{ "cells": [ { "cell_type": "markdown", "id": "n8xo4ceg28", "metadata": {}, "source": [ "# Empirical Bayes for Automatic Hyperparameter Tuning\n", "\n", "When using `NormalRegressor` in a contextual bandit, you must specify two hyperparameters:\n", "\n", "- **`alpha`** (prior precision): Controls how strongly the model regularizes its coefficients toward zero. Too high means the model is slow to learn; too low means it overfits early data.\n", "- **`beta`** (noise precision): Controls how much the model trusts individual observations. Too high means updates are too aggressive; too low means the model learns too slowly.\n", "\n", "In practice, these are rarely known upfront. Misspecified hyperparameters lead to poor exploration-exploitation tradeoffs: the bandit either over-explores (wasting pulls on suboptimal arms) or under-explores (locking in on a suboptimal arm too early).\n", "\n", "`EmpiricalBayesNormalRegressor` solves this by automatically tuning `alpha` and `beta` via **evidence maximization** (MacKay's update rules). Each time the model is updated, it adjusts the hyperparameters to maximize the marginal likelihood of the observed data.\n", "\n", "In this notebook, we demonstrate that:\n", "1. A `NormalRegressor` with badly misspecified hyperparameters suffers high regret\n", "2. An `EmpiricalBayesNormalRegressor` initialized with the **same bad hyperparameters** recovers and matches or beats a well-tuned baseline\n", "3. Adding a **decay rate** enables the EB agent to **adapt to non-stationarity** — when the environment shifts, the EB + decay agent tracks the new optimal hyperparameters while stationary agents stay stuck with outdated estimates" ] }, { "cell_type": "markdown", "id": "whesvt9qddd", "metadata": {}, "source": [ "## Setup: Linear Reward Oracle with Regime Shift\n", "\n", "We simulate a contextual bandit where the expected reward for each arm is a linear function of user features. The oracle knows the true coefficients and noise level; the bandit agents do not.\n", "\n", "To demonstrate the value of decay, we use a **two-regime data-generating process**:\n", "\n", "- **Phase 1** (rounds 0–2499): Large coefficients, low noise (`sigma=0.5`). The optimal EB hyperparameters are low alpha (weak regularization) and high beta (trust data). Arm 0 is best on average.\n", "- **Phase 2** (rounds 2500–4999): Smaller coefficients, higher noise (`sigma=1.0`). The optimal alpha is higher (more regularization) and optimal beta is lower. Arm 1 is now best.\n", "\n", "This creates an environment where the stationary agents' converged hyperparameters and learned coefficients are wrong for the second phase, while the EB + decay agent can adapt.\n", "\n", "We use a **shared model** (`LipschitzContextualAgent`) so that all arms contribute data to a single learner. This is essential for empirical Bayes — the hyperparameter updates need sufficient data from every observation, not just the 1-in-K observations that happen to land on a particular arm." ] }, { "cell_type": "code", "execution_count": 1, "id": "5hnvpftwzto", "metadata": { "execution": { "iopub.execute_input": "2026-03-08T20:09:17.068429Z", "iopub.status.busy": "2026-03-08T20:09:17.067890Z", "iopub.status.idle": "2026-03-08T20:09:17.373527Z", "shell.execute_reply": "2026-03-08T20:09:17.372613Z" } }, "outputs": [], "source": [ "import numpy as np\n", "from numpy.typing import NDArray\n", "\n", "rng = np.random.default_rng(42)\n", "\n", "# Phase 1: strong signal, low noise\n", "TRUE_COEFS_1 = np.array(\n", " [\n", " [1.0, 0.5, -0.3, 0.2], # arm 0: best on average\n", " [-0.5, 0.8, 0.1, -0.4], # arm 1\n", " [0.3, -0.2, 0.6, 0.1], # arm 2\n", " [0.0, 0.1, 0.1, 0.7], # arm 3\n", " ]\n", ")\n", "TRUE_NOISE_STD_1 = 0.5 # beta_true = 4.0\n", "\n", "# Phase 2: weaker signal, noisier — different optimal hyperparams\n", "TRUE_COEFS_2 = np.array(\n", " [\n", " [0.0, 0.1, 0.1, -0.1], # arm 0: was best, now weakest\n", " [0.3, 0.2, 0.3, 0.2], # arm 1: now best (moderate, uniform)\n", " [-0.2, 0.1, -0.1, 0.1], # arm 2\n", " [0.1, -0.1, 0.2, 0.1], # arm 3\n", " ]\n", ")\n", "TRUE_NOISE_STD_2 = 1.0 # beta_true = 1.0\n", "\n", "SHIFT_ROUND = 2500\n", "\n", "N_ARMS = TRUE_COEFS_1.shape[0]\n", "FEATURE_NAMES = [\"intercept\", \"feature_1\", \"feature_2\", \"feature_3\"]\n", "\n", "\n", "class LinearRewardOracle:\n", " \"\"\"Generates linear rewards with Gaussian noise.\"\"\"\n", "\n", " def __init__(self, coefs: NDArray[np.float64], noise_std: float):\n", " self.coefs = coefs\n", " self.noise_std = noise_std\n", " self.context: NDArray[np.float64] = np.zeros(coefs.shape[1])\n", "\n", " def set_context(self, context: NDArray[np.float64]) -> None:\n", " self.context = context\n", "\n", " def expected_reward(self, arm_idx: int) -> float:\n", " return float(self.context @ self.coefs[arm_idx])\n", "\n", " def best_expected_reward(self) -> float:\n", " return float(np.max(self.context @ self.coefs.T))\n", "\n", " def generate_reward(self, arm_idx: int, rng: np.random.Generator) -> float:\n", " return self.expected_reward(arm_idx) + rng.normal(0, self.noise_std)\n", "\n", "\n", "oracle = LinearRewardOracle(TRUE_COEFS_1, TRUE_NOISE_STD_1)" ] }, { "cell_type": "markdown", "id": "uhx1zzc0jp", "metadata": {}, "source": [ "## Defining the Arms and Agents\n", "\n", "We compare four agents, all using Thompson Sampling with a **shared model** (`LipschitzContextualAgent`). The arm featurizer creates **block-sparse** features: for arm $a$, the user features are placed in the $a$-th block of a $K \\times d$-dimensional vector (zeros elsewhere). This lets the shared model learn arm-specific coefficients — equivalent to a disjoint model, but with a single set of hyperparameters.\n", "\n", "1. **Default `NormalRegressor`** (`alpha=1.0, beta=1.0`) — reasonable hyperparameters that happen to work okay\n", "2. **Misspecified `NormalRegressor`** (`alpha=10.0, beta=0.1`) — tight prior and underestimated noise precision; the model barely updates from data\n", "3. **`EmpiricalBayesNormalRegressor`** (`alpha=10.0, beta=0.1`) — same bad initialization, but EB corrects the hyperparameters automatically\n", "4. **`EmpiricalBayesNormalRegressor` + decay** (`learning_rate=0.995`) — same as above, with exponential decay to forget old data. This agent pays a regret cost in the stationary first phase, but after the regime shift it adapts while the others cannot. The estimator uses **stabilized forgetting** (Kulhavý & Zarrop, 1993) to continuously re-inject the EB-tuned prior, preventing it from decaying to zero." ] }, { "cell_type": "code", "execution_count": 2, "id": "0hlapbfixn1g", "metadata": { "execution": { "iopub.execute_input": "2026-03-08T20:09:17.376865Z", "iopub.status.busy": "2026-03-08T20:09:17.376083Z", "iopub.status.idle": "2026-03-08T20:09:17.917891Z", "shell.execute_reply": "2026-03-08T20:09:17.917019Z" } }, "outputs": [], "source": [ "from bayesianbandits import (\n", " Arm,\n", " EmpiricalBayesNormalRegressor,\n", " FunctionArmFeaturizer,\n", " LipschitzContextualAgent,\n", " NormalRegressor,\n", " ThompsonSampling,\n", ")\n", "\n", "N_FEATURES = len(FEATURE_NAMES)\n", "\n", "\n", "def block_sparse_featurizer(X, action_tokens):\n", " \"\"\"Create block-sparse features: arm a gets user features in the a-th block.\n", "\n", " For arm a with user features x ∈ R^d, the output is:\n", " z_a = [0...0 | x | 0...0] ∈ R^(K*d)\n", " where x occupies positions [a*d : (a+1)*d].\n", "\n", " This lets a single shared model learn arm-specific coefficients.\n", " \"\"\"\n", " n_contexts = X.shape[0]\n", " n_arms = len(action_tokens)\n", " # Output shape: (n_contexts, K*d, n_arms) — one feature vector per (context, arm) pair\n", " result = np.zeros((n_contexts, N_ARMS * N_FEATURES, n_arms))\n", " for i, token in enumerate(action_tokens):\n", " start = token * N_FEATURES\n", " end = start + N_FEATURES\n", " result[:, start:end, i] = X\n", " return result\n", "\n", "\n", "arm_featurizer = FunctionArmFeaturizer(block_sparse_featurizer)\n", "\n", "\n", "def make_agent(learner, seed):\n", " \"\"\"Create a LipschitzContextualAgent with the given shared learner.\"\"\"\n", " return LipschitzContextualAgent(\n", " arms=[Arm(action_token=i) for i in range(N_ARMS)],\n", " learner=learner,\n", " policy=ThompsonSampling(),\n", " arm_featurizer=arm_featurizer,\n", " random_seed=np.random.default_rng(seed),\n", " )\n", "\n", "\n", "# Agent 1: Default NormalRegressor (reasonable hyperparameters)\n", "default_agent = make_agent(NormalRegressor(alpha=1.0, beta=1.0), seed=0)\n", "\n", "# Agent 2: Misspecified NormalRegressor (bad hyperparameters)\n", "misspecified_agent = make_agent(NormalRegressor(alpha=10.0, beta=0.1), seed=0)\n", "\n", "# Agent 3: Empirical Bayes (same bad initial hyperparameters, but self-correcting)\n", "eb_agent = make_agent(EmpiricalBayesNormalRegressor(alpha=10.0, beta=0.1), seed=0)\n", "\n", "# Agent 4: Empirical Bayes + decay (defensive against non-stationarity)\n", "eb_decay_agent = make_agent(\n", " EmpiricalBayesNormalRegressor(alpha=10.0, beta=0.1, learning_rate=0.995),\n", " seed=0,\n", ")" ] }, { "cell_type": "markdown", "id": "rt5xffflrp", "metadata": {}, "source": [ "## Running the Simulation\n", "\n", "We run all four agents for 5000 rounds on the same sequence of contexts. At round 2500, we introduce an abrupt **regime shift**: the true coefficients change (making a different arm optimal) and the noise level doubles. Each round, we sample a random user context, let each agent choose an arm, observe the reward, and update the agent." ] }, { "cell_type": "code", "execution_count": 3, "id": "3pf27jxnova", "metadata": { "execution": { "iopub.execute_input": "2026-03-08T20:09:17.920842Z", "iopub.status.busy": "2026-03-08T20:09:17.920194Z", "iopub.status.idle": "2026-03-08T20:09:33.021026Z", "shell.execute_reply": "2026-03-08T20:09:33.020257Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Round 1000/5000\n", "Round 2000/5000\n", "Round 3000/5000\n", "Round 4000/5000\n", "Round 5000/5000\n", "Simulation complete.\n" ] } ], "source": [ "N_ROUNDS = 5000\n", "\n", "# Pre-generate contexts so all agents see the same sequence\n", "context_rng = np.random.default_rng(123)\n", "contexts = np.column_stack(\n", " [\n", " np.ones(N_ROUNDS), # intercept\n", " context_rng.standard_normal((N_ROUNDS, 3)), # 3 user features\n", " ]\n", ")\n", "\n", "agents = {\n", " \"Default (α=1, β=1)\": default_agent,\n", " \"Misspecified (α=10, β=0.1)\": misspecified_agent,\n", " \"EB (α=10, β=0.1)\": eb_agent,\n", " \"EB + decay (lr=0.995)\": eb_decay_agent,\n", "}\n", "\n", "# Track results\n", "optimal_rewards = np.empty(N_ROUNDS)\n", "expected_rewards = {name: np.empty(N_ROUNDS) for name in agents}\n", "\n", "# Track EB hyperparameters over time (both EB agents)\n", "eb_alpha_history = []\n", "eb_beta_history = []\n", "eb_decay_alpha_history = []\n", "eb_decay_beta_history = []\n", "\n", "for t in range(N_ROUNDS):\n", " # Regime shift at the midpoint\n", " if t == SHIFT_ROUND:\n", " oracle.coefs = TRUE_COEFS_2\n", " oracle.noise_std = TRUE_NOISE_STD_2\n", "\n", " context = contexts[[t]] # shape (1, 4)\n", " oracle.set_context(contexts[t])\n", " optimal_rewards[t] = oracle.best_expected_reward()\n", "\n", " for name, agent in agents.items():\n", " reward_rng = np.random.default_rng(t * 1000 + hash(name) % 1000)\n", "\n", " (action,) = agent.pull(context)\n", " arm_idx = action\n", " expected_rewards[name][t] = oracle.expected_reward(arm_idx)\n", "\n", " reward = oracle.generate_reward(arm_idx, reward_rng)\n", " agent.select_for_update(arm_idx).update(context, np.array([reward]))\n", "\n", " # Record EB hyperparameters from the shared learners\n", " eb_learner = eb_agent.learner\n", " eb_alpha_history.append(eb_learner.alpha)\n", " eb_beta_history.append(eb_learner.beta)\n", "\n", " eb_decay_learner = eb_decay_agent.learner\n", " eb_decay_alpha_history.append(eb_decay_learner.alpha)\n", " eb_decay_beta_history.append(eb_decay_learner.beta)\n", "\n", " if (t + 1) % 1000 == 0:\n", " print(f\"Round {t + 1}/{N_ROUNDS}\")\n", "\n", "print(\"Simulation complete.\")" ] }, { "cell_type": "markdown", "id": "ukjlzlsasls", "metadata": {}, "source": [ "## Results: Cumulative Regret\n", "\n", "Cumulative regret measures the total cost of not always choosing the best arm. Lower is better. A well-calibrated agent should achieve sublinear regret (the curve flattens over time)." ] }, { "cell_type": "code", "execution_count": 4, "id": "ht9mecx2oqn", "metadata": { "execution": { "iopub.execute_input": "2026-03-08T20:09:33.023465Z", "iopub.status.busy": "2026-03-08T20:09:33.023052Z", "iopub.status.idle": "2026-03-08T20:09:34.005748Z", "shell.execute_reply": "2026-03-08T20:09:34.004952Z" } }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABW0AAAHpCAYAAAD5+R5uAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQABAABJREFUeJzsnQV8FFcXxU/cPSEhENzdvThF20Jp0RYtVqC4fjgtFHdt0RarAAUKFG0p7l7cI8Tdd/f73RdmshviunL/dJqZt7Mzb2Z2kztn7jvXSKVSqcAwDMMwDMMwDMMwDMMwDMNoBcYF3QGGYRiGYRiGYRiGYRiGYRgmGRZtGYZhGIZhGIZhGIZhGIZhtAgWbRmGYRiGYRiGYRiGYRiGYbQIFm0ZhmEYhmEYhmEYhmEYhmG0CBZtGYZhGIZhGIZhGIZhGIZhtAgWbRmGYRiGYRiGYRiGYRiGYbQIFm0ZhmEYhmEYhmEYhmEYhmG0CBZtGYZhGIZhGIZhGIZhGIZhtAgWbRmGYRiGYRiGYRiGYRiGYbQIFm0ZhmG0hH79+qFEiRK5us2tW7fCyMgIL168yNXtMgzDMAzDMHkHx4W6zy+//AJnZ2dERkYWdFe0igYNGmDixIkF3Q2G0QlYtGUYRq94+vQphgwZglKlSsHS0hL29vZo3LgxVqxYgZiYGOgr8+bNw/79+6EtSGKxNJmamqJIkSLiBsTb2xvaTnR0NGbNmoW///67oLvCMAzDMEw24bhQOzDEuFChUGDmzJkYOXIkbG1t5XZK0OjUqRO06bpcvXo1X/c7adIkrFmzBn5+fvm6X4bRRUwLugMMwzC5xZ9//onPP/8cFhYW6NOnD6pUqYL4+HicPXsWEyZMwL1797Bx40a9FW0/++wzdO7cWaP9yy+/RI8ePcQ5KQjmzJmDkiVLIjY2FhcvXhTBIV2Pu3fvClFdm4Pz2bNni/nmzZsXdHcYhmEYhskiHBdyXFiQceHBgwfx8OFDDB48ONf6oS988sknIrFm7dq14l6BYZi0YdGWYRi94Pnz50KcLF68OE6dOoXChQvLrw0fPhxPnjwRwbuhYWJiIqaCon379qhTp46Y/+qrr+Dq6ooFCxbgwIED6NatW771Q6VSCeHYysoq3/bJMAzDMEzBwHFh6nBcmH9x4ZYtW8RoP8ooZjQxNjYWySbbt28XYjhl+zIMkzpsj8AwjF6wcOFC4Re1adMmDcFWokyZMhg1apSYJ39XCg4o6zMl1E7DnyRontoePXqEL774Ag4ODnBzc8P06dNFwPf69Wv5abGHhweWLFmSKU9ZGl5F7RkNs1q8eDEaNWoEFxcXEVjWrl0bv/3223t9joqKwrZt2+RhZzTcLLX903Asso5IjYYNG8oCq8TPP/8s9kn7Jk8uEsbpmLPLBx98IA9XVOfBgwcieKN9UAYu9YOE3ZTcvn0bzZo1E/0pWrQovv32WxEUpzzH0tCzv/76S2yL1t+wYYN4LTQ0FKNHj4aXl5fIQKbPBgnJSqVSvE7boWtMSIFkys8FwzAMwzDaC8eFHBcWZFxIgvDRo0fRunXrbH1+ExMTMXfuXJQuXVr0ifo/depUxMXFaaxHfaR+eHp6wtraGi1atMD9+/fF+tK9QG5w48YNkYhB9ztk9dCqVSsxgi67cTrRpk0bvHz5Ejdv3sy1fjKMPsKZtgzD6AU0BInESBI484Lu3bujYsWK+P7770XGLgUhJDBSwNeyZUsR3O3YsQPjx49H3bp10bRp01zZL3nxfvzxx+jdu7eweti9e7ewgDh06BA6duwo1vnpp59EFmu9evXkIVgU5KV1HGQdceXKFdFPCQqaKPhatGiR3Pbdd98JcZoyYmn7AQEBWLVqlTg2Ct4cHR2zfDxSwObk5CS3kW2FlIkwefJk2NjYiMINZPXw+++/o0uXLmI98jyjYJQCvylTpoj1fvzxxzStH2hIWs+ePYXH8aBBg1C+fHkxvI2CSdoWtRcrVgznz58X2/P19cXy5ctFYL5u3ToMGzZM7PvTTz8V26tWrVqWj5dhGIZhmPyH40KOCwsyLrx27ZqI22vVqpWtzy/F3ZSMQQkN48aNw6VLlzB//nz8999/2Ldvn7we9ZMeUHz00Udo27Ytbt26JX6SaJxbUJxOSRck2FLxMDMzM3H/QzYR//zzD+rXr5+tOJ2SQohz586hZs2audZfhtE7VAzDMDpOWFiYin6dffLJJ5la//nz52L9LVu2vPcatc+cOVNepnlqGzx4sNyWmJioKlq0qMrIyEj1/fffy+0hISEqKysrVd++feU22ge9n/apzunTp0U7/ZSg9xUvXlxjvejoaI3l+Ph4VZUqVVQtW7bUaLexsdHYb1r7p3NlYWGhGjdunMZ6CxcuFMfz8uVLsfzixQuViYmJ6rvvvtNY786dOypTU9P32tPa74kTJ1QBAQGq169fq3777TeVm5ub2D8tS7Rq1UpVtWpVVWxsrNymVCpVjRo1UpUtW1ZuGzlypOjjjRs35LagoCCVs7Pze+eYziO1HT16VKNfc+fOFefq0aNHGu2TJ08Wx/vq1SuxTH1O+VlgGIZhGEb74biQ48KCjgt//PFHsT7FzSmhvnTs2DHN9968eVO896uvvtJoHz9+vGg/deqUWPbz8xMxeefOnTXWmzVrllgvtfuCtOL1K1eupLkObd/c3Fz19OlTuc3Hx0dlZ2enatq0abbidAna7rBhwzLsJ8MYMmyPwDCMzhMeHi5+2tnZ5dk+6Im3uh8YDa0ijXfgwIFyO2We0lP7Z8+e5dp+1b22QkJCEBYWJp52X79+PVvbo6fkNLyJMlmTNOok9uzZgwYNGogMA2Lv3r1iyBVl2QYGBsoTWUCULVsWp0+fztT+aFgYZSjQkDPKFqCn7mR7QEOmiODgYOFBTPuJiIiQ9xMUFCQyBR4/fixXFaZhZmThUKNGDXn7lO1MWcipQQXQaBvq/Prrr+L8Uaav+nFRP6nK75kzZ7JxVhmGYRiG0RY4Lsw8HBfmTVxIcWzKkWWZ5fDhw+Ln2LFjNdop45aQanScPHlS2Ch8/fXXGuuNHDkSuQWdg2PHjonRb+r2amRF16tXL1FcWPq+ZTVOJ6TzzjBM2rA9AsMwOg8FnASJfnmFJGZKkLctea9SYa2U7VKglhuQDQJZMZDfk7qPVU4M+8kiYf/+/bhw4YKwkyB/WRrGRUPAJEgsJVGXBNrUoKFRmWHNmjUoV66cEJs3b94sgl/1YVJUII72QzYMNKWGv7+/sE4gCwcKBlNC3mNpibYpoeMivy3Jmyy1fTEMwzAMo7twXJg1OC7Mu7hQPUEis1C8S4W6Usa3lDhBCSL0urQekXI9EkrVxWISXsniLOU65ubmGfaF3kcWEpSUkhKyjaMED6p1Ubly5SzH6dL54SJkDJM+LNoyDKMXwTkZ8N+9ezdT66cVHFBQkxaUXZuZtpQBWnb2JfHvv/8KP1vykF27dq14qk1iKRn679y5E9mFfK+oWAFl25JoSz8pOCSvXAkKwqjvR44cSfU4qQhBZiCfXam4GT2lb9KkiXgyT75itA2pyAN5AafMis1MsJceqVUEpv1R4QPy5EoNEpgZhmEYhtFdOC7MGhwX5n5cSAWEpVFy0uiyrJJbYiaJqikTGWjEHHnSFjRUBC5lAgzDMJqwaMswjF5AFWE3btwoskdTe8qrjvT0mQIFdaQn1rlJTvZFRbgom5cq3apnp5Jom5PAjiwK6HyRVcDSpUuFNQINDSPhW4IKmZH4TEFebgmZJP5SEQUqUrB69WpRdEwaakVidEYVdosXLy4yc1OSWlta0HFFRkZmuC9+6s8wDMMwugvHhRwXFmRcWKFCBfHz+fPnqFq1apbeS/EuJRnQ6DDKZpV4+/atuJ+g16X1pDhYXZSlEX8kFqtn6B4/flxjH9WrV89UXygDmRI9KNkiJQ8ePBBJH2SBlp04nezPqFib+jEyDPM+7GnLMIxeQE/ISYwk71kKalJCFgArVqyQMzDoqW5KnyrKZs1tKBgk1PdFWbYkMGdG5KQgUT0r98WLF8LaICV07CmF4YyGwvn4+IiqrlRplpbVocq4tP/Zs2e/N7SLlrNrAUFP9Sn7lqwYqLJtoUKFRBtVoaUqvSlRH85FmbgkypNVhAR54u7YsSPT+yfvXNoGCeEpofNH3mAEBahSG8MwDMMwugXHhRwXFmRcWLt2bWE/cPXqVWSVDh06iJ/qtmUEJVoQHTt2FD9btWoFU1NTrFu3TmM9SoxQhxJASJRWnzLrtUv3Ah9++CH++OMPcQ8iQfdaNOqPRtBJdiRZjdPJmo2gUX8Mw6QNZ9oyDKMXkDhKwQOJj/TEtk+fPqhSpYp4gnv+/HmRVdqvXz95fRJ3v//+e/GThu+TqPro0aNc7xd5PFGBrylTpojAhTykdu/eLQeB6UFBGQVo7dq1E5YC5KtFHrFkF0C+rCmDwxMnToj1KWOWnrjXr18/3YCQCreRLQEFZF27dn3vfJKXLvWbgjSyNqD1KWNg3759GDx4sHhvdpgwYYKwYti6dSuGDh0qjomCPspEGDRokMi+pWCQAr83b94IUVm6Afv555+FvQEVWSChmkRn8humc5uZLAjaNxVCowwc+jzQeYuKisKdO3fw22+/iWMlQZ+sFSpVqiSykCnTmK4bfZ5oYhiGYRhGu+G4kOPCgowLSSglsZNi8zlz5qSafUpxdkpq1qwp4v++ffuKBA8SiZs1a4bLly9j27ZtIh6nEWuEu7s7Ro0ahSVLlgg7NbpfoJiZrM2oz1nJDqa6E1RILCW0feonZepSrE5Fz0gopmQLqrWxcOFCed2sxum0TXqNjplhmHRQMQzD6BGPHj1SDRo0SFWiRAmVubm5ys7OTtW4cWPVqlWrVLGxsfJ60dHRqoEDB6ocHBzEOt26dVP5+/tTSqlq5syZ8no0T20BAQEa++nbt6/Kxsbmvf03a9ZMVblyZY22p0+fqlq3bq2ysLBQubu7q6ZOnao6fvy42O7p06c1tlm8eHGN927atElVtmxZ8d4KFSqotmzZIvdJnQcPHqiaNm2qsrKyEq/Rtghan5afP3/+Xl979+4tXqO+pcXvv/+uatKkiThWmqgPw4cPVz18+DDN96jv98qVK++9plAoVKVLlxZTYmKifI769Omj8vDwUJmZmamKFCmi6tSpk+q3337TeO+NGzdUH3zwgTgfRYsWVc2fP1+1cuVKsS8/Pz95PTqPHTt2TLVvERERqilTpqjKlCkjPiOurq6qRo0aqRYvXqyKj4+X1zt//ryqdu3aYp2UnwuGYRiGYbQfjgs5LiyouHDv3r0qIyMj1atXrzTaqS/0/tQmujchEhISVLNnz1aVLFlSxMVeXl6ij+r3MgTF0dOnTxfxM90DtGzZUvXff/+pXFxcVEOHDs3w+yHF62lNr1+/Futdv35d1bZtW5Wtra3K2tpa1aJFC3E+UpLZOJ3uBQoXLqyaNm1ahn1kGEPHiP6XnqjLMAzDMNrM6NGjxRN/8iRLqzgcwzAMwzAMo/9oS1xI9maUnUsWDHPnzs23/VJ2LtkfUIbs//73P2jj9SCrNxpFSPZ1VGiZYZi0YU9bhmEYRmeIiYnRWCZv3Z9++kkM2WLBlmEYhmEYxnDQ5riQ9k/WCGQDRoJlfhy/uhcu1YzQ1uuxYMECjBgxggVbhskEnGnLMAzD6Aw1atQQQSj5FpPv7aZNm0RBtZMnT6Jp06YF3T2GYRiGYRgmnzD0uJDqQ9BEtSpsbW1x9uxZ7Nq1S/jpplZcLa8x9OvBMHkBFyJjGIZhdAYKSqkwBBVnoIIGtWrVEgEhB4IMwzAMwzCGhaHHhdWqVROFwaggWHh4uFycLLUiZ/mBoV8PhskLONOWYRiGYRiGYRiGYRiGYRhGi2BPW4ZhGIZhGIZhGIZhGIZhGC2C7REygVKpFF4sdnZ2Is2fYRiGYRiG0Q5UKhUiIiLg6ekJY2POR5Dg+JVhGIZhGEa341cWbTMBCbZeXl65eX0YhmEYhmGYXOT169coWrQon9N3cPzKMAzDMAyj2/Eri7aZgDJspZNpb2+P/MiMCAgIgJubG2eM6Ch8DXUbfb5+8fHxWLJkiZgfN24czM3NoY/o8zU0BPj66Tb5ff2o+Ao9XJfiNUNnzZo1YkpMTBTLHL8y+gD/XWD488XoKob0+8tQ7jXzM35l0TYTSJYIJNjml2gbGxsr9qXvX2p9ha+hbqPP10+hUKBt27Zi3snJCSYmJtBH9PkaGgJ8/XSbgrp+bGGVxPDhw8VENwMODg4cvzJ6Af9dYPjzxegqhvT7y1DuNfMzfmXRlmEYxoCgP5zNmzcv6G4wDMMwDMMwDMMwegTfa+Y++i3zMwzDMAzDMAzDMAzDMAzD6BicacswDGNgVSrJU4kgXyUeTswwDMMwDMMwDMPwvab2waJtLvt3JCQk5IrnCW2HfE/03fNEX+FrqBuQMbqhfcfod8vatWvF/NSpU9kcnmEYxsDh+JXRBTi2TsLMzIw9IhmG0Vr4XjP3YdE2lzLX/Pz8EBoammvbo8AkIiKCs+B0FL6GugEJtiVLlmThkmEYhjE4OH5ldAmOrZNxdHSEh4cH3ycyDMMYAAUq2p45cwaLFi3CtWvX4Ovri3379qFz586prjt06FBs2LABy5Ytw+jRo+X24OBgjBw5EgcPHhQCTNeuXbFixQrY2trK69y+fVtU0b1y5YoYDkzrT5w4MdeOQxJsCxUqBGtr6xz/AaWgJDExEaampvzHWEfha6j90IMRHx8f8bunWLFi/F1jGIZhDAqOXxldgmPrpHMQHR0Nf39/cU4KFy5c0JeFYRiG0WfRNioqCtWrV8eAAQPw6aefprkeibkXL16Ep6fne6/17t1biC7Hjx8Xqdj9+/fH4MGDsXPnTvF6eHg4PvzwQ7Ru3Rrr16/HnTt3xP7oCSWtlxtDyiTB1sXFBbkBByW6D19D3YAe4pBwSw9JaLgZwzAMwxgCHL8yugbH1klYWVmJnyTc0v0nVWpnGIZh9JcCFW3bt28vpvTw9vYWmbF//fUXOnbsqPHaf//9h6NHj4oM2jp16oi2VatWoUOHDli8eLEQeXfs2IH4+Hhs3rxZDIGuXLkybt68iaVLl6Yp2sbFxYlJgoRfKTOPppTrUhBBf0DpZ24hbSs3t8nkL3wNtR8Sauk60QMf9aCXvufSMDx9Qzo2aV4fj1Hfr6EhwNdPt8nv68ff86wj1WCgEWIMw+gW0vc2ZfzKMAzD6B9a7WlLQfiXX36JCRMmCLE1JRcuXBAZs5JgS1BGLdkkXLp0CV26dBHrNG3aVMOzsm3btliwYAFCQkLg5OT03nbnz5+P2bNnv9dOFdepOJg69MeS+kkZC5StlxvQjQ5tj+DK7roJX0PdgL5n9P0NCgrSyLSltrCwMHEd9a1QGT3EolEOUpaG+u9GfUKfr6EhwNdPt8nv60c1AJjswXEmw+ge/L1lGIYxHLRatCVhlXxdv/nmmzS9uGhYiDq0vrOzs3hNWocKDanj7u4uv5aaaDtlyhSMHTtWI9PWy8tLDKW2t7fXWJdEXLpZoP3SlJvwcG3dh6+hdkPfWRIUyNrE0tJSQ3CggJi+8/oo2trY2Ih5+v2pz6Ktvl5DQ4Cvn26T39dP/fc3wzAMwzAMw+gLWivaUnEyKih2/fr1fH+aaGFhIaaU0I1HypsPWqb+SVNuQJkp0rb4SapuwtdQN5C+t6l9t9Nq14cHCY0bN5bn9e34DOEaGgp8/XSb/Lx+/B1nGIZhGIYpeMiypVGjRvI8k3O09k7233//FUN3qaq7lMX68uVLjBs3DiVKlBDreHh4yNUzJciiIDg4WLwmrfP27VuNdaRlaR0md9i4caPISKabp+XLl+fKNl+8eCFu/MiHODuQvca8efOgC/z0008oV66c8KmqUqUKfv/99zzd3/3791G0aFF5qDxjGNAfTyrOSBP/IWUYhmEMGY5ddS92pXsNjl0ZhmG0E77XNCDRlsS227dvC7FOmqiwGPnbUlEyomHDhggNDRVZuRKnTp0Sw/Lq168vr3PmzBm54AJx/PhxlC9fPlVrBEOiX79+cqYhZdyRbUSbNm1E0basFvUgC4kRI0Zg0qRJonhcWkXecsrff/8t+kvXPSNu3bqFw4cPp2mvkVt899134mkSBazksZwdqKhe//790bt3b9y5c0cU3+vVqxdevXqVZ/2qVKkSGjRoIIryMQzDMAzDaDscu+YOHLsyDMMwjG5QoKJtZGSkLMgSz58/F/MkVJHHJD2xVZ9IWKTsWBJciYoVK6Jdu3YYNGgQLl++jHPnzgnhsEePHkLgJUj4Is/GgQMH4t69e9izZ4+wXVD3rDVk6Pz5+vqKjNYjR46gRYsWGDVqFDp16pSlwmp0zUgY79ixIwoXLqwV1YhXrVqFzz//HLa2tnnuEUr7GTZsWI7EaPrMz5gxA6VLl8aQIUPg4OCAq1ev5mm/SChet25drhXRY3TDuoMeetBE8wzDMHnxeybxv//4xDJ5Aseuhh27knDPsSvDMIx2wveaeiba0h/1mjVriokgIZXm6Y9/ZtmxYwcqVKiAVq1aoUOHDmjSpIkY6iRBwcOxY8eEIFy7dm1hr0Dbz6tMUF2DvHtJCC9SpAhq1aqFqVOn4o8//hAC7tatW+X1SOD56quv5GJsLVu2FJmsBK1XtWpVMV+qVCmRCUsi8NOnT/HJJ5+IDF4STuvWrYsTJ05o7J/W3b9/v0YbZYWq71uCtkmiMkFZ0vReCtxSQ6FQ4LfffsNHH330njVGly5dhKhcvHhxMazr/PnzwoYju8yePRtjxoyRz0F2IJsPErslD2MqcEfnPCfF7TLTL8qsJjuRf/75J9v7YXQLerhC9iU0qY9AYBiGyQ1USiXezpyJiOEjEHnqFJ9UJtfh2JVjV45dGYZhtBNdvdeMf/ECL3r2wuMWLRF1/jy0iQItRNa8efMsZXqRaJcSZ2dn7Ny5M933VatWTXjk5icfrTqLgIi4bL9fBRWMkPXCZm52Fjg4sglyAgmy1atXx969e4VQS9BTbysrKyHmkhC+YcMGIZQ/evQI3bt3F/5SrVu3FhnPNE/i7t27d4WQTkOwKMDevn27EFEfPnyYLZGUtkteWV27dhXbIPGY+pQaZK0RFhaGOnXqaLRTBjH1hfpJthqUgf3ZZ5+hc+fO8jrkgZuRDy55auVE6E2J+vcgKCgIEydOFP2UCkblVb8oC71GjRri+0HXk2EYhmFyQuC6dQj7LcnX0mfCRFifPAFTZ2c+qTpATmPX7MavHLsmwbFr5uDYlWEYhslNYu7ew5uvv0biu3pZyricxUJ6JdrqMxT0+oXHQleh7GUKHomzZ88KkZOyQUlIJBYvXiwyZCmblbKWaXgUQWKtVOCNhF+aJObOnYt9+/bhwIEDwsYiO6bWJNIThQoVStc/lorW0fq0ngSJtJTdTZm1kuXG+vXrsWvXLpw8eVJeb+jQoejWrVu6fZHsN3IbEsVJ6CbIi1c63rzsF72HzhfDMAzD5ITQ3/cicNVqedl95gwWbHUIjl05ds0OHLsyDMMwukrkv//Ce8xYKCMj5TYjc3NoEyza5hGUNZATcpJpm1uZn9JQfbJBIP9hSZiViImJERYIaUHvmTVrFv7880/hm0u+qfSenBTXyiy0HxKYpWMgHj9+LITcevXqyW00T9nCTZs2ldtIKFUXS/MT6selS5eEuEznjmweVq9enaf9omzl6OjoXN8uwzAMYzhEXb4Mv9mz5WXL/v3g8PHHBdonJv9jyOxm2uYGHLty7MowDMMwmSX86FF4T5hIng5i2apmTRRdtRKmrq7QJli0zSNyYlEgCngkJgo/U3XRMT/577//ULJkSVl8Jb9VKjiQkvSyXcePH4/jx4+LrNwyZcoIcZCsCKjIgAQdX0qLjNzwPnF1dRVCJO2LhlERJOKSaEuTBGUG169fX8M7tiDsESRsbGyEkEwT2T9QoTDyg6H+5VW/yBeMCkgwDMMwTFahv+EBK1YgaP0Guc2ha1cY9enDJ1PHyKm9VkHHrxy7cuzKMAzDMJmKXZcuQ9APP8htts2bo8jSJTC2toa2waIt8x6nTp3CnTt3RBErggqU+fn5iSC8RIkSmT5j586dE4XCqPCXJP6m9CUm0ZSycNWzYdPL+pQEWMpATQ/yaZVETGmehEkScckKgIqQEWTVQP1Sp6DsEehGRx0zMzMxSSJzXvWLvIdJTGcYhmGYrECeXy+690Dcgwdym3X9+ig0dQoCIyL4ZDL5BseuHLsyDMMwTGYK5vovWIDgbdvlNvuOHeH5/XwYmZlBG2HR1sCJi4sTgiyJoG/fvsXRo0cxf/58UbCrz7ssGSow1rBhQ1Gsa+HChShXrhx8fHyE7QEJsimLfUmULVtWFDOj4mOUcTF9+nQolcr3ip7R8H/aPvVh0qRJQqhMCxJbaVuHDh0S3q+UvWtra/veeiQGk9hMfrySaEsF6erWrYuZM2di06ZNOHbsmBCRSbQliwQ6ruzYEJDdA2Wr0k86hps3b4qnNyRwp5eJnJJ79+5h8uTJovgbbY+uQ9u2beVsldzoF0FZz9I5o+P39vYW15hhGIZhslJl99lHH0OlNjqm0PhxcO7fHyr6u8WiLZNHcOyaN7FryhhR22JXGo1GcOzKMAzDZAeKWX1nz5YL5hKFJk2Cc7++4u/WsRfHUMSuCCq7VIZWoWIyJCwsjMbvi58piYmJUd2/f1/8zC2USqUqPj5e/MxL+vbtK46LJlNTU5Wbm5uqdevWqs2bN6sUCoXGuuHh4aqRI0eqPD09VWZmZiovLy9V7969Va9evRKv37hxQ2zn+fPn8ntovkWLFiorKyux/urVq1XNmjVTjRo1Sl7H29tb9eGHH6psbGxUZcuWVR0+fFjl4OCg2rJli7wN2i5tX2LOnDkqDw8PlZGRkTiGtFi7dq2qQYMGGm2PHj0SfXBychLbOHLkiGrBggXi2I8ePZrj86g+HT9+XL6GxYsXV82cOTPNbdBrRYoUEeeLzoW7u7s4v2/fvs1Wn9Lr1+nTp+V15s2bp2rbtq3KUEnr+0uff19f3/e+B/pAQkKC6tChQ2KieX1Fn6+hIcDXT3tRxsWpfL/9TnW/fAWNKXj3ngK7funFaYaMPsavHLvmbeyqHiNqW+wqfb6+++47g45d8+r7a+hw3MHw50u/7zUTQ0NVL/r2S45dK1ZShfz2u/z6H0/+UFXfVl3VcGdD1d3Au1oVvxrR/wpaONZ2wsPD4eDggLCwMOEzqk5sbCyeP38u/F8tLS1zZX8F7QmmL1AxsvLly2PPnj0ikzc/Ub+G1A8q4kbVdZs3b57q+lR0bP/+/XJGQX5AVhGUDb1z5040btwYhkha31/KCPf390ehQoVgbGxcoH1ksgdfQ92Gr592En39BrxHjUJiQIBGu8fMGXDq2bPArl96cZohw/Gr7lGQsas6ZFWmbbErxdbUr0qVKhl07JpX95+GDscdDH++9BdFWBhe9umLuIcPkxrMzFBkwfew79BBLG67tw2Lry6W1/+y0peYWHei1sSvbI/A6C1knbB9+3YEBgYWaD9Onz4tbCDSCnoLChp6NnXqVIMOehmGYZjMEXn2HF4PGkTKSVKDqSmcevSAy1cDYebhwaeRYXIBjl0zjl2nTJnCsSvDMAyTKVQKBbzHjJUFWxNnZxRdvRrWtWqK5R9u/4CVN1bK68eHNIBrXFdoEyzaMnqNNgilHTt2FJO2Qf5gNDGGhZSpQlhbW3M2P8MwGRL/8iXefPONLNiauLnCa9UqWL3zjGcYJvfg2DVtKG6tUKECf9wYhmG0FG2611SpVPBfuBBR58/Lgm3xHT/DomRJsXz0+VENwTYuoA3iA1vC3d4a2gSP+2UYLYCGmOWnNQJjuCQkJGDRokVionmGYZj0iH/1Cq+/Hg7VuwDcslo1lD54kAXbfCA0NFQUe6WCqlWqVMEPP/zAH1ZGa+DYlWEYhtHWe00VCbYLFiJ42/akBhMTFFm+TBZsL/hcwJSzU+T14/zbIj6wFSa0rYCO1QpDm+BMW4ZhGIZhGOY9Ym7fxushQ6EICRHL5iVKoNjmTTDJQnV5JvvY2dnhzJkzIlMlKipKCLeffvqp8BplGIZhGIZhUidw7VoEb92atGBkBI9ZM2FTr55YvBVwC6NOj0KiMlEsx4fWQXxQc3SrUxRfNy8NbYNFW4ZhGIZhGEYjOyFo/XoErF4DKBSizax4MXitX8eCbT5iYmIiBFsiLi5OXBeuH8wwDMMwDJM2QT/+iMBVq+XlwnPnwPGzz8T8o5BH+PrE14hJjBHLCRGVEOfbBR+UdcO3natqpXUg2yMwDMMwDMMwAhIF3879FgErVsqCrWXVqiixe7fItGUyD2XJfvTRR/D09BQ3Afv3739vnTVr1qBEiRKiAnz9+vVx+fLl9ywSqlevjqJFi2LChAlwdXXlS8AwDMMwDJMK4YcPw3/xEnm50KRJsmD7OuI1hhwfgvD4cLGcGFUasd49UdzFDqt71oK5qXbKo5xpyzAMwzAMw0AVHw+fqf9D+KFD8tlwHT4crkOHwMjMjM9QFiFLAxJcBwwYIGwNUrJnzx6MHTsW69evF4Lt8uXL0bZtWzx8+BCFChUS6zg6OuLWrVt4+/at2MZnn30Gd3f3VPdH2bg0SYSHJ92UKJVKMalDy1Lmbm5m70rb4oxgJi/gz1fyeaApte82kz2k34l8Ppm8wJA+X9KxSvP5ecxR58/DZ8pUedn1m2/g1LeP6ENgTCAGHRskfhKKmKKIedMHbra22Na/LuwsTfL9+mR2fyzaMgzDMAzDGDjxL1/Ce8xYxN6/L7e5jR0L18GDCrRfukz79u3FlBZLly7FoEGD0L9/f7FM4u2ff/6JzZs3Y/LkyRrrklBLAvC///4rhNvUmD9/PmbPnv1ee0BAAGJjYzXaqDgI3SwkJiaKKTegmzTFu+xsbRxeyOg2/PlKhr6z9P0NCgqCGT9QyxXofIaFhYnPmbGxdmbbMbqLIX2+4uPjxUNrwt/fH+bm5vmy34SbNxE5aTI9wRbL5m3bIrFLZ9EHOu8zbsyAd6S3eE0RVwjRr/vDwsgSizqVhGViJPz9I5HfREREZGo9Fm0ZhmEYhmEMlARfX/jNnoPIf/4hVUS0GZmbo/C3c+Hw8ccF3T29vqm5du0apkxJrlxMN3KtW7fGhQsXxDJl15KnLRUko5s9slsYNmxYmtukbVHmrnqmrZeXF9zc3GBvb6+xLom4dLNgamoqptyERSQmL+HPF8R3ln5fUFFCslZhckdUo4dN9PtS30U1Jv8xpM8XxTc2NjZinkYN5YdoG3PjBl5PniILttZNGqPI/Hkwfvf78fDzw7gYcFHMqxLtEPNqIKCwwYLu1dC0qicKisz+/tbvTwyTqzRv3hyjR4/WibPar18/dO7cWV6mpyuDBw+Gs7Oz+IV58+bNXDmerVu3iqGLGbFp0yZ8+OGH0Bd++uknlCtXTtxMUjXr33//HdpGYGCg+EPx5s2bgu6KVkGBQo0aNcSk70EDwzDpE/fsOZ5364bIv/+WBVtjGxtRcIwF27z/G0VZqSmtDmjZz89PzL98+RIffPCByLClnyNHjkTVqlXT3KaFhYUQZ9UncU2NjVOdKB7KzYlQ/6kNU4sWLTBmzJgC70dmJsq47tKli8b5HDJkiBDm6HqRTUZuHM+2bdvg5OSU4XqU8U12HQV9XtTPh/rPrE4///wzypcvL8QE+h7t3bu3wI8p5USZs/Q7wNvbO1Prp/Xd5il754DPKX928vK7YyifL3qoVLNmTTFJD5jyclIEBMD7m1FQvRtRZF2/PrzWrIGptbV4PToxGguvLJRjpVi/T6BKdMDXzUujc82iBX6+MgPfsRswJGzSL4+hQ4e+99rw4cPFa7SOBAU3c+fOhS6wYsUKIahKHD16VCwfOnQIvr6+QmjMr+OhbJbp06dj5syZebofOh4Shim4l4Tp1PpC15bWsbW1RdeuXUUmT1b477//xI1F7969cefOHXET2atXL7x69SrbfQ8ODhbboxtMEsEHDhyIyMj0hyhs3LhRCO/0HjpeKtaiDhVr6dOnT56fd12D/njSAw2acju7imEY3SExOBjPOnaEIiDJ28vYzg4ugwahzOlTsGnUqKC7xwCoV6+e+FtOYt3t27eFgGfocOxquLErxXUkNpM9CMeuDMMw2kl+3msqIiPx5uvhUAQFyYJtsR82wtjCQl6HBNuQuBAxnxBeBYkRVdC6YiGM/7A8dAUWbQ0cGja3e/duxMTEaARHO3fuRLFixTTWpSxVGqKnCzg4OGhkwD59+hSFCxdGo0aN4OHhIX6B5Nfx/Pbbb0JYbNy4cZ7uh7xjmjRpggULFqS5DmVnHDx4EL/++iv++ecf+Pj4pFocJT3+/vtvEVzPmDEDpUuXFjeRdL6vXr2a7b6TYHvv3j0cP35cCOs0BJQyo9MjOjoa7dq1w9SpyWbjKSFxeceOHUIUZhiGYZJQJSTgeecucnatmacnSh89gkLjxsIkxTB6Jm8gAcrExOQ98YmWKU5h0oZjV8OMXX/55RecPHmSY1eGYRgGisgovBk6DLH37omzYVq4MIosWSwsviROvDyBfU/2iXmV0hxx/h1Qys0GK3rUhLGx7njvs2hr4NSqVUsEv/SkW4LmSbCllHZ1UtoJrF27FmXLlhVeHDSUR70wBgV7NPTIyspKCHzk0SYZUkvWBVQsQ/JZo2xf8j9R932hgholS5YU26ChgbRNdUjk69Spk3g/ia80dJDEWfV9SPOUDUqZoPQUv0SJEqkeD1VcHj9+PIoUKSKGTlElZxIo1aFsXTo3ZAtAQ9hoGFNGkCj+0UcfvVdAgPZF54bEY/KgI+GcjkU6hqzy5ZdfCiGVznVqkB8e2TRQ4ZOWLVuidu3a2LJlC86fP4+LF5M8XjIDmXmTAC4NUSORn7Jcs/skjTJ3KRP6xx9/FOecgvdVq1aJ80aiclrQtaNCLQ0aNEhzncqVK8PT0xP79iX9smaSrELou0YTV/dmGMMj6uIlPGryARL9/eW2IitWwNTFpUD7ZWiQxxv9HSYRSj32oeWGDRvmaNtr1qxBpUqVULduXegjHLsabuxK155sGzh2ZRiGMdx7TVViIrzHjkH0u6QxE0dHkWFr6uoqr/M26i1mXZglL8f6fQRTpQtW96wFGwvdGm2qW73VIZ53/QyJgUlDDrMDfcAlUSwr0Ae15O+a4mZGDBgwQIh3lO1IUDBEGYopgz51KKvym2++Ed6mlL1KmYxU0Zgg+4GePXti4cKFIjikQhf0mvqXlm5KSOylfbx48ULsj4LA7777TrxOgi15T1ElZRKGKfPyiy++ECJvs2bNhNdT06ZNhfB66tQpETCeO3cu1QrIZJVAGaE0nP7KlSsisyU1RowYgfv374tAVRL6KJOTLACoD5cuXRLD9qlvJAiT0JiZYWNnz54VQak6JDZu375deMFSQQXaHgXRJChTXwk6Z+lVnSY2bNggX7eMoIInVC1aPTCuUKGCCOSp6El64qc66teR+jxx4kTho6eejTFv3jwxpQeda2nflBVdp04d+TXqI3m80Dmnz1BOh5fSuaRrxyRVDJeuDWUp51dFT4ZhCp7YR4/wSs32iPBctBBWVasUWJ/0GbL5efLkibz8/PlzMfybBC/6+0eiV9++fcXfP/pbtXz5cvGAm2KinEBDyWmiQmQ0Eia/Ytfsxq8cu3LsmhYcu3LsyjCMbpHX95oqlQp+336LqDNJ2pOxvT28fvgBFmXKyOsoVUpMOzcNYXFhSX0Kr4zEsDr4unkpVPLUvRFlLNrmERT0JmbRK7SgIDGUKg5TwQuCxE8SLtMTbSlrlZ7oU6YrZbkWL15czswl0ZbEUxp2T+1EysIZ9OUlcZie+lM25Jw5czBhwgTh0yV90U+cOCFnm5QqVUqInyRSkmhLWSR0I0L9lKrIUmGs1KD1qI8k1qY15JCOh4Rr+kmCLUHZBCTMUjv1h8RfEnFJpJT2R0/6aZ20oAxUyhKQtill0vzwww9CuCXRmfjqq6/w/fffC/8wCbqJS83bS52UBUzSgwqb0HlPWThNvehJVjhy5Ag6dOgg5knAp5tQCcqc7tatW7rvl84J7ZsKhqkj2Vdkp1+p7efGjRs53g7DMIwuowgNxcsvkh8gUlZC4XnzYNeyRYH2S5+hB9xUOEqCRFqChFrKfuzevTsCAgJEpiH9vaMCkRRTZOVve27CsSvHrtocu6onDXDsyjAMY5gEb96C0N17khbMzFB01ar3kg9+uv8TLvomjSRWJtgj1u9TFLKzxLDmSclxugaLtnl1YtVSs/M70zarUPZqx44dxQ0E7ZfmyWstPdq0aSMEWRJTScikiTIiSYQlK4NWrVoJoZaqzlKBAbJOoOIBErQOrStB4ixlpLx+/Vr8JL9S2oc6lGIvCcMUEJIdgiTY5hTKpqUqzimFXxp2RhnA0jD+lFmf1O/0RFvJK5iyitXtBSj7RX34I2XYEOrbJ1uIMmpPjLQNynSmTFi6KZ01a5Y4f6tXrxavkeCqLuIWJHQe6fPEMAxjqERduIBXA78i5SWpwdQUpf86CpMsZGEyWYfErYyGBtIoH5r0IXbNSaZtVuHYlWPX7MCxK8MwjO4S+99/CFyzJmnByAiF58yBTf0kHUXidfhrrLy+Mvk9vp/DSJnkY2tnmTvaUX7Dom0ekVWLgpQBL2WqUrZhdoTb7EAWCdJNA2WxZgRlrl6/fl1k4x47dkxkiZBwR/YD9DScCkpRFiq9Rv6k//vf/4TARx61GUGiLfHnn38Kjy51aBi+JMTlJrRPysSlYVgp7RNsbW2zvV2pGm5ISFLFQvVjUB8qQDcfJGKr+wjntj0CZRmT8E0ZFOrZttktekKZ1iQ200T2FDSck4Z20uc2K/YItG8SstWhzz9ZbuRGMRbaDp1fhmEYQyPu6VO8XbBAHkImMDNDiZ07WLBlcjV2LYj4lWNXw4pd1a0+OHZlGIYxLBIDA/Fm5DdQvkvGcuzWDY5dkmoYqcch5GMbr0yqlRQf3AiKqLIY3qI0GpbW3doNLNoyAsqUpaCIgjTKjs3Uh8fUVHiP0kTeriQEkr8s2SLQdsjjlCYSdCkrlzxipaGBt27dElmokvhKhbAowKSiaJShScEhWRWQFUJqVKtWDdu2bRNWCrmRbUsBJ2WKknhIGbypUbFiRSE8q5NRAS8KbqkYCAmUlHFMUMYxTY8fPxZ+wMSBAwdENijtX7IKyO0hZlTwhM4V+Ql37dpVtD18+FCc56wWPUnpHUzbpUm6aciKPQLtm4JxuumgPhL0OaKheFRQI6fcvXtXHsrHMAxjCFDQGvLTz3j7/ffJ2bUkvFSsCI8Z02GVwrKI0S/o4TtNFNfoMxy7GlbsSvcXBMeuDMMwhhfXeo8Zi4Q3b8SyZeXKcJ+UZFmpzp/P/8Rlv8tiXpnggLiAtmhVoRDGtikPXYZFW0ZAYhsN/5fmM+LQoUN49uyZGGZEQdzhw4eFyFa+fHkRHFJwRYEeBXG0TJ5tFDhKkEBMhaGmTZsmCpGR6EuZvlR8irJ4yU92zJgxYptNmjQR3lrktUsZneQFR+tSBm+PHj2EHy89facglLI+qQ9ZhWwR6Kl/nz59sGTJEiHiUp/pOEggJssI8m0lEXrx4sX45JNP8Ndff6VrjSBBIjj58Y5Wq/ZLx06VcKkyLwW85JtLWcUHDx6UC2Zl1R6BMkpJgPXx8ZGDWilLgSY6R7RtEs5JGKdzOXLkSCGaZrYImcS9e/eErxn5mdF+qTgbHaeUWZMVewT6XNCN16BBg0ThORLi6frStZWEXSo8R5YbVLxNspIgnzOapCIvZHFBnx3K3pX2TeeWxOCMsn4ZhmH0BZVSiYDlKxC0caPcZupZGM5ffAmn3r1g/C5jjtFfsluITNfg2NWwYle636BRXnR/wLErwzCM4RCyaxeir1wR86YeHii6ZjWM1aw2ibdRbzHvUvI9f6zfJ3C0tMX8rlVhYpw/o9fzCuOC7gCjPZCIR1NmoKzavXv3omXLlkJ0I7Ft165doqgYbePMmTOiSBWJoSTMUjCpPlyKBLiyZcsK0ZcKcXz88cfCXkGCCpJRYQMSAyVRj+wSJHsFGrpF2Zhka0DZuPQkngok5CTrloJPCnzHjRsnhN/OnTsLuwcSAQkSNmkfVJCMPHnJ+oGOLSMo2CRRm4RnCTpWKjhC54C2NWrUKJE5TOK1+nnICpTxQAE7BekEiZ60TNdGYtmyZaJ4HGXa0rmngJiuozolSpTIsA8UpF++fFkcA127KlWqiOFu2WXHjh2oUKGC+FzQ54aE+o1qggMJuRTIq3vT0nHR8ZHYS9Dx0DKdB4k//vhDXL+0MlAYhmH0ifg33njZq7eGYGvbuhVKHTgIlwH9WbBl9A6OXQ0ndqX6GHTfwbErwzCM4RBz8ybezv9eXi48exbMUlgoKlVKTD83HRHxEWI5IawaFJGVMLVDRVGATNcxUmVUIYGRMxUocEkpasbGxuL58+dCTFQvNpUTCsLTNj/p16+fGA6/f/9+6Cspr+Hnn3+OWrVqiaxgbYZEURLEjxw5kqalAAXmdO0yGv6mDdDNCmWZ9OrVK9XX0/r+Uoa3NNyPsr/1CfpcSkI9DTWkz6g+os/X0BDg65c1FBERCFy9BsHbtmm027VtiyLLl+V7LJHf1y+9OM2Q4fg19zCE2DUl2ha7pnV/ZGixa17dfxo6HHcw/PnSvnvNxKAgPO/6GRL9/MSyc9++cJ8y+b319jzYg28vfSvmlYk2iHo2FlU9PPHH8MYw1uIs28zGr/p5t84wWsaiRYvE8DFt5/Tp0yKLQR88YAMDA8Ufip49exZ0V7QK+sOZkd8wwzC6Q+Q//8Bv9hwkvBteLDA1Fd61jp9/rpcPfxmGyXs4ds1/OHZlGEbXya17TXpQR8V0E98JtlZ1aqPQ+HHvrRcYE4gV11fIy7E+PQCFDcZ9WE6rBduswKItw+QDZDlA/rHaDg1Pk4ao6Tqurq6YOPF9g3KGYRh9IXj7T3ibwrPbpllTFJ45E2bvPMEZhmGyA8eu+Q/HrgzDMEmE/vorwg8kJb0Z29mhyJKlMErFCnPZtWWISHhnixBaC4qosvi8dlE0L59UIFMfYNGWyXe2bt3KZ12HoSFm2fUuYxiGYXKH4J07NQRby2rV4DnvO1hkoQgQo5+sWbNGTAqFoqC7ojdw7KrbcOzKMAyjO8Tcvg2/OXPlZffJk2Dm/r4I+zD4IQ48Tapno1JYIs6/A0q52WDmx5WhT7DRH8MwjAERHx8v37zQPMMwukfEyZN4qxbMOnzyCUrs3sWCLSMYPnw47t+/LwpSMQzDMAzD6Mq9piIyCt6jx5A5rlh26t0bjl27pmqf8P3l5AJlcQGtYaS0xaqeNWFroV+5qSzaMgzDMAzD6AjRV6/Ce/wEedmyejUUnj8PRlxwj2EYhmEYhtFhAletkus00Cgy98mTUl3v1OtTuPr2qphXJjgiIbQ+utQsisqeDtA39EuCZhiGYRiG0UMogA3cuBGhu/fIbZaVKqH49u0s2DIMwzAMwzA6TezDhwj++Wcxb2RhgSKLFqbqY5ugSBBethJxbzvB1Mgco1uXhT7Coi3DMAzDMIwWE37kCLzHjNVos/ngAxRdvQrGFhYF1i+GYRiGYRiGySlkd+A3cxbwzo/fZchgmBcvnuq6626tw8vwl2I+MboEEiMq4/PaReDlbK2XF4LtERiGYRiGYbSUyH/+0RRsjY2FvxcLtgzDMAzDMIw+EH74MGJu3hTz5iVKwGXgwFTXu+J3BT/e+VHMq1TGIsvWyswU49uWh77CmbYMwzAMwzBaSIKvL3ymTZOXzYoXQ/HtP6VaQZdhGIZhGIZhdDHLNmj9BnnZfeqUNEeSrbm5BiqoxHx8QCsoY4vim3Zl4W5vCX2FM22ZPOPkyZOoWLEiFO9S3HWdo0ePolq1arC2tkaZMmWwdu1aaBtUobFEiRK4ejXJlJthGIbRTeJfvcKL3r2hCAgUyyaurih14AALtkyGrFmzBpUqVULdunX5bGURjl3zH45dGYZhDJvwQ38i7vFjMW9VvbqwAEuNu4F3ce3tNTGviHNDfFALVCpsj8FNS0GfYdHWgOnXrx+MjIzem9q1ayevQwKg1G5iYgJPT08MHDgQISEhGW5/4sSJmDZtmnhfXnHmzBl89NFHol/Ux/3796f65GbGjBkoXLgwrKys0Lp1azx+90shswQFBaFbt25o1KgRbt++jUWLFmH06NG4ePFitvseGxuL4cOHw8XFBba2tujatSvevn2b7nv27t2LDz/8ULyHjvfmuyEEEubm5hg/fjwmTUq9yiLDGBsbo2zZsmKieYZhtI/IM2fw/PNuSPTxFcvGDg4o/tN29q9lMgXFFvfv38eVK1f07oxx7Jp5OHZlGIZhtP1eUxEWhrfz58vLLsOGCp0jNTbf3SzPJwQ3EXLmtE4VYWKc+vr6At+xGzgk0Pr6+mpMu3bt0lhnzpw5ov3Vq1fYsWOHEEq/+eabdLd79uxZPH36VAiReUlUVBSqV68uskrSYuHChVi5ciXWr1+PS5cuwcbGBm3bthWiaWa5fPkyIiMjsWTJEpFl26VLF9SoUQPnz5/Pdt/HjBmDgwcP4tdff8U///wDHx8ffPrppxkeb5MmTbBgwYI01+ndu7c4//fu3ct23xj9xdTUVHxGaKJ5hmG0B3rI6L9kKV4PGQplWJhoM/PyQsnffoVFyZIF3T2G0Qo4ds0cHLsyDMMw2n6v+XbefCiCg8W8XZs2sGvePNX1XoS9wImXJ8S8MtEOCWG18GEldzQq7Qp9h+/YDRwLCwt4eHiku46dnZ28TpEiRdC3b9/3hN2U7N69G23atIGlpaa3CImNy5cvF+Ljxx9/jE2bNqFq1apYt24dWrVqleX+t2/fXkzp3QDT/ijj95NPPhFt27dvh7u7u8jK7dGjR6b24+/vD3t7eyH4SlBWbHZFr7CwMHHsO3fuRMuWLUXbli1bhJ0EZe82aNAg1fd9+eWX4ueLFy/S3LaTkxMaN24srsHcuXOz1T+GYRgmf1HGxeH5p10R//Sp3GbbuhU8582Dib09Xw6GeQfHrhy7MgzDMLpPxOnTCPvjDzFvbGsL9ymT01x3672tspdtQlATmJtYYN6nVWEIsGibR3Q/1B2BMUk+dNmCPo/ZyPJ2tXLFnk57kFd4e3uL7ND69eunu96///6LXr16abStXr0as2fPFkIl2S50794dgwcPRmBgIJo1aybWoWxe8mFLj6lTp4opMzx//hx+fn7CEkHCwcFB9P/ChQuZFm1J/JWgjFsSn+lckDAtQVnIQ4YMSXc7R44cwQcffIBr164hISFBo18VKlRAsWLFRL/SEm0zS7169cQ1YBiGYbSfBB8fIdgqQkPlNucBA1Bowvg0h4gxjNbFrtmMXzl21YRjV4ZhGEbfIVsEvxkz5WUSbM08PVNdNzw+HH8++1PMqxQWiA+tjx61i8DVNvViZfpGgYq2NMyevEFJwKLh9/v27UPnzp3FayRoUXbk4cOH8ezZMyG0kcD1/fffC/9SieDgYIwcOVIIieSZQcPxV6xYITxCJciDlPy9yNvLzc1NrE9+q3kJBb3+0f7Qdg4dOqRxrlITRckfla4FFRQjSwESPJcuXZrudl++fKlxnYgNGzaITFHpGo8dOxZDhw4VbVLGKr0npU9rSpydnTN9fCTYEpRZqw4tS69lBfKIq1y5spgnKwOySpCgzGFJzCaRNzExURyX+g03ZSpL/SL/WUdHx1zpV0roPNI1YJjUCn7Q711iwoQJ4nPIMEzBEffsmRBsVWqWPe4zpsM5xYNPhslrOHY17Ng1LTh2ZRiGYXL7XpNsERIDAsS8TdMP4JCOTeThZ4cRq0iKk8kWwRRWGN4i+W+ZvlOgoq3kRzpgwID3vDyjo6Nx/fp1TJ8+XaxDha9GjRolgourV6/K65FXBgm+x48fF0Jv//79RfYmZXMS4eHhonATCb7kaXrnzh2xPxLLaL28grIGckQOMm2zQosWLYQ1QXqBJX3ZqPADCZGvX78Wgm7Hjh2F6J5WkbGYmJj3rBGo+Ne4ceM0skEJ8oeVoABYPZjUNqhv9PmjzxFZD1DfSfiWbCRoSk+0zS+o4Bp9hxgmNeh3JcMw2pFl8LL3F7JgS/61hWfPgk2jRgXdNcYAyXHsmoNM26zAsWvexK4FDceuDMMwhnGvqWGLYGeHwnPmpKmZqFQq/P749+Rth9ZD9zpF4eVsDUOhQEXb9PxIKbOWhNiUw+tJ6KMh9DSM/L///sPRo0dFBm2dOnXEOqtWrUKHDh2wePFi8eSbhqyT2r9582ah8tOTZnoaTpmiaYm2cXFxYpIg4ZdQKpViUoeW6YMkTRK7O+7O8QfdzMwsW+9V70dGkEdr6dKl092Gi4uLvA4FfsuWLUOjRo1w6tQpjeH96ri6uoosaPXtkAcZHZPURusQtC2pja6tlA2QFlOmTEnTHiHldZCyFCgzQd27l/xo6WFAZs+VtB71v1atWmKic9K0aVMxjI3sHuizRtkX6UGZ42SPQP2izyU9jFDPtqV+0WsZ9Ut6PeXxqlcMpqzyrHwWDBHp/KX8bkvf65Tfd31AOjZpXh+PUd+voSFgCNcv5s4dvOquadFT7OefYOrmpvPHnd/XT9fPl7aQU3ut/HpgTbFrRg/4KcaU1qEK1lTfoGHDhjh9+nS6sSvFZepQ7KqeJUSxFUG1AyRy29pLilcpJixcuLDcTstUBDerUP9r164tJjonFIdSbQQpds2stRf1i2LX0NDQ92LXjOpjZAa6b5DOL8MwDKOfJIaEwHfGDHnZffJkmKXzN+SK3xU8CH4g5hUxXjBK8MTXzfMmye/RFT84edjAzUs7HmbqpKctFW+iIFAKFMj7k+YlwZagQIxsEi5duiSeJNM6JKypB1xt27YVnqQUmFHRppTMnz9feK+mJCAgQNgDpBRX6WaBglSacgMKesmKgMjLoFcSbDLqd1rrkLdrWu+loPLevXsar5cqVQoPHz6U26gQGPH06VM5u7dQoUJChE8PWjet/dJ5U3/Ny8tLBJL0AKBKlSqyCE+fj0GDBmX6mknniq63dE3oc0bQzQlthx4WSH2XblZpnZT2CLQuCcYkAB87dkzOMqdzQ4E/PZjIqF/S62l97iibgvaRW59JfYXOD10nErnVH5JQG/2+oesoXWd9gW64aJSDVGBPX+0R9PkaGgL6fP2UYWGI3bETcb/8ktxoZgbbRQsRTA9U/LXfWknbrl9ERESe70OXWLNmjZikWJKBPDKMRoKlRc2aNYWVgDr0gJ5GikkcOHBA/CTRk2LWvLBHKFmypIhdT548KYu0Uuw6bNiwbD1AkWJRKdaRRsNlxR6BRF96P/WL7OjUY1cSxHPK3bt3xTVgGIZh9JeApcugCEjyz7f5gGwRkkddp4T+fq27lTwqPD64MVqUd8uTLFsSbE9svg9zK1N0GlkdHiUdoC3ojGhLYil5q/bs2RP276ooU/akFDBJkIBGgZHk+UQ/KfhRRz37MjXRljI5ybNKggIlEv/o6a+0b/V+0c0C7Vfytsotsptpm1noRopESCoEpg4dh5QFS5DAQ+tI9gh0fuhc0FP3tI6ZhPHt27drvD5w4EDhSUz2FHTeKSu6aNGiIvtUCvZofSrIlVlIOH7y5Im8TIEjBX30GaBsbIJsNUiIL1++vPgszJgxQwTYFHBm9prRuaLrTNnZ5IdM5422Q8ElHQNBx6T+eUovW5qyl+k80Lakz9U333wjzoN69kbFihUxb9482UKCshDoGH18fGTBm46Bgnv1LIdz585hzpw5uf6Z1Dfo/NC1peuhbudBNzl0g0PXRh9FW8pSIuj3pz6Ltvp6DQ0Bfb1+ITt2iGBVlUI4Kr5rFywrVYS+kN/XL6Udk6FDdRxooviVRq7pGzQaLqW3a8rYlWI2WkeKXaV4i0Z3pQXFrtu2bdNoo9iVYkiyCaMYjzJ2Ke6jWhqSzVdWrb1Sxq40YotEXyl2pe/O6NGj8e2334osYYpdyS6OYlfJWzez0Hkg6zi6h6K4lLJ9KXaVYsas2CPQZ4nOB90jUV8pdqU6IRS7qhfQpTiezllasSsJvUTK2JUK6JJ9A8MwDKOfJLx9i9C9e8W8sbU1Cn/3bbpJitf9r+Pq2yRrVGW8CxLDq2JAN01tL6coEpQ4s+cR7p9N+hsVF52Ip9cDWLTNKhRkdOvWLUlpT+G/mhfQUCiaUkI3HilvPqRMSmnKDdSfiOe1HyrZS6QsukDi5oMHSSnoxMyZM8VEUMBbt25dkSGqHhyn5IsvvhAB4qNHj8T2CMoOoKJyFORS9sfnn38u2kg8JV/iTZs2Zbn/VMSOvM0kJM/cvn37YuvWrWKe+kH+rjT8i4Z0NWnSRBw3eWdJNG/eXAwTk96TEroOtD69n4JTCtBJXN24cWOq1ygz15ACf8r8+Oyzz8QNCN0srF27VmN9Cmzppktqo5sECr4l6CEGQddn1qxZYp6yyynDic4vVx1PH+l7m9p3O612XUc9+1sfj88QrqGhoE/XTxERAf8lSxC6W3P4uW2rVvCYMR1mKQoO6ToKpQKPwh+Jh+T5cf304TPCZB6K4dRtA1KLXenBOk0pY1d6SJsWVCeDxF2KvaTYlWyv6AE5vT9l7EoiZHZiV/KXVY9dpUQR9diV+kFJE5QsoB67qj+gyCh2JaTYlURV9dg1u5BFmlT4WT12VYfOH8Wh6tnJ6rFrjx490oxdKSZmGIZh9JPA1WtoaLSYd/ryS5ilSMBMycGnB+X5uMAWqOzphIal0/47nlVioxJwaPUtvH2eZIdKlK5VCI26vG8fWpCY6opg+/LlS+Ghqp7pSk9naXhvyuHO9ERXenJLP8lrSR1pOTf8l3QZCvLSC/Sk4V/ZgZ7AjxgxQngHU+VdggRKWqZJHfVsg6xCAWtGvq10409ZpzSlBWU5UBZFelBGomTpkBtQ4C0NYUyLlMdGfcyonyQGU/E4dVGaYRiGKRiir16F97jxSFSLRazr1kXh+fNg/m6khj4Rr4jHN6e/wUXfi1hiuQStircq6C4xegTHrslw7MowDMPoClHnzyP011/FvJG1NZy/6J3u+mFxYTj8/LCYVyksRJbt4Palci0pjYTaY5vuIjwwyf7UxMwYdTuWQI3WxWBknP+F5NPDWBcEW/KSOnHixHtPxynjkZ4eU7alBAm7NCxP8meidc6cOaNRwY78TekJemrWCEzu8b///Q/FixfX+gIh5L1LQ7769OkDfRj6XrVqVYwZM6agu8JoKfSHjjJzaOJMbIbJOxRhYfCZ+j+87NNXU7CtUwfFNv2ol4KtX5Qf+v/VH+d8zkGhUmDyv5MRFBNU0N1imEzDsWv+w7ErwzCMft9rKuPj4TtturyO+4TxovBuevz26DfEJCbZiSWE1UQRB0d0rKo5yiY7UFLcv3se4bcFV2XB1sLaFJ+Or4Xa7UrAxFT7JNICzbRNz9OJhj3REJnr16/j0KFDYkiS5F9Fr1PWI/l9tmvXThSUWr9+vRBmKbuTht1IQ/579eolioqRBxMNkye/0xUrVojhPUzeQkXiMlsptyCpXLkybt++DX2AvhfTpk0r6G4wWgz5LGeUrc0wTM6CwYgTJ/B2/nwk+vjK7Va1asF94gRYZaP6u7aTqEzE1ntbsfH2RjnANoYx5jWZBxer3BvGxjB5Dceu+Q/HrgzDMPp9rxl+4AAS3vmaW9erB8fu3dPdRkR8BJZfXy7mVSojJIQ0Qp9WxWFqkjNBNTFBgYv7nuH26Tdym7OnDdoMqATXopnzdzc40TY9TyfyOJIqtEqVUyVOnz4thsUTO3bsEEJtq1atZI+llStXyutSBiV5WFFBBqp6Sj6s5HFFHlEMk1kyY0vAMAzDMN5jxyLiyFGNE+H69TC4DB0KYz0s/EfZtX2O9IFvVLJAbWpkipk1ZqJVMbZGYJiCgmNXhmEYpqBRxsUhYNVqedn1669hlEEtgr2Pk4qVEYnhNWCq9EC3Ol456kdEcCz2L70uZ9cS9T4qiVpti2tldq3WiLYZ+ZFm5FUqZd3u3Lkz3XWqVasmKpIyDMMwDMPkFcHbt2sIthYVK6LIkiWwKJW7lW61hR3/7cD3l7/XaKvoXBFzG82FQ4JDgfWLYRiGYRiGKXiCt26TbcJsmzeHTYMkG9P0NMD9T5LrCMUHN8bHVTzgZJP9xIfwoBj8uea2hmBb/5NSqNO+BHQBrS9ExjAMw+SudxwVqyNGjx4thiUyDJNzgrZshf+CBfKyc98+KDRhAoxM9S/U+uXhL1h9YzVC4kI02je22YiGng2Fl33KQrEMwzAMwzCM4dxrjvjiCwS9K0oPY2O4ZaLuzqOQR3gSmmShqogtDFVcUQxuWjrb/Xl63R+ntv+H+FiFWLayM0O7wVXhWdYRuoL+3UkwDMMw6RIdHc1niGFyCcoICFy7FoEaQ7+GwXXkSL0r9nfB5wIWX10sAmp1qrtVx5zGc1DKoVSB9Y1hGIZhGIbRnnvNgDVroXw379i1KyzLl8vwvUeeH5HnE0LroUOVwqjkaZ/lPigUSpz75THu/OMtt1HBsc5ja8G5sA10CRZtGYZhGIZhskng6jUIXLNGXrZu0EDvBFv/aH9M+XcKLvtd1mi3M7fD5HqT8XHpjwusbwzDMAzDMIz2edmGHTgIM0qytbGB6/CvM5UIIYm2KpUxEsOrYkTLMlned6h/NP764S4CX0fKbUXKO6Flnwqwd7GCrsGiLcMwDMMwTDaIOHVKQ7B16t0b7lMm65VgGxYXhk77OiEmMUZuc7Z0xrg64/BRqY/06lj1hTVr1ohJoUgaCsgwDMMwDJOfxD97BlViImBkBIeun8LMwyPD99wMuAmfKB8xr4gqg3rFiqFi4axl2YYFxODAipuICEryrzU2NkKzXuVRsXFhnY1ZWbRlGIZhGIbJhmDrM2GivOwyZAgKjRmtV+cxThGHoceHagi2zb2aY2HThbAy1b1MBUNh+PDhYgoPD4eDAxeEYxiGYRgmf4l9+FCed+jQIVPv+fPZn/J8QlgNdP2waJb2+fp+MA6uvgWVUiWWbRzM0X5oNbiXzLq9gjZhXNAdYJjMQk9G9u9PriSorQQFBcHd3R0vXrwQy3///bfoe2hoKAyFo0ePokaNGqIYDcMwjD5BQ7cCVq3Gm6+HQxkVJVsiuH0zEvpEVEKUsES4G3RXLJsamWJli5VY1XIVC7YMo4exa6FChTh25diVYRgmx8Q+eoSEN2/EvJmnJyyrV89UfH369emkeaUJjKOr4sNK7pnaH7335olXOLRGTbB1tMCnE2rrvGBLsGhrwPTr108Ekymndu3ayeuUKFFCbjcxMYGnpycGDhyIkBDNitFMMt999x0+/vhjce7yktjYWJFJ4+LiAltbW3Tt2hVv375N9z30Ol13uo7W1tbiWj9+/FhjnadPn6JLly5wc3ODvb09unXr9t521T8X0vT999/Lr9N2zczMsGPHjlw+aoZhmIJDFR+Pt3O/1bBEsKpVC0WWLIaRiYleedh2PdAVx18eF8vmxubY3n47WhRrkaXtUBDNMLkJx655F7t+8sknWh27FilSRGSOt2/fnmNXhmEYLSMxIABhBw/Ce/IURBw7Jre7jRmdKVuCJ6FPRPxJKKJLo2V5Lzham2f4vvjYRBzffB/nfnsCpSIp7ixW2QU9pteDvat+jApj0dbAIXHN19dXY9q1a5fGOnPmzBHtr169EiLcmTNn8M0332RpP/RFlTJP9b1S4qZNm4SwnVnIcy47GaljxozBwYMH8euvv+Kff/6Bj48PPv3003Rvnjt37oxnz57hjz/+wI0bN1C8eHG0bt0aUe+yxejnhx9+KK7XqVOncO7cOcTHx+Ojjz56r4/S50KaRo7UzDKjAHvlypVZPi4mb6FrS6I9Tbrq68MwBYFKqYTv9OkI2blTbnPs2QPFt2+DqYuL3lwUhVKB9r+3h3ekt5xhu7jZYlR1q5qp90fFJWL8r7fRedMdzDvyII97yxgiHLsaZuxKGcuXL19GsWLFOHZlGIYpIFQKBSJOnEDg+g142b8/Hn/QFI/qNxA/yTYs4s8/UcjUTEy2TZrA4aOPMrXdc97n5PnEqDL4uIZnhu+JDInDD2PO4PGV5Id/lT7wRMevq8LShkqg6Qcs2ho4FhYW8PDw0JicnJw01rGzsxPt9IS7RYsW6Nu3L65fv56n/aLsz6ZNm8LS0hKVKlXC8eNJ2T7qvH79WmSBOjo6wtnZWWQIpBSGN2/ejMqVK4vjLFy4MEaMGCG/tnTpUlStWhU2Njbw8vLC119/jcjISFm8pCzT3377TWN7FDDS+hEREan2+/Dhw2JfDRo0SPPYtm7dKvp84MABcWy0PgniWSEsLEwE2HQMLVu2RO3atbFlyxacP38eFy9eTPOc0mvr1q1D3bp1Ub58eTEfExMjC/Uk0tI5pD7SuaFp27ZtuHr1qhBxU/tcSBOdF3VI6KX3UeYuoz1QBvTgwYPFRPMMw2QMCQfeo0Yh7I8DYtnIzAweM2eg8MyZMDLVn/IA9wLv4ZM/PkG8Ml4s25vbY3en3VnKsP3fvjvYe8MbfhHxePw2uWovw+QWHLty7MqxK8MwTN6R4OODkD2/IGDlKvhMmoTXQ4biRfceeNKyFR5UroI3I0YiYPlyRF+4KDJsFWFh8nvNjIzQzdER/Tt2RLF532V6nxd8kzUMy8QKaFWxULrrhwfFYN+Sa8C7QV3GpkZo2aciWvSuAGMT/ZI59edOQ8v4Zd4VRIcn3fRkD/r0ZT0LztreHN2m1kVe4e3tLZ6Q169fP8/2QU/u6ak7+cJeunRJCJSjR2sWd0lISEDbtm3RsGFD/PvvvzA1NcW3334rsi9u374Nc3NzIUiOHTtWDNunoVS0HRIlJYyNjUUmaMmSJcUTfBJtJ06ciLVr1woBskePHkII/eyzz+T3SMskWKYG9YUE1MxkNSxYsAA//vijGCJGPmKUxTxkyJB033fkyBF88MEHuHbtmjgHlCUrUaFCBZF9cOHChVRF47i4OPGThHD1c0A3P2fPnsVXX30l1qHsS2qToPVpPVpHfX90XufOnSv22atXL5E9QddBgtrpGtI5KV26dIbnhGEYRlsFW18a6nX8hNxW+Ltv4fDxx9AXXoW/wpKrS3DqtebDuf/V/x/KO5fP1DaUShWWnXiE/TeTqv4SpiaczW9YsWv24leOXTl25diVYRgm/1ElJCDm5k3E3L6NuOfPhQAbe+8+FIGBWd6WRcWKsCxXFrbNmsG6Xj2Yurpm+r1U8PaK31Uxr0y0RcsS1WBhmrbtmN+zMBzZcAfRYckxS6fh1eFV0Rn6CIu2eQQFvVGhSSKZNnPo0CHhKaXO1KlTxSQxadIkTJs2TQyFIi8qEmwpwzOvOHHiBB48eIC//vpLDOEm5s2bJ4RXiT179ghxl0RPaYg3CaqUwUqFv2iIP4m448aNw6hRo+T3UYaphLoQTB5etP7QoUOFaEuQiNmoUSMx9J+ydP39/UUmLfUvLV6+fCn3OT1IcKX9VFcz5SYf3IzEcMp2Jvz8/IQwTcerDomk9FpqSKLulClTsGHDBiFML1u2DG/evBHHSJDYS+10zemck1gxefJkce2ldQiyx6hVq5bIcKbsXtomvZ7yc0Hngs4JwzCMrvJ60GBEnT0rL7sMGaJXgq1flB/6HOmDoNggua2EfQnMbjQbtdxrZWobzwOjMOm327j8Ilhum9q6OL5qWSlP+szkDRy7Zh+OXfM+dl2/fr1IKliyZAnHrgzDMFmA7unj/vsPbxctEs9WFcHBIjs2MQNP8ZQYOzgISzBT90KwqlIFtk2bwqx4cZgVSj8rNiMu+FxAgjJJO0uMqIjWtT1SPw6lChf/eIbrfyXrCzYO5mg/rBrcS+h+wbG0YNE2j6CsgZyR/UzbrEB2B5SRqg4JcepMmDBB+JPSl50sCUjQ7dixo/C2peJkqUECK2VYqkM2BZLASl6q9+7dS/W9//33n7ArUBc/KaNWnVu3buHJkyfvZbySqEzD8UlgJZ+sVq1apRtgz58/XwjE4eHhSExMFO+nLFgq0lWvXj3RZ7IHIOHy559/Fv0m24a0oOFa6pmsaUFBa7Vq1TTa6FjSyuDNDWgo/N69e4VnGV1junaUOUvXSioWQ8XHyGds2LBhIguZMmx79uwpBFqal6AMZgk6DjoeyhKm86mepWtlZSXOJ6M90AODNe+KKFExELZIYJi0Cfn1Vw3B1mP2bDh176Y3pyw0NhRfn/xaFmyp4Fjfyn0xuNpgWJpm/LeMOHDLBxN/u4XYhGR/y1GtyuCjyvobPOsrOY9ds59pmxU4djXM2JVGpnHsyjAMkz50Xx97/z5i79xF3KOHiH3wEDFZsbY0M4N5kSIwK1IE9p06waJ0KZgVLQoTOzthD5YX95qnXiWP9EqMrIymZd1SLTj255rb8HkcKrcVKmGPTiOqwco2N+IX7YVF2zwiJxYF9EWjIIyGmud1oSDKqixTpky667i6usrrlC1bFsuXLxci6unTpzWGy6tDGbAUBErQ+yhLVXranlOhiLxnyYaALAVSQsKjusCYGuTb2qlTJyFOUsVcEjFp+D8FhVR4iwJfKduWfulQ4EuZvP3790/3mtC5CgkJybD/JGam3E5W7BHIQ5b6GRoaqpGxQBV26bW0oHN28+ZNYRVB76dzRdm9derUkdehLGUSvgMDA8VnkLZP2yxVqlSa26Vt0GeWzit55UoEBweLfTDaA/1+oc+NNM8wTOokhoTAf/ESednx88/0SrCNU8Rh5KmReBzyWCwXsS2Cn9r/BDfrzP3Opt8fcw/9h83nnsttxZyt8X3XqmhQ0lk8PGV0i5zaa+VX/Mqxq2HGrvQ+Eqdp9BuNDOPYlWEYJulvryIoSAi1kX//jfDjx6EIyNjewLRQIRjb2MDUzU1YGlhVqwqzYsXEslEGWkpu3msmKhPx9+t/ktZXmKOsXU042WiKsJEhsTi46haCfZKKpxPVWhRFw09Lw9QsbRsFfYFFWybLSNm16qJsSiRxVh160k82BBlRsWJFkdEr2RIQKYtrUeYnWSSQFywVDEsN2tfJkydFRkZKyBOW7BVoiJUk8P7yyy/vrffFF18In1vKOr1//74owpYeNWvWFFkN2SEr9ggUwJLwTcfXtWtX0fbw4UNR0CxlVnJqODg4yMXJqFgYedOmFsQTVICMbr6pf2lBwTSdR7oeKbOe6ZwwDMPoGn4zZkD5rrCCXbt2KJzK70ldZv2t9bgZcFPM25nZYWHThVkSbIf+fA1/3UseVteglDM29qkDe0uzbFWVZ5i8hGNX/YhdSbDn2JVhGEMniBLk7twVYm38mzdITMNiRsLE1RXmXl4wcXCA42ddYduqVZ4nB2aWG/43EBafJPImRpVHw1LuGq/7PA7B0Y13ERORIJZNLUzQpn8llKphOIlhLNoaOFR4KqWPFGVISIIdERERIdaR7BFIxKTsSfJ7zQsoe7dcuXJCIF20aJEY/vW///1PY53evXuL1z755BPMmTMHRYsWFd6pNISK+kfLs2bNEh61JCSSBQAdBxUiGzlypMgcptT9VatW4aOPPhLt5JWVEicnJ1EUjSwiKAOVtpseVByNfLcoYyGrw8WyMsSMAlfKrCCbAsq0IOGajouCXvUiZOQFRpYFXbp0EctkfUDXjvzB7ty5I/x+O3fuLI5NgrIySDin9agwBK1DRcakDFpqowJxJIZTf2mZXieBm86XBAntZJWQmUCcYRhGmwj55Re58JiRlRUKjdEshqnLRCdEY92tddh6b6tYNjUyxQ9tf0Bll8qZev+roGj023IZzwKjNATbnV81gLGxdtwAMPoNx655F7uqx3HaFruSdRolCVC9Co5dGYYxNFRKJYI2bULguvVQZWA/aFW7NuxatoBFuXIwL1kS5hn8HShITr8+Lc8nRlRG11rJfX16wx9HN9yVl22dLdBuUFW4lzQsCy4WbQ2co0ePytmsEiTOkVeWxIwZM8REUNBExbyOHTsmvKXyAsrY3LdvnwjsyJuLMmYp07Vdu3byOjQEjDx1qWAWBaYkyNKTfPKwlTJvSfSlbE8qtjV+/HghRH/22WfiNSoARkWzFixYIAJV8vqiALFPnz7v9Yf6sXPnTgwYMCDDvletWlVkAVPWLr0vL6HjonNF2Qp0A0NBt1RETYIyGMgKQYKylylYpqFodN3peKdPn/7ee+ickLUBnXsSzEmUlSAhdvfu3UIUp/2WLFlSvK7uc0vs2rVLiOvScD2GYRhdIGTXLvjNSc6qdZ80EebFi0MfCIsLQ98jffE07Knc1q18t0wLtne9w/DFpksIjU7KdiCal3fD1v718qS/DJMaHLvmXeyakdWBNsSuX375pXxfov4ejl0ZhtF1FJGRUMXECIsuyp5N8PVD3JMniLl5E7F37qT5PssqVWBZsQIsKlaETf36sChdGroAJQWeePH3u3kjlLSujSpFkkYE3z3jjTO7H8nrepZ1RPshVWFpmzObTV3ESMWmhhlCmZ70dJgCiJRD8UkUfP78uRCuMmPir22etkzG/PTTT0KUpMJmVIQhI/7880+R3XDjxg2xviFeQ/LDJfGfrBfou6GtpPX9paG9ZAlBWdoZ+SPrGuQnN2/ePDFPRQUz85nWRfT5GhoCBXH96G+v/6LFCN68WW5z6tUT7tOn68Xv8edhzzHgrwEIjEn2OWtXoh1mN5oNazPrTAm2vX+8hLCYJMHW1dYcHaoWxuT2FWBtblqg1y+9OM0QIT9TmhQKBR49esTxqwGS3dj17t27Wvs3M6/vj3Qlds2r+09Dh+NGJj8/X4rwcITs2g1lTDQS/QOEzUHM7dtQZMJfnKDsWRJp3adNg2mKIvK6dK95L+geehzqIeYTo0ri08Lf4bsuVfH6QTAOLE+y8CIKl3bAR9/UgJmFfvnXZjZ+5UxbhkkDKnZAT/e///57kXmQWXGrY8eO4ibJ29tb64O+vIIKklHmhKEeP8MwukfA0qUagq3DJx/rjWB7L/Ae+v/VHzGJyV70i5stRtsSbTP1/sdvI9Bj40VExiWK5RpejtjSr+57hSIY7YCqNdMk3QwwhkNOYlfyiqXYlWwIDBGOXRmGyWsUYWF4M2Eios6ezfJ7TQsXhkWZMrBr3VpvCuOeeJlkRUYkRlTF4O6l8PjqWxz78Z7cXrqmG1oPqGQQBcfSgkVbhkmDhQsXiuq8NPyMhlxlhdGjR4tsAEOFKvqqV/VltAcSoMjmRJpnGENHFR8Pn8mTEX74iNzm8MknKDx/nl58R448P4I5F+bIgq2xkTFWtVyFpkWbZur9j95GYOC2K7JgW7eEE7b0rwdbCw4hGUbfYldDhmNXhmHyArI6iLl3H9GHDiFk//501zVxcYFV1aowtrWFqYsLTFxdYO5VTNgfmBd9v9C7rt9r/v0qWbw2iqqOqAdhOP1Tsk1nscrOaPNVZZiYaOcIkPyCI26GSQPybKWJYfQJqtxMGVgMwwDRV6/Cd9p0xL94IZ8Ot1HfwHXYMJ0/PfGKeCy6sgi7H+6W26q5VcOSZkvgYeOR8fsTlVj/z1OsOPkYCqVKtLnbW2Bzv7os2DKMlsKxK8MwTMFbuSS8fImIU6cRvGULEgMCUl3P1LMwHDt3ht2HH8LU1RUmjo4wMjU1mHvN4NhgPAl7KOYVsYUxqmQ5DcHWo5Q92g+tavCCLaH7nwqGYRiGYZgsBtRvv/0OITt2aLS7T50K5z5f6vy5jE2MxdSzU3H85XG5rVahWljdajXszDOu9E5ZtV/8eAk3X4fKbeXd7fBj3zqwszS8AhAMwzAMwzAZEf/yJV4N/AoJb96kuY6JgwO8fvwRVlWrGPQJvehzkSJymCrMUPtFFyAkWH6t8geeaNarvF6MeMsNWLRlGIZhGMZgUCUkwHvsWEQcT/bRMnVzQ9HVq2BVvTp0nTcRbzDy1Eg8CX0it3Uu0xnTGkyDhYlFpgTtbw/d1xBse9UvhhmdKsHSgP3EGIZhGIZhCEVkJJQREUgMpAJit6AMD0fkmX8Rc+NGqifItlUrKOvVhUfLlrAwUN/wlPz1/B84xLihzaN+cI0uKrd7VXTCBz3KsWCrBou2DMMwBkRCQgI2btwo5gcPHiyGsDCMoZDw1h9vhg9H7N27cptNs6YoumoVjDNZsEdbiU6Ixrh/xuGsd7I/mKmRKb5t8i06luqYqW2QYDv74H3svvJablv/RS20q1I4T/rMMAzDMAyjjcS/8YYiMAAKEmf9/RH/+jVi790XMaQiJCTD99t37AjbFi1g+0ETGNnZwd/fH2aFCkHfycy9JsWbd5//h853R8Mq0Va0UVJtvY9Konb7EizYpoBFW4ZhGAOC/kgGvPNWonmGMRSUcXF43rUrFIGBcptjzx4oPHMmdJ17gfcw8cxEvIp4pdFOdgiNizTO1DbIt3b2wXvYfuGl3Db/06os2DIMwzAMYxDE/vcfYh88RNgffyD6Ig3fzzpGlpYoNHasht2WUqmEoZCZe82HQY/Q9lZ/WbC1dDBHxyFV4VHKIV/7qiuwaMswDMMwjF5DRSBejxghC7ZGZmYoum4dbJtkTtDUZrbd24bFVxdrtFVwroCVLVaisG3mMmQP3/HFhF9vISpeIbct/KwautXhIXwMwzAMw+gfJCjG3r+PuP/+EwVpoy5dRuydO5l6r5mnJ8xLlYKpizNMqIiYnR3MS5eGVaVKMPX05EzRDPj7wF3YJDiJ+SjTOPSf1hjWdhlbeBkqLNoyDMMwDKO3xD1+jFeDhyDR1zepwcQExXftglWVytBlEhQJ6Haom4Z3bQn7EphSfwoaeTbK9Hb23/DG6D03NdqGtyjNgi3DMAzDMHpH3NOnCNq0GTHXrwuxNj2sataEdf16MHN3h4mzCyxKl4J58eLi4T+TPUL8oqC4nCTYEv5ljVmwzQDjjFZgmJyydetWODo6at2JpGqE+/fvT/P1v//+W6wTGppcjIXWL1OmDExMTDB69Oh86inDMAyTHSJOnsSLnr2SBVsARRYv0nnBNiQ2BJ8e+FRDsG1etDl+//j3LAm2Oy+90hBsyxSyxdretTChbYVc7zPD6BIcuzIMw+gPCT4+8F+2HA/r1cezjp0QtndvmoKtZeXKKLJ8Gcr88zdK7NqJQqNHw6lnT9i3/RAWZcqwYJtDTuxKzma+43ITdRpWzOkm9R7OtDVg+vXrh23btol5U1NTFC1aFJ9//jnmzJkDS0vLXNtP9+7d0aFDB+gajRo1gq+vLxwckr1VhgwZgv79++Obb76BnZ2dOIck6qYn/jIMwzD5izI+Hr6TpyD88GG5zaxIEXj9sBEWpUrp9OWIiI/A0BND8SI8+WZjXO1x6Fu5b5aG4/11zw9T9yUHzi3Ku+GHPnVgasLP84m4uDhYWPBQPW2DY9f04diVYRhDJDEoCCG7dkMZFQVldDSi/v0XqsREqOLioAgLS9d/1rpWTdh37ASLCuVFFq2JbZLPKpP7BL6JhP/DaDEfbRaBsw5+mFvRnU91BrBoa+C0a9cOW7ZsEVX+rl27hr59k276FixYkGv7sLKyEpOuYW5uDg8PD3k5MjJSVH1s27YtPD09C7RvDMMwTNoEb92mIdjaNm8OzwXfw0TtIZwueq9tvbcVP9z+AREJEaLN1NgUsxrOwidlPsnStu77hGPIT9fk5Z71iuHbzlVgYpx50VffOHLkCHbv3o1///0Xr1+/FkVDbGxsULNmTXz44YfigS3/7dcOOHZNG45dGYYxhMKyCa9eiXoF0VevIXDt2ixvw75DBzj16insD4xMTPKkn8z7XNz/VJ6/VvQoSjs3ght72WYIp1PkIfHx8WlOiYmJmV6XBNXMrJsdKIuEhEkvLy907twZrVu3xvHjx+XX6aZl/vz5KFmypBBeq1evjt9++01jGwcOHEDZsmVFdm6LFi1E9q66rUDKIWazZs1CjRo1sHnzZhQrVgy2trb4+uuvoVAosHDhQtGfQoUK4bvvvtPYD23vq6++gpubG+zt7dGyZUvcunUr3fM/YsQIFC5cWPStePHi4ljUCQwMRJcuXWBtbS2OgY4lNXsEmqfMWoL2S+3NmzcXx/rHH3+IZZpoPYbRZuhzSt9HmrKSlccwuoBKqUTQ5i0IWLlSbnOfMR1F163VacE2JjEG9XfWx9JrS2XB1tHCEXs67cmyYEseth1W/isvNynjiu8MWLDdt28fypUrhwEDBohRR5MmTcLevXvx119/4ccff0SzZs1w4sQJlCpVCkOHDpUrIusrHLty7MowDKONxNy6hefdu+Nh9Rp49tHHeDVgYIaCLY2ysihXDlZ1asOhSxcUmjgR5W/eQJGlS2Bdpw4Ltvl4r+n9MAQv7waJ+UjzEDx0u4wO5erlRRf0Ds60zUPmzZuX5mskEPbu3VteXrRokSzOUjYNiaXGxsbig16iRAkxHExi+fLliI5OSitXh8TQnHD37l2cP39eiJsSJHL+/PPPWL9+vejzmTNn8MUXXwjhlG5knj9/js8++wyjRo0SguqNGzcwfvz4DPf19OlTkdVy9OhRMU/bePbsmbhx+ueff0Q/6AaKROT69euL95B1AwnH9D6yLNiwYQNatWqFR48ewdnZ+b19rFy5Uoiwv/zyixCHKXOGJnVmz54thGI6/6tWrRLX5OXLl+9tj4abPXz4EOXLl8fvv/8ulknoHTRoEMLDw0W2MpFaPxhGmzAzM2M/ZkZvCd6+Hf4LF8rLlEXh3KsXdJnQ2FB8dvAzIdxKeNl5YV3rdShun/z3OjNcfRGs4WHrbGOOxZ9Xh7GBCrYExQDLli1D+/btRdyVkm7duomf3t7eIk6gmGjMmDHQV7Ibu6aMX+lhP8euHLsyDMPkBGVsLKIvXULg+g2IuXEj3XXNvLxg37EDbJs2JekQFmVKw8Teni+AltxrPrrsJ89fK/oXEoyUaFWBRy9nBhZtDZxDhw6JTFfK/CX/Ngq0V69eLV6jZQreKcOkYcOGoo0yTc6ePSsEUxJt6ScJmRS4EzRP4m/KLNmUUFBPmbaUvVqpUiWRoUui6OHDh0UfaDtk0XD69Gkh2tI+L1++LOwJJI+5xYsXCy9ZyvwdPHjwe/t49eqVuMFo0qSJEL/VxWgJuqHo2bOnmKdjJaGX9kND71ION6PsX0mYlWwTSESm86Ruo8AwDMPkP7H37yNg6TJ52anPl3CfNEmnL0V0QjQ6/9EZQbFJmQnEp2U/FZYIWc2UfxkUhR4bL8rLVYrYY+OXdeDhkHse9rrIhQsXMrVekSJF8P333+d5f5iM4diVY1eGYfSXxJAQRJ46hahz5zWsrtQxcXSEfadOsKxQHpbVqol6BUamLG1pK0qlCs9vB4r5BOM4PHa9CqOIxijtxv7BmYE/2XnI1KlT03wtZTbHhAkTNDIVSESlYXrSsHt1UntykV1ILF23bh2ioqJEpgnts2vXruK1J0+eiIzeNm3avDd0jjzeCBJa69atq/F6vXoZp7lT9rBkN0C4u7vDxMRE47xQG4m0BNkgkKesi4uLxnZiYmJEpm5qkCBLfScBmETYTp06CV86dapVqybPk3cd2S5I+2QYhmF0g5g7d/Cq/wCo3lkFkWDrkc7fYF0gQZmA7oe6awi2q1quQnOv5lne1vH7bzFo+1V52dHaDL8NbQRLM/ZxSw2Kc2gkUenSpUVcZEhkN3ZNGb+mXJdj12Q4dmUYhkm9iGzQho0I2rwZqpjk0UXqmLq5JVkbpNAfGO0m8HUEYiKSRua8dnyARJMEFLUuy1Z9mcSwItF8hrIzs7MuBb0U7EqibU62mxEkVJYpU0bMU+YredZu2rQJAwcOFCIp8eeff4oME3VyWlGZ0ubVoeNMrY0ycgnqC3nTpuYZq+6Xq06tWrXETRfZKVC2MA1xJLsFdU/e9PbJMPoIDWWV7DyosE7K7wDD6BL09zJ40yb4L14it1lWr4ZCmbDp0WYCYwIx6cwkvAh/IbfNaDgjW4Lt7suvMGXfHXnZ2twEvw9jwTY16EH1yJEjhV89QfZLNMKI2igOmjx5MvSd7MauGcWvHLtqwrErwzBMEnGPHyP62jUErFgJRUiIxmkxsrKCeYkSsCxfHk49e8CqenU+bTp4ryl52RK+9kkJd9XcKhdYH3UNFm0ZGQq0KcNi7Nix6NWrl7AtIHGWbAbICiE1KIuVLA3UuXLlSq6fVRJg/fz8xI0AZelmFsqc7d69u5jIN5cyboODg3PNe5ZuQqiAGsPoCnRT7ePjI88zjC4XHXvzzTeIPHFSbjMrWhRe69fDOBcfbuY3T0KeYNDxQUK4lZhUdxI+L/d5lrYTFpOAqfvu4M/bvppFx7pUQXEXm1zts74wZcoUMbKHHhCr2yTRA1+qG6Arou2aNWvEZAjxCceuWYdjV4ZhCgKVQoHY/x4kibSXLyPi5Ekow8NTXZcyaZ3794NNw4YwtrLK974yuXevqVQocefvN/LrPvZPoEy0QT2vcnyaM8n7FRcYg4aKfZFNAQX7ZF9ARcWo4AZlnZANwfXr10UhDikLZciQIXjw4IGotkwZKVT0a+vWreK13KxMTzdM5KvbuXNnHDt2DC9evBDFyv73v//h6tXkIZ/qLF26FLt27RL9o779+uuvwns2reyG7EAC8u3bt4VNRGBgoEZBDoZhGCbvClN4jxunIdhaN2yAUn/sh6mTk86e9j0P9uDzQ5/Lgq2ThRM2tNmALyp9kaXt3PMJQ9OFpzUE2w5VPbC1f10WbNOBfPLJ11/ywpeoXLlymlZM2sjw4cNx//79PHmIro1w7Jo1OHZlGCY/SXj7Fq8GDMSj+g3w4rPP4DtlCsL27UtVsDXz9ETRNatR/KftsGvZkgVbPeDti2RrhFcO/yHIxhuK6JKoU1zT9pJJG860ZTQ/EKamGDFihKimPGzYMMydOxdubm6YP38+nj17JgRPynqVPM+oOjDZDYwbNw4rVqwQwioJqfTenFooqEM3T5TRS9umNPuAgAAhwDZt2lR436YGic50HI8fPxZCNHnvSoXOcotBgwaJjJw6deoICwcqnNa8edaHrzIMwzCZQxEejhefd0P8y5dym/uUyXDu21enT+Gam2uw/tZ6ednTxhPr2qxDKYdSWdpOTLwCw36+LjJtCWMjYELbChj0QUmYmvCz+vSg2EIqOqoO+f7n5oNoJnfh2DVrcOzKMExekBgcjJhbt6CKjYUyLg6KkFBEnT+PqH//Tfd95mVKw7J8BTh+/jms69TmgmJ6xqt7ydYIT11viJ+JkRU4iSALGKl4fGyGhIeHw8HBAWFhYWK4vTqxsbHCN5XES0vL3KnAnLIQma7x3XffYf369Xj9+jUMFV2/hoZCWt9f8jWmgnR0856bIr+2FNiZN2+emKeHL7npM6hN6PM1NATSun5UpOLll18i9tZtuc1z0UI4fPQRdBWFUoFt97dh2bVlclv7ku0xq+EsWJtZZ+nvzq9X32DxsYfwj4iT2//8pgkqezpAn79/6cVpWYEeBFPWJnnY0oNfGklDfx9omR4AHz16FLoEx6+Zh2NX7Ydj67y9/zR0OG7MutVB/IsXCNywAYrAQCjCwhF7716G77OqVQvWtWvBokwZmDg5wapWbZjY6r9lkyF9vlLea/6x5Bb8X0aI5e21pyPaPBz1sB6b+jaGoROeyfiVM22ZHLN27VqRxeri4oJz585h0aJFIluXYRiGYXIzw/b14CEagq3Xph9h27ixTosQcy/Oxe+Pf5fbXK1cMa/JPJgaZz5Ee+Ifgen77+HCs+RsBuLngfXzXbDVZegmo3379sJagB680ggimic7pn/++aegu8fkIhy7MgzDZJ3o6zfgN3s2Et68gTIqKnNvMjaGTaNGcBsxHFY1avBpNyAigmNlwTbQ2lsItopYD3zZvGxBd02nYNGWyTGUffLtt9+KAl/FihUTVglUzINhGIZhcoPYh4/gPXYs4iVfUSMjFF23VqcFW2L7/e0agm2fSn0wvs74LI3QOH7/LUbuuo7YBKVGO/nXNinrmqv91XfIy/bmzZv4/vvvUbVqVeGhT5ZQFy5cEMuM/sCxK8MwTOZQhIYi6tJlBG/dipgbScPbU8PY2hrWjRrCunYdGFtawMjKCmYehUVmrZGZGZ9uA+TxFT95/oXTHfFTFeeBOiVypyi8ocCiLZNjli1bJiaGYXQDa+vMD7lmmIImMSQEzz/5RF42cXRE0bVrYV2rJnQVpUqJDbc3YO3NtXLbtPrT0L1C9yxtZ+/1N5i89w7iE5MEW0drM/RtWALDW5SBual+D7/LK0qXLo0ffvihoLvB5DEcuzIMw6RuexB24KDwoY2+eQPKsPA0M2ptGjWEeYkSMC9REk69e8HIxIRPKaNxr/n4ir98RvztXomfzubFYGvBMmRW4LOVS7A1MMPoHob4vSUP24kTJxZ0NxgmU8S/eYPnH7bVaNN1wfac9zksvLIQz8KeyW29KvTKsmD747/P8O2f/8nLdYo7YeuAehwIZ8NPLLPkxC9XWzHEv4MMo+vw95bJ9c+UUomoCxfgM268yKxND/NSpeC54HtY8QgUJp17Tfo9tWXiWbn9tWNSzGqpKsLnLYsUaBrGmTNn8NFHH8HT01MMBdy/f7/G63ShZ8yYgcKFC8PKygqtW7cWw5nUoSH5vXv3FoG0o6MjBg4ciMjISI11qJDEBx98IIzavby8sHDhwlw7BrN3qf7R0dG5tk2GYfLPKJ0w4SfDDKOVNxC+4ydotBXbvk2nBdtfH/2KoSeGagi2dT3qYlK9SVnazj+PAjQE2xpejtg5qAELttmAYkcnJ6d0J2kdfYLjV4bRXaT7Tul7zDDZITE4GEGbNsF77Dg8qFQZrwd+9Z5ga+rmBoty5eDwycdwnzoVpQ4fRunDf7Jgy2RIVGg8YiISxHy4ZRBURkmjwjpVqM5nT5cybaOiolC9enUMGDAAn3766Xuvk7i6cuVKbNu2TVTHnD59Otq2bSuKQkiVMkmw9fX1xfHjx5GQkID+/ftj8ODB2Llzp5xB8eGHHwrBd/369bhz547YHwXgtF5OIbGHtkXVAKVU8Kx40aUGV0fVffga6kYVz4CAAPGdNTXlQQcMo23EbNyIuNvJRceKrFoJm3r1oKv8/fpvzLkwR14u5VAK39T8Bi2LtcxS3PA6OBp9N1+Wlws7WGLvsEYwNs5Z7GGonD59GoYIx6+MrsGxddI5IMGW7jvp/pOTDpjsfI/C/vgDESdOIPLEyTTXc/isKwqNGwdTPXtgyeQfwb7JiZSvnB/I8/W8yvFlyCIFqlRQhV6a0vqFsnz5ckybNg2fvPOy2759O9zd3UVGbo8ePfDff//h6NGjuHLlCurUqSPWWbVqFTp06IDFixeLDN4dO3aIbLrNmzeLVO3KlSuLIhNLly5NU7SNi4sTU8qhcyTy0JSSQoUKif6+ffs2V86LtC9jY/aj02X4Gmo/9B0rWrSo+P6qDzWja0fLqX3fdR16uEW/F6WHXvqapaHP11DfIT8132nTEPfHgaQG+p5u3Ci803TxeorhYfe2YMWNFXJbw8INsbrlapgam773+yc9Dt32xbhfb8nLthYmODGmKe0FSqXKYL9/OdlPs2bNYKh4eHiIn1LiQU6Rrjn9bc1pAgPD8OcrbUiwlb6/DJPW7+O4Bw+gjIlBwps3CFi9Bolv30KlpnG8h5ERrGrWROFvv4VFqZJ8Ypkc3Wu+uh8MR0UVmJiYIsDqZdLnUmmKioUd+MxmEa1NL3v+/Dn8/PxEhqyEg4MD6tevL6r4kmhLP+mPliTYErQ+BYuXLl1Cly5dxDpNmzYVgq0EZesuWLAAISEhqQ53mz9/PmbPnv1eO2XlxcbGptpf2if1RaFQ5Mov2YiICNja2nLQq6PwNdR+6IaSMhRCU/FtopvOsLAwcR317eEJPcSi0QoEPWhS/92oT+jzNdR3YrZtQ6wk2AKwGjIEUWVKIyqXhKX8JCA2ADOuz8CTiCdyW0WHiphZZSaCA4MzvR36HI8/8BTnnofJbfaWJtjYrQIiQoMQAcP+/lHMlJtQJturV69kCx2JatWqQd/+DpIFGSUf0E1Wblz3oKAguLi48O9dJtfhz1cS9LCdM2yZ9CwPIo4dQ8Cq1VAEBWV8ooyM4NijOwqNHQtj1h6YXIBivxcvXuDZowDUdqss2oKtfcTPhNB6cLbRz3tPgxRtSbAlKLNWHVqWXqOfFGiqQ8OcnZ2dNdYha4WU25BeS020nTJlCsaOHauRaUteuG5ubvlShEIatk37Y7FBN+FrqPvXj25m9fE7SCKEjY2NmKffn/os2urrNdRnIv/+GyHbf5KXHb/4AoVGDNfJB5ivwl9h/Nnx8IlKClSJD4p8gEVNF8HK1CrT21EoVRiz55aGYFvE0QrbB9RFSdek77Khf/8ky6ycQrEX2WwdOXIk1ddz48G8NkICUG6IQHTdSVCi68G/d5nchj9fDJM6sQ8fImzvPkRduYy4+8l+96lB/rRGpqawbdECLoO+gpGZGYy4tgeTB0SHxQFugMpIhWBrX9HmaVWaz7U+ibYFiYWFhZhSQgFofgWhdLOTn/tjch++hrqNvl4/9WGr+nh8hnAN9ZHYBw8Q/NNPCPt9r9xm3qE93KdO0cnr9zD4IYYcH4Kg2KQsF1crV0xrMA2tirXK8rY2nX2GQ3eSgl2icRkX/NinLqzMcy6y6cv3L7f2MXr0aDH6gkZrNW/eHPv27RMjEr799lssWbIkV/bBMAzDMLmVVftm5DeIuXYt1ddNHB1h16Y1jG3tYOLsBMdPP4WpiwuffCbPUSqUiIlMGsETbRsOhXGimG/sVYPPvj6JtpJPDwXLNHRLgpZr1Kghr5PShysxMRHBwcHy++lnSq9ZaZm9gBiGYRimYIdQBa5eg8C1a2lBbrf9sA1MR43WyUtz8OlBzDg/A4nKpAC1jGMZbGyzEW7Wblne1h83vfHd4eSsmTGty2FU67K52l8mmVOnTuGPP/4QtlskBBcvXhxt2rQRo6zIOqtjx458uhiGYRjt8P+fMjVVwda6Xj24DhsqfnIWLVMQREckyGH9W6vX4qdKaYIGXpX4gmQDrU1fIUsDElVPnjypYVNA2Q8NGzYUy/STMiKuqf2yooCbhs+Q9620zpkzZzS8uo4fP47y5cunao3AMAzDMEzeo1Iq8XbefASuWSMLtuSn5jpiBDyXLIGRhe5ZdzwIfoA5F+ZoCLZb223NlmD7Ojga/9t3V17u27A4C7Z5TFRUlGy7RTEi2SUQVatWxfXr1/N69wzDMAyTIeFHjuBR/QaI/Ocfuc2ifHkU37UTpY/9hWLbtsKmYUMWbJkCIyIouQ6Uj/VD8VMZ74ZqRZz5quhapm1kZCSePHmiUXzs5s2bwpO2WLFiYpgaDUkrW7asEHGnT58OT09PdO7cWaxfsWJFtGvXDoMGDcL69euFMDtixAhRpIzWI3r16iWKig0cOBCTJk3C3bt3sWLFCixbtqzAjpthGIZhDB3fqf9D2P798rJTr54oNG4cjG1sxMNXXePa22vod7SfvFzeqTy2td8GG7Os+84evOWDOYfuIzIuSfwt5WaDSe0r5Gp/mfehB/oPHz5EiRIlUL16dWzYsEHMU4ypPuqLYRiGYfILZXw8/GbPRtSFC0j0SbZLkrBp3BjFNv3IF4TRGiKCYuR5X/un4qex0g5FnTJf04HREtH26tWraNGihbwsFf/q27cvtm7diokTJ4qsh8GDB4uM2iZNmuDo0aMaBSd27NghhNpWrVqJoWxdu3bFypUr5dcdHBxw7NgxDB8+HLVr14arqytmzJghtskwDGOIUKEYhinIIX0+EyYi/PBhuc2+Uyd4zJihsxflUcgjjP07uYBpKYdS2NJuS5YFW7KLGPfrLey97i23mZsYY/fgBrA211pHK71h1KhR8PVNuiGeOXOmSAygOJMKNlJcyjAMwzB5DcUCCd7eiHvyBJEnTyJ07z6qhJnqus59+8B15Ei+KIzWoFAoERGYABNjU6jMFQiyToppXcxKwNhY9woLawMFegdARR7ol1J6RSzmzJkjprSgrNydO3emu59q1arh33//zVFfGYZh9AESH/73v/8VdDcYA8Z3xgwNwdbhs64oPHcudJXQ2FBM/GcigmODxbKnjSd2dtyZZcE2KDIOHy47g6CoeLmtXglnLO1eHYXskh9WM3nHF198Ic/Tg/6XL1/iwYMHYvQXPfRnGIZhmLwkMSAAb0aPSbO4mFmRIsJKyqJ8OXh+9x2MOBGD0RJe3g3CtSMv4Ps0DO1q9RZt0U5hwDudtoxjxYLtoA7DaRsMwzAMw+Q59JDWZ+IkhB88mNRgYoLC334Lxy5Jlke6iFKlxIQzE/A07Kks2G5utznLgu2bkGg0X/Q3EpXJD7KHNCuFye0qiAfYTMFgbW2NWrVq8elnGIZh8hRFZBT8Fy5E6C+/vP+ikZGwQCAbKbuWLflKMFpFXEwiXt4JxPHN9997zdf+lTxfr3D1fO6Z/sCiLcMwDMMweQoVy3gzZixU0dFyW+HZs3RasCV+ffgrLvpeFPNOFk5Y02oNitgWydI2nvhHotcPFzUE24WfVUO3Ol653l8mfchiq169eqIGgjoLFy7ElStX8Ouvv/IpZBiGYXJdsH09ZMh72bXWdevCpukHcOzcGaZuWS9oyjB5nYyxd9E1+D0Lf+81Jw9reJZzwp6ENXJbHa8SfEGyCYu2DMMwBkRiYiL27Nkj5rt37w5TU/4zwORtQBe6ezf8ZmvaHNm2bgXHzz7T6VP/LOwZ5l2eJy9/2+RblHEqk6Vt+ITG4OPVZxEdn+xVt653LbSvykWvCoIzZ85g1qxZ77W3b98eS5YsKZA+MQzDMPrt8/+qTx/E3k/OUiSh1mPqVJiXYJGL0S6UCiWOb7kP/5cRCA9ILjYmYWJqjD7zG2Dfgd/hrQLCE98KxVGZaIsybk4F0md9gO/WGYZhDAilUonHjx/L8wyTV6ji4/FqwEBEX72q0e4+fRqcevXS6RP/JuINhhwfIuwRiHYl2qFp0aZZ2saft30xfOd1jbZ/J7aAl7N1rvaVyTyRkZHC9zu14o3h4e9nkjAMwzBMTh5sv+rbT0Ow9Zg9G07du/FJZbSO83uf4MaxZLuDlFRq4okqzYrA1MJY3GtSjJzgFQpjGMFY4QgHKy6EnV1YtGUYhmEYJleJffQIvlP/h9i7d+U263r14PXjDzBORRTTJW4H3MbIUyPlwmOuVq6YWHdipt8fn6jEzAP3sOtycuDrbGOOvcMasWBbwFStWlWMRJgxY4ZG++7du1GpUqUC6xfDMAyjf0SdOaPxYNtj5gwWbBmt5NFlv1QF22KVXWBlZ4aWX1aAsYmxaIuPTyqoGxkfBRiR9ZcRrIxd8r3P+gSLtgzDMAzD5BqJQUF4PfArUQFZwmXoELiNGqXzRbXuB93HgL8GIE4RJ5ap4Nj61uvhZp05r7nw2AT03XwZN16Fym2UeXBgRGMUdeIM24Jm+vTp+PTTT/H06VO0fFfs5eTJk9i1axf72TIMwzC5RuS5c3gzYqS87Pj5Z3Dq2ZPPMKNVRIfH49IfT3H/nK9Ge+PPyqBaSy8YG6cd1/tHhsnzjmbsyZwTWLRlGIZhGCZXiDx7Dq+/+kpeNnFwQJFlS2HTqJHOn+HI+EiM+3ucLNhWcqmEVS1XoZB1oUy9/214LPpsuoyHbyPktv6NS2BSuwqwNDPJs34zmeejjz7C/v37MW/ePPz222+wsrJCtWrVcOLECTRr1oxPJcMwDJNjIk6dwpuvh8vLxg4OcJ8+nc8so3XsX3odIX7JRYSJvvMbwdbJMsP3+kclJyi4ZTJWZlKHRVuGYRiGYXIMebK9HjpUo63Ent16UUiDhNrx/4zHm8g3Yrmic0VsbbcVVqZWGb43QaHEor8eYuOZZ3KblZkJVvWsidaV3PO030zW6dixo5gYhmEYJreg0Udvv1+AuEcPEff4icZrhWfO0HnrKEa/iI1KwN87Hrwn2Lb4skKmBFsiPD5Sni9m75nrfTQkWLRlGIZhGCZHJHh74/WIEUBiotxWcu/veiHYEnMuzME5n3Ni3tbMFstaLMuUYHvpWRCG7biO4Kgkfy/C1dYCuwc3QJlCtnnaZybrvH79Wlh4FC1aVCxfvnwZO3fuFH62gwcPLpD+fPnll/D394epqamwb/j888/zvR8MwzBM9oh7/hyB69Yh/MDB916zql0bXuvXwcTOjk8vozUE+URi95zLGm1VmxVBk+7l0rVDSG2EmkRxhyK52kdDg0VbhmEYhmGyjSI8HC96f4FEPz+xbGRhgdLHj8GskH4MhVp9YzUOPD0g5i1NLIUlQhHbjIPPu95h6LP5MuISlXJbywqFMLVDRRZstZRevXoJcZaEUj8/P7Ru3RpVqlTBjh07xHLKAmV5DQm1y5cvR40aNcT+a9eujQ4dOsDGxiZf+8EwDMNkHVV8PF4PGYqEV5oFnIxtbWFesiSKLFnMgi2jFfg9D4PPo1AEvonE4ytvNV6zcbRA425lsyTYEjGKKEg5uSWdONM2J7BoyzAMY0CYm5tj1qxZBd0NRk9QJSQIXzZJsDXz8oLXxg16I9i+jniNTXc2ycvDagxDHY86Gb7v+P23GL37hizYlnK1wfxPq6J+Ka6eq83cvXsX9erVE/O//PILqlatinPnzuHYsWMYOnRovou2hQsXFhPh4eEBV1dXBAcHs2jLMAyjxaiUSsTeuw+f8eM1BFsSat2nTIZt06YF2j+GiQqLw8s7QfB9GooHF5Ji+NSo1bY4GnQulaVCwtK95l/briMSz6FSGaGcS9IIJiZ7sGjLMAzDMEyWUalU8B47FtFXr8qZI0VXroBFyZJ6c3wjT45EoirZ8qFvpb4Zvm/sLzex97q3vFyzmCN2ftUAVuZcbEzbSUhIgIWFhZin4mMff/yxmK9QoQJ8fTUrJ2eGM2fOYNGiRbh27Zp4/759+9C5c2eNddasWSPWoUza6tWrY9WqVbJwrA5tQ6FQwMvLK9vHxzAMw+Qdcc+eI2DVSsRcvSY8bNUptnmTXhRlZXSby4eew/thCHweJxcJS4uPv6kBr0rO2d5XtOotYASoEu1R2IFHCOWraEsBaKNGjcSQLXUSExNx/vx5NOUnRwzDMAyj16gUCrzq1x/RV64kNZiZwWvjRlhWrAh9Ycb5GXga9lTMu1q54mDngzAxTl94PXrXT0OwbV2xEJb3qMmCrY5QuXJlrF+/XhQiO378OObOnSvafXx84OKS9SzpqKgoIcQOGDAAn3766Xuv79mzB2PHjhX7rF+/vrBCaNu2LR4+fIhCatnqlF3bp08f/PDDD+nuLy4uTkwS4eHh4qdSqRRTXkP7oIcd+bEvxvDgzxej7Z8v35kzESPFRWo4DxoEqwYN+HejAaMNv78uH3yOa0depvl6xcaFUaKqC6wdzOFa1BbGJsbZ7m9MYgyURklFzIwVzjA3MeLPfypk9vxmWbRt0aKFyBZQDyaJsLAw8RplATAMwzDaCT1g27t3r5gnESHlAziGyQhVYiJeDx+eLNjS0O3//Q/WtWrqzcn75eEv2P9kv7w8se5E2JrbZijYjtx1XV7uWa8Y5nWpkqUhZUzBsmDBAnTp0kVkvvbt21cIrsSBAwdSzX7NiPbt24spLZYuXYpBgwahf//+YpnE2z///BObN2/G5MmTRRuJsJSdS8uUNJEe8+fPx+zZs99rDwgIQGxsLPLj5oPuB+jG1NjYOM/3xxgW/PlitPnzFbNtG2LV4iKzRo1g1rABzJo1g8reXhSUZAyXgvr99fRyMB6eDYQiQYX4GE2dztTcGOUaucC1hDWci1rBxJT6RSJiLAKDYnN0r/nT3p8Q9jAM9rXtYaFy5M9/GkRERCAzZPlunT5oqd2ABAUFsccWwzCMDgQN9+/fF/Mph+kyTGZ4O28eov45Iy+7T50Cpx7d9ebk/fP6H8y/PF9ebl2sNdqXTFt4I47c8cWwHcmCbfPybizY6iDNmzdHYGCgyFB1cnKS26k4mbW1da7uKz4+XlgeTJkyRW6jGzkqfnbhwgU55u7Xrx9atmwpiqNlBG2LMncl6DjITsHNzQ329vbIj78vdI9A+2PRluHPF6NL5OT3V8Tx4wjZslVedhs/Hs4Dkh7GMUxOP19ZJcQ3Cuf3PsWbByFQKlSprvPVsg9gZpE3tl0U39x9/BCx3rGwq2UHRwu39xI+mSQsLaVSbbkk2krDuujDRgGk5PlFUHbt7du3M8wAYBiGYRhGN1HGx8P3f9MQfvCg3OYxayacevSAvvD7o98x60Jyob6uZbtiZsOZGRYdG7X7przcsJQL1vWuzRm2OoqJiYmGYEuUKFEi1/dD4jDFz+7u7hrttPzgwQMxT0XQyEKhWrVq2L8/KfP7p59+EgXSUoNic/X4XIJuEPNLRKX7hPzcH2NY8OeL0abPlzI2FtFXr8Fn1Gi5zbxMabj0+RJG/DuQyeHnKyuolCpcP/YSL+8GwfdJWKrrGJsawdTUGF3G14aFlVmeXR86Pt/IQHm5iE0JjgnSILOfhUyLtg4ODvJTfzs7O1hZWWlUiGvQoIEY4sUwDMMwjP5VQn47f76GYGvfqZNeCbZkiTD3YpKHKdG4SGNMazAtXfH12ssQDNqeVIiNqFjYHtsH1oOZCQtWTM5p0qQJe8AxDMNoqbf/6yFDEX3pktxm6uaGUvv2wcgs7wQxhpFQJChx48QrBL6KwLNbgUK4TYm9qyUc3a3Rul8lWNmZ59vJ84kIkucru5XNt/3qK5kWbbds2SJnG4wfP56tEBiGYRjGQAhYsRKhu3bLy67Dh8N12FDoA/Qwes3NNdhwe4PcVqtQLSxrvgymxmmHSeefBqLXD8k3ay3Ku+GHPnVgyoItkwlcXV1FVu/bt2812mnZw8ODzyHDMIyWxgyKkBC8/OJLxD97pvGa26hvWLBl8pyEeAX+3vEAjy5pxg8paTOgEsrVK5h4Ik6ZVISMqOZeskD6oE9k2dN25syZwlz4xIkTePr0KXr16iUyb6myLvll2dqmX6iDYRiGYRjdQBkXh7A//kDQhmRB033GdDj36gV9ufn69uK3+OXRL3Ib+dd+/8H3MDZKO1t27/U3mLz3jrxc2s0Gy3vUZMGWyTQ0Sq127do4efKk7C9Onne0PGLEiBydyTVr1oiJiwMzDMPkDqqEBPjNm4ew3/dCFR+f/IKREVwGD4Zl+XKwS6fwJMPkBglxCuyYcQFRYWqfwXfYu1mhaAUnNO5aBuaWBVtoOv6daKtSmaBuseIF2hd9IMtX8+XLl2jXrh1evXolKtq2adNGiLZUcZeWqfItwzAMwzC6L9hSJknsnWRx0q5tW70RbIll15ZpCLYNCzfE/Cbz0xRsSeRd9NdDrP37qdxWwcMOuwY1gEMe+oMxuklkZCSePHkiLz9//hw3b96Es7MzihUrJoqG9e3bF3Xq1EG9evWwfPlyREVFoX//nBWwGT58uJioEJlkb8YwDMNkn5DdezRGHEm4fTMSrsOG8all8ozEBAUu7n8G3yeh8H8Z8d7rlZsWEUJtXhUWyw6JqmiA3MUS7eFi877XPpPHou2oUaNEcHnr1i24uLjI7V26dGFPW4ZhGIbRA0J/3wv/RYugCA2V28xLlYLH9GnQB0h8XXhlIX7+72e5bVj1YWJKz8P250uvNATb6l6O2PFVfdhaFGxGA5M7rFy5MtV2+kxQhd8yZcqgadOmwtYgM1y9ehUtWrSQl0mkJUio3bp1K7p3746AgADMmDEDfn5+qFGjBo4ePfpecTKGYRim4EgMDETg6tUaxcbMPArD9ethsK5Viy8Nk+solSpEhcYh4GUEjmxITp5Qx9bJAl982xAmBWDLpVQpccHnAm7430C8Ih4JygSEx4cjNC4U3qHeUBkliPXMVM5cmDcXyPJdxr///ovz58+LYV3qkNett7d3bvSJYRiGySPMzMwwdepUeZ5h1FFGReHtgoUI/SU5+xRmZig0biyce/fWG6+29bfXawi2X1f/GsNqpJ8p88Q/Et8f/k9eHtC4JKZ3qsjBqB6xbNkyIaJGR0fDyclJtIWEhMDa2lrYf/n7+6NUqVI4ffo0vLy8Mtxe8+bNxQOC9CArhJzaITAMwzB5Q9iff8Jn3Hh52bZ5c3itX8enm8kzYiMT8Ov3VxAeGJvq69YO5ihb1x31OpbME8E2ODYYT0Ofip8vw18KITYwOhBRiVFCoA2LC8N/wcnxcEoo7nH7yE3M24EfQheIaEt+W6l5ZL1580bYJDAMwzDaC2WMpXzoxjBE3NOn8Jk8RcMOwapObRQaNw7WNWvqzUm6H3Qf2+9tl5eHVh8qpvR4ERiF7hsuICo+Kf75qLonZnxUKc/7yuQv8+bNw8aNG/Hjjz+idOnSoo3sDYYMGYLBgwejcePG6NGjB8aMGYPffvuNLw/DMIweooyORvhfxxB96RLC9u+X2+nBNRViZZi8IDFegbiYRNw49ipVwZaE2ua9ysPcKvuju+IUcQiNDcXdwLuIVcSKDNnohGgExQYhMj4Sf7/+Gz5RPjm+1zQyNYIyzhVVnTrlaFtMElm+4h9++KHw3KKgVroo5NlFBco6dOiQ1c0xDMMwDFOA0BPxkO3b8Xb+9xrtriNHwPXrr/Uqk3Tz3c1YcX2FGNZFlLAvIbJs0zvGez5h6LjyrLxc1MkKcz6unC/9ZfKXadOm4ffff5cFW4IsERYvXoyuXbvi2bNnWLhwoZjXZrgQGcMwTNZRRUcjcNUqBG/bLuZTUmz7NlhVrcKnlslVG4TA1xE4tf0Bgn0ikXJwjpW9Oep1KoniVVxg52yZpW1TtuwVvytYcnUJ7MztEJMYg8iEyNzpuMoYKoUlFDHFEB/SSMwDJlAl2kKlNEMpZzcUdbbGxBYcLxeIaEuBKxUiq1SpEmJjY9GrVy88fvwYrq6u2LVrV650imEYhskbEhMTcejQITHfqVMnmJqyF6ehC7ZBP/6IgCVL5Taz4sVQZMlSWFXRr0Br6bWl2HJ3i7xcza0aNrTekK5g+zY8Fj02XtQQbH8aWB9ONpytro/4+vqK35EpoTbynCU8PT0REfF+IRBtgguRMQzDZD27NmLkSCiePnvvNSNLSxTb9KNejTpiCpbwoBj8Mu8K4qLejznU+WJOA5hbZnyvRrYFj0IeITwuHM/CnmHBlQUar8fGpG61kB5xgc2hUlhDGecBVYIDVAobIchCRf0hWwbN+HlSuwqo5GmPRiUd8eeff1KFDHg5sudzbpDlu3Xy8KIiZHv27BE/Kct24MCB6N27N6ysrHKlUwzDMEzeQBY3VL2c4NERjM+kSQg/cFA+EbbNmsFzyWKY2Nrq1cnZ82CPhmDbpngbfNfkO1iZph+3jPvlFiJikwPqnV81QDEX6zztK1NwUNEwskIge4Sa727Ob9y4gWHDhqFly5Zi+c6dOyhZsiRfJoZhGD0hMSAAbxcu0hBsLcqVEyOOLCtWgqmzE4yt+W8/k3MUCiX2fHsFwT5Raa5TurYbYKqCR00r3Am7hZigGGFh8DriNe4F3UOCIgGJqkQxakyhVCAiIQIPgh/Io8jSwsLYCpbwQEK8LWLjzGAEM0SFF4aKRFilGVSJDlApLYUwq0pwfk+U1diWqTFcbMyRqFShV/1i+KJBcbjaWojX4uPj+V6zIEXbhIQEVKhQQWRpkUhLE8MwDMMwukf4kSMagq1T795wn/Y/vbJDIFbfWI0NtzfIy+1LtseCDxZkeJy/Xn2Ns08C5eUDIxqzYKvnbNq0CV9++SVq164tF2qkLNtWrVqJ1wgqSLZkyZIC7inDMAyTUxKDgxG2/w8ErFwJVWxyJmKRpUtg17693sVDTMEQH5uIwDcReHDFH//9c/+9181sjaEsH4TY0m9xMuwIXoS/SHrh1rspBygTbRHn3x6KyPKIUGQ9IaNWMUe0r1IYpQvZwMLUBMVdrOFhbwnTPCiAxuSSaEsBLFkiMAzDMAyju4T88gv8ZsyUl10GfSUKjukTlHGw5uYabLyd5MFPdC3bFTMazsjwRuxVUDQm/X5bXp7/aVVUK+qYp/1lCh4PDw8cP34cDx48wKNHj0Rb+fLlxaSejcswDMPotjWU79T/IWzfvvdecxk2DPZcp4fJIQlxChxceRO+T8PSXe942a146nojaeF19vZlBCPYm9ujrGMV2MALf92OgTLeBcpYT6gyEGrJ9otE2OIuNmhXxQNlCtmKLFpLMxM4WZvxgwtdtUcgn6wFCxaIoWPshcgwDMMwukXUpcsagq11nTpwGzsW+gR5ew34awBuBSSnKAysMhDf1PoGxkbpZwcERcahz+ZLUL4rBvFhJXf0qOuV111mtAgaVSYJtZxpxTAMoz/Ev/FG4KqVCPvjgEa7XadOQKeOcG3atMD6xugGZE/wNOwpfCJ9EJUQhURlIuIUcXgZ/lJYGIQHxqD+6fRHpL9xeIhDFdem50CAFl4t4GrlChcrF1iaWMIMdnAwd0ZJ+3IwUpnigV8U7nlH4OAtf7yJVOBNBv0e3bosbC1MYW1uihYV3FDYga1N9Va0vXLlCk6ePIljx46hatWqsLGx0Xh97969udk/hmEYhmFyKbMk4thx+EycKLfZNGuKosuW6Z0wtfnuZlmwpQyECXUn4MtKX2b4vtfB0eiy9hwCI+PFsru9BWZ+XFnvzg+TNtu3b8eiRYtEkV2iXLlymDBhgrBN0BXWrFkjJoVCUdBdYRiGKXAS3voj/MhhRJw4gZir1zReMy9eHI49e8Dxyy8REBBQYH1ktIvQ2FD8+uhXBMUGCVGWkgGehz3H8/DnCItLO3vWMdodn90e/157hHkIAmxf4bnzbfjZPUOEZbBor+BcATam9ohLMEI561YwUznDWGkLY6UL3vhF41ZYLIKj4vE8UPLAjSN3/Uwfx/efVkWDUi4o4aqp2TF6Lto6Ojqia9euedMbhmEYhmHyBL85cxC6a7e8bFW7NrzWrYORsX75Ut0JuINNd5L8R4llLZahVbFWGb7viX8kPl59FtHxSUKXm50FfhnSEEUcORPBUFi6dCmmT5+OESNGoHHjxqLt7NmzGDp0KAIDAzFmzBjoAjQqjqbw8HA4ODgUdHcYhmEKDFVCAl717Yv4F+98QtVwGTIEhcaMlgv1MvoNCa+UESsKeKkUciEvaZlej1fG43bAbcy9ODdb+ygWWhGmKnN5OdE4ATsafgeVSgkPk1qwMy+DSsYN8PytEYzjyuDuk3jEJCTFnReS3kGS8bspa3xQ1hXl3O1gY2GK+iWd0ai0CycdGKpou2VLcvVlhmEYhmG0G0VYGF5+8QXiHj+R26xq1kTR5cv0TrD95/U/GPv3WBF0Ex+V+ihTgu35p4EYsPUKYhOSbtrIz+ungfWExxdjOKxatQrr1q1Dnz595LaPP/4YlStXxqxZs3RGtGUYhmGSCD98WEOwNbaxgW2rlnDp3x+WFSvyadIxJKE1ODYYkQmRSFAmCLH1pv9NnPU+K8RXmmh0Gf2k9aITokVcGBEfkeP9Wxk7wRKuMDOyhZWxG6yMnAGlFWLigZhYUxQPKCuve98sEcetExF/b5JYfj+POybT+6VEAnMTY1QsbA9LM2OYGhvBxNgY5qZGaFnBXQi0JNYy+glfWYZhGAOCCkrSUF9pntFvlPHxeN6tGxJevpLbHD//DB4zZ8LIVL9CgMCYQEw/N10WbCs6V8S0BtMyfN/2Cy8w++B9KN6Z2FJm7W/DGrLXlwHi6+uLRo0avddObfQawzAMo1u2CG8XLJSX3UaPhsuQwZx9qAOQwDri5AiExoUKUZZivFhFbL73QxHjhVi/zoDKGCqFDSIS7eXXrJVAyQQTUF6tpcoIjeNMYK9KToa4YpGI+Ey4a9Ur4YySrjYo7moNdztLOFiZwcLMGE7W5ihkbwFna3OYmuhOkgXfa+Y+Wb5jq1mzZqq/6KjN0tISZcqUQb9+/bi6LsMwjBZCv6tTepEz+knU+fPwmTwFif7+cpv7jOlw7tUL+gb5i/U72g8hcSFiubFnYyxvsRyWppbpvm/Z8UdYcTLJu5Qo7GCJfV83QiH79N/H6CcUw/7yyy+YOnWqRvuePXtQtmxy9gzDMAyjvaji4xF1+QoCli2DIjjJO9S2WTMWbLUQyoi9HXgbgdGBIobb93ifWM4tTI1MRSEvc2NzmBpbws7UEcZGJjCCMYyMjMVPwFj8S0g0RkCEEqFRCgRHmCExvBqU8YU0tmeiAsolmMBdYYS6cWknvyiNgEY1PeDpYo0mZV3pQBEWGgpnZychwNL9GImx+ug1y/eaWiDatmvXTgwdoyJk9erVk4uT3b59W4i19+/fR+vWrUVBsk8++SQPuswwDMMwTHoErl+PgOUrNNrcxo3VS8GWKvV+efhLUSyCKGRdCHMaz0lXsI1NUODrHddx6kGyoN22sjtW9KgJSzOTfOk3o33Mnj0b3bt3x5kzZ2RP23PnzokCvCTmMgzDMNqP7/TpCPvjgLxs6uGBwt/P5wxbLYIsC355+AuWXFuSqfXdrd3hZuUGU2NTmBibwMLEAq5WrlApzRCXAARHGCEgxAFuJlVhCmuERivwKjgGYTEJeJqoIs00R1TwsEMLD0fY/h2YqfW/nNUAju7W8jJ5Jvv7K1CokDOM9cyajNFC0ZYKMYwbN04UalDn22+/xcuXL3Hs2DHMnDkTc+fOZdGWYRhGy0hMTMRff/0l5tu2bQtTPRsizwDBP+/QEGxNHB3huXABbJs21bvTs+rGKmy8vVFetjK1wg9tfhDCbVokKpT4ZtcNDcG2fRUPrOlVC8bGmRjHxugtVGj30qVLWLZsGfbv3y/aKlasiMuXL4uRZgzDMIx2ZmuSKhd98SLC/zqmIdgaOzig6IrlMHVyKtA+GvK1eRr6FEOOD5GLfyWqEjPlL9u0aFN8WvZTtPRqqSG404N3srXacTLZ+isZslDIuY0ChYPkIzuuTXmUNjdH3NMI3DiW2v6A6i294FHaAabmxnAsZK0h1hoifK+Z+2T5bp0yDa5du/Zee48ePVC7dm388MMP6Nmzp6jAyzAMw2gX9KSXRkcQbdq0KejuMLlM7IMH8F+8WF6m4YBFli+DsZWV3t0ELLu+DFvubtEQbJc1X4ZSjqXSfN/b8Fj023IF//mGy21DmpXCpLYVWLBlBBTL/vzzzzp9NtasWSMmhSKpIjXDMIy+kRgYiLfz5iP+9WvEPXgAVULCe+uYeXmhxK6dMHV1LZA+GhJ+UX64F3RPFAaLSYgRtlVX317FP2/+yfQ2JtWdBHMTc5R3Lo+qrlWhUhmJTNmAiDhExiXCPyIOodEJGPrz+1pUWrjamqOIk7UoMGtrYSqWzU2NYWJkJOI+KuhFP2mZ5kmobVelMIzCE/D3zw8QsOUJAt7VPFCnUAl71G5bHJ7lHGFpwzVC1OF7TS0Qbcm39vz588L3Sx1qo9ekCyXNMwzDMAyT9yhjY/H8s8/pEbdYNitaFEXXr9O74YAk2E46MwlHXhyR2yq7VMaiZovgZeeV5vsSFEr02HgRzwOj5CyKZd1r4JMaRfKl34x2Eh6eLOBnhL19cgESbWb48OFiomNzcHAo6O4wDMPkKqrERHiPHoPoq1fTXMfExQUlftnDGba5ed5VKrwIf4EXYS9w4tUJIcwGxQThbtDdTG/DztwOLpYuMDEygb2FPUo7lka3ct1QwbmCHK8+fhuBib/dwYn/3gqRNiOcrM3wZYPiQnCtU8IZLjZUGgywMDMRRb2yilKhxK8rbyHwdeR7r5lZmKD90Krwquic5e0yTL6JtiNHjsTQoUNFtm3dunVFG2Vt/fjjj3LhBhp6W6NGjWx3imEYhmGYrOG/ZKks2FqULYviu3bpnWBLfHfpOw3BdkCVARhTe0y67wmLTsDwnddlwdbO0hTbB9RDzWI8XNLQcXR0zPB7QjeqtA5nrjIMwxQ84X/+qSHY0kNqyqY1trMTI4xM3QvBqnp1FmxT4BPpg+Mvj+PUq1NCcFVJ/94ZvkrzZGFA86JNJa8F/2h/8Vp2qOhcEUOqD0GrYq1SfZ1GQD0NiMSOi69w4VlSjYKMaF3RHeu+qAUzk+x7xCqVKry+H4yI4Fj4PQuD/4twhPhFa6xj72oJYxNjNOhcCiWqucIkB/tjmHwRbadNm4aSJUti9erV+Omnn0Rb+fLlhS1Cr3cFTkjUHTZsWLY6xDAMwzBM1oi+dg0h7/4mE+7TpsHEVv8q0h58ehB7Hu6Rl3tW6InRtUan+543IdEYuPUqHr5N9k/b3K8uC7aM4PTp03wmGIZhdIiwg4fkea+NG/TSsz83C37tfbwX62+vF0JtXmCEpAef1d2qo45HHVEgjCZjI2PYmtmirkddUUAstRFQt16H4n/77mrEaOqUKWSL0m42MDUxFhm1dpZJmbO1ijmhTSX3LPfV/2U4bp16jbioRMREJgiRNj2KV3FBpxHVs7wfhslNslWBpnfv3mJKCys9885jGIZhGG2EMiDC9u6D74wZcpt9h/awqV8P+sajkEeYe3GuvPxV1a8wqtaodN/zx01vTNt3FxFxSRnIxIKuVVG3BA9rY5Jo1qwZnwqGYRgdIfLcOUSdPSvmzTw9YfPBBzBEpIxY+kdJsZQJG5MYgyPPj8A70lus8yD4AS76XkxzGzZmNkJwFaKr+M9ICK2i5d0IFGle+uds5SwsqQrbFEYx+2JoWawlLEwsMuxvYGQcvENiEB2vwI3XISKj1js0Jt33LP68Oj6rXRQ5xe95GK4dfoHYqAT4PcucJZKNowVK1XQTvrUMo5OibWhoKH777Tc8e/YM48ePh7OzM65fvw53d3cUKcLecAzDMAyT16iUSvhO/R/C3lW5JywqVEDhefP07uRfe3sN/Y72k5epmnB6gi0VrPh6x3WceRQgtxVxtMKGL2ujShH2+GSSefXqFYoVK5bpU+Lt7c2xLsMwTD4SdfEiIo4dQ/ybN4g686/cbt+xg17ZQEmesVEJSVZOJMKefHUSr8Jf4WX4SyhUScUl6WdgdCASVckPpDNLkyJN0KlUJ7Qu3jpTYmtO2X7hBdb//RQ+YbEZrjvog5JwtDZHaTdbtKpYKEu2ByTIPr7yForEJPuGMP8YhAXGIDFeAd8naWcYG5sYwcTUGHYulqjVtjhsHS3gWdYRRlT4gGF0VbS9ffs2WrduLQobvHjxAl999ZUQbffu3SsC3+3bt+da58g7bNasWaKKr5+fHzw9PdGvXz9h0SD9gqZfbjNnzhT2DCQmN27cGOvWrUPZsmXl7QQHBwsv3oMHD8LY2Bhdu3bFihUrYGtrm2t9ZRiGYZj8FGy9x45DxNGjcpt9x47wmDUTxnpWCJSyRAYdG6Thiza53uR037Pk2EMNwfaDsq5Y3bMWHKy5wi+jCdVn6Ny5s4hnpVoNKQkLC8Mvv/wiYsfBgwfjm2++4dPIMAyTxwT/vAOBG9ZDERD43mvkWesyeLDeXIOz3mcx6/wsvI1+myfb716+O/pX6Y8itrmbYEc1A0Ki4/E6JBoP/SIQr1AiOk6BoKg47Lr8OlPbaFmhEL76oCQalXbNVh9un36Df/c8yvT6JMh2GlENRco6wcSM/WkZPRRtx44dK4TThQsXws7OTm7v0KGD7GmbWyxYsEAIsNu2bUPlypVx9epV9O/fXwjGUsBM/Vi5cqVYh7x2p0+fjrZt2+L+/fuwfHfjSlYOvr6+OH78OBISEsQ2KOjeuXNnrvaXYRhG2zEzM8Po0aPleUb3oIeVPhMmJgu2RkZwnzIZzn36QN848+YMxvydXGTMzNgMa1uvhZWpVZrnZvbB+9h6/oXcNq9LVfSs56VX2ThM7kHx4nfffYc2bdqIuLF27doiSYDm/8/efcA3VX1xAP9lp3vvyd57g0xRliCICogyVESWCshfQIY4EJXlABdLEcSNmyGoyN4byirde6UjO/l/7g19JbRlpk2TnK///PPuy2ty+15p807OOycvL48/fubMGbRu3Zq/52TvdwkhhFQtfXo6MhYtEhqsCsRi+I8Zg+Dp0yCSSJziMPxw4Qe8tu+1W27nr/QX6se6y9wRoAzg5QwYXtZAJEKwezC6RnTl5QsYiViChv4NbZJVy2rQ/ng0GVmFWpjMwLaz6TidcnvlBko92TEavm5ytI7xRaMwb4T53F1ZTZ3GgMtHs6Ap0mPvj5duui17+xcc6437HquHoCgvCtRWMTrXrAFB20OHDuHTTz8tt56VRWDZsLa0d+9ePPzwwxgwYAAfx8bG4uuvv8bBgweFk7Ply5fzzFu2HcMyfVmZhs2bN2P48OE4d+4ctmzZwufdtm1bvs2HH37I33QvXryYvzG/kVar5bdSKpXll5HJZOK3qsZeg9epqYbXIlWDjqFjc/bj5+3tXdYR9lrHWGfjrMeQl0R4eYZVwDZs0SJ4D3zIqb5X9r2kFqdiyu4pwjp2wrGh3wb4K/wr/V43H0+xCtjO7NsAw9tFOvXPek1U3f/+7uV1AgICsHTpUh64/f3337F7924kJCRArVYjMDCQf/DPkgGaNm0KR7FixQp+Y1esEUKIIyn6bzcKfvoJuoQEq4Bt0NSp8OzRHdKAAEgD7y4jsyY5mnGU15/VmXS8Udj16vjUQYewDnyZNfBqGdySlzWo7APr6rJ0+wV8/M/lO/46b6UU/87oCT8PuU3m8d+3F3ByZ3KFjzXrHoGwer58mZU9CK/rC5mbBJI7KLVA7g378MDX13IMiJ2CtgqFQghiXu/ChQsICgqCLXXu3BmfffYZf+769evjxIkT/M00e3PNxMfH80AxK9dQimXhdujQAfv27eNBW3bPfmhKA7YM256VSThw4ACGDBlS7nXffvttLFiwoNz6rKwsaDS3rsdii5MPdikeO+Fh8ySOh46hY6Pj5/ic8RiygG3xgteh//dfYZ3i8ceh6dAemsxMOBMWsB2zu6yGbXO/5ni33buQ6CXIrOR7zSrS4c3fzgnjCZ3DMbihZ6XbE+f591dYWHHX6TvBmug++uij/OboJk2axG/s/Tp7X0wIIY5AffIkkidOhFmvF9aJZDLU3bkDUhvHGaoa+/t3Kf8SVLqyuEmWOgvxBfHI1+Rj4/mKr/jdNnQbwjwtWbLVjTUG23o6Hat3x6NArYeBJayxz7uvNTrTGyv+8NtDLsGDTUL599y1XhD8PeQI8lJAIhbBUyFFlL+7TeZnNJqwa9MFnP0vtcLH73u8Hlr0irLJaxHi0EHbQYMG4fXXX+e1vUoj6ayW7SuvvMJrxdrSzJkz+RvOhg0bQiKR8IwBlgnBsh6Y0sxelll7PTYufYzdBwcHWz0ulUp5Hd7KMoNnzZrFy0CUYnOIioriQenSDLWqPtlh+5W9nrMEG1wNHUPH5szHj/0e3bFjB1++//77+e9WZ+Rsx9CYn4/k8c9Df+qUsC541kz4PfUUnE2uJhfj/xrPTxBKvdHtDYR5h9208dgLX+1FboklK+eBRsGY8VCLapkvsf+/v9JyWIQQQhyTqbgYKS9NtQrYMgHjxjlMwJY1ENt8aTNvnronZQ9KDCW3/bVKiRLvdnu3WgK2+SU6rPj7Ei9zwOKwmSoNtAYTjifl3/ZzrB7dFl5KGZpF+MBNXj3nEhcOZJQL2DbrGYmohn7wD/eAT5BtgsPk3rjKuWaNDtouWbKEZyGwQCi7dKx79+48+NmxY0ceULUlFhjesGEDrz3LatoeP36c12JkJQ1Gjx6NqsKyidntRuzEo7pO/tnJTnW+HrE9OoaOzVmPn8FgwP79+/lyr169nO77c8ZjaMjKwqWu3azWBc94GQFV+HfQXliWxszdM6Exll3VsqbPGtT2rV3p18RnF2PEZ/uRrrJ8TaSfG94a0szhj7ujq85/f3SsCSHEsWV9tAL6VEtAzq11a0QsWQyxuzsk1Xi1QHJxMub+NZd/eJxVkmX1XqQiN5ZdutX2Ffmg5weI9YnltWg9ZB64V+w90cwfTiI5T33TjNrb4aWUItLPnVfRZX/KWT1dbzcppj1QH21i/FHd/v06zmo8amFnePnTh7Y1MWjLypwyPXr0oKCtPYK27DIr1tCLlSk4efIkioqKeHOG60sU2MqMGTN4ti0rc8A0a9aM1xlj5QtY0DY0NJSvz8jIQFhY2adSbNyyZUu+zLa58dJIFrTIzc0Vvp4QQgipiXTJyUgcM9ZqXfh778Jn4EA4o2VHluFA2gFhvLbPWrQNLStvdKM/TqXh5e9OoERnqd3pJpNg3dj2CPamN/GEEEKII9DGxyP3yy/5skguR/jbCyG77ty+qrEA7fIjy/HLlV9s+ry1fWqjVXAreMuvXakrAqK9ohHqYYlB1Perz4O1d+tSZhHmbD6FArVBCCKfT7+3ckEyiQiv9G2Ivk1DecDWntj3k5NSBE2xAXt/uASjvqx+/VNvdaKALXEZdxy0LXXffffxW6mjR49i3rx5+O2332w1N5SUlJTLnmDp1aUNJ2rVqsUDryz9ujRIy0oZsFq1EyZM4ONOnTohPz8fR44c4R2BmZ07d/LnYLVvCSGEkJrIbDAg8ZlnoE8ua7YQ+80muLVwzsv+N57biLVn1grjWc1moXVw6wq3NZrM+HDnRSz/66KwjtVN++rZDqgb7Fkt8yWEEELI3QfktOfO8YZjWe9/wNLz+PqA556DPCbGZrs1W52Nr85+hXxtPkxmVqPVxMsvlS7nafKwL21fua8LdAuEr8K3wqtIrMY8D7WMn8IPQ+sPRc+onlBKq+4D5JX/XMK7W6wzT2/E6spWRiYWYWibSAxtHYlALwUUUkvMRSoWlfse7cFsMmPDa/tRkFlBVrAI8A6wb1M2Qmps0Hbr1q08y1Yul+PZZ59F7dq1cf78eZ4N++uvv/LuurY0cOBAXnIhOjqal0c4duwYb0L29NNP88fZLxRWLuHNN99EvXr1eBB37ty5vHzC4MGD+TaNGjVC3759MW7cOHzyySfQ6/WYPHkyz95l2xFCCCE1DTuZSX/zTegTEi0rxGLEbPjKKQO27HudtXsWfr/yu7BudOPR6BXeq8Lti7UGPP7pPpxJLWvuwWqqffxka7tnhRDHtmvXLt4El/U+uPEKLXapX7du1mVKCCGE3L7CnX8ja9kymNRqmDQaGLOzrR6XBAYi4Gnrq4vu1TsH38GWq1tue/uWQS2xqNsiRHhGoKYwmcwo1BpgMJpgMJl5LdobA7as6RcLuLL7JuHeWDikGeqFeMERZScX4a+1ZysO2AJ4fHa7ap8TIQ4RtF29ejUPfLIGXnl5eVi1ahUPoE6ZMgXDhg3D6dOneYDUlj788EMehJ04cSIvccCCrOPHj+cZvaX+97//obi4GM899xzPqGXZv1u2bLFqSsHq4rJALSuEzDJ3WcO0Dz74wKZzJYQQQmwlb+NG5G/6RhhHff4Z3Fu1crodrDPqMP3f6fgn6R9h3QMxD2Bq66nIysoqv73BhCdWHbAK2D7eNhJvP9Kcn6gQci969uyJtLS0cg1sCwoK+GOsThshhJDbxzJp9SkpMBYVIXXmLJhLKm7OxcoihL02n9extRW9UY9dybtua9sWQS0wsd5EdKzTsUbVSf/1RCrm/3IGucW6Srf5dnwntK9V/TVmbfnhfUa8CgVZapiMJuz88ny5bdoPrAXfYHfUbh0EiaTmHB9CalTQ9v3338c777zD68z+8MMPeOyxx7By5UqcOnUKkZGRVTI5Ly8vLF++nN8qw7JtX3/9dX6rDAs0s2ZmhBBCSE2X9913yHizrLFnyKyZ8OzSBc5GpVPhxZ0v4nDGYT4Wi8SY0XYGRjYaWa65R2mmydh1B3Hiuu7Gnz7VBn2aUH16Yhvs566iy0JzcnLg4XHvDWIIIcRVsCaq6a+/jsLtf5V7TOzpyYO08thYeHa9D9LgYHjcdx9kISF39VqH0g/hp4s/WTUCUxvU2J2yWxjfH30/prSaYmmSyf4Tifky+89N6sbLGtzYB6e6JeWW4KOdl6DS6IWmYreqUfvNcx0dOmCr1xmxeclRZCZU/H0GRnnikRltIJNLqn1uhDhc0Pby5cs8UMs88sgj/NKx9957r8oCtoQQQoirKd6/H+mvLWDRIz72HT4M/qNHw9mU6Evw9JanEZdXdnnfgs4LMLiupbTRjUFbdgLDuiHvuZQjrFvxRGsK2BKbYO9rGXYCP2bMGCgUZXUAWXYta7zLyiYQQgi5NZNOh6QJE6E5fbrcY/K6dVDr22/vKaOWvUfYkbgDZ3LOIKUoBVuvbuX1aW+GBW3r+NapfM7XeubYAyt7wJqqbj6eWuk2Pm4ytInx4yUQZBIx3OQSDG8XhbaxjhmwNeiMSDybi51fnoO2xNJI7UbtBsSi/cDa1T43Qhw2aKtWq+F+7Zcre1PL3tCGVWNXR0IIIfdOJpPxkjOly6TmUG3ZgtQZ/xOacXh074bQOXPgbDQGDcZtG2cVsH2l3StCwLaikggjPttvVRLh/eEtMaA5vQchtuHj4yMEAthVXm5uZQ1OWB+Hjh078hJhjmLFihX8RuUcCCH2kPneYiFgK/bxgc+gQRB7uEOsdIPPI0PuuQTCntQ9mPrP1Ftup5QoIRPL0DqkNfrE2rb3ji2kFaiRmFOClf9cxr8XypeEKhUT4I5fJt0HH3fHfN/O/ramXsznNWqzkgqRdqkAOSlF5bZr3jMSviGWn43gGG+E1PK2w2zJvaJzTTs3ImN1bD09PYWmDOvWrUNgYKDVNi+88IJtZ0gIIcRm2IduN9ZrJPanOXcOKS+VnYB49uiBiGVLIbqhIZKjM5gMGPnHSFzIu8DHUpEU8zrNw5B6QyoN2D65+qAQsJVJRHj94aZ4uGXNaRBCHN/atWv5fWxsLF5++WWHL4UwadIkflOpVEJAmhBCqkPJ0WPIW7+eL7MSCDHr1kJp474319fBZ1h5gxENR2B4g+FCiRtW/iDQLZDf18Qg5qSNR/HHqfRyj7FM2s9GtUHjMMvvbrlUDH8PORxVyoU8bF99BsUFldfkZdy8ZOg6rH61zYtUHTrXtL3bPhuMjo7G559/LoxDQ0Ox/tov5OsPEAVtCSGEkNunPnUaiWPLuiV79uqFyPeXQ+SEmdCzd88WArbsRGpF7xXoHF75ZefPbziKwwl5wnjlyDZ4oPHd1bwj5Fbmz5/PkxL++usvXhbsiSee4Jm3qamp8Pb2FhIXCCGElMn/aTPyNn0N6A3Qp6UJ64OnT7N5wJY5mnlUeB+xru86NPBrAHeZ7RqY2ao+7c/HU5BbbKlPy8RnF+FCRhFS8tUVfk2dIA/8Na17hbXVHaXkgdFohtlk5oHpuP3p2PP9pUq3lykkCIr2QuMuYYhpZp0ISAi5i6Dt1atXb3dTQgghNRS7XPa///7jy127doVEQoX97Ul7JR5Jzz8PU5HlMjGJjw9C589zuoBtVkkWZv03CwfSDwjrVt6/8qYB291X8vFPXNnlgi/eXw+9G1GWOKk6CQkJ6Nu3LxITE6HVavHAAw/woC1rxMvGn3zyCe1+Qgi5TtF//yFt9myhFn8pReNG8Bs50ub7KludjYt5F/lyQ/+GaBXcqsYcj+S8Enzy72X8dzEbCTklt/U1rFbt8PZRiA3wwCOtIxwqYJubVoyTO5NgNJhwfl/5rOGK9BjZADKlBDFNA6Fwc66ryYgFnWvaHv1LIYQQF/tD+s8/lsvKWGMdCtraj/rUKSQ8MRJmvSULQxoWhlo/fA+pv2M2lahMviYfo/4cheSiZGFd14iu6BLRpdKvScgpxsu/XBbGE3vUwdQH6LI5UrVefPFFtG3bFidOnEBAQICwfsiQIQ5V05YQQqqDPjUVKS/PKAvYspJOIhFkISEIf/vtOyrx9OGxD/HL5V9u2RAsU50pLLcObo3qpNYZse1sOnKLdcK3zO7i0lXYfyUXibm3F6gN8VZABBHaxPph6eMtoJA6RgJFdnIhtnx6GsUFWhh0d9a4beSCjvAJdnOooDS5O3SuaXsUtCWEEEKqWc66dch89z3WrtjyxzgsDLGbvna6gO2h9EMYv3089CZLYNpD5oFJLSdhZKPKs280eiP6vL9bGHvIJRSwJdWCXYWwd+9e3nzseqzWbUpKCh0FQgi5hl3+njZ3HkwFBXzsef/9iPzwA4jEd15D9mzOWXx28rM7+hpWx/axBo/Z/HiYTGbojEaeKVuo0fPa+pezi/HriVQcjM+97ecJ91GiW/0g9GoYjEAvBV/HwpUNQr3gLnfMEMyBX+JRkFVxaYdSCncpQmr5gJUS5s3r3aRo+UC00GCMEHLnHPM3BiGEEOKATDodslesRM6nnwrrFI0aIfLDD3lmijM5lnkMT2992uoEa9WDq9A0sOlNv27d3qv8JIl/jUyCbdO6QyapeY1EiPNhGV4sQ+RGycnJvEwCIYQQi4Iff0Lxnj18WcozaxfeVcCWWXVqlbDsr/SHXHLzxlsRnhGY3mY6avvUtunhOJpciFmfnkSBuqwO7Z3wUkrRKNQb/+vbAG1jHe9DeFaLNie1GHu+v8iDs2x8vaI8rbDsH+7BA7ReAUpE1PNDWF0fuHnKofR0rvJehNQEFLQlhBBCqoFRpULCU6OgjYsT1nn17YuI9951uhq2/yX/h4k7JgpjH4UPPnvgMzQOaHzTr9t1IQtLtpXtn5UjWyHC161K50pIqQcffBDLly/HZ59ZMr5YllBRURFvUNa/f3/aUYQQwt7PFBUhc+lSYV+ELngNEm9vq31jMpuwP20/VDpVpaWTfrvyGy7lX0Kxvpiv85J54c9H/rRLU7Gv9idg3i+WRqk307VeIHo2CEaAp1y41J/9f91gTzQKs94H9pCTUoQrx7NQotLd1vaaIj0vd2DUm5CZUHhbXxMc44XHZrW7x5kSQm4XBW0JIYSQariMMGPh22UBW6kUgc+NQ+DzzztdwJaVRHjp75es1q1+cDUa+De46eWIf55Ox0vfHIPeaMns6N/IH93rB1X5fAkptXjxYt6IrHHjxtBoNHjiiSdw8eJFBAYG4uuvv6YdRQhxaawGv/rkSWQuXgJjTo7w4bNXjx7ltl1yeAm+PPvlHT3/k42frPaArdFkxryfT2PDgUSr9f2ahvIPjeVSMb/aJ8xHiUEtw2tUaYO89GL89tEJIQPWdO39ky15+FhnPcvdZWg/0LYZzoSQm7ur3zqXL1/G2rVr+f3777+P4OBg/Pnnn4iOjkaTJk3u5ikJIYQQpz3JSVuwAAWbN1tWyGSI+eILuLeuOR2PbYU1EXl196vCONwjHKv7rEakV+RNa9iOXXsI+65YTgCZ1tG+eLF7VJXPl5DrRUVF8SZk33zzDb9nWbbPPPMMRo4cCTc3yvgmhLgOQ14e8r/5BtpLl2HWamDSaKE5fRrGvLyyjWQyBE+fVu5rz+ScwVfnvrqt12HlEKK9ojGwzkA8Vt/2NWorcjZVhZe/O8EbiqWrNFaPBXkp8N34TogN9EBNdua/FPyzoezKpHvFkoYlUjEkMjHqtA5GVCN/1GkdRI3DCHHEoO2///6Lfv36oUuXLti1axfeeustHrRlb25Xr16N77//vmpmSgghhDig1FmzofrtN2EcOnuWUwZsj2cex9w9c4VxjHcMNvTfwEsjVKagRI9hn+3D+fRCq0sPV41qg7yc7CqfMyGl9Ho9GjZsiN9++40HadmNEEJcre5+/qZNKDl0GEV79sBcUnLT7f2ffBLyqKhyVxYt3L+Ql0dgBtUZhAZ+FV9pU9e3LjqFd6qWwCDLqN18LAUp+Wos3V5xGYRnOoRh9qAWkEgkqOn2bb5sNWaNvmQKCQ++evgq0Pi+cHj4WBqg3YpPsBvkypqTQUwIsXbH/zpnzpyJN998E9OmTbNqytCrVy989NFHd/p0hBBCqpFUKsW4ceOEZVK1VFu2lAVsJRLeqMNn0CCn2+3nc8/jhZ0vCCdpdXzq4LuB30Emqbz0Q1JuCUauOoDEXMtJIbsE8X99GmBM51iIq/78jRArMpmMl0QghBBXxIKtKS++hKK//650G5FCAc+ePaGoUwfSoED4Dh1abpv/Uv7DyeyTwnuB1zq/Bpm4espAJeeV4EyqijczvZhRCLXeyEsuGUwm/BOXheQ8dbmvCfVWQiET45kusXiwtluNzCzVluiRl1EC9haLNQdLOp8LbbFBeHzApOaIbRZo1zkSUorONW3vjs/YT506hY0bN5Zbz7Jts7MpK4YQQmoysViMiIgIe0/DJWjiLiB15ixhHPzyy04ZsGXNRJ764ylojJaAV8uglljTd81NT9IyVBo8ubosYMusG9sOnetYTjpMJkvwl5DqNGnSJLzzzjtYtWoVfahFCHEpeevXWwVsxR4e8Hn4YfgOHwaJjy/ECjlfV1Ed/lNZp/Dz5Z+hN+l5XftSU1pNqbaAbXx2MR764D8U64y3tT2Lza4Z0443FSt935GZmQl7Y0HZ6yvTXj2RjW2rz8BoqPh9UbOekRSwJTUKnWvWgKCtr68v0tLSUKtWLav1x44do0AAIYQQwi7DKyhAyrRpMF/L3PPs0QP+Y0Y73b5hXaEn7pgoBGwjPCPwYa8Pb3qSdiIpn5dE0OjLTkA+H9VWCNgSYi+HDh3Cjh07sG3bNjRr1gweHtY1DX/88Ue7zY0QQqpCydGjyPnscxTt2iWsi1i+HF4P9IboNsoEZKuzMf6v8SjUlZU5Yur71UfP6J63PY9LmUX4+3wm9CYTzOay7F/LPXgg03JvFsYJOcUo1FgyTq/mFN8yYMuu4hnePhq9GwUjys8d9ULKrhq2NxaUvXI8C39/dR56ze0Fnpn2D1nHZAghzueOg7bDhw/HK6+8gu+++45fPsA+ldqzZw9efvlljBo1qmpmSQghxCaMRiP279/Plzt27OgQdbscjSEnB4nPPAvdZUu9MVl0NMIXv1cjL7m7F+dyzmHCXxOQo7E0EPOUeeLj3h/DV+lb6df8E5eJMWvLsnC8FFKse7o92sT4VcucCblVYsLQCi73dTQrVqzgN/b7nhBCKqM+cQIJo0YDhrJL7f3HjoV33z4Vbl+gLcCu5F0o0ZddJbMzaWe5gK1EJMH0ttMhFolva+cXaw0Y9uk+5BTr7vlg+XvIMa5rbcQGuPOmYhKxCDKJGFKJCAEeCr6uJji6NQFndqfCzMo36I1QF+pvur1MKUHjLuEQiUUQiwGJTIL67UKg9KieTGZCbheda9aAoO3ChQv55WOswy47II0bN+b3TzzxBObMmVMFUySEEGIr7Pf19u3b+XK7du0oaGtjmrNnkTBmLEwqFR9L/P0RvepzSDw94UxYZs247eP4CRzjIfPAsp7LUMun4oyPTJUGC349i99PpQnr6gZ7Yu2Ydojyd6+2eRNyM2vXrnWKHcTep7ObSqWCj0/ljQAJIa7LkJeHlKnThICtNDQUfsMeR8Czz/JxniYPX537CvEF8TzjtUBXgJNZJ6E1ait8Pl+FLz7p/QkUEgUC3QJv+gHujXacz7RJwFYmEWHeQ40xuFXNLAOm1xmRfqkAhXka7PvJupHYjcLqlP3uVnrK0OXRevAJcquGWRJyb+hcswYEbeVyOT7//HPMnTsXp0+fRlFREVq1aoV69epVwfQIIYQQxzkBShr/vBCwFXt5IfLDDyCPjoYzYTXrXt39qhCw9Vf68xO1RgGNKq1f+/BHe5CuKmvypJSJsXlSF3gqqBkeIYQQUl0Kfv0NGYsWwZhjuUqGcWvdGjFfrEOGLgdzDryGrJIsnM4+jUK9dQbtzcxsPxNNApvc0VxYSYQLGYX4an+CsO7V/o0QE+DOr05i1yeVXqTE7vkay//44wEecoT7uvEx/z7kEihl9r2CzKAzokSlQ+LZXJzdncrHDCvpkJ9RlqF8PXdvOTx8FZApJDxzttOQOvANoQ+0CSEWd3y2tHv3btx3332Ijo7mN0IIIcTVlRw9htQZM2DIyuJjaVgYYjd8BVl4OJzNR8c+wt7UvUJmzQ+DfuBZNRX55lAiXvnhlNW6GX0a4OkutfjJFSE1CUtCqKiMCVunVCpRt25djBkzBj173n6dRkIIqSl0iYlImztXqLfPSHx9EbFkMUwSMab/O51n01bGU+aNCOl9cDdbxwDcxSHYeyIae09Y/72/maQ8NXZdsLxnKhXspcDT99XiJQ1qGlbCQFNUWkairNbu9c7vS8Ph36/CZLrhgcqIgMdntUNQdM2prUsIcYKgba9evXjDsREjRuDJJ5/k5REIIYQQV1W8dy+SJk6yOgmK+uQTpwzYrj29FmtOr+HLUrEUy3osqzRgu+1MulXA1kMuwYZxHdEy6vYvmSSkOvXt2xcff/wxb0LWvn17oTnZyZMnebD27Nmz6N27N29I9vDDD9PBIYQ4BLPJxMsgpM2dJ7xXkdetA2X9BvAfOwaysDBsOLfBKmArFcsQJOoEUUEPmI1KFJZIkJZvQpqQ13qjxHue52NtI2tMwLYwVwO91sgzZeNPZuPolgSYjLcZjL2BXHntQ2qWHRzhgfB6vhBLxAir7UMBW0KI7YO2qamp2LRpE77++mssWrQIzZs3x8iRI3kQNzIy8k6fjhBCCHFIZoMBWcuXI2fVamGdLCYake+/D2WD+nA2my9txtIjS4Xx8AbD0Ta0bYXbfns4Cf/7vuzkr06QB9aMaYeYAI9qmSshdyM7OxvTp0/nJcCu9+abbyIhIQHbtm3D/Pnz8cYbb1DQlhDiEIr+/RcpL8+AqbCs1AH7UDl20zeQeFr+Jl/Iu4D3j74vPN5c8j8cOueHPOP1AVQWsLRtQNVdLsHjbaMQ4euGQC85+jcLu6vnyU0tRtyBdBiNJmGdulCHnOQiIdDKsmLNJrMlR9ZshtnE7q49xtabLbU4RaKLMBlM0GnurZFj/Q4h8PRVoE2/WMiVVAqKEHL37vg3SGBgICZPnsxv8fHx2LhxI7744gvMmjUL3bp1w86dO+9hOoQQQkjNZ9JoeP3akgMHhHVubdsg+rPPIHZ3vjpkv17+FXP3lAWy+tXqhxdbv1jhtn+dzbAK2LKsmZ8n30f1a0mN9+233+LIkSPl1g8fPhxt2rThPR1YksLSpWUfXhBCSE3OsM1Y9A4KdYUovq4nYfjcF5COfOgKMnE5/zJe2/ca1AY1f0yX1wF70v2FbVniq1QshrebFBF+7nisTSRaRdvmipnYAA94XKttb9SbkJVYeAfZrGaUqPQoyCrB/s1XYDs3D9bWbhVkCV0L8WvLQmllHQ8fBdoPrAW5GwVqCSG2cU+/TWrVqoWZM2eiRYsWPCvh33//tdG0CCGEkJrJpNMhYdRoaE6WBSYDxj2LwMmTIVYo4GzSitIwe/dsYTyk7hC83uX1Crdd8fclvLc1Thh3qh2A1WPawl1OJy+k5mN1a/fu3ctr116PrWOPMSaTSVgmhJCarHjvPuxwv4qPXpLAfH3ZgcQ5FVYzMKqjoM3sx5dZo9CnOsVgcs+6QmC1qrAasN+/exjZSUU2fV4WSJXKJZYmZuz7v9bMTCS2PMjXX7tnj7Hf71KphD/m7iWHb4gbxGIRbxKmcJehUecwCsYSQqrdXf8G3rNnDzZs2IDvv/8eGo2GXyb29ttv23Z2hBBCbEoqlfLajKXL5M6lz39NCNiK5HJErvgInl27OuWuvJR3CU/88YQwbh7YHK91fq3CbbefzbAK2DYM9cIXT7eHXMrOjgip+aZMmYLnn3+eZ9u2a9dOqGm7atUqzJ5t+eBi69ataNmypZ1nSgght5b91Xps6ia2DthWQmmoh6zEJwGTgjcMHd+tNqSS6vn7nRGvuueArVQmxoPPNoHSU87HLNjqH+EB2W02PWUB28zMTAQHB0MspvcthNz1v0U617S5Oz5jZ2UQWE1bVtv2gQcewPvvv88Dtu5OeDkoIYQ4G/ZGNDY21t7TcEis9lnGwrdR8NNPwrqI95c7bcA2pSgFz2x7RrhkkpnfeT7EPEXFmkZvxMI/zgnjKH83fPt8JwrYEocyZ84cfhXZRx99hPXr1/N1DRo04GURnnjC8uEFC+pOmDDBzjMlhBALU0kJ1CdOWJqNXYfVsN2btAtZnSxBy2ivKDQJaFruPWGIWwiOX1Hi32PhgFmKBiFe1RqwZRJOZQvLtVsGwSfI7ba+Tu4uhZunjGfRRjcOgKef813tRIijoXPNGhC03bVrF2bMmIHHH3+c17clhBBCnJ1RpULy5CkoOXhQWBcyeza8evaEMwanvzz7JW9Kojfp+TovuRdW3L8C9f0qbrC26M/ziM8uFpqO/TWtO7/kkBBHw5rrsltl3NxuL5hACCFVTXP2LBKfeRbGvLwKH9/2aFngNRrD4FNkuYLAYDIjs1CLnCItTmkMOJOqEmrQvzG4abUEbFnzr4tHMpCVWIRLRzKE9d1G1Od1YQkhhNxl0JaVRSCEEOKYWGfc0kY7rLGORHJ7l425MmNRMZKeGw/18ePCupBZM+E/6ik4ox2JO7D48GJh7K/0x3cDv0Owe3CF2289k451e6/yZalYhMWPtaCALXFY+fn5vPTXlStX8PLLL8Pf3x9Hjx5FSEgIIiIi7D09QgjhtJcu8YCtujAPmQHld0qRG3C0zrUPT3Xe+POgH4D4Svce+5x16eMt0L5WWROyG2mK9bhyPAs6taHcYwadEfmZan5/O7KTi1CQWXYlDxMU7UUBW0IcHJ1r2ilo+8svv6Bfv36QyWR8+WYGDRpkq7kRQgipgj+kf/zxB19mdRkpaHtrqTNfKQvYSiQIW/AafB991Cl/NpMLkzH1n6nCuFtkN8ztOLfSgO2lzELM+O6EMJ7Yow5aRbMTQ0Icz8mTJ9G7d2/4+Pjg6tWrePbZZ3nQ9scff0RiYiK+/PJLe0+REOLC8n/8CblffgmzXg9DRgbSJUV47TkJsn1ufmWLtqA9ewNT6eNBXgq82r8hBjQO5YHZtMsFyE4q5NmwjNkMlKh0uHw0E9qS8gFbW2nanT4YI8TR0bmmnYK2gwcPRnp6Oi/MzZYrwy6FZAeJEEIIcQa5Gzei6K8dfFmkVCJ24wYoGzeGM7qQdwFTdkwRxm1D2uKjXh9VmjV7Mjkfwz7dD7Xe8ne/fognJvWqW23zJcTWpk2bxhs1vvvuu/Dy8hLW9+/fX6hpSwgh9lD4999Ie/VVSwQVgMoNeOuJWwdszSYp9PntMahFOJ7qFMPXsa8I9FQgyEsOo9qIy4czceKbBHyafwHVycNHjrb9Y+Ef4Ql3Lzl8Q6hHDiGE3FXQlnVTrGiZEEIIcVYFv/+OjNffEMZhb7zutAHbuNw4PPprWfZwiHsIXu/yeqUB27QCNZ5cdUAI2NYK9MAPEzpDIaVyG8RxHTp0CJ9++mm59awsAkteIIQQe9DExSF1+su4EmLGzhZiGBRSXAg1Iy3AEsCN9IxEu1BLvdrr7TyfBVVSE/RWBaBzuhnZW1P4epPRjMt5Gqiy1NBpbj/hir0laNgpDFGN/cu9PxCLRfAOcoPS4/arL7r7KPjXEUIIsWFNW3Zp2LBhw6BQWBcI1+l02LRpE0aNGnWnT0kIIYTUKJrz55E2c5Yw9nn4YXg/9BCcUVZJFp7Z9owwruVTCyvvX4lIr8gKt88v0WHgh3ug0lgukfRSSLFmTDt4KWXVNmdCqgJ7b6tSWRryXO/ChQsICgqinU4IqXb6tDQkTZiAVEUJ3hguQbEbC3KWJVH5KwIxr90HCHYLs/q6TJUWX/66HwOLZWiolyInLh85t3it0NrekMgk8At1R0R9P8iV1z6IFQFKDxm8A1lQlv7WE0JIjQ7ajh07Fn379uWlEq5XWFjIH6OgLSGEEEemz8hE/GOPA3o9H7u1bImwhW85ZXOtpMIkTPhrAgq0BcK6VQ+uqrSGrdlsxoSvjiK7SMvHCqkYW6Z2Q4SvW7XNmZCqwvoyvP766/j222/5mP2bZ7VsX3nlFQwdOpR2PCGkWmkvX+bNxopy0vHe6NKAbRmTwQNJV57EiOOsrEH50gZyM1BHX/EVMCzD1dNfwQOxfqEeaNI1HAERnlX2vRBCCKmmoC07YavoxDU5OZk3biCEEEIcFWvukfrKK0LAVhIQgKhVn0Mkcb7L/kv0JXhh5wtIUCXwsZvUDb8M/qXSgK3JZMaEDUew70pZrs7GcR0pYEucxpIlS/Doo4/yxAS1Wo3u3bvzsggdO3bEW2+9Ze/pEUKcXEbKBfz81wqotCpWjBaaU6dhqqvG+R5iJAVZzr+N2iBoUkYAZglMel/AbH316/Xq6iWQ8Qq2QKMuYWj/UG2+LBIDbp4yiCXiavrOCCGEVHnQtlWrVjxYy273338/pNKyL2XNx+Lj43kGLiGEEOKIzCYTkiZMRMn+/XwsUigQ9cknkHg6Z+bJWwfewqX8S3w5xjsGi7svRqhHaIXbFmsNeOmb49h+NkNYN757bbSJ8au2+RJS1Vjywfbt27F7926cPHkSRUVFaN26NXr37u1QO3/FihX8Rs2BCXEMeVlJ+OSHWfhBegJaVn2gNA7Ly9SWfWhsNiqgSRmFgY1aQCq+dcA19KSKfULLl1ktWk+/ygO8hBBCHDxoO3jwYH5//Phx9OnTB57XncTK5XLExsbSpWOEEFLDsQ/cSrugX//hm6tjAdu0WbNRvHs3H4tkMkQsWwa3Zk3hjD469hF+ufwLX5aL5fig1weo7WPJwKnItG+tA7ZzBjTC011qVctcCalu9913H7+VOnr0KObNm4fffvvNIQ7GpEmT+I3V56Wr4Aip2bbt/AnzL8xD0S2qDJnNYqhTh+PFrl3wYu96t3xeTbEea//ZzavfsmBtWG26IpYQUvXoXNP2bvuMff78+fyeBWdZIzKlUlkF0yGEEFKVxGIx6tevTzv5hrI/KdOno/DPLaU7CZErV8Kza1nQxpn8fuV3fHryU2E8pumYSgO2bN+8uzUOW89YArYyiQjvD2+F/s2sG54Q4ui2bt3Ks2xZIsKzzz6L2rVr4/z585g5cyZ+/fVXnrBACCG2lJCZiyXHXkdRoGUsNZhR50oYijRNePcvtUSBZM8gVmAbJl0AutdqiKHRgTj8x1WkXymAtsTSELQiWrWBlzVi6rYJhkjsfHX5CSE1D51r2t4dp1mNHj26CqZBCCGE2EfOZ5+XBWwBhMx51SkDtiwAu+b0Giw/ulxYN7TeUExsMbHS7V/YdBy/nkgV1rHsWgrYEmezevVqjBs3Dv7+/sjLy8OqVauwdOlSTJkyhScqnD59Go0aNbL3NAkhTqSgRI+5Hz6P1LqWwGtYjgRuxVOQ61dH2IYVRoi5tlwn2gOD1Ar8vPTYHb9WvXYhNps3IYSQGh60ZfWxli1bxjvrso66Op3O6vHc3Fxbzo8QQogNsd/hp06d4svNmjWDxAkbbN2J7E8/Q9ayZcI4ZPZs+F8rH+FMWAB21J+jcDzruLDugZgHML/T/AqbiybmlODVzafw38VsYV2vhsGY9iBlaRPn8/777+Odd97BjBkz8MMPP+Cxxx7DypUr+e/KyMhIe0+PEOJkcot1eGnllzgdc46PRSYzZrd+FT16PFbp15z6Jxm7Nl24o9dhf94bdAxFULTXPc+ZEEJuB51r1oCg7YIFC3gGwvTp0zFnzhy8+uqruHr1KjZv3szrfRFCCKnZf0jZ72umcePGLh201V65gqz33xfGAc8+A/9RT8HZ5Gvy8dz253Au13JyyAypOwSzO8yuMGB7NbsYj326D1mFWmHdjD4NML5bbUip0zRxQpcvX+aBWuaRRx7h9djee+89CtgSQmyKlSvYdCgJ72w5i2jZp9CzpmMABqjq3zRgm59Rgr0/WBqHMl0erYvY5oHwDnTDrYoeUFkEQkh1onPNGhC03bBhAz7//HMMGDAAr732GkaMGIE6deqgefPm2L9/P1544YUqmCYhhBBiO7qrV5H49DPsDIqPlU2bImj6dKfbxbuSd+Hlf1+G2qAW1s1qPwtPNKo4m5gFanss/kcYsxq2/+vTEOO6Vd6kjBBHp1ar4e7uzpfZBxkKhQJhYVS3mRBy51e1rPznMradSce1crJWCtR66DKOIdb/O3iKOqLrlSC46UXoVX8g/tkYV+nzpl7Mh0Fveb/StHsEWvaOpkNDCCEu4o6Dtunp6fySWsbT0xMFBQV8+aGHHsLcuXNtP0NCCCHEhgzZ2bjct58wltepg+h16yrMOnVURpMR7xx6B1+f/9pq/fQ20ysN2J5KLsCkjUeFcYSvGzY91xFR/pZgFiHOjF1Fxt7XMgaDAevWrUNg4LXuQNdQYgIh5Ka/R/6Lx7e/fY1A6WFAVD5qGywpgSbCFx0TpsFbGyCsv7D/9soL+gS7ofMjdekgEEKIC7njoC2r7ZWWlobo6GieYbtt2za0bt0ahw4d4pkJhBBCSE2WvfJjq3H0559B4ukBZwrYPvnHkzidc1pYF+UVhTe6vIE2IW0q/JoDV3Iw7LP9wthbKcXqMW0pYEtcAntPy64iKxUaGor169dbbcM+1KGgLSGkMn+eSsPXP3+BwkY/IlfGYrZiROU3RExeU8iMcr6NUu+J6At319RQqpDggbFNIFO4blkrQghxRXcctB0yZAh27NiBDh068K66Tz75JO+6y5qSTZ061eYTTElJwSuvvII///wTJSUlqFu3LtauXYu2bdsKl6HMnz+fv9nOz89Hly5d8PHHH6NevXpWzdHYXH/99VeIxWIMHTqUN50ozagghBDiGrJWrkTexo3COGbjRsjCw+FM3jrwllXAtm9sXyy8byFkkmvF826w93I2Rq85KIybRnjj45FtKGBLXAbrzUAIIXfrWGIeXv/qZ4hr/wiFyQ9tr3ZHvew2cNd7V/o1oXU80WlI/dsOwnr5K6H0qPjvOCGEEOd1x0HbRYsWCcvDhg3j2Qn79u3jQdKBAwfadHJ5eXk8CNuzZ08etA0KCsLFixfh5+cnbPPuu+/igw8+wBdffIFatWrxEg19+vTB2bNnoVQq+TYjR47k2cHbt2+HXq/H2LFj8dxzz2HjdSfuhBBCnJv20iVkf/ChMA6cOBHurVvB2WrYfnfhO2E8seVEPN/8+UpLP2w4kIBXfyoL8MolYqwd0x5BXnTlDCGEEHIjg9GEj/6+hMtZxcK6w3Hx8A76FEXSADx6YjqUxsqv3pEpRbjvsQZo1DnMqcoyEUIIqSFB2xt16tSJ36rCO++8g6ioKJ5ZW4oFZkuxLNvly5djzpw5ePjhh/m6L7/8EiEhIbw7+vDhw3Hu3Dls2bKFl28ozc798MMP0b9/fyxevBjhFWRYabVafiulUqn4vclk4reqxl6DfW/V8VqkatAxdGzOfPxKv7fSZWf8His7hqlzyuquK1u1hP/ECU71/RdoCzBpxyRh/ED0AxjfbDzfD6XH/HoXMgqtArb1gz2xeVJnKGUSu+8XZ/436Aqq+/jRzwkhpLqs35+A5X9dFMYiswEt3BcjwV+MIaeeFQK2YokIsc0D0bBjKPzCPKwyZiVSMR0wQgghtgva/vLLL7f3bAAGDRp029vezuuyrNnHHnsM//77LyIiIjBx4kSMGzeOPx4fH88bo/Xu3Vv4Gh8fH166gWX/sqAtu/f19RUCtgzbnpVJOHDgAC/3cKO3334bCxYsKLc+KysLGo0G1XHywRq8sRMeNk/ieOgYOjZnPn7se+vVq5dQOsbZvr/KjqFu505ojh8XHle8+iqysrPhTJacXiIstw5ojekNpiMzM7PCbY8mF2Len/HCuEstHywcUBuqvBxYPqa0L2f+N+gKqvv4FRYWVvlrEEKIyWTGl/sShB0RoT8D/4DvcDlcg94XRyNAbUkG8gt1x+BpreHuballSwghrkIqlfL4XekyuXe3tRcHDx58W0/GLvEwGo2wlStXrvD6tNOmTcPs2bN5tixrAiGXyzF69GgesGVYZu312Lj0MXYfHBxs9Tj74fH39xe2udGsWbP4a16facsyfll5Bm/vymsT2fJkh+1L9np0suqY6Bg6Nmc/fqzJjisdQ0NKCq6+t1h4zP/ZZxHUuDGcyf60/diSskUYz+8yH2E+YRVuezK5AJN+OILS5FtfNxk+erIdfNxqTq08Z/836Oyq+/iVlsMihJCqtPdyDuKzi+FpzEErz4042SAZKrEILVJ7om5Oa76NTClBv+ebUcCWEOKS2Pu+Jk2a2Hsarhe0tddlZ+x1WYbswoUL+bhVq1Y4ffo0PvnkEx60rSoKhYLfKvoBrK6TR3ayU52vR2yPjqFjo+PnHMfQkJCAhCdGwqxW83XKFs0RPG0qRE70u/X3K79j5n8zhfH45uNR2692hdvuv5KDJ1cdEAK2MQHuWP90B/h51LwatvRv0LFV5/Gj90qEkOqwfv9VROiS0EsEeGWPR7NrF+y4672EbXqPaQy/0Mpr2hJCCCF3okaftYaFhaHxDdlQjRo1QmJiolW2WEZGhtU2bFz6GLu/8fJQg8HALwt2hWwzQgi58cOwM2fO8Juz14E0JiQicfQYmAoK+FgSGIioTz5xqoDt6ezTVgHbpgFN8Vzz5yrc9sMdFzH8s/0wmCwR24ahXtg2tRuiA9yrbb6E1HSXL1/mvRJGjBghvH9kzXDZ70xCiOtKK1Bj+9kMdDZnIbS4Pjz0PsJNdO2Uuu2AWNRuGWTvqRJCiN240rlmdbnjIhOvv/76TR+fN28ebKVLly6Ii4uzWnfhwgXExMQITclY4HXHjh1o2bKlUMqA1aqdMGECH7Mmafn5+Thy5AjatGnD1+3cuZP/ALHat4QQ4krYh1bfffcdX2ZlZ1i5GWdkyMtD0ezZMF2rW6uoVw9Rqz6H1M8PzuJ45nGM2TLGKmC7us9qyCXljyk70Vz61wVh3CTcG5+PaguFVFJt8yWkpmP9E/r168fff+7atQtvvfUWL7F14sQJrF69Gt9//729p0gIsZOvDyRCaihBkMFy2a8ZJrh7SyGRyviVBbHNAtF+QFnDbEIIcUWucq5Zo4O2P/30k9VYr9fzhmCsTmydOnVsGrSdOnUqOnfuzMsjPP744zh48CA+++wzfmPYH8iXXnoJb775JurVq8eDuHPnzkV4eLhQh5dl5vbt25c3L2NlFdh8J0+ezJuUse0IIYQ4F2NBAZKefAqmlBQ+lteujegv1kHq7w9noTao8dSfTwljf6U/Prz/Q7jLrLNmdQYT5v9yGl8fTBLWdasfhM+eagOljAK2hFxv5syZ/D0l62vg5VV2uTNr3vjRRx/RziLERbG/pV8fSkJn3Wl46LvzdcWelzH53fH2nhohhBAnd8dB22PHjpVbx7Jbx4wZgyFDhsCW2rVrx4PErDEYy/BlQdnly5dj5MiRwjb/+9//UFxcjOeee45n1N53333YsmWLVVOKDRs28EDt/fffz+ueDR06FB988IFN50oIIcT+dMkpuDJwoFDDVhIQgOjPP3OqgC0zd89cYVkEEb4e8DUC3QKtttl9MRsvf3cC6SqNsO6h5mH4cEQr/qEnIcTaqVOnsHHjxnK7hWXbZl/L2ieEuJ5tZ9ORVahFpNhHWNe4edkyIYQQUmOCthXx9vbGggULMHDgQDz1VFnmjy089NBD/FYZduLJAro3K9vg7+9f4ZtwQgghzsNUXIyUF18UAraQyRC++D3IIiLgTFYeX4mtV7cK4zV91iDc0/rKkRV/X8J7W63LCw1sEY5FjzSjgC0hlfD19UVaWhpPErgxYSHCyX6PEEJu3/p9CQjVpyJIYymNoJUUoc+wR2gXEkIIqXI268ZSUFDAb4QQQkh1Kzl6FBe7doPmumZBXh99CHcnq13+yYlP8PGJj4Xxo/UfRdvQtlbbfLH3qlXA1tddhrVj2/EMWw+FTT6rJcQpsdJZr7zyCtLT0/mHG6z/wZ49e/Dyyy9j1KhR9p4eIcQOzqaqcCA+F+2NiZCaZXyd2D8RcgXVaSSEEFL17vjs7cayAmazmWclrF+/njdvIIQQQqpT8f4DSBxT1pCLCX3rTWgbNHCqA7H29FqsOL5CGPer1Q9zOsyx2uZEUj5e+7UscN2zQRBWjmwDNznVryXkVlgPhUmTJiEqKgpGoxGNGzfm90888QTmzLH+t0YIcX4avRHTvzsBmE0IQm1hfc++7ew6L0IIIa7jjoO2y5YtsxqzGrFBQUEYPXo0rz1LCCGEVBf1yZNIev55YSwNDUXUZ59CXrcuMjMzneZAbDq/CUuPLBXG/Wv1x9td34ZYVHbBjFpnxMMr9gjjKH83rBnTjsohEHKbWIfjzz//nDe1PX36NIqKitCqVSve7JYQ4nre+v0czqWp0FJ7Cv6ajnxdkSIVLbo8ae+pEUIIcRF3HLSNj4+vmpkQQgipchKJBIMHDxaWHRkriZD03HiYNZZGW/I6dRC74StIfH35Zc3O4kz2GSw8sFAYD6ozCG92ebNcMHbp9rKSCPVDPPHjxC4UsCXkDuzevZs3tI2OjuY3Qojr+vNUGtbvT+DLDcRl7ylCYkrsOCtCCKnZnOlcs6ag4naEEOJC2B/Pli1bwpGxsjyZ7y1G7po1wjq3tm0Q/emnEHt4wJnkafIwY9cMmGHm4ycbPYlX2r9Sbn/8cDQFn/9X9qHqe4+2gCfVryXkjvTq1Ys3HBsxYgSefPJJXh6BEOJaCvOzsWbtR4hLLkBfs+Vvb5CuB783igwY+MQQO8+QEEJqLmc413T4oK1Go8GHH36Iv//+m196emM209GjR205P0IIIcRK+oIFyN/0jTD26NwZkSs+gtjNzan2FAvGTvxrIpIKk/g42C0YU1pNKbfN+PVHsO1shrBueLsotIjyrfb5EuLoUlNTsWnTJnz99ddYtGgRmjdvjpEjR/IgbmRkpL2nRwipBkvf/QSBud3QpILHNJ5XEBD6IB0HQgghNTdo+8wzz2Dbtm149NFH0b59e7r0khBCHAj7oO3SpUt8uW7durwuuaNgAcrcdV9YBWy9+vZF+DuLIFYo4Gz+SfoHp3NOC+P3ur8Hd5m7MDaZzBj/1RFsvy5g26VuAF4d0Kja50qIMwgMDMTkyZP5jZUD27hxI7744gves6Fbt27YuXNntc9pyJAh+Oeff3D//ffj+++/r/bXJ8SVbPtpA/zyOlT6eMduZc3ICCGEONe5ptMEbX/77Tf88ccf6NKlS9XMiBBCSJUxGAw8EMHMnj2bN95xlIBt9scfI/uDD4V1svBwRCxdApETvhlIL07HnD1l3eoX3rcQrUNaW23z7JeHsfN8WbO1F3rVxdQH6tOHqYTYQK1atTBz5ky0aNGCNyb7999/7bJfX3zxRTz99NM8eEwIqTo6rRp7d6sQZA7j40xZHKKivOGltJwu124Ui3YPVB7QJYQQ4rjnmk4VtGW1vry8vKpmNoQQQkgF0l9/HflfbxLGsogI1PrxB6cM2LI6ttP/mQ6VTsXHncM746HaD1lts/titlXAdkT7aEx7sEG1z5UQZ7Rnzx5s2LCBZ7aysmAPP/ww3n77bbvMpUePHjzTlhBStdZ+/D6Citvz5RJpLhS9u+DZh5vTbieEEGJXd3y2u2TJErzyyitISLB00ySEEEKqUuay5VYB26Bp01B3x1+Q+Pg43Y5nGcUv/f0STmaf5OMQ9xC82+1dq+zZo4l5eGrNAWE8uGU43n6kmV3mS4gzYWUQWIYta0iWmJiI999/H+np6Vi/fj369u17x8+3a9cuDBw4EOHh4fzf8ObNm8tts2LFCsTGxkKpVKJDhw44ePCgjb4bQsjtSku6hLwrZaUP4pSZmNKHSg0RQghxwKBt27ZtedZB7dq1ecatv7+/1Y0QQgixlaJdu5Dz6afCOGTuHAQ+N85pd/A3cd/gaGZZQ8/F3RfDR1EWnN53OQePrNyLaw2t0SraF0sepw6thNgCC7LOmDEDKSkpvBwYa0Dm7l5WR/pOFRcX8/IKLDBbkW+++QbTpk3D/PnzeSNftm2fPn14o19CSPVZ89lGeOkC+XKuLB69HhsAL6WMDgEhhBDHK4/A3sCyN7MLFy5ESEgI1c4jhBBSJfK+/hrpC14Xxn5PPAH/kSOddm/H5cZhyeElwvjtrm+jZXBZQHbDgQS8+lNZY7Jof3d8+mQbSMRlWbiEkHsri2BL/fr147fKLF26FOPGjcPYsWP5+JNPPsHvv/+ONWvW8Hq6d0qr1fJbKZVKJTQFYbeqxl6DXS1QHa9FXE9V/Xwd+PcPeOS0s7wGTIgL9MTMNpH0c+xi6PcXoZ8v2/5bKl2m9wSVu919c8dB271792Lfvn08G4AQQgipCqotW60CtspmzRAy8xWn3dlGkxEv/v0iNEYNHw+uO9iqju33R5KtArYRvm74ZXIX+LpTcX9C7sUvv/zCA6symYwv38ygQYNstrN1Oh2OHDnCSzKUYh2We/fuzd9n3w1Wd3fBggXl1mdlZfGr5Krj5KOgoICfrFG3aOIIP1/FBVps/e0KgkxN+ThFdhrD+/dHTnaWTZ6fOA76/UXo58t272/YlUYMu3KIGpFVrrCwEFUStG3YsCHUavWdfhkhhBByW0oOH0bKyy8LY3lMDGLWfwmRk3YfZSegr+55FSlFKXzc0L8h5nScIzx+OqUA//v+hDB+sHEI3nusBXzc6NJNQu7V4MGDed3a4OBgvlwZVpPWaDTabIdnZ2fz52NXrV2Pjc+fPy+MWRD3xIkT/AQoMjIS3333HTp16lThc7IAMCu3cH2mbVRUFIKCguDt7Y3qCHqw/cRej4K2pCb+fBn1JmRcVSHxTC4Sz+YgJ7kYQbAEbLWSIuQ1a43+beraeObEEdDvL0I/X7YL2np4ePBl9t6KgraVY/0MqiRou2jRIkyfPh1vvfUWmjVrxjMTrlcdbwoJIYTcHYlEgv79+wvLNS14mf3xx8j+4ENhndcDDyBi+TKIathcbenXK7/i9yu/C+MZbWdAIVHw5ewiLaZ/ewKmazVsR7SPxsIhTak0ESFVcGlaTbyE76+//rrtbRUKBb/diAW4qiuIyoJq1fl6xLXcyc9XiUqH9CsFyEkpQm5qMb/Pz1TDXPoH9TqsLMI56SXMfmQ8/ey6MPr9Rejn696x+OCAAQOEZXo/ULnb3Td3HLQt7Z57//33lzvZtnUWAiGEENtigdr27dvXuN1qzM9Hyoz/ofi//4R1Hp07I2LJYqcO2J7JPoO5e+YK4xdavYD2Ye2Fv6vPfnEYcRmWS2dCvZWY3b8hBWwJqSJffvklhg0bVi7wybJGNm3ahFGjRtnstQIDA/nv44yMDKv1bBwaGmqz1yHEFbEA7XeLDvPM2pvJ9EhEku856DWX0aDT66gVaMkOI4QQ4lznmo7sjoO2f//9d9XMhBBCiEsy5OUh4YmR0MXHC+t8hw1D6KuznbYkArM3dS/Gbx8vjAfVGYRxzccJ438vZOF4Uj5fVsrE+OSpNtTNmpAqxBqCseQEdjnfjTXH2GO2DNqyywXbtGmDHTt2CGUZWKYvG0+ePPmennvFihX8RokUxFWd+S+1XMBWLBXBL9QDARGeOJy6Fz+GrIZGVgQ3DdDe/Dpe7tPAbvMlhBBCbBa07d69+51+CSGEkBqCBQUSExP5cnR0tN0vWdElpyB+yBCYrivEHv7eu/AZOBDO7GLeRbz090vCuK5vXbza4VVhnKHSYNq3ZXVsFz3SHC2jfKt9noS4ktKrxm6UnJwMHx+fO36+oqIiXLp0SRjHx8fj+PHj8Pf3579/Wf3Z0aNHo23btjwrZfny5bx2LQsQ34tJkybxG6tpezfzJsSRsfIHl49l8mWJVIz7xzTigVqfYDcYzGbM2Lgfpz0+hEZmCep2L+iCd14eDLG4/L99Qgghjn2u6ZJB2127dt308W7dut3LfAghhFQhg8GAdevW8eXZs2fbtTg8C9gmjh5tFbCN+vQTeDr5h4MX8i5g1J+joDZYmnpGeEZgxf0r4C5z5+Ok3BKMXXcIucU6Pu5Qyx+DWoTbdc6EOLNWrVrxYC27sfJfUmnZ22OWrcqCraXlwe7E4cOH0bNnT2Fc2iSMBWrZ72FWiiErKwvz5s3jzdBatmyJLVu2lGtORgi5fayObUmB5e9nUD0fXHUHfj+TgjPbVTiZnI9I1QfIamQJ2NbJVmLR9I8oYEsIIU54rumyQdsePXqUW3d9VgJdikUIIeRWNHFxiH/0MUCvt/wxCg5G1OefQdnAuS9P3J+2H+O2lZVAiPWOxdcDvoan3JOPjSYznlt/BJcyi/g4xFuBFSNb0wklcRh6rRGXj2ci/nQ6ousb0aRrBGq60vIELAu2T58+8PS0/Htk2MlGbGwshg4dirt5z8yyd2+GlUK413IIhJAyl45asmyZ1QkZOPtlqjAO1cfhSuMr7OwVEqMZb/ZaCInkjk+HCSGEkGpzx3+l8vLyrMZ6vR7Hjh3D3Llz8dZbb9lyboQQQpyQPjMTiWPGCgFbea1aiP5iHWQ31JF0NnG5cVYB2yivKLzf830hYMuCO3N/Po1zaSo+dpdL8M1znRDoWb4bPCE1TXJcHk79k4zE0zkwXKsladZJHCJoO3/+fH7PgrMs+1WpVNp7SoSQu6Aq0eHonlR+gmuEGVekWkTo4xAoOg+pIgm5YZkollqSjR7TNkPTVg/QfiaEEFKj3XHQtqLaWA888ADPRGCXfR05csRWcyOEEOJkWGAyZdo0GK99ACgJCkT02jUuEbAd/ttwYdwprBMW91gMb7m3sG7G9yfx/ZFkviwVi7BqVFvEUidrUsOYjCZoig1QZauh0xhQmKPB5WNZSDqbW25b9pgjYWULCCGOaffFbLy34SQe1FrGuYo4eNb7GCo5YPkotExwoRgvjV5pj2kSQgghd8Rm14Ow+ltxcXG2ejpCCCFOSPXLL1Aftny4J1IqEbNmDWShoXBmV/KvYPSW0TCYDXwc7hGOD3p9AKXUks1nMJowas1B7L2cI3zNksdboHPdQLvNmZDrm/rkphcjJS4fFw+lI/NqIUymm1/yH1LLG7FtvNCwTbRD7UhW4mvZsmX49ttveRMNnc5SF7NUbm75wHRNtGLFCn6jkmXEmRRrDUjO10ItKbZqbMPKCq3ZE4+v9ieiRwk7tZXx9acjj0BbQSnFmAI53ui4AB5eftU5fUIIIaR6grYnT54slzWVlpaGRYsW8QYKhBBCSEUMWVlIf/0NYRyxfBkU9eo59c5ifyMX7FuAYn0xH4d5hOH9Xu8LAVvmx6MpVgHb8d1r4+GWNf+ScuKcdGoDUi7m49LhDF7yQFOsh8lw8yAt4+YlQ4v7o9C4SzgUHlJkZmbC3cexSnssWLAAq1atwvTp0zFnzhy8+uqruHr1KjZv3sybhTmKSZMm8ZtKparwCjlCHM3B+FyMXXsQxTpjpdt46bPQ1OgFwA9GkRHx/qfgWyJCU20QGnnVRbOY9mjR/AH4BzvWh0mEEEJc2x0HbVlgljUeu7GxQseOHbFmzRpbzo0QQoiT0KenI3Hs0zAVW4KX3g89BK8KGls6mw3nNuBo5lG+HOoRih8H/SjUsGWOJubhjd/OCuMnOkRjZt+GdpkrcV3sPV1+RgkuHs7E8b8SoddUHhgRi0WIbOgHrwAl3LzkULhL4RvijvB6vpArLW8rTSZLTVtHs2HDBnz++ecYMGAAXnvtNYwYMQJ16tRB8+bNsX//frzwwgv2niIhLicptwTPf3XkpgHbhvq/IA+7ALdL0/k4xScOvYtDMHfMOnh6B1TjbAkhhBA7B23j4+OtxuzylKCgIGraQAghDkAikfA65KXLVc1sMiF75cfIWbMG5pISvk7s5YXgly0nVs7sr4S/8M6hd4Tx/9r9zypg+8epNEzcYAnoMvc3DMbCIc2qfZ7ENUseFGSpkR5fgHN70pCZWAiDtnxARCoXwzvQjd/C6vrAN9gdMc0CIJGUXZrsTNLT09GsmeXfoKenJwoKCvjyQw89xBvuEkKqV6FGj2e+OAR9QQraKL+E2icfYpGlkVgpk8iMpEA9OiYMEta1ivbC8Im/0uEihBAnP9d0BXcctI2JiamamRBCCKly7I9nly5dqmVP61NTkfH2IhRu3y6sk4aGIurTT52+jm22OhtvHXhLGA+uOxgPxJR1qd57ORsvbjomjGMD3LH4sRbVPk/iWtKvFODo1gSkXsyHtsRSY/lGLB5Su1UQ6rcLRUzTAEhkzhmgrUhkZCQv+RUdHc0zbLdt24bWrVvj0KFDUCgcq9QDIY6O1ap9adNxpKXEIzRsOS4EVp5pCzNQO6eF8Dts8Kgnqm+ihBBC7HKu6SpuO2i7c+dOTJ48mV8e5u1d1u2aYZkInTt3xieffIKuXbtWxTwJIYQ4EPWZM0h6dhyMeXnCOq++fREyayZkISFwZhnFGXji9yd44JZpGtAUr3V6TXj877hMjF17SBg3j/TBqlFt4edRQccUQmxQo/bc3jRcOJiOzITCCrcRS0SIbOCHsHq+qN8uhGfWuqIhQ4Zgx44d6NChA6ZMmYInn3wSq1ev5k3Jpk6dau/pEeJS3t16HgfPnEFY2PtIuy5gK7qhRB8TmxMNb62leScr36L0tDQjI4QQQlwmaLt8+XKMGzeuXMCWYU0Oxo8fj6VLl1LQlhBCajBWa5JlkjFhYWFWHZhtpWj3HiRPmQKzWs3HYm9vhL35BrwffBDOjgVqe3/fWxizOrZLeiyBRGy5POir/QmY9/Np4fHW0b7Y9FwnyKWuk81IqjZAe/FwBrISC5GbVozc1OJKM2qDor14Ji2rRRvRwI/XqnV1rKluqWHDhvGM23379qFevXoYOHAgHMWKFSv4zWi8SWYiITXY90eSsWHnIYSHsICt5XeYX7EI7zV/C+06Dyj33mXfT5dx9GICX67TOtgucyaEEFI955qu5raDtidOnMA775TV5rvRgw8+iMWLF9tqXoQQQqqAwWDgjXaY2bNnQy63bXZnyZEjSJ44EWadjo9l4eGI+Wo9v3d2WqMWo/4cZbXu4/s/Rrin5XvfcjodczaXBWw71PLHhyNaUcCW3LXiAi0yrqh4gDbuQDpvJnYrLXpHocOg2pDJqc7YrXTq1InfHM2kSZP4TaVS8cQKQhzJkYRcvPXd34gIWYbUoLKA7WddP4JvUP0KGyleOppZVt6lZVC1z5kQQkj1nGu6otsO2mZkZEAmq/xSE6lUiqysLFvNixBCiIPRZ2QgYdRo4Fp2l2fv+xHx3nsQuzn/pdZqgxqz/5uNpMIkPnaTumFj/42o61eXj3ecy+Ddr0uN7RKLuQMaU3YjuS06jYHXoU2Oy0NhjgZ5acUoytNCX0HzsOt5+il4qQOvACUi6vvx0geuVKP2dvzyyy+3ve2gQWWNjgghtpecV4IX12xDcMBSIWDrWyLC590+Qr3G9yEz0xKcvV52chFUWZYre8Lr+8HNiwIEhBBCXDBoGxERgdOnT6NuXcsJ6I1OnjzJ058JIYS4HpZZm/DUKCFgq2zSBBGLF0OsVMLZ5WpyMfGviTiTc0ZYt6zHMiFg++PRZEz79oTwWLMIHwrYklv/XKUV4+KhDKTE5SEjXgWTqXwdxxuVZpk16BiKsLq+UHpQXcdbGTx48G39NIpEIio3QEgVKtEZMHHVNnh4voPU4OsCtvd9gAZNu/FLbity+VqWLVO3NWXZEkIIcdGgbf/+/TF37lz07dsXyhtOwtVqNebPn4+HHnqoKuZICCGkhstZsxb6xERhHLF0iUsEbA0mA57f/jzO5Z7jY6lIioVdF6JLhKVraoZKg9d/Oyts3y7WDx+MaEUZtqRCJqMJaZcLcPyvJFw9aWlkd7MsWp8gNwRGecE3xB2htX3gG+wGKZU9uCOVBYIIIdVr6eaDaJR5DgrzFCADkJhEqO0Zg1O/e+LU74fZx8PQ6/WQydgVLWU1uPPTiy0L7EOrVlTPlhBCiIsGbefMmYMff/wR9evXx+TJk9GgQQO+/vz580Kzg1dffbUq50oIIaQGyvv2W2QtXy6Mw95+G/KYGLhC0zEWsI3Li+PjQLdAfNDzAzQLaiYEbB/7ZB/yS/R8HOylwDfPdaKALamwNu2xrYk4sdNSXuNGLCgbUd8XwbHeCKnlDU8/JRRut/0WjhBCarTTKQUo3LsL0ZoOVutVJYAKqhu21lT4HOF1feHuTaURCCGEOJfbfscfEhKCvXv3YsKECZg1axYv+l56uVifPn144JZtQwghxHXkffcd0ufNF8Ye990Hn8EPw9mdzTmLsVvGosRgafwkEUmwpPsSIWCrN5oweeNRJOZaHg/0VOCHCZ0pYEtQnK/lDcOKVVokn8tDVlIh8tJKYDRYZ3xK5WK07B2NJl3DeZCWVL3XX3/9po/PmzePDgMhNmYwmvD+pxvRXNPmurVmS72XG7HTzwpWu3vJ0f6hWnRsCCGEOJ07StOIiYnBH3/8gby8PFy6dIkHbuvVqwc/P7+qmyEhhJAaqfCvv5A+tyyI4TN4MMJeX8A/zHP2DNsX/35RCNj6K/3xZpc30TqkNR8XlOh507FDV/PKMmzHd0KUv7td502qF6tBm51UiLRLBdCqDTBojUi9lI+MqypL4KESfqHuaNYjEnVaB1PWWDX76aefrMbsUuz4+HjebLdOnToOE7RliRSlV8ERUtOt2XURMQVKiCHhY0P4cbw4b1qFpUxYI7Lg4GCIxdRQkRBCiGu4q2vrWJC2Xbt2tp8NIYSQKiWRSNCjRw9h+W4V79+P5MlThLF3/34Ie3uh0wdsi3RFGPH7CKQXpwsB200DNiHMM0y4xPPJ1QeEkghyqZjXsK0V6GHXeZPqw7JoT/+bgvP706AtsTTTuRW5UoLY5oFoP7AWfIIouG8vx44dK7dOpVJhzJgxGDJkCBzFpEmT+I3N3cfHx97TIaRSSbklOP3zd2ig7czHhfJ0TJn6LO0xQghx8XNNUoYKohFCiIv+Ib1b2Z99jqxly4SxW8uWCH/vPacP2OZp8jBpx6RKA7bZRVqMXXfIOmA7vCU61g6w67xJ9WTVsg7mLFibejH/ptsq3KWIaRYA/zAP+IV4IKKBLxTuMjpMNZS3tzcWLFiAgQMH4qmnnrL3dAhxGuyKzTe/3IoG+pbCuuadxfDw8rbrvAghhNj3XJNYo6AtIYSQOwvYLl0qjN3btUPkio8gcvJPUlmg9qk/nxICtszHvT8WArbn0lSY+s1xZBVq+TjMR4n1z3RA3WBPu82ZVE/Q4fLRLOz/+TIKMtXlHo9uEoBazQPgFeAGmUIMd28FvIPcqLaxgykoKOA3Qojt/HoyDUFXr0JhbGH5d+Z9Cv2Hv0i7mBBCCLkOBW0JIcTFgkxZWVl8OSgo6LazY9nXZSx8G3nr1wvr/EY9hZCXX4ZI7tzdmtn3/u6hd4WArbvUHct6LkPjgMZ8/MepNLy46Rj0RrOQYfvl0+0pYOtkzCbL8TXoTTjw8xWkXc5HZkJhhZm0TbpGoFmPCGog5mA++OCDcv/209LSsH79evTr189u8yLE2eSX6LB5/Qa0udZ8TCspxlPj+tp7WoQQQux0rkkqR0FbQghxIayxzsqVK/ny7NmzIb/NgGvOp59aBWxZSYSQWbNc4g/xl2e/xPaE7ULA9tuB3yLGO4aPvzuchBnfnxS2jfJ3w+rR7VAvxMtu8yW2kXIhD8nn85CVVMizaFXZapiuBeYr4hfmgS5D6/JyB1KZc2eeO6tl15V9YVizI3bCMXr0aMyaNctu8yLE2Sz86Qjq66KEsQ5cTIsAAGRGSURBVHftq4iqN9CucyKEEGK/c01SOQraEkIIuamCX35B1vL3hbHPww8j7M03XCJgezHvIhYfXiyM3+jyBg/Ysk+R5/9yBl/uSxAee6h5GJY83gIKKQXsHLEmbVGuBhlXVUi/UoBze9Kg1xpv+XVuXjKeTdugQyia9YyksgcOLj4+3t5TIMTp7bucA+PBHfDSWZqPFSgT8L8Xnrf3tAghhJAayaGCtosWLeKZDi+++CKWL1/O12k0GkyfPh2bNm2CVqtFnz59eGQ/JCRE+LrExERMmDABf//9Nzw9PXnGxNtvvw2p1KG+fUIIqVYsMJn6yitQ/fKrsM7viScQOm+uSxyJ1KJUPPbrY8L40fqP4sHYB/l+mfL1Mfx2Mk14bEDzMCx+jAK2jqK4QIukc7lIu1SAjPgC5GWUwGSoPItWLBHB3VsOD18F/7DC00+Bxl3CEdXYv1rnTQghjkyjN+LDLzajg64dH5tgRM8B4ZDKqBkjIYQQUhGHiVoeOnQIn376KZo3b261furUqfj999/x3XffwcfHB5MnT8YjjzyCPXv28MeNRiMGDBiA0NBQ7N27l9cmGzVqFGQyGRYuXGin74YQQmq+9NcWWAVsZTHRCJ7xMlzF8qPLYTRbsi3DPcLxUuuX+PJHOy9ZBWwXDGqC0Z1j7TZPcvuB2vgT2bh8NJOXPbgVsVTEa9O26BUJ70A3l8gsd3UsEeDDDz/kH/JnZmbCZDJZPX706FE4ghUrVvAbew9MSE2yYscFNFDpITVbgrTawNPo9MBUe0+LEEIIqbEcImhbVFSEkSNH4vPPP8ebb74prGedfFevXo2NGzeiV69efN3atWvRqFEj7N+/Hx07dsS2bdtw9uxZ/PXXXzz7tmXLlnjjjTfwyiuv4LXXXquwxgbL2GW3UiqVit+zN+83voGvCuw1WCZXdbwWqRp0DB2bMx+/0u+tdLmy77Hgp5+Q/803wjho+nT4jR3Dijw6xH6512P48+Wf8Wf8n8J4eY/lcJd44rVfTmPd3rKSCAsGNcZTHaMdYp+42r/B/IwSJJzJQfplFTKuFKC4QFfptn5h7vANdodPkBv8wz0QXMsbvsFlgVo2l9J/N6R6jt+dsNXrPPPMM/x946OPPor27ds7bKB+0qRJ/Mbev7KEBkJqgkuZRTi/5Qe01FjKIpTI8vDsCyPsPS1CCCGkRnOIoC1748myZXv37m0VtD1y5AgvdMzWl2rYsCGio6Oxb98+HrRl982aNbMql8BKKLByCWfOnEGrVq3KvR4rnbBgwYJy61kXPJaFUR0nHywgzU54WBMM4njoGDo2Zz5+Op0OxcXFfJllklX0wZUpvwAFr84Rxm7PPQfDwIeQlZ0NVziGl1WXMW/fPGE8s9lMuKl9MGzTbhxNLhLWj2obij613fh+JPY7fmaTGYU5OqgytShI1/DlvBQ1Sgr0Ny13ENPSB8G1PRFc2wNyN+s6xHoUISur7FiTqjt+tlBYWGiT5/ntt9/wxx9/oEuXLjZ5PkKIBftd8ObXu9DE0ETYJbEtCuAXHEq7iBBCCHHkoC2rVcsuR2PlEW6Unp7OAw6+vr5W61mAlj1Wus31AdvSx0sfqwirmztt2jRhzDIVoqKieAdhb29vVMfJDsvuYK/nbAEjV0HH0LE58/FjQVsPDw++HBwcXGHQNvWdd4VlrwH9Ef7Si3CVY5hRnIHXd78ujPvF9sOIliPw6ubTVgHb+Q81opIIdjx+JqMJmQmFuHw0i5c7KM6vPIu2NEgbFO2F8Hq+iGrsh9A6PpBInOvftiv/DlUqlTZ5noiICHh5ednkuQghZX49mYaQKyfgbujAx/kecZj07ATaRYQQQogjB22TkpJ407Ht27fb7A357VAoFPx2I3biUV0BHHayU52vR2yPjqFjc9bjx+p5l2aRseUbv7+So0dR+GdZWYCgCRMcdh/c6TE0mAyYvms6UotT+TjCMwLzO8/Hj8dS8fXBJGG7VaPaondj6w8DSfUcP4PeiHN70rD/5yvQqQ03/XoWpI1pGoCYZgEIjPSEVGadTUuc53eorV5jyZIlvHzWJ598gpiYGJs8JyGurlCjxxdf/4LuujZ8bBDpMHiEZZkQQohzkUgk6Ny5s7BMnDxoy8ofsMtOW7duLaxjTRV27dqFjz76CFu3buVZY/n5+VbZthkZGbzxGMPuDx48aPW87PHSxwghxJWwP54PPvhgpZcvJk95QRj7P/00FHXrwlUsOrgIp7JPCY3H1vdbj30XizDzh5PCNu8ObU4B22qmKdIj7kA6ks/nIvViPnSa8s2Vwur6IKK+Hw/Uevop4OGrgIdP+Q9fCbmZtm3b8jJYtWvXhru7O/9g63q5ubm0Awm5Q0u3nEdzjRkSs+W00xxyFk3all3RSAghxDXONYkTBm3vv/9+nDplOYEuNXbsWF63lmVCsJIF7A31jh07MHToUP54XFwcEhMT0alTJz5m92+99RYP/rJLgRmWucvKHDRu3NgO3xUhhNRMmYsXw5iTw5flMTEInu46J1XfnP8G38RZGq9JxVK8ed+buJQmxrNfln3o91ibSDzeLsqOs3Ru7EMDVbaalz1Iu5SPtCt5UBdcqrSBmFQmRuehdVGndTDcvcuX+SDkTo0YMQIpKSlYuHAhL6XlqI3ICKkpzqaqcPWfX9FW01FoPvbclNH2nhYhhBDiMGp00JbVFWvatKnVOlaLMSAgQFjPOv2y+rP+/v48EDtlyhQeqGVNyBgW5WfB2aeeegrvvvsur2M7Z84c3tysohIIhBDi7IEx1iCIYV3FS4MSBb/+htzVa4TtgmfNhMhFLmmJy43DkiNLhPGrHV5FqLwJHvp0t7Auyt8Ns/s3stMMnVd2ciGuHM9GRrwKGfEF0JbcvOSBWCxCbPNANOocxksfiMQUVCO2s3fvXt7AtkWLFrRbCblHJpMZC747iNbGsit2IpvkwCcggPYtIYS42LkmcdKg7e1YtmwZr2XGMm21Wi369OmDlStXWqVns27AEyZM4MFcFvQdPXo0Xn+9rNEMIYS4Cr1ej+XLl/Pl2bNnA6mpyFm3DvmbLFmmjNcDD8CrRw+4gmx1Np7b/hzUBjUfP1T7ITwY/TC6vL0ThVpLAFEpE2PbS93hJneNIHZVY3VpLxzMwJldKTyr9laCY73hH+6Bem2DEdHAjxqIkSrDruRSqy2/Cwgh9+b7o8mIuHIAnnrL1Y/57pfx/LhnaLcSQogLnWtW1PSaOHnQ9p9//rEaswZlK1as4LfKsGYSf/zxRzXMjhBCHEfu+vXIW7qMFQsX1rm1aIGIpWVZp87MaDJizJYxyNVY6lQ2C2yG2e3nYvKGY0LA1kspxc7pPShgew/yM0uQnVSEEpUOZ3enICeluNJtQ2p5I7yuL/zD3aEzl6Bxu1hIZQ73VoU4qEWLFmH69Om8rFazZs3K1bRlV3QRQm4tv0SHdT/+hQf1lr4kJhhxf/9YakpDCCGE3CE6EyKEEBekOXceGV9vgvzaJSsiuRw+Q4Yg5NXZEN0QqHBW3174FgmqBGG8uPtiLPrjMv69kCWse/uRZgjyolI6d3JJVG5aMfIzLIFa1jiM3SrjHahEy97RvJGYf7gnL3/AmEwmXoteLBHf9fEl5E717dtX6Klw4881u7yPNcN1BKXJDI4yX+J83t0ahzYlBZCZYvhY7X8aHXtPtfe0CCGEEIdDQVtCCHFiLNhg1mqhPnkSRTt2oiQzE3lbt8CQmQkEBPJtvB7ojdAFCyD194erYNm1y44sE8ZvdnkTfxzTYMOBRD6WS8VYN7YdOtex7CNSOaPehIwEFS4eysDVk9koytPedHdJpGJE1PdF3bYhaNAhhAKzpMb4+++/4QxY3wZ2U6lUvJ4cIdXpRFI+rvz7B7pq2vGxRlKE0c8PoYNACCGE3AUK2hJCiAMz6XRQHz8OU0kJjDm5MBaqYMjIhDEvD7qEBKhPnQIMZc2ddGYzDDnZwtjn4UEIW7TIpYrEG0wGdP+muzC+P/p+KDQd8ObvR4R1s/s1pIBtBTTFemQnF0GVpYYqW42k83nISS6C0WC6aTYtC9D6BLnBw1eB8Hq+kFF9YFIDde9e9nuBEHLnjCYz5v1wFB1MEcI6n1oJCIkeRLuTEEIIuQsUtCWEkBqWGcvp9VBt3Qpjbq5lnckMs8EAY34+zDodzHo9DFlZUJ84AWNOzl29ls/AgS4XsGXWnVknLEtEEjxV70WMWXVCWNejQRBGd4610+xqppyUIuzbfBkJp279s8bq0rKGYSxIGxTthYBwD8qmJQ5h165dN328W7du1TYXQhzRxgMJiIrfAx9dFz5WKZMxfcpz9p4WIYQQ4rAoaEsIIdWIZcSq/vgDusQkmI0GHozVp6VBd/UqjKoCGLOyeUDWliQ+PnBr2xZuTZtAft998P/iS4jkMoS/9prLBWzTi9Px/tH3hfET9SZgzKoLKNRYspF7NQzGqlFtXW6/VERTpMfVU9lIOJ2DS0cyK99QBNRuGYSYpgGIauQPL39ldU6TEJvp0aNHuXXX/y6gGrGEVC67SItVv+zBw4aWwrq2XT0gV1BdeEIIIeRuUdCWEELuAst61cTFwZCWxoOsZoPREoQ1GPiyMS+XB2JZkJaNWU1ZVlvWVFRk2/0tFsO9fXu4tWoJsbs7ZGHhkHh5Qhoaxu/Fnp6QXNfx3GAwoGNPS2BCLHatJk96ox5jtowRxm2C2+HbHXVQqLHUYI3yd8O7jzYXmmG5GlbiIPFMDjKuqpB5VYXk83koTfy+Xv0OIfAP84C7twKBkZ7wClBC6eEazeuIc8vLy7Ma6/V6HDt2DHPnzsVbb71lt3kRUpMDtYev5uHw1Vz8HZeJzuoUKIxh/DGV92k8MPQFe0+REEJINWLnl+3atXPJc82qQkFbQohLM5tMKDl0GNrz52BSa2A2GXmQtSQ1Bak6HW/YpU9MgkmrZWlWrK29pbmXWm3zuUj8/CBSKiENCIBIKgWkEsijY+DZ9T5AJGYpXxB7eEDi7QWRTAaRQgF5dLRl29sklUoxYMAAuKIvz36JlKIUviwTy6DMH4nsIg0f+7rL8MOEzgj0dJ2MIFablmXSZicVITu5EKps9vNfQZSWZRuKRQiI8EC/8c3gHehW7XMlpDpU1LTrgQcegFwux7Rp03DkSFnda0JcDXvvcyW7mAdoeaA2IQ/x2cWQGXXwMReinj4N4do2fFu9WIPHR/W095QJIYRUM1c+16wqFLQlhDg1lgXLsl11SckwZGfx7FdDZhZLK+RNvPK/3lTp11ryL21LUa8eFPXrQxocDM/u3XjwlWXLKuvX5wFZUjUKtAVYfnS5MPZTD8eWK5aALUus/XZ8JwR7Of9l/SUqHS4dycCFgxnIiFfddFupTIzarYNQp2Uwbx6m9KRsWuKaQkJCEBcXZ+9pEGIXK5Z/jMxENWCSQQw5JGYFAswKDDApITO5QWpiH3awW6TwNdKIONRp2p+OGCGEEHKPKGhLCHFIRpXKcsvLhzbuPA/Kai9ehD4piWfPsuu6jYUqGLNz+PI9EYkgDQuFWOkGkYRlvIp5oFUaGMiDsCwAK5JIIJJJAXYvlUEkl0MWGgJZRIRlnUgEib8/387emTIlJSV82d3d3WVqt751oOzSZoXYHZevNBbGcx9qjPohXnAm6iIdclKKoSsxID+rBIXZGp5Nm36l4kCtRCqGd6AS/uGeqNM6CCGx3rw2LcuwJcRVnDx5stzvy7S0NCxatAgtW5bV6STEFajVGrwz70MEFbZB8B18XZE8E5NfeKYKZ0YIIaSmctVzzapEQVtCSI2oD6u9fAWGzAxe95WN2b2xoIAHYE3FxTDmF1iyY7VaFO3YCWN+vk3nIA0PQ+icOTwYy4K0+SUlCGrQAFIvL97IiwVhnQGr0fjee+/x5dmzZ/PLfp39jcOGcxvwZ/yfwjpNwiRL9ywA/ZqGYmyXWnB0rKxBdkoRshIKcea/FGQmFN7ya1gd2gadQhHZwA8RDfwgk9v3AwVC7I0FZtnJBfu9cb2OHTtizZo1dpsXIdUt/ko81n+wBUEaS7mD65lhgl6shUGigVGsgUmsBcQ6QGyAVK5H3wHt4OFVVkufEEKI63C1c83qQEFbQohdShaUHD4MQ1YW9KmpyFr+vk2fX+TmZqnzyrJbfX0hj4qCxMcb0jDWnMsb8ugoyMLDWXohxG5KSENCIPEqy7Q0mUwozsyEPDiYCqg7ML1Jj3n75uGXy78I60K1z+BiUYBl2VuJlSNbw1GZTGZcPpqJi4cyePOwkgLdLb/G3VuO6KYBqNc2mAdrxSxznBDCxcfHW+0J1kAjKCgISqXzl04hpNRf23bi8G8ZCNA14GOTyIii4OMY8nAv+IeGwD8oBFJW2okQQgghVY6CtoSQKmPIy0Pxnr3QXr4EQ2oaDPl5MObkQnP6tE2eXxIQAM+uXXkDL2XjxpBHRfJ6sWJ3d5s8P3FcLFPuzRNvYm/mXmFd58BHsfW/unyZXanz+ai2DnPJjlFvQvqVAmQlFfLGYapsNc+s1WuMlZY7YFm0Xn5KKD2k8Ap0Q2CEJzx8XafRGiF3KiYmhnYacWmrVq6G6kwwvIwhfKyVFMO3eQ6mjJ9h76kRQgghLomCtoSQu8ZKFZQcPIiSQ4dhyM6GMTcXmrNnYSwshFmnA4wVB5Qq4/Pww5DXqcOzZEVSCcSeXjwTlpH4B0Ds6cEfYw27WB1ZMV1uQSrxxv43rAK28zq8gVfXs4CliY8ndK+DZpHlO8XXtJIHmhI9jm1NxLHtiTfdVq6UILSOLyIa+MI/zANRDf0hkVEWLSG3Y+fOnZg8eTL2798Pb2/ry7oLCgrQuXNnfPLJJ+jatSvtUOKUDAYD3n19GbwyW0IBS6kclTwdnQaFo0fvgfaeHiGEEOKyKGhLCLmjzFnNqVMoOXwERbv/g/bsubvaeyKFgmfDenbrBmXzZrxxl2e3rpCFhdHRIPecYbvs6DL8cOkHYd0bXd7AXwejoTOmCeteuL9ejdvTmiI9Lh3NxNWT2Ui9mA+99uYfenj6KRAU7YUm3SIQ3cifmoYRcpeWL1+OcePGlQvYMj4+Phg/fjyWLl3qMEHbFStW8JvxDj84Ja6pIL8Ay95YhaDisvq12W7nMebFAYiJpexzQgghxJ4oaEsIKRf00l64yOvN8pqzSUnQnD8PbVwc9Ckpt95bMhlkoaG8RqzI3Q2KWrV46QJl06aQRURAzAK2Hh6010mV2Jm4E2tPrxXGTzQYjX3Ha+O3k0nCut9fuA9Kmf2bbhkNJmQlFuLCwQwUZJUg8UzuLb+mTb8YxDYLhH+4B+RK+hNOiC2cOHEC77zzTqWPP/jgg1i8eLHD7OxJkybxm0ql4kFnQipz7vQ5fL9qN4I0rYR12d5H8L8FU+B27UonQgghhNgPnfER4oIMOTko3LYNZr0BusRE6NPTYC5Rw6TRQH3yJLtO7raehzf5io2FR5cu8Ox6HySBgZAGBkJMTVuIHSSoEjB371xhPDD8Cfy4owWyCssCtksea4Em4fYLYlw9lY2E0znISS7izcNMRusu9aVkSgkvcyCVi+ET6IbQOj5o0CGUGocRUgUyMjIgu0ljJalUiqysLNr3xKn8+tNvOLuzBP76OnxsEOlhiDqL+bOpfi0hhBBSU1DQlhAXoU9LgyEjA4X//IOcTz69q+dgDb8k/v7wGzYMHl3v41m0xLGwbugtW7YUlp3FmZwzGLtlLNQGNR839GmF7/5pDr1Rx8cecgnmD2yCoW0iq31umQkqXDiQgSvHs1CYq6l0O6WnDDFNAlC/QwgiGvhBInGe40NITRYREYHTp0+jbl1Lo8IbnTx5EmFUvoc4kRVLP4b+Ugw8TYF8rJYWIKK9FiNGTbX31AghhDgwZz3XtCcK2hLi5FhJg8Snn4EuIeG2thd7ekJeuzbcW7WCLCoK0sAAyCIioWzUkDcBI46NZYwNHjwYzqREX4Lhvw0Xxl7SQBw+3Afma+Ucw3yU+Oa5TogOcK+2BmL5mSU4uycN5/el8Vq1FZFIxQit443AKC807hwOvzB3iESiapkjIaRM//79MXfuXPTt2xfKG64UUavVmD9/Ph566CHaZcQpvDnnPfhmt4IMlpPpfEUyeg9viA6d2tt7aoQQQhycM55r2htFYAhxEmajEbqrV6G9eAmaM2dgyMuFWa1B4fbtMOss2YY3Cpw8GfKYGCga1IcsPBxiuRwiubza507I3dIYNBi/fbzVurRzz8Js8OXL9UM88fW4jgjwVFTZTtbrjDizKwVplwqQl16Mgix1pWUPAqM8Ed3YH7HNgxAS60XlDgipAebMmYMff/wR9evXx+TJk9GgQQO+/vz580JDr1dffdXe0yTknm366hv4ZV/XcMz9NJ6fMRwhYcG0dwkhhJAaiIK2hNTghmBmtRqG3Dw2AkwmGAtUMObmwGwwQBcfD11yMg/MltaiNaSl3fQ5fR9/HNKQYMjCI+Dd50GI3asn85DUrJ8rvd6S+clqODpyZicL2E79ZyqOZx3nYxHEMKeNEwK2HaK9sXRE6yoJ2LL9qMpW49zeNJzYkQSDzlTptixQGxDuiaY9IhBai5oCEVLThISEYO/evZgwYQJmzZrF/30z7Pdjnz59eOCWbUOIo7twOA8BCOLL2d6H8erCaTwrihBCCLEFZzrXrCnorzQhdmLmQdgCXmfWVFgIY2EhbxBmKiiALjEJ+d99x37r2eS15HXqIHL5Mijq1bPJ8xHHxf6ILly4kC/Pnj0bcgfNrE4vTseMf2cIAVuZSI6CxOEwFlnqLHeqHYClA2MQ4uNm09fNSSnC+f3puHwks9L6tN5BbvANckNYPV806hQGD9+qy/IlhNhGTEwM/vjjD+Tl5eHSpUv8pKNevXrw8/OjXUycwpdr1iNAU58vF8ty8NLs8RSwJYQQYlPOcq5Zk1DQlpBqZNbrob14EZlLl6F4926bP797x45wa9ECssgIKBs2hMTHB2IPD948jD7lIs5Uw3bQ5kFC0zE3iTtyLj8Fo9oSsK0X7Inlw1rArC6459cyGU24fCwLiWdykHAmF2pVxaVGwuv5ok2/GETU84NERkX3CXFULEjbrl07e0+DEJtLOKmB/7VlUXASfHzpyg9CCCGkpqOgLSHVwKTVIufTz5C7bh1MJSV39LWeve+HWK5g7RcBsQjSoCBIfHx5UzB5TDTktWpBrFBA7OMDiadnlX0PhNQUM/+bKQRslRIlpFnjYVRbLl32cZNh86QucJOJkWnZ5I6w7LrEs7k4+18qbyZWmKOBXnuto9kNgqK9ENM0gN/Xah4IkZgu/yGEEFLzrP5kDfw1dfhyoTwTU6Za14InhBBCSM1EQVtCqhALABXu2IHkSZMrfJzVlPXs2RPSkBBIvDwh8fODxM8fYjclD8bKIiMpQ5aQ6/49LT68GH8n/S3sE0Pqs8jKtQRsvRRSbJvaDR4KKUymymvM3khdqMPlo5m4dCQT6fEqGPWVf62nnwL124eiWY9IvkwIIYTUZAaDAWnngNJCH7KwdHh4eth5VoQQQgi5HRS0JcSGDLm50CcnQ5+VhcJ1XyD/xAmYddaXU4vkcvgOHwbfoY9C2cBSW4wQcuuA7TuH3sGGcxuEdbrczijMjeTLkX5u+P75zgjxVt72rsxKKsR/31xA2qXKyyh4BSgREOGJRp3DEBjpCe9A29bIJYQQQqrS5ytXw09r6WmgkqfjpZeepx1OCCGEOAgK2hJyj/QZGdBeuICsDz6E5tSpm24bMH48gl56kbJnCblDHx770CpgayiqB23GQ3w5yEuB9c90QKhP5QFblj2bdD4XqRfyoS7SIf2KCvkZ5UuVuPvI4R/mgahG/mjeKxJSmYSOFSGEEIfNss29rITvtbF7VC7c3G7/w01CCCGE2BcFbQm5Q2ajEaotW1C8axeK/tsNY25upduKPT2hbNoUirp1ETRlMm8MRgi5M79c/gWfn/pcGGsyBkCf25UvD2gWhiWPt4DyhuAqayCmKTTg1NlkXDqchYyrKphN5kpfg2XSNusZybNpqWkfIYQQZ/Dx+5/BV9uQLxcoUvDyC8/Ze0qEEEIIuQMUtCXkDuWuX4/MRe9U+rhH505QNGkKjacnwh97FDL/0l69hNifWCxG48aNheWaLkGVgFd3vyqMdTldhIBtmxg/fDiiFQ/GXjiUjqyEQmQlFaEwVwNV1q27kPmGuKN5z0g06BgKuZL+HBJCCHEeOq0ORQk+8L429q1VDLlCbudZEUIIcWaOdq7pCOgslZCbMBYWwlRUBGN+Pgp++w2m4mLkb/rGeiORCO4dO8Cjc2d43ncflI0a8SZImZmZkPiWXpBGSM0glUrx+OOPwxGU6Evw9NanhbGhqC4MGQMQYBKhjlKBabXCsfWz07h6Mhumm2TRMhKZGJEN/RBR3w+RDfzg4auAuzedvBJCCHFOK5Z9Am9dU76cp0jAzMnP2ntKhBBCnJwjnWs6CgraErh6cyNTQQGK9++H5swZGAtUMKnVMKSnQ3vpEox5eTf9+vqHD0Hs7g4RfYpEiE0VaAswccdEZJZk8nGtrHaonfA4autlkEAEFAJHNl+p8GulMjF8Q90hkZsRHOXL69PGNg+ksgeEEEJcQnFRMbSpQSj9aDK4noGfSBNCCCHEsdBfb+L0zAYD1CdPwZiXC2N+AYwqFfSpqdAnJaHk8GGeSXvHZDJELF0CiadnVUyZELj6hylz9szB2fTzqJPXCq1SHkBgSUSl2ys8pDyDNqZpAGKbBfIM2tJs9+DgYLo0hxBCiMvVsvXSteDLucormPX8WHtPiRBCCCF3gYK2xGkZi4qgT0xEysszoLtScUbeTUkkELu5QV6nNuRR0RApFXBv2RJurdtAHhtD2bXEIel0OixcuJAvz549G3J5zSsRsGbvV1Dv9sLTqYsghnWDMZlCgoj6vvAL84BPkBuCY7wtzcPEIrvNlxBCCKkpCgsLYUwv+6AzqomMsmwJIYRUC0c413Q0FLQlTseQl4f4IY/wEge3IpLLoahXD/LatSGPjoZ7u3aQBgZA4ucHaUBAtcyXEALeTOzMnhRs/f0glPkRaAXrzFqdGOg9vD6a3BcBMQVoCSGkUitWrOA3o9FIe8kFrVj6Gbz0rfhyjvIiJo0bb+8pEUIIIeQuUdCWOLzCv/9G0T//onDrVl6P1qzVVrid90MPwa11K0j9/CD29IQ8NhaysDCIqMYXIXZjNJhw4WA6Dv0Rj8JsLZSwbt6XKjEhw0+CWeNbo06Uj93mSQghjmLSpEn8plKp4ONDvzddSV5eHsTZscK4bms6/oQQQogjo6AtcUhmnQ7pb7+N4n938fq0lXFv3x5uLVvCd+gjkMfEVOscCSE3z6xNOJ2DfzfFoSjX+oMWtbQQZ93zccwchMAwD2wc1xFBXgranYQQQshNfLxsFXz0bfhyttt5TBo1kfYXIYQQ4sAoaEscii4hAWmvvYaSffsr3catRQtIQ0Ph/+RIXu6AEFKznN6VgmPbEqDK1litL5YV4HDUFhwzhkJf0B5eSglWjW5LAVtCCCHkFjLSMiHPrSuMm3YMoX1GCCGEODgK2pIa2z2+5OAhaM+fg9lghCEzA6otW2HIyCi3rcTfH8pmTRH80ktQ1K0LkUxmlzkTQm6uuECLg7/F4+x/1tnxuW5p2Bu7GUneF6FOHAejOhah3kqsGdMOMQEetFsJIYSQW1j10Tr4Gtry5Wy3s5g0bDLtM0IIIcTBUdCW1JhyBwW//Y7cdetgLCyEIS3tll8ji4xE0NSX4DNgQLXMkRByd9RFOpz+NwVHtyTAoDeVrQ/MxS7/n3DV/xRMZgnUSWN4wLZHgyAsH9YSvu7UbZQQQgi5lZTkVCjzG/JlM0xo1Y1KghFCCCHOgIK2xG6MRUXQXU1A0d9/I3vFitv/QpkM0Z99Co9OnapyeoQ4JbFYjHr16gnLVSk5Lg+n/03GlePZvIZt2RxEULdJwBfSJcI6dfJIGEvqoV/TUHw4ohWkkqqdGyGEEOIs1qz4Cv7Ga1m27mcweciL9p4SIYQQF1Sd55qugoK2pNpp4uKQMn06dJcuV7qNJDAQYoUCikYN4fPQQEAkgiwslJc/ELu7V+t8CXEmUqkUI0eOrNLXYAHaY9sTse+nG/6Ni4Cohn7IaH0Sa69YArZmswialJEwFjXCK30b4vnutSESiap0foQQQoiziL8SDw9VI75sggmdH7AsE0IIIc54rulqKGhLqpXqzz+RMnVahY+Jvb3h3a8fgiZPgjQoiI4MIQ7IaDBh6+enEX8i2yqztlnPSDTrEYEStwIM2vyc8Jg2YwAMhU0xo08DTOhRx06zJoQQQhzT+s++Q8C1LNsc91N4sN9Ue0+JEEIIITZCQVtSLU3FCn78CZmLF8OYl2f1mFvr1vDu2xceXTpDUYcCNoQ4Mr3WiJ+XH0NGvEpYV69dCHqMbAC5Usp/F0zf/gI0Ro1le1VT6PO6YMGgJhjdOdaOMyeEEEIcz/mz5+GlasqXTSIjeg5obe8pEUIIIcSGanSRibfffhvt2rWDl5cXgoODMXjwYMTFxVlto9FoMGnSJAQEBMDT0xNDhw5FRkaG1TaJiYkYMGAA3N3d+fPMmDEDBoOhmr8b18MCNHmbNuFK335Ie/XVcgHbqM8/R+zGDfAf9RQFbAmpJjqdDm+99Ra/sWVbZthuevOgELAVS0Xo93wzPPhMEx6wZX6+/DMOpB3gyya9NzRpQzG4ZQQFbAkhhJC78M26XyA3KflyjvtJdL+/O+1HQgghTneu6cpqdND233//5QHZ/fv3Y/v27dDr9XjwwQdRXFwsbDN16lT8+uuv+O677/j2qampeOSRR4THjUYjD9iyH5i9e/fiiy++wLp16zBv3jw7fVeuwWwyIeWFF5H+2gLoEhKsHpNFRaH+gf3w7Hqf3eZHiCtjv0vZzZY1bHd/exGqLDUfi8Qi9J/QHLVbBgkf4Px6+VfM3ztf+Bpt5gB0qxuNZcNa2mwehBBCiKs4cewkfIqa8WWjyIC+j3Sx95QIIYQQm59ruroaXR5hy5YtVmMWbGWZskeOHEG3bt1QUFCA1atXY+PGjejVqxffZu3atWjUqBEP9Hbs2BHbtm3D2bNn8ddffyEkJAQtW7bEG2+8gVdeeQWvvfYa5HJ5udfVarX8VkqlsmSOmUwmfqtq7DVYkKM6Xqsq6BITkTJpMnSXy5oQSQIC4Dd6FHwfewwSHx++zlG/P1c4hq7OmY9f6fdWunwv36PJaMLJv1Nwfl8a8tJKhPW9nmqAqEZ+/LnzNHkYs3UMrqquCo8biurD29gayx5vzudSOh9bcuZj6Aro+Dm26j5+9O+cuKKfNm5FkKkNX871OImOXV6295QIIYQQ4kpB2xuxIC3j7+/P71nwlkXwe/fuLWzTsGFDREdHY9++fTxoy+6bNWvGA7al+vTpgwkTJuDMmTNo1apVhWUZFixYUG59VlYWL8dQHScf7HtlJzxicY1OhhaY2L7ZsAH6I0dhSkqyekzWvTs8XpsPo0iEHBYMz8yEs3PEY0hc4/ixqw5Kr1bIzMys8IOr27X/2ySknC20Wle3oz98a4n5c18ouIB5x+YhR5sjPM7q2GpSnsCr/aOhL8pHZhGqhDMfQ1dAx8+xVffxKyy0/j1EiLM7tP8w/Iqa82WDSIdBI+6395QIIYQQ4spBW3YC8NJLL6FLly5o2tRScD89PZ0HHHx9fa22ZQFa9ljpNtcHbEsfL32sIrNmzcK0adOsMm2joqIQFBQEb29vVMf3KhKJ+OvV9GBD/jffInftWugTEyt8PPDFFxEwvqxTvKtwpGNIXOv4saCth4cHX2ZXLtxN0LYwV4M9312yCtgGx3qh/UO1ENXY8qHaT5d+wmv7X7P6OnXqYzAUtMajbSIxvHMDiMUiVBVnPoaugI6fY6vu46dUWmp6ElKTLZz3Htyzm0JmUtjk+aSQ8fs8z1No3aavTZ6TEEIIITWLwwRtWW3b06dPY/fu3VX+WgqFgt9uxE48quvkn53sVOfr3SlTSQlSpr+Mor//rvBxVgbBf+wYKGrXhquq6ceQuObxY98P+95Kl+/0+yvK0+LnZcdRmFN21UGHQbXRtn+sMN5ydQsW7Cu7WoE1HVMnPwWTJgpPd6mFeQMbozo46zF0FXT8HFt1Hj/6N05qur2798EjqzmkZkug1Vb0Yi0eG9XPps9JCCGEkJrDIYK2kydPxm+//YZdu3YhMjJSWB8aGsqzxvLz862ybTMyMvhjpdscPHjQ6vnY46WPkdtjLCpG4ZY/UfDb79CcOQNTBZci+j7+ONxaNIfPwIEQ3cMl14SQmun0rhTs++kydGoDH4slIrQfWAut+8QI2+xN3YsZ/84QxobiulAnPwmxWYmNz3ZA57qBdpk7IYQQYi/bftqHILOl8WaRLAt6cVkd+LtlFhngG6VDk2YUtCWEEEKcVY0O2rJaaFOmTMFPP/2Ef/75B7Vq1bJ6vE2bNpDJZNixYweGDh3K18XFxSExMRGdOnXiY3b/1ltv8fqK7FJgZvv27bzMQePG1ZPt5cj0GRnI+2oDcj7/vNJtFPXqodaPP0Aks232ACGkarLfYmMtWbGlGbe387t4++ozuHi4rB61p58CQ//Xlt+XYk3HXtn1ijDW5beFNm0oWkT54YPhLRETYCnLQAghhLiKLb9vRUCxpf6sTlKCIZNboX6D+vaeFiGEEFIjzjWJAwdtWUmEjRs34ueff4aXl5dQg9bHxwdubm78/plnnuH1Z1lzMhaIZUFeFqhlTciYBx98kAdnn3rqKbz77rv8OebMmcOfu6ISCKSMsagIl7r3qHCXKJs1g7xWLHyHPgqPDu1ptxHiINgHXWPGjLnt7Yvztfjj45PITCjLro9pFoBuw+tbBWyNJiOe3vo08rX5lrEmBNq0wXiuWx3M6teQ/mgTQghxSQf+uoBANOHLRd5nUb/BQ/aeEiGEEFIjzjWJgwdtP/74Y37fo4d14HDt2rXCD8KyZct4LTOWaavVatGnTx+sXLlS2FYikfDSChMmTODBXNaAZ/To0Xj99der+btxHJq4OOT/8APyvlxvtV4SEICAsWPgN2IExNcaGRFCnFd+Zgl+XnaM17EtxcohtBtgfdUD8/Pln3Ep/xJfNpvF0KQ8gfHd6mMmBWwJIYS4qG82fItAtSVgq5aq8OyUp+w9JUIIIYQ4kBodtGWX5N5Ox+AVK1bwW2ViYmLwxx9/2Hh2zilt7lzkf/d9ufUhs2fB76mnKFuOEBfBMmx/WnIUJQU6PpYpJWjQIRRt+5U1HCu1+dJmzN87XxhrM/uhU3RjvNKXMmwJIYS4rrhDeQiApZa7PuASwsIH23tKhBBCCHEgNTpoS6qPSadD5qJ3ygVspUFBCJ0/D169e9PhIMQJsOaNy5cv58svvfQS5BU0DTTqTfjz01NCwNYrQIlHXm5jVQ6h1I7EHZi7Z27Z16ojoM/tghWTW0MspjpGhBBCXNPqT9YgQFOPLxfJsjFx6jh7T4kQQgix+7kmuTMUtCXQXonHlf79rfaEPDYW4e+9B7dmTWkPEeJkSkpKbnqFw79fxyEjXsXHnv4KPDazLdy8yv/BTVQl4rW9rwljQ2FDqFMfx8ZnO8HXnf5AE0IIcU0GgwFp58TwuzaWhqXAx9fHzrMihBBC7HuuSe4cBW1dmD4jAxlvvonC7X9ZrRd7eaH2r79AJJPZbW6EEPuI25+Oc3vT+LJUJkb/55tXGLAt0BZg9JbRQuMxs1EBdcoTGNqqNjrXtVwKSgghhLiilcs/hZ+2EV9WydPw0tQJ9p4SIYQQQhwQBW1dkNlgwNURT0Bz6lS5xwLGjUPglMkUsCXEBRXlabDji3PCuOeohgiK9iq33YW8C3hh5wvIVmfzsUnnh5KrkzDjwWaY2KNOtc6ZEEIIqUnUag2KE/3gfW3sXUsFNzelnWdFCCGEEEdEQVsXU1EpBEZRrx7C330HykaWrABCiGvRqg3Y9MZBYRzdxB/124VWGLAd+cdIaAwaPjYb3KFOGYmR7RpjUs+61TpnQgghpKb5aMlKeOta8uU8xVXMnEK1bAkhhBByd8R3+XXEARny8soFbKXhYYhcuQK1fvmZAraEuChVthqbXj8AbYmBj2UKCboOq19uu/TidDy79VkhYMsybIsTJmBU6/vw5uBm1T5vQghxdr/99hsaNGiAevXqYdWqVfaeDrmFvLw8mDOihXFkUwmkUsqRIYQQQsjdoXcRLiTnc+s3+z6DByN80dt2mw8hxL40xXoc3JyAM3tSYdAahfWDp7WCb7C71bZJhUno/6P1hz4lCRPwaIvGmDOAMvQJIaQqmllNmzYNf//9N3x8fNCmTRsMGTIEAQEBtLNrqJVLV8FX34Yv5yjjMGkc1bIlhBBCyN2joK2LUG3bhtw1a4RxwPPjEfzSS3adEyGk+olEIoSHh0OnNuDH946iOFcvPCaRijFkemsEx5RW4rNQG9QY8vMQq3XF8RPx+oBOeLJjDH9OQgghtnXw4EE0adIEERERfNyvXz9s27YNI0aMoF1dAyUmJUOZV3aVSpPOIXadDyGEEGKvc83SZXLvqDyCC8hasQIpL7wojKVhYQicQJ/8E+LsTCYzjHoTdBoDtCV6ZFxVIeV8Abo1HwhlQlOrgG2z7hEYvagzQmpZB2wTVAkY9ecoaI1aYV3J1ecxqnU3PNUplv4YE0JIJXbt2oWBAwfykxd24rJ58+Zy26xYsQKxsbFQKpXo0KEDD9SWSk1NFQK2DFtOSUmh/V1DrVu5AUqDpXlnlvtpDH38EXtPiRBCCKlWMpkMzz33HL+xZXLvKNPWyWUuX46cTz61Whe9ehXECoXd5kQIsY2iPC2unspG0rlcXt7AaDTzDFpWo5Y1FoP51s8hU0owZFprBEVbTjSvF5cbh0d/fdRqnTrpKYzvcD9m9GlAh5EQQm6iuLgYLVq0wNNPP41HHikfwPvmm294+YNPPvmEB2yXL1+OPn36IC4uDsHBwXe8b7VaLb+VUqlU/N5kMvFbVWOvYTabq+W1appzZ87Bq8BS290EI7o82Ngl90NVcuWfL1L16OeL0M8XqW63+/eMgrZOTJ+ZWS5gW2/vHkj9/e02J0LIzem1Rh50ZbeLhzNhNpn5jWXNms3gJyxsrCnSIzOh8J52p5uXjJdD8Av1KPdYka4IL/09VRib9F7QZg7AoPoP4n99G9JhJISQW2DlDNitMkuXLsW4ceMwduxYPmbB299//x1r1qzBzJkzeYbu9Zm1bLl9+/aVPt/bb7+NBQsWlFuflZUFjcbSQLKqTz4KCgr43ymx2LUu5vv2i98RbGrLl7M9TqJZy5HIzMy097Sciiv/fJGqRz9fhH6+SHUrLLy9c3kK2jqxjDffsho3OHIYYo/ywRlCiO2wUgSFORrodUYYdJaskPjj2SjM1cCoN8KgN8FouBaINbKbid8bDSaeHavXlDUEu5dgrIevAlKZBGKJyHITiyBVSKD0FOPXfzfxS3Vnvji9woAts/DAQiQXJfFls1GBkvgXMbpDU8x7qPE9z48QQlydTqfDkSNHMGvWLGEdC0T17t0b+/bt42MWoD19+jQP1rJGZH/++Sfmzp1b6XOy52KZu9dn2kZFRSEoKAje3talb6oq6MH+trDXc6Wg2v69++Ff3IIvG0Q69H+0211lSpObc9WfL1I96OeL0M+Xbej1eqxcuZIvT5w4kUok3AQrjXU7KGjrpEw6HQq3bRPGtf/8gwK2hFQBdZEOR7cmIie5EIW5WuRnlFTbfma13f3DPdHl0boIjvGCWCrmAVqJRHzTQMG/FyV82SfYrcJt9qbsxa9XfhUCtpqEF7FmVE/0bEAnoYQQYgvZ2dkwGo0ICbFuVsXG58+f58tSqRRLlixBz549eUDhf//7HwICAip9ToVCwW83YgGu6gpysaBadb5eTbDtp/0IMrfky/leJ9Gh0//sPSWn5Yo/X6T60M8XoZ8v2/w7YldFXP9vilTsdvcNBW2dVMmBA8KyJCAAilq17DofQhwVa+B15Xg2L1fAMmJZ5izLoC3K0/Cs2pS4/Lt6Xpb5KrouC5bdy5VSXmPW008JTz8FZHIJIhv6wT/cAyK2vUgEkbj0D6AIYqmIZ9PaksFkwML97wljXXYvfDbiQQrYEkKIHQwaNIjfSM207c9tCChuzpd1YjWGjaFjRQghhBDboaCtk0qdPVtY9hs2zK5zIcQRsHIFqhwNb+RVotIiK0OFC/8U4NQ/KXeVAdu4awQUblIeZGU8fBSo3yEUUpklG5YFXmuib85/i4SiS8K4b8wQ9GponQlGCCHk3gQGBkIikSAjI8NqPRuHhobS7nUQe7fHIQhN+HKh9xk0bDzA3lMihBBCiBOhoK0TSn/zLRizsoWx79DyHYsJcVUsOzb9SgEKMtW8vmxxgRaqLDUy4lUoUenu6jlZILbtgFi06BUFiGDz7NfqklqUikWH3hbGxVcnYMHLbew6J0IIcUZyuRxt2rTBjh07MHjwYL6OlUBg48mTJ9/Tc69YsYLfWPkFUnW+2/QDgkosAVu1tBDPTH6SdjchhBBCbIqCtk5Gn5GJvI0bhbFHt66QRUTYdU6E3AsWVGUBVpPJ0ryLN/Bi92ZAXaiz3FR6XsaAbavXWhp6sYxZvr352va86ZeZB2rvFCtV0LJ3NAIiPXkWLbvJlFL4hrhDKhXz0gWOjpVFGPn7U8JYr2qCDU8Og5+H3K7zIoQQR1VUVIRLl8quXIiPj8fx48fh7++P6Oho3jRs9OjRaNu2LW86tnz5chQXF2Ps2LH39LqTJk3iN9aIjDUwI1Xj3P4sBMCPL+v8LyAi8mHa1YQQQgixKQraOhEWnEqfN4+lagjrIpcvt+ucCLkZo9GE4jwtDv9xldeILS7Q8cAqC7KaDCaUFOpgMpirbScGx3ojMNITbl4yaPVqhET4I7y+H7wDKm7Y5SxMZhNe2PkisjWZwrp2XmPRqU7lDW8IIYTc3OHDh3kTsVIsSMuwQO26deswbNgwZGVlYd68eUhPT0fLli2xZcuWcs3JSM2z5tN1CNDU58vFshxMmPqsvadECCGEECdEQVsnUrx7D4r+/VcY19r8E8Tu7nadE3F+ep0RmiI9jHoTjAYT9FojNMV65KYWoyhXA6PRzMsOaIv1PDuWfbjA7lkmbH5GSZXMSSITQ+kh4/VkecMv1rTr2j27+Qa5ITDaC17+SsgUEn7Pmn1JpGLhEtXMzEwEBwc7XcdLVks3KChIWGbm73kN/6XsErYpSRyLhZO62m2OhBDiDHr06MH/5t0MK4Vwr+UQSPUyGAxIPWu+lmMLiEOT4edXOiKEEEJcV0XnmuTeUNDWiWR//LGwHDDheSgbNrTrfIjjMRlNPMhaGoDVFhugKdYhP0ONgqwS6DVGmMyAXsvKDxh5dmx2chFgw2RYhbuU/4JnAVdWgoDVi2WB1fB6vpYGXqUBWJEIYqkIbp5yuHvL4e4j582+lB5SiCXOFWi1JZlMxi+bZdKL0/G/bXNxLHu/8LgmbTA+fuQJRPnTBz6EEELIjT5+/1P4aRvxZZU8DS9Nm0A7iRBCCLnhXJPYBgVtnYQ+IwPqo0eFsd+wYXadD6m5WED22PYEFGRreJ1X1pgrP1PNA7AsKFstWF1YdicRwc1DxrNca7UIQuP7woVsV1K18jX5GLr5SagMZZ3LdXntsaTfBDzYhDqXE0KIo6JGZFVHrdagKMEP3tfGnjH5cHNTVuErEkIIIcSVUdDWSSSMeEJYloaGQhZKQRdXw+rDppzPQ0GWmjfi4tmyehMvX6BW6Xh92JICHX/clliQNbSONzx8Fbwpl1QugdxNyjNmfYLd4eWv4NmxPoFuloAsb+RFl0rcC95czWyC+dp/JfoSGM1GS+mJa2nPpctsu+vH7D+tQYun/5woBGzNRiVkhQ9g6YNT8EDjMBv8VBBCCLEXakRWdVYs/RjeuhZ8OU+RgJkvjq/CVyOEEEKIq6OgrRPQJadAn5oqjKNWrrDrfEjVKc7X4r9vLkCVo+HlC0pvrIwBqyV7Lzz9FLy2q1gqhkQq4gFWFnyVKyRQeMjgF+oO7yA3SCRiXqaABWVZGQKFh5Svs3cAk/9nNsFoMvLAZGkQkz9WGuA0m5GlzsLVgquWx9l2JlbywSRsz+7ZOlWhCqI0EQxmg/AcwnbXgqHX3/QmvVVgVAig8v+VrTOYDCjQFVheFyb+eI4mB/nafKsAq+W1Sp/n2vpr9/e0vwxm5P6Ty5f9e/gDIgXaSt/Ah+Pvh5dSds/PTwghhDijvLw8mNIjhXFEExGkUjqVIoQQQkrp9Xp89tlnfPm5557j5RLIvaF3Gg7ObDTi8gMPCGO3Nm2gbNzYrnMit2YymZFxpaAs+Ko38TIF6kI91IU6FOZqeBDWZDTDbDLz7dm9KltzT7uXBWGVnjK4ecoQ3dgf9duH8ixYFoBljbsqwwKMRfoiHnBkQUudyYgSk4EHNHUFOhTqCpFZksm3yVZn8+3YYywDVGfUIVeTKwQlrw+OVjSubF25wOm1x0szS8ntYfvLUGi49nMoQQfPGVgzvA9lPxNCCCE38fGyVfDRt+HLOcoLmPTc87S/CCGEkOvPNVmSVFaWsEzuHQVtHVzehg3sX4NlIBYj7M037D0ll8V+KWmK9dCWGGA2mZCTVILitGz+WE5KMfIzSpCbWoSifC3PjL3n15MaYRabAYkJJpkBJqkBBqUG6sAcaD0KYZDqYRDroEEJCsS50MlLYJRYskb51xvNMO0zCb9MS7NRr8/2ZMFYtVHNg7Ck6pjNIpgNXjAb3a6t4RV/rz0ormCdZdly5ESASQaYJdbrrttOGLPH9UaYtGf5Y22U07Fm+DAK2BJCCCE3kZKcCnlufWHcqKOlMzYhhBBCSFWioK0DM2RnI2Ph28I4cMIEKGrVsuucaqo8TR6K9cWWDNBrWaDsXmPQQG1Q80vb2Y2tY9tpjVqeXVq6HcvoLN3eZDIB2UqIC5SQZHhDlukFcYkCIr0EIiHAdu/YpfAGsR4mkQlmkclyDxOK5fnYVftbZHsm3/wJboyz3luSrs2Y2T4Sgolly3w9G7P15usfs9ybS4OPfLvr7q8t88dLn/eGxyzBzLLnNRu8YdIFWG1nef3rn5+tk1sCouXmeV0wtPQ5edC0NFDK3HAvfM+A2eR27Xkt61iJ32AvBfw9FJa9wkv/iiARixDt746eDYMgFrHtRBCLwJfZPRuz7QM85fBWyvj27CaViOEpl5ZN4To6nQ5Lchbx53ht2HAK2BJCCCG3sPqjrxBgaMuXs9zPYNLwKbTPCCGEEFLlKGjrwFKmTReWWUmEwMmTUFOxgCe7PJ4FQktrhurNeqj1loApe5wFSjNKMvi9dV1PS8CU3bMAa2kN0uuzQq+vWVp6z+qFFuuK/9/enUA5VV8PHL/ZZ2MywwwzA8halqogyCqK4kIFtO57+QuixT92tPLHqqCt2NPTA65VERW1QBcVaut2VKgUARUXFgVkKaLsO8MyG7Mm73/ubyYxw76ESd7k+zkn5L3JS/Lyfi/h5ea+e2Xe5nmyv3r/yb8IS8QbSJbz1l0rHQtqmlCciCpnpRT5CqTMUyLlnhLZ4v/OBGdNVqynVMrdJVLu3i9l3uKTX+c6gdJQsFEigo2106EgZeS8CvrECtYELq1gksnmrAlaumoeT4Olln6MaKZomgSrMsQKpNVmfuptHhG9vzgkWNn4mD5yvFpTNzI46ayZ1mCkBim9LqckeZzmNv1bk0Y+Sfa4wvO6vMfpkEZJ7nBAUq80vFlzXTtfG/D0eVwmYKqP63Hr47skzeeUosJCaZyZaer1msetvU8oYFoTRA0FUB3iT/aIx1UzHXr8H+8T8fyh+2pwNmLZ+lLpsiTJUxtgBgAAR/Tflf+VRkVnmumgBOTcn3VkiwEAgHpB0NamKjdvlv0LFoTnmz35RMwy5jT7NBRQ1QZPC3csNH/TjNXNJZtlR+kO2VS8KRyMjRuWSEZ5juQWtxZ/eY74qlPEGXRJekWWZJTliDvoMZmzTsupOY/iCmVSHsau1E0m4FrmKa2tHCpS6imWCne5aEi52hGQDf7vpcRTKlYgxWRbWhrc1ICoyehMF0uzP8uSxIrI7KxZV5cEq9NFTOA0lFV6YFaqBlTdtX8PBVRrp4O+iIDtiQsFHDWIaYKUToc0TvNK66xUSfW6TcBVg6Wa6amBSl2mWUayuJ0OyU1PklSf+8egaW0QM/S4uv9mpXrjIqCo2dQ7d1qSk5MlTk17BQDAJiZOnGgugcDJNShFjelT35PsYE2W7e7Ub+XSQaPYNAAAoF4QtLWhQGGhrLvyqvB8+pVXiK9t2xN+vMiM1lDjJ81sDf1Ns1/3lO0xHe/1WoOxxVU1jae+3fWtbC3dKvFIA7CaGduk5DTpUNBT3FWNxFGdKsmBZPEFvJIc8InPZIGeuFmNdkmJKyDrPOUSdGigNE3E8pvT70026oElCop+bBqnAc4kt1M8Lqe4XQ7xuV0mY9TjdZi/hU53D2V2hqY1wPnj32tu02Coz1OToRo6Rd5kq+qp8qFlIzJWw1mpEUFYvWSmeM1jZafVBE/1vhqA1QzSJLfLPAYAAIhf+fn55lJUVCR+vz/Wq2NrC79cJBnFZ5npakeVDLq+b6xXCQAAJBCCtja0e/IUCe6vPd3f6ZTc++8/6n20pICWCNASBRp03VO+R1buXimLdyw21zp/qmi2p1XZWIKV2bV1PyNOrdegae3f9DYNdlqBpPB93UG3uSQFPNKkKlW8AY8kB73ir/aK12TAOsRhOSTJckmjgFvcllNclubFiviOkhl7NC6/Rxx6Sn74IubaneoRX5ZP8jpnyTlNupoAa54/SXyu2ufT5Ndg0HRNbJLTxGRqRoY6Q6fla3AUqG+6/2VkZISnAQDAoX3w5jxpYp1tpvemLZPefY5+zA0AQKLiu2b0ETWymWBFheyeNCk8nzV8uLibHLqD7caijfLkoidlRcEKKSgvCNd/jfo6VWlTpyYiltuc9q8NnoIVObUBWJ3Prm0wVVdKUCQz6JTOGSnSssIpyWVBcQVFXFpXIGiJBCyR6poqq9HmS3WLL9kt6dnJktUsTRo3T5XMvFTxJrvE5XaKPzvZBGhP5vT68iS3yZbl9HrEE4/HIyNHjoz1agAAENc+mjlLskprsmwrneVy09Cfx3qVAACIa3zXjD6CtjZuPuZp0UJy/m+krN23VpYVLDPZtJoxu2rPKhOoPZ6yBRmutuKwvFJRZUl5lSWBoEggUFs7NZBUU09VA7AaiNXO94EksYK+moCsNps6hEY+t/hTPdK5g1+6tMiQvPQkczp+8aLdUrR8rwTLa2utleh1dOqupaR7xe11msCrN9ktyWke8STVBGg79WsuKX6veLwucXtjXzcVAAAA8enzj1ZJE+lkpov938rpnS6L9SoBAIAEQ9DWRqr37JGS2bPNtOafvnP3WTLrnwOOKTib5WsuVlWm7NznlKAGXAPJIkGvBKsypbq4kxRrI6wT0CorRTo395tmU1oDVQOzfq9L2vt8kuZxSdHuctm+tkg2zdoh2/ZVSPAYMmdTM3zicjvCgdekNI+ZTmnklYy8FElKcYsvxSOZTVNMNqzWWdU0/KRUj7g8NI0CAADAifvX9H9Jk/01Adtyd7Hc9qvBbE4AAFDvCNrahNZIXXPueWZ6e4bI7//HLbv3/fuI9/E4kkzZgqJN10lxRd4xP5dmw7bOThWP02maU+Wm++SGHi3MtT/ZKz5tSuVyScn2UincXCrFu8ulek9QKvZXyr4de2X9llJZf4zP1aZLtjTvkCmeJJcJurbqnCUuF4FX4FSpqqqSKVOmmOlhw4aZU1gAAMCPVnyxU7Ik00xXZH4nLVv82AAYAADwXbO+ELS1iT1/+avJrp3b2SEv/vzgU/vbp1wonqqfyOa95bK9IFUCZa1qGn4dyBJpnZksfVtnSdMUr/h9HrEsS9K9LsltlCRep0PSxSmV5QHZX1QpaxbuECm0pGDNOtlUXCmBgCWW1ps9AcmNPKaGrJYqyG3rl+4DWpEZC9Qzfb9v3bo1PA0AaFgmTpxoLoFAdEpPJZqpr/xFsso7mulSzx65a9QvY71KAADYAt81o4+grQ0EKytl2/PPym+Gu2RLdt3mWN6Si2X3pkvk60MFaJUl8pO0JOnTJF1aVTsluK5UqtdXi6zfbW4urF2sSEQ2R2l9tWRBi9MzJat5mplv3CzVZNM2apwUpWcAAADAoeTn55tLUVGR+P1+NtJxqK6uls3LA7U5tiKO3I2SmRmaAwAAqF8EbW1g85OPyajBVbItKyJga4k4tl0jVXvOMQeWSZZIWtAhSZZDWiZ5paXHKyn7qsRZadVEZrcUS+VJrIOWL9AmX94kt6k363Q5zd9yWjaSnNbpkpaZJG6P0zT60mUAAAAAO5k04RXJrKjJsi3ybpeR9/0q1qsEAAASGNG1OFe5fbv8T/p0KU7xSub+xtJt8wDJK2onyVXp4pa6WbdhZeaeh31Mt9dpMl99KW5xup2mkZfT5Qhfa5MvLWWgzb1CWbMalAUAAAAaorKycilaly7ptfOpLfdKcjLHvwAAIHYI2sa5oS8PlPNK7pamxW3FZZ3YcOW19Ys/J1nSMn2SmZcq7XvkmExZAAAAACIv/OlFSa/sYjbFXt9GGT3yf9ksAAAgpgjaxrGXXnlc+m59+rC357RqZBp7aVkCp8cpyake8aa4JdXvk4ycFEnN8Epqhs9kzAIAAAA4WOG+Qqne1jw83/T0oLjdfE0CAACxxdFInNKOv9VLzq7TXiw50yc/7ZlrGny175lryhkAwPFKSUlhowEAUOuFP70i6VXdzPTupDXyyAiybAEAOBF814wugrZxas6/PxGXlRueb929sVz2yy5kzQI4KV6vVx544AG2IgAAIrJt6zbx7G4X3han92rMdgEAgO+acSGhCptOnDhRWrduLUlJSdK7d29ZsGCBxKut3/yYY1ua8Z1cPrwrAVsAAAAc9Xj3jDPOkJ49e7KljsGrE/4mydU17ccKklfIDb+4ge0GAADiQsJk2k6fPl1GjRolL730kgnYPvPMMzJgwABZvXq15OTkSDz5/YNPSk5Vj/D8tcMuiun6AAAAwB7y8/PNpaioSPx+f70855z/zJXP5y6QYDAoTqfmhNilhJclaUVnmKmgBKV3/w6xXiEAAIDEC9o+/fTTMnz4cBk2bJiZ1+DtBx98IJMnT5bRo0fXWbaiosJcQvSgV+mBqF5ONf/+08PT+5K+l7btL6yX50X06HhZlsW42VRDHr+qqip57bXXzPTgwYPF4/FIQ9SQxzARMH72Vt/jx/s89pYvXSEZBT8mHNjR7tRlcs/lo2K9GgAA2FaifNesTwkRtK2srJTFixfLmDFjwn/TLID+/fvLF198cdDy48aNk9///vcH/X3Xrl1SXl5+ytc34Kgy11XOcrn81u6yc+fOU/6ciP4XyMLCQvOltSbjBHbSkMdPPw9Xrlxppnfs2GFq3DZEDXkMEwHjZ2/1PX7FxcWn/DnQsFW69svAa8+N9WoAAGBreuy3fv368DROXkIEbQsKCiQQCEhu7o+NvZTO//e//z1oeQ3uaimFyEzbFi1aSJMmTSQ9vabm1amU1Ga7lFdukEFXDpSfnt7xlD8fTs0XVofDYfYZAkb205DHT4O2qampZlpLwzTkoG1DHcNEwPjZW32Pn/YqQGydc15v+axqvlRXV4vb7bZReYSaVe3Vp7ucc945sV4TAACAxAvaHi+fz2cuB9IvHvXx5SP//0aY7FoNqBBssC/9wlpf+wyir6GOn74efW2h6Yb2+hJhDBMF42dv9Tl+vMdjr+c5PaR7r24cvwIAAERRQnyTzc7OFpfLZU4FjqTzeXl5MVsvAAAAAAAAAEjIoK2e/tu9e3eZPXt2nVP3dL5Pnz4xXTcAAAAAAAAASMjyCFqjdujQodKjRw/p1auXPPPMM1JaWirDhg2L9aoBAAAAAAAAQOIFbW+66SbZtWuXPPLII7J9+3bp2rWrzJw586DmZADQ0Hk8nlivAgAAAACggeG7ZnQlTNBW3X333eYCAIlKy8U8/PDDsV4NAMApMnHiRHMJBAJsYwAAUG/4rhl9CVHTFgAAAEgE+fn5snLlSlm4cGGsVwUAAAAngaAtAAAAAAAAAMSRhCqPAACJrrq6WqZPnx6u9e12898AAAAAAIDvmvGGb+sAkECCwaCsWbMmPA0AAAAAAN814w/lEQAAAAAAAAAgjhC0BQAAAAAAAIA4QtAWAAAAAAAAAOIIQVsAAAAAAAAAiCMEbQEAAAAAAAAgjrhjvQJ2YFmWuS4qKqqX59OO7sXFxZKUlCROJ3F1O2IM7a0hj19lZaVUVFSEP9O8Xq80RA15DBMB42dv9T1+oeOz0PEaanD8ioaE/xfA/gW7SqTPr0T5rlmfx68OiyPco9q8ebO0aNEiKgMDAACA6Nu0aZOcdtppbNpaHL8CAADY+/iVoO0x/jKydetWadSokTgcDqmPiLsGiXXw0tPTT/nzIfoYQ3tj/OyPMbQ3xs/e6nv8NP9AM1iaNWvW4DNYonH82rNnT1m4cOFh73eit9v5fXu01xyvz3Wij3W89zvW5Y9lOfavhr1vnch92b/s8fnF/pV4n1+J9n9jfX9+HevxK+URjoFuwFhkbugb2k5vahyMMbQ3xs/+GEN7Y/zsrT7Hz+/318vzNITjV5fLdcRxOdnb7fi+PdpritfnOtHHOt77Hevyx7Ic+1fD3rdO5L7sX/b4/GL/SrzPr0T7vzEWn1/HcvxKOgIAAACQIPLz80/p7XZUn68pms91oo91vPc71uWPZTn2r4a9b53Ifdm/7PH5xf6VeJ9fifZ/Y6w/vw6H8ghxSNPnNeJeWFhoq19i8CPG0N4YP/tjDO2N8bM3xi8xMe5g/4Jd8fkF9i/EKzJt45DP55OxY8eaa9gTY2hvjJ/9MYb2xvjZG+OXmBh3sH/Brvj8AvsX4hWZtgAAAAAAAAAQR8i0BQAAAAAAAIA4QtAWAAAAAAAAAOIIQVsAAAAAAAAAiCMEbQEAAAAAAAAgjhC0BQAAAAAAAIA4QtA2Dk2cOFFat24tSUlJ0rt3b1mwYEGsVykhffLJJ3LFFVdIs2bNxOFwyDvvvFPndsuy5JFHHpGmTZtKcnKy9O/fX9asWVNnmT179sjgwYMlPT1dMjIy5I477pCSkpI6yyxbtkzOP/98M94tWrSQxx9/vF5eX0M3btw46dmzpzRq1EhycnLk6quvltWrV9dZpry8XPLz8yUrK0vS0tLkuuuukx07dtRZZuPGjXL55ZdLSkqKeZz7779fqqur6ywzd+5c6datm/h8PmnXrp1MnTq1Xl5jQ/biiy/KWWedZd47eunTp4/MmDEjfDtjZy/jx483n6MjR44M/40xjG+PPvqoGbPIy09/+tPw7Ywfjsf7778vHTt2lPbt28urr77KxkNUXXPNNZKZmSnXX389WxZRtWnTJrnwwgvljDPOMMelb775JlsYUbNv3z7p0aOHdO3aVTp16iSvvPIKWxcHsxBXpk2bZnm9Xmvy5MnWihUrrOHDh1sZGRnWjh07Yr1qCefDDz+0Hn74Yeutt96y9K3y9ttv17l9/Pjxlt/vt9555x1r6dKl1pVXXmm1adPGKisrCy8zcOBAq0uXLtaXX35pffrpp1a7du2sW265JXx7YWGhlZubaw0ePNhavny59cYbb1jJycnWpEmT6vW1NkQDBgywpkyZYrbrkiVLrMsuu8xq2bKlVVJSEl5mxIgRVosWLazZs2dbixYtss455xzr3HPPDd9eXV1tderUyerfv7/1zTffmH0iOzvbGjNmTHiZtWvXWikpKdaoUaOslStXWhMmTLBcLpc1c+bMen/NDcl7771nffDBB9Z3331nrV692nrooYcsj8djxlMxdvaxYMECq3Xr1tZZZ51l3XvvveG/M4bxbezYsdaZZ55pbdu2LXzZtWtX+HbGD8eqqqrKat++vbV582aruLjY6tChg1VQUMAGRNTMmTPHHDdcd911bFVE1datW813AKX/DzZr1qzOdwngZOh3zdLSUjOt+5UeL/P/Iw5E0DbO9OrVy8rPzw/PBwIB85/DuHHjYrpeie7AoG0wGLTy8vKsJ554Ivy3ffv2WT6fzwRelQbw9H4LFy4MLzNjxgzL4XBYW7ZsMfMvvPCClZmZaVVUVISXefDBB62OHTvW0ytLHDt37jTjMW/evPB4aRDwzTffDC+zatUqs8wXX3xh5jVI63Q6re3bt4eXefHFF6309PTwmD3wwAMmsBHppptuMkFjRJe+V1599VXGzkY0QKPBmlmzZln9+vULB215/9kjaKs/Oh4K44fjMX/+fOvqq68Oz+vnwOuvv85GRNQDtwRtcarpD9AbN25kQyPqdu/ebbVq1arOD+SAojxCHKmsrJTFixeb0+xDnE6nmf/iiy9ium6oa926dbJ9+/Y6Y+X3+005i9BY6bWWRNBTHkJ0eR3Tr776KrzMBRdcIF6vN7zMgAEDzGn8e/fuZbNHUWFhoblu3Lixudb3WlVVVZ0x1FN/W7ZsWWcMO3fuLLm5uXXGp6ioSFasWBFeJvIxQsvwno2eQCAg06ZNk9LSUlMmgbGzDy0/ouVFDnyPMIb2oCV/tERQ27ZtTakfLRejGL/EcrRyUUcr7bV161Zp3rx5eF6nt2zZUm/rj4a9fwH1tX/p/316TKrl7IBo7V9aIqFLly5y2mmnmTJ82dnZbFzUQdA2jhQUFJj/CCIDRErnNUCI+BEajyONlV5rDdRIbrfbBA0jlznUY0Q+B05eMBg0tTTPO+88Uy8otH01WK6B9SON4dHG53DLaGC3rKyM4TsJ3377rak1rLWCR4wYIW+//bapKcbY2YMG2r/++mtTX/pAjGH80y8WWp975syZpsa0/lip9deLi4sZvwSjP5jpF0r94nko06dPl1GjRsnYsWPNe16X1R8vd+7cWe/rCvth/4Id9i/tUzJkyBB5+eWXGTBEdf/S76JLly41x1mvv/76Qf1VADebAEAiZPstX75cPvvss1ivCo6DNq1ZsmSJyZL+5z//KUOHDpV58+axDW3SuOPee++VWbNmmcwC2M+gQYPC09p8RYO4rVq1kn/84x+m+SYSa1+I3B8O9PTTT8vw4cNl2LBhZv6ll16SDz74QCZPniyjR482GUiRmbU63atXr3pZdzT8/Qs41ftXRUWFaWis8+eeey4bHKfk80uTfjSo++mnn9JUEXWQaRtHNBXe5XId9OuKzufl5cVsvXCw0Hgcaaz0+sBfaaurq80vtZHLHOoxIp8DJ+fuu+82XavnzJljTjuJHEMtSaKnpBxpDI82PodbJj09ncDGSdJM6Hbt2kn37t1NtqYeyDz77LOMnQ3oKYT6+detWzdzhoFeNOD+3HPPmWk9MOX9Zy+aCdKhQwf5/vvveQ/iuEp7aYBWfzjVYG1JSYnMmDHDZBoBR0PpOMR6/9LWJrfddptcfPHFcuuttzIgiOr+pd8Z9QwmpUkqWm5Bk1aASARt4yxAocGJ2bNn1zmtW+e1jiPiR5s2bcyX1six0tPhtVZtaKz0WgOC+mEd8vHHH5sx1Yyl0DL64ay1VUM0M00/rDMzM+v1NTU0epClAVs9pV63u45ZJH2veTyeOmOotYS1ZmPkGOop+pHBdx0fDcjqafqhZSIfI7QM79no0/eOZjswdvHvkksuMe8dzZQOXbS+t9ZFDU3z/rMXDbb98MMP0rRpU96DOK7SXvpDzVNPPSUXXXSRdO3aVe677z7JyspiKyJqpeM0CHLDDTfIhx9+aH6gp68AorV/zZ8/35zirrVK9fNLL3p8A0Rj/9qwYYMpPaWJKXp9zz33mH4qQB30Y4sv06ZNs3w+nzV16lRr5cqV1p133mllZGTU6V6P+ut6/s0335iLvlWefvppM71hwwZz+/jx483YvPvuu9ayZcusq666ymrTpo1VVlYWfoyBAwdaZ599tvXVV19Zn332memifsstt9TpwJ2bm2vdeuut1vLly834p6SkWJMmTWKYT9Jdd91l+f1+a+7cuda2bdvCl/3794eXGTFihNWyZUvr448/thYtWmT16dPHXEKqq6utTp06WZdeeqm1ZMkSa+bMmVaTJk2sMWPGhJdZu3atGbP777/fWrVqlTVx4kTL5XKZZXHiRo8ebc2bN89at26deX/pvMPhsD766CPGzqb69etnusaH8P6Lb/fdd5/5/NT34Pz5863+/ftb2dnZ1s6dO83tjF9i0uOht99+Ozy/ZcsW87fPP/+8znL6f2KvXr1isIawM/YvsH/Brvj8wqlC0DYOTZgwwQSSvF6vOeD98ssvY71KCWnOnDnmw/fAy9ChQ83twWDQ+t3vfmeCrhpov+SSS6zVq1fXeYzdu3ebIG1aWpqVnp5uDRs2zASDIy1dutTq27eveYzmzZubYDBO3qHGTi9TpkwJL6MB9l/96ldWZmamCbxec801JrAbaf369dagQYOs5ORkE7DQQEZVVdVB+0rXrl3Ne7Zt27Z1ngMn5vbbb7datWpltqkGyvX9FQrYMnYNI2jL+y++3XTTTVbTpk3Ne1D/b9L577//Pnw745eYDvxSWlFRYX6ojPybGjJkiHXllVfGYA1hZ+xfYP+CXfH5hVPFof/Uzb0FAAAAgLocDocpO6RNeUK05JPWrZ0wYUK4lE3Lli1NiSIaReF4sH/hVGL/AvsX7Mgd6xUAAAAAEL/1jLUBXci6detMberGjRub4OyoUaNk6NChpla1Bm+feeYZKS0tDXfLBti/ECt8foH9C3ZHpi0AAACAQ5o7d65pInYgDdROnTrVTD///PPyxBNPmOYq2qjnueeeCzddBY6E/QunEvsX2L9gdwRtAQAAAAAAACCOOGO9AgAAAAAAAACAHxG0BQAAAAAAAIA4QtAWAAAAAAAAAOIIQVsAAAAAAAAAiCMEbQEAJ+XCCy+UkSNHshUBAAAAAIgSgrYAYAO33XabOBwOc/F4PNKmTRt54IEHpLy8PNarBgAAAAAAoswd7QcEAJwaAwcOlClTpkhVVZUsXrxYhg4daoK4jz32GJscAAAAAIAGhExbALAJn88neXl50qJFC7n66qulf//+MmvWLHNbRUWF/PrXv5acnBxJSkqSvn37ysKFC8P3nTp1qmRkZNR5vHfeeccEfUMeffRR6dq1q/ztb3+T1q1bi9/vl5tvvlmKi4vDy5SWlsqQIUMkLS1NmjZtKk899VS9vHYAAAAAABIJQVsAsKHly5fL559/Ll6v18xrqYR//etf8pe//EW+/vpradeunQwYMED27NlzXI/7ww8/mGDu+++/by7z5s2T8ePHh2+///77zd/effdd+eijj2Tu3Lnm+QAAAAAAQPQQtAUAm9Agqma4aiZt586dZefOnSaIqtmvL774ojzxxBMyaNAgOeOMM+SVV16R5ORk+fOf/3xczxEMBk1WbqdOneT888+XW2+9VWbPnm1uKykpMY/35JNPyiWXXGLWQYPE1dXVp+gVAwAAIBEa1l5wwQXy+uuvh+f1bDBNJLCz430No0ePlnvuueeUrhMAeyFoCwA2cdFFF8mSJUvkq6++MvVshw0bJtddd53JjtU6t+edd154WW1W1qtXL1m1atVxPYeWRWjUqFF4XksgaHBY6fNUVlZK7969w7c3btxYOnbsGJXXBwAAgMRrVvvee+/Jjh07TFmuRPab3/zGJESsXbs21qsCIE4QtAUAm0hNTTVlD7p06SKTJ082wdtjzaR1Op1iWVadv2mg90B6sB9JvwBo9i0AAADit1nttm3bTLDvT3/6k0yaNEnGjh0rdvHcc8+ZZAQ9Xk1k2dnZpryZnkEHACqxPxUBwKb0oPahhx6S3/72t/KTn/zE1LadP39+nYCsNiLTUgmqSZMmpqGYllII0azd46HPo0FdDRaH7N27V7777ruovCYAAABEt1ltvDes3bVrl3z88cdyxRVXHHG5b7/9Vi6++GJT/isrK0vuvPNOU7orRMt16WvU16G3P/jgg+bMNN0eh7NhwwbzvJmZmSY54swzz5QPP/wwfPuKFSvk5z//uaSnp5sz0bR0mJ55pnT7/exnPzOBVt0W/fr1O2qfh02bNsmNN95o1lHPVrvqqqtk/fr1dZbR9Zk2bdpRtxuAxEDQFgBs6oYbbhCXy2V+jb/rrrtMfduZM2fKypUrZfjw4bJ//3654447zLJa0iAlJcUEevVgU2uG6QH68dADcH08fR49uNZmaHpKXqJnRQAAAMRrs9p4b1j72WefmWPU008//bDLaDBY11eDqxosffPNN+U///mP3H333eFlHnvsMXnttddkypQpJpGhqKjoqPVk8/PzTUD7k08+MUFhfQw93lVbtmwxdXY1IK7HvYsXL5bbb7893MtBg9UaFNb1//LLL6V9+/Zy2WWX1QliR9KECn0NGvz99NNPzTrqc2mWtJYfC9HyZps3bz4omAsgMbljvQIAgBPjdrvNwerjjz8u69atM2UMtHGYHiz26NFD/v3vf5uDW6W/5v/97383B9PapEwbiWnGhGYpHA9tdqZZDZoFoAed9913nxQWFjKEAAAAMW5WqwFFDULqD+rPP/+8uS3UsFZ/rNeGtUqPBTUTV8ts6bHh8TasDfU/CDWs/eMf/xhuWKvHm3qcqTRIfNpppx3xMTXbNTc394hJAJpsoDV6//rXv5qMWKWvT49HNdCq958wYYKMGTNGrrnmmvDtkVmzh7Jx40bTH0Kb66q2bduGb5s4caLJoNWs11D5sA4dOoRv16zfSC+//LLJoNWgtWbnHmj69Olm+7366qvhDGYNMOt9NLh96aWXmr81a9YsvF00oxlAYiNoCwA2cLisWO0yq5dQPTC9HI6eHnbgKWKakRuiQVy9RNJuv5Edf/ULgZ4Wp5eQ4znYBwAAQPSb1WpgVgO0WtNWf9jXYKSK94a1ZWVlpmTDkeh6ak+HUMBW6evRIOjq1avN/bWRmb6mED0brXv37kfszaDlFPRsNc0K1pISus3OOuuscBkxLYdwYL+HEH0+LVOmAVfdBoFAwJzlpoHgQ1m6dKl8//33dbaf0mB0qOSC0vIPSh8LADinFQAAAACABGxWG+uGtVoTVnskxMIvf/lL07xNM4a1PIKeqaYZu5HB08PR0gga2H322WdNOQqd1lq6kaUOImkmsgaRdbnIi/aG+MUvfhFeLlSyQvtRAABBWwAAAAAAGlizWs1ijfeGtWeffbZs3779iIFbrXermaqR66evR1+rZvJqGQMtkRDZXE0zX49WT1dp87YRI0bIW2+9Zcp+aekIpRm3Wnv2UMHr0PNrpq7WsdUGZlr7tqCg4LDP061bN1mzZo1pBqcB9siLrn9kTWLdjvqYAEDQFgAAAACABtasVuuyahZuPDes1aCtZttGBpUPNHjwYFMCQbNb9XHnzJkj99xzj8mQ1WCt0vlx48aZJmhaMuHee+81geBQ/dhD0RJg2gNCe0NogFcfN9QQTftGaDOzm2++WRYtWmQCrloeTB9baeMxndfSDRqo1nU8Unau3q6v86qrrjLBYH1OLa2ggV9tPBait2lZhqNl+gJIDARtAQAAAABogM1qNTt1/Pjxpl6rBjk141Nrqx6qYa027tKmXG+88cZBfQ6OtWGtBhy1QZjWiO3bt68pCXAkGlweNmyYvPbaa4ddRgPKur5aOqBnz55y/fXXm2ZnoWZr6sEHH5RbbrlFhgwZIn369DFB5AEDBhyxXq5m4+bn55tA7cCBA02jsRdeeMHcpqUONPisZQ369etnXodm4YZKRGj5CQ0K6/bU7arBV82iPdJr+OSTT6Rly5Zy7bXXmufUILfWtE1PTw8vp43PIntOAEhsDuvA4jUAAAAAAAD1QMsjaDkAzXZt1apVVB5Ta+1qYPTGG2+UP/zhD2IHM2bMMCUali1bZgLvAMAnAQAAAAAAiIm8vDyTubpx48YTDtpu2LBBPvroI5MVW1FRYbJwtQRBZJOveKdZ0VOmTCFgCyCMTFsAAAAAAGBbmzZtMvVnteatnkzcqVMnUxbiggsuiPWqAcAJI2gLAAAAAAAAAHGERmQAAAAAAAAAEEcI2gIAAAAAAABAHCFoCwAAAAAAAABxhKAtAAAAAAAAAMQRgrYAAAAAAAAAEEcI2gIAAAAAAABAHCFoCwAAAAAAAABxhKAtAAAAAAAAAEj8+H8MNGQZWino/AAAAABJRU5ErkJggg==", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stdout", "output_type": "stream", "text": [ "Final cumulative regret:\n", " Default (α=1, β=1): 1137.7\n", " Misspecified (α=10, β=0.1): 1371.9\n", " EB (α=10, β=0.1): 1183.2\n", " EB + decay (lr=0.995): 491.4\n" ] } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 5))\n", "\n", "colors = {\n", " \"Default (α=1, β=1)\": \"C0\",\n", " \"Misspecified (α=10, β=0.1)\": \"C3\",\n", " \"EB (α=10, β=0.1)\": \"C2\",\n", " \"EB + decay (lr=0.995)\": \"C4\",\n", "}\n", "\n", "for name in agents:\n", " regret = np.cumsum(optimal_rewards - expected_rewards[name])\n", " ax1.plot(regret, label=name, linewidth=2, color=colors[name])\n", "\n", "ax1.axvline(\n", " x=SHIFT_ROUND, color=\"black\", linestyle=\"--\", alpha=0.5, label=\"Regime shift\"\n", ")\n", "ax1.set_xlabel(\"Round\")\n", "ax1.set_ylabel(\"Cumulative Regret\")\n", "ax1.set_title(\"Cumulative Regret\")\n", "ax1.legend()\n", "ax1.grid(True, alpha=0.3)\n", "\n", "# Log-log scale to see sublinear behavior\n", "rounds = np.arange(1, N_ROUNDS + 1)\n", "for name in agents:\n", " regret = np.cumsum(optimal_rewards - expected_rewards[name])\n", " ax2.loglog(rounds, regret, label=name, linewidth=2, color=colors[name])\n", "\n", "ax2.axvline(\n", " x=SHIFT_ROUND, color=\"black\", linestyle=\"--\", alpha=0.5, label=\"Regime shift\"\n", ")\n", "ax2.set_xlabel(\"Round (log scale)\")\n", "ax2.set_ylabel(\"Cumulative Regret (log scale)\")\n", "ax2.set_title(\"Cumulative Regret (Log-Log)\")\n", "ax2.legend()\n", "ax2.grid(True, alpha=0.3)\n", "\n", "plt.tight_layout()\n", "plt.show()\n", "\n", "# Print final regret values\n", "print(\"Final cumulative regret:\")\n", "for name in agents:\n", " regret = np.cumsum(optimal_rewards - expected_rewards[name])[-1]\n", " print(f\" {name}: {regret:.1f}\")" ] }, { "cell_type": "markdown", "id": "1cboap3fes6", "metadata": {}, "source": [ "**What to notice:**\n", "\n", "- **Before the regime shift** (rounds 0–2499): The stationary EB agent (green) performs best — it quickly corrects its hyperparameters and exploits the strong-signal, low-noise environment. The default agent (blue) does well too. EB + decay (purple) pays a visible regret cost due to its smaller effective sample size. The misspecified agent (red) struggles throughout.\n", "\n", "- **After the regime shift** (rounds 2500–4999): The three stationary agents' regret curves steepen — they're pulling the wrong arms with outdated coefficients. The default and EB agents have accumulated so much data from phase 1 that they can't adapt; their learned coefficients are locked in. The **EB + decay agent adapts**: its exponential forgetting discounts the stale phase-1 data, and the EB mechanism re-tunes the hyperparameters to the new regime. Its post-shift regret slope is noticeably lower than the others.\n", "\n", "- The **key tradeoff**: decay costs you regret in stationary phases but buys you adaptation when the environment changes. In this experiment, the regime shift is dramatic enough that the adaptation benefit dominates." ] }, { "cell_type": "markdown", "id": "qumuxibv0d", "metadata": {}, "source": [ "## Hyperparameter Evolution\n", "\n", "The key advantage of empirical Bayes is visible in how the hyperparameters evolve. Despite starting from badly misspecified values (`alpha=10`, `beta=0.1`), both EB agents quickly adjust them toward values that better reflect the data.\n", "\n", "In phase 1, the true noise precision is `beta_true = 1 / 0.5^2 = 4.0`. In phase 2, it drops to `beta_true = 1 / 1.0^2 = 1.0`. We expect the EB + decay agent's `beta` to track this shift, while the stationary EB agent's `beta` remains near its phase-1 value.\n", "\n", "The EB + decay agent uses **stabilized forgetting** (Kulhavý & Zarrop, 1993): at each step, a fraction of the EB-tuned prior precision is re-injected into the model, preventing the prior from decaying to zero over time. This ensures that the EB-estimated `alpha` always has a channel to influence the model — and crucially, allows the hyperparameters to adapt when the optimal values change." ] }, { "cell_type": "code", "execution_count": 5, "id": "5rwmm2sp8k4", "metadata": { "execution": { "iopub.execute_input": "2026-03-08T20:09:34.008620Z", "iopub.status.busy": "2026-03-08T20:09:34.007952Z", "iopub.status.idle": "2026-03-08T20:09:34.334686Z", "shell.execute_reply": "2026-03-08T20:09:34.333773Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 5))\n", "\n", "# Skip the first few rounds where alpha is wildly unstable (insufficient data)\n", "SKIP = 50\n", "\n", "ax1.plot(\n", " range(SKIP, N_ROUNDS), eb_alpha_history[SKIP:], linewidth=2, color=\"C2\", label=\"EB\"\n", ")\n", "ax1.plot(\n", " range(SKIP, N_ROUNDS),\n", " eb_decay_alpha_history[SKIP:],\n", " linewidth=2,\n", " color=\"C4\",\n", " linestyle=\"--\",\n", " label=\"EB + decay\",\n", ")\n", "ax1.axhline(y=10.0, color=\"C3\", linestyle=\":\", alpha=0.7, label=\"Initial value (10.0)\")\n", "ax1.axvline(\n", " x=SHIFT_ROUND, color=\"black\", linestyle=\"--\", alpha=0.5, label=\"Regime shift\"\n", ")\n", "ax1.set_xlabel(\"Round\")\n", "ax1.set_ylabel(\"alpha (prior precision)\")\n", "ax1.set_title(\"Prior Precision (alpha) Over Time\")\n", "ax1.legend()\n", "ax1.grid(True, alpha=0.3)\n", "\n", "ax2.plot(\n", " range(SKIP, N_ROUNDS), eb_beta_history[SKIP:], linewidth=2, color=\"C2\", label=\"EB\"\n", ")\n", "ax2.plot(\n", " range(SKIP, N_ROUNDS),\n", " eb_decay_beta_history[SKIP:],\n", " linewidth=2,\n", " color=\"C4\",\n", " linestyle=\"--\",\n", " label=\"EB + decay\",\n", ")\n", "ax2.axhline(y=0.1, color=\"C3\", linestyle=\":\", alpha=0.7, label=\"Initial value (0.1)\")\n", "# Show true noise precision for both phases\n", "ax2.axhline(\n", " y=1.0 / TRUE_NOISE_STD_1**2,\n", " color=\"gray\",\n", " linestyle=\":\",\n", " alpha=0.7,\n", " label=f\"Phase 1 true β ({1.0 / TRUE_NOISE_STD_1**2:.1f})\",\n", ")\n", "ax2.axhline(\n", " y=1.0 / TRUE_NOISE_STD_2**2,\n", " color=\"gray\",\n", " linestyle=\"--\",\n", " alpha=0.5,\n", " label=f\"Phase 2 true β ({1.0 / TRUE_NOISE_STD_2**2:.1f})\",\n", ")\n", "ax2.axvline(\n", " x=SHIFT_ROUND, color=\"black\", linestyle=\"--\", alpha=0.5, label=\"Regime shift\"\n", ")\n", "ax2.set_xlabel(\"Round\")\n", "ax2.set_ylabel(\"beta (noise precision)\")\n", "ax2.set_title(\"Noise Precision (beta) Over Time\")\n", "ax2.legend()\n", "ax2.grid(True, alpha=0.3)\n", "\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "o6loesga7ml", "metadata": {}, "source": [ "**What to notice:**\n", "\n", "- **Phase 1** (before the dashed line): Both `alpha` traces drop rapidly from 10.0 and both `beta` traces climb from 0.1 toward the true noise precision of 4.0, as in the stationary case.\n", "- **After the regime shift**: The stationary EB agent's (solid green) hyperparameters barely move — 2500 data points of phase-1 evidence anchor its estimates. Its `beta` stays near 4.0 even though the true value is now 1.0.\n", "- The **EB + decay agent** (dashed purple) adapts: its `beta` drops toward the new true value of 1.0, and its `alpha` rises as the weaker phase-2 signal requires more regularization. This is the mechanism behind its lower post-shift regret — the hyperparameters track the new regime because old data has been forgotten.\n", "- The EB + decay traces are noisier throughout, reflecting the smaller effective sample size (~200 observations). This is the cost of adaptability." ] }, { "cell_type": "markdown", "id": "l8jzg5lkq7", "metadata": {}, "source": [ "## Discussion\n", "\n", "### When to use `EmpiricalBayesNormalRegressor`\n", "\n", "- **Unknown noise scale**: You don't know how noisy your rewards are (common in most real applications)\n", "- **Unknown signal strength**: You don't know how large the true coefficients are, making it hard to set regularization\n", "- **Sensitivity to hyperparameters**: Your problem is one where bad `alpha`/`beta` values lead to poor performance\n", "- **No offline data for tuning**: You can't cross-validate hyperparameters before deployment\n", "\n", "### When plain `NormalRegressor` suffices\n", "\n", "- **Well-understood domains**: You have strong prior knowledge about the noise level and coefficient scales\n", "- **Offline tuning available**: You can cross-validate `alpha` and `beta` on historical data before deploying the bandit\n", "- **Computational budget is tight**: EB adds a MacKay update step per `partial_fit` call. For most problems this overhead is negligible, but in extremely latency-sensitive applications it may matter\n", "\n", "### Decay and stabilized forgetting\n", "\n", "As this notebook demonstrates, decay is not just \"insurance\" — it is **actively necessary** when the environment is non-stationary. After the regime shift at round 2500:\n", "\n", "- The **stationary agents** (including stationary EB) accumulate regret at a steep rate because their learned coefficients and hyperparameters are locked into the phase-1 regime. With 2500 observations already incorporated, new data barely moves the posterior.\n", "- The **EB + decay agent** adapts: exponential forgetting discounts stale phase-1 observations, and the MacKay updates re-tune `alpha` and `beta` to the new noise level and signal strength. The hyperparameter evolution plot shows this clearly — `beta` drops from ~4 toward 1 after the shift.\n", "\n", "A naive implementation of decay causes the prior to vanish over time — after enough steps, `gamma^n * alpha ≈ 0`, leaving new or dormant coefficients with no regularization. `EmpiricalBayesNormalRegressor` addresses this with **stabilized forgetting** (Kulhavý & Zarrop, 1993): at each decay step, a fraction `(1 - gamma) * alpha` of the EB-tuned prior precision is re-injected into the precision matrix. This ensures:\n", "- The prior precision converges to `alpha` in steady state (instead of decaying to zero)\n", "- The EB-estimated `alpha` always influences new/dormant coefficients\n", "- Cold-start features receive the current population-level regularization, not a decayed remnant\n", "\n", "The tradeoff is visible in phase 1: EB + decay pays a regret cost compared to stationary EB because its effective sample size is smaller (~200 vs 2500). In practice, the right `learning_rate` depends on your expected rate of environmental change — faster change warrants more aggressive decay.\n", "\n", "### Key takeaway\n", "\n", "The empirical Bayes approach provides **robustness to hyperparameter misspecification** at minimal computational cost. Pairing it with decay and stabilized forgetting adds **robustness to non-stationarity** — the model adapts its hyperparameters and coefficients to track a changing environment, rather than remaining anchored to stale data." ] } ], "metadata": { "kernelspec": { "display_name": "bayesianbandits", "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": "3.10.13" } }, "nbformat": 4, "nbformat_minor": 5 }