{"cells":[{"cell_type":"markdown","metadata":{"id":"3JhHIQqNe4Qs"},"source":["# Lecture 16 - Convolutional Neural Networks with Keras-TensorFlow"]},{"cell_type":"markdown","metadata":{"id":"kPG5INqg-_qn"},"source":["[![View notebook on Github](https://img.shields.io/static/v1.svg?logo=github&label=Repo&message=View%20On%20Github&color=lightgrey)](https://github.com/avakanski/Fall-2024-Applied-Data-Science-with-Python/blob/main/docs/Lectures/Theme_3-Model_Engineering/Lecture_16-ConvNets/Lecture_16-ConvNets.ipynb)\n","[![Open In Collab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/avakanski/Fall-2024-Applied-Data-Science-with-Python/blob/main/docs/Lectures/Theme_3-Model_Engineering/Lecture_16-ConvNets/Lecture_16-ConvNets.ipynb)"]},{"cell_type":"markdown","metadata":{"id":"mU2KzrFhAmHe"},"source":[""]},{"cell_type":"markdown","metadata":{"id":"iEkmemKte4Qv"},"source":["- [16.1 Introduction to Convolutional Neural Networks](#16.1-introduction-to-convolutional-neural-networks)\n","- [16.2 Loading the Dataset](#16.2-loading-the-dataset)\n","- [16.3 Creating, Training, and Evaluating a CNN Model](#16.3-creating,-training,-and-evaluating-a-cnn-model)\n","- [16.4 Introduce a Validation Dataset](#16.4-introduce-a-validation-dataset)\n","- [16.5 Dropout Layers](#16.5-dropout-layers)\n","- [16.6 Batch Normalization](#16.6-batch-normalization)\n","- [16.7 Data Augmentation](#16.7-data-augmentation)\n","- [16.8 Transfer Learning](#16.8-transfer-learning)\n","- [References](#references)"]},{"cell_type":"markdown","metadata":{"id":"gTFgm_bJe4Qw"},"source":["## 16.1 Introduction to Convolutional Neural Networks "]},{"cell_type":"markdown","metadata":{"id":"vntv_OZeLW9X"},"source":["**Convolutional Neural Networks (CNNs)**, also known as **ConvNets** have accelerated various computer vision tasks, such as image recognition and classification, image segmentation, and object detection. CNNs have been used in related applications, including autonomous vehicles, medical image diagnosis, intelligent robots, and others.\n","\n","A typical architecture of CNN is shown below and consists of 3 main types of layers: convolutional layers, pooling layers, and fully-connected (dense) layers. CNNs typically have multiple blocks of convolutional and pooling layers, followed by fully-connected layers. The optimal number of layers is task-dependent, and it is a hyperparameter that needs to be tuned by the user during the training and model selection."]},{"cell_type":"markdown","metadata":{"id":"PJCris5Cb1_g"},"source":["\n","\n","*Figure: Architecture of a CNN.*\n","\n","Each layer in CNNs transforms the input images into a higher-level representation. The initial layers in CNNs learn low-level features (such as edges and lines), the middle layers learn mid-level features (e.g., textures and parts), and the last layers learn high-level features (such as objects). The fully-connected layers use the high-level features to classify the input image."]},{"cell_type":"markdown","metadata":{"id":"ElfqfEcdoMGI"},"source":["\n","\n","*Figure: Extracted features in a CNN.*"]},{"cell_type":"markdown","metadata":{"id":"kMvSZgqZmHcD"},"source":["### Convolutional Layers\n","\n","The process of convolution is to pass a convolutional filter to each pixel in an image, multiply the corresponding pixels, and calculate the sum. This process is repeated until the filter is slid over all image pixels.\n","\n","\n","\n","*Figure: Convolution process.*"]},{"cell_type":"markdown","metadata":{"id":"0nPXLn9g5zY7"},"source":["Modern deep learning libraries, such as TensorFlow and PyTorch, allow to create convolutional layers in one line of code, as follows.\n","\n","\n"," tf.keras.layers.Conv2D(.....)\n","\n","\n","Beside 2D convolutional filters that are used for processing images, 3D convolutional layers are used for analyzing videos, and 1D convolutional layers are used for analyzing time-series data.\n","\n","The output of a convolutional layer is high-dimensional feature maps, and its dimension depends on the number of convolutional filters used in the layer. For example, if the layer has 32 filters, then we will have 32 feature maps at the output.\n","\n","### Pooling layers\n","\n","Pooling layers are used to reduce the size of the feature maps output by convolutional layers, which reduces the number of network parameters and the computational cost for training the model. Pooling layers are also helpful for selecting more imporant information from the feature maps in the previous layer, they reduce the sensitivity of the model to the exact location of objects in inputs images, and they reduce the impact of noise in input data.\n","\n","There are different pooling options, although the most common is the Maxpooling layer, which reduces the image sizes by keeping the pixels with the maximum intensity.\n","\n","Implementing a pooling layer in most frameworks is very simple. The following figure depicts Maxpooling with a pool_size=2, which means that the output is the maximum value for each square of 2x2 pixels.\n","\n","\n"," tf.keras.layers.MaxPooling2D(...)\n","\n","\n","\n","\n","*Figure: Pooling layer with a pool_size=2.*\n","\n","The portion of the network that consists of convolutional and pooling layers is often called an **encoder**, since it is used for encoding the information in input data into a representation that is more useful for the task at hand. Similarly, the portion of the network that consists of fully-connected layers is referred to as *classifier* (*head*, *top*) of the network."]},{"cell_type":"markdown","metadata":{"id":"IfklXOJdmwvS"},"source":["\n","### Fully-connected Layers (Densely-connected Layers)\n","\n","The last layers in ConvNets are fully-connected layers, that match the produced feature maps from the previous layers to the labels of the original image.\n","\n","\n","```\n","tf.keras.layers.Dense(....)\n","```"]},{"cell_type":"markdown","metadata":{"id":"vTEkWfuxAg1d"},"source":["## 16.2 Loading the Dataset "]},{"cell_type":"markdown","metadata":{"id":"C56dWwwDkeJ1"},"source":["In this notebook, we are going to use one of the most well-known datasets for image classification called `CIFAR-10`.\n","\n","CIFAR-10 consists of 60,000 color images in 10 categories. The 10 classes are: `airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck`.\n","\n","Each class contains 6,000 images. In the dataset, 50,000 images are allocated for training, and 10,000 for testing. The top accuracy on the test dataset is about 90%.\n","\n","You can learn more about the dataset [here](https://www.cs.toronto.edu/%7Ekriz/cifar.html).\n","\n","Note that there is a larger version with 100 classes called CIFAR-100.\n","\n","CIFAR-10 is available in Keras built-in datasets, so we can simply load the data by using the helper function `keras.datasets.cifar10.load_data()`."]},{"cell_type":"code","execution_count":1,"metadata":{"executionInfo":{"elapsed":7812,"status":"ok","timestamp":1731443892536,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"SZ2qJTZms7fl"},"outputs":[],"source":["# import required libraries\n","import tensorflow as tf\n","from tensorflow import keras\n","\n","import numpy as np\n","import matplotlib.pyplot as plt"]},{"cell_type":"code","execution_count":2,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":6,"status":"ok","timestamp":1731443892537,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"x_ley-yjto1E","outputId":"66c3c169-18f8-4023-fda9-6cb8e6fea8c1"},"outputs":[{"output_type":"stream","name":"stdout","text":["TensorFlow version:2.17.0\n"]}],"source":["# Print the version of tf\n","print(\"TensorFlow version:{}\".format(tf.__version__))"]},{"cell_type":"code","execution_count":3,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":15198,"status":"ok","timestamp":1731443910141,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"0I77B_touxef","outputId":"12cf381b-523f-4c54-e6b3-340f065e23e0"},"outputs":[{"output_type":"stream","name":"stdout","text":["Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n","\u001b[1m170498071/170498071\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 0us/step\n"]}],"source":["# Load the train and test images and labels\n","(train_data, train_label), (test_data, test_label) = keras.datasets.cifar10.load_data()"]},{"cell_type":"markdown","metadata":{"id":"h1D1boUL4FCX"},"source":["As usual, we can inspect the shapes of the training and testing datasets.\n","\n","Note that each image is a 32x32x3 tensor, having a size of 32x32 pixels and 3 channels corresponding to the RGB (red-green-blue) channels."]},{"cell_type":"code","execution_count":4,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":4,"status":"ok","timestamp":1731443910142,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"90ULfCPD4I9D","outputId":"d8598995-8601-4852-b16e-e521b2e940ce"},"outputs":[{"output_type":"stream","name":"stdout","text":["Training images (50000, 32, 32, 3)\n","Training labels (50000, 1)\n","Testing images (10000, 32, 32, 3)\n","Testing labels (10000, 1)\n"]}],"source":["print('Training images', train_data.shape)\n","print('Training labels', train_label.shape)\n","print('Testing images', test_data.shape)\n","print('Testing labels', test_label.shape)"]},{"cell_type":"markdown","metadata":{"id":"a5D5z_0kjLUy"},"source":["### Data Preprocessing\n","\n","The values for the pixel intensities in the images are in the range between 0 and 255. The type is `uint8` which stands for unsigned integer with 8 bits, that is, the possible values are between $2^0=1$ and $2^8=256$. This means that each pixel has an intensity value in that range, where 0 is a black pixel, 255 is a white pixel, and all other colors are in between."]},{"cell_type":"code","execution_count":5,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":232,"status":"ok","timestamp":1731443910371,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"euPUiiMZnctz","outputId":"81cef021-3299-4d38-b5e0-d5843bfed471"},"outputs":[{"output_type":"stream","name":"stdout","text":["Max pixel value 255\n","Min pixel value 0\n","Average pixel value 120.70756512369792\n","Data type uint8\n"]}],"source":["# Display the range of images\n","print('Max pixel value', np.max(train_data))\n","print('Min pixel value', np.min(train_data))\n","print('Average pixel value', np.mean(train_data))\n","print('Data type', train_data[0].dtype)"]},{"cell_type":"markdown","metadata":{"id":"S4yw3K7Ktadx"},"source":["When processing image data with neural networks, the pixel values are commonly normalized to the [0, 1] range. This allows the networks to train faster and it usually leads to better results. Let's normalize the images by dividing them with the maximum intensity of 255, and check again if the scaled data look correct. Note that the data type after the normalization is `float64`."]},{"cell_type":"code","execution_count":6,"metadata":{"executionInfo":{"elapsed":520,"status":"ok","timestamp":1731443910888,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"NZdRD_HRh7l7"},"outputs":[],"source":["# Normalize the images\n","train_data = train_data / 255\n","test_data = test_data / 255"]},{"cell_type":"code","execution_count":7,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":248,"status":"ok","timestamp":1731443911134,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"bgwijQULiPw5","outputId":"a4a058f6-fcbf-42cd-85d2-5d6745a3721e"},"outputs":[{"output_type":"stream","name":"stdout","text":["Max pixel value 1.0\n","Min pixel value 0.0\n","Average pixel value 0.4733630004850874\n","Data type float64\n"]}],"source":["# Display the range of images (to make sure they are in the [0, 1] range)\n","print('Max pixel value', np.max(train_data))\n","print('Min pixel value', np.min(train_data))\n","print('Average pixel value', np.mean(train_data))\n","print('Data type', train_data[0].dtype)"]},{"cell_type":"markdown","metadata":{"id":"5YCLmkNjDqWM"},"source":[" The labels for the 10 classes in CIFAR-10 are shown below.\n"," \n","| Label | Description |\n","| ----- | ----------- |\n","| 0 | airplane |\n","| 1 | automobile |\n","| 2 | bird |\n","| 3 | cat |\n","| 4 | deer |\n","| 5 | dog |\n","| 6 | frog |\n","| 7 | horse |\n","| 8 | ship |\n","| 9 | truck |\n"]},{"cell_type":"markdown","metadata":{"id":"YSy_y4nAt-xI"},"source":["Let's inspect the first 10 values in the labels. We can see that each label corresponds to one of the 10 categories.\n","\n","Also, in the next cell we plotted the first 100 labels, and we can notice that they have values between 0 and 9, so everything looks good."]},{"cell_type":"code","execution_count":8,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":9,"status":"ok","timestamp":1731443911134,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"SMZDOLzuihpG","outputId":"ec2032cc-6487-4931-8be7-96c1bea22a6c"},"outputs":[{"output_type":"stream","name":"stdout","text":["[[6]\n"," [9]\n"," [9]\n"," [4]\n"," [1]\n"," [1]\n"," [2]\n"," [7]\n"," [8]\n"," [3]]\n"]}],"source":["print(train_label[:10])"]},{"cell_type":"code","execution_count":9,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":430},"executionInfo":{"elapsed":220,"status":"ok","timestamp":1731443911349,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"QNzq_v1Pnnec","outputId":"671c014c-7f83-43c2-bc53-7e01e8f2a3d6"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["plt.plot(train_label[:100], 'o')\n","plt.show()"]},{"cell_type":"markdown","metadata":{"id":"FLxJ51fl2NG6"},"source":["Let's visualize some images to see what they look like. We can notice that the resolution is quite low, since they are 32x32 pixels."]},{"cell_type":"code","execution_count":10,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":752},"executionInfo":{"elapsed":804,"status":"ok","timestamp":1731443912144,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"y5xezSpVicHx","outputId":"c0b9eb90-ae11-456b-bba2-723eaa5e5c8f"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAA3AAAALfCAYAAAAg1yBYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAC5OklEQVR4nOz9eZxlVX3v/7/3mWuu6uqq6pGeGRoaISAoYEBFWwUTFANi/GoHgj5QgzFXo95rRG9uYkw0oqgRNQxeYq4yXK9RRI2ARoJMMkPT8zzWPJ9x//4g1M+2P5/VVUXb9O5+PR8PHw/5nFprD2eftfeq0/VeURzHsQAAAAAAh73US70DAAAAAIDJYQIHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAncQbNq0SVEU6XOf+9xB6/Pee+9VFEW69957p9X+vPPO00knnXTQ9gcADrYkj50v7PtNN900re28IIoifepTn3pRfQBIpiSPgXhpHbUTuJtuuklRFOnhhx9+qXcFABKDsRPA0YwxEIeDzEu9AwAAJM2CBQs0NjambDb7Uu8KAOAoc9R+A4cXr1KpqFQqvdS7AQCHXBRFKhQKSqfTwZ8bGRk5RHsEAMnA8+OLxwQuoFQq6ZOf/KROO+00tbS0qKGhQa961at0zz33uG2+8IUvaMGCBaqrq9O5556rp556ar+fWb16td72trdpxowZKhQKOv300/X973//gPszOjqq1atXq7u7e9LH8Mwzz+jVr3616uvrNXfuXP393//9fj+zZ88eXXHFFerq6lKhUNDLXvYy3Xzzzfv8zG/+O+1rr71WS5YsUT6f1zPPPCNJuu6663TiiSeqvr5ebW1tOv300/Xtb397nz62b9+uyy+/XF1dXcrn8zrxxBN1ww03TPpYACTDkTB2PvLIIzrrrLNUV1enRYsW6Wtf+9o+r1t/A7dq1So1NjZq/fr1etOb3qSmpib98R//sSSpWCzqQx/6kDo6OtTU1KQ/+IM/0LZt2ya9PwCS40gYA3l+PLzxTygDBgcH9c1vflOXXXaZrrzySg0NDemf//mftXLlSj344IM65ZRT9vn5b33rWxoaGtL73/9+jY+P64tf/KJe85rX6Mknn1RXV5ck6emnn9bZZ5+tuXPn6mMf+5gaGhr03e9+VxdddJFuv/12veUtb3H358EHH9SrX/1qXXPNNZP6o/e+vj694Q1v0Fvf+lZdcskluu222/TRj35UK1as0Bvf+EZJ0tjYmM477zytW7dOH/jAB7Ro0SLdeuutWrVqlfr7+/XBD35wnz5vvPFGjY+P6z3veY/y+bxmzJihb3zjG7r66qv1tre9TR/84Ac1Pj6uJ554Qg888IDe8Y53SJJ2796tV7ziFYqiSB/4wAfU0dGhH/3oR7riiis0ODioP//zP5/8GwPgsHYkjJ1vetObdMkll+iyyy7Td7/7XV111VXK5XK6/PLLg20rlYpWrlypc845R5/73OdUX18vSfrTP/1T3XLLLXrHO96hs846S3fffbcuuOCCA+4LgOQ5EsZAnh8Pc/FR6sYbb4wlxQ899JD7M5VKJS4Wi/vU+vr64q6urvjyyy+fqG3cuDGWFNfV1cXbtm2bqD/wwAOxpPhDH/rQRO21r31tvGLFinh8fHyiVqvV4rPOOitetmzZRO2ee+6JJcX33HPPfrVrrrnmgMd37rnnxpLib33rWxO1YrEYz5o1K7744osnatdee20sKb7lllsmaqVSKX7lK18ZNzY2xoODg/scY3Nzc7xnz559tvWHf/iH8YknnhjcnyuuuCKePXt23N3dvU/97W9/e9zS0hKPjo4e8JgAvPSOlrHz85///EStWCzGp5xyStzZ2RmXSqV99v3GG2+c+Ll3v/vdsaT4Yx/72D59PvbYY7Gk+H3ve98+9Xe84x2T3i8Ah4ejZQzk+fHwxj+hDEin08rlcpKkWq2m3t5eVSoVnX766fr1r3+9389fdNFFmjt37sR/n3HGGTrzzDN15513SpJ6e3t1991365JLLtHQ0JC6u7vV3d2tnp4erVy5UmvXrtX27dvd/TnvvPMUx/GkI6cbGxv1zne+c+K/c7mczjjjDG3YsGGiduedd2rWrFm67LLLJmrZbFZXX321hoeH9fOf/3yfPi+++GJ1dHTsU2ttbdW2bdv00EMPmfsRx7Fuv/12vfnNb1YcxxPH3d3drZUrV2pgYMA8nwCSKeljZyaT0Xvf+96J/87lcnrve9+rPXv26JFHHjlg+6uuumqf/37hOK6++up96kftb46BI1zSx0CeHw9/TOAO4Oabb9bJJ5+sQqGg9vZ2dXR06Ic//KEGBgb2+9lly5btVzv22GO1adMmSdK6desUx7H+6q/+Sh0dHfv875prrpH0/L8nPljmzZunKIr2qbW1tamvr2/ivzdv3qxly5Ypldr3UjjhhBMmXv9NixYt2m87H/3oR9XY2KgzzjhDy5Yt0/vf/37dd999E6/v3btX/f39+vrXv77fcf/Jn/yJpIN73ABeekkeO+fMmaOGhob99kfSxD55MpmM5s2bt09t8+bNSqVSWrJkyT7144477sXvLIDDUpLHQJ4fD3/8DVzALbfcolWrVumiiy7SRz7yEXV2diqdTuszn/mM1q9fP+X+arWaJOnDH/6wVq5caf7M0qVLX9Q+/yYvHS2O42n3WVdXt1/thBNO0HPPPacf/OAHuuuuu3T77bfrq1/9qj75yU/q05/+9MRxv/Od79S73/1us9+TTz552vsE4PCS9LHzxcjn8/s90AA4uiR9DOT58fDHBC7gtttu0+LFi3XHHXfs85uIF37b8dvWrl27X23NmjVauHChJGnx4sWSnv+K+fzzzz/4OzwNCxYs0BNPPKFarbbPQ8fq1asnXp+MhoYGXXrppbr00ktVKpX01re+VX/zN3+jj3/84xOpa9Vq9bA5bgC/O0kfO3fs2KGRkZF9voVbs2aNJE3s01QsWLBAtVpN69ev3+dbt+eee+5F7yuAw0/Sx8DJ4PnxpcWvCQNe+A3Eb/7G4YEHHtD9999v/vz3vve9ff4N8oMPPqgHHnhgIrGns7NT5513nq6//nrt3Llzv/Z79+4N7s90YmAP5E1vepN27dql73znOxO1SqWi6667To2NjTr33HMP2EdPT88+/53L5bR8+XLFcaxyuax0Oq2LL75Yt99+uxmLe6DjBpAsSR87K5WKrr/++on/LpVKuv7669XR0aHTTjttUn38pheO40tf+tI+9WuvvXbKfQE4/CV9DJwMnh9fWkf9N3A33HCD7rrrrv3qH/zgB3XhhRfqjjvu0Fve8hZdcMEF2rhxo772ta9p+fLlGh4e3q/N0qVLdc455+iqq65SsVjUtddeq/b2dv3lX/7lxM985Stf0TnnnKMVK1boyiuv1OLFi7V7927df//92rZtmx5//HF3X6caAzsZ73nPe3T99ddr1apVeuSRR7Rw4ULddtttuu+++3TttdeqqanpgH28/vWv16xZs3T22Werq6tLzz77rL785S/rggsumGj/d3/3d7rnnnt05pln6sorr9Ty5cvV29urX//61/r3f/939fb2HpTjAXBoHMlj55w5c/TZz35WmzZt0rHHHqvvfOc7euyxx/T1r39d2Wx2cifoN5xyyim67LLL9NWvflUDAwM666yz9LOf/Uzr1q2bcl8ADg9H8hg4GTw/vrSO+gncP/3TP5n1VatWadWqVdq1a5euv/56/fjHP9by5ct1yy236NZbb9W99967X5t3vetdSqVSuvbaa7Vnzx6dccYZ+vKXv6zZs2dP/Mzy5cv18MMP69Of/rRuuukm9fT0qLOzU6eeeqo++clP/q4O01VXV6d7771XH/vYx3TzzTdrcHBQxx13nG688UatWrVqUn28973v1b/8y7/oH//xHzU8PKx58+bp6quv1ic+8YmJn+nq6tKDDz6o//k//6fuuOMOffWrX1V7e7tOPPFEffazn/0dHR2A35Ujeexsa2vTzTffrD/7sz/TN77xDXV1denLX/6yrrzyymn3ecMNN6ijo0P/8i//ou9973t6zWteox/+8IeaP3/+QdxzAIfKkTwGTgbPjy+tKH4xf5EIAAAAADhk+Bs4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASYtILeUdR9LvcjwNv/yXdOqYqeL04Kw/Gzguh9z7t1AvZ3FQ2LUkar5TMei3QxuuvVmN5xaPJ+963yqy3t7e7bbwVOFPOr9UqlbLbV0tLi1nPZPwhfuPGDWZ9Zoe9z5mM92mTxseLziv2QZZKFbevpqYms16tVt02c+bMMeuFjD0OuCcfh8S7Lr/6pd4FHELX/aLfrOci/+6ac4ablPNAEPpIZ52+8v6QpqY6e+xsbSyY9UzgmadSs48zStvbGC3642O54t04/IOJa/a+1Wr2mOo9i0n+tz452X015N2ulM9nzXo18o+l6DxbjZbt++OIe2+Sxnv3mPX/vOuHZr1l4YluX3NXnGbWy1X/vh05Z/N9Z3W6bX4T38ABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJMSkUyiPJJGTaxhK3jlceccSckiOM7AJL6EymsZuecFPZSexL5Qo6W5jGm1wdOnr6zXrg0ODbpt8zk5ITDkxlN42JGnBwgVmfWb7TLfNrt12AtfI6JhZz2XtxDBJGh4eMetpJ7kyCnzYd+3abdYbGxvcNnOdFMqXOj0ZgNRYsB81Zzb6Y0qjEyCbcZIrM2k/uTCXscfUbOArjKwzDqdTztgVGNO85Mg4stMmC4F0zKITajjmpVNKip19LjgpkKm0P26Wx+wdiMt2CmU6FZhmRPZrXmqnJFXL9v0pVRq1Gwx0u33d/2//x6yvefCXZv3Ck052+yo4UaeplJ+enAtcs5PBN3AAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAAS4rBaRuBgBz5PNWI/9PMv9RID01kuINSbberH6O3XtOK7p9HGe1+q8XTWJLDLod9yJG/hCfwunH3OK8360NCw28aLxfc+O06qtSQpnbaHci/GX5Je/epz7RdiLw7Z34GaE/ucTtttYvnRyl5fOWfZhdBrcdX+hLK4AHDo1Dsf3XzKjtGXpEYnlr0+Z9ezaX9JAu/Rolb1tx9Xi2bda1ILPE5Hsvc5nXKWtArE6Ndie+xM5fxRzR27I7uv0PaVdfY5U2fWS87+StJ4yV6SoFp11kqQlCrby+k01ux77a9+8QO3rwd/dKtZnz+jyaxnR/e6fXXah6/hmn8PrpTta2yy+AYOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICF+pymUU04iDET6vdSpYQc3BfIQCJx795WDGKkYTycFchpqU9zp0LsYOV2lA+cykNWEo0hLa4tZb2xqdNv4Ca7eVeVf697nLfQxzDXVO21e2t/refeN0JjiveZ+dImPBQ6ZpryXtlhy24yX7HEo5Ty2xmn/Q+0NA7Wa36bmjin2fkWBlF7vGSLjtEkFjiXnpHBWUv7TiJcq6R1j7KRTSlKU8ZKF7b7KtUBf1XGzXhrrcdsUqnYSZL42Ztazw7vcvvLO9Tc82G/Wn7zvLrevtJMqPZKz65JU8W5Qp8xx2/wmvoEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASIjfaQrllIUiAkkNM6WcFKOakzok+WlBiUvalBRPdZcD15H324zQeUneGcPvQq1aMete+pcUzJSccgv3Gg1coLWqnQ4Wy08NO3jsJLWQKacaS9w3gMNA1nnSjGL/EbTqfN5Lzk0/EMKoTMYeb0KBuyknbTKddlIYYz9RUzX7/lB10yHtn5ekmpPqWK6W3TbFsv1auWz35f28JMWlUXu/ikNmfXio3+2ro8l+X1qy/rks9m8z66lM1qzPasm5fTXU15n1wVH7GLu3rHH7Sg9uN+vtC05w2xRTL24KxjdwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEuLwWkbgKDedmOyKEykbBebm6WksPXC4mmpK+HRWqvCWXZjWDuCIlHbyqKe1zETsXFTT6Wxa1+fBvKi9nT40SwKwzAfw0qs5jyPVkh+XX3ai7MfH7Ppg4PElcmL8M5G/fe/hOHIGomp1xO0rLhWduh2XXx4PLAlQtCPupXG3zep1m8x6tn6GWV+w9ES3r3zKfuZsydrbz6nb7au5Yp//9gY73l+SdowPmvW9g2NmvTQ+7PZVSztLatU3mPVszl9ip7nOvgBndM102wzXWEYAAAAAAI4KTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAkx6QiU6SQkHg2mdVZSTisnHVKSak6KUuvMDrO+9IST3L66d+4w61vXrTHrofc+juwUnzgOtInT3gv29uUn/7hbcbYfCrJzQ6wIocQBRM6VGAU+B6HeTFxsABJo/eO/MutNjY1um5TzbOGNqJWq/zgbV+0UyHzaT3vMOc9jhVTWrNdnnOcaSbmc3UZ1BXsbLTm3r5aMfZzN9XZfkrRn74BZr0/bZ/PYFv9Yik6qY0PWTsfMNwee3yp2m7FBP1Fz2EmujJpbnJ/f7fYl51m02G8f40jGTy297xe/MOuvX/h7/vYzftrmZPANHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgISY9DICRz0vu9aJ9k6Fll1w4mljJ9JVkn7v9Jeb9ZUX/ZHdV77B7euH3/0X+4V19sGkAoslVGrOCZjW+gpOo0O0goWX0u6H4AIAgJA9a58w64vPOcdvFNnR+2kner2W8qP3vRT/xjo/Lr8+l7f7kt0mXfQj5svOEkllZ/OpnP/dSpS2H9u95Q0kqc1Zumrzr//DbrDpUbevlg576Ycly+eb9Y457W5fe/bYyzvs2NXntpmz5HizPl61j3HvvY+4fc3snGX3NbLTrA/128sxSNKuLdvN+tY1z7htGmZ2Oq+8zG3zm/gGDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhSKGcJCdEKBBO6WUaSlknRejUM09327zhzX9g1stZO20yXfBTKF/+cjvRcu0jD5n11pYWt6/6RjsRasvmrW6bWmznOsbO2fTPpBTFhyiiEgAATFlzU71Zb2u1Ew0lSRm7TbVqPz/l0v6TQmuTnSjZ3Fhw26RTTtqk85zSu9dOLpSkgT09Zr2pYB9jZftet69MrWTWe4Z63TZDj99t1kfXPWnW93bZ6YyStOyUC8367KV2OmR9W5vbVzq2ny0Xz/Pb5Cp22uct37zerPcNDLl9nXi2/SzctqjbrPesftrtK3Kuix9971a3ze+d9Qr7hTfb5/i38Q0cAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIcXimUobjBw1QU2ckz1bjmtmlrspOXjjvOTvGRpOfWbTDri1fYyZUrXnaK21djPG7Wm520yfYOP5Fo6dLFZn3H9l1um3KpbNa9FJ9q4PcMXtpnlMSLCQCAI0z33m1mffOGJ9w28+YvMevtrbPNekernYgtSfV19rNFlLITHSWpXLKfIXI5O9Ey0+Unfzc3221S4/az0J3f+99uX73b1pv1bNZO95ak8a12m7zsNiOBdO/CzAVmva7Nfr+qefvYJal90Ry7r7yfTrrhV78w6+tWr7a33zHT7avQbm9nZsZ+5uzdZJ9HSRrotlNAB9Xvttm40U4BnSy+gQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQk15GII5/97HsXoz84cw7L6EjGRsbM+tPPPG422bWomVmfe78Y8x6W2ur21dvW5tZb2y16wOjfjxt5zEnmPX5C/241Q1rnjbrzooMB1heYlqNAADAIbB36zNmfW20123Tmhox6yfNnWHWo6r985I01D1o1isVO8Zfkqq1tFmf3TXPrKfL/tJRLY3NZr2YsZd0Ov6Vp7p9FYftuP5Czl/G4PE77zDrkbOMwMLff63b13DZfuYa6K+Y9aaOVrevVNp7TvOfOSNnVYJcR5NZH8r47/HO7RvN+vDeIbPe12MvFSBJXV2dZv3MkwPLg22xPxeTxTdwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAkx6RTKQyOUHPjSJlROdeupQAsvhfK51c+5bV7zpj8w64sXLTDrQ4MDbl/ZnB3jky3YKT6qc+qSGjvtRKSuY45126xfu9qsR7JTnELnnqxJAJ44cpLhIn/kSMXOiBPbv+88VGPQdJKgIzfaFzh0SsN2CuTWDX5y5HGL7GeLvbvs5MC9e/e4fY2M2NtvLNS5bbo3bjfrzxbtMWVs3E87XPLy0+z6qSvM+nEvO9Hta8uWnWZ94bLfc9vUNbaa9XTKTns87vdf5fbV22+fy2xUb9YjObGRkjsOl8ftbUjS4888bL/QaKdg1qftNFFJ6t2x1t5+X9Gsv2zpbLevU3/vbLO+8NST3DY9P9jsvjYZfAMHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBCHWQplEtkpOlHkz40rNTvFaM7cuW6bla9/nVnvnNlu1p9ds8bta3h41Kync3YiUznd4vb19MbdZn13n70NSYojOxUoju1EpBAy1gB4Yjcj0kmnlBQ7v9cMJQtPdfvTSZQEkiqTtz9ThfpGt83IqJ3quHHTVrO+e+c2t6/BgX6z3pTJuW3W/upRsz60bYdZz+b8x+nVzz1u1l89fKFZf/bJp9y+8nn7nI339rltHr3vF2Z9wYL5Zn0wkAK55NRTzXp9u/2cWKv4Y23KScndvmWT2+aJxx8z6+WKnUKZCiQON9TZ73+pZF+vAwN+uvvd9/zSrC8e9M9lU4v/bD0ZfAMHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgISa9jEB0UAPbvVjPwzkU/ncf+/zyl5/hvrZs2bFmfbhk71dn+wy3r+eefMLuq2LH+De21Lt97dqx3azv2LrRbVOTt1yA/f7HgRhYzzSaADji2L+jrNXspUwkKYrt15zVX6RUYLBxB6LQvY7BC0eWbMG+plMpP2J+1047rj+Xsh9bC7ms21e+rc2s//Knd7ttdq7dbNbrnTGloeY/Tqe7u836fT+6y6zv2eIvidDa3GTWh7Y957bZudF+HutZYz8L9o4W3b5e+/bLzPqr3/pHZj0V2ctBSFI8Yr+2/rGH3DbVkt2mqbnD/vnAkjG5yFk6a8iO/t8+MOz2lc7ZY/o8ZxuSlEo1uK9NBt/AAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACTEFFIoD6bDOW1yarwZcC32k2+6OmeZ9T+86C1um0LBTqvpHbTTjdKBJLMNTiJR3+CAWY/TPW5fY4O7zfpwn12XpMg7N7GTQhn4NUMUk9gGHO1id7yzb3GpKOf2NTZkJ7BVSna9scVPv3MC80TSJI4mlbhk1kdGh9w2e3fZzxAzGlvN+oqTlrp9NdfZSdoP/uw/3DbDo+NmvZxyEmwz/nNtNGSnFxbXbDDr9Sm/r+Fu+3lseNh+fpOkqGaPN/2DfWa9b8h+vyTp0fseMOunn3eOWW9pbXf7WvOA3dcjv/iZ26Z7sNesZ2t2OmeU8R8gK1HFrA8M2O/9SNUf60849kSzPmvBQrfNAw897b42GXwDBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmvYwAHJEd91oLxNsvXWbH3Z728tPdNpVq1aynUvYcfON6O55Wkp57+hmzPtpjR8qWBu1tS1Im60S3BmKynRBeectLxCwVABz1guOAMw6nUvZyAeWiHwe9ZdM2sz48ZMdXLzthjttXQ1PerGcy3Hpx9CjaK3CoOjLqtqlE9nJDg+32c0pcKbt9jQ4OOm38Z5soa48Rw2U7ej4q+eNTdcSO5a9zlh4oR35fKW9tkpo/pnlLMRWa28x6S52/jEF9rs6sl4fsJSF2DNjjpiQ98LOfmPVNm9e7bboz9rmpOUtSpLL+WJtL2cdSKtrXRc25z0hSod7uq6XNXgJMkvr6/eW2JoNv4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASgiisSXKzZ5xUmlBiWiZnpwWl034+Yz7vtbG3v3H9OrevMSeRKVO197m+zk83Gi+NmfWU/HQnZ5flnbJQ+FzspjX5aUEADl/TyZyNa3ZiXcoZU9NO+pgkFQqtZj2btm+XVWfcfP41exwkhRJHk7Ex+/NZdtIZJWnISRWc19Vl1ndu3+rvQNFOjuzeu9dtEqecZ4i8nWw7VLK3IUklJ+2ymLW/Q3ECMCVJqdjezoxso9umUnUSOp1HzpkzWty+GnP22LXxubVmvW98wO3riSceM+ulwPGrYG8/dsbacs2/xmoVe+yuOG9lXc4/x/V5+72st4OIJUkN9f4z/2TwDRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQhCFtQ8/TSxyEolqTpvZ8+a6fb3xzRea9YbGQIpQyU4R2rRhg1nfsc1PZBoZslMo02l7Pt/UWO/v19CwWS8Fs+S8hEin7iZN+u8YGZQ4HITSaCMnwfZoEDryyPtUR6HfN9qxZZWi3VdgSNGMtlazXo3t22VTi99ZOjOdTM2p4RrD4a6hxb4O9/QV3Ta9u/vNet98O9Vw/caN/vazdursWMVPKEw32fs8+5jZZn3Hxj1uXwM99jPXSMlOIcxn/OjCvPOZrhZ73TZ1zfbY1VHfYdZTgZTc3hHn/G+wk89b589x+9ot+/zn6+2kT0nyQtG9FPco7Seiq2a3GXbSKUf77PdRksZ7Npv1nk3+uZxVaPD3bRL4Bg4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCsIzAPvzI5bhmx4rm6gpm/ZLLLnX7esMFbzLroyOjbpudu7vN+pOPP27WtweWEfi9l59m1ufNtbdRjZ3cVkmnz7YjdZ997D/cNuuffsqsxzX756cTxB1Po1XE4gM4hLz496Mi+j2Q41+rVuwmVT9au7lhplnfsbPHrDc0+NHSba32mD5UtJdMSWX99+tQvJOhZQQ8R8U1hsNGKmcvF1CJ7eWRJGl4zI6Y37p9h1lvbPC/j9g2uM2sN8xsdts0Ntn1fKsdPT8nY49BktTTZO/bcO+IWR8bHXf7Gi3bn/dSYOmBOGsvBbV7uM+sZ0v2fklSZ2enWZ/ZZR9/Y0ur21eh1T7JuZS/vMTcJvs9q8vb05mas1TB8y/az7a7R+ztb6/a9wBJ2rHxabPe3mjfzySpLeMvlzAZfAMHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBCkUP6GUDCXF/SVzdopNoPDflrN+g0b7Pqan7pthofHzPripcvM+oJFC92+Xr9ypfOKfSx9/X46ZsfsLrP++AOL3DZf/dznzPr27bvMejAwzXlfSJQEXnpeGmwtkEJZHLeT6WojfpvRPXvsvip2Algu8pN127ra7O1n7HTKmuyxGcDzUs6TZpTxv0MoxXZS7EjRTmjMZf1Ev84l9vPIlj1+WncpZ49DkRP2WGj1nznmNM6wtzHcaNbHBvwxJa7a28nm/OOPnLTPiuy0yQYn0VGSmprscXB4cMCsjw4PuX0tabfPSzbykxtnNrXYL9TsNrH8pNPiuB19XknZbQYD12t9zj5nuciJV5c0XvbTRieDb+AAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEmLSKZRevo6fCzaNvg5RcKC3z6HNp51Xx0fttKCf/vgnbl9DQ3byzxO/fsxtU7MDmfTf/vIjZv0P/+BCt6+WjplmPZOtN+szxp2NS6pW7eSf0848w21zyu+dbta3b/++WU8Hfs/gvpeEUAKHjvdBTNkfxNB9Y2zYHlMGtw+6bUZ7d5v1ijN0pAMpmNFxC8x63GKPg5kGf3wKbOagiRjscJhrrLejG6OUn1BYdR56as6DYufMTrevc197jlm/467b3TY9Y91mvc45lmq65Palqj0QtBTsRMfszAa3q6hiH3+65o9DzQ32dloa7O1EZb+vihMpOtTXa9ZTFf+8zGqos9s4z5WSVB63kxtjZ7AtB7ZfLtnXWEOjnejpBE1KktI1O20yHboHlP3jnAy+gQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQk19GwInonE6AceS0CqYhH4Ko5FDkczqy57rFih0DWkhn3b52bN5i1ge6e9w2o8PDZv3mr3/NrF+26p1uX0tOXG7W45QdNVvf2Or2lSvYcatx3l6SQJKWn7TCrP/sLnvphbhcdPvyrouaE1TuXXtBoSaHICYcyXUkRbzH8TQudifyOy75vzscH7Sjnbt39rlt9mzba9ZHnMjpljp7rJOk+rS9/fYT2s16U2Oz25fsZGnFgYHDu2amdf6Bw0Bdxv68ZwOfg5Rz4005z2K9vf748PDDD5j1OLbHB0mqdx4I88WyWW/MN7p95dL2o3ZDzh6HGvP++FQZs585i4P2klaSVMjb209V7QEqFflTg0xjk1nvWnCMWd/83FNuX0OD9nvW2OCfS+9rp4yzJEM19Pgoe4mBfD5t1rN5+3lXksaK9vtS32CfL0kqjvlL40wG38ABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJMSkUyi9RKAQN4DNeSET2EbaSTM7mIIpX1NM4axV7HQbSZLz2jmvPMNtsnXDJrP+zCOPmPWbx+zUSkk69cwzzfoJLzvdrJ925tluX40tdgLbzm47FU6SWmbMMOszuzrM+u7tW92+Us61dMiy/46ckEEgKJSoGcd2mlkmstN4t2/c7fa1fYM9dqRq/u8bc3X2rSxKNZj1rB0YJkkqDY7aL1RbnBZO1KSkGsGRgPLO56Ax66d1Z72ba8X+vI2OOJ9bSdu2bjLruTr/s9voJNW25O0xpSnrJ2+raifbOoG3KpT8sbZQZ6caDlb98bEW2QPe8IidXJkv5N2+GuvrzHrXTPu5rn+7n6hZbbb7Klb9ATr20klr9nsZVf1BOJ+y0ya9lOBMxr9eh0btuMuK8x5LUpzy+5sMvoEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEJNeRkBOxH4oWjpyM9ad6PdA5HLKifU8qAKR8HHK3n4+ZTfq2bPT7Wuov8es5wInIK6WzXrkxK2uf2a129funbvMenv7LLPe8ho7Nvf57dsZqX099jFKUiZvX3YnnbLCrO/Zvd3fvnddTuNy8VeR8Dvzr3HgyBIHP1T27wK3bbY/u08+9KzbU33WjusfHhpy2/T02ssSpMbt/VrUNdvtqxbbY20m50ROB5afCa1Mc7AE78GB14BDpSHOmfUWJ6pfkvIZ+9mmVrPj2js6/eeUhbPtMaU04j+n1Gr29itV+zMVF/28+NLouFlPOeuM1PL2+ZKk8azdJg58H1PI23H13vA0VvKXV2hubTTr2Zx9vrLO/kpSU7PdVzwy4rbJR8447Lxfcdm+XiSp4jy/5uvsJSFyGX/KVF/nvC+xv6RYLv/ivkPjGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhJh0CmU6msZczwmfiZwXQoFdVWfzwZQtJwKs5tRDaWLusTibrxX95Jlx57VHH37IbZNO2W9VlHF2IJAYNzg0YNZTWbuvphY/3WnPrj1m/cFf3e+2OWbBXLN+7LHLzPov/v2nbl8pJ0VoOvFv3nURdghi5oDfAW/krDljR5z2r/XRfntMe/axdWZ97y47CVeSzjhpjlkf6dnrtimP2ClvnQU7fS6X9xPjMm32zabQbCeTheKLvVfigzhukDSJw92MOvtzONhkJwdKUmPerjc4abA5+X1lS/b4kK34n8Nx59mi5j1yOUnlklRXZ48p1bLznJj1j6VcsxMiU34T5evsFMq6OjvtspB1Tr6k/l47uTNO2fu1Zauf9JlL29uP0vb1Ikm5nP0G1NXZz8j1jX5fe3v7zHqxbB9LIe+npnrvflwdc9tEceBNmwS+gQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEmnUB7UnCsvriUwnYy9pK3AjkUpu8OM01cqHdgBJ5nNCx7KxFM/Y7VAMlnVSUhM2YFMcsqSpEqlbNa3b91i1kvlotvXU088Ztafe/YZt82b3vg6s75zvZ1YlwukQ3phmzWnHsp+SznXRfCdJAEOCeUnITrpayV/fNz07DazvmeznfI1a0an21fPHjttshxI9l2wYJFZr43bCWDVBn8k6Fo8y6xHTkpvzUmFk6SUM0BHB/eO6vISKoOJy8BBlpWdalif9VP92hsbzXpz3u6rMmI/10hSrcmuF7J1bptMwd5O3hkGy07SpSQ11tt9Vct2OmSU8x/NS05CovzNuynmqYydAhlX/fFpZMR+Hkzl7PPf1++/L31OoqUCKZjZtH0fcE6x8nV+ivreHvv+lC/Y14UzpXh++06iZybtv5eV8ou7D/ANHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIR48csITCcF00kwDkXfezuazdgxrJKUz9uxnvmcXY/S/h6UnZ3OOHPgfOT35cXVF514f0kqORG1qVrFrKcDMdH5lB0TvnuzvYzA2MCA29dQv/3aogUL3TaLjjnGrA848eEFe3clSZm0/f7HGfv4vaUlpFDkth8T7i1XABzuYme5gLQzdvXs7Hf76t8xZNYbMnZ+d/fOHW5f3U69sbnNbTM0OmzW8832naNj2Ry3r9bZ7Wa9EtnnKwrcubwxZXoYbJBMa9duN+v9/SNum3zGXmIg7SzRVBr3HxSqsf2cMDDuL01Scx4Vxiv280Cl7D9bFDP2aylnv0YCSxJs2rHLbjPqL/fU2uhE6TvjU/fgqNvXq159tlnvmN1q1rMFfwysa7KfxeubW9w2qZr9npVGB836wIBdl6T6evu8zOzoMuuDQ35f3d3282tDYBmDlPxny8ngGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhJh0CmXNCZIJZWxFTmiWl8JYF0gIbMzZaT2FfN5tk83abbxksNExP8VnYHzMrFerdopMKIWy4KRjVgNph5WynbzT4pyXtvpGt6+2Bvs8Z8bt5KN4xE54k6SZM1rNei2e57YpjtrJU/19fWZ93vz5bl+Rc8mUqnaiZxxI55yO2kHuDziYajV/TEml7DFqdMAeB9Y+udbta9sWOxktm3LGurSdnitJmZw9pg+X7KRLScpm7TbHL1lq1pcca9clKXJuXN7YEQXugm6baaVTTr3NwR7vgOn49eqtZj12Eh0lKeskTI8N2QmJu/v73b7q2+znoceeeMxtMzpmp1rWYntMKweOpblgJxHmMvYjeP+wP9YNOmnl41X/s14X2emJcztmmPVW57lOktpnd9rbr9jPdems/1x9/DI77bF95ky3TXHUfhZXbZa9X0U/abTkvGfePSiT9vvKpOxzOTLov5dxHIhYnwS+gQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEmnUFacZK50IOTKmx1m0vYrjYWC21djxk5MSweSK1NOqmPKSYjMyD8YZ5eVcVIwc04CpiRFzj4XnESi5/trNevNBXv7s1vtRBxJanISKrMtdn37to1uX06gphbMn+O2KTmJnstPOdmsL1yyyO1rdMxOPhpx6jUnNVSSopSd8pYJvS9OWhFwKE0n7XB8xE7U2rRmp1nfvbPb7Wv2Ajs1bMEyO0G2UvJTKPP5OrM+0uenebUVmsx6zhkfx/r9ZN102T5n2SY7SS6URDutsEmH/x6HWnkvevtM0iUOvqGq/TyU9qLKJZWc6POyM3ZUA4nk9Z122uGmXjvRUpJKJS/F3GkQOJb+MXvsyjhJwNXAs2g5bT901QLbr1ScFPOGerN+7NwOt6/hvh6znqm3z9dJJ65w+xrctcOsj+6165KUd8b0TMY+L6Wak1opqeI8i6pmz0Xy/mO9CgX7OXFgYMBtEwXe58ngGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJMellBNy4y1DqsPNaOmtHpza0t7hdtTbaEc75QHSs91pDvR2dGgeWJKg4fXXMnmXWZ7S1uX1lnCUGsoGlB+qb7Yj/+oIdud2Q8vuqlu0Y3p/+4h6z/v2f/tjt661vudisd3V2um3yztoDr33d+Wa9XC67fY0X7RjYYmncrJdKdpyuJNVq9hID6bR9vUoSCdo4HHjLBZQrflz/1s12VPNQnx2tffoZL3f7mn+cHdOt+qpZTtX8cTuXsSOcd2/Y7rYZ2LTbrO/YuMWs17L+ciKzlthLoMxbZt+fMjn/JhjLPv+hpQe8l0LLmXi8MS2K7Hud9/PPt7GPMzQ+hpaxwNEjn7fHgbEx/95eTtljRCm2r7etu/vdvk7N22NKV9dst832rfayKU0N9ucwF3h+bHSeeTvb28363j4/er5veGrPPJLU1mA/J+acz+6WTVvdvmbMt8f6VMH+rPcHln9R1W5THPevi5GxollvbLKfkctV+9qTpNFxu69UxWmT8sfglDOmRil/+5XAvk0G38ABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJMSkY60a6+1EoFzBTveRpPoGO+2xpcVO85o/f57b17x59mttgbTHtrZWpz7DrDc1N7t91TU2mfXWGU5fTfbPS1LBSbTM5ux0RklKFezX0hm7ng2E22zcsMGs//ujD5v1gnMeJWnJ0qVmPZ/1jyWdtn9vUFdnJyWVyn5yZKVipxWVyna6ULFo1yVpdNRO3xsbs1OfpHBCJjAtkRNDGEg8jWSnicVV/3d042N2alnnPHtMm3ucn9imlP0ZjcecuvwBatNae3xa8/gat0195NyHnCS7oWE/nXNszR6zPl60Exrrm/17YLrOPv+tnf59a3hoxH7BCYisVP0xyEs2rjrpZ15dkmY497rQ+OiNzzi6vPaUhWZ92+4+t02+wf6M7O7uN+vVcedzI0llO9XxvDNPcJusa9lo1js6Ws16fSBFPF+zPwftLfZz4si4n+K9q6ffrBeddPHn2TePqpe8nfKf33bt2GbWZ5bt8aE4MOj21ZBzxifnuVaSvLDPcs1OtKxF/jQnW2cnVxZL9vtVDqRjpjP2scSpULKvn+A7GXwDBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmvYzAWWe/0qw3BqL3G5rtiNQWJ5Y+39Dg9tXYai890NU1y23T6mynscGODq2vt5c9kKT6Ovu1uoIdfe/FN0tSOmOf9qxTl6Qobb+Wiu05eCbr99XY3GrWX376mWb92OOPdfvKOUsiVCt+pG21ZkdVV5w2xWJoGQH7teksI+DFYQ8PDwfa2EsP4OgSx3ZMs1cPc2KHYzsmWZJiJyY6tPX2Djv2eUaHHd9dTfufw6yzLEBB9j5nnaVEJKnb+ezOnzXTbZOJ7HFo45adZn3Hrl6/r6w91u/t2WvW0wU/Jnru0jlmvb7Nv9d5Y0p/rx2FPl70Y/y9pVm8sa7BuTdK/pIpg4N+THho7MTR46R59md32Zx2t006by/PMThsj0+Vqv85rKvan/fWRnvckKT2k46xX4jsz0Eq5Y/PFWc5lUhDZj2f8UduZ3hWNfKfOVNO9n4qZcfYj4/7Y8rocI9Zj5rsMa0xsNRYsWiPdUX5z4+FtH2cZWcZhWLJv2/l8vb4mHGWwRoILIlQc5ZkyAaW1CoFl344ML6BAwAAAICEYAIHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEiIKJ5eTBoAAAAA4BDjGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4H6HNm3apCiK9LnPfe6g9XnvvfcqiiLde++9B63P3xZFkT7wgQ8c8OduuukmRVGkTZs2/c72BcCRifERwJEqqeObJH3qU59SFEWJ6fdoxQTut7xw03344Ydf6l0BgMMK4yOAIxXjG5KECRym7f/7//4/jY2NacGCBS/1rgDAYYXxEUASfeITn9DY2NhLvRs4gMxLvQNIrnQ6rXQ6/VLvBgAcdhgfASRRJpNRJhOeHtRqNZVKJRUKhUO0V/htfAM3DaVSSZ/85Cd12mmnqaWlRQ0NDXrVq16le+65x23zhS98QQsWLFBdXZ3OPfdcPfXUU/v9zOrVq/W2t71NM2bMUKFQ0Omnn67vf//7B9yf0dFRrV69Wt3d3Qf82bVr1+riiy/WrFmzVCgUNG/ePL397W/XwMDAfj/7ve99TyeddJLy+bxOPPFE3XXXXfu8bv2Nx8KFC3XhhRfqJz/5iU455RQVCgUtX75cd9xxxwH3DUDyMT4+j/EROPIkeXz7j//4D/3RH/2RjjnmGOXzec2fP18f+tCH9vu2zfpbtRf+9vdf/uVfdOKJJyqfz+uuu+7a52/9JnOcv+3GG2/Ua17zGnV2diqfz2v58uX6p3/6p/1+7oWx85e//KXOOOMMFQoFLV68WN/61rf2+9n+/n79+Z//uebPn698Pq+lS5fqs5/9rGq12gH3J0mYwE3D4OCgvvnNb+q8887TZz/7WX3qU5/S3r17tXLlSj322GP7/fy3vvUtfelLX9L73/9+ffzjH9dTTz2l17zmNdq9e/fEzzz99NN6xSteoWeffVYf+9jH9PnPf14NDQ266KKL9H//7/8N7s+DDz6oE044QV/+8peDP1cqlbRy5Ur96le/0p/92Z/pK1/5it7znvdow4YN6u/v3+dnf/nLX+p973uf3v72t+vv//7vNT4+rosvvlg9PT0HPD9r167VpZdeqje+8Y36zGc+o0wmoz/6oz/ST3/60wO2BZBsjI9hjI9AciV1fJOkW2+9VaOjo7rqqqt03XXXaeXKlbruuuv0rne9a1LHfvfdd+tDH/qQLr30Un3xi1/UwoULp3Scln/6p3/SggUL9N//+3/X5z//ec2fP1/ve9/79JWvfGW/n123bp3e9ra36XWve50+//nPq62tTatWrdLTTz898TOjo6M699xzdcstt+hd73qXvvSlL+nss8/Wxz/+cf3FX/zFpI4zMWLs48Ybb4wlxQ899JD7M5VKJS4Wi/vU+vr64q6urvjyyy+fqG3cuDGWFNfV1cXbtm2bqD/wwAOxpPhDH/rQRO21r31tvGLFinh8fHyiVqvV4rPOOitetmzZRO2ee+6JJcX33HPPfrVrrrkmeGyPPvpoLCm+9dZbgz8nKc7lcvG6desmao8//ngsKb7uuusmai+cq40bN07UFixYEEuKb7/99onawMBAPHv27PjUU08NbhfA4Y3xkfEROFIdyeNbHMfx6OjofrXPfOYzcRRF8ebNmydq11xzTfzb0wNJcSqVip9++ul96lM5Tqtfa59WrlwZL168eJ/aC2PnL37xi4nanj174nw+H/+3//bfJmp//dd/HTc0NMRr1qzZp/3HPvaxOJ1Ox1u2bNlve0nFN3DTkE6nlcvlJD3/74B7e3tVqVR0+umn69e//vV+P3/RRRdp7ty5E/99xhln6Mwzz9Sdd94pSert7dXdd9+tSy65RENDQ+ru7lZ3d7d6enq0cuVKrV27Vtu3b3f357zzzlMcx/rUpz4V3O+WlhZJ0o9//GONjo4Gf/b888/XkiVLJv775JNPVnNzszZs2BBsJ0lz5szRW97ylon/bm5u1rve9S49+uij2rVr1wHbA0guxscwxkcguZI6vklSXV3dxP8fGRlRd3e3zjrrLMVxrEcfffSA7c8991wtX77cfO1AxzmZfRoYGFB3d7fOPfdcbdiwYb9/ur58+XK96lWvmvjvjo4OHXfccfuMu7feeqte9apXqa2tbeJcdnd36/zzz1e1WtUvfvGLAx5nUjCBm6abb75ZJ598sgqFgtrb29XR0aEf/vCH5t9KLFu2bL/ascceO/G3EevWrVMcx/qrv/ordXR07PO/a665RpK0Z8+eF73PixYt0l/8xV/om9/8pmbOnKmVK1fqK1/5irnPxxxzzH61trY29fX1HXA7S5cu3e/fTx977LGSxJpIwFGA8dHH+AgkWxLHN0nasmWLVq1apRkzZqixsVEdHR0699xzJcnc99+2aNEi97UDHafnvvvu0/nnn6+Ghga1traqo6ND//2//3dznyYz7q5du1Z33XXXfufy/PPPl3TwzuXhgBTKabjlllu0atUqXXTRRfrIRz6izs5OpdNpfeYzn9H69eun3N8Lf1j54Q9/WCtXrjR/ZunSpS9qn1/w+c9/XqtWrdL/+3//Tz/5yU909dVX6zOf+Yx+9atfad68eRM/56WnxXF8UPYDwJGJ8RHAkSqp41u1WtXrXvc69fb26qMf/aiOP/54NTQ0aPv27Vq1atWkAj5+89uyg2H9+vV67Wtfq+OPP17/+I//qPnz5yuXy+nOO+/UF77whf32aTLjbq1W0+te9zr95V/+pfmzL/yy7EjABG4abrvtNi1evFh33HHHPr9JfeG3Jb9t7dq1+9XWrFkz8QegixcvliRls9mJ3xL8Lq1YsUIrVqzQJz7xCf3nf/6nzj77bH3ta1/T//pf/+ug9P/Cb5R+89ysWbNGkvb5o1cARx7GxzDGRyC5kjq+Pfnkk1qzZo1uvvnmfUJLDlZ40oGO0/Jv//ZvKhaL+v73v7/Pt2uhRM8DWbJkiYaHhw/JveKlxj+hnIYXfgvwm7P+Bx54QPfff7/589/73vf2+TfMDz74oB544AG98Y1vlCR1dnbqvPPO0/XXX6+dO3fu137v3r3B/ZlsjOzg4KAqlco+tRUrViiVSqlYLAbbTsWOHTv2SU4aHBzUt771LZ1yyimaNWvWQdsOgMMP42MY4yOQXEkd36z9juNYX/ziF4PtJutAxznZfRoYGNCNN9447f245JJLdP/99+vHP/7xfq/19/fvN8YnGd/AOW644Yb91vWRpA9+8IO68MILdccdd+gtb3mLLrjgAm3cuFFf+9rXtHz5cg0PD+/XZunSpTrnnHN01VVXqVgs6tprr1V7e/s+X/F+5Stf0TnnnKMVK1boyiuv1OLFi7V7927df//92rZtmx5//HF3Xx988EG9+tWv1jXXXBP8Q9a7775bH/jAB/RHf/RHOvbYY1WpVPS///f/Vjqd1sUXXzy1ExRw7LHH6oorrtBDDz2krq4u3XDDDdq9e/eL+lACOHwwPk4f4yNweDsSx7fjjz9eS5Ys0Yc//GFt375dzc3Nuv322yf1d7uTMZnj/G2vf/3rlcvl9OY3v1nvfe97NTw8rG984xvq7Ow0J7OT8ZGPfETf//73deGFF2rVqlU67bTTNDIyoieffFK33XabNm3apJkzZ073MA8rTOAc1kKCkrRq1SqtWrVKu3bt0vXXX68f//jHWr58uW655Rbdeuutuvfee/dr8653vUupVErXXnut9uzZozPOOENf/vKXNXv27ImfWb58uR5++GF9+tOf1k033aSenh51dnbq1FNP1Sc/+cmDckwve9nLtHLlSv3bv/2btm/frvr6er3sZS/Tj370I73iFa84KNuQnv9j1uuuu04f+chH9Nxzz2nRokX6zne+4/77bgDJwvg4fYyPwOHtSBzfstms/u3f/m3i73oLhYLe8pa36AMf+IBe9rKXvej+J3Ocv+24447Tbbfdpk984hP68Ic/rFmzZumqq65SR0eHLr/88mntR319vX7+85/rb//2b3XrrbfqW9/6lpqbm3Xsscfq05/+9ETa8JEgivmraxxECxcu1EknnaQf/OAHL/WuAMBhhfERwJFk06ZNWrRokf7hH/5BH/7wh1/q3Tmq8DdwAAAAAJAQTOAAAAAAICGYwAEAAABAQvA3cAAAAACQEHwDBwAAAAAJwQQOAAAAABKCCRwAAAAAJMSkF/J+61d/ZNbjmv8ndP6f13n1yO0r8l86JLxDqcU15+f98xIFjtPdvuztRM65jAInLHbaePsc2l+3r+lcF85mUin/9wyxc5y1aBp/2unsV6inlPPiDz/41qlvH4k1UqyY9Wq16rYJfUanKq7Z40MtMA6lnN/fZaK0/fPexS5Jkb19/9Njb+O/OjOrtcC5DI0RlkrFfr8kaXR01KyXSiWzng5sur6Qn/L2h4aGzPqevXvN+oyOmW5feWf7mbR9628KLXLrXRfj/rHUnOuysb3J3w6OOP/yr98265WK/5n2xg5/2Aw8c01jrI2dJt6YGo6TsF+rOsNmJcq5PbV32gtjp+Wfy/492+29ist2X874IEmlkv15j5wxPXTmvddC75f/2lTnG1O/LqpVf6xz74HehSSpVrNfu/KKKye1P3wDBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEmHmKQje64Xp7w/XpdqTpBF7Pzp4kucUxIUuX+gONU/3g/94eTUw0K8Jl64SZC3X8E/KHWuC+eP16VwKIj584E/Dq454S7ebyaCf2Y85dCdA3SIo4YXouGFOExH8I/knc9oKvTZdf642g1sCgQTRc59wBvr0il/v7w/FE8F29h/jD86Om7WvUASyd/nuro6s15f8AMHRobtQJJczm8zXiza26mvN+sz29vdvrq7e8x6VLB/vjg25vaVy9mNxmv+uQwmvOCoUV9vXzvTGR8PZvhTiPfMVZ1GiIn3WtUZU3PNM9y+OufYISb9vXbIkSRlnTCjbNqup1N+yFQmY4+1XohJaASInLC5cCCMzbsHh/KtvHtarWYHwkSBcBnnUViVciioJxTmdWCMrgAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABJi0ssISE4Uphuj77/mhugH4+qdpQcOYqRsqCd3pjud6P3p7EHKfqtqXpNpRI67Xfk9+W2ciHJJqjrRwdXq1K+xqb77wZ+fxqUUSDbHUcSNifauaU19TJtOTHWQ87Hyor3TgcTjlBMHXanYkdM9fX7k9bp168z6smXL3DblshNtHdnjZkN9o9uXF/GfdiLxR4eH3b68c5nP2/HdkjQyMmLWveuiWraXXfivHbD7ci6X0SH/WOo77GUUUvX+sYwHlmvA0SOO7XHQi2sPOzTfO3gjas15ZTpjsNemuanJbdPY0GDW9+7cNuXtpLxloKZ1r/GWDQudF3t8CrXwlsaJUt5SDf74ONVlLKLIvwmWnXG4VvWPxlv6YLL4Bg4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgISadQplJO3F7gRi+OLYTWw5mcmQwxWWKsYqhvcp6++wk8mQCkW1p57VKxU/L8V6pTuNUTjV5J5iu5LwWTB5yItAi7xoLnMvaFLcfOl2Rcy2FTzExlPA/u6HPjjd2eW1Cn1vvtVCbtJOolXISZLu7e92+Bof6zHpLs52Ytm3jZrevzRvWm/WGOj/tcPasOWY95RxjJvC7y+LomFkfHR0163UN/n7V1dnJjaH3xUvB3L7NTpkbHbZTKyVpzhz7vMQVO/0vdGvu6bGTQ0eK9vmSpGrRTgdtb2n1N4QjkHefntqzSKArTeteHHq08R75DmYKZc1JoQyk5M5oajbrWwPJ3yknCTFOO4ng00k1ds5/7CXYS4qcZ0EvHdPfyvNbsrfhNlDKmb949+ZqNZS6b5dD851QSvVk8A0cAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEJMPoXSTeQJRLx4wZVe2l8grcVN+AkmJHr7ZW8nCiXGxU5qV2yn0qQDx1Kft1PLRmslt42XfuOm9YRChJx9dn880Jn/ir+NlJs8NPl9eoETrhQ8fE/svMfhvkihhJ8mFUobjNzUroOXQhlKRhseGTbr2ZR9W9i4cYPb1/YdW8367FkdZr3iJDpKUiGftbexxU+uHB4YMOuR7ETHpqZWt69SqWjW+/vtbSxcssDtq6XNTowLpScPDQ6a9bER+5wN9va7faWc97+urt6sF0vjbl+9I/Z2Rsf8FMqmlP1eLj9phdsGR56am0I5DVMLBA81Uew8i0hSzUmIDCZse9v3njmdel2h4PdVc55TKnbiqxRINo6n/vwYOS+mvOfK4BRh6m+mlzzu3k+Db5fdxrvV1rwHTkm1Kd7PpfD1Nxl8AwcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhJr2MgJcr6kWKBntyYlC9eP+QULT2VPcsdCxR5Oyz0yRdq7h9ZWTHvXpLFUhSpWL3V5tGir0fLT71ztzo/+AyBi8uOnWf7XvxuNMJKHabHLz9xZEqbVa95T8k/3OY9sba0JIETl8jwyNum8cefcysLzzmGLO+dMkSt6+uDnu5gHLZjuTvi/e4fbVE9rksl/2Y7D09vWa9UrHPS3HrFrevshPHXXGWiujuDxxLU5NZzztLyUjSjh07zPqevT1mfWzcj/5PpZzlLZzrslj270HD4/ayE2NFf/t1zucCRxdvuSknqT/Ie0ycznJHoWVWvMchN/o+9PzqbCeVtj8fDU0Nbld9fd1mfWjIXuYktG/e+U+n/M9tXPGWtHI2HVgyJXaWHgi9L+6SDPHUlgR4vi+7XqlMfY7i7bG37IF0gOtvEvgGDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhJp9CGbk5PlPeqJscGexq6nGHfsCL/UIoxajsvFZxTmEm9lMoczUntctL55TkvRI72wkGejqv1ZxIolSos5r9O4DQ5v3fGkw9BdPbtWmEcwYusUDy0LQ2hCONNzzmMlMI+v0vcbk0pbokDQ3ZaZO9/X4y2fjoqFn30i5bm1vcvupyBbPe0FBv1te78bVS2klma2pudtuMO0mM2XzObhAY0waHBs366OiYWe/eucvta6jPPv/btm5z2+zcudOsj5ft96USuNcMj9rbj5xbcNEODZUkDTgpd1u2bHbbeIlxgHSg68N5TnMe7A5iuLWkqV+7oURB77VMxn4aKhTq3L669+w167XA82PKuUHFsgcC71lQCjxzOQ92XtLk868dvDfN72vqcwT/GjvISacvEt/AAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASIhJZ1xXnWjp2FsSQIHwTidWMxTbOp004tjL9nbKwUBb58XYaRUKDs1m7Dbu8gqSFNnR2pGX3Rrgnct0INrb5SfXurxzNh0151oKpOAG+gqcfyAgo7JZ37XLj5j34oXHRuy4+tERe6kAScrn82Y960TyS9LC+XPNeqViL1ewe5cdby9Jra2tZr1WsyPuh4aG3b76+nrN+ooVK9w2TU1NZn1k3D6XoQG6Lsqa9Y4Z9jIGy5csc/uqpe3xOXRd/Od/3mfWTzh+uVkfHRpy+3rmmV+b9XWrV5v1+oJ9HiWpfc4cs97aOsNtk8s6yzjgqDK9KHXn2cp7fgs+P3p9TT36/2DKZu2xJhNYfmZk1L4PBHfXi/53npFDz9vuS87zU2iO4G0o9F6mUvaYGnx+dkx1uQDveVOSYudZPHQsL/Ya4xs4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgISYdAplpWbHDQZTKKeYFpQKpCC6QS6BEJep5rtEgRbu5p2DTDtJOZKfGBdFdpKdJHmnuebtc/C82C+mnBSdUFJOKnaOMxRjNMXkn5CqcyxeamjovHjJlYcijQrJVi2Nm/U9O7e7bbxxcHjE7qt7b4/bV8b7GMZ+TGy5YidEjjv1trY2t6+uri6zns/ZY92aNc+5fW3dutWse+ljklTf0GD3tXOHWa+V7WOUpGzF/rx79ddf/AduX/mGOruvvJ/O2DLDPs+zZ9nnODOr0+1r/YZnzXqlYt9rfu9lJ7l9nXHe+WZ9xw4/nXTHDvv84+jipvpNIzkwlfKSdaee9jete/t0ni2c4/RSKONAInbZGbumk+LuDaleOqUkxdUpzgWcBPVp895L58dD56XqHYuzjdA9qDqNSynU36Tav6jWAAAAAIBDhgkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhJp1CGU0nLeUgpvd5yYmhFCGvjRcjFEqhTE0x7NFLt5H8VJxAbqObYhSnpx6JVHPiFquy9zkK7Jmb5RZ6771EKqceOi/eWfbf+wDnfYlD6VbTSNHCkWe7k8TX1Nzitkln7OG3Gg+Y9Xnz692+6rP2NTo+1O+2eXr1arO+aedus3788ce7fW3bZidHemPdxo2b3L5yOTuhcfu2bW6bweFhsz48PmrWy8WS21e6ZI8q6bL9WZ//7LFuX00ddqLk2JidNCpJnR12quTwYL9Zr4yOuH0ds2ChWd+5ZZNZj8v+ftXlC2a9s3OW26ZS9u+DOIo4t2Mv+fr5JlN76AolN/ppl6HkxtCTh7V9/7Va1d6+lyhZDIxPXuJx2ntIlX/85aK9/fp6+7MuScWqnWCbdt5LL91bkuSmXfrn3n0nnfcrlHTqJ+J7xzL12PvQ3Gmq19hv4xs4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACTHpZQRqXox/IK499iJCnXItEAPrbiYU6+nF9U89eV/Vmh2HXI3tOXDVDTuVGurrzHomO+Zvv2hHt8bOsYRT/J24fC9uNRWI0Z9i/fntO/WpJ/2616VXD8YGO68Fl6o4eCtlIME2bNlh1isV+3MrSem0PfyOjRXNemNd3u0r12BH76dr/vYj57XZs+0Y++bmBrevwcE+s1514rPzef9YTjn1VHu/Zvlx9U8++aRZ7+7tNeuLlyx2+6o60d6lov2+PPDwg25foyN2xP/cOXPdNiedfLJZ3+Ms1bB54zq3r2OWn2jWc84SFv177SUkJGl4yFmSIbRUgHN/xNElch9UAm2c+3FoiSa/L0/o3u4989rXdDAR3jn+dNoet+sL9jOiJOUi+/jzKf+8DDr3lNERe8ma9paFbl9l2UscuM+VwYekaTyMT7HJdK4XT3BJAud+Hloq4MXuG6MrAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEJMOoWy7KWlBJIj/fAZL91n6kIJL1PtMAqk5aSc19zkzMDUuL6+3qxn03YikCRVSnaaWZyzN+QlAklSbarRiVX/56vyUnRCuU9TSycN7e7UUyj9zrxrKZiiVCOGEtKcY5aY9ZFxO/1LkqLY/ux0Vuw02rhqp39JUrloJwQODg27bWbMnGnWxyN7TNmyzU7alKTBPnt8am23t9HY7J+Xsb5us7634h//np49Zn3pkgVm/dXnnO321d/bb9ab29vN+uqnn3X76t613azX5/0bRFy2jzOXtpPp2prtcyxJlaI9Po07qcbNs+wEUkkqOemco2U/6bQ26ScMHNmc+3Qo1c95hAg+802Rm7wtKfaeYaaR/O3tc13BSeN1Us8lqVSyP4cZJwVRkvbu2WjWR0ft+8PixQvdvtxHy2k8CqWck5ly7kGSFDvPXF5SfDhFfGo7Hdov74INXeNT3f5++/OiWgMAAAAADhkmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhJh8CqWTQBVKUZlqWpCfHOgn/IQ2UfMSAt0UoVCMkH2qsk5ntZLfV53s5KGWop3kJkmdpX6zPppuNuv9spMuJakS2ak4sez0s0ooudE9xYFEzyht1mveuZT985JUdt7LSuxcr7LrkpRy4pUyge1PKzoVR5zRcfu6KsZOypikXNr+HOZTTkJgU4vbV6Vmb7/Q4I8DJSchc6Tq9FVodPuaXbCTM+tmzjDr5Wqb21f72KBZ373dTlKTpLa8PT43jtnpnDv+8363r5199vbzi+aa9e6tW92+ZuTsc1kZ6XHbrH/Cfm3u3OVmfc68xW5ffUN2snHGeY9bG+ykS0mqOommI+Vxt02UPXiJgUgye6yrBpJ102nvvut97xD6PmLqD5DplJPw7Wwn+Czs9FVzkhN3797p9jUwZI+1Y6N+ivn9999j1le87FSznsnl3L6UclKSncNPB94XL4UymBzppU1OIxHcm6P4cxf/eilV7DE1lEKZcq6LyeIbOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAkx+WUEqk4UZiDuc6oRmbVAX9WDmNeeirx42EAbp56L7fNSHrCjqCVpeMdusz435UfqLjjGjt3ubphj1v9js7/9muyYcu/tqjlLKEhSqWY3ipxYc0lKO7Gq0TQiZbMpu00+Y7epOXHGklSx02kVVQPLCPA7EEgqjg6Z9VH5sezjTkx2LbavqaY2e8kQSaqP7CVI6pr8a3e8zl5iYEbBHmvKdfa4JUmDI6vNes/Wp8z6aLbB7au1aMfVNw/b51iSWmfZEf+1sv2hrvX1u30NbLKXBVjz5K/NemOlz+1r0YmzzXo6649D/Xvt5R0efOoJs14+ZoHbV7Mzds5L2+e4rX+b21ff0/aSGMN5f6mKlvn+vuHokc3Z41C1Flpmwr52U07Ee+wsA/T8a1N/fhwbtePyd+/tNuujo/aSJZI0OGg/jw0P2/XBQX+sK9S1mvVa1X/mi2v2/WHW7JlmveJE9UtSKmO/l0POMiPeUgGSlMs6yxUEVx9x3kvnUSyuTn2pM2/uEloSwJNyl8NQcP40qb5fVGsAAAAAwCHDBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBCTTqF0EyIDKSrTSf7xeGkx02nj7ZUTaChJqovtZLDMuJ0cOdqz0+1r1xY7yea0Y+1ESUlKZ+2da5rRbtYf3bjL7atBdsLQvFl2ItHGvQNuXz1OCqXkJ++kq+N2vWKnGBUiP9EyGreTn8aH7b6KkX9N1jV32Nso2HVJqqScFCUcVZoK9uezNO5fu1u27zXr89J2Alkc+SljQ6N2EmL/gJ8qOFizP6NNrQvN+uZnn3b72r11vVlfMGeeWZ/Z2er21f/sZrOeL9pJapI000m5q7XbiZr5ij/YN23dbtYra+39qm/33+OObJdZTznpxZJUHzsJcG2tZvnRvh63r9H+XrN+bJ29/T3rn3H7enaX/f5nFhzntjmlc777Go4e3rNgKKl81y77GaaxscXuK5CWnXaSAEPbL1fstO6REXsc2unsryTt2LHDrPf22PeAUScBU5JSOfv4I/nj48tPtdNwK05a+FjJft6VpPpGO0G430le73fOlyTlC3ayrfd+Sf57lnaiK/OpqV8XnlBSvpeiHvJip0h8AwcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhJr+MgBO+H7uh/FIUiEqequlFdDr77NRDSxVUS3a0c6Vox7Cm0wW3r7GqEzmurNtG43a0eGvG3q+lrX5f9Vn7tTrZ0ftDY93+flXs4y9X7QheScpH9tILKWc72VK/21e5117ioGfrbrM+Eoga7lr+crOemW3H9kpSJbBcAo4eG5551Kx3D/oR8zv77M9IobrWrLc8Z39uJGm8ZscxNyxc5LbZsXOr3deja8z6nl57+Q9J2ugc5sLTTjHrc4+xo6glqW9sj1kf2mN/piXp6S12xH9p2L4H9e2247slaW+vPQ5VW+wxvWnhLLev9X32vaZ3TyD6v8+OEG+b32nW80X/97Bbd9nv2XDKvp/MbPTHs901+1yO7PGXqhh68hGz/uo3vMptgyNPS0uTWX/2mdVumzWr7XEwV1dn1vNOXZIKBfuzm8/5ywClc3ab5pkzzHpTu12XpKUnHG/W6xvqzXrfoB+9f889vzTrxTH/2aaxfYFZr2u29zkOrKk1WrLHlKozDOXygXM8xRh/yY/yj7xM/sDXVCnnOKtVe6xLBeYI3lzoYC6n9tv4Bg4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgISadQuml0oSCJr3kSk8qlDTpJc8E+nPTX5wkmXLZT06Mh+1ksKb6RrOeb/WTd8Yq9kkb8jeveW12EuIxbXaK0QVn2KlHklTs3WLW1z37hFnvGLVT4SSpJSqa9b0Dg26b9pl2mtrQ2C6zPrbXTsuTpKaqnai5uMU+L6MZuy5JcWRfL6WUfx1XnTY4umxd87hZz2b8tMWTF5xo1ptH7MS20W77cytJ4yP253Ak5actFkt2m3jQTihsSPspbzPnzHH2y04s615nf9YlqXePvc87e+zEWUnqrtn71jM4atYLDX5K8PHHvdKsD+2yUzDXbPdTGB+r2YN6acxPJx0bs9+X6q7HzHo58u81pbR9r8vk7PTiPVV/PBuv2b/vHRrw35enfvojs/7Jj/y52wZHnnTKvnbaZ8x02yxdaj8n7emxU2KHhu1xS5J6euzU13Ta/w4jV2c/K0Rp+5kjFUhU9NIL4x67vn23n1JbqdhpxK0trW6bGTPt8dl7Fss4xyhJvU7yd7lkj3WhpM/RUXt8TjnXS+g1b+iqOknpkhTbj/WqVu02NSed8vlGXl/OC5JmzvSv/8ngGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhJh0CmW1ZqevuEmPB3jN5CT1SFImZSf8BLcRSLIxNx/oq75mp6m1RXn75wt+IlFpsNesD/X5yWiNs48x6w0Z+5zNXein2+wcftKsF7N22mTbHD8esxbZ5+XxMfsYJalWdZL5GtrM8oy8nfQpSUN77LSmWPY+V8b9YxkfsSOJShW/zXB60h8hHMGyKTtpKh72UyC719upr5pvJ4P1yf6sSVJLxh67hrb5CbILT7VTMHf1rzfr9YGh9sST7JSzXffdZ9af3rHT7WukYifYdlf9+0OpbZ5Z31u208RmttupvpJUarJTQFc/9phdX7/B7csJOVM1cOutOfetOGtfY5WUn7JWq9lpcnnnfu7fgaSi7DS5OOPfZwtOCiaOLiNOinfKea6TpDlz7M9016zZZn3Dpo1uXz/72c/Menv7DLeNt53RcTvtMgo8v3ovVZxUw6G+YbevASdRszzin8unH3/MrM+fa99rOjtnuX1lnfesVLTvTwO9fW5fpaKduBtKoaw5Y5eXep/N28/okp9iXyrZSZ+hFMp8ZI/p6cA13t7e7r42GXwDBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmnYHuJUiHYvxr7mt2PY79GNbYie8MpRSnvb2O7HlrLuVHhGb6Npv1Lc/acdgNWTtyWZLqM3bsc19sR7pK0siMUbO+c2ybWd8w5ke3FkY3mfXx7tVmvX9kxO1r8dIFZn1+lx/Pu7rbPs/FyF5GoHvI3/6uXXbkeHtLs1kfDVxjQ250rB/TXUzb0d44umze4y0NYkdOS1JTk708RanJvnY37bDjoyUpV7Kv0eKwv/1Nv7Jjn0d6+816NbCcRrbHHofG9w6Y9dj+qD3/mpwlGTL+7Wp8xF76YDhnR0jv6dnt9vWfv/pPsz46Ykd7p2qB9RWcl2qxP27ETuZ4zXmPqyn/fUk5v6ONnWjrsXJgPIuc8TH2t//qN13o94ejxtiYHRdfDVxusRMXn8vbS2P0dne7ff3qP+836687/3y3zaknn2zWd++2x44777zT7cuLpZ/ZYT/z7dzrP7/t3rbDrLe21Llttsh+fuzday9zM3NGh9tX1lk6aWzYfk7bs8tfyqa+YO9zLu8/P2cz9vufLdhjfd+QfQ+SpGzW7qujwz5+bx4iSXUZe/spZ74hSeWyP3ZOBt/AAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACTEpFMoVXPigpykIElKyUv8s6O5qoH5pPtKv58m1rNxjVmf1TnTrLe3tbh9lQfstMly3wazXoz95MRMxk4EajnmBLdNU9FOk2sas89/9xY7XUiSSplGs94/7vTV77/H6V32e7x9u3/8Q31OItKovc8j42NuX8WUfS0NN9jpRpVGP92oXNdk1mtOKp4kpQNpcjh6bCvaaVbljJ2sKknDcYNZ37V2u1kvjvlph5nIHsprKTvRUpIq/XYCVlywPwfBxOGSPUbUmu00r1ogPVg1J22xZKdmStLYiD0+DjuJmsVxOxVPkmol+7xknDSx0LHUnHtdFPnnMnbapJ0E3WzFv407gZaqyxXMejGQKFlx3n/7yn/e8mVLA6/iaFGpOM8QgVToyPm8xU5XpcBnOp1y+vI6k9TabI+dxVH7eWTDWjsJV5IqFXtMO+uVrzTrjz/+fbev4qj9bDWW8tOy566wP4eLFy4y61VnfyVpfMw+/rq8/czV3GjfTyTp6aeeNustToq4JL3yFfY5K9TbY9oTjz3u9tXYaN+DT1iyzKyHUiNTspN9g22c63Ky+AYOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmnUJZrdppe6FkMi9hxcsd8vOApKqznWJvr9tmaMsms57rtZMrG+fPcfvqam816zPqF9oNKoNuX51t9mmvb5/httk0WDLr8YCdzNbb46c7jQzabSrjdiJPj3+KtW73VrM+WAwk1mVbzXpzh51Yp6KdWilJ6Zx9LoeyebM+Ij+FslKwj181//ccsZeuhaPKaNq+djKNfkbfUMVJ46059YxzfUrKOmNtyklplaSaHZqlYsq+pkNpWpWqnVoWVe2+Mk7SoySVx+2+Rob8ZNviqDNG1OxkulQgfS7nRDd6p7IcunM55zL24iHl3x9j57rwW/ivFYv2/aQW+J1utWpvv3OGn95c6vHTkHH0qDlp5TX3mpa8gL6qM9bUFfx7+9Ili836vLn+M5/3qdq7176mh4f98Wn2rFlm/U9WXW7WV6/Z6Pb1i/+4z6ynsv695oyXv9ys/94pJ5n159b629/tpIJ76e61kp8O+uzTT5j1U099mdumkLePc3zYvgdEFf/5ceG8uWa9PGK/l5mMnbQpSVUvPTjlT7NIoQQAAACAowQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEmPQyAhUnVNUPgZVSwXjj/QUSZVUZs6NIxwcDEaHzF5n14t6dZn332nVuX4VFdgxscdyOY64U7WUXJKmarjfrY4E46v51dvT/7p4Bs14b97dfGrJjYOd32JG61aofTzsgezvzjjvRbTMcF8z6SMbeTt+ovyTDUGxvv5qxlxGopOxtS5KyjWY5ln/8zuoaOMp0NtnxwqFlVhoa7DbV2BlrA+NjzYnlLzvjpiQVq3aboZI91mRjO75bUuBXgfaYVqr6kdvDwz32CyV/rM/IPs6RlP0BrQVOZuTdiJx6xok1l/wlCULXReS0iVL2ug/TGYJSkd1XHBjQ0s5h1qf93wP/+le/nNJ+4UjlXO+R/8wTO228FTgWLlrg9jV37myzvnix/Yz4/Ibscm9vv1mvVPzPTpezjMCcOXaM/StecYbb168eedCsn37m6W6bs895hVlva7WfeebN7nL76u+2n8e8pcaGh/rdvt648jVm/Q1vXOm26esdNuvlMXscamn2o//nzWk369mUvSRFHPtjXc2514XGen9pmMnhGzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhJh0CmW1YqelhBJWihU7GaxSsdPPFEh40ZidjLZ3Z7e/fSfNbEZk9zXev9vta7zf3k41tpNnKhU/mWyt06azs9NtUxq3+xsr2imYdc1+2mKuwX6tOqvBrNdSfpro6KCdqLkjkFzZV7b76y/ax1jM2tuQpGrGvoQjL0Uo8i/5VNpOroxCyUOhaEAcNf7g98826zt3bXPbpFJOypozPhSdMVCS4rL92cmmW/zt1+zP4fioPW6PjvopkLWak8DmRMaVKq1uXz3d9vhUHB1y21SK9j5XS/Znt1IKpN8557LqJH2OxX7SZy1rv8cVZxuSVK3Z+1aot8/LmJMaKkmlkn1/KBTs8bE07tyb5Ybyqd4OtJQkNbQ1+S/iqJdK+fdW99nSKTc3N7t9nXbaaWZ9xoy2KW9/fNxO8R53nsUkP+2y5iQB19f7zylNjfZzyoJj/OTI5iZ77PDG7WLRHwcqVW/stEeI4RH7fElSY0urWZ87f77bpq9/jVkv1NuJmsuWLnf7KpXswSvrjI/+XSOcyP+7wjdwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEmLSywgUh+wo0NiJvJakohOrWnSijaNADmemYsedNjXPcNvs3GDHjXYP7zDrueqg21dctI8zk7UjXVN5O4ZU8qNTuwcCMd1F+/x3dtgx4R2L/EjZqM2OlO3J2HHUgyU/8nqs0GrW9zpR5JI0Gtm/N6g12DGw1byfUx3Lfi0VO21Sfl8pZ7/STty7JLGKACRp/hw7JrqlyY/xr8X2OBgXR8z6SF+v21dp3P68NRTq3DZx2h7+h5xo6+69e92+BgftsbPkxFFHFf9z2JC3lzOJAtH7JWcVg9j5TEc5/3eXkRPjH1XtD3s6cjYuyVnlRH4ovxQ5Sy/UZe3OMs77FerLG+sqzrFLUilj97W5b6fbZku3vzQPUAtcb961GzuB7em0P6accMLxziv+DTxyXuvsnGnWj5k/2+3rlWedadYrVfseUCr6z4IZZ+iKYn8cqsvbyzpt3LDJrN96251uX14sf6HgLV3lvy9RZL9WDDxzlpy5QCT7Wbxz1jy3Lzljd61m75d3z36+jbNfzlgrhZdhmwy+gQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEmnUEZOzFetGkjgcgKGosjZbCiRyElrqWv1UyijxXYyXN+GbrPemGty+ypk6s16Nu/Um/z0t/o2O5kuCiRXlp3QslKdvZ1tqUDyjRNwFKftY4mzdoKRJBUb7O1Xnb4kKU7ZKXOVjJ0i5F9hUtq5hFNOulEg/E01JwY1jvyUNwUShnD0KDTYY0ec8i+42EkNS5XtD2iTk84oSZWSfY3G1UBKsJOaVa3YCWAdHf440NTUZtaHx+xk2/5xO1VXkiqN9jgQDfrj48igndzZP27XS04SsuQffzWy63knAVSS6sadZLJA+FjNuWTqnUTRXM0fgzJpu7OCk0oXB8ZHFew22ZmdbpO2Dj8NGZiOyLmJh/L8vLTLUHKl96xw7HHLzPoH/uwqt6vly+0UzOHhIbNeLvpPPSND9v1h7277uVaSsml77Ny08Tmzns95iZJSOmWPA2VnjjBrtp/OOTJmH//WHX6ybdUZpLzU0Fj+PVA156qJ7bG+Jvt+9jz7vHhpqhIplAAAAABw1GACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEmnUFadlLNa1U5rkaSak7DiJUp6SUGSJKfNWDWQENhkp7Z1HX+CWW9rCCTCpOxktIqT9lhO+SlCfc5Zr0T+8ZdSTtpizt6vxqyfWOelbeYzdspZJZCiU87Y56zsJY1KqjoJkXFq6smRXmyaeyYDqT9V7/oLxGDGTpIfji7Vip1EWA6kHXrjXVS028Rlf3yInd/FVQPZbF5AZdr5HOadsUaSUk4aaybrJEqmA8mJJTuhMq5vdNvU0vZrtcE+sz6WstMpJT+hMirY+xy1+wPU8NCgs5HAfTNjb2fQOZeVsn8uK865zJad5EwnGVWS0lU7gW3B/IVum7N+/7Xuazh6eGl7sZcCKCntjBF+2qDPS5sMpVB6zwN1BftzuHjRQrev+gb7eWx02E6UzGf8xN+5XXPMesZ5RpWkLVt3mfVSyT7GxYuXuH3VnPfMey+7uvwk2tFxe9zu6xtw20SRc25i+1hip/78a942nPTgUHyw+2zptwk8Wk8K38ABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiEkvI1Cu2rHDtUCMeuzkv7sxsIFo6VTWfq3mxPhLUrlqx6rGhXazvjfylyQYqdr7XHESQmuxv19pJ0Y/k/Xfjrp8zqy31NnxtK2FZrevQrpg1iveMQZWd8jU7EjXVOB98ZYLqDmRqrGzVIDkx6TXnCUZglHDzn55EemSVIn9OHAcPcbG7Lj2YtGJa5cfrZ2u2mNqteJfa+WKPXZVK/74XHbGbq+Fd4ySfyxyxoFioK/hQTtCuuyMT5KUkj0+ppxPfCYVWBrFOZaSs8+tDfayLJLU1jHTrFdG7Uh+SYqdZQQqGfsYK5VAFHrJHuvTzrscisnO5OztNNbVu20qzpIYOLqknLz00L017Tw/1pxY+CiQye695o5bkmre2OGMz1FgGaxMxv7sjOy1x7pF82e7fb3vPVea9cAqM9q4eY/9gjM+h1b0cu8QzumPAutA1eXtpatCy0tE7nJbTj32x0d3z5zlVGo1vy9v+6Fr7EWuIsA3cAAAAACQFEzgAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJMekUylqdnTYYB1JZvCRCrx5KCIydV1OBtBpFdjrW0JCdJuYEz0iSKjkn5cxJF8o5SWKSVJ+3z2Wdsw1JaqyzEzWb83bKWEPGTveRpJSTkFmq2YlhFSf1SZKqbvRQ4HcDTgKc9x4Hs3q85CmnHkr9qTnXUhw4/ih0nDhqpNP2OJDN2p91KZBCmbavt2wqkMJYtrdfCqRg1qr2pyEdSJD1FMft7XhhvNnA+FjI2belKJBCmXaSdVOZNruvqMXtq7/fTobr7++3t+Gk0kmSnHTQysiI22SkbCdUltP2/WG0HEjSK9vvSy49tYRoKXAPrvijalwNxtnhaOEmRAauNy8K8SDecmuBhEBv+96I3ljwn99anefnobEhs16Xt5/3JKm+0R7TimV/HCpVvTTYabwvgXNmqQbGR/ce6CSCS4FEUS/dfmq7K0mqOe/9VI/9v1pN45XJ4ekTAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmnUKpeieFMdDES/ipOvVgwouTXFizQ74kScWcnWRTbbATGqtVP/mmmraPtOAkRxYy/qltcto0BBLr6pzkSi/lLor8Y4lrTvKQ01dNfopQzemrEsjXqTqpjrGzz8HfMjjXTChdyu9q6m28RFVA8tMpg22cz0fKSQ6UpFTWSW50xi1Jqo7ZaYfeZycXSMmtVCpmfXRs1P75kp+O6SVUeuOGJMXOuUmV7Tb5nD/WFpxzmXPOZXPBT4xrcO4DgyX/xlV1kiMj5x4Y1fnvS9W5DcUVO5Uu7aYFSvUNzWa9vW2G2yYdzP3FUcO5tbpJk5K8V9zPQeDa9e7tXtqg5D93ZJzttDU1uH3NaLLHm96s/WzVFxi3x4tOWnggkd07Nd5YP51nIa9N6H1xXwsMG9NLgvzd9/VS4OkTAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQkx6GYFS5ESEhho5ca+ecKKnHfdaCcStlp1Y/FLVjn2OnPhoSUpnnYh7Z/uZjB9Tnc04SzKEolOdJH9vRYBq5MfjVmK7UdmpVwIxsN51YYeKP6/m9Bd7ywvE/nvsxsB6+xy4yKYTKFut+kss4OiRiuxrdDrLCKjqfA5rgeU8vOs6sMxF5LxWLdkx1cVxe9kBSYqdCOuMc/yBYVtyjrO5qdFtEtW32tsZtM/l8PCQ21epNGbWW1qa7Hq9v4xA5CzVUEj595r5s2aZ9dGUfS67R/z3Rc7yM6ra73EmcBNqaGw16y2N/vuSnc71j6OGtySAJKWcp0s3Ej9wB/eWCwjFyMfOmO499La01Lt9NTTYjZob7G1Eg25X7j6HjqXmLt009WU+phq9f7CXdziYprr90LFMdRv/9eqU+/tNfAMHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBCTTqFMOwmNU02kkfwkFy+FUJJqTspZHEgxir0kyHzBa+H2lc3Y28k620+nA6fWeS2Yu+MkxqWdZLLQW+sl1lW8pM9A8k7FTcF0m/jJR04KphN0GeRtI9RVzUuUDBx/ahqpRDjy1JyY2FogOdLvzP4cVp0xUJIqFXs7xWLRbVOt2lmxKW+syfhjSmncTm702tTV+YlttaKdqjg+7mfbFksDZr3iHGPeSSiWpMb6BrNeLpen9POS1D5zplkfafJj5kbL9nvmjfTlQKKll7pbGR826011fqJmXUOLWffSTKUD3NNw1KjW7M9hOpRj7j1bOONjKKXX3UTg/u2lWla9p4iUPz6XiiPONuwxpVzxx7pqzR4JYuf5SZIfF+/scjCdcxrP/J7ppIMeTAczhdLf5d/dsfANHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgISY9DIC3kyvFoj+92I1vVDN0DIC8l4LRGtHXnZq5AQyB6JLU852Us42QomuJS9cOZCXn3bOWs57ZyL/XFaczZSdc+zVQ31VAvHANW+5AK9NHPg9g/Oexd45DsbTTvGC1aGLu8XhrVyxo+/LZT8O2oskTnuXYc3/HPhjbWBpFu9zHdnbyeZzbl9yllPxxvS6uMntKpuyt1MKpIQPOhH/ubL9vuTS/rlsaWw16+WKvY2GQuA26ixjUN/Q6DbJyl5ioTpuR5FniiV/8845K2Tsc9yQ9hYrkPI5u00qsCRDaTrLaOCIk8160ffTeH6r2n2lQstpuBvxn1NykX3tZlP2mNbdbX8+JWl01B47+vrsbVTkL+cROUtaZQLPXLHzqO8tKRV7A4f8pZO89zL0XJ92liAJP1YdvGcubzupwPPzwfRij4Rv4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASIoqJ0QMAAACAROAbOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgIZjAAQAAAEBCMIEDAAAAgIRgAgcAAAAACcEEDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgfsc2bdqkKIr0uc997qD1ee+99yqKIt17773Tan/eeefppJNOOuDPvbDvN91007S284IoivSpT33qRfUBIJmSPAYCwKFwOI6ThwPGah8TOMNNN92kKIr08MMPv9S7AgCHHGMgAIQd6ePkt7/9bV177bUv9W7AkXmpdwCHrwULFmhsbEzZbPal3hUAAAAcIt/+9rf11FNP6c///M9f6l2BgW/g4IqiSIVCQel0OvhzIyMjh2iPACAZKpWKSqXSS70bAPA7Nz4+rlqt9lLvxlGFCdw0lUolffKTn9Rpp52mlpYWNTQ06FWvepXuuecet80XvvAFLViwQHV1dTr33HP11FNP7fczq1ev1tve9jbNmDFDhUJBp59+ur7//e8fcH9GR0e1evVqdXd3T/oYHnnkEZ111lmqq6vTokWL9LWvfW2f162/gVu1apUaGxu1fv16velNb1JTU5P++I//WJJULBb1oQ99SB0dHWpqatIf/MEfaNu2bZPeHwDJcSSMgc8884xe/epXq76+XnPnztXf//3f7/cze/bs0RVXXKGuri4VCgW97GUv080337zPz/zm369ce+21WrJkifL5vJ555hlJ0nXXXacTTzxR9fX1amtr0+mnn65vf/vb+/Sxfft2XX755erq6lI+n9eJJ56oG264YdLHAuDwk9Rx8rzzztMPf/hDbd68WVEUKYoiLVy4UNL//2/r/s//+T/6xCc+oblz56q+vl6Dg4P61Kc+pSiK9uvvhX9uumnTpn3qP/rRj3TuueeqqalJzc3NevnLX77f2PjbfvKTn6i+vl6XXXaZKpXKAY/5SMU/oZymwcFBffOb39Rll12mK6+8UkNDQ/rnf/5nrVy5Ug8++KBOOeWUfX7+W9/6loaGhvT+979f4+Pj+uIXv6jXvOY1evLJJ9XV1SVJevrpp3X22Wdr7ty5+tjHPqaGhgZ997vf1UUXXaTbb79db3nLW9z9efDBB/XqV79a11xzzaQCQ/r6+vSmN71Jl1xyiS677DJ997vf1VVXXaVcLqfLL7882LZSqWjlypU655xz9LnPfU719fWSpD/90z/VLbfcone84x0666yzdPfdd+uCCy444L4ASJ4jYQx8wxveoLe+9a265JJLdNttt+mjH/2oVqxYoTe+8Y2SpLGxMZ133nlat26dPvCBD2jRokW69dZbtWrVKvX39+uDH/zgPn3eeOONGh8f13ve8x7l83nNmDFD3/jGN3T11VfrbW97mz74wQ9qfHxcTzzxhB544AG94x3vkCTt3r1br3jFKxRFkT7wgQ+oo6NDP/rRj3TFFVdocHCQf8IEJFRSx8n/8T/+hwYGBrRt2zZ94QtfkCQ1Njbu8zN//dd/rVwupw9/+MMqFovK5XJTOjc33XSTLr/8cp144on6+Mc/rtbWVj366KO66667JsbG3/aDH/xAb3vb23TppZfqhhtuOOC/EDuixdjPjTfeGEuKH3roIfdnKpVKXCwW96n19fXFXV1d8eWXXz5R27hxYywprquri7dt2zZRf+CBB2JJ8Yc+9KGJ2mtf+9p4xYoV8fj4+EStVqvFZ511Vrxs2bKJ2j333BNLiu+55579atdcc80Bj+/cc8+NJcWf//znJ2rFYjE+5ZRT4s7OzrhUKu2z7zfeeOPEz7373e+OJcUf+9jH9unzscceiyXF73vf+/apv+Md75j0fgE4PBwtY+C3vvWtiVqxWIxnzZoVX3zxxRO1a6+9NpYU33LLLRO1UqkUv/KVr4wbGxvjwcHBfY6xubk53rNnzz7b+sM//MP4xBNPDO7PFVdcEc+ePTvu7u7ep/72t789bmlpiUdHRw94TAAOrSN9nLzgggviBQsW7Fd/oY/FixfvNzZdc801sTW1eOFcbdy4MY7jOO7v74+bmpriM888Mx4bG9vnZ2u12sT/P/fccyfGz9tvvz3OZrPxlVdeGVer1QPu/5GOf0I5Tel0euK3DbVaTb29vapUKjr99NP161//er+fv+iiizR37tyJ/z7jjDN05pln6s4775Qk9fb26u6779Yll1yioaEhdXd3q7u7Wz09PVq5cqXWrl2r7du3u/tz3nnnKY7jScf1ZzIZvfe9753471wup/e+973as2ePHnnkkQO2v+qqq/b57xeO4+qrr96nzm+OgSNT0sfAxsZGvfOd75z471wupzPOOEMbNmyYqN15552aNWuWLrvssolaNpvV1VdfreHhYf385z/fp8+LL75YHR0d+9RaW1u1bds2PfTQQ+Z+xHGs22+/XW9+85sVx/HEcXd3d2vlypUaGBgwzyeAw1/Sx8mQd7/73aqrq5tW25/+9KcaGhrSxz72MRUKhX1es/4J5r/+67/q0ksv1Xvf+15df/31SqWYvnAGXoSbb75ZJ598sgqFgtrb29XR0aEf/vCHGhgY2O9nly1btl/t2GOPnfj3wOvWrVMcx/qrv/ordXR07PO/a665RtLzf4txsMyZM0cNDQ377Y+k/f6N8m/LZDKaN2/ePrXNmzcrlUppyZIl+9SPO+64F7+zAA5LSR4D582bt9+DQltbm/r6+ib+e/PmzVq2bNl+DwsnnHDCxOu/adGiRftt56Mf/agaGxt1xhlnaNmyZXr/+9+v++67b+L1vXv3qr+/X1//+tf3O+4/+ZM/kXRwjxvAoZXkcTLEGu8ma/369ZI0qTXeNm7cqHe+8526+OKLdd1115kTvKMRfwM3TbfccotWrVqliy66SB/5yEfU2dmpdDqtz3zmMxMX5lS8kN7z4Q9/WCtXrjR/ZunSpS9qnw+WfD7Pbz+Ao1zSx0DvbyfiOJ52n9Zvo0844QQ999xz+sEPfqC77rpLt99+u7761a/qk5/8pD796U9PHPc73/lOvfvd7zb7Pfnkk6e9TwBeOkkfJ0Os8c6bXFWr1WlvZ/bs2Zo9e7buvPNOPfzwwzr99NOn3deRhAncNN12221avHix7rjjjn0u2Bd+A/Lb1q5du19tzZo1E6k+ixcvlvT8P885//zzD/4O/5YdO3ZoZGRkn2/h1qxZI0kT+zQVCxYsUK1W0/r16/f51u2555570fsK4PCT9DFwMhYsWKAnnnhCtVptn19arV69euL1yWhoaNCll16qSy+9VKVSSW9961v1N3/zN/r4xz8+kdpbrVYPm+MGcHAkeZyczjddbW1tkqT+/n61trZO1H/7Xyu88K+1nnrqqQNOOAuFgn7wgx/oNa95jd7whjfo5z//uU488cQp79uRhq9RpumF397+5m9rH3jgAd1///3mz3/ve9/b598lP/jgg3rggQcm0s46Ozt13nnn6frrr9fOnTv3a793797g/kw1QrtSqej666+f+O9SqaTrr79eHR0dOu200ybVx2964Ti+9KUv7VO/9tprp9wXgMNf0sfAyXjTm96kXbt26Tvf+c5ErVKp6LrrrlNjY6POPffcA/bR09Ozz3/ncjktX75ccRyrXC4rnU7r4osv1u23327GhR/ouAEcvpI8TjY0NJj/zDPkhYnZL37xi4nayMjIfkuvvP71r1dTU5M+85nPaHx8fJ/XrH8F0dLSoh//+Mfq7OzU6173uml9e3mk4Ru4gBtuuEF33XXXfvUPfvCDuvDCC3XHHXfoLW95iy644AJt3LhRX/va17R8+XINDw/v12bp0qU655xzdNVVV6lYLOraa69Ve3u7/vIv/3LiZ77yla/onHPO0YoVK3TllVdq8eLF2r17t+6//35t27ZNjz/+uLuvU43QnjNnjj772c9q06ZNOvbYY/Wd73xHjz32mL7+9a8rm81O7gT9hlNOOUWXXXaZvvrVr2pgYEBnnXWWfvazn2ndunVT7gvA4eFIHgMn4z3veY+uv/56rVq1So888ogWLlyo2267Tffdd5+uvfZaNTU1HbCP17/+9Zo1a5bOPvtsdXV16dlnn9WXv/xlXXDBBRPt/+7v/k733HOPzjzzTF155ZVavny5ent79etf/1r//u//rt7e3oNyPAAOviN1nDzttNP0ne98R3/xF3+hl7/85WpsbNSb3/zmYJvXv/71OuaYY3TFFVfoIx/5iNLptG644QZ1dHRoy5YtEz/X3NysL3zhC/rTP/1TvfzlL9c73vEOtbW16fHHH9fo6Oh+Ez5Jmjlzpn7605/qnHPO0fnnn69f/vKX+wS+HHVeiujLw90Lcafe/7Zu3RrXarX4b//2b+MFCxbE+Xw+PvXUU+Mf/OAH8bvf/e59YldfiIb9h3/4h/jzn/98PH/+/Difz8evetWr4scff3y/ba9fvz5+17veFc+aNSvOZrPx3Llz4wsvvDC+7bbbJn7mYERon3jiifHDDz8cv/KVr4wLhUK8YMGC+Mtf/vI+P+ctI9DQ0GD2OzY2Fl999dVxe3t73NDQEL/5zW+Ot27dyjICQMIcLWPgb/vtfY/jON69e3f8J3/yJ/HMmTPjXC4Xr1ixYp8x8beP8bddf/318e///u/H7e3tcT6fj5csWRJ/5CMfiQcGBvbbzvvf//54/vz5cTabjWfNmhW/9rWvjb/+9a8f8HgAHHpH+jg5PDwcv+Md74hbW1tjSRP7+0Ift956q9nukUceic8888w4l8vFxxxzTPyP//iP+y0j8ILvf//78VlnnRXX1dXFzc3N8RlnnBH/67/+68Tr1li9bt26ePbs2fEJJ5wQ792794DHcaSK4vhF/MU2AAAAAOCQ4W/gAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABIiMxkf/CGv/kfZr1SrbhtosiuV6s1s55Kpd2+0mn7Na8eeq1Ssfe5Wq26fXmvlYpFu14uT7mvWs0+L6HXosieg2ezWbevlPO+pNL2C+mUP8/PZpz3JdAmkr2duGYvSVgs2edYkioV+1yWnfe4HHhfxst2m2Kp5Lapyt7nz910q9sGR55/v91+v9Mpf4jd27/LrG/a8KhZ39O91+2rlG0y6+s3b3Xb1NUVzPriuXPMesYZHySpnLfr9c4wkEuPun31Dg6a9V29fW6bTM4+lsgZUjMFZ4cl7dq7x6wvW7DQrA8G9mvzLruv8dg/lzM7Ws16nXOriwK/h606r7U12ttIlf1lYbfuto9lcKjfbZPP29f/T/7f424bHIns66oaeH7E4Sl23svIqzvPqP/VyBR6Fi6X7Ge+YtF+Tgs9C+dyObOecp9f/f2ajnTaez7w7w+/iW/gAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABJi0imUVScVJo5DqSx2koqXMFNzUghDbeLYbzPV5MrIi81UKEnHSW5002Wkas1O0fGTb6Sac5xxzU5xCqVzxvJSMO16Ou3vVyZjH2cohdJLIS1X7YTIUigF0kn0jL1rL3C9uAJtvPcSR5cnnrvbrMcV/3O4s7vbrEe1cbO+20kBlKTRqp2EODY85rYpOMPd7i1b7P2a0eD2tezUk8z6SV12ouWTT93n9hU5Y1c+X+e2GS/bn8O9u3vNenNLs9tXS1OrWS85ibcDRfv9kqRcg73PuUAyW8q5D2br7fM/MuZvv31Gi1kvD9tjbY9zviSptd3uy436lDQ2NuK+hqPJ5FL1cHgIPQunU36qoyWUVD80OGTWe3rse6Mk7dq526x7CePLly93+2pvb3dfs4Wu42k8W75IfAMHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBBM4AAAAAAgISa9jIBkRwV7kfCSv8SAF/0ekkrZ0dKh6H1vO9msHYOaCkSnZtJ2m0rNWaogEBNdcPK7o2wg+t+Jso+cGHtvSQBJqjhLD1SdYwmdYy9uNhWIWy07catV5xqL/NPiXmNl572vhJYEqNgxtFXnfEnh84yjR3dvj1nv7x1022zebi8L0N5pR9ynm+rdvvq22NsfceL1Jamj1d7O3LZOs/7snu1uX08/8ZxZbzrevsWUqgW3rz3du+wXAr9uzDhLD3TNtZcxKBXtz7okRbG9z9t32ue4FIiP9u4p87u63Da9e+wI7bEhZ7mAvNuVcjl7+z1lO94/1+wvFVGu2NvP5f3HiHI55+8cjnqhuHr87nnLc42N+cvP9A/a97Txkj2m9vf3u33t2WPfA0NtSuP2OHT8sceZ9cbGRrev0LPtofHirv+Xeu8BAAAAAJPEBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEgIJnAAAAAAkBCTTqH00lpCIUKVipMEWHFSEAPplJmMkxzppFNKUjptH146bSfvxDU/Tay/d69Z37xpk1kv1NW5fS1atMisNxf8tJyac268FMaKk6goSXHZSYF0EhXLzvslSbHsNtVAuk6xavdXiZ3rJZDyNu4cZ8nZ50rJT02tlp0UysDxV53zj6NLc0urWT/hBDsZS5KO2d1r1oslOwFsuHfI7as3spPBslk/BfCY2fPN+vkvP8esp+77T7evBzauN+v3dP/arGfS/vgwNFo06/Pmtrlt0jknwTZvRzTmM+1uX6ufXGfWuwdGzXpdix8DWZ8p2fvlJKlJUl1s39NammaY9YIfTqpZ7S1mvS1r7/P2vXY6pSRVa/b7Uhqyj1GSZszwzzOAl5b3LL5p00a3zaNPPGXWs3n7mXfASa2UpIZ6e/Dq6+t32zQ6z9Zz5tiJw17q/JGAb+AAAAAAICGYwAEAAABAQjCBAwAAAICEYAIHAAAAAAnBBA4AAAAAEmLSKZSxkwTo1SWp7KQNlp3kwGrVT/TztpLJhA7B7i/lRGeOFO30N0na3d1t1ucvmGfWC3V++lvW2ed84Fgqzj5Xak7apB/oKaWcvpzz7yUVSVLFeWNSgRTKSs3uz02nrPqJmuNlOxmt7Cag+tdr7GzfS+eUpErgNRw9RivDZj3Kz3TbnLBigVkf67b72tm/xe0r1WUncO3oH3DbdLR2mPVq1f69XkPsp3l1NNlpg+W0nXaYTvljXSayx6GBgUAabspOiOwf3WrWW+pb3b5aW+zkxv4RO20xHUgvXjqj06zPTvnJlZW8ff7z4/Z5qQZS3vLOmJYftu91xZ5+t69SvX1Py+X8JOg5s0ihBF5qsTNEDQ/Z95r16/0Uyiiyx6dKxR6fRkf85+pM2r6neH1J0tJly8z6jBl2Sm8oKT/p+AYOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQkx6GYGiE/3vLRUgSZXYiaV36jWnLkmp2I5rr8X+9ms1O/a5Z3e/WT9mrh/5Pd7ZatYXLDjGrLc3+JHbm7f1mPVq1Y+kr6Wcc+lF7wfOS9U5z2UnEj/0Hrv5tAEV5zhLJXtJgGroGivb12XNOZY4Dixv4MSXVwNrMnjbwdGlZ3jIrO/s63XbDI7Z8cp1Y/bv1ebk7Xh7SVK93df2XjteX5I2bths1ouDdrT01u1r3b5KdXYsfqpi32KqgeVn2lvnmvWxsr1fktTbb4/1xbK9/fZ6P8b/pOOXOq+Mm9XOuoLb1/JUnVlvK/r3unze3rds2W4zFPnHMrbbHh9Pap9t1uvm+9fYk4N7zXrHPDu+W5JmtvA7Yrz0Yuc5xatLUuTkz3v1Q8bb5cCjWGncHge2bt1l1nfutJ9RJSmTt5cTGSva43M26z8Lj4yMmPWuri63zbHOMgKFgrd0V+j98sfhJGB0BQAAAICEYAIHAAAAAAnBBA4AAAAAEoIJHAAAAAAkBBM4AAAAAEiISadQlit2EmAtkOKTStnzw1TaToVJpfy0mEwmHdg72/CInYqzYfV6s152UrYkqX2WnbTVWLLTekol/7wU8k1mvVrzE3Fqkd2fl9wZSkf03suKUy+V7IS3kFA4Zc05Ti+Fs+wkTYZei92kU/86mk5SFSmUkKShETsF8v6Hn3LbzJrZadbnNXeY9bNmLXH7ikYGzHp3vT/ED9XsfV63eaNZL5ftxDBJ6h7rN+sz2u2ExGrZT5Yd3m0nehYaGt02bc12alnvqH0P2r3LPl+S1Fix0zZbnZi3WUV/fFpQsFMoZ+T9ZLYossfbeufnRwv2/USSVo/a41O381a+4uxz3b7aB+zz0p8edNtkGv3jBA6m8H166mmDL3na5BRFzvO2JG3YssmsP/rkk2Y9TvvPSZHzWktLs1n35gGSVCjYCb5Llvj3umzOvqcl7f06GPgGDgAAAAASggkcAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEJMehmBuGrHsIaiW+XELmfSzvICKT+6NJOydzWb8Q8hrtmxou2z7Mjp3SNFt6+erfZyATMa7Wjr+QvmuX2N99qx0wNDdqy3JEXOXDuuOcsLBGJzq1U7wtuL8fd+XjrQ+2/z9q1Ssc+Lt7yBJDmHL8l+78NRs3ZnVQXOJcsIQJKXir9jtx9XPzZkXzutS+045tzsBrevGTk7rn9Je5vbZqB9jln/zydXm/VsLrDMi3MCuvt2mvUWZ9yUpErV/hyODfmfw3jEblOp2ftVKPvHUtnda9bbnSUZ5rT4961ZztI4+cCqOLWMPdbXO0vpFFN+VP/GHjvi/5m9fWZ9uG2229eZL5tr1jeX/DHwmb7t7mvAwRS6t3tR9tN5fnnJOV+7lGP/OalvuN+s5xrs+0Zdm30PkqTIWwaqOG7WB4fsbUtStWIvjpJK+e/LQL/dX2OdvSRBOrAkQtLxDRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQkw6hTLtpPp5ZUmKvRRKJzkynfZ3Jx3Zc81QCmWtZrcpOMk79S1+YltvX79Z7zxmqVnP1dvpOpIU9+w1615SkiRVnJS3mpOCOK0USiexrRYH0t+c9zgObL9Wsfe54uyXEyb6PCd5Ko6dpNNppFBWYj9lrRr5x4mjh3NJa1annfQoSfVO6uoZS+wxZfYxdgqgJP36qYfNenOX36bWNsus948+ZNbbG/zxqSOfM+ubBu0UxGLOH7fzOTtts5CztyFJoyP2dhrGR8z6MY11bl+dssehec32ftWVA+nBzvAQNfopb1ln37wxLZNrcvsa7rWTlQdb7GN8escet6+ZdfZxdi3y73Xtzf5xAoeKl1AZTqU+PI2N2Z/Dx558zG3zzGo7WXho2E6OHB6z65I0NjRk1ovDw2Y9lCI+c+ZMsz6jtcVtc/KKk8z66OioWW9q8sfHpOMbOAAAAABICCZwAAAAAJAQTOAAAAAAICGYwAEAAABAQjCBAwAAAICEmHQKpSeVTruveQE/mbSTppXJun1lnYTKXDCF0o4AmznT3k4c+8fS1WGn5fT128k7o4N2+pkkjRft1zJZP2WtVC6Z9Th2UiCduiTFNfs1L48plfKTmqpOX6FsxprTn1cPHIpqsZMuFWgzVV7SpiTV0qRQQurp7jXrtar/2Wmrt8eh9lE75Wvvjo1uX91ZezuprJ8C+Ksn15j1sZo91pTL/rEsrLcTfEdKdjznUNlPds2m7M9U2Rk3JSlVscfheXV2X3MiP2VtbrOdWtbVZtf3bu9z+xpyxvTWhce5beqc5NDymJ1amgncxts32emchTE7CbkUGOseenqzWX9d/Ty3Tb6N3xEDB1PsJI/v3LrVbfPEI4+Y9WLFeX5JhRLhnXuNk2KeDswRimP2mP70k4+7bebNsdOTc1n7fprP26nzB3otCRhdAQAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJwQQOAAAAABKCCRwAAAAAJMSklxGoxk7caNWPHc5k7PlhKrI3m0r5caOhKHt/+/Z24tiZt9b8+WxLU6NZ37TFjm4tpP39HSyOmfW2tk63jRdlX/UC+0PnyznNcdp5L2t+X7Hz/gej953uopR9/iMnNlfyDzOW18Y/lsh5LR24LuPM1K9LHHlGx+3PdMqJ0ZekmXOWmvXePXvMemum4PY1VmfH1a/bvMFts2fvLrPe2dFg1uPxottXuWq/tqy91aw/0T3q9jWesuOoFdvLG0jSbGfoXpq37wF1KX98mj3DPv68cz+rVQO/B821muX6ucvcJnV5Ow67f3CTWR/q7Xf7qo3aSxyMDNhtqgX7OpakwUF76YEtm/zjLwTuacDBNZ17cajNFNciCnbl9OWsd+Qt9SRJDfX2+HTuq17ltnnqySfN+pZtO8x6oc6fGsRV757mHYt/D6xU7KVR9uzZ7bZ57rnnzHpTc4tZjyJ/fOro7LDbuC0OL3wDBwAAAAAJwQQOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQk0+h9BJmyk5imKRslLfbONGBVTfdRvJCHWuBvJiUk2roJQJlc/b+SlJj60yzXujvNutrn3nW7WvdVjthZ2a7vQ1J+r3TVpj1dN4+xnLNf1+UcpIrI6ceSGNKOQk/VS91SaFEUbtNFOir6uxznLKvpTjwO4uUkyiZDiR6xoGEShw9Zs9tM+uNKX9MKTvX1QObtpj1JXXNbl9rd/eY9cGyn9w4t9NO1m2dae9zNWOnI0rS0AY7zWymM3DPz3pjjbS3OGjWs3n/c3hswd63rqz9eS95qcqSWpzkypGiPaZGgTGlPm8nxtUCt95dTz1q1vu2PWPXh4fdvoaGRsx6xUli7h/1Uygzzvuyrd9PR11am+O+BsSBe/uURYH79HT6i6eYRRg4liiyn0ciJ0U8FQWSr536zC4/8fWMV5xl1nf8v++b9WJx3O0r7TxXp5zzX636Y22pZKdQ5rJ2qrIkrX5ujVmfO+8Ys75g4UK3r2LRvj/m895923+P45p9nFHgunyx+AYOAAAAABKCCRwAAAAAJAQTOAAAAABICCZwAAAAAJAQTOAAAAAAICEmnUJZcxJWaoHkyFLZTphJO6kscaCvWsVuk8/6KW/ePsc1u6+6+vrA9u20mu6ddmJcod5P5prV3m731WenfEnS6mfXm/Wlxy8y61EgHdFLdcw6yUehMKbYSbRMh1KUnP68dKOak3QpSW4IqXOModQrL7Q0qgbSraaaVIUjUkO9nYKYy/jjwM49fWa9Mlg062sftdMJJUl1LWY5G/lD/Lw5dkJga7s9bkdN/men2GanLXY/ucusL230x4d5JfuDWCj4bY6rt49zaNw+l/lAOmhjzd5O97id9liM/KTPuGqnQMbDvW6b8l77nBV77DbFip/yFlec9DunSa1iv/eSVIzsa3zLgJ+CWVm91n0NCCX0ec9vXpM48Dn00rJT8pN1D+b3G9Wa/7myRN7DiPzn12rNH5/POvtss755i/38+otf/IfbV31dnb1fzjkOvcdeCmQmkHjcPzBg1tesXWfWZzv3OUkaHbNTd9MZ+36SCbwvL4XDa28AAAAAAC4mcAAAAACQEEzgAAAAACAhmMABAAAAQEIwgQMAAACAhGACBwAAAAAJMfllBMp2HHHZqUtSxYk3zjhtYvl9lZ09reb8NrlczqynU3ZE6VB/v9tXPGK/1tlib2PPoB+dumypHfldW2NHukpST7cdOd6yy16SYP6CTrev8ciJTnXiXjOZituXl0edCryXsbzYa3v7tZQfj+vG+Md2FHg1CvQVOdell7ktKQ4sMYCjx64dQ2Z9dMz+3EpSY67RrjfaSw/MnjfL7Wv5sSeb9Xh03G3T1WovmzJQ2m7WR9N+9P1gzd7OQGyPHV05P8b/1EVLzHp9zl9GIDu016zvHdpq1lsK9rmXpKhs77OXIJ3KBqLQy6NmfbzHXipAkuJBOyY7Hrcjt8ux/3tYL8I7jp2xruaP28rb8eE7Svb9RJIW5P3zDIR4S/6k3PV+/M9h/7C9RNPYqB/vXyra9/1KyW5TyPrjU0eH/ZxWdJbaKpbsceN59vHnneddSWputp85zzvv9836o4895vY1MmzvW13eH9M9VWfpsPFx/76Vy9v3xw0b7KW2Fi+2l9qSpDpnSQRv7pCpt39ekuQ/Wv7O8A0cAAAAACQEEzgAAAAASAgmcAAAAACQEEzgAAAAACAhmMABAAAAQEJMOoVSVTtipeKkyIRei2p2mlY65af9pZ09rVUDCYFOilE2Y6cV9ffZSWaSVGq0kytbGuwkN3Xv8fcr1WD3Ve+/HR0tdsJPNG6n3EXjduqRJLU1zzTr3f328acjf57vBe94iZaSJCftseq9l17SpKR02tk3b8cCiZbV2EnUDBxLKvKTp3D06B2wr+lK1U9wzUR2All9vT3WNNTb6VuSVBsbMevZip+yNjBopweO5+zrfXDI/+xUvDDYRifRsWKnoknS3DkLzXoukGw7ULSPXzV7fKgEPtOjsX3OUhX7WNrqm9y+yqP2Of7/tXdnP5Kd93nHf7VXV1VX9d49PT3TPQspUZQocZFkURZIaqNhOwHiixixYgSG/xHnDwiCXDuAY1hBAmQVEgORpSBGbCYURVHiOkPNiLP2Mr13dXXtSy5yYyPPc9KSLBtH8/1c/qrf99TZ3nPeLuB5j3e3fJuWTjQdmHHwzKTiRUScdnu6zcA8G3MJ46Op7w/08zwiYnP32H4GjMf+/c09d5un+v547Y3XbV9vvfW2rN++7ZO/W0197zSbOtHyS198zvb13Aufk/X7Dx/JeteMNRERu9u6zZPrl22bf/Bbr8q6S2h89eu/Zvv61n/+lqz3+3ocsKmh4d/Rk66Lgkm77PX0+bp9+7bta3l5WdaLBZ1CWZnyz2C3n24f/ybwCxwAAAAApAQTOAAAAABICSZwAAAAAJASTOAAAAAAICWYwAEAAABASjCBAwAAAICUOPcyAmMThTlKWEagb+KNJ0PdJp+QyJ4zm8kkxEEXizoKNFyThLj8ncO2rLfyZl9KZnmBiCiUdBTp6gUd7x8RsbO9I+u3Pron6+2Oj3aeX9Vxs32TvZ8v+ONiE2IzCcs7mP8buL6SzrGN+M+ayO2BjpqNiBhPzLnM+NtknPnFRcQiPXpmaZRi0Y+P7a5etqPb1+PW/oEegyIint7QcfWXVhZtGzdEHR+dyHo/Idp5GHr/Fy7qiP3FoV/mJGti/I8PE5Z5GehjU8jqh0p77M/LqKJjqrNnOr684NZQiIieGYdHJgo9IqLf0efSLRbgxu2IiJ75bBDmXCYse9E3Md3thKVZ9o71vgD/Pxlz7/7gB2/K+p/+t/9o+2p39X3Y6fsxrdnWY8re4YGsDzJ6+ZeIiP/yZ6/JerEyp/tKuKe+/5re/62PHto2X/3ai7I+XdfPh5deesn29aMfviPrH958X9YLhYT3p/FPv4xA3izNMjevnyn37/ulItbW1mS9VqvJeunMn+NaTS8Plsyd54RluP4KfoEDAAAAgJRgAgcAAAAAKcEEDgAAAABSggkcAAAAAKQEEzgAAAAASIlzp1AOxjoDa5wUlmJiBScmscylE0ZE5HL6qxYKPrqyUDRtyrpey+r0sYiIfFt/dmV9WdYnQ5+i0zdJiGOT2hkRsW/SvFYv6eSbjg8Tizffek/3tab3ZfXSjO0rk9HnMpOQzjiZ6IsmY5J3JgkplDmTkJk1h3+Q8cc4TOpVJKRQjib+POPxkc92ZX1pRacwRkS0mvp6ax7oa2qqptNrIyLGeX2PdEyiY0REMTcl6/22TmicT0i0PDEBmSfNQ1kvLOiUr4iIQbsp693DbdumndNf4MykBB/1Wrav6UN9nBdzOh00O/JJi928PpcVk74WEdENPXhP5Uyi5MQP9oOsTk2rmeul1T+zfY1MMlxxbNKeI+KQFEpEhEvbS3i0x3Co3y0+/PENWe8Pj21fWTMOzC3WfRtzvy+tXJf1S1ev2b5e/+G3zfb1+0tuyt9TuZy+p9ttf+92u/qdc2LS5efmZm1fX37lZVm/e/cnsn7W8unJ5bJ+r26bBNCIiKyJqz84cEn5/h3thz98S9aXl5f0tk26eUREyexLPpcQr+/SgCOpzV/5Puf6KwAAAADA3zkmcAAAAACQEkzgAAAAACAlmMABAAAAQEowgQMAAACAlDh3CqXLXinmEpIjTWLLOKsTdpLSWiqViqw3Gj5FaKqiU9Zcik9uSv99RMRm91jWX3/rHVnfuLBi+xoMdSJSr+OTd+o1ndq2tjYn629/cMv2NZro7Zyencr67R/7dKNrT6zKukuUjIgYmv2frulz2enrBKWIiHZX78t4YNIxTTJqRETeXBfDpKTJgt4XPF6uXZmX9flln7Z486ZOVRyaa7dU9ONj8+xA1k9OH9g2Swt67Fg1Y1e94cfHTLEh69ubW7L+3on/XisLa7JeGPt7LWNSkjvm3t3pJSQ37h7L+pNm/8cJ/wadNqesnjCm5U3aZKZgkkZHPmm01ddtRn39pQcjn445MmNnLmF4XFnS9wUQEZFLeB5/cON9Wf/+D97UfZX8+Di0yd8+jbY2pd9H1i9/TNbPWvr9KSLi4b2bsv7RLb2PFZNoGBGRNemc3YJPCR4O9HiXDf3OEyZxNiLis597TtZvfPiBrP/pf/0z29fYpCT3+j69tmQSOnM5PdblEuYVW1v6OfTRR7dlvTGnn5kRESdn+j25Me3fAXIJ78nnwS9wAAAAAJASTOAAAAAAICWYwAEAAABASjCBAwAAAICUYAIHAAAAACnBBA4AAAAAUuLcywgUTNzlMHzcaN7EeparM7I+PT1t+6pM6WUEaiZ6PiIimzERygP9nUcJccjPfPoZWb/1Yx3Duvnwru3r9FTHjbbOfHTqXFVHWD/c2pX1Qd9HS+fMvP1of1/WDw/98gYzs1VZX1jw5yUz0eelVpuR9e6BjkiPiBgO9Ukzm4hcQUfQRkRk8+Yad51FRMGngeMxsrqiI5wPT/y1mwsd2V6f1tdouerjkLtDHWG9vKLj/SMiGvM63nj3SI8DffN9IyLaHT12DQZ6fHxw6iO3Txo68rqRsGJHLacfZcWifm7sHOh9jIjYN7d7z/y7czT24/ZTeT1uXxz6ZQzqFR0hfjrWg81ZwjInHfPRxMS3TxIegm4ZgWzG78vSkn4+4HGjbyo3bkREfPs735X1e/c2ZX1+0S9Z0R/osStf8NduPqvfexYX9Vh/58FD29fhvl4y5tr6hqx/7StfsX3Nz87I+pXLl2ybORN/3+vpJQnKZf+eVKvq58Yrr3xZ1v/if75u+9ra0svM1Bt63I7wy1ANzFIJOfNsiIjodvX+v/eeXt5hbf2q7Stf0M+6qlnOLCKilPfH+Tz4BQ4AAAAAUoIJHAAAAACkBBM4AAAAAEgJJnAAAAAAkBJM4AAAAAAgJc6fQlnQc73xyCeTzc3MyvrGpXVZn4x9ApbbSqHk02o6fZ1K0xnpdKFeXyfSRERcrOnt/MN/9NuyvvXgx7avD97+kaz/j794w7b54e2P9PdaWpL1pVmf6Dlb18lgRZPQ2Lnk5/nDrj5nhUzZtsnldELm6XFTf6+cTveJiJir6rTL/b1Hsp4d+yS/nElNjby/TaoFf5zx+Lj3oU6Dbfd9guvC/IKs98t63FpfX7Z99Qd6OyenR7bNyWlX1scDPQ7cvX9s+xqbBLC+SQzLm7+PiLi9r4/lEyZVOCIiY0J3sxl9754lJA43R7qz44MTWR+Hj6It5nQC2RWTqhwRkR/q7fcz+inYyfrjctLTicf9cOOzPzA5l0I58eeyn3D9Azu7e/azv3xNpxdu7ehk380dn2zb7+uxrlLTia8REY2GfraPzD2ysKCTHiMiXv7i52X9qy+9JOu/843ftX1lsvodppTwntLp6nHAJTdOTfnjMp7ocejKFZ3Q+PVXv2r7+uY3vynr/YR38bzZz9FIj8OThBTxUknv59b2jqy/b9IpIyLqDZ34nE1Iby4t6eTUvHlu/L99AwAAAABSgQkcAAAAAKQEEzgAAAAASAkmcAAAAACQEkzgAAAAACAlzp1CWZ3WaX/5nk+LWZ5flPWFWZ3Ws7ur08ciIgZDnZYzDp8q2Grr5KFOT6fVNE99itHWwy1Zr1Z1mlcuIXnmK196UdavX9mwbf7lH/8HWb99956sd/o6ATQiYn1Vn5dnPvmUrI+G/hi//97bsn60t2/blCs6+Wds0oUGA51kFxGRMclDtZzuK1Pyl3yupNP38gnJbPVazX6Gx8fpkU7bGyYkFI7qOtXwY09ek/WEMK3Y2tVpk3v7x7ZNr6ev63pVp2lNTDpjRER2rMfnokkza57pxNmIiBuP7st6puHHtOh2ZHmY1fd7pexTvrZMOufQpGAOMj6x7eZJS9av1XxK74WxPs49Mw61E5411dkZWW/um2ddJmGsm9KpfOOE8Xll9aL9DNjd14mSERGb2zqh8qSp7/VC0b+nTMY6ybrZ9GPazKx+H5iYcWCq7MeBf/zbvyXrX3vlZVkvFPy+jCb6d5eBSa+NiBia92e3nfHY39MjM0TkzFj7iU983PZ19apOpL9xw6e4T8z+53K63u8nnOOGbpM3u//e+z6FcuPqhqxnL6zYNnfv6efD9av+mP21vs/1VwAAAACAv3NM4AAAAAAgJZjAAQAAAEBKMIEDAAAAgJRgAgcAAAAAKcEEDgAAAABS4tzLCFxbvyrrZy0fvV/I6O73d3V07NbOI9tXtVKR9Vzfxx63OzoOujfQSx8UEmK6hx0dE37/zl1Zn6roCNqIiEebermEel3vY0TEF579pKwvL+rlHfZPz2xf2wc6nve1N78n6+trT9i+uhkdzztbn7FtWs1DWc9OdBz27LSONY+IqM7qY9YZ6ajhXEF/34iIQln3Vcz7/3Pkcz7uF4+P6aoePC5f10sCRESU63rsGkz0uLVtYrUjIh5sbuptTOnxISIik9fbn13SEfsXltZsX5Wiidau6L5u3/ZLxlS6uq9WU4/BERGtMx3HXGnMy/ooIS7ffeKWERiacSsiYtc9N06PbZulho7rH5g47KJ5NkZE1PP6+G8f6uOVT/ifrvtkkvFtdhOWsQAi65+fC4sXZH2SNXH95v6MiBj09T2acOvG+sZ1WV+7dFnW3/nBD2xfSyW9odbBjqwXazr2PyKiP9HvMLuP/PvzaUsvWXPtmn6vr1b9Mis5884zGut9THpF+sY3fkfW//AP/5Vt88H7N2V9fl6P9T2zxExExGRsnltTev9bHd/XO++8I+sXlvWyXRERj8zSaSwjAAAAAAC/ZJjAAQAAAEBKMIEDAAAAgJRgAgcAAAAAKcEEDgAAAABS4twplPO1qqyXxjoZKyLirK3T1Fot3Wb/8Mj2lc/pr9oo+LSccl7H3/Q7OoVyblrvY0TEwkxZ1k9aTVnPhk8G29vTyTPdO/p4RURsm5S5mJjMtKHex4iIqZJOcVqcn5P1atVfJs/9yudl/b13dVJQRES7qb/bypw+/oWc/z/DpK/TmrKmSSZ8UtVCQ6cFFUzCXkRE11xLeLxcv6avnW5CzNk77+px4LSrr6lczl+Hy8sLsn5xVdcjInIFPT72+j1Zn1us+e0v6gSw+3fuyfr+iU8vXqwtyfqnVn0a7cOi3pcfnegUyL2u334m9PiYGeuxdmFWPxsiIjbm9b6sz/l00GxLJ0S2RyO9/SU9bkdEDMv6u93f0c+tzNCn35119fMpl5Do+ZE5/0BERLHo07qvmwTf1Ys6DTebS0iWHZvkxJEfU596Sm9/49IlWb+46NMGZ0z49e6jfVkvnOl7PSIiivo9KW/G84iIUkkf561t/V45nvj3mk2Tol4o6HFzbs6PT6urOmn0a1/9sm3T7+nv1m7rhMhi0aSWhp9XuPfEXs+/o9/44ANZX1+7aNt02zot/osv2iZ/Db/AAQAAAEBKMIEDAAAAgJRgAgcAAAAAKcEEDgAAAABSggkcAAAAAKQEEzgAAAAASIlzLyNwdKijQwd9H6vZMRHrnb6eNxYKPlJ2ONJLD0xP+2UEZss6dvroUP/9oKMjPSMiKiamOj+j46CPTvySCAWTXLtgIqcjIuarOg6619fHJfPhR7avxSUdd3v9Y9dl/ehox/a1WteRtqMnn7ZtvvPfv2s+0dHeuayPxx0c6jjs1RUdXXtyZk5+RNSrev/7A79URslcF3i81MwSJDd+5GPUm01zT2f0ADEOPz6V8/qzw20dox8R8fkvPC/rAxOh3Gj4ZQRqdT0Oj7I62vvoVMdnR0TcP9XPlPxVfX9GRKzW9Dj89pEeU7J5v7xD0fxfMzvU+1Kr+jFgfWNF1tdmZmyb0U09dre6bnkBH5+eKehH/Cj0/o/GfhmBoVkyKOceaBGxvOKXsQBGZsmSiIgY6/fHfFZH7Bcy/veIgVnqYjjycfnjrn63qJvrffHiVdtX1ozpo6G+p8YJyx1NTemxNmmZmeOTY1n//vffkPXvfvfbtq9eXx//Tz/zrKw//8JnbF/tth6fly/4pQeuXtXLONy7q5dEyOfNGg4R0e3o6293b0/WRwnLKyzM6flG8/jEtima8fm8+AUOAAAAAFKCCRwAAAAApAQTOAAAAABICSZwAAAAAJASTOAAAAAAICXOHYGytbUl64WsT3jJ5Uuy/mDzoawXywkplCat6KO7d22bclVvPzvRSVtZk3QZEbH5QG+nUtfpZ/2EdCWXf1Yq+WM5N6MTKrd3dDrok1c3bF8uhTJrEshyGX+Z7D3QKXvzMz5F6KUXPyvrf/m9t2S9+9An1j391KqsD/I6KSkz6Ni+Fub19dJt+3Sns9ax/QyPj4NDnTq7vKLHh4iI4USnCrZMsmou58fHalGnYC7M+DGlZsbbSdFtx49p+/s6AWzt0rKsP/ecTzs8vKmfD/eOtm2bpVFF1ufNM2Chqv8+ImK/pVMluy39fJiZ9X21ujr19uGBP5bVnj42nZH+f+sg58fnwyOdgHba1umkeZMaGhHhQv4yiYl5+vgDERGjkX5OR/h3qFbLJOv6YNmYmM8Kef8bxv6DO7L+vT/XKdpfevXXbV+LS3oczOb1vZPL+e+1ad6fb9y4Yds0mzrtsWjG+he/+Ku2r75JoaxW9DPonXfetX3dvPm+rP/v1/+XbfPG6/o9cePyNd3AD09xsK/fLas1vS9JSZ/PPatTOH/j1/11MR778fY8+AUOAAAAAFKCCRwAAAAApAQTOAAAAABICSZwAAAAAJASTOAAAAAAICXOnUI5GugYn+HIp2nNzukkl3ZXp6xlsz6xKlvWX3Xz4Z5tU2no7WddXNHYJyJ1ezrtMVvQfa1c0OmIERHdkU45G2Z8CmajOyXrGbMvq5dnbV/Lq/OyfnrclfULVy7avvqnuk1Wf92IiNi4rhMq16/qpKb/9O++ZftqNc9kfVjQx/LjT5ikoogYj8qyXq/p4xURsdhYsJ/h8XHW1umml66t2DaVhk4brB/o/6vd+YlOFIyImIQeU6p1nTgbEbF3rL/zSUuPqWuXZmxflZIenzMm8fepjQu2r8Oivg9bW49sm/K+HocWZ/Qz4OkL/ri8/xOdKLp5ptPvLl/048PlNT3WZfb1uBURsXtPJ6Mdm8CyTMank05MsvJkqJ91E/M8i4jI5XSiaUKQYEzXGv5DPPZ6XX3fRkRsbupkWzfWZhPSWDMmQrWekJLa3dXjbe9Ej4/XP/WM7WswMsmyJg1236QjRkT80R/9say/8cabts3MjH4ffPrpp2U9n9fPk4iIR4/0u3CrpVOVDw4ObF+9nj7/rZZOzYyI6JzpMc0HOiYl6+pUyd///d+T9Vdeecn2VavVZH1uzieyZ7PnnoLp9j9XawAAAADA3xomcAAAAACQEkzgAAAAACAlmMABAAAAQEowgQMAAACAlGACBwAAAAApce4My4aJ5G81dQxqRESvp5cYyGX1vHGcEONfLuk2Vy77uP5JTkfEDic6VvTg2MedNrs69nmpPiPrjdq07StaOlL2oOljwnd3d2R9xkRI1xo++v+odSjrt+48kPW1vo73j4i4sKAjUquzM7bNuKeXkbg0r8/xr738GdvX/Vv3Zb3T0tdSp+1jsrfv69jibMbfJkUTu33FtsAvo9kFvZxEq9X3jcyyFZ98ak3Wsxl9fUZE7O3paO3tpl+a5M7elqyXizouvl7398Gl6/o7nx3paOn5uo+XX5nXx3Kw6pdGOf7zD2W9PKW/87UNPz7+ZFNHWBdLOlr7akJfjRm9/Qe7Cc+avB5TmqEjr4f7etmDiIhmT18XRROrHhP/DDYrEsRk7CPHH9zXzy0gImLilnSKiJ1H+trZ39PvL4WEJQHK5Yqst4q+jbl1o2HGrn/zb/+97at1pt8fi0X9/tbtJC2vYN4FZ3xc/eGBPmbvvfu+rJ+Z7xsRYYahmJvT4/PEn+I4OdbbGQz0O3JERK9nlmTo6PlGpeLP8fKyXk7miSf1clMbG1dtX+Ox/l6TpAPwc+IXOAAAAABICSZwAAAAAJASTOAAAAAAICWYwAEAAABASjCBAwAAAICUOHcK5VlXp+KUSjpFJyIiYxKt8ibVbzD2XyeX14lt87Pzts1wovvrD3Uy3CTrE9tWlvR26vW6rGeHPs1rpqZjfMYdE+8TEWOTNlke6vq9n+zZvib39We90Od4a3Pf9nX39o9l/Ynr67ZNqaCP80lTJ9ZNev66KBR0AtrIXJc72zoBMyKiNNafjUc6tTQiotLw1z8eH8Wcvne3D3yy7FFLtzk+25X1Uc+nWY1HOjlyc+vYtnHJaCvLeqzbP9KJhhERtx9sy/q8SeNt5BMS48yY2s34/b83pVPmJhN9jC/kfXLi1Rmdgtk26XfZQkJyo4lsyzV8SvG4odPU4lSPj4sNnxK8/0gnlxZK+noZD/0xHoz1Z7mi/z/wYdMnZAJXrvhUv/XL+h1ie/uRrPfPfOLvaVu/23SqPtl2UNLvnLNDPXa1tv21fnyinwN9k9ReMEnAERFrFzdk/eTk2LZpneqx++REjykuHTMiYjA0x7KjE+nL5Snbl/usWPBtTk/0++vxkX5/e+GFl2xf03V9jpcW3bzCj/UubTKT8e/1Py9+gQMAAACAlGACBwAAAAApwQQOAAAAAFKCCRwAAAAApAQTOAAAAABIiXOnUB4f67SapbmEFJ++SdMaDmW5N0pIwDKfDcc+IbBQMmmXfZ2+Nun7lLWiSTnb3THJZCOfVlPWXUWv6xMSVxdWZL1amJH19969Yfu6YxLj1q5elvUra5dsX2/v6oTKe6+9Y9tUi/pcThV18tDqwqrta35pRtbv3L4v69mMTzfKFfT/M7pDfb1ERETbJ1/h8ZGZuFRD/z+yo7Yebx7unsr6uOfHuknWpJZ1/LVbmdJtFmZ1QmKh6B8XR22dLDu3qPuamHs9ImJgng9TU1XbppnXqWlnA/29npjRSZcREc+t12R92NLPwHLVH5dKTQ/2g+yxbTM1r5+pS0P9nfeaOhUuImL3WF9L/bE+LlmTNBkRkTOX8mDstz+w9wUeL/q6Wpyfsy1+75/8rqx/7rPPy/rWjn6viYh4+70PZf3Wrbu2zcAkV1amdHLhdFXXIyK6Xf0u3HJjintJjIheV9+7Zy0/1rscxHZbJ0e6RMWIiF5PP7e65lnTaPg5wvPPPyfr2YwfNz68ect8or/zeKKfJxERn/nMp2V9cXHxp9pGxC82bdLhFzgAAAAASAkmcAAAAACQEkzgAAAAACAlmMABAAAAQEowgQMAAACAlGACBwAAAAApce5lBGKkI6yPTExxRES3q2NYxyaqeKau45sjIqZNrGqv5aP/26f6s4mJUM6Fjy49O9IRqUPdVeztHdi+ckV9LBs1EwUeEdWyjtC+sHpR1rttf14GQ31c8ll9vo72N21f9ZKOAz94dGTbbB0ey3qloiPHux0fz3p1VS8xUJrS11K17M9xuazPy7Dsb5PC+e8g/BLLZUqyPlP1EcqFkb4PM+bfapmyjzCulfU9srzkY7ovmCU4PvmJDVmfm/fjc8HcCFNlF63t/3c4Mkuw9BKWZmmbqOjToV7m48yMgRER7aNjWV9fWZL1XEL0fqdrYrrHfvmRwxO9nMxgqJdKuLW7Y/s66eiY8on5zklB2OOJHh/d+YqIGA75HzEiIvS1Uyz66+PVr78i619++VdlvdMxL2MR8cEHN2X9D/7gn9o2t2/flvXjQ33vfOyJz9m+Lq6uyfq3v/0dWW+e+Penh+ZY1qb90gML8wuyPj83L+vPv/CC7evTn35a1ut1/Y5arfrlXy5fXpf1fN6/WO3u7sr62Zl+R69U/JI1165dk/V6XT+3RwlLnbGMAAAAAADAYgIHAAAAACnBBA4AAAAAUoIJHAAAAACkBBM4AAAAAEiJc2fotbom4afvE6iqUyaZraHTzBqzPjGtWNAJXI9292ybdlcnfV1Y0QkzhbzeRkTEnfuPZD2b1XPg1qlOEouIWF1blPXZhBTOzplOiGy1TmR9vq4THSMiXvqCThgal3S60SAheWe5pRN+lqZc+pxPjLt4YUPWm8f+HE8V9XYuXNDH+ORo3/Y1yunrZe+RPvcREReWG/YzPD7WLs/I+mjk78NcUY+de8f6nnbpuRERzzz9hKwvz/kxpTFbl/ViRbfJZX1K7nCox4izpk7DzZR8X5mJ7quWkCaWqej+mkc6za3Z9Sm9rTM93kwXdJpadqKPY0TEwMQUNxp+X46mdELmowNdb418ouXEJKNlMyaFMiFJzSVXJoRwRiGv3wHwuEm4SGwT/T5SMAmFpYZPYXzxC78i6//in/8z2+bW7Vuy7t75nn32WdvX4qJOsHWJ7H/yJ//a9lWt6rFjfl4nXUZEfOpTn5L1V155WdbX13U6ZETE4qJOrvzZ/PTXxcbGlb+5rU/0M9gl6yaOj+a59YvEL3AAAAAAkBJM4AAAAAAgJZjAAQAAAEBKMIEDAAAAgJRgAgcAAAAAKcEEDgAAAABS4tzLCOzsHsq6X0Qg4trli7K+tKQjVXsdHRsbEbG1tS3r/aH/Bv2+jnBut9uy/mDTx9Vvbu/IesGkYU8XE+KTB3r7w4GP3u8N9H4e7O7KeutQx2dHRFy+pCNiC1kdT9tuHti+8iZVdTLS+xgRMTOj434rRR2Tvbihr5eIiN6gJ+vt+/p8be7pekRE+Uz3NVXy8e1VE/mNx0utrseuWklHz0dEXLz4pKwf7B/LeqmYs31dWluQdZMWHxERbTN2Ns3SJMWCj8t3S7D0+3r5k2Hfjw/Vkh6HTFf/t820Ps6FnH7EjUb6Xo+ImF3Uyyhk2vpgVqo+vvykd6z7yuilVCIiKiYOffhQjzWZ8NHWeXP+x6bN0MRqR/gI7WzC9rtu+SE8Zn76iPVsVo93Lq19bJYnivD3yNOffMq2cZ/ZPTHLHkREZDJ6X/7e3/8NWX+UsHRRoahfOj/+cf08iYj4zd/U29nY2JD18djvy3DojrNukxSv78aUnyWS37fx45NbFeBn+V5JSwz8ovALHAAAAACkBBM4AAAAAEgJJnAAAAAAkBJM4AAAAAAgJZjAAQAAAEBKZCY/S9wLAAAAAOBvHb/AAQAAAEBKMIEDAAAAgJRgAgcAAAAAKcEEDgAAAABSggkcAAAAAKQEEzgAAAAASAkmcAAAAACQEkzgAAAAACAlmMABAAAAQEr8H94XQmk98YwYAAAAAElFTkSuQmCC\n"},"metadata":{}}],"source":["label_names = ['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']\n","\n","# Plot a few images to check if the labels are correct\n","plt.figure(figsize=(12, 9))\n","for n in range(9):\n"," i = np.random.randint(0, len(train_data), 1)\n"," ax = plt.subplot(3, 3, n+1)\n"," plt.imshow(train_data[i[0]])\n"," plt.title('Label: ' + str(label_names[train_label[i[0]][0]]))\n"," plt.axis('off')"]},{"cell_type":"markdown","metadata":{"id":"giLJg2i4GkCU"},"source":["## 16.3 Creating, Training, and Evaluating a CNN Model \n","\n","### Create the CNN"]},{"cell_type":"markdown","metadata":{"id":"i4EJITkbGzdZ"},"source":["Next, we are going to define the network architecture with the Keras library. Recall from the previous lecture that the main data structures in Keras are **layers** and **models**.\n","\n","In the first cell below we imported the layers, and afterward we defined the layers.\n","\n","We need to first introduce an **Input** layer to provide the size of the data, and in this case, the shape of the input data is set to (32,32,3). This is because the images are arrays with shape 32x32x3.\n","\n","Next, we will include a convolutional layer **Conv2D**. The arguments of Conv2D layer in Keras are:\n","\n","- `filters`: define the number of filters in the layer.\n","- `kernel_size`: the height and width of each filter, defined as an integer such as `kernel_size=3`, or a tuple such as `kernel_size=(3,3)`.\n","- `padding`: it is not a very important argument, when it is equal to `'same'` it means that the output images are the same size as input images; otherwise, the output images can be slightly smaller.\n","\n","Next, we will include a pooling layer **MaxPooling2D**. We can specify the pooling size with the `pool_size` argument, such as `pool_size=3`. Otherwise, the default pool size is 2.\n","\n","ConvNets typically consist of several blocks of convolutional and pooling layers.\n","\n","And **Fully Connected Layers** are used for matching the compressed feature maps to their labels.\n","\n","And there is one more layer that is used called **Flatten**. This layer is needed because the outputs of the convolutional and maxpooling layers are 3-dimensional tensors, but the dense layers require one-dimensional data (vectors). The Flatten layer just concatenates (flattens) all dimensions of a tensor into an 1D array.\n","\n","For example, the following layer that we defined below,\n","\n","```\n","conv1b = Conv2D(filters=32, kernel_size=3, padding='same')(conv1a)\n","```\n","\n","is a convolutional layer, which takes as input the previous layer named `conv1a`. The layer `conv1b` has 32 convolutional filters of size 3, and padding is applied to preserve the size of the images."]},{"cell_type":"code","execution_count":11,"metadata":{"executionInfo":{"elapsed":4,"status":"ok","timestamp":1731443912144,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"e_YS-JqfkyiK"},"outputs":[],"source":["# import the layers and the model\n","from keras.models import Model\n","from keras.layers import Input\n","from keras.layers import Conv2D\n","from keras.layers import MaxPooling2D\n","from keras.layers import Dense\n","from keras.layers import Flatten"]},{"cell_type":"code","execution_count":12,"metadata":{"executionInfo":{"elapsed":1245,"status":"ok","timestamp":1731443913386,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"gjqyblPbHmmc"},"outputs":[],"source":["# Define the layers in the model\n","inputs = Input(shape=(32, 32, 3))\n","conv1a = Conv2D(filters=32, kernel_size=3, padding='same')(inputs)\n","conv1b = Conv2D(filters=32, kernel_size=3, padding='same')(conv1a)\n","pool1 = MaxPooling2D()(conv1b)\n","conv2a = Conv2D(filters=64, kernel_size=3, padding='same')(pool1)\n","conv2b = Conv2D(filters=64, kernel_size=3, padding='same')(conv2a)\n","pool2 = MaxPooling2D()(conv2b)\n","conv3a = Conv2D(filters=128, kernel_size=3, padding='same')(pool2)\n","conv3b = Conv2D(filters=128, kernel_size=3, padding='same')(conv3a)\n","pool3 = MaxPooling2D()(conv3b)\n","flat = Flatten()(pool3)\n","dense1 = Dense(128, activation='relu')(flat)\n","dense2 = Dense(64, activation='relu')(dense1)\n","outputs = Dense(10, activation='softmax')(dense2)\n","\n","# Define the model with inputs and outputs\n","cifar_cnn = Model(inputs, outputs)"]},{"cell_type":"markdown","metadata":{"id":"2nL-Znyr1WKt"},"source":["Also note that in multi-class classification problems, the last Dense layer has\n","**softmax** activation function. Softmax activation outputs a multiclass probability distribution, that is, it computes the probability that an image belongs to one of the 10 classes.\n","\n","After we define the layers, we create the model as an instance of the `Model` class, for which the arguments are the input and output layers.\n","\n","We can inspect the architecture of the CNN with `cifar_cnn.summary()`."]},{"cell_type":"code","execution_count":13,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":577},"executionInfo":{"elapsed":8,"status":"ok","timestamp":1731443914133,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"nVvtKIGF1gnc","outputId":"31be19f3-c85b-4b1f-c335-f7751e7bb414"},"outputs":[{"output_type":"display_data","data":{"text/plain":["\u001b[1mModel: \"functional\"\u001b[0m\n"],"text/html":["
Model: \"functional\"\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n","┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n","│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m3\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m896\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m9,248\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m32\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m18,496\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_3 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m36,928\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d_1 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_4 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m73,856\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_5 (\u001b[38;5;33mConv2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m147,584\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d_2 (\u001b[38;5;33mMaxPooling2D\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m4\u001b[0m, \u001b[38;5;34m4\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ flatten (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m2048\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m262,272\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,256\u001b[0m │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m) │ \u001b[38;5;34m650\u001b[0m │\n","└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n"],"text/html":["
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n","┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃\n","┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n","│ input_layer (InputLayer)             │ (None, 32, 32, 3)           │               0 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d (Conv2D)                      │ (None, 32, 32, 32)          │             896 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_1 (Conv2D)                    │ (None, 32, 32, 32)          │           9,248 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d (MaxPooling2D)         │ (None, 16, 16, 32)          │               0 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_2 (Conv2D)                    │ (None, 16, 16, 64)          │          18,496 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_3 (Conv2D)                    │ (None, 16, 16, 64)          │          36,928 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d_1 (MaxPooling2D)       │ (None, 8, 8, 64)            │               0 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_4 (Conv2D)                    │ (None, 8, 8, 128)           │          73,856 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ conv2d_5 (Conv2D)                    │ (None, 8, 8, 128)           │         147,584 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ max_pooling2d_2 (MaxPooling2D)       │ (None, 4, 4, 128)           │               0 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ flatten (Flatten)                    │ (None, 2048)                │               0 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense (Dense)                        │ (None, 128)                 │         262,272 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense_1 (Dense)                      │ (None, 64)                  │           8,256 │\n","├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n","│ dense_2 (Dense)                      │ (None, 10)                  │             650 │\n","└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Total params: \u001b[0m\u001b[38;5;34m558,186\u001b[0m (2.13 MB)\n"],"text/html":["
 Total params: 558,186 (2.13 MB)\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m558,186\u001b[0m (2.13 MB)\n"],"text/html":["
 Trainable params: 558,186 (2.13 MB)\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"],"text/html":["
 Non-trainable params: 0 (0.00 B)\n","
\n"]},"metadata":{}}],"source":["# Model summary\n","cifar_cnn.summary()"]},{"cell_type":"markdown","metadata":{"id":"XDYdR1SG1zh4"},"source":["### Compile the CNN"]},{"cell_type":"markdown","metadata":{"id":"GrJHF5Qe2wwr"},"source":["Next, we need to compile the model, and provide the loss function, optimizer, and metric.\n","\n","From the previous lecture, we know that the following three crossentropy losses are commonly used in Keras:\n","\n","- `binary_crosssentropy` used for binary classification (i.e., there are only 2 classes).\n","- `categorical_crossentropy` used for multiclass classification (3 and more classes) and the target labels are encoded in one-hot matrix format.\n","- `sparse_categorical_crossentropy` used for multiclass classification (3 and more classes) and the target labels are encoded in ordinal format.\n","\n","Since the target labels are encoded in ordinal format, we will use sparse categorical crossentropy.\n","\n","Let's use the Adam optimizer, and accuracy metric."]},{"cell_type":"code","execution_count":14,"metadata":{"executionInfo":{"elapsed":186,"status":"ok","timestamp":1731443917389,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"OZuZ18XZ2zVV"},"outputs":[],"source":["cifar_cnn.compile(optimizer='adam',\n"," loss='sparse_categorical_crossentropy',\n"," metrics=['accuracy'])"]},{"cell_type":"markdown","metadata":{"id":"f7Bqx-N03K1D"},"source":["### Train the CNN\n","\n","Training CNNs is similar to training Fully-connected NNs. We use the `fit` function and list the train data and labels, number of epochs, and batch size. Here we used a batch size of 128 images, hence, since the training dataset has 50,000 images, the updates of the model parameters will be repeated 50,000/128 = 391 times (notice this number under the Epoch in the cell output below). Since we selected 10 epochs, this means there will be 3,910 training steps in total.\n"]},{"cell_type":"code","execution_count":15,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":67130,"status":"ok","timestamp":1731443986323,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"zmkoLSQJ39IZ","outputId":"8e4ab10b-7145-4404-c7e7-4d925de89092"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m17s\u001b[0m 19ms/step - accuracy: 0.3784 - loss: 1.7053\n","Epoch 2/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m11s\u001b[0m 11ms/step - accuracy: 0.6625 - loss: 0.9644\n","Epoch 3/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 10ms/step - accuracy: 0.7361 - loss: 0.7610\n","Epoch 4/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 10ms/step - accuracy: 0.7787 - loss: 0.6301\n","Epoch 5/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 10ms/step - accuracy: 0.8139 - loss: 0.5329\n","Epoch 6/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 10ms/step - accuracy: 0.8480 - loss: 0.4248\n","Epoch 7/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 10ms/step - accuracy: 0.8763 - loss: 0.3537\n","Epoch 8/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 10ms/step - accuracy: 0.8935 - loss: 0.3009\n","Epoch 9/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 10ms/step - accuracy: 0.9071 - loss: 0.2639\n","Epoch 10/10\n","\u001b[1m391/391\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 10ms/step - accuracy: 0.9224 - loss: 0.2225\n"]},{"output_type":"execute_result","data":{"text/plain":[""]},"metadata":{},"execution_count":15}],"source":["cifar_cnn.fit(train_data, train_label, epochs=10, batch_size=128)"]},{"cell_type":"markdown","metadata":{"id":"dkO2ycHHrbgT"},"source":["### Evaluate on Test Data\n","\n"]},{"cell_type":"markdown","metadata":{"id":"ZT0dJ-vxs3pV"},"source":["We can first use the `predict` function to output the class for each image in the test dataset, and afterward use `accuracy_score` to calculate the accuracy."]},{"cell_type":"code","execution_count":16,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":5342,"status":"ok","timestamp":1731443991652,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"K79im89prfzv","outputId":"94c1c09f-c719-4f83-ae86-eb915b537470"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step\n","The test accuracy is 72.66 %\n"]}],"source":["from sklearn.metrics import accuracy_score\n","\n","preds = cifar_cnn.predict(test_data)\n","\n","accuracy = accuracy_score(test_label, np.argmax(preds, axis=1))\n","print('The test accuracy is {0:5.2f} %'.format(accuracy*100))"]},{"cell_type":"markdown","metadata":{"id":"mYF9aQTj-GTa"},"source":["We can check that the shape of the predicted outputs is `(10000,10)`, since there are 10,000 test images and 10 classes.\n","\n","Let's also display the predictions for the first 5 test images in the following cell. The model outputs probabilities for each of the 10 classes. The probabilities sum to 1. For instance, for the first test image, the model assigned the highest probability of 0.801 to the class with index 5."]},{"cell_type":"code","execution_count":17,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":20,"status":"ok","timestamp":1731443991652,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"Ai46Kj1d7p1m","outputId":"143ef0e6-b524-40fe-c7be-0314047199d5"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["(10000, 10)"]},"metadata":{},"execution_count":17}],"source":["# check the shape of the predictions\n","preds.shape"]},{"cell_type":"code","execution_count":18,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":18,"status":"ok","timestamp":1731443991652,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"Ph2r53Kt77Hg","outputId":"46822860-31f8-47e3-89d4-c25c24278022"},"outputs":[{"output_type":"stream","name":"stdout","text":["Predictions for first 5 test images:\n"," [[0. 0. 0. 0.616 0.007 0.231 0.141 0. 0.005 0. ]\n"," [0. 0. 0. 0. 0. 0. 0. 0. 1. 0. ]\n"," [0.004 0.002 0. 0. 0. 0. 0. 0. 0.988 0.006]\n"," [0.998 0. 0. 0. 0. 0. 0. 0. 0.001 0. ]\n"," [0. 0. 0. 0.032 0.951 0.002 0.014 0. 0. 0. ]]\n"]}],"source":["# display the predictions for the first 5 test images\n","print('Predictions for first 5 test images:\\n', np.around(preds[:5],3))"]},{"cell_type":"markdown","metadata":{"id":"bOG64FhY-zdH"},"source":["Next, let's output the indices with the highest probability for each image with `np.argmax`, and compare them to the ground-truth labels."]},{"cell_type":"code","execution_count":19,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":15,"status":"ok","timestamp":1731443991652,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"C7bmDQ2k8191","outputId":"af0a4581-7b43-4b4b-f7ed-6ece1bad18da"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([3, 8, 8, 0, 4])"]},"metadata":{},"execution_count":19}],"source":["# print the index with highest probability for the first 5 test images\n","np.argmax(preds[:5], axis=1)"]},{"cell_type":"code","execution_count":20,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":11,"status":"ok","timestamp":1731443991653,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"5r3luKAi9IJp","outputId":"dd7a905c-80bc-45f7-815a-5b04660dbbd1"},"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([[3],\n"," [8],\n"," [8],\n"," [0],\n"," [6]], dtype=uint8)"]},"metadata":{},"execution_count":20}],"source":["# print the ground-truth label for the first 5 test images\n","test_label[:5]"]},{"cell_type":"markdown","metadata":{"id":"AzPfBj7dtLWy"},"source":["And, a simpler way to calculate only the accuracy is with `evaluate`."]},{"cell_type":"code","execution_count":21,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":3138,"status":"ok","timestamp":1731443994783,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"BKW5SWlbqMBb","outputId":"17acdd7e-fc41-4f74-c47b-e0f12f8217e4"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 4ms/step - accuracy: 0.7241 - loss: 1.1521\n","Classification Accuracy: 0.7265999913215637\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"XYoQL6RTy8TX"},"source":["The important thing to notice is that the accuracy on the training dataset reached about 90%, while the accuracy on the test dataset is 73%. This means that the model overfits the training data. That is, the model begins to memorize the training data, and it learns to predict on the training dataset very well, but the generalization ability on unseen images is quite low. In the next sections we will learn how to deal with that. "]},{"cell_type":"markdown","metadata":{"id":"5D24LVEj0B6G"},"source":["### Plot the Predictions\n","\n","But first, let's plot a few images and the predicted class labels by the model."]},{"cell_type":"code","execution_count":22,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":521},"executionInfo":{"elapsed":500,"status":"ok","timestamp":1731443995280,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"GUzSpQw30BHp","outputId":"33a169e9-93ab-4e9a-d972-859a8d7f3677"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# Plot a few images to check them\n","plt.figure(figsize=(12, 6))\n","for n in range(9):\n"," i = np.random.randint(0, len(test_data), 1)\n"," ax = plt.subplot(3, 3, n+1)\n"," plt.imshow(test_data[i[0]])\n"," plt.title('Label: ' + str(label_names[test_label[i[0]][0]]) + ' Predicted: ' + str(label_names[np.argmax(preds[i[0]])]))\n"," plt.axis('off')"]},{"cell_type":"markdown","metadata":{"id":"4NBsfPgbt8v0"},"source":["## 16.4 Introduce a Validation Dataset "]},{"cell_type":"markdown","metadata":{"id":"bZkS3-A4z2ia"},"source":["To be able to observe if the model overfits, one more dataset is introduced, referred to as a **validation dataset**. The original training dataset is typically randomly split, and approximately 70-80% is used as a training dataset, and 20-30% is used as a validation dataset.\n","\n","This way, at the end of every epoch, we will calculate the accuracy of the model on the validation dataset. As the model is trained, if the training accuracy increases, but the validation accuracy decreases, it means that the model begins to overfit.\n","\n","In Keras, the `fit` function has a `validation_split` argument, which allows us to use a percent of the training data for validation. In the next cell, we will use 20%, meaning that out of the 50,000 training images, the model will use 40,000 for training, and 10,000 for validation. An alternative way to introduce validation dataset is to first manually split the training dataset into two subsets of data, and in the `fit` function to pass the validation features and labels as a tuple named `validation_data`. However, the first option with using a `validation_split` argument is much simpler, and therefore preferred.\n","\n","Note also that in the cell below we will just continue training the same model. If we re-defined and re-compiled the model, we would have begun the training from scratch.\n","\n","Now we can notice the overfitting, because the training accuracy continues to increase, but the validation accuracy decreases."]},{"cell_type":"code","execution_count":23,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":51369,"status":"ok","timestamp":1731444082072,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"X2fLe3YVx6aR","outputId":"cf83d041-1232-4c3a-90f1-ee94d0ee1de1"},"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 21ms/step - accuracy: 0.9382 - loss: 0.1795 - val_accuracy: 0.9135 - val_loss: 0.2485\n","Epoch 2/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 12ms/step - accuracy: 0.9444 - loss: 0.1594 - val_accuracy: 0.8927 - val_loss: 0.3135\n","Epoch 3/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9490 - loss: 0.1492 - val_accuracy: 0.8968 - val_loss: 0.3319\n","Epoch 4/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 11ms/step - accuracy: 0.9557 - loss: 0.1306 - val_accuracy: 0.8728 - val_loss: 0.4574\n","Epoch 5/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 11ms/step - accuracy: 0.9525 - loss: 0.1483 - val_accuracy: 0.8756 - val_loss: 0.4272\n","Epoch 6/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 12ms/step - accuracy: 0.9625 - loss: 0.1081 - val_accuracy: 0.8640 - val_loss: 0.4886\n","Epoch 7/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9616 - loss: 0.1162 - val_accuracy: 0.8437 - val_loss: 0.5513\n","Epoch 8/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9550 - loss: 0.1342 - val_accuracy: 0.8410 - val_loss: 0.6122\n","Epoch 9/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9614 - loss: 0.1148 - val_accuracy: 0.8337 - val_loss: 0.6248\n","Epoch 10/10\n","\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9702 - loss: 0.0892 - val_accuracy: 0.8316 - val_loss: 0.6833\n"]},{"output_type":"execute_result","data":{"text/plain":[""]},"metadata":{},"execution_count":23}],"source":["cifar_cnn.fit(train_data, train_label,\n"," epochs=10, batch_size=128,\n"," validation_split=0.2)"]},{"cell_type":"code","execution_count":24,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1767,"status":"ok","timestamp":1731444083836,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"S9wQ3UyltaZS","outputId":"939a69cc-1a33-405c-92a7-523f48383fbf"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - accuracy: 0.7112 - loss: 1.6430\n","Classification Accuracy: 0.7105000019073486\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"1WdoDXEKBiIu"},"source":["## 16.5 Dropout Layers "]},{"cell_type":"markdown","metadata":{"id":"pEGOb-cbBtws"},"source":["One way to deal with overfitting in neural networks is to introduce **Dropout** layers. The idea of dropout is very simple: during the training, at each step of processing a batch of images, a portion of the neurons in a layer are randomly disabled. This introduces randomness in the network, and it helps to improve model performance and reduce overfitting.\n","\n","Dropout can be considered to be similar to the ensemble methods, where during each iteration, a slightly different model is trained, which uses only a portion of all neurons.\n","\n","In the next cell, we add several dropout layers, where for instance, dropout of 0.2 means that 20% of the neurons in that layer are randomly dropped."]},{"cell_type":"code","execution_count":25,"metadata":{"executionInfo":{"elapsed":3,"status":"ok","timestamp":1731444083836,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"kFvKugKnuOAK"},"outputs":[],"source":["from keras.layers import Dropout\n","\n","# Define the layers\n","inputs = Input(shape=(32, 32, 3))\n","conv1a = Conv2D(filters=32, kernel_size=3, padding='same')(inputs)\n","conv1b = Conv2D(filters=32, kernel_size=3, padding='same')(conv1a)\n","pool1 = MaxPooling2D()(conv1b)\n","dropout1 = Dropout(0.2)(pool1)\n","conv2a = Conv2D(filters=64, kernel_size=3, padding='same')(dropout1)\n","conv2b = Conv2D(filters=64, kernel_size=3, padding='same')(conv2a)\n","pool2 = MaxPooling2D()(conv2b)\n","dropout2 = Dropout(0.2)(pool2)\n","conv3a = Conv2D(filters=256, kernel_size=3, padding='same')(dropout2)\n","conv3b = Conv2D(filters=256, kernel_size=3, padding='same')(conv3a)\n","pool3 = MaxPooling2D()(conv3b)\n","flat = Flatten()(pool3)\n","dense1 = Dense(128, activation='relu')(flat)\n","dropout3 = Dropout(0.2)(dense1)\n","dense2 = Dense(64, activation='relu')(dropout3)\n","dropout4 = Dropout(0.2)(dense2)\n","outputs = Dense(10, activation='softmax')(dropout4)\n","\n","# Define the model with inputs and outputs\n","cifar_cnn_2 = Model(inputs, outputs)"]},{"cell_type":"code","execution_count":26,"metadata":{"executionInfo":{"elapsed":269574,"status":"ok","timestamp":1731444353407,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"wisTs_NTurjP"},"outputs":[],"source":["# compile model\n","cifar_cnn_2.compile(optimizer='adam',\n"," loss='sparse_categorical_crossentropy',\n"," metrics=['accuracy'])\n","\n","# fit model\n","history = cifar_cnn_2.fit(train_data, train_label,\n"," epochs=40, batch_size=128,\n"," validation_split=0.2, verbose=0)"]},{"cell_type":"code","execution_count":27,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1562,"status":"ok","timestamp":1731444354966,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"eQZY8Xaaurng","outputId":"058bbff3-1208-4bbe-f757-266333a1d6ec"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.7554 - loss: 1.1086\n","Classification Accuracy: 0.7544000148773193\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn_2.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"1KDHTF0vmKhx"},"source":["The performance was increased to about 75% from the initial 73% for the model without dropout layers.\n","\n","Let's create a figure with the **learning curves** to show the accuracy and loss of the model. The blue lines indicate the model performance on the training dataset and the red lines indicate the performance on the validation dataset.\n","\n","We can notice that overfitting begins around epoch 15, when the training accuracy increases, but the validation accuracy stays about the same. Similarly, the training loss decreases, but the validation loss increases after epoch 15. The plots of the accuracy and loss are fairly correlated, however as we can see the accuracy can stay constant when the loss is changing."]},{"cell_type":"code","execution_count":28,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":410},"executionInfo":{"elapsed":259,"status":"ok","timestamp":1731444355440,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"1SLzE4yuPAEK","outputId":"27570f63-e47b-4bcf-fdb4-75b91a48f13c"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# plot the accuracy and loss\n","train_loss = history.history['loss']\n","val_loss = history.history['val_loss']\n","acc = history.history['accuracy']\n","val_acc = history.history['val_accuracy']\n","\n","epochsn = np.arange(1, len(train_loss)+1,1)\n","plt.figure(figsize=(12, 4))\n","\n","plt.subplot(1,2,1)\n","plt.plot(epochsn, acc, 'b', label='Training Accuracy')\n","plt.plot(epochsn, val_acc, 'r', label='Validation Accuracy')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('ACCURACY')\n","plt.xlabel('Epochs')\n","plt.ylabel('Accuracy')\n","\n","plt.subplot(1,2,2)\n","plt.plot(epochsn,train_loss, 'b', label='Training Loss')\n","plt.plot(epochsn,val_loss, 'r', label='Validation Loss')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('LOSS')\n","plt.xlabel('Epochs')\n","plt.ylabel('Loss')\n","plt.show()"]},{"cell_type":"markdown","metadata":{"id":"sj7hlnWZby9V"},"source":["## 16.6 Batch Normalization "]},{"cell_type":"markdown","metadata":{"id":"CMwSro-YBWNB"},"source":["Another type of layers that are used to reduce overfitting are **Batch Normalization** layers.\n","\n","Similarly to scaling the input features to range [0,1] or to a Gaussian distribution with 0 mean and 1 standard deviation, Batch Normalization layers scale a batch of data to have 0 mean and 1 standard deviation. This reduces the instability in training the network due to differences among batches of data, and also reduces the overfitting."]},{"cell_type":"code","execution_count":36,"metadata":{"executionInfo":{"elapsed":167,"status":"ok","timestamp":1731444878936,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"QalCw2dJboMS"},"outputs":[],"source":["from keras.layers import BatchNormalization\n","\n","# Define the layers\n","inputs = Input(shape=(32, 32, 3))\n","conv1a = Conv2D(filters=32, kernel_size=3, padding='same')(inputs)\n","conv1b = Conv2D(filters=32, kernel_size=3, padding='same')(conv1a)\n","bn1 = BatchNormalization()(conv1b)\n","pool1 = MaxPooling2D()(bn1)\n","dropout1 = Dropout(0.2)(pool1)\n","conv2a = Conv2D(filters=64, kernel_size=3, padding='same')(dropout1)\n","conv2b = Conv2D(filters=64, kernel_size=3, padding='same')(conv2a)\n","bn2 = BatchNormalization()(conv2b)\n","pool2 = MaxPooling2D()(bn2)\n","dropout2 = Dropout(0.2)(pool2)\n","conv3a = Conv2D(filters=256, kernel_size=3, padding='same')(dropout2)\n","conv3b = Conv2D(filters=256, kernel_size=3, padding='same')(conv3a)\n","bn3 = BatchNormalization()(conv3b)\n","pool3 = MaxPooling2D()(bn3)\n","flat = Flatten()(pool3)\n","dense1 = Dense(128, activation='relu')(flat)\n","dropout3 = Dropout(0.2)(dense1)\n","dense2 = Dense(64, activation='relu')(dropout3)\n","dropout4 = Dropout(0.2)(dense2)\n","outputs = Dense(10, activation='softmax')(dropout4)\n","\n","# Define the model with inputs and outputs\n","cifar_cnn_3 = Model(inputs, outputs)"]},{"cell_type":"code","execution_count":30,"metadata":{"executionInfo":{"elapsed":199,"status":"ok","timestamp":1731444369940,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"vJZRw2DH3wls"},"outputs":[],"source":["# compile model\n","cifar_cnn_3.compile(optimizer='adam',\n"," loss='sparse_categorical_crossentropy',\n"," metrics=['accuracy'])"]},{"cell_type":"markdown","metadata":{"id":"wBNiUaNgxTdw"},"source":["Also, let's use the argument `verbose=0` to not display the loss and accuracy after every epoch. Instead, after the training is complete, we will display the learning curves to observe the performance of the model. We increased the number of epochs to 60 for this model.\n","\n","And, we will use the `datetime` python library to measure and display the training time of the model."]},{"cell_type":"code","execution_count":31,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":437731,"status":"ok","timestamp":1731444809108,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"HBeizSe3_pQf","outputId":"086e74fa-ea53-4694-b8ae-48abb05d0c36"},"outputs":[{"output_type":"stream","name":"stdout","text":["Training time: 0:07:17.549027\n"]}],"source":["import datetime\n","now = datetime.datetime.now\n","\n","t = now()\n","# fit model\n","history = cifar_cnn_3.fit(train_data, train_label,\n"," epochs=60, batch_size=128,\n"," validation_split=0.2, verbose=0)\n","\n","print('Training time: %s' % (now() - t))"]},{"cell_type":"code","execution_count":32,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":410},"executionInfo":{"elapsed":523,"status":"ok","timestamp":1731444809628,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"_rkzWwx842Tq","outputId":"55bcbee3-cf25-4c3c-a701-57d4f097cec7"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# plot the accuracy and loss\n","train_loss = history.history['loss']\n","val_loss = history.history['val_loss']\n","acc = history.history['accuracy']\n","val_acc = history.history['val_accuracy']\n","\n","epochsn = np.arange(1, len(train_loss)+1,1)\n","plt.figure(figsize=(12, 4))\n","\n","plt.subplot(1,2,1)\n","plt.plot(epochsn, acc, 'b', label='Training Accuracy')\n","plt.plot(epochsn, val_acc, 'r', label='Validation Accuracy')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('ACCURACY')\n","plt.xlabel('Epochs')\n","plt.ylabel('Accuracy')\n","\n","plt.subplot(1,2,2)\n","plt.plot(epochsn,train_loss, 'b', label='Training Loss')\n","plt.plot(epochsn,val_loss, 'r', label='Validation Loss')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('LOSS')\n","plt.xlabel('Epochs')\n","plt.ylabel('Loss')\n","plt.show()"]},{"cell_type":"code","execution_count":33,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1551,"status":"ok","timestamp":1731444811173,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"UvcbDLp-9COb","outputId":"d1fbf3f0-9452-4565-ad01-2093ee2d6c83"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.7699 - loss: 0.9293\n","Classification Accuracy: 0.7695000171661377\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn_3.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"ADsuQtiRE0RR"},"source":["With the Batch Normalization layers in the model, the test accuracy increased to 77%."]},{"cell_type":"markdown","metadata":{"id":"Np9mB-EFcvy4"},"source":["## 16.7 Data Augmentation "]},{"cell_type":"markdown","metadata":{"id":"RiFBo4vJ1BkV"},"source":["One of the main reasons for overfitting is the lack of sufficient training samples, or lack of diversity in the training samples, for training a model.\n","\n","To deal with overfitting we can use **data augmentation**, which refers to expanding the existing dataset by applying different image transformation operations. Examples are flipping the image vertically or horizontally, cropping the image, changing the contrast and color of the image, adding noise to the image, rotating the image at a given degree, and similar.\n","\n","![Image Augmentation](https://cdn.hashnode.com/res/hashnode/image/upload/v1623166213173/JTRR1Btgm.png)\n","\n","By doing data augmentation, we are synthesizing new images from existing data, as well as introducing some diversity in the images.\n","\n","Keras provides the function `ImageDataGenerator` for data augmentation. Among the available operations for data augmentation are:\n","\n","- **rotation_range** is a value in degrees (0–180) to randomly rotate images.\n","- **width_shift** and **height_shift** are ranges of fraction of total width or height within which to translate pictures, either vertically or horizontally.\n","- **shear_range** is for applying shearing randomly.\n","- **zoom_range** is for zooming in pictures randomly.\n","- **horizontal_flip** and **vertical_flip** are for flipping images.\n","\n","In the example below we applied width shift, height shift, and horizontal flip."]},{"cell_type":"code","execution_count":34,"metadata":{"executionInfo":{"elapsed":207,"status":"ok","timestamp":1731444866559,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"IQD2s2UAQQO5"},"outputs":[],"source":["from tensorflow.keras.preprocessing.image import ImageDataGenerator\n","\n","# create data generator\n","datagen = ImageDataGenerator(width_shift_range=0.1, height_shift_range=0.1, horizontal_flip=True)"]},{"cell_type":"markdown","metadata":{"id":"cHYVQidO7iNS"},"source":["ImageDataGenerator returns both the augmented images and their labels for each batch of images.\n","\n","Let's plot a few images to see what they look like. They don't look much different, but that is because the applied shift and flip are not very large, and we didn't use other operations."]},{"cell_type":"code","execution_count":35,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":367},"executionInfo":{"elapsed":778,"status":"ok","timestamp":1731444868872,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"U2M9gFJ37QMG","outputId":"1020aaed-4b66-49b8-f7ad-2d8754cf0240"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAzAAAAFeCAYAAABaafoYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB8VElEQVR4nO39eZBc1Xk//j/33r6993TPvkmaGe0Li7DYjRE4AQzYfEmCHDv+GglDgZcvKhLsFA6/GLAdk4odQwqXnZDPx4BjKlUGb5XELCEGjG0MAktsAjFaZiTNaPbpWXrvvvf3Bx/Nh+d5DmiQAXGZ96sqFZ+rc7tvt+jT56rP+zyW7/s+AQAAAAAABIB9rC8AAAAAAABgvnADAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAIDNzAAAAAAABAYuIEBAAAAAIDAwA0MAAAAAAAEBm5gAAAAAAAgMHADY3DzzTeTZVnH+jIAAICIuru76aMf/egR+z322GNkWRY99thjc8e2bNlC3d3d79zFAcCCtm3bNjrzzDMpkUiQZVm0Y8eOY31JC0LoWF8AAAC8+7773e9SPB6nLVu2HOtLAQAIpEqlQps2baJoNEq33XYbxeNx6urqOtaXtSDgBgYAYAH67ne/S01NTe+rG5izzz6bCoUChcPhY30pALAA7Nmzh/r7++lf//Vf6aqrrjrWl7OgYAnZe0QulzvWlwAAEGi2bVM0GiXbxlcbALzzRkZGiIgok8m8aT/M8d5+C36U//Wvf02nnHIKRaNRWrZsGf3Lv/yLsd8Pf/hD2rBhA8ViMWpoaKBPfOITdODAAdXvqaeeoo985COUTqcpHo/Txo0b6Te/+Q3rczhjs3PnTvqLv/gLqq+vp7POOusdeX0AEHz9/f30+c9/nlatWkWxWIwaGxtp06ZN1NfXx/q9UX7v7rvvJsuy5vp3d3fTSy+9RI8//jhZlkWWZdE555wz13/v3r20adMmamhooHg8Tqeffjr913/9F3vMw3mTH/3oR3TLLbdQZ2cnpVIpuuyyy2hqaopKpRJdd9111NLSQslkkq644goqlUrsMarVKn3ta1+jZcuWUSQSoe7ubvqbv/kb1e+whx9+mNavX0/RaJTWrl1LP/nJT4zX9PoMjInneXT77bfTunXrKBqNUmtrK11zzTU0OTn5pucBABy2ZcsW2rhxIxERbdq0aW4c3bJlCyWTSdqzZw9ddNFFlEql6FOf+hQRvXYjc/3119PixYspEonQqlWr6Fvf+hb5vs8eu1Ao0NatW6mpqYlSqRRdcsklNDAwQJZl0c033/xuv9T3pAW9hOyFF16g888/n5qbm+nmm2+marVKN910E7W2trJ+f/d3f0d/+7d/Sx//+MfpqquuotHRUbrjjjvo7LPPpu3bt8/def/yl7+kCy+8kDZs2EA33XQT2bZNd911F334wx+mJ554gk499VT2uJs2baIVK1bQN77xDfUfLwDAYdu2baPf/va39IlPfIIWLVpEfX199L3vfY/OOecc2rlzJ8Xj8bf0eLfffjtde+21lEwm6cYbbyQimhv3hoeH6cwzz6R8Pk9bt26lxsZGuueee+iSSy6h+++/n/7kT/6EPdatt95KsViMbrjhBtq9ezfdcccd5Lou2bZNk5OTdPPNN9Pvfvc7uvvuu6mnp4e+8pWvzJ171VVX0T333EOXXXYZXX/99fTUU0/RrbfeSi+//DL99Kc/Zc/T29tLf/7nf06f/exnafPmzXTXXXfRpk2b6MEHH6TzzjvvLb3+a665hu6++2664ooraOvWrbRv3z76zne+Q9u3b6ff/OY35LruW3o8AFh4rrnmGurs7KRvfOMbtHXrVjrllFOotbWV7r33XqpWq3TBBRfQWWedRd/61rcoHo+T7/t0ySWX0KOPPkpXXnklrV+/nh566CH60pe+RAMDA3TbbbfNPfaWLVvoRz/6EX3605+m008/nR5//HG6+OKLj+GrfQ/yF7BLL73Uj0ajfn9//9yxnTt3+o7j+Iffmr6+Pt9xHP/v/u7v2LkvvPCCHwqF5o57nuevWLHCv+CCC3zP8+b65fN5v6enxz/vvPPmjt10000+Efmf/OQn38mXBwDvE/l8Xh178sknfSLyf/CDH8wdOzy2SHfddZdPRP6+ffvmjq1bt87fuHGj6nvdddf5ROQ/8cQTc8dmZmb8np4ev7u726/Var7v+/6jjz7qE5F/3HHH+eVyea7vJz/5Sd+yLP/CCy9kj3vGGWf4XV1dc+0dO3b4RORfddVVrN8Xv/hFn4j8X/7yl3PHurq6fCLyf/zjH88dm5qa8tvb2/2TTjpp7tjha3r00Ufnjm3evJk97xNPPOETkX/vvfey533wwQeNxwEA3sjhMee+++6bO7Z582afiPwbbriB9f3Zz37mE5H/9a9/nR2/7LLLfMuy/N27d/u+7/vPPvusT0T+ddddx/pt2bLFJyL/pptuemdeTMAs2CVktVqNHnroIbr00ktpyZIlc8fXrFlDF1xwwVz7Jz/5CXmeRx//+MdpbGxs7v/a2tpoxYoV9OijjxIR0Y4dO6i3t5f+4i/+gsbHx+f65XI5+qM/+iP61a9+RZ7nsWv47Gc/++68WAAItFgsNve/K5UKjY+P0/LlyymTydDvf//7t/W5fvGLX9Cpp57KlrUmk0m6+uqrqa+vj3bu3Mn6X3755ewXi9NOO41836fPfOYzrN9pp51GBw4coGq1Ovc8RER/9Vd/xfpdf/31RERqyVpHRwf79aeuro4uv/xy2r59Ow0NDc379d13332UTqfpvPPOY2P6hg0bKJlMzo3pAAB/iM997nOs/Ytf/IIcx6GtW7ey49dffz35vk8PPPAAERE9+OCDRET0+c9/nvW79tpr38GrDZ4Fu4RsdHSUCoUCrVixQv3ZqlWr5r5ce3t7yfd9Yz8imvvi7u3tJSKizZs3v+FzTk1NUX19/Vy7p6fnqK8fABaOQqFAt956K9111100MDDAlpxOTU29rc/V399Pp512mjq+Zs2auT8/7rjj5o6//h+AiIjS6TQRES1evFgd9zyPpqamqLGxkfr7+8m2bVq+fDnr19bWRplMhvr7+9nx5cuXq3zPypUriYior6+P2tra5vX6ent7aWpqilpaWox/fjiUCwBwtEKhEC1atIgd6+/vp46ODkqlUuz468fWw//ftm01R5Rj5UK3YG9g5svzPLIsix544AFyHEf9eTKZnOtHRPTNb36T1q9fb3ysw30Pe/2/qgIAvJFrr72W7rrrLrruuuvojDPOoHQ6TZZl0Sc+8Qn2y+4bFeCt1Wrv2LWZxsU3O+6LvN+7XTTY8zxqaWmhe++91/jnzc3N7+r1AMD7TyQSwW6I77AFewPT3NxMsVhs7peT19u1a9fc/162bBn5vk89PT1z/9pnsmzZMiJ6bVnDH//xH7/9FwwAC9b9999Pmzdvpn/8x3+cO1YsFimbzbJ+h3/hzWazbFtP+WsG0RvfOHR1dbEx8LBXXnll7s/fDl1dXeR5HvX29s79CyTRa5sIZLNZ9Ty7d+8m3/fZdb/66qtE9NquavO1bNkyeuSRR+iDH/wg/hEJAN41XV1d9Mgjj9DMzAz7FUaOrYfHxn379rHVP7t37353L/g9bsHeHjqOQxdccAH97Gc/o/37988df/nll+mhhx6aa//pn/4pOY5Dt9xyi/qXQ9/3aXx8nIiINmzYQMuWLaNvfetbNDs7q55vdHT0HXolAPB+5ziOGn/uuOMO9cvK4X9I+dWvfjV3LJfL0T333KMeM5FIqBsgIqKLLrqInn76aXryySfZY9x5553U3d1Na9eu/UNeCnseotd2RHu9b3/720REasedwcFBtjPZ9PQ0/eAHP6D169fPe/kYEdHHP/5xqtVq9LWvfU39WbVaNb4nAAB/qIsuuohqtRp95zvfYcdvu+02siyLLrzwQiKiuRz2d7/7XdbvjjvueHcuNCAW7C8wRES33HILPfjgg/ShD32IPv/5z1O1WqU77riD1q1bR88//zwRvTYh+PrXv05f/vKXqa+vjy699FJKpVK0b98++ulPf0pXX301ffGLXyTbtul//a//RRdeeCGtW7eOrrjiCurs7KSBgQF69NFHqa6ujv7jP/7jGL9iAAiij370o/Rv//ZvlE6nae3atfTkk0/SI488Qo2Njazf+eefT0uWLKErr7ySvvSlL5HjOPT973+fmpub2T/UEL32jy7f+9736Otf/zotX76cWlpa6MMf/jDdcMMN9O///u904YUX0tatW6mhoYHuuece2rdvH/34xz9+25ZFnHjiibR582a68847KZvN0saNG+npp5+me+65hy699FI699xzWf+VK1fSlVdeSdu2baPW1lb6/ve/T8PDw3TXXXe9pefduHEjXXPNNXTrrbfSjh076PzzzyfXdam3t5fuu+8++qd/+ie67LLL3pbXCABw2Mc+9jE699xz6cYbb6S+vj468cQT6eGHH6af//zndN111839A9SGDRvoz/7sz+j222+n8fHxuW2UD//i/G4vu33POlbbn71XPP744/6GDRv8cDjsL1261P/nf/5n41akP/7xj/2zzjrLTyQSfiKR8FevXu1/4Qtf8Hft2sX6bd++3f/TP/1Tv7Gx0Y9EIn5XV5f/8Y9/3P+f//mfuT6HH390dPRdeY0AEGyTk5P+FVdc4Tc1NfnJZNK/4IIL/FdeecXv6uryN2/ezPo+++yz/mmnneaHw2F/yZIl/re//W3jNspDQ0P+xRdf7KdSKZ+I2JbKe/bs8S+77DI/k8n40WjUP/XUU/3//M//ZM9j2j7U9//vls3btm1jx03jXqVS8W+55Ra/p6fHd13XX7x4sf/lL3/ZLxaL7Nyuri7/4osv9h966CH/hBNO8CORiL969Wr13PPZRvmwO++809+wYYMfi8X8VCrlH3/88f5f//Vf+4ODg6ovAIDJG22jnEgkjP1nZmb8v/zLv/Q7Ojp813X9FStW+N/85jdZ+Q3f9/1cLud/4Qtf8BsaGvxkMulfeuml/q5du3wi8v/+7//+HX1NQWH5PiooAgAAAAC8V+3YsYNOOukk+uEPf0if+tSnjvXlHHMLNgMDAAAAAPBeUygU1LHbb7+dbNums88++xhc0XvPgs7AAAAAAAC8l/zDP/wDPfvss3TuuedSKBSiBx54gB544AG6+uqrVY2thQpLyAAAAAAA3iP++7//m2655RbauXMnzc7O0pIlS+jTn/403XjjjRQK4bcHItzAAAAAAABAgCADAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDMOwl0/fV/y9qZhgbVJ9PYzNphN8LbIV09tG9fL2u3tTerPqeffjprP/nkb1l7x3PPq3PaOjpZu0Zl1q6WZ9U5sYjL2uVaVfWpWRXernmsHQ3XqXPCTpK19/buZu18Tl/L6jXreJ98UfU5cLCftROJMGuHHP1+16o11i6Xxeup6tdsi7+3eKZe9Wlo6WDtv7/x/6f6ALxf/cW/fOgtn2PbDms7jqP6yIrLkUhE9fE8Pga5Lh/HqoZxrOLxccAN8bHDJz12lEt8rCC/pvqEwiJSKV5Tqcqfh4jIEt8TLh+qqY6i6pzm5lbWroV0lHM6N8Xa5SK//kxUj9Vtaf64VTFeHpgeUufkqlnWtj39fteJ5/rG//O/VR+A96vVpyxj7aJhLnjOp/6ctTvX8TlQ1RPjDxHN7ubzx3bDnOfVl/n8MDuVZe36ej2fKfn8Mxxt4Z/f4bFRdc7w+BhrJ5Jx1ac+GWNt2+fXOzmr54IU59P0zq7jWNuq8MckIqpz+Ti75/mXVZ/dr77I2l09bfwxmlvUOStOP421c/lxfi2lA+qchigfU91CWvWRbrriq0fsQ4RfYAAAAAAAIEBwAwMAAAAAAIGBGxgAAAAAAAiMeWdg5DpruTaYiKhS4WsUbYvfHzmk13jXxFps0+NWRS7D9/m1+OLaiIg82Yf4Gmn5ekzHTCVy5LEjtU3H5POYrkW+Rs/Tj6uvV55jyMDU3vy5PU+//5bniD76euXfI8BCEg7rbMeRmD7TkszA1Gr6cyY/j/IcXw8DamyuifXlcswlIqoRf56aY7gWh78PtSofO1oae9Q5jTG+Br0tzNdIr+1crs5pqmti7YN5vXb8ly/tYO2GZv64p4q15ERE3Sn+uCWvwNo7D+m15C/18eep1fS12KTHTICFYu0SnoN4bmCv6nNg/0us3djNc7WxeEqdU/FLrF2sllWftlae1y5XeJ5Y5oCJiOJp/lydHUtYOyMyv0RE9ZM8czczNa367N/Jcyck5k2tPYvUOe3LeX4o09TF2rWc/h2ib8dO1t7++O9Un4h4OzNNGdZOZnTO3XF4vpIq/PpN41w4wq+vrbld9Tla+AUGAAAAAAACAzcwAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABMa8Q/xDIwdZe09fr+pjhXjAp66Oh4DSSV3AZnximLX3H9SFcBxRlGdg4BBrj46O6AsWhZIicR4mHR3er06Jhvn1J9M6NDZbyLO2ZfG3MJXUwVzb50HQ4aFB1i4W+WMSES2a4WGuqWkdDB0d4UXVcjl+/Z6hgJ0MrMmNF8iwCYET5q/RmRxTfWZyM+oYwEIhg/Pz84dvAEKkC2DWxOe+ZtgswBLniEw/OYZvBl88ru8Yxgqfj9UJlxdDawzzAsNERMuaeUj1uHYe2u8UwXoiooTFi1+mIvp9ec7hIdreAV58bkWDLjS3qoGPu1HiIeGVHbqophvij7On/0XVp1TJqmMAC0Xj0gxrL0rpguVk8znEgb2vsHbPMl7YkoioWOHzoqGinptERcHgnMdD/LMzep60okUUtBWbH1mGAXLJIh70L0R1iH/3L5/mzz2dZe3Vy9eqc5a38fFxYJy/5oirC/KWKnwzA9tVXai5lc/HQ2H+JRBN8ALsRGrPASrO8LlrOKQ3RJAbQ1nuvG87jgi/wAAAAAAAQGDgBgYAAAAAAAIDNzAAAAAAABAYb6GQJV8nWK3qtW5VkaeoVPkaaZkXISKKxPj64VK5qPrs7eN5lclRvpY5Go2pcxyRgalWeQ4lEtGF58JibV4kHFF9POKLCWWmxLQWvlzir8knviYwnjCsqxbrESMRvYgxJs6zbP64XkWv7fR9cczn54Rc/TwhUaSvXC2pPlOThhwSwAKhsmTzYIu12aaxIxTiY5KpkKUu9OuLPzcUnBTnxOL8Mx6O6LHProjrq+p//2pJ8QzJB1ady9pJm68tJyKamuHrtXMlMd4k9VrsmMXHviZHF0n+o3VnsPbwU79i7Sd2PKfOSdoJ1j5+0WLWTjk6j7OyjV9vyNNj6L4B/VwAC8Wwm2Xt1pWNqk9JROoO9e9h7cYGXTyyRnzcrUX0nCebz7F2wePzl4kxXoCSiGi6iR8rlHnupGiY/y7v4GNFY16Puydm+Pjh1/H8yqqIzuXVjfEszYEDPAMe6dTjzVJRENPP62xNcwsfQ0XNYWPOR0wXaWpskl9rUs8Niwk+vvuG+4CjhV9gAAAAAAAgMHADAwAAAAAAgYEbGAAAAAAACIx5L0ZrSLextuvkVB9PFBNIpfk+07G4zqpMT42zdk3sX01ElJvhawALqhaLrkdAxPftLhR4DsXz9XrzSJivf04b6g/UifoDo+N83/FSSa8BtC2+uNAVmZJiQb+Xvb27WHt6WteKqYosSljkZkw5JYv46w6J3I8bNmRgRP4mP62vd3JCHwNYKEwj0JF4Ph+jXEd/9mSNF1MdGFMu5vVsR491ts2Pyf39LV9nYKwaHyva0otVn40nfoS1V7etZ22/pL9ynpnmY92gWJOeTOrcjB3nj5Mx1NBZmuF1yP5o/Ums/cAzv1Xn/Pr5baydEH8nPR38e5CIqFri6+6Xtur15pWcruMFsFBUYny8qUvpsS5c42NdOcc/4xVDvTw5n0kmdO2+7Ayf+7kim+1N6xp2w7t4ncO2jnrWbq7Tz5MZn2DtxRGd3Vuymtd0ccP8+t24zmbPHOSZlzWeqMs3w/PdRERNNp8LNi/tVn1IlGV86RB/HllLkYgoJN7v6qzIrEf0OFzI8XlqsfT2/W6CX2AAAAAAACAwcAMDAAAAAACBgRsYAAAAAAAIDNzAAAAAAABAYMw7xO9XeVinXNDBp2lxbDbPC06Gw7pgY6XMg1kz2azqMzU2xNoyyJpM6eI/7qwItMsCjmQoOFnmSdawrTcdiMR4eGt6RgTLqvpxXRHEdaM8IDszqwspTYzzzQ2qFR3elUUnvZoIUJmSxeLybBH2ChtC/NE4f389wwNPTUyoYwALRcRQ+PFISkUebpQFKInMxS0lOR7Kc3zSY4csomk7/Pot0mNqY6aZtVev+KDq09TAA+wRhwfpo3FdcHL1khWs/cL+PtYeFIWLiYgSHXxsjtn6cV3xuk9s72Tt2obT1DlPv/Qia2/vfYG1M0n9Hdac4sXoqr7+3ji+5yR1DGChiDo8MR4KGQqJi81G4s18rmWXDBthFPicrb1Bb/iRCPHNjNJJfk59s96AKCEC7J0xfr1tSf6ZJyLKiMlVW1iP55U0Hz/CMRHINxRPHx4Tc8w8D+1HLV5MkojIEY874+mp/qFJ/n7OijlnsjmrzgkT/w6oc8XrMXzXVIr8evv6Dqg+Rwu/wAAAAAAAQGDgBgYAAAAAAAIDNzAAAAAAABAY887AlKo8Q1Ku6IKNboivLbQtXpRyOqvXGop6ahSP6bWR1RJ/3JIoShmJ6JdRLfHnLpT5OjyL9JrpCPH1h66hAJxl8eeuVHjuxzfkQ3KzfK1hVVyLbyhE19zEC6Y1NeoCauNizeL4BM8KWba+P5VL6qvi71UW/CQiskKi2GVIv9+JhC7aBLBQyIKw81Gtys+9Hm9kkUpj0UrxoZbFaE0xGk8cc8RXQSKmi/guWryOtSdyeqx+aR/Pq0S7Glm7NaHHpJY6vj5+Rfsi1h405OvGxbFoc6PqExfvZ4r49Z7Y1qPOSaX4tQwMDbB2rqDX4S+q58+dsPU6dj+s1+YDLBQDO/m4UGzX84UGUYh7cULMO3xe0JyIqJzl869Mgx631rTyuVPB5eOA3yAqOhKRVRPzXVH80jEULE8neG4w5ui5oGeJxxU5yJChULElCpIXRE487OtraYyJ8bCqH3cmx6+lKc/75AZ1VqXcxDMwyUb+3nnqO43It/i8f2hsl+pztPALDAAAAAAABAZuYAAAAAAAIDBwAwMAAAAAAIGBGxgAAAAAAAiMeSdPkykeuiqJIDoRkeuK8KLFQ0wzk8PqnFiSB58SMR2AHC/wwKYvgucysE+kg/GJGH8e19UFySKOeDs8HUhKRHmRMpt4wCqXE0WHiMgWYVLL5+fUqjqEFY/z5+nq7lJ9HIfff2aneFDO9/X1h0QA3xLX5htCZCURWPMNoWBb7sYAsIAUDWPQkXji34/skCH8LT5Wvm8oShnij+OHRODUUEDYFwV3HeJ9Fi9ao86JxZaw9sysLnq78+BBfm2igNrGlavVOWkx7HbV82BoxFDgk8TY58hdCYhoqsTH2awotDw0ozeV8aL8uXraW/i1RTLqnJgYQz3TAGm99UKnAO8XB3fxQPjofv1v5+vbefB8xfH1rO2W9ee1MCOKgA8OqD6Le3jwvJLkcysrpsfHwUN8rjo1zTfviNh66jwtNrZyPMNGV2Ke5Im5bLFs2KSlxs9pSPDrNU3iy7P8vbKK+vupXWw2cloDH99fnNUFMnOj/O+xoYMXB64UefFRIqKyx/9OytNvX9Fz/AIDAAAAAACBgRsYAAAAAAAIDNzAAAAAAABAYMw7AzM6cYi1y0Vd8DAerWPtSoWvu/NEcSAiokScr+dzLL3euVLiz2WJdeDFvM7jzE7zQpUdi/lavY6OdnVO1OVrujN1utiSZfPHlaudqxV9LY6oJFcRfVIpviaTiCidTrB2saQLqBWKfE13tcbXXFYMxUZDNf5X7jji9RiKX/piDXq1ptfhkyE7A7BQeIZsypHUZFbCkLmriTFTxvSIiCyRgbFEp1pFf6bDIZ4z6VjEsykli4/lREQzk3w8zxX0uuqsyPc9tftF1m6K6vXmp3bxtddRcf3NDbpI5ajIGm576SXV58DYGGsPFfm68L4hnhkkIkqIofjkZR2s3br6A+qcOpGn9Cz9fpu+1wAWiuPWrmLtsZFB1WdZhn/OV4vi2LPTvHA3EVF7lY9BkRk9TyqOifyZmM9Yhn/H9yu8T9Ti86Swoch5ucTHpJKeIlMyIwo/imGhoqfIFIvweWk4xAtxmmoblwsij2PIEbpi9tqV5u+/U9EZx8FJXkw02cQzgvHFK9Q5+VneZ2oChSwBAAAAAGABwg0MAAAAAAAEBm5gAAAAAAAgMOadgSmJnIlvyLNEwnydYCzCa6/Up/W66kwdXxNYKOg1jER8fXksxtcAOhHeJiIqlPkawFd7e1n7wMH96pzO9jbWjq7VtRDqwnxdpisyJJ5hQWJerL12XH7fuG71ieqcE447ibUL5YrqM5Pje6C7g3zN4uwsX69IRFQV+45HInx9qC1eDxER1fj1umG9NjJsWpwPsECEDWPQkciaSz7pscN2+TrlsKvzclTjn8cQ8c901OF5OiKi+qblrJ2t8XXKew7qBdxhcb0VQ+ynZPHXUCvy8fzXr76gzulq5Guvm9K8/sMrw7z2ABHRo89uY+1hw9r3/hFey2FCjJfTU3p8rIvzv8clLfw7q+bqsa/q8PXlPunvxkKeP3c8qTOYAO9XJ566nrUnJvR//6vFGBoWgZCUrcfYpWledyQU1uNjIc/nTrJ+Xsgw5YmLrF5Y1DiMRk05N3GgZqpfxZ+sIubRJT3NIyvEB1pPZFeKVT0Q50U+sT6hvwMsm1+fLeauS2MZdU6dwy9wdz+v8TI4pr/D1q44lT9up+py1PALDAAAAAAABAZuYAAAAAAAIDBwAwMAAAAAAIGBGxgAAAAAAAiMeSevG5t46KpW1iHPhnQDa1vEQ0IzWR20jIZ5OCo3O6P6yLpgmXoe/PcNwfPZoigoKSoGlQv6+g8N8OJKrq2LFaVEpbPsDA9nFgyPmxdF10Ii8D54UBd1GjjAA6iNLc2qT1psgNBUx8OvuWn9Xspb1poIbhnLrXn8/fWrOqRaNWwyALBQOIYCsEdii/El5OrxhsL8cX3SY13I45ulNMZaWfuE7hPUOdkCP+elg/zzO1bRXw0p4n1MG7mUPD7+FbITrJ00BE6HZ/kmJ4emefvRF36vznly105+vbmc6jOV4983XoVfmy2/I4hoaQvf3GBxWxdru2FdiLNa4e/L6Kgez/f3v8Laf/zBS1UfgPcrS3xuWjo7VJ9YfpK1yzn+uYqm+PyGiKg9xsegmiE4L8t5O7bYPKWmC34nImKctfn1h0J6HKuKoprFsi70OyPmgqUKH/Nrvh7fPVE4vCo2U8kZil9OzvLnibh6AwTX4Y9jl/j1xjxdyD2d4PP1/CgvcP/sc3vUOZkof5ym9pTqc7TwCwwAAAAAAAQGbmAAAAAAACAwcAMDAAAAAACBMe8MTDzF162Nj+RVn4MHR1i7XeQ2VixfqS/A5Wv+xsdHVJ9Ugq+hkyvFJycnSQqLQkRN9Tyf41X1+sRolK8TTEb0eudahWdGLI9fTbWsC/lUymJdpsfXduZmR9U5daIIaKZOr//sbuZF1rx6Xohz6AAvMkRENCXWfXtiPaVtWMsvC9iFIrqYG9UMVe0AFohK5a1nwCIiuxfy9PrnqlgjbZFey5wK8bHhhCUfYO1Tlp6sznl+b5a1vQofd8uGrE2uwvN+oZIhCyfWUc8M8wxMqkEXsKuF+Hjy5I4drP2r53ibiGh0ij9uyZDLs8TfiSyjfMbJ+n35yMY/Zu0Vi5aydnlKf+/t6t/H2pOTh1QfJ4V/K4SFa6qf58IaUzrvF03wMacWEp+ZqJ532FU+t6oYMsiWCFFXxNyvWtEZmIjIZls+H19mc/qcaVk43NJ5nESUz+viEZ6prurpI1lhPk0vVfm4ljdkbXyfv7+OISdeFcXeSbyXw/lxdU6pwL9/KnlRGN1QFHTPXl50OF9r0p2OEkZVAAAAAAAIDNzAAAAAAABAYOAGBgAAAAAAAmPeGZhyie+zX60Z9rjOi72nY3wd4ZKuxeqcVIKvCcyO6zxI2OYL6yayfC120bDuMSbWFna28NoIDQ1yRTTRyBivvTIyrtcApkWWJizWNBZzev1zfpavNWxLZVj7A9096pw1bZ2s3dnSqfrUN/HXREW+gPIp63l1TiGfZW1P1MfxDZVgXFHvwa3otaiuLNYDsICY1hgfiWvz8THk63xLTZQqScf1+uHTVp/F2uuXnMLaDeFGdc6SJl5DqiXGx+7hKT7GEhGVxAJtuf6ZiGh2jGdTrBk+Nvc085weEdGegYOs/djvn+Z/fqBPneOI+lX6nSNa3NTC2hd96GzWPvtU/j4REdWn+PsyPcrfh21PP6POqRX5d+P6Dxyv+titOk8JsFAc2snrIHUua1F9YjE+NhREJsMr6FpPNVl7xZBNIYtPc2seHzsKJUNNwxAfUSIOP6fm6XlSTeROXENdr1iMz50ilqgDY6hj48b590RBvMTZKV3PKu7y1+x7Olxji++sYoWP51OGmoxk8Xm1K6LPdRF+rUREMxN8DPWtt69mIGadAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABAZuYAAAAAAAIDDmHeLft4eHsIqG4jk1kVlyozwkdHBQF1ZMxXgoqD6jCzZGwjxQJUP8jq3DUk0NPGzf1soDYlMzPGxKRLR73x7WLtV08CndzItzOj5/jW6Ivx4iIltcX0mEsGxfX78twvSlsg5qDU2NsfbeUb6BwHRJn+OTLHAkizzpYG5ZFKwLhfV/NuGYKUYLsDBY1lsP8TsiEEll/RlqifEA/mnrPqT6rOtaz9rpED8n5OtgZXsmwdobuvj42L9jSJ0zIaushfQ4UCzzsWJ52yLWzqT05ikP/PZXrN17aD9r+74ek9JiHDt5zRrV5+I/Po+11y5dwdoyxEpE9MquV1l7365e1n755V3qnJExvvFMIaW/A1rDXay9/u2r5QbwnlcWRRNDZV34emqCh8YTophkrajPsUQB7WJBB8SrIugfEfMXy1B8PBnlzx0RRTU9Q8VJP8o/99G4HncjET5uhcT1yw2riIhqVf7e+QW+4UrUNxTMFCH+qqHQspx2lmpyQyd9LV6Bj8WNYqzrNHwPvrSPz0sHD2RVn6OFX2AAAAAAACAwcAMDAAAAAACBgRsYAAAAAAAIjHlnYNJpXuDLLeZVn5kcX8NYEsUvR2d0YcihYV5EqKUurfos7uBFHMvLeOHHmZxeULx4MV9zPC2KbL7a16/OKYn8h+/rNZeVIl+P2LFoJWsvWnKqOmc4y5/7+Wd/y9q7h3Qep7WFF3oK23rNZW6Kv5+Ts/x5qpaheJrN10I6jniNIV2k0hL3uQlXr/FOxPR6T4CFwnH05+aI5xD/fLq2zoecvIpnXo5ffJLqk3Iy/HFq/HE9T69LjoiR/ySRgXlxUF/L7w7wXEzeEPuJ1PPzjj/hONaenMmqc/Yc7GPtisXH4SWtPHdIRPSnG85k7Y0nb1B9Glt4DjI/w8fu5/fsVueMTvBcYTXEx93eMZ3j3DvI35fqiy+oPscl+Bt+2QrVBeB9q66Zz+s8Q+6kUuSf+6yIWZeKOnfdmORzkXA8pfo4IgOTTvCsocwxExHFwnxeVBbFLislPf+NOvxa4iE9/wqL3KDr8rmVY8id5MScM2Tz60/G9DkFEbSuejpn7YkMYMXj77+no4dEooBnQgTfm8I6x5kf59fSP6zvA44WfoEBAAAAAIDAwA0MAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgTHvEH9GFIYM53RotVzmQae4KAZk+Tq4JZNC9ZmM6tK1eAlrL1rEQ/3ThmBodoZvKNB3aIS1SzWdUKo5/O0IuwnVp7GVpy+Xrz2dtZ2ILsTpH+Qhz87JSdaO5XmhHyKiXfv3svar/XtVn/aOZfzaGvn71N2j/46GxgdZOzfNn9s1bBYQDvFgVtjV/9noiBjAwuGEjmITixL/fK5cfZbqsrKHh9XrHT0myYpknm2/afv/HGWtxUkesj1r6XJ1xt5hPob2zcyqPl3pDGufvpJvcrJv13PqnIYwv/71a3gg/yNnn6vOWdPOvwNChn+L693Px7pX9/ONW2qGcSzZ1sraz2z7HWuP13SQuK6dnzMwPqb60Isv8/Z5ugvA+1acf8Zjhk1/mtJ87nRwjBcsf6WfF4wlIkqn46zd2aw3gWpvzrB2OM7Hirijg+fpFH/cothAoCSrkRNRqcDnv15Vh/hLef4++CJI7xmKUvpiHuqJYphVQ5FzT2yEUiU9F6yJAL4vgv6u4TstGuWvqSoKekaqenxsaeKbbJXiSdXnaOEXGAAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAJj3hmYaISvh5ueyqo+tRpfDxcWRW38qs5X1KX4msX6hlbVpybW7yVSGf4YaZ7PISJqEEU027uXsnbPhC6ms7efr5GOxvTjLl99CmtXiK+VHBzU658rFb7evKVtEWtnSnqtZKrMczKzNb02MhvmReNKUX4tp6w4UZ3j1/jazd9v+x/WHtwv1moTkWPx64/G9JrRWGze/ykBvO+ETEVjj6CztZu1E6l21WeyyNclt7p6XbJr8XFVLG2mqq3HDkestY74/N+yTlrMiwUTEe0b52vQZ5/drvqctITn8LpEdjLSxQsMExH9fx//FGt3LuKPkUzo3E92khf/ff4lPW5NV0XR3gwfL3v37VPnyCzfzr38O6GusVGd09jI13i/tPMl1edgv14bDrBQRDJ8zuAaCv8mRCatvZWPh69O6NzJcwcPsnbOkPWoaxaF0EV2ry6k56VtHfxz7sb4WBIy5OeyY3zOJjMxRETTUzybHRFzNjuqi4SXLf7cQ5N87lorGXIzDv+e8G39uLlp/j5ExXdANKqzKr7NR8iQxd/vpKXf/5Z2/rjpVj2GHi38AgMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBgzDu4MD09zdqFot572rL4+rhcLs87eHzPayKiRIKvH57WywYpWubnFcti/+qqXhuZivF100taWlh73boPqHM+9EG+d/Z0QdeKmZjix/Yf4nuVF/L6fZF1U2IuXw9azOkqKmFRT8aK16k+0Tr+murrMqydSqTUOSGHP1elzLNCfkn8nRHRxBhfZ2oZir5YuBWGBSwRbTpyJ6GtiedMxg5Nqj5Tk7z+U/Px61WfVlHnxRb5Fp8M9bdEbsbx+GM0hHXO7Y+P45k6x/DvX2ta+Lr1jFhHnexepc4JiTo2tQofY19+ZZc659nndvDHEHUbiIisej5m7tzNH+eVXp2Baarn67Nnc/y7pbtHZzQtUTurUs0Z+ui14QALRbqTz2cmBvVEb0TU82vr5Hm5kz5wgjonW+VjRzipx+GZEs9/FMVnuhzR4+PMDB+DXJePqdGkzoc0h/lnfOjgsOrjVvnjpBtFvi+sM46HslnW7h3cz9r5qh6HU2JenTTMz6LiWNLlOU7H0rcHk6L+jR/ifSxbnxOO89ccatRz2aOFaScAAAAAAAQGbmAAAAAAACAwcAMDAAAAAACBgRsYAAAAAAAIjHmH+HM5XvSmUq6oPj6vd0jVKj9Ql+ZBLiKi5vbF/IJiOnjuh0SxnDpe/NIhXeisLK53YpQXPpuc1IHZsseDW5ari/94Hg+3Vot8c4OSIcTvxPlriobF257RBTMr4jWHU7pPfSMPlLo23+wgX9CB/KzYfMG3+etpbtYh1Wopy9u+DuB5nqeOASwUTenFR+4ktDXw4rotYhMOIqJfv8pD/AOGzVMaw3w8DJHcfERvRkLEP68W8bEjUtXF0RaJImsXnnCS6hMRu3m4MtRp638zmxoeYe0XX3yetfcNDahzmjo6WLsS0TuL/OZFXmhzRrymZav0hgL5LB8zS0X+Ped7OvA7nc+ydiKpN0BYtVoXBgVYKCoh/vl8cWRC9RmdGGTtc6I8VL5q7Xp1TnPrR1i7OKMLxk5n+RxtRlT6rYT4XJGIaHiUF+2tT/OxLxLRG1Kl0zycXt+sN5eKxPj4EhXjVrmqr7/B5XPBerGBwJ4hvVnAAZ+PU21hPWdbVsfn47UQH5u9qp7TVcvyesX3iKNvKXyxQczkmP67P1r4BQYAAAAAAAIDNzAAAAAAABAYuIEBAAAAAIDAmHcGJiyK9FSrep2vKwo2ZlIZ1q6v0zmOzk6+djwa0QXJZrI8rzK8P8vatqXX6sXE9UbSPIdSKet14eUqXzcYMRQfs8VTJcJ8TaDTklHnRBLimMfXRtq+zgaRzddcuhGd80nW8dfkE1/jnS/qdY+OWIOeTvP1lJUmXWSoUuLHKp4uthSRuR6ABSSeaDlyJyEW4+dkos2qT+MEX799YGJc9ekRRXvr5EdRhhOJyLdExkWMoQ7pTElcHGuK6u+AWo0/Tl5kdl7d36/O+f32Z1m7IrKI9Yt4cUwiorEyf9zZKZ33kynNxd3drN3e3kXSM795mrWzU7xQ8cCgzuOUq7zP2nUrVZ8TTtRF+AAWioEDPOe259X9qs9LB2ZYu+Lx8eb8lJ6brF5zHGvPGOaC+Vk+D7IcPjKUDRnBrMgPO2IqGKnp+Y4b5rkYO6RzMlaNP1dhWrzmmh53fVEAvqu5k7WHSvp3iP4SL6Y76M+oPvL3i1aXZympojM8rjgnJPIt44YM+JgoUDpVMFRCP0r4BQYAAAAAAAIDNzAAAAAAABAYuIEBAAAAAIDAmHdwoSGTYe2oq/MhZbFmLhrha6SnpvR+2337+Jro1av13vyZNM/FjAzy9ZOOr/fmj2fEer4qv96Qrdcnzhb4+r1DBw+pPr54y5oXLWftnsX6+kMRnlWpenyv76ph7+9CXtQfMKxJr4nMy2yOX78T1n9HTpWv7cxl+Z7crqXfy+Z6/l76ll4z6oTfvnWNAEGTzLz1DIwV4tmViBtVfVYt4hnBX/7616pPi8jqHb+Y10iJmv6dSuRiqqJ+S81Qr8UTY5Dv6XFgaJTXJHjquRdYu29Ij6mJJp4B9MQ69l/87gl1Tl4898ply1Sf1WuP59crvrOe275NnfPcdp6BmZnh4+PIiK5/dtKGday9fv0HVJ+6VFodA1goitN87Jie1lkJN8XnSc+KeiH53/xGnfPB6SxrL23rUH1ci39m6+N87ufU9FzQcXjexonxz+/0DM+9ERFNjPOxryJyKEREtqiXFxJj/lRRjy9ZkeGpVnk+2tHTagqV+Nww6+o8dLYxw9p1Tbw9sKdPndPo8++FZjG3fdWQERwL8/uCWL0h832U8AsMAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgYEbGAAAAAAACIx5h/gdUfcsZMhs+6KT5/Hg0OSkDm7lS/yYR7p4TkOKh11lND1drwtkNjbxY44IvBdL+nmoygNU44MHVZexMV5IbnSIB80aBnmbiCjT0MraSRFiiokim0REsi5kuaKD/lVxrFziAf1KRQfyyeOv0a7x9z/sGDZEqOOBtpqvw7u+zsEBLBhl661/AOQ5VUNBta4G/tlLOvp5XhDFIXva2lg7EdKbeZDNP+cVcS0lWw/wxQofO/a++orq8zsRjPciorhxmw5wvtTPr3/PQT7uer748iGiEz+wnrXbm3V4d2ycj8XPPM83QOjbra/fqvC/g6XL+SYK69frgpRr165m7Xg8pvqQr4t+AiwUqRgfk8LxPtVnyUq+IVKmiRfZ7n1llzpn6Fd8040zV65WfZY08flXRBT+jYT15zUc4+NUJcQ3kjowrDcjGRcblHgVHZyPirmsHeXj7GhOz5H3j4yxdjyZ4dcyOqrOGS7wZH8+pgt8Jlv5mGmlGll7sPyyOmdKzJFnXP49csjSOwr4Mf7e2e7bNxbiFxgAAAAAAAgM3MAAAAAAAEBg4AYGAAAAAAACY94ZGN/ja90cR9/7xON8zWIyxde+VfUyPIrEM6ydzsR1nwhfo9jcwIvGLe3uUee0tTSx9vQkXw89sH+fOicZ4+v71qxYqvpMN/NsTbnGX5Rf4kU2iYiKo0O8PRXh50T4+0ZEFIllWNsJ6XWDRVE4tOqJvxNbF8ajKl9jGXX49TsxfU4qxt9/11Ags1AzZIoAFoipvKGa2BHMxnmGrSiDb0SUsfga6XNOPVX1efj3O1h7QGQ/2lv4GnAiIvL541ZEkcqh/Iw65bfPPsPaL217WvVpXcTXund0L2Lt3X171DkDYu340h4+7rY06SKhaZHL27tLj+fbtu9g7dnSJGsv6dDvy7rVK1l7UXc3a2cyfJ04EVFYFMi0LNPXKkKCsHAtX3ESax/o1fnihkb+2Vp1PC8QG4vorPBLv+Nj0M9+83vVpzPF8yyNdfxxOlub1Tmt4lpiLp9bjQ7yopVERJUczyC31GdUn5DLMzA1h8/rQklDfm5W5E7ENG/C1RnBIXEOVfT4MzTJMzqJkWnWHs7qQpxWlX/PZUVmvRLTz+OK15gbe/vmivgFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMDADQwAAAAAAATGvEP8daKgWq2sCyuGRXXLREqE0y0dNpJFymKGImBtnctYu7WFF+Cpa9QhTzvCg0ORBA81xWI6OF8p8HBRnQiKvnYeD2wWyzwUb1n6njDk8Le5JoL/lZoOAHviWnzS751TFIUryyK45ehAvi2uz63yIJdr8w0GiIhCIpTqGor2hcL6/QRYKGazU2/5nGySf34bYjqkmhDtZkMw9MTuLtauj/LPvakQZE0UuR0QxdDu/59fqnP++4knWLtc0ePWhV18bC6I57YienxftZKHdVvbePHI3Kx+nicef5K1+/v15gBOiI+zJ67lhfJOWs+fl4iooYGHd21R5M6yDWF88fYa3m6yDMV/ARaKrp5VrL1mvQ7xV4v8cz64l2/uUasa5lZRPkKOlvXmI2P7+SZK8RB/3PT+PnXOYhHi72lOs3ZLo54b1qf5ZgAtCUOBcvEa/CgfX1rq9DnVIh87Do3xsTpZ00WH68Q8zrL1VD/i8g2zyqJYp016Q62JCb55gW8aD4UZsYFWY8KwqcxRwi8wAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABMb8MzBpvuYvN6vXGvpVnsGQeRCL+LprIqJyha9Tnp4pqj6pGf5cYYuvw5saGVHnhMWtWUwU+ykV+PpzIqJSkT+35+mCO06IP07I5Vkb31CsUxaNc8R9Y9jV95FhkeHxDfmhaowfq9R4PqdS1edUxNr3aoj/J+AbFnBb8kXV9N9j2NWFNgEWipFhXdjsSGJRfk7a1euq4ym+Rjpk68/e6m6eGalz+PrnaUORzVdeeYm1/+epbaz9yLbt6pz9Yu11vqoLne09yNeXpxp4Ebm0oRBkPMXXUb/yyi7Wfv7559Q5U1O86FpLs0wLEW3YsJa1ly7jBTJTMqNJRHZIFBlWBYT1enNLFButVU15F0MwBmCBiMV4nmLJqhWqz749z7J2375e1vZ43JiIiEb6+dxvdlKPSWGZ4a3jGcG2RRl1zoomXpC3OcLHhZZmwzgW4vO4SFmPA1WRbZ4WGepSWc9LK7O8T1EUmKw3ZJ3rm/m4W43o8cduyLB2op5neNoX60LuxSn+fk9P8Cx8Syt/TCIiW4yPa44/TvU5WvgFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMCYdwYmKmoLVEXehYioWBD5Cl9kYgzrgCfz/B4qOzGh+oyOZFm7NcUvO2bIXyTjfM1lY4bvr+0Ybt2qVbHI0pA7CYv15ZEovxaZMSEiqpT4OkG5JtCy9cWYajdItqgvY8lsSk3/HdnE8ywhkUvyfH395PPr8zy9DrxqXPcNsDAcGhg6ciehWN3N2k3pNtWnlBvnB2p6IbgjsobLRB2VV3Y8o87Z9eKLrP3SHn4t48WsOqe+k6+RLg3q8WUqy/M2ns/zLZPjk+qcbb//PWv39e9nbccwPnb3LGLt9evXqj49S3mfmPgOi4T194Yn3ssa8es3jco1Me7KTAyR/jsCWEhyZT4uuCmd27BECZTMIp5RK4/rOUZI5IcrRZ2hLpdErcEInwOlUnycICJavngJa7cn+PXalg47V0TWsJjTeZahcZ7n7h3nmZIpT48w8WiDOMBr0kRUTo+oXORjs20YuLwafz+9EB/H4m38eYiIWku8hsu+V8X1D+u8Zedy/l52re5SfY4WRlUAAAAAAAgM3MAAAAAAAEBg4AYGAAAAAAACAzcwAAAAAAAQGPMO8dsihBgyBCCtGg+rFz3e9jxeaJGIaKLCw1FDU/q5c1NZ1g6XeSIp2tykr8XiBeAKJR5+DRlSTW6YBzZdV19vKMSPybC95xkqWYo+tSoPS1XK/H0iIqqJuKhj2HVAXq8vnrtS1IU4qyUe3LJsHjit1gzFl0QhSzeqA3gRR79XAAvF+JjefORIDk3uZO3Vq09UfdKiINnslC7UNjbKC2JWxMYck1UdJo028KKZkyUeLm1Z2q7OaV7Sw69lVl9LUWxY8vLLvCjljud4YJ+IaGKKF8hsaOSvec0aHdBft44fa2puUH1cVxTpFX9eqRmKu4lidHJ8NwX0FVMXa95ftQDvO6PTfJMT3zBdsMP8oBPjn0UnoTcYalvO535hR3/4JgazrO2F+LxpYJj/ORHR07N8k5MVizOs3WDYOKpW4BsITBiKve8Z4SH3Fwb5++LF+LyViKghza/XExuLlA0bR8lCv7my3vwlXeZj/LrVq/jz8P2qXnuuGN8cIJLh78PBJw+oc1zx95rNZ/UDHyX8AgMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBgzHthbqHE1/eZshKWWD9cE2uMCzW9PrHk8GI5eU/nNgqTvLBZA18CSK0NuuBOWVxvpcLbsYh+6YmGDG/HE6pPVeRX5JpvmXchIvLF6uuqxx/DN5wjsyk1T7/fXkkUkhMPY5nuT0Ufr8oPhGzx5hKRLfM3tv57rJkKYAIsELWqIft2BCOjfP3zCy+9pPp0bvwQa6dbdLHLiBi3qkWeebHTvCAcEdHg7j2sPTTDw4edXR3qnHCCL4o25fJefP4F1t71Cn9NobA+Z9mybtY+Yf061u7u5tkbIqJ4ghcqDhvWpMt/n5PFl31j0V7etES2z/eP/PfsOPq7xZpHYWKA96uRyQHWzhhyy+EIr2RZ9XiRxLKli1Q2L+LZt6Y6PRccrucZu3iGZ3gbYnpMKk3z+dbL42OsXefrc2wxlxrJ6aKOs2KcKqT42FwwjEnZaX79NTF/JFuPSeEUH7eKs3peXR7mj1ua5O+v5+i5YG6W95FzzJAh65ef5vmbwYlx1edo4RcYAAAAAAAIDNzAAAAAAABAYOAGBgAAAAAAAmPeGZjZIl/HVvUNmQyxfq9a5uvlcp6uH+I7vB5BrTyg+gwN8AxMbVLUlynqegRNaf64yRRfX5nJZNQ58Qi/vpBjWlfN6TSIXuss80KeJXInriF3IvrIPBERUaXKH7cm6sBYll5PaYvaMdUKf4xSVa+VdGP8fbBC+lVP5afVMYCFIhmvO3InITQxydrPPPW06tOa5mu8V61ervp4If4ZnhBrm2fH9JrjPSO8T1XUt0ok9espiLov1apek17z+dicSWZYe83aleqcdSfw+gNNbY2sHYvq2giyHotaF/5aJ9F2RFP/+52s+2Krui+mIi/zqAU2n/oxAO9TuQKvk5Ws6Fxec8Mi1p6e5PWtZqt8vCQiCoucRlNHi+rTuqiTtZ0k/yyWswfVOYuJZ2lmq/yciVE9pk5N8DnQaFXX92tcxMe2JUt51rDvoJ7/OmIel0zz+Vg4oeePYZHxzk/qWmBjA/xYYZqP79FGPmcmIrKq/HFDNr+2jm6dbUpm+HdJ6W2siYVfYAAAAAAAIDBwAwMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBgzDtNUyzzcLcM7BMRhWUY3ebhy5Kng6EVn59TqxRUH1sETCsuLzA5ktMBSc/jhdnI50XMXEMQfTLMn6dW06/RdXmfkLg2U36zWOBhrlxebjpgOEkUTDPUjiQ3xN87R3SyQzrc5Ys+RbHBQL6k33+H+HtXmJ1SfaZmdUgMYKGIx+JH7iTPcWZYe2C/DnA++vCvWDs7pT9nkTQfg6pl/rjTQ7wIGxHRrv5B1s6XeRC9ktObtIyO8sJytZre8KOrm4dS168/nrVXrFyqzkmKQpshMcaagvM1UVDYVF7SCfGvN1eM1aZgvS3GXfkdYNuGAna2LOCsr6bmoZAlLFy2LcYTQ2Hu9ja+QUl2nBf6HT6ow/Z2iAfaG1o6VZ+G1lbWnijycaz30C51TrjEP/chUXAy3KHnsmWPb2pSLuvP/JTP534d7XzjglXtuhBndpxvZiA3RAg36+8eV8zs0zG9aYItNq0qO3zuF3Z0iN91eTHjUJx/B3SuW6zOSaT434mVbFB9jhZ+gQEAAAAAgMDADQwAAAAAAAQGbmAAAAAAACAw5p2BceU6ZF/f+zhi3aDl8KxKxNXr+6oFvu4uEqqoPt1dy/i1iPV8xYLOZMxU+LrvqMidmIuY8ddYqehrkUXV3DBfE2iqp1Yo8HWCMhNTrernsSy+jjpkuNW0I3xNt+3yTp4l15IT+aI4Z0WsHi/4+lrys7OsPVvUa9+dUEQdA1goLO+tFypMirEjash67H5lN2sXyjpf0drDi7flZ3mRtQO796hzxgZ5IcuQw78K9uzS5xTzPFuztGeJ6nPKqSexdkcHX38ejetixhGRH/LFOGzKIspClral836WyKZYouidqeDk0URV5PXJgsKvPS7+rRAWLs/jc55K2ZS15Z/Pliaep9gbe0mdEw3zLEpdQ5vqUxVFweWHvFLRn83dA3x8jDfwIpXRjM6HjI3zPnlDnrta5GNDbILnfBradYYkXs+LQ06XeNZmZETnIuMR/hrDth4fnRSfn1OEz/2qhrEwFuevu60hw9qNpLNBfl07a+d9PS89WhhVAQAAAAAgMHADAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDMO8RfF+bhdarqQkSVHA/Kh5P8nIqtwztDRR62j1ll1Sca5yFPO17PL8VQ5bE4w4P9Q5OTrJ3L6yB6Ps+fu1jSfeIx/ppCDg/iuq4OqdoiQBWO8MdIZ3RhH1GDjUq5ad1HBONqZd72fB0mdWI8xC8rHlXL+p62Ju5zU431qk8qqTdoAFgwxGd6PsJJPl4m0zo1OZ3n41Zf74zqc7BfbCRS4eNWuahDnrUyH7+tEH/u6Uk99n3gA7wo5foNJ6s+Ta3NrO2I8cUyFNf1ZMBUbohgGN9tccyUvQ858jw+HnqGYnqyQKbcnMEzXEtNjLPGcddQABNgoSiKjX8KJT0mlcQ4FY/ywLicBxIR1Sr8k+9G9fxrpsg3NSnm+VyqnNfjgCcKNlbFuDU+rjeOGhnMsrYVL6o+oTR/nKIoCl6YzqhzYmm+SUuqjs/zDh7oU+cUa3zjpVhUT/WjGbEhVY2H+C3DxiPRKN9Aq7udX689ocf3cbGZVDmv5/hHC6MqAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgWH5vn8UpbsAAAAAAADeffgFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMDADQwAAAAAAAQGbmAAAAAAACAwcAMDAABvatu2bXTmmWdSIpEgy7Jox44dx/qSAADg/+ju7qaPfvSjR+z32GOPkWVZ9Nhjj80d27JlC3V3d79zF/cOCR3rC3g/GxwcpDvvvJMuvfRSWr9+/bG+HACAt6xSqdCmTZsoGo3SbbfdRvF4nLq6uo71ZQEAvKO++93vUjwepy1bthzrSwED3MC8gwYHB+mWW26h7u5u3MAAQCDt2bOH+vv76V//9V/pqquuOtaXAwDwrvjud79LTU1N76sbmLPPPpsKhQKFw+FjfSl/MCwhAwCANzQyMkJERJlM5k375XK5d+FqAADgaNm2TdFolGw7+NP/4L+Cd8jAwABdeeWV1NHRQZFIhHp6euhzn/sclctlmpiYoC9+8Yt0/PHHUzKZpLq6Orrwwgvpueeemzv/scceo1NOOYWIiK644gqyLIssy6K77777GL0iAIC3ZsuWLbRx40YiItq0aRNZlkXnnHMObdmyhZLJJO3Zs4cuuugiSqVS9KlPfYqIXruRuf7662nx4sUUiURo1apV9K1vfYt832ePXSgUaOvWrdTU1ESpVIouueQSGhgYIMuy6Oabb363XyoABFh/fz99/vOfp1WrVlEsFqPGxkbatGkT9fX1sX4333wzWZalzr/77rvJsqy5/t3d3fTSSy/R448/Pjd/O+ecc+b67927lzZt2kQNDQ0Uj8fp9NNPp//6r/9ij3k4b/KjH/2IbrnlFurs7KRUKkWXXXYZTU1NUalUouuuu45aWloomUzSFVdcQaVSiT1GtVqlr33ta7Rs2TKKRCLU3d1Nf/M3f6P6Hfbwww/T+vXrKRqN0tq1a+knP/mJ8Zpen4Ex8TyPbr/9dlq3bh1Fo1FqbW2la665hiYnJ9/0vHcTlpAZDA4O0qmnnkrZbJauvvpqWr16NQ0MDND9999P+Xye9u7dSz/72c9o06ZN1NPTQ8PDw/Qv//IvtHHjRtq5cyd1dHTQmjVr6Ktf/Sp95Stfoauvvpo+9KEPERHRmWeeeYxfHQDA/FxzzTXU2dlJ3/jGN2jr1q10yimnUGtrK917771UrVbpggsuoLPOOou+9a1vUTweJ9/36ZJLLqFHH32UrrzySlq/fj099NBD9KUvfYkGBgbotttum3vsLVu20I9+9CP69Kc/Taeffjo9/vjjdPHFFx/DVwsAQbVt2zb67W9/S5/4xCdo0aJF1NfXR9/73vfonHPOoZ07d1I8Hn9Lj3f77bfTtddeS8lkkm688UYiImptbSUiouHhYTrzzDMpn8/T1q1bqbGxke655x665JJL6P7776c/+ZM/YY916623UiwWoxtuuIF2795Nd9xxB7muS7Zt0+TkJN188830u9/9ju6++27q6emhr3zlK3PnXnXVVXTPPffQZZddRtdffz099dRTdOutt9LLL79MP/3pT9nz9Pb20p//+Z/TZz/7Wdq8eTPdddddtGnTJnrwwQfpvPPOe0uv/5prrqG7776brrjiCtq6dSvt27ePvvOd79D27dvpN7/5Dbmu+5Ye7x3hg3L55Zf7tm3727ZtU3/meZ5fLBb9Wq3Gju/bt8+PRCL+V7/61blj27Zt84nIv+uuu97pSwYAeEc8+uijPhH5991339yxzZs3+0Tk33DDDazvz372M5+I/K9//evs+GWXXeZbluXv3r3b933ff/bZZ30i8q+77jrWb8uWLT4R+TfddNM782IA4H0pn8+rY08++aRPRP4PfvCDuWM33XSTb5r63nXXXT4R+fv27Zs7tm7dOn/jxo2q73XXXecTkf/EE0/MHZuZmfF7enr87u7uufnh4bHzuOOO88vl8lzfT37yk75lWf6FF17IHveMM87wu7q65to7duzwici/6qqrWL8vfvGLPhH5v/zlL+eOdXV1+UTk//jHP547NjU15be3t/snnXTS3LHD1/Too4/OHdu8eTN73ieeeMInIv/ee+9lz/vggw8ajx8rWEImeJ5HP/vZz+hjH/sYnXzyyerPLcuiSCQyt36wVqvR+Pg4JZNJWrVqFf3+979/ty8ZAOCY+NznPsfav/jFL8hxHNq6dSs7fv3115Pv+/TAAw8QEdGDDz5IRESf//znWb9rr732HbxaAHi/isVic/+7UqnQ+Pg4LV++nDKZzNs+L/vFL35Bp556Kp111llzx5LJJF199dXU19dHO3fuZP0vv/xy9ovFaaedRr7v02c+8xnW77TTTqMDBw5QtVqdex4ior/6q79i/a6//noiIrVkraOjg/36U1dXR5dffjlt376dhoaG5v367rvvPkqn03TeeefR2NjY3P9t2LCBkskkPfroo/N+rHcSbmCE0dFRmp6epuOOO+4N+3ieR7fddhutWLGCIpEINTU1UXNzMz3//PM0NTX1Ll4tAMCxEQqFaNGiRexYf38/dXR0UCqVYsfXrFkz9+eH/79t29TT08P6LV++/B28YgB4vyoUCvSVr3xlLnt3eF6WzWbf9nlZf38/rVq1Sh2X49xhS5YsYe10Ok1ERIsXL1bHPc+bu97D46QcF9va2iiTyajnWb58ucr3rFy5kohIZYHeTG9vL01NTVFLSws1Nzez/5udnZ3b2OVYQwbmKHzjG9+gv/3bv6XPfOYz9LWvfY0aGhrItm267rrryPO8Y315AADvuNf/Eg0AcCxde+21dNddd9F1111HZ5xxBqXTabIsiz7xiU+weZkpwE/02mqad4rjOG/puC82PHmja36neJ5HLS0tdO+99xr/vLm5+V29njeCGxihubmZ6urq6MUXX3zDPvfffz+de+659L//9/9mx7PZLDU1Nc213+3/6AAAjqWuri565JFHaGZmhv0K88orr8z9+eH/73ke7du3j1asWDHXb/fu3e/uBQPA+8L9999Pmzdvpn/8x3+cO1YsFimbzbJ+9fX1RPTafO31W8PLXzOI3ngO19XVRbt27VLH5Tj3hzo8Tvb29s79ukP02iYC2WxWPc/u3bvJ93123a+++ioRvbar2nwtW7aMHnnkEfrgBz/Ilua91+CfzwTbtunSSy+l//iP/6BnnnlG/bnv++Q4jrpDvu+++2hgYIAdSyQSRETqAwQA8H500UUXUa1Wo+985zvs+G233UaWZdGFF15IREQXXHABEb1WKO717rjjjnfnQgHgfcU0L7vjjjvULyvLli0jIqJf/epXc8dyuRzdc8896jETiYRx/nbRRRfR008/TU8++SR7jDvvvJO6u7tp7dq1f8hLYc9D9NqOaK/37W9/m4hI7do4ODjIdiabnp6mH/zgB7R+/Xpqa2ub9/N+/OMfp1qtRl/72tfUn1Wr1ffMnBa/wBh84xvfoIcffpg2btxIV199Na1Zs4YOHTpE9913H/3617+mj370o/TVr36VrrjiCjrzzDPphRdeoHvvvZeWLl3KHmfZsmWUyWTon//5nymVSlEikaDTTjtNrfsGAHg/+NjHPkbnnnsu3XjjjdTX10cnnngiPfzww/Tzn/+crrvuurnJw4YNG+jP/uzP6Pbbb6fx8fG5bZQP/2shfr0GgLfiox/9KP3bv/0bpdNpWrt2LT355JP0yCOPUGNjI+t3/vnn05IlS+jKK6+kL33pS+Q4Dn3/+9+n5uZm2r9/P+u7YcMG+t73vkdf//rXafny5dTS0kIf/vCH6YYbbqB///d/pwsvvJC2bt1KDQ0NdM8999C+ffvoxz/+8du2tPbEE0+kzZs305133knZbJY2btxITz/9NN1zzz106aWX0rnnnsv6r1y5kq688kratm0btba20ve//30aHh6mu+666y0978aNG+maa66hW2+9lXbs2EHnn38+ua5Lvb29dN9999E//dM/0WWXXfa2vMY/yDHcAe09rb+/37/88sv95uZmPxKJ+EuXLvW/8IUv+KVSyS8Wi/7111/vt7e3+7FYzP/gBz/oP/nkk/7GjRvVlns///nP/bVr1/qhUAhbKgNA4LzRNsqJRMLYf2Zmxv/Lv/xLv6Ojw3dd11+xYoX/zW9+0/c8j/XL5XL+F77wBb+hocFPJpP+pZde6u/atcsnIv/v//7v39HXBADvL5OTk/4VV1zhNzU1+clk0r/gggv8V155xe/q6vI3b97M+j777LP+aaed5ofDYX/JkiX+t7/9beM2ykNDQ/7FF1/sp1Ipn4jY/G7Pnj3+ZZdd5mcyGT8ajfqnnnqq/5//+Z/seUxjp+//3y2bZamOw1s8j46Ozh2rVCr+Lbfc4vf09Piu6/qLFy/2v/zlL/vFYpGd29XV5V988cX+Qw895J9wwgl+JBLxV69erZ57PtsoH3bnnXf6GzZs8GOxmJ9Kpfzjjz/e/+u//mt/cHBQ9T0WLN8Xv7kBAAAcAzt27KCTTjqJfvjDH9KnPvWpY305AADwHoUMDAAAvOsKhYI6dvvtt5Nt23T22WcfgysCAICgQAYGAADedf/wD/9Azz77LJ177rkUCoXogQceoAceeICuvvpqVR8BAADg9bCEDAAA3nX//d//Tbfccgvt3LmTZmdnacmSJfTpT3+abrzxRgqF8G9rAADwxnADAwAAAAAAgYEMDAAAAAAABAZuYAAAAAAAIDBwAwMAAAAAAIEx76TkLX9zDT/RELKU1ZNrNe9N/9x0rFKpqD46psPPsSx9H6bP4W3Ttfi+p45JtVrtTR6VyPQQ8lixlGft1Wv1jjtdy1azdu++CdXHTbayth2Ns7ZjuD91yGHtmscvzvgOiLfKm0ds6ppP/MkR+wC8X/zrE//9ls+ZT/jQEp2M/+IkxszK9BRrz4wNq1OyYyOs7Xk10ebjBBHRoYEh1p4cGVN9rBq/YK9SZu1SsajOSSQSrB2NRlk7NzOrzqlW+fW64bDq44TdN+1jOfo7rFgssfbUFH8vbcN3zZLlS1l71RkbVJ/GpXyMv/rsi1QfAHhj76W49vzG7iP38sXkyjM8shpxLHnOkenR/O1hmkcfDfl3O9/HxS8wAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABAaqhQEAHANvnuz7P8fE2uC8IQ+SHRxg7dLQId6ezKpzpmenWTsnsh+1mmktNv/3rpitV1aXqiV17PWi0Zg65ro882KJFdsRkYkhInJlZsfwXJUaz98Uc/za5POYRKM8N+Mb3peZiUnWzo6Oqz51i9qO+FwA8O6bT7bGF6O15+lzZB/L06OSLQ5VLH6gZsjAREQe0XLEuOUYfocQXybzyuy8TXmWdxN+gQEAAAAAgMDADQwAAAAAAAQGbmAAAAAAACAwcAMDAAAAAACBgRA/AMAf7MgFYX2VkeT/fmTX9FmTB3lAv//lnbrPIA/t2+Uqa7uG7QFm87yYrm/zPlFRFJeIqCSC/nnxGES6yG0oxItJOrb+yqlU+fWWq7wwp6nAcK1WFe2a6lMVx2TRXlkQjojIsvnfSVgUv4zF9IYClojIVgoF1YfEawSA9whDwt2XBb7FWFIq6c1K5Ofen86pPrUZXsi3SuJxHT0mhcJ8zImmkqwdq+NtIqKw2HzEMwT9ZWj/SO33IvwCAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGAgAwMA8AeSRczmwxLrqkf7B1SfwZ088zJ5sF/1KczwtdY1j69dLpV4QUciIlusiW5paeXnlPU5YxMT4nkMhdrEumk3xF9jPK6zNZ7IuMhsjanQnDziGa5FnifXdDuGdeG2zAuJXFKV9Psii13ans7jeEV9HgC8zeYxVqg/r+rPa0mMQeNjvDjt/v371TkTIq8YGpxQfari2FRuhrXHQvpaKmmecWlc1M7ai1csVecsWdrD2m3tupCuHIvnk3l5r+Vi8AsMAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgTHvDIxc+2ZalxwK8YerirWFpr365T778jGIiDyxplg+tWlZnlwS7ftyj2vniOcY6wSItqxR4Pv6NVq2XHvN2+NjWXXOilX8fWioT6g+MyW+p7jtx3gHQzEK3+Jv3uzMLP9zW7/mmFy3/t5aBglw7InPhGVYeG2JPMX4fl6/pffZHeqc2eEh/hhVnaWoqBoFvOZIpaJrkMjx/NAhfi2yhgqRHvNdw1gt+8jxPZ1OH/FaolFe98D0XSMzL6bMjqzVUKnw+jKm3Izj8O8Fef2uq1+z4/Jz8rO6/sP0+KQ6BgB/IMPYIFmij8zu5bJT6pzdL+9i7Sd/+1vW3r5jhzpndop/xtvssOrTY9ex9tDoMGu/nOVtIqIxl1+/neDzvEyHzrcct2E9a3/knD/SfdauY+26On5tpoygdHRTwbdvAolfYAAAAAAAIDBwAwMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBgzDvEL4OU8wlW2iIQXqkYCnyJxzGF+Eslfp4MfZo2B7Asfm8mw5mG/CaRCPr7hj41EcT1iT+33HCAiMj3RYjWkoXbdMh2dJQXTkql61WfsQO8KFKl4orn5W0iIjfM39+pKRFgs/U9rRvRYTTJ9N8DwEIh/+t3DEHF3OQ0a/e9wItUju7t0w9c4UF00/hSqvFwummDEqlQKLB2tcrHoIgI0hMRuS4fT2zDWCGPyXE3l9MBd/m9URaBfNP4LjdYMfWRr0k+j6komwz6yxB/IsmLyhERRaIR1i7l8qrP1MiYOgYAb2xecwo5LzWUrbRq/Nj0KP8s7nhmuzrnl//zS9be/vwO1p6Y0Jty+GE+9g0b5rK5GB/bonE+ps6M681IJsV3gO/zMWrykB77hn7Fv2smhsdVn0su+ihrb/jAetbO1GfUOW6Iv0bbkt8BpoC+2vrK0GceD2OAX2AAAAAAACAwcAMDAAAAAACBgRsYAAAAAAAIjHlnYGQ2xbTmWK4flmsY5XpoIiJPPE7FGE55c6aCZHJ5tszEmKo8ev6R8yyeyLPIh3XDevGe6/I10pEIX1+eSLaoc8oV/jgzI3oNY1/fftYulUdYO1XXqM7p7GxnbblmvVzlf4dE81s7DgD/l28YkyYGh0WbF4/0ynytMxGRL4oBV2r681kjmU8UY5Th+pIiy+GKrIfp+mueXLtsWsvMz6uIPEvO8L0hxxfZDkf4+ElElEqlWNv03ZLP85yP/M4yDWPFIi8OLPM4sk1E5IZF1rCm35fyrM7FACwU80rIHqGTKRNTFXM2u6rzxLUcH1ef/+3vWPvxx55Q50yJz3nbimWs7Y7qTFs0ygtMRhO6+LhV5I8bLvDrb52ZUedMjB9g7aJ4jUUxzhERFUUx4+25F1SfmhinigU+Rn3g5JPUOa0tfE4ZdUU+2pS/nEcm82hnlPgFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMDADQwAAAAAAATGvEP8Muwti4QR6XC3bM+nMJEpkC/P06Fy0yPJc3j4tVrTYUxZlNKN6PCRK0JLiQQPbiWTOriVSMRZOxbjAdRSTZ9DDg+ujg4Pqy6FIn8fcnkeAIvF+fMQEXnem2+sUMrrQnPy71oWdwNY8MQY5BnGx1yWF43NicBmqaJD/LYcDg2DnS/D9WK87O7pUee0tray9sGDB1l72DDe1MS4ayo6bImvFDl2lEr6NcoxKCHCr/X1uohvIiEKShq+A6KiGKcM6MtinkT6O0t+10xP8wJxRET5Ah8zLd/wd2R4rwDgTchhzZDyL5R58Nya1Z/pid39rP3yU9tYO24o2tu6cjlrz4r546u9r6pzGtMNrF3f2KT6zEzzAphTe/hGTHUpPWdrmeRzwcEyf42+pd+XUI2Ps+Wi3vyl/2VeSPkpMWxVs3rjqDPPOou1m1v494gV0e8liY2tHFuPj7Z4CdY8Y/34BQYAAAAAAAIDNzAAAAAAABAYuIEBAAAAAIDAmPfCXLk22LT+WWZVZORFFlgznXM0bFtnVRxHrLsL8eeJp3h2hYgoFuPZjnhcF1DTeRbeVrXeiKgqCnzmi3zNdNXS6wbjsTr+vCn9Glvb+N/B2ARfsygLcxIRzczydfcy22RSk+v5DRkYFLeEhUyt2TV8rixR8LAoxsN8Refy6kXWoy5Zp/p4Yg10bpaPL/kcz34QEY0M87FidmaWP6YhiyizKqGQHgfk94IcDmWOhojIdfn7khTrwGWBOCIiS/zbW8lQYHJqmo9109P8NdYMhYqdkMxt8j5VQ0YwKtZ9h0Ku6lN7G77nAILqaP7rVzOKqh47Jnv3sfbMK/tUn5EXd7O2PciLUDZ3L1LnuA4fxywxDiRVYXSiepGlqY/pcSsvMjCv9Pfya6noedTyJM/WuLP8+suG4sB1Ymof8vT1JqZ5liZ+gBdWnpjV3xsHxKFCdzd/jJZmdU5mUQdrW3VJ1YfC/HptZ36/reAXGAAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAJj/hkYj6/Nsw2nVny5RpGvfLQdneOQfSwyrUvm54UjPJsicyhERMkkP2aH+LXFE3qdsjxWMdRymBXrAkcH+LrByUm+7pqIqFKW9WX4GsCWRXq/8IjF15eHo/q9iyfFuu8pvr6yZsjAlMU6e/neOYa1+7J2Qzyu69YgAwMLmSXCb1ZIf17bV3Sz9nKxHvrgS6+oc5IiZ9JY36D6VKr8M13I87XNhw7pmi62NcrajhgO7ZD+PIfE+G36yOuoh8giGr4D4nE+BkXE+G5aQK/qyxR1BqZQ5OOWJ67FNmRVqjU+vluWGLvlG0VErhgzLcNrjKZ1fQeAheJoss5yeKkZsnxDTzzD2rMi70JEFBbTuEVxPv+ayevH9QZH+GOIeVNrSc+tYuO8zpdd0q/ZzfIx35/MsvasoV7LmV3rWLvH5VmbSlHn8upE/iYe0XPkiM3HMrfM5/TpScP88fFnWXvf08+zdqxVZ2Da1qxk7czxq1Wf1Gpedyfk6nHWBL/AAAAAAABAYOAGBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMCYd4hfhrA8Q/EcyxZBVhFudA2hyUSCh1SjMd0nluCBpFiMh8hDYV0IslDgwayp6Sxr21V971Yt8LTXRFaHo4YO8WJok5PTrF02hLvKJbmBAA+pNnTqgkeeFRJt/bjhKH+v6up4kTtTsLVa49dSqYoimwUeACYiiogCTYakriosB7CQ6ZJrRKnGetZeuXYNaxdHJ9Q52YEh1j5UPKT6FEv8M1sq8c99JKqH+USCj6GWxcc+y9Gf8YgoKOnYeqzOi0BsqczHF1MBZP0Yedb2I/pa5GYA5UpJ9ZHp/3CIj1EhR+9CkG5s5H1c/jyFgn6eaoX/bRtqxlEspTc+AVgo3o4yrjOj4+pYdsdO1m6q6Q+fKzYJccRnOqanVmRN8jGoJDZKabL1nK2aFwH8nB7Pk2LIiXTx8PrsJA/5ExGtW8QLbTphvgnB7JDepKUqN9AyTPWtMu8zXuFz2UJRv98tNf4ComLjq9m+fnXOnld2sXasf7/qs7aJj7ux5PzGS8w6AQAAAAAgMHADAwAAAAAAgYEbGAAAAAAACIz5F7J0+DrreET3SSb5ukC5zjqW4BmN187h6xNDYV0ETBYkKxT5osWRIb1Wb2CQF2qbyfE1jfXNfD06EZFc0p0r6KJCuSLvZLm8sFzE1feETkTkTsSay+mcXlfd0s6fp1DIqz45kevJiPd7tKSv3xPFRm1D4UppPn1MuRiAhco3rvrm64fLZVmAUn/GZ2d55s70OZOf6ZDDP6+pOr1ee+nSLtGHj8NJQ2ajoYGPdUVDxu7FF/ia9N29/PorFT0m1USeUhbFDdXr7wRPBE3y+VnVp1SWBer4+xSL6tfY0cYLsVXEteXLY+qcqCgG3NTervq0dOpjAAvF0cwO5Bg6aSjI++qvn2LtQzWdPrQSPMPrinlq1DANjpEo2ivGVJnLJiLybD5ueRV9LQkxlCVEFrsY1mNSUozv5TjPjY/XdCFO8vm1pAzjeaHK552j8uIihjzRMC/W6Y5lWbsY5ddGRBSq8e+J/b/fofpETzqRtVu6F6k+JvgFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMCYdwamuyvN2ulISvVxyny9sFw/nK437O0s1g2Ojet9sPv7+drH7DRfuzed0xt5D43wx/FsvjavbOk8jhvjb4cb1jmZRJqvd7bEuvZqlddTICIql/j15kWNmskp/ZqrYt1g1NV/VZ5YQ1+t8DWjjmHlaSrF9xCPx/nfiakOTDjM3zu5Rh0AuJDhM+LafI1xbnqGtaen+PpiIiJfZtYM/+YUCfNAYqOoN5NI6nXJHR1NrL12Ha9JUyrpXJ4cm0dG9Jr0ySzvUyqL/KKhjooj3pdUin+3xGQdKiKqVPjYV6vpPI6sS+a4PFfY3NGqzomn+fdCXoyxKzs71TnFCh/zo3H9PVff0qyOAcCbEHm/4YFB1aX/0AHWDnl6/uWJmlEZkW9JVXVWxa2JuZOYp1JMj6meyArbFUP9KjWei3othnneS6I+WH+S9+k31GvpSvG84vnLl6k+I4N8/J4hXgemuV2PdVOVvaw9OSxqxxjqQzbG+dg3LXOdRPT8E79l7dP/n/NVHxP8AgMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBg4AYGAAAAAAACY94h/g5RcLI0qcNSuSFeiK0gCod5IR1SnZzmIaDBYR1oHzwkgkIlHqiKJngwnYioJgqd+SLE71u6EJEb4wHOeCKj+zgyUMpDWJ4hRJbL87BuUYTGpmf0+3LwAA+nLW5vU316lixm7f59PGAVj+rX2NjIw7uqaJwIuhIRWfMoZIlgPyxklsfHgeKsoShljm+QUc3ztukzJGOgkYiuIFwvCkyuWr2ctVta9YYl9fV8zJSbhrzau0ed89xzz7P2uGHDlaIo/lsu83atqkOexQrf1MTz+PhYnzZsuOLyIG4yqb8DyjU+Fjui+nK4znBOiD9u2+IVrN3Y063Omc7x76ep8QnVp1zVBTwBFoqjmR34IsQ/NqGLyA77fAzNhnQgn8QmIatrfC4YMWy85IswuhyHqxW92ZEM8bu+oai5GNvkRkum74ADBR56fyQiCiAbxrHJHH/uNaMjqk9CfP/YZT4PjbXq4rtWhm/m5fl87M7W9DjnpvmmLLOG741927erY/OBX2AAAAAAACAwcAMDAAAAAACBgRsYAAAAAAAIjHlnYGb6s6w9PaLXAJbzfD1fJcbXHB/qPaTOmZzJsXbBsFR4cpqvGyyWRQaG9LVYovhippEX06lv1oXFLJH/sGydB3FC/DXZosBRpaYLtYVEobl4khc6q5R0EbYpUdQuGdPF3FobGll78ZIu1q45+q83ns6wdnYyy9qmtapuSD6OLtAkC+4BLCRTo7yY2EFDhmSs/yBruyJvETaMNwXxsbIs/W9OvlhXPXSIj7OzM3rteFQUYpsVmZ0DB3TRuPEJUaSyZMp18OuLx3kOr86QVbF9MYZW+OOODOvvjboMz8XUNTWqPq2N/FjjYl6YLZHm67mJiDyx9j0mMoNeWGeQYm6GHxBr7omILGveX7UAYFD19BxjyuKf14Ou/uxFRYZuIsvnaOmSftyKyG+LOI7K6REReWIcyxvGatviDxQS46Wsl0lENOLxnEleFMptX3u8Oicrin7+cttvVZ+NUZ6dtAr8O2ysqMf3ZS2L+PU282LA42O6uHFeFC92DO/LxMABdWw+8AsMAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgYEbGAAAAAAACIx5JwuzB3gxnVJJF/+Z8vmxiRIPhuZ8HbCqiEBPyFB8sWzzkHs1ZIk/148bifHCm4l0vfhzHoQiIvJFgkoWUiLS+UzP5yEyyzKcIx7XdUUhJVcHQ8MO7zM9a9iowMqydlIEZskU+BWXl8vrgnvqWsSGCChaCcCNiNDkyN5+1Sc7yMPoiSQv8OWV9ZhaE0VvCwW9SYjj8I1QJidGWbtc4iFQIqJIlG8KIse6guEc2+HjiWMocJupz7B2WwcvhmZV9fhIFf4a83n+3DVLvy8dPTxM2tTdrfrEW3nx34jYwMRUoNer8WuxHT7g+4axz7b4+BhPZVQf0+YLAPBmxLxJzm+IaFZk6SNNTarP4lWrWTv37Iu8ndObKMln8mSo37CRUU0cqti6aK8tXpOt5o96nJgSzx2ry7B2d/dSdc6oKM65v09/Hw2LHWK8Mv8eydKQOueQyPVnxSYKM55+L9tXrWLtuKXn6/v3vKqOzQdGVQAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAJj/hmYab4+znfjqs+oWI84Lor91Cq6+E+uxB83YViXTDG+Zi7i8MyILC5JRBSK8uurVPnjFvK6SE+qjq9Jdw1FkWQGRsdk9Fvqe7xAXVUUaquW9FpJN8bXSlblAksims7x/MpMgedkImFd/DIsskvFIl9vbsq3OGIduOPgvhfg9WTxxXJZrwXOzcywtvyshQ1ZuEiYjx2mx/V9/jmX2Y5iWY91JbFG2g2JIr6GdcqNoqhjNKrHuowoMJlM8TF1JsuzlERE2Rx/X7wIf9wusYaaiKh75XJ+LfX1qk/N5e9LVbzfhjQO2SExtolzLGOpX/FIxkKWyA3CwnU0//XLj0xb1xLVJ9rcws/J6IK2zYsXs3b/K7tZe2JyWp3TIArP1uREz5CPltdr6EKe6eDr/9xQfDxv83PcKP+eUNlnIhoTRdnHfP0dMFjmY7FdEXmWvM5d58d5ocrBohi7W/X7v+g4PlZ7hsxRzZB9nw/MRAEAAAAAIDBwAwMAAAAAAIGBGxgAAAAAAAiMeWdgDoile65tyHok+Prn8hRfW1giXcOgKPa9LuV0n7oGvvbasmWmRGdrZJ+yWKNeq+naAnKdpmOb8iD8emMRnrXxfH0tcu/visihlPJ8HSERUbEo8iwRvT5eXrBc+27ZhroSHn9cWf8hHtfZJlkHxvTeASxksRSvKxVJ6s+RJbJk+Rxfg+yk9L8nxWIix1E1rB/2RJ5FjBWRmB47HFFLy3HEmGrI5dUq/HnqmnXupKGBH5uZ4q9xYmpKnVMJ8e+S5m6+Zr1pxQp1TkjUe6gavo884u+3zKFYhhSMT3r8PjL5PWFaH48MDCxcR/Vfv/jMLFujs3DnXnopaz9pqCeSr/KxbETMX2qWHutK0bA4IsYOw+dZjkCOIXtoiflWSMx/q47Oz00Tv96qeAyrpsesWZGPnvD1axwWFxyr8teUdwzzXzEnnhXXYhvmv7kK/86azenMUbvINM4XfoEBAAAAAIDAwA0MAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgTHvEP+k6OoaAuLJMD8WjfJAUsHTASVZcNJyDMXcYnxzAM/nQSKvpoP/VRE4teS9miFsJOszRiO6EGRUFBGyRAjLFHCPiSI901M8tC+L4BER2SKQn8vlVB+poaFBXKu+fpkvDYkArSnEL6+FfP33CLCQpZv4Zy/dqAPukyLYnxeFLQsF/RmXm3e4YT1kyzEnGuKf+3Qmrc5pEQXHLBHOHBocVefIcWp8YlL1yYtiuhVRHM1NJtU5PWtXs3br0h7WjohimERENVnwzT9yTFiF9i1D2P7o4sb8MRDYB2Dm9Zk4QpHH+pZmdeyS//dTrJ396X2qz/DkBO8jQuXJFr4hCBGR297O2rUaH8c8T4fii2J8HBk6pPrkpnmAvb6Oj83xhGGsK/Ni43Vio5S4YYOnQom/xpyhMPFUhj+XTWIjl7DcyICIEvy7pezx5xkbG1en/Pw//4u1V69eq/pc9KlP6ueaB/wCAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDMOwOTn+YFyBKRmOpT8WTBHZ4zqXp6HWQsztdE+4a1er7H10a6Ib42z3d1nqVc4rkYuQRTFqQkIorF+GtKJHQeJJXkBescm19vThQQIiKqiPXaSfG409O6sI9cb16t6myNzKaUy3w9Yjymr18+juvydY+ZTEad44jiSqWSzhwdaf0qwPtZoo6vJ04acicxkS8r5/lYYfqMh0L8sxcOu6pPTXz0fJFnMeXaOpfwNd6ZNL/+pnp9/b7Hx5vJCT1uHRri2Rk3ztdM9/R0q3Paly1l7ZDIvPjWkf+dzTLkWUgUpZQ9zHkX5FcA3m7Gj6cgu/hy0hbS48CilctYe8nixarP6NgIa9fE+Nh9ysnqnDPPOJO1SwWeQylVDEXZxbHHfv2o6rNv+3Os3bKOF+dcu+54dU6nfG6Xfyf4hjxOSRTRdML6O2DpOWez9qIMz3F6js7A+HGetxn/1WOsPfDsNnVOtq+PtT/4kYtUn5PPPlsdmw/8AgMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBgzDsD097M95626/Sp2QKvazAywdd4NzTztc5ERKk6/riT2SnVx/f4WuaIyG2EQ/paKmIPa7m+XOZFiHTuRGZviIjC4nFdkW+pVPTjhkTeJilyNGFXr2sviUxJIpFQfeT1yrovKUP9hKGhIdaW6+Nl3QmTqqFujY8MDCxgrvjsxev0Zy8kPlv1om5Tpao/V7JeVcKQZymLsS0S51m+po4Wfb1ibEg08rowPWI9NBHRTJbXqSn7etyNTPE6MLaoWxOt4zW9iIhccb0kMndqcTyZkiqmTjIchHwLQFDppDORLTKBYVdnqKfHxli7tYWPh3UtreqcvMg2y/lkxVDXxhfz0MXLVqk+dpjP4zqXdLG2m8qoc1JJnnEpTPLXMzBwUJ1TLPLcTHNLh+pD4vtnNsXz6K6tMzCWqIPYuYrX8Do5rOePixYvYu0TNmxQfdyEoWbhPOAXGAAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAIDNzAAAAAAABAY8w7xt3XyoFP/xKjqs7d/nLXdKA+GtrXrIkMhEWAfG8+qPoUcD4ZGXR76DEd1cEgWaJSB95Ah+C+D6KbIp7zjkwXUfF9HzUKiAJMjCin5hgCqLFJpul4Z1PLEZgf5vC6qKV+jfB75GKY+juFaTBlagIXCFsHzeJ0uBBlO8OD89Cwf10z7YMhPWnNjRvVJp3kwPlfmof5Yiw7xJxYtYe1SiI+hnqyOSUROhL+mUE4Xc7MHD/G2HOtkQJ908WLHl+OjiS5LecQuKFIJcGzM56N3hDmE6SHkBkknb9BFKdMJHk6XQ1tzKy/qS0QUE0XA5bypbNjIqCY2YTntZH0tXo0H8tVc0PAeVMQcc2KCh/j39u5V52Qam1h7Ufsi1Wfxcr6BQNjlof14VBerj4pi72edcTprm34RyYiizi1tbapPJHbkzaNM8AsMAAAAAAAEBm5gAAAAAAAgMHADAwAAAAAAgTHvDMysx9fqxTK6+E+dqH0Wi/MMTDLNi1YSEdmiIJBt6XsquW6wUubZj7q0Lo5mHSFD0igKt5mOJRO6aJxFb56TiUd1QR5HrPvOiz5y7SERUanEi8aZ1kbKopPT09OsLbMrRERNTU3q2Ovlcjl1TBa7tG29jh1gQZOFZw2FLJMNfPwb3n+AtT0xrhERUZUXxh0O68+0E+F5Py/CxxdHrAEnInJT/FpCrliDbCj66JV4Zifk6q+PcIiPDSFZ+NdQaE4ufldFcQ1F4+TIK8dlU5+3/ucAcKzM59PpiF4rVqxUfbq7e/jjivFE5qWJ9NxJDkmmrHBNzFNNw9aRXpOpILh8Jlksffq0aZJk4fZ4TBdCj8b494Ql8jhuSL8vci4r5++mF+iI+aLt6O8w+XcyX/gFBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMDADQwAAAAAAATGvEP8i9ZsYO2ao8NGI9O/Z203xMOjlqUDnNUqDz6ZCqi5IoAv4z6VcpmkiCi4ExZhUhlMJ9Ih/qghpFqp8CCr5/GwlCGfRLncLD8gil2agvU1ERKTATEiomSSv7+ysGWppAvNyaC/LPApQ1pEOrCWMISCTRsGACwUnsfHrUhMb+bR2cOLR44NDLJ2fowXAiYiskQovmQooJYXoc6OpUtZu3FRpzrHERuHeL7cTEWPw5aIk9qGor1REdIPiQ0G3JAeX8KOKOYmBvjavAL6COQDvJ/NJ+gtNzYyHTvawPjrmUL8nnfkzY3elnmSmNumDUWT5WYAcp5nIufIpvfpSO+daRMC07G3C2adAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABMa8MzAdPWtYu3fvTtWnWOaZi7DLcya5fF6dI5daxxO64E5YZFFCojCOKeshryWV4oXlYoY16skkf+6Qrdf71aoiZyLyN7LoJpEuPCRzJh0dHeqcsFi3OTQ0pPrI9YiyWGc+z/M6RET9/X2snZ2YYO36TEad07loMWt39SxVfeRzAyxkvmGtcMuSRay9Yv0JrN2743l1TnlmirXjcV30Vi4xLpf4GOVV9fpnV+RXfJEhCRlyJ7kcH79LhvHcFhdTFWPz+OAhdY4nnjtZzysiR+pMhYpRTBdgoTuaPMt8zjma3MbUFB+rTXkXmTORfUxZZ1Pe5vXm83rGx3W+Uj7ukiU8ozmfOZ165nnkZkzXe7Q5GfwCAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDMO7hgizXHk6NZ1adW4TVRalWe/SgX9JrpkKhHYMrA1GfqWVvWfZnJ8domRESFfI61Y3GeKTEtoa7V+FrxSkWvRyyI55b5G1MdlZh4TfEkz+NEo/o1p1UWRa8blOsabZuvaaxW+d8HEdHsFH+vhg7uZ+1DB/r0ObMzrN3YqOvWyLWdAAuJ/HTKWiZERBTjn5HOdStZ23P1vyf17XiOtUuzeqwLi3F1eN9e1s7lRR0qImpfsYq1U/W8Btb0FP/MExEd6tvH2pNjo6qPK2q6eGIMGtl/QJ0zPMjzfW3d3azdtW6tOicixtR3sNQAALwL3o76LO8UmdEwZVWeeeYZ1jbOBY9Qn7BsqGk4Njb2pueY8iOuy+tvZQzZZlkTUNZBTKd1fRlF5luOfMbbCr/AAAAAAABAYOAGBgAAAAAAAgM3MAAAAAAAEBi4gQEAAAAAgMCYd4jfr/EwZnYyq/oUReHEiMODRE4DD+MTEWXSvEhZrarDUTKAJINOliGQ77qOaPNrMRUZKonil8YCQiL9n0jx649FdYHMUIif44T4tYQcHYCvq+ObDnR2dqo+Mryl3hd1BpG9mBfTc3y+ccHu3b3qnNFhHrItFHSBzKMtRATwfqBC/IY+NdHJFxuLLF7NQ/1ERCEx7o707tbPXeaFK22xmcr4wYPqHEeMU1FRONcr8sck0uN5JB5XfUgU/42KITQZ0xuW1Cw+Fhdn+QYss9N6Q4FwXD8OAMA7YT7FGGWw3xTif/55Xaz49SJiHCYi2r+fb7TULTY5MRU5lwH81atXqz5FMcabNhB4r8MvMAAAAAAAEBi4gQEAAAAAgMDADQwAAAAAAATGvDMwcpV3zVDk0RZhlGqF5yvckF4T2CRyMaWCXnvthsRlenyFeUN9g75c8bhyzWLY1WsN5TGZmyHS2Rm5ztFx9D2hLCgp4yK+r99L+VdjKioksyi5HF87nhBFk4iI8nVJ1m5p5sWL6gzPU67wheymnA8KWcJC5s+jhJctqltaYiAIhfV409jRwZ+nqnN5ITG2+SHezouivkRErsXHF1+Mu5nWFnVOqq2dtZvzS1WfYoE/lyO+E2KG3Izl8GvJiXHNjuqxWmbu3stF8ADgveto8rumfHRcjG2mYpcyzy3ncKZz5NjW0qLHZkm+ptFRXXRYzinl9QcBfoEBAAAAAIDAwA0MAAAAAAAEBm5gAAAAAAAgMOadgbHFOryYIV/heXz9XnYqy9pTWd4mIvKX8LWEYddwSb5cb8jX99XV1ZHU2MSzHTKHEjfUEYiKbEdIZm9I51dsW6691muxLbEO3Bevx/P0GkzL4u+lKY/T1NQkHpc/jun9dsN8PXkiyTMxza18zT2R3h9cnvPa9WINOixkb/2///nUjomk+Gct3qjzflGxdjlRn2FtT9S3IiKaHhlm7dkcX4vtpPVa7HCSP0/EkNkJi3XV8jVZpvdJjB3JRPDWYgPAwmGqIyhrrZjyLCtXrnzTPjMzuubVvn37WPvkk09mbVMep1TiY77peuUxOf8NAvwCAwAAAAAAgYEbGAAAAAAACAzcwAAAAAAAQGDgBgYAAAAAAAJj/iF+UbDRFJyfnp5mba/Gg/OHDg2oc2RRRFOByWiUbxggC1s6ti6QGQrxwoqJZIp38I8ctjfd31kWj6VWRWE5HeonknFdXz23Kb575OJKMnRVV8df44zYRIGIyFEhfn5OLKaDXDWxAUKtWlF9ABa2I4f4dY/5FFATxS8NYUx5TI7VJNtEZFv8nBrxMKmptps6ZNi444ibeZhesiWb2BAE4P3kaIpFvpeZNnhqbW39gx+3UtFzqyVLlrB2KsXnbKaA/tG830ezEdOx/nvFLzAAAAAAABAYuIEBAAAAAIDAwA0MAAAAAAAExrwzMDJ30t7RrvrU19ez9uz0FGtPTkyqc16cfZ61m5v1OsKYyMA0iT7Vqi4Y1L9/P2u3NLewdloUXCOa7xrANy9caVoSaFnyPlEXHpLkukbTWkNZwEhmhdoMazJr4hxbZI5M68/lc1dNGZj32RpXAAAAgPl4OzIkpoLlpmPvxLUEEX6BAQAAAACAwMANDAAAAAAABAZuYAAAAAAAIDBwAwMAAAAAAIEx7xC/I0L8He0dqs+Gk09m7Ref287a+/v61DmVUom1qxUdyI9FeHHFxoZG1k4kk+qc/ESRtbPZLGubQvzzCeRL8wnbq2cRAStT4OroQlj8uWWhSyIiT1xfxedF7mo1wwYDlnxfTFXuEOIHAAAAgHcefoEBAAAAAIDAwA0MAAAAAAAEBm5gAAAAAAAgMCx/PqENAAAAAACA9wD8AgMAAAAAAIGBGxgAAAAAAAgM3MAAAAAAAEBg4AYGAAAAAAACAzcwAAAAAAAQGLiBAQAAAACAwMANDAAAAAAABAZuYAAAAAAAIDBwAwMAAAAAAIHx/wct51/gAO6dOAAAAABJRU5ErkJggg==\n"},"metadata":{}}],"source":["# Get images in batch of 20\n","aug_iterator = datagen.flow(train_data, train_label, batch_size=1)\n","\n","plt.figure(figsize=(12,4))\n","for i in range(6):\n"," ax = plt.subplot(2, 3, i + 1)\n"," plt.imshow(aug_iterator[i][0][0])\n"," plt.title(str(label_names[aug_iterator[i][1][0][0]]))\n"," plt.axis(\"off\")"]},{"cell_type":"markdown","metadata":{"id":"DctH3QgL6XMd"},"source":["When we use data augmentation, the arguments in the `fit` function are slightly different. First, we need to provide the name of our defined generator with `flow` to yield a batch of augmented images. In this case, we will need to use `datagen.flow`, since we assigned the image generator to the name `datagen`.\n","\n","Also, the generator in Keras does not accept a `validation_split` argument, and therefore we need to manually create a validation dataset, and then use it. This has been done in the next cells."]},{"cell_type":"code","execution_count":37,"metadata":{"executionInfo":{"elapsed":161,"status":"ok","timestamp":1731444887002,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"yBTx9rXUEFCI"},"outputs":[],"source":["# Define the model with inputs and outputs\n","cifar_cnn_4 = Model(inputs, outputs)\n","\n","# compile model\n","cifar_cnn_4.compile(optimizer='adam',\n"," loss='sparse_categorical_crossentropy',\n"," metrics=['accuracy'])"]},{"cell_type":"code","execution_count":38,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":373},"executionInfo":{"elapsed":704,"status":"error","timestamp":1731444904281,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"uFEyAsjcW4bw","outputId":"9af4cd71-7f6b-41fa-9117-1cac1b612dce"},"outputs":[{"output_type":"error","ename":"ValueError","evalue":"Argument `validation_split` is only supported for tensors or NumPy arrays.Found incompatible type in the input: []","traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mValueError\u001b[0m Traceback (most recent call last)","\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[0;31m# fit model\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m history = cifar_cnn_4.fit(datagen.flow(train_data, train_label, batch_size=128),\n\u001b[0m\u001b[1;32m 3\u001b[0m epochs=120, validation_split=0.2, verbose=0)\n\u001b[1;32m 4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0;31m# note that the generator does not work with validation split\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.10/dist-packages/keras/src/utils/traceback_utils.py\u001b[0m in \u001b[0;36merror_handler\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m 120\u001b[0m \u001b[0;31m# To get the full stack trace, call:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 121\u001b[0m \u001b[0;31m# `keras.config.disable_traceback_filtering()`\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 122\u001b[0;31m \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwith_traceback\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfiltered_tb\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 123\u001b[0m \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 124\u001b[0m \u001b[0;32mdel\u001b[0m \u001b[0mfiltered_tb\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/usr/local/lib/python3.10/dist-packages/keras/src/trainers/data_adapters/array_slicing.py\u001b[0m in \u001b[0;36mtrain_validation_split\u001b[0;34m(arrays, validation_split)\u001b[0m\n\u001b[1;32m 470\u001b[0m \u001b[0munsplitable\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mt\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mflat_arrays\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mcan_slice_array\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 471\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0munsplitable\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 472\u001b[0;31m raise ValueError(\n\u001b[0m\u001b[1;32m 473\u001b[0m \u001b[0;34m\"Argument `validation_split` is only supported \"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 474\u001b[0m \u001b[0;34m\"for tensors or NumPy arrays.\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mValueError\u001b[0m: Argument `validation_split` is only supported for tensors or NumPy arrays.Found incompatible type in the input: []"]}],"source":["# fit model\n","history = cifar_cnn_4.fit(datagen.flow(train_data, train_label, batch_size=128),\n"," epochs=120, validation_split=0.2, verbose=0)\n","\n","# note that the generator does not work with validation split"]},{"cell_type":"code","execution_count":39,"metadata":{"executionInfo":{"elapsed":1055,"status":"ok","timestamp":1731444907830,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"U2S2jV_aXIV9"},"outputs":[],"source":["from sklearn.model_selection import train_test_split\n","\n","train_data_1, validation_data_1, train_label_1, validation_label_1 = train_test_split(train_data, train_label, test_size=0.2, random_state=20, stratify=train_label)"]},{"cell_type":"code","execution_count":40,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":3,"status":"ok","timestamp":1731444908237,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"-BTZ3qaXXnq6","outputId":"d5942beb-1d7f-4ba6-a4af-94b332d522d7"},"outputs":[{"output_type":"stream","name":"stdout","text":["Training images (40000, 32, 32, 3)\n","Training labels (40000, 1)\n","Validation_images (10000, 32, 32, 3)\n","Validation labels (10000, 1)\n"]}],"source":["print('Training images', train_data_1.shape)\n","print('Training labels', train_label_1.shape)\n","print('Validation_images', validation_data_1.shape)\n","print('Validation labels', validation_label_1.shape)"]},{"cell_type":"code","execution_count":41,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"3JBjOhQcFsIe","executionInfo":{"status":"ok","timestamp":1731448213650,"user_tz":420,"elapsed":3299972,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"31529719-af2c-4c09-c9ea-5dc215019d72"},"outputs":[{"output_type":"stream","name":"stderr","text":["/usr/local/lib/python3.10/dist-packages/keras/src/trainers/data_adapters/py_dataset_adapter.py:121: UserWarning: Your `PyDataset` class should call `super().__init__(**kwargs)` in its constructor. `**kwargs` can include `workers`, `use_multiprocessing`, `max_queue_size`. Do not pass these arguments to `fit()`, as they will be ignored.\n"," self._warn_if_super_not_called()\n"]},{"output_type":"stream","name":"stdout","text":["Training time: 0:54:59.929572\n"]}],"source":["t = now()\n","# fit model\n","history= cifar_cnn_4.fit(datagen.flow(train_data_1, train_label_1, batch_size=128),\n"," epochs=100, validation_data=(validation_data_1, validation_label_1), verbose=0)\n","\n","print('Training time: %s' % (now() - t))"]},{"cell_type":"code","execution_count":42,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":410},"id":"YSuGN-gMLjVG","executionInfo":{"status":"ok","timestamp":1731448214227,"user_tz":420,"elapsed":580,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"194ad47d-1a75-4e47-dbb9-02af747436ca"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# plot the accuracy and loss\n","train_loss = history.history['loss']\n","val_loss = history.history['val_loss']\n","acc = history.history['accuracy']\n","val_acc = history.history['val_accuracy']\n","\n","epochsn = np.arange(1, len(train_loss)+1,1)\n","plt.figure(figsize=(12, 4))\n","\n","plt.subplot(1,2,1)\n","plt.plot(epochsn, acc, 'b', label='Training Accuracy')\n","plt.plot(epochsn, val_acc, 'r', label='Validation Accuracy')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('ACCURACY')\n","plt.xlabel('Epochs')\n","plt.ylabel('Accuracy')\n","\n","plt.subplot(1,2,2)\n","plt.plot(epochsn,train_loss, 'b', label='Training Loss')\n","plt.plot(epochsn,val_loss, 'r', label='Validation Loss')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('LOSS')\n","plt.xlabel('Epochs')\n","plt.ylabel('Loss')\n","plt.show()"]},{"cell_type":"code","execution_count":43,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"RhiFueRJISfR","executionInfo":{"status":"ok","timestamp":1731448215905,"user_tz":420,"elapsed":1683,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"e04c22ab-e0df-47ec-e096-af3dfdbbcceb"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.8109 - loss: 0.5873\n","Classification Accuracy: 0.8141000270843506\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn_4.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"fzfn60wHE9wC"},"source":["The data augmentation was helpful to further improve the performance of the model to about 81% accuracy. We trained the model for 100 epochs, which took about 50 minutes. From the learning curves we can conclude that the accuracy was still increasing at epoch 100, therefore we could continue training the model, or we could have selected a higher number of epochs initially."]},{"cell_type":"markdown","metadata":{"id":"CPmTsXovhNQq"},"source":["## 16.8 Transfer Learning "]},{"cell_type":"markdown","metadata":{"id":"k29r7L-LAeAG"},"source":["**Transfer learning** uses pretrained models that are trained on very large datasets to improve the performance on smaller datasets.\n","\n","There are many pretrained models available in Keras Applications.\n","\n","Typical steps in transfer learning include:\n","\n","- Initialize the pretrained model (base model) and weights, and remove the top layers (fully-connected layers) of the pretrained model.\n","- Create a new model by adding new trainable fully-connected layers on top of the primary model.\n","- Train the new model, and evaluate the performance.\n","\n","In this case we will use the VGG16 model, which is popular for image classification. The model is imported with the weights pretrained on ImageNet, which is a very large dataset with 1.2 million images in 1,000 classes. Since our task has only 10 classes, we will remove the dense layers which are very specific to ImageNet's 1,000 classes, and add our own custom layers with 10 classes.\n","\n","Transfer learning is also often referred to as **model fine-tuning**, because we start with a pretrained model and we fine-tune its parameters to our custom dataset.\n"]},{"cell_type":"code","execution_count":53,"metadata":{"id":"CiKULXs8e4RB","executionInfo":{"status":"ok","timestamp":1731449514923,"user_tz":420,"elapsed":385,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}}},"outputs":[],"source":["from tensorflow.keras.applications import vgg16\n","from keras.layers import GlobalAveragePooling2D\n","\n","base_model = vgg16.VGG16(weights='imagenet', include_top=False, input_shape=(32,32,3))\n","\n","# Add a global spatial average pooling layer\n","x = base_model.output\n","x = GlobalAveragePooling2D()(x)\n","\n","x = Dense(128, activation='relu')(x)\n","x = Dropout(0.2)(x)\n","x = Dense(64, activation='relu')(x)\n","x = Dropout(0.2)(x)\n","predictions = Dense(10, activation='softmax')(x)\n","\n","# The model we will train\n","cifar_cnn_5 = Model(inputs=base_model.input, outputs=predictions)"]},{"cell_type":"markdown","metadata":{"id":"dJ1NMvpmPXKs"},"source":["### Early Stopping"]},{"cell_type":"markdown","metadata":{"id":"Cnd2oWhIOjdO"},"source":["Keras has implemented several callbacks that allow to monitor the training process and have more control over it. We will study the various callbacks in the next lecture, and in this lecture we we explain how the callback Early Stopping works.\n","\n","**EarlyStopping** monitors a metric (e.g., validation loss) and if the metric doesn't improve for a certain number of epochs (a.k.a. patience), terminates the training. By improve, we mean the decrease if the metric is loss, or increase if the metric is accuracy."]},{"cell_type":"code","execution_count":45,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"NwEj4sNZ-RKK","executionInfo":{"status":"ok","timestamp":1731448957234,"user_tz":420,"elapsed":453949,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"4e69ad1d-9eb3-4ac1-995d-f77399dac4ec"},"outputs":[{"output_type":"stream","name":"stdout","text":["Training time: 0:07:33.763617\n"]}],"source":["from keras.callbacks import EarlyStopping\n","\n","cifar_cnn_5.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics = ['accuracy'])\n","\n","# fit model\n","t = now()\n","history = cifar_cnn_5.fit(train_data, train_label, validation_split=0.2, batch_size=128,\n"," epochs=100, verbose=0, callbacks=[EarlyStopping(monitor='val_loss', patience = 10)])\n","\n","print('Training time: %s' % (now() - t))"]},{"cell_type":"code","execution_count":46,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":410},"id":"a01FmtI2t1QO","executionInfo":{"status":"ok","timestamp":1731448958180,"user_tz":420,"elapsed":949,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"ad17be99-462b-486b-e84a-e1ac7242307d"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# plot the accuracy and loss\n","train_loss = history.history['loss']\n","val_loss = history.history['val_loss']\n","acc = history.history['accuracy']\n","val_acc = history.history['val_accuracy']\n","\n","epochsn = np.arange(1, len(train_loss)+1,1)\n","plt.figure(figsize=(12, 4))\n","\n","plt.subplot(1,2,1)\n","plt.plot(epochsn, acc, 'b', label='Training Accuracy')\n","plt.plot(epochsn, val_acc, 'r', label='Validation Accuracy')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('ACCURACY')\n","plt.xlabel('Epochs')\n","plt.ylabel('Accuracy')\n","\n","plt.subplot(1,2,2)\n","plt.plot(epochsn,train_loss, 'b', label='Training Loss')\n","plt.plot(epochsn,val_loss, 'r', label='Validation Loss')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('LOSS')\n","plt.xlabel('Epochs')\n","plt.ylabel('Loss')\n","plt.show()"]},{"cell_type":"code","execution_count":47,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"MVq2V_pItfXA","executionInfo":{"status":"ok","timestamp":1731448963559,"user_tz":420,"elapsed":5382,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"bba02b5a-d9d4-43df-e6d3-f4ccdf0176af"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 7ms/step - accuracy: 0.7904 - loss: 0.8536\n","Classification Accuracy: 0.7901999950408936\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn_5.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"PawwVj2rjZsK"},"source":["Although the accuracy is reduced, note that the model was trained in about 15 epochs, in comparison to 100 epochs or more epochs when trained from scratch.\n","\n","The following figure illustrates the early stopping callback. When the validation loss starts increasing, it terminates the training. In our case, the validation loss reached minimum at epoch 6, and since we specified a 10 epochs patience, the model stopped at epoch 16.\n","\n","\n","\n","*Figure: Early stopping.*"]},{"cell_type":"markdown","metadata":{"id":"_uad8ljQEp07"},"source":["Let's try to use a pretrained model and combine it with data augmentation and early stopping."]},{"cell_type":"code","execution_count":54,"metadata":{"id":"c4Wa4EWvRVkQ","executionInfo":{"status":"ok","timestamp":1731449521374,"user_tz":420,"elapsed":175,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}}},"outputs":[],"source":["# The model we will train\n","cifar_cnn_6 = Model(inputs=base_model.input, outputs=predictions)"]},{"cell_type":"code","execution_count":55,"metadata":{"id":"TQ0Fs2KtRgCp","executionInfo":{"status":"ok","timestamp":1731449522629,"user_tz":420,"elapsed":189,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}}},"outputs":[],"source":["# compile model\n","cifar_cnn_6.compile(optimizer='adam',\n"," loss='sparse_categorical_crossentropy',\n"," metrics=['accuracy'])"]},{"cell_type":"code","execution_count":56,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"OhYu6897Axla","executionInfo":{"status":"ok","timestamp":1731451054881,"user_tz":420,"elapsed":1530965,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"}},"outputId":"24d37b2f-a268-48b1-b910-b5e2fa17afb8"},"outputs":[{"output_type":"stream","name":"stdout","text":["Training time: 0:25:30.829182\n"]}],"source":["t = now()\n","# fit model\n","history= cifar_cnn_6.fit(datagen.flow(train_data_1, train_label_1, batch_size=128),\n"," epochs=100, validation_data=(validation_data_1, validation_label_1), verbose=0,\n"," callbacks=[EarlyStopping(monitor='val_loss', patience = 10)])\n","\n","print('Training time: %s' % (now() - t))"]},{"cell_type":"code","execution_count":57,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":410},"executionInfo":{"elapsed":358,"status":"ok","timestamp":1731451055236,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"OnrPyR1atOWL","outputId":"bfe25026-1fc9-4bc8-f437-6d0e65c4aea4"},"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{}}],"source":["# plot the accuracy and loss\n","train_loss = history.history['loss']\n","val_loss = history.history['val_loss']\n","acc = history.history['accuracy']\n","val_acc = history.history['val_accuracy']\n","\n","epochsn = np.arange(1, len(train_loss)+1,1)\n","plt.figure(figsize=(12, 4))\n","\n","plt.subplot(1,2,1)\n","plt.plot(epochsn, acc, 'b', label='Training Accuracy')\n","plt.plot(epochsn, val_acc, 'r', label='Validation Accuracy')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('ACCURACY')\n","plt.xlabel('Epochs')\n","plt.ylabel('Accuracy')\n","\n","plt.subplot(1,2,2)\n","plt.plot(epochsn,train_loss, 'b', label='Training Loss')\n","plt.plot(epochsn,val_loss, 'r', label='Validation Loss')\n","plt.grid(color='gray', linestyle='--')\n","plt.legend()\n","plt.title('LOSS')\n","plt.xlabel('Epochs')\n","plt.ylabel('Loss')\n","plt.show()"]},{"cell_type":"code","execution_count":58,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":3130,"status":"ok","timestamp":1731451058361,"user":{"displayName":"Aleksandar Vakanski","userId":"07675307153279708378"},"user_tz":420},"id":"ThYRPaEWWGQw","outputId":"d314aafe-8a1d-4af7-e209-a74196528cb7"},"outputs":[{"output_type":"stream","name":"stdout","text":["\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m2s\u001b[0m 7ms/step - accuracy: 0.8457 - loss: 0.5439\n","Classification Accuracy: 0.8471999764442444\n"]}],"source":["# Evaluate on test dataset\n","evals_test = cifar_cnn_6.evaluate(test_data, test_label)\n","print(\"Classification Accuracy: \", evals_test[1])"]},{"cell_type":"markdown","metadata":{"id":"48ygKtOVWS0Z"},"source":["Combining transfer learning with data augmentation increased the accuracy to about 84%."]},{"cell_type":"markdown","metadata":{"id":"vweobvFVe4RB"},"source":["## References \n","\n","1. Complete Machine Learning Package, Jean de Dieu Nyandwi, available at: [https://github.com/Nyandwi/machine_learning_complete](https://github.com/Nyandwi/machine_learning_complete).\n","2. How to Develop a CNN from Scratch for CIFAR-10 Photo Classification, Jason Brownlee, available at: [https://machinelearningmastery.com/how-to-develop-a-cnn-from-scratch-for-cifar-10-photo-classification/](https://machinelearningmastery.com/how-to-develop-a-cnn-from-scratch-for-cifar-10-photo-classification/).\n","3. Python Machine Learning (2nd Ed.) Code Repository, Sebastian Raschka, available at: [https://github.com/rasbt/python-machine-learning-book-2nd-edition](https://github.com/rasbt/python-machine-learning-book-2nd-edition).\n"]},{"cell_type":"markdown","metadata":{"id":"bDbzcXmWe4RB"},"source":["[BACK TO TOP](#top)"]}],"metadata":{"accelerator":"GPU","colab":{"provenance":[{"file_id":"1zniWJ7YmbvOhFjB9LWDHtZG0V7nQOn6B","timestamp":1665618878004}]},"kernelspec":{"display_name":"Python 3 (ipykernel)","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.9.13"}},"nbformat":4,"nbformat_minor":0}