward: 138.9456296231756.\n",
            "# Episode: 256, Reward: 261.83742292217414, Mean reward: 141.48122948680984.\n",
            "# Episode: 257, Reward: 278.8055244498257, Mean reward: 142.14609004339917.\n",
            "# Episode: 258, Reward: 266.3079031852393, Mean reward: 141.70989581137002.\n",
            "# Episode: 259, Reward: -153.0939509424878, Mean reward: 137.51240941500447.\n",
            "# Episode: 260, Reward: 267.3918608759608, Mean reward: 138.05946766346992.\n",
            "# Episode: 261, Reward: 254.22780272933716, Mean reward: 141.08727123310697.\n",
            "# Episode: 262, Reward: 66.97525580473618, Mean reward: 138.85648073133763.\n",
            "# Episode: 263, Reward: 264.3547862058284, Mean reward: 138.7724816775207.\n",
            "# Episode: 264, Reward: 256.2582857958642, Mean reward: 138.68000836723277.\n",
            "# Episode: 265, Reward: 243.55432036860188, Mean reward: 143.55513260262606.\n",
            "# Episode: 266, Reward: 39.43074682215479, Mean reward: 141.42607331965948.\n",
            "# Episode: 267, Reward: 252.22550849597903, Mean reward: 144.27606814669582.\n",
            "# Episode: 268, Reward: 254.62102299159682, Mean reward: 144.03405726501938.\n",
            "# Episode: 269, Reward: 283.9802595691933, Mean reward: 144.18324278815805.\n",
            "# Episode: 270, Reward: 6.862836523049737, Mean reward: 144.34662671597462.\n",
            "# Episode: 271, Reward: 310.962211117787, Mean reward: 145.00045096130853.\n",
            "# Episode: 272, Reward: 26.51955856372369, Mean reward: 142.89666335414725.\n",
            "# Episode: 273, Reward: -188.64352163898673, Mean reward: 141.4262885793566.\n",
            "# Episode: 274, Reward: 250.32666217009793, Mean reward: 143.558879656354.\n",
            "# Episode: 275, Reward: 299.17476643045717, Mean reward: 150.74851110650627.\n",
            "# Episode: 276, Reward: 304.0101477501148, Mean reward: 153.62014141836866.\n",
            "# Episode: 277, Reward: 273.03873064955985, Mean reward: 156.17744086877835.\n",
            "# Episode: 278, Reward: 242.0018106094555, Mean reward: 158.63957631244807.\n",
            "# Episode: 279, Reward: 237.14594014716715, Mean reward: 161.48798480095093.\n",
            "# Episode: 280, Reward: 267.11942848961667, Mean reward: 161.54320420258205.\n",
            "# Episode: 281, Reward: 287.30552488703006, Mean reward: 161.5687283936259.\n",
            "# Episode: 282, Reward: 258.55129224939134, Mean reward: 161.89891312366535.\n",
            "# Episode: 283, Reward: 271.2417703962688, Mean reward: 164.50280660052096.\n",
            "# Episode: 284, Reward: 257.10919514126607, Mean reward: 169.97132393169298.\n",
            "# Episode: 285, Reward: 235.35181197598575, Mean reward: 169.8098554131494.\n",
            "# Episode: 286, Reward: 275.54947313889477, Mean reward: 172.0979644856212.\n",
            "# Episode: 287, Reward: 289.67835443726176, Mean reward: 172.59543698304321.\n",
            "# Episode: 288, Reward: 252.58048386578076, Mean reward: 172.3540993400234.\n",
            "# Episode: 289, Reward: 293.62001312639916, Mean reward: 176.56638042666486.\n",
            "# Episode: 290, Reward: 282.4733968431755, Mean reward: 179.1968707425669.\n",
            "# Episode: 291, Reward: 39.65437066713278, Mean reward: 176.6120589812532.\n",
            "# Episode: 292, Reward: 297.9017198494957, Mean reward: 179.81351857117988.\n",
            "# Episode: 293, Reward: 317.1942923108396, Mean reward: 180.78293751639504.\n",
            "# Episode: 294, Reward: 262.25144496655764, Mean reward: 180.68051588571115.\n",
            "# Episode: 295, Reward: 313.9951187216442, Mean reward: 181.39290063118244.\n",
            "# Episode: 296, Reward: 289.5770310181008, Mean reward: 182.3766553844924.\n",
            "# Episode: 297, Reward: 266.59159913449275, Mean reward: 185.2425568169365.\n",
            "# Episode: 298, Reward: 249.34753318142026, Mean reward: 185.4497769475577.\n",
            "# Episode: 299, Reward: 281.70983093351856, Mean reward: 188.1153341193466.\n",
            "Saving models ...\n",
            "# Episode: 300, Reward: 278.37913480775126, Mean reward: 188.41814161347753.\n",
            "# Episode: 301, Reward: 228.53108308625622, Mean reward: 187.86490529791718.\n",
            "# Episode: 302, Reward: 298.84596041649615, Mean reward: 191.2583374907539.\n",
            "# Episode: 303, Reward: 251.32050704134167, Mean reward: 190.75855853967113.\n",
            "# Episode: 304, Reward: 253.66518384378728, Mean reward: 190.4075315103875.\n",
            "# Episode: 305, Reward: 292.528186470421, Mean reward: 197.07867838672627.\n",
            "# Episode: 306, Reward: 275.27996361232294, Mean reward: 201.9371675395808.\n",
            "# Episode: 307, Reward: 280.11565129622136, Mean reward: 204.12972266534615.\n",
            "# Episode: 308, Reward: 276.5459283463827, Mean reward: 206.72651875338255.\n",
            "# Episode: 309, Reward: 265.5606700785799, Mean reward: 207.21066109206294.\n",
            "# Episode: 310, Reward: 241.86098241099444, Mean reward: 206.99984059368893.\n",
            "# Episode: 311, Reward: 255.74316251665155, Mean reward: 207.96379350572056.\n",
            "# Episode: 312, Reward: 255.72596207624977, Mean reward: 207.99920466564583.\n",
            "# Episode: 313, Reward: 294.0712555591008, Mean reward: 208.2794535262476.\n",
            "# Episode: 314, Reward: 255.18829641607826, Mean reward: 214.20855974079277.\n",
            "# Episode: 315, Reward: 276.66613493357295, Mean reward: 214.89023457810174.\n",
            "# Episode: 316, Reward: 233.37932181610861, Mean reward: 214.6706300461071.\n",
            "# Episode: 317, Reward: 258.42570946142314, Mean reward: 214.57505785013032.\n",
            "# Episode: 318, Reward: 283.72476064904333, Mean reward: 214.59004376509907.\n",
            "# Episode: 319, Reward: 266.0558443614178, Mean reward: 214.26381961158927.\n",
            "# Episode: 320, Reward: 276.77626654587186, Mean reward: 215.82160222352374.\n",
            "# Episode: 321, Reward: 259.42279421048823, Mean reward: 216.55333565828096.\n",
            "# Episode: 322, Reward: 246.65470802937998, Mean reward: 216.44140353835934.\n",
            "# Episode: 323, Reward: 301.28708939370347, Mean reward: 219.70839507784115.\n",
            "# Episode: 324, Reward: 278.02743845337636, Mean reward: 222.2332533095688.\n",
            "# Episode: 325, Reward: 275.2978176863519, Mean reward: 222.6636113339117.\n",
            "# Episode: 326, Reward: 270.1470580248672, Mean reward: 222.48323806838084.\n",
            "# Episode: 327, Reward: 267.30000791003897, Mean reward: 227.51547892182137.\n",
            "# Episode: 328, Reward: 281.44320028016176, Mean reward: 227.20384527661736.\n",
            "# Episode: 329, Reward: 254.71384454464967, Mean reward: 226.98720495352418.\n",
            "# Episode: 330, Reward: 256.1801172474028, Mean reward: 226.8150119868293.\n",
            "# Episode: 331, Reward: 262.6212301064545, Mean reward: 230.21459886595568.\n",
            "# Episode: 332, Reward: 26.928456854999908, Mean reward: 227.9464874618643.\n",
            "# Episode: 333, Reward: 264.9311579177372, Mean reward: 228.35059544851177.\n",
            "# Episode: 334, Reward: 284.73130333657605, Mean reward: 228.67703445304267.\n",
            "# Episode: 335, Reward: 265.3386042547016, Mean reward: 228.74458693164564.\n",
            "# Episode: 336, Reward: 267.3305268398667, Mean reward: 228.3985292723854.\n",
            "# Episode: 337, Reward: 269.76141246795095, Mean reward: 228.58559339104144.\n",
            "# Episode: 338, Reward: 288.7157590775905, Mean reward: 228.8113351036237.\n",
            "# Episode: 339, Reward: 289.974962018923, Mean reward: 229.13066751746516.\n",
            "# Episode: 340, Reward: 259.3462867389672, Mean reward: 232.58458289563438.\n",
            "# Episode: 341, Reward: 290.4434281349759, Mean reward: 232.70834118637217.\n",
            "# Episode: 342, Reward: 280.6280500970514, Mean reward: 232.83796722006042.\n",
            "# Episode: 343, Reward: 260.40024594842816, Mean reward: 232.64083807627082.\n",
            "# Episode: 344, Reward: 283.24811970059375, Mean reward: 232.67778646231736.\n",
            "# Episode: 345, Reward: 263.425411957228, Mean reward: 232.9839284593031.\n",
            "# Episode: 346, Reward: 277.6667700869442, Mean reward: 236.8309211810971.\n",
            "# Episode: 347, Reward: 278.71563837940516, Mean reward: 236.54698029834438.\n",
            "# Episode: 348, Reward: 262.92477815604184, Mean reward: 239.75079527361925.\n",
            "# Episode: 349, Reward: 271.3407469456791, Mean reward: 239.99582864641036.\n",
            "Saving models ...\n",
            "# Episode: 350, Reward: 234.18460735551633, Mean reward: 239.69601427506893.\n",
            "# Episode: 351, Reward: 273.9843669166363, Mean reward: 242.44230100473217.\n",
            "# Episode: 352, Reward: 295.0045967294483, Mean reward: 242.83876767709754.\n",
            "# Episode: 353, Reward: 244.75708773720777, Mean reward: 246.89843587601584.\n",
            "# Episode: 354, Reward: 299.36145185384447, Mean reward: 247.31920058140201.\n",
            "# Episode: 355, Reward: 268.51025063464124, Mean reward: 247.4021886870461.\n",
            "# Episode: 356, Reward: 266.8224971927194, Mean reward: 247.45203942975152.\n",
            "# Episode: 357, Reward: 288.37485932424397, Mean reward: 247.5477327784957.\n",
            "# Episode: 358, Reward: 251.3047946163591, Mean reward: 247.39770169280692.\n",
            "# Episode: 359, Reward: 265.8107165308721, Mean reward: 251.58674836754054.\n",
            "# Episode: 360, Reward: 279.96262912705254, Mean reward: 251.71245605005143.\n",
            "# Episode: 361, Reward: 34.095106380591034, Mean reward: 249.51112908656395.\n",
            "# Episode: 362, Reward: 289.86137391224145, Mean reward: 251.739990267639.\n",
            "# Episode: 363, Reward: 293.12904044643756, Mean reward: 252.02773281004508.\n",
            "# Episode: 364, Reward: -151.20554576496804, Mean reward: 247.95309449443678.\n",
            "# Episode: 365, Reward: 280.57044543578957, Mean reward: 248.32325574510867.\n",
            "# Episode: 366, Reward: 255.00162700972342, Mean reward: 250.47896454698432.\n",
            "# Episode: 367, Reward: 71.99783707672148, Mean reward: 248.67668783279177.\n",
            "# Episode: 368, Reward: 242.19704963020192, Mean reward: 248.55244809917778.\n",
            "# Episode: 369, Reward: 260.48968225461806, Mean reward: 248.31754232603205.\n",
            "# Episode: 370, Reward: 57.700204281136024, Mean reward: 248.82591600361292.\n",
            "# Episode: 371, Reward: 243.28416587826325, Mean reward: 248.14913555121765.\n",
            "# Episode: 372, Reward: 273.8513460236304, Mean reward: 250.62245342581676.\n",
            "# Episode: 373, Reward: 245.01993316344834, Mean reward: 254.9590879738411.\n",
            "# Episode: 374, Reward: 242.48147193248005, Mean reward: 254.8806360714649.\n",
            "# Episode: 375, Reward: 230.1740336033593, Mean reward: 254.19062874319388.\n",
            "# Episode: 376, Reward: 281.0425667424446, Mean reward: 253.9609529331172.\n",
            "# Episode: 377, Reward: 255.68567042834934, Mean reward: 253.7874223309051.\n",
            "# Episode: 378, Reward: 265.592021893534, Mean reward: 254.0233244437459.\n",
            "# Episode: 379, Reward: 249.38718460169164, Mean reward: 254.14573688829117.\n",
            "# Episode: 380, Reward: 12.241604364698247, Mean reward: 251.596958647042.\n",
            "# Episode: 381, Reward: 42.487006085593436, Mean reward: 249.1487734590276.\n",
            "# Episode: 382, Reward: 71.19467294947037, Mean reward: 247.27520726602842.\n",
            "# Episode: 383, Reward: 290.5295624400628, Mean reward: 247.46808518646634.\n",
            "# Episode: 384, Reward: 280.0255799252849, Mean reward: 247.69724903430654.\n",
            "# Episode: 385, Reward: 51.34068722345009, Mean reward: 245.8571377867812.\n",
            "# Episode: 386, Reward: 290.8368303258207, Mean reward: 246.01001135865047.\n",
            "# Episode: 387, Reward: 266.9087209831739, Mean reward: 245.7823150241096.\n",
            "# Episode: 388, Reward: 251.89296708267764, Mean reward: 245.77543985627852.\n",
            "# Episode: 389, Reward: 230.40109774644964, Mean reward: 245.143250702479.\n",
            "# Episode: 390, Reward: 215.99643993117303, Mean reward: 244.47848113335897.\n",
            "# Episode: 391, Reward: 245.57356310470828, Mean reward: 246.53767305773476.\n",
            "# Episode: 392, Reward: 288.9303478000724, Mean reward: 246.44795933724058.\n",
            "# Episode: 393, Reward: 267.51360012886425, Mean reward: 245.95115241542078.\n",
            "# Episode: 394, Reward: 290.0632633816501, Mean reward: 246.2292705995717.\n",
            "# Episode: 395, Reward: 282.9890007996357, Mean reward: 245.91920942035162.\n",
            "# Episode: 396, Reward: 294.59040216848604, Mean reward: 245.96934313185545.\n",
            "# Episode: 397, Reward: 51.548604392016614, Mean reward: 243.8189131844307.\n",
            "# Episode: 398, Reward: 222.57826869936832, Mean reward: 243.55122053961023.\n",
            "# Episode: 399, Reward: 38.1930135055608, Mean reward: 241.1160523653306.\n",
            "Saving models ...\n"
          ]
        }
      ],
      "source": [
        "agent, scores, avg_history = train()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "QtqayElp5fe9",
        "outputId": "c79d40f7-25e8-4d02-9cc9-a2beabe4e9b6"
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlAAAAGwCAYAAABmTltaAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAClwUlEQVR4nO2dd3gU1frHv7ubZFMgIUBCKKEjvaMQVIoiAbHg9WJDBEUULngFsYAFxYY/C8q9Fq5XBb127AICAQUpQXrvNZQECJBGIG3n98dhds7MzuzO7M7WvJ/nyZPd2Slnys757vd9zzsWQRAEEARBEARBELqxBrsBBEEQBEEQ4QYJKIIgCIIgCIOQgCIIgiAIgjAICSiCIAiCIAiDkIAiCIIgCIIwCAkogiAIgiAIg5CAIgiCIAiCMEhUsBsQiTgcDpw8eRI1a9aExWIJdnMIgiAIgtCBIAgoLi5GgwYNYLW695hIQPmBkydPIj09PdjNIAiCIAjCC44dO4ZGjRq5nYcElB+oWbMmAHYCEhMTg9wagiAIgiD0UFRUhPT0dGc/7g4SUH5ADNslJiaSgCIIgiCIMENP+g0lkRMEQRAEQRiEBBRBEARBEIRBSEARBEEQBEEYhAQUQRAEQRCEQUhAEQRBEARBGIQEFEEQBEEQhEFIQBEEQRAEQRiEBBRBEARBEIRBSEARBEEQBEEYhAQUQRAEQRCEQUhAEQRBEARBGIQEFEEQBEEQhEFIQBEEQRARR0UFIAjBbgURyUQFuwEEQRAEYSa5ucB//wv06gV07QpERQHJycFuFRFpkANFEETAKSoC8vKC3Yrw5I8/gHfeAUpKgt0S7xAEIDsbOHLEf9v46SfA4QDWrAE+/BD46CP2niDMhAQUQRABZ+ZMYPZs4OzZYLck/FixAigoYELKLCoqgKws4Ngx89apxenTwOLFwIIF/tvGmTPS64oK4MIFutYI8yEBRRBE0Dh6NNgt8J1Ll4CLFwO/3fPntT8rKQEOHNCfA3TwILB6NbB8uSlNc0tZmfy/yM6dwH/+47uIq6xUd5tyc31br7ecO0fuV6RCAoogiIDCd+qXLgWvHUpKS4Ft25hjoZeqKuCDD4C33gLKy33b/p49wMaN+ucvLNT+7JdfgM8/1y9QxX02su/eIooJXlQcOADMm8dEjpFjoMaJE+rT1QSUIACrVgF79/q2TS127QL+9S9g0SL/rN8IxcXsGPszdFrdIAFFEERA4TvpUBJQf/4J/PADE1F6KS9nQqaykgkgbykrA77+Gvj1V/25TQUF2p+JISy9YSs1UeMvRAHNb+uvv6TX7oShFsXFUk6dEQGVkwMsXQp89ZVr+7QoKGDXiJ5j9fvv7P+6dZ7n9RfHj7Pr6uefmcs3d67nZQoKAiOmjXDgALB1a7BbIYdG4REEEVB4p0YMfVVWAtu3A82bA0lJwWlXaan8vx74TnTbNqBTJ++2feiQ9Fqr41J27FVV2vMVF7PXesWYuO5ADPtXE2tiewEW8tJLbi5gtzPhe/w4MGqUthOYm8v2z2KRpvHHp6qKiaktW4AxY4DatdXXs2ABsH8/EBsLXHGF+/YlJQH5+dL6bTa9eybn1Cng5EmgRQvmKtrtnrct8tFH6tMFgf1ZORvF4QAWLgQ2bABatQKGD/euvVqcOwfs2AH07AlER7Pj3awZ25aSggJ2/HbvZsf688/Z9ObNgZo1zW2Xt5CAIogw5NIllq/SsSPQsKH6PFVVrDNo2hSoU0f+mSCwG9nWrUDv3uymFCj43Bex49yyBZg/n91Un3nGvG1VVrIONT7e87zeuDD8vAcPMvGlZ1tKDhyQXldWqm/nk0+AmBjWCYviSSkIACZKxXVcuKBv+4F0oNS2xQuZwkJ2zmJi2Pu//mKJ50OGyDv7/HyWM2W3S9fU77+z612NsjKWN8YLI6WYys5mrxcvBu6+W309oruXn+9ZxERHS69Pnwbq13c/PyDVrxL3f9s2JhAB9l0XHbaJE4FatdyvS+t8VlSw66mqChg7Vjque/Yw8QQwd85sfvuNic+9e4FrrmGjJNesAV54QT7fjh3Ad98xYbV/v/yzsrLQEVAUwiOIILF3L7upesPy5cDatazWjRYbNzLr/t//dv1s3Trg++9Zx71pk3dt8BbeISgqYv9PnWL/KyrUBYS3fPghG/KvJ8nbiIgoKmK5LatXS9MEQb9g4REEuYBSc5bOnGEOy6FDcoHGOzd820TcOVD8dtTCamZz9CjwxhtSiJQ/3srjJrpQlZWs0924ETh8WD7P9u3sPy/Ic3LUj4lIbi778bF1K7sm+BAyv9zp0+rHwuGQji8fanQ4WAj255/l8/NuJh9CLClRFygOBxOF778vOZG7d0uf8+HJNWuAfftY+FHruuNHI/Ls38/ac/q0FAoWBPnxSEhgbTCS2+dwsPVqXUfiAIETJ+TfSaXrm5UltVNtG6ECOVAEEQSOH5fyLpS/vvSgJ8xx8qS+5c0ULHpQE1B160rTjh+Xuwh5eUws9u9vLLxXWioJ1NxcdZft0iV2Q46PNyagFi1ix3DtWvl0b0Jgp0/LO2M1AaUlAAsKgMRE+TReCCgF1K5dQFwcE83btzNncsQI4yG81avZcbrmGlcHTIs5c9h/UfiIx7m0VHLSGjRgnevZs0BamvwaVuZGqXWuAHMClSQmsmstN5f9rVrF3KEOHaR5eIFz/jzw4oss1DRwoBR6Ky6W2l1QIIWZ8vKkHLiMDBZySkyUC5tjx4Bu3djrjz5iy44eDaSnS/OcOiWF/E6cAJo00R4IsGkT29ezZ4H165ljlZsLdO4MtGzJ5jl+XH1ZPpeosJCJ06VL5Y6aILB25ucDTz7JnD5PrFrFXMC+fdn3VUmTJlLC/s6d0vS8PPn3MyFBOxdOK3QdDMiBIogg4E7c6EHPzczdPHxHGehfdLyAunCB3RD59vD5QAALp2zZwhwfI/DHWCss9tprwOuvs8/FG7N4PMS2qWFmTSHeYdBqK/8LXZl8rXQIeAeK78BLS4FvvwU+/VQSMWfPMvfLiHg8f545BMuW+VYM1eFg510UeQkJkpAWjy9f0oB3UwoKtL9Dao6JKChEAQUwd2XzZmketeTzv/5iDpgI36nv3cvczdWr5U7y++8Ds2Yx4cEf/82b2ejIc+ck10c5+o93pY4eZftcWsrEHh8OBNh1Ih6n5cuBJUvYeV24UJpHS0Dx2z1/noUuL16Uj9BzOJigq6pSF6UAE428eyomza9YoT4/f33xx0yZ4B+lsHYGD1ZfR7AhAUUQJrB1K/Dmm9o3LCW+/oqKjfU8Dy+glM4CfxPy1w2poID9MlbuK9/BiR0o34Zdu+TLiCGaqirtkIQa/E1ZLTGcd2ouXpSLiMJCdj4/+0x93bxI4fHGgdq1S/5e7drg289/vnOnqwumFcJT1l3i5zHiQPEC19dRUbyAqlHDVUDxgiI7m4Ws8/Pd13QSBSjfCYtJyrm52vWzlN/ddu3Y/y1bpOOvNvJx6VJXIVlVxY6NGBLr3Jk5bJs2yX8IKB1Cfn9zciRxkp4OpKaqt1vJuXOstMYvv+irfXXggOR68QMY+O+k2vdOFORffy0t5ylJnr92+W0p26n8fnXoIOWukQNFEBHGjz+ym+Gvv+qb31cB404cifAiSzmyKxAO1PffsxFLixfLp6s5Jnx78vNZfocInwDPOwae4B0KNQHFuwkWi3QcBIG5QoLAXAC146sVUvMkQI4fZxXExfORn89+iVut0rPaPIXwxHY2aaLeFl4YlpWxc3D+vOt5FrfHh6XE/1VV2oKLdyO2b2ci7v33pTy20lL3opnH4ZDaW6OGdK7PnmXHUllU88QJJqS0BCwgCSg+tNmsGTvGpaXa4W9RRHToAEyaBAwbxkKKlZVSnqBWWEncdx7RWbRagaFDgb/9zXUe3oURBLmAOnSIuUpi+/nvgdZoz7Q0qT2bNun7wcE7oFoCSi1XMz+fzcM7YTVqSJ+vWycdU7V1agkofhQpwEKkCQlSojs5UAQRoejNB9H6JaYXXkBpdXT8r0GlgOBvQmb/olu1iv36FTs/ZQ0cZWfKd+Bim3lXhW+rkYrfnhwo3k1wOOQigg+XGKlLpOb0ffMNE01iTsmKFVKnJYbSWrSQOh+1EJ6agBKvAeU2leJi/Xrm3PDHsW5d4Kqr2OviYnkSuSAwB+Odd1zP1ZkzcgfqwgVWnPH0abaM6Nx98om8XVohz6oquQPFC6jiYvXztnu35CIpQz084kgtu539mKhXT3tenuRk1mlbLMCVV7Jp4nlSuxbsdvVQpige4uPZujp2BO64Qy4yzpxhgm7ePLaN4mImFOLipBIDXbqwhyLzeYJduqi3vXNn+Xu177YoskT488Tfi/hlRQElCFKCPS+OxNf8NbZwoby+lnKd/OuzZ6XrlhfgVis7boB0bwglAUVJ5AShE4dDPoxajbg4fetSCig9OU08fDsuXlQP6fHbuHhRPuTZVwfq4kXJ4Vm6lCWMXnEFW+/Spa7z88P7lZ3yhQtSG5o2ZQ4HP483bl1FhVwgeRJQgiAXUPz8+fnyY+dO8CrFzMGDrMPfvVsKCQFMJAmCFALr1EmqwF1Vxdq2cSNLYq5RQz2Ep/WLXM2dKS2VthUfD0yYIOXB8CFUQWAdutghFhQAKSlsm+JoOHEdVqtrCOrECbauEyfYaDuxQ1c6ESIOh1xAiWGaixclN6ZOHbkAKy2VBE2TJpIjZrXKj0VKCnMQRVHWrJkkqsWkcjX4gQpi7tTp0+yHitpjZsQfMFYrcP/97Fj98IO0/oQEad527YA2bdg19MYb7L8Y0hOFaWwscPvtbFtdukjXnrgfdjv7nojlG1q3ZqPx7rqLicTff5dfo7GxTMBdvMj2++GH2fE8fx744gv1YwC4hvAEgf2wWbyYJYnzQl/8XDka8OxZJjrFY+rux9q+fUCPHvLj9sQT0ufi9R5KITwSUAShA7EWy9//Lh+5A8hvJHoFFO8aeeNA8R31xYtSOIaHvwGa7UAtWMBqtYgsWcIEFH8DTUiQ3m/ZwkRE587qAkrcH/FXpiCwY1Ra6p2AUro4yiRsq9W9A8XvR36+1JEC7iuAKwUULzD4UUcxMayDLChgHWGbNpLAqapiIcx169hn11wjd6CUx6qyklVRb9aMbe/0adZhKtsini+xIxIdGqUDxTtMGzeyHK0OHSTx1Lo1qx0m1iZKSpKcGT4ctGwZcw+qqrQflaIUUNHR0vrEZerWZcJg1y4mis6ckc4PL6CSk+VCKyWFFdYUBUjz5lJouEED7UECvFiuUYP9lZQAM2ao74NI3brSiLpWraTjxQsogB1/u521j3dJeSHWogX742ncmImhVq3YPDfcwEbPDR3KrgXxvD76KHOA9+1j75OS2HG+eJG1xWJhbZX96LpwgX0QH89el5aiKrU2AMnGPn9eCsevWMGuA1RUAPPnI/+dhSitvRYOPMF2LjqaHdzKShw4WYTub9wFJCW5/f7u2cMElBi+U44uJQeKIMKIS5dYZ1+zptRZ/PKLq4Dif8m6Cyko1y3izTPU9IS13M3jaw6W0lHIz2c3WF4Y8usVczmOHWOuCk9pqXQz5wXU22+z48QXPtTbVuV8ooDKzWXhpf793TtQvIBShp/cPcRXKVr488wLTkGQwj7NmrH+hhdEoqgQrw01B02cf/t2SYSLnfXVV7NQKo84j1JAlZTIC3PyOU7iI1bEApPt27PcIEAKV/PHmh9yX1TExNiOHfJ95+EFlNieOnVcBVSfPsB11zHXU/wuAkxUiGJRKaCsVnk5DDFnTKRGDfWQnLLQZv368pIJKSlMGIphWRE+RNixoySgxIKYSlq0kAuo2Fh2nrUSsWvWBB5/XPq8Rw/2p6RGDZZwzguo8nImPONjHUAxU44JAhB10YbKH39Fwv7NuIAEILUecJoldFXVrgNkZrIVFRaiyFIJ7OwB2GyohQKcPRIFZG0GTp7AGUTjQuEpAEx9J4LdFIuQiAP7dqP7908DU6eiquIhoOSyVVizJmCxIC2NfRcOH2btFO+nSgFFDhRBhBGvv86+52KHAbAvuDKUxwsovW4S37F640DxnZbW8+T4G42yAzYribxTJxbyOX6cdZJ8foVWfSlRFNSsyX5tXrggdTK8gBL3i0/6Fd2Z+vWZ8NBCS0D9+is73kuWuAozXkDxzpFSLBpxoPhzwwsvPlFW7CjEfa+qkpbjSyooEefnr58LF5gY69+fHd/Fi5nrt3mzdD5E4SO6EXzopbLS/cNm+c5dXI/aiEmRbdvktb7UkopFcS+6t3XqMOElXid16zJTw253DaHXrQtcfz07XsrvgVKIREdLYa8mTVi7RAEVF8d+/PTr5/oIF15AdesG3HILe52RwZ4rJ5Y/4K99XqxpCe7+/dn6jh9nolApcNXQ+wONd9GSkoCLPy8BvtiKhNK1wCimQC0A2mMoctAYrbAf63CVUzzBFgWcOwt89aVzPSdfWAKAVTNNwX4cRAsAVsAei7O3/xOFN44CvrMjxV6I8T3W4VhpHXyc3Q7HNsYDZ74FnngCVcgBcPkA100B6tZFaq0cXCxNQaE1GSeLT6A4xwb8XoWacTuBN7LZzSEmBrazNwGxbeFomQ7UbaB/WKIfIQFFVEt+/ZV1zPfc41pfZd8+1qmJndeyZfLPX3yR3bSvvZa953/FeiOgvHGglCE8NfwZwhP3s3t3qUr24cPycIUnAZWcLAkoMUfCk01/8qRUsPDxx+UJuTxaAorvnPjzJghyF4YXLMqRTO7Ol1JA8etRfsaPPgOkzrGqSjqnegRUTIy8TSkp7LOePZlDsWcPE1DK3CmrlZ2vkhLpWGgNSBDhj59aDpbYjthYdo3v3i0lPzdsqC6gxOMirk/52CE+eVoZqk5IYCFOQF6vSdlWkXHjpFwb0aEB2GNRYmLUB4HwzhKfxxYdze4TooDi57NYgAEDWD6geJ9QYrMxsSY6kXpLAXgkNxe1DpwAclKBHTuQ9MHHiD5dDKA3akKe+HUbfgKaNsXKqQuBHfWYcm3cmCnK+fOZpZeaCtSrhwPnagAFdYDKSpxLqgdHYRSi69eF9fr+KIurhS/2A+gM1GgGYGQfxJ8F4AAqrukB1K8LzJqFqm02MOkGIP8MkH8GNmxGQ9hRiHY4vmUpCpACoDMSsR3IkYbdWtEEQCmqRjwLYBtw553sZqz3oYB+gAQUUS0R7fWVK1loQOT4ceDLL+XzqnXmy5ZJN8ZgO1DehPB8daDENkdHS53ahQty50ZrSL/YydauzRKF+SRyT50Hvx9//gnceKP6fMp9Eus88Z2q8lEmWiG8khImBnmBoxetx4rw9Y/E0BUfwuMdqMpKdWGjlUTO/zC32SRRoHSgxG2XlLgvC8DDL6vmQInnvHFjVl+ookLaT+UPFbHtYvvF9TVoIJ+HF1QWi1ww8u1RCiY1AVWrljT6kL/WtMSTuC/i+pSuJx9mUo5uu/pqlgSuzIFSoqfdHikqYjeuffuAf/8btSqTAEwAACRhH9pbTqJq+ChcOW080KiG3MqKioJ1jQU4Cbn9pngY4KmEvsDl70WJHUAZkFCLRfl+/VX6kSLur/P6hA144AHggQfgeK0CuBjFLuj9+4GyMljTr0Sj+EvYtakuTlSloLjUBjS7FnUHDwIaFLCTXV4O67I04M9cOE7+CRSADW8VBPY/SJCAIqodfMe+di2z7cUvu7IKNiB1EGKsHpBuEidPyqvu6nWTfBVQZjpQvgooseMpLXX/HDIR3oEC1JPIteBdrc2bPQuo6Gjp4ayXLml3TnxHXlnperx4l8zd8dJymdTmEz9TCiilA6WW/8TPr3T6UlLk79110KL75Y2AcpeTYrOxP4dDOt9aAkrpQKWnM4dowwZ2bJQPZ65XT300nPLa8XQt8RrCXfmRmjXZA3ftdtd1iudOTDbnsVi0HVJ37dRst8PBiqsdOSI9NFL8//PPsgz+pCbJQGEyUKMGkl58B8k3dsBgN3Uc9Ig2XsSL5zQqCmjbll2vv/zCpon7zF/PIlXWaGZAxcY66xPYrgIatgcwBzhesy+7t1wC6t4EgP8xUAmgHuC4eRgQtRV47jlg+nTPDfcjJKCIaocy5HDkiPQcJrWkUlEsXHkl+9G0Zw9zuCsrgc8/l3dgesWQ2s3ICHocKGUZAx6+o9dyVI4eBX76CRgyRD4KDZALKDF/Sa+AEvddFFClpdIx9NTpKWvWiM9Q05ovKoqt89Ilth2tjpJ3oMTK3BaLFOISHSmx79LTPkCfA6UM4SkdKE8CSokyNUSPgNLahhJ+WbUkcr5tNpv8YbSeBJS4PouFXXPp6a75SADLQfrySyl0p9Y2tfdqbdSL0l0S4cOTuikvZ2InIQGoWRPW5TuB/x5kO9ygAaz7zwLn1rKDEx8vWYTr1jHbVYsGDdivwVtvRfSwYag324Jz54C6dwKI114M0Hcs+Puc8gdP+/aSgBKvXd4hFb9Pavcaq5U13WqVvi8Wi+u5lwn27p2lDQYRElBEtUP5JT5+XBJQao8+EG8cMTHMlt+zh90U9u937Xj0CCjlU89D1YH63/8kkfjCC/Jti8eEF1BVVfqeESd2qGKxQl5MGM3/cDjUlxH3SRwyLgooPQ6UeBOPj5dCXCUlbCTa9u3qnbqIMjSqrJMkUlkpiTKlA8UvI4bw1NA6Vp4cKF5E6k1KVlvWnWsjCihAOiZ6Q3jia2VRSJGUFDZUX22bPJ4EVIMG8tIS3nDFFay0iRjmc4tYIKtvX5YweBkbmgAYBZw8cfl9DoA56uuIi2MlzcVSAVFR7H+tWsBDD7GM98vcfz/7rikdPDW8ChtCOuZ2O8uH3LhRKvLJnw/xe6oltqOjWdPFfLLkZNdrk8oYEEQIoCagxOlqj2QQOzCrVf6rSqzbk5HB6vjMmaNPDJWXyzvaQJQxUCYh63GgtDpufrr4kNOoKDZdz6MjxP2126WyM6Jo8YeAEkskuAvh8Q6U2JaEBClUe+GCdAzduWz8ceVDk0rEdVitUgenJaC0zo/avsTGygtBAq5CRy0RXC96l+UFlIjeEJ63KJf3dC316sXcUPE5ed5uU1nWREZVFbsxvP46yw8QTyZXidRqswKdurELpqIC1tp1gU6XV37hAlPYiYlMJA0bdrkAk2diY/U9MxPwPnGdFzlDhrByE+L1x58P8Xuq5UAB7B4qCih+4IByvnXr2EjcESNcfywEGhJQRLVD+Qvm+HHpEQXuEoR5AVVZKQ1t7tJFqiSuRwwph1sHooyB+LgUtcRjrV90asUYAXl7xRtofLz+PBrxGMXESMU2RdFgtBPVajsvoPhcDD0OlLjPCQlSiIuv1q03hOdOaPEj8ESBIx5LXuxWVWlfk2qd3m23uRdMyvd6Hz2kNr9eB0rEiAPlDUYdKJtNPoDEMOIvLpuNOUOxsWwnd+1io1OOH2cht5Ur5cu1asWG5zVuDDgcsB2tBD6VikXZWgAY4UO7vMBXB0pcBy/e+c+qqtj1reVAASyXShzxrJY7Js4nljU5epQEFEEEHL5DstlYaKegwH19H0AuoMrLpfUkJ0udqviIDnedgRkCiu+otZwipUgqKpLqw+gZhWezqa9bbG9UlHQ8jAgocXlRQAHeO1Ba4kJLQOnJgRKJipI7UHpCB/xxdZdXJB4rMXwHGHeglMeqXTt1c8JdCM9ox6l3Wb4ytoiagKqqcs2B8hZTRrPp5eOPWRxRae0qnycDsIvopZfYg/BsNqnOxOX5rbExLqsINGYIKHfr5IWy1ny866QWJle2MRQKagbhVPmPGTNm4Morr0TNmjWRmpqKoUOHYq/iGQKXLl3C+PHjUadOHdSoUQO33347TiniNjk5ORgyZAji4+ORmpqKJ554ApVavRQRdohf5KgoKTn05EnPHSTfKfCXg9UqrzasFESHDwOLFrH8gNOnXQWUryE8Ty6MiPLRJSLiA12V7dC6qfIJ5CJ68ixExA4zJkZaTu8oPCX+cKBEbDa5A6Xnhq1H2ALuBRTfJxsRUFr75q8QnlkOlL9CeKYIkVOnWDn2v/4Cvv0WmDmTjf56+GF2opQbcTjYhT1wIPCPf7CHuW3ZAkyaxDLN09JcDozR0YP+wNttehJQfKkLPaHo0aNZaFX5tAK1bYVClxxRDtSKFSswfvx4XHnllaisrMTTTz+NgQMHYteuXUi4/FNy0qRJWLBgAebNm4ekpCRMmDABf/vb37B69WoAQFVVFYYMGYK0tDSsWbMGubm5uO+++xAdHY1XX301mLtHmIT4ReY7yLIyeTVsrVi9WghM+cWuqJALqkWL5LlVnTq5zm8UvqPWyrNR7sP589KjLZRi4c03WRTiqaekaVo3R18FlIg4Qo7HnwJKWQeKhy+kKSIWmgS8c6DczS+KJL5GkBjCU4pjT7/ctd5rTTfLgXInoNTOrVrCuj9DeD4Jkd27gWnTgO++057nnnuAzz5jB7GsjCUjXrrEVLGe+gWXCahzprMNevF0jK1W9r3S+0MgPV16nqCnNoaCAxVRAmrRokWy93PnzkVqaio2btyIPn36oLCwEB9//DG+/PJLXHc5+D1nzhy0bdsWa9euRa9evbBkyRLs2rULS5cuRb169dClSxe89NJLeOqpp/DCCy8gRuvBRkTYoNa5VlZKgiApSf74EBFeQIlYLNKNX6w5VF4u7xiV1ZmVI/1EQVJczIbJt2vn+cZkxIESowrKZ78pUSajGxFQeh+izGOxGE/8VWKmA6U8JrzAVqsE7gl3N3hexPPbU2uXXgdKS4CYmQPlzyTykAnhvfoqc5kcDtao9HT2PzWV5S/Fx7MH2f3zn9JOGsnYVhCpDpT4ueg+GRkMoWc+ElB+pvByUZ/alwOqGzduREVFBQYMGOCcp02bNmjcuDGys7PRq1cvZGdno2PHjqjHFR3LzMzEuHHjsHPnTnTt2tVlO2VlZSjjCvsU6U0GIYIC33nx1aXFG3mtWiy/c8sW+XJqAop/LwoopaMkduZ167K8UmWnLwqs2bOl0V69ernfBz1Ohzi9Th02Ok4rhKeFv0J4gFQhWy0MpZW8roanfedFmjsBpRYO4B2okhJ9+6hndKOyfSK+Cii9DpQvITx/JJGbFcIzmkSuyr/+BTzzDHs9dCjw4ovOgo/+IpwdKE9lMHjH3l2upR6U84WCgIqoHCgeh8OBiRMn4uqrr0aHy+NM8/LyEBMTg1r8kxYB1KtXD3mXS0zn5eXJxJP4ufiZGjNmzEBSUpLzL13LgyRCAl5A8Q4UfyMfOpT9yOTh3Qx+mohoTmoJKK1HgYjziy5HTo7nfdDjQInbER+FoXyYrSe8DeHpuSFyObQyeEdPD0YdKK11a1XTFh2oixf15VzoFVDiZ55qMvkjB8qXEJ4/HCizQnheu5kbNwILFgA//ghMnsymvfwye+9n8QREvgMFkAMVdowfPx47duzAqlWr/L6tqVOn4rHHHnO+LyoqIhEVwvCdq5oDJd7I1X4ZurtJix2EloDSKgR3+VFPTvQMzTXiQIkjWow6UN6G8FJT1QuSqq1b7Rj7S0C5y4HScqDi4iRHTE+V9UA6UN7mQIWqAxWwMgZbtgBz57I4fUkJE0o8t98OPP20b40xQDg7UHpyoABjPwT0bouSyP3EhAkTMH/+fPz5559o1KiRc3paWhrKy8tRUFAgc6FOnTqFtMvDsdLS0rBu3TrZ+sRRemka9fztdjvsYiEgIuRRc6C8FVDKEB7gOppNKaDE7YsPRa2okBegdBcqOnoUWL1aXjJATwgPYAKgooK1U48DZSSEx4vG2rV9E1Bi4qketPaddxP15EC5GzQQFcX2z+goPP4601o22CE8X+pAmZ0D5ddReCdOsOem/fYbe82fKIuFOU35+WyU3OzZvqs5A4SCA+UvAcV/98TXyu8DOVAhgiAIeOSRR/Djjz9i+fLlaKZ4dHb37t0RHR2NZcuW4fbbbwcA7N27Fzk5OcjIyAAAZGRk4JVXXsHp06eRevmhUllZWUhMTES7du0Cu0OEX9BKIvckoNRq26gJKKUDpUwYFrdvt0sCih+l507czFF5woOnEF7NmsxJuXiRPWolLc24A8XXtuLrQImI4c46dYyF8LRyoPSipw6UtzlQfBv1jpRUE1BxcVJtJ7Fiu4iZITy9SeRmhfCMOFBq+W6AySG83BPA7iJ2ga9fD+uyWUDlOVZHRPmIgWHD2BOLCwuBQYOAa6/1beM+EAoOlL9CeGoOVEyMfMBKOOdARZSAGj9+PL788kv8/PPPqFmzpjNnKSkpCXFxcUhKSsLo0aPx2GOPoXbt2khMTMQjjzyCjIwM9LqctTtw4EC0a9cOI0aMwOuvv468vDw8++yzGD9+PLlMYY4gAJs2yZ+75i6Ep9a5uwvhecqBUhNQxcVMRJ0+7d0+8evTmm61srBgTg5zutLStEUaL5T4fauqYvslujGA3FGoVYuVuomLAxYu9NzmcMiBEttmpHNRE1CxsdI1Z7drCyhPAkNJMOpAeRvCU/vxAciPu9cCShCAGTNge+EzoOIuaZtYDqBUWvmVVzIXqkMHgItMBJtwdqA8JZGrub/R0XIBRQ5UiPDBBx8AAPr16yebPmfOHIwaNQoA8Pbbb8NqteL2229HWVkZMjMz8f777zvntdlsmD9/PsaNG4eMjAwkJCRg5MiRePHFFwO1G4QGZ8+yDtqbmkMAe/TKr79K75UOlPKXsLchPE8Civ8lJs6v14FSQwyDKDsgfru8gOI/U1uXmrgpLgZmzWJtvmzWuoRk1J6BpUWwcqC08ORA6UXLgRKx2+UlEfj9NxLCUysBoTeEF6gkcuW+eRJQutty5gxw7Bjw0UesXtP588DWrbCiCRAVDVRWAA0awjr9Q6BOPHtKbYcOhmozBRJPj94JBIFwoMTvn/K+oXfbJKD8jKCj54mNjcV7772H9957T3OeJk2aYKGen9FEwDh9Gnj/fZag/I9/eLeO/Hz5e29yoNzd7NRyoPhHhCgLJYoCqqpK3qkaFVDiMsq2iTcY0YECJAGltQ1eQPGIIwPFkCOgntMibs8TwRBQ7hworSRyvq168CSglKWCPIXwtEYvqTl1gc6BMsOB0nLjNHn5ZeD5511PvNUK20svAcV3sAS89HRYR9iAMAgaiOFNtdpggcKIaOPLjBjJgeK/T1ar/DuqB0oiJwgvEfP6fQl1qVnlRkfhib/81b78ag4U36EqnRC+PXx/4I2AUhthpgzhAfocKLXX/D55ElCBzIHSI6D4HCittmmVMRDX4Q1aDhSPt0nkag6U1nHzVwjPSBK5Vw5URQUbKVdWxirQfvstK3ApMnQo8Pe/swuxXTtYkzsA/4Wz3H4whIi38AIq1EfhRUdLPxK9yYESf9So3QeNtJEcKILQydGjvq9DTRBp1YHSmh/Q/vKr5UC5e+QLv34zBJTWNDGEB7D+iA9XKtEqj2BEQOm5GWsdY3/nQBlJIjczB0rEnQMVbiE8Iw6U2o8PQMOBKi8H3n2XPYBX7QnfTz0FvPaa6zYVZfqCIUS8xRdRawZ6r/GoKPanV0Cpub/itSHeR7x1oEhAEYQOHA75MH9vMcOBUk7nX4vr0hJQyhCNvwUU/4uvZk2pUnpxsfsQntprswWUpzIGejGaA6X3uYF828wI4UVHs/fKR954OwrPFwEVrCRycXv8OXNJIl+4kI1G2LdP+iA2lll39esDU6YA992nuk29jlwoonasAokREaN2z/O0Xj4HSi0/zps2koAiCB3wCdaevrDu0OtAeSugxNdaLk6wHCjx17/4q095M+NvRFqdGy+gRNfADAHl7xAeP3yeF8lK/JlEHh0N3HEHe690UrUcKFFouBNQwciBMpJEriWgeJzHvbAQ1lvvBRbOZ+9TU9kz6UaN0n0ClJ1yOAmocHKgjIg9NQeKv++K771pI+VAEYQO+EeQ+ILyi2o0iVytE1DWugFCT0CJ2+XbJ24jKkpbQPGv+cT4UEoiN1oHyoiAMsuBstnYM2gB18f0aAmohATmFGqVMTDiQKk+ykUQgNJSWC8BqLTr/mVihgMFwDnqoaoK7EB99RUsp+azdkycyHKdEhN1tUm5Ddl2woRQc6B4p1CZ4G6krWo5UEoHSu+5IgeKILxAq4M3itpNwswQnjsB5anD83cIT9k+fmQg9xzsgOVAaQkorRwotWOrbKPadGUOlKeio3rayLdJ2R4tAcUvo1yHCK9heAHlaxK5y3W3+Ddg3FggJwdWtAasw4GrrwaaNQPq1XNbJ8RiAZCXB2zZAuv6ZKCoNSsCBsDqqITDynbCZhVg+/ZrYFMd9tnBlcBbX8Na8BBQHsV2rm5dVP2tKbAJwKk8WGrXBlatAtq21dy+O4Lt4vhCsNuu3Caf5ySGn8XpvjpQRkWYVhtJQBGEDpSuiNqQfT24c6D0hPDUpqvd7NREiJqzYvYoPK1pagKKd6C01hOsHCi1c8u7SFrt5VF7lIu7HChvHCh+NJJyu3xb9QooUTwKgqRhzAjhyeYrKIDlX+8AYFaYBQLgqAJW/sn+YuxAgwbsgIjx3po1gd69gePHYd24GFj0BlBZCQsGA5ZeTHyVlCBqyzqUN2sNNGwI246vYft+O4BB7BjgJHD6AKw4BeByPaYzp1H5n6UAOsEKB/D2216LJ4AcKLO2r3zP3yPMcqDMCOGRgCIIHSi/KN4KKCWeQnhqo4jE1/w6RNw5UGrJ0cEM4SlrU6mtx5OA0or6+CMHyqiA0nKglMdWDFV4kwPF/0oX8VVA2WysLQkJ0n74mkTOCzNs387EyjXXAPPnw5pjB6btBf788/IIgyLgyGH5Cs6cBg4dZOvCLwAqgbZtYb2QDuQ4gFUrAQDRKEf54UPA4UOw4RtY0QawWIHUVNhuGwwMvx+2r+sDZbEsm37vXlQdqQAuNIIl4yrgvk7qO6ATbxKTQ4VgO1CyawTy7zb/QykqylgSOf/dE1HeC8mBIgg/otZpenOTUes89Ybw9LxW6/jdOVBaAsob1EJJyrIMZjlQngr+KY+JmiA0OgpPFAxaVd6V6M2Bio5mIUxvRuGpdR6+CChxnWYLKACwWgRU7T8I/L6MCajRo4GkJFjjAXTuzP4EgT037sIFWGKiINii2YZWrgSOHgFS68F664PAPV2Afv1g+c0BvL4S2LULSExEdJ9OwMGzwKVLsHUaDVvrW4HifuwYtABwDWDdCEDMaczIQOWoDGAvYDGhJyIHyvc2iD8k+Gtb6UAZET+8A6UVwqMkcoLwI3pdB08oO08jITwjAkpvCI//1ccvY4YDpVagUNy+u9IKagIAUH+YrpYLqLT/1ZY1mgOlJax8daDcCShPDpRWCFPEGwElzmtmCA/nzsG6cDWq/trIlo2NBS4/TF22jMUCNG8OgD0RxXnemjYFTp4EGjSA5R4bcDnKZrFZgb592R+AqLoA0i/vx5RBsO0B8JO8bVoughmCJ9guji+EQtu1hJHSgfJUu4xHzYGKpDIGYXaZEdURfwkobx0orV9P3goofj61dupBeUzURv+J2+KPp14HSvl4Gnfosfi9yYFSOx++1oES2+cuB0qrQ/PkQKkVW/UkesR5RQcK0BavbpPIS0tZzaSUFKBOHVj/WsOmd+kK6//NYHlNKu0RkYlDmw1ITwdsNrcdvZ5cGeUy4nE3KySv9jocCIW28z+0+PPEXwtm5EDx30l+Hk9QDhRBeIFZAkq5nBEHSusLr3YDMSKg1JbRU+TS02f8e3cOlLL9ekJ4yuOkRI+A0gqP6RVQonNkhgMFeJ9ErsSMEB7gWUC5zak7dw4YOBDYuFHaTr16QL9BQJs2sHZRWUaB1nlzV8ZAT0er5SKYIaB8qbAebELBgeLPkdZ3WHldGHGg+OfnmTEKz5cBRWYRZpcZUR1xJw6MoOZAqQ1xD1QIT0ssmCGg3IXwvHGgtJ7vp0YgHCjeOVRDbw6Uu/XoSSJX4quA6tqVmT2NG0vT1JL2NUN4S5cCPXsy8VS3LvD990B+PqyPTQTatHHZpi8CSrmsKKDEthkRUGaIBt45CTcBFUoOlM0mP89mOVBalcj1niu1+YLtQpEDRYQ8/syBCmQSud58HyOPG1FuR2274jb0CChx2/xIPcB8B8rXHCjxpu6rA6UnhKfWSWjlZPkqoK69lv2pHXsxwRyCAMuaNbD+7zHgzMNARgZQUQHroQXA58+wRjRoACxZArRvL9sXd6959Iyw1HLPxA440CE8cTveDjIJJqHmQCnznMRczago+TVuZBSeWmkR8b2R9vFUVvr2dApfIQFFhDyBCOEB0s1c7VesnnwoX3KgeMwWUMr2uUsiFz9TtiGQDpTaDVXLgfI1B0oUYkYdKH8JKOf0zZuA3wuA+HhUdG4LxCUhyuYAYAWWLoVlza+wYgOAwcCxHNYmfAVAAMaMAd58U1bJWyv0pisHSqO9WiE8tePmKYncLAHlKewaqoSSA6W8ti0WaTCIKFBFjDhQfAiP35bec08OFEF4gT+TyPkbgCgSAj0Kz1M7RbRGYyndIn5etfbpCeEpt8U7NGY6UL7kQPHtVaL3WXjeJpErwxAiPgmoqipgyxZg5kxYvvwSVjwHB6yoXJYF1K6LqPOHgAqWHGWBAMv06cCvXYEN64FaybD0Gwrc0QW4+27NfXH3mkePA6WVRK52brXOt/JHi69oCfNQJxQcKK0QnvgdqqhwffSTkRwofv2ewuPu1sVDAoogPBCIEB7gGk7Qspk93ewCKaDE2i3eOFBqI7nEz9wd42DlQCnLI7hrp1YOlHJ+d0nkRhwom81VoBkSUHv3ArfdBuzeLW2jbRs4ikrhOHEcOH0KUbgI4LKA6tULlueGAw4AffoACQmw3mcFWrq2VWyvp9c83iSR6xFQgXKgwk1AhYIDpRXC4797yrZ540Dx6zNynsiBIggv8KcDJYoIvhp1IAtpGhFQap28mBejJaDUOm++01LeAMVt6xFQehworRtsqORAuVuPu3CQ8le61apDQDkcsPz8C7DUBlx3HXscSu4F4EA2K2xZWMjKC1x/PfDcc7Au6gaUCcCJE8ClS4iqJwA//wWUlsJy263StXu5JIG7zkjrOvVFQCm/B8pOVu06CEQOFP8/XAglB0othCceT+V14amt/HeP/2HqTaiVHCiC8AKzRuGp5UAB0iM53IXwtH4hesqB4kNpesWCFu7ydIyE8NwltpvhQCmPjyhQ1ebRG8JTumWeHCi9z8Jzl4Bq1IGqqHAjoHbuBMaMgSXbAWAQsOlyUct3vgewg8149dVs1Fy9emwbS8B2vFEj1tZGAGq2YsvVkNqh5jYq0RI+Wtef0SRyMU8G8C4HyizRQA6U723QCuGJn/FPMfB0/+IdKF5AmeVABbsaeZhdZkR1xJ8OFCB9mc1yoAIZwtMSEu5CeFoF7fjlzHKgtJKtzSpjYDSJXCuEp4YnB8qd0yYInIDauI6VFsjOhiUuDoiNuzyXBZaoKCA1FfjnP4Hff3eKJ377IsoyBvx/5WutfXH3WmtbPFrbU3OgKISnn1BzoJTnVrweoqKMuXy8A8X/oPDGgaIQHkF4gT9zoPj/4ZhEruVAuQvhBTIHyhsBpTa/VgjPSB0otfndOVDuOmPlesXXLiHQvDzYbrsFuHAB6NsXlslfAwsEYOtWoGtXWMY/D7R3v321tnq6Rt2tS08IT0tYurv2vRFQytprvhKuIbxQcKD0hPD4UXh62sk7UPw0bxwotWuEBBRBeMDfAkrsmLwRUGo3kVBwoPSMwlNrk7htdzcmMxwoLXGiNwfKGwcKcLX83TlQ7jpjFwfqwF7gxAUIfboCsLDjl5cH/O9/sJWeBXr3BhYsgGVXApAGIC0NgH7XCPBNQLnLXVLDqAPlrYDy1A6jkAPlPVohPKW7aERA8Q4UP82s0ZIkoAjCA/6qA+WvEJ7aNoPlQLnLgRJLIKgtFygHSs0h81cSOSAXUGpJ9MptKtvobH9VOazH84DKBkBBAaz//Q9QlQih7i6g6xBUTX8d+KQmUFEOW5dOwIIFQEKC6kACLZTb9RTC88aBMpoDZSSJXE1AaR1vsx2ocBNQoeZAaYnjqCjfHShvQ3hqBDsHigQUEfL461Euypt6oHOgtPJ9zMyBcjcKz2r1TUDpdaDU5lPr6Nyt1ymgLl0CTp5E1LL/ASVd4GjYB0Cyy/zK4y6KRf6Gq3SRVNt44QJsH30ObGgKdO4MxMUBx4/D+vP7sFQcBWrdANjtsFWVsX2YMwf4/VlUHc0H8DgsTZvC8vsyoFYt1X0LlANlVg6UOyHmTkB5cobIgVJ/HUj476TyPPM5UOL9Q08FcDUHytsQnhrkQBGEB/ydRG5WCM+TgNIbrjLDgdITwlNzoIyUMdDClxwovp2ydVaWw/btN8CCw4DgQDT+ACDAMX4esLUWq7wdG+ucXymgrFa230oB5VbAHD4IDB4I26EYAHcCa7OlZXERVqsFKDjP3sdEAa06QNi5Cjh6FFXJTYGBd8HW7gpYkqWd9FZAKd0yNaFgZhK5nkrkyvV4UwdKbb2+QDlQ3qPlQFmtQJcu7P7YrBmwfz+bbtSBKi1lr+PizHOgSEARhAf8+SgX/r87B0qt81K+DuYoPC0nyV0IL1AOlM8CShBg/eA9WNevBtABqJmIqL6Dgc114TixB3jvPeZM3XAD8NJLQJ8+cDR/GUBttv4VK2D7KxpV7buiMi7OuVq3DlRJCaz9+wInT8Bavx8Q25zVY6qoAAQHrF27wDr1fuD9TYDDAdsN7YCKVAh9o3BpswMbhr8PnGkDq+IO662AUgurKJfXmwOlZxlvHKj69dlyl6suqH5//C2gwt2BMlraxB9tUCtj0KkT+xM/5/+7Q5ynvBw4z35rICUFyM2Vb9NbSEARhAeC5UDpEU1qN4BQH4XHO1Bay3l7jJXbNEVArV4N6/q1sEZZgb/fDbRujegbASwQ4DjXHnj3G+Djj9kfAOzcCUeMFWh4Faz/+wo4tAg2PAlkrUDlvXcA6S2cbVDtrASBJXyfPAG0aQPbJ98Ci1PYZyUlwIkTsA1sAWtSFNC3L9uP+gCOAcKtQzG//1Ds2OF6LNT2Ta+AUoo9s0J4vuRAKV83aABMmSItK557h4OSyD0RCrlb7kJ4PEbcI3Hes2fZ1yo2FkhIMM8pJAFFEB4IVBkDpbOiJ2znqTPihUyojMLT40DpuTGZ4UBp5Ts5OX8eWL4cVjhgfeghIKU1gMv7bbHA0TMDaDgDeP55trKHHgI2boRjTQVw+BAsOADYbLDVSgbOXkLlJ58B7ToAjRvDdngzrNO/AWreA9x6K6vHBLDHquzZzTby9dewJaZI7alRA2jdGtYY9WtCEOAUT/x0rWNmRECpiZdg5kCpvVarVO1weBY21T2EZ1ZIy4w2qIXweMRnVHPPqtZEmQNVt64UUldbt1EoiZwgPMAnLao9tkQvngppKqebnQOlN4lcC7NG4fmaA6VcnxJ3zomy3W7Xe/Ik8MMPQFUlrO3bwTpwALCZfSR21FVVYLbHk0+yRl8eJuQYvwPIdcD6txuAAc1g/V8K8O5nEI4fA3btBHbthA1nYUEhUHwC+PBD4Ior2PLbt7GVP/EE0LkzbDnq7de6JuLjpXwPTx2irwLK1xCemQ6UGjYb+85SCM89oeRAaV1rIk2aAPfd56zE4Rbl/tSt67otXyAHiiA8IHbm0dG+CSh3j3Lh8dcoPF+TyJW/tvjEYm8KaXqbA+UJIw6UGhYLgKIi4LPPgLJLQHwCLA/fA6vN4vzcZb/5DVmtEDp2AuoB1msBNABssQDuvRfYvp2te/du2Jo1hWX8GOClbcDBA8DuXVIjatQEnn5Ys61K8cs7UDVqaAsofzpQ3iSRm1nGQA13I/O02ucLoSBEvCHUHCh359ZiAZo317dO5f6IAqpZM1Z0v0MHY20cPJg9bzsujv0nAUUQHlAOm/VXCE8kUGUMfA3hCYK0fbMcKCMCyowQngtHj8IycSZwIpm9r98AGDEC1vQ42Q1eSziKKPffZgNgtwM9erAJ112H5r0u5y4N7wgcOwbk5LCK4UePsof91qih2ValA8ULqIQE7f0MhRCeHrfHmyRyNZTfMXKg1AkF4acVwvPl3Gg5UMnJwLhxxtfXsyf7+/VX9p4EFEF4QPySeCqe6Am9ITxvyxiobcffSeRaAkpvDpSSQDlQqh1FaSlwww2wnOgAIBmIigZuuw2Ii5Othxcvx44By5YB/fvL16kqoDhGjmS/grdtAzswjRuzPw/7o7Vf4mulgCovly8XCgJKTyctPihWeS16GwoMlIDi6xWFE6HgQHn6QenLOkVEAeUravWlgkGYXWZEdcQsAeVLCM9IGQO1bWoJKDXMEFB6C2leey2waxczXvhtB9yBEgTgp5+A/fthqd0HGHg38/gvF6DkXSflOleuZEOjxWHWfPu1OiYjv/iNOlD8cRFDeSJmCyhvcqD07LN4jJXXnJ4fFTy9egGHDrESB/x8NptrcUUz6NaNHfN27cxZX6AINQdKbbov6xRfJyd7vy4eZemZYEECigh5/O1AaXVqoR7C47fvjQNlsbCRNI8/DixcCKxf7zoKT82F8ITy+Kjto0yUlJYCv/wC7N8H2O2wPvs0UChPslCkOLnc1EUBKKIUUGp5HPx/d2jlQGk5UPy5EEtjKLer9V65Df61WTlQevZZS0AZdaDEkIuyHXxFa71t0kOjRsBdd5mzrkASCg5U/frsPNSvzx7jKGKWA1W7tnn7d/XVLBofH2/O+ryFBBQR8vhLQGn96gtUErmWuDBDQBUWsv+X03hk7VOKK35YsTKEJ46iUsMUB2r1auD9bOBCCWC1AR9+CEut5sAm13WqhfBEYmLk7z05UGrnWAujDpS761NtJKae7fqrDpS7+dXmM5pErtWOqCigrEx9vdWRUCi/0K4dG9Bqt7MfVCJm5UCZFb4D2H2Nv7cFiyAahgThGYdDEhRmJ5F7cif05D2Fag7U6dPsv1jaSNweP6+ag6AmoIzC76vbHKh9+4Cbb2biqW4KMGYMcN99qsfEkwPlrYDy1oFShtS8FVB6XaPoaN9CeN44UFrOm9p69HayvIDSWm91JD2d/XXtGtx22O3svzfiWA3+GjJTQIUK5EARIQ3fGfkrB0qrUzNaiTxUQniC4F5AqYX3+BAUvz5d9ZpUEMM/mk7Gpo1MPJ0/DzRKZ4VlLp9gbwRUQQHw5ZcslHfddeY6UFoCUE8IT4m3AqpGDd9CeEZzoHhXUms9RtfJz6d81l51d6BiY4HRo4PdCgl/jMJLSdGeL1whAUWENLxo8FcOlJ78GCMCiieQAmr9epawm5kJXLzI1s3/6lMrY6C2Hv6/tyEFtwIqN5epnKIioH17YNCdsh7Vk4BSC+Ft3gycO8der1njvxwoq1Wq16nlSvLXp7JejrcCKiFB/Xzx/43kU3nCrXOo8lpvJyseJxJQoY034liNSHegqrlxqs17772Hpk2bIjY2Fj179sS6deuC3aRqiT8FlFYnyk9X64C1Ok6zHCgtPAmoBQtYcbmsLDatTh15qEQZplPbJ6WAcjccXE+HbT1/FtZFC4E//gDmzQN++w349FMmnvr0YWpHkcygx4FSCjs+WfvECVeRaFYOVI8eQNu2QMeO2jlQ4ra7dwfuvFN9u1rvefQ4UHqFoDchPCMOFIXwIgtvxLEaUVHsvh0VFZkCihwoFb755hs89thjmD17Nnr27Il33nkHmZmZ2Lt3L1L5mAjhd/iO0FPxRE+IwqZBA6B/f2m6ljshfsY/y0s5vy8hPC3XwIgDpRRFAHDkCPuvvFTdOVDK9Yjz+OJA4dgxWO+4D9ZzKQAU48ozMlg1PJUHaml13O5CePyxuXTJdV2+OFCi0BUEoGFDoHNnNv3ECdft8CG8Fi2knBLldrXe8ygFVHGx63Ja+6e2D2rrdbdtT/WAvMmT4R0o8ZiqrZsILmaF8Gw24J572HlWfhciAdL9KsycORNjxozB/fffj3bt2mH27NmIj4/HJ598ojp/WVkZioqKZH+EOfChJK2EaaPr6tQJaNVKmu5JQInbV5tfrePgBRCfb6S2HX/kQIlD+pW/+PQ4UKblQOWeAL74AtZzZ2Bt2ADo1h3JN12D4YPOYdykOGD5cs2nkXoTwtO6JrQEhl7hIbZHbcSmlgPlrlCpWSE8owLKjBCeu++J3k62eXOgTRvgqqvU3VsiNDArhAewgrV6H/0SbpADpaC8vBwbN27E1KlTndOsVisGDBiA7Oxs1WVmzJiB6dOnB6qJ1QreCfFVQGn92tXzyzqQITxvBJTaZ8phvgHLgXrySVj/5QCEGrB26wrr5NeAfbGw1gFaPTLA4+JaAkqrkCagXZHYDAdKaz1qx48XUGrHzhcHShwYwC+nNxTpTQhPObLRDAcqPl6q08SXyKAQXmhhVggv0qHLVkF+fj6qqqpQr1492fR69eohj68uxjF16lQUFhY6/44dOxaIplYL/CGgjPyy9iSgAplErlaPSdw+X1NHJDbWdXvKNinXY0RAqd5Ys7KAN96AVagCWl0B6/vvwhIXqz2/zvUqc6C0BFSDBq7L8f+V2zAaetJyIv3hQPGfJSSoX2v+dKBuuAHo21d7OV/DPORAhS5mhfAiHXKgTMBut8MeiQHeEEAtBOZrCM9TJ+argPKXA6V8rhq/fTUBFRfnuj1AXw6UVw7UuXPAo4+ydvW5Fuj/N9hqA9Y8eVs94cmBUo6A49tbuzZw8qR8ObVtG3WgPIXweAdKrUyEcrta73n4XC5PITxP++GuQ+RzkUSsVqBpU/a3ahXbJ3dt96aTJZcjdDEzhBfJ0KFRULduXdhsNpw6dUo2/dSpU0hLSwtSq6ovgQjhGXWgtFwIfwso5XPV+O3zna2Itw6U+AcYGIW3fz978Nnu3UDdurDdmOlcr5F8I5f1cu1r0YKNgOvZU3tdSUmuy6lt20wBxZ9bMx0o/nwrr59AOFDK1+6cW286Wa3vERF8SNzqgy5bBTExMejevTuWLVvmnOZwOLBs2TJkZGQEsWXVk1AP4RmpRK62Xb0CShBcn/fGt8VICM9dIc0dO4B33mF1pACdDtTu3Uw87d8PNGkC/PEHomslAGCjrcwSUPHxrCzAFVdor6tmTdfl1LZtZgiPF8JmCijxHKjN60sOlFEBpSU2KYQXuZADpQ8K4anw2GOPYeTIkejRoweuuuoqvPPOO7hw4QLuv//+YDet2mHmKDy9SeT8e72hG345fzhQZWXuc6DUHCitEJ5aKJPfj6IiYM8e9tpjDlRlJasifu4ccOWVrDRBvXq4IZ6VU2jUCNi503Ub7lCbTzlNq102G/tThin1OlBxca7CBWADBs+dkws0NQcKMCag3B0TPQJKrzg1KnbUrnEzksh5yIEKXSgHSh8koFS48847cebMGUybNg15eXno0qULFi1a5JJYTvgfMx0orRwoPQ6UJ+dBbb38Nn0VUKL7FBMjz4US26KWH2XEgVK2Q1cO1KVLwPjRwIYNQK1awE8/AZe/I82bS0OXzXKg3L3nl61RQ3qYsicBpTwGI0YAv/zCEqh5/v53JqDUKrtrhfB8HYVXpw5w5oz6vEZzoNyJHbUcKHfbcjePEShMFLrQudEHCSgNJkyYgAkTJgS7GWFLcTHrV7t31yz3o4tglDHQG8LTKjToj0rkooBKSFAXUErE6r9q7XNXxkBE3AfZPpaVsYViYoCSElgybwTWrWQz/fe/rkPgFOsOlIBKSJAElNb8ag5UVBTbhbFjXder9vR3f4fwhgxh27zyStd5/R3Cc7ctd/MYgUJ4oQuF8PRBAorwCxs3AitWsM7k+uu9X4+Zo/C8yYESXRx+kKWnvBp/hPB4AXX+vHbble1Wa5+7JHJlG2zHjgBfrmEO0+bNLEt71Cjgs89gObMWSE4GvvuOPdtOAzMElFY4Um1bSqGjtm014eEuYV5rW+J//tybNQqvZk3gppvU5/UlidxIG9xtg0J4kQuF8PRBAorwC2JOjlpysxGCPQpv8GAgJwdo3FiaJooprcoV/niUiyigtFwQJUrBAehz2kSqKgVgx05ELZoOXGgvfXA2H3jrTfa6fgNg+RKW1e0GXwRU27ZA69aeH8rLT09I0G6DcnmlA2UELQFllgOltT3ltrW2pbYdI66n2rJq09U+0wM5UKELhfD0QQKK8AtiwrO3YkckEDlQ7t6npbE/nqQk5gzUqqW+nN5ReN46UDzeOFBqy1oqyoHzF1jSzcGDqNizBygqhA2FTCgJArO+yi4r4xg7LD//5FE88dvxRkAlJwNduuhbTtyGHgdKrU1GBZT4gF/+Qb/+FFBmhPB8yVUyO4mcOunQhUJ4+iABVQ3Jzwc+/xy49lqWo+QPKirYf18FVCBG4blzoLTo0cN1WiCSyPUKKEMO1MKFsN7zMlA4UD5DjB3R99wLNLibJVU5HMD33wMlJcAttwBd6qhvXGO7ejtJvR3z3//OoofKbRkJ4fniQCUmAqNHs1Db0aNsmplJ5O6W9TaJ3JvO0F9J5BTCC10ohKcPElDVkHnzgIICNuLcXwIq0A5UYSHr0Nx92bVyoHzp1NSWU3OgbLbACihdDlRJETB8HPDll7CiNWCLYrZaejp74muLFogZHAVkcRsbNkxzfVp46ryvugpYt05ymvTevDt0AA4cALZskc/foQOwZAl7iKmyDcr1+uJAAaxMA7++QDtQRkN4etqk9bmRELAeKIQXuvjqLlYXSEBVQxRF1v2C2QLKXRL5oUPAZ58BV1/tOvycx5syBkbwJolcjYA4UIWFsEyaBBz/ErBYYL3nHqDZVBfbxIwnFHnq5DMzgfbtgYYNXdupt3Pnt5WQAEyZ4t7hMMOBUoN/qHGkhfDcJZFTDlRkQQ6UPkhAEX7BLAElDtnnK1or13n2rPy/FnpDeL7+4vL3KDweQw5UXi5wvApISQG++ALWM0eZ2/Tjj7Akdge+cF0mJkZ9/WL79SCG1NRCawDrSJs0UV+v3tCUcv7oaH3z8dMNPfdPgbi+QAsoo6PwfAnhmZ1ETi5H6EICSh8koKoZvPjwpT6TJ8wSUEVF7H9ioraAUj4AV4tghvD4DsydqFIuLyJWpdY7nN/Fgdq+HZaHnwYudgdi7EB5GSzJycDq1UB6OqwH1dfjTkDppX175mTxIxnd4YsDpTW/Hgequgoofr1/+5u8WCi/jNlJ5ORAhS4kbvVBAqqaUVAgvVY+eNVMxCRyvkPxBn8IKH+H8NTaptYJGXGg+FwwHq2bm8ypOnUKuPlmWC5ergVQXgYk1ID1lZeYA+VmPe5CeHqPk83GShHoxdcQnhp6HChfOopACCi1kJne0JzeRP5OnbSX9WcSOQmo0IIcKH2QgKpm8I+G8Ceh6EB5U8bACHocKD0CSg2txGTl+2uuUQiWCxeA224Djh6FpeG1wE1jWZGu9HRY29s01yNihgNlFCOiRq/4DVUHyohoMyMHypdyA/5MIieXI7QgAaUPElDVjNOnpde+iht3+ENAiWGsUHWglNvh2+SrA6U1NF7Z9iZNgFatLr8pLWUFq7KzgeRkWGa8ChyUilrpuUmakQNllGCF8Mx0oLTCs74cMzNCeN60yZ0DZbGw69XXHCjqpEMLCuHpgwRUNWHvXmDPHuDECWlaqAuoykopeToxUapqbnYOVKBG4SnX7U0Iz1PbnYJHEIAHHwSWL2eFihYsgKU8HTiovqw3ITx/EawQnpkOlFY7vHUg1ZYF2Knl/3ta1swkcnGatwKKQnihCzlQ+iABVU349lvXfKRQF1DFxex/dDRLivZXDpTZITx+e8oOVel46BFQ7moLKd9H2xzA78uBN94AFi1iY/N//RXIyIDlT+32ehPCCwUHypsQnpbzEQwB5c02+Ndt2gAjRkj1qLTwRwhP/MzhoBBepGGWMxvp6BJQXbt2hUXnt33Tpk0+NYjwD2rJ3P4UUGZUIi8sZP/FApn+yoHylwNVVSW9FofVKztvZacoCOoCSkTZySckMJFTfjIf2LwZMR9OBnK3sw+jooAPPgD69lXdL08OlMXiWg4gEPg7hKe1frNDeO7mU742sg1+/VYr0KKF52XFUZlq5S30btddThc5UJEFhVf1oUtADR061Pn60qVLeP/999GuXTtkZGQAANauXYudO3fiH//4h18aSfiGILCblXhzb9sW2L3bfwJKEMxxoPj8JyB8HChxO6KIBKQije5COFarXHSJKBOT69cHcnOBGgkCYr7/Gtf9vBaL1iUDAKJxlA2vvOMO4MkngZYtNffLU2ceHa19LPx5UzUiarwJ4Wntty8OlIh43Wuty2wHSi9paWwsgfK5jnrW5cmB4v8bgTrp0IVCePrQJaCef/555+sHH3wQ//znP/HSSy+5zHPs2DFzW0eYwqVLUid8661A7dr+FVAOhyQCQlFA+SsHSrkdsTO1WKQO1RsBxe+X1QrcdRcr39Rzw/vAPRNwFSzIw62o6N4LiVM+AW4aomo1GHWg3Akof2KkY/XVgdKaxyjBCuEZWbZzZ2PLKLflTweKwkShBYXw9GE4B2revHnYsGGDy/R7770XPXr0wCeffGJKwwjzEBOxY2OBrl2lRHJ/CShROPi6DTEHSkyQDbcQnuhA8ULEk4Dil1e2GwBsjgokHdiGGw+sBqZPYstNfBRDJ0zwGMtxJzbUbpIxMcF3oMzKgdKzTjNyoLTaZaQdWgTLsdEzgo9CeJEFuYP6MCyg4uLisHr1arRyjpVmrF69GrHeBNgJv6N8DIhWR20WegRUTg6wciUwaBBQp476PKIAEUeChUsIT0RsP/+MNd0CKjcXeOop4PRpVHW9Fjg6DMjPh6XTA8C+vdJCo0cDM2fqarQ3DpS4nL+uFTV8yYEyGsLTs6weAiGg/BVW0XuM1faJQniRCYXw9GFYQE2cOBHjxo3Dpk2bcNVVVwEA/vrrL3zyySd47rnnTG8g4TslJey/UkD5y4Hic3+0KpGLRmVBATB+vPo8ohATBUioh/CUDpTYfj4RWylaVAVUcQlw/fUszgrAsXgNgApY4YAFe5kll5bGDtwjj+husLtOXisHSvxMKaBCxYEyM4QX6g6Uvzq19HRg/37tkhXuXCYK4UUmFMLTh2EBNWXKFDRv3hyzZs3C559/DgBo27Yt5syZgzvuuMP0BhK+cfAgsP3yoCzxYa7+ElCnTrG8peRkaZqnbbirjK5MyPWXA+VJUOnFXQhPOY/42uVGVVwMfPExcGo3kJoK/POfcCxZB+ypB1tqLWD4DGDsWKBWLa/bJ9ueymsRXkAFEiM3b73nLtAOVKglkbtj6FBWa7VrV/XP/ZVETiG80IXcQX0YElCVlZV49dVX8cADD5BYCgMqK4H//U96728H6oMP2P+//12a5mshTcA8B8rfj3IRcedAuRNQlu3bgC/nQ6goZkPtVqwAWrVC1VgA/wasdgBTvG+XNzlQastpTTMLcqDMX9YdCQnAgAGet0tJ5NUHcqD0YejQREVF4fXXX0cln+RChCxi5W4Rfwoofl1nz6pPN4o7AXXxomsyfLAf5cIvJwgGc6DKy2H9fh5QUQ6hUWNWQfxynqFWEU1f2qdcn9q61UYOaq3LTIIloAKdA2V0e8HKS/FXEjm5HKEL5UDpw/At4/rrr8eKFSv80RbCZPwpoCoqgA0bpJFyYqI6v35329BToFFLQAHAv/8N/Pe/wJEjvudA+TuJXJcDtX49rCWFQK1kCGMeAq64wjmf1nPwfG2fpw7Ml87RF3wpY+BLCK+6OlB6t2t2EjmF8EIXErf6MJwDNXjwYEyZMgXbt29H9+7dkSD2ype55ZZbTGsc4Rvl5fL3Yg6U+IXwRUAtWwasXQusWgVMnCgXUPx2tbYRFycJDK1nabkTUKWl7P/+/aHrQOkO4Z06BaxYASscQJ8+EKzyntxTbSFv2qdcn9HwTKg4UHrPnad9BbxKK9MkEgUUJZFXHyiEpw/DAkqsNj5z5kyXzywWC6q0hl0RAUcpoJSPE/FFQO3bx/4XFLD/4kg/QO58aW0jNlYqlFlWpv6ICaWAiolhbefXabeHXh0oQDuEx4dDLJUVsDw6GdhZm312RUugc2fNOlChJKD8iZGbtxkhvLvvBk6elJl+hgmEAxWKdaDIgYpMKISnD8OXvcPh0Pwj8RRa8ALKbgcaNmSv+Zud2FmfOgXs3Kl/3cpwh1EBxX8pefeKRymgLBYgPl4+jxEB5e86UMptuQvhWSwAfvsNlp2Xh0impML6wP2A1aopoMwO4fHvo6NZ8j8/ACAcHCgzQnitWwP9+/u2T8plw2kUnt7tmi2yKUwUutC50YdhB4oIH0QBlZ4O3Huva0FKgHXONps0gi4hAWja1PO6lTdTXgRpCai1a9lnffvK60OVlKgX01QKKIAJKF6sRUfLBZRWOJB/SK+/60CJ23MXwrNYAMydCwsEoFt34KabYKllAS5qVyL3pwMFAB06sP/ffafS1gDi7xCeP/ZHr5Bzt4yR+QMZVgmEA0VhotCCHCh9eCWgLly4gBUrViAnJwflijjRP//5T1MaRviOKGRiY+VF8tQElEhurj4BZcSBEgT2f9EiNq1zZ3m1cn5ZHi0BxSOum3+vJaBE/OVAaSWRq43Cs1woAebPB9Ab6NkTsFg0K8QHIgfK3fzh5EDpEVD+oLomkZuVA0WddGjBn1c6N9oYFlCbN2/GjTfeiNLSUly4cAG1a9dGfn4+4uPjkZqaSgIqhBC1rVjPR8TdKDm9eVF6HShAHs4S26V0oNRQE1CKMQtwOORtdjjUb/TuBJS/c6BUHaiVfwIVFbC0a8cKZsI1tKpM9vfnKDwjy/kbf4/C8wdGjq3Foi309W4jVHKgSEBFJr44i9UJw4dn0qRJuPnmm3H+/HnExcVh7dq1OHr0KLp3744333zTH20kvEQUUMpHNJgtoC5dAs6fl7/nqaqS52NVVnovoJQOVFWVq4BSIxgCSvNRLgUFsP6VzZZ54H7VdvDtDVQIT4mejtMfGAlVeRPC88dz/YwKKLVljGwjVEbh+dLR6hkZSQSH2Fh2TpT3W0KO4ct2y5YtmDx5MqxWK2w2G8rKypCeno7XX38dTz/9tD/aSHiJlgPF3wi9FVD8r8eZM4Fjx6T3SgfK4ZALqPJyeQhPLEmgXEZsi9kCyl91oJTbcwnhORywHDkMfPklLFUVQN++sFzZQ7VdCxawkWGA/0J4WvvZpw87zn37up/PX/g7hBcIAeXOLQxXAaV2/XXtynIsGzc2vl5yoEKX2Fhg5EiWO0toYziEFx0dDevlb1JqaipycnLQtm1bJCUl4RjfixJBx52AEsMIaiEwPfA3U2W5BDUBxYfwysrkDpTaNnmB5S6Ep1dA8dP9dbO2WOCMIbqE8M6fB+68E5asRgAawxITA3z8MSy5UmP4Y7pxI/t74YXAh/Cuu04+Ki2ccqD0hPDIgTKGO5epe3f25w0koEKbJk2C3YLQx7CA6tq1K9avX49WrVqhb9++mDZtGvLz8/G///0PHcRhPERIIAoZpYAC2M1QFB+8WDl4EDh8GLjpJmdqjpPKSpbrlJTk/oanfNKP0oG6dMl7AaV0oMQkdf797t1s+Y4dpenuQnimUFkJyzPPATNrAL16AU9eh8pKtqHo44eBezKB/fthiRoDpDaE5ca+QIsWsOR5blcwQnjBHIUTjg6Uu+0p8VZABWtoefv27LFJbduau14SUES4Y/iW/Oqrr6J+/foAgFdeeQXJyckYN24czpw5gw8//ND0BhLeo+VAAfJimryYOX4cyMkBvv3WdZkPPwTefhvIy3MVSe5QCihlyM6dgLJa5R2HpxBeRQXwzTfA99+z5+WJuAvh+Ux5OXD77bC8/hpQWQGsWgnhvpGo2LkP2LwZUffexUqmN2kCy3PPAmPGwHK5aqOnfJ+KiuCNwtNaTmuaWfijjIEyP81svBGnRs9nsERtejowerRUR84sAi1qCcJsDDtQPXpIORupqalYJI5NJ0IOrSRyQLp58cnOPOfOuU47fZr937HDdZlOnZhQ27DBdTllCM+IgIpSXKGeQnj8aEC+jX4L4ZWXA8OGAb/8AthjgY5XAhs2QPj+e1QgFkADROM8cO21wHffwfJbKnBY3YVQ61DPnqVReO7wdp/MJpJDeP6Cv55JQBHhiOHbzSeffILDhw/7oy2Eyeh1oNQElLtcKLVl4uO1O3hfHCilgPLkQPEj+vibsl9CeOXlwB13MPEUGwvLLz8DQ4YAY8ZAuCETlXXrA/UbIHrCWGDpUiA11SWfxJOAOnPGfyE8vcchlEfhebNPoeJAkYCSXvvyWCmCCBaGHagZM2ZgzJgxaNiwIfr27Yu+ffuiX79+aNmypT/aR/iAXgFlFC0BpSxfwM/PCyjlo1vEPKbFi9l6+vTxXkDx61YW2AT8IJ5+/plZfD//DAwcCEs2IDRoAOGtH1DxCYDzQNRoADHy7et1oM6ckfY5lEJ4/sQfITweGoUXGlAIjwh3DN+S9+/fj5ycHMyYMQPx8fF488030bp1azRq1Aj30pjHkMJTEjmg7UC5Q22ZuDj3DpSnEN6RI+xRL7//zqZpCSibTS6ilKMIeQeKz+3SeoyLV2zaBNx6qySefvkFGDgQgLxj8/gsPHgWUPn5oRnCC5UcKOW8oSKgyIHyjLtyKgQRDnjVnTRs2BDDhw/H22+/jVmzZmHEiBE4deoUvv76a7Pbp5sjR45g9OjRaNasGeLi4tCiRQs8//zzLo+a2bZtG6699lrExsY661cpmTdvHtq0aYPY2Fh07NgRCxcuDNRumIovITx3VFX5FsJTc6AOHpTe83lZSgEFADffLIkodyE8tfIGPnc8c+awcduLFsmcJxFx/bqehQfPAurcueAV0tRazt/4IqD0QAIq9KhRI9gtIAjjGL4lL1myBE8//TR69+6NOnXqYOrUqUhOTsZ3332HM2fO+KONutizZw8cDgf+85//YOfOnXj77bcxe/ZsWXHPoqIiDBw4EE2aNMHGjRvxxhtv4IUXXpCNHlyzZg3uvvtujB49Gps3b8bQoUMxdOhQ7NixIxi75RPuksjFG7ByFB6PVkdj1IFSViLXcqD47boTUG3bSkUe9QooU0J4v/4KPPgge33rrcCqVUBmpuqsqoU0YVxAXbgQ+EKanpYzsqw3eCug9B6fSBBQkVK5e/hw4JZbgJSUYLeEIIxjOAdq0KBBSElJweTJk7Fw4ULUqlXLD80yzqBBgzBo0CDn++bNm2Pv3r344IMPnI+Y+eKLL1BeXo5PPvkEMTExaN++PbZs2YKZM2fioYceAgDMmjULgwYNwhNPPAEAeOmll5CVlYV3330Xs2fPVt12WVkZyrjqkUVFRf7aTd3wYTNvHahLl5gwUlu3mgOldVNXhvD48gIA++z4cfX1qwkovv3ucqDUQnhed/wbNgB33cU2dv/9wMcfq65MTZi6PMqF++9JLJSWSusJpRCePzHyiA9vaiOFq4AKVh0of9KqVbBbQBDeY/h3zMyZM3H11Vfj9ddfR/v27XHPPffgww8/xL59+/zRPp8oLCxE7dq1ne+zs7PRp08fxHCKIjMzE3v37sX5yw9zy87OxoABA2TryczMRHZ2tuZ2ZsyYgaSkJOdfenq6yXtiHN7xMVIHiod3ivhOx9cQnpK8PPl7PQJK3JZRB8qrX+5HjrDKoqWlzHH6z380ezBxMi8Y9Ybw1HJ4qqokwRlKIbxQdKBCSUBREjlBRD6Gb8kTJ07EDz/8gPz8fCxatAi9e/fGokWL0KFDBzRq1MgfbfSKAwcO4N///jcefvhh57S8vDzUq1dPNp/4Pu9yL641T56yl+eYOnUqCgsLnX+h8EgbUbBYreo3c091oADtEW28OBDxlETuTkApt+8phAdI29KbRO51DtT588CNNwKnTgGdOwPz5skVkQJx/fz+6g3haSVBFxez/ySg3M8bzBCeEsqBIojIx6tbsiAI2LRpE7KysrB48WL88ccfcDgcSPFDIHvKlCmwWCxu//bs2SNb5sSJExg0aBCGDRuGMWPGmN4mJXa7HYmJibK/YMMnkKvdbPWE8HgHihcjamIoOtpVQInCQRnC84QvDhTfNp9zoObNA7p1Y8+GadSIPeG3Zk23iygdKJtNveMzIqDEiHB1CeF5W8YglBwofwgod9sjCCLwGM6Buvnmm7F69WoUFRWhc+fO6NevH8aMGYM+ffr4JR9q8uTJGDVqlNt5mjdv7nx98uRJ9O/fH71793Z5tExaWhpOnTolmya+T0tLczuP+Hm4IAoeTzlEekN4vBhRPixYRNnB2+1MCHlyoJT4kgOlXI+I4RDewoXAnXeyBdPSmHgy8CwLrbylUHKgIi2JPJQElL9CeBYLaz8JKIIIPoYFVJs2bfDwww/j2muvRVJSkj/aJCMlJUW3s3XixAn0798f3bt3x5w5c2BV9DgZGRl45plnUFFRgejLYZisrCy0bt0aycnJznmWLVuGiRMnOpfLyspCRkaGOTsUIDw5LnocKK2EbC0xpOw0YmLYOrQElNWqLn6MhPDclWDwOoS3Zw9w992sIaNGAe+951rBUwNx/WK7IlVA+RMjYblQGZnmjYDypr0koAgidDD8FX7jjTdw0003ISkpCZe0Sk8HgRMnTqBfv35o3Lgx3nzzTZw5cwZ5eXmy3KV77rkHMTExGD16NHbu3IlvvvkGs2bNwmOPPeac59FHH8WiRYvw1ltvYc+ePXjhhRewYcMGTJgwIRi75TWeBIMvITx+FF1KilQGSU1AidtQC+GpJbcr2+RJQLkLDXoVwrt4kY2rLipiz6/7z390iyd+/VqlBzyNwnPXPjNDeFZrZDhQoRrC07pu+Xm9OYZmhP8IgjAHwwLK4XDgpZdeQsOGDVGjRg0cOnQIAPDcc8/h448/Nr2BesnKysKBAwewbNkyNGrUCPXr13f+iSQlJWHJkiU4fPgwunfvjsmTJ2PatGnOEgYA0Lt3b3z55Zf48MMP0blzZ3z33Xf46aef0KFDh2Dsltd4EgzidL0hPDUBVaMGMH480Ls3e+9OQKk5UFoCyogDZaqAcjiA778H9u8HGjcGvvtOu5Ea+OpA8cTGyt+b6UAZWVegO2tvHyZcHQSUmvAmCCI4GA7hvfzyy/j000/x+uuvyxK0O3TogHfeeQejR482tYF6GTVqlMdcKQDo1KkTVq5c6XaeYcOGYdiwYSa1LDh4yvkRp3//vXYO0cWLTKBERamHw9Qes8Kvn08i10o8V8MsB0r3o1yKi4EDp1itp4MHmOP0889Aaqr2yjVQOlC+hPBq1pQ/XzCUBFSoOFDhHMIjB4ogwhvDt5zPPvsMH374IYYPHw4bd5fo3Lmzy2g4IngYyYHS4swZ4LXXgG+/VZ/PnYCKipLenz0rCajLqWYAtM2dHTuAdevUt6Fsv7scKF2Pctm5E+jaFfjic2Dv5et37lygSxftFetAK4TnrgNUE1A8ZobwjHTAoTwKz9t9MhsK4RFE9cOwgDpx4gRatmzpMt3hcKDCyFh1wq/ozYFSQ7z5nz/PhMDu3ephPk8CStzGH39I07m6ppoO1LJl2ttQbsunEN6ZM8CgQexBfLFxQFQ0cP0AwAf30cwkcuXzwaqjA2WkEnkoOVD+ElANGgAJCUAIVEohiGqP4RBeu3btsHLlSjRp0kQ2/bvvvkPXrl1NaxjhG3pDeGrEx0u1h0SMCqjoaPVt1K4tPThYT3qRVifDF9LUwu2jXBwOYMQI9gyZK64Abn3EULK4p/ZqhfCMJJErHahQElD+JBwdKCX+CuGNGMGuLTe1XAmCCBCGBdS0adMwcuRInDhxAg6HAz/88AP27t2Lzz77DPPnz/dHGwkv0BvCU6NXL2DJEvk0XxwoHv6Xs55OQEsg6Qlnua0D9e67wOLFLFP7+++B73wXT4CrA+UphOeutEBCgvwzs0fhebOcu2lmYbGwuqVaz2LUakd1COFZrZHzIGGCCHcMfxVvvfVW/Prrr1i6dCkSEhIwbdo07N69G7/++ituuOEGf7SR8AJvQnh9+wIPPAD06OH6mRkC6sYb5WEpTw5UYiJLT1JDTyeimQO1axfw1FNswptvAh06mNb5mplErjTEqpMD9cADwD/+Eb4hPH85UARBhA6GHSgAuPbaa5GVleUyfcOGDeih1vsSAccbB6pGDTZ6XxCYAOITtNVG0RkRUD16AFddBRw4IE3z5EDdc492VM2lg+KrC166BFitqKqKkX0MAJaSYlYk89Illv/0j3+w6RZzh7frTSJ3J6Di4uTtMlNA+ZpE7u/O35vHzFQHB4ogiNDBsIAqKSmBzWZDHOetb9myBc899xwWLlyIKq2iQkRA8dTpqt28RVFisbDOW6yADag/vsVIDpRY04jP6/HkQKm2vbISyM6G7a3ZwMpeQJMmbOUbNzIFGBMD5OUBNhscu+KA082BW2+FIFiBoiJY/z0DOLcNqFsX+OQT03sxM5PIrVa2a2LdrVAK4YUKJKAIgggWum+jx44dQ0ZGBpKSkpCUlITHHnsMpaWluO+++9CzZ08kJCRgzZo1/mwrYQBvHCj+pq/MPVErOm/EgRIFFB/C8yQIXNr46adAs2ZAnz6w/fw9cO4ssHkTkL0GKC9j7/NyAQhAVSWqfv4V+NvfgH794Jj1b+Dbb2E5lw+0aAGsWAFwRVZDMYRntcrPQyiF8EKl8/d2n/wNhfAIIvLR7UA98cQTuHTpEmbNmoUffvgBs2bNwsqVK9GzZ08cPHgQjRo18mc7CYN4kwPlTkAZdaCUAspuZ//5kJzWQ4lV2/jrr+y5dACQnAzbTbcBZXcCR44A+flAx45MpVksQL16wNmzcBTWApatB1auhLDyNIC7YLHbgUWLAJVSHGbgKYnc0yg85bx8NfJQElChgjePcgkEehyoUBJ8BEEYR7eA+vPPP/HDDz+gV69euOOOO5CWlobhw4fLHrpLhA7elDHgBZDZAkoUAnwnp1tAXboEPPooe/3QQ8CsWbBGxwIvAWjbVn3hWrXguGYiMPMm4H//g7DHAuzqCEtmW1XxFKoOFC+gQqmQZqiIlVAJ4SkhB4ogIh/dAurUqVNo1qwZACA1NRXx8fEYPHiw3xpG+IavITylODIrhOduec02vvEGcPgw0LAh8NZbQGwsrDoSvquqwMTS9OkQdgH4FrA28bSUOZiRRE4OlGdCNYTn7piRgCKIyMDQLcfK3aGsVitiDD5olQgc3oTweAGkHJGm5hZxKUTObYnbi46Wr48XAkOHAunprGyCOywWsGF7M2awCW++6Uyislg8OzK6HuUC99ONYnYSOeVAuSdUQ3juIAFFEJGBbgdKEARcccUVsFz+1peUlKBr164yUQUA586dM7eFhFf46kApC1iqCShl9EwUNZWV7h2oLl3Yn1BUDOw6xhLDAfYQvMREoHVr1sbtW4G/D2bD0Pr2Be68U7Y9m029PlVMDCu74PFRLn4gXEJ4kehAhXI7eUhAEURkoFtAzZkzx5/tIEzGmxwodwJKGcLr00fbxRIFFN9BuITw/vgDlvvug/X4/XDE12SKp/Lyg+16Xw2UlsI6YyRQXgi0bw988YVLj6O1b3Y7W50oYioqgOxs9jrQDpTRJPJwCeGFSucfqiE8d5CAIojIQLeAGjlypD/bQZiMJ8fFXR0ogFUA37tXei86UG3bAldeKZlGWuuIipIX37TbweozffMNsH498O23QEUFrBbAUXqBzVSjJlBSDKxZDQCw4iJw003AZ58Byclu28tjt7MaVqII3LABOHmSvU5KUl/G7M7MHw5UdSqkqRcjIbzevYE1a4CePf3bJk+QgCKIyMCrSuRE6ONrGYPWrdmAt8WLgaNHJQFVowbQvLn2dnkBVVJyeWJVFaJ/+A54+CH5U4rvuAPWlk8Bm3cCqamsKOayZcCpU0D9+rA+3hO4KVNTObgTUIB0DMRClImJ7HEy/kQZwlM2XRREYhuN5EBRCM8VI6LwhhuATp3YpeZPPB1bElAEERmQgIpQvAnh8flEFgvQoIFUt0kUUJ46cfHz6OOHUf6v74D8RsDRo7CUTWUfdO4MDBsGdOgA3HwzrK9b2TNeRAYMkNo4BG6HOXgSUOL+iP87dNB+fEygksg7dmTCTswfC5YD5etyodL5GxGFFguQlubf9gCevyMkoAgiMiABFaEYTSJPTQVq13adT3SlDAmoqipEPTsFFQcEAJfDc3XrAg8+CLz4okzFuOv0PHWIeh0orXCaP/CURG63y/ViuIbwQoVQTCInAUUQ1QMSUBGKkRDeXXcBV1yh3kGLWkcUZG47B0GAraQIWPgnog7sRjk6AX37sZjff59TXdjds/o8dTDuksgBVwdKT3FDX/EUwnO3XSqk6R1WK7veQyWJnEJ4BFE9MHzLefHFF1FaWuoy/eLFi3jxxRdNaRShTtn5UpzfdBjFe054nNeIA6V88C+PstilZudw/jyQkQHr008BmzYiCpUoH/so0K8f0LixZu9vJMSoJBQdKBGtEJ4Sd6LGapX2xQwi0YECQk+QkANFENUDwwJq+vTpKHFmB0uUlpZi+vTppjSKUGfto19hVvdPsXzyLx7nNZIDpZUXpPaZaudQWgrcfDPw119IQiFQoyaSPn4b3cdeCQBo2lR7/aEioAJVxsDddi0WeQFTq1X+7EBf69ZGqgMVaoJEr4AKFceMIAjvMBzCEwTBWUyTZ+vWraitlkRDmEZMShKAY6g45ypglRhxoNw9UsXd41oAsCJLd94JrF4N1KqFWxc/g/y0DmiQDjQAkJLC/rRQDkM3UvBSq6MSw15VVWwEXijlQGnNL75WCiibDfjnP9l0d0LXSNsCtVygUNbWCjbkQBFE9UC3gEpOTobFYoHFYpFVJAeAqqoqlJSUYOzYsX5pJMGISWVFjMrPX/A4r5FHl3jtQDkcrNbB/PlMtfz6K+Kv6oDG3CwNGrhvp1LIVVS4TtfCkwN19Cjwf//Hyhe4mx/wXw6UEdGmJqAA9eR+X9qmfG1kOW+W9zehJkgoB4ogqge6BdQ777wDQRDwwAMPYPr06UjiKhLGxMSgadOmyMjI8EsjCUZ0CismWV7omoOmxEgIzysH6vx55jxlZbGJ334LXHONx3Z5aocRAeUpiVxELD0VSAdKxEgID1AXUGZBOVCBgRwogqgeGK5E3qxZM1x99dWIctfrEn4hJo1ZERWFFz3OazSJXAtVB0oQgNGjmXiKiwM++YTlQHmBlpAzw4HSO7+4zAXPxp5hfA3hmUmkO1ChEsIjB4ogqgeGbzk1a9bE7t27ne9//vlnDB06FE8//TTK+Wd3EKYTU78OAKD8YqXrw+kUeBJQ/LPujDhQ1r27gREjgB9/ZOpq+XJWB8FLfBFQWu3WEi3uxMwdd7BSVYrnFRtGebx9EVD+7GAjyYFSe65gMCEHiiCqB4YF1MMPP4x9+/YBAA4dOoQ777wT8fHxmDdvHp588knTG0hIRNdJBKw2lCMGOHPG7byecqB4AaXLgRIE4OefYbt7GHuwL8ASjPiqkF7gi4DSeiSHNwIqLQ2YMEGqEO4tvoTwlALKbCKxEjkQeoLEk4Bq0YI9EqlJk8C0hyAI/2A4Drdv3z506dIFADBv3jz07dsXX375JVavXo277roL77zzjslNJERi7BYgIQEVxYXseXHp6ZrzesqB4h/botnJ79qFqFufAApuZE/hPXgANiuAu+4BHnkE6NXLq/3g8UVANW6sPl1r2WDkQPniQPmTSHKgQi2E5+mcd+rEHukT6seVIAj3eFXGwHHZvli6dCluuukmAEB6ejry8/PNbR0hIyYGQEICyotjmIByg6cQnvK5dy4Lf/UV8OijiMqvCeBKIJ85XranHgdeHeVN81XhOxujAqpRI/XpoSSgQsmB0tquN/OGUucfaiE8PdduqLSVIAjvMSygevTogZdffhkDBgzAihUr8MEHHwAADh8+jHr16pneQEIiOhrMgUI0hFOn4e4ebCSE58LUqSw8ByC641VAu7uA3FwgNha2B3x3nXh8caC0Qo9aQikY4x7IgfI/4RbCIwgiMjDcpbzzzjsYPnw4fvrpJzzzzDNo2bIlAOC7775D7969TW8gIRETA6BGDQiwoDL3DNzVVTQSwpPx7bdO8YTnn0fUiCeB/8UDbdqw9ZncOfgioACgfXtg506ga1cgLw/o0CG0HKhIFVChIlaA8AvhEQQRGRgWUJ06dcL27dtdpr/xxhuw0Z3Dr4gOFACU557VJaAMOVCXLgGTJ7PXTz0FvPACohVRWbNPsa8CauhQlk/SooXkSGlFkimE5x2hJJbUCLUQHt0GCaJ64NVvtoKCAnz00UeYOnUqzp07BwDYtWsXTp8+bWrjCDkWCxCdGAcAKM8753ZeTwKqRw9W+6hrV27iu+8Cx4+z5KIXXgCg41EuPqIloPR2htHRzBzjw3nejMIzC3KgAk+ohfBCxQkjCMK/GHagtm3bhuuvvx61atXCkSNHMGbMGNSuXRs//PADcnJy8Nlnn/mjncRlYmrFowJA+anzbufzlANVsybw5JNcB79+PfDss+z19OnOB8rFxck79lBzoDytk4ccKPXtmjlvMKAQHkEQwcDwLeexxx7D/fffj/379yNWfGorgBtvvBF//vmnqY0jXImpycpsV3h4nIunHCiAu9ELAjByJFBWBtx6KzBqlHMeux24nOYmX8YPRIKAMrrNcBVQoSSqQs2BIgFFENUDw93U+vXr8fDDD7tMb9iwIfLy8kxpFKFN9GUBVV7kWyVyGX/+Cezezar7zZ3rokB69JBe+/NXvlkCKpxDeG5HRwaJUBEmWojXSrAdqD59mLPbp09w20EQRGAwHMKz2+0oEp/OyrFv3z6kpKSY0ihCm5iazPXzJKA8hfBkfPgh+3/PPUCtWi4ft2oFJCczg6pmTQON1QHvuESCAxWJITx/LG8moeJAXXcd0L9/8NtBEERg0N1N5eTkwOFw4JZbbsGLL76IiooKAIDFYkFOTg6eeuop3H777X5rKMGISWJJ5BXF+hwoj0Lk00+Br79mrx98UHUWqxUYNw6YOPFyKQU/EYkCikJ4/idUBFSotIEgiMCgu5tq1qwZ8vPz8dZbb6GkpASpqam4ePEi+vbti5YtW6JmzZp45ZVX/NlW3ZSVlaFLly6wWCzYsmWL7LNt27bh2muvRWxsLNLT0/H666+7LD9v3jy0adMGsbGx6NixIxYuXBiglnvGOQqv0uL2gcK6QnjbtgEPPMDsqrFj5bE6BTEx/hVPQGSG8CLBgQp1URAqITyCIKoXukN4wuU7e1JSErKysrBq1Sps27YNJSUl6NatGwYMGOC3RhrlySefRIMGDbB161bZ9KKiIgwcOBADBgzA7NmzsX37djzwwAOoVasWHnroIQDAmjVrcPfdd2PGjBm46aab8OWXX2Lo0KHYtGkTOnToEIzdkRGTGAvAwh4oXFjoHC2nRJeAeuYZJp5uvRV4//2g9JS8YOA7QH84UMHoYMmB8j+h5EARBFF9MJQDZeHuUNdccw2uueYa0xvkK7/99huWLFmC77//Hr/99pvssy+++ALl5eX45JNPEBMTg/bt22PLli2YOXOmU0DNmjULgwYNwhNPPAEAeOmll5CVlYV3330Xs2fPVt1mWVkZysrKnO/VcsTMIibWCsTEoKI8mgkojcfnuM2BOn0aGD4cWLqU9fD/938h0fvwTfBF7Kjtis0WmF0MlxCeEULg0nALCSiCIIKBIQH13HPPIT4+3u08M2fO9KlBvnDq1CmMGTMGP/30k2o7s7Oz0adPH8RwsajMzEz83//9H86fP4/k5GRkZ2fjscceky2XmZmJn376SXO7M2bMwPTp003bD3dERwOIjUV5+WUHSgO3OVCffy6Jp5dfBlq39ktb9eAPB0qNQA0try4hvFASK6FWB4ogiOqBIQG1fft2mfhQYgniXVUQBIwaNQpjx45Fjx49cOTIEZd58vLy0KxZM9k08QHIeXl5SE5ORl5enstDkevVq+e2RMPUqVNloquoqAjp6ek+7I02MTEA7HYWwiso0JzPbQhv3Tr2f/p0YMoUs5voNZEmoCyWyBVQoUSoPcqFIIjqgSEB9eOPPyI1NdVfbVFlypQp+D/x4bYa7N69G0uWLEFxcTGmTp0aoJZJ2O122O32gGwrJgZAbCwqEO3WgXIbwlu/nv3v2dP09vmCUkyYSSgW0QTCV0CFklihEB5BEMFAt4AKlrs0efJkjOIqY6vRvHlz/P7778jOznYRMj169MDw4cPx6aefIi0tDadOnZJ9Lr5PS0tz/lebR/w82ERHQ3KgvAnh5ecDhw6x125G3QWKSA7hhXJIKZIcKArhEQQRDAyPwgs0KSkpugp0/utf/8LLL7/sfH/y5ElkZmbim2++Qc/LTktGRgaeeeYZVFRUIPry02ezsrLQunVrJCcnO+dZtmwZJk6c6FxXVlYWMjIyTNwr7xEdKL0CyqXzE92nK65QLZoZTPwpPoIhoLxxoEKRUHegKIRHEEQw0N1NzZkzB0lJSf5si080btwYHTp0cP5dccUVAIAWLVqgUaNGAIB77rkHMTExGD16NHbu3IlvvvkGs2bNkuUvPfroo1i0aBHeeust7NmzBy+88AI2bNiACRMmBGW/lIhJ5J5CeB4F1JVX+qV9RvGXA5WRASQmmrc+vXgjAoMRgookB6ppU/bMxgYNgt0SgiCqE7q7lZEjRwYsz8dfJCUlYcmSJTh8+DC6d++OyZMnY9q0ac4SBgDQu3dvfPnll/jwww/RuXNnfPfdd/jpp59CogYUcLnYpN2OKti8y4HavJn9D4HwnRIzBVRmJjB+vG/r8AajDhS/TDgJqFASVVdfDTz1FBAiUXaCIKoJhp+FFy40bdpUNezYqVMnrFy50u2yw4YNw7Bhw/zVNJ+w2QDExqISUd7lQIkCqksXfzTPJ8wO4fGVzYMRgQ5lAWWkqnwoiSUtKP+JIIhAQ7edMMNmgy4HSjWEd+4ccPQoex0iAsqfSeT8OoIxui0UQ3iDBwNNmgC9eulfJtQdKIIgiGAQsQ5UpBIVBSA21rsQnvhcwGbNQi6BHDBfQAWjk+e3GaXz2xVIAdWzp/HqFSSWCIIgXPGqmyooKMBHH32EqVOn4ty5cwCATZs24cSJE6Y2jnDFqAMlEyJi+K5rV7+1zyi8M+TPUXjBcKBCUUB5AzlQBEEQrhh2oLZt24YBAwYgKSkJR44cwZgxY1C7dm388MMPyMnJwWeffeaPdhKXEXOgeAG1bRuwdy8wdOjlUXrQCOGJDlSIhO+U+LMOFAko7wnVdhEEQQQTw93UY489hlGjRmH//v2IjY11Tr/xxhvx559/mto4whWbDUB0NEsiLy0FAPzwA7BzJxNSIqoCKgQdqMxMIDYWuO668ClCqRe9AqpbN6BFC+ByKbKQgxwogiAIVww7UOvXr8d//vMfl+kNGzZ0+7w4whxsNgBRUcyBunRJ5qzwr11yoC5eBPbsYa9DSEClpABPPskE0+HD0nSzO+hgOFB6R+FlZvqnLWZBYokgCMIVw7/z7XY7ioqKXKbv27dPV8VwwjeioiAJqLIyFBdLnyUkSK9dcqC2bweqqphiCbGKg2Ibq2sIj+f++1k9owceMK9NvkIOFEEQhCuGu6lbbrkFL774IioqKgCwZ+Tl5OTgqaeewu233256Awk5vAMlXLyEggL1+VxCeHz+U4j2fpEQwvNVQDVpAowdCzRubF6bfCVELxeCIIigYribeuutt1BSUoLU1FRcvHgRffv2RcuWLVGzZk288sor/mgjwSEKKABwOAQU5Fc6P6uqkuZzCeGFYP6TEnKgQhNyoAiCIFwxfItPSkpCVlYWVq1ahW3btqGkpATdunXDgAED/NE+QgEvoCoRhYLT5RBPoyiaAJUQ3tat7H+IjsADIkNA8ZCAIgiCiFy8vsVfc801uOaaa8xsC6EDmw3O7OQq2GQOlJqAslguv9m9m01o3z4wDfUCCuGFJiSWCIIgXDF8i//Xv/6lOt1isSA2NhYtW7ZEnz59YNM7BIkwhNUKWKOscFisqBJsKDirQ0CdOQMUFLA3rVoFtL1GiAQHqroIKBJVBEFUdwzf4t9++22cOXMGpaWlSL5cuOb8+fOIj49HjRo1cPr0aTRv3hx//PEH0tPTTW8wwQwoR1QUqipsKDjrcGayaeZAieULmjQB4uIC2lYjVNdK5KEOiSWCIAhXDHdTr776Kq688krs378fZ8+exdmzZ7Fv3z707NkTs2bNQk5ODtLS0jBp0iR/tJeAlAdViSiUFEqqSTMHau9e9qZNm4C10Rv84UC1aMH+9+hhzvo84U0dqFCHHCiCIAhXDP9GfvbZZ/H999+jhdgzAWjZsiXefPNN3H777Th06BBef/11KmngR2SlDCrdCyiZA9W6dcDa6A3+EFB33gkcOwY0bWrO+owQKQ5UUlKwW0AQBBF6GO6mcnNzUVlZ6TK9srLSWYm8QYMGKOYrPBKmwhfTFCrUc6BkIbwwcaD8EcKLiWEuVKDcoEgM4dWpA/zjH8DVV0vTyIEiCKK6Y7ib6t+/Px5++GFsFusKAdi8eTPGjRuH6667DgCwfft2NGvWzLxWEjJkDlSFeh0oVQcqxAUUL5rCtYOORAEFAKmp7JmFBEEQBMOwgPr4449Ru3ZtdO/eHXa7HXa7HT169EDt2rXx8ccfAwBq1KiBt956y/TGEgwtAaWaA+WoBI4cYW9atgxYG73Bn6PwAkWkCigl4SpwCYIgzMLwLT4tLQ1ZWVnYs2cP9u3bBwBo3bo1WnP5Nf379zevhYQLfBK5xxyo3JPMmrLbQ+4ZeEoioQ4UT6QJKBJNBEEQEl7f4tu0aYM2IR4SilTEYpq6cqByjrIXzZqFvCohByp8IDFFEER1x6tb/PHjx/HLL78gJycH5eXlss9mzpxpSsMIbcQk8kpEAZUeQng5R9iL5s0D1j5vIQEV2pBoIgiCkDB8i1+2bBluueUWNG/eHHv27EGHDh1w5MgRCIKAbt26+aONhAI+hMcLKNUk8qNH2IswEFCREMKLxDpQapCYIgiiumO4m5o6dSoef/xxbN++HbGxsfj+++9x7Ngx9O3bF8OGDfNHGwkFooCqQLRMNamG8MJIQJEDFdqQaCIIgpAw3E3t3r0b9913HwAgKioKFy9eRI0aNfDiiy/i//7v/0xvIOGKlgOlmkQeRgIqEhwonkgTUDwkpgiCqO4Y7qYSEhKceU/169fHwYMHnZ/l5+eb1zJCE0M5UEcOsRdhIKDIgQptSDQRBEFIGL7F9+rVC6tWrULbtm1x4403YvLkydi+fTt++OEH9OrVyx9tJBTIQnjucqAuXYKl4BybEAaFTSPBgYpkAcVDYoogiOqO4Vv8zJkzUVJSAgCYPn06SkpK8M0336BVq1Y0Ai9A6AnhORwAzp+HBQIrI12jRsDbaRSLhf0JQvh20Pw5iDQBFa7nhCAIwh8YusVXVVXh+PHj6NSpEwAWzps9e7ZfGkZoI9aB8hjCO38eVjjCInwnYrUyJy1cHSjeBYw0AcVDYoogiOqOoW7KZrNh4MCBOH/+vL/aQ+jAkwMlCJKAskAIKwHVtClQuzZQs2awW+IdvICKtDIGJJoIgiAkDP9G7tChAw4dOkQPCw4iYhK5Mwfq4EHgyBFUNe8HwOZMIA9HAXXvvUz8RYIDFcmCI5L3jSAIQg+Gu6mXX34Zjz/+OObPn4/c3FwUFRXJ/gj/I3OgqqqAX38FVq2EY+NmAAhrAWWxhK94AmSGYMTBiyYSUARBVHcMO1A33ngjAOCWW26BhbuLCoIAi8WCKv4nOOEXZAKqoAAoLAAAOA4eBtBDJqDCLQcq3KHLnyAIonpgWED98ccf/mgHYQCZgMrLc053HGYPDhYEsISogoKwc6DCnUgWUORAEQRBSBgWUH379vVHOwgDyAppQnBOrzpyDBAEOBwWoKgIEBywREcDDRoEra3VjUgWUARBEISEV9kmK1euxL333ovevXvjxIkTAID//e9/WLVqlamNI9SRFdLkcJReBA4dco7AAwBr08aRNxwshIlkAUUOFEEQhIRhAfX9998jMzMTcXFx2LRpE8rKygAAhYWFePXVV01vIOGKrA6USIwdDliBv/6SCShLcxotGUgiWUARBEEQEl6Nwps9ezb++9//IjpackCuvvpqbNq0ydTGEerIcqAAwGoDOndmAmrZMhJQQSSSBRQ5UARBEBKGBdTevXvRp08fl+lJSUkoKCgwo00+sWDBAvTs2RNxcXFITk7G0KFDZZ/n5ORgyJAhiI+PR2pqKp544glUKsaeL1++HN26dYPdbkfLli0xd+7cwO2ADlxCeGlpQOvWTEAtWgRHlSAJqBaUQB5ISFgQBEFUDwwnkaelpeHAgQNo2rSpbPqqVavQPMijvb7//nuMGTMGr776Kq677jpUVlZix44dzs+rqqowZMgQpKWlYc2aNcjNzcV9992H6OhoZ/jx8OHDGDJkCMaOHYsvvvgCy5Ytw4MPPoj69esjMzMzWLsmQ55EDqBhQ6BJE1RFxwInT0LYuYtKGASJzEzg9Gng6quD3RLzIQeKIAhCwrCAGjNmDB599FF88sknsFgsOHnyJLKzs/H444/jueee80cbdVFZWYlHH30Ub7zxBkaPHu2c3q5dO+frJUuWYNeuXVi6dCnq1auHLl264KWXXsJTTz2FF154ATExMZg9ezaaNWuGt956CwDQtm1brFq1Cm+//XbICCiXEF7DhkBUFBwtrgD2AMJz04CTnVgJA6oYH1Dq1AEmTgx2KwiCIAh/YziEN2XKFNxzzz24/vrrUVJSgj59+uDBBx/Eww8/jEceecQfbdTFpk2bcOLECVitVnTt2hX169fH4MGDZQ5UdnY2OnbsiHr16jmnZWZmoqioCDt37nTOM2DAANm6MzMzkZ2drbntsrKygFZkdxFQl8sUOK7pw4TU0mUABBJQhKmQA0UQBCFhWEBZLBY888wzOHfuHHbs2IG1a9fizJkzeOmll/zRPt0cOnQIAPDCCy/g2Wefxfz585GcnIx+/frh3LlzAIC8vDyZeALgfJ93uSCl1jxFRUW4ePGi6rZnzJiBpKQk5196erqp+6bEZgMQHw/UTARq1wHq1gUAOK5oA6xZAwGsd7PCASQl+bUtBEEQBFEdMSygPv/8c5SWliImJgbt2rXDVVddhRo1avijbQCY42WxWNz+7dmzBw6HAwDwzDPP4Pbbb0f37t0xZ84cWCwWzJs3z2/tA4CpU6eisLDQ+Xfs2DG/bk8sY4AJE4Bx45x2QFUVgCuvhDDhnwAAiyJPjSB8gRwogiAICcM5UJMmTcLYsWNxyy234N5770VmZiZsfizUOHnyZIwaNcrtPM2bN0dubi4Aec6T3W5H8+bNkZOTA4AlwK9bt0627KlTp5yfif/Fafw8iYmJiIuLU92+3W6H3W7Xv1M+4uy8YmJk0y9rSAjTngdO7oSlWf2AtYkgCIIgqhOGBVRubi4WLVqEr776CnfccQfi4+MxbNgwDB8+HL179za9gSkpKUhJSfE4X/fu3WG327F3715cc801AICKigocOXIETZo0AQBkZGTglVdewenTp5GamgoAyMrKQmJiolN4ZWRkYOHChbJ1Z2VlISMjw8zd8gmrhm8oCigHrEDHjrCo6z2C8ApyoAiCICQMh/CioqJw00034YsvvsDp06fx9ttv48iRI+jfvz9atGjhjzbqIjExEWPHjsXzzz+PJUuWYO/evRg3bhwAYNiwYQCAgQMHol27dhgxYgS2bt2KxYsX49lnn8X48eOdDtLYsWNx6NAhPPnkk9izZw/ef/99fPvtt5g0aVLQ9k2JsvMSDUBRQJWXs/9aQosgCIIgCN8w7EDxxMfHIzMzE+fPn8fRo0exe/dus9rlFW+88QaioqIwYsQIXLx4ET179sTvv/+O5ORkAIDNZsP8+fMxbtw4ZGRkICEhASNHjsSLL77oXEezZs2wYMECTJo0CbNmzUKjRo3w0UcfhUwJA0BdQFVVMQF15Agg1v0kl4AwE3KgCIIgJLwSUKWlpfjxxx+dhSbT09Nx991347vvvjO7fYaIjo7Gm2++iTfffFNzniZNmriE6JT069cPmzdvNrt5pqF0lvgUtF9/lV5TJ0cQBEEQ/sGwgLrrrrswf/58xMfH44477sBzzz0XUvlB1QGlMIriziIvriiER5gJOVAEQRAShgWUzWbDt99+qzr6bseOHejQoYNpjSPU0cqBAuRiijo5giAIgvAPhgXUF198IXtfXFyMr776Ch999BE2btyIqkh+HH2IoHSWeNHEiylBCEx7iOoBOVAEQRASXgd5/vzzT4wcORL169fHm2++ieuuuw5r1641s22EBsrOixdU/GelpYFpD1H9IAFFEER1x5ADlZeXh7lz5+Ljjz9GUVER7rjjDpSVleGnn36SFbAk/Iuy87JYpJF4ly5J0ysqAtsuIrIh0UQQBCGh24G6+eab0bp1a2zbtg3vvPMOTp48iX//+9/+bBuhgZqAEl0ojcf1EYSpkJgiCKK6o9uB+u233/DPf/4T48aNQ6tWrfzZJsIDyhwoElBEICDRRBAEIaHbgVq1ahWKi4vRvXt39OzZE++++y7y8/P92TZCA3cOFOXwE4GAxBRBENUd3QKqV69e+O9//4vc3Fw8/PDD+Prrr9GgQQM4HA5kZWWhuLjYn+0kOLRyoAjCn5BoIgiCkDA8Ci8hIQEPPPAAVq1ahe3bt2Py5Ml47bXXkJqailtuucUfbSQUuAvhEUQgIDFFEER1x6dut3Xr1nj99ddx/PhxfPXVV2a1ifCAuxCeu/kIwhfoeiIIgpDw6WHCIjabDUOHDsXQoUPNWB3hAU8CymIBbrsNaNQosO0iqg8kpgiCqO6YIqCIwKIWwuNzoGJjgU6dAtsmIvIh0UQQBCFBmTNhiCcHKiYmsO0hqh8kpgiCqO6QgApDPAkouz2w7SGqBySaCIIgJEhAhSEWi7wzs1pJQBH+hx4mTBAEIUECKkxRdmYkoAiCIAgicJCAClOUAopPIicBRfgDcqAIgiAkSECFKe4cKEoiJwiCIAj/QgIqTFHWfeIFFTlQhD8gB4ogCEKCBFSY4q4zowKaBEEQBOFfqJBmmKIUUO3bA2fPAn36sNcEYTbkQBEEQUiQgApTlCG8bt3YH0EQBEEQ/odCeGEKuQFEoKFrjiAIQoIEVJhCnRkRTOiaIwiiukMCKkxRhvAIwt/QdUYQBCFBAipMIQeKCCZ0zREEUd0hARWmkIAiAg1dZwRBEBIkoMIUCuERwYSuOYIgqjskoMIUcqCIQEPXGUEQhAQJqDCFBBQRTOiaIwiiukMCKkwhAUUEGrrOCIIgJEhAhSmUA0UEE7rmCIKo7pCAClPIgSICDV1nBEEQEiSgwhQSUEQwoWuOIIjqDgmoMIVCeESgoeuMIAhCggRUmEIOFBFM6JojCKK6QwIqTCEBRQQaus4IgiAkIkpA7du3D7feeivq1q2LxMREXHPNNfjjjz9k8+Tk5GDIkCGIj49HamoqnnjiCVRWVsrmWb58Obp16wa73Y6WLVti7ty5AdwLfVAIjwgmdM0RBFHdiSgBddNNN6GyshK///47Nm7ciM6dO+Omm25CXl4eAKCqqgpDhgxBeXk51qxZg08//RRz587FtGnTnOs4fPgwhgwZgv79+2PLli2YOHEiHnzwQSxevDhYu6UKOVBEoKHrjCAIQiJiBFR+fj7279+PKVOmoFOnTmjVqhVee+01lJaWYseOHQCAJUuWYNeuXfj888/RpUsXDB48GC+99BLee+89lJeXAwBmz56NZs2a4a233kLbtm0xYcIE/P3vf8fbb78dzN1zge/MrBFzFolwgcQUQRDVnYjpeuvUqYPWrVvjs88+w4ULF1BZWYn//Oc/SE1NRffu3QEA2dnZ6NixI+rVq+dcLjMzE0VFRdi5c6dzngEDBsjWnZmZiezsbM1tl5WVoaioSPbnb8iBIgINXWcEQRASUcFugFlYLBYsXboUQ4cORc2aNWG1WpGamopFixYhOTkZAJCXlycTTwCc78Uwn9Y8RUVFuHjxIuLi4ly2PWPGDEyfPt0fu6UJ5UARwYSuOYIgqjsh70BNmTIFFovF7d+ePXsgCALGjx+P1NRUrFy5EuvWrcPQoUNx8803Izc3169tnDp1KgoLC51/x44d8+v2AHKgiMBD1xlBEIREyDtQkydPxqhRo9zO07x5c/z++++YP38+zp8/j8TERADA+++/j6ysLHz66aeYMmUK0tLSsG7dOtmyp06dAgCkpaU5/4vT+HkSExNV3ScAsNvtsNvt3uye15CAIoIJXXMEQVR3Ql5ApaSkICUlxeN8paWlAACrIqPaarXC4XAAADIyMvDKK6/g9OnTSE1NBQBkZWUhMTER7dq1c86zcOFC2TqysrKQkZHh876YCYXwiEBD1xlBEIREyIfw9JKRkYHk5GSMHDkSW7duxb59+/DEE084yxIAwMCBA9GuXTuMGDECW7duxeLFi/Hss89i/PjxTgdp7NixOHToEJ588kns2bMH77//Pr799ltMmjQpmLvnAjlQRDCha44giOpOxAiounXrYtGiRSgpKcF1112HHj16YNWqVfj555/RuXNnAIDNZsP8+fNhs9mQkZGBe++9F/fddx9efPFF53qaNWuGBQsWICsrC507d8Zbb72Fjz76CJmZmcHaNVVIQBGBhq45giAIiZAP4RmhR48eHgteNmnSxCVEp6Rfv37YvHmzmU0zHQrhEQRBEETwiBgHqrpBbgARaOiaIwiCkCABFaZQZ0YQBEEQwYMEVJhCITwi0JBoJwiCkCABFaZQZ0YQBEEQwYMEVJhCAooINHTNEQRBSJCAClOoMyMIgiCI4EECKkyhHCgi0JBoJwiCkCABFaZQZ0YQBEEQwYMEVJhCAooINHTNEQRBSJCAClMohEcQBEEQwYMEVJhCbgARaOiaIwiCkCABFaZQZ0YQBEEQwYMEVJhCITwi0JBoJwiCkCABFaZQZ0YQBEEQwYMEVJhCAooINHTNEQRBSJCAClMohEcQBEEQwYMEVJjCiyYrnUUiAJADRRAEIUFdb5hCnRlBEARBBA8SUGEKCSgi0NA1RxAEIUECKkyhHCgimNA1RxBEdYcEVJhCbgARaOg6IwiCkCABFaaQgCKCCV1zBEFUd0hAhSkUwiMCDV1nBEEQEiSgwhRyoIhgQtccQRDVHRJQYQoJKCLQ0HVGEAQhQQIqTKEQHhFM6JojCKK6QwIqTCEHigg0dJ0RBEFIkIAKU0hAEcGErjmCIKo7JKDCFArhEYGGrjOCIAgJElBhCjlQRDCha44giOoOCagwhQQUEWjoOiMIgpAgARWmkIAiggldcwRBVHdIQIUplANFBBq6zgiCICRIQIUp5EARwYSuOYIgqjskoMIUElBEoKHrjCAIQoIEVJhCITwimNA1RxBEdYcEVJhCDhQRaOg6IwiCkCABFaaQgCKCCV1zBEFUd8JGQL3yyivo3bs34uPjUatWLdV5cnJyMGTIEMTHxyM1NRVPPPEEKisrZfMsX74c3bp1g91uR8uWLTF37lyX9bz33nto2rQpYmNj0bNnT6xbt84Pe+QbFMIjAg1dZwRBEBJhI6DKy8sxbNgwjBs3TvXzqqoqDBkyBOXl5VizZg0+/fRTzJ07F9OmTXPOc/jwYQwZMgT9+/fHli1bMHHiRDz44INYvHixc55vvvkGjz32GJ5//nls2rQJnTt3RmZmJk6fPu33fTQCOVBEMKFrjiCI6o5FEAQh2I0wwty5czFx4kQUFBTIpv/222+46aabcPLkSdSrVw8AMHv2bDz11FM4c+YMYmJi8NRTT2HBggXYsWOHc7m77roLBQUFWLRoEQCgZ8+euPLKK/Huu+8CABwOB9LT0/HII49gypQputpYVFSEpKQkFBYWIjEx0YS9duXoUWDOHPb60UeB5GS/bIYgZHz+OVBaCjz4oNwFJQiCiASM9N8RcwvMzs5Gx44dneIJADIzM1FUVISdO3c65xkwYIBsuczMTGRnZwNgLtfGjRtl81itVgwYMMA5jxplZWUoKiqS/fkbcqCIYDB8ODBmDIkngiCIiLkN5uXlycQTAOf7vLw8t/MUFRXh4sWLyM/PR1VVleo84jrUmDFjBpKSkpx/6enpZuySW/gOjDozIlBYLCTYCYIggCALqClTpsBisbj927NnTzCbqIupU6eisLDQ+Xfs2DG/b5McKIIgCIIIHlHB3PjkyZMxatQot/M0b95c17rS0tJcRsudOnXK+Zn4X5zGz5OYmIi4uDjYbDbYbDbVecR1qGG322G323W10yxIQBEEQRBE8AiqgEpJSUFKSoop68rIyMArr7yC06dPIzU1FQCQlZWFxMREtGvXzjnPwoULZctlZWUhIyMDABATE4Pu3btj2bJlGDp0KACWRL5s2TJMmDDBlHaaBZUxIAiCIIjgETbZMzk5OdiyZQtycnJQVVWFLVu2YMuWLSgpKQEADBw4EO3atcOIESOwdetWLF68GM8++yzGjx/vdIfGjh2LQ4cO4cknn8SePXvw/vvv49tvv8WkSZOc23nsscfw3//+F59++il2796NcePG4cKFC7j//vuDst9akANFEARBEMEjqA6UEaZNm4ZPP/3U+b5r164AgD/++AP9+vWDzWbD/PnzMW7cOGRkZCAhIQEjR47Eiy++6FymWbNmWLBgASZNmoRZs2ahUaNG+Oijj5CZmemc584778SZM2cwbdo05OXloUuXLli0aJFLYnkoQQKKIAiCIAJL2NWBCgcCUQfq1Cnggw/Y6ylTgNhYv2yGIAiCIKoN1bIOVHWGHCiCIAiCCCwkoMIUyoEiCIIgiOBBAioCIAFFEARBEIGFBFQEQAKKIAiCIAILCagIgAQUQRAEQQQWElARAAkogiAIgggsJKDCFKpEThAEQRDBI2wKaRJy6tQBWrQA4uJIQBEEQRBEoCEBFaZYLMCIEcFuBUEQBEFUTyiERxAEQRAEYRASUARBEARBEAYhAUUQBEEQBGEQElAEQRAEQRAGIQFFEARBEARhEBJQBEEQBEEQBiEBRRAEQRAEYRASUARBEARBEAYhAUUQBEEQBGEQElAEQRAEQRAGIQFFEARBEARhEBJQBEEQBEEQBiEBRRAEQRAEYRASUARBEARBEAaJCnYDIhFBEAAARUVFQW4JQRAEQRB6EfttsR93BwkoP1BcXAwASE9PD3JLCIIgCIIwSnFxMZKSktzOYxH0yCzCEA6HAydPnkTNmjVhsVhMXXdRURHS09Nx7NgxJCYmmrruUCDS9w+I/H2M9P0DIn8fI33/gMjfx0jfP8A/+ygIAoqLi9GgQQNYre6znMiB8gNWqxWNGjXy6zYSExMj9ksBRP7+AZG/j5G+f0Dk72Ok7x8Q+fsY6fsHmL+PnpwnEUoiJwiCIAiCMAgJKIIgCIIgCIOQgAoz7HY7nn/+edjt9mA3xS9E+v4Bkb+Pkb5/QOTvY6TvHxD5+xjp+wcEfx8piZwgCIIgCMIg5EARBEEQBEEYhAQUQRAEQRCEQUhAEQRBEARBGIQEFEEQBEEQhEFIQIUR7733Hpo2bYrY2Fj07NkT69atC3aTvOKFF16AxWKR/bVp08b5+aVLlzB+/HjUqVMHNWrUwO23345Tp04FscWe+fPPP3HzzTejQYMGsFgs+Omnn2SfC4KAadOmoX79+oiLi8OAAQOwf/9+2Tznzp3D8OHDkZiYiFq1amH06NEoKSkJ4F5o42n/Ro0a5XJOBw0aJJsnlPdvxowZuPLKK1GzZk2kpqZi6NCh2Lt3r2wePddlTk4OhgwZgvj4eKSmpuKJJ55AZWVlIHdFEz372K9fP5fzOHbsWNk8obyPH3zwATp16uQsrJiRkYHffvvN+Xm4n0NP+xfu50/Ja6+9BovFgokTJzqnhdQ5FIiw4OuvvxZiYmKETz75RNi5c6cwZswYoVatWsKpU6eC3TTDPP/880L79u2F3Nxc59+ZM2ecn48dO1ZIT08Xli1bJmzYsEHo1auX0Lt37yC22DMLFy4UnnnmGeGHH34QAAg//vij7PPXXntNSEpKEn766Sdh69atwi233CI0a9ZMuHjxonOeQYMGCZ07dxbWrl0rrFy5UmjZsqVw9913B3hP1PG0fyNHjhQGDRokO6fnzp2TzRPK+5eZmSnMmTNH2LFjh7BlyxbhxhtvFBo3biyUlJQ45/F0XVZWVgodOnQQBgwYIGzevFlYuHChULduXWHq1KnB2CUX9Oxj3759hTFjxsjOY2FhofPzUN/HX375RViwYIGwb98+Ye/evcLTTz8tREdHCzt27BAEIfzPoaf9C/fzx7Nu3TqhadOmQqdOnYRHH33UOT2UziEJqDDhqquuEsaPH+98X1VVJTRo0ECYMWNGEFvlHc8//7zQuXNn1c8KCgqE6OhoYd68ec5pu3fvFgAI2dnZAWqhbygFhsPhENLS0oQ33njDOa2goECw2+3CV199JQiCIOzatUsAIKxfv945z2+//SZYLBbhxIkTAWu7HrQE1K233qq5TDjtnyAIwunTpwUAwooVKwRB0HddLly4ULBarUJeXp5zng8++EBITEwUysrKArsDOlDuoyCwDpjvrJSE2z4KgiAkJycLH330UUSeQ0GQ9k8QIuf8FRcXC61atRKysrJk+xRq55BCeGFAeXk5Nm7ciAEDBjinWa1WDBgwANnZ2UFsmffs378fDRo0QPPmzTF8+HDk5OQAADZu3IiKigrZvrZp0waNGzcO2309fPgw8vLyZPuUlJSEnj17OvcpOzsbtWrVQo8ePZzzDBgwAFarFX/99VfA2+wNy5cvR2pqKlq3bo1x48bh7Nmzzs/Cbf8KCwsBALVr1wag77rMzs5Gx44dUa9ePec8mZmZKCoqws6dOwPYen0o91Hkiy++QN26ddGhQwdMnToVpaWlzs/CaR+rqqrw9ddf48KFC8jIyIi4c6jcP5FIOH/jx4/HkCFDZOcKCL3vIT1MOAzIz89HVVWV7IIAgHr16mHPnj1BapX39OzZE3PnzkXr1q2Rm5uL6dOn49prr8WOHTuQl5eHmJgY1KpVS7ZMvXr1kJeXF5wG+4jYbrXzJ36Wl5eH1NRU2edRUVGoXbt2WOz3oEGD8Le//Q3NmjXDwYMH8fTTT2Pw4MHIzs6GzWYLq/1zOByYOHEirr76anTo0AEAdF2XeXl5qudY/CyUUNtHALjnnnvQpEkTNGjQANu2bcNTTz2FvXv34ocffgAQHvu4fft2ZGRk4NKlS6hRowZ+/PFHtGvXDlu2bImIc6i1f0BknL+vv/4amzZtwvr1610+C7XvIQkoIuAMHjzY+bpTp07o2bMnmjRpgm+//RZxcXFBbBnhLXfddZfzdceOHdGpUye0aNECy5cvx/XXXx/Elhln/Pjx2LFjB1atWhXspvgNrX186KGHnK87duyI+vXr4/rrr8fBgwfRokWLQDfTK1q3bo0tW7agsLAQ3333HUaOHIkVK1YEu1mmobV/7dq1C/vzd+zYMTz66KPIyspCbGxssJvjEQrhhQF169aFzWZzGWlw6tQppKWlBalV5lGrVi1cccUVOHDgANLS0lBeXo6CggLZPOG8r2K73Z2/tLQ0nD59WvZ5ZWUlzp07F5b73bx5c9StWxcHDhwAED77N2HCBMyfPx9//PEHGjVq5Jyu57pMS0tTPcfiZ6GC1j6q0bNnTwCQncdQ38eYmBi0bNkS3bt3x4wZM9C5c2fMmjUrYs6h1v6pEW7nb+PGjTh9+jS6deuGqKgoREVFYcWKFfjXv/6FqKgo1KtXL6TOIQmoMCAmJgbdu3fHsmXLnNMcDgeWLVsmi32HKyUlJTh48CDq16+P7t27Izo6Wrave/fuRU5OTtjua7NmzZCWlibbp6KiIvz111/OfcrIyEBBQQE2btzonOf333+Hw+Fw3gTDiePHj+Ps2bOoX78+gNDfP0EQMGHCBPz444/4/fff0axZM9nneq7LjIwMbN++XSYUs7KykJiY6AyxBBNP+6jGli1bAEB2HkN5H9VwOBwoKyuLiHOohrh/aoTb+bv++uuxfft2bNmyxfnXo0cPDB8+3Pk6pM6hqSnphN/4+uuvBbvdLsydO1fYtWuX8NBDDwm1atWSjTQIFyZPniwsX75cOHz4sLB69WphwIABQt26dYXTp08LgsCGqTZu3Fj4/fffhQ0bNggZGRlCRkZGkFvtnuLiYmHz5s3C5s2bBQDCzJkzhc2bNwtHjx4VBIGVMahVq5bw888/C9u2bRNuvfVW1TIGXbt2Ff766y9h1apVQqtWrUJmmL+7/SsuLhYef/xxITs7Wzh8+LCwdOlSoVu3bkKrVq2ES5cuOdcRyvs3btw4ISkpSVi+fLlsCHhpaalzHk/XpTh8euDAgcKWLVuERYsWCSkpKSEzRNzTPh44cEB48cUXhQ0bNgiHDx8Wfv75Z6F58+ZCnz59nOsI9X2cMmWKsGLFCuHw4cPCtm3bhClTpggWi0VYsmSJIAjhfw7d7V8knD81lCMLQ+kckoAKI/79738LjRs3FmJiYoSrrrpKWLt2bbCb5BV33nmnUL9+fSEmJkZo2LChcOeddwoHDhxwfn7x4kXhH//4h5CcnCzEx8cLt912m5CbmxvEFnvmjz/+EAC4/I0cOVIQBFbK4LnnnhPq1asn2O124frrrxf27t0rW8fZs2eFu+++W6hRo4aQmJgo3H///UJxcXEQ9sYVd/tXWloqDBw4UEhJSRGio6OFJk2aCGPGjHER96G8f2r7BkCYM2eOcx491+WRI0eEwYMHC3FxcULdunWFyZMnCxUVFQHeG3U87WNOTo7Qp08foXbt2oLdbhdatmwpPPHEE7I6QoIQ2vv4wAMPCE2aNBFiYmKElJQU4frrr3eKJ0EI/3Pobv8i4fypoRRQoXQOLYIgCOZ6WgRBEARBEJEN5UARBEEQBEEYhAQUQRAEQRCEQUhAEQRBEARBGIQEFEEQBEEQhEFIQBEEQRAEQRiEBBRBEARBEIRBSEARBEEQBEEYhAQUQRAEQRCEQUhAEQRBXObIkSOwWCzOZ4j5g1GjRmHo0KF+Wz9BEIGBBBRBEBHDqFGjYLFYXP4GDRqka/n09HTk5uaiQ4cOfm4pQRDhTlSwG0AQBGEmgwYNwpw5c2TT7Ha7rmVtNhvS0tL80SyCICIMcqAIgogo7HY70tLSZH/JyckAAIvFgg8++ACDBw9GXFwcmjdvju+++865rDKEd/78eQwfPhwpKSmIi4tDq1atZOJs+/btuO666xAXF4c6dergoYceQklJifPzqqoqPPbYY6hVqxbq1KmDJ598EsrHjzocDsyYMQPNmjVDXFwcOnfuLGsTQRChCQkogiCqFc899xxuv/12bN26FcOHD8ddd92F3bt3a867a9cu/Pbbb9i9ezc++OAD1K1bFwBw4cIFZGZmIjk5GevXr8e8efOwdOlSTJgwwbn8W2+9hblz5+KTTz7BqlWrcO7cOfz444+ybcyYMQOfffYZZs+ejZ07d2LSpEm49957sWLFCv8dBIIgfEcgCIKIEEaOHCnYbDYhISFB9vfKK68IgiAIAISxY8fKlunZs6cwbtw4QRAE4fDhwwIAYfPmzYIgCMLNN98s3H///arb+vDDD4Xk5GShpKTEOW3BggWC1WoV8vLyBEEQhPr16wuvv/668/OKigqhUaNGwq233ioIgiBcunRJiI+PF9asWSNb9+jRo4W7777b+wNBEITfoRwogiAiiv79++ODDz6QTatdu7bzdUZGhuyzjIwMzVF348aNw+23345NmzZh4MCBGDp0KHr37g0A2L17Nzp37oyEhATn/FdffTUcDgf27t2L2NhY5ObmomfPns7Po6Ki0KNHD2cY78CBAygtLcUNN9wg2255eTm6du1qfOcJgggYJKAIgogoEhIS0LJlS1PWNXjwYBw9ehQLFy5EVlYWrr/+eowfPx5vvvmmKesX86UWLFiAhg0byj7Tm/hOEERwoBwogiCqFWvXrnV537ZtW835U1JSMHLkSHz++ed455138OGHHwIA2rZti61bt+LChQvOeVevXg2r1YrWrVsjKSkJ9evXx19//eX8vLKyEhs3bnS+b9euHex2O3JyctCyZUvZX3p6ulm7TBCEHyAHiiCIiKKsrAx5eXmyaVFRUc7k73nz5qFHjx645ppr8MUXX2DdunX4+OOPVdc1bdo0dO/eHe3bt0dZWRnmz5/vFFvDhw/H888/j5EjR+KFF17AmTNn8Mgjj2DEiBGoV68eAODRRx/Fa6+9hlatWqFNmzaYOXMmCgoKnOuvWbMmHn/8cUyaNAkOhwPXXHMNCgsLsXr1aiQmJmLkyJF+OEIEQZgBCSiCICKKRYsWoX79+rJprVu3xp49ewAA06dPx9dff41//OMfqF+/Pr766iu0a9dOdV0xMTGYOnUqjhw5gri4OFx77bX4+uuvAQDx8fFYvHgxHn30UVx55ZWIj4/H7bffjpkzZzqXnzx5MnJzczFy5EhYrVY88MADuO2221BYWOic56WXXkJKSgpmzJiBQ4cOoVatWujWrRuefvppsw8NQRAmYhEERVESgiCICMViseDHH3+kR6kQBOEzlANFEARBEARhEBJQBEEQBEEQBqEcKIIgqg2UsUAQhFmQA0UQBEEQBGEQElAEQRAEQRAGIQFFEARBEARhEBJQBEEQBEEQBiEBRRAEQRAEYRASUARBEARBEAYhAUUQBEEQBGEQElAEQRAEQRAG+X/k1OoG9nZf+wAAAABJRU5ErkJggg==",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "show_curve(avg_history, scores)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "pwkOuaN5ZOLk"
      },
      "source": [
        "In this case the training curve shows a more stable training which seems to indicate that the chosen parameters are more suitable for problems of this complexity. As can be seen, a stable reward of 200 has been reached with about 400 episodes. This seems to me to be a very positive result. \n",
        "\n",
        "The main problem with this environment is that by the time the agent learns to stay in the air, the episodes can be quite long-lasting, which means that training can be quite time-consuming. With this in mind I decided not to run the experiment again with more episodes even though the reward was still rising. Even so, this experiment allows us to see that the algorithm works as it should and that it is most likely that with more episodes the agent will be able to solve the environment."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 494
        },
        "id": "-FCs5Lin5i5U",
        "outputId": "0764c196-ebaa-4c86-af16-9a7979887d9d"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/gym/wrappers/record_video.py:78: UserWarning: \u001b[33mWARN: Overwriting existing videos at /content/video folder (try specifying a different `video_folder` for the `RecordVideo` wrapper if this is not desired)\u001b[0m\n",
            "  logger.warn(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Loading models ...\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<video alt=\"test\" autoplay \n",
              "                loop controls style=\"height: 400px;\">\n",
              "                <source src=\"data:video/mp4;base64,AAAAIGZ0eXBpc29tAAACAGlzb21pc28yYXZjMW1wNDEAAAAIZnJlZQAAYm1tZGF0AAACrgYF//+q3EXpvebZSLeWLNgg2SPu73gyNjQgLSBjb3JlIDE1NSByMjkxNyAwYTg0ZDk4IC0gSC4yNjQvTVBFRy00IEFWQyBjb2RlYyAtIENvcHlsZWZ0IDIwMDMtMjAxOCAtIGh0dHA6Ly93d3cudmlkZW9sYW4ub3JnL3gyNjQuaHRtbCAtIG9wdGlvbnM6IGNhYmFjPTEgcmVmPTMgZGVibG9jaz0xOjA6MCBhbmFseXNlPTB4MzoweDExMyBtZT1oZXggc3VibWU9NyBwc3k9MSBwc3lfcmQ9MS4wMDowLjAwIG1peGVkX3JlZj0xIG1lX3JhbmdlPTE2IGNocm9tYV9tZT0xIHRyZWxsaXM9MSA4eDhkY3Q9MSBjcW09MCBkZWFkem9uZT0yMSwxMSBmYXN0X3Bza2lwPTEgY2hyb21hX3FwX29mZnNldD0tMiB0aHJlYWRzPTMgbG9va2FoZWFkX3RocmVhZHM9MSBzbGljZWRfdGhyZWFkcz0wIG5yPTAgZGVjaW1hdGU9MSBpbnRlcmxhY2VkPTAgYmx1cmF5X2NvbXBhdD0wIGNvbnN0cmFpbmVkX2ludHJhPTAgYmZyYW1lcz0zIGJfcHlyYW1pZD0yIGJfYWRhcHQ9MSBiX2JpYXM9MCBkaXJlY3Q9MSB3ZWlnaHRiPTEgb3Blbl9nb3A9MCB3ZWlnaHRwPTIga2V5aW50PTI1MCBrZXlpbnRfbWluPTI1IHNjZW5lY3V0PTQwIGludHJhX3JlZnJlc2g9MCByY19sb29rYWhlYWQ9NDAgcmM9Y3JmIG1idHJlZT0xIGNyZj0yMy4wIHFjb21wPTAuNjAgcXBtaW49MCBxcG1heD02OSBxcHN0ZXA9NCBpcF9yYXRpbz0xLjQwIGFxPTE6MS4wMACAAAAFMWWIhAAz//727L4FNhTIUJcRLMXaSnA+KqSAgHc02r/DxT6sALbNRu7OGINqSD8x1Uimje2Xh7p1M29foX75HxK8LFLK8btH3f16Zj/Z8ZQaCcFO/pkiAd6fr2Gd+OJNnqFJr894fK+y5tpcRumc7eaEKoP33w2z3gt8TfskPml8Id7kmZQ5vo6L9UBWcV5Re4bT2FNvskmDrFXJ0k41gIGAEFKFD/QiOgjAt8V/l7XCD8TSaa/uDMzwTlxSYb8AVAD2Q04krboRELsvYfB71xwRmnbYrTm1Nylyest2pLZb7CwAAAMAAA+Md4K4FW+Lfrg2nh3+aAAAcALX5OFNESEsE9GqKKRknZLi+HaB7u/NTmZ5Isqkgr1naocF0EdEfvVzGjfo2yqy5pfO7r80dGkXrpzrdkB8k9R3/uVo2l96VcDtcoCVD8G2N7sB9AHGzzJToa6NWgA/etcWEMMP1RM/fQ8i7IGe4pLZVblMdY/YRq6UT82LMNJK0NkLmEKfhlCORR7pAX9yox2G/0rTScxmC+J2rog/ncRRABKhXTYcaaPMD6TK6DQ1Z1PTjqKtIcS8+br7x6Lx2e+xymKA6xEUhssuaz4LPvPlT8J6OKIr9v7QBBIe4AILkUWjJ3b0vrNT/avab42eixyJZTLMlLc2D2qptmpKkfGHIL8+Ys6XH4tgdqcmzRMdaQOYIgFt+ORzjW/NCDrdvNobkda2f/OXCrte+lcR3JG2ZqdXGzpnbwFMcotUNnYGwcZwzuQyjZnErhyKmEun7aH/asycKbIazzQw3WkJdBe5YpGHfyWV6h5q0WcSmcQd7CFt7m/tYHK4JCe2QRPBSk9kf/gp6resJRjh/T9Bi/e+nnyujA3wIuV2xnWDL2p/pc1R/cuRARwJRJ3obPl6HA7RtyFoTlv+xHiZq+x1BAP4NmrRtMOrnk0Bz5XF/xY8rWhK4p0F073EF3bpYeZXUDntuzU/l2bZd6fZfuvX44WQm3L/lN6S/s0B4zBYi5vE3pfmsI3jYNj0u5WP8kvE1M2hbQYT2KdFVdrTbtOZUYl45xjP9kqW3MiSGXQrvtzemt3WWdn/G4x9fVEf6UCOen5aO183hLqY7gWFM4l79W7BhiWAjOwWymiPTvjPPOncK/R+0rlcY+f3Za3vMIOy61ClojOcruvhzZhZDkgEUD8LByIhG3qg1xWGD52hxZApN1JAGfByJ5SQ9eKWmb8zWrKYCy0vhOKbQIsDTn3LlFeLNt1VUlTK0pbsyuX84WkBJVbLIBzlGdusZq8TMtz2aKzhj+7ETUeodYu+O4ZJI57gDoDdxwHPay5qzcwRrxDeZAVS5QzFYQO7CCV63GPNDAUUkhuWMZtdEQFh/qVg7lVjurZ9Vo+Mzvef2Qz4eFs+OSa7JskYT+uiT6FgVYR/AJDoKFHoDc4p60UNRMvmVJH3AxqWB09H+0oyz/zbnop39d3+UIymQw9Os8ycxe4GiRk2f4D95aPSUBMSiIMFLowHaulHYf6RYegVSBSRSvdzq7/Sa1TZ02dM9OabkdShT5HPLKynlVBStd0SDRm3KJLmOzIyi8yudQ1I3j/RV8MPs5Kwu+Qi7EQzmP0n9jebOYXhp5JrJvGQ1cQdV+eV7Hf1y2Tp80v/lB92DP2qU2l07kvCanLREbh/9z5Nuipw+Cz2rvYR8BG6VV6cbS4Dv/4G/35SobPhbtugkpiAizJy6ahVG+/Tz4t0LbO59EjUfavLF4g/Z96grqO4hFSQAAADAAADAAA2YQAAAQtBmiRsQv/+jNMzf5+OF7Xw4GAB7tC6YpItEHq59f1MNDcN4NpvC77IglAx8XGKHcuIHfFuzFoqAbODbv9vE5zovv5QYXijkCmybCAAH/J4+3+921TLOumE1g+8Bc7HAa+nufbeyUyNHEXkobixVTlhkjyoNXzAckNP7X3dyhmug0j2hASUUAuuQRpbkdvrppHi60HM3hE5Yvq7AXZ/r2DyNxGFgudy6LQZh2bqpv7MlsuPcDsBQb31yZGEPphT7NEc4ZBST5a/iO1oB3/9ejior6zyNOTGJmRLAAAU392J1vtxwAHv6fFw+AWstSu/ZQAyHon3Ca/ddezsPoEdStTTUF3d34yjYn+mAe0AAAA0QZ5CeIR/Dz5VkAwIrf7GF7UmvUr3Lm1SAYdcAAQH4SJcaWQxFsF97MW8EitRiAQfVFRnwQAAAC8BnmF0R/8Umz+0ytT/Gp5gAl3b5xqVCRDPDAFkTh4BbwAmwse+7Gu6RnQFeE3BxwAAACYBnmNqR/8T88wPYTzMqNcPeBMDJgPTZ4AByF+p+rqfSLWwaDACVwAAAKFBmmhJqEFomUwIX//+jMPoyzuaNmjq+9ACYJV2JdlpmogJmqqWJ0j26M6E2gmiGi1PebZ03YYiOGfng2xbLE6BYzaz0zWqzg72W78HiBkWPa+LVc+IW7pjKfQBgkIfPDS72aUt9EXTCiAAAG8uQ10309wvfdrUqsxku1wgcS4tONG81yo0QsNqYnuwVph44cHckEnWObGhCvlebfbxZeAJWQAAADJBnoZFESwj/wz8navmKGgAiy0iPZgGYcO8PwAAXSbBVnr/eH/u6lSyKqD4MUYU1OYBWwAAACUBnqV0R/8S2A2UW3SzctxgL5KTGvAANQXwyYNiZlAA72z8ACLhAAAAIgGep2pH/xMDqo1JRRW0MhHUzlsVRvy/J2PKAAlunoeuG0kAAAC4QZqsSahBbJlMCF///ozGAwQiv+ocmADgSuBNmuVHEX1GH1AyoNK+3owQUCoaLI/9NMbP59qZN7GuMEaN52bD157kj8OqnQuPJphn1E7dkOb3nmuQE6x8Cp6dU8tz3/JLvYRZ70EPsVHNuDGZxp6kKAAAAwBVc20mfB70GtM+6I6W/F0WTRWS7KXrcXbdevu2qZwybqdZgXexpJB4JuQg8xhOoU+WquMZcrRLH6p2NYAd5TKxoOhRgAAAAD5BnspFFSwj/w3QZe6DuGwAG/nP6MGuJiTmuZd31/RCK3O7qKlzMYEr4qXBTvzAAMkFbrvzNrTRaXTyn2AfMQAAABkBnul0R/8S+slEr9cEo3SWFAAw5RVvuAl4AAAAKQGe62pH/xLqK2/FsfZ3rn0Q7ezlftDhFOe48AZgAT6BkVmpig8AF/WVAAAAnEGa7kmoQWyZTBRML//+jMRzyoL/qYdH5PXHQAmrJsh/skmp6c8jHyP/srfWDYQguhhrHgbemYqqtLmka1UL2pUsspXUq/nbAZhF29jTQQh2J4QhRpQ9rx935ssBrEw+1CmgpgJEAAADAL7+HlckqsDemfcm/OseuXAVAL5qGPWmZFfXiKqSt7XB33bVMuFDXTmIBJ595GuEyUNBSQAAACgBnw1qR/8TIklaZ6W72iGb51KVSITdKb8R6JAAhMhYAAJJ3WQBKyO5AAAAxkGbEknhClJlMCF//ozGumzP0stSnv670sBYje1AVT8XJUBefUriTQp1JzwLOmlCV0N7/6Ia1CaRZkEzPLmHrwjszlSMclSoR0XJcfHjyjRct5PgJuVlMS6+SVp8wrNQ6pClhWeIrbIAAAMAP/S4X0NHqHmpR89ApA2dX0SNBKZGzmGSWWjnjN6BoszBVQadwtgEM07XGC88g2/33bVM5Z1B17HIw7QRD9K0cmRur9mLh6S8mVSAxtqDF0dXm4zxFCl2OTAIeQAAAEJBnzBFNEwj/w3wzIZ7RiLEt9MMF0vAVny2ryADt1vAu+/TLgY7Baq6WNoRNIAACCz6Tn67HHPw/ahGHfTO8KQAm4AAAAA0AZ9PdEf/EwQit48Mp/6rKx/1BMHo3ERtkGEbyd6SJZAA2V/v/DOx7iTAAahJzgx0v+gr4AAAACsBn1FqR/8RfIZGDiKI8RWFsxzfyM+74FtrxxAAfOod2DYGGCIJYzt4EBbRAAAAwUGbVEmoQWiZTBTwv/6Mv2Cf0ObN9DtfgEZjQb3A6sPP1GaEJuKIZ/lwcCdaoezDa0NtTCJ2+WuwIvtf8C9LF2wonZSym71EHjS644AAASJxuPoG5ch/FyBa6dIESA9TTIlmrDGpA8/Hqi7zx93i97tqmdhHB752u8rV6+/1+ycIXHTRCOOmjI9ZNLnlRkXg3gTetdFIee/xT8iZLKm0Yg5AeG3t+lpttoiLPFolw3QmeJYJV/uz99vUiBzzlJPAf4AAAAAuAZ9zakf/EV793txZUstHnCDgoc4dBxOALnzN14TBwhF0p/WzY1H9OEcpmwAhYAAAANxBm3ZJ4QpSZTBSwz/+nhhTnnnQcAHCmdS2/xJJ4ExTTfeQw/bWTawM+3H9Se7t5EqlBRxs1uek/dAsolujGdJGasjMIV1iN38Dt7cvgrNCNDXaTAOWUAAAapulwGV3vpx1H6But6r5mNDObfVHs2BX/dEgY+LOX7FM8uHj6ejgvPVdHtkRu6AEudd+0zm3VzAQMd921TRtfJw7I72vDSzsTWMazR7CnqGHz7sQmVCi6XYMJAn+s556LhRqqJqUv4FbGZfGXR4f7+zcz2D/dE3qYvplgL7EDkm/44YxAAAAMgGflWpH/xGItrMgL1vFayv9CV0dRNUnij6KST5sTNYPLudkwYoAn9fOLZxAj/Ihc2NqAAAA00GbmknhDomUwIZ//p4ZEbhQ5G1CzQALeIequcK7Ts/HFGRnw6JirpKASrdsIdapmvhEQO3HK3kb7MP5rEM27jLdJVTDgAAD+VW/FJP4qf/jeSIvW5j1bz0dRiIjB8Ha3E4kyfFmZnTylKN1Ur/ofyV73bVNPv0ItxeodmGypZDc/vScqyMp/qghUVyV6Kb9HfyEHBSUo4THVL+6iCdtBvkEoMzHhLal8BmcbVU5O7KpY8LQXVVFukDQPlkGKmeLv/7vEI+rrySk4oBm5GhvgECAtIEAAABOQZ+4RRU8I/8LfsTqRTegBWBs6e+MZvaeiA1wGr4gE9tN6J6b3MobVL23ER/NcPcspOVoKT7our+gAJbdWBqAACAwBaOAF6aooAiLgD0hAAAALAGf13RH/w/TpedEpmFKKVvi4XoysMsaaBhBvEOxLhAGXb4pUgAG4nYhiMqAAAAAMgGf2WpH/w/F78YdvdsuDwU8L0hncRquFQxOu8lU2F6hCwRmqIACVbgAAAQYIyp3wMCBAAAAuUGb3kmoQWiZTAhf//6Muq2p2H+M936Brwa2AF1AD1gdh6/YZ937Z34tFcbtUBaPffVJiyv52HSMykSi8wnDye8Qz/nHWNuYA+n6pkeD+sqJZuU2z6pJaNlzZ8We/ziPEfEYD2wGJ2ISozK9YB7uNealxq2uo6GVR6mT1KFMZJEdPzNzOBWaQMilpHTmd0Ho9jtqXEMX9Zx1j/PYteWSV991epLz8uQ+eLmXpFAEMtffbmFKruHEALSAAAAAYEGf/EURLCP/CwD/1/cnRYD8iFpWKFuyCDD153dTNMWSH7reOe0axWvbi2VLnWYNapJcLjUxvlV6UrubT6o4MHW3rRRGzBsz7uRxeJURQ9aSx75B5pwCs4cYqlQQKvrpGQAAADEBnht0R/8PRrCrpIz7GPhaRmpnBOHjRkhalwmo5hSZ6YvtFMdgf/W/AA9iCVgUhGXhAAAAQgGeHWpH/w61BhUd41YACw7kNCHDJYLafjZJkh3f2Uk2tgylBM0etSIoRVSWQiAUem2zw4H9jmw9ADnoDxwuXsA6YAAAAMdBmgFJqEFsmUwIZ//+nhk+OgQYK5ZbtcAJahdLJ3v8dU4YMPTwlFaM6z74Umeln+29hhdDmDZH7oCYHNZl8td+SPgWTseB9QTwjgWdqupOSw2jUUPY0XqVbDkRy1i1jVCYnwi3kl2AqqBIRcU2b7hCQFCCgD/ZB3oQHwZeVODhvWa7G5TgFBF3PA0oaDItNR4k/UT9sOO9U1rTY+hg+h5n5MaH4nz+gUj6ZxDqdl7fLkmvU6orywfgy0+ykFhgldOUGU7PTWHwAAAAVUGeP0UVLCP/C3cVqSHEwBW+5QbjHtW2EfY3ZZtz5G5k030zpsKyafofpV/wVFnIVjRT/HXSf54mee6DgYfAtLdz0py5qMoKP03eAYArFfbP782Zj/EAAABFAZ5Aakf/A7GBgsvEhAHR47y45KCPoPseUSGR6SkYln3rqN1d1ZxAqdbvfWaT9kcES4kWyFdMswqDmUqB10ADrnTCUq2YAAAA/EGaRUmoQWyZTAhn//6eEAgvrZrpq7n5lkSNY/vXQEFQb3oGXButscUk/7RO6bSQ+O1jMJMHceP+hrZ1ryKY0fUPfebfunX8TxnB1ySWeKe5+VTXHL0C3TRHkMdzY+cmkHy1UuaHKPAhhFet1HNK7j/PyozetnYuOfBKj6j3YcoV+nioyH0pk8KX7Q3m3bBTtdn4y7VgqQ2lo3j3nsf2jzB+j7PsxK3s7px6mXCnMAc8dkPVFTSP4NLuFViYpF6GFv7bDOuRYVdH6bJaupisy6yJK+BJfDLsFaX1l8/Vws3XuHZ+wktbxxkg7VXaoQmhkBlExSz4Nb2oqazegQAAAGZBnmNFFSwj/wI6/samU4A9Tw+/wAV032ard1v9evUsDHEOJgWXNPIHxqEqZ0++TEE5M1wV5SH4j5qN3/lSEF0MHBmC3tsbwDrsS6fb8XzfpvLAzDgFuqrw87LPw8R/8WRdH2+AErAAAABDAZ6CdEf/A4UPvuVMO3MDbqiMZF3Qt7egHN13UXgAdInt8ex8VM6ycTd/MDUmab4/h/c4r+CHEzrmiFv5EatORLABDwAAAEQBnoRqR/8BWcuzIQAUaONVATEcuhuX33TNAdagE7f1cm7uEE+e9zZBkx8ZGeFUA6lb4MaHtNB7u2WkzkocE4A1noJ4QQAAAQFBmolJqEFsmUwIX//+jLACqe5nqeXXx0K32Lveg+n+gigX28d1zG6y0gY98FTpJ98JexZQH4AfC6JeAuhilGnU3f9/TGj93B48GTSfnJvxDuyds0xy/N2lrAhpTF77zv52noWJM7wq1tbkt2SXXbklTcKF9ei6LuNzW780fi4lMZhxIyX/XYVxmCqnP8ACiQWeBdI59GFMn+FYkDgPjy7uZHhFN9Vn2ceCgSwI8Jn7ReD6NN0iIyzYPFbA8oYxFV2xAJLd1vQ7O7yqQAn93IRPhttWJrAHt8ATZ8pfoPCu/aaOLBN0GFLGk+DiP40kmzx/q5DVQrGX1k5PHlQ/99gFtQAAAHVBnqdFFSwj/wDYE06X0FJOvB9NHzfl2vyuQ6wfDmwmvg6KxELkFL1Ka7ZT9K/JnKj57RRF3SI3jJtkJUVrVuAkIuSFw2xsRzMYbS0H0v9Q51mvUgMHC6R2T8b1gaMiakExNqs9xTCl3/XJo9h/VlpREaTgIeEAAABSAZ7GdEf/AVkUlhTOo0fS/4hQFYyIJlsMBLlJ8g4MZgBF5M6L42eVACEMqw+b4xHgIKeKcFl3q0ousFy7GXV0/wkiObHymGqpDOoSg4miPgoCXgAAAGEBnshqR/8BR89ThzIA3pr8xxmTYbGY1jUHBNZ9T76xDj53BxdIASasERssV67UoUlGNeRA+D9aAyDzkZzecgjAwb9eoB5yG0CSRsTyFUIJkGSHOdGkhIe/RpPfshcVeqx8AAAA1kGazUmoQWyZTAhf//6MsADv+924jndWGePQj+QFI+okyoDSlL57L+APjjSZERQM+sL4HBShnAR/EnEau0vfs8QHtxBoEML0uhqX+7/AMMFrChBsFBmkGbXwRt6GuTynYhZuAso67OUO2BR7nYe5StI4TMa39B9z8OFAbYirgjn3TYr30PsXce7j+8rxu46SiwNqzkknmScK8cPZiD0LzYFFKLLxAlaQcx+k3R4AjTbAZjAfOTb2T1i0OtuhN5WJJho/Rc8R/egwx85q9nd7vUTT+VHey8EAAAClQZ7rRRUsI/8Azm87znQ7QjBB6POywgBJ7lPEukcAHeX9Lui6461b5jZhXX2pW1jblgd3nruQIWwdfvOlEGEzmdn40lUvea5+WHiXZpoDdo7Wrdb2UlfMeO4CoSj109QiRt6cbn7dzI8nyXUPbT8mRcbSivUXQAtZ0WcR3Bvria9d1qrC8tXP3fwjSfvgMe5TSybg7lTp30wKvHO9rfKaqUMBfwoIAAAAZgGfCnRH/wFHGXtvbcY04bXrA3a+RSWQqeODBDoLLxW+itbEMQAmV13h3bCWe9NDvteUp8UZojhJCtR2c42SZO/qGeXvqwqfdZ6yWwsx8Wc9zW3gqjdEqobnoBl9qnum+H1SkACHgAAAAHkBnwxqR/8BR89TiAPJUay0JVliyznJimrIlY7licjFObOAwx8QAl9NJV9Suk5NdEmlDkf8Zit3W2Shd+oyUEm/gwpeGfGxqvsWb4uTmLa0dgefEBVnXuZeZZJn3nlWSMlDKk/bbTfFOwtskpMX4IeBqUme67ywQJ8vAAABEUGbEUmoQWyZTAhf//6MsABb/czf+XGarrGH4AOOwTGet19tfTyahDvNa5VzQLQKrjQOoj9HstNScsphUU/Uk17mBUlvVL5Ll1ChABmUlzQ3+nzZYDL8EDxpNd8TvfAujIMmdlzKcB1v+blnmSmauqj+yMxEFJmX2BSrmtk3O4qQgFc9jedODMhfm4G9d4HMinpn+v/bTpX3DRSlt6GHnpIW8xi05mWj6CqHC4XkRue1qg6XzIvClMCJamg/Tcp38pY1h5Jp8ZpUT3yk2/iuB5uzvp/YIuB1JAOk3eNxU6cUaBg7UKBoPSaWo7CUo79bGlTey0/RcbvIyy9swd0T6F5+sPDMJNnPDXWnCDBEjwAoYQAAAKBBny9FFSwj/wDObzvOhNQ7jVNx4Q6QE2AFcLqqsF9cNCFMc+qeeAxrhNL6RH4UiKtofKoTGZCPgjpK5MmWEvgyR8CGUOkARJ73fJPnUuMtbLs2I3om58T/GpA7SVj8SO6isoIa781AgPhboGGI5njSjZZXR7FDiPxLJlyTR5sHiOQ35jn/F6y3wcW3JAkWmwY7VaVUZawcj1LDTB6UkD0hAAAAkQGfTnRH/wFHGXtxO2GwHBwiTu43O9Fb3y2RX11yTKfxUJcKf50oB/bOB5c08Kpc5zgBYe0hwNTkS7x5hmaG58g6EkVopxtQulyriGXYQUJbRcb70HGxUldDnttnYz4mVLuOF2nBfMSHX5FH5pjyDqf2EVkJdZ6gmiiG/fOY28OZgaLPhpwV3dmbCG/cz4bABjQAAACPAZ9Qakf/AUfPU4hVPlOeAPpC1QetHwBVm/HfE8wURwck1LFY7zroLdhsGIB6IdHRIBsd2F+DVXg2cB2+mGQ0KxMb5gdvvPa4Ka39L8SK6ySO3zq8NVd+1z8hqNWDK/dwlFlb9ZsI31j8PBzxXinktyAbmQLYaGsqo8BBE9PEPOCgO+EdjFnEOqrfBRIwM+AAAAEpQZtSSahBbJlMCGf//p4QACPcefLI5JiAIkNwrxXhCtP5EcovhMBaXqJp1oiAGnMSjFuEP4pfrF4TEmoe+nZq7dInK+HsMumVSgyBq3lLZAWToRnCDTtwKvIcTCZLWzKarwamuz6yBHYtYS0wlr1F2choKOddXHCuRlkEJoy5J40Pvv5jdz5bwJI4eV3CXm/v8DmpXFTsrDj6WB8MUu5zP8aSLSs16YnOBmRVVpc4C+wIBN3yLvVtoeukviXvzB51mvkxjekeUmvUXK33ltwNjZPFFwnr9pJO6jdQoUuMBws8ejoB5SyXt0qbRktksM2NSVtNlwOpd48pGtrUmvuEXHPE79pksV7O6mhbpe0qjz0E44i7owKuJb2COad16FX6KopdqlAjIwKTAAAA4UGbdknhClJlMCF//oywACQ/E5Oc+AHspFDKeAKtZXtJckHYqT6P/81soe/ygs3wbBCASHNmcTnBQx2MH2GOasaUwiQt4O7fVm9R9JCJatLZy1y1VG30tZH2vybjQZVEexUqfrvGhw6HaVdF4nKzzbouhMmPV4xIrFnEQeBLXZimqMMJViA4cTsS5eYlKBE8gxLx/eSME/Vy3lG+6KUE/gdTNJvJdjryOoHhPL/Nvynk87+Kur8Nkz091fLOFmRQWCjGyw1xAlTYu8Etx/Bn5jIbFxDFjD1HJGlSyx/tBsewuAAAAL1Bn5RFNEwj/wDOX1xGETUHEdPWF76Jx////+UgBoEEI4BUgNvnXBlkB/ziQqbJof+YQasDHEOb2GKD3+VNxHjY2G+uZD5Ok7l6icqmQzDeQJNUJic9Ica3kiT+0NQVjGmsJaL2xdDFXt0H+Rf3rDeFnSh0GEU25MR+4RmWD+Z2fIWwYz2VMOdxrgp0YFAY2CpLtusKwQ6K6KmoaZ4ciNfKn/NTPrBkFesaUBfAY2XhReV8+VsfG0sMDPgAPmAAAACYAZ+zdEf/AUcZe3EVwp16bUgrthADZ6EP/Hr7ZTZZ8OXimSFuYy63bZM+ebE2VCMy6+ljJ3XCLaFoc/QIOuJT+lpksEvoY0n4QOqdB3zfaFitTVlluvf6MgSo6Z5r0xdIWUUD5/12OodqvcRm/sKGsJPKq65UiRLIRiKpImd9GMG8WyYP8A0+qgJS/lJBXdY7WMVTY+gAOOEAAACCAZ+1akf/AUfPU4g61DJEvVACw0d1b2uD+bpHzr5gPAaacWYyG8uhn/7qnxvu8TLrlSyUoIcp8b7xb2HnWRpvJ8gR1AjDdOajCNS0Tm+tX8bV8A/yQGu9guaUxBiWPAmYYpIoBaOZ0WKjenrfB2jnwQKsQ2N5Od+jGPTxdaZHRcAWUAAAARJBm7pJqEFomUwIX//+jLAADfet259R6fmAzjPtVwAJ1dmVqQbtP+FeoX+vjEGda4zOnsn+eRQyRSx5xX+k95tgqDOkOXVImT5df6Mfp0z6LnldukGF/AJxPhVcS058g2MvZ0VDSydYX6/Ffop5On0SX+PHsG1kYDTZ94B5JNRYgux5zKJqiVg1Ymhy3W8rj6mq89MtL43gBZi5zHDBpnXS8I1AWIEob/zWq1pL7qYT1yvBg+0eyH3rK+DB/IWPclJ2RQjlhMkig+dfNkiww/GLtHw3bTLs9eEVu9NuMmlBnWAmGltnBoqlICx4iJKuCl+LL45TLxWYtcgG8KimvfsvXHezBTcYUr+1od5vAAywJWXzAAAAmkGf2EURLCP/AM5vO86E1DrU9Sna9mcrrSW0AL+qotvf5vrMb/sVGJlSqjdHg2ZqnrmnPM5d19aiTEzPv6jqjHg5fjyHIU3f8M0M0CjYIr19h9ik3P7s4EwJdAH83fcQ1AiewG3XLHxa1me59rSmh4bbt7kWrxAIEIa6n3JQu7QqnmRhuanp8sMwk0oP6Xebb7USaFZ15jpYKCEAAAB1AZ/3dEf/AUcZe3EF7qANRDK3oB3A6TAAoxmfUpwIOfTBY0OrKo4cx4DmUCGfyoFAZAgVJkGcB2I6ITS1Ua1fqpO1CGTsJCRUrA01sY4h+WD7epoVnbxZPDyt8PI4HvJDaqfoqT8ozryuqN0QV7tQpF2A1IakAAAAqgGf+WpH/wFHz1OIK7NadboA1vUla/3tf2AbKZ6R7jYgqIGY78Atbd/yYRVOYwDRNR75EsiXLPgfJjc7mOsHYJqLkJcOCjzD+Jg5F2QLLg2dlbZzt4kpN9+AG4tj1NH0EYn51d7Cs2h3ware7iKfJlSflX9XZxfCvFMUuzZXUM311XT07uXstAoQR2NN2X9oBDX5tZo28/U6NHUMxAP2w8xaXMMPpyvdz9gRAAAA70Gb/kmoQWyZTAhf//6MsAAFL9zTPVN6/6lxKpACdXZlakG7T/hecs5KVieWgPtJLTkA8juUpK1NSoS+9xGZLTbjkQ/lk6D3lYJpa7B+34UoQUCV9xJ/+NDYGDR1I9tmq06JDgGeZfOrPKyR/4vZdeoBNVdub5ZOj4XiaDHZ9g3e32mnQvnRxXZ76NtQATninvrbFWF5AYbiCrDQGXrNLFkA1UbwPsnNfowtObRBF5z/4ncdf0jEbV42xeqeC7U0H2L4NFNrj/L/qdUCSuxVGRNpKsoYMy9wAIkjDEZT1KQOsRkw+DyhjVA6BZNc8BSQAAAAo0GeHEUVLCP/AM5vO86E1DrCRMk92obMJFqRQARiJ1mYgOwyYHzcBVemkyFzpmX9mD5OWHORAFPXS2gemOtB68fDIqinNXvzHfW1T0BmTOX0h1ujMxrclZ7nQdLFmgMKLNW8bDt6yrFVXKH/jMu/yKdu5DzOBWo+nZOBsNo1UoNq6wu9USCTcjcQEHDhKtaM1J8hWNEFZh8xmMBKhyyFwloMAZkAAACiAZ47dEf/AUcZe3D/yid64B/l6HA1HVNUm4lBkFr82YDa0hQgMkuzt4QdmrDcWLDMJea0mn/bNBNmaB1WKxYheQHQ694QlAh44y9CD3inxBNaaynEA76VN3xNl1Uesd0ohOu/9+IUfEV+7DfbDSAVCNbCjRhfPlZ54y1w/LMPwb4DSqj1RtZgsC/xJoheinIl0kZ2osuFq1smmmGjEVKFJA1JAAAAtwGePWpH/wFHz1OIK7NlO8g3AgALOiTHvT/cENfA58v4WJo+ez8fjeSOrkKJCsP9imV98WWqUo5Pu6qADTGc6ByYP7NlwUmQWzFNFGbgJJzER/SwCy/zFWijsWT/5hv7zSKOK2gDq1xrx8vVVBbP8FSL6POZS2q7qB2ThwbvzaFYaDZ3To+Wy+vYasL4sASSd/12O4Gfy6h72UYJLG1c89PrmlEU+LdJtoZ58wUfKWJZO2tizDYB4QAAAP1BmiBJqEFsmUwUTC///oywAAHq5FQAK6noR9zD/e87/rl4yAYHZ+w6O/9haJ8ssw5fIAhGuhVQkhl++GW2wFlt65ACCUEZcUr9ZewpOGGHZMCWzPb1XFuEULZcBAcn/vTF79jPBYW+upLv5OtkFzHvckXC9vL7tl2GgDtujKSolM7v0h0koh5YR9yPqJ5KbJpunwyrLUpPnIKFS9eLWbxgg6zU9883OL229UYOWP6Y5lR9DbbQ/tK7EvqGfsr+huwWy6H7Z056vP5b6uV0/+lx9LOASHe+Zyhw0H4Q1fVHnYrQQNVlw3D9NvZ6bFvrN2B1QDq51sE2b9XmWoKGAAAApgGeX2pH/wFIISFVGaiwGVqMzWvU43MAOUG3SNFi7T89z0jbVF1QaMQe46YhofLrVtXRTV94n174cN0vrmOyE4nG8LJOF2e2Ah6eGPRd6q3VXIG6EYDk3QJ9bO3n7TfCkR+Dq5pUBHTiPjPtrHSbiRPfHTL3iG/1UCoihbPuul60g3qxuUrAUOIo6ovXvkY3NA9wFyQm7dOMCzRluAxZW5fHPTxpA48AAAEGQZpCSeEKUmUwUsM//p4QAAHn9jy72GF/3CD/+nYAStRvegZcG62xxST/s3OBwpYzMkPik3x6erhgVLGSlKoPhUIKnBD3SWwFJsh0tdmJxGQfrnmn3/rcWHiWC/Vy74pV4o6wc16s7l1KkHEQ02plufqoYbWrEdRkTvnfek5baKjFzokYGfFc/EFHu5+Bgi5tuL0DdbqFMs/7lNRDuxs+baGUVUp8BiHGNB75/sOfpNE2o/WFJkQwjmchCRIFXoNKWHeiYJgtSWbvrMS5cp7sKqBSp1gV0oWl8M0GnErPjY4KwG224whI+2fm6nGcnll635soEvoVNS8iJdHg3nQjttcWhLAEvAAAAKoBnmFqR/8AAP4+/LhH8LhtIwAcKl3+GdUweSezmA2mJLvnD18dxOWBNnQavwpL9h4bGMe9Ye2jvoGUXr+3UL84ZshGG8WzT9vmB3kXLi68lwZZ78bG5Db9XKSGG2QHhD+cZRSDrCg4vSgcd5q4AlIgA/DLFLGhZc5ZbTOeHjgOiLpeRcIdslyeBF/KhQ5oVlKsx/iB58NE3eVDBdbhJIi7e0I4uKEnluQ3oQAAARVBmmZJ4Q6JlMCGf/6eEAAAvtK3wbpM2AWCP3y6I8t3S5DYAWYwiphP6SNKX4aEx0xWEmFLsmUoSpdnEvt23yRrVXul7HG5lxaxo7a5obagXCqilKVUgLSXlE5Ytyc3TdxW3+RbV72dTfZai+sPOTA2FMf/2KFY+pDq6TzQGNhhWlaouRi/HwkmtTGbj55/24+filqZfe8+Ib2s4ovotWQeyDeIe+JBtOzNp7CG1Es1v9Nr5UA0lAzZ0CbQCD9mUjcsagtaq9dbsWHL12lXoXH5VYUC0iaschFnY+uzDSt3nQoT6Yq5z3dQWA5F7DCFObWgJr/UjjdwDiMqz0fwMrckC/3eQmR+T7oPsK8VpFmrZ5tLqPmAAAAAp0GehEUVPCP/AACkBQKTKV2MIZZc+RiEHAAqKctVMmdQBElqfPE3JiSi1Hz+8wx4u1qyswY8ECOBVjw+OnoiJ/lYX+NQhNbAjC9iigJgtO4hWpPVeWG6gM0wxBReal5Hm0xnQkl59gH3ietRipuZiGoKpkVhO84lI6yVgiQ7/j5MC31XhbF02Zlo8PBmhHTU0HwWwMGg+XJMkzFHPVCzIPksDDZXvEDBAAAAbQGeo3RH/wAA/cU8J0RwktRn4XQ5NPUwSIbjiJf8OdcBXI5HZd/DVq8j8dABrVpW1EGcnezAuDNj8qig+7rdIZpZ5Tr6Lv3zLFBcDoTbZUrmwUeNnLz5EdflJRzgVyY73qszCDZcvu1oUnkwM+EAAABuAZ6lakf/AAD+QAl8+15rVAPnGvFgiAQ+oO+4dl3j1ofREPlwxPSjTOjMFahuaxLhBC+QxvsSrgvK1MY08HjiHzbat/CxHsCK6vDjznAJ3a1+eYWywOgKOMkYT9aNmbY5u4uqmJmB9M/AFBWwUEEAAAEFQZqqSahBaJlMCGf//p4QAAA/fbkrF/EQAUa731fgEJkmiSHmjHcyghGe+3Mh4cUIRB/tMQ6dhGw+JI8bpU9WXPIR8PVFtS/6YxaY831kuY9Cmh59YwS4zfHGQY+PHUvh1WtDd2SyEeCZvliooAwj9I0gPsnVuhC9Fd58TnnrPIux5ZzcRh84F8e3boIBxwSqWuhYbe2mzRfyD9icRNJfqfBaXEN9GpwbykwPMDP9hOV0zn+V4grhXmHlqy4QoTHAPhCr4nq3GougBAaCIz7+e+CHEGzdldvCs4H77/qg3ZApOE2N+W6sabmSwrp+3zLV0LbIKZLUDZkh3NHmT1oxvAh6x2rBAAAAjEGeyEURLCP/AACkBQKTNWy2vIgFe2aqoAWFO+w62UqIAgcVGqENUlptWpZftmcKDOf1tGbo2jGpzIYUhvvecXawkCeD8H/25QSBG+pG5SgKIWJ8KDRq7QYfMnWAH0QH5yXich7Ih7nTV13Ngbb4RQtzHOOPofh4ONnUFbNbyVthmmW4Reb8PRTHyGrAAAAAYAGe53RH/wAA/cU8KIcY5OF2X9vcCxYJ43q1Ofkf4IuC+EQhZCo7SYUNAaR3VvYiumWMcjE5lTEANP23v32whKBFi6+NSrT+/oDnbMjIwKsRm/uc5CsACaLadDT6B+ALuAAAAHgBnulqR/8AAP5ACXzZnA6J6edNnD7PwAFAKLYoNWk7tE7+rXrMeZ8UFIMrz1z+sP+Ze9mL+16icSCjx6X8cEqwsDSFgFZwD6ZGmGcspvbMkarxvV354i96DrGydBFXCYjASoeDEQDVncGp05kACGlol9d2chKWsoEAAAEdQZruSahBbJlMCGf//p4QAAA/vsd1MRfp58VuirjbQAs/XoZZ2qJs2Sd46JDqrBGq+8eYVpfucBkOe5uMe8Q7Tdx4RKelt7dfuEj95/7jPkd2LP60ctIZKi0VJTZIS2pbkmuUdwO9pi4PwjE5GyqPArSCCR+O0V7lctQgy3MvjpX1RNoUxjDktN02fuF7SjfrQuMrUgjtMn8AHFfQXe/Ur/WxuzbVlJ2an5fPPO9aurycMUlt+8UNvaf5Rvh03FGqjHnbDEjne0ZD0Fn2urHJh1NhfEqHIRpwGHIH5kLxpRZPYPsCOT08NQcZtv538BctiX02zNuJ1FdiDnFhYS7CT4xE6QW3N8404J9FG3suf2HW7J/EIFAqprhbEVJAAAAArEGfDEUVLCP/AACkBQKTNWz/RlZKnHAFVYIJA3ioNmXAlX0tBvnGoqJ5tJSISUgvxt7Eg3IQJCLgQZzyNNK1SPmg1vC1t3JjF0tPlMTlBPWTmxgwOvsINQksEtMcMJKq2QV2lLgSB1kkATVZeIeKb/nb7VlHCwOAK5rNA7ZR1qJab4JFpmriBYEofOhTIiZyqwq8xKI2EC5s7XCrItGc2178wEuCG5T3og/tpSQAAAB/AZ8rdEf/AAD9xTwohzFzKejTwLgBtXvcGx5hpZzXqhPCOtkauFUC+eFocbobaejpr4/XL7NUckA9s+qR/ea7WTk4UidXp0nOxSPS7F/b3R593Kfj3ygu/g4zYYNFjwi2CzYun/jsTCTgplN03saDQ8v2LMy+i/Mc4hhFzoqXZwAAAIUBny1qR/8AAP5ACXzZXtWlM/hHSPCzZD7gnPlIACo+fjGNktMPCMea4kI3TbP+CoHpvnpHVSw62f4/wd7jnCv1WKgwD/wRPuX+t69eamvuKEAN1TiXN2HQ/pH5W6ydX3WbLxKm73W3xHkyQZeIa1ksBeOtaZby9WYjqjk8ufsis1woL3dBAAABCkGbMkmoQWyZTAhn//6eEAAAGJ4S2pO8klGJRQU8wAmmYFFheNefsMwW2vdD/PQAdpOPbrYhMVbBMmADaNeZB1j3eMWSABhKQthmBTJfGI/RJ1T4brpT33lryG9MjmrvHaAEVMMWNCUpSTrHNGSXMC/sPOX4uoGUKLyon8u/ETp3Pbp1+C567fiHUG7cFms9TyjuFfsQHrXEWV/9kiK0z5iEw7+b1FXh91s6CSGO0O+1NDsQ3rBTUxqEoRgbMvMY5ubMCePE4Sxrm0AOKnuil67gWMaXZ+3hxOGM/eSB3vr1mKHF2zvCwbv04hlptfz8TF9fqCtmlzhkL7ZUXdt81Nebn6J/xogKYKmBAAAAp0GfUEUVLCP/AACkBQKTNTtGrizt0QQduZ3wAiYbmU6BLUInSi0f8ZmGA8T98rPIxGxOfZHFMCbkrnIhFy8FvqwKhPd15ptjKZy+iDN1vs8lgH5IaFYjyWJ1MBDMcnDs4tpxWgPVjaBIdCKbUscFux4xe7CU119KmSwGdAFZC0WKrzuH5N8E49ueuDZEJNkMJphPJtxkNPXg/HtXdtGZaWQiDgy2ArKAAAAArQGfb3RH/wAA/cU8KIIstDvTqY52dRlh9j2ZgBxYgjEji1CHIQxB3bKWQmoRFZD6ur9HVIs5BmokXy3BLn4Z0XB5/vjK1au4ZnT6tMayFYkdgk63Co9Ft7j2kuUCuJo8ZpgRUYvjxaJHEnoDEz2lgnn5UxqyTEGGyg93qgl3FZDbRHnJuHTtjiNbE8bqOjmK7MdX9k96Y2ZoYlPU77fyqMAPT0pbSML6+1eA6bKAAAAAeAGfcWpH/wAA/kAJfNcMpcaxeVmAHJ0hjdhiOez/cxUGf+0zr7YJuvwN69OM9oZgyhe+yKDuExvKaepLB3Cgpm4Y2ZoJz3THM8ak2lBHLeNPCtQe+TcsKJkew8BTP4WdJezM1xJSf30M9xXrLyqO/rc/oAGuaCV+EQAAARJBm3ZJqEFsmUwIZ//+nhAAABY/d+sGhy1TWEZJ0bPh+b3cx0AJX3PB0xBer31d7gdhej+7rCdctawuEPU9d/YgvX/xOjt7nctt1TbW2MZmlTUfG5LkEON65Pf2iYewa24xX/TpnnCUQR9WdDRR7ecVqM4OS0h2YqveA+3z21Y+p/LbmnFqzzsQeU8fAMq3uqwsLS2JtYjl/o2rspgmbyrrQ/mHDrSxiDfIzjYRLqbU7pXC3GkJH6LpvTXSK13kv7zXgTEF9YS6zjqbDiOeXXT4BHc3VIEe3bkhByFb2MF6VRRrrPd2A5BssFuXi9nDjDUql0noFVDBOTOwgEPcGejBXaABF/D00tIaemJrEwqNicy4AAAAlkGflEUVLCP/AACkBQKTNTfo2Yidpk1qXAALjqUL+jOGYZ6UEhPuxhApWXsrQd8SbTmZuOiek883icE63VDo1UhRRh1zvOlofPgJS6cpwBWSxQI5Liud35GXQsshPt2AUvi9BNXIG9oyoNKja6WauedCrSEt1PR+Gd0AtC+WGUouZRAdWij3NAQ49zFfkDarnncRI5gTQgAAAIEBn7N0R/8AAP3FPCiBy6oTkAOYV8JnscAj+hN0ilbmE8zzJDn7B1gxoT/BA3WSQaandQRNyTKKLbyDO3en0oXymoeB28qPy6ExM2vBlWUOqR27tAS0155pLI7PGXJValS7J5pm85cp8Yko+6g2LqipyiPMlsQWzWZWDABRoN0IMUkAAACdAZ+1akf/AAD+QAl81g/GDYMpFx+wACQx36vcPUxWM+eHpNi7Xyh5aoKMxUzb9OCYeIQa4ISRf3DhF54wR9ry/8eQ9F+f5gT4Asta5JSAtN/hETxTpzR1YdbiwNEqymdqyp3b741friymRl2lopL93CzBGEUuL/9IXjcRNkPqG1BFvWTmtiY7VY1aYA9FCe8tugbBmpROtQV6+pFNWAAAAOpBm7pJqEFsmUwIZ//+nhAAAAk3+F0DY0qhamBjwKgd2rPo5DjftmXvjmNw5OZEfwEQtLdVC8oBRaaTDqqOSCifFKUhL3CDrsLF8FBga8XVbH79x4JxcNo01NAY0IwvV7vN2JIiwffJDx+8vRjAKVBM+doNuPR4LbxN55jLmTnyiveiChv7qc7g/jBSXPiFikUdtkY6jdbcMAKtyX+3JbXCqw7U+06+lEjZh3podkfMP6o80V0K3nTTT8iishNGfcZXlIgnU5IpAbLt6w/dE3RRTtObYW0JkBdq5BqSUmX3oLtzepvBQ2wA6YEAAACzQZ/YRRUsI/8AAKQFApM1KFq7dI9rG8tEmgiLABToQD87PfiWEmqmk/W1wD4blVywlPzjjTwBx+gh6nVEupUZtvAntGrsCUVAnBVghTPi7uXhk7kz8cjaX8JYe424XD6D02YQHeb4xnahTRQN7lk0Iwmqbqtps3LrJKfvQVKMcfb8yOmaLoUXV1hIR8CbXe26YZlYBg0x3Z7gQB9L6NKgi2emDcIPguaTnULwFfN6gnStVsEAAACGAZ/3dEf/AAD9xTwogB+yWiHxWanc1fIAFnJfq9w9TFYzXkn3TGA5zS1Fn91USTd69LiCQ5ga1zz1BayiGbHxTk34fOlId2a0PKLcIIG6Wn94z0wU31JIWNzT6HMoSCPWTBGZrXaWReqch4eOOjjgZKXvxRJscEVWjtJnc23JMOBnmrpFXegAAACTAZ/5akf/AAD+QAl81hBnzLrehNAABr6YLhVQbghsx2DJp064Dd9OjdguxLcLfZNaMd9xQatJ2xIEWSzYR66T64l/qN+gjkNY6SXgjoC8LC9zKx44RPMewDlVwQtfxu4aqstu6ozvsHMOq48ALBZNOAumOFKpL42az7+9G3FIvqYTZIA8VpMug5dslLBEmQkhGj/pAAAAyEGb/kmoQWyZTAhn//6eEAAAAwNyx9c+4LYPrmnMfAQz6VdXppQAbUFrqWiZU5piXvobTvaSSfEAitkKuwPOSWMSx4lLLxVG1yLXEnrovS9/EMyZrsR77CpBWiF+5obD77+OAxmYKNIJjwzU2bD+ZSf0jJNTMW0E0N5lvGRi5QMgHyStsThveiV1Y1msVMrjB8GyyZPTefgrX/iyIaB7axhYECpDLx3djpsACVL2tzfrSxtDSMZ/2V2bzT06oE0HZp0VSIUTIE3AAAAArEGeHEUVLCP/AACkBQKTNSEzANAJcbcbeKcyz7Fra8AVvGB05d+EkxAi3O+0gAoTV2unIb4jhwe5upOevfcuqa2l1suNX5juESSmvYAh9VDrBWxlvMWv7N8A4dJtDF6eGyBDF/W9R62wj1drjRt1F3WJ1/QywI2YZDAnCYvtC8k8M4sLAm+JVlPx83opcHJmFXYAsVZKoz2juBhEhrh684er59w2S0NI1W+RJh8AAACAAZ47dEf/AAD9xTwof2Hbhoy1cN+j+j5j4gBYeg1YdDh8fcLz0mXqUfi8sK9uZHckiGOIYGyZCbJE0JxBK8rcu7oqyobEq+oEjMIr7I6U6DBRrOb0Xi0ED89zTycGfyVvq5iZLLli4bEDHJN2kmR/qrKA+q1KfI7KSo7zPzI3KSEAAACGAZ49akf/AAD+QAl81bEC1b6+0h+HdpZRAAV6fntHlNdQY9oSfPrPK6FUr6vWsYqsg837KyRCJrbiL/dVVEWRIDQaITj9qLpgf+vonkkoDPh5b8RGO5/2oJAdlHEtkTf0GRQcvK+abHWTBuFFty4H4OOO6jZw+AWYVVprMJLFvMgBAxB+FNwAAADmQZoiSahBbJlMCGf//p4QAAADA3LH1vrBYASWb1bp/IdBATWk3ZDW2z0eJE2/Q02KLhmt2jsrdIBC+0aL+pz07PCImq+wgB9TTwdKihOpatovlaTf40LvaQQowfG1RfkKGufQsgnrGQL/w2J+OIfx3qw34S+hACsBGSN297jYnpEgGRAxcrDUcYNRGI6uYLEi+y8tl//pQ4NCrpBL3zT279G84yiT2HNxNtkqcf55yadWvtXgJSTTodNCU2BpA9iDbasWZzf63Q3+ozLh5V+t8jyp5fLJH6OLqWIIiVjtRcrQlwi72LAAAACLQZ5ARRUsI/8AAKQFApM1ITFP71LeRtWVZUAEDggNy5n3QJCQg+MZ2oU0No6thUIOLitmiKu9WJQdOTkXltwFwfdlV9EH/Um8HPuN6j4YNQtAV6zHv01/ZVQUQS/l7bShnQVCX9HgWvsdNEXUJO3YJD5QpvmyZ+mNs5YPMXVZ88zHQWOUMEnbfb2NOQAAAHoBnn90R/8AAP3FPCh/YW7+cy6/p7ikwiANvdbhsuI8kAoRNKsOMPQpnuRLm3uOEHc4ihBEn7rD3138d2N0nfLvP+TfisqmKd4Xwg+KH3RVbAiwTy57CHuufZjEH8OfOuJtJt+3IhoDf4FHSJBQuon2e+u5HuiablOO2AAAAG4BnmFqR/8AAP5ACXzVsLpFTW2T6LXLsABw0hUYGOUKc6GeWA0ialG4FGYoQE3n6IVLzTkUz8QFggEgzD2jH+JXvdwVwm1NcWslnEWdO8d43Sb4sfGONsa2NQeAcxIMqxyXeeyZLO958QXEe4Oc+QAAAPJBmmZJqEFsmUwIZ//+nhAAAAMDcyFgwNi6pH+osrfBqx50rCRPNKli7m3acERaiSlqlHyU4264EdVRGXkgJnyPjcLYqU1zD7L7JhB1foroXhoDpO572PIyFBHZylwK3yiMDp6QMeDJdwDP8IFdchyGajF4qXAXRId4cYnBIvR0iNLSKcRdSgo7UY7CnrzVZatXfCF8PJVSXTS/kpZ2E1Q4IDRyw5UZiHN+pK1i3Bxkgfgl2nyoNoQlH8Y5OlqaPPqOzzfMBg86bNm/6fPHSjRWVob2WgbM+b1alV6fVUyFBB5VrPnjwOdJhUTWbdMjbNrHwAAAAHFBnoRFFSwj/wAApAUCkzUhMoC0z3SmEuwIqABC2W5hmdDskLy8HV/AFoS9+igRlfX5E/mys/+O35h4uOBkWN4GfqtcyqVIcBscI3zO7IS2OXxKbxKIZpMCFiOFPF4OdQC7OlqDj09QLfsXj08+xRBYZQAAAGYBnqN0R/8AAP3FPCh/YduILwfWlgGF81ZCqiBUDiVTv/5YFwUI9cEoFzuya4pr8scaX7Y7bo7nAPuI/qfc4m/nZW5T41eDD5DBLNNSnjVkBeEq9UPTaARPI9JRo2uktUvQ0wVr4KkAAABIAZ6lakf/AAD+QAl81bERHPUARQYCPWhC7dr4JSbOBcBZBTC4HADWPU7tBy+1kBc3YA2Y9YjRx1CUeRc/59NJBlAJc8WD6FSRAAAA4EGaqkmoQWyZTAhf//6MsAAAAwN9wnT4hi5AR8C7DtwVQCp8PydHkNY99xEAESAWs7JBkRvDJTF3lxocNJfackjx8vPAHvQwvk2gtaRMdlcN9OXuatA1/FIdpn1cNzd7w4ObmiIGEwLdP664KaCCrT37NM4A2+CQUp1DOq+yLFDodznFT5L4uC4yHmps4yumA//IaMOevPyU9d86yoWTJnLqiuhNYDldZ5Q/Mj5vWoL+CpkrUwkFZSAqVfUetpXhF7RULzVVn/OYKGU0W4VyBsZh0erwzWPvX4v/ExwNWsGBAAAAcUGeyEUVLCP/AACkBQKTNSE1NfWHE6V7dS5N733Bs+ABaj2HBmSqgr3AwjIb2wwC0dYA6a0Vpzn8Kobg5QneQXk/qmqELQ8EIhxndUH3G/6kPF7uqHIeP8AmRYUiNoOLwy7woHNL8OECH0vGa9gb4KmAAAAAUwGe53RH/wAA/cU8KH9hxktxpaHwPIWRqQcgpACv9dy8pZDnw2Set9TTabdgCJZnzeGfqYollNvZySFZevsZsJRQXidfj1LTYjCrscFB7IIvHMGfAAAAXQGe6WpH/wAA/kAJfNWNzd9s+I+cs/yngAKUu/mDXYRMZ4RCtLzrGAC7yz75b87dZaD7KfZXTZSu9vkYClVHzJe2tlS0BvtC9qpdbxxt0Z6kzMcWtXSVfTNzQYxuWQAAAPRBmu1JqEFsmUwIZ//+nhAAAAMBUN35xXOl8xBxw996gATzYTlyVav/9kQbOCSTVthTBjbky2GzvAI0rRknSZnY77Y/6ku89xeFFwIKAY8lHZ2a7sf1Fyu9US9/yvNvAyidAr9Vkb3tkBAyFYSzZpiCKOMYRmKYO4QyhgtkwkR9IOUjTn0OA4lNmeVpkY1Hm1kZDAML/Iw9R9Jzk0M3iO7Xu7gshVK7sGGOblEYeyBEO/B/6IqjmA8GSpAZQUK7heRxbfDe5We7vd54y8u258TcrUrqLEQI/Eb38nRnd/+JcvDh/TOQ6jE8ygh1dt5s1TBRp0TAAAAAZkGfC0UVLCP/AACkBQKTNSD25tmu0/2sOeERDQohHM/NMTPX0ABp66CYWAfl8mqMAn99UfCUsnHalR6ufwaC5kqZS0mi1fW+E9MLqx3HnQbReWupPrMgKkq5PlGkPR9YvdAuGuD0gAAAAFoBnyxqR/8AAP5ACXzVjdPR++bx29PHlSVFv1jFuzmhIOOHHWsolqYw+YASmC9dLHXXA+BB0X4G8N0MKH25SMpYlTFFNlyS/HrML+3KwUMwpzPB2SUleHG0b6UAAAEJQZsxSahBbJlMCGf//p4QAAADAU+1i0YfUAOM5qs6IF+O7R16V/Vcmd39DDdFr0SaSwetTj9Ok9AvYdqETvxpUEfiB8U9BZgoMplXrb9M81CsptFoTlycy3/2yE/1ZglFEItlzMWwqXBfKXep0V2bG2dv/AFO15gYLe4waRzBVbZXxBQTEbSm/Gg8G5fobuDobJg+JmjOv26N0XmxdcxwoU89k4CnhVRq74QXcxJQQsBg9HVAhSNvuW/N0hscpNlcLRisMF628mxqQdFLv+wDV9quMye6Y8ijSzJ7XpPq8+zGqhFHM9t8SEMkIJfQfqDago12V5EwgBS3UerrBEu8LY9seDIhtXzEwQAAAJ9Bn09FFSwj/wAApAUCkzUenM+sqLBo4hq/VvwALZ1JkJXGNVOVf59vYMalHml4umxH0hcfDk5a/PshtvJIeh7qpFbMNyZyySJZBOrEjS15VpJr/IjbWhoXDt50spUG05HY6SLeMFbw/iInnZwR2oObOoqlbqTcAXA2xHmMRM05jIzQ1ODxmsmuzv4oOZ1ZRMnVySs2SVIRXywnTijrDaEAAABHAZ9udEf/AAD9xTwofxy+x/EQZFW6qd0ADZc1Y6mlwCv1AIjsIR5s6dmHxkrvANQz8lr+buLAu39if2OkayeQFBC28J5+/CYAAABmAZ9wakf/AAD+QAl81Y5iyjdrwHloiDt5Z/F1Men3sOrAKsPcZSWtFLAeeACkfh+cUfKvkgAAfmYEHVTjybXw2wwfV1vD3sP+NWkezN2txte/tSGKPHEBodv4C/fedCfWMI9RZqjAAAAA60GbdUmoQWyZTAhn//6eEAAAAwFR5l0eiZ4sR1e1KgAmr0/unvJRbcrDGnTdNKjPdMHoHAQxY6vfGoPSXdPTwXgp++F/KTJRcPHtlFitw/kKv4Zmpu2Z1b8RF7SxNs4Ykc4S9w2lQ9BMrHCzMz3qw6zX9llWJEdQ3mr9Vx5WsIb/icLHyBiFE4bHhWZNm2ZUikOxM/vBuxsE9Np3vsjg6Rk/StuGqCEb9W9npFRD97esXq7fM6AxHy805wd8/lvJSyHPztwhlXo1mdYTh4SXO5THGU3v3L6k98OwlClHNzdFRx1jWqb9qgATCi0AAACBQZ+TRRUsI/8AAKQFApM1HqbV8geb6NwnhSczaMDVARIAR3+RTrSzs6z2Bu+XXNtuv6+mst6E7S8xNq4omeFSe7TiQGG1Zel5WyXX6v31vQijUkuyEp/rLB/tO6sl3hcmq0Yq+kgD1sA9LEBWGukDXy4FWQ9SVDzghY7OT4gPg/9YAAAAYwGfsnRH/wAA/cU8KH8cmpPibTPcIfdiBPlHxNBzAB/1fRY+PtwTzJqVvJycNBnIuZeH9WUPpSkDIueDbMtV17ao3eQEufYb7/+EBau1gtqrCKNv3t1Ha71pAr5bEpa9tQ7ZUAAAAGoBn7RqR/8AAP5ACXzVjltCV4/CTChAr2KislBABxfzVnPb1CFGKLjSg6ulkvHqRyGFA0VkC1YOJZAnHPYfOuUxOgfnrHNcNgt7xF5YYPO32pVVD/VSkJCQJxNUaZG8xrKTSl9sajjK76cdAAAA4UGbuUmoQWyZTAhf//6MsAAAAwFL9zQ1nNkcUjACGvjOstpiTE/otjC/SZ4ukiOwJsF5uG28rOHT69hBGS73Qgqb08n2WqjGVb2wTITC/rvUCeeZ8y+th8z2jn/nufOtwxXmFaadMbHu/pvWRxK0IJxBysK1oCFMWDmjK4VaGzZx876izbl5U06wMyh1KmYpZim8N8NeGrXCRK4GmDbYX1aMZ9jkwcmn07ik/y9f1lnMSA6tlll8W/haXXjvPESFhDJgvtMVzJ76wQDcVK7s1MI26Gr5+NZ14cyaAVAZbwDbMAAAAHdBn9dFFSwj/wAApAUCkzUepqhmSYeeKZcgeI3n/NVt3yQAHZ9sx/S+Chle2s7lXMaaMe/ylnN748v7pQ6lM3KZid0PSb2BeFBjWE2EHOMku9GKzSk/mA+YCK+AwFsDHQaph38KEmt7AtKSKJbDuOT0hY5SxFqtgQAAAF8Bn/Z0R/8AAP3FPCh/HJxemODY22QAEtTdCONMZZ1mntj0x9iGZmw9G1q79oQ/8R0ByqpkYY4hMk1r5xih2hOWYkmBovcg8a5gEb0xB2rHPxbqVBXjZSplEEoR91hxwQAAAGUBn/hqR/8AAP5ACXzVjlrskpS8404CgA/DqijdIqIRbGhsYKl0gxZnghBi3nYhS5gX9F7rG++Ry9W0nnr/XBrxPpvEBnUJvSqVbXaG7TG08OfNyJhJChPr0ScO/l6s5QyE9wrw8AAAAN5Bm/xJqEFsmUwIZ//+nhAAAAMAfs7vtBNABQKdbtzN3VWf8ouIY/RgYC5sKS3Py05UKEenZkOBRUOg3tSmSs8KPWVdXte3KmXT/JG7ANNhzscA0skJov8I+/weXh3X5HfKPj6oBpK4UpORihFifk6M0/vWa/O+eGma8ecf4Jd2spF0GK1M3hVing6W7DOgHIBDU+sLP5VMtrJY5ibdz46BRJlaWsNIpLF/qk+iec/CBntwWju9rRQGjhxJ2s4LfCs7Ke6TKI2LACPB2o748tp3gcY3pnzDp+/NPfWxoWEAAAB1QZ4aRRUsI/8AAKQFApM1HqalpYW7Y6MG2U5gypAEYidZmH3bS/uETx00o7nOXf10r6N5SrstpvgZD43EQSFn1ceBHnWc8W/r9P0/OBXqZhlaDhQ14wGpOlc4gxlbQ+5+q37rqbr70BIIclkPJAXVYlJtCK7gAAAAagGeO2pH/wAA/kAJfNWOWu8UQ65gH0QCgv6JAosrPGHj32pxwKZwL1J0whmuu8mv0remNWFsvmzLhi7W50/65DeW37gly3uxHxooM+C51CnqIml/kOPyWm9Vsyk9Vny3aQPIMhoyw/miu4EAAADkQZogSahBbJlMCGf//p4QAAADAJJ9xrgCOtGl1nuf0IePcaoHIgdF2td3PPweqwmJla6DsoWk4lqiYZVQEnwStBqaRPAArGgxEiHK9sQo4hCmppDNZO7Hn2cVOGbWF7bf9WaIaOd8p0fKA/Z7YZhnxVUpd2RT0BQ+q1PErt1PrxiYZEQ1e2Pr9fBWqUcBdUxAP7vOIrlYtgn96d/cWY58k6l+2xoxgmUSnDsgKxfdrefGxg4yAaKry6xSAzQIVCEsnLPd6LGsIJxOAl1jH3FsXM0VKOXLDoH/42rYhwAPW9hvpLphAAAAcEGeXkUVLCP/AACkBQKTNR6mpjgOpWM9o5M1QMw9zHSs3J+gHMbABdElMJIC/vhZCuevv4o8X3U7AnbiJcFm/5qo+JN0ndnVl1gDC/rORwF+cIbhW+MphTG4qHDaBJKSUO4G0fZH5B55Z73PNRbW4XoAAABPAZ59dEf/AAD9xTwofxybY7yUmCUedVyCgDW/MUwwBuLGhgAA/Wv+iWjYKoqiLMZ55UpI26s0mWzj5D+i1IwBHRlQDOnWGal0sV1ECgc80wAAAFcBnn9qR/8AAP5ACXzVjlsAA96/DHBbljs/4rZ/sRoAVaXofEDTxra/6PUUPRUVqx/G5bUyBIXPJRsz5RucOFR5n26UrRT9shoDks5VwhaO6dlGQmUML0EAAADRQZpkSahBbJlMCGf//p4QAAADAH94TpDF3TetatmS8Tyq+aAe+D84NR09mAE0rB0h5MjVozgWVomg7tiJ7xisiTzqtYFzpwiGeFzdtqpfydpDCM540FaQPlheagp5HlJzj0Hw6Z1Y2hsne3Q2p6Noo5tHSSSLCz0jPvNyCbixWSKArtgFWocdF+jGsbDz9AQCktSvZQzJ4LFUut+sBw3p2RIwpFJLwl5tvAv1HfjJt2hlpHQZpm4szRaLs5/t5/5w1qR32sKtyPhqXke5+D/3VBwAAABoQZ6CRRUsI/8AAKQFApM1HqamOEuRWezsF8zQdFRAAAP0UDJ0U3JR4i4MUYP5FVYdvFtS1iN7pKSuDBdN3l6gipr2GJ18mIy+VGvrFjLlCAeips8VIFzo1MjoBUPp2pbaYL83NTLNRuEAAABAAZ6hdEf/AAD9xTwofxybYZsJdf3wl9j4Xd/EQ0iV3yzgE9Uo1MdUwARAYsCwKBEyNHeMWDm1XEWlXfQps/Y70AAAAEQBnqNqR/8AAP5ACXzVjlsABA4jnLuTl+ZmTgAuoXmnzrhPCiGMtXs8vbyB0k2MmTCk9H27SzMgfk45pT42X1HKEnEnLQAAAURBmqhJqEFsmUwIZ//+nhAAAAMAkqP0TGgDk7aSNO1AZs+S3V371MOEuhvL63vglC9No3FxZbkkPxEaI9y8TLPXsq3x2fHOBsWqtiD2G07hP2G8Tfg9IS3IQKB/Oxbj67tNVJwiO226YkoBFyRBbXX6enyyTqfSVvNm8+yFeP506on621htgEXVTNslxsbbNgLSrC191ukRBrmoBo+c4GVAyaxs38mfiIy5wBxYGBSRpRSPIVouR8gGqTwL8H5yXPST+0zkqOTyOvj4gBGNIrnkVgaZWu7XboM+i6G+RLZ9Xrj7r9ouFwfN4/pJtiFgOLnMFZKDzJ9QASyvmU9DzX3RdFZC2xb17WwEB4+NZ7vX+v8r472xaauA0V5MJIDFZGJf9WEETf5uptqFk0BkOewRhY2AvacUKeuDLqXWmFbShMZ944EAAACQQZ7GRRUsI/8AAKQFApM1HqamN3Pu/bAAX1Tb5uGZFlyQLz/+RVQ2qJm7hQbkMMi6qD//GtzuZvpEyYczaMT1M8IXp80PINtELQSFXLA/j3eizEFJN8IxyKgFMu++KQw2m+yW2MO2MhA2Dut1A7YDpmo+U7xa4eqsMlKZFXrEHVK0v0XBvUwWNtmh2TxjD5VRAAAAfAGe5XRH/wAA/cU8KH8cm5Oy8x4LzEiNjlDIrwATrB3zfdbdzefyG1s7kpcAD3sJZYNwQ4clJQWnYwK7pLlCI+SEPijkuPM7/6pLK0CZOhNPtLsjsXhUdF4MX4KgJXUaHLzMyTJ/Y1BBU3rVBragQLw6dHbJ7pofdtijZVUAAABZAZ7nakf/AAD+QAl81Y5a/J4d/1bMtJLgDzKOXXiQANqL7D2yjBC9ccdLv7tT7KG6rotg9L9yB9533B9hXaMOQ0VHYyutGc9zgbsC2onx/Cfmyq+fLual+oIAAAGSQZrsSahBbJlMCGf//p4QAAADADSvOjAC3iiEv9EPIhOHJGo0qaMcErBvvmbEK3gjpCFtsNUc8GT7yrx/lKbpQvV7QAsSWxDvyZJuIHrKnodWUP0f7Bp07iTsnodpk5MAGU8dk1mJAR59nwEzatNvdI4wLuQukDBm9gysLMZV1L2c9NIoggknUrbbf0HH0nVto1wetzYLKsjP8/Tn+M/VQbkfSf2U88SodMwfjSUObLLkmlNoj+FmwDNwU1Ennr0d5td/ri01lqlzpuJLKlOZRoYCK75d71xi1T+MCGf9DVKv3GRgwyVdtX08ISGgnclJnRH5lzM3+QTCrJu0pyfSV8MPawyC1eLW1WCpR1J8ZkGlhKYZtUXvk4TaBAalFcxdmOJAxhHnJEq0KWaEYgAvyAcLtjI1DvYk93FAiH+nTuKvInCNDpKVdPjxSPYc6bvUvOTvdHXQPdda2XS3excgQutkeC8xVz2pPQw5kOlUPA5N+kgGn8DQPaKVo5IrfQbIKXqaWdnQqIjTh0s4wIpLmaRSAAAAjUGfCkUVLCP/AACkBQKTNR6mpfCxOfQsBofAAIyLjlk858QZ0CWrzhPAyLKXLEfQTRwDWFAh1ph1czQfPASgjOKCUnTFp97W9tERwV6Zj1ah7teeTpslsyFJusl7n+FNgGJfnSR2euoXFBIgE2Xd/yIU5mETaUTMR7gliIGDntdeM0lUURKdQ4+QINoAgQAAAHMBnyl0R/8AAP3FPCh/HJt5lHPkFuMoD9kZagAAbFtTcqtP1Slzfa0Qp0Lws85/MmnbX2KGnUK2FCLFO1oWlAue6oGCgYjH/MhWUeaQixwTVMgMNo66xt4bSYzTQBxOcDbCWWwdLgHiXY8wCQuosK7RjEE0AAAAUQGfK2pH/wAA/kAJfNWOWvcWgR6w/CpITYjAdtbAANW7rthsP7dqF8+UZRUDHyjEis/nimU4f2eU4RyjI5T4pYYc1SOdQz+rWNNVLyryvJ/yoAAAAP9BmzBJqEFsmUwIZ//+nhAAAAMAONyUgAKzn6RiHBDHa0iXFu+pybbniYfjpdJmbFfMSrCactlAkAtB130cDnVeQNYRD1ulUPl6Ptjkw5sBQX8jJbTN2AMtijkvCYEkBAdOEvnyfd5upDJdRwTNDpZIJUwd3EsddTbKqMIuXkm6xMFyrrok/u0VKAHkw7INcWo3ukIf/rP+g/iZhHeVpLu0RvcTsF1FaAaJ9Qs5Gq686LTOF0c8hxAsbaI5hEVOEJhCZVzfynr0MY0RxF3KVmKiIwp83wyy+YpJr2MUqhO3YmYo18AEinRke9JqIvxx3q+cggyE2qANwoSfUMvb31EAAABcQZ9ORRUsI/8AAKQFApM1Hqal87Gpq4ANoCkFFO93VJYlY9SAwv3P2FBEkAuXdGntD4e21X3o1z8ndldB26GCOwLsQ4oQLSDIOq8bI4QQwLcj+PxpAKYNhq94D5kAAACTAZ9tdEf/AAD9xTwofxybemLw/GGmbxoWH49+0ALQ7QUDXc7Hy285ZZ1FUfLqisuOJSZcC2SDI48fK2VcgUTDSbdB8YD5thdoOFv9oGHEoQ8u3Aok4UJM6p9e1CKfrezWE+xDfPxss8c25JTWuWnN2UocvtjvH4Jmr+tayhdVHtroG+99b+CdJIHSDZAczS5WYHHBAAAAVgGfb2pH/wAA/kAJfNWOWvd2ceBgAIU7xIGFX8o3MW8Fh8Jw4JQlsUo2pnxHYxUEg/XNnl9Rq/QGfC3XACtRtOgQWtIU9mzBilsTCDmMUzjfgUG5YR/gAAABRUGbdEmoQWyZTAhn//6eEAAAAwCSfYZACtJ/dHZLmGKaJytV8d/cpG6ZS7DQPZFzJWiab2VHPY+K27KQzewiWbLOH5PKyzweoUN7Gi5012iUs4lUk3dwlAuqkxL+JS99o+zHmp+kEFAPTEEHeXSOIe6Zi2sk04T3iE6x9XX0yfs8QmgpjyJO/L90vGi6Nqm39nIXcx5WOjoQQ62jC2ghf0y3/WXOJagfzM4NkFqdF1E8tU3DQ+NDyno01BwrLYONaQ5QSltF3yvzPBNugd5sgdXwMIdNl4PM1iKSHBcvOrQUnjQOPnc8czH41hkV6BxZBeMrdYRBONWTecAtTO33clla71PvKP0bYyWubgX4pqSSszEYbFifnLJNFi65wi8f/aRwJ+UOCdBaSg/SvqSirnouvV7y3uIgTGN2n7Tn5yNxMOBbeHcAAABGQZ+SRRUsI/8AAKQFApM1HqamN7OeO3PVGWwNKBwjAH3MG3G0fHCTOZwoJ7o4VAC1GUrlzI5Gn1J6ZOkneJeJnqHQHwmvxwAAADgBn7F0R/8AAP3FPCh/HJuT+qpTxn7MEp3ACI9BO/o3CLQM6cKBu68sWKTFvV6kgsSpwOQ/SD/TugAAADABn7NqR/8AAP5ACXzVjlr3MzUc7ESACcXmhak7upZKX4m80aoI0/vOMwfAvGnsqoAAAAC8QZu4SahBbJlMCGf//p4QAAADAI6qK6bMtNANRuK9nzaNSl6J57lLYsdnWjs21aFqo3mUQq+JTlsRQhWMJ7xjCEf4H3ghaM0LsspJh5GU6KFTfvquKLZb7PhRqDjCUmvZHS9ZeVwDhLMeMtawH0Jbtc1CDvNIz2LgOjmISaq2gPS9DuA/Nvy31Zm294GF32QJNWlq1/lexPHJO3NPoScq6T9lE193l87jYPURmmMlwOkjlBnNmnOinTB944EAAABIQZ/WRRUsI/8AAKQFApM1HqamHzyL+HEdJwAIyLVpbr0WytQRszCiUUfQDqnfIJQlJr8U3kKeNL3gSAQ8XDeFGJW8hMh5cC2gAAAAMwGf9XRH/wAA/cU8KH8cm3ncpucYq4ABsfyY96f7FdrykF9n4oBvFm4kb1xnYIo3GpM/4QAAAD8Bn/dqR/8AAP5ACXzVjlr8YiATFvAAEdLlGhqzfLIJRPmBBKGJd09JxaZTGk+mizbNqXkM+yglBUiVJHLaAoMAAAB3QZv8SahBbJlMCGf//p4QAAADAJLOekU3B/vSgBDVWCJB74gHYc6x0zwHOYsaELIqOgumdenWy5ndZ7L7ev9Xn6eExcdTQH0YYPquXQVV8i+/aj7SVMRi94hcPnSKGVJgF9u6yhr1a0u+sPCn5/UxL3nh0fmkwIAAAABEQZ4aRRUsI/8AAKQFApM1HqamOAwb9xCtPMqluABQFRdxjqtoGsimA5BUgKEJ3uNIcZb3h36m2eEvfAZg8vL8NiAApoEAAAA0AZ45dEf/AAD9xTwofxybeiSTvbAA2tJrMmh1TVbUjwgBpNwObU2Wt9x6HaerpGfdpwBKGAAAAB4BnjtqR/8AAP5ACXzVjlr8HiCyTOtmmGza4OUkEvEAAAAaQZogSahBbJlMCGf//p4QAAADAAa+3nT0i4EAAAAbQZ5eRRUsI/8AAKQFApM1HqamHzyI8kolnZqYAAAAFAGefXRH/wAA/cU8KH8cm2/86q3dAAAAEwGef2pH/wAA/kAJfNWOWvMAP8EAAAAXQZpkSahBbJlMCF///oywAAADAAADA0IAAAAZQZ6CRRUsI/8AAKQFApM1HqamHzyI8kaBvwAAABMBnqF0R/8AAP3FPCh/HJtugLaAAAAAEwGeo2pH/wAA/kAJfNWOWvMAP8EAAAAXQZqoSahBbJlMCF///oywAAADAAADA0MAAAAZQZ7GRRUsI/8AAKQFApM1HqamHzyI8kaBvwAAABMBnuV0R/8AAP3FPCh/HJtugLaBAAAAEwGe52pH/wAA/kAJfNWOWvMAP8AAAAAXQZrsSahBbJlMCF///oywAAADAAADA0IAAAAZQZ8KRRUsI/8AAKQFApM1HqamHzyI8kaBvwAAABMBnyl0R/8AAP3FPCh/HJtugLaAAAAAEwGfK2pH/wAA/kAJfNWOWvMAP8AAAAAWQZswSahBbJlMCFf//jhAAAADAAAMqQAAABlBn05FFSwj/wAApAUCkzUepqYfPIjyRoG/AAAAEwGfbXRH/wAA/cU8KH8cm26AtoEAAAATAZ9vakf/AAD+QAl81Y5a8wA/wAAAABZBm3NJqEFsmUwI//yEAAADAAADAMCAAAAAGUGfkUUVLCP/AACkBQKTNR6mph88iPJGgb8AAAATAZ+yakf/AAD+QAl81Y5a8wA/wAAAC2Ntb292AAAAbG12aGQAAAAAAAAAAAAAAAAAAAPoAAAOEAABAAABAAAAAAAAAAAAAAAAAQAAAAAAAAAAAAAAAAAAAAEAAAAAAAAAAAAAAAAAAEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACAAAKjXRyYWsAAABcdGtoZAAAAAMAAAAAAAAAAAAAAAEAAAAAAAAOEAAAAAAAAAAAAAAAAAAAAAAAAQAAAAAAAAAAAAAAAAAAAAEAAAAAAAAAAAAAAAAAAEAAAAACWAAAAZAAAAAAACRlZHRzAAAAHGVsc3QAAAAAAAAAAQAADhAAAAIAAAEAAAAACgVtZGlhAAAAIG1kaGQAAAAAAAAAAAAAAAAAADIAAAC0AFXEAAAAAAAtaGRscgAAAAAAAAAAdmlkZQAAAAAAAAAAAAAAAFZpZGVvSGFuZGxlcgAAAAmwbWluZgAAABR2bWhkAAAAAQAAAAAAAAAAAAAAJGRpbmYAAAAcZHJlZgAAAAAAAAABAAAADHVybCAAAAABAAAJcHN0YmwAAACYc3RzZAAAAAAAAAABAAAAiGF2YzEAAAAAAAAAAQAAAAAAAAAAAAAAAAAAAAACWAGQAEgAAABIAAAAAAAAAAEAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAY//8AAAAyYXZjQwFkAB//4QAZZ2QAH6zZQJgz5eEAAAMAAQAAAwBkDxgxlgEABmjr48siwAAAABhzdHRzAAAAAAAAAAEAAAC0AAABAAAAABRzdHNzAAAAAAAAAAEAAAABAAAFkGN0dHMAAAAAAAAAsAAAAAEAAAIAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAAAwAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAAAwAAAAABAAABAAAAAAEAAAMAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABAAAAAACAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAAAgAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAADAAAAAAEAAAEAAAAAAQAAAwAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABAAAAAACAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAQAAAAAAgAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAUAAAAAAQAAAgAAAAABAAAAAAAAAAEAAAEAAAAAAQAABQAAAAABAAACAAAAAAEAAAAAAAAAAQAAAQAAAAABAAAFAAAAAAEAAAIAAAAAAQAAAAAAAAABAAABAAAAAAEAAAQAAAAAAgAAAQAAAAAcc3RzYwAAAAAAAAABAAAAAQAAALQAAAABAAAC5HN0c3oAAAAAAAAAAAAAALQAAAfnAAABDwAAADgAAAAzAAAAKgAAAKUAAAA2AAAAKQAAACYAAAC8AAAAQgAAAB0AAAAtAAAAoAAAACwAAADKAAAARgAAADgAAAAvAAAAxQAAADIAAADgAAAANgAAANcAAABSAAAAMAAAADYAAAC9AAAAZAAAADUAAABGAAAAywAAAFkAAABJAAABAAAAAGoAAABHAAAASAAAAQUAAAB5AAAAVgAAAGUAAADaAAAAqQAAAGoAAAB9AAABFQAAAKQAAACVAAAAkwAAAS0AAADlAAAAwQAAAJwAAACGAAABFgAAAJ4AAAB5AAAArgAAAPMAAACnAAAApgAAALsAAAEBAAAAqgAAAQoAAACuAAABGQAAAKsAAABxAAAAcgAAAQkAAACQAAAAZAAAAHwAAAEhAAAAsAAAAIMAAACJAAABDgAAAKsAAACxAAAAfAAAARYAAACaAAAAhQAAAKEAAADuAAAAtwAAAIoAAACXAAAAzAAAALAAAACEAAAAigAAAOoAAACPAAAAfgAAAHIAAAD2AAAAdQAAAGoAAABMAAAA5AAAAHUAAABXAAAAYQAAAPgAAABqAAAAXgAAAQ0AAACjAAAASwAAAGoAAADvAAAAhQAAAGcAAABuAAAA5QAAAHsAAABjAAAAaQAAAOIAAAB5AAAAbgAAAOgAAAB0AAAAUwAAAFsAAADVAAAAbAAAAEQAAABIAAABSAAAAJQAAACAAAAAXQAAAZYAAACRAAAAdwAAAFUAAAEDAAAAYAAAAJcAAABaAAABSQAAAEoAAAA8AAAANAAAAMAAAABMAAAANwAAAEMAAAB7AAAASAAAADgAAAAiAAAAHgAAAB8AAAAYAAAAFwAAABsAAAAdAAAAFwAAABcAAAAbAAAAHQAAABcAAAAXAAAAGwAAAB0AAAAXAAAAFwAAABoAAAAdAAAAFwAAABcAAAAaAAAAHQAAABcAAAAUc3RjbwAAAAAAAAABAAAAMAAAAGJ1ZHRhAAAAWm1ldGEAAAAAAAAAIWhkbHIAAAAAAAAAAG1kaXJhcHBsAAAAAAAAAAAAAAAALWlsc3QAAAAlqXRvbwAAAB1kYXRhAAAAAQAAAABMYXZmNTguMjkuMTAw\" type=\"video/mp4\" />\n",
              "             </video>"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "test_behavior(agent)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "MO1fMyuRaq49"
      },
      "source": [
        "The video shows the behaviour learned by the agent after the training. As can be seen, the agent has learnt to descend with speed until a few moments before reaching the ground where it uses his propulsion to slow down his descent and to be able to land perfectly. At this point in the training, the agent needs to centre the landing a little more to the left. This is the point that I think it could learn with more episodes, as I have already explained."
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "jv18bAnob1gu"
      },
      "source": [
        "# **Conclusions**"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "JxcQBASxb8kG"
      },
      "source": [
        "The TD3 (Twin Delayed DDPG) algorithm for reinforcement learning was implemented and evaluated in environments such as Pendulum-v1 and LunarLanderContinuous-v2. The experiments conducted demonstrated that TD3 can achieve high performance and generate robust results across a variety of tasks. The results indicate that TD3 is effective in learning policies that can successfully solve the given tasks. The agent trained using TD3 consistently achieved high rewards and demonstrated the ability to control the pendulum or land the lunar lander smoothly.\n",
        "\n",
        "However, although the results show that the algorithm can learn useful policies with the set parameters, it is still possible for the algorithm to learn useful policies with the set parameters, it is important to note that the performance of TD3 can be sensitive to the choice of hyperparameters. This implies that fine-tuning of hyperparameters may be necessary to achieve optimal performance on specific tasks or environments.\n",
        "\n",
        "Overall, the implementation of TD3 showcased its capability to learn effective policies in reinforcement learning tasks. The robustness of TD3 across different environments highlights its potential as a reliable algorithm for various real-world applications. Further research and experimentation can be conducted to explore additional environments and optimize the hyperparameters for improved performance.\n",
        "\n"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [
        "A0ELfvjv9cZj",
        "p_03BMxoTWgl"
      ],
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}