{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Eligibility traces\n",
"\n",
"The goal of this exercise is to study the effect of eligibility traces in the simple Gridworld environment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"try:\n",
" import google.colab\n",
" IN_COLAB = True\n",
"except:\n",
" IN_COLAB = False\n",
"\n",
"if IN_COLAB:\n",
" !pip install -U gymnasium pygame swig\n",
" !pip install -U moviepy==1.0.3\n",
" !pip install gymnasium[box2d]\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"gym version: 1.0.0\n"
]
}
],
"source": [
"import numpy as np\n",
"rng = np.random.default_rng()\n",
"import matplotlib.pyplot as plt\n",
"import os\n",
"# os.environ[\"IMAGEIO_FFMPEG_EXE\"] = \"/opt/homebrew/bin/ffmpeg\" # if moviepy complains about not finding ffmpeg, put its path here\n",
"from IPython.display import clear_output\n",
"\n",
"import gymnasium as gym\n",
"print(\"gym version:\", gym.__version__)\n",
"\n",
"import pygame\n",
"from moviepy.editor import ImageSequenceClip, ipython_display\n",
"\n",
"class GymRecorder(object):\n",
" \"\"\"\n",
" Simple wrapper over moviepy to generate a .gif with the frames of a gym environment.\n",
" \n",
" The environment must have the render_mode `rgb_array_list`.\n",
" \"\"\"\n",
" def __init__(self, env):\n",
" self.env = env\n",
" self._frames = []\n",
"\n",
" def record(self, frames):\n",
" \"To be called at the end of an episode.\"\n",
" for frame in frames:\n",
" self._frames.append(np.array(frame))\n",
"\n",
" def make_video(self, filename):\n",
" \"Generates the gif video.\"\n",
" directory = os.path.dirname(os.path.abspath(filename))\n",
" if not os.path.exists(directory):\n",
" os.mkdir(directory)\n",
" self.clip = ImageSequenceClip(list(self._frames), fps=self.env.metadata[\"render_fps\"])\n",
" self.clip.write_gif(filename, fps=self.env.metadata[\"render_fps\"], loop=0)\n",
" del self._frames\n",
" self._frames = []\n",
"\n",
"def running_average(x, N):\n",
" kernel = np.ones(N) / N\n",
" return np.convolve(x, kernel, mode='same')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Q-learning in Gridworld\n",
"\n",
"### Random interaction with the environment\n",
"\n",
"The goal of this exercise is to solve the **Gridworld** problem using Q-learning. The code is adapted from \n",
"\n",
"The agent is represented by the blue circle: the **state** $s$ of the agent is its position in the 5x5 grid, i.e. a number between 0 and 24.\n",
"\n",
"The agent can move either to the left, right, top or bottom. When the agent tries to move outside of the environment, it stays at its current position. There are four **actions** $a$ available, which are deterministic. \n",
"\n",
"Its goal is to reach the green circle, while avoiding the red ones. Actions leading to the green circle receive a reward $r$ of +100, actions leading to a red square receive a reward of -100. The episode ends in those states. All other actions have a reward of -1. An episode stops after 100 steps if a goal has not been reached.\n",
"\n",
"The code of the environment is provided below. It is not important to understand it completely, but feel free to have a look at it. Rendering uses the `pygame` library, a well-known framework for simple games."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"class GridWorldEnv(gym.Env):\n",
" metadata = {\"render_modes\": [\"human\", \"rgb_array\", \"rgb_array_list\"], \"render_fps\": 2}\n",
"\n",
" def __init__(self, render_mode=None, size=5, rewards=[100, -100, -1]):\n",
" self.size = size # The size of the square grid\n",
" self.window_size = 512 # The size of the PyGame window\n",
" self.rewards = rewards\n",
" self._step = 0\n",
"\n",
" # The state is the flattened (x, y) coordinate of the agent\n",
" self.observation_space = gym.spaces.Discrete(size**2)\n",
"\n",
" # Goal location\n",
" self._target_location = np.array([3, 2], dtype=int)\n",
" self._distractor1_location = np.array([3, 1], dtype=int)\n",
" self._distractor2_location = np.array([2, 2], dtype=int)\n",
"\n",
" # We have 4 actions, corresponding to \"right\", \"up\", \"left\", \"down\"\n",
" self.action_space = gym.spaces.Discrete(4)\n",
"\n",
" self._action_to_direction = {\n",
" 0: np.array([1, 0]), # right\n",
" 1: np.array([0, 1]), # down\n",
" 2: np.array([-1, 0]), # left\n",
" 3: np.array([0, -1]), # up\n",
" }\n",
"\n",
" assert render_mode is None or render_mode in self.metadata[\"render_modes\"]\n",
" self.render_mode = render_mode\n",
"\n",
" if self.render_mode == \"rgb_array_list\":\n",
" self._frames = []\n",
" self.window = None\n",
" self.clock = None\n",
" self.font = pygame.font.SysFont(None, 16)\n",
" self.Q = np.zeros((self.observation_space.n, self.action_space.n))\n",
"\n",
"\n",
" def _state2coordinates(self, state):\n",
" \"Returns coordinates of a state.\"\n",
" return (state % self.size, int(state/self.size))\n",
"\n",
" def _coordinate2state(self, coord):\n",
" \"Returns the state with the coordinates.\"\n",
" return coord[1] * self.size + coord[0]\n",
"\n",
" def reset(self, seed=None, options=None):\n",
"\n",
" self._step = 0\n",
"\n",
" # Initial location\n",
" self._agent_location = np.array([0, 0], dtype=int)\n",
"\n",
" if self.render_mode == \"human\":\n",
" self._render_frame()\n",
" \n",
" if self.render_mode == \"rgb_array_list\":\n",
" self._frames = []\n",
" self._render_frame()\n",
"\n",
" return self._coordinate2state(self._agent_location), {}\n",
"\n",
"\n",
" def step(self, action):\n",
"\n",
" # Map the action (element of {0,1,2,3}) to the direction we walk in\n",
" direction = self._action_to_direction[action]\n",
" \n",
" # We use `np.clip` to make sure we don't leave the grid\n",
" self._agent_location = np.clip(\n",
" self._agent_location + direction, 0, self.size - 1\n",
" )\n",
" \n",
" # An episode is done if the agent has reached the target or the distractors\n",
" if np.array_equal(self._agent_location, self._target_location):\n",
" terminal = True\n",
" reward = self.rewards[0]\n",
" elif np.array_equal(self._agent_location, self._distractor1_location) \\\n",
" or np.array_equal(self._agent_location, self._distractor2_location):\n",
" terminal = True\n",
" reward = self.rewards[1]\n",
" else:\n",
" terminal = False\n",
" reward = self.rewards[2]\n",
"\n",
" if self.render_mode == \"human\" or self.render_mode == \"rgb_array_list\":\n",
" self._render_frame()\n",
"\n",
" self._step += 1\n",
" if self._step == 100:\n",
" truncated = True\n",
" else:\n",
" truncated = False\n",
"\n",
" return self._coordinate2state(self._agent_location), reward, terminal, truncated, {}\n",
"\n",
" def render(self):\n",
" if self.render_mode == \"rgb_array\":\n",
" return self._render_frame()\n",
" elif self.render_mode == \"rgb_array_list\":\n",
" f = self._frames.copy()\n",
" self._frames = []\n",
" return f\n",
"\n",
" def _render_frame(self):\n",
"\n",
" if self.window is None and self.render_mode == \"human\":\n",
" pygame.init()\n",
" pygame.display.init()\n",
" self.window = pygame.display.set_mode(\n",
" (self.window_size, self.window_size)\n",
" )\n",
" if self.clock is None and self.render_mode == \"human\":\n",
" self.clock = pygame.time.Clock()\n",
"\n",
" canvas = pygame.Surface((self.window_size, self.window_size))\n",
" canvas.fill((255, 255, 255))\n",
" pix_square_size = (\n",
" self.window_size / self.size\n",
" ) # The size of a single grid square in pixels\n",
"\n",
" # First we draw the target and the distractors\n",
" pygame.draw.rect(\n",
" canvas,\n",
" (0, 255, 0),\n",
" pygame.Rect(\n",
" pix_square_size * self._target_location,\n",
" (pix_square_size, pix_square_size),\n",
" ),\n",
" )\n",
" pygame.draw.rect(\n",
" canvas,\n",
" (255, 0, 0),\n",
" pygame.Rect(\n",
" pix_square_size * self._distractor1_location,\n",
" (pix_square_size, pix_square_size),\n",
" ),\n",
" )\n",
" pygame.draw.rect(\n",
" canvas,\n",
" (255, 0, 0),\n",
" pygame.Rect(\n",
" pix_square_size * self._distractor2_location,\n",
" (pix_square_size, pix_square_size),\n",
" ),\n",
" )\n",
"\n",
" # Now we draw the agent\n",
" pygame.draw.circle(\n",
" canvas,\n",
" (0, 0, 255),\n",
" (self._agent_location + 0.5) * pix_square_size,\n",
" pix_square_size / 3,\n",
" )\n",
"\n",
" # Add some gridlines\n",
" for x in range(self.size + 1):\n",
" pygame.draw.line(\n",
" canvas,\n",
" 0,\n",
" (0, pix_square_size * x),\n",
" (self.window_size, pix_square_size * x),\n",
" width=3,\n",
" )\n",
" pygame.draw.line(\n",
" canvas,\n",
" 0,\n",
" (pix_square_size * x, 0),\n",
" (pix_square_size * x, self.window_size),\n",
" width=3,\n",
" )\n",
"\n",
" # Print Q-values\n",
" for x in range(self.size):\n",
" for y in range(self.size):\n",
" s = self._coordinate2state((x, y))\n",
" \n",
" # Up\n",
" val = f\"{self.Q[s, 3]:+.2f}\"\n",
" text = self.font.render(val, True, (0, 0, 0))\n",
" canvas.blit(text, \n",
" ((x + 0.5) * pix_square_size - 6, \n",
" (y) * pix_square_size + 6)\n",
" )\n",
" # Down\n",
" val = f\"{self.Q[s, 1]:+.2f}\"\n",
" text = self.font.render(val, True, (0, 0, 0))\n",
" canvas.blit(text, \n",
" ((x + 0.5) * pix_square_size - 6, \n",
" (y+1) * pix_square_size - 12)\n",
" )\n",
" # Left\n",
" val = f\"{self.Q[s, 2]:+.2f}\"\n",
" text = self.font.render(val, True, (0, 0, 0))\n",
" canvas.blit(text, \n",
" ((x) * pix_square_size + 6, \n",
" (y+ 0.5) * pix_square_size - 6)\n",
" )\n",
" # Right\n",
" val = f\"{self.Q[s, 0]:+.2f}\"\n",
" text = self.font.render(val, True, (0, 0, 0))\n",
" canvas.blit(text, \n",
" ((x + 1) * pix_square_size - 32, \n",
" (y+ 0.5) * pix_square_size - 6)\n",
" )\n",
"\n",
"\n",
" if self.render_mode == \"human\":\n",
" # The following line copies our drawings from `canvas` to the visible window\n",
" self.window.blit(canvas, canvas.get_rect())\n",
" pygame.event.pump()\n",
" pygame.display.update()\n",
"\n",
" # We need to ensure that human-rendering occurs at the predefined framerate.\n",
" # The following line will automatically add a delay to keep the framerate stable.\n",
" self.clock.tick(self.metadata[\"render_fps\"])\n",
"\n",
" elif self.render_mode == \"rgb_array\":\n",
" return np.transpose(\n",
" np.array(pygame.surfarray.pixels3d(canvas)), axes=(1, 0, 2)\n",
" )\n",
" elif self.render_mode == \"rgb_array_list\":\n",
" array = np.transpose(\n",
" np.array(pygame.surfarray.pixels3d(canvas)), axes=(1, 0, 2)\n",
" )\n",
" self._frames.append(array)\n",
"\n",
" def close(self):\n",
" if self.window is not None:\n",
" pygame.display.quit()\n",
" pygame.quit()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's now define a random agent that just defines the interaction loop:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"class RandomAgent:\n",
" \n",
" def __init__(self, env):\n",
" self.env = env\n",
" self.Q = np.zeros((self.env.observation_space.n, self.env.action_space.n))\n",
" \n",
" def act(self, state):\n",
" \"Selects an action randomly\"\n",
" return self.env.action_space.sample()\n",
" \n",
" def test(self, nb_episodes, recorder=None):\n",
" \"Runs the agent on the environment for nb_episodes.\"\n",
" # Returns\n",
" returns = []\n",
" steps = []\n",
"\n",
" # Fixed number of episodes\n",
" for episode in range(nb_episodes):\n",
"\n",
" # Reset\n",
" state, info = self.env.reset()\n",
" done = False\n",
" nb_steps = 0\n",
"\n",
" # Store rewards\n",
" return_episode = 0.0\n",
"\n",
" # Sample the episode\n",
" while not done:\n",
" \n",
" # Select an action \n",
" action = self.act(state)\n",
"\n",
" # Perform the action\n",
" next_state, reward, terminal, truncated, info = self.env.step(action)\n",
" \n",
" # Append reward\n",
" return_episode += reward\n",
"\n",
" # Go in the next state\n",
" state = next_state\n",
"\n",
" # Increment time\n",
" nb_steps += 1\n",
"\n",
" # Terminal state\n",
" done = terminal or truncated\n",
" \n",
" # Pass the Q table to the GUI\n",
" self.env.Q = self.Q \n",
"\n",
" # Store info\n",
" returns.append(return_episode)\n",
" steps.append(nb_steps)\n",
"\n",
" if recorder is not None:\n",
" recorder.record(self.env.render())\n",
" \n",
" return returns, steps"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can now run the interaction loop. You can also set the `render_mode` to `'human'` and remove the recorder to see it in a window. Setting the parameter to `None` would switch off rendering."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Create the environment\n",
"env = GridWorldEnv(render_mode='rgb_array_list')\n",
"\n",
"recorder = GymRecorder(env)\n",
"\n",
"# Create the agent\n",
"agent = RandomAgent(env)\n",
"\n",
"# Perform random episodes\n",
"returns, steps = agent.test(10, recorder)\n",
"\n",
"video = \"videos/gridworld-random.gif\"\n",
"recorder.make_video(video)\n",
"ipython_display(video, loop=0, autoplay=1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Q:** Adapt your Q-learning agent from last exercise to the problem. The main difference is the call to `self.env.Q = self.Q` so that the GUI displays the Q-values, the rest is similar. Train it for 100 episodes with the right hyperparameters and without rendering."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"class QLearningAgent:\n",
" \"\"\"\n",
" Q-learning agent.\n",
" \"\"\"\n",
" \n",
" def __init__(self, env, gamma, exploration, decay, alpha):\n",
" \"\"\"\n",
" :param env: gym-like environment\n",
" :param gamma: discount factor\n",
" :param exploration: exploration parameter\n",
" :param decay: exploration decay parameter\n",
" :param alpha: learning rate\n",
" \"\"\"\n",
" self.env = env\n",
" self.gamma = gamma\n",
" self.exploration = exploration\n",
" self.decay = decay\n",
" self.alpha = alpha\n",
" \n",
" # Q_table\n",
" self.Q = np.zeros([self.env.observation_space.n, self.env.action_space.n])\n",
" \n",
" def act(self, state):\n",
" \"Returns an action using epsilon-greedy action selection.\"\n",
" \n",
" action = rng.choice(np.where(self.Q[state, :] == self.Q[state, :].max())[0])\n",
" \n",
" if rng.random() < self.exploration:\n",
" action = self.env.action_space.sample() \n",
" \n",
" return action\n",
" \n",
" def update(self, state, action, reward, next_state, done):\n",
" \"Updates the agent using a single transition.\"\n",
" \n",
" # Bellman target\n",
" target = reward\n",
" \n",
" if not done:\n",
" target += self.gamma * self.Q[next_state, :].max()\n",
" \n",
" # Update the Q-value\n",
" self.Q[state, action] += self.alpha * (target - self.Q[state, action])\n",
" \n",
" # Decay exploration parameter\n",
" self.exploration = self.exploration * (1 - self.decay)\n",
" \n",
" \n",
" def train(self, nb_episodes, recorder=None):\n",
" \"Runs the agent on the environment for nb_episodes.\"\n",
"\n",
" # Returns\n",
" returns = []\n",
" steps = []\n",
"\n",
" # Fixed number of episodes\n",
" for episode in range(nb_episodes):\n",
"\n",
" # Reset\n",
" state, info = self.env.reset()\n",
" done = False\n",
" nb_steps = 0\n",
"\n",
" # Store rewards\n",
" return_episode = 0.0\n",
"\n",
" # Sample the episode\n",
" while not done:\n",
"\n",
" # Select an action \n",
" action = self.act(state)\n",
"\n",
" # Perform the action\n",
" next_state, reward, terminal, truncated, info = self.env.step(action)\n",
"\n",
" # Terminal state\n",
" done = terminal or truncated\n",
" \n",
" # Append reward\n",
" return_episode += reward\n",
"\n",
" # Learn from the transition\n",
" self.update(state, action, reward, next_state, done)\n",
"\n",
" # Go in the next state\n",
" state = next_state\n",
"\n",
" # Increment time\n",
" nb_steps += 1\n",
" \n",
" # Pass the Q table to the GUI\n",
" self.env.Q = self.Q\n",
"\n",
" # Store info\n",
" returns.append(return_episode)\n",
" steps.append(nb_steps)\n",
"\n",
" if recorder is not None:\n",
" recorder.record(self.env.render())\n",
" \n",
" return returns, steps\n",
" \n",
" def test(self, recorder=None):\n",
" \"Performs a test episode without exploration.\"\n",
" previous_epsilon = self.epsilon\n",
" self.epsilon = 0.0\n",
" \n",
" # Reset\n",
" state, info = self.env.reset()\n",
" done = False\n",
" nb_steps = 0\n",
" return_episode= 0\n",
"\n",
" # Sample the episode\n",
" while not done:\n",
" action = self.act(state)\n",
" next_state, reward, done, info = self.env.step(action)\n",
" return_episode += reward\n",
" state = next_state\n",
" nb_steps += 1\n",
" \n",
" self.epsilon = previous_epsilon\n",
"\n",
" if recorder is not None:\n",
" recorder.record(self.env.render())\n",
" \n",
" return return_episode, nb_steps"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"epsilon = 0.1\n",
"decay_epsilon = 0\n",
"alpha = 0.1\n",
"nb_episodes = 100\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(render_mode=None)\n",
"\n",
"# Create the agent\n",
"agent = QLearningAgent(env, gamma, epsilon, decay_epsilon, alpha)\n",
"\n",
"# Train the agent\n",
"returns, steps = agent.train(nb_episodes)\n",
"\n",
"plt.figure(figsize=(12, 5))\n",
"plt.subplot(121)\n",
"plt.plot(returns)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Return\")\n",
"plt.subplot(122)\n",
"plt.plot(steps)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Steps\")\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Q:** Train a Q-learning agent with rendering on. Observe in particular which Q-values are updated when the agent reaches the target. Is it efficient?"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"epsilon = 0.1\n",
"decay_epsilon = 0\n",
"alpha = 0.1\n",
"nb_episodes = 10\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(render_mode='rgb_array_list')\n",
"recorder = GymRecorder(env)\n",
"\n",
"# Create the agent\n",
"agent = QLearningAgent(env, gamma, epsilon, decay_epsilon, alpha)\n",
"\n",
"# Train the agent\n",
"returns, steps = agent.train(nb_episodes, recorder)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"video = \"videos/gridworld-trained.gif\"\n",
"recorder.make_video(video)\n",
"ipython_display(video, loop=0, autoplay=1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Q:** Modify your agent so that it uses **softmax action selection**, with a temperature $\\tau = 1.0$ and a suitable decay. What does it change?\n",
"\n",
"If you have time, write a generic class for the Q-learning agent where you can select the action selection method flexibly."
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"class SoftQLearningAgent(QLearningAgent):\n",
" \"\"\"\n",
" Q-learning agent with softmax or e-greedy AS.\n",
" \"\"\"\n",
" \n",
" def __init__(self, env, gamma, action_selection, alpha):\n",
" \"\"\"\n",
" :param env: gym-like environment\n",
" :param gamma: discount factor\n",
" :param action selection: exploration mechanism\n",
" :param alpha: learning rate\n",
" \"\"\"\n",
" self.action_selection = action_selection\n",
" \n",
" super().__init__(env, gamma, action_selection['param'], action_selection['decay'], alpha)\n",
" \n",
" def act(self, state):\n",
" \"Returns an action using epsilon-greedy or softmax action selection.\"\n",
" \n",
" if self.action_selection['type'] == \"egreedy\":\n",
" # epsilon-greedy\n",
" if rng.uniform(0, 1, 1) < self.exploration:\n",
" action = self.env.action_space.sample() \n",
" else:\n",
" action = rng.choice(np.where(self.Q[state, :] == self.Q[state, :].max())[0])\n",
" else: \n",
" # softmax\n",
" logits = np.exp((self.Q[state, :] - self.Q[state, :].max())/self.exploration)\n",
" probas = logits / np.sum(logits)\n",
" action = rng.choice(range(4), p=probas) \n",
" \n",
" return action\n",
" "
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAA/kAAAHACAYAAAD5r6hAAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAC1QUlEQVR4nOzdeXxcVd0/8M+9M5PJnjRpkzRtugClC2UpLZSyiVBBBBXhQVFQQATBokIVhJ9SBJciPiDig6I8rAoiPAICIogFWaS0tKWFltKFljZdki5p9mW2+/tj5tw59869M3cmM5k7yef9evXVZmYyuZ2i537PdzmKpmkaiIiIiIiIiKjgqfm+ACIiIiIiIiLKDgb5RERERERERMMEg3wiIiIiIiKiYYJBPhEREREREdEwwSCfiIiIiIiIaJhgkE9EREREREQ0TDDIJyIiIiIiIhomGOQTERERERERDRPefF9AoYlEIti1axcqKiqgKEq+L4eIiAiapqGrqwuNjY1QVe7fZwPXeyIicpN01noG+WnatWsXmpqa8n0ZRERECZqbmzF+/Ph8X8awwPWeiIjcyMlazyA/TRUVFQCiH25lZWWer4aIiAjo7OxEU1OTvkbR4HG9JyIiN0lnrWeQnyZRsldZWclFn4iIXIVl5dnD9Z6IiNzIyVrPxj0iIiIiIiKiYaKggvzXX38dn/3sZ9HY2AhFUfDMM88Yntc0DYsWLcLYsWNRUlKC+fPnY9OmTYbXtLW14cILL0RlZSWqq6tx2WWXobu7ewj/FkRERERERES5UVBBfk9PD4488kjcc889ls/ffvvtuPvuu3Hvvfdi2bJlKCsrwxlnnIH+/n79NRdeeCHWrVuHl19+Gc8//zxef/11XHHFFUP1VyAiIiIiIiLKGUXTNC3fF5EJRVHw9NNP45xzzgEQzeI3Njbie9/7Hr7//e8DADo6OlBfX4+HHnoIF1xwAdavX48ZM2bgnXfewZw5cwAAL774Ij7zmc9gx44daGxsTPlzOzs7UVVVhY6ODvboERGRK3Btyj5+pkRE5CbprEsFlclPZuvWrWhpacH8+fP1x6qqqjB37lwsXboUALB06VJUV1frAT4AzJ8/H6qqYtmyZZbvOzAwgM7OTsMvIiIiIiIiIjcaNkF+S0sLAKC+vt7weH19vf5cS0sL6urqDM97vV7U1NTorzFbvHgxqqqq9F88M5eIiIiIiIjcatgE+bly4403oqOjQ//V3Nyc70siIiIiIiIisjRsgvyGhgYAQGtrq+Hx1tZW/bmGhgbs2bPH8HwoFEJbW5v+GjO/36+fkcuzcomIiIiIiMjNhk2QP3nyZDQ0NGDJkiX6Y52dnVi2bBnmzZsHAJg3bx7a29uxcuVK/TWvvPIKIpEI5s6dO+TXTERERERERJRN3nxfQDq6u7uxefNm/eutW7di9erVqKmpwYQJE3DNNdfgpz/9KaZMmYLJkyfjpptuQmNjoz6Bf/r06fj0pz+Nyy+/HPfeey+CwSCuvvpqXHDBBY4m6xMRERERERG5WUFl8lesWIFZs2Zh1qxZAICFCxdi1qxZWLRoEQDg+uuvx7e//W1cccUVOOaYY9Dd3Y0XX3wRxcXF+ns8+uijmDZtGk477TR85jOfwYknnog//OEPefn7EBERkdHrr7+Oz372s2hsbISiKHjmmWcMz2uahkWLFmHs2LEoKSnB/PnzsWnTJsNr2tracOGFF6KyshLV1dW47LLL0N3dPYR/CyIiovxRNE3T8n0RhYTn5hIRkdsMp7XpH//4B/7zn/9g9uzZOPfcc/H000/rFXkA8Itf/AKLFy/Gww8/rFftvf/++/jggw/0Tf0zzzwTu3fvxu9//3sEg0FceumlOOaYY/DYY485vo7h9JkSEVHhS2ddYpCfJi76RETkNsN1bVIUxRDka5qGxsZGfO9738P3v/99AEBHRwfq6+vx0EMP4YILLsD69esxY8YMvPPOO5gzZw4A4MUXX8RnPvMZ7Nixw3F73nD9TImIqDClsy4VVE8+USHb1z2AG596H58/qhFnHzG4GRD9wTB++dIG7OsewE/PmYmKYp/j793fPYCnVu3EvzfuwaH1FfjysRNwaH2F/vxHe7vxl3easXZnB044ZDTOnz0edZXFSd5x6ITCEfzPq5vx0rpWpLM/ObWhAl+a04TjDqqFqioAop/DX1ftwD/WtqAvELb93mKfB2cc1oDzZo9DXYX15xCOaHht4x48uWIHtu7rSe8vRSPahJpS/OFrc/J9GQVj69ataGlpwfz58/XHqqqqMHfuXCxduhQXXHABli5diurqaj3AB4D58+dDVVUsW7YMX/jCFyzfe2BgAAMDA/rXnZ2dufuLFIhXPmzFI0u34RfnHYF6l6wDRESUGoN8Knh7uwYQ0TTX34D84fUtePmDVqzf3YmzDh8LRVEyep/mtl5c+aeVWLcregPa2RfEfV+bA6/HOGJj855ubG+LB5x9gQheeH83/vlBC4LhaID8n8378eB/PsbRE6oxf0Y9/r1hL5ZvbdO/562P9uPOlzfitGl1OPvIRpT7PfpzNWV+HDm+KuXfQ9M0rNvViXHVJRhVVpTR3xmIBuXf/vO7eOuj/Wl/74ctXfjb6l2YWFuKc2eNx8bWLsPnkMrq5nbc8c8NOG16Hc4+ohFlsc9B04A1ze14YsUOtHT2p31dROEIi+nS0dLSAgCor683PF5fX68/19LSgrq6OsPzXq8XNTU1+musLF68GLfcckuWr7iwPbasGf/esBevbdiLLx7TlO/LISIihxjkU0F7fPl2LPrbOpT5PXjle6cMKojMpb5AGH95pxkAsONAH7bs68HBY8rTfp83Nu3Ft//8Ltp7gxhV6kNvIIxXN+zFz15Yj5s/exgAIBLR8JtXNuOuJRthl+w+cnwVzj6iESu3HcC/1rdi1fZ2rNreDgBQFeCTU+sw7+Ba/GNtC1ZuO4B/ftCKf37QmvA+B40uw5eOacJ5s8djdLnf8Nzermim/C/vNGPrvh6Mqy7B0wuOt82GJ/PejnZc+ceV2NXRj9IiD3541nRMrClz9L3BcAT/Wt+Kv63ehW37e/Grf200fA5fPKYp6XvtONCLJ1Y0Y9X2dry0rhUvrUv8HABgVKkP5x49HicfOgaeDDdwaOQpKfKkfhENiRtvvBELFy7Uv+7s7ERT08gObIPhCAAgEPudiIgKA4N8KkgDoTB+/Ow6/Hl5NHAO9EbwxIpmfPMTB2f157y+cS+eXbMLP/7cYSj3Z/4/l7+t3omOvqD+9Wsb9qYV5Guaht+99hH++6UNiGjAEeOr8LuLZmNNczu+9egqPPifj3HQmHJ87shGLPzLaiz5cA8AYFpDBYq80Qy/AuDIpmpccMwEzGiM9vFcDmBPVz/+unIn3t6yH3MmjsJ/zRmPsVUlAIBvnHQQNrZ24fHlzVi5/YBeIq9p0bL+Lft6sPgfH+K//7kBM8dVwRsrhQ9FNLy/owMhKUu5s70Plz+yEn+54jgU++KBzcptB/C7f2/GkeOrcf6cJjRUxTcBmtt68Zd3mvGHN7YgEIpg8ugy/P6rsw3tBU58clodfnjWdPz9vd14aV0rxlUX40vS55DKBcdOwIaWLjz+znas2t5uaBWoLSvCF44ejzMOq4ffy4CNKJcaGhoAAK2trRg7dqz+eGtrK4466ij9NXv27DF8XygUQltbm/79Vvx+P/x+v+3zI1Ek9v91EY5vIiIqKAzyqeDsau/DVX9aiTU7OqAowImHjMYbm/bhj29vwzdOOggeNXtZ1P/+5wa8t6MDh4+rwsXHT8roPTRNw8NLtwEAJtWW4uP9vXht4158/cTJCa/bsq8HE2tKDaX3Xf1BfP/JNXoG+UtzmnDL5w9Dsc+DcdUluO6MqfjlSxvw42fX4d5/f4Sd7X0o8qr42Tkzcf6c1FmouopiXHXKwbjqFOsNkkPrK7DoszMSHu8eCOG5Nbvw+PLtWLOjA+/GKgFkRzZV4yvHNmHmuCpc+L/LsKa5Hd97cg1+c8EsKArwp2Xbcetz6xAMa/jX+j341b824tRpdfjEoWPwzw9a8ebmfXo1wvzp9bjzS0eiMo35A7LSIi/On9Pk6DOxMrWhQq+WIKL8mDx5MhoaGrBkyRI9qO/s7MSyZctw1VVXAQDmzZuH9vZ2rFy5ErNnzwYAvPLKK4hEIpg7d26+Lr0giXaSkMPWJiIicgcG+VRwLn9kBdbt6kR1qQ+/vmAW5k6uwXGLl2DHgT688uEefGpGfeo3cSAS0bCxtQsAsGLbgYyD/BXbDmD97k4U+1Tc/l9H4ou/X4q3t+xHfzBsyGjf8+pm/Pc/N6K+0o/zZzfhS8c0YSAUxhV/XIkte3tQ5FHx488dhq/MnWB4/2+dcjC27O3BX1ftwM72PoyrLsG9F83G4eOrMv67O1Hu9+LLx07Al2NZ7q37jGdQTx5djqkN8Yz77y+ajYvuX4a/v7cb40eVYH93AP+3cgcA4NRpdejqD+Kdjw/gX+v34F/r41m4Ew8ZjQuObcJnZo7Vh+YR0fDV3d2NzZs3619v3boVq1evRk1NDSZMmIBrrrkGP/3pTzFlyhT9CL3GxkZ9Av/06dPx6U9/GpdffjnuvfdeBINBXH311bjgggscT9anKBHkc3YEEVFhYZA/TIUjGm5/6UMcd1AtPjm1LvU3FIiWjn6s29UJVQH+tuAETKyN9lJ/aU4Tfv/6Fjyy9OO0gvzmtl7c8+pmXH7yQQnl880HetEfjPYhrvy4zerbHXnorY8BAOccNQ7HTBqFsVXF2N3Rj7e37McpsX+b/mAY97+5FQDQ2jmA/3l1M/7n1c0o8qoIhCIYW1WM3100G0c1VSe8v6Io+Pm5M6Eq0TL5m86egZohnk0wtaHCENBbmXtQLX7+hcNx3f+9h9+/tgVAtP//hjOn4fKTDoKiKNi8pwt/eacZK7cdwLyDa/GlORMwobZ0KP4KROQSK1aswCc/+Un9a9Enf/HFF+Ohhx7C9ddfj56eHlxxxRVob2/HiSeeiBdffBHFxfFWn0cffRRXX301TjvtNKiqivPOOw933333kP9dCp0o0w8xyCciKigM8oepFR+34fevbcGTK3Zg+f87LWHyeqFasS0abE9rqNQDfAC46LiJ+MMbW/DGpn3YvKcbh9Q563f/0TNr8drGvQiGNdzxxSMNz21sjWemd3X0Y1d7HxqrS9K63paOfry0NjrN+WvzJkFRFHzi0DF4/J1mvLZxrx7kP7dmFw70BjGuugQ3fmYa/vJOM97YtA+BUATzDqrFb74yK2Gwnczv9eCX5x9p+7xbnD+nCVv29eB3//4INWVF+J8vz8Lxh4zWnz+krgI/PCuxNYCIRo5TTjkl6RGZiqLg1ltvxa233mr7mpqaGjz22GO5uLwRJZ7J5+A9IqJCwiB/mOrqDwEA2noCWLHtAI47qDbPV5QdKz4+AACYM2mU4fGmmlKcNq0e/1rfij8u/Ri3fH5myvdaua0Nr23cCwBYu7Mj4XlRqq//7G0H8Lk0g/zHlm9HKKLh2Ek1+pC3U6bGg3xA9Ox/DAD46ryJOPuIRpx9RCOa23qxoaULp0wdM2w2aQDg+jOm4qQpo3FofUXSjQsiIsovvSefmXwiooIyfCIHMhgIxXfdX1pnfy6wGzS39eKW59ZhZ3tfyteu3BYN8mdPHJXw3MXHTwQA/N/KHejqDyY8b/arlzfpf960pwt9gbDh+U2xIF9Mp0+3ZD8QiuCxZdsBAF+LXRsAHH/IaHhUBVv29qC5rRertrdj7c5O+L0qviQNhWuqKcX8GfXDKsAHolm44w8ezQCfiMjlwhp78omICtHwih5INxCKB6z/XNeatPRxKPQFwrbX8D+vbMaD//kYf4xNoLfTMxDCB7s7AQDHTKpJeP7EQ0bjoDFl6AmE8fS7O5O+17It+/Hm5n3weRRU+L2IaMCHLZ2G12yIleufdXj0mKZ3YlUETq3d1YF93QOoKSvCGYfFj22qLPZh9oToJsW/N+7FI7Es/ueObMSoIe6lJyIishOO5QuYySciKiwM8ocpMTAOiJ5PvnZnZ5JX59bmPV046tZ/4qa/rbV8/p1Yn32qTP6a5naEIxoaq4ote+MVRcHF8yYBAB76z8e2mQdN03DHyxsBAF+c04SjY1UBa3fFP6NwRMNHe6NBvphm/2FLJ7oHQkmvUdbeGwAAjKsugc+Ujf/E1DEAgKdW7cAL7+8GgIyn9xMREeVChNP1iYgKEoP8PHrh/d34f0+/jxfX7s76e8uZfCC/JfsbWroxEIrg2dW7Em4U2noC2LK3BwDQ0pE8yF8hSvUtsvjCebPHo6rEhy37evD8e7ssX7P0o/1YvrUNRR4VCz55CGaOi/bKr5P68rft70EgFEGJz4PZE0ZhXHUJIhrw7nbn2fz23mjLQHVp4rnunzg0GuS/u70dwbCG2RNHYea43B55R0RElA5Rrh8KM8gnIiokDPLzaOW2A3hs2Xa829ye9fcWPfnl/uhsxXwG+eImobM/hPW7jRUFosceAHa19yd9HxHkz7HoxxfK/V5ccfJBAIBf/2sTQmHjRGBN03BnLIv/5WOb0FhdgpmN0eB67a54kC+G7k2pL4eqKvqgvxVplOyLIL+qJDHInzG20tCT/rV5ExNeQ0RElE8ikx/Jc8sfERGlh0F+HnlVBQAQzsEOeX8wmsn/5LQ6+DwKNu3p1svPh1pEyt4v/Wi/4bkV0jC71s5+w2tl4YiGd5MM3ZNdfPwkjCqNZvP/ttqYzf/7+7uxYtsBFHlVfOuThwCAnkHf0NKFQGxzRByfN6Uueva72FiQNyVS6eizD/JVVcHJh0aPjhtT4ceZM8c6fl8iIqKhoGfyeYQeEVFBYZCfR6oI8nOwQy4y+WPK/Zh3cDSYzFc2Xy7RX7rFFORLQXMoomFfz4Dle2xs7ULXQAhlRR5Ma6hI+vPK/V588xMHAwDufmUTgrFs/vs7OvD9J9cAAC47cTLqK4sBAONHlaCy2ItgWNMz+OL3Q+vLAQBzYi0C724/YKgOeH3jXvzjfet2CxHkW5XrA8CFcydiVKkP3z/9UH2CPxERkVuE2ZNPRFSQGFnkkUeJBvl22evBGIgN3vP7VJxxWD0A4KV1rVn/OU7ImxjLt7bpQXJ/MIz3d0RL5Itig+laOqxL9kXGf9aEUY6OlPvavImoLSvCtv29eGrVDuzu6MNlD7+D/mAEnzh0DL73qUP11yqKomfz18VK9vUgP7ahcGh9BSr8XvQEwviwJfrcS+tacPGDy7HgsVXY3524OZEskw9EKxLeXXQ6vnTMhJR/HyIioqEm7k/Yk09EVFgY5OdRLjP5/bHBe36vik/NqIeiRKfT2wXRuSRvYnQPhPQp9mt3diAQjmB0eRGmN0aH39n15a9wWKovlBZ5cdUpsWz+ks247KEV2NM1gEPry/Gbr8xK2Cg4PBbkr93ZiWA4gq37osMAD62PBvkeVcEsqWR/7c4OXPP4amgaENGA1s7EIF9M168u4bF4RERUeMT9CTP5RESFhUF+HolMfthhq9u6XR2OS+5FJr/Y50FdRTGOjp3L/s8Phr5k37yJIfry5cC9sSpaOm83YV8MvBMD8Jy4cO5EjKnwY2d7Hz7Y3YnasiLcf/ExqCxOzKwfJoL8XR34eF8PgmEN5X6vfl1AvC//pXUt+MbDK9AXjJ9gcCAW0Mv0TL5NuT4REZGbieA+xCCfiKigMMjPI5FMdlqu/+0/v4tv/nEldhzoTfnaASmTDwCfmhEt2X9z074MrnRwzH8/0ZevB+4Ta9AQC6Z3dyZm8ls6+rGzvQ+qEi3Xd6qkyINvxbL5RV4Vf/jaHDTVlFq+dmaskmD97k58EDsBYEp9OZTYRkz0OqM/+62P9qOlsx+H1JXjsNj3WQX57SnK9YmIiNyMPflERIWJQX4epVuuL45kE78n0y968r0eAMDYWBDd1R9K+zoHS9wcHFIXHWK34uM2BEIRrNouzr0fhcaqEgDWPfkrtkX78aePrdSPBHTqouMm4vunH4qHLj0maan/pNoylBV50B+M4MW10WqHQ+uMA/6OmlANT+zfrKasCA9cfAwmxDYNDvRYZPJ7kw/eIyIicrN4Jp/T9YmICgmD/DxKd/CeGFjn5Lxakckv9qmx36PBvujVH0piXs/0sZUYVepDbyCMv63eibaeAPxeFTMbq+KZfIue/HjG33kWX/B5VFx96hQcHzthwI6qKjisMVqyv2T9HgDRTL6stMiLU6fVodzvxR++OhsTaktRXRrtt2/rMW68aJqWcvAeERGRm4nbE2byiYgKS3ppUcoqkRV22uuWzmIrjtATmfwSEeQHh343Xmxi+FQFcyfX4sV1Lfjtvz8CABw5vhpFXlWvNNjdmdiT/25zOwDg6AyC/HQcNq4Syz9uQyC2mTLV4qi+3180G/2hMEqLov/TqSmLBvDmcv2eQFj/d+XgPSIiKkTsySciKkzM5OeRJ81yfVEu5yiTHzT25OuZ/ODQZ/LFzYGqKph3cC0A6NPrZ8cG6YlMfmvHgKGyIRLRsDF2ZJ3ItOfKTNP7i8n6MlVV9AAfAEbFMvnmIF9k8Ys8ql5NQUREVEg4XZ+IqDAx+sgjEeQ7LdcXLXFOpvGLTL4I7kWgmY8gX2xKeJR4kC+IEvz6ymIoChAIR7Bf6m/f2d6HvmAYRR4Vk2qth+Zly8xx8SC/stiLugp/yu8ZpZfrG4N8cXxeVanPMLyPiIioUEQ4eI+IqCAxyM8jVT9CL71Mflrl+rHgviSPmfywlMmfUleO0eXx8nUxDM/nUTGmPBpUy8P3NsSy+AeNKUs42z7bDh5Tplc+TG2ocBSc15Qlz+RXsx+fiIgKlMjks1yfiKiwMMjPIz2T76D8XtM0vSffyev7bcr1+/IY5HtUQFEUHHdQNJs/pa5cH1wHxE8A2N0R78vfuCca5Fv1x2eb16Ni+tjokXhTLEr1rYwSQb5p8J6YrM+he0REVIg0TYPGwXtERAWJQX4eedLI5MuvSSeTL4J7v16uH4HmcAZAtsjl+gBw9hFjAQBnHNZgeJ3oy2/pjGfyN7V2A7Duj8+F06bVAQBOnjLG0etHlVoP3mvv4/F5RERUuOR7DWbyiYgKC6fr55GqD95L/Vp5OJ+TQX3mwXuiXB+IbgAUS1/nmlyuDwCfnjkWr1/3SYytLja8bmxVCQBgV3tiuf6UOuNxdrly1SkH45xZ49BU46z/X2TyewNh9AfD+ucqyvUrmcknIqICZLjviAz9yTxERJQ5ZvLzSLSYOxm8FzZNnE+l33SEnhzUDwzxMXriRsGrxnvcJ9SWwmfqsRfl+i2xcv1wRMPmvdFM/lCU6wPRkn2nAT4AVPi9+t+rvTdesi/+zOPziIioEBky+U6yEURE5BoM8vPIo0Y//pCDHfJ0yvVD4Yj+GjFV3+dR9RkAQ92XHzFl8u006D350Uz+tv09CIQiKPapaBqV28n6mVIURZ8rIE/YF5l89uQTEVEhSrdNkIiI3INBfh6JHnUnVXCGTH6Kcn3Rjw/EM/kAUOzNzzF64sg/T4pp9aJcX/Tkb4z14x9SV55ygyCfasoS+/I7+qJ/Zk8+EREVIvnehEE+EVFhYZCfR6Ja3UmPvXFHPflr5SBe9OQDQElR7Bi90BBn8sXgvRSB+lgpk69pGja1Rvvxh2roXqZGlSYeo6eX6zPIJyKiApTuLCAiInIPBvl5pGY6Xd9hJr/Ioxoy4CKr3xcY6kx+rFw/RSa/vjIa5AdCEbT1BLChQIL8Gv0YvcRyfQ7eIyKiQsSefCKiwsUgP49EZtvJufdyYJ9q8N6APnTP+M9bLB2jN5TCDjP5RV4Vo8v9AKLZfHF83lSXB/nxnnyrwXsM8omIqPBENPbkExEVKgb5eaQfoedg8ZR30VO9XpTr+33Gf95cluvf/+ZWPLGi2fI5sSmRKsgH4iX7Ow70Ysu+aJA/pX5ojs/LlFVPficH7xERUQEzZPIZ5BMRFRQG+XnkSaNcP5JGb9yA6fg8oTj29UCWB++19wbwk+c/wI+eXgvN4tqclusD8SB/6Uf7EQxrKCvyYFx1SVavN9vMPfmhcARdAyEA8Sw/ERFRITHOAhraCkAiIhocBvl5lE65vryLnrJc3yaTX+yL9eRnOcgX5f+BcARWlxYv10/9XiLIf23jXgDAlPoKKA42B/JplOkIvc7+kP5cZbE3L9dEREQ0GPK9CTP5RESFhUF+Hokg38niGUlj8F6/XSY/Rz35IWmH36oqIZ1MfkPsGL2P9/cCAA51eak+IA3ei2Xy22O/V/i98DrZ2SAiInIZYyafQT4RUSFhBJJHeibfSU9+Bpn8YptMfn+WM/ny4m9VlRDOoCdfcPtkfQAYpU/Xj/bht4t+fB6fR0REBYo9+UREhYtBfh7pR+g5ma6fxo66/XT93JTrp7oRiDicrg8UaJBfahy818Ghe0RElCPPv7cLtzy3zvHxuz9+dh1eeH+35fMtHf24/v/WYO3OjsTv5XR9IqKCxSA/j+KZ/NSvNQT5Kdbafj2TPzTl+qk2INIbvGccslcQQX4sk98bCKM/GEaHOD6PmXwiIsqy/35pAx78z8f4YFdnyteu3dmBh976GL98aYPl88+t2YUnVuzAH5duS3jOvLZbDdYlIiJ3YpCfR+lM15d31FOW69tk8kt8uZmun6qVIBzbU3CSya+r9Ot/riz2ol762q0q/F54Y3+3A70BZvKJiChnRDXegIPjcMWmf/dAyPJ5cRKM1XuZExBM5hMRFQ4G+Xmkxj79tMv1Mz1Cbwh68pOW6zvI5Bf7PKiNZcYPLYDJ+gCgKIp+VN6BniDae0WQz+PziIgou0Kxcj6n5foA0B+wXvfF/YBVhaD5XiPEY/SIiAoGg/w8SmfwXjo9+f0pBu9luyc/lMXBewAwtjralz+lAEr1hZqyeF++yOSzXJ+IiLItGCuPcxLki/W53ybrrwf5FgG8+f3Zl09EVDgY5OeRJ8PBe87L9e0y+UPbk5/O4D0AmFBTCgCYMbZwgvxRsUx+W08A7X3RAXws1yciomwLikx+GvcOwbCGUDhx7RdBfsgilW/etOeEfSKiwuHN9wWMZCLoDaeapAfj4pq6XD+6aCdO1xeD93JXrp908J7DIH/hp6ZiWkMlzj16fHYucAiIIL+9NxAfvMcgn4iIskyUzTsJuuXX9IciKPcY7wv6Ypv+yarw9K8d3KsQEZE7MJOfR3qQ72A3PpJOJj+2aCdM1/fmqlw/nh1IFuQ76ckHgEPqyvGd06agzF84e1Biwn5bT5CD94iIKCc0TdMz+c5a/eLrs9UGv57Jt6rCizCTT0RUqBjk55GaxnT97GTyxXT9HJbrW1xbvFw/qz/WVeSe/HYR5LMnn4iIsijVoFsz+TV9FsP34j35Fhv0pvWcPflERIVjGIdd7qcP3kt3un6KGF0E8X7T4L2Soli5voNjd9KR+gi9WLl+AUzKz5Tck89MPhER5UKq9dZMvnewOiYvWZBv3kTgdH0iosLBID+P9HL9NBfqVJsCIoi3K9fPdk9+JEVmQbTxOR28V4hEkH9A7skv5RF6RESUPUFpl99JJl++d7Aauisesxyay+n6REQFi0F+HonMdkSL9tklI5fNpVpo9Uy+qVzfPwRH6CW7UXA6eK8Q1cR68ne19yEQuwnj4D0iIsomeQq+kypAQ7m+xdrfl6xcnz35REQFi0F+HsmZ7VRrZzjFcDtZv96Tb8zklwzBEXpWNx3ixsDp4L1CVB3rv9/e1gsA8KoKSos8yb6FiIgoLYZMvoNp98ZMfpJy/STzdPSvGeQTERUMBvl5JAe9qQJ3uQ8/1e59fLr+0ByhF0pRri9uDIZzub7I5Iupx9WlPijDeFODiIiGXjCNIbzAIAfvmfIBzOQTERUOBvl55PGkE+Q7z+QPhES5vqkn35ebnnz52iwH72kjYPBembH/vpKl+kRElGWhsPN7AQAIS6/vD6XXk8/p+kREhYtBfh4ZMvmpevLTyOSLIN7uCL1gWMvqYi1fW7Ke/OGcya/we+GV/n7sxyciomwLSiX66R6/259mJt+8ac9MPhFR4WCQn0eq9OnnJJPvs+7JB7KbzU91bWIDwzOM/2tTFMUwTZ/H5xERUbaF0rgXML/GfHxuMBzRA3cng/fCPEKPiKhgDOOwy/3kTH6qgTbyYmvukzMbCFln8uWvsxnkh1L0CIprH87l+gBQUxYP7Hl8HhERZVtoMJl807ovf+2kXN/JoD8iInIHBvl5JJevpyrXlxfq1OX61oP3VFVBUSzQz+YxesYNiJFZrg8Ao5jJJyKiHAqm2ZMvJxD6AsYMgXwfkGxobjo/j4iI3IFBfh4pigKR3E6VyZcD+9Tl+tZH6AG5OUYvVWYhXq7PIJ+IiChThp78NKfrm8v1B4LJNwzMgT978omICgeD/DwTJfvpZPKTvVbTNKknP/GfNxfH6KXagBCJh2Ef5EsT9qtLGeQTEVF2pT1dP8kReqnK9c1Vg8zkExEVDgb5eabGAt9Ui6ec6U+W9Q+EIxDrslUmPxfH6KVqJRCPeUZQTz4z+URElG3BFO1xZvL6PGDK5Mvl+snm6Vi9FxERuRuD/DzzOgzyQw4X9gHpHFxzTz6Qm3J9+XqsbgL0wXvDPZNfykw+ERHljpzJdxJ0yxPxzet+f4py/cTp+gzyiYgKBYP8PNPL9dPJ5Ccp15d77Ioszqzz5yKTn6InXx+8N8wz+ezJJyKiXJJ78lPN8gGMGwEs1yciGjmGVZD/4x//ODbMLv5r2rRp+vP9/f1YsGABamtrUV5ejvPOOw+tra15vOJ4djvVxHynmXyxaPu9KhSLoLo4Nl3fPIBnMORMgdXfY6QM3qspk4N8HqFHRETZFYqkm8m3H7zXl+oIPVPBn/yziYjI3YZVkA8Ahx12GHbv3q3/evPNN/Xnrr32Wjz33HN48skn8dprr2HXrl0499xz83i18cDXvJiayf1yyY6qFeX6ovfeTDxu3tEfDPnarM7RHSnl+nKJPsv1iYgo2+Q1NlVyAGAmn4hopPLm+wKyzev1oqGhIeHxjo4O3H///Xjsscdw6qmnAgAefPBBTJ8+HW+//TaOO+64ob5UAIDqsFw/7LBEL358nvX+jd6TH8riEXocvAfAnMlnkE9ERNkVkHvyk+34x8j3DuZ1P1WQz8F7RESFa9hl8jdt2oTGxkYcdNBBuPDCC7F9+3YAwMqVKxEMBjF//nz9tdOmTcOECROwdOlS2/cbGBhAZ2en4Vc2ibb5VDvy4RTH1AlikI7V8XlAfBjfQBZ78sOGnvzE50N6Jj9rP9KVxo8qxTGTRuEzhzfAZzEPgYiIaDDSzeTL9w7mdd8weM/BdH1m8omICsewyuTPnTsXDz30EKZOnYrdu3fjlltuwUknnYS1a9eipaUFRUVFqK6uNnxPfX09WlpabN9z8eLFuOWWW3J2zU4H78nPWy3GgsjkF1scnwfk/gi9sKlnT9M0/Ui/4Z7J96gKnrzy+HxfBhERDVPGnvzUFXnyvUNfQpBvzORrmmaY5WPeRGAmn4iocAyrIP/MM8/U/3zEEUdg7ty5mDhxIp544gmUlJRk9J433ngjFi5cqH/d2dmJpqamQV+rIPrUkwXugHGhTl6unyqTH+vJz2YmP8lQQPnr4T54j4iIKJeCKSrnzOTA3Ly5b74PiGiAR1qmzUF92MkPJCIiVxjWNcXV1dU49NBDsXnzZjQ0NCAQCKC9vd3wmtbWVssefsHv96OystLwK5u8apYz+fp0/VSZ/Owt1smGAsrPDffBe0RElHvhcBg33XQTJk+ejJKSEhx88MH4yU9+Ak1abzRNw6JFizB27FiUlJRg/vz52LRpUx6vOjtCYflseyeZ/PhrEgfvGb/fXBlgTigwk09EVDiGdZDf3d2Njz76CGPHjsXs2bPh8/mwZMkS/fkNGzZg+/btmDdvXt6uUc0gyHeSyS9O0ZOfzXL9ZEMB5XuG4V6uT0REufeLX/wCv/vd7/A///M/WL9+PX7xi1/g9ttvx29+8xv9Nbfffjvuvvtu3HvvvVi2bBnKyspwxhlnoL+/P49XPnjG9jgHr3c4eC/6fsl78J3MACAiIncYVuX63//+9/HZz34WEydOxK5du3DzzTfD4/Hgy1/+MqqqqnDZZZdh4cKFqKmpQWVlJb797W9j3rx5eZusD8QD32SBO5BOJj9Wrp8ik5/Ncn35piOhvE9juT4REWXPW2+9hc9//vM466yzAACTJk3Cn//8ZyxfvhxANIt/11134Uc/+hE+//nPAwAeeeQR1NfX45lnnsEFF1yQt2sfrGDamfz4GhwIRRCJaHpyIWWQz558IqKCNawy+Tt27MCXv/xlTJ06FV/84hdRW1uLt99+G2PGjAEA/OpXv8LZZ5+N8847DyeffDIaGhrw1FNP5fWaPU578jVnu/f9Do/QG8hmub50o5HsXF2VmXwiIhqk448/HkuWLMHGjRsBAGvWrMGbb76pz+XZunUrWlpaDKfpVFVVYe7cuXk9TScbDEG+g5jbHJiLewQgdZBvTj6EnfxAIiJyhWGVyX/88ceTPl9cXIx77rkH99xzzxBdUWqqw+n6Iafl+kFRrm+Xyc9+uX4oyeC9CAfvERFRFt1www3o7OzEtGnT4PF4EA6H8bOf/QwXXnghAOgn5tTX1xu+L9+n6WRDyDB4L71MPhDtwy8tiv7ZXNHHTD4R0fAxrDL5hUgEvql63SJOy/VTZPL1wXuh7AX58rUnK9dnjE9ERIP1xBNP4NFHH8Vjjz2GVatW4eGHH8Z///d/4+GHHx7U+954443o6OjQfzU3N2fpirPHOF0/ddBtHqYnB/bmwXuJPflI+jwREbnXsMrkF6L44L3kr3Oaye9PmcmP9eQHspjJTzp4L/q1R1UM5+8SERFl4rrrrsMNN9yg99Yffvjh2LZtGxYvXoyLL75YPzGntbUVY8eO1b+vtbUVRx11lO37+v1++P3+nF77YMlBu5OgOzGTH1/7EzL5WvJyfWbyiYgKBzP5eSbOpM3aEXpOM/lZ7cm3vzbxNSfrExFRNvT29kJVjWucx+NBJBYAT548GQ0NDYbTdDo7O7Fs2bK8nqaTDelm8pMF+QOmID8UTl6u76Q9gIiI3IGZ/Dzzxm5U0gryHRyhZxvkxx7PZrl+sp588bXK7SQiIsqCz372s/jZz36GCRMm4LDDDsO7776LO++8E1//+tcBAIqi4JprrsFPf/pTTJkyBZMnT8ZNN92ExsZGnHPOOfm9+EEKSWV/TjLryYJ882a/uW1QZPJ9HgXBsMZMPhFRAWGQn2ci+E05Xd9xuX4sk29Trl9SFMvkZ7FcP9kGhNj4ZyafiIiy4Te/+Q1uuukmfOtb38KePXvQ2NiIb37zm1i0aJH+muuvvx49PT244oor0N7ejhNPPBEvvvgiiouL83jlg2do3XNwbn3CdH0psDeX69vN1CnyqAiGw+zJJyIqIAzy80wfvJe1cv0UmXx98F6OyvVtbhJUTt0jIqIsqKiowF133YW77rrL9jWKouDWW2/FrbfeOnQXNgQCcibfwZF25jVZnsdjPmXHrge/yKuiJxBmJp+IqICwiDrPnB6hZwyk7V8njtCzy+QXe0VPvnFxX761DV+9fxm27O1Oec3Jr826h4/H5xEREQ2OXK6fTiZfFNPJrXqpMvkRKciXvyYiIvdjkJ9nIvhNWa6vOSvREwt4sW0mP9aTHwxDk97nsWXb8MamfXhuzW5nFy4xTPtNGNQT/Z3l+kRERIMjZ+/T6ckvK4oWbsqZ/IGUR+gZg3xm8omICgeD/DwTwW+qHfJkw+1kKTP5sZ78iGYs+9vXHYj9PuDgqo2SzQuID95jkE9ERDQYQYfzeQSxCV9aZGzVC0c0/R5APJcwUye2ae/3Wj9PRETuxSA/z1SHmfyIw4U95RF63njwLw/gEcF9JkG+vAGRUO7HI/SIiIiyIu3p+rHMf3lxNJMvjs2TW/bK/F7L99Mz+R5m8omICg2D/DzLKJOfrFw/FrgX22TyfR4FIqkun5ErMvn7Y7+nw0kmnz35REREgyOX6zvJrIt7h3K/sVzfEOTrFX7mwbnR30W5fjiSvYG9RESUWwzy88zjiQa/qXbII07L9VNk8hVFQUlsA0AM3QlHNLT1ZJ7JTzb5Pz5dP+23JSIiIklQnoHjIMgXgbsI8sXcHrH+F3lVeEWm3jSt3zx4z8k0fyIicgeGXnnmcThdXx5ul6yyP9UReoB0jF4s69/eG4D48YMN8u2m87Jcn4iIaHCCYftBt1bEmlymZ/Kj3y/W/xKfB15xlG/C4FxjuT578omICgeD/Dzz2CyuZvLaOphyffk5Ua63vydeot/ZH0IglF5JXshBuT4H7xEREQ1OuuX6ek++KZMv1v9in6of5ZvQk69xuj4RUaFikJ9nqp7JT/66kMMSvVTl+gDgl47RA4B9Xcbs/f6e9LL54SStBGEO3iMiIsoKQyY/g578/qA5yPfA67GeDRRhJp+IqGAxyM+z2NqZOpMfMX9t/Xq9XD9JJt/ck7+vxzhsL93he8mCfHHdHLxHREQ0OE6P0zW/piwhyI+X69tl8kPmnnwO3iMiKhgM8vNMBL/p9OQD1iX7kYiml9oXp9GTb87k702zLz/Z5H89k88gn4iIaFDSn64fXefL/cZ1X2zy+6We/IRNelO5PmN8IqLCwSA/z1SHg/fM5fxWrw9IL0qWyS+OleuL0n5zeX76mXz78kF98B6DfCIiokFJp1w/EtH0eT5lNkfolfhUfWZOQrsdM/lERAWLQX6e2U21NTOfT2v1+oGgFOQnyeTr5foB0ZNvDOrTnbBvGLxnM51XZU8+ERHRoCSrnDOTny+zHbwnZfJt1m9xP8GefCKiwsEgP8/EDnqqqbV2O+wysXh7VAU+T7LBe+bp+tGgviJ2E7A/zSDfcIRemOX6REREuRCUTr8Jpzi3Xl6bK/SefHGEXizI93qktkHrZAKn6xMRFR4G+Xkmps7bDdIT7AbayUQmP1kWH4gu6gDQH7tZ2Bcrz5/aUAFgcIP3zJl8vVyfmXwiIqJBCcrtcSkrAC0y+ebBe0VykG/9/X5O1yciKjgM8vPM6eA9u4F2MifH5wFASZHpCL1Y5l4E+ekO3ks2XV/s/Kv8L42IiGhQ5Gq5VJl1+fnyYmOQ36eX66vS4D1zJj/6OzP5RESFh6FXnqk2vXBmjsr1YzvzxUmG7gHxTL5Y5PcPIpOvaVrSI30iLNcnIiIaNPN6m04FYLkYvBdM7Mm3O0JPfL+PmXwiooLDID/PMi7XH0QmX2wCDAQj6BkI6Yv+1PpokJ/O4D3zZdsN7uHgPSIioszZnWNv//roxr+qxAfumsv1i30eeD3W9yGcrk9EVLgY5OeZk0y+pmmJwbTF4j4QEj35KTL5vni5vsjaF/tUTKgtBQC09QRSbjoI5kXfrqePmXwiIqLMmQfbAskTBGL99aqqvrnfH4xA0zR9c78kSSbfPHgv1aA/IiJyDwb5eSYy+ebgWGYV0FuX68d77JIRi31fMKz339eW+VFb5gcQXeg7+oKpLx6JAwDtpvNy8B4REVHmghaZ9GTZfLEp4FEVw33BQCiCAcuefOtKgSJRrp+irZCIiNyDQX6e2ZXJyawWcetyfaeZ/HjZnjgub3SFH0VeFVUlPgDxY/VSSczkm8v9or+rzOQTERFlTD4+T7C6FxDkSjp5Vk9/MCwN3vPEKwoTTvExZfLZk09EVDAY5OeZXZmcTF7EfR77afx6T77DTH5/MIL9PdFy/dFlRQCA2vLo73u7nA3fS5wVYHqemXwiIqJBs7pPSJrJl4J8nyeese8Lhg2D97w2bYPiaz+n6xMRFRwG+XkWq4JLuhsvL6yibM7q9WKQTqpMvjyAZ19XLJNfHi3VHx0r2XeeyTeX95nK9dmTT0RENGjBcHyQnpAsux7vyY9+g7zBLw/e04/yDZsr8diTT0RUqBjk55mq2GfmBbmU36eXzSW+TvTYpc7kS4P3Ypl8kcEfXRH93ekxeonlfdbPs1yfiIgoc6LHXi69dxLke0xBfl8gbBi857HJ5It7D5E4YCafiKhwMMjPM7vFVSYvrMnOq4335Dsv19cH78Uy+WL4ntNj9BJ78K2n83oZ5BMREWVMVMoVeVWIDrj0MvmxDf5Q2DCo12OTbAibp+szyCciKhgM8vNMBPnJBu/JJe9iMU5Wri/v8luRF3p98J7I5JeLID+zTL55p1/P5LMnn4iIKGPBcPxIPLvAXCY2BTyxWT5yq57ck+9RrYN4UZknkgvmdjwiInIvBvl55qRcXx+eoyjxzH+ywXsOM/l9gbAezIvgXpTtO83k252rK+iD9/hfGhERUcZEub7PoziqAoxn8qMLcLEhyI8G7NFyfePr9e8XmXx9FlDyhAQREbmHN98XMNLpmXyHx+CoYjFOcoRe6kx+fKEXg3z0wXux3/c7LtdPfoQeB+8RERENXiC2XnvlID/JMLxQQk++mMcTQX9IKte3yeSbB+8B0XsPFVzPiYjcjkF+niXLzAthq3J9i9eL8junmfzeQFi/CdAH7+mZfGfl+nbl+fGvo7+zXJ+IiChzodiC6lPVNDP5FoP3AolH6MnruXyPId9ThCMaUuQRiIjIBVhEnWciaHd61q2atFw/vSP0Qnq/PDCqNBrc16adyU8V5Ef0ayciIqLMiDXbkMlP0icfX+NNQX4wbKj8s7qvMBzdawryiYjI/Rjk55mTcv2IlpjJT16u7+wIPaGmrEi/DpHJ75F2+pOx6+Ezf81MPhERUeZEe51XVfXsu9VxuoLYAPCaBu919AX118iZfHn9lu9JiqShOjxGj4ioMDDIz7NkmXlBDNvxqIo0jT/xdfFy/RQ9+abnxbF5AFDu9+q79k6G74VMPXt25frM5BMREWVOH7znVfWN82QT7+V7ByC+wd/eG2/HK/aqlv39YWbyiYgKGoP8PIv32Nu/Rs/kK0p8Gn+STH6qnnxVVQyL9uiKIv3PiqJgjCjZ70ndly8WfL/H5ggezXiTQUREROkTAb1PVfTse7J7B7ue/AO90Uy+z6PA67Hu75f/7FEViCWcx+gRERUGBvl55kkyLV+Qe/LjmXyLIF869zaVYinIlzP5gHSMXpeDTH5s598vtQDI1xaOsFyfiIhosILheE++qqbO5IdNm+yiXF9k8sW9gsfiKF95Hfeqin4MHzP5RESFgUF+nqkWi6tZOJK4sFu9vt9hJh8wbgSIY/OE2rJokL+/J3WQHzGdowuYsgH6BkXKtyIiIiIboiff54n35Ds5flcE6H49yI9m8vUg3+K+Qv6zfHxvKMmRfURE5B4MvfLMyeA9PVBWFMTm51iX64ue/BSD9wBjkC8y94II+p0coyeqDHw2PXtyqwERERFlRgTYXumknWRBt7knv0Qv1xeZfNXwvCHIj63dihJt42Mmn4iosDDIzzOrxdUs7LBcPyAdiZNKiSGTbwzya/UgP3UmX0zvNZ+ja/6zyp58IiKijAX1afmq5UR8s8SefDF4L5rJL0mSyRddAGKD3qO3BzDIJyIqBAzy8yzdID/Z4L34dH0nmXxp8J6pXF8E/fudZPLDxun65mtjJp+IiGjw9On6nvi9QLI5ePI8HyCeAGjvsy7XD1lk8sUGvdfBvQoREbkHg/w8sxp4YyYPz0m2KRCfrp86k+83lOubg/x0MvkWPfkWx/Awk09ERJQ5Q0++x8Hgvdhz5nJ9sS6bg3y5bTASMW7QexwM+iMiIvdgkJ9nqqOSu/hCnayHf0Av10/9z5q8XN95Jl9ct8928F70dy+DfCIiooyFpEF6+vG7Dk/mARLvDczT9UMWrXYeUyafMT4RUWFgkJ9nyXrsBREoG8r1LRbaeLm+gyP0fPZH6GWSyfd6rOcFRExH+BAREVH6giGRyY+vt8kG7yX25BvvDcRRuqIqQF67RcAvlm6Pg8oBIiJyDwb5eZasx17QM/mK/eC9UDiiL8rpHKFXVuRBSZFx4ReZ/LbeQMr+u/j0XjVpNkBlTz4REVHGglab6o4y+dF7AnOQL9Z+q1J88wY9p+sTERUWBvl55nFQAmeZyTct7AEpte9kun5xLNs/usKf8FxNaREUBdA0oK0necm+nClIdgwPM/lERESZC8XWea+qOpp2nzqTbyzXtzoZR/wcTtcnIiosDPLzzMngvZChJx+Wr+8PxoP8IgeZfLGDX1tWlPCc16NiVGmsL78necm+3POXbHgPB+8RERFlTqy3crm+o5N5PMbBe4I5k295/K0YvOfgXoWIiNyDQX6e6YtrkpK7iMV0fXOJ3kAo2o8vL/7J+GM9+ebj8wSnx+iFpSPyxI8NWfT18Qg9IiKizInp+l6PqpfgJ08QmDP5xls+cR9gFeSLewwvM/lERAWJQX6eOdmNj/e9y4P3TEF+0PnxeQBQXRIN4sdWFVs+X1MmMvkpgvzYTYfHo8AbKzMwDN7TM/2OLouIiIgsiHsBn0dFLDmfIpNvfYSeoJfrWyQbzMffiuF8YQ7eIyIqCN58X8BIZ1d+L5N31O02BURJv1iIU/ninPEIhCL4rznjLZ8XmwWBUPIFXc4UWM0LEH/m4D0iIqLMiXXepyoZZfL9qQbvhRMz+Qk9+Umm+RMRkXswyM8zEfwmO0IvJPXG2Z2NK5+f60RtuR/fnT/F9nmfOC7H6qw+SdjQkx+7lnBiNoCD94iIiDIXCInp+mo8QZDsZJ6wMRtvLtcXR+hZtQHqA38VMV2fPflERIWERdR55qgnXzo2R9UXWuNrRGDtzVIwLTYLgimCfDlTIL7HMHiP0/WJiIgGTc/kexRHR9qZM/lFHhXyUqxn8pMdf2vK5Ce7VyEiIvdgkJ9ndj32srQy+Q7L9VPxeUWQn3xBj0iZetWi9cA8oZeIiIjSJ2/mqw4y6/FKuujirCiK4Rg98Wdx32CYpyMN1Y3+zNSbCkRE5B4M8vPMblq+TD7r1m5hj5+fm6UgX5+k6yyT71EVyyN2xLczk09ERJQ5ebq+k/J5cyYfMA7fE0G+apHJD9lk8tmTT0RUGBjk51k6Z92qUt974uC97JbFi539VJn8sDQLwGoDgoP3iIiIBk+s89Gy+9T3DhGL+wLLTL6a+mQc9uQTERUWBvl5Fs/kA5pNNj8sT9e3K9eXjtbJBvE+TnvyPaoSvwmwOIaHmXwiIqLMxTP58fU22bn1Vpl8vzR8TwzeE612Vj354p7D4+DnERGRezDIzzOPlOG22yEXE3I9ycr1TefhDpYI8lOV5okzcw1H6Fn19fG/NCIioozpPfmeeOVcspN5whb3BXK5vhi8ZzU0V6/CMw/eS9HCR0RE7sDQK89UafG1m1oblibUeyzOogeksvksRdNi5z+YYkEXif5oK4FFub5p8A8RERGlT2TyfWrmmXyrcn2PRSY/YvpeZvKJiAoLI688k3fY7eJpuWzOY7N7H8z2EXqiXD/kPJPvVRNbCcwlf0RERJS+oLSZn87QXo+0+W/I5OtBfmxyftgik69P12dPPhFRIWGQn2eGcn27TL6UDY+X61u/JltBfpEn/en64trkEv+IXvKXlcsiIiIakUJST76TzLp1Jj++GIv+fCfzdMRGADP5RESFgaFXnsnBr21PvjTl1nbwXiS++GeDnsl3PF3feiggM/lERESDpw/YVVXbqj6Z1frrtzpCz2LDIKIZg3xm8omICguD/DyTF1+7xTpkmcm3nq6frd73+BF6TjP5qtSTH3+e0/WJiIgGT8zI8aWZybcdvKcfoZe4YaDP2xHT9T0M8omICgmD/DyTF9/U5fqwHbwnMvm+rJXri+n6yYN8+Sxdj1XJn2lCLxEREaVPnq7vsTjNxixsUeEnyvU9qqKfoiMCeavBe+ZMPsv1iYgKw4gN8u+55x5MmjQJxcXFmDt3LpYvX56X61AUBSL+tcvkhw3ZcuvXWu3YD0Z8un7yBd06kx/fGBB/ZLk+ERFR5sSmu5zJTx7k22fy5Yy+3LMv7i3E2q5n8nmEHhFRQRmRQf5f/vIXLFy4EDfffDNWrVqFI488EmeccQb27NmTl+tJVXanH6GnKLbl+uJrX7aO0NOn66c6Qk/qyWe5PhERUU4E9FN0VMvKOTOrgbyiD18ewCdX2on7kPjRvTC8BzP5RESFYUQG+XfeeScuv/xyXHrppZgxYwbuvfdelJaW4oEHHsjL9agpyu7CeomeYluuHwxnN5j2eZwt6KJNwCMP3pP7+kzH8BAREQ3Wzp07cdFFF6G2thYlJSU4/PDDsWLFCv15TdOwaNEijB07FiUlJZg/fz42bdqUxysevJBFT344yXBcudJOEEG+32uTyY+t2eZyfdXBzyMiIvcYcUF+IBDAypUrMX/+fP0xVVUxf/58LF26NOH1AwMD6OzsNPzKtlTn3cqBst1EXaveu8Hw6dP1nWfyLSf0MpNPRERZdODAAZxwwgnw+Xz4xz/+gQ8++AB33HEHRo0apb/m9ttvx9133417770Xy5YtQ1lZGc444wz09/fn8coHx9CTP8hMfklRPMj3WGXyTeX6zOQTERUWb74vYKjt27cP4XAY9fX1hsfr6+vx4YcfJrx+8eLFuOWWW3J6TakG6BgCaT2Tb3xNMJy4mA+GVx+85+wIPVVVrM/aNZX8ERERDcYvfvELNDU14cEHH9Qfmzx5sv5nTdNw11134Uc/+hE+//nPAwAeeeQR1NfX45lnnsEFF1ww5NecDWLTXT6yNllPfsiiwk+U6cvl+oYBwLH3Mx+hJ6oBOF2fiKgwMPRK4cYbb0RHR4f+q7m5Oes/Q02VyZcCaftMfnaP0BNT+p0eoSdn8o3H8LBcn4iIsufZZ5/FnDlzcP7556Ourg6zZs3Cfffdpz+/detWtLS0GCr2qqqqMHfuXMuKPWEoKvcGQ6y3RV41rcF78ua/1eA9eTBu2JTJ53R9IqLCNOKC/NGjR8Pj8aC1tdXweGtrKxoaGhJe7/f7UVlZafiVbVYD62RWJfHmhV2eupsNerl+igVdvhHwODiGh4iIaDC2bNmC3/3ud5gyZQpeeuklXHXVVfjOd76Dhx9+GADQ0tICAJYVe+I5K4sXL0ZVVZX+q6mpKXd/iTRpmmYz6DZZT370vkAerHfE+CqU+Dw4ZlKN/piqKhBxvh7kSwN/AU7XJyIqNCMuyC8qKsLs2bOxZMkS/bFIJIIlS5Zg3rx5ebmmlIP3LAJpcx9e1o/QE4P3HGfyVX2nn4P3iIgoVyKRCI4++mj8/Oc/x6xZs3DFFVfg8ssvx7333juo9x2Kyr1MBaXWOUNPfpqZ/EPqKrD65k/h+k9PM7zWXP5v3qBnJp+IqLCMuCAfABYuXIj77rsPDz/8MNavX4+rrroKPT09uPTSS/NyPd4U5fpyAC962+3K9bN1hF66g/c8cpWBJmfyoT9PREQ0WGPHjsWMGTMMj02fPh3bt28HAL0qz2nFnjAUlXuZktdiebp+sqDbbvNfnqwvxN8v+nPEjxPrul2rIBERudOIG7wHAF/60pewd+9eLFq0CC0tLTjqqKPw4osvJpT2DZVUi7U8AEcs1bk/Qi+9wXuGKgOLTD6DfCIiyoYTTjgBGzZsMDy2ceNGTJw4EUB0CF9DQwOWLFmCo446CgDQ2dmJZcuW4aqrrhrqy80KeS32qvFMvl1yAJAz+ak3/72qggHEN+bN5frM5BMRFZYRGeQDwNVXX42rr74635cBABDrr13Znb4bryT2zQn6EXpZLtcPpui/SzUvwDy8h4iIaDCuvfZaHH/88fj5z3+OL37xi1i+fDn+8Ic/4A9/+AMAQFEUXHPNNfjpT3+KKVOmYPLkybjpppvQ2NiIc845J78XnyF5LfZ5rGfgmKWz/qoJmfyI4Xs9Hk7XJyIqJCM2yHcTsVjb7ciL8jivNFTP/NpgGjv2TvhUZ5l8uRzQq1r39AHG6b1ERESZOuaYY/D000/jxhtvxK233orJkyfjrrvuwoUXXqi/5vrrr0dPTw+uuOIKtLe348QTT8SLL76I4uLiPF555uTj8BRF0e8HkpXPhy3uHeyY2wb1cn1m8omIChKDfBewm5gv6BNyLY650b8OO1/MndAz+Sl78qVze01/D7mlQGUmn4iIsuTss8/G2Wefbfu8oii49dZbceuttw7hVeVO0HSCjqoYM+9W0hnIa24bjLcJxp5PMSCYiIjchUG+C+iZfLue/Nga7lUVaKbHBPm8+myID95znslXTZP/5ZsBlusTERFlRqy1osoufpqN/fdYTde3k7BJH/vdPHiPmXwiosLAIN8F9MXVdrp+/Kxb8ZLEI/SM/XOD5XN4hJ7cSmAuHwyzXJ+IiGjQxFos1llzD73l96RxX2DO1Ov9/KJc3yOeT35PQERE7sAg3wXUFGVwIpnuVRWIl5hfq+/yZ+kIPW+amXxVUaTyQaty/axcFhER0YgT0IN8YyY/2RKdzuA9j8d4HyLK9b3mTH6KewIiInIHBvkuoGfAbTL5YSmTr2ia5WvFLn+2M/nOp+ureu8eB+8RERFljwiufaoxk58ss55WT75NJl/8HPNgXSIicreMgvyenh7cdtttWLJkCfbs2YOIaZHZsmVLVi5upNAz4DY75GGpJ1/8OfEIvSz35MdS75oWfW+7mwT5JsIT+54Ie/KJiIiyRpTeJ2TybWL8SETT2/ucnLpj7snXB+8pIpMfO0LPJhlBRETuklGQ/41vfAOvvfYavvrVr2Ls2LFQmKUdFLG4psrkexQFMGXLhaA+XT9b5frxf9NgOAKP6rG5tnhPvvncXnEzoCjgfyNERCNQe3s7qqur830ZBS9oOkEnnnm3jvJDaW6ypxq8x0w+EVFhySjI/8c//oG///3vOOGEE7J9PSNSfLG2ft7QVxd7jXlDIOuZfGmzIBiOoNiXPMiPZvJj16aX60e/Zqk+EdHw94tf/AKTJk3Cl770JQDAF7/4Rfz1r39FQ0MDXnjhBRx55JF5vsLCJSr9imILrTkoN5PvEZxN1zdm6sX9iIc9+UREBSmjtO+oUaNQU1OT7WsZsUQlnV0ZnBxIqzYLe7yUL/tBfrJFXZ7Aq98kmDL5Kkv1iYiGvXvvvRdNTU0AgJdffhkvv/wy/vGPf+DMM8/Eddddl+erK2xB0xqfKshPP5Nv/D5DBSGYySciKjQZBfk/+clPsGjRIvT29mb7ekYkvVzfdrq+nC0Xpf3G14hAPFuZfI+qQCTgkw3fk4/o8Zg2KyLSBgAREQ1vLS0tepD//PPP44tf/CJOP/10XH/99XjnnXfyfHWFLb7GmzL5dsmBcIaZ/LDYpI8+LjbpnRzZR0RE7pFRuf4dd9yBjz76CPX19Zg0aRJ8Pp/h+VWrVmXl4kaKlEfoheNBfqoj9DxZPKvOp6oIhCNJj9GTe/LNf490ju8hIqLCNmrUKDQ3N6OpqQkvvvgifvrTnwIANE1DOBzO89UVtmCsft5nzuTbrM9yMO5sun7098RN+ujjzOQTERWWjIL8c845J8uXMbKl3JHXnAT52S3XB6I3E4Fw/Hg+K/J0ffNNgF6uzxifiGjYO/fcc/GVr3wFU6ZMwf79+3HmmWcCAN59910ccsgheb66wiaCfMeZfGltdjL41mtutzNt0us9+QzyiYgKQtpBfigUgqIo+PrXv47x48fn4ppGHG+qcn1psY3YDN7Ldrk+ICb1h20z+eYjesynBESYySciGjF+9atfYdKkSWhubsbtt9+O8vJyAMDu3bvxrW99K89XV9hC5un6DnvynbbLJUzXN83UMW8CEBGRu6Ud5Hu9Xvzyl7/E1772tVxcz4ikKsl3yOXJ+WE18Qx6+XudnIfrlCgLtOvBkzMIHiU+FDAUNt4kZLOFgIiI3Mnn8+H73/9+wuPXXnttHq5meBHrsJiun6p8Pt12OXOQb56pw0w+EVFhyahc/9RTT8Vrr72GSZMmZflyRiZzBtxMLKqqoti+Vu6NzxaxYRAMJb+JAACPJ16uH9GMGxEexvhERCPChg0b8Jvf/Abr168HAEyfPh3f/va3MXXq1DxfWWELmjL5qZIDISk54IRdJl88Ln4uM/lERIUhoyD/zDPPxA033ID3338fs2fPRllZmeH5z33uc1m5uJHC7lg8ISJl6c39coLek5/F0nifN/pedtP15ZsLr5pk8B6n6xMRDXt//etfccEFF2DOnDmYN28eAODtt9/GzJkz8fjjj+O8887L8xUWLjEbx6tn8qO/27f5xU6+cbjxnxDk2/TkM8gnIioMGQX5orfuzjvvTHhOURRO0U2TJ8V0fT2Tr9pvCJiP18kGX+y9QjY9+fJUX/l4v5DpJkFlTz4R0bB3/fXX48Ybb8Stt95qePzmm2/G9ddfzyB/EMS66tOPtDM+bibm5aadyRczdcyZfAb5REQFJaOIMBKJ2P5igJ++VOX6YSmTLzYEzOtsyLTrng2iPC9oM13f3JOfMHhPy/41ERGRO+3evdtyXs9FF12E3bt35+GKho+AXSbfts0vlsl3GuQrNpv0CT359qftEBGRe7Bb2gXiZe7Wz8en3NqXzIVMZ+hmgy92M2EX5IvFXlGi2frEcr/o61iuT0Q0/J1yyil44403Eh5/8803cdJJJ+XhioYPUVEn1vjUmfz0qvtEWb8o/xexfDyTz+n6RESFJKNyfXMpntmiRYsyupiRSgyms9qRNx9TFxTT9c1H6OUkk5+iXN802MfcdsByfSKikeNzn/scfvCDH2DlypU47rjjAER78p988knccsstePbZZw2vJef0nnzVmMnXtOh9gnmdTfeewJzJF5v4IgmRalOBiIjcJaMg/+mnnzZ8HQwGsXXrVni9Xhx88MEM8tPkSbJDnlASrxh324X4Ln82e/KTl+eJn+nRewRNR/CIcn1m8omIhj0xr+e3v/0tfvvb31o+B3B2TyaCEeMaL6+rYU2DCuM6a96ET0U/HUefrh993JzJt9tUICIid8koyH/33XcTHuvs7MQll1yCL3zhC4O+qJHGk2SH3HxMnRq7LzJn8tM9E9cJ0ZMfSJnJN53bqxmf580AEdHwF2G/ds6YW/LkqfnhiAafx/z69NZfVTVm8iP6PQViv8ffJxTRUMR1nYjI1bKW9q2srMQtt9yCm266KVtvOWLYZecBU5AvZfI1DdCkQF8/Qi8HPfmhFIP3EjP5EdPzWbskIiIqAP39/fm+hGElKE7QEUG+YgzyzTLN5Ovrt2nwnvw+7MsnInK/rIZfHR0d6OjoyOZbjgh6cGzRky9n9+Vj6gCpLD6i6dP2s3qEnsOefE9CTz7065IfJyKi4SscDuMnP/kJxo0bh/LycmzZsgUAcNNNN+H+++/P89UVNn0jP7bGG+4FLO8d0puuH9+kj35tPh3HmMlnxQYRkdtlVK5/9913G77WNA27d+/GH//4R5x55plZubCRJFkmP2IK8lXTwu5F4kZAtoid+4DddH1TT35CTx/L9YmIRoyf/exnePjhh3H77bfj8ssv1x+fOXMm7rrrLlx22WV5vLrCFgwZp+sbgnyLjfhsZfLN67v8HBERuVdGQf6vfvUrw9eqqmLMmDG4+OKLceONN2blwkYSu2PxAGMAryrGrLjYTJd31bN6hJ43Rbm+6SYi3tMXfT0H7xERjRyPPPII/vCHP+C0007DlVdeqT9+5JFH4sMPP8zjlRW+oN6SF12X5djdKpOf7pwej6miMGxav62qCImIyL0yCvK3bt2a7esY0ZKV68slc4qiWJbo5SqTH5+ub72gm8sBxe/i5WJvgJl8IqLhb+fOnTjkkEMSHo9EIggGg3m4ouFDVM6JTXVxPxCOaCl68p218JmP0IuYNglS/TwiInKXjBq4v/71r6Orqyvh8Z6eHnz9618f9EWNNMnK9c1n3aoWw3bknnlfFnvyRcYgaNOTLzYgvKYgX1yXORNARETD14wZM/DGG28kPP5///d/mDVrVh6uaPgQm+pF3vga76QK0HEm32M+Qi+x3c6TYuOfiIjcI6NM/sMPP4zbbrsNFRUVhsf7+vrwyCOP4IEHHsjKxY0U5jI5WSRiHUjLz4nFX1GymzUXpf/BFD354mfGB+8ZMwHZnPhPRETutGjRIlx88cXYuXMnIpEInnrqKWzYsAGPPPIInn/++XxfXkHTp+tLG/nmNVcWTnP9TczkGx+P/mwFAZufR0RE7pJW2rezsxMdHR3QNA1dXV3o7OzUfx04cAAvvPAC6urqcnWtw5aj3XhFZPLjz+m9c2kO2HEq5RF6NhsQeibfdAQPERENX5///Ofx3HPP4V//+hfKysqwaNEirF+/Hs899xw+9alP5fvyCppYh+Wg3ZvFTL55cK7VdH7zRsBI9sGuTvx5+XbDUcZERG6SVia/uroaihLtDT/00EMTnlcUBbfcckvWLm6kSBbk68NzPPG+OFWJ9r1HTOX62Tw+T36/oG1PvriJMB7pY958yOacACIicq+TTjoJL7/8cr4vY9gR6608XFdNUj4f1o/cS+8IvZC+SR97XNqkF/chYR6hhx8+8z7e3d6OaQ0VmDVhVL4vh4goQVpB/quvvgpN03Dqqafir3/9K2pqavTnioqKMHHiRDQ2Nmb9Ioc71XS+vCxsyuQD0aA5EtYSBu9lP5MfK9cPZZjJ15jJJyIaKQ466CC88847qK2tNTze3t6Oo48+Glu2bMnTlRW+QEgE7fHNfD37bpFNDqVZSWd+L3nor/k1zOQDB3oCAID2Xg6UJCJ3SivI/8QnPgEgOl1/woQJUBi8ZUWsKt5yobbKhkcXbU0avJdYxpcNerm+zYJuvjbV1B8Yfz6rl0VERC708ccfIxwOJzw+MDCAnTt35uGKho+kmXyL4bjp9uSb38tq/fYk+XkjzUBs02UglPjfOxGRG2Q0eG/ixIl444038Pvf/x5btmzBk08+iXHjxuGPf/wjJk+ejBNPPDHb1zmsmYNjmVWQrx9VF0uwm8vms8WbavCe6drMPX1WmQAiIhpenn32Wf3PL730EqqqqvSvw+EwlixZgkmTJuXhyoYPfTPfaSY/nN59gdfUbhexqAQQP5uD94D+YDS4H7CpdCQiyreMgvy//vWv+OpXv4oLL7wQq1atwsDAAACgo6MDP//5z/HCCy9k9SKHu2TT9cMWgbI+UdfU++7LVSbfZtfevAFh15PPcn0iouHrnHPOARCdGXPxxRcbnvP5fJg0aRLuuOOOPFzZ8CGm6/ukI/TUJIPw0h3Im1CJpx+Rm3hkH8v148F9gEE+EblURqnfn/70p7j33ntx3333wefz6Y+fcMIJWLVqVdYubqTwmDLgsrDFhFvV1PsuMu3ZzpiLmwP7TL5xsE/i4B5m8omIhrtIJIJIJIIJEyZgz549+teRSAQDAwPYsGEDzj777HxfZkET661P7pH3JKkCTLOSztxvr2/Sq4mvsaocGGni5foM8onInTIK8jds2ICTTz454fGqqiq0t7cP9ppGnOTT9Y2vkf8cGaIj9Oym65uDeNtyfWbyiYiGraVLl+L555/H1q1bMXr0aADAI488gsmTJ6Ourg5XXHGFXvFHmdFP0ZGa5D0OWv2c3heYkw1W7XbsyY8KhSP658tMPhG5VUZBfkNDAzZv3pzw+JtvvomDDjpo0Bc10iRbqPWzaqVA2VxWF7RY/LNBlP+HbDL55psINaGNIHa9zOQTEQ1bt9xyC9atW6d//f777+Oyyy7D/PnzccMNN+C5557D4sWL83iFhS8YSRywmyxBEO/JdxrkGwft2p3sY/fzRpJ+KbAP2NwfERHlW0ZR4eWXX47vfve7WLZsGRRFwa5du/Doo4/ie9/7Hq666qpsX+Owp5p62WURy0x+9HfzFPtsZ/LFpoFdub5dT76mRbMBzOQTEQ1/a9aswWmnnaZ//fjjj2Pu3Lm47777sHDhQtx999144okn8niFhS8YivXkW/TIW2fyje10qciVeJqmQbylapXJj4zswHYgGJb+PLI/CyJyr4wG791www2IRCI47bTT0Nvbi5NPPhl+vx/XXXcdvvGNb2T7Goc9R5l8i8F7Ioi22uHPBr1c36Y0zzxdXw7mw5om9fQxyCciGq4OHDiA+vp6/evXXnsNZ555pv71Mcccg+bm5nxc2rARSpbJt5qun+apO/JMHflexGOYrs9MPmDsww9YHBlJROQGGWXyFUXBD3/4Q7S1tWHt2rV4++23sXfvXlRVVWHy5MnZvsZhz9xjL4toiVl68+C9cJpH5Till+vb7NrHKwiiP9cj3XyEpRuFLHcREBGRi9TX12Pr1q0AgEAggFWrVuG4447Tn+/q6jIM6aX06dP1PVaZ/MQ1Ot31Vx6qJ28aWGfyGeTrf2Ymn4hcKq3wa2BgADfeeCPmzJmDE044AS+88AJmzJiBdevWYerUqfj1r3+Na6+9NlfXOmyZg3aZ6KuzWmjFBoBYcH3ZLtdXB5HJZ7k+EdGI8JnPfAY33HAD3njjDdx4440oLS3FSSedpD//3nvv4eCDD87jFRY+MRvHZ9mTb/H6TDP5YQ3ynoFcRSjuCUZ6Jr9fKtdnTz4RuVVa5fqLFi3C73//e8yfPx9vvfUWzj//fFx66aV4++23cccdd+D888+Hx+PJ1bUOW3r5vcVaYZXJj5f3R7+2KunPBlEWaN+Tbz5CT3qO5fpERCPCT37yE5x77rn4xCc+gfLycjz88MMoKirSn3/ggQdw+umn5/EKC5845cZ6ur59Jj/dnvywKZNvNXiPmXypXJ/T9YnIpdIK8p988kk88sgj+NznPoe1a9fiiCOOQCgUwpo1a6AwW5sxfZBekr46VZ6uby7X1xf/7P4bFMUuzO64HPP0fK8U5Uci8RuFbA8EJCIi9xg9ejRef/11dHR0oLy8PGGz/8knn0R5eXmerm540DP5FlV91pn89Db/5VN75Ey9vHkv7jGsNhVGEsPgPQb5RORSaQX5O3bswOzZswEAM2fOhN/vx7XXXssAf5A8SUrgrAL4hMF7YWNvfLakncmX/jMIRTR9VgAz+UREw19VVZXl4zU1NUN8JcNLtP0t+mevRU++1dwcsWynncmPaIjYDN5LtqkwkvQzk09EBSCtqDAcDhtK8LxeL3fns8B8JJ4s7CiTn95ROU7Fe/KtFzFzT76iKHqgL2fy2ZNPRESUGXkNtpqubzW0V9wXeBxW+MnH8RnK9Q09+czkA8zkE1FhSCuTr2kaLrnkEvj9fgBAf38/rrzySpSVlRle99RTT2XvCkcANckRelZ9debyfj2Tn+1yfW/y/jura/OqKgLhCMJaPBuQ7VkBREREI4W8BvtUi0y+RUtdKM2efI9FJl9VYKjUZE9+lGG6fohH6BGRO6UV5F988cWGry+66KKsXsxIlXw3PjFQjg/qM/XkZ7tcX03ek281vVdVAYSj3yM2IVRm8omIiDISkjL58nR9r6N7B2f3BVaZfPMGPafrR3HwHhEVgrSC/AcffDBX1zGieZJk8s0l8UBiub7Va7IhVU9+PFMff0yeFyC+jZl8IiKizMjH2BruBRT7zHqmmfxQJGLZJggYj9kbyfpZrk9EBSC7qV/KiB60W+zGRyx21M2D98Quf66m66fuyU8sH5RL/hjkExERZUYM1vN5FEP5vFjzI1atfmkOvo1XFFpXEALG4XwjGTP5RFQIGOS7gL64Wu3Gh61K4o0TbtPdsXfKm/IIPat5AYklfyzXJyIiykzI5gSdXGfyzUNz2ZMfJffhB0b6UQNE5FoM8l1AH7yXLJMvrbUe0+v1GwBPtnvyY+X6NpN0rc7h9UhVCVbl/ERERORcwKZaL1lmPWyxPicTn/UTv+8wVwFwun7UQJCD94jI/Rh+uYA+PMdi3UxWEh8fvJebI/R8erk+M/lERET5IDbyfaYdc/N8HsP3DCqTb3ws8TUjO5PfL2fyWa5PRC7FIN8F5MXVLGyRDTcv7MEcTdcXU3zDEQ1akum98m6/PETQrq+PiIiInBFzccwBu1d1UAXocP316us9bAfvsSc/Ss7kM8gnIrdikO8Cerm+xVphdQyOqNgTC7ueUc/y4D25/N8qm2+VKZA3INK9ySAiIiIjsdaaM/l65ZzV+mzTx28nvkEfkdZu02t4hB4A4+A9TtcnIrdikO8C8am2yc66tXi9OEIvnJtgWj6P12rCvlWmXj631y4bQERERM6IE3R8HuvyeatMfrqVdIZWO5vqQK+HmXzANHiPQT4RuRSDfBcQAbz18JzExdY8qE8/XidHPfmA9YT95Jl82Pb1ERERkTNBm+G6cnucWaY9+YZ5OgmZfPbkA8Zy/ZC0KUJE5CYM8l1AVYyZeZnV8LqETL5FSX82yDcHVhP2RYmgxyO3EsTnC8RPBmCQT0RElImQzXDdZOXzeibfYRuf5ck45iP0kmwqjCTmifrM5hORGzHIdwEnJXdyv71qWmhDNsfrDJaiKPpNhVUmP2wRxMc3IKwH8xEREZFzQb1c39yTH/3d6t4hZBOo27Eq1zev3cmGBI8k5j58BvlE5EYM8l3AHLTLrPra9ZL42MvTLctLh9g4SNaTb3mEnqZZPk9ERETOxcv1bTL5VpvwaR6taxXkmzcIOF0/qj9ozOSbM/tERG7AIN8FxMKdbPCeIZCO/VGU1OXyqDqRObAK8kMWPzd+oxBhJp+IiGiQRCWdT80gk+80yBdtg5r994rSf6vKvpHEnMnnhH0iciMG+S4Q72N3NiFXNZX36zcA5vNuskC8p/W1JbYJyMcBWpXzExERkXP6cF1v+j35Ttv45OG+YlPffDIOM/lR8uA9AAhYnX9MRJRnDPJdQATtmgZoph15y2y5uSc/dgOQi0y+WNStes6sju6TbwL04T38r4yIiCgjQZsz75MNwosnCJwtwPLLxHqfkMlX7Tf9RxJzeb456CcicgOGXy4gZ7rNi3UkSUm8Pl1fz+TnrlzfalG3mp6vyn19FicDEBERZdNtt90GRVFwzTXX6I/19/djwYIFqK2tRXl5Oc477zy0trbm7yIHIaQP3jNl1pOcW5/uTBx5A0Fkps2tdszkR/Uzk09EBWBYBfmTJk2CoiiGX7fddpvhNe+99x5OOukkFBcXo6mpCbfffnuerjZOXkjNvXVWmfyEcv0cHaEHxG8qQk578pX4tVltUBAREWXLO++8g9///vc44ogjDI9fe+21eO655/Dkk0/itddew65du3Duuefm6SoHJxixzuQnG9qbbk++ZSbf9K0eBvkAeIQeERWGYRXkA8Ctt96K3bt367++/e1v6891dnbi9NNPx8SJE7Fy5Ur88pe/xI9//GP84Q9/yOMVGxdh88k0IlvutQikI6Zy/Vxk8r364D1nPX/6EEE5k88gn4iIsqy7uxsXXngh7rvvPowaNUp/vKOjA/fffz/uvPNOnHrqqZg9ezYefPBBvPXWW3j77bfzeMWZCYasj8lNllkfTCY/aNGKJ78Xy/Wj/x5lRZ7Y15yuT0TuM+yC/IqKCjQ0NOi/ysrK9OceffRRBAIBPPDAAzjssMNwwQUX4Dvf+Q7uvPPOPF6xqVzfJpOvWpxFbx68l8uefMvp+uHECgJVGiIovoWD94iIKNsWLFiAs846C/Pnzzc8vnLlSgSDQcPj06ZNw4QJE7B06VLb9xsYGEBnZ6fhlxvEN/JNmXzTvYDhe8LpzeqRXxaIBa2JPfnx03NGMhHkV5b4ADCTT0TuNOyC/Ntuuw21tbWYNWsWfvnLXyIUCunPLV26FCeffDKKior0x8444wxs2LABBw4csHy/oVj05YXUriffboK9/D25OI++yCt68hMXMcvj/aR5ASzXJyKiXHj88cexatUqLF68OOG5lpYWFBUVobq62vB4fX09WlpabN9z8eLFqKqq0n81NTVl+7IzErSZu5Mss57u0bqKouivtcvke5jJRzAcPx64otgLgEE+EbnTsAryv/Od7+Dxxx/Hq6++im9+85v4+c9/juuvv15/vqWlBfX19YbvEV/bLfxDsegby/XNmfzEo2zEZr4o5bfr18sGr2nRl1kN1pOrDDh4j4iIsq25uRnf/e538eijj6K4uDhr73vjjTeio6ND/9Xc3Jy19x4MUTXntcnkm+8bgPR78uXXBmyO0GNPfjyLDwAVxb6Ex4iI3ML1Qf4NN9yQMEzP/OvDDz8EACxcuBCnnHIKjjjiCFx55ZW444478Jvf/AYDAwMZ//yhWPTlNdi8Qy6y9XK2XDUttKJ0zpPTnvwkmXzp53qkcn1m8omIKNtWrlyJPXv24Oijj4bX64XX68Vrr72Gu+++G16vF/X19QgEAmhvbzd8X2trKxoaGmzf1+/3o7Ky0vDLDfRy/TR65MMZbP6L9XvA5gg9L4/Qw0Aw3n/PTD4RuZk33xeQyve+9z1ccsklSV9z0EEHWT4+d+5chEIhfPzxx5g6dSoaGhoSjtARX9st/H6/H36/P/0LT4OiKFAVIKLFs/OCCODl4XXms3H1I/RykMkvEkfoWWTyxY2H3fF+IpPvcf1WEhERFYrTTjsN77//vuGxSy+9FNOmTcMPfvADNDU1wefzYcmSJTjvvPMAABs2bMD27dsxb968fFzyoARtMvnmIbyCpmkZZfLFpkF8uj4z+WZiA6TIq6LYy8F7RORerg/yx4wZgzFjxmT0vatXr4aqqqirqwMAzJs3Dz/84Q8RDAbh80XLrF5++WVMnTrVMJk3HzyqgkhYS1g8RWxt2feecIReLjL59oP3wuLGw+LawpH434Xl+kRElC0VFRWYOXOm4bGysjLU1tbqj1922WVYuHAhampqUFlZiW9/+9uYN28ejjvuuHxc8qCIIXrm6fp2PfLyl+nM6lFV43pvPhmH0/WB/lgm3+9V4fdFN11Yrk9EbjRscqxLly7FXXfdhTVr1mDLli149NFHce211+Kiiy7SA/ivfOUrKCoqwmWXXYZ169bhL3/5C379619j4cKFeb56+/NuwxbZcvNrxQ1ATo7QS1KeZ7W5IAf5LNcnIqJ8+NWvfoWzzz4b5513Hk4++WQ0NDTgqaeeyvdlZUQE3eZqPfOGvyDfR6TTxpcyk+/hdH0R0Pu9Hr3SMWCRBCEiyjfXZ/Kd8vv9ePzxx/HjH/8YAwMDmDx5Mq699lpDAF9VVYV//vOfWLBgAWbPno3Ro0dj0aJFuOKKK/J45VGpFmvLkvghyOT7kmTyI1riz9U3IKRyQWbyiYgol/79738bvi4uLsY999yDe+65Jz8XlEVBi/k3gJTJD9sH+Zlk8gO2PfnGk31GoniQr+qnDw0ER/AHQkSuNWyC/KOPPhpvv/12ytcdccQReOONN4bgitJj7rMX9CDfaoK9Pngvd9P1ffrgPftMvnwT4ZUz+Zr1jQkRERE5E6/Ws87kh03JAfnI20x68u3K9eP3HiM3qBXl+sU+Ff5YTz4z+UTkRsOmXL/QiTI4J5n8eLl+9Ov4UJ7c9eSHkvTke6TNBdWiJ99c8kdERETO6MN1bTL5dskBIL3Nf3FvMRAW5frG5zld31SuH8vkc7o+EbkRg3yX8CjWZXDW5frR38WGgNhVT6cszynRA2hVrm+VyRfXZhi8x558IiKijARtqvXsKgDlIDyd5VcftKuX65srB2D580YScYSe3yeV63O6PhG5EIN8lxCBcMhUBhdO1vduOkLPfLxONvi8onzP/hxewwaEEq9IEPcBzOQTERFlxm64bnwQnn0FoJLG+ivWalF+br6lEEG/1ZG6I4Xck+9nJp+IXIxBvkvEz7s1Ph4KW2XyjX14Vhn1bImX51mU61tsQIibAEO5PjP5REREGQnabOSnyuSnu/Z6TD359oP3Rm6QH+/J9zDIJyJXY5DvEnYDdJJO1xeZfItj9rLFp/fkG69L06yDeEO5vsZyfSIiosEQQbd5I98u6BbzctLd+BdruQhazSfj6NP8R3CQbzldn0E+EbkQg3yXUG163fRsubTYJpTr2xyvkw1em3Ng7Y7okQfviU0IlusTERFlRmzkm6frqzbJgUw3/s1Bvn0mPztB7UAojCdWNGN3R19W3m8oyIP3mMknIjdjkO8Sci+7LGwRwOuZfC0aSItvyeUReuZMvryTL98IeKWbjngmP+uXRURENCIEQ2K6vs0ReqbkgH58bZpBvnj9wBBl8l9c24Lr/+89/PKlDVl5v6Eghuwxk09EbsfwyyVUu7I7MaFeWmzlPrygtKOei0y+z24goM0RPeLaQuH45gMz+URERJnpGggBACqKvYbH7YL8eE9+erd4asqe/PjMnWzY0zkAANjVXjiZ/P5g9LOJ9uR7ADCTT0TuxCDfJbymPntBz+RbnUWv2ZfNZ+26RLl+yLqNADBN/rfYFODgPSIiosx09gUBAJUlPsPjtoP3MuzJF68P2AT5dtP8M9UTiG5etPcGs/J+Q8GQyY/dHw1YHDFMRJRvDPJdQu+ztynXlzfkRcVeJKIZjrbLTbm+TSY/nLxcXy5f4+A9IiKizHT2x4J8h5n8TE+2Efchoj3AXK6f7en6vYFowNzRV0BBfiyT7/dJ5fqxiftERG7CIN8l7HrdLDP50u59rjP5Tnry5R8bL/eTNgFYrk9ERJS2SERDd6xcPyGTb3PfkOkwXvH6eCbf+Lz88zRt8IF+T+zvdaA3MOj3GioigVEsD95jJp+IXIhBvkuYj8UT4mfRJ742rGkIxRYXRclNxlws+kGb6fpeVYFiMS8gGGK5PhER0WB0DYT0+Tbmnny9zc/u6N00N9hFAkGfrm8evCd9nY1kvsjk9wcj+vnzbiey9nImnz35RORGDPJdwnwsniDK4j0Ww+0iEU3fsfflaIS9eF9zkG93RI/H1NMHJJb8ERERUWqiH7/Yp+qD3gR9Bo7D9TmVxJ580zR/qTLA3MKXCZHJBwqnZF8+Qo/T9YnIzRjku4THbkdeS9yRl8/GzbT3zimf17ocUKzv5hYB8zm7ubw2IiKi4UwEv5XFvoTn4pl84+OZ3heY129zub683mejL79Pyt4XyvA9efAep+sTkZsxyHeJ+JRc4+P6UTiexJL46OC96Dfk4vg8ID4LwC6Tb24RsM7k5+TSiIiIhjV96F5JYpAvquTMWfVMe/LNmwJ267v8MwZDzuS3F0hfvvEIPZbrE5F7Mch3CVEVZ56uH7HorfNYZPJzMXQPiE/XD4btBgImz+R7TD37RERE5ExnX2zonqkfH4gH8ebKef2+Ic02voT2u4Tp+vH3C4cHH+SLnnwAOFCAmfx4uX5hzBMgopGFQb5LiMXTPHgvZFF2p5frR+LBt9dcV5cl8en61pmChJ49MXgvbD24h4iIiJxJlsn3pMrkp12ub1rPzZl96cusZPIDck9+YWTy9Z58n6pn8iNa4j0SEVG+Mch3iXjgHl845YBfXmzlcv1cZ/LF5oHTTH78CD1Rzp+TyyIiIhr2OpP05Huknnz5SLuMe/JNLzcPzVUURV/zs9GT3ztQgD35wcTBewCH7xGR+zAEcwmxuMoLZ8gmyJdL+zOdouuUT7XOFNjdRHjN5frM5BMREWWksz9Wrl+SWK7vsRmEl6tMvvyYubUwE3Imv71Apuv3x0rzi30qiqQKSvblE5HbMMh3CauFU56077XJ5OtH6OWoXN8uk2/VRiB/LV5vHtxDREREzjjJ5APGe4fwII/QE6zWbz2TP8ie/HBE04fYAYWZyfd6VMthw0REbsAg3yVUJbEEzi6TL28IhMI5PkJPH7xnnclPuClQzOfsMsgnIiLKRNKefLtMfob3BQnT9C0q8VSb6r50ycfnAYXUkx8fvAdAz+YPBBnkE5G7MMh3iXhvXWJfnfw8YOzfFwtt7qbri8F75ky+dRCfMF2f5fpEREQZiU/Xdx7kZzqrJ/G0HPvXDLYnv1c6Pg8ADvQUSCY/FM/kA9EBfAAQCHPCPhG5C4N8l7AavGcI8uUj9CzK9dM9D9cpb4pMvn25vhi8xyCfiIgoE/FMvkVPvmKTyc908F7Cep54iygeG+x0/Z6AMSguhJ58TdPQH4z35ANSJp89+UTkMgzyXcJjUa4v/qwoxmDZulw/t0fo2Zbre6zL+3iEHhER0eA47sm3zOSnd1+QGOQnviZbmfweUya/o9f95fqhiAbx1xaZfDFhn0E+EbkNg3yX8CYp1zcHyvH+/fiAHV+uyvVtdu3trs08eI89+URERJnp0qfrJwb5iqLoZ9dbBfmDzeSbj9CTXzPYTH5vLJMv5v4UQiZfDuRFmb7ozed0fSJyGwb5LhEv148/Jqbl2pXERzQt47I8p0SmPrEnP/m1CTkqMCAiIhr24pn8xHJ9IJ6tD1skCNI/Qi/5eg7E7wnCgxy8J47Pa6gqBhAN+sVQO7cakIYFijL9olhGn5l8InIbhmAuES/Xjy8UYZtsuCihC0fi5fq5OkJPvG8gHIFmeROR/FxdlusTERGlLxzR0BUra6+yyOQD8Y10eSM+4558c2Veskz+II/Q64tl8hsqi/VqhA6XH6PXHwvki7yqnpgpYiafiFyKQb5LpJPJVy0G7+X6CL3otaW+iTCX93HwHhERUfq6++N96xUWPflAfKPd2OoXO3UnzYG8iZV4Fpl8aSbQYIie/HK/V9/AcHvJvsjkixJ9+c8M8onIbRjku4SenbdYqO1K6KKD92I9+Tmbrh//T0TuwYvYDN6zG8RHREREzonJ+iU+j54xNhO3B6FcTNe3WL9LfNHy9LaewQ3KEz35pX4vqkuLAADtLs/km4/Pi/5ZDN5zd6sBEY08DPJdQj4WTxBZfXNfnSpN4s95T770vgGpzED83ITMvc0gPiIiInKuo8/++DxBbMRHrAbvpbnJ7qQnf9aEUQCApR/tT+u9zURPfqnPE8/ku3zCfj8z+URUQBjku4RqUQIXimXy7QLniJTJT/eoHKfkXn+5B08vB0xxU8Agn4iIKH0ik291fJ4g7g+sM/mDO0LPqlz/xENGAwD+s3lfWu9t1jsQDZjL/F5UlxZIuX4skC/2xT9XvSc/zCCfiNyFQb5LWGXyIzaZfL1cX8rkp9t75/i61PgRPSGLTL75piBV0E9ERESpdfbZH58nWJ1bH87wviBh/baoBJh7UA08qoKP9/eiua03rfeX6Zn8Ig+qY38/tw/esyrXF1P2B4IM8onIXRjku4THk7hQ65l8u8F7Wua9d+kQ5YBBBzcRqcr3iYiIKLV4Jt++XN9jEeSHbE7mScXJEbgVxT4c1VQNYHDZfGMmP9qTf8Dl5fr64D2fXK4fDfiZyScit2GQ7xL6EXpSub6YlpssOx6M7Sz7cnggvdipDko9Z2GbckCW6xMREQ1ep96Tb5/J91i0+tm106WSWJlnfV8hSvbfHEyQHwuYS4s8BVOu369n8hPL9QfYk09ELsMg3yWS7cabM/lyCZ1YWDw5KtcH4tl6UVkgX6f5xzqZzktERETJdcaO0EvWk29172B3/G4qiZv01q87cUo0yH/ro/2GFsN09MaO0Csr8hZOuX5sY6LYJ5Xrc7o+EbkUg3yXkCfmC3pJfJISukA4sx37dIjd/GA49WAfJ+V+RERElFyng+n6lkG+zb1DKuZNebt2u6OaqlFW5EFbTwDrWzrT+hmC6MkvKfLEj9Drc3m5vkUmn9P1icitGIK5hDwxXxC78cmOpRM7y7marg8Avli6PhhOzOSnGtTDcn0iIqL0OZmu77FIEMR78gc3Xd9u/fZ5VMw9qBZA5n35vQHRk+9BlSjXd3sm32rwHsv1icilGOS7hGW5voPhdnomP4fl+uIYPUMmX9xEmH6u+WsO3iMiIkqfk+n6Wc3k2wz5tXKC3pe/P62fIfQMiOn68XJ9twf5/Xq5vsURegzyichlGOS7RLxcP/5YRO97T5bJH4JyfdGTL2fy7YYCMpNPREQ0aI4y+UkSBIPvybf/fjF8b/nW/Rn1o+uZ/CIvRolyfbdP17fI5OvT9RnkE5HLMMh3CTHgRi7Xt1uoDYP3hqAnX0zuD0USp/cmHJln+i+Kg/eIiIjSN9ie/FwG+YfWl2NMhR/9wQhWbWtP6+cA8SC/1B+frt8TCLs6WBabGdbT9Tl4j4jchUG+S4jeOXmhjtgs1PK0fbEgeu3G4GaBzxv9efI5sCGHPfnmkwGIiIgota40puuHDJn8iOE5pxKG/CbZpFcURTpKb29aPwcAegPx6foVxT6IH9Xh4mP0ROWkXyrX93tYrk9E7sQg3yVEK3vYQSZffkw/Qm8IpuuHpJ78sE1PvnkAIDP5RERE6Ytn8lMH+YahvRn25Ccb8msl0778QCiiz/gp9XvgURV9I6PDxRP2Rba+WC7XjwX8chKEiMgNGOS7hF5yJwXSkSRn3YrgORAS0/VzOXgvsSffLpOfUK7PTD4REVFawhENXQMik5+kXF+xyuRnVq5vHuCbapP+hEOiE/bf39GeVgZeZPEBoDR25nx1AUzYt8rkF8Uy+eI5IiK3YJDvEqKs3ZDJT3IMjnhoYAjK9UV2Xt6pjm9AmDL3SVoLiIiIKLXu/nggXJGkXF8E5hGr6fppnrqTkMlP8f1jq0pQV+FHRAO27+91/HN6Yv34RV5Vv3cRE/YPuDnItxq8x0w+EbkUg3yXEDvmhoVaBNIW62w8kz8Eg/e8ieX6IQeT/3N9XURERMORmKxf4vPow92sqBaZ/PjgvfRu8TJptxtT4QcA7OsecPxzegdEP348WK4ugAn74gg9w+A9D6frE5E7Mch3CatMfrKFWjX15Ke7Y58Onz7YRzpCL2ydKUgYvMeefCIiorR0OJisD8Q30iMW5frpbrInVuKl/p7a8vSDfJHJLy2K/91Eub6rB+/F7reKffHNifh0fQb5ROQuDPJdQgTH1sfgWLxeHbpMvgjkA1aZ/BRH7uSwi4CIiGhYEpn8ZJP1gfiGv3Umf5BH6DnYpB9dHs3A7+t2noHXJ+v7pUx+SQH05FscoSf+zEw+EbkNQzCXSD4hN/GfKbFcP4dH6HlEub6UyY9Yby4oigL5IQ7eIyIiSk9nX2zoXpLJ+kB8DbY6mWewmXwn6/eYDDL5vQOJmfwqUa7v6un6FoP39Ex+OC/XRERkh0G+S+jT9S12462G18XL9WPT9XNZru9J7MkXf0x2vB/Acn0iIqJ06cfnJZmsD0itfhab8OkOvs1kcO7ojMr1oxsYpUWJmXw3D96L9+SzXJ+I3I9BvkvomXxpnUh21q05k5/LjLn4+cFI4k2E1c+VA3tm8omIiNKjl+s7zuTHHxMb8ulm8s2vd1KuXxsr199vUa6vaRpueW4d/vLOdsPjvcl68ocoyO/sD+J7T6zBm5v2Of6eeE8+y/WJyP2SbxHTkIlPyJUC6Vj5nVU23KNn8nNfri+OuAmGrI73S7w2r6pA7Okzk09ERJSeeCY/eZAfn+cjb8Jn1pOfcITeIDP5H+zuxIP/+RhVJT586ZgJ+uM9A4k9+aOGuFz/Xx+04q+rdmB3Rx9OnDLa0fcMBBOP0JMz+ZqmQeE9DxG5BDP5LuGx2I1PlskXMX2mvXfpKPJYbEAkvTZm8omIiDLV2S968pPnYuKtfvHHks3zScbc9udkkz5ZkL+7vR9AdGK+KHUHrDP5VaVDO3hvb1f0ett6nG8qWA7e88QD/qB8A0dElGcM8l1CTKGPWE3Itei3N5fR5bInX8/kW07XtxgKyCCfiIjIVqrybseZfNUik68NZSY/moFv6wkYZgoBwO7Ofv3PezrjmwAiyC+z6MkfqnJ9Edyns6nQb5HJl4fwBcIs2Sci92CQ7xKqxRF6eiBtsZtuHoiT23L9WE++tICJUwCsMvnyYyzXJyIiilvd3I6ZP34J97y62fY1TnvyrTL52erJd/LtNWVFUBQgogEHeo1Z8dYOKcjviv9ZHKFX6pd78qObBV0DIcO9Rq7sF0G+w/YATdP0TL7ck18knRM8EOSEfSJyDwb5LmF1hF4kSV+dOfDPZca8yOIIPXETYTn53zB4L2eXRUREVHDe3LQXgVAEy7a22b5GP0Ivk0x+hj355io8J/3lXo+q99ObS/ZbpEy+KI8HgJ6BxEy+fIpAR1/us/kik98fjBhaCeyEIhpEDkbO5KuqAl8sEcJMPhG5CUMwl/Aky+SnOKYOgL7I5IKoEghatBJYTv6XbxSYySciItJtb+sFAHT12wez8Uy+w558LfHeId02vkzX7toy6wn7LYZMvlyun3iEntejoiIW6A9FX/5+qRffyaaCfESeXKIPxBMhnLBPRG7CIN8lrBbqZCXxmfTOZUrcKBgy+UmO0JMfS/ecXiIiouFMBPndseF6VtKdrh8ybMKLU3cyD/LT6QC0G74nZ/Llcv0ei8F7QHzCfscQTNhv64lfq7nNwIqc7S8ylSjKE/aJiNyCQb5L6OX6ciY/SUl8YiY/d/+UPr0nn5l8IiKiwWhu6wMAdA8kCfL16fopgnyPxb1DxP743aTvleHaPboiGuTLJfmAqSdfHrxncYQeAFQP4YT9NqnqwMnPEwF8kVdNuCcT5fvM5BORmzDIdwmxaMi78Ukz+erQZfJ9+nR9Z9N75ZsDZvKJiIiiAqEIdnVEg/wum0x+KBzRNwDkXnUr1pn8zI7Qy3TtFhP290mBc/dACF3SJoZcrm+Xya8qGZogvz8Y1q/B6c8TQ/Xk4/OEeCafg/eIyD0Y5LuEWFyNu/HRoNpqN97capfTnnx98F5ilYFVkK+ahvcQERERsLO9D6Irr3sgZFjzBTnDnyqT77WqAkxy/G4y8uvTWbtFuf5+qVxf7scHjFn+voBdJj+6WdCe48F75vJ8J+0BVsfnCSzXJyI3YpDvElY9+SJxnqokPvp1Dsv11cQj9JJN75WvN92eQCIiouFK9OMLPYHEbL6YrF9a5EnZimdVBZisnS4Zb6bl+uWJ0/VbY/344i2dZPKr9Ux+bnvyzQMCnZXrJx6fJ4jsPsv1ichNGOS7hKpP148/JobnpDqmDshtMK2X6zssB5SvLd2eQCIiomQWL16MY445BhUVFairq8M555yDDRs2GF7T39+PBQsWoLa2FuXl5TjvvPPQ2tqapyuOMwf5ViX7+mT9FEP3ACmTH0sQaJqW8RF6asbl+mLwXjx4Fpn8Q+srAAD7ewb04b2iJ1+erg8Ao8SU/p7cBvltpvd3UjkgsvTJy/UZ5BORezDIdwmPaaEG0svkp3tUTjqsp+s7O96P5fpERJRNr732GhYsWIC3334bL7/8MoLBIE4//XT09PTor7n22mvx3HPP4cknn8Rrr72GXbt24dxzz83jVUdt399j+Npq+J4+WT/F8XmAlMmPtdDJ1f9DlcmvtSrXj2XyZ4ythKoAmhYN3iMRDb1B60z+GJsBftmWEOQ7qByIB/kW5fo8Qo+IXCj1CkJDQlTkhS2OwUkVSNu9Jlt8Fj35Tqfrc/AeERFl04svvmj4+qGHHkJdXR1WrlyJk08+GR0dHbj//vvx2GOP4dRTTwUAPPjgg5g+fTrefvttHHfccfm4bABWmfzELHImmXzR6idm+QDp3xdkukEvD97TNA2KouiZ/MbqEowu92NP1wD2dA6gotirzyQw9+TXxYL8PTkO8s2VAlbl+pqmoTcQRpk/epssjtDzW5Xr+zhdn4jch5l8lxA99ZbDcxyU6/ty2JMvbiICDnvyeYQeERENlY6ODgBATU0NAGDlypUIBoOYP3++/ppp06ZhwoQJWLp0qe37DAwMoLOz0/Ar27bHjs8TLMv1+5wdnwfIrX6a4Xcg/en6iqLoPfSZDN4LhCP60X8ik19fVYy6yliGvrsfPQPh2M8Cir3WQf7eTuPQvmxr64luIoxKcmTfn5c347CbX8Lz7+0CEM/km68ZiGfyWa5PRG7CIN8lRDAsD96LJDumzpzJz2G5vi/WbyZnCJKW6ytyNiBnl0VERCNcJBLBNddcgxNOOAEzZ84EALS0tKCoqAjV1dWG19bX16OlpcX2vRYvXoyqqir9V1NTU1avVdM0NMcy+SL7bVmur2fyUxdb6pn8iMjkx+8hMqnwE9+TzvcW+zwoj2W8xfA9MXivobIYY2KbAHs6B9Anhu75PAmVfnq5fvcANC3x1IFsEeX6B40pB2Ddk790y34AwP1vbgUgHaFnmckX5fo8Qo+I3KNgQrCf/exnOP7441FaWpqwcAvbt2/HWWedhdLSUtTV1eG6665DKGRcQP/973/j6KOPht/vxyGHHIKHHnoo9xfvgNhwlxfopMfUDWEmX7y3sVw/GvCnLNdnJp+IiHJkwYIFWLt2LR5//PFBv9eNN96Ijo4O/Vdzc3MWrjDuQG9QD+qnj60EYJfJFz35qTP5HlOQHw5nJ8hP91vFpoWYXL87Vq4/tqoYdRXFAKJl+OI0gVJ/4gaGCPKDYc3RxPtMiWs8eEwZAKDDoid/T2yT4t3t7di2vyfp4D2/6MkPM5NPRO5RMEF+IBDA+eefj6uuusry+XA4jLPOOguBQABvvfUWHn74YTz00ENYtGiR/pqtW7firLPOwic/+UmsXr0a11xzDb7xjW/gpZdeGqq/hi198F7EIpNvESibM+S57MkXg/cCHLxHREQucfXVV+P555/Hq6++ivHjx+uPNzQ0IBAIoL293fD61tZWNDQ02L6f3+9HZWWl4Vc2iX78hspi1MYmyXdbTtePles76MkXrX5WmfxMll9xv5Hu2h2fsD+AYDiiZ/TrK+Pl+nu6+tEbC/LLihLL3v1eD6pjJfSZ9OW/83EbXt+4N+XrzJn8AxYbCvLwv7+t3hXvybcq1xfT9YMM8onIPQomyL/llltw7bXX4vDDD7d8/p///Cc++OAD/OlPf8JRRx2FM888Ez/5yU9wzz33IBCI/h/6vffei8mTJ+OOO+7A9OnTcfXVV+O//uu/8Ktf/Woo/yqWrMr1nQbSwNAcoSdn8iPJjtBjkE9ERDmiaRquvvpqPP3003jllVcwefJkw/OzZ8+Gz+fDkiVL9Mc2bNiA7du3Y968eUN9uToR5E+oKUV5rBS/y6JcvyOWya9wUK5vHtorznMv8qhQMqiki2fyMw/y93YNQNMAn0dBbVlRfKBe54Dek19SZP13iw/fS68vvz8YxsUPLMclDy7Hx/t6kr5WD/JHRzP5fcGwHsQL8ibDM6t3xnvyrcr1vczkE5H7FEyQn8rSpUtx+OGHo76+Xn/sjDPOQGdnJ9atW6e/Rh7EI16T70E8QDww1jTovWjJhtsZz6LP7RR7n+kIvUhE0xczq1kAXpbrExFRjixYsAB/+tOf8Nhjj6GiogItLS1oaWlBX190qF1VVRUuu+wyLFy4EK+++ipWrlyJSy+9FPPmzcvrZH3Rj99UU4qKWJbearq+CPJHlRalfE89kx+7b9h5IPoZNFYXZ3SNmfTkA0CtNGFfDN2rqyiGqiqGXvtkmXwgXrK/pzO9TP66XR3oDYQR0aKZ92TEdP1Jo8v0aodOqS+/NxDS2yp8HgVb9vZg1bYDAFJk8jl4j4hcZNgE+S0tLYYAH4D+tRi0Y/eazs5O/ebALNeDeAS5JN88JTdVJj/dCbrpEu8fjF3Pxj1dCIY1lBZ5UB9bkGWqwkw+ERHlxu9+9zt0dHTglFNOwdixY/Vff/nLX/TX/OpXv8LZZ5+N8847DyeffDIaGhrw1FNP5fGqge37pUx+rCfdqlz/QKxHvKrUSbl+9Hdxv7Bd2kjIhNg0GEy5vjg+r6EqutEwRvTkS5l8q558AHr//t7u9IL8d7e363/+2+qdtoP7guGIvolSW1aEqtjcA3n4nijVL/F5cPqMaHvHqxv2ALDuyRdBPo/QIyI3yWuQf8MNN0BRlKS/Pvzww3xeYs4H8QhyJj7sIJMvbwp4czhZH4hn8oOx7P2Kj6M72rMmVMNrMT5ffoiZfCIiyiZN0yx/XXLJJfpriouLcc8996CtrQ09PT146qmnkvbjDwW9XL+2RC/Ft5qu3xHrEa92NHjP2JPfLLUEZEKs32mX68c2/Pd1SUF+ZTRg14/Gkwbv2WXy6zLM5K/Z0aH/ecu+Hry/s8PydWIDRVGA6tIiVMeqJeRBf6JUv67Sj88f1QgAEKMOLKfrx7L7zOQTkZukbvjKoe9973uGRdnKQQcd5Oi9GhoasHz5csNjra2t+nPid/GY/JrKykqUlJRYvq/f74ffn5itzja5xF2cVBfW+94tyvWHsO/d3JO/Mla2NntijeXr5coCZvKJiIiMPfniz1bT9UVWudpJuX4sGA+ZMvmZBvneTDP5sUGC+3sC+vF59ZUikx+9hwqEI/rU/VKbnvwxGfbkr26O3pc0VhVjV0c/nnl3F44YX53wOtGPP6q0CB5ViWfypQn7YoOhrsKPU6bWoarEp2f/i5OW6/MIPSJyj7xm8seMGYNp06Yl/VVUlHqRA4B58+bh/fffx549e/THXn75ZVRWVmLGjBn6a+RBPOI1+RzEI3isMvmx36121A2Z/BwH0l5zJn9bGwBgzsRRlq83bkDk9NKIiIhcLxCKYHdHtC2wqaYU5f5YT74pkx+JaHrAWe2oXN94Ms+2QWfyYz35mWbyuwcMx+cBQLHPowfTYihemd8mk18ZP27Pqf3dA2hui3621316KgDgufd26YkSWVvs+Lya2KbEqFIR5MuZ/Oj1j6nwo8ir4jOHj9Wfs8rkF3lYrk9E7lMwIdj27duxevVqbN++HeFwGKtXr8bq1avR3d0NADj99NMxY8YMfPWrX8WaNWvw0ksv4Uc/+hEWLFigZ+KvvPJKbNmyBddffz0+/PBD/Pa3v8UTTzyBa6+9Np9/NQDGQF6cc9sTW/yLfYmLoRxIW5XMZ5PI5AfDEezp7EdzWx9UJVqub0XuHmC5PhERjXS72vsQ0aLT2ceU+6WefOPgve5ASC8Nr3JUrm/M5DcPuic/Nl0/zdsKvSe/a0AfvFdfFR/+JzL0H8fmEpTYDd6T3sepNTvaAUTPvT/r8EZUl/qwt2sAb320L+G1YuieCPL1cv2+eCZf9OSL+QDnxEr2AevBeyLwZ5BPRG5SMEH+okWLMGvWLNx8883o7u7GrFmzMGvWLKxYsQIA4PF48Pzzz8Pj8WDevHm46KKL8LWvfQ233nqr/h6TJ0/G3//+d7z88ss48sgjcccdd+B///d/ccYZZ+Trr6UzZ/I1TdN3psePSmwlkOP6XGfyRZAf0YDlH0ez+FMbKvXpwInXFr+4XA8FJCIicjs5w64oit6Tby7XF/34JT6P5Qa/mVj/I5qGnoEQ9sUy1RNqBxfkp7t2i+n6PYGwnq0XPflAvNd+2/5YJt/uCL1KUa7vPMhfHRu6d1TTKBR5VZwVy7w/827ilH1Rrl8bC/Lj5fqJPfliY+KYSTVo1KsS7DP57MknIjfJa09+Oh566CE89NBDSV8zceJEvPDCC0lfc8opp+Ddd9/N4pVlhxynhyMa9nUH0BcMQ1GAcVZB/hAO3pPf/+0t+wEAx0yyLtUHTIP3GOMTEdEIZ+6Vtxu8dyCNUn0gXtUXjmhoPtCrf2+lzSZ8KuLeIt1jeSv8XhR5VQRCET1ItgryewOx6fopBu91D4TQGwjZ9u7LVseG7h3VVAUAOGfWODy6bDteWteCnwVnGjZLEjP5idP19cF7sWtRVQWLPjsDD7+1DadNN57QBAD+2Pszk09EbsIQzCUURdED/Yim6TcEjVUlluVh6hAeoeeT3n/pR9Egf7ZNPz5grEpIt6+PiIhouDGX0Yty/d5A2NA7LjLKTkr1gXgmPxzRDEf0ZSrek5/e9ymKopfaCyIrH/1zseG5Mpsj9Mr9XpTEgmYnE/Y1TcOa5nYA0Uw+AMyeMArjqkvQPRDCkvV7DK9v64m+p8jkixMMOuRMfqzdQL7mT88ciz9fcZzeliCLZ/I5eI+I3INBvot45B15/YbAeur/UA7e80mr/Ud7o6V2cyZZT9YHjH34nK5PREQjnTkALy+OB7ndUsl+fLK+w0y+IgX5g+zHB6QgP4O1W5TsA9FMuZxBN28A2GXyFUWRJuynDvI/3t+Ljr4girwqpjZUAIgmQcTRd8+s3ml4fZtNT/4Babr+vu4By2u2449N1w+E7TP5f3p7G97anDgjgIgoVxjku4i8WG9LsSNvyJbnOJA2v//YqmKMq7befACMmw7plvwRERENN+Zyfb/Xox+91jUQzyJ3iHL9EmcnC4l2Ojk5kI1MfiZDc+Usd70pcy9n9QH7nnwgXia/10GQL47Om9lYqX+eAPD5o8YBAF7bsNeQYd8vpuvHrrXaNF0/FI7oJf3ma7ajB/k25fqbWrvwo2fW4juPvwtNS5z4T0SUCwzyXUQeoJPqrFvjdP3cBtKKohiy+clK9QHTEXos1yciohEsOkg3cU2v8Cf25Ytgc1RZmpl8B/cNTgwmkz9ayuQ3mAJkkZ0X7DL5gDx8rz/lz1zTHO3HP7Kp2vD4ofXlqCkrQiAcwfrdXfrj5sF7IpPfEaug2NcdgKZF78dqSp1ttIjNBbvBe+JEgX3dAbQ6aEEgIsoGBvkuolqU60+oLbN8rbFcP/f/jPLPmJMiyPewXJ+IiAhANHDvigXycim91YR9Ua5f5TSTL47QC2c3yM+kCk/O5DdUGav9xHF0QqlNT778Wifl+u/q/fjVhscVRcGR46OD+ETPPhAvy68x9eS3xx4XGwujy/2OP4OiFJn8nbGBiADwYUuno/ckIhosBvkuIvfkp5XJH4JAWs7kJ+vHBwCP9NpMSv6IiIiGC7Ge11f6DX3qoi/f0JPfm15PvrhvCEUiaD4QPXZ3UEG+ohh+T0etHOSnLNe3z+TrPfkpst4DoTDW74oGzeYgP/pYNCGxOhbkRyIaDsQ+31rTdP2eQBiBUERvETBXHiQjhiPbBfm7OuIVCR+2dFm+hogo2xjku4hYVHsDYbTEprva9uQP4RF6AOCLTY8tLfJgWmy4jR1m8omIiKLsNu3FhP3O/nhPfrvek59ekL+3awCBUAQeVcHYquIU32VP3E8Muly/yhgkV/i9eu86kDyTLwLsvd3Jg/z1u7sQCEcwqtRnea90ZJMxk9/RF9RPMhBl+hXFPohblo6+YMLxeU6kKtff2d6n/3kDg3wiGiKpDyClISOy89tiNwTlfi9G2ezmy2fRD0m5fmzhP3rCKHg9yX+ecShgTi+LiIjI1c6c2YDXr/sk+k1HrFXEzrI39OSnOV3fo8/yiX49rrok5RqdjKi+y6QKb0ySwXuKoqCu0o/mtmjAmyyTX6dn8pP35Ivg/cimaigW1yuy+1v29aCjN6gP1Kso9uqBuUdVUFXiQ3tvEO29Ab16wOnQPcBJuX48yF+/m+X6RDQ0GIK5iMiAf7wvekzdhJpSy4ULMA23G5Jy/eh/KqmG7gHG62G5PhERjWRej4oJtaU4tN5YBacP3jOU60cDUac9+eb1fzCl+kC8/S+TNkC5XH9sVeIJPHJffmnS6frR16Wari/K8I8cX235fHVpESbVRj+P1TvaE4bu6a8Tffl9Qb0nf0yF82oI+Qg9q+n5u6RM/kd7uxFMctQeEVG2MMh3EbFYb5WCfNvXSsGzbwjK9UVZ4dzJyfvxAZbrExERpWI1eE9MeXc6Xd/cO980yCBfHcTgPbmP3dyTD8Qz9D6PYjjuLuF1sSz6/p5A0oB4zY52ANb9+IJ4bk1zO9p6opsGNaYgvypWut/eGy/XT6cnX/67mEv2B0Jh/T19HgXBsKbf4xER5RKDfBcRVfd6kF+bJMgf4kz+rZ+fiR+dNR3zDq5N+VqVmXwiIqKk9MF7sXJ9TdPig/fynMnPpOK/pqwI3zhxMr75iYNQZdFuIALnEp99qT4A1JQW6X8vca69WV8grN8rzRxXZfte4mi91c3terl+TZkxgJcn7O/NoCdfnjUQMG1KtMSG7hX7VBweu06W7BPRUGBPvouI3vqP90cXrmQ78qph8F7u92qOnVyDYx1k8QFjmR8z+URERInK/dHgUmTyewJhhPTBcOn15AsTkyQH0nm/TKbrA8CPzp5h+5wInMuSDN0DoomC0eVFaO0cwJ6ufjRYDBLc2NoFTYuW3ifLusuZ/FmxPyeU68c+646+YEZBfpF0DzYQjADS5Yqhe43VJZg2thKrtrdz+B4RDQlm8l1ErNViJz9puf4QH6GXjqGuMiAiIio08XL96Jp/IJZp9ntVw1F7yZjX/8Fm8j2DKNdPRfTalyYZumd+rd0xeiJQnpritJ/pYyvh8yjY3xPAmh0dAICacuue/ANyJt+i3cCOoih6oG/O5O9qj2byx1WX6CcT8Rg9IhoKDPJdJGFHPlkm3xDku+ufUWVPPhERUVIVpnL9jjQn6wOJwfhge/IHm8lPZvKYMgCJk/et6BP2bYbvrW+JlrxPa6hM+j7FPg9mjI2+5j+b9wGwyuRHv/54f68epI8ud9YuIZT6oxsXHb1Bw+Ni6F40yI9eBzP5RDQU3BUdjnBycKwq0fIuO/IC7LZMvlcaBJiLGwUiIqJCJwbaiiBfVPGNKnUeYMrrf1WJD1UlzjcIrIg1Oxcb9HMmjsJvLzwavzjviJSvFcP3xLR7MxEoT0uRyQfiffl9wegRhubBe2JTZVNrl/613+uskkI4ZEx59D32GAN4cXxeY3UJpsZOV9jZ3qdv6BAR5QqDfBeRF9WxVSVJp8/KbfieIZiunw7DZoXLNiCIiIjcoKLY2JPf3ieOz8sskz/YUn0gvkmfi7VbURR85vCxjqoNxsSO47M6Rk/TNL3kfdrY1EG+efq+XZAvBvml048viLaB9buNQf6ujniQX1XqQ2NsvsDGVmbziSi3GOS7iCeNxVoOpH0uC6TZk09ERJScyOTrQX5v+uX63iwH+eLeIt9VeGNiJf1W5fp7uwfQ1hOAogBT6pxn8oXahOn60aA/GI4OPRTzANIhKgo2tBgn58cH70XfU2wGfMgJ+0SUYwzyXUQO3FNNyDUG0u76ZzRcG8v1iYiIEsR78qPBvd6T7/D4PMB43zDYfnwAmD1xFIo8atKz54dCsp58Uao/ubYMJQ6G+E2uLUNlcXyiv3nwnvm4v0wy+dPGJvbba5qm9+SPry41vI7D94go13iEnovIwXGqxVp+rc9l5foeQ7l+Hi+EiIjIpUSQ3x+MIBiO6NP185nJP/fo8Tj7iMak7YJDQQTaezsTe/I/3O1ssr6gqgqObKrGG5usB++ZZyAkO5LPzqGxfvtdHf3o6A2iqtSHtp4A+oMRKApQXxV9z3jGn0E+EeUWQzAXybRc320l8czkExERJSefF9/dH0J7LJNvziwnk859g1P5DvCBeKC9t3sAmqYZnvvQ4fF5MlGZUFrkSTiesNo0AyGTIL+qxIdxsWHJG2L99uL4vDHlfn2QnzxhX/y9whENq7YfQCAUMb9twduytxvb9/fm+zKIRqT8/z856eSAONViLS/sXo+7/hlV9uQTEREl5fOoKPZF1+/ugVBG0/XlM9pTtfkVEhFoB8Oa/rkIG1qdHZ8nE0H+6PLEAL7SFOTXOTjiz4rebx/ry9/ZHg1ux42Kn5R00Jgy+DwKugZC2Nneh0Aogu/8+V2c+9u3cPVjqzL6uW61v3sAn/3Nmzjr7jdsT0kgotxxV3Q4wskBcarFWnXzEXqx61GU6A0IERERJRIT9jv7g+iITdc3Z5ZTuf7TU7HgkwdnpSffLfxej962IPflh8IRbGztBuDs+Dzh5EPH4IJjmvC90w9NeM6jKoae/Ux68uXrEZUGO2OZfPk4ZJ9HxcGx4/ZWN7fjij+uwN/f3w0A+OcHrXjn47aMfrYb/f393egJhNE1EML/vLI535dDNOIwyHcRkQGvKPamPELHzRPs3TKdl4iIyM0qYiX73f3xTH465foA8I2TDsJ1Z0zL+rXlW3z4XjwL/PH+XgRCEZT4PGm1J/g8Km477wh8/qhxls9XS9UTmQb55sn5YujeOCnIB+KbAdc9+R7+vWEvin0q5k6uAQD84h8fJrQnFKpn3t2p//mxZdtZtk80xBjku4iYnzehpjRlBlyu0Hfd4D01d+fsEhERDRfl+oT9eE9+OtP1hzNxlN2ezngmX5TCH9pQkdV7DHnYYSY9+QAwPTY5f2Nrt2GyfkKQH3tdXzCMCr8Xf7psLn59wSz4vSpWbDuAVz7ck/TnvLlpH3787Dp0D4Qyus6hsH1/L1Ztb4eqRFslQhENd768wfCa/d0DuPlva/Hi2pY8XSXR8MYg30VEcOxkd9o4eM9d/4yiXJ+ZfCIiIntiwn5XfwjtvelP1x/OmmqiwbEc9Iqp9NPTKNV3QmTyS3welPszO3hq8uhov333QAg7DvRhZyzIbzQF+UdPGAUgOuX/z1cchzmTatBQVYxLTpgEALj9xQ0IR6yz+ZGIhuv/bw0eeutjV5fA/211NIt//MGj8dNzZkYfW7MLH+yKbtLs7ujDF3+/FA8v3YYr/7QSD/1na96ulWi4cld0OMKJwN1JkO/mI/TE7rrb2giIiIjcRASUrZ39CIajgR2D/KivzZsERYn2dr+/owNAZpP1nRBzEOoq/RnPEpL77T9s6dIz+Y3VxkF+x06uwR8vOxYvfPckzBxXpT/+rU8cgspiLza0dulBstk7H7dhV0e0feHB/2xFS4f7BtppmoZnYtf/+aMaMXNcFc4+Yiw0DfjlSx9i674e/NfvluKjvT0oK4qeOvDj5z7Ab5ZsGjatCkRuwCDfRUSJ2IzG1BNjPW4+Qi92bS67LCIiIlcp90eDy+YD0X7lIq+KEtMRbyPV9LGVOCfWQ3/7Sx8CiJfrZz3Ij22sZNqPL4iS/TXN7djXHa3MMJfrA8BJU8ag3jTFv6rUhytPORgAcOfLGzEQCid83zOrd+l/HghF8OslmwZ1vZnoD4bRMxDSf4XCxqP/1u3qxEd7e+D3qvj0zAYAwPdPnwqvquDVDXtxzj3/wc72PkweXYaXrj0Z186PDkO84+WN+PkL6xnoE2UJg3wXuf7T0/DAJXNw1uFjU75W7kVz23R9DzP5REREKYly/R0Holnf6hIfT6WRLPzUofB5FLyxaR/+ua4FzW3Rzymd4/OcEJn8TPvxBbH58OqGaItBWZEn5SBl2aXHT0ZdhR87DvThsWXbDc8FQhG8EJvELwLjJ1Y0Y8ve7kFds1PBcATff3INpi96EYfd/JL+67jFS/DW5n3660QVwvzp9frpEZNGl+FLxzQBADr6gpg+thJPfHMexo8qxXfnT8Gis2cAAO57Yyvu+tfQb1wQDUcM8l2kqsSHU6fVOzr33mMI8t31zxgP8t11XURERG6SEOSzVN+gqaYUF86dCAC44an3AUSz7TVl2R1OeMIho1FZ7MWp0+oH9T5icv66WO95Y3VJWps2JUUefHf+FADA/7yy2TBc798b9qCjL4i6Cj+uPvUQnDatDuGIhjv+uXFQ1+xEfzCMb/5xJf5v5Q6YE+37ugO45KF38PIHrQhHNDy7Jlpt8LmjGg2v++5pU3BIXTk+cegYPH75cYYNla+fOBk//8LhAIB7X/vIlW0IRIWGUViBkgfveV3Wky/2KBzsVRAREY1Y8SA/Wq7PyfqJFnzyEJQWedDWEy1/F9Pps2nuQbVYc/Pp+K/Z4wf1PuYKA/PQPSe+OKcJk0eXYX9PAPe/ER9I97dYqf7njmyER1Vw3aen6jML3tvRPqjrTqarP4iLH1iOVz7cg2KfigcumYMPf/JpfPiTT+P9H5+OMw6rRyAUwZV/WolbnluH1s4BVBZ7ccrUMYb3qassxsvXnoyHv36s5TGRXz62CcdOqom1IeR+44JouGMYVqDcncmPXg+n6xMREdkTPfn9wWhfs1XwM9KNqfDjGydO1r+eluV+fCEbbRL1lX5DNca4UekH+T6Piu+dHi3Hv++NLdjfPYCu/iD+tb4VAHDOrOicgmkNlfhCbGbBzc+uw6PLtum/Vje3W763pml4Y9NeNLdZn1nfHwzj2TW7DO914f8uw7KtbSj3e/HI1+fi1Gn1KPZ5UOzzoKLYh3u+cjTOnTUO4YiGR5ZuAwCcdcRY+L2JsyWSfcaKouD6T08FADyxYgc+GmQbwrpdHVhj8zmkYvU5PLtmF/qDiXMSiNwqs3NCKO/cPHjv0PpyTGuowAmHjM73pRAREblWebHxNqw6jf7tkeTykw/CH9/ehgO9QUwfm5sgPxsURcHU+gos29oGwHronhOfmTkWM8d9hLU7O/Hbf3+EaQ0VGAhFcPCYMhwmDWe+9lOH4rn3duHd7e14d3u74T1u/fxh+Nq8SfrXkYiGW55bh4eXbkOF34v7LzkGx06u0Z/v7A/iGw+twPKP2xKup6asCA9feiwOH1+V8JzXo+K/zz8SFcVePBwL8j8f23xI15xJNZg/vQ7/Wr8Hd/5zI+658OiM3uevK3fg+r++h4im4dbPHYavSp9DKh19QVz20DtYse1AwnPHTqrB/ZfM0WcNELkZg/wCJSfv3XaEXmmRFy9ec3K+L4OIiMjVKkxB/qgs95oPFxXFPvz2wtn41/pWfMbBcOJ8mtYQD/LNx+c5paoKrj9jGr72wHL8cek2HFIXPZrvnKPGGbLhTTWl+OV/HYl/rN2tP9bRF8TbW9qw6G/r0NUfwrdOORjhiIbr//oenloVHYrXNRDC1x5Yhnsvmo1TptahrSeArz2wDGt3dqKi2IvjD67V36+i2IcrP3Gwfg121/vjzx2Gg+vKcaAniLnS5kG6vn/GVCz5cA/+/v5ufHNHO44YX53W9z/81se4+dl1+tc3/W0dugZC+NYph6T83n3dA/ja/cvxwe7Ez+Gtzfux/OM2fOW+ZXj468dmfS4EUbYxyC9QcvbebZl8IiIiSq3Cb7wNS2cS+0gz7+BazJOCLreSZwaMqy7N+H1OmjIa8w6qxdIt+/HB7uggP/MwOyBavi9K+IFoSf6v/rUJdy/ZhF++tAEdfUFs29+Dl9a1wqMq+PkXZuKlda145cM9uPyRFVh09gw8vHQbNu/pRm1ZER7++rGYOS4xY5+KoiiGyoFMTWuoxBdmjcNTq3bi9hc34E/fmOvo+zRNwz2vbsZ/xwYRXnrCJJT7vfjNK5tx+4sb0NkXwg8+PdW2ZWBXex8uun8ZtuztwejyIjzy9bmGI63X7uzA1x5Yjvd3duCLv1+KP102Fw1VmW3iEA0FBvkFSi7Xd1tPPhEREaWWUK7PnvyCN1WaGZBpJh+IBs0/OHMazrnnPwCAWROqMbG2zNH3LfzUoags9uKnf1+PP7y+BQBQ5FVxz1eOxqdm1OMLs8Zj4ROr8fx7u3HT36JZ77FVxfjTN+bi4DH2Gfuhcu38Q/Hcml14c/M+nPvb/zi6z+0PhfHejg4A0Un+18yfAkVRUFHsxc9f+BD3vvYR3ti0F2VF1qHPln092Nc9gMbY53CQ6XOYOa4KT3xzHr56/zJs3tONs3/zJg4anfrfg0j43UVHo7Z8cMd0poNBfoFSVfdO1yciIqLUzL29nK5f+KY1VKDC74XXo6C+cnCZ3qOaqnHWEWPx9/d244LYOfNOfeOkg1BR7MWNT72PYp8H931tjj4rqcir4tcXzEJFsRd/Xt6MyaPL8MfLjsX4UZlXHmRTU00pvjZvEu5/cytWmWYNpPKjs6bjGycdpH99xckHo6LYh//39Pv60YZ2Dhpdhj9+Y67tLIVD6srx5JXzcNH/LsPH+3uxr3sgrWujkS0Y1lK/KIsY5BcoYyafQT4REVGhKfczkz/clBZ58ey3T4RHUeDLwlnCd5x/JC45fhLmTByV9vd+6ZgJmD1xFMr8XoytMgau0dL9w/HFOU2YUl+R8N9ivv3g09NwwiG1+skTTkyqLTOU2AtfPnYC5kwchU177Cf2e1UFxx8yOuXnMH5UKZ7/zkl4a/M+hCJDG7RRYRvq/3931/+iyTHDEXo8kJ6IiKjgMMgfniZnsYy72OfBMZMyH2R3SJ39aQSKomDWhPQ3D4ZCkVfFqdPqs/Z+U+orMKU+OyczlPu9OP2whqy8F1GuMDosUIZyfWbyiYiICo5HVVBWFD9PvLqU5fpERDR4DPILlKFcnz35REREBUkevlfN6fpERJQFDPILlDxolJl8IiKiwiRK9n0eBaVSVp+IiChTDPILlJzJ9/AIPSIiooIkJuxXlRTZnuFNRESUDkaHBcrDnnwiIqKCVxEr1+fQPSIiyhYG+QXKMHiPPflEREQFSZTrj2KQT0REWcIgv0AZBu+xXJ+IiKggiUx+VQkn6xMRUXYwOixQLNcnIiIqfOX+aAaf5fpERJQt3tQvITcq8qior/QjGNZQ5uc/IxERUSE6pK4cAHBofXmer4SIiIYLRocFSlUVvPCdkxDWNBR5WZBBRERUiL58bBPmTBqFg8cwyCciouxgkF/Aasv9+b4EIiIiGgRFUXBofUW+L4OIiIYRpoCJiIiIiIiIhgkG+URERERERETDBIN8IiIiIiIiomGCQT4RERERERHRMMEgn4iIiIiIiGiYYJBPRERERERENEwwyCciIiIiIiIaJhjkExEREREREQ0TDPKJiIhoWLrnnnswadIkFBcXY+7cuVi+fHm+L4mIiCjnGOQTERHRsPOXv/wFCxcuxM0334xVq1bhyCOPxBlnnIE9e/bk+9KIiIhyikE+ERERDTt33nknLr/8clx66aWYMWMG7r33XpSWluKBBx7I96URERHlFIN8IiIiGlYCgQBWrlyJ+fPn64+pqor58+dj6dKllt8zMDCAzs5Owy8iIqJCxCCfiIiIhpV9+/YhHA6jvr7e8Hh9fT1aWlosv2fx4sWoqqrSfzU1NQ3FpRIREWUdg3wiIiIa8W688UZ0dHTov5qbm/N9SURERBnx5vsCCo2maQDAMj4iInINsSaJNWqkGz16NDweD1pbWw2Pt7a2oqGhwfJ7/H4//H6//jXXeyIicpN01noG+Wnq6uoCAJbxERGR63R1daGqqirfl5F3RUVFmD17NpYsWYJzzjkHABCJRLBkyRJcffXVjt6D6z0REbmRk7WeQX6aGhsb0dzcjIqKCiiKMqj36uzsRFNTE5qbm1FZWZmlKxz++Lllhp9b+viZZYafW2YG87lpmoauri40Njbm6OoKz8KFC3HxxRdjzpw5OPbYY3HXXXehp6cHl156qaPv53qfX/zMMsPPLTP83NLHzywzQ7XWM8hPk6qqGD9+fFbfs7Kykv/jyAA/t8zwc0sfP7PM8HPLTKafGzP4Rl/60pewd+9eLFq0CC0tLTjqqKPw4osvJgzjs8P13h34mWWGn1tm+Lmlj59ZZnK91jPIJyIiomHp6quvdlyeT0RENFxwuj4RERERERHRMMEgP4/8fj9uvvlmwzRfSo2fW2b4uaWPn1lm+Lllhp/b8MV/2/TxM8sMP7fM8HNLHz+zzAzV56ZoPG+HiIiIiIiIaFhgJp+IiIiIiIhomGCQT0RERERERDRMMMgnIiIiIiIiGiYY5BMRERERERENEwzy8+iee+7BpEmTUFxcjLlz52L58uX5viTXWLx4MY455hhUVFSgrq4O55xzDjZs2GB4TX9/PxYsWIDa2lqUl5fjvPPOQ2tra56u2J1uu+02KIqCa665Rn+Mn5u1nTt34qKLLkJtbS1KSkpw+OGHY8WKFfrzmqZh0aJFGDt2LEpKSjB//nxs2rQpj1ecX+FwGDfddBMmT56MkpISHHzwwfjJT34CeZYrPzPg9ddfx2c/+1k0NjZCURQ888wzhuedfEZtbW248MILUVlZierqalx22WXo7u4ewr8FDQbX+uS43g8e13rnuNanh2u9M65c6zXKi8cff1wrKirSHnjgAW3dunXa5ZdfrlVXV2utra35vjRXOOOMM7QHH3xQW7t2rbZ69WrtM5/5jDZhwgStu7tbf82VV16pNTU1aUuWLNFWrFihHXfccdrxxx+fx6t2l+XLl2uTJk3SjjjiCO273/2u/jg/t0RtbW3axIkTtUsuuURbtmyZtmXLFu2ll17SNm/erL/mtttu06qqqrRnnnlGW7Nmjfa5z31Omzx5stbX15fHK8+fn/3sZ1ptba32/PPPa1u3btWefPJJrby8XPv1r3+tv4afmaa98MIL2g9/+EPtqaee0gBoTz/9tOF5J5/Rpz/9ae3II4/U3n77be2NN97QDjnkEO3LX/7yEP9NKBNc61Pjej84XOud41qfPq71zrhxrWeQnyfHHnustmDBAv3rcDisNTY2aosXL87jVbnXnj17NADaa6+9pmmaprW3t2s+n0978skn9desX79eA6AtXbo0X5fpGl1dXdqUKVO0l19+WfvEJz6hL/z83Kz94Ac/0E488UTb5yORiNbQ0KD98pe/1B9rb2/X/H6/9uc//3koLtF1zjrrLO3rX/+64bFzzz1Xu/DCCzVN42dmxbzwO/mMPvjgAw2A9s477+iv+cc//qEpiqLt3LlzyK6dMsO1Pn1c753jWp8ervXp41qfPres9SzXz4NAIICVK1di/vz5+mOqqmL+/PlYunRpHq/MvTo6OgAANTU1AICVK1ciGAwaPsNp06ZhwoQJ/AwBLFiwAGeddZbh8wH4udl59tlnMWfOHJx//vmoq6vDrFmzcN999+nPb926FS0tLYbPraqqCnPnzh2xn9vxxx+PJUuWYOPGjQCANWvW4M0338SZZ54JgJ+ZE04+o6VLl6K6uhpz5szRXzN//nyoqoply5YN+TWTc1zrM8P13jmu9enhWp8+rvWDl6+13ju4y6ZM7Nu3D+FwGPX19YbH6+vr8eGHH+bpqtwrEongmmuuwQknnICZM2cCAFpaWlBUVITq6mrDa+vr69HS0pKHq3SPxx9/HKtWrcI777yT8Bw/N2tbtmzB7373OyxcuBD/7//9P7zzzjv4zne+g6KiIlx88cX6Z2P1v9mR+rndcMMN6OzsxLRp0+DxeBAOh/Gzn/0MF154IQDwM3PAyWfU0tKCuro6w/Nerxc1NTX8HF2Oa336uN47x7U+fVzr08e1fvDytdYzyCfXW7BgAdauXYs333wz35fies3Nzfjud7+Ll19+GcXFxfm+nIIRiUQwZ84c/PznPwcAzJo1C2vXrsW9996Liy++OM9X505PPPEEHn30UTz22GM47LDDsHr1alxzzTVobGzkZ0ZEGeF67wzX+sxwrU8f1/rCxXL9PBg9ejQ8Hk/ClNPW1lY0NDTk6arc6eqrr8bzzz+PV199FePHj9cfb2hoQCAQQHt7u+H1I/0zXLlyJfbs2YOjjz4aXq8XXq8Xr732Gu6++254vV7U19fzc7MwduxYzJgxw/DY9OnTsX37dgDQPxv+bzbuuuuuww033IALLrgAhx9+OL761a/i2muvxeLFiwHwM3PCyWfU0NCAPXv2GJ4PhUJoa2vj5+hyXOvTw/XeOa71meFanz6u9YOXr7WeQX4eFBUVYfbs2ViyZIn+WCQSwZIlSzBv3rw8Xpl7aJqGq6++Gk8//TReeeUVTJ482fD87Nmz4fP5DJ/hhg0bsH379hH9GZ522ml4//33sXr1av3XnDlzcOGFF+p/5ueW6IQTTkg4smnjxo2YOHEiAGDy5MloaGgwfG6dnZ1YtmzZiP3cent7oarGJcTj8SASiQDgZ+aEk89o3rx5aG9vx8qVK/XXvPLKK4hEIpg7d+6QXzM5x7XeGa736eNanxmu9enjWj94eVvrMxrXR4P2+OOPa36/X3vooYe0Dz74QLviiiu06upqraWlJd+X5gpXXXWVVlVVpf373//Wdu/erf/q7f3/7d1ZSFVrGMbxZ5u429syTcWkFIoksyBsxIaLksouIsWbwmJnQTQiVggVTYQYURJ0YQhlRFFQFNFMUQQZaYMNlElR2o1CNFg22OB7Lg5ncXbZ0ROJtvz/YIGu9a1vva4Ln/2y1/DBGbN48WJLTEy0S5cu2c2bNy0tLc3S0tI6sequ6d9P3DXjvLWmsrLSQkNDrbCw0B4/fmwHDx40v99vBw4ccMZs3brVIiMj7cSJE3bv3j2bNWtWt3tFzL8FAgHr37+/81qdY8eOWUxMjBUUFDhjOGd/P/26qqrKqqqqTJIVFxdbVVWV1dXVmVn7zlFGRoalpqZaRUWFXb161ZKSkniF3h+CrG8bef97kPVtI+v/P7K+fbpi1tPkd6Jdu3ZZYmKihYWF2dixY+369eudXVKXIanVpayszBnz8eNHW7p0qUVFRZnf77esrCyrr6/vvKK7qO+Dn/PWupMnT9rw4cPN6/VacnKylZaWBm1vaWmx9evXW1xcnHm9XktPT7eamppOqrbzvX371vLy8iwxMdF69uxpgwYNsnXr1llzc7MzhnNmdvny5Vb/lwUCATNr3zl6+fKlzZkzx3r16mURERGWm5tr796964S/Br+CrP9v5P3vQda3D1n//5D17dMVs95jZvZr1wAAAAAAAICuhHvyAQAAAABwCZp8AAAAAABcgiYfAAAAAACXoMkHAAAAAMAlaPIBAAAAAHAJmnwAAAAAAFyCJh8AAAAAAJegyQfwW9TW1srj8ejOnTsddoz58+crMzOzw+YHAAA/R9YDfwaafACS/g5Vj8fzw5KRkdGu/RMSElRfX6/hw4d3cKUAAOBXkPVA9xDa2QUA6DoyMjJUVlYWtM7r9bZr3x49eqhfv34dURYAAPhNyHrA/fgmH4DD6/WqX79+QUtUVJQkyePxqKSkRDNmzJDP59OgQYN09OhRZ9/vL+F7/fq1cnJyFBsbK5/Pp6SkpKAPFffv39eUKVPk8/kUHR2tRYsWqampydn+7ds3rVy5UpGRkYqOjlZBQYHMLKjelpYWFRUVaeDAgfL5fBoxYkRQTW3VAABAd0PWA+5Hkw+g3davX6/s7GzdvXtXOTk5mj17tqqrq3869uHDhzp79qyqq6tVUlKimJgYSdL79+81ffp0RUVF6caNGzpy5IguXryo5cuXO/vv2LFD+/bt0969e3X16lW9evVKx48fDzpGUVGR9u/fr927d+vBgwfKz8/X3LlzdeXKlTZrAAAAPyLrARcwADCzQCBgPXr0sPDw8KClsLDQzMwk2eLFi4P2GTdunC1ZssTMzJ49e2aSrKqqyszMZs6cabm5ua0eq7S01KKioqypqclZd/r0aQsJCbGGhgYzM4uPj7dt27Y52798+WIDBgywWbNmmZnZp0+fzO/327Vr14LmXrhwoc2ZM6fNGgAA6G7IeqB74J58AI7JkyerpKQkaF3fvn2dn9PS0oK2paWl/fQJu0uWLFF2drZu376tadOmKTMzU+PHj5ckVVdXa8SIEQoPD3fGT5gwQS0tLaqpqVHPnj1VX1+vcePGOdtDQ0M1evRo5zK+J0+e6MOHD5o6dWrQcT9//qzU1NQ2awAAoDsi6wH3o8kH4AgPD9fgwYN/y1wzZsxQXV2dzpw5owsXLig9PV3Lli3T9u3bf8v8/9zTd/r0afXv3z9o2z8PEOroGgAA+NOQ9YD7cU8+gHa7fv36D78PHTr0p+NjY2MVCAR04MAB7dy5U6WlpZKkoUOH6u7du3r//r0ztry8XCEhIRoyZIj69Omj+Ph4VVRUONu/fv2qW7duOb+npKTI6/Xq+fPnGjx4cNCSkJDQZg0AAOBHZD3w5+ObfACO5uZmNTQ0BK0LDQ11HmBz5MgRjR49WhMnTtTBgwdVWVmpPXv2tDrXhg0bNGrUKA0bNkzNzc06deqU8yEhJydHGzduVCAQ0KZNm/TixQutWLFC8+bNU1xcnCQpLy9PW7duVVJSkpKTk1VcXKw3b9448/fu3VurV69Wfn6+WlpaNHHiRDU2Nqq8vFwREREKBAL/WQMAAN0RWQ+4H00+AMe5c+cUHx8ftG7IkCF69OiRJGnz5s06fPiwli5dqvj4eB06dEgpKSmtzhUWFqY1a9aotrZWPp9PkyZN0uHDhyVJfr9f58+fV15ensaMGSO/36/s7GwVFxc7+69atUr19fUKBAIKCQnRggULlJWVpcbGRmfMli1bFBsbq6KiIj19+lSRkZEaOXKk1q5d22YNAAB0R2Q94H4es+9eRgkArfB4PDp+/LgyMzM7uxQAANAByHrAHbgnHwAAAAAAl6DJBwAAAADAJbhcHwAAAAAAl+CbfAAAAAAAXIImHwAAAAAAl6DJBwAAAADAJWjyAQAAAABwCZp8AAAAAABcgiYfAAAAAACXoMkHAAAAAMAlaPIBAAAAAHAJmnwAAAAAAFziL9wVHNc26cVeAAAAAElFTkSuQmCC",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"#action_selection = {'type': \"egreedy\", \"param\": 0.1, \"decay\": 0.0}\n",
"action_selection = {'type': \"softmax\", \"param\": 1.0, \"decay\": 0.0}\n",
"alpha = 0.1\n",
"nb_episodes = 100\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(render_mode=None)\n",
"\n",
"# Create the agent\n",
"agent = SoftQLearningAgent(env, gamma, action_selection, alpha)\n",
"\n",
"# Train the agent \n",
"returns, steps = agent.train(nb_episodes)\n",
"\n",
"plt.figure(figsize=(12, 5))\n",
"plt.subplot(121)\n",
"plt.plot(returns)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Return\")\n",
"plt.subplot(122)\n",
"plt.plot(steps)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Steps\")\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**A:** The agent explores much less at the end of training, as the difference between the Q-values becomes high enough to become greedy. In particular, it quickly stops to go to the red squares. In this environment, there is no real need to decay tau."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Eligibility traces\n",
"\n",
"The main drawback of Q-learning is that it needs many episodes to converge (**sample complexity**).\n",
"\n",
"One way to speed up learning is to use eligibility traces, one per state-action pair:\n",
"\n",
"```python\n",
"traces = np.zeros((nb_states, nb_actions))\n",
"```\n",
"\n",
"After each transition $(s_t, a_t)$, Q($\\lambda$) updates a **trace** $e(s_t, a_t)$ and modifies all Q-values as:\n",
"\n",
"1. The trace of the last transition is incremented from 1:\n",
" \n",
"$$e(s_t, a_t) = e(s_t, a_t) +1$$\n",
" \n",
"2. Q($\\lambda$)-learning is applied on **ALL** Q-values, using the TD error at time $t$:\n",
" \n",
"$$Q(s, a) = Q(s, a) + \\alpha \\, (r_{t+1} + \\gamma \\, \\max_{a'} Q(s_{t+1}, a') - Q(s_t, a_t)) \\, e(s, a)$$\n",
" \n",
"3. All traces are exponentially decreased using the trace parameter $\\lambda$ (e.g. 0.7):\n",
"\n",
"$$\n",
"e(s, a) = \\lambda \\, \\gamma \\, e(s, a)\n",
"$$\n",
"\n",
"All traces are reset to 0 at the beginning of an episode.\n",
"\n",
"**Q:** Implement eligibility traces in your Q($\\lambda$)-learning agent and see if it improves convergence. Train it with rendering on and observe how all Q-values are updated."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"class QLambdaLearningAgent(SoftQLearningAgent):\n",
" \"\"\"\n",
" Q(lambda)-learning agent with softmax or e-greedy AS and eligibility traces.\n",
" \"\"\"\n",
" \n",
" def __init__(self, env, gamma, lbda, action_selection, alpha):\n",
" \"\"\"\n",
" :param env: gym-like environment\n",
" :param gamma: discount factor\n",
" :param lbda: trace\n",
" :param action selection: exploration mechanism\n",
" :param alpha: learning rate\n",
" \"\"\"\n",
" self.lbda = lbda\n",
" \n",
" # Traces\n",
" self.traces = np.zeros([env.observation_space.n, env.action_space.n])\n",
" \n",
" super().__init__(env, gamma, action_selection, alpha)\n",
" \n",
"\n",
" \n",
" def update(self, state, action, reward, next_state, done):\n",
" \n",
" # Bellman target\n",
" target = reward\n",
" if not done:\n",
" target += self.gamma * self.Q[next_state, :].max()\n",
" \n",
" # Update ALL Q-values\n",
" self.Q += self.alpha * (target - self.Q[state, action]) * self.traces\n",
" \n",
" # Decay exploration parameter\n",
" self.exploration = self.exploration * (1 - self.decay)\n",
" \n",
" \n",
" def train(self, nb_episodes, recorder=None):\n",
" \"Runs the agent on the environment for nb_episodes.\"\n",
"\n",
" # Returns\n",
" returns = []\n",
" steps = []\n",
"\n",
" # Fixed number of episodes\n",
" for episode in range(nb_episodes):\n",
"\n",
" # Reset\n",
" state, info = self.env.reset()\n",
" done = False\n",
" nb_steps = 0\n",
" \n",
" # Reset traces\n",
" self.traces *= 0.0\n",
"\n",
" # Store rewards\n",
" return_episode = 0.0\n",
"\n",
" # Sample the episode\n",
" while not done:\n",
"\n",
" # Select an action \n",
" action = self.act(state)\n",
"\n",
" # Perform the action\n",
" next_state, reward, terminal, truncated, info = self.env.step(action)\n",
" \n",
" # Terminal state\n",
" done = terminal or truncated\n",
" \n",
" # Update return \n",
" return_episode += reward\n",
" \n",
" # Increment trace\n",
" self.traces[state, action] += 1\n",
"\n",
" # Learn from the transition\n",
" self.update(state, action, reward, next_state, done)\n",
" \n",
" # Update all traces\n",
" self.traces *= self.gamma * self.lbda\n",
"\n",
" # Go in the next state\n",
" state = next_state\n",
"\n",
" # Increment time\n",
" nb_steps += 1\n",
" \n",
" \n",
" # Pass the Q table to the GUI\n",
" self.env.Q = self.Q\n",
"\n",
" # Store info\n",
" returns.append(return_episode)\n",
" steps.append(nb_steps)\n",
"\n",
" if recorder is not None:\n",
" recorder.record(self.env.render())\n",
"\n",
" return returns, steps"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"lbda = 0.7\n",
"#action_selection = {'type': \"egreedy\", \"param\": 0.1, \"decay\": 0.0}\n",
"action_selection = {'type': \"softmax\", \"param\": 1.0, \"decay\": 0.0}\n",
"alpha = 0.1\n",
"nb_episodes = 100\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(render_mode=None)\n",
"\n",
"# Create the agent\n",
"agent = QLambdaLearningAgent(env, gamma, lbda, action_selection, alpha)\n",
"\n",
"# Train the agent\n",
"returns, steps = agent.train(nb_episodes)\n",
"\n",
"plt.figure(figsize=(12, 5))\n",
"plt.subplot(121)\n",
"plt.plot(returns)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Return\")\n",
"plt.subplot(122)\n",
"plt.plot(steps)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Steps\")\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"lbda = 0.7\n",
"#action_selection = {'type': \"egreedy\", \"param\": 0.1, \"decay\": 0.0}\n",
"action_selection = {'type': \"softmax\", \"param\": 1.0, \"decay\": 0.0}\n",
"alpha = 0.1\n",
"nb_episodes = 10\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(render_mode='rgb_array_list')\n",
"recorder = GymRecorder(env)\n",
"\n",
"# Create the agent\n",
"agent = QLambdaLearningAgent(env, gamma, lbda, action_selection, alpha)\n",
"\n",
"# Train the agent\n",
"returns = agent.train(nb_episodes, recorder)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"video = \"videos/gridworld-traces.gif\"\n",
"recorder.make_video(video)\n",
"ipython_display(video, loop=0, autoplay=1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Q:** Vary the trace parameter $\\lambda$ and discuss its influence."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"#action_selection = {'type': \"egreedy\", \"param\": 0.1, \"decay\": 0.0}\n",
"action_selection = {'type': \"softmax\", \"param\": 1.0, \"decay\": 0.0}\n",
"alpha = 0.1\n",
"nb_episodes = 100\n",
"\n",
"list_returns = []\n",
"\n",
"for lbda in np.linspace(0.1, 1.0, 10):\n",
"\n",
" # Create the environment\n",
" env = GridWorldEnv()\n",
"\n",
" # Create the agent\n",
" agent = QLambdaLearningAgent(env, gamma, lbda, action_selection, alpha)\n",
"\n",
" # Train the agent\n",
" returns, steps = agent.train(nb_episodes)\n",
" \n",
" list_returns.append(returns)\n",
" \n",
"plt.figure(figsize=(12, 6))\n",
"for idx, lbda in enumerate(np.linspace(0.1, 1.0, 10)):\n",
" plt.plot(list_returns[idx], label=str(lbda))\n",
"plt.legend()\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**A:** $\\lambda$ should not be too high nor too low in order to speed up learning. It controls the bias/variance trade-off."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Q:** Increase the size of Gridworld to 100x100 and observe how long it takes to learn the optimal strategy using eligibility traces or not.\n",
"\n",
"```python\n",
"env = GridWorldEnv(size=100)\n",
"```\n",
"\n",
"Comment on the **curse of dimensionality** and the interest of tabular RL for complex tasks with large state spaces and sparse rewards (e.g. robotics)."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Parameters\n",
"gamma = 0.99\n",
"lbda = 0.7\n",
"#action_selection = {'type': \"egreedy\", \"param\": 0.1, \"decay\": 0.0}\n",
"action_selection = {'type': \"softmax\", \"param\": 1.0, \"decay\": 0.0}\n",
"alpha = 0.1\n",
"nb_episodes = 100\n",
"\n",
"# Create the environment\n",
"env = GridWorldEnv(size=100)\n",
"\n",
"# Create the agent\n",
"agent = SoftQLearningAgent(env, gamma, action_selection, alpha)\n",
"#agent = QLambdaLearningAgent(env, gamma, lbda, action_selection, alpha)\n",
"\n",
"# Train the agent\n",
"returns, steps = agent.train(nb_episodes)\n",
"\n",
"plt.figure(figsize=(12, 5))\n",
"plt.subplot(121)\n",
"plt.plot(returns)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Return\")\n",
"plt.subplot(122)\n",
"plt.plot(steps)\n",
"plt.xlabel(\"Episodes\")\n",
"plt.ylabel(\"Steps\")\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**A:** When the Gridworld is too big, the likelihood to hit the target per chance when exploring is very low. There are a lot of unsuccessful trials before learning starts to occur. But it happens after a while."
]
}
],
"metadata": {
"kernelspec": {
"display_name": ".venv",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.7"
}
},
"nbformat": 4,
"nbformat_minor": 4
}