diff --git a/TD2 Deep Learning.ipynb b/TD2 Deep Learning.ipynb index d344fe77b722139d1c6378d5b246cd985a56d137..c86f3a01b79fbd92c55dd7ea035376f86cc959e3 100644 --- a/TD2 Deep Learning.ipynb +++ b/TD2 Deep Learning.ipynb @@ -2310,161 +2310,150 @@ ")\n" ] }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - " -the code conducts fine-tuning on a pre-trained ResNet18 model using the Hymenoptera dataset for binary classification. Over the 10 training epochs: \n", - " The initial epoch reveals moderate training accuracy and lower validation accuracy. \n", - " Subsequent epochs exhibit improvements in both training and validation accuracy, peaking at 94.77% in validation Training loss fluctuates, yet the model effectively generalizes to the validation set." - ] - }, - { - "cell_type": "markdown", - "id": "bbd48800", - "metadata": {}, - "source": [ - "Experiments:\n", - "Study the code and the results obtained.\n", - "\n", - "Modify the code and add an \"eval_model\" function to allow\n", - "the evaluation of the model on a test set (different from the learning and validation sets used during the learning phase). Study the results obtained.\n", - "\n", - "Now modify the code to replace the current classification layer with a set of two layers using a \"relu\" activation function for the middle layer, and the \"dropout\" mechanism for both layers. Renew the experiments and study the results obtained.\n", - "\n", - "Apply ther quantization (post and quantization aware) and evaluate impact on model size and accuracy." - ] - }, { "cell_type": "code", - "execution_count": 109, + "execution_count": 129, "metadata": {}, "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "c:\\Users\\ZINEB\\anaconda3\\lib\\site-packages\\torchvision\\models\\_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.\n", - " warnings.warn(\n", - "c:\\Users\\ZINEB\\anaconda3\\lib\\site-packages\\torchvision\\models\\_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=ResNet18_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet18_Weights.DEFAULT` to get the most up-to-date weights.\n", - " warnings.warn(msg)\n" - ] - }, { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/10\n", - "----------\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "c:\\Users\\ZINEB\\anaconda3\\lib\\site-packages\\torch\\optim\\lr_scheduler.py:136: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`. Failure to do this will result in PyTorch skipping the first value of the learning rate schedule. See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n", - " warnings.warn(\"Detected call of `lr_scheduler.step()` before `optimizer.step()`. \"\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "train Loss: 0.6425 Acc: 0.6680\n", - "val Loss: 0.3366 Acc: 0.9346\n", + "----------\n", + "train Loss: 0.3783 Acc: 0.8197\n", + "val Loss: 0.1810 Acc: 0.9477\n", "\n", "Epoch 2/10\n", "----------\n", - "train Loss: 0.4859 Acc: 0.7582\n", - "val Loss: 0.2422 Acc: 0.9412\n", + "train Loss: 0.3373 Acc: 0.8689\n", + "val Loss: 0.1897 Acc: 0.9346\n", "\n", "Epoch 3/10\n", "----------\n", - "train Loss: 0.4214 Acc: 0.8279\n", - "val Loss: 0.2159 Acc: 0.9346\n", + "train Loss: 0.3545 Acc: 0.8402\n", + "val Loss: 0.1725 Acc: 0.9412\n", "\n", "Epoch 4/10\n", "----------\n", - "train Loss: 0.3997 Acc: 0.8361\n", - "val Loss: 0.2607 Acc: 0.9085\n", + "train Loss: 0.3417 Acc: 0.8320\n", + "val Loss: 0.1810 Acc: 0.9412\n", "\n", "Epoch 5/10\n", "----------\n", - "train Loss: 0.4037 Acc: 0.8115\n", - "val Loss: 0.2860 Acc: 0.8824\n", + "train Loss: 0.3439 Acc: 0.8443\n", + "val Loss: 0.1987 Acc: 0.9346\n", "\n", "Epoch 6/10\n", "----------\n", - "train Loss: 0.3597 Acc: 0.8443\n", - "val Loss: 0.2590 Acc: 0.9085\n", + "train Loss: 0.3766 Acc: 0.8238\n", + "val Loss: 0.1841 Acc: 0.9412\n", "\n", "Epoch 7/10\n", "----------\n", - "train Loss: 0.2971 Acc: 0.8770\n", - "val Loss: 0.2063 Acc: 0.9281\n", + "train Loss: 0.3908 Acc: 0.8115\n", + "val Loss: 0.1838 Acc: 0.9412\n", "\n", "Epoch 8/10\n", "----------\n", - "train Loss: 0.3260 Acc: 0.8566\n", - "val Loss: 0.1808 Acc: 0.9412\n", + "train Loss: 0.3553 Acc: 0.8402\n", + "val Loss: 0.2205 Acc: 0.9281\n", "\n", "Epoch 9/10\n", "----------\n", - "train Loss: 0.3064 Acc: 0.8730\n", - "val Loss: 0.1748 Acc: 0.9477\n", + "train Loss: 0.4331 Acc: 0.8033\n", + "val Loss: 0.1657 Acc: 0.9542\n", "\n", "Epoch 10/10\n", "----------\n", - "train Loss: 0.2438 Acc: 0.9139\n", - "val Loss: 0.1790 Acc: 0.9477\n", + "train Loss: 0.2844 Acc: 0.8525\n", + "val Loss: 0.1845 Acc: 0.9542\n", "\n", - "Training complete in 8m 28s\n", - "Best val Acc: 0.947712\n", - "Evaluation Loss: 0.1748 Accuracy: 0.9477\n" + "Training complete in 5m 51s\n", + "Best val Acc: 0.954248\n" ] }, { "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA90AAAHqCAYAAAAZLi26AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAADyUUlEQVR4nOzdd1xV5R/A8c/lsjcOhmwnLnArjtTcua3cmqVmaZlZ/crScuWoLDNHy5mmpFZWjsTU3FvBlVsZggrKUDac3x9HbiGogMBhfN+v13294LlnfM8FPXzP8zzfR6coioIQQgghhBBCCCEKnJHWAQghhBBCCCGEEKWVJN1CCCGEEEIIIUQhkaRbCCGEEEIIIYQoJJJ0CyGEEEIIIYQQhUSSbiGEEEIIIYQQopBI0i2EEEIIIYQQQhQSSbqFEEIIIYQQQohCIkm3EEIIIYQQQghRSCTpFkIIIYQQQgghCokk3aLY0Ol0uXrt3Lnzic4zefJkdDpdvvbduXNngcRQ3A0bNgwvL6+Hvn/r1i1MTU3p37//Q7eJi4vD0tKSHj165Pq8y5YtQ6fTcfXq1VzH8l86nY7Jkyfn+nyZrl+/zuTJkzlx4kS2957k9+VJeXl50a1bN03OLYQoXeQeW3zIPfZfWt5jM6WmpuLs7IxOp2PdunWaxiJKL2OtAxAi0/79+7N8P23aNHbs2MH27duztNeqVeuJzjNixAg6d+6cr30bNGjA/v37nziGkq5ixYr06NGDX3/9lTt37uDg4JBtmzVr1pCYmMjw4cOf6FyTJk3ijTfeeKJjPM7169eZMmUKXl5e1KtXL8t7T/L7IoQQxYXcY0sOuccWrT/++IMbN24AsHjxYp577jlN4xGlkyTdotho1qxZlu8rVqyIkZFRtvYHJSQkYGlpmevzuLm54ebmlq8YbW1tHxtPWTF8+HDWr1/PqlWreO2117K9v2TJEpycnOjatesTnadKlSpPtP+TepLfFyGEKC7kHluyyD226CxevBhTU1Nat27N1q1bCQsL0zymnKSnp5OWloaZmZnWoYh8kOHlokRp06YNderUYdeuXTRv3hxLS0teeuklAAICAujYsSMuLi5YWFhQs2ZN3nvvPe7du5flGDkNZcocxrtlyxYaNGiAhYUFPj4+LFmyJMt2OQ19GzZsGNbW1ly8eJFnnnkGa2tr3N3deeutt0hOTs6yf1hYGM899xw2NjbY29szaNAgDh8+jE6nY9myZY+89lu3bjF69Ghq1aqFtbU1jo6OPP300+zevTvLdlevXkWn0/HZZ5/x+eef4+3tjbW1Nf7+/hw4cCDbcZctW0aNGjUwMzOjZs2arFix4pFxZOrUqRNubm4sXbo023tnz57l4MGDDB06FGNjYwIDA+nZsydubm6Ym5tTtWpVRo0aRVRU1GPPk9PQt7i4OEaOHEn58uWxtramc+fOnD9/Ptu+Fy9e5MUXX6RatWpYWlri6upK9+7dOXnypGGbnTt30rhxYwBefPFFwxDLzCF0Of2+ZGRk8Mknn+Dj44OZmRmOjo4MHTqUsLCwLNtl/r4ePnyYVq1aYWlpSeXKlZk1axYZGRmPvfbcSEpKYsKECXh7e2NqaoqrqytjxowhJiYmy3bbt2+nTZs2lC9fHgsLCzw8PHj22WdJSEgwbLNo0SL8/PywtrbGxsYGHx8f3n///QKJUwhR/Mk9Vu6xULbusdevX2fLli10796dd955h4yMjIf+rvz444/4+/tjbW2NtbU19erVY/HixVm22bJlC+3atcPOzg5LS0tq1qzJzJkzs8Tcpk2bbMd+8OeQ+Xv2ySefMH36dLy9vTEzM2PHjh0kJSXx1ltvUa9ePezs7ChXrhz+/v5s2LAh23EzMjL46quvqFevHhYWFtjb29OsWTN+++03QH24U65cuSx/C2R6+umnqV27di4+RZEbknSLEiciIoLBgwczcOBANm3axOjRowG4cOECzzzzDIsXL2bLli2MGzeOn376ie7du+fquEFBQbz11lu8+eabbNiwAV9fX4YPH86uXbseu29qaio9evSgXbt2bNiwgZdeeokvvviC2bNnG7a5d+8ebdu2ZceOHcyePZuffvoJJycn+vXrl6v4bt++DcBHH33Exo0bWbp0KZUrV6ZNmzY5zn9bsGABgYGBzJ07l1WrVnHv3j2eeeYZYmNjDdssW7aMF198kZo1a7J+/XomTpzItGnTsg03zImRkRHDhg3j2LFjBAUFZXkv84+EzD/WLl26hL+/P4sWLWLr1q18+OGHHDx4kJYtW5Kampqr68+kKAq9evXihx9+4K233uKXX36hWbNmdOnSJdu2169fp3z58syaNYstW7awYMECjI2Nadq0KefOnQPU4YyZ8U6cOJH9+/ezf/9+RowY8dAYXn31Vd599106dOjAb7/9xrRp09iyZQvNmzfP9kdOZGQkgwYNYvDgwfz222906dKFCRMmsHLlyjxd96M+i88++4whQ4awceNGxo8fz/Lly3n66acNf5BevXqVrl27YmpqypIlS9iyZQuzZs3CysqKlJQUQB2qOHr0aFq3bs0vv/zCr7/+yptvvpntD2ohROkm91i5x5ale+yyZctIT0/npZdeon379nh6erJkyRIURcmy3YcffsigQYOoVKkSy5Yt45dffuGFF17g2rVrhm0WL17MM888Q0ZGBl9//TW///47Y8eOzfawIC/mzZvH9u3b+eyzz9i8eTM+Pj4kJydz+/Zt3n77bX799VdWr15Ny5Yt6dOnT7aHOsOGDeONN96gcePGBAQEsGbNGnr06GGY1//GG29w584dfvzxxyz7nTlzhh07djBmzJh8xy4eoAhRTL3wwguKlZVVlrbWrVsrgPLXX389ct+MjAwlNTVV+fvvvxVACQoKMrz30UcfKQ/+6nt6eirm5ubKtWvXDG2JiYlKuXLllFGjRhnaduzYoQDKjh07ssQJKD/99FOWYz7zzDNKjRo1DN8vWLBAAZTNmzdn2W7UqFEKoCxduvSR1/SgtLQ0JTU1VWnXrp3Su3dvQ/uVK1cUQKlbt66SlpZmaD906JACKKtXr1YURVHS09OVSpUqKQ0aNFAyMjIM2129elUxMTFRPD09HxvD5cuXFZ1Op4wdO9bQlpqaqjg7OystWrTIcZ/Mn821a9cUQNmwYYPhvaVLlyqAcuXKFUPbCy+8kCWWzZs3K4Dy5ZdfZjnuxx9/rADKRx999NB409LSlJSUFKVatWrKm2++aWg/fPjwQ38GD/6+nD17VgGU0aNHZ9nu4MGDCqC8//77hrbM39eDBw9m2bZWrVpKp06dHhpnJk9PT6Vr164PfX/Lli0KoHzyySdZ2gMCAhRA+fbbbxVFUZR169YpgHLixImHHuu1115T7O3tHxuTEKJ0kHvso8k9tvTfYzMyMpSqVasqrq6uhp9lZjz//Tdw+fJlRa/XK4MGDXroseLj4xVbW1ulZcuWWX7eD2rdurXSunXrbO0P/hwyf8+qVKmipKSkPPI6Mn9Xhw8frtSvX9/QvmvXLgVQPvjgg0fu37p1a6VevXpZ2l599VXF1tZWiY+Pf+S+Ivekp1uUOA4ODjz99NPZ2i9fvszAgQNxdnZGr9djYmJC69atAXUo1uPUq1cPDw8Pw/fm5uZUr149y1PMh9HpdNme9vv6+mbZ9++//8bGxiZbwZABAwY89viZvv76axo0aIC5uTnGxsaYmJjw119/5Xh9Xbt2Ra/XZ4kHMMR07tw5rl+/zsCBA7MM7fL09KR58+a5isfb25u2bduyatUqQ4/p5s2biYyMNDyBB7h58yavvPIK7u7uhrg9PT2B3P1s/mvHjh0ADBo0KEv7wIEDs22blpbGjBkzqFWrFqamphgbG2NqasqFCxfyfN4Hzz9s2LAs7U2aNKFmzZr89ddfWdqdnZ1p0qRJlrYHfzfyK7O35MFYnn/+eaysrAyx1KtXD1NTU15++WWWL1/O5cuXsx2rSZMmxMTEMGDAADZs2JCrYYlCiNJH7rFyj4WycY/9+++/uXjxIi+88ILhZ5k5BP6/Ux8CAwNJT09/ZK/vvn37iIuLY/To0QVajb1Hjx6YmJhka1+7di0tWrTA2tra8DNfvHhxls998+bNAI/trX7jjTc4ceIEe/fuBdTpBT/88AMvvPAC1tbWBXYtZZ0k3aLEcXFxydZ29+5dWrVqxcGDB5k+fTo7d+7k8OHD/PzzzwAkJiY+9rjly5fP1mZmZparfS0tLTE3N8+2b1JSkuH76OhonJycsu2bU1tOPv/8c1599VWaNm3K+vXrOXDgAIcPH6Zz5845xvjg9WQW3sjcNjo6GlBvWA/Kqe1hhg8fTnR0tGF+0NKlS7G2tqZv376AOp+oY8eO/Pzzz/zvf//jr7/+4tChQ4a5b7n5fP8rOjoaY2PjbNeXU8zjx49n0qRJ9OrVi99//52DBw9y+PBh/Pz88nze/54fcv49rFSpkuH9TE/ye5WbWIyNjalYsWKWdp1Oh7OzsyGWKlWqsG3bNhwdHRkzZgxVqlShSpUqfPnll4Z9hgwZwpIlS7h27RrPPvssjo6ONG3alMDAwCeOUwhRcsg9Vu6xZeUemzkfu3fv3sTExBATE4OdnR0tW7Zk/fr1htoot27dAnhkcbXcbJMfOX0OP//8M3379sXV1ZWVK1eyf/9+Dh8+zEsvvZTl38StW7fQ6/WP/X3r2bMnXl5eLFiwAFCH3N+7d0+GlhcwqV4uSpycniBu376d69evs3PnTsOTdyBbMSktlS9fnkOHDmVrj4yMzNX+K1eupE2bNixatChLe3x8fL7jedj5cxsTQJ8+fXBwcGDJkiW0bt2aP/74g6FDhxqejp46dYqgoCCWLVvGCy+8YNjv4sWL+Y47LS2N6OjoLDfbnGJeuXIlQ4cOZcaMGVnao6KisLe3z/f5QZ33+ODN9fr161SoUCFfx81vLGlpady6dStL4q0oCpGRkYbiNQCtWrWiVatWpKenc+TIEb766ivGjRuHk5OTYS3YF198kRdffJF79+6xa9cuPvroI7p168b58+cNvSZCiNJN7rFyjy0L99jY2FjWr18PkOVe+V8//vgjo0ePNtxfw8LCcHd3z3Hb/27zKObm5lnm/Wd62OiynP49rly5Em9vbwICArK8/2BhwYoVK5Kenk5kZGSOyXsmIyMjxowZw/vvv8+cOXNYuHAh7dq1o0aNGo+8FpE30tMtSoXM/3QeXEbhm2++0SKcHLVu3Zr4+HjDcJ9Ma9asydX+Op0u2/UFBwdnW3s1t2rUqIGLiwurV6/OUjDk2rVr7Nu3L9fHMTc3Z+DAgWzdupXZs2eTmpqaZdhbQf9s2rZtC8CqVauytD9YBCTz3A+ed+PGjYSHh2dpe7CH4lEyh10+WKTl8OHDnD17lnbt2j32GAUl81wPxrJ+/Xru3buXYyx6vZ6mTZsanmgfO3Ys2zZWVlZ06dKFDz74gJSUFE6fPl0I0QshSgq5x+ad3GP/VRzvsT/++COJiYmG9eoffFWoUMEwxLxjx47o9fpsD2T+q3nz5tjZ2fH1119nK8L2X15eXpw/fz5LghwdHZ2n3wmdToepqWmWhDsyMjJb9fLM4nePijvTiBEjMDU1ZdCgQZw7dy7HZerEk5GeblEqNG/eHAcHB1555RU++ugjTExMWLVqVbaKn1p64YUX+OKLLxg8eDDTp0+natWqbN68mT///BNQnzQ+Srdu3Zg2bRofffQRrVu35ty5c0ydOhVvb2/S0tLyHI+RkRHTpk1jxIgR9O7dm5EjRxITE8PkyZPzNPQN1OFvCxYs4PPPP8fHxyfLfDUfHx+qVKnCe++9h6IolCtXjt9//z3fw5Y7duzIU089xf/+9z/u3btHo0aN2Lt3Lz/88EO2bbt168ayZcvw8fHB19eXo0eP8umnn2Z7el6lShUsLCxYtWoVNWvWxNramkqVKlGpUqVsx6xRowYvv/wyX331FUZGRnTp0oWrV68yadIk3N3defPNN/N1XQ8TGRnJunXrsrV7eXnRoUMHOnXqxLvvvktcXBwtWrQgODiYjz76iPr16zNkyBBAnae4fft2unbtioeHB0lJSYY/Jtq3bw/AyJEjsbCwoEWLFri4uBAZGcnMmTOxs7N7aC+AEKJskHus3GNL2z128eLFODg48Pbbb2ebugAwdOhQPv/8c4KCgvDz8+P9999n2rRpJCYmMmDAAOzs7Dhz5gxRUVFMmTIFa2tr5syZw4gRI2jfvj0jR47EycmJixcvEhQUxPz58wF1Ktc333zD4MGDGTlyJNHR0XzyySfY2trmOvZu3brx888/M3r0aJ577jlCQ0OZNm0aLi4uXLhwwbBdq1atGDJkCNOnT+fGjRt069YNMzMzjh8/jqWlJa+//rphW3t7e4YOHcqiRYvw9PTM9aoEIg+0rOImxKM8rLJq7dq1c9x+3759ir+/v2JpaalUrFhRGTFihHLs2LFsFTMfVlk1pyrRD1aZfFhl1QfjfNh5QkJClD59+ijW1taKjY2N8uyzzyqbNm3KVmE0J8nJycrbb7+tuLq6Kubm5kqDBg2UX3/99aEVLz/99NNsxyCHyqPff/+9Uq1aNcXU1FSpXr26smTJkmzHzI369evnWElbURTlzJkzSocOHRQbGxvFwcFBef7555WQkJBs8eSmsqqiKEpMTIzy0ksvKfb29oqlpaXSoUMH5Z9//sl2vDt37ijDhw9XHB0dFUtLS6Vly5bK7t27c6weunr1asXHx0cxMTHJcpycfo7p6enK7NmzlerVqysmJiZKhQoVlMGDByuhoaFZtnvY72tuP19PT08FyPH1wgsvKIqiVgB+9913FU9PT8XExERxcXFRXn31VeXOnTuG4+zfv1/p3bu34unpqZiZmSnly5dXWrdurfz222+GbZYvX660bdtWcXJyUkxNTZVKlSopffv2VYKDgx8bpxCi5JF7bFZyj/1Xab/HBgUFKYAybty4h26Teb2vv/66oW3FihVK48aNFXNzc8Xa2lqpX79+torsmzZtUlq3bq1YWVkplpaWSq1atZTZs2dn2Wb58uVKzZo1FXNzc6VWrVpKQEBAnn7PFEVRZs2apXh5eSlmZmZKzZo1le++++6hn+UXX3yh1KlTRzE1NVXs7OwUf39/5ffff892zJ07dyqAMmvWrId+LiL/dIryiDEQQohCN2PGDCZOnEhISEiBF+AQQgghyjK5xwqRO2+99RaLFi0iNDQ0xwJ14snI8HIhilDm8CIfHx9SU1PZvn078+bNY/DgwfLHgBBCCPEE5B4rRN4dOHCA8+fPs3DhQkaNGiUJdyGRpFuIImRpackXX3zB1atXSU5OxsPDg3fffZeJEydqHZoQQghRosk9Voi88/f3x9LSkm7dujF9+nStwym1ZHi5EEIIIYQQQghRSGTJMCGEEEIIIYQQopBI0i2EEEIIIYQQQhQSSbqFEEIIIYQQQohCIoXUcpCRkcH169exsbFBp9NpHY4QQogyRFEU4uPjsbGxwdbWVu5DjyH3bCGEEFrJvGdXqlQJI6NH9GdruUh4cRUaGqoA8pKXvOQlL3lp+oqNjdX6lpjFggULFC8vL8XMzExp0KCBsmvXrkduP3/+fMXHx0cxNzdXqlevrixfvjzbNnfu3FFGjx6tODs7K2ZmZoqPj4+ycePGXMck92x5yUte8pKX1q/Q0NBH3qukpzsHNjY2AISGhmJra6txNEIIIcqSuLg43N3dCQ0NNdyPioOAgADGjRvHwoULadGiBd988w1dunThzJkzeHh4ZNt+0aJFTJgwge+++47GjRtz6NAhRo4ciYODA927dwcgJSWFDh064OjoyLp163Bzc8vzdcs9WwghhFYy79mPu2/JkmE5iIuLw87OjtjYWLmBCyGEKFLF9R7UtGlTGjRowKJFiwxtNWvWpFevXsycOTPb9s2bN6dFixZ8+umnhrZx48Zx5MgR9uzZA8DXX3/Np59+yj///IOJiUm+4iqun5cQQojSL7f3ICmkJoQQQohHSklJ4ejRo3Ts2DFLe8eOHdm3b1+O+yQnJ2Nubp6lzcLCgkOHDpGamgrAb7/9hr+/P2PGjMHJyYk6deowY8YM0tPTC+dChBBCCA1I0i2EEEKIR4qKiiI9PR0nJ6cs7U5OTkRGRua4T6dOnfj+++85evQoiqJw5MgRlixZQmpqKlFRUQBcvnyZdevWkZ6ezqZNm5g4cSJz5szh448/fmgsycnJxMXFZXkJIYQQxZkk3UIIIYTIlQergyuK8tCK4ZMmTaJLly40a9YMExMTevbsybBhwwDQ6/WAWnnc0dGRb7/9loYNG9K/f38++OCDLEPYHzRz5kzs7OwML3d394K5OCGEEKKQSCG1J5Cenm4YIidEQTExMTH8QSqEEMVBhQoV0Ov12Xq1b968ma33O5OFhQVLlizhm2++4caNG7i4uPDtt99iY2NDhQoVAHBxccn2f17NmjWJjIwkJSUFU1PTbMedMGEC48ePN3yfWcRGCCGEKK4k6c4HRVGIjIwkJiZG61BEKWVvb4+zs7OsOSuEKBZMTU1p2LAhgYGB9O7d29AeGBhIz549H7mviYkJbm5uAKxZs4Zu3boZ1jJt0aIFP/74IxkZGYa28+fP4+LikmPCDWBmZoaZmVlBXJYQQghRJCTpzofMhNvR0RFLS0tJjESBURSFhIQEbt68Cai9QEIIURyMHz+eIUOG0KhRI/z9/fn2228JCQnhlVdeAdQe6PDwcFasWAGoyfOhQ4do2rQpd+7c4fPPP+fUqVMsX77ccMxXX32Vr776ijfeeIPXX3+dCxcuMGPGDMaOHavJNQohhBCFQZLuPEpPTzck3OXLl9c6HFEKWVhYAOqwTUdHRxlqLoQoFvr160d0dDRTp04lIiKCOnXqsGnTJjw9PQGIiIggJCTEsH16ejpz5szh3LlzmJiY0LZtW/bt24eXl5dhG3d3d7Zu3cqbb76Jr68vrq6uvPHGG7z77rtFfXlCCCFEoZF1unPwqPXWkpKSuHLlCl5eXobkSIiClpiYyNWrV/H29s625I4QonSTdafzRj4vIYQQWpF1uguZDCkXhUl+v4QQQgghhCgdJOkWQgghhBBCCCEKiSTd4om0adOGcePG5Xr7q1evotPpOHHiRKHFJIQQQgghhBDFhSTdZYROp3vka9iwYfk67s8//8y0adNyvb27u7uhAE9hkuReCCGEEEIIURxI9fIyIiIiwvB1QEAAH374IefOnTO0PVgULjU1FRMTk8cet1y5cnmKQ6/X4+zsnKd9hBBCCCGEEKKkkp7uMsLZ2dnwsrOzQ6fTGb5PSkrC3t6en376iTZt2mBubs7KlSuJjo5mwIABuLm5YWlpSd26dVm9enWW4z44vNzLy4sZM2bw0ksvYWNjg4eHB99++63h/Qd7oHfu3IlOp+Ovv/6iUaNGWFpa0rx58ywPBACmT5+Oo6MjNjY2jBgxgvfee4969erl+/NITk5m7NixODo6Ym5uTsuWLTl8+LDh/Tt37jBo0CAqVqyIhYUF1apVY+nSpQCkpKTw2muv4eLigrm5OV5eXsycOTPfsQghhBBCCCFKL+npLgCKopCYmq7JuS1M9AVW6frdd99lzpw5LF26FDMzM5KSkmjYsCHvvvsutra2bNy4kSFDhlC5cmWaNm360OPMmTOHadOm8f7777Nu3TpeffVVnnrqKXx8fB66zwcffMCcOXOoWLEir7zyCi+99BJ79+4FYNWqVXz88ccsXLiQFi1asGbNGubMmYO3t3e+r/V///sf69evZ/ny5Xh6evLJJ5/QqVMnLl68SLly5Zg0aRJnzpxh8+bNVKhQgYsXL5KYmAjAvHnz+O233/jpp5/w8PAgNDSU0NDQfMcihHi49AyF8DuJeJS31DoUIYQQQjxKTAjEhmsdRe65NQL940f2FgRJugtAYmo6tT78U5Nzn5naCUvTgvkxjhs3jj59+mRpe/vttw1fv/7662zZsoW1a9c+Mul+5plnGD16NKAm8l988QU7d+58ZNL98ccf07p1awDee+89unbtSlJSEubm5nz11VcMHz6cF198EYAPP/yQrVu3cvfu3Xxd571791i0aBHLli2jS5cuAHz33XcEBgayePFi3nnnHUJCQqhfvz6NGjUC1B78TCEhIVSrVo2WLVui0+nw9PTMVxxCiMebueks3++5wvdDG9G+lpPW4QghhBAiJ5Gn4Pt2kJakdSS5985lsCpfJKeSpFsYZCaYmdLT05k1axYBAQGEh4eTnJxMcnIyVlZWjzyOr6+v4evMYew3b97M9T4uLi4A3Lx5Ew8PD86dO2dI4jM1adKE7du35+q6HnTp0iVSU1Np0aKFoc3ExIQmTZpw9uxZAF599VWeffZZjh07RseOHenVqxfNmzcHYNiwYXTo0IEaNWrQuXNnunXrRseOHfMVixDi4e4mp/HjoRAA/vrnpiTdQgghRHGUngYbRqsJt1VFMLfTOqLcMSq6mdaSdBcACxM9Z6Z20uzcBeXBZHrOnDl88cUXzJ07l7p162JlZcW4ceNISUl55HEeLMCm0+nIyMjI9T6Zw+X/u8+DQ+gVRXnk8R4lc9+cjpnZ1qVLF65du8bGjRvZtm0b7dq1Y8yYMXz22Wc0aNCAK1eusHnzZrZt20bfvn1p374969aty3dMQojsfg+6TkKKOnUnOCxG22CEEEIIkbN9X0JEEJjbwyt7wUYekj9ICqkVAJ1Oh6WpsSavgprPnZPdu3fTs2dPBg8ejJ+fH5UrV+bChQuFdr6HqVGjBocOHcrSduTIkXwfr2rVqpiamrJnzx5DW2pqKkeOHKFmzZqGtooVKzJs2DBWrlzJ3LlzsxSEs7W1pV+/fnz33XcEBASwfv16bt++ne+YhBDZrTn8b62Ec5HxJGlUO0MIIYQQD3HzH9g5S/26y2xJuB9CerrFQ1WtWpX169ezb98+HBwc+Pzzz4mMjMySmBaF119/nZEjR9KoUSOaN29OQEAAwcHBVK5c+bH7PlgFHaBWrVq8+uqrvPPOO5QrVw4PDw8++eQTEhISGD58OKDOG2/YsCG1a9cmOTmZP/74w3DdX3zxBS4uLtSrVw8jIyPWrl2Ls7Mz9vb2BXrdQpRlZyPiCAqNwUSvw9xET3xSGmcj4qjv4aB1aEIIIYQAyEiHDWMgPQWqdQTfflpHVGxJ0i0eatKkSVy5coVOnTphaWnJyy+/TK9evYiNjS3SOAYNGsTly5d5++23SUpKom/fvgwbNixb73dO+vfvn63typUrzJo1i4yMDIYMGUJ8fDyNGjXizz//xMFB/YPe1NSUCRMmcPXqVSwsLGjVqhVr1qwBwNramtmzZ3PhwgX0ej2NGzdm06ZNGBXhvBAhSruA+73cHWo5kZCSzs5ztzgZHitJtxBCCFFcHFgI4UfAzBa6zYVCHIFb0umUJ5kcW0rFxcVhZ2dHbGwstra2Wd5LSkriypUreHt7Y25urlGEokOHDjg7O/PDDz9oHUqhkN8zUZYlpabT5ONtxCWlsfylJhy9ept52y/ybAM35vT10zq8Qveoe5DITj4vIYTQQPQlWNRcLZ7W4ytoMFTriDSR23uQ9HSLYi8hIYGvv/6aTp06odfrWb16Ndu2bSMwMFDr0IQQhWDLqUjiktJwtbegZdUKpKapRRVPhsdoG5gQQgghICMDNrymJtyV20L9IVpHVOxJ0i2KPZ1Ox6ZNm5g+fTrJycnUqFGD9evX0759e61DE0IUgjWH1WXC+jZyR2+kw9dNXXrk4s273EtOw8pMbl1CCCGEZg5/ByH7wNQaesyTYeW5IH+5iGLPwsKCbdu2aR2GEKIIXIm6x4HLt9Hp4PlGbgA42prjbGtOZFwSp6/H0cS7nMZRCiGEEGXUnauwbbL6dYcpYO+hZTQlhlR+EkIIUWxkFlBrXb0ilewtDO117/d2y3rdQgghhEYUBX57HVITwKsVNHxJ64hKDEm6hRBCFAup6RmsOxoGQP/GWZ+c+xmS7qJdPUEIIYQQ9x1dCld2gYmlOqxcVu7JNfmkhBBCFAvb/7lJ1N1kKlib0q6mY5b36rrZA9LTLYQQQmgiJhS2fqh+3e5DKFdZ23hKGEm6hRBCFAtrDqkF1J5t6IaJPuvtyddV7em+Gp1AbEJqkccmhBBClFmKAr+/ASnx4N4UmrysdUQljiTdQgghNHc9JpG/z98Csg8tB3CwMsW9nDrH+2S4DDEXQgghisyJVXDpLzA2h54LwEivdUQljiTdQgghNLfuaBgZCjT1Lod3Basct/HNHGIu63WLwnJ0OSTe0ToKIYQoPuIiYMv76tdt34cK1bSNp4SSpFvkSZs2bRg3bpzhey8vL+bOnfvIfXQ6Hb/++usTn7ugjiOEKF4yMhRD1fL+Tdwful3mEPPgUOnpFoXg5Dr4fSx88xSEH9M6GiGE0J6iwB9vQnIsVGoAzcZoHVGJJUl3GdG9e3fat2+f43v79+9Hp9Nx7Fje/8g4fPgwL79csPM6Jk+eTL169bK1R0RE0KVLlwI914OWLVuGvb19oZ5DCJHVnotRhMckYmtuTJc6Lg/dLrOnW4aXi0JRvirYe0JMCCzpBIe+U//gFEKIsurkOji/GYxMoNdC0BtrHVGJJUl3GTF8+HC2b9/OtWvXsr23ZMkS6tWrR4MGDfJ83IoVK2JpaVkQIT6Ws7MzZmZmRXIuIUTRyezl7l3fFXOTh88Tq+Nqi04H4TGJRN1NLqrwRFlRqR6M2gU+3SA9BTa9DeteguR4rSMTQoiid/cmbH5H/br1u+BYU9t4SjhJusuIbt264ejoyLJly7K0JyQkEBAQwPDhw4mOjmbAgAG4ublhaWlJ3bp1Wb169SOP++Dw8gsXLvDUU09hbm5OrVq1CAwMzLbPu+++S/Xq1bG0tKRy5cpMmjSJ1FS1GvGyZcuYMmUKQUFB6HQ6dDqdIeYHh5efPHmSp59+GgsLC8qXL8/LL7/M3bt3De8PGzaMXr168dlnn+Hi4kL58uUZM2aM4Vz5ERISQs+ePbG2tsbW1pa+ffty48YNw/tBQUG0bdsWGxsbbG1tadiwIUeOHAHg2rVrdO/eHQcHB6ysrKhduzabNm3KdyxClAbRd5PZeiYSgH45FFD7LxtzEyrfn+99UtbrFoXBwh76rYSOH4ORMZz+Gb5tA5GntI5MCCGK1sa31BoXzr7QcpzW0ZR4MkagICgKpCZoc24TS9DpHruZsbExQ4cOZdmyZXz44Yfo7u+zdu1aUlJSGDRoEAkJCTRs2JB3330XW1tbNm7cyJAhQ6hcuTJNmzZ97DkyMjLo06cPFSpU4MCBA8TFxWWZ/53JxsaGZcuWUalSJU6ePMnIkSOxsbHhf//7H/369ePUqVNs2bKFbdu2AWBnZ5ftGAkJCXTu3JlmzZpx+PBhbt68yYgRI3jttdeyPFjYsWMHLi4u7Nixg4sXL9KvXz/q1avHyJEjH3s9D1IUhV69emFlZcXff/9NWloao0ePpl+/fuzcuROAQYMGUb9+fRYtWoRer+fEiROYmJgAMGbMGFJSUti1axdWVlacOXMGa2vrPMchRGny87FwUtMV/NzsqFXJ9rHb+7rZc+nWPYLCYmjr4/jY7YXIM50Omr8Gbo1h3YsQfRG+bwdd50D9wVpHJ4QQhe/0r3D2N/XhY88FoDfROqIST5LugpCaADMqaXPu96+Dac6Vfh/00ksv8emnn7Jz507atm0LqEPL+/Tpg4ODAw4ODrz99tuG7V9//XW2bNnC2rVrc5V0b9u2jbNnz3L16lXc3NwAmDFjRrZ52BMnTjR87eXlxVtvvUVAQAD/+9//sLCwwNraGmNjY5ydnR96rlWrVpGYmMiKFSuwslKvf/78+XTv3p3Zs2fj5OQEgIODA/Pnz0ev1+Pj40PXrl3566+/8pV0b9u2jeDgYK5cuYK7u1rs6YcffqB27docPnyYxo0bExISwjvvvIOPjw8A1ar9W+ExJCSEZ599lrp16wJQuXLlPMcgRGmiKAprDqtrcz+ulzuTr5sdvxwPl55uUfg8msKo3fDLy3BxG2wYA9f2wTOfgWnRTKsSQogidy9a7eUGaDkeXHy1jaeUkOHlZYiPjw/NmzdnyZIlAFy6dIndu3fz0ksvAZCens7HH3+Mr68v5cuXx9ramq1btxISEpKr4589exYPDw9Dwg3g7++fbbt169bRsmVLnJ2dsba2ZtKkSbk+x3/P5efnZ0i4AVq0aEFGRgbnzp0ztNWuXRu9/t85oi4uLty8eTNP5/rvOd3d3Q0JN0CtWrWwt7fn7NmzAIwfP54RI0bQvn17Zs2axaVLlwzbjh07lunTp9OiRQs++ugjgoOD8xWHEKXFkWt3uHTrHhYmerr7PbyA2n/5uqkjX4LCYlGkyJUobFblYeBaeHoi6IzUtWq/bwe3zmsdmRBCFI7N/4OEKHCsBU+9o3U0pYb0dBcEE0u1x1mrc+fB8OHDee2111iwYAFLly7F09OTdu3aATBnzhy++OIL5s6dS926dbGysmLcuHGkpKTk6tg5/QGse2Do+4EDB+jfvz9TpkyhU6dO2NnZsWbNGubMmZOn61AUJduxczpn5tDu/76XkZGRp3M97pz/bZ88eTIDBw5k48aNbN68mY8++og1a9bQu3dvRowYQadOndi4cSNbt25l5syZzJkzh9dffz1f8QhR0q05pBZQ6+7ngo157oau1XKxQ2+kI+puMpFxSbjYWRRmiEKAkZH6h6d7U1g3HG6ege/aQvcvoe5zWkcnhBAF55+NcGqd+pCx5wIwNtU6olJDeroLgk6nDvHW4pWL+dz/1bdvX/R6PT/++CPLly/nxRdfNCSMu3fvpmfPngwePBg/Pz8qV67MhQsXcn3sWrVqERISwvXr/z6A2L9/f5Zt9u7di6enJx988AGNGjWiWrVq2Sqqm5qakp6e/thznThxgnv37mU5tpGREdWrV891zHmReX2hoaGGtjNnzhAbG0vNmv9WdKxevTpvvvkmW7dupU+fPixdutTwnru7O6+88go///wzb731Ft99912hxCpEcReXlMrGk+r/FbkdWg5gYaqnmqNaCyFI1usWRcn7KXhlD3i1gpS7sH44/DEe0qSSvhCiFEi8o/6fBtB8LLjmfVUj8XCSdJcx1tbW9OvXj/fff5/r168zbNgww3tVq1YlMDCQffv2cfbsWUaNGkVkZGSuj92+fXtq1KjB0KFDCQoKYvfu3XzwwQdZtqlatSohISGsWbOGS5cuMW/ePH755Zcs23h5eXHlyhVOnDhBVFQUycnZ/6AZNGgQ5ubmvPDCC5w6dYodO3bw+uuvM2TIEMN87vxKT0/nxIkTWV5nzpyhffv2+Pr6MmjQII4dO8ahQ4cYOnQorVu3plGjRiQmJvLaa6+xc+dOrl27xt69ezl8+LAhIR83bhx//vknV65c4dixY2zfvj1Lsi5EWbLhxHWSUjOo5mhNAw/7PO3rZ1ivO6bA4xLikWycYMiv0Op+/ZMji2FxR7h9RdOwhBDiiW15H+5GQoXq0GaC1tGUOpJ0l0HDhw/nzp07tG/fHg+Pf3uYJk2aRIMGDejUqRNt2rTB2dmZXr165fq4RkZG/PLLLyQnJ9OkSRNGjBjBxx9/nGWbnj178uabb/Laa69Rr1499u3bx6RJk7Js8+yzz9K5c2fatm1LxYoVc1y2zNLSkj///JPbt2/TuHFjnnvuOdq1a8f8+fPz9mHk4O7du9SvXz/L65lnnjEsWebg4MBTTz1F+/btqVy5MgEBAQDo9Xqio6MZOnQo1atXp2/fvnTp0oUpU6YAajI/ZswYatasSefOnalRowYLFy584niFKIkC7hdQ69/E46FTRR6m7v153cFSTE1oQW8M7SbBoHVgUQ4iTsA3reHsH1pHJoQQ+XMhEIJ+BHTqsHITc60jKnV0ilSiySYuLg47OztiY2Oxtc26hE1SUhJXrlzB29sbc3P5hRSFQ37PRGl2KjyWbl/twVRvxIH321HOKm9zxk6GxdJ9/h7sLEw48WGHPCftxd2j7kEiO00/r9gwWPsihB1Sv/d/DdpPluV1hBAlR1IsLPSHuHBoNgY6z9A6ohIlt/cg6ekWQghRpDKXCetY2ynPCTdADWcbTPVGxCamEnI7oaDDEyL37NzgxU1qsg2wfz4sfUZNxoUQoiQI/FBNuB281ZUaRKGQpFsIIUSRSUxJZ8NxtYDagCa5L6D2X6bGRtR0sQFkiLkoBvQm0Olj6LcSzOzUXu+vW8GFbVpHJoQQj3Z5Jxxdpn7dcz6Y5m1VJJF7knQLIYQoMhtPRhCfnIZ7OQv8K5fP93H+ndcdU0CRCfGEanaHUTvBxQ8Sb8Oq52D7dMh49GocQgihieS78Nv9ZWsbjwSvltrGU8rJOt1CCCGKjKGAWmMPjIzyPxfb180eCJGe7iK2cOFCPv30UyIiIqhduzZz586lVatWD91+wYIFzJ8/n6tXr+Lh4cEHH3zA0KFDc9x2zZo1DBgwgJ49e/Lrr78W0hUUsnKV4aWt8Of7amXzXZ9CyAF4drFa+bysiQmFmGuP3644sCwPFX3yvBSryEFqEkQEQUaq1pE8nt4UKtUvm3UY/poCMSFg56HWohCFSpJuIYQQReLizbscvnoHIx0819DtiY7le7+n+1R4LOkZCvonSOBF7gQEBDBu3DgWLlxIixYt+Oabb+jSpQtnzpzJshJGpkWLFjFhwgS+++47GjduzKFDhxg5ciQODg507949y7bXrl3j7bfffmQCX2KYmEO3z8HDH35/A67uhm9aqYm3dym4vsdJSYCzv8Hxleq1lySOtaH+IPDtB1YVtI6mZFEUtZL/8ZVwch0kxWgdUe45+0Lf5epDs7Li6l449K36dY95YGatbTxlgCTd+ZSRkaF1CKIUk98vURpl9nI/7eOIk+2TVeWvWtEaCxM991LSuRJ1l6qONgURoniEzz//nOHDhzNixAgA5s6dy59//smiRYuYOXNmtu1/+OEHRo0aRb9+/QCoXLkyBw4cYPbs2VmS7vT0dAYNGsSUKVPYvXs3MTExRXI9hc73eXWo+U9D4dZZWNED2n4ALceDUSmb3acoEHoITqyEU79ASvz9N3RQvgro9JqG93gK3LkGN0+roxQCP4TqnaH+YKjaQV0mTuTsXhQEB8DxVernl8mqorqkXnEXHwGRweqyfz0XQK0eWkdU+FISYMMY9esGL0CVttrGU0bI/yJ5ZGpqipGREdevX6dixYqYmpqWuuVqhHYURSElJYVbt25hZGSEqWneKzsLURylpGWw/lg4oA4tf1LGeiNqV7LlyLU7BIXGStJdyFJSUjh69CjvvfdelvaOHTuyb9++HPdJTk7OtuShhYUFhw4dIjU1FRMTdTjn1KlTqVixIsOHD2f37hLWM/o4FavDyO2w8S11Ddzt09Th5n2+BcsSkJA8TlwEBK9RE67oC/+2O3hBvUHgNwDs3TULL08S76g9tCdWwfXj8M8f6svKEfz6Qb3B4OijdZTFQ3qquq7ziVVwfgtkpKntejOo2U392VduA0bF/WELEBsO616C0APw0xBoNhraTwHjUvz3146P4c4VsHWFjtO0jqbMkKQ7j4yMjPD29iYiIoLr169rHY4opSwtLfHw8MCotPWGiDJr29kb3L6XgqONGW1qVCyQY/q62XPk2h1Ohsfy7BMOVxePFhUVRXp6Ok5OWeclOzk5ERkZmeM+nTp14vvvv6dXr140aNCAo0ePsmTJElJTU4mKisLFxYW9e/eyePFiTpw4ketYkpOTSU5ONnwfFxeXr2sqMqaW0HsReDaHTW/DxUC1uvnzS8G9idbR5V1aMpzbrCZcF7eBcn9klokl1Oqp9g57NC95vfkWDtBkpPq6cUa9vqA1cO8m7PtKfbk2Uoef13kWzO20jrjo3fxHHc0QFKB+LpkqNfj3c7Fw0C6+/LBzhWF/qPOb930FBxZC2GF4bmnJeWCUF6GHYP8C9evuX5bN32ONSNKdD6ampnh4eJCWlkZ6ulQlFQVLr9djbGwsIyhEqbL6kDq0/PlGbhjrC+aP8cx53UFSwbzIPPj/kqIoD/2/atKkSURGRtKsWTMURcHJyYlhw4bxySefoNfriY+PZ/DgwXz33XdUqJD7+bMzZ85kypQpT3QdmmgwRC3Y9NNQuH0JlnaBDtOg2aslo3hXRJDao33yJ7VXOJN7MzXhqt0bzErJiBOnWuoycO0nw/k/7/fo/gnhR9TXlglqtfp6g8C7dcl7wJAXiTFwar36GYQf/bfdqqI6973eIPXzKsn0JtBxulqH4ddX1aT7m1bQ+1uo3lHr6ApOatL9YeUK+A2Eah20jqhM0SmKomgdRHETFxeHnZ0dsbGx2Nraah2OEEKUaKG3E3jq0x0oCux6py0e5QtmHdDLt+7y9Jy/MTM24tSUTpgUUDKvteJ4D0pJScHS0pK1a9fSu3dvQ/sbb7zBiRMn+Pvvvx+6b2pqKjdu3MDFxYVvv/2Wd999l5iYGIKDg6lfvz56/b9DUDPrWRgZGXHu3DmqVKmS7Xg59XS7u7sXq8/rkZLi4PexcPoX9XufbupcUgt7TcPK0b1oOLlWLY514+S/7TYu6tDxeoOgQlXt4itKd2/+O3f51tl/2+08oN4AqDdQHVZfGmRkwJW/1UT77O+QlqS2GxlDtU7qQ5ZqHUtnxe87V2HtMHWKAag1GNp+UDrm9W+bDHu+AGsnGHOw5I1KKKZye88uBb9BQgghirO1R8NQFGhRtXyBJdwAXuWtsDEzJj45jfM34qldSYbJFRZTU1MaNmxIYGBglqQ7MDCQnj17PnJfExMT3NzU4f9r1qyhW7duGBkZ4ePjw8mTJ7NsO3HiROLj4/nyyy9xd895aKeZmRlmZmZPeEUaMrdVh656tlB7TP/5A26cgueXQ6V6WkcH6Wlw6S810T63+d9ln/SmUOMZqD9ELbxUEubrFiRrR2j+Ovi/BteP3e/1XwexIfD3bPXl1Up9EFGrB5haaR1x3t2+Aid+hKDVEBv6b3vFmuq0Ad++6udQmjl4wUt/wp8fwOHvYM/n6pDs5xaDjbPW0eVf+DHYO0/9utsXknBrQJJuIYQQhSY9Q2HtEfWPt34FUEDtv4yMdNR1s2PfpWhOhsVK0l3Ixo8fz5AhQ2jUqBH+/v58++23hISE8MorrwAwYcIEwsPDWbFiBQDnz5/n0KFDNG3alDt37vD5559z6tQpli9fDoC5uTl16tTJcg57e3uAbO2ljk6nzh12baD2qt25Cos7QpdZ0PBFbYabR11QE+2gNXD3P/P0XfzUImJ1nysdxd+elE4Hrg3VV6eP4Z+N6ud2eae6RNrV3bDpHajTW/3c3JsU7+kDKffgzP0l3q7t+bfd3A7qPKf2aldqULyvoaAZm0HXz8DTH34bq34uX7dSE2/vp7SOLu/SUtRh5Uq6+jP16ap1RGWSJN1CCCEKza7zt4iITcLe0oROtZ0ev0MeZSbdQWGx9C+BNalKkn79+hEdHc3UqVOJiIigTp06bNq0CU9PTwAiIiIICQkxbJ+ens6cOXM4d+4cJiYmtG3bln379uHl5aXRFRRDrg1h1C74dTSc2wR/vAnX9kG3uUWzbm5SHJz+We21DTv0b7tl+X/n6zqX8gcgT8LEQn0YUfc5iAlVH1icWKVWhj62Qn2Vr6YOPfcbALYuWkesUhQIPagm2qd/zbrEW5W26s/dp5u65nxZVudZdQ3vn15Ql0Nb0RPavA+t3ipZ8/h3fwY3z4BlBejyidbRlFkypzsHxXE+nRBClESjfjjCn6dv8FILbz7sXvDFdjadjGD0qmPUcbXlj9dbFfjxtSD3oLwpFZ+XoqiVk7dNVnujKlSHvivAsWbBnysjQ+2NPbFK7eFMS1TbdXq1sFK9Qeoa1aV5yaTCpCjqg5MTq9R5+6kJarvOCKq0U4dp1+ii9qYWtbjr6tDxEz9C9MV/2x287y/x1r90Vux+UikJsPkd9SEFqD/HPt+CVe4LQGomIhi+a6su6/b8MrXgoShQub0Haf6YZuHChXh7e2Nubk7Dhg1zvUbn3r17MTY2pl69etneW79+PbVq1cLMzIxatWrxyy+/FHDUQgghHudmfBJ/nVWXlenfpHD+kKvrqg4p/ycinqRUWU1ClFA6HbQYCy9uAptKEHUevm0LJ1YX3DnuXIOds2CeH6zooRYFS0uECjWgw1QYfwYGBqjzkSXhzj+dDrxaQK+F8PZ56DFfrYqtZKjLxa19AebUgE3/UyvCF7a0ZDX5X/kcfFEb/pqqJtwmVmqiPWwTjD0Ord+RhPthTC3VYoc9F4KxhVrz4OtWEHJQ68geLT1VHVaekQY1e0jCrTFNh5cHBAQwbtw4Fi5cSIsWLfjmm2/o0qULZ86cwcPj4XP/YmNjGTp0KO3atePGjRtZ3tu/fz/9+vVj2rRp9O7dm19++YW+ffuyZ88emjZtWtiXJIQQ4r71R8NJy1Bo4GFPdafCWUrIzcGCclam3L6Xwj+R8dRzty+U8whRJDyawSu74eeRcGk7/PoKXNsLz3yqDmXOq5QEtfr0iZVwZde/7Wa2UKePWhTNtWHZmq9blMxs1KXiGgyB6Etq7/eJ1RB/HQ59o76c6qrzpuv2BavyBXfuiCC1Z/bk2qxLvHn4q73ttXqWniXeikr9QWqxw59egOgLsOwZdVk5/9eK57+hvXMhMlgtmtZ1jtbRlHmaDi9v2rQpDRo0YNGiRYa2mjVr0qtXL2bOnPnQ/fr370+1atXQ6/X8+uuvnDhxwvBev379iIuLY/PmzYa2zp074+DgwOrVuXtiXCqGqgkhhIYURaHtZzu5Gp3AJ8/60rdx4fWgvLDkEH+fv8XUnrUZ6u9VaOcpKnIPyptS+XllpMPuObBjBqCAUx11uHn57EuoZaMoEHYEjv+g9nAmx/37nndrNeHy6ab23omil5EOl3aoD0L+2QjpKWq7kYk67Lz+YHX4cn6WqLoXra6jfnzVA0u8Vbq/rNmg3P0OiUdLjoff31DXLweo0RV6LSheFcFvnoVvnlJ/v/p8p1aeF4Wi2C8ZlpKSwtGjR3nvvfeytHfs2JF9+/Y9dL+lS5dy6dIlVq5cyfTp07O9v3//ft58880sbZ06dWLu3LkFErcQQojHO3D5NlejE7Ay1dPVt3CLB/m52fH3+VsEh8UW6nmEKDJGemj9P7Xy9foR6pJi37SGnl89fIhofOS/hbyizv/bbu+pJlv1BoB9wa4gIPLBSA/V2quvhNtq4nZ8JUScgLO/qS9rZ/Drp1Y/r1j90cdLT4OL29Qk/tyW/yzxZqZWqa4/CCqXwSXeCpOZDTy7GDybq8v+ndsI35yCvsuhUn2to1N/J34drSbc1btA3ee1jkigYdIdFRVFeno6Tk5Zq9k6OTkRGRmZ4z4XLlzgvffeY/fu3Rgb5xx6ZGRkno4JkJycTHJysuH7uLi4h24rhBDi8QIOq1Wse9RzxcqscG81dd3sAQgOiynU8whR5Cq3gVG7Yf1wdZj52mFqka6O09VCXGkpcH6z2rN5cZtahA3Ueae1eqq9pp4tSlal5bLEspy6dFyTkXDjtPpzDL6/ZNveL9WXW2P151i7j7rGe6Zb59VEO2gN3P3PVEuXeur2dZ6VJd4Kk04HjUeo0zN+egFirqnL/nWaobZrOdz8wAJ1LXkzO+j2efEc+l4Gab5kmO6BXwRFUbK1gbr0yMCBA5kyZQrVqz/6qV9uj5lp5syZTJkyJQ9RCyGEeJjYhFQ2nVIfdPYvxGHlmXzd1GJqF2/e5V5yWqEn+UIUKVsXGPob7PgY9nwOh75Vh4+7N1WHEidE/7ute1O1V7t276wJmij+nGpD5xnqHOELf6oJ+IWtEHZYfW1+Ty1yV6m+2jsedvjffS0rqEu81R+kHkcUnUr1YdTf8OsYtcd709sQsh+6f6nNnPmoC7D9Y/XrzjPAtlLRxyBypNlfJhUqVECv12frgb5582a2nmqA+Ph4jhw5wvHjx3nttdcAyMjIQFEUjI2N2bp1K08//TTOzs65PmamCRMmMH78eMP3cXFxuLtLBUchhMiPX46HkZKWgY+zjSEhLkxOtuY42ZpxIy6Z09fjaOItvTuilNEbQ/uP1EJrP7+s9mJdP6a+Z+OiLvVUbxBUqKZtnOLJGZtCze7qK/6GWmX+xCq49Y/6dXCAup1OD9U7qT/3ah2l4ryWLByg/yrYvwC2faQ+FIkIUuswFOVDkIx0tVp5erJaF6DeoKI7t3gszZJuU1NTGjZsSGBgIL17/zs/KTAwkJ49e2bb3tbWlpMnT2ZpW7hwIdu3b2fdunV4e3sD4O/vT2BgYJZ53Vu3bqV58+YPjcXMzAwzMw3WSxRCiFJGURTWHA4FYEATj0eOMipIvm72BJ65QXBYjCTdovSq3kmtbr75XXV4ud9AqPJ0/opuieLPxkldSq756xB+TB1OHnVBTbJ9+6nvi+JBp4Pmr6nTAda9qC7L9l07tWp4/SJKfg99C6EHwdRG7WmXYeXFiqb/S48fP54hQ4bQqFEj/P39+fbbbwkJCeGVV14B1B7o8PBwVqxYgZGREXXq1Mmyv6OjI+bm5lna33jjDZ566ilmz55Nz5492bBhA9u2bWPPnj1Fem1CCFEWBYfF8k9kPKbGRvSq51pk5/V1tbufdEsxNVHK2XvAgAJcv1sUfzoduDVUX6J482iq1mH4eaS6nveG0Wodhmc+LdwVA25fhm33p8p2nCprrhdDmibd/fr1Izo6mqlTpxIREUGdOnXYtGkTnp6eAERERBASEpKnYzZv3pw1a9YwceJEJk2aRJUqVQgICJA1uoUQogisuV9A7Zk6zthZmhTZeX3vr899MlySbiGEEBqyKg+D1qnL/u2coY5QuH5crW5eGFNAMjLgt7GQlgjeT0HDFwv+HOKJabpOd3FVKtf8FEKIQnYvOY0mH2/jXko6a15uRrPK5Yvs3LfvpdBgWiAAQR91xM6i6BL+gib3oLyRz0sIUWxd/ltd9u/eTTC1hh7z1MryBenw97DxLTCxhFf3QTnvgj2+eKTc3oNkDQkhhBAF4o/g69xLScervCVNi3hedTkrU9zLWQBwSnq7hRBCFAeVW6t1GDxbQspdWPeSmiCnJT9+39yICYHAj9Sv20+WhLsYk6RbCCFEgcgsoNavcdEVUPsvX1d7AIJkvW4hhBDFhY0zDN0Ard5Svz/8vbqm952rT3ZcRVGHlafcBQ9/aDzyiUMVhUeSbiGEEE/sXGQ8x0NiMDbS8WzDoiug9l+Zy5OdlGJqQgghihO9MbT7UJ3rbeEAESfg66fgn435P+bxH+DyDjA2h54LwEjSuuJMfjpCCCGeWGYBtXY1HXG0Mdckhrr3k26pYC6EEKJYqtZBrW7u1hiSY2HNQPjzA0hPzdtx4q6r+wE8PRHKVyn4WEWBkqRbCCHEE0lKTeeX4+EA9G/soVkcdV3VpDs8JpGouwU0X04IIYQoSPbuMGwTNBujfr9/PizrCrHhudtfUeD3cZAcB66NoNnoQgtVFBxJuoUQooCE3k5g6u9nmPTrKZJS07UOp8hsPXODmIRUXOzMeap6Rc3isDE3oXJFK0CGmAshhCjGjE2h8wzotxLMbCH0IHzTCi5ue/y+wQFw4U/Qm94fVq4v/HjFE9N0nW4hhCgNTl+P5Zu/L7PxZATpGeoqjHcSUpjXvz5GRkVfUKyorTmkDi1/vpE7eo2v18/Nnsu37hEcFktbH0dNYxFCCCEeqWZ3cKoNP70AkcGw8jl46m1oMyHnZDr+Bmx+V/26zXvg6FO08Yp8k55uIYTIB0VR2HcpiqFLDtF13h5+C7pOeoZCs8rlMDbS8UdwBJ8Hntc6zEJ3Lfoe+y5Fo9NB30ZuWodjGGIeLBXMhRBClATlKsPwQGj0EqDArk9hRU81wf4vRYGN4yEpBlz8oPlYLaIV+SQ93UIIkQfpGQpbT0fy9d+XCLo/hNlIB119KzHqqcrUcbXjpyOh/G9dMPN3XMSjvCV9G7lrHHXh+emIukxYq2oVcXOw1Dga8HO/n3SHx6IoiiZLlwkhhBB5YmIO3b4Aj+bw+xtwdbc63Py5JeDVUt3m9C/wzx9gZAI9F4LeRNuYRZ5I0i2EELmQnJbOz8fC+W7XZS5H3QPAzNiIvo3cGdmqMh7l/004+zZyJyQ6gfk7LvL+zydxs7egedUKWoVeaNLSM1h7JAyA/o2Lx4OFWi526I103IpPJjIuCRc7C61DEkIIIXLH93lw8VWHm986C8u7q9XJ6w+FTW+r2zz1NjjX0TZOkWeSdAshxCPEJaWy6kAIS/Ze4Va8WhHbzsKEof6evNDciwrWZjnuN75Dda7dTuD3oOuMWnmUX0Y3p6qjTVGGXuh2nLvFzfhkyluZ0r6mk9bhAGBhqqeaozX/RMYTHBYrSbcQQoiSpWINGPkXbHwLglbDX1Nh31eQeAec6kDL8VpHKPJBkm4hhMjBzbgkluy9yqoD14hPTgPAxc6c4S296d/EA2uzR//3aWSk49PnfLkek8jRa3cYtvQwv4xuQUWbnJP0kijg/trczzZ0w9S4+JQI8XWzu590x9CptrPW4QghhBB5Y2oFvRaBZ3PY9I6acOv0arVyY1OtoxP5IEm3EEL8x+Vbd/lu92XWHw0nJT0DgGqO1oxqXYUefpXylFyam+j5dkhD+izax7XoBEauOMKal5thblLyl/eIjE1i+z83AYrdnHVfN3t+OhJGsCwbJoQQoqTS6aDBUKhUH/6aBjW7QaV6Wkcl8kmSbiGEAIJCY/j670tsOR2Joq76RUNPB15pXYV2Po75XvqrvLUZS4Y1ps/CfZwIjWH8TyeYP6BBiV9KbN3RUDIUaOJVjqqO1lqHk4Wvm1pM7aQUUxNCCFHSOdeFQT9pHYV4QpJ0CyHKLEVR2HUhiq93XmL/5WhDezsfR15pU4XGXuUK5DxVKlrzzZCGDFl8kE0nI/mk3Dne61Jy19bMyFAIuF+1vF8xKaD2XzWcbTDVGxGTkEro7cQsRe6EEEIIIYqaJN1CiDInLT2DjScj+Obvy5yJiAPA2EhHj3qVGPVUFWo4F3zBs2aVyzP7WV/G/xTE139fwrO8JQOaeBT4eYrCvkvRhN5OxMbcmGfqumgdTjZmxnp8XGwIDoslKCxGkm4hhBBCaEqSbiFEmZGYks7ao6F8t/syobcTAbA01dO/sQfDW3njal+4la77NHDjWnQCX/51gYm/nsLNwYJW1SoW6jkLw5r7BdR61XPFwrR4zk/3dbMjOCyWk+GxdPerpHU4QgghhCjDJOkWQpR6MQkp/LD/Gsv2XSX6XgoA5axMGdbci6H+nthbFl0l0HHtqxFyO4FfjoczeuUx1r3avFB61gvL7XspbD19AyieQ8sz+braAyEEhcZoHIkQQgghyjpJuoUQpdb1mEQW77nC6kMhJKSkA+DmYMHLT1Xm+YbumvTS6nQ6Zj1bl/CYRA5duc1Lyw7zy5jmONqYF3ks+fHzsTBS0jOo42pLHVc7rcN5KF93NbZT4bFkZCglvnCdEEIIIUouSbqFEKXOhRvxfP33ZTacCCctQy1FXtPFlldaV6ZrXReM9dquKW1mfH8psYX7uBx1jxHLjxDwsn+xHaqdSVEUAg6rBdT6Ny7e89GrVrTG3MSIeynpXI66S1XHkjOaQAghhBCliyTdQohS48jV23z99yW2nb1paGtWuRyvtK5C6+oVi9XSUfaWpiwZ1pjeC/cSHBbLuIDjLBzUEH0x7pE9FhLDhZt3MTcxoke94j1P2lhvRJ1Kdhy5dofgsFhJuoUQQgihGW27e4QQ4gllZCj8dfYGz3+9j+e+3s+2szfR6aBzbWd+HdOCNS/706aGY7FKuDN5VbDiu6GNMNUb8efpG8zcdFbrkB5pzSG1gFrXupWwNTfROJrHq3t/ve7gsFiNIxFCCCFEWSY93YVs94VbLN93lYWDGmJqLM84hCgoqekZ/HbiOt/susT5G3cBMNUb0aeBKyOfqkyVitYaR5g7jbzK8enzvryx5gTf77mCZ3lLhvh7aR1WNvFJqfwRHAHAgCbFt4Daf/m52QMQHBajaRxCCCGEKNsk6S5Ed5PTGLv6OHcSUvn0z3/4oGstrUMSosS7l5zGmsOhLN59meuxSQDYmBkzsJkHL7Xwxsm2ZBQk+6+e9VwJvZ3AZ1vP89Fvp3FzsKStj6PWYWXxW9B1ElPTqVLRioaeDlqHkyuZPd2nr8eRmp6BicZz+YUQQghRNslfIIXI2syY2c/6AvDd7iv8dfaGxhEJUXJF303m863naD5rO9P+OMP12CQq2pjxbmcf9k54mgldapbIhDvTmLZVea6hGxkKvPbjMc5cj9M6pCz+W0CtOA7Vz4l3eStszIxJTsvgwv3REEIIIYQQRU2S7kLWsbYzL7bwAuCttUFcj0nUNiAhSqBzkfG0//xv5m2/SGxiKt4VrJjZpy67/9eWV9tUKRHzix9Hp9Mxo3dd/CuX515KOi8tO0zk/Z58rZ2+HktwWCwmeh19GrhqHU6uGRnpDMuayRBzIYQQQmhFku4iMKFLTXzd7IhJSGXs6uOkpmdoHZIQJUZIdAKDFx/kTkIq1Z2sWTSoAdvGt2ZAEw/MTYr3Elt5ZWpsxNeDG1KlohWRcUkMX36Ye8lpWodl6OXuWMuZ8tZmGkeTN5nrdQeHSzE1IYQQQmhDku4iYGpsxFcD6mNjZsyRa3f4IvC81iEJUSLciEti0OID3IpPxsfZhrWjmtOlrkuxXlbrSdlZmrB0WBPKW5ly+nocY1cfJ/3+WuNaSEpN55fj4QD0a1wyCqj9l6+rPSA93QVl4cKFeHt7Y25uTsOGDdm9e/cjt1+wYAE1a9bEwsKCGjVqsGLFiizvf/fdd7Rq1QoHBwccHBxo3749hw4dKsxLEEIIIYqcJN1FxLO8FbPuz+9euPMSf5+/pXFEQhRvd+6lMGTxQUJvJ+JZ3pIVw5tgZ1nyh5Hnhkd5S757oRFmxkb89c9Npv1xRrNYNp+KID4pDVd7C1pWraBZHPnle7+Y2rnIeJJS0zWOpmQLCAhg3LhxfPDBBxw/fpxWrVrRpUsXQkJCctx+0aJFTJgwgcmTJ3P69GmmTJnCmDFj+P333w3b7Ny5kwEDBrBjxw7279+Ph4cHHTt2JDw8vKguSwghhCh0knQXoa6+Lgxu5gHA+IAT3IgrHvM1hShu7ianMWzZYc7fuIuTrRkrhzfF0abkFknLjwYeDnzetx4Ay/ZdZeneK5rEsfqQOrS8X2N3jErgCAM3BwscLE1ITVf4JzJe63BKtM8//5zhw4czYsQIatasydy5c3F3d2fRokU5bv/DDz8watQo+vXrR+XKlenfvz/Dhw9n9uzZhm1WrVrF6NGjqVevHj4+Pnz33XdkZGTw119/FdVlCSGEEIVOku4iNrFrLWq62BJ9L4U31mg7bFSI4igpNZ2XVxwhKDQGB0sTVg5vins5S63D0kRXXxfe7ewDwLQ/zrDtTNGugHD51l0OXbmNkQ6eb+RWpOcuKDqdDt/763WflCHm+ZaSksLRo0fp2LFjlvaOHTuyb9++HPdJTk7G3DzrwzILCwsOHTpEampqjvskJCSQmppKuXLlCiZwIYQQohiQpLuImZvoWTCwPpameg5cvs28vy5oHZIQxUZaegZjVx9n36VorEz1LHuxCdWcbLQOS1OvtK5M/8buZCjw+urjnCrCgmABR9Re7jY1HHGxsyiy8xa0zCHmQWFSTC2/oqKiSE9Px8nJKUu7k5MTkZGROe7TqVMnvv/+e44ePYqiKBw5coQlS5aQmppKVFRUjvu89957uLq60r59+4fGkpycTFxcXJaXEEIIUZxJ0q2ByhWtmdG7LgDztl9g38Wc//gQoizJyFD43/pgtp65gamxEd+/0Bg/d3utw9KcTqdjWq86tKxagcRUdSmxolh6MCUtg/VHw4CSWUDtv/7t6Zak+0k9uEa7oigPXbd90qRJdOnShWbNmmFiYkLPnj0ZNmwYAHp99pUHPvnkE1avXs3PP/+crYf8v2bOnImdnZ3h5e5esn8/hRBClH6SdGukV31X+jZyQ1HgjYATRN1N1jokITSjKApT/zjDz8fC0RvpWDiwAf5VymsdVrFhojdi4eAGVHO05mZ8Mi8tO8zdQl5KbPs/N4i6m0IFazOe9nEs1HMVtsye7gs340lI0X4JtpKoQoUK6PX6bL3aN2/ezNb7ncnCwoIlS5aQkJDA1atXCQkJwcvLCxsbGypUyFqU77PPPmPGjBls3boVX1/fR8YyYcIEYmNjDa/Q0NAnuzghhBCikEnSraEpPepQzdGaW/HJvBlwggyZ3y3KqLnbLrBs31UAPnvel/a1cv4jviyzNTdhybDGVLA245/IeF778Rhp6RmFdr4199fmfr6RGyb6kn2rcLI1x8nWjAwFTl+Xocj5YWpqSsOGDQkMDMzSHhgYSPPmzR+5r4mJCW5ubuj1etasWUO3bt0wMvr3d+rTTz9l2rRpbNmyhUaNGj02FjMzM2xtbbO8hBBCiOKsZP8lVcJZmOpZMKgB5iZG7L4QxaK/L2kdkhBFbsmeK3x5v7bB1J616V2/ZBbsKgru5Sz5/oVGmJsYsfPcLSb/fhpFKfiHdeExiYZlDfs2Kh1Dd+veX687KDRG0zhKsvHjx/P999+zZMkSzp49y5tvvklISAivvPIKoPZADx061LD9+fPnWblyJRcuXODQoUP079+fU6dOMWPGDMM2n3zyCRMnTmTJkiV4eXkRGRlJZGQkd+/eLfLrE0IIIQqLJN0aq+5kw9SedQD4PPA8h6/e1jgiIYrOuqNhTL2/BvVbHaoz1N9L24BKgHru9sztVx+dDlYeCGHxnoJfSmztkVAUBZpVLod3BasCP74W/O4PMT9ZhIXoSpt+/foxd+5cpk6dSr169di1axebNm3C09MTgIiIiCxrdqenpzNnzhz8/Pzo0KEDSUlJ7Nu3Dy8vL8M2CxcuJCUlheeeew4XFxfD67PPPivqyxNCCCEKjU4pjG6SEi4uLg47OztiY2OLZNiaoiiM/ymIX46H42xrzqY3WlHOyrTQzyuElv48HcnoVcdIz1AY3tKbiV1rPrQgk8ju+92Xmb7xLDodfD24IZ1qOxfIcdMzFJ76ZAfhMYl82b8ePeu5Fshxtbbz3E2GLT2MdwUrdrzdRutwHqmo70ElnXxeQgghtJLbe5D0dBcDmdWJK1ewIjIuibfXBsn8blGq7b0Yxes/quvUP9/QTRLufBje0pvBzTzUYoxrjhNcQGtQ775wi/CYROwsTAoskS8OMiuYX4m6R2xizmtECyGEEEIUBkm6iwlrM2PmD2yAqbER2/+5WShDRoUoDo6H3GHkiiOkpGfQubYzM/vUlYQ7H3Q6HZO716Z19YokpWbw0rIjhN1JeOLjBtwvoNa7vivmJtmXdSqpylmZ4uagrjVelGudCyGEEEJI0l2M1Kpky4fdagEwe8s/HA+5o3FEQhSsc5HxDFt6mISUdFpVq8CXA+phXMIrY2vJWG/E/IH18XG2IequupRYXFL+e3Gj7iYTeOYGUPLX5s6J3/3e7mBZr1sIIYQQRUj+2i1mBjX1oKuvC2kZCq/9eJzYBBkGKUqHkOgEhiw+SGxiKvU97Pl6cEPMjEtPT6pWbO4vJeZoY8b5G3cZs+oYqflcSmz90TDSMhT83O2p6VL65sbWvV9MraCG4gshhBBC5IYk3cWMTqdjZp+6eJSzJDwmkf+tDyqUJYGEKEo34pIYtPgAN+OT8XG2YdmwJliZGWsdVqlRyd6CJcMaY2GiZ/eFKCb9eirP/28oimIYWt6/FPZyA/gakm7p6RZCCCFE0ZGkuxiyNTdh/sD6mOh1/Hn6Bsv3XdU6JCHy7c69FIYsPkjo7UQ8y1uyYngT7CxNtA6r1KnjasdXA9SlxNYcDuWbXZfztP/hq3e4HHUPS1M93f0qFVKU2qrjqibd4TGJRN9N1jgaIYQQQpQVknQXU75u9rz/TE0AZmz6h5PSMyNKoLvJaQxbdpjzN+7iZGvGyuFNcbQx1zqsUqt9LSdDXYhZm/9h08mIXO+75pC6vnJ330pYl9JRCLbmJlSuqK47HizF1IQQQghRRCTpLsaGNfeiYy0nUtIzeG31MeKfoECSEEUtKTWdl1ccISg0BgdLE1YOb4p7OUutwyr1XmzhzbDmXgC8GXCCY7koyBibmMrG+wl6/yalc2h5Jt/7vd3BoZJ0CyGEEKJoSNJdjOl0Oj59zg9XewuuRScw4eeTMr9blAhp6RmMXX2cfZeisTLVs+zFJlRzstE6rDJjUrdatPNxJDktg5HLjxB6+9FLif12IpzktAxqONlQz92+aILUSOZ63SfDYzSNQwghhBBlhyTdxZydpQlfDayPsZGOP4IjWH0oVOuQhHikjAyFd9efZOuZG5gaG/HdC43wK+WJXHGjN9Ixb0B9aleyJfpeCsOWHnroSgiKohj+X+nX2L3Ur5meWUwtKCxWHmIKIYQQokhI0l0CNPBw4J1ONQCY8vtpzkbEaRyREDlTFIVpG8+w/lgYeiMdCwY2oHmVClqHVSZZmRmz+IXGONuac+nWPV5ddZSUtOxLiZ0Kj+NMRBymxkb0aeCqQaRFq3YlO4x0cCs+mRtxUkxNCCGEEIVPku4SYmSryrStUZHktAzG/HiMe8lpWockRDZf/nWBpXuvAvDZ8750qOWkbUBlnLOdOUuGNcbKVM++S9F88Ev2KSqrD6sF1DrXdsbe0lSLMIuUhame6venOgTJet1CCCGEKAKSdJcQRkY65vSth7OtOZdv3WPShlNahyREFkv2XGHutgsATOlRm9713TSOSADUqmTL/IENMNLB2qNhLNx5yfBeQkoav524DpTetblzkjnEXFaFEEIIIURRkKS7BClnZcq8AfUx0sHPx8JZe0Tmd4viYd3RMKb+cQaA8R2q88L96tmieGjr48iUHrUB+PTPc/wWpCbaG4MjuJuchmd5S5pVLq9liEWq7v1iatLTLYQQQoiiIEl3CdPEuxzjO1QH4MMNp7lwI17jiERZ9+fpSN5dHwzA8JbevP50VY0jEjkZ4u/F8JbeALy9NogjV2+z5rD64K5vI3eMjEp3AbX/8svs6Q6XYmpCCCGEKHySdJdAr7apSsuqFUhMTWfMj8dITEnXOiRRRu29GMXrPx4nPUPh+YZuTOxas9RXvy7J3n+mJh1qOZGSlsGLyw5z9Nod9EY6nm9YtqYC1HC2wUSvIyYhldDbiVqHI4QQQohSTpLuEkhvpOOLfvWoYG3G+Rt3mfL7aa1DEmXQ8ZA7jFxxhJT0DDrXdmZmn7qScBdzeiMdX/avR11XO+KT1GKMT/s44mhrrnFkRcvMWE9NF1sAgmW9biGEEEIUMkm6S6iKNmZ82b8eOh2sORzKhhPhWodUJP6JjOOrvy6w6/wtMjJkWKhWzkXGM2zpYRJS0mlVrQJfDqiHsV7+OykJLE2NWfxCI1ztLQAY3MxT44i0UddVHWIeLMXUhBBCCFHIjLUOQORfi6oVeP3pasz76wLv/3ySuq52VK5orXVYheJmfBJfBJ4n4HAombm2dwUrBjfz5LmGbthZmGgbYBkSEp3AkMUHiU1Mpb6HPV8PboiZsV7rsEQeONqas+G1Fly8ebdMFVD7Lz83e1YdDCFYiqkJIYQQopBJ11QJ90a7ajT1Lse9lHRe+/E4Samla353Umo6C3ZcpO2nO1l9SE24m3qXw9rMmCtR95j2xxmazfiLCT8Hc+Z6nNbhlno34pIYtPgAN+OT8XG2YdmwJliZybO7kqiCtVmZTbgB6t4vpnYqPE5GzQghhBCiUGmedC9cuBBvb2/Mzc1p2LAhu3fvfui2e/bsoUWLFpQvXx4LCwt8fHz44osvsmyzbNkydDpdtldSUlJhX4om1Dma9SlnZcqZiDg+3nhW65AKREaGwoYT4Tz92U4+/fMc91LS8XOzY90r/gSM8ufA++2Y1qsO1Z2sSUxNZ/WhUJ6Zt5vnFu1jw4lwUtIytL6EUicmIYWhiw8RejsRz/KWrBjeBDtLGWEgSqZqjtaYmxhxNzmNy1H3tA5HCCGEEKWYpl1UAQEBjBs3joULF9KiRQu++eYbunTpwpkzZ/Dw8Mi2vZWVFa+99hq+vr5YWVmxZ88eRo0ahZWVFS+//LJhO1tbW86dO5dlX3Pz0lsoyNnOnM/7+jFs6WF+OHAN/yrleaaui9Zh5duRq7eZtvEsQaExAFSyM+fdLj50961kWNbI2syYIc08GdzUg4NXbvPD/mv8eTqSI9fucOTaHaZZn2FAEw8GNvXAxc5Cw6spHe4lpzFs6WHO3YjHydaMlcOb4mhTev9NidLPWG9E7Up2HL12h+CwGKo6ls6pOUIIIYTQnk7RcJHSpk2b0qBBAxYtWmRoq1mzJr169WLmzJm5OkafPn2wsrLihx9+ANSe7nHjxhETE5PvuOLi4rCzsyM2NhZbW9t8H6eozdr8D1//fQkbM2M2jm2FR3lLrUPKk5DoBGZv+YeNJyMAsDLVM7ptVYa39Mbc5PFzhm/EJbH6UAg/HgzhZnwyoI4E6FDTiSH+njSvUl6qa+dDUmo6w5cfZu/FaOwtTfhplD/VnWy0DkuIJzbl99Ms3XuVYc29mNyjttbhGJTUe5BW5PMSQgihldzegzQbXp6SksLRo0fp2LFjlvaOHTuyb9++XB3j+PHj7Nu3j9atW2dpv3v3Lp6enri5udGtWzeOHz9eYHEXZ291rE5DTwfik9N4bfWxEjPEOjYxlRmbztL+87/ZeDICIx0MaOLOjnfaMKZt1Vwl3ABOtuaMa1+dve89zYKBDWjqXY70DIUtpyMZ9P1B2n/+N8v2XiE+KbWQr6j0SEvPYOzq4+y9GI2VqZ7lLzaRhFuUGr5umRXMY7QNRAghhBClmmbDy6OiokhPT8fJySlLu5OTE5GRkY/c183NjVu3bpGWlsbkyZMZMWKE4T0fHx+WLVtG3bp1iYuL48svv6RFixYEBQVRrVq1HI+XnJxMcnKy4fu4uJJZkMtEb8S8AfV55svdBIfFMnvLP0zqVkvrsB4qNT2D1YdC+CLwPHcS1ES4ZdUKfNC1pmEN3fww0RvR1deFrr4unIuM54cDV/nlWDiXbt1j8u9n+OTPc/Su78pQfy9qOEsC+TAZGQrvrj/J1jM3MDU24rsXGuHnbq91WEIUGF83ewBOX48jLT1Dlr0TQgghRKHQvOzwg8N9FUV57BDg3bt3c/fuXQ4cOMB7771H1apVGTBgAADNmjWjWbNmhm1btGhBgwYN+Oqrr5g3b16Ox5s5cyZTpkx5wispHlztLfjseT9GrjjC4j1XaFa5PB1qOT1+xyKkKAo7zt3k441nuXRLLWBUpaIVE7vWok2NigU6BLyGsw3Te9Xl3c4+/HwsnB8OXOPizbusOhjCqoMhNPEux1B/TzrVdsZE/uA2UBSFaRvPsP5YGHojHQsGNqB5lQpahyVEgfIub4WNmTHxyWmcv3GXWpVkaLIQQgghCp5mSXeFChXQ6/XZerVv3ryZrff7Qd7e3gDUrVuXGzduMHnyZEPS/SAjIyMaN27MhQsXHnq8CRMmMH78eMP3cXFxuLu75/ZSip0OtZwY3tKbxXuu8PbaIDa90QpX++JRTOzs/Qrrey5GAVDOypQ321ejfxOPQk16bcxNeKG5F0P9Pdl/KZofDlxj65kbHLpym0NXbuNoY2YovOZkKwXCvvzrAkv3XgXgs+d9i92DGyEKgpGRjjquduy/HM3J8BhJuoUQQghRKDTr2jM1NaVhw4YEBgZmaQ8MDKR58+a5Po6iKFmGhuf0/okTJ3BxeXg1bzMzM2xtbbO8Srp3O/vg52ZHbGIqr/94jNR0bed334xP4r31wTwzbzd7LkZhqjdiVOvK7HynDUP8vYqsl1mn09G8agUWDW7InnfbMvbpqlSwNuNmfDJf/nWB5rO2M3rVUfZfikbDGoOaWrLnCnO3qQ+ppvSoTe/6bhpHJEThyZzXHRQWq3EkQgghhCitNB1ePn78eIYMGUKjRo3w9/fn22+/JSQkhFdeeQVQe6DDw8NZsWIFAAsWLMDDwwMfHx9AXbf7s88+4/XXXzccc8qUKTRr1oxq1aoRFxfHvHnzOHHiBAsWLCj6C9SQqbER8wc24Jl5uzkWEsOcred5r4tPkceRlJrO97svs3DnJRJS0gHo6uvCe519cC+nbXV1FzsLxneswWtPV2PL6Uh+2H+Vw1fvsOlkJJtORlLdyZohzTzp3cANazPNZ2IUiXVHw5j6xxkAxneozgvNvbQNSIhCljmv+6Qk3UIIIYQoJJpmEv369SM6OpqpU6cSERFBnTp12LRpE56engBEREQQEhJi2D4jI4MJEyZw5coVjI2NqVKlCrNmzWLUqFGGbWJiYnj55ZeJjIzEzs6O+vXrs2vXLpo0aVLk16c193KWzH7Wl9GrjvH135doVrkcbWo4Fsm5MzIUfgu6zidb/uF6bBIAfu72TOpak0Ze5YokhtwyNTaih18levhV4sz1OH44cI1fj4dz/sZdJm04zewt5+jTwJUhzTypVoord/95OpJ31wcDMLylN68/XVXjiIQofJk93f9ExpGclo6Zce5WSxBCCCGEyC1N1+kurkrbmp+Tfj3FDweuUc7KlE1jW+FsV7hzlg9fvc30P84Yhmu62lvwv8416O5bCSOjkrFOdmxiKuuPhrHywDUuR90ztPtXLs9Qf0861HIqVZWO912MYtjSw6SkZ/B8Qzc+ec5X1jQXZYKiKDSYFsidhFQ2jGlRLCr0l7Z7UGGTz0sIIYRWcnsPKhtjZsu4D7rW5Oi1O5yJiGPsmuP8OKJpoSSM16LvMWvzP2w+pRbHszLVM7ptVYa39M71WtvFhZ2FCS+19GZYcy/2XYpmxf6rbDt7g/2Xo9l/ORpnW3MGNvWgfxN3HG1KduG1E6ExjFhxhJT0DDrXdmZmn7qScIsyQ6fTUdfNnl3nbxEcFlMskm4hhBBClC6SdJcB5iZ6FgxqQLd5uzl05Tbz/rrA+I41Cuz4sYmpLNhxkWV7r5KSnoGRDvo19mB8h+pUtDErsPNowchIR8tqFWhZrQLhMYn8ePAaaw6FEhmXxOeB55n31wU613FmqL8Xjb0cNE1WFUUhMTWdOwmpxCSkEJuQSkxiKncSUohJSCU2UW2/k5B6/z21PfpeCukZCq2qVeDLAfVKVQ++ELnh52Z3P+mWed1CCCGEKHiSdJcR3hWsmNGnLm+sOcFXOy7StHJ5WlR9snWXU9Mz+PFgCHO3nedOQioArapV4IOuNfFxLn1D/FztLXinkw9j21Vj88lIVuy/yrGQGP4IjuCP4Ah8nG0Y4u9Jr3quWD1B4TVFUUhISVcT5nsp95Plf5PkmPtJdEyimjzfSUgxfJ2Szyr1zSqX4+vBDWU+qyiT6rqq87ol6RZCCCFEYZA53TkozfPD3lsfzJrDoVSwNmPzG63y1ROtKArb/7nJx5vOcvmWOt+5qqM1H3StSZvqFcvU0ORT4bH8sP8aG4LCSUpVE14bM2OebejGEH9PnGzNDUly7EN6ndXv1fY7979OTc//P0sTvQ57S1McLE2wtzDFztIEewsTHKxMsbMwwf5+u4OlifqepSmV7MzL1M9NiP+6EZdE0xl/YaSDU1M6YWmq7fPo0nwPKgzyeQkhhNBKbu9BknTnoDTfwBNT0um1YC/nbsTTsmoFlr/UBH0eipuduR7Hx5vOsPdiNADlrUwZ16E6Axq7l+lhybEJqaw9GsrKA9e4Gp1QIMc01RthZ2mSJXl2uJ8kZybPDpam2Fv8mzw7WJpgYaKXBFqIPGry8TZuxiez9hV/Gmu8wkJpvgcVBvm8hBBCaEUKqYkcWZjqmT+wPj3m72XPxSgW7bzIa09Xe+x+N+OSmLP1PD8dDUVR1ITwpZbejG5bBVtzkyKIvHizszRhRKvKvNTCm90Xo/hh/1X++uem4bOyt/y3h9nw9f3k2cHyftv95Dnze0mehSg6vm72bDt7g+CwWM2TbiGEEEKULpJ0l0HVnGyY2rM276wL5vPA8zT2KkfTyuVz3DYxJZ3vd19m0d+XSEhJB6CbrwvvdvbBvZxlUYZdIhgZ6WhdvSKtq1fkbnIaRjokeRaiBPB1s7ufdMdoHYoQQgghSpmyOx64jHu+kTt9GriSocDYNceJvpuc5f2MDIVfjofx9JydzAk8T0JKOvXc7Vn/qj/zBzaQhDsXrM2MsTQ1loRbiBLA100tpnZSiqk90sKFC/H29sbc3JyGDRuye/fuR26/YMECatasiYWFBTVq1GDFihXZtlm/fj21atXCzMyMWrVq8csvvxRW+EIIIYQmJOkuw6b1rEPlilbciEvmrbVBZGSo0/sPXblNr4V7eTMgiIjYJFztLZg3oD6/jG5OQ08ZdimEKH183ewBuBx1j9jEVG2DKaYCAgIYN24cH3zwAcePH6dVq1Z06dKFkJCQHLdftGgREyZMYPLkyZw+fZopU6YwZswYfv/9d8M2+/fvp1+/fgwZMoSgoCCGDBlC3759OXjwYFFdlhBCCFHopJBaDspSUZazEXH0WrCX5LQMRj1VmZDbCWw+FQmoPbWj21bhpRbemJvIUlJCiNKt5ezthN1J5McRTWn+hEsqPonieg9q2rQpDRo0YNGiRYa2mjVr0qtXL2bOnJlt++bNm9OiRQs+/fRTQ9u4ceM4cuQIe/bsAaBfv37ExcWxefNmwzadO3fGwcGB1atX5yqu4vp5CSGEKP1yew+Snu4yrqaLLR91rw3AN7sus/lUJEY6GNjUgx1vt2F0m6qScAshyoTMIeZBMsQ8m5SUFI4ePUrHjh2ztHfs2JF9+/bluE9ycjLm5uZZ2iwsLDh06BCpqepogv3792c7ZqdOnR56TCGEEKIkkqRbMKCJO33quwLwVPWKbH7jKWb0rpuvNbyFEKKkyhxifjI8RtM4iqOoqCjS09NxcnLK0u7k5ERkZGSO+3Tq1Invv/+eo0ePoigKR44cYcmSJaSmphIVFQVAZGRkno4JajIfFxeX5SWEEEIUZ1K9XKDT6ZjT1493OtfAxc5C63CEEEITvq73e7pDpaf7YR4sDKkoykOLRU6aNInIyEiaNWuGoig4OTkxbNgwPvnkE/T6f0dQ5eWYADNnzmTKlClPcBVCCCFE0ZKebgGof/RIwi2EKMvq3B9eHh6TmG1Fh7KuQoUK6PX6bD3QN2/ezNZTncnCwoIlS5aQkJDA1atXCQkJwcvLCxsbGypUUOfMOzs75+mYABMmTCA2NtbwCg0NfcKrE0IIIQqXJN1CCCEEYGtuQuUKVgAEh0tv93+ZmprSsGFDAgMDs7QHBgbSvHnzR+5rYmKCm5sber2eNWvW0K1bN4yM1D8//P39sx1z69atjzymmZkZtra2WV5CCCFEcSbDy4UQQoj7fN3suBx1j5NhsbSt4ah1OMXK+PHjGTJkCI0aNcLf359vv/2WkJAQXnnlFUDtgQ4PDzesxX3+/HkOHTpE06ZNuXPnDp9//jmnTp1i+fLlhmO+8cYbPPXUU8yePZuePXuyYcMGtm3bZqhuLoQQQpQGknQLIYQQ99V1s+fXE9cJDovROpRip1+/fkRHRzN16lQiIiKoU6cOmzZtwtPTE4CIiIgsa3anp6czZ84czp07h4mJCW3btmXfvn14eXkZtmnevDlr1qxh4sSJTJo0iSpVqhAQEEDTpk2L+vKEEEKIQiPrdOdA1vwUQoiy6cjV2zz39X4cbcw49EF7TWKQe1DeyOclhBBCK7JOtxBCCJFHtSrZYqSDm/HJRMYmaR2OEEIIIUoBSbqFEEKI+yxNjanuZAMgQ8yFEEIIUSAk6RZCCCH+o+799bqDw6SCuRBCCCGenCTdQgghxH/4utsDsmyYEEIIIQqGJN1CCCHEf/gaerpjkFqjQgghhHhSknQLIYQQ/+HjYoOJXkdMQiphdxK1DkcIIYQQJZwk3UIIIcR/mBnr8XFWl/0IkmJqQgghhHhCknQLIYQQD/B1U4eYn5RiakIIIYR4QpJ0CyGEEA/ITLqlp1sIIYQQT0qSbiGEEOIBvm72AJwKjyMjQ4qpCSGEECL/JOkWQgghHlDN0RpzEyPuJqdxOeqe1uEIIYQQogSTpFsIIYR4gLHeiNqV7s/rDo/RNhghhBBClGiSdAshhBA5qHt/ve6gUCmmJoQQQoj8k6RbCCGEyIGfe2ZPtyTdQgghhMg/SbqFEEKIHNR1tQfg9PVY0tIztA1GCCGEECWWJN1CCCFEDipXsMLazJik1Awu3LyrdThCCCGEKKEk6RZCCCFyYGSko46rLQDBJXS9bi8vL6ZOnUpISIjWoQghhBBlliTdQgghxEP43V+vOzisZM7rfuutt9iwYQOVK1emQ4cOrFmzhuTkZK3DEkIIIcoUSbqFEEKIh6jrphZTK6lJ9+uvv87Ro0c5evQotWrVYuzYsbi4uPDaa69x7NgxrcMTQgghygRJuoUQQoiHyOzp/icyjuS0dG2DeQJ+fn58+eWXhIeH89FHH/H999/TuHFj/Pz8WLJkCYqiaB2iEEIIUWpJ0i2EEEI8hJuDBQ6WJqSmK/wTEa91OPmWmprKTz/9RI8ePXjrrbdo1KgR33//PX379uWDDz5g0KBBWocohBBClFrGWgcghBBCFFc6nY66bvbsOn+L4PBY/NzttQ4pT44dO8bSpUtZvXo1er2eIUOG8MUXX+Dj42PYpmPHjjz11FMaRimEEEKUbtLTLYQQQjyCr+v9ed2hMdoGkg+NGzfmwoULLFq0iLCwMD777LMsCTdArVq16N+/v0YRCiGEEKWf9HQLIYQQj+B7v5jayfCSV0zt8uXLeHp6PnIbKysrli5dWkQRCSGEEGWP9HQLIYQQj+B7v5ja+RvxJKSkaRtMHt28eZODBw9maz948CBHjhzRICIhhBCi7JGkWwghhHgEZztzHG3MyFDgzPU4rcPJkzFjxhAaGpqtPTw8nDFjxmgQkRBCCFH2SNIthBBCPEbmEPOgErZe95kzZ2jQoEG29vr163PmzBkNIhJCCCHKHkm6hRBCiMfIHGJ+MixG0zjyyszMjBs3bmRrj4iIwNhYyroIIYQQRUGSbiGEEOIx6t7v6Q4uYT3dHTp0YMKECcTG/ht3TEwM77//Ph06dNAwMiGEEKLskMfcQgghxGNkLht2OeoecUmp2JqbaBxR7syZM4ennnoKT09P6tevD8CJEydwcnLihx9+0Dg6IYQQomyQpFsIIYR4jPLWZrjaWxAek8ipsFiaV62gdUi54urqSnBwMKtWrSIoKAgLCwtefPFFBgwYgIlJyXhwIIQQQpR0knQLIYQQueDnbkd4TCLB4SUn6QZ1He6XX35Z6zCEEEKIMkuSbiGEECIX6rras+lkJMElrJgaqFXMQ0JCSElJydLeo0cPjSISQgghyg5JuoUQQohc8CuBxdQuX75M7969OXnyJDqdDkVRANDpdACkp6drGZ4QQghRJuSrenloaChhYWGG7w8dOsS4ceP49ttvCywwIYQQojipfb+YWtidRKLvJmscTe688cYbeHt7c+PGDSwtLTl9+jS7du2iUaNG7Ny5U+vwhBBCiDIhX0n3wIED2bFjBwCRkZF06NCBQ4cO8f777zN16tQCDVAIIYQoDuwsTKhcwQqAk+Elo7d7//79TJ06lYoVK2JkZISRkREtW7Zk5syZjB07VuvwhBBCiDIhX0n3qVOnaNKkCQA//fQTderUYd++ffz4448sW7asIOMTQgghio2Stl53eno61tbWAFSoUIHr168D4Onpyblz57QMTQghhCgz8pV0p6amYmZmBsC2bdsMhVh8fHyIiIgouOiEEEKIYsTXzR4oOUl3nTp1CA4OBqBp06Z88skn7N27l6lTp1K5cmWNoxNCCCHKhnwl3bVr1+brr79m9+7dBAYG0rlzZwCuX79O+fLlCzRAIYQQorjwNfR0x2gbSC5NnDiRjIwMAKZPn861a9do1aoVmzZtYt68eXk+3sKFC/H29sbc3JyGDRuye/fuR26/atUq/Pz8sLS0xMXFhRdffJHo6Ogs28ydO5caNWpgYWGBu7s7b775JklJSXmOTQghhCiu8pV0z549m2+++YY2bdowYMAA/Pz8APjtt98Mw86FEEKI0qZ2JVuMdHAzPpkbccU/MezUqRN9+vQBoHLlypw5c4aoqChu3rzJ008/nadjBQQEMG7cOD744AOOHz9Oq1at6NKlCyEhITluv2fPHoYOHcrw4cM5ffo0a9eu5fDhw4wYMcKwzapVq3jvvff46KOPOHv2LIsXLyYgIIAJEybk/6KFEEKIYiZfSXebNm2IiooiKiqKJUuWGNpffvllvv766zwdKy9Pzffs2UOLFi0oX748FhYW+Pj48MUXX2Tbbv369dSqVQszMzNq1arFL7/8kqeYhBBCiJxYmhpTzdEGgKDQGG2DeYy0tDSMjY05depUlvZy5coZlgzLi88//5zhw4czYsQIatasydy5c3F3d2fRokU5bn/gwAG8vLwYO3Ys3t7etGzZklGjRnHkyBHDNvv376dFixYMHDgQLy8vOnbsyIABA7JsI4QQQpR0+Uq6ExMTSU5OxsHBAYBr164xd+5czp07h6OjY66Pk9en5lZWVrz22mvs2rWLs2fPMnHiRCZOnJhlqbL9+/fTr18/hgwZQlBQEEOGDKFv374cPHgwP5cqhBBCZJE5xLy4VzA3NjbG09OzQNbiTklJ4ejRo3Ts2DFLe8eOHdm3b1+O+zRv3pywsDA2bdqEoijcuHGDdevW0bVrV8M2LVu25OjRoxw6dAhQ1xXftGlTlm2EEEKIki5fSXfPnj1ZsWIFADExMTRt2pQ5c+bQq1evhz7xzklen5rXr1+fAQMGULt2bby8vBg8eDCdOnXK0js+d+5cOnTowIQJE/Dx8WHChAm0a9eOuXPn5udShRBCiCwyk+6gElBMbeLEiUyYMIHbt28/0XGioqJIT0/HyckpS7uTkxORkZE57tO8eXNWrVpFv379MDU1xdnZGXt7e7766ivDNv3792fatGm0bNkSExMTqlSpQtu2bXnvvfceGktycjJxcXFZXkIIIURxlq+k+9ixY7Rq1QqAdevW4eTkxLVr11ixYkWuC7Pk56n5g44fP86+ffto3bq1oW3//v3ZjtmpU6dcH1MIIYR4lMwK5ifDYlAURdtgHmPevHns3r2bSpUqUaNGDRo0aJDllVcPDktXFOWhQ9XPnDnD2LFj+fDDDzl69ChbtmzhypUrvPLKK4Ztdu7cyccff8zChQs5duwYP//8M3/88QfTpk17aAwzZ87Ezs7O8HJ3d8/zdQghhBBFyTg/OyUkJGBjo85p27p1K3369MHIyIhmzZpx7dq1XB0jP0/NM7m5uXHr1i3S0tKYPHlylqIskZGReT5mcnIyycnJhu/lqbkQQoiH8XGxwUSv405CKmF3EnEvZ6l1SA/Vq1evAjlOhQoV0Ov12e6lN2/ezHbPzTRz5kxatGjBO++8A4Cvry9WVla0atWK6dOn4+LiwqRJkxgyZIjhPl63bl3u3bvHyy+/zAcffICRUfa+gQkTJjB+/HjD93FxcZJ4CyGEKNbylXRXrVqVX3/9ld69e/Pnn3/y5ptvAurN19bWNk/HystT80y7d+/m7t27HDhwgPfee4+qVasyYMCAfB9z5syZTJkyJU9xCyGEKJvMjPX4ONtyMjyW4LDYYp10f/TRRwVyHFNTUxo2bEhgYCC9e/c2tAcGBtKzZ88c90lISMDYOOufGXq9HsAwQiAhISFbYq3X61EU5aGjCMzMzDAzM8v3tQghhBBFLV/Dyz/88EPefvttvLy8aNKkCf7+/oDa612/fv1cHSM/T80zeXt7U7duXUaOHMmbb77J5MmTDe85Ozvn+ZgTJkwgNjbW8AoNDc3VNQghhCib6paw9boLwvjx4/n+++9ZsmQJZ8+e5c033yQkJMQwXHzChAkMHTrUsH337t35+eefWbRoEZcvX2bv3r2MHTuWJk2aUKlSJcM2ixYtYs2aNVy5coXAwEAmTZpEjx49DAm6EEIIUdLlq6f7ueeeo2XLlkRERBjW6AZo165dlifgj5Kfp+Y5URQly9Bwf39/AgMDDb3voD4MaN68+UOPIU/NhRBC5IWfmx0/HoTgYl5MzcjI6JEjvfJS2bxfv35ER0czdepUIiIiqFOnDps2bcLT0xOAiIiILKuPDBs2jPj4eObPn89bb72Fvb09Tz/9NLNnzzZsM3HiRHQ6HRMnTiQ8PJyKFSvSvXt3Pv7443xcrRBCCFE86ZQnrAITFhaGTqfD1dU1z/sGBAQwZMgQvv76a/z9/fn222/57rvvOH36NJ6enkyYMIHw8HBDpfQFCxbg4eGBj48PoK7bPW7cOF5//XWmT58OwL59+3jqqaf4+OOP6dmzJxs2bGDixIns2bOHpk2b5iquuLg47OzsiI2NzfNweSGEEKXfmetxPDNvNzZmxgR91BEjo7yve/0wBXkP2rBhQ5bvU1NTOX78OMuXL2fKlCkMHz78iY5fHMg9WwghhFZyew/KV093RkYG06dPZ86cOdy9excAGxsb3nrrrYcWPslJXp+aZ2RkMGHCBK5cuYKxsTFVqlRh1qxZjBo1yrBN8+bNWbNmDRMnTmTSpElUqVKFgICAXCfcQgghxONUd7LGzNiI+OQ0rkTfo0pFa61DylFOI8eee+45ateuTUBAQKlIuoUQQojiLl893RMmTGDx4sVMmTKFFi1aoCgKe/fuZfLkyYwcObLEDwuTp+ZCCCEep8/CvRwLieGLfn70ru9WYMctinvQpUuX8PX15d69e4Vy/KIk92whhBBaKdSe7uXLl/P999/To0cPQ5ufnx+urq6MHj26xCfdQgghxOP4utlzLCSG4LDYAk26C1tiYiJfffUVbm4lJ2YhhBCiJMtX0n379m3DvOr/8vHx4fbt208clBBCCFHc+RoqmBffYmoODg5ZCqkpikJ8fDyWlpasXLlSw8iEEEKIsiNfSbefnx/z589n3rx5Wdrnz5+Pr69vgQQmhBBCFGe+bvYAnL4eS1p6Bsb6fK3CWai++OKLLEm3kZERFStWpGnTpjg4OGgYmRBCCFF25Cvp/uSTT+jatSvbtm3D398fnU7Hvn37CA0NZdOmTQUdoxBCCFHsVK5ghbWZMXeT07hw8y41XYrffOJhw4ZpHYIQQghR5uXrsXzr1q05f/48vXv3JiYmhtu3b9OnTx9Onz7N0qVLCzpGIYQQotgxMtJRx1VNtE8W0yHmS5cuZe3atdna165dy/LlyzWISAghhCh78j0WrlKlSnz88cesX7+en3/+menTp3Pnzh25iQshhCgzMoeYB4XFaBrHw8yaNYsKFSpka3d0dGTGjBkaRCSEEEKUPcVvApoQQghRQmQWUzsZXjx7uq9du4a3t3e2dk9PT0JCQjSISAghhCh7JOkWQggh8snX1R6AsxFxJKelaxtMDhwdHQkODs7WHhQURPny5TWISAghhCh7JOkWQggh8sm9nAX2liakpiuci4zXOpxs+vfvz9ixY9mxYwfp6emkp6ezfft23njjDfr37691eEIIIUSZkKfq5X369Hnk+zExMU8SixBCCFGi6HQ66rrasftCFEFhsYY53sXF9OnTuXbtGu3atcPYWL3lZ2RkMHToUJnTLYQQQhSRPCXddnZ2j31/6NChTxSQEEIIUZL4udmz+0IUJ8NiAE+tw8nC1NSUgIAApk+fzokTJ7CwsKBu3bp4ehavOIUQQojSLE9JtywHJoQQQmRV934xteBiumwYQLVq1ahWrZrWYQghhBBlkszpFkIIIZ6A3/0h5edvxJOYUryKqT333HPMmjUrW/unn37K888/r0FEQgghRNkjSbcQQgjxBJxszahoY0aGAqevF6/e7r///puuXbtma+/cuTO7du3SICIhhBCi7JGkWwghhHgCOp0Ov2I6xPzu3buYmppmazcxMSEuLk6DiIQQQoiyR5JuIYQQ4gnVvb9ed3BYjKZxPKhOnToEBARka1+zZg21atXSICIhhBCi7MlTITUhhBBCZNe5jjOuDhY08LDXOpQsJk2axLPPPsulS5d4+umnAfjrr7/48ccfWbduncbRCSGEEGWDJN1CCCHEE6rhbEMNZxutw8imR48e/Prrr8yYMYN169ZhYWGBn58f27dvx9bWVuvwRD6FxyRiZarH3jL71AEhhBDFjwwvF0IIIUqxrl27snfvXu7du8fFixfp06cP48aNo2HDhlqHJvIhPCaR9nP+pvv8PdxLTtM6HCGEELkgSbcQQghRym3fvp3BgwdTqVIl5s+fzzPPPMORI0e0Dkvkw6/Hw0lMTSf0diLzd1zUOhwhhBC5IMPLhRBCiFIoLCyMZcuWsWTJEu7du0ffvn1JTU1l/fr1UkStBPvtxHXD19/vvsyzDdyo6mitYURCCCEeR3q6hRBCiFLmmWeeoVatWpw5c4avvvqK69ev89VXX2kdlnhC/0TGce5GPCZ6Hf6Vy5OarjD5t9MoiqJ1aEIIIR5Bkm4hhBCilNm6dSsjRoxgypQpdO3aFb1er3VIogBk9nK3qeHI7Gd9MTU2Ys/FKDafitQ4MiGEEI8iSbcQQghRyuzevZv4+HgaNWpE06ZNmT9/Prdu3dI6LPEEFEXhtyA16e7hVwmP8pa82roKANP+OCNF1YQQohiTpFsIIYQoZfz9/fnuu++IiIhg1KhRrFmzBldXVzIyMggMDCQ+Pl7rEEUeHQuJIexOIpametrXdALg1TZVcC9nQURsEl9tl6JqQghRXEnSLYQQQpRSlpaWvPTSS+zZs4eTJ0/y1ltvMWvWLBwdHenRo4fW4Yk8+O1EOACdajtjYapOFzA30TO5e20AFu+5zMWbdzWLTwghxMNJ0i2EEEKUATVq1OCTTz4hLCyM1atXax2OyIO09Aw2nowA1KHl/9WuphPtfBylqJoQQhRjknQLIYQQZYher6dXr1789ttvWocicmnfpWii7qbgYGlCy2oVsr3/UffahqJqm05KUTUhhChuJOkWQgghhCjGNtyvWt7V1wUTffY/3TzKWzK6jRRVE0KI4kqSbiGEEEKIYiopNZ0/T6u91z38XB+63Sut1aJqkXFSVE0IIYobSbqFEEIIIYqpHf/c5G5yGpXszGnk6fDQ7f5bVO373Ze5eFMq1AshRHEhSbcQQgghRDGVObS8e71KGBnpHrltu5pOtK/pSFqGwkdSVE0IIYoNSbqFEEIIIYqhuKRUtp+7CWSvWv4wH3ZTi6rtvRgtRdWEEKKYkKRbCCGEEKIY+vNUJClpGVR1tKaWi22u9pGiakIIUfxI0i2EEEIIUQz9FqQOLe/pVwmd7tFDy//rldZV8ChnSWRcEvO2Xyis8IQQQuSSJN1CCCGEEMXMrfhk9l6MAqB7LoeWZzI30fNR91oALN59RYqqCSGExiTpFkIIIUSuLFy4EG9vb8zNzWnYsCG7d+9+5ParVq3Cz88PS0vL/7d332FNnV8Ax78hbEREkaGCoLgVtW4U995d2tZZUWvVOjq01mqtrVrbam214k/r6HCPVq0TrHtLxb1FcYA4AUEBSX5/XIlSUUETbgLn8zx5DDc3955E4HLyvu85eHl58e6773Lz5s0M+9y5c4cBAwbg5eWFvb095cqVY+3ataZ8GRZhzeGr6PRQ2bsAvm5O2X6+FFUTQgjzIUm3EEIIIZ5r8eLFDBkyhJEjR3Lw4EGCgoJo1aoVUVFRme6/Y8cOunfvTnBwMMeOHWPp0qXs37+f3r17G/ZJSUmhWbNmXLhwgWXLlnHq1ClmzZpF0aJP70edV6x8OLU8qwXUMvNFuwrYPSyqtuZItLFCE0IIkU2SdAshhBDiuSZPnkxwcDC9e/emXLlyTJkyBW9vb0JCQjLdf8+ePfj6+jJo0CD8/PyoV68e7733HgcOHDDsM2fOHG7dusVff/1F3bp1KV68OPXq1aNy5co59bLMUtTNJA5G3UGjgXYBXi98HO+Cjrz/sKja13+fkKJqQgihEkm6hRBCCPFMKSkphIeH07x58wzbmzdvzq5duzJ9TmBgIJcvX2bt2rXo9XquXbvGsmXLaNOmjWGfVatWUadOHQYMGICHhwcVK1Zk/PjxpKWlPTWW5ORk4uPjM9xym9WHlVHuwJKFcM9v/1LHkqJqQgihPkm6hRBCCPFMN27cIC0tDQ8PjwzbPTw8iInJvBd0YGAg8+fPp3Pnztja2uLp6UmBAgWYOnWqYZ/z58+zbNky0tLSWLt2LZ9//jmTJk1i3LhxT41lwoQJuLi4GG7e3t7GeZFmZFXEy08tT2dvo2VMeymqJoQQapKkWwghhBBZ8t+2VXq9/qmtrI4fP86gQYMYPXo04eHhrF+/nsjISPr162fYR6fT4e7uzsyZM6lWrRpvvfUWI0eOfOqUdYARI0YQFxdnuF26dMk4L85MnIyJ59S1BGy1VrSs8OJTyx/XuOyjomqjV0pRNSGEyGnWagcghBBCCPPm5uaGVqt9YlQ7Njb2idHvdBMmTKBu3bp88sknAAQEBODk5ERQUBBff/01Xl5eeHl5YWNjg1arNTyvXLlyxMTEkJKSgq2t7RPHtbOzw87OzoivzrysfDjK3bBMYVwcbYx23C/aVWD7mRvsOqcUVWsb8PKj6EIIIbJGRrqFEEII8Uy2trZUq1aN0NDQDNtDQ0MJDAzM9DlJSUlYWWX8MyM9uU4faa1bty5nz55Fp9MZ9jl9+jReXl6ZJty5nV6vfzS1vIpxk2Lvgo70b+gPwFd/H+euFFUTQogcI0m3EEIIIZ7rww8/5JdffmHOnDmcOHGCoUOHEhUVZZguPmLECLp3727Yv127dqxYsYKQkBDOnz/Pzp07GTRoEDVr1qRIESWhfP/997l58yaDBw/m9OnTrFmzhvHjxzNgwABVXqPa/o26zZU793Cy1dKkbOYzCF7Gew1K4FPQkWvxyUzdJEXVhBAip8j0ciGEEEI8V+fOnbl58yZjx44lOjqaihUrsnbtWooXLw5AdHR0hp7dPXv2JCEhgWnTpvHRRx9RoEABGjduzMSJEw37eHt7s3HjRoYOHUpAQABFixZl8ODBDB8+PMdfnzlIn1reooInDrba5+ydfelF1XrNO8DsHZG8Wb0Y/u7ORj+PEEKIjDR6qabxhPj4eFxcXIiLiyN//vxqhyOEECIPkWtQ9uSW9+tBmo5a4zdxMzGFue/WoFEZd5Odq/evBwg7cY3AkoWY37vWU4vhCSGEeLasXoNkerkQQgghhMp2nrvJzcQUCjrZUs/fzaTn+qJdeeysrdh17iZ/H4426bmEEEJI0i2EEEIIobqVEVcAaFPJCxutaf88e7yo2tdrpKiaEEKYmiTdQgghhBAqup+axsZj1wDjVy1/GimqJoQQOUeSbiGEEEIIFf1zMpa7yQ8oWsCBaj6uOXLO9KJqALN3RHLmWkKOnFcIIfIiSbqFEEIIIVSUPrW8XeUiWFnlXFGzxmU9aFrOgwc6PV+sOobU1hVCCNOQpFsIIYQQQiVx91LZfOo6AO0r58zU8sdJUTUhhDA9SbqFEEIIIVSy4VgMKQ90lHLPRzmvnO+ZLUXVhBDC9CTpFkIIIYRQyepDVwFllFutftlSVE0IIUxLkm4hhBBCCBXEJtxn59kbQM5VLc+MFFUTQgjTkqRbCCGEEEIFaw5Ho9NDFe8CFC/kpGosjct60Ky8UlRt9EopqiaEEMYkSbcQQgghhApWPTa13ByMbqsUVdt9/iarpaiaECKXOnDhFneSUnL0nKon3dOnT8fPzw97e3uqVavG9u3bn7rvihUraNasGYULFyZ//vzUqVOHDRs2ZNhn3rx5aDSaJ27379839UsRQgghhMiSqJtJHIy6g5UG2gZ4qR0OoBRVG9BIKao2ToqqCSFyodQ0He/9Hk6NcWEcvRKXY+dVNelevHgxQ4YMYeTIkRw8eJCgoCBatWpFVFRUpvtv27aNZs2asXbtWsLDw2nUqBHt2rXj4MGDGfbLnz8/0dHRGW729vY58ZKEEEIIIZ5r1SGlN3dgSTfc85vP3yh965egeCGlqNpPUlRNCJHL7Dx7g5uJKeS3t6GsZ851jFA16Z48eTLBwcH07t2bcuXKMWXKFLy9vQkJCcl0/ylTpjBs2DBq1KhBqVKlGD9+PKVKlWL16tUZ9tNoNHh6ema4CSGEEEKYA71ez8oI85pans7eRsuYdhUAmCNF1YQQuUz6sp42AV5Ya3MuFVYt6U5JSSE8PJzmzZtn2N68eXN27dqVpWPodDoSEhIoWLBghu13796lePHiFCtWjLZt2z4xEi6EEEIIoZaTMQmcib2LrdaKFhXNb2CgUVl3KaomhMh17qemseFoDAAdcrhjhGpJ940bN0hLS8PDwyPDdg8PD2JiYrJ0jEmTJpGYmEinTp0M28qWLcu8efNYtWoVCxcuxN7enrp163LmzNOnSCUnJxMfH5/hJoQQQghhCumj3I3KFsbFwUblaDInRdWEELnNphOxJKakUbSAA6/4uObouVUvpKbRaDJ8rdfrn9iWmYULFzJmzBgWL16Mu7u7YXvt2rXp2rUrlStXJigoiCVLllC6dGmmTp361GNNmDABFxcXw83b2/vFX5AQQgghxFPodHpWG6qWF1U5mqeTompCiNwmvZZG+ypFspRvGpNqSbebmxtarfaJUe3Y2NgnRr//a/HixQQHB7NkyRKaNm36zH2trKyoUaPGM0e6R4wYQVxcnOF26dKlrL8QIYQQQogs+jfqNlfu3MPJVkuTcu7Pf4KKpKiaECK3iLuXyuaT14Gcn1oOKibdtra2VKtWjdDQ0AzbQ0NDCQwMfOrzFi5cSM+ePVmwYAFt2rR57nn0ej0RERF4eT29HYednR358+fPcBNCCCGEMLb0qeUtKnpib6NVOZpns7fRMqb9o6Jqp6WomhDCQm04GkNKmo7SHvko65nzuZ6q08s//PBDfvnlF+bMmcOJEycYOnQoUVFR9OvXD1BGoLt3727Yf+HChXTv3p1JkyZRu3ZtYmJiiImJIS7uUY+1L7/8kg0bNnD+/HkiIiIIDg4mIiLCcEwhhBBCCDWkpulYe0RZH21uVcufplGZx4uqHZWiakIIi5RetbxDFXWW9aiadHfu3JkpU6YwduxYqlSpwrZt21i7di3FixcHIDo6OkPP7v/97388ePCAAQMG4OXlZbgNHjzYsM+dO3fo27cv5cqVo3nz5ly5coVt27ZRs2bNHH99QgghhBDp0vvDFnKypa6/m9rhZFl6UbU9529JUTUhhMWJTbjPrnM3APU+8NTo5SPLJ8THx+Pi4kJcXJxMNRdCCJGj5BqUPZb0fn24JIIV/16hW+3ifNWxotrhZMvUTWeYFHoad2c7/vm4IfnsrNUOSQghsmTuzki+XH2cqj4F+LN/XaMeO6vXINWrlwshhBBC5HZq9oc1hj4Pi6rFJiTzY9hptcMRQogsS6+l0UHFZT2SdAshhBBCmJia/WGN4fGianN3XpCiakIIi3DxZiIRl+5gpYE2AZJ0CyGEEELkWun9YdtVLoKVVc72hzWWRmXcaS5F1YQQFmT1wwJqdf3dKOxsp1ocknQLIYQQQpiQ2v1hjWnUY0XV0qsBCyGEOdLr9fz1cGq52h0jJOkWQgghhDChjP1hndUO56V4F3RkYCN/AMatOUHC/VSVIxJCiMydiE7gbOxdbK2taFHRU9VYJOkWQgghhDCh9BHh9pWLoNFY5tTyx/WpXwLfh0XVftp0Ru1whBAiU+m/exuXcSe/vY2qsUjSLYQQQghhIrHxj/eHLapyNMZhb6Pli4dF1ebsvMCpGCmqJoQwLzqd3rCe2xyW9UjSLYQQQghhIn8fjkanh6o+BfAp5Kh2OEaTXlQtTYqqCSHMUHjUba7cuUc+O2salXVXOxxJuoUQQgghTOXxqeW5zai25bG3sWJvpBRVyympaTq1QxDCIqx6WECtRQVP7G20KkcjSbcQQgghhElk7A/rpXY4Rudd0JEBDaWoWk6IibtPt9l7qfzlRsKOX1M7HCHMWmqajjVHogHzmFoOknQLIYQQQphE+khLXX833J3tVY7GNB4vqvZjmBRVM4X1R2No+eM2tp+5QVJKGoMWHeT41Xi1wxLCbO04e4NbiSm45bMlsGQhtcMBJOkWQgghhDA6vV7PyodTrtvlwqnl6exttIx5WFRt7i4pqmZMSSkPGLHiMP3+COdOUiqVirpQy68gSSlp9P51P7EJ99UOUQiztPrhB55tKnlhrTWPdNc8ohBCCCGEyEUe7w/bUuX+sKbWUIqqGd2Ry3G0/WkHC/ddQqOB9xqUYPn7gczsVp0ShZ24GnefPr+Fcz81Te1QhTAr91LS2HAsBoD2VcynY4Qk3UIIIYQQRpZeWKxRmcKq94fNCVJUzTh0Oj0ztp7jtZCdnL+RiGd+e+YH12JEq3LYWlvh4mjDnB41KOBow6FLd/hk2WH5kEOIx2w6eY3ElDSKuTrwik8BtcMxkKRbCCGEEMKIMvaHNZ+RFlPyLujIwEZSVO1lRMfdo+vsvXyz7iSpaXpaVvBk3eAgAv3dMuzn6+ZESJdqWFtpWH3oKj9ukrX0QqRLr6XRvnIRNBqNytE8Ikm3EEIIIYQRPd4ftrEZ9IfNKVJU7cWtPxpNyynb2XXuJg42Wia+XomQrq/g6mSb6f51ShZi3KsVAZgSdsbwIY8QeVncvVS2nLoOQHszqVqeTpJuIYQQQmTJ9OnT8fPzw97enmrVqrF9+/Zn7j9//nwqV66Mo6MjXl5evPvuu9y8eTPTfRctWoRGo6Fjx44miDxnpY+0NK/gYRb9YXOKnbUUVcuupJQHfLr8MP3++Je4e0qxtDWD6tG5hs9zR+k61/ChT5AfAB8vPUTEpTs5ELEQ5mvD0RhS0nSU8XCmrGd+tcPJQJJuIYQQQjzX4sWLGTJkCCNHjuTgwYMEBQXRqlUroqKiMt1/x44ddO/eneDgYI4dO8bSpUvZv38/vXv3fmLfixcv8vHHHxMUFGTql2FyGfvD5o2p5Y9rWMadFhWUomqjpKjaMx2+fIe2P+1g0X6lWFq/BiVZ/n4gJQrny/IxPm1VjiZl3Ul+oKP3rwe4eueeCSMWwrytPHQFML9RbpCkWwghhBBZMHnyZIKDg+nduzflypVjypQpeHt7ExISkun+e/bswdfXl0GDBuHn50e9evV47733OHDgQIb90tLS6NKlC19++SUlSpTIiZdiUun9YQs52VLXTPrD5rT0omr7pKhapnQ6PSFbzvHa9F2PiqX1rsWnrcpia529P821Vhp+fLsqZT2duXE3meBfD5CY/MBEkQu1yIdXzxcbf59d55SZVO3NsE2jJN1CCCGEeKaUlBTCw8Np3rx5hu3Nmzdn165dmT4nMDCQy5cvs3btWvR6PdeuXWPZsmW0adMmw35jx46lcOHCBAcHZymW5ORk4uPjM9zMiaE/bID59IfNacVcHxVV+/yvo/y8+SzxUlgNUIqldfllLxPXn+SBTk+rip6sHxJEYEm35z/5KfLZWfNLj+q45bPjRHQ8gxdFkKaTJC23OBEdT5WxoXz+1xG1QzFrfx+ORq+HV3wK4F3QUe1wnpA3rwZCCCGEyLIbN26QlpaGh4dHhu0eHh7ExMRk+pzAwEDmz59P586dsbW1xdPTkwIFCjB16lTDPjt37mT27NnMmjUry7FMmDABFxcXw83b2/vFXpQJPN4ftoMZTm/MSX3qlyCgmAsJ9x/w3YZT1J3wD99tOMmNu8lqh6aadUeUYmm7zz8qlja9yysUcMy8WFp2FHN1ZGb3athaWxF24hrfrj9phIiF2nQ6PSP/PELcvVTm743i/PW7aodktlaaeccISbqFEEIIkSX/Leyk1+ufWuzp+PHjDBo0iNGjRxMeHs769euJjIykX79+ACQkJNC1a1dmzZqFm1vWR/lGjBhBXFyc4Xbp0qUXf0FGlrE/rKva4ajKzlrL8vcDmdypMv7u+UhIfsDPm89Rb+I/jFl1LE+tPU5MfsDwZYd5f75SLC2gWNaLpWXHKz6ufPdGAAD/23aeJQfM52dDvJhl/17m36g7AOj1MGv7eXUDMlMXbiRy6NIdrDTQupKX2uFkylrtAIQQQghh3tzc3NBqtU+MasfGxj4x+p1uwoQJ1K1bl08++QSAgIAAnJycCAoK4uuvv+batWtcuHCBdu3aGZ6j0+kAsLa25tSpU5QsWfKJ49rZ2WFnZ2esl2ZU6VXL25lZf1i12GiteO2VYnSsUpTQE9f4efNZDl+OY96uC8zfe5FXqxalX4OS2SocZmkOX77D4EURRN5INBRLG9q0dLbXbmdVhypFOXc9kZ82nWHkn0fwKehI7RJ5s7aApYtLSmXiOmXGQpsAL9YcjmZ5+BWGNi2Ne357laMzL+kt8+r6u1HY2TyvDzLSLYQQQohnsrW1pVq1aoSGhmbYHhoaSmBgYKbPSUpKwsoq458ZWq3SPkuv11O2bFmOHDlCRESE4da+fXsaNWpERESEWU0bz4q4pEf9YfP61PL/srLS0KKCJysH1OWP4FrUKVGI1DQ9Sw5cpsnkrQyY/y9Hr8SpHaZRpen0TN9yltem7yLysWJpw1tmv1hadg1pUoo2AV6kpunp90c4F28mmvR8wjQmhZ7iZmIKpdzzMaVzFaoXdyUlTcfsnZFqh2ZW9Ho9f0UoVcvNdWo5yEi3EEIIIbLgww8/pFu3blSvXp06deowc+ZMoqKiDNPFR4wYwZUrV/jtt98AaNeuHX369CEkJIQWLVoQHR3NkCFDqFmzJkWKKElpxYoVM5yjQIECmW63BOuPRZttf1hzodFoqFfKjXql3Ai/eJuQLWcJOxHLmiPRrDkSTcMyhRnQyJ8avgXVDvWlRMfdY+jiCPacvwVAq4qeTHitklHWbmeFlZWGSW9W5vKtJA5djqPXvP2s6F8XFwebHDm/eHlHr8Txx56LAHzZoQI2Wiveb1iS4F8PMH9PFP0b+sv/50PHo+M5dz0RW2srWlTIfOaVOZCkWwghhBDP1blzZ27evMnYsWOJjo6mYsWKrF27luLFiwMQHR2doWd3z549SUhIYNq0aXz00UcUKFCAxo0bM3HiRLVegkmlt8Yyx/6w5qhacVd+6VGDkzHxhGw5x+pDV9ly6jpbTl2npm9B+jcqSYPShS1umv66I9F8ukIpfOVoq2VMuwq8Wb1Yjr8Oexsts7pXp8PPOzl3PZGBC/5lbs8aebaiviXRPexxr9MrS1XSK9s3KuNOaY98nL52l/l7L9K/ob/KkZqH9N+9Tcq642xvvh9EaPTS+O0J8fHxuLi4EBcXR/788mm1EEKInCPXoOwxh/crNv4+tSZsQq+H7cMamWW7GnN38WYiM7aeZ3n4ZVLSlLX9FYrkp39Df1pW9ERrZd7Jd2LyA8auPs7ih8XLAoq58ONbVfFzc1I1rmNX43hzxm6SUtLoVrs4X3W0vFkkec2SA5cYtuwwTrZaNn3UEE+XR+u3V/x7mQ+XHMItnx07hjfC3karYqTq0+n01Jv4D1fj7jOj6yu0rJjzRdSyeg2Sj7uEEEIIIV5Cen/YqmbaH9YSFC/kxITXKrF9eCP6BPnhaKvl2NV4Biz4l2aTt7LkwCVSHujUDjNThy/foe3UHSw+cAmNBt5vWJJl/QJVT7gBKhRxYUrnKmg08Puei/y664LaIYlniEtK5ZuHxdMGNy2VIeEGZeS7aAEHbtxNZvm/l9UI0awcuHibq3H3cbazpmEZd7XDeSZJuoUQQgghXoKhP2xlmVr+sjzy2zOyTXl2Dm/M4CalcHGw4fyNRIYtO0zD7zYzb2ck91LS1A4TyLxY2oLetXOkWFp2NK/gyfCWZQH4cvUxtp6+rnJE4mm+33iKWw+Lp71b1++Jx220VvQOUrbP3HaeNF3enrC86pBSQK1FRU+zH/U3n98IQgghhBAW5vH+sG0CJOk2FlcnW4Y2K83OTxvzWeuyFHa242rcfcasPk69if/w8+azxN9PVS2+q3fu0eWXPXy7/hQPdHpaV/Jk/ZAg6pQ0z/Zc79UvwZvViqHTw8D5/3LmWoLaIYn/OHoljvl7MxZPy0znGt64Otpw8WYS645G52SIZiU1Tceaw8rrt4SOEZJ0CyGEEEK8IEvoD2vJ8tlZ07d+SbYPa8TXHStSzNWBm4kpfLfhFHUn/MO3609y425yjsa09kg0rX7czp7zt3C01fLtGwH8/M4rOVad/EVoNBrGvVqJmr4FSUh+QPCvB7iVmKJ2WOKhpxVPy4yjrTU9An0BCNlyjrxanmvHmRvcTkrFLZ8tdSygF70k3UIIIYQQL+Dx/rDtZWq5SdnbaOlauzhbPm7IlM5VKO2Rj4TkB0zfco663/zDmFXHuHLnnkljSEx+wLBlh+g//1/i7qUSUMyFNYOC6FTd2yKqrNtaWzGjWzV8CjoSdSuJfr+Hk/zAPKbq53XLwi9zMOoOTrZaRrYu99z9e9TxxcFGqXuw4+yNHIjQ/KRXLW8bUMQiqvKbf4RCCCGEEGYoQ3/Yip5qh5MnWGut6Fi1KOsH12dmt2pU9i5A8gMd83ZdoMG3m/l46SHOxt41+nkPXbpDm5+2s+TAZTQa6N+wJMvfN49iadlR0MmW2T2q42xnzb4Ltxj559E8O1JqLuKSUvlmvVI8bUjT0k8UT8uMq5Mtb9X0BpTR7rzmXkoaG47FAMrMAEsgSbcQQuRFUXthxxRIU29NpBCWLn2kpXEZd/KbcX/Y3MjKSkPzCp781T+Q+b1rEViyEA90epaFX6bZD1vpPz+co1fiXvo8aTo9P28+y+shu7hwMwkvF6VY2rCWZZ+65tbclfJwZlqXV9BaaVgWfpn/bTuvdkh52uPF03rW9c3y83oHlcDaSsOuczc5dOmOyeIzR5tOXiMpJY1irg684lNA7XCyxDJ/WwghhHhx0Yfh944Q9gXs/lntaISwSDqdntURD6uWW0ARn9xKo9FQ19+NBX1q82f/QJqV90Cvh7VHYmg7dQc95uxjX+StFzr21Tv3eGfWHr7b8KhY2rrB5lssLTsalC7MF+3KAzBx/UnDqKHIWUevxPFHFoqnZaZoAQfaP/zdM2Nr3hrtXvnY715LWNoBknQLIUTeknANFr4NqUnK19snQeJNdWMSwgI93h+2UVnz7g+bV1T1cWVW9+psGFKfjlWKYKWBraev0+l/u3lzxi42n4rN8lTqNYejaTllG3sjLadYWnZ1r+NL9zrF0ethyKIIo8wMEFmXXjxNr1dqQjyreNrT9GtQEoD1x2I4d934yyrMUVxSKltOxQLQvnJRlaPJOkm6hRAir0i9D4u7QPxlKOQPHhUhOR62fqN2ZEJYnPT+sM0rmH9/2LymjKczU96qypaPG/FOLR9stVbsv3Cbd+fup81PO/j78NWn9jdOTH7AJ0sPMWDBv8Tff0BlCyuWll2j25YnqJQb91LT6PPbAWLj76sdUp6RoXham+cXT8tMaQ9nmpZzR6+HWXlkmcD6Y9Gkpukp6+lMGU9ntcPJMkm6hRAiL9DrYfVguLwf7F3g7cXQYrzy2IE5cOOMuvEJYUEsrT9sXuVTyJHxr1Zi+/BG9K1fAkdbLcej4xm44CBNJ29lyf5LpDzQGfZPL5a2NFwpljagUUmWWWCxtOyw1lox7Z1XKFnYiei4+/T57QD3U6WiuandSUrJUDzNI//zi6c9zfsNldHuFf9e4Voe+NAkfWp5ewv73StJtxBC5AU7f4TDi0CjhTfngZs/lGgApVqA7gGEjVE7QiEsxuP9YQNzwfre3M4jvz2ftS7HzuGNGdK0FC4ONkTeSGTY8sM0+G4zc3ZEPlEsbWGf2nzSwnKLpWWHi4MNc3rWwNXRhkOX4/ho6SF0T5kJIIxj0sbTL1Q8LTPVihekhq8rKWk65uyINE6AZupa/H12n1eWxLULkKRbCCGEOTm59lFS3fIbKNn40WPNv1IS8ZN/w4UdqoQnhKVJr1reppKXRfSHFQpXJ1uGNC3Nzk8bM7J1Odyd7YiOu8/Yv48biqW1qeTF+sH1qV0ib32YUryQEzO6VsNGq2HN4WimbJLZT6Zy5PKj4mljO1Q0ygc76aPd8/dGEXcv93Yl+ftwNHo9VCvuindBR7XDyRa5UgghRG527Ris6APooXovqNkn4+OFy0C1Hsr9jZ+DTvfEIYQQjzzeH7Z9Fcsp4iMeyWdnTZ/6Jdg2rBHjX62ET0FH8tlZ8+0bAUx7pyoujnmz/VutEoUY92olAH7adIaVEVdUjij3+W/xNGNVwm9Uxp0yHs7cTX7AH3suGuWY5mjVw+9JS1zWI0m3EELkVok3YOFbkHIXfIOg1beQWSGghiPANh9cPQhHl+d8nEJYEEvsDysyZ2+j5Z1aPmz9pCHho5rm2mJp2dGpujfvNSgBwCfLDhN+8bbKEeUuy8IvE3Hp5YqnZUaj0dCvofL/NndnZK5clx95I5FDl+PQWmloXclL7XCyTZJuIYTIjR6kwOKucCcKXP2g02+gfcroTT53qDdUub/pS0i9l3NxCmFhDEV8KltOf1jxbBqNBjtrqUCfbniLsjQr70HKAx3v/X6Ay7eT1A4pV3i8eNrQZi9XPC0zbQOKULSAAzfuprAs/LJRj20OVj9c1lPX3w23fHYqR5N9knQLIURuo9fDmqEQtRvs8sM7i8Gx4LOfU7s/5C8KcZdg74yciVMIC/N4f9gOMrVc5FJWVhqmdK5COa/83LibQu9fD3A3+YHaYVm87zee4lZiCqU98tEj0Nfox7fRWtEnyA+AmdvO8yAt9ywX0+v1/JU+tbyy5U0tB0m6hRAi99n9Mxz8AzRW8MYcZd3289g6QuNRyv3tk5Wp6UKIDNL7w5bxsKz+sEJkl5OdNbN7VKewsx0nYxIYvPDgU3ubi+c7cjmO+XujAPiyvXGKp2WmUw1vXB1tiLqVxLqjMSY5hxqOXY3n/PVE7KytaF7BQ+1wXogk3UIIkZuc3gihD5Pn5l9DqWZZf25AZ/AMgOR42PKNaeITwoJZan9YIV5EkQIOzOpeHTtrKzadjOWbdSfUDskiPV48rUMV4xVPy4yjrTU9A5XR7pAt59Drc8cHJelTy5uUc8fZ3jILHUrSLYQQuUXsSVgeDHodVO2mTBnPDisraDFOuX9gDtyQljFCpHu8P2x7C53eKER2VfEuwPdvVgZg1vZIFu+PUjkiy7M0/JKheNpnrY1XPO1putcpjoONluPR8Ww/Y/mz1nQ6vaFNY/vKlrusR5JuIYTIDZJuwcLOyii1TyC0mZx5pfLn8asPpVuBPg1CRxs/TiEsVHp/2Fd8Clhcf1ghXka7ykUY0rQUACP/PMruczdVjshy3ElKYeL6U4BpiqdlxtXJlrdr+gDKaLel23/hFtFx93G2s6ZhmcJqh/PCJOkWQghL9yAFlnSH2xeggA90/h2sbV/8eM2+BI0WTq2FyO1GC1MIS/aoP6zljrQI8aIGNylF+8pFeKDT0++PcCJvJKodkkUwdfG0p+kd5Ie1lYbd528ScelOjp3XFNJHuVtW9MTexnK7DEjSLYQQlkyvh3WfwIXtSq/ttxeDk9vLHbNwGajWU7m/8XPQ5Z4KqEK8CEvvDyvEy9JoNHz7RgBVvAsQdy+V4Hn7iUtKVTsss/Z48bSxHUxXPC0zRQo4GD4gnGHBo92paTrWHokGLL+WhiTdQghhyfbNhPB5gAZenw0e5Y1z3IYjwNYZoiPgyFLjHFMIC5VexCewZCEKO1tef1ghjMHeRsvM7tUo4mLP+RuJ9F8QTmouaktlTP8tnla7hOmKpz1NvwYlANhwPIZz1+/m+PmNYceZG9xOSsUtnx11VHgPjUmSbiGEsFRnN8H6T5X7zb6EMi2Nd+x8hSFoqHJ/01hIvWe8YwthQTL0h5Wp5SKPc3e2Z3bPGjjaatl59iZjVh3LNRWyjSm9eFo+O+scKZ6WmVIezjQt54FeDzO3nlclhpe18uHv3rYBXljn4EwBU7Ds6IUQIq+6cQaWvqtUKq/8NgQOMv45aveH/MUg/jLsCTH+8YWwAOn9YW2trWhhof1hhTCmcl75+emtqmg0MH9vFPN2XVA7JLNyJymFb9adBGBI01I5Ujztad5vWBKAFQcvExN3X7U4XsS9lDQ2Hr8GWP7UcpCkWwghLM+927CgMyTHQbGa0O7HF6tU/jw2DtDkYQXz7ZPh7nXjn0MIM2foD1vWcvvDCmFsTct7MKJVWQC++vs4m0/FqhyR+fh+4yluJ6XmePG0zFQr7kpN34KkpumZszNS1ViyK+zENZJS0vAu6EBV7wJqh/PSJOkWQghLkpYKS3vCrXPg4g1vzQdrE64xrfQmeFWGlATY+o3pziOEGXq8P2yHXDDSIoQx9QkqQafqxdDp4YMFBzkVk6B2SKpTs3ja06SPds/fc9Giit+tjHj4u7dyUTSmGFjIYep/JwghhMi6DZ/B+S1g4whvL4R87qY9n5UVNB+n3D8wF66fNu35hDAjGfvDmvhnTQgLo9Fo+LpjJWr5FeRu8gOCf93PjbvJaoelmseLp3VUqXhaZhqWKUxZT2cSU9L4Y+9FtcPJkjtJKWw9rcyeyA1Ty0GSbiGEsBz7ZyvVygFemwmelXLmvH5BUKY16NMgdHTOnFMIM5A+yt3CwvvDCmEqttZWzOhaDd9Cjly+fY9+v4eT/CBN7bBUYQ7F0zKj0Wjo10AZ7Z6zI5L7qeb//7P+aAypaXrKejpT2sNZ7XCMQpJuIYSwBOe3wtpPlPuNR0G5djl7/qZfgkYLp9dB5LacPbcQKni8P6xMLRfi6VydbPmlRw2c7a05cPE2I5YfyXMVzf9bPM1dxeJpmWkb4EXRAg7cTExhafhltcN5LsPU8lzUMUL1pHv69On4+flhb29PtWrV2L59+1P3XbFiBc2aNaNw4cLkz5+fOnXqsGHDhif2W758OeXLl8fOzo7y5cvz559/mvIlCCGEad08B0u6KyPNld6EoI9yPobCpaH6u8r9jZ+DTnqzitztUX9YW4vvDyuEqfm75yOkSzW0VhpWHLzC9C3n1A4pR323QSmeVsbDWfXiaZmx1lrRt77St3vmtnM8MOP+6jFx99kTeROAdpW9VI7GeFRNuhcvXsyQIUMYOXIkBw8eJCgoiFatWhEVFZXp/tu2baNZs2asXbuW8PBwGjVqRLt27Th48KBhn927d9O5c2e6devGoUOH6NatG506dWLv3r059bKEEMJ47t2BhW/B/TtQtBq0n2qaSuVZ0XAE2DpD9CE4skSdGITIIY/6wxax+P6wQuSEeqXcGNO+AqAkob9sP49Ol/tHvI9cjmPBvvTiaRXMonhaZjpV96agky2Xbt1j7dEYtcN5qr8PX0Wvh+rFXSnm6qh2OEaj6nfF5MmTCQ4Opnfv3pQrV44pU6bg7e1NSEjm/WCnTJnCsGHDqFGjBqVKlWL8+PGUKlWK1atXZ9inWbNmjBgxgrJlyzJixAiaNGnClClTcuhVCSGEkaQ9gGW94MZpcC4Cby1Q2nipxckNgj5U7m8aC6n31ItFCBPKbf1hhcgp3WoX5926vgB8veYEPeftJzbBsvpDZ4dOp+fzx4qn1TLjWTEOtlp6PhyFD9lyzmyXAOTWjhGqJd0pKSmEh4fTvHnzDNubN2/Orl27snQMnU5HQkICBQsWNGzbvXv3E8ds0aJFlo8phBBmI3QUnNsE1g5KpXJnT7UjgtrvK63K4q/AnulqRyOESeS2/rBC5KTRbcvzVYcK2Flbse30dVpN2c6mE9fUDssklhy4xCEzLJ72NN3rFMfRVsuJ6Hi2nbmhdjhPiLyRyOHLcWitNLSulHumloOKSfeNGzdIS0vDw8Mjw3YPDw9iYrI25WHSpEkkJibSqVMnw7aYmJhsHzM5OZn4+PgMNyGEUFX4r4+S2ldDoEgVVcMxsHGAJg8rmG//Ae5eVzceIUwgvYhP+8pFckV/WCFykkajoVsdX1Z/UI+yns7cTEwh+NcDjF551CIqZ2fVnaQUJq433+JpmSngaMvbNX0ACNlyVuVonrTq4e/eev5uFMpnp3I0xqX6ooP/Xsz0en2WLnALFy5kzJgxLF68GHf3jL0zs3vMCRMm4OLiYrh5e3tn4xUIYcF0OjgbBrEn1Y5EPO7CTljzsFhawxFQ4VV14/mvim+AVxVISYAtE9SORgijerw/bG6qnCtETivt4czKgXUJrucHwG+7L9Ju6g5OROeOwS1zL572NMH1/LC20rDn/C0ORt1WOxwDvV7PykNKLY3cNrUcVEy63dzc0Gq1T4xAx8bGPjFS/V+LFy8mODiYJUuW0LRp0wyPeXp6ZvuYI0aMIC4uznC7dOlSNl+NEBbocjjMaQ5/vA7Ta8HCd+BKuNpRidsXYHFX0KUqyXaD4WpH9CQrK2gxTrkfPg+un1I1HCGMKTf2hxVCLXbWWka1Lc+vvWpS2NmOM7F36TBtJ7N3RFp0kbXDl+9YRPG0zBQp4EDHqsoHijO2mk+V+WNX4zl/PRE7ayuaVzCD5XRGptp3iK2tLdWqVSM0NDTD9tDQUAIDA5/6vIULF9KzZ08WLFhAmzZtnni8Tp06Txxz48aNzzymnZ0d+fPnz3ATItdKiIG/+sMvjeHyfmW9MBo4tQZmNYbfOkLkdjDTAhu52v14WPAW3LuljCR3mK5epfLn8a0HZdoobcxCR6sdjcgh2WnzCTB//nwqV66Mo6MjXl5evPvuu9y8edPw+KxZswgKCsLV1RVXV1eaNm3Kvn37TP0ynskwtTwXjrQIoZYGpQuzfnAQTcq6k5Km46u/jytF1uItr8iaTqdn1MpjFlE87Wn6NVDah208fo2zsXdVjkaRXkCtaTkP8tlZqxyN8an6scyHH37IL7/8wpw5czhx4gRDhw4lKiqKfv36AcoIdPfu3Q37L1y4kO7duzNp0iRq165NTEwMMTExxMXFGfYZPHgwGzduZOLEiZw8eZKJEycSFhbGkCFDcvrlCWFeHiTDjh9gajWImK9sq/wODDoIA/Yp9zVaOL8Zfm0Lc1rA6Y2SfOcUXRqs6APXT0A+T6Vwmq2Zt8po9iVYWcPp9XB+q9rRCBPLbpvPHTt20L17d4KDgzl27BhLly5l//799O7d27DPli1bePvtt9m8eTO7d+/Gx8eH5s2bc+XKlZx6WRlk6A8bIEm3EMZUKJ8dv/SozlcdKxqKrLX80fKKrFla8bTM+Ls706y8B3q90rdbbTqdntUPk+52lXPn715Vk+7OnTszZcoUxo4dS5UqVdi2bRtr166lePHiAERHR2e4mP/vf//jwYMHDBgwAC8vL8Nt8ODBhn0CAwNZtGgRc+fOJSAggHnz5rF48WJq1aqV469PCLOg18PJNfBzLQgbAyl3oWh16P2PUqArvxcULq3cH3QQavQGrR1c2gsL3oT/BcGxP5WkUJhO2BglebW2V1qD5beAi45bKajeS7m/8XOlRoDItbLb5nPPnj34+voyaNAg/Pz8qFevHu+99x4HDhww7DN//nz69+9PlSpVKFu2LLNmzUKn07Fp06acelkZpPeHrVbcFe+CZv6hlxAWSKPR0K12cf7+oB7lvPJz62GRtVF/HeVeivn/nfF48bShzUpbRPG0p3m/YUkA/jx4heg4dVuA7r9wi+i4+zjbW9OwTGFVYzEV1Rcg9O/fnwsXLpCcnEx4eDj169c3PDZv3jy2bNli+HrLli3o9fonbvPmzctwzDfeeIOTJ0+SkpLCiRMneO2113Lo1QhhZmJPwO+vwqJ34HakMoL66v8gOBSKVXtyf9fi0GYSDDkMgR+AjRPEHIGlPZWkPWIBpKXm+MvI9SIWwK6flPsdfs78/8ZcNRgOdvkh5jAcXqx2NMJEXqTNZ2BgIJcvX2bt2rXo9XquXbvGsmXLMl0ali4pKYnU1NQMrUD/y5QdR3Jrf1ghzE0pD2f+GhBoKLL2+56LtJu2g+NXzbvIWobiaXWKqx3OS3nFx5WafgVJTdMzZ0ekqrGsfPi7t1VFT+xttKrGYiqqJ91CCBNIugVrh0FIXWW6uNYWgj6CD8Kh8ltKIaxncfaE5l/D0KNKUmXvAjfPwF/vw9RXYP8vkGp567DMUtReWP1wtk79T6DSG+rGk11ObhD0oXL/n68gJUndeIRJvEibz8DAQObPn0/nzp2xtbXF09OTAgUKMHXq1Kee59NPP6Vo0aJPFEl9nKk6juTm/rBCmKP0Imu/PSyydjb2Lh1/3skv28+bZZG1/xZPs7ag4mlPkz7avWBvFHFJ6gyqpDzQsfZINADtK+fejhGW/90ihHgk7QHsm6Ukxvv+pxS5KttWWbPdZDTY5cve8RwLQqPPYMhRaPolOBWGO1FKO6sfA2DnT5BsHgU4LNKdKFjcBdJSlP+nhp+pHdGLqfU+uHhD/BXY87Pa0QgTyk5LzuPHjzNo0CBGjx5NeHg469evJzIy0lC35b++/fZbFi5cyIoVK7C3f/qUTVN1HNlx9gYAdf3dcMtl/WGFMGf1HxZZa1pOKbL29ZoT9Ji7z6yKrD1ePO3VqkUtsnhaZhqWLkxZT2cSU9L4fc8FVWLYcfY6d5JScctnR52SueN9zYwk3ULkFue3Kuuv134M926De3novhLemg8F/V7u2Pb5od4QGHIEWn0H+YvB3WsQOgqmVIQtE5XRdZF1yXdh4duQeB08KsFrM58/A8Fc2dhDky+U+zumwN1YVcMRxvcibT4nTJhA3bp1+eSTTwgICKBFixZMnz6dOXPmEB0dnWHf77//nvHjx7Nx40YCAgKeGYupOo50q12cTR81YHjLMkY5nhAi6wrls2NW90dF1rafuUHLH7cTdtw8iqylF09ztrNmROuyaodjNBqNxjDaPXfnBe6n5vy6+vSOEe0qe6G1MtOOLUZgoX/hCSEM0vs6/9YeYo+Dgyu0/h7e2w4lGhr3XDYOUKuvUnCtw89QsKSS4G8ZD1MqKa2jJOF6Pp0O/nwPrh0FJ/eHlcqd1I7q5VR8HYpUVQr1bZmgdjTCyF6kzWdSUhJW//kgSatV1urpH+uK8N133/HVV1+xfv16qlevbuTIs6dk4XxUKOKiagxC5FXpRdbWDHpUZK33bwf4/K8jqhZZu534qHjakGalcXe23OJpmWlTyYtirg7cTExh6QHjzBzKqqSUB2w8pnyw0j6XVi1PJ0m3EJYq+S5sGgvTasKJ1Uq7r5p94YN/oWYf0Jqwx6G1LVTtCgP3wxtzwKOikmzt/FFJvtd+Andy9he3Rdn8NZz8W1lr/9Z8KGCcNamqsrKC5uOU++HzIPakquEI48tum8927dqxYsUKQkJCOH/+PDt37mTQoEHUrFmTIkWUP66+/fZbPv/8c+bMmYOvr6+hFejdu7JsRYi8yt9dKbLW+2GRtT/2RKlaZO27jUrxtLKell88LTPWWiv61lf6dv9v23kepOVcJ5KwE7HcS03Dp6AjVbwL5Nh51SBJtxCWRqeDQ4uUftvbJ0FaMvg1gH47oPV3yjrsnGKlVUY4++2AtxdDsRrw4D7smwk/VYG/BsCNszkXjyU4vET5fwNoPxW8a6objzH51lXWput1yqwHkatkt81nz549mTx5MtOmTaNixYq8+eablClThhUrVhj2mT59OikpKbzxxhsZWoF+//33Of76hBDmw85ay+dmUGTt8OU7LDQUT6uYK4qnZebNat4UdLLl8u17rDkS/fwnGMmqiCuA0jHiafVBcguN/vE5XgKA+Ph4XFxciIuLM9paMSGM4nI4rB8Ol/crX7v6QovxUKY1mMMvK70eIrfB9u+VfwE0VlC+o1I93bOiquGp7vIBmNta+aCk7hBo9qXaERnfjbMwvRboHig1BYy9xCEPkGtQ9mT1/UpLSyM1VVoeitzH1tb2iaUcucmtxBSGLTtM2AllGnJQKTcmvVnZ5D2ydTo9r07fyaHLcbxatSg/dK5i0vOpbeqmM0wKPU05r/ysHVTP5EnwnaQUaowLIzVNT+jQ+pTycDbp+Uwlq9cgE84/FRbh9gU4EwpnN0HcZajaBWr0Bq2N2pGJxyXEQNiXcGiB8rWNE9T/GOoMAGszqrKr0UCJBsrt0n5lRPf0Oji2QrmVbglBH4N3DbUjzXlxl5V+6WnJULqVUk0+N3Lzh+rBSvX8jZ9D363KjAghVKLX64mJieHOnTtqhyKESVhZWeHn54etra3aoZhEQSdbZnWvxvy9UXy95jjbz9ygxZRtfPtGZZqVz7yQozEsPnCJQ5fjcl3xtKfpVqc4IVvPcSI6nq2nr9OwjLtJz7fuaAypaXrKeeW32IQ7OyTpzmtS78HFnXAmDM6Gws3/TP1d/6nSg7nZV1CmlXmMnuZlqfdhz3QleU15uMax8jtKwpbfzPvIeteAdxZBzFEl/mN/wun1ys2vvpJ8+9XPG99jKYlKpfK715Sq8q/Pyt2JaIPhcGghxByBw4uhyjtqRyTysPSE293dHUdHx1w/hVHkLTqdjqtXrxIdHY2Pj0+u/f7WaDR0rV2c2iUKMmhhBMej4+nz2wG61PLh8zblcbA17jX1dmIK3z4snjY0FxZPy0wBR1veqenDLzsiCdlyzuRJ98rHppbnBTK9PBO5bmrfzXMPR7PD4MIOeHDv0WMaLfjUBv+mSvXkbd8pLYwAfIOUqctez27fIkxAr4dTa2HDZ8psBICi1aHVt1CsmqqhvbAbZ2HnD8p6dN0DZVuxGsq089Itc2/yrdPBsp5wfCU4FoI+m8E19xViecKOKRD2BTgXgQ/CwdZR7YgsRq67BpnYs96vtLQ0Tp8+jbu7O4UK5d7+ryJvi4uL4+rVq/j7+2Njk/tnKiY/SOP7DaeYtT0SAH/3fPz4VhWjdh747M8jLNgbRVlPZ/7+oF6uXcv9X9Fx96j/7WZS0/Ss6B/IKz6uJjlPTNx96nyzCb0edn7amKIFHExynpyQ1Wt23vgOymtSEuH0BljzMfxYBaa+oqwDPhuqJNz5i8Ir3aHT7zA8Et5dC0EfQq33lMrX9T4ErR1c2A7/qw8rB0B8zhVVyPNiT8DvHZWpyLcvQD5PePV/EBxquQk3KNOOO/wMgyKUKuvW9sra9IVvwYx6cHQ56NRrCWIyWycqCbeVDXT+I28k3AC1+oGLDyRchd0/qx2NyKPS13A7OsqHPiL3Sp9WnpaWC6+hmbCz1jKyTXl+D66J+2NF1mZtM06RtbxSPC0zXi4OdKxSFIAZW86Z7Dx/H76KXg81fF0tOuHODhnpzoTFjTLo9XDj9KPR7Iu7lHWj6axsoHgd8G+mjGi7l3v+qOKdKGUN8dFlytc2TlBvCNQZKCNWppJ0S+lvvH826NOUDz4CByofgtjlUzs647sbqyRj+395NHW+YEmoNxQCOittySzd0RWw7F3lfvtp8Eo3dePJaUeWwfJg5ffHoIPgbLq1d7mJxV2DVPas9+v+/ftERkbi5+eHvX3unx4q8qa8/H2eWZG179+sjMcLFll7vHjaa1WLMjmXF0/LzNnYBJpOVorhhn1YH39346+3bjd1B0euxPFVx4p0q23ZgxEy0p3bJSfAyTWweghMCYCfa8LGkXB+s5Jwu/hA9V7w1kIYfgF6rIa6g8CjfNam8RbwgTdmQ3AYFKsJqYmweZzSpurQImXKrDCOtAewb5YyI2HfTCXhLtcOBuxV1m7nxoQbIJ+7Ur176FFo+Bk4uMKtc7BqIPxUFfbOVGoQWKor/8Jf7yv36wzMewk3QIXXoMgryu+PLRPUjkaIPK1hw4YMGTIky/tfuHABjUZDRESEyWIS4mWlF1kb92pF7G2s2H7mBi2nbCP0+LUXOt7jxdM+zQPF0zLj7+5M84cF6v639bzRj3/++l2OXIlDa6WhdUVPox/fXMlIdybMcpRBr4fY449Gs6P2gO6x1idaO6VHbvpotlsp462R1euVytOhYyDuYQ/WIlWV9d7FA41zjrzq/FaleF3sceVr9/LQckLebLOUfBfC58KuqUrBMQCnwkqF9urBYG8mP4tZER8NsxpBQrTyM/nO4txdOO1ZLu6Cua2U1nHv71Jm2uRWKYlKbYyXZJbXIDOW20a6n1cIq0ePHsybNy/bx7116xY2NjY4O2dt1CotLY3r16/j5uaGtXXO1N1t3rw5mzZtYufOndSuXTtHzpkbWOL3uSmcjU0wFFkDsl1k7XZiCo0mbeFOUiqj25anVz0/U4Zr1v6Nus1r03dho9WwbVgjvFyMNwV8SthppoSdoWGZwsx7t6bRjquWrF6zJenOhNn8wXPvDpzfoiTZZzcpayMfV7CEkmD7NwPfeqaf9p16H/aGwLZJkJKgbCvXXhmtLFjCtOfObW5FKu2UTv6tfO3gCo1GQrV3QZvHmwqk3oeIP2DHj48+5LF3UdYIl++gLJfQWCkfKllpH95/wZsp+pqm3lN6cV/9F9zKQO9QJf68bFEX5Xu9VHPoslTtaIxPr1dqEqwfAe1+hLKtX+pwZnMNshC5LemOiYkx3F+8eDGjR4/m1KlThm0ODg64uDz6nZKamporimdFRUVRoUIFevXqRVJSErNmzVI1Hkt6Xy3x+9xUkh+kMWnjaWZuU0ZoSxZ24se3qlKx6POvw3m1eNrTdP7fbvZG3qJ3PT8+b1veKMfU6/U0mbSV8zcSmdypMq+9Uswox1WT9Om2RDodXDvyqG/2pb3KVON01g7gF/RwNLsJFCqZs/HZ2Cvrbat0hS3jIXwenFgFp9YpRdjqfwIOBXI2JkuTfFdpn7V7GqSlKNXja/SGhp+CY0G1ozMPNvbKe/JKD2VN8I7JSs2CrROVm7G9bNL+320piRB3Sfkg5Z1FknADNBurtIo7sxHObYaSjdSOyHhunoM1HylLewD2z3rppFvkbZ6ej6Zburi4oNFoDNsuXLiAl5cXixcvZvr06ezZs4eQkBDat2/PwIED2b59O7du3aJkyZJ89tlnvP3224ZjNWzYkCpVqjBlyhQAfH196du3L2fPnmXp0qW4urry+eef07dvX8O5/Pz8OHjwIFWqVGHLli00atSIsLAwhg8fzvHjx6lSpQpz586lTJkyhvN8/fXX/PTTT9y7d4/OnTvj5ubG+vXrnztNfe7cubRt25b333+fmjVrMmXKFJycHs0cuXPnDsOGDWPlypXExcXh7+/PN998Q9u2bQHYuXMnn332Gfv378fOzo6aNWuyaNEiXF1d8fX1ZciQIRmm11epUoWOHTsyZswYQJlhEBISwrp16wgLC+Pjjz9m9OjR9O3bl3/++YeYmBh8fHzo378/gwcPzhD7nDlzmDRpEmfPnqVgwYK8/vrrTJs2jV69ehEbG8vff/9t2PfBgwcUK1aM8ePH06tXr+d8N4jssrPW8lnrctQvVZgPl0Rw7noir07fybAWZQmu54eVVeYzSQ5dyrvF057m/YYl2Rt5iwX7ohjY2J8Cji9fa+fY1XjO30jEztqK5hXyztRykKRbfUm34Nw/SpJ9NgwSYzM+7lb64Wh2UyheV0lI1JavMLT9AWr0UUZrz21SksiIBdDoM6jWE7SW8elwjtHp4MgSCP0C7j4cxSjREFp+k7un274MrQ1UeRsCOsGJ1cq081vnQa9TRhb1usduaRm/zo4Xec5zY7eFTr/JDJB0hUoqH6TsnaH8znhvm+VPt3+QrLRF2z5JqaOhtVM+eKw7SO3IxDPo9XrupapT4dnBRmu0HsrDhw9n0qRJzJ07Fzs7O+7fv0+1atUYPnw4+fPnZ82aNXTr1o0SJUpQq1atpx5n0qRJfPXVV3z22WcsW7aM999/n/r161O27NPXso4cOZJJkyZRuHBh+vXrR69evdi5cycA8+fPZ9y4cUyfPp26deuyaNEiJk2ahJ/fs6fp6vV65s6dy88//0zZsmUpXbo0S5Ys4d13lUKUOp2OVq1akZCQwB9//EHJkiU5fvw4Wq3yeyQiIoImTZrQq1cvfvrpJ6ytrdm8eXO2q3l/8cUXTJgwgR9++AGtVotOp6NYsWIsWbIENzc3du3aRd++ffHy8qJTp04AhISE8OGHH/LNN9/QqlUr4uLiDO9H7969qV+/PtHR0Xh5eQGwdu1a7t69a3i+MI16pdxYP6Q+w5cfJvT4NcatPcHW09eZ1OnJImtpOj2jVh5Fr4fXqhalpp8MggA0KF2Ycl75OREdz++7L/JBk1Ivfcz03txNy3uQzy5vpaF569WaA50Oog/CmTAlyb5yIOMf/DZOUKLBw0S7Cbj6qhbqc3mUh24rlNeycSRcPwlrP1aKgTX/WplKmlt7L2fH5QOwbrjyfw3K/2mL8VCmtbw/WWGlhQodlVtWPZ6U6/6TkBtu+swT9scff+pz05//lMfdSkP+IqZ6RyxT/WEQsRCuHVWKMVbtonZEL+78FmV0++ZZ5euSjaH19zk/+0hk273UNMqP3qDKuY+PbYGjrXH+7BoyZAivvfZahm0ff/yx4f4HH3zA+vXrWbp06TOT7tatW9O/f39ASeR/+OEHtmzZ8syke9y4cTRo0ACATz/9lDZt2nD//n3s7e2ZOnUqwcHBhmR59OjRbNy4kbt37z7z9YSFhZGUlESLFi0A6Nq1K7NnzzYcJywsjH379nHixAlKly4NQIkSjz7U/Pbbb6levTrTp083bKtQocIzz5mZd95554nR5y+//NJw38/Pj127drFkyRJD0vz111/z0UcfZRj9rlGjBgCBgYGUKVOG33//nWHDhgHKiP6bb75Jvny5tEiqGSnoZMvMbtVYuO8SY/8+xo6zSpG1ia8HZBhlXbz/EofzePG0zGg0Gvo1KMHgRRHM3XWB3kElsrw+PjM6nZ7Vh5QWxO0r572/kSTpzgmJNx6NZJ/bBEk3Mz7uXl5JsP2bgU9tsLZTJ84XVaqpMmr776+webwyFXhBJ2Vb83HgWVHtCNURHw2bvoRDC5WvbZyg/sdKYTBL+z+2NBqNMnUfrcy6MBdOhZTv/9BR8M9XyocoRig6lqMSrikfMB55uC49n4cyW6XCq/IBmshR1atXz/B1Wloa33zzDYsXL+bKlSskJyeTnJycYXp2ZgICAgz306exx8bGPuMZGZ+TPnobGxuLj48Pp06dMiTx6WrWrMk///zzzGPOnj2bzp07Gwq2vf3223zyySecOnWKMmXKEBERQbFixQwJ939FRETw5ptvPvMcWfHf9xVgxowZ/PLLL1y8eJF79+6RkpJClSpVAOV1X716lSZNmjz1mL1792bmzJkMGzaM2NhY1qxZw6ZNm146VpE1Go2Gd2r5UNOvIIMXHeTY1Xj6/h7OO7V8GNWmPPdT0/h2w0kAhjYrjbuzGcwoNSNtKnnx/cZTXLp1j6Xhl+hex/eFj7Xvwi1i4u/jbG9NwzKFjRekhZCk25T0eqWgUtRu4LF6dXb5H45mP1yb7WL5RQTQWkONYKj0hjLdck+IMhr0vyCo2hUafZ43evTeOv/oA5bzW+HBw5ZXld9R2n/l91I3PiHUVLOvsub5TpTSo73BMLUjyhpdmlJZP2wsJMcBGuW1NB4pa/YtjIONluNjW6h2bmP5bzI9adIkfvjhB6ZMmUKlSpVwcnJiyJAhpKSkPPM4/y0UptFo0D2nJejjz0mfLv/4c/47hf559Xpv3brFX3/9RWpqKiEhIYbtaWlpzJkzh4kTJ+Lg8OzKyc973MrK6ok4UlNTn9jvv+/rkiVLGDp0KJMmTaJOnTo4Ozvz3XffsXfv3iydF6B79+58+umn7N69m927d+Pr60tQUNBznyeMy989Hyv6BxqKrC3YG8We8zcp4ZaPO0mplPV0pnsdy+4XbQrWWiv6BpVg1MpjzNx2nndq+rzweveVEUpB6NYVvbCztvAlZi9Akm5T0mgejmjqwbPSo0rj3jVz7+ibvYtSNKl6LwgbA8f+hH9/g6MrlCJsdQaAjfHaDqgu+S5c2PGwwnwY3I7M+HjR6tDqWyhWTZ34hDAnNvbQ5AtYHqysh36lh/l/GBd9CP4eClfCla+9qig1LYq+ompY4sVoNBqjTfE2J9u3b6dDhw507doVUJLgM2fOUK5cztYMKVOmDPv27aNbt26GbQcOHHjmc+bPn0+xYsX466+/MmzftGkTEyZMYNy4cQQEBHD58mVOnz6d6Wh3QEAAmzZtyjAV/HGFCxcmOjra8HV8fDyRkZGZ7vu47du3ExgYmGH0/ty5c4b7zs7O+Pr6smnTJho1yrxAZKFChejYsSNz585l9+7dhinzIuc9XmTto6URnL+eyPnriYAUT3uWN6t7MyXsDJdv32PNkWg6VCma7WOkPNCx9sjDqeVV8t7UcpCk2/RafqNU9HbOWxX6cPWFN+dBrfdhwwjlD9Z/voIDc6HpGKj4umnaNZlaer/09DZuUbuVKuTprKzBu/bD5QJNwDNApp0K8biKryszYa4cgM3joP1PakeUueQE2DxBaZOo14GtszJbpUaw5ReBE7mOv78/y5cvZ9euXbi6ujJ58mRiYmJyPOn+4IMP6NOnD9WrVycwMJDFixdz+PDhDOuv/2v27Nm88cYbVKyYcSla8eLFGT58OGvWrKFDhw7Ur1+f119/ncmTJ+Pv78/JkyfRaDS0bNmSESNGUKlSJfr370+/fv2wtbVl8+bNvPnmm7i5udG4cWPmzZtHu3btcHV1ZdSoUYYibM/i7+/Pb7/9xoYNG/Dz8+P3339n//79GQrDjRkzhn79+uHu7m4o9rZz504++OADwz69e/embdu2pKWl0aNHjxd4Z4Ux1SvlxvrBSpG1jcev8XZNbyme9gz2NlrerevL9xtPE7LlHO0rF8l2UcjtZ64Tdy+Vws521C5RyESRmjdJuk3NPY8XZPCpBcFhSg/bsDEQfxlW9Fb+kG0xXlnDbu6SbilT5c9tetgvPTrj4wV8HlWY9w0Ce+mrK8RTaTRKocW5LeHg70r/dQ/j9P80Cr1eaYW47lNIUKbCUeE15feVLA8RZmrUqFFERkbSokULHB0d6du3Lx07diQuLi5H4+jSpQvnz5/n448/5v79+3Tq1ImePXuyb9++TPcPDw/n0KFDmfbkdnZ2pnnz5syePZsOHTqwfPlyPv74Y95++20SExMNLcMASpcuzcaNG/nss8+oWbMmDg4O1KpVy9AybcSIEZw/f562bdvi4uLCV199laWR7n79+hEREUHnzp3RaDS8/fbb9O/fn3Xr1hn26dGjB/fv3+eHH37g448/xs3NjTfeeCPDcZo2bYqXlxcVKlSgSJG8OcpnblydbPlft2pcvn2PogVy0QxME+lW25eQLec4GZPAltPXaVTGPVvPT59a3i6gCNqntG3L7TT65y22yYOy2uRcZFPqPWUd544fIOVhJdMKryoj3+ZUpV2XBlcjHk0Z/2+FeWsH8K33KNEuVFJGs4XIrsVdlVZw/s2g6zK1o1HcvgBrP1H6iQO4+kGb75Wf8xwk16Dsedb7df/+fSIjI/Hz88PeXgokqaFZs2Z4enry+++/qx2KapKSkihSpAhz5sx5ouq8Mcj3ucgJ49YcZ9b2SGr6FWTJe3Wy/LyklAdU+yqMe6lp/DWgLlW8C5guSBVk9ZotI90i59g4KNWLq3ZTppUe/F1Z831yDdR+H4I+Uq8oUULMw37pYcq/925nfLxw2Udt3HwCzaNfuhCWrOmXcGodnA1VfuZKNlYvlgcpsHsqbP1OKX5oZaPUoAj6MHfVoBDCxJKSkpgxYwYtWrRAq9WycOFCwsLCCA0NVTs0Veh0OmJiYpg0aRIuLi60b99e7ZCEeGHB9Uowb9cF9kXeIvzibaoVd83S80KPX+NeahrFCzlSuVjeLT4qSbfIec4eyjrOmn1hw2cQuRV2/ggH/4BGn8ErPZVq6Kb0IAUu7X20NvvakYyP2+VXWp6lJ9q5ocK8EOakUEmo0UdZarJxFLzXQJ210hd2KoXSbpxSvvYNgjaToXDmrYmEEE+n0WhYu3YtX3/9NcnJyZQpU4bly5fTtGnOzhYxF1FRUfj5+VGsWDHmzZtnaIkmhCXydLHn1apFWXLgMjO2nmNW9ydb7GVm1cOp5R1eYC14biI//UI9nhWh+0plKufGz5X+3ms+gn2zlP7epYx8kb594WGS/Y+S6KdPcU9XpKqSZJdsAsWq594K80KYiwbD4NACuHZU6WdftWvOnTvxJoSOhog/lK8d3ZR12wGdZLmIEC/IwcGBsLAwtcMwG76+vs9tmSaEJelbvyRLwy8TevwaZ2MT8Hd3fub+txNT2Hr6OpB3q5ank6RbqEujgdItlKml4fNg83i4fhLmv64kv82/fvEiSylJSjuvcw/7Zt88m/Fxp8LKef2bQolGkK/wS78cIUQ2OBaE+p8oH7pt+kqp8WDr9PznvQydTkm0Q0c/WkZS7V1o+gU4ZG2qnBBCCJEX+bvno3l5DzYcu8aMref5/s3Kz9x/3dEYHuj0lPfK/9wEPbeTpFuYB60N1OwDld6Abd/D3v8pyfKMzUov30Yjn58U6/Vw/dSjAmgXd0Fa8qPHNVqlWnp6ou0ZYJlty4TITWr2hX0z4U4U7JoGDYeb7lzXjitTyS/tUb72qKj03PauabpzCiGEELlIvwYl2XDsGisjrvBhs9IUeUb195URVwDokMdHuUGSbmFuHFyhxTilF27oF0rrnvC5cGQZ1P9I6fv9eBGze3eUqeLpa7Pjr2Q8nov3w57ZTcGvvnqF2oQQmbO2UzoYLOul1Hao1gOcPY17jpRE2DpR6Z6gewA2Tkr9iFr9TF8/QgghhMhFqvq4UrtEQfacv8XsHZGMapv5jNTouHvsu3ALgLaVJemWvzaEeSpYAjr/roxWb/gMrh5U+nzvn6NUQL8bq4yEX9oH+rRHz7O2h+J1H7Xzcisl6zOFMHcVXoPd05X2fJvHQfupxjv2qfVKG7C4KOXrcu2g5TdSHFEIIYR4Qe839GfP+X0s3BfFB439KeBo+8Q+fx+KRq+Hmr4FpRc6knQLc1c8EHr/A0eWQNiXyh/Oqwdl3Met9KMq48XrSosfISyNRqPMcJnTQuliUKsfeFR4uWPGXYZ1w+Hk38rXLj7Q+jso0/Ll4xVCCCHysPql3CjvlZ/j0fH8tvsig5qUemKflYeU2aftZGo5IEm3sARWVlD5LSjXHnZPg6PLoZC/kmSXbAKuxdWOUAjxsnxqKz/jJ1YpLcS6rXix46Slwt4ZsHkCpCaClTXUGahUSjd1kTYhhBAiD9BoNPRrWJJBCw8yb9cF+gSVwMH2UdvPc9fvcvRKPNZWGtpU8lIxUvMhVaSE5bB1VP5wHrAX3poP1XtJwi1EbtJ0DFjZPOo4kF2X9sHMhko19NRE8KkD722HZl9Kwi3ynIYNGzJkyBDD176+vkyZMuWZz9FoNPz1118vfW5jHUcIYb5aV/TEu6ADtxJTWHLgUobH0ntzB5Vyo6DTk1PP8yJJuoUQQpiHQiWVLgYAG0eDLu3Z+6e7dxtWD4bZzZSe3w6u0H4a9Fz74i0HhVBJu3btaNq0aaaP7d69G41Gw7///pvt4+7fv5++ffu+bHgZjBkzhipVqjyxPTo6mlatWhn1XE9z7949XF1dKViwIPfu3cuRcwohwFprRd/6JQGYue08qWk6APR6PasOKUl3Xu/N/ThJuoUQQpiP+p8oXQZij0HE/Gfvq9fDocUwtTqEz1O2Ve0KA8PhlW7SElBYpODgYP755x8uXrz4xGNz5syhSpUqvPLKK9k+buHChXF0dDRGiM/l6emJnZ1djpxr+fLlVKxYkfLly7NixQsuSzESvV7PgwcPVI1BiJz0ZrViuOWz5cqde6w5HA3A0SvxRN5IxN7GimbljdyNxILJXyRCCCHMh2NBqD9Muf/POEi+m/l+10/Dr+3gz76QdAMKl4V310GHn8GpUM7FK4SRtW3bFnd3d+bNm5dhe1JSEosXLyY4OJibN2/y9ttvU6xYMRwdHalUqRILFy585nH/O738zJkz1K9fH3t7e8qXL09oaOgTzxk+fDilS5fG0dGREiVKMGrUKFJTUwGYN28eX375JYcOHUKj0aDRaAwx/3d6+ZEjR2jcuDEODg4UKlSIvn37cvfuo5/tnj170rFjR77//nu8vLwoVKgQAwYMMJzrWWbPnk3Xrl3p2rUrs2fPfuLxY8eO0aZNG/Lnz4+zszNBQUGcO3fO8PicOXOoUKECdnZ2eHl5MXDgQAAuXLiARqMhIiLCsO+dO3fQaDRs2bIFgC1btqDRaNiwYQPVq1fHzs6O7du3c+7cOTp06ICHhwf58uWjRo0ahIVlXDKTnJzMsGHD8Pb2xs7OjlKlSjF79mz0ej3+/v58//33GfY/evQoVlZWGWIXQm32NlreresHwIyt59Dr9Ybe3E3LeZDPTsqHpZOkWwghhHmp2QdcfeFujFI88XGp9+CfryEkEC5sB2sHaPKFsna7eKAq4QoLotcrfdvVuOn1WQrR2tqa7t27M2/ePPSPPWfp0qWkpKTQpUsX7t+/T7Vq1fj77785evQoffv2pVu3buzduzdL59DpdLz22mtotVr27NnDjBkzGD58+BP7OTs7M2/ePI4fP86PP/7IrFmz+OGHHwDo3LkzH330ERUqVCA6Opro6Gg6d+78xDGSkpJo2bIlrq6u7N+/n6VLlxIWFmZIbtNt3ryZc+fOsXnzZn799VfmzZv3xAcP/3Xu3Dl2795Np06d6NSpE7t27eL8+fOGx69cuWL4YOGff/4hPDycXr16GUajQ0JCGDBgAH379uXIkSOsWrUKf3//LL2Hjxs2bBgTJkzgxIkTBAQEcPfuXVq3bk1YWBgHDx6kRYsWtGvXjqioKMNzunfvzqJFi/jpp584ceIEM2bMIF++fGg0Gnr16sXcuXMznGPOnDkEBQVRsmTJbMcnhCl1rVUcJ1stJ2MS2HQiltWHH04tl97cGcjHD0IIIcyLtZ1SVG1pT9j5I7zSA/J7KcXV1nwEty8o+5VqrrQBc/VVL1ZhWVKTYLxKfwh+djXLBf169erFd999x5YtW2jUqBGgJF2vvfYarq6uuLq68vHHHxv2/+CDD1i/fj1Lly6lVq1azz1+WFgYJ06c4MKFCxQrpvSsHz9+/BPrsD///HPDfV9fXz766CMWL17MsGHDcHBwIF++fFhbW+Pp+fQppPPnz+fevXv89ttvODkpr3/atGm0a9eOiRMn4uHhAYCrqyvTpk1Dq9VStmxZ2rRpw6ZNm+jTp89Tjz1nzhxatWqFq6srAC1btmTOnDl8/fXXAPz888+4uLiwaNEibGxsAChdurTh+V9//TUfffQRgwcPNmyrUaPGc9+//xo7dizNmjUzfF2oUCEqV66c4Tx//vknq1atYuDAgZw+fZolS5YQGhpqWL9fokQJw/7vvvsuo0ePZt++fdSsWZPU1FT++OMPvvvuu2zHJoSpuTja0KV2cWZuO8+IP49wPSGZ/PbWNChTWO3QzIqMdAshhDA/5TtCsRpKkrT+UyUB/+N1JeF2LgKdfod3lkjCLXKlsmXLEhgYyJw5cwBlRHf79u306tULgLS0NMaNG0dAQACFChUiX758bNy4McNI6rOcOHECHx8fQ8INUKdOnSf2W7ZsGfXq1cPT05N8+fIxatSoLJ/j8XNVrlzZkHAD1K1bF51Ox6lTpwzbKlSogFb7qOWQl5cXsbGxTz1uWloav/76K127djVs69q1K7/++itpaUoRxoiICIKCggwJ9+NiY2O5evUqTZo0ydbryUz16tUzfJ2YmMiwYcMoX748BQoUIF++fJw8edLw3kVERKDVamnQoEGmx/Py8qJNmzaG//+///6b+/fv8+abb750rEKYQq+6fthoNVxPSAagdSUv7Ky1z3lW3iIj3UIIIcyPRgPNx8Gc5nD8r4fbtFD7fWj4Kdg5qxqesFA2jsqIs1rnzobg4GAGDhzIzz//zNy5cylevLghQZw0aRI//PADU6ZMoVKlSjg5OTFkyBBSUlKydGx9JlPdNRpNhq/37NnDW2+9xZdffkmLFi0MI8aTJk3K1uvQ6/VPHDuzc/43MdZoNOh0uqced8OGDVy5cuWJKe1paWls3LiRVq1a4eDg8NTnP+sxAKuHhRgff6+etsb88Q8UAD755BM2bNjA999/j7+/Pw4ODrzxxhuG/5/nnRugd+/edOvWjR9++IG5c+fSuXPnHCuEJ0R2ebrY81rVYix+2DpMppY/SUa6hRBCmCefWlDp4chO0erQdwu0GCcJt3hxGo0yxVuN21MSz6fp1KkTWq2WBQsW8Ouvv/Luu+8aktTt27fToUMHunbtSuXKlSlRogRnzpzJ8rHLly9PVFQUV68++gBi9+7dGfbZuXMnxYsXZ+TIkVSvXp1SpUo9UVHd1tbWMKr8rHNFRESQmJiY4dhWVlYZpnpn1+zZs3nrrbeIiIjIcOvSpYuhoFpAQADbt2/PNFl2dnbG19eXTZs2ZXr8woWVqbHR0dGGbY8XVXuW7du307NnT1599VUqVaqEp6cnFy5cMDxeqVIldDodW7dufeoxWrdujZOTEyEhIaxbt84wy0EIc/VegxLYWVvhW8iRWiWkoOl/SdIthBDCfHUMgb5bITgUvALUjkaIHJMvXz46d+7MZ599xtWrV+nZs6fhMX9/f0JDQ9m1axcnTpzgvffeIyYmJsvHbtq0KWXKlKF79+4cOnSI7du3M3LkyAz7+Pv7ExUVxaJFizh37hw//fQTf/75Z4Z9fH19iYyMJCIighs3bpCcnPzEubp06YK9vT09evTg6NGjbN68mQ8++IBu3boZ1nNn1/Xr11m9ejU9evSgYsWKGW49evRg1apVXL9+nYEDBxIfH89bb73FgQMHOHPmDL///rthWvuYMWOYNGkSP/30E2fOnOHff/9l6tSpgDIaXbt2bb755huOHz/Otm3bMqxxfxZ/f39WrFhBREQEhw4d4p133skwau/r60uPHj3o1asXf/31F5GRkWzZsoUlS5YY9tFqtfTs2ZMRI0bg7++f6fR/IcxJicL5CB3agGXvB6K1yt6HjHmBJN1CCCHMl9YGilSRntsiTwoODub27ds0bdoUHx8fw/ZRo0bxyiuv0KJFCxo2bIinpycdO3bM8nGtrKz4888/SU5OpmbNmvTu3Ztx48Zl2KdDhw4MHTqUgQMHUqVKFXbt2sWoUaMy7PP666/TsmVLGjVqROHChTNtW+bo6MiGDRu4desWNWrU4I033qBJkyZMmzbtiX2zKr0oW2brsRs1aoSzszO///47hQoV4p9//uHu3bs0aNCAatWqMWvWLMNU9h49ejBlyhSmT59OhQoVaNu2bYYZA3PmzCE1NZXq1aszePBgQ4G25/nhhx9wdXUlMDCQdu3a0aJFiyd6q4eEhPDGG2/Qv39/ypYtS58+fTLMBgDl/z8lJUVGuYXF8CnkiFs+O7XDMEsafWYLe/K4+Ph4XFxciIuLI3/+/GqHI4QQIg+Ra1D2POv9un//PpGRkfj5+WFvb69ShEK8mJ07d9KwYUMuX778zFkB8n0uhHqyes2WoQMhhBBCZMn06dMNf9hXq1aN7du3P3P/+fPnU7lyZRwdHfHy8uLdd9/l5s2bGfZZvnw55cuXx87OjvLlyz8xhVmIvCY5OZmzZ88yatQoOnXq9MLT8IUQ5kOSbiGEEEI81+LFixkyZAgjR47k4MGDBAUF0apVq6e2kNqxYwfdu3cnODiYY8eOsXTpUvbv30/v3r0N++zevZvOnTvTrVs3Dh06RLdu3ejUqRN79+7NqZclhNlZuHAhZcqUIS4ujm+//VbtcIQQRiDTyzMhU/uEEEKoxVyvQbVq1eKVV14hJCTEsK1cuXJ07NiRCRMmPLH/999/T0hICOfOnTNsmzp1Kt9++y2XLiltZTp37kx8fDzr1q0z7NOyZUtcXV0zXR+cGZleLvI6+T4XQj0yvVwIIYQQRpGSkkJ4eDjNmzfPsL158+bs2rUr0+cEBgZy+fJl1q5di16v59q1ayxbtow2bdoY9tm9e/cTx2zRosVTjwnK1Nv4+PgMNyGEEMKcSdIthBBCiGe6ceMGaWlpT6wt9fDweGqrqsDAQObPn0/nzp2xtbXF09OTAgUKGFoyAcTExGTrmAATJkzAxcXFcPP29n6JVyaEEEKYniTdQgghhMgSjSZj71W9Xv/EtnTHjx9n0KBBjB49mvDwcNavX09kZCT9+vV74WMCjBgxgri4OMMtfar6s8hKOpGbyfe3EObPWu0AhBBCCGHe3Nzc0Gq1T4xAx8bGPrWy8oQJE6hbty6ffPIJAAEBATg5OREUFMTXX3+Nl5cXnp6e2TomgJ2dHXZ2WesDm96POSkpCQcHhyw9RwhLk5KSAoBWq1U5EiHE00jSLYQQQohnsrW1pVq1aoSGhvLqq68atoeGhtKhQ4dMn5OUlIS1dcY/M9KTgvSRuTp16hAaGsrQoUMN+2zcuJHAwECjxK3VailQoACxsbEAODo6PnMUXQhLo9PpuH79Oo6Ojk/8vAkhzIf8dAohhBDiuT788EO6detG9erVqVOnDjNnziQqKsowXXzEiBFcuXKF3377DYB27drRp08fQkJCaNGiBdHR0QwZMoSaNWtSpEgRAAYPHkz9+vWZOHEiHTp0YOXKlYSFhbFjxw6jxe3p6QlgSLyFyG2srKzw8fGRD5SEMGOSdAshhBDiuTp37szNmzcZO3Ys0dHRVKxYkbVr11K8eHEAoqOjM/Ts7tmzJwkJCUybNo2PPvqIAgUK0LhxYyZOnGjYJzAwkEWLFvH5558zatQoSpYsyeLFi6lVq5bR4tZoNHh5eeHu7k5qaqrRjiuEubC1tcXKSso0CWHOpE93Jsy1R6oQQojcT65B2SPvlxBCCLVIn24hhBBCCCGEEEJlknQLIYQQQgghhBAmIkm3EEIIIYQQQghhIlJILRPpy9zj4+NVjkQIIURek37tiY+Px9nZWSoSP4dcs4UQQqgl/drzvDJpknRnIiEhAQBvb2+VIxFCCJFXeXt7S3GwLJBrthBCCLUlJCTg4uLy1MelenkmdDodV69ezVMjDPHx8Xh7e3Pp0iX5A89I5D01Lnk/jUveT+Mz1nuq1+tJSEjA2dmZ/Pnz55nr0IuSa7b8/L4seT+NS95P45P31LiM+X6mX7OLFCnyzNZ9MtKdCSsrK4oVK6Z2GKrInz+//DAbmbynxiXvp3HJ+2l8xnhPn/VpuchIrtny82ss8n4al7yfxifvqXEZ6/3MyjVbCqkJIYQQQgghhBAmIkm3EEIIIYQQQghhIpJ0CwDs7Oz44osvsLOzUzuUXEPeU+OS99O45P00PnlPRU6R7zXjkvfTuOT9ND55T41LjfdTCqkJIYQQQgghhBAmIiPdQgghhBBCCCGEiUjSLYQQQgghhBBCmIgk3UIIIYQQQgghhIlI0p3HTZgwgRo1auDs7Iy7uzsdO3bk1KlTaoeVa0yYMAGNRsOQIUPUDsWiXblyha5du1KoUCEcHR2pUqUK4eHhaodlkR48eMDnn3+On58fDg4OlChRgrFjx6LT6dQOzSJs27aNdu3aUaRIETQaDX/99VeGx/V6PWPGjKFIkSI4ODjQsGFDjh07pk6wIleR67VpyfXaOOR6bTxyvX555nTNlqQ7j9u6dSsDBgxgz549hIaG8uDBA5o3b05iYqLaoVm8/fv3M3PmTAICAtQOxaLdvn2bunXrYmNjw7p16zh+/DiTJk2iQIECaodmkSZOnMiMGTOYNm0aJ06c4Ntvv+W7775j6tSpaodmERITE6lcuTLTpk3L9PFvv/2WyZMnM23aNPbv34+npyfNmjUjISEhhyMVuY1cr01HrtfGIddr45Lr9cszq2u2XojHxMbG6gH91q1b1Q7FoiUkJOhLlSqlDw0N1Tdo0EA/ePBgtUOyWMOHD9fXq1dP7TByjTZt2uh79eqVYdtrr72m79q1q0oRWS5A/+effxq+1ul0ek9PT/0333xj2Hb//n29i4uLfsaMGSpEKHIzuV4bh1yvjUeu18Yl12vjUvuaLSPdIoO4uDgAChYsqHIklm3AgAG0adOGpk2bqh2KxVu1ahXVq1fnzTffxN3dnapVqzJr1iy1w7JY9erVY9OmTZw+fRqAQ4cOsWPHDlq3bq1yZJYvMjKSmJgYmjdvbthmZ2dHgwYN2LVrl4qRidxIrtfGIddr45HrtXHJ9dq0cvqabW30IwqLpdfr+fDDD6lXrx4VK1ZUOxyLtWjRIv7991/279+vdii5wvnz5wkJCeHDDz/ks88+Y9++fQwaNAg7Ozu6d++udngWZ/jw4cTFxVG2bFm0Wi1paWmMGzeOt99+W+3QLF5MTAwAHh4eGbZ7eHhw8eJFNUISuZRcr41DrtfGJddr45LrtWnl9DVbkm5hMHDgQA4fPsyOHTvUDsViXbp0icGDB7Nx40bs7e3VDidX0Ol0VK9enfHjxwNQtWpVjh07RkhIiFzEX8DixYv5448/WLBgARUqVCAiIoIhQ4ZQpEgRevTooXZ4uYJGo8nwtV6vf2KbEC9DrtcvT67XxifXa+OS63XOyKlrtiTdAoAPPviAVatWsW3bNooVK6Z2OBYrPDyc2NhYqlWrZtiWlpbGtm3bmDZtGsnJyWi1WhUjtDxeXl6UL18+w7Zy5cqxfPlylSKybJ988gmffvopb731FgCVKlXi4sWLTJgwQS7iL8nT0xNQPj338vIybI+NjX3ik3QhXpRcr41DrtfGJ9dr45LrtWnl9DVb1nTncXq9noEDB7JixQr++ecf/Pz81A7JojVp0oQjR44QERFhuFWvXp0uXboQEREhF/AXULdu3Sfa4pw+fZrixYurFJFlS0pKwsoq469+rVYrLUiMwM/PD09PT0JDQw3bUlJS2Lp1K4GBgSpGJnIDuV4bl1yvjU+u18Yl12vTyulrtox053EDBgxgwYIFrFy5EmdnZ8P6BhcXFxwcHFSOzvI4Ozs/sb7OycmJQoUKybq7FzR06FACAwMZP348nTp1Yt++fcycOZOZM2eqHZpFateuHePGjcPHx4cKFSpw8OBBJk+eTK9evdQOzSLcvXuXs2fPGr6OjIwkIiKCggUL4uPjw5AhQxg/fjylSpWiVKlSjB8/HkdHR9555x0Voxa5gVyvjUuu18Yn12vjkuv1yzOra7bR66ELiwJkeps7d67aoeUa0oLk5a1evVpfsWJFvZ2dnb5s2bL6mTNnqh2SxYqPj9cPHjxY7+Pjo7e3t9eXKFFCP3LkSH1ycrLaoVmEzZs3Z/o7s0ePHnq9XmlB8sUXX+g9PT31dnZ2+vr16+uPHDmibtAiV5DrtenJ9frlyfXaeOR6/fLM6Zqt0ev1euOn8kIIIYQQQgghhJA13UIIIYQQQgghhIlI0i2EEEIIIYQQQpiIJN1CCCGEEEIIIYSJSNIthBBCCCGEEEKYiCTdQgghhBBCCCGEiUjSLYQQQgghhBBCmIgk3UIIIYQQQgghhIlI0i2EEEIIIYQQQpiIJN1CCLOh0Wj466+/1A5DCCGEEM8h12whsk6SbiEEAD179kSj0Txxa9mypdqhCSGEEOIxcs0WwrJYqx2AEMJ8tGzZkrlz52bYZmdnp1I0QgghhHgauWYLYTlkpFsIYWBnZ4enp2eGm6urK6BMIwsJCaFVq1Y4ODjg5+fH0qVLMzz/yJEjNG7cGAcHBwoVKkTfvn25e/duhn3mzJlDhQoVsLOzw8vLi4EDB2Z4/MaNG7z66qs4OjpSqlQpVq1aZdoXLYQQQlgguWYLYTkk6RZCZNmoUaN4/fXXOXToEF27duXtt9/mxIkTACQlJdGyZUtcXV3Zv38/S5cuJSwsLMMFOiQkhAEDBtC3b1+OHDnCqlWr8Pf3z3COL7/8kk6dOnH48GFat25Nly5duHXrVo6+TiGEEMLSyTVbCDOiF0IIvV7fo0cPvVar1Ts5OWW4jR07Vq/X6/WAvl+/fhmeU6tWLf3777+v1+v1+pkzZ+pdXV31d+/eNTy+Zs0avZWVlT4mJkav1+v1RYoU0Y8cOfKpMQD6zz//3PD13bt39RqNRr9u3TqjvU4hhBDC0sk1WwjLImu6hRAGjRo1IiQkJMO2ggULGu7XqVMnw2N16tQhIiICgBMnTlC5cmWcnJwMj9etWxedTsepU6fQaDRcvXqVJk2aPDOGgIAAw30nJyecnZ2JjY190ZckhBBC5EpyzRbCckjSLYQwcHJyemLq2PNoNBoA9Hq94X5m+zg4OGTpeDY2Nk88V6fTZSsmIYQQIreTa7YQlkPWdAshsmzPnj1PfF22bFkAypcvT0REBImJiYbHd+7ciZWVFaVLl8bZ2RlfX182bdqUozELIYQQeZFcs4UwHzLSLYQwSE5OJiYmJsM2a2tr3NzcAFi6dCnVq1enXr16zJ8/n3379jF79mwAunTpwhdffEGPHj0YM2YM169f54MPPqBbt254eHgAMGbMGPr164e7uzutWrUiISGBnTt38sEHH+TsCxVCCCEsnFyzhbAcknQLIQzWr1+Pl5dXhm1lypTh5MmTgFKldNGiRfTv3x9PT0/mz59P+fLlAXB0dGTDhg0MHjyYGjVq4OjoyOuvv87kyZMNx+rRowf379/nhx9+4OOPP8bNzY033ngj516gEEIIkUvINVsIy6HR6/V6tYMQQpg/jUbDn3/+SceOHdUORQghhBDPINdsIcyLrOkWQgghhBBCCCFMRJJuIYQQQgghhBDCRGR6uRBCCCGEEEIIYSIy0i2EEEIIIYQQQpiIJN1CCCGEEEIIIYSJSNIthBBCCCGEEEKYiCTdQgghhBBCCCGEiUjSLYQQQgghhBBCmIgk3UIIIYQQQgghhIlI0i2EEEIIIYQQQpiIJN1CCCGEEEIIIYSJSNIthBBCCCGEEEKYyP8B9geWrXnI10QAAAAASUVORK5CYII=", "text/plain": [ - "(0.1747902406309579, tensor(0.9477, dtype=torch.float64))" + "<Figure size 1000x500 with 2 Axes>" ] }, - "execution_count": 109, "metadata": {}, - "output_type": "execute_result" + "output_type": "display_data" + }, + { + "ename": "NameError", + "evalue": "name 'train_losses' is not defined", + "output_type": "error", + "traceback": [ + "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[1;31mNameError\u001b[0m Traceback (most recent call last)", + "\u001b[1;32mc:\\Users\\ZINEB\\Documents\\GitHub\\mod_4_6-td2\\TD2 Deep Learning.ipynb Cell 54\u001b[0m line \u001b[0;36m3\n\u001b[0;32m <a href='vscode-notebook-cell:/c%3A/Users/ZINEB/Documents/GitHub/mod_4_6-td2/TD2%20Deep%20Learning.ipynb#Y210sZmlsZQ%3D%3D?line=26'>27</a>\u001b[0m model, epoch_time \u001b[39m=\u001b[39m train_model(\n\u001b[0;32m <a href='vscode-notebook-cell:/c%3A/Users/ZINEB/Documents/GitHub/mod_4_6-td2/TD2%20Deep%20Learning.ipynb#Y210sZmlsZQ%3D%3D?line=27'>28</a>\u001b[0m model, criterion, optimizer_conv, exp_lr_scheduler, num_epochs\u001b[39m=\u001b[39m\u001b[39m10\u001b[39m\n\u001b[0;32m <a href='vscode-notebook-cell:/c%3A/Users/ZINEB/Documents/GitHub/mod_4_6-td2/TD2%20Deep%20Learning.ipynb#Y210sZmlsZQ%3D%3D?line=28'>29</a>\u001b[0m )\n\u001b[0;32m <a href='vscode-notebook-cell:/c%3A/Users/ZINEB/Documents/GitHub/mod_4_6-td2/TD2%20Deep%20Learning.ipynb#Y210sZmlsZQ%3D%3D?line=30'>31</a>\u001b[0m \u001b[39m# Plot the curves\u001b[39;00m\n\u001b[1;32m---> <a href='vscode-notebook-cell:/c%3A/Users/ZINEB/Documents/GitHub/mod_4_6-td2/TD2%20Deep%20Learning.ipynb#Y210sZmlsZQ%3D%3D?line=31'>32</a>\u001b[0m plot_curves(train_losses, val_losses, train_accuracies, val_accuracies, num_epochs\u001b[39m=\u001b[39m\u001b[39m10\u001b[39m)\n", + "\u001b[1;31mNameError\u001b[0m: name 'train_losses' is not defined" + ] } ], "source": [ - "# Download a pre-trained ResNet18 model and freeze its weights\n", - "model = torchvision.models.resnet18(pretrained=True)\n", "\n", - "# Replace the final fully connected layer with two layers and dropout\n", - "num_ftrs = model.fc.in_features\n", - "model.fc = nn.Sequential(\n", - " nn.Linear(num_ftrs, 256), # Add a middle layer with 256 units\n", - " nn.ReLU(), # ReLU activation for the middle layer\n", - " nn.Dropout(0.5), # Dropout with a probability of 0.5\n", - " nn.Linear(256, 2), # Output layer with 2 classes\n", - ")\n", - "\n", - "# Send the model to the GPU\n", - "model = model.to(device)\n", + "# Function to plot training and validation curves\n", + "def plot_curves(train_losses, val_losses, train_accuracies, val_accuracies, num_epochs):\n", + " plt.figure(figsize=(10, 5))\n", "\n", - "# Set the loss function\n", - "criterion = nn.CrossEntropyLoss()\n", + " # Plot losses\n", + " plt.subplot(1, 2, 1)\n", + " plt.plot(range(1, num_epochs + 1), train_losses, label=\"Training Loss\")\n", + " plt.plot(range(1, num_epochs + 1), val_losses, label=\"Validation Loss\")\n", + " plt.title(\"Training and Validation Loss\")\n", + " plt.xlabel(\"Epoch\")\n", + " plt.ylabel(\"Loss\")\n", + " plt.legend()\n", "\n", - "# Observe that all parameters are being optimized\n", - "optimizer_all = optim.SGD(model.parameters(), lr=0.001, momentum=0.9)\n", - "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_all, step_size=7, gamma=0.1)\n", + " # Plot accuracies\n", + " plt.subplot(1, 2, 2)\n", + " plt.plot(range(1, num_epochs + 1), train_accuracies, label=\"Training Accuracy\")\n", + " plt.plot(range(1, num_epochs + 1), val_accuracies, label=\"Validation Accuracy\")\n", + " plt.title(\"Training and Validation Accuracy\")\n", + " plt.xlabel(\"Epoch\")\n", + " plt.ylabel(\"Accuracy\")\n", + " plt.legend()\n", "\n", - "# Train the model with the new architecture\n", - "model, epoch_time = train_model(model, criterion, optimizer_all, exp_lr_scheduler, num_epochs=10)\n", + " plt.tight_layout()\n", + " plt.show()\n", "\n", - "# Evaluate the model on the test set\n", - "test_dataloader = torch.utils.data.DataLoader(\n", - " image_datasets[\"val\"], batch_size=4, shuffle=False, num_workers=4\n", + "# Train the model\n", + "model, epoch_time = train_model(\n", + " model, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=10\n", ")\n", - "eval_model(model, test_dataloader, criterion)\n" + "\n", + "# Plot the curves\n", + "plot_curves(train_losses, val_losses, train_accuracies, val_accuracies, num_epochs=10)\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + " -the code conducts fine-tuning on a pre-trained ResNet18 model using the Hymenoptera dataset for binary classification. Over the 10 training epochs: \n", + " The initial epoch reveals moderate training accuracy and lower validation accuracy. \n", + " Subsequent epochs exhibit improvements in both training and validation accuracy, peaking at 94.77% in validation Training loss fluctuates, yet the model effectively generalizes to the validation set." + ] + }, + { + "cell_type": "markdown", + "id": "bbd48800", + "metadata": {}, + "source": [ + "Experiments:\n", + "Study the code and the results obtained.\n", + "\n", + "Modify the code and add an \"eval_model\" function to allow\n", + "the evaluation of the model on a test set (different from the learning and validation sets used during the learning phase). Study the results obtained.\n", + "\n", + "Now modify the code to replace the current classification layer with a set of two layers using a \"relu\" activation function for the middle layer, and the \"dropout\" mechanism for both layers. Renew the experiments and study the results obtained.\n", + "\n", + "Apply ther quantization (post and quantization aware) and evaluate impact on model size and accuracy." ] }, { @@ -2749,6 +2738,137 @@ ")\n" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + " -It is noticeable that the train accuracy improved after adding eval_model and the loss values are significantly better\n" + ] + }, + { + "cell_type": "code", + "execution_count": 130, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch 1/10\n", + "----------\n", + "train Loss: 0.6302 Acc: 0.6148\n", + "val Loss: 0.3467 Acc: 0.8758\n", + "\n", + "Epoch 2/10\n", + "----------\n", + "train Loss: 0.4843 Acc: 0.7787\n", + "val Loss: 0.2343 Acc: 0.9346\n", + "\n", + "Epoch 3/10\n", + "----------\n", + "train Loss: 0.3667 Acc: 0.8566\n", + "val Loss: 0.1599 Acc: 0.9477\n", + "\n", + "Epoch 4/10\n", + "----------\n", + "train Loss: 0.4856 Acc: 0.7910\n", + "val Loss: 0.1757 Acc: 0.9346\n", + "\n", + "Epoch 5/10\n", + "----------\n", + "train Loss: 0.3042 Acc: 0.8730\n", + "val Loss: 0.2384 Acc: 0.9150\n", + "\n", + "Epoch 6/10\n", + "----------\n", + "train Loss: 0.3596 Acc: 0.8566\n", + "val Loss: 0.1889 Acc: 0.9412\n", + "\n", + "Epoch 7/10\n", + "----------\n", + "train Loss: 0.3749 Acc: 0.8074\n", + "val Loss: 0.2144 Acc: 0.9150\n", + "\n", + "Epoch 8/10\n", + "----------\n", + "train Loss: 0.2367 Acc: 0.9180\n", + "val Loss: 0.2013 Acc: 0.9216\n", + "\n", + "Epoch 9/10\n", + "----------\n", + "train Loss: 0.2649 Acc: 0.8893\n", + "val Loss: 0.1969 Acc: 0.9216\n", + "\n", + "Epoch 10/10\n", + "----------\n", + "train Loss: 0.3034 Acc: 0.8770\n", + "val Loss: 0.1970 Acc: 0.9281\n", + "\n", + "Training complete in 8m 52s\n", + "Best val Acc: 0.947712\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "<Figure size 1000x500 with 2 Axes>" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Test Loss: 0.1599 Test Acc: 0.9477\n" + ] + }, + { + "data": { + "text/plain": [ + "tensor(0.9477, dtype=torch.float64)" + ] + }, + "execution_count": 130, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# Download a pre-trained ResNet18 model and freeze its weights\n", + "model = torchvision.models.resnet18(pretrained=True)\n", + "\n", + "# Replace the final fully connected layer with two layers and dropout\n", + "num_ftrs = model.fc.in_features\n", + "model.fc = nn.Sequential(\n", + " nn.Linear(num_ftrs, 256), # Add a middle layer with 256 units\n", + " nn.ReLU(), # ReLU activation for the middle layer\n", + " nn.Dropout(0.5), # Dropout with a probability of 0.5\n", + " nn.Linear(256, 2), # Output layer with 2 classes\n", + ")\n", + "\n", + "# Send the model to the GPU\n", + "model = model.to(device)\n", + "\n", + "# Set the loss function\n", + "criterion = nn.CrossEntropyLoss()\n", + "\n", + "# Observe that all parameters are being optimized\n", + "optimizer_all = optim.SGD(model.parameters(), lr=0.001, momentum=0.9)\n", + "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_all, step_size=7, gamma=0.1)\n", + "\n", + "# Train the model with the new architecture\n", + "model, epoch_time = train_model(model, criterion, optimizer_all, exp_lr_scheduler, num_epochs=10)\n", + "\n", + "# Evaluate the model on the test set\n", + "test_dataloader = torch.utils.data.DataLoader(\n", + " image_datasets[\"val\"], batch_size=4, shuffle=False, num_workers=4\n", + ")\n", + "eval_model(model, test_dataloader, criterion)\n" + ] + }, { "cell_type": "markdown", "id": "04a263f0",