CNN_final.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "**Mounting drive** connecting to have access to our database"
      ],
      "metadata": {
        "id": "xke1j7pDcn2Z"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sUvH9NOVcWi9",
        "outputId": "128cd1c8-a623-466c-dea9-78bfee9efbfa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Required for the code"
      ],
      "metadata": {
        "id": "SCYEwkg9cvV2"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "CxfiFw3Mav0S"
      },
      "outputs": [],
      "source": [
        "#!pip install light-the-torch >> /.tmp\n",
        "#!ltt install torch torchvision >> /.tmp\n",
        "#!pip install fastai --upgrade >> /.tmp"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Uncomment the following lines if GPU doesn't seem to be working, may be your version of tensorflow is not prepared to work with GPU"
      ],
      "metadata": {
        "id": "JKKGo-Tqc2g4"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OlrwiSskm3zB"
      },
      "outputs": [],
      "source": [
        "#!chronic pip install tensorflow-gpu==2.0.0\n",
        "#!pip uninstall tensorflow"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#testing if there is a GPU or not. if you don't get \"Found GPU at: /device:GPU:0\" then you have to activate your GPU before starting the work \n",
        "# to activate the gpu : select \"Runtime\" in the google colab's toolbar then \"change runtime type\" then choose \"GPU\" and save changes\n",
        "%tensorflow_version 2.x\n",
        "import tensorflow as tf\n",
        "device_name = tf.test.gpu_device_name()\n",
        "if device_name != '/device:GPU:0':\n",
        "  raise SystemError('GPU device not found')\n",
        "print('Found GPU at: {}'.format(device_name))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xmBLvsuJQZsT",
        "outputId": "c07a38b4-0702-4d72-f36f-582548d03c00"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found GPU at: /device:GPU:0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BmCxS6QCP_mJ"
      },
      "source": [
        "**Import the required libraries**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "L9ko7vbOPool"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import tensorflow as tf\n",
        "import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Conv2D , MaxPool2D , Flatten , Dropout \n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from sklearn.metrics import classification_report,confusion_matrix\n",
        "import cv2\n",
        "import os\n",
        "import numpy as np"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dZVHGLC-RQ1_"
      },
      "source": [
        "**Loading the data** "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LJ1dhPW6QSsU"
      },
      "outputs": [],
      "source": [
        "labels = ['NORMAL', 'PNEUMONIA'] #Labeling data\n",
        "#Without having a fixed input shape, you cannot define architecture of your model. It is therefore necessary to convert all your images to same size.\n",
        "img_size =224\n",
        "def get_data(data_dir):\n",
        "    data = [] \n",
        "    for label in labels: \n",
        "        path = os.path.join(data_dir, label)\n",
        "        class_num = labels.index(label)\n",
        "        for img in os.listdir(path):\n",
        "            try:\n",
        "                img_arr = cv2.imread(os.path.join(path, img))[...,::-1] #convert BGR to RGB format\n",
        "                resized_arr = cv2.resize(img_arr, (img_size, img_size)) # Reshaping images to preferred size ( the '-1' reverse the order of colors channels )\n",
        "                data.append([resized_arr, class_num])\n",
        "            except Exception as e:\n",
        "                print(e)\n",
        "    return np.array(data) # numpy makes it easier to manipulate image data and takes less amount of memory compared to python's lists"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Dg9onZM9Rbn3"
      },
      "source": [
        "\n",
        "\n",
        "> *specifiying data path*\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-xg_2IGkRO_q",
        "outputId": "bb008ba6-abd5-4379-8d70-cd4eb3a6901c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "'NoneType' object is not subscriptable\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:15: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
            "  from ipykernel import kernelapp as app\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "'NoneType' object is not subscriptable\n",
            "'NoneType' object is not subscriptable\n"
          ]
        }
      ],
      "source": [
        "with tf.device('/device:GPU:0'):# ensure you are working with your GPU \n",
        "#loading the validation and training data \n",
        "  train = get_data('/content/drive/MyDrive/Kaggle/pneumonia_data/chest_xray/chest_xray/train')\n",
        "  val = get_data('/content/drive/MyDrive/Kaggle/pneumonia_data/chest_xray/chest_xray/val')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9we5l37VSDnG"
      },
      "source": [
        "**Visualizing Data**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 337
        },
        "id": "YDcdawBuSOdO",
        "outputId": "a66375cf-dcd3-4afc-8a43-95c94792287c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/seaborn/_decorators.py:43: FutureWarning: Pass the following variable as a keyword arg: x. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
            "  FutureWarning\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fc150362c50>"
            ]
          },
          "metadata": {},
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "l = []\n",
        "for i in train:\n",
        "    if(i[1] == 0):\n",
        "        l.append(\"NORMAL\")\n",
        "    else:\n",
        "        l.append(\"PNEUMONIA\")\n",
        "sns.set_style('darkgrid')\n",
        "sns.countplot(l)\n",
        "# visualizing data is an important step: although it's preferred to have equal size of data for each class , it's understandable to have more images \n",
        "# for different forms of this anomaly "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        },
        "id": "VYkxD3aRVuC4",
        "outputId": "1db23a20-c7ca-4a8f-8caf-d166ab055f1a"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'NORMAL')"
            ]
          },
          "metadata": {},
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "#example of the data ( ensure that images are quite visible and detectable, even if they are smaller then first, to make the model work effeciently )\n",
        "plt.figure(figsize = (5,5))\n",
        "plt.imshow(train[1][0])\n",
        "plt.title(labels[train[0][1]])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        },
        "id": "JOBJAexxWlGs",
        "outputId": "4ae16b21-f497-4366-9327-e1f6b5f2a943"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'PNEUMONIA')"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "#example2\n",
        "plt.figure(figsize = (5,5))\n",
        "plt.imshow(train[-1][0])\n",
        "plt.title(labels[train[-1][1]])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PgiOpQomWvEz"
      },
      "source": [
        "**Data preprocessing**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5yPw9M5sWoug"
      },
      "outputs": [],
      "source": [
        "#x and y represents two lists of each dataset ( training or validation , x_train/x_val is a list of images and y_train/y_val is a list of the corresponding label)\n",
        "x_train = []\n",
        "y_train = []\n",
        "x_val = []\n",
        "y_val = []\n",
        "\n",
        "for feature, label in train:\n",
        "  x_train.append(feature)\n",
        "  y_train.append(label)\n",
        "\n",
        "for feature, label in val:\n",
        "  x_val.append(feature)\n",
        "  y_val.append(label)\n",
        "\n",
        "# Normalize the data : all pixel values (between 0 and 255) will be values from 0 to 1 to make the work easier\n",
        "x_train = np.array(x_train) / 255 \n",
        "x_val = np.array(x_val) / 255\n",
        "#changing the order of parameters to reverse the order of channels colors -> working with numpy arrays \n",
        "x_train.reshape(-1, img_size, img_size, 1)\n",
        "y_train = np.array(y_train)\n",
        "\n",
        "x_val.reshape(-1, img_size, img_size, 1)\n",
        "y_val = np.array(y_val)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0hbXcUKeZMvk"
      },
      "source": [
        "**Data augmentation** \n",
        "\n",
        "> Data augmentation is a set of techniques to artificially increase the amount of data by generating new data points from existing data. This includes making small changes to data or using deep learning models to generate new data points.\n",
        "\n",
        "> for images's datasets it can be done by zooming images/ increasing brightness/ rotating the image in different directions / whitening \n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PjEL5VwqZR7T"
      },
      "outputs": [],
      "source": [
        "def gpu_datagen():\n",
        "  with tf.device('/device:GPU:0'):\n",
        "    datagen = tf.keras.preprocessing.image.ImageDataGenerator(\n",
        "            featurewise_center=False,  # set input mean to 0 over the dataset\n",
        "            samplewise_center=False,  # set each sample mean to 0\n",
        "            featurewise_std_normalization=False,  # divide inputs by std of the dataset\n",
        "            samplewise_std_normalization=False,  # divide each input by its std\n",
        "            zca_whitening=False,  # apply ZCA whitening\n",
        "            rotation_range = 30,  # randomly rotate images in the range (degrees, 0 to 180)\n",
        "            zoom_range = 0.2, # Randomly zoom image \n",
        "            width_shift_range=0.1,  # randomly shift images horizontally (fraction of total width)\n",
        "            height_shift_range=0.1,  # randomly shift images vertically (fraction of total height)\n",
        "            horizontal_flip = True,  # randomly flip images\n",
        "            vertical_flip=False)  # randomly flip images\n",
        "    return datagen.fit(x_train)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "gpu_datagen()"
      ],
      "metadata": {
        "id": "7WFVXHH2T4hl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xXtPfRr8cf4S"
      },
      "source": [
        "***Define the Model***\n",
        "\n",
        "Let’s define a simple CNN model with 3 Convolutional layers followed by max-pooling layers. A dropout layer is added after the 3rd maxpool operation to avoid overfitting.\n",
        "---\n",
        "**Overfitting** is \"the production of an analysis that corresponds too closely or exactly to a particular set of data, and may therefore fail to fit to additional data or predict future observations reliably\".\n",
        "\n",
        "**MaxPooling for 2D spatial data** Downsamples the input along its spatial dimensions (height and width) by taking the maximum value over an input window (of size defined by pool_size ) for each channel of the input.\n",
        "\n",
        "**Dropout layer** The Dropout layer is a mask that nullifies the contribution of some neurons towards the next layer and leaves unmodified all others.Usually, dropout is placed on the fully connected layers only because they are the one with the greater number of parameters and thus they're likely to excessively co-adapting themselves causing overfitting.\n",
        "\n",
        "**Conv2D layer** creates a convolution kernel that is convolved with the layer input to produce a tensor of outputs.\n",
        "\n",
        "**Flatten layer**  A flatten layer collapses the spatial dimensions of the input into the channel dimension. For example, if the input to the layer is an H-by-W-by-C-by-N-by-S array (sequences of images), then the flattened output is an (H*W*C)-by-N-by-S array -> you want to make it linear to pass it onto a Dense layer\n",
        "\n",
        "**Dense layer** The dense layer is a neural network layer that is connected deeply, which means each neuron in the dense layer receives input from all neurons of its previous layer. The dense layer is found to be the most commonly used layer in the models. visit : https://machinelearningknowledge.ai/keras-dense-layer-explained-for-beginners/"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PTDjwwSMcor6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "69d14e09-15ed-4cb9-938d-cb47d66aae80"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 224, 224, 32)      896       \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2D  (None, 112, 112, 32)     0         \n",
            " )                                                               \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 112, 112, 32)      9248      \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPooling  (None, 56, 56, 32)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 56, 56, 64)        18496     \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPooling  (None, 28, 28, 64)       0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 28, 28, 64)        0         \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 50176)             0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 128)               6422656   \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 2)                 258       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 6,451,554\n",
            "Trainable params: 6,451,554\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(32,3,padding=\"same\", activation=\"relu\", input_shape=(224,224,3)))\n",
        "model.add(MaxPool2D())\n",
        "\n",
        "model.add(Conv2D(32, 3, padding=\"same\", activation=\"relu\"))\n",
        "model.add(MaxPool2D())\n",
        "\n",
        "model.add(Conv2D(64, 3, padding=\"same\", activation=\"relu\"))\n",
        "model.add(MaxPool2D())\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(128,activation=\"relu\"))\n",
        "model.add(Dense(2, activation=\"softmax\"))\n",
        "\n",
        "model.summary()\n",
        "# after training the model you can add layers or change their hyperparameters to make your predictions better if you are not satisfied with the accuracy"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Optimization function : Gradient Descendent"
      ],
      "metadata": {
        "id": "HLTSoY0SGEyg"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cn6A5HkZcweg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "026b12db-cf15-4889-f83c-502c7c9e0768"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/optimizer_v2/adam.py:105: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  super(Adam, self).__init__(name, **kwargs)\n"
          ]
        }
      ],
      "source": [
        "#we will be using the gradient descendent function (Adam in python) with a learning rate of lr=0.000001\n",
        "opt = Adam(lr=0.000001)\n",
        "model.compile(optimizer = opt , loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True) , metrics = ['accuracy'])"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**fitting the model : training**"
      ],
      "metadata": {
        "id": "l8YATt-MGLHl"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wZCJAYyMc6Th",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c65283e1-1c15-44a6-8ed0-b8aa2ae9f21b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/500\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/dispatch.py:1082: UserWarning: \"`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a sigmoid or softmax activation and thus does not represent logits. Was this intended?\"\n",
            "  return dispatch_target(*args, **kwargs)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "110/110 [==============================] - 18s 50ms/step - loss: 0.6662 - accuracy: 0.6186 - val_loss: 0.7041 - val_accuracy: 0.5000\n",
            "Epoch 2/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.6498 - accuracy: 0.6186 - val_loss: 0.6938 - val_accuracy: 0.5000\n",
            "Epoch 3/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.6327 - accuracy: 0.6197 - val_loss: 0.6843 - val_accuracy: 0.5000\n",
            "Epoch 4/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.6056 - accuracy: 0.6368 - val_loss: 0.6577 - val_accuracy: 0.5625\n",
            "Epoch 5/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.5812 - accuracy: 0.6968 - val_loss: 0.6455 - val_accuracy: 0.5625\n",
            "Epoch 6/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.5503 - accuracy: 0.7412 - val_loss: 0.6171 - val_accuracy: 0.6250\n",
            "Epoch 7/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.5194 - accuracy: 0.8083 - val_loss: 0.6191 - val_accuracy: 0.6250\n",
            "Epoch 8/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.4869 - accuracy: 0.8316 - val_loss: 0.6251 - val_accuracy: 0.6250\n",
            "Epoch 9/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.4601 - accuracy: 0.8572 - val_loss: 0.6083 - val_accuracy: 0.6250\n",
            "Epoch 10/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.4289 - accuracy: 0.8737 - val_loss: 0.5660 - val_accuracy: 0.6250\n",
            "Epoch 11/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.4075 - accuracy: 0.8783 - val_loss: 0.5744 - val_accuracy: 0.6250\n",
            "Epoch 12/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.3822 - accuracy: 0.8877 - val_loss: 0.5947 - val_accuracy: 0.6250\n",
            "Epoch 13/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.3649 - accuracy: 0.8905 - val_loss: 0.5449 - val_accuracy: 0.6875\n",
            "Epoch 14/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.3467 - accuracy: 0.8985 - val_loss: 0.5585 - val_accuracy: 0.6250\n",
            "Epoch 15/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.3301 - accuracy: 0.8996 - val_loss: 0.5672 - val_accuracy: 0.6250\n",
            "Epoch 16/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.3178 - accuracy: 0.9044 - val_loss: 0.5734 - val_accuracy: 0.6250\n",
            "Epoch 17/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.3042 - accuracy: 0.9096 - val_loss: 0.5458 - val_accuracy: 0.6875\n",
            "Epoch 18/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2948 - accuracy: 0.9081 - val_loss: 0.5532 - val_accuracy: 0.6875\n",
            "Epoch 19/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.2855 - accuracy: 0.9124 - val_loss: 0.5589 - val_accuracy: 0.6250\n",
            "Epoch 20/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2749 - accuracy: 0.9201 - val_loss: 0.5719 - val_accuracy: 0.6250\n",
            "Epoch 21/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2670 - accuracy: 0.9215 - val_loss: 0.5597 - val_accuracy: 0.6250\n",
            "Epoch 22/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2607 - accuracy: 0.9187 - val_loss: 0.5721 - val_accuracy: 0.6250\n",
            "Epoch 23/500\n",
            "110/110 [==============================] - 5s 46ms/step - loss: 0.2510 - accuracy: 0.9221 - val_loss: 0.5216 - val_accuracy: 0.6875\n",
            "Epoch 24/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2450 - accuracy: 0.9229 - val_loss: 0.5538 - val_accuracy: 0.6875\n",
            "Epoch 25/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2419 - accuracy: 0.9189 - val_loss: 0.5429 - val_accuracy: 0.6875\n",
            "Epoch 26/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2340 - accuracy: 0.9275 - val_loss: 0.5411 - val_accuracy: 0.6875\n",
            "Epoch 27/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2288 - accuracy: 0.9275 - val_loss: 0.5239 - val_accuracy: 0.6875\n",
            "Epoch 28/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.2233 - accuracy: 0.9272 - val_loss: 0.5184 - val_accuracy: 0.6875\n",
            "Epoch 29/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2173 - accuracy: 0.9320 - val_loss: 0.5464 - val_accuracy: 0.6875\n",
            "Epoch 30/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.2132 - accuracy: 0.9283 - val_loss: 0.5724 - val_accuracy: 0.6250\n",
            "Epoch 31/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2096 - accuracy: 0.9332 - val_loss: 0.5457 - val_accuracy: 0.6875\n",
            "Epoch 32/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2026 - accuracy: 0.9337 - val_loss: 0.5323 - val_accuracy: 0.6875\n",
            "Epoch 33/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.2016 - accuracy: 0.9366 - val_loss: 0.4890 - val_accuracy: 0.6875\n",
            "Epoch 34/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1979 - accuracy: 0.9354 - val_loss: 0.4754 - val_accuracy: 0.6875\n",
            "Epoch 35/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1934 - accuracy: 0.9369 - val_loss: 0.4970 - val_accuracy: 0.6875\n",
            "Epoch 36/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1905 - accuracy: 0.9391 - val_loss: 0.4869 - val_accuracy: 0.6875\n",
            "Epoch 37/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1867 - accuracy: 0.9377 - val_loss: 0.5473 - val_accuracy: 0.6875\n",
            "Epoch 38/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1840 - accuracy: 0.9406 - val_loss: 0.4705 - val_accuracy: 0.6875\n",
            "Epoch 39/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1777 - accuracy: 0.9437 - val_loss: 0.5263 - val_accuracy: 0.6875\n",
            "Epoch 40/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1765 - accuracy: 0.9403 - val_loss: 0.5024 - val_accuracy: 0.6875\n",
            "Epoch 41/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1700 - accuracy: 0.9454 - val_loss: 0.4381 - val_accuracy: 0.6875\n",
            "Epoch 42/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1729 - accuracy: 0.9440 - val_loss: 0.4963 - val_accuracy: 0.6875\n",
            "Epoch 43/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1698 - accuracy: 0.9434 - val_loss: 0.4501 - val_accuracy: 0.6875\n",
            "Epoch 44/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1660 - accuracy: 0.9454 - val_loss: 0.5006 - val_accuracy: 0.6875\n",
            "Epoch 45/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1628 - accuracy: 0.9457 - val_loss: 0.4365 - val_accuracy: 0.6875\n",
            "Epoch 46/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1630 - accuracy: 0.9488 - val_loss: 0.4970 - val_accuracy: 0.6875\n",
            "Epoch 47/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1574 - accuracy: 0.9514 - val_loss: 0.4227 - val_accuracy: 0.6875\n",
            "Epoch 48/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1562 - accuracy: 0.9514 - val_loss: 0.4521 - val_accuracy: 0.6875\n",
            "Epoch 49/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1541 - accuracy: 0.9497 - val_loss: 0.4524 - val_accuracy: 0.6875\n",
            "Epoch 50/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1542 - accuracy: 0.9488 - val_loss: 0.4199 - val_accuracy: 0.6875\n",
            "Epoch 51/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1506 - accuracy: 0.9494 - val_loss: 0.4255 - val_accuracy: 0.6875\n",
            "Epoch 52/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1475 - accuracy: 0.9528 - val_loss: 0.4262 - val_accuracy: 0.6875\n",
            "Epoch 53/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1468 - accuracy: 0.9522 - val_loss: 0.4458 - val_accuracy: 0.6875\n",
            "Epoch 54/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1432 - accuracy: 0.9516 - val_loss: 0.4600 - val_accuracy: 0.6875\n",
            "Epoch 55/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1432 - accuracy: 0.9516 - val_loss: 0.4260 - val_accuracy: 0.6875\n",
            "Epoch 56/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1415 - accuracy: 0.9551 - val_loss: 0.4598 - val_accuracy: 0.6875\n",
            "Epoch 57/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1412 - accuracy: 0.9553 - val_loss: 0.4684 - val_accuracy: 0.6875\n",
            "Epoch 58/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1385 - accuracy: 0.9551 - val_loss: 0.4447 - val_accuracy: 0.6875\n",
            "Epoch 59/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1350 - accuracy: 0.9568 - val_loss: 0.4251 - val_accuracy: 0.6875\n",
            "Epoch 60/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1366 - accuracy: 0.9551 - val_loss: 0.4229 - val_accuracy: 0.6875\n",
            "Epoch 61/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1346 - accuracy: 0.9568 - val_loss: 0.3978 - val_accuracy: 0.6875\n",
            "Epoch 62/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1318 - accuracy: 0.9568 - val_loss: 0.4338 - val_accuracy: 0.6875\n",
            "Epoch 63/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1298 - accuracy: 0.9556 - val_loss: 0.4313 - val_accuracy: 0.6875\n",
            "Epoch 64/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1290 - accuracy: 0.9571 - val_loss: 0.4063 - val_accuracy: 0.6875\n",
            "Epoch 65/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1271 - accuracy: 0.9593 - val_loss: 0.4189 - val_accuracy: 0.6875\n",
            "Epoch 66/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1263 - accuracy: 0.9579 - val_loss: 0.4082 - val_accuracy: 0.6875\n",
            "Epoch 67/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1272 - accuracy: 0.9556 - val_loss: 0.3966 - val_accuracy: 0.6875\n",
            "Epoch 68/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1240 - accuracy: 0.9588 - val_loss: 0.4077 - val_accuracy: 0.6875\n",
            "Epoch 69/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1228 - accuracy: 0.9593 - val_loss: 0.3974 - val_accuracy: 0.6875\n",
            "Epoch 70/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1236 - accuracy: 0.9590 - val_loss: 0.4135 - val_accuracy: 0.6875\n",
            "Epoch 71/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1226 - accuracy: 0.9573 - val_loss: 0.3972 - val_accuracy: 0.6875\n",
            "Epoch 72/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1184 - accuracy: 0.9608 - val_loss: 0.3943 - val_accuracy: 0.6875\n",
            "Epoch 73/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1182 - accuracy: 0.9605 - val_loss: 0.4141 - val_accuracy: 0.6875\n",
            "Epoch 74/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1182 - accuracy: 0.9602 - val_loss: 0.3538 - val_accuracy: 0.7500\n",
            "Epoch 75/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.1174 - accuracy: 0.9608 - val_loss: 0.4145 - val_accuracy: 0.6875\n",
            "Epoch 76/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1150 - accuracy: 0.9625 - val_loss: 0.4208 - val_accuracy: 0.6875\n",
            "Epoch 77/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1140 - accuracy: 0.9625 - val_loss: 0.3201 - val_accuracy: 0.9375\n",
            "Epoch 78/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.1164 - accuracy: 0.9630 - val_loss: 0.3986 - val_accuracy: 0.6875\n",
            "Epoch 79/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1123 - accuracy: 0.9613 - val_loss: 0.3755 - val_accuracy: 0.7500\n",
            "Epoch 80/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1124 - accuracy: 0.9630 - val_loss: 0.3540 - val_accuracy: 0.7500\n",
            "Epoch 81/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1105 - accuracy: 0.9630 - val_loss: 0.3639 - val_accuracy: 0.7500\n",
            "Epoch 82/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1122 - accuracy: 0.9605 - val_loss: 0.3927 - val_accuracy: 0.6875\n",
            "Epoch 83/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1082 - accuracy: 0.9627 - val_loss: 0.3701 - val_accuracy: 0.7500\n",
            "Epoch 84/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1122 - accuracy: 0.9625 - val_loss: 0.3591 - val_accuracy: 0.7500\n",
            "Epoch 85/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1089 - accuracy: 0.9639 - val_loss: 0.3971 - val_accuracy: 0.6875\n",
            "Epoch 86/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1070 - accuracy: 0.9642 - val_loss: 0.3578 - val_accuracy: 0.7500\n",
            "Epoch 87/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1056 - accuracy: 0.9644 - val_loss: 0.3430 - val_accuracy: 0.7500\n",
            "Epoch 88/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.1067 - accuracy: 0.9667 - val_loss: 0.3466 - val_accuracy: 0.7500\n",
            "Epoch 89/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1062 - accuracy: 0.9653 - val_loss: 0.3047 - val_accuracy: 0.9375\n",
            "Epoch 90/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1036 - accuracy: 0.9630 - val_loss: 0.3141 - val_accuracy: 0.9375\n",
            "Epoch 91/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1027 - accuracy: 0.9647 - val_loss: 0.2994 - val_accuracy: 0.9375\n",
            "Epoch 92/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1024 - accuracy: 0.9656 - val_loss: 0.3214 - val_accuracy: 0.9375\n",
            "Epoch 93/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.1027 - accuracy: 0.9639 - val_loss: 0.3184 - val_accuracy: 0.9375\n",
            "Epoch 94/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1028 - accuracy: 0.9619 - val_loss: 0.3012 - val_accuracy: 0.9375\n",
            "Epoch 95/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1014 - accuracy: 0.9670 - val_loss: 0.3787 - val_accuracy: 0.7500\n",
            "Epoch 96/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.1001 - accuracy: 0.9656 - val_loss: 0.3885 - val_accuracy: 0.7500\n",
            "Epoch 97/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0992 - accuracy: 0.9664 - val_loss: 0.3379 - val_accuracy: 0.7500\n",
            "Epoch 98/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0986 - accuracy: 0.9681 - val_loss: 0.3591 - val_accuracy: 0.7500\n",
            "Epoch 99/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0986 - accuracy: 0.9667 - val_loss: 0.3155 - val_accuracy: 0.9375\n",
            "Epoch 100/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0961 - accuracy: 0.9650 - val_loss: 0.3526 - val_accuracy: 0.7500\n",
            "Epoch 101/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0964 - accuracy: 0.9664 - val_loss: 0.3342 - val_accuracy: 0.8125\n",
            "Epoch 102/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0984 - accuracy: 0.9644 - val_loss: 0.3512 - val_accuracy: 0.7500\n",
            "Epoch 103/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0983 - accuracy: 0.9633 - val_loss: 0.2904 - val_accuracy: 0.9375\n",
            "Epoch 104/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0959 - accuracy: 0.9664 - val_loss: 0.3549 - val_accuracy: 0.7500\n",
            "Epoch 105/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0949 - accuracy: 0.9679 - val_loss: 0.3396 - val_accuracy: 0.7500\n",
            "Epoch 106/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0946 - accuracy: 0.9667 - val_loss: 0.3156 - val_accuracy: 0.9375\n",
            "Epoch 107/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0947 - accuracy: 0.9693 - val_loss: 0.3254 - val_accuracy: 0.9375\n",
            "Epoch 108/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0919 - accuracy: 0.9707 - val_loss: 0.3133 - val_accuracy: 0.9375\n",
            "Epoch 109/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0906 - accuracy: 0.9679 - val_loss: 0.3402 - val_accuracy: 0.8125\n",
            "Epoch 110/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0953 - accuracy: 0.9656 - val_loss: 0.3392 - val_accuracy: 0.7500\n",
            "Epoch 111/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0915 - accuracy: 0.9693 - val_loss: 0.3151 - val_accuracy: 0.9375\n",
            "Epoch 112/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0930 - accuracy: 0.9684 - val_loss: 0.3440 - val_accuracy: 0.7500\n",
            "Epoch 113/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0915 - accuracy: 0.9684 - val_loss: 0.3431 - val_accuracy: 0.7500\n",
            "Epoch 114/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0928 - accuracy: 0.9679 - val_loss: 0.3661 - val_accuracy: 0.7500\n",
            "Epoch 115/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0906 - accuracy: 0.9699 - val_loss: 0.3651 - val_accuracy: 0.7500\n",
            "Epoch 116/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0890 - accuracy: 0.9690 - val_loss: 0.3754 - val_accuracy: 0.7500\n",
            "Epoch 117/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0913 - accuracy: 0.9670 - val_loss: 0.2811 - val_accuracy: 0.9375\n",
            "Epoch 118/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0894 - accuracy: 0.9690 - val_loss: 0.2858 - val_accuracy: 0.9375\n",
            "Epoch 119/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0881 - accuracy: 0.9681 - val_loss: 0.2782 - val_accuracy: 0.9375\n",
            "Epoch 120/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0900 - accuracy: 0.9693 - val_loss: 0.2555 - val_accuracy: 0.9375\n",
            "Epoch 121/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0886 - accuracy: 0.9699 - val_loss: 0.3445 - val_accuracy: 0.7500\n",
            "Epoch 122/500\n",
            "110/110 [==============================] - 5s 48ms/step - loss: 0.0899 - accuracy: 0.9690 - val_loss: 0.3027 - val_accuracy: 0.9375\n",
            "Epoch 123/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0857 - accuracy: 0.9690 - val_loss: 0.2919 - val_accuracy: 0.9375\n",
            "Epoch 124/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0866 - accuracy: 0.9693 - val_loss: 0.3485 - val_accuracy: 0.7500\n",
            "Epoch 125/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0854 - accuracy: 0.9704 - val_loss: 0.2892 - val_accuracy: 0.9375\n",
            "Epoch 126/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0863 - accuracy: 0.9693 - val_loss: 0.2994 - val_accuracy: 0.9375\n",
            "Epoch 127/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.0834 - accuracy: 0.9687 - val_loss: 0.2743 - val_accuracy: 0.9375\n",
            "Epoch 128/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0868 - accuracy: 0.9710 - val_loss: 0.2458 - val_accuracy: 0.9375\n",
            "Epoch 129/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0841 - accuracy: 0.9701 - val_loss: 0.2886 - val_accuracy: 0.9375\n",
            "Epoch 130/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0855 - accuracy: 0.9727 - val_loss: 0.3055 - val_accuracy: 0.9375\n",
            "Epoch 131/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0825 - accuracy: 0.9693 - val_loss: 0.2916 - val_accuracy: 0.9375\n",
            "Epoch 132/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0805 - accuracy: 0.9735 - val_loss: 0.3188 - val_accuracy: 0.8750\n",
            "Epoch 133/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0864 - accuracy: 0.9721 - val_loss: 0.2844 - val_accuracy: 0.9375\n",
            "Epoch 134/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0827 - accuracy: 0.9721 - val_loss: 0.2993 - val_accuracy: 0.9375\n",
            "Epoch 135/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0827 - accuracy: 0.9721 - val_loss: 0.2908 - val_accuracy: 0.9375\n",
            "Epoch 136/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0850 - accuracy: 0.9707 - val_loss: 0.2756 - val_accuracy: 0.9375\n",
            "Epoch 137/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0839 - accuracy: 0.9707 - val_loss: 0.2485 - val_accuracy: 0.9375\n",
            "Epoch 138/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.0806 - accuracy: 0.9718 - val_loss: 0.2662 - val_accuracy: 0.9375\n",
            "Epoch 139/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0806 - accuracy: 0.9704 - val_loss: 0.3078 - val_accuracy: 0.9375\n",
            "Epoch 140/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0844 - accuracy: 0.9701 - val_loss: 0.2924 - val_accuracy: 0.9375\n",
            "Epoch 141/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0783 - accuracy: 0.9741 - val_loss: 0.3422 - val_accuracy: 0.8125\n",
            "Epoch 142/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0828 - accuracy: 0.9724 - val_loss: 0.2477 - val_accuracy: 0.9375\n",
            "Epoch 143/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0831 - accuracy: 0.9701 - val_loss: 0.2784 - val_accuracy: 0.9375\n",
            "Epoch 144/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0791 - accuracy: 0.9718 - val_loss: 0.3062 - val_accuracy: 0.9375\n",
            "Epoch 145/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0795 - accuracy: 0.9693 - val_loss: 0.3010 - val_accuracy: 0.9375\n",
            "Epoch 146/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0794 - accuracy: 0.9724 - val_loss: 0.2638 - val_accuracy: 0.9375\n",
            "Epoch 147/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.0797 - accuracy: 0.9710 - val_loss: 0.2633 - val_accuracy: 0.9375\n",
            "Epoch 148/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0800 - accuracy: 0.9721 - val_loss: 0.2862 - val_accuracy: 0.9375\n",
            "Epoch 149/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0816 - accuracy: 0.9693 - val_loss: 0.2834 - val_accuracy: 0.9375\n",
            "Epoch 150/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0765 - accuracy: 0.9730 - val_loss: 0.2634 - val_accuracy: 0.9375\n",
            "Epoch 151/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0784 - accuracy: 0.9718 - val_loss: 0.2820 - val_accuracy: 0.9375\n",
            "Epoch 152/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0788 - accuracy: 0.9741 - val_loss: 0.2436 - val_accuracy: 0.9375\n",
            "Epoch 153/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0780 - accuracy: 0.9738 - val_loss: 0.2478 - val_accuracy: 0.9375\n",
            "Epoch 154/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0763 - accuracy: 0.9747 - val_loss: 0.2881 - val_accuracy: 0.9375\n",
            "Epoch 155/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0761 - accuracy: 0.9727 - val_loss: 0.2708 - val_accuracy: 0.9375\n",
            "Epoch 156/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0786 - accuracy: 0.9730 - val_loss: 0.2961 - val_accuracy: 0.9375\n",
            "Epoch 157/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0757 - accuracy: 0.9716 - val_loss: 0.2640 - val_accuracy: 0.9375\n",
            "Epoch 158/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0748 - accuracy: 0.9758 - val_loss: 0.2336 - val_accuracy: 0.9375\n",
            "Epoch 159/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0742 - accuracy: 0.9724 - val_loss: 0.2487 - val_accuracy: 0.9375\n",
            "Epoch 160/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0753 - accuracy: 0.9713 - val_loss: 0.2768 - val_accuracy: 0.9375\n",
            "Epoch 161/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0735 - accuracy: 0.9744 - val_loss: 0.3304 - val_accuracy: 0.8750\n",
            "Epoch 162/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.0754 - accuracy: 0.9724 - val_loss: 0.3051 - val_accuracy: 0.8750\n",
            "Epoch 163/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0756 - accuracy: 0.9727 - val_loss: 0.2318 - val_accuracy: 0.9375\n",
            "Epoch 164/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0766 - accuracy: 0.9741 - val_loss: 0.2129 - val_accuracy: 0.9375\n",
            "Epoch 165/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0778 - accuracy: 0.9721 - val_loss: 0.2520 - val_accuracy: 0.9375\n",
            "Epoch 166/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0739 - accuracy: 0.9741 - val_loss: 0.2559 - val_accuracy: 0.9375\n",
            "Epoch 167/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0744 - accuracy: 0.9724 - val_loss: 0.3010 - val_accuracy: 0.8750\n",
            "Epoch 168/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0758 - accuracy: 0.9727 - val_loss: 0.2418 - val_accuracy: 0.9375\n",
            "Epoch 169/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0748 - accuracy: 0.9741 - val_loss: 0.2009 - val_accuracy: 0.9375\n",
            "Epoch 170/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0743 - accuracy: 0.9767 - val_loss: 0.2625 - val_accuracy: 0.9375\n",
            "Epoch 171/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0740 - accuracy: 0.9755 - val_loss: 0.2086 - val_accuracy: 0.9375\n",
            "Epoch 172/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0723 - accuracy: 0.9761 - val_loss: 0.2823 - val_accuracy: 0.9375\n",
            "Epoch 173/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0728 - accuracy: 0.9724 - val_loss: 0.1925 - val_accuracy: 0.9375\n",
            "Epoch 174/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0737 - accuracy: 0.9716 - val_loss: 0.2419 - val_accuracy: 0.9375\n",
            "Epoch 175/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0711 - accuracy: 0.9744 - val_loss: 0.2198 - val_accuracy: 0.9375\n",
            "Epoch 176/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0720 - accuracy: 0.9772 - val_loss: 0.2700 - val_accuracy: 0.9375\n",
            "Epoch 177/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0742 - accuracy: 0.9750 - val_loss: 0.2837 - val_accuracy: 0.9375\n",
            "Epoch 178/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0723 - accuracy: 0.9772 - val_loss: 0.2709 - val_accuracy: 0.9375\n",
            "Epoch 179/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0727 - accuracy: 0.9755 - val_loss: 0.2140 - val_accuracy: 0.9375\n",
            "Epoch 180/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0715 - accuracy: 0.9724 - val_loss: 0.1997 - val_accuracy: 0.9375\n",
            "Epoch 181/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0708 - accuracy: 0.9750 - val_loss: 0.2498 - val_accuracy: 0.9375\n",
            "Epoch 182/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0702 - accuracy: 0.9753 - val_loss: 0.2623 - val_accuracy: 0.9375\n",
            "Epoch 183/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0714 - accuracy: 0.9735 - val_loss: 0.2753 - val_accuracy: 0.9375\n",
            "Epoch 184/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0700 - accuracy: 0.9761 - val_loss: 0.2461 - val_accuracy: 0.9375\n",
            "Epoch 185/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0708 - accuracy: 0.9741 - val_loss: 0.2463 - val_accuracy: 0.9375\n",
            "Epoch 186/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0669 - accuracy: 0.9747 - val_loss: 0.2430 - val_accuracy: 0.9375\n",
            "Epoch 187/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0674 - accuracy: 0.9750 - val_loss: 0.2514 - val_accuracy: 0.9375\n",
            "Epoch 188/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0710 - accuracy: 0.9747 - val_loss: 0.2443 - val_accuracy: 0.9375\n",
            "Epoch 189/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0689 - accuracy: 0.9753 - val_loss: 0.2224 - val_accuracy: 0.9375\n",
            "Epoch 190/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0666 - accuracy: 0.9761 - val_loss: 0.3219 - val_accuracy: 0.8750\n",
            "Epoch 191/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0713 - accuracy: 0.9727 - val_loss: 0.2360 - val_accuracy: 0.9375\n",
            "Epoch 192/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0687 - accuracy: 0.9764 - val_loss: 0.2475 - val_accuracy: 0.9375\n",
            "Epoch 193/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0693 - accuracy: 0.9741 - val_loss: 0.3124 - val_accuracy: 0.8750\n",
            "Epoch 194/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0697 - accuracy: 0.9735 - val_loss: 0.2451 - val_accuracy: 0.9375\n",
            "Epoch 195/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0669 - accuracy: 0.9770 - val_loss: 0.2199 - val_accuracy: 0.9375\n",
            "Epoch 196/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0664 - accuracy: 0.9767 - val_loss: 0.2580 - val_accuracy: 0.9375\n",
            "Epoch 197/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0677 - accuracy: 0.9750 - val_loss: 0.2677 - val_accuracy: 0.9375\n",
            "Epoch 198/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0670 - accuracy: 0.9738 - val_loss: 0.2309 - val_accuracy: 0.9375\n",
            "Epoch 199/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0659 - accuracy: 0.9730 - val_loss: 0.2745 - val_accuracy: 0.9375\n",
            "Epoch 200/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0698 - accuracy: 0.9770 - val_loss: 0.1893 - val_accuracy: 0.9375\n",
            "Epoch 201/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0694 - accuracy: 0.9761 - val_loss: 0.2214 - val_accuracy: 0.9375\n",
            "Epoch 202/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0683 - accuracy: 0.9770 - val_loss: 0.2495 - val_accuracy: 0.9375\n",
            "Epoch 203/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0671 - accuracy: 0.9778 - val_loss: 0.2370 - val_accuracy: 0.9375\n",
            "Epoch 204/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0681 - accuracy: 0.9738 - val_loss: 0.2132 - val_accuracy: 0.9375\n",
            "Epoch 205/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0662 - accuracy: 0.9758 - val_loss: 0.2386 - val_accuracy: 0.9375\n",
            "Epoch 206/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0652 - accuracy: 0.9772 - val_loss: 0.2089 - val_accuracy: 0.9375\n",
            "Epoch 207/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0680 - accuracy: 0.9747 - val_loss: 0.2083 - val_accuracy: 0.9375\n",
            "Epoch 208/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0652 - accuracy: 0.9778 - val_loss: 0.2040 - val_accuracy: 0.9375\n",
            "Epoch 209/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0665 - accuracy: 0.9753 - val_loss: 0.2788 - val_accuracy: 0.9375\n",
            "Epoch 210/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0637 - accuracy: 0.9778 - val_loss: 0.2471 - val_accuracy: 0.9375\n",
            "Epoch 211/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0666 - accuracy: 0.9770 - val_loss: 0.2476 - val_accuracy: 0.9375\n",
            "Epoch 212/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0634 - accuracy: 0.9792 - val_loss: 0.2239 - val_accuracy: 0.9375\n",
            "Epoch 213/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0644 - accuracy: 0.9770 - val_loss: 0.2821 - val_accuracy: 0.9375\n",
            "Epoch 214/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0630 - accuracy: 0.9801 - val_loss: 0.2986 - val_accuracy: 0.8750\n",
            "Epoch 215/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0658 - accuracy: 0.9764 - val_loss: 0.2149 - val_accuracy: 0.9375\n",
            "Epoch 216/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0630 - accuracy: 0.9787 - val_loss: 0.1893 - val_accuracy: 0.9375\n",
            "Epoch 217/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0637 - accuracy: 0.9790 - val_loss: 0.2209 - val_accuracy: 0.9375\n",
            "Epoch 218/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0640 - accuracy: 0.9787 - val_loss: 0.2008 - val_accuracy: 0.9375\n",
            "Epoch 219/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0622 - accuracy: 0.9775 - val_loss: 0.2403 - val_accuracy: 0.9375\n",
            "Epoch 220/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0622 - accuracy: 0.9787 - val_loss: 0.2881 - val_accuracy: 0.8750\n",
            "Epoch 221/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0646 - accuracy: 0.9781 - val_loss: 0.2915 - val_accuracy: 0.8750\n",
            "Epoch 222/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0638 - accuracy: 0.9758 - val_loss: 0.1959 - val_accuracy: 0.9375\n",
            "Epoch 223/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0610 - accuracy: 0.9787 - val_loss: 0.2007 - val_accuracy: 0.9375\n",
            "Epoch 224/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0651 - accuracy: 0.9770 - val_loss: 0.2591 - val_accuracy: 0.9375\n",
            "Epoch 225/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0607 - accuracy: 0.9772 - val_loss: 0.2471 - val_accuracy: 0.9375\n",
            "Epoch 226/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0613 - accuracy: 0.9784 - val_loss: 0.2639 - val_accuracy: 0.9375\n",
            "Epoch 227/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0610 - accuracy: 0.9778 - val_loss: 0.1999 - val_accuracy: 0.9375\n",
            "Epoch 228/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0610 - accuracy: 0.9809 - val_loss: 0.2516 - val_accuracy: 0.9375\n",
            "Epoch 229/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0627 - accuracy: 0.9772 - val_loss: 0.2487 - val_accuracy: 0.9375\n",
            "Epoch 230/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0606 - accuracy: 0.9781 - val_loss: 0.2454 - val_accuracy: 0.9375\n",
            "Epoch 231/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0638 - accuracy: 0.9753 - val_loss: 0.2745 - val_accuracy: 0.9375\n",
            "Epoch 232/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0604 - accuracy: 0.9807 - val_loss: 0.2124 - val_accuracy: 0.9375\n",
            "Epoch 233/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0626 - accuracy: 0.9781 - val_loss: 0.2305 - val_accuracy: 0.9375\n",
            "Epoch 234/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0606 - accuracy: 0.9792 - val_loss: 0.2439 - val_accuracy: 0.9375\n",
            "Epoch 235/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0584 - accuracy: 0.9804 - val_loss: 0.3010 - val_accuracy: 0.8750\n",
            "Epoch 236/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0608 - accuracy: 0.9795 - val_loss: 0.2123 - val_accuracy: 0.9375\n",
            "Epoch 237/500\n",
            "110/110 [==============================] - 5s 47ms/step - loss: 0.0615 - accuracy: 0.9792 - val_loss: 0.2555 - val_accuracy: 0.9375\n",
            "Epoch 238/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0615 - accuracy: 0.9795 - val_loss: 0.2359 - val_accuracy: 0.9375\n",
            "Epoch 239/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0604 - accuracy: 0.9792 - val_loss: 0.1759 - val_accuracy: 0.9375\n",
            "Epoch 240/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0607 - accuracy: 0.9784 - val_loss: 0.2598 - val_accuracy: 0.9375\n",
            "Epoch 241/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0582 - accuracy: 0.9790 - val_loss: 0.2432 - val_accuracy: 0.9375\n",
            "Epoch 242/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0612 - accuracy: 0.9770 - val_loss: 0.2497 - val_accuracy: 0.9375\n",
            "Epoch 243/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0607 - accuracy: 0.9792 - val_loss: 0.2909 - val_accuracy: 0.8750\n",
            "Epoch 244/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0603 - accuracy: 0.9792 - val_loss: 0.2187 - val_accuracy: 0.9375\n",
            "Epoch 245/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0608 - accuracy: 0.9784 - val_loss: 0.2392 - val_accuracy: 0.9375\n",
            "Epoch 246/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0576 - accuracy: 0.9778 - val_loss: 0.2192 - val_accuracy: 0.9375\n",
            "Epoch 247/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0591 - accuracy: 0.9795 - val_loss: 0.2174 - val_accuracy: 0.9375\n",
            "Epoch 248/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0583 - accuracy: 0.9795 - val_loss: 0.2031 - val_accuracy: 0.9375\n",
            "Epoch 249/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0600 - accuracy: 0.9801 - val_loss: 0.2458 - val_accuracy: 0.9375\n",
            "Epoch 250/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0590 - accuracy: 0.9795 - val_loss: 0.2436 - val_accuracy: 0.9375\n",
            "Epoch 251/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0612 - accuracy: 0.9787 - val_loss: 0.1977 - val_accuracy: 0.9375\n",
            "Epoch 252/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0569 - accuracy: 0.9804 - val_loss: 0.2190 - val_accuracy: 0.9375\n",
            "Epoch 253/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0566 - accuracy: 0.9804 - val_loss: 0.3022 - val_accuracy: 0.8750\n",
            "Epoch 254/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0563 - accuracy: 0.9809 - val_loss: 0.2344 - val_accuracy: 0.9375\n",
            "Epoch 255/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0589 - accuracy: 0.9772 - val_loss: 0.2012 - val_accuracy: 0.9375\n",
            "Epoch 256/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0577 - accuracy: 0.9784 - val_loss: 0.2172 - val_accuracy: 0.9375\n",
            "Epoch 257/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0579 - accuracy: 0.9792 - val_loss: 0.1965 - val_accuracy: 0.9375\n",
            "Epoch 258/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0584 - accuracy: 0.9787 - val_loss: 0.2043 - val_accuracy: 0.9375\n",
            "Epoch 259/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0555 - accuracy: 0.9818 - val_loss: 0.2267 - val_accuracy: 0.9375\n",
            "Epoch 260/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0593 - accuracy: 0.9798 - val_loss: 0.2686 - val_accuracy: 0.9375\n",
            "Epoch 261/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0576 - accuracy: 0.9815 - val_loss: 0.2531 - val_accuracy: 0.9375\n",
            "Epoch 262/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0570 - accuracy: 0.9784 - val_loss: 0.2285 - val_accuracy: 0.9375\n",
            "Epoch 263/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0538 - accuracy: 0.9798 - val_loss: 0.1804 - val_accuracy: 0.9375\n",
            "Epoch 264/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0585 - accuracy: 0.9781 - val_loss: 0.3208 - val_accuracy: 0.8750\n",
            "Epoch 265/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0562 - accuracy: 0.9787 - val_loss: 0.2005 - val_accuracy: 0.9375\n",
            "Epoch 266/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0590 - accuracy: 0.9778 - val_loss: 0.1878 - val_accuracy: 0.9375\n",
            "Epoch 267/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0557 - accuracy: 0.9818 - val_loss: 0.2849 - val_accuracy: 0.8750\n",
            "Epoch 268/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0553 - accuracy: 0.9778 - val_loss: 0.2539 - val_accuracy: 0.9375\n",
            "Epoch 269/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0548 - accuracy: 0.9824 - val_loss: 0.2412 - val_accuracy: 0.9375\n",
            "Epoch 270/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0552 - accuracy: 0.9812 - val_loss: 0.2145 - val_accuracy: 0.9375\n",
            "Epoch 271/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0571 - accuracy: 0.9778 - val_loss: 0.2474 - val_accuracy: 0.9375\n",
            "Epoch 272/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0527 - accuracy: 0.9809 - val_loss: 0.2016 - val_accuracy: 0.9375\n",
            "Epoch 273/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0553 - accuracy: 0.9807 - val_loss: 0.1735 - val_accuracy: 0.9375\n",
            "Epoch 274/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0562 - accuracy: 0.9790 - val_loss: 0.2584 - val_accuracy: 0.9375\n",
            "Epoch 275/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0556 - accuracy: 0.9798 - val_loss: 0.1947 - val_accuracy: 0.9375\n",
            "Epoch 276/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0554 - accuracy: 0.9818 - val_loss: 0.2419 - val_accuracy: 0.9375\n",
            "Epoch 277/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0544 - accuracy: 0.9798 - val_loss: 0.3082 - val_accuracy: 0.8750\n",
            "Epoch 278/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0533 - accuracy: 0.9821 - val_loss: 0.2022 - val_accuracy: 0.9375\n",
            "Epoch 279/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0541 - accuracy: 0.9812 - val_loss: 0.2427 - val_accuracy: 0.9375\n",
            "Epoch 280/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0521 - accuracy: 0.9815 - val_loss: 0.2523 - val_accuracy: 0.9375\n",
            "Epoch 281/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0538 - accuracy: 0.9821 - val_loss: 0.2295 - val_accuracy: 0.9375\n",
            "Epoch 282/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0537 - accuracy: 0.9812 - val_loss: 0.2152 - val_accuracy: 0.9375\n",
            "Epoch 283/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0549 - accuracy: 0.9827 - val_loss: 0.2210 - val_accuracy: 0.9375\n",
            "Epoch 284/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0550 - accuracy: 0.9795 - val_loss: 0.1884 - val_accuracy: 0.9375\n",
            "Epoch 285/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0520 - accuracy: 0.9832 - val_loss: 0.1974 - val_accuracy: 0.9375\n",
            "Epoch 286/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0547 - accuracy: 0.9801 - val_loss: 0.1960 - val_accuracy: 0.9375\n",
            "Epoch 287/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0517 - accuracy: 0.9827 - val_loss: 0.1918 - val_accuracy: 0.9375\n",
            "Epoch 288/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0548 - accuracy: 0.9818 - val_loss: 0.2225 - val_accuracy: 0.9375\n",
            "Epoch 289/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0538 - accuracy: 0.9835 - val_loss: 0.2611 - val_accuracy: 0.9375\n",
            "Epoch 290/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0531 - accuracy: 0.9815 - val_loss: 0.1842 - val_accuracy: 0.9375\n",
            "Epoch 291/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0534 - accuracy: 0.9827 - val_loss: 0.2286 - val_accuracy: 0.9375\n",
            "Epoch 292/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0529 - accuracy: 0.9827 - val_loss: 0.2236 - val_accuracy: 0.9375\n",
            "Epoch 293/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0543 - accuracy: 0.9829 - val_loss: 0.2599 - val_accuracy: 0.9375\n",
            "Epoch 294/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0528 - accuracy: 0.9809 - val_loss: 0.3012 - val_accuracy: 0.8750\n",
            "Epoch 295/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0540 - accuracy: 0.9807 - val_loss: 0.2083 - val_accuracy: 0.9375\n",
            "Epoch 296/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0516 - accuracy: 0.9827 - val_loss: 0.2222 - val_accuracy: 0.9375\n",
            "Epoch 297/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0550 - accuracy: 0.9807 - val_loss: 0.1804 - val_accuracy: 0.9375\n",
            "Epoch 298/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0530 - accuracy: 0.9829 - val_loss: 0.1678 - val_accuracy: 0.9375\n",
            "Epoch 299/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0554 - accuracy: 0.9809 - val_loss: 0.2298 - val_accuracy: 0.9375\n",
            "Epoch 300/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0509 - accuracy: 0.9849 - val_loss: 0.1974 - val_accuracy: 0.9375\n",
            "Epoch 301/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0510 - accuracy: 0.9812 - val_loss: 0.2159 - val_accuracy: 0.9375\n",
            "Epoch 302/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0497 - accuracy: 0.9849 - val_loss: 0.2918 - val_accuracy: 0.8750\n",
            "Epoch 303/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0532 - accuracy: 0.9812 - val_loss: 0.2306 - val_accuracy: 0.9375\n",
            "Epoch 304/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0507 - accuracy: 0.9835 - val_loss: 0.2609 - val_accuracy: 0.8750\n",
            "Epoch 305/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0491 - accuracy: 0.9829 - val_loss: 0.2318 - val_accuracy: 0.9375\n",
            "Epoch 306/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0512 - accuracy: 0.9809 - val_loss: 0.2316 - val_accuracy: 0.9375\n",
            "Epoch 307/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0512 - accuracy: 0.9846 - val_loss: 0.2127 - val_accuracy: 0.9375\n",
            "Epoch 308/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0509 - accuracy: 0.9815 - val_loss: 0.1663 - val_accuracy: 0.9375\n",
            "Epoch 309/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0499 - accuracy: 0.9838 - val_loss: 0.2182 - val_accuracy: 0.9375\n",
            "Epoch 310/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0518 - accuracy: 0.9815 - val_loss: 0.2167 - val_accuracy: 0.9375\n",
            "Epoch 311/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0509 - accuracy: 0.9815 - val_loss: 0.2063 - val_accuracy: 0.9375\n",
            "Epoch 312/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0488 - accuracy: 0.9855 - val_loss: 0.2257 - val_accuracy: 0.9375\n",
            "Epoch 313/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0486 - accuracy: 0.9827 - val_loss: 0.2252 - val_accuracy: 0.9375\n",
            "Epoch 314/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0491 - accuracy: 0.9846 - val_loss: 0.2186 - val_accuracy: 0.9375\n",
            "Epoch 315/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0524 - accuracy: 0.9824 - val_loss: 0.2121 - val_accuracy: 0.9375\n",
            "Epoch 316/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0500 - accuracy: 0.9821 - val_loss: 0.1820 - val_accuracy: 0.9375\n",
            "Epoch 317/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0513 - accuracy: 0.9846 - val_loss: 0.1924 - val_accuracy: 0.9375\n",
            "Epoch 318/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0501 - accuracy: 0.9844 - val_loss: 0.1760 - val_accuracy: 0.9375\n",
            "Epoch 319/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0477 - accuracy: 0.9844 - val_loss: 0.2078 - val_accuracy: 0.9375\n",
            "Epoch 320/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0501 - accuracy: 0.9829 - val_loss: 0.2152 - val_accuracy: 0.9375\n",
            "Epoch 321/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0500 - accuracy: 0.9821 - val_loss: 0.2007 - val_accuracy: 0.9375\n",
            "Epoch 322/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0489 - accuracy: 0.9824 - val_loss: 0.1629 - val_accuracy: 0.9375\n",
            "Epoch 323/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0502 - accuracy: 0.9824 - val_loss: 0.2372 - val_accuracy: 0.9375\n",
            "Epoch 324/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0485 - accuracy: 0.9844 - val_loss: 0.2238 - val_accuracy: 0.9375\n",
            "Epoch 325/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0493 - accuracy: 0.9827 - val_loss: 0.2843 - val_accuracy: 0.8750\n",
            "Epoch 326/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0452 - accuracy: 0.9849 - val_loss: 0.2487 - val_accuracy: 0.9375\n",
            "Epoch 327/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0498 - accuracy: 0.9824 - val_loss: 0.2799 - val_accuracy: 0.8750\n",
            "Epoch 328/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0480 - accuracy: 0.9832 - val_loss: 0.2338 - val_accuracy: 0.9375\n",
            "Epoch 329/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0474 - accuracy: 0.9841 - val_loss: 0.1935 - val_accuracy: 0.9375\n",
            "Epoch 330/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0478 - accuracy: 0.9855 - val_loss: 0.2003 - val_accuracy: 0.9375\n",
            "Epoch 331/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0468 - accuracy: 0.9844 - val_loss: 0.2106 - val_accuracy: 0.9375\n",
            "Epoch 332/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0488 - accuracy: 0.9841 - val_loss: 0.2040 - val_accuracy: 0.9375\n",
            "Epoch 333/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0511 - accuracy: 0.9824 - val_loss: 0.2333 - val_accuracy: 0.9375\n",
            "Epoch 334/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0461 - accuracy: 0.9844 - val_loss: 0.1989 - val_accuracy: 0.9375\n",
            "Epoch 335/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0464 - accuracy: 0.9838 - val_loss: 0.1973 - val_accuracy: 0.9375\n",
            "Epoch 336/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0462 - accuracy: 0.9832 - val_loss: 0.1602 - val_accuracy: 0.9375\n",
            "Epoch 337/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0496 - accuracy: 0.9829 - val_loss: 0.2298 - val_accuracy: 0.9375\n",
            "Epoch 338/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0496 - accuracy: 0.9818 - val_loss: 0.1945 - val_accuracy: 0.9375\n",
            "Epoch 339/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0462 - accuracy: 0.9829 - val_loss: 0.2111 - val_accuracy: 0.9375\n",
            "Epoch 340/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0464 - accuracy: 0.9841 - val_loss: 0.2347 - val_accuracy: 0.9375\n",
            "Epoch 341/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0483 - accuracy: 0.9818 - val_loss: 0.2935 - val_accuracy: 0.8750\n",
            "Epoch 342/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0450 - accuracy: 0.9835 - val_loss: 0.2394 - val_accuracy: 0.9375\n",
            "Epoch 343/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0450 - accuracy: 0.9838 - val_loss: 0.2603 - val_accuracy: 0.8750\n",
            "Epoch 344/500\n",
            "110/110 [==============================] - 5s 43ms/step - loss: 0.0478 - accuracy: 0.9849 - val_loss: 0.1989 - val_accuracy: 0.9375\n",
            "Epoch 345/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0470 - accuracy: 0.9835 - val_loss: 0.2050 - val_accuracy: 0.9375\n",
            "Epoch 346/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0469 - accuracy: 0.9846 - val_loss: 0.2578 - val_accuracy: 0.9375\n",
            "Epoch 347/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0485 - accuracy: 0.9829 - val_loss: 0.1787 - val_accuracy: 0.9375\n",
            "Epoch 348/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0475 - accuracy: 0.9827 - val_loss: 0.2649 - val_accuracy: 0.8750\n",
            "Epoch 349/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0451 - accuracy: 0.9849 - val_loss: 0.2163 - val_accuracy: 0.9375\n",
            "Epoch 350/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0474 - accuracy: 0.9849 - val_loss: 0.1945 - val_accuracy: 0.9375\n",
            "Epoch 351/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0457 - accuracy: 0.9849 - val_loss: 0.1817 - val_accuracy: 0.9375\n",
            "Epoch 352/500\n",
            "110/110 [==============================] - 5s 46ms/step - loss: 0.0459 - accuracy: 0.9844 - val_loss: 0.1954 - val_accuracy: 0.9375\n",
            "Epoch 353/500\n",
            "110/110 [==============================] - 5s 46ms/step - loss: 0.0444 - accuracy: 0.9858 - val_loss: 0.2072 - val_accuracy: 0.9375\n",
            "Epoch 354/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0460 - accuracy: 0.9844 - val_loss: 0.2513 - val_accuracy: 0.9375\n",
            "Epoch 355/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0459 - accuracy: 0.9844 - val_loss: 0.2525 - val_accuracy: 0.9375\n",
            "Epoch 356/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0474 - accuracy: 0.9832 - val_loss: 0.2098 - val_accuracy: 0.9375\n",
            "Epoch 357/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0453 - accuracy: 0.9841 - val_loss: 0.1860 - val_accuracy: 0.9375\n",
            "Epoch 358/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0444 - accuracy: 0.9863 - val_loss: 0.2092 - val_accuracy: 0.9375\n",
            "Epoch 359/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0465 - accuracy: 0.9858 - val_loss: 0.1671 - val_accuracy: 0.9375\n",
            "Epoch 360/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0439 - accuracy: 0.9829 - val_loss: 0.2062 - val_accuracy: 0.9375\n",
            "Epoch 361/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0460 - accuracy: 0.9863 - val_loss: 0.1969 - val_accuracy: 0.9375\n",
            "Epoch 362/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0458 - accuracy: 0.9852 - val_loss: 0.2217 - val_accuracy: 0.9375\n",
            "Epoch 363/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0435 - accuracy: 0.9844 - val_loss: 0.1508 - val_accuracy: 0.9375\n",
            "Epoch 364/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0452 - accuracy: 0.9855 - val_loss: 0.2206 - val_accuracy: 0.9375\n",
            "Epoch 365/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0459 - accuracy: 0.9844 - val_loss: 0.2051 - val_accuracy: 0.9375\n",
            "Epoch 366/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0416 - accuracy: 0.9861 - val_loss: 0.2048 - val_accuracy: 0.9375\n",
            "Epoch 367/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0462 - accuracy: 0.9844 - val_loss: 0.1912 - val_accuracy: 0.9375\n",
            "Epoch 368/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0436 - accuracy: 0.9861 - val_loss: 0.1789 - val_accuracy: 0.9375\n",
            "Epoch 369/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0459 - accuracy: 0.9846 - val_loss: 0.1919 - val_accuracy: 0.9375\n",
            "Epoch 370/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0441 - accuracy: 0.9863 - val_loss: 0.2283 - val_accuracy: 0.9375\n",
            "Epoch 371/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0440 - accuracy: 0.9841 - val_loss: 0.2293 - val_accuracy: 0.9375\n",
            "Epoch 372/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0438 - accuracy: 0.9858 - val_loss: 0.1427 - val_accuracy: 0.9375\n",
            "Epoch 373/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0438 - accuracy: 0.9855 - val_loss: 0.2304 - val_accuracy: 0.9375\n",
            "Epoch 374/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0442 - accuracy: 0.9863 - val_loss: 0.2131 - val_accuracy: 0.9375\n",
            "Epoch 375/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0459 - accuracy: 0.9858 - val_loss: 0.1876 - val_accuracy: 0.9375\n",
            "Epoch 376/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0448 - accuracy: 0.9855 - val_loss: 0.2376 - val_accuracy: 0.9375\n",
            "Epoch 377/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0449 - accuracy: 0.9824 - val_loss: 0.2048 - val_accuracy: 0.9375\n",
            "Epoch 378/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0448 - accuracy: 0.9861 - val_loss: 0.2490 - val_accuracy: 0.9375\n",
            "Epoch 379/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0438 - accuracy: 0.9852 - val_loss: 0.1880 - val_accuracy: 0.9375\n",
            "Epoch 380/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0438 - accuracy: 0.9863 - val_loss: 0.2179 - val_accuracy: 0.9375\n",
            "Epoch 381/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0456 - accuracy: 0.9846 - val_loss: 0.1251 - val_accuracy: 0.9375\n",
            "Epoch 382/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0440 - accuracy: 0.9861 - val_loss: 0.1489 - val_accuracy: 0.9375\n",
            "Epoch 383/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0423 - accuracy: 0.9844 - val_loss: 0.2125 - val_accuracy: 0.9375\n",
            "Epoch 384/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0431 - accuracy: 0.9852 - val_loss: 0.2033 - val_accuracy: 0.9375\n",
            "Epoch 385/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0444 - accuracy: 0.9832 - val_loss: 0.1748 - val_accuracy: 0.9375\n",
            "Epoch 386/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0424 - accuracy: 0.9861 - val_loss: 0.2545 - val_accuracy: 0.9375\n",
            "Epoch 387/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0426 - accuracy: 0.9858 - val_loss: 0.1989 - val_accuracy: 0.9375\n",
            "Epoch 388/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0426 - accuracy: 0.9869 - val_loss: 0.1849 - val_accuracy: 0.9375\n",
            "Epoch 389/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0427 - accuracy: 0.9849 - val_loss: 0.2338 - val_accuracy: 0.9375\n",
            "Epoch 390/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0430 - accuracy: 0.9855 - val_loss: 0.1818 - val_accuracy: 0.9375\n",
            "Epoch 391/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0434 - accuracy: 0.9835 - val_loss: 0.2072 - val_accuracy: 0.9375\n",
            "Epoch 392/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0381 - accuracy: 0.9872 - val_loss: 0.2592 - val_accuracy: 0.8750\n",
            "Epoch 393/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0403 - accuracy: 0.9878 - val_loss: 0.1814 - val_accuracy: 0.9375\n",
            "Epoch 394/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0404 - accuracy: 0.9869 - val_loss: 0.2189 - val_accuracy: 0.9375\n",
            "Epoch 395/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0430 - accuracy: 0.9852 - val_loss: 0.1855 - val_accuracy: 0.9375\n",
            "Epoch 396/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0422 - accuracy: 0.9858 - val_loss: 0.2304 - val_accuracy: 0.9375\n",
            "Epoch 397/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0395 - accuracy: 0.9886 - val_loss: 0.1773 - val_accuracy: 0.9375\n",
            "Epoch 398/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0391 - accuracy: 0.9886 - val_loss: 0.1477 - val_accuracy: 0.9375\n",
            "Epoch 399/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0411 - accuracy: 0.9861 - val_loss: 0.2126 - val_accuracy: 0.9375\n",
            "Epoch 400/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0414 - accuracy: 0.9846 - val_loss: 0.2222 - val_accuracy: 0.9375\n",
            "Epoch 401/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0420 - accuracy: 0.9855 - val_loss: 0.1947 - val_accuracy: 0.9375\n",
            "Epoch 402/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0407 - accuracy: 0.9849 - val_loss: 0.2000 - val_accuracy: 0.9375\n",
            "Epoch 403/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0418 - accuracy: 0.9852 - val_loss: 0.1612 - val_accuracy: 0.9375\n",
            "Epoch 404/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0416 - accuracy: 0.9846 - val_loss: 0.1843 - val_accuracy: 0.9375\n",
            "Epoch 405/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0396 - accuracy: 0.9875 - val_loss: 0.1580 - val_accuracy: 0.9375\n",
            "Epoch 406/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0413 - accuracy: 0.9849 - val_loss: 0.2252 - val_accuracy: 0.9375\n",
            "Epoch 407/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0400 - accuracy: 0.9844 - val_loss: 0.1227 - val_accuracy: 0.9375\n",
            "Epoch 408/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0414 - accuracy: 0.9869 - val_loss: 0.2305 - val_accuracy: 0.9375\n",
            "Epoch 409/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0402 - accuracy: 0.9878 - val_loss: 0.1858 - val_accuracy: 0.9375\n",
            "Epoch 410/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0404 - accuracy: 0.9869 - val_loss: 0.1993 - val_accuracy: 0.9375\n",
            "Epoch 411/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0390 - accuracy: 0.9878 - val_loss: 0.2005 - val_accuracy: 0.9375\n",
            "Epoch 412/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0387 - accuracy: 0.9869 - val_loss: 0.1895 - val_accuracy: 0.9375\n",
            "Epoch 413/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0419 - accuracy: 0.9861 - val_loss: 0.1946 - val_accuracy: 0.9375\n",
            "Epoch 414/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0416 - accuracy: 0.9855 - val_loss: 0.2027 - val_accuracy: 0.9375\n",
            "Epoch 415/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0397 - accuracy: 0.9886 - val_loss: 0.1813 - val_accuracy: 0.9375\n",
            "Epoch 416/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0395 - accuracy: 0.9881 - val_loss: 0.1975 - val_accuracy: 0.9375\n",
            "Epoch 417/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0385 - accuracy: 0.9855 - val_loss: 0.1625 - val_accuracy: 0.9375\n",
            "Epoch 418/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0413 - accuracy: 0.9852 - val_loss: 0.1599 - val_accuracy: 0.9375\n",
            "Epoch 419/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0408 - accuracy: 0.9844 - val_loss: 0.2025 - val_accuracy: 0.9375\n",
            "Epoch 420/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0412 - accuracy: 0.9835 - val_loss: 0.1554 - val_accuracy: 0.9375\n",
            "Epoch 421/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0413 - accuracy: 0.9866 - val_loss: 0.2399 - val_accuracy: 0.9375\n",
            "Epoch 422/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0392 - accuracy: 0.9861 - val_loss: 0.1864 - val_accuracy: 0.9375\n",
            "Epoch 423/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0421 - accuracy: 0.9858 - val_loss: 0.1930 - val_accuracy: 0.9375\n",
            "Epoch 424/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0407 - accuracy: 0.9846 - val_loss: 0.1576 - val_accuracy: 0.9375\n",
            "Epoch 425/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0379 - accuracy: 0.9881 - val_loss: 0.1466 - val_accuracy: 0.9375\n",
            "Epoch 426/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0409 - accuracy: 0.9849 - val_loss: 0.2376 - val_accuracy: 0.9375\n",
            "Epoch 427/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0400 - accuracy: 0.9858 - val_loss: 0.1816 - val_accuracy: 0.9375\n",
            "Epoch 428/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0373 - accuracy: 0.9869 - val_loss: 0.2367 - val_accuracy: 0.9375\n",
            "Epoch 429/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0397 - accuracy: 0.9858 - val_loss: 0.2158 - val_accuracy: 0.9375\n",
            "Epoch 430/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0394 - accuracy: 0.9869 - val_loss: 0.1676 - val_accuracy: 0.9375\n",
            "Epoch 431/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0368 - accuracy: 0.9875 - val_loss: 0.2358 - val_accuracy: 0.9375\n",
            "Epoch 432/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0370 - accuracy: 0.9883 - val_loss: 0.2544 - val_accuracy: 0.8750\n",
            "Epoch 433/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0389 - accuracy: 0.9863 - val_loss: 0.2042 - val_accuracy: 0.9375\n",
            "Epoch 434/500\n",
            "110/110 [==============================] - 5s 46ms/step - loss: 0.0405 - accuracy: 0.9866 - val_loss: 0.2360 - val_accuracy: 0.9375\n",
            "Epoch 435/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0373 - accuracy: 0.9892 - val_loss: 0.1718 - val_accuracy: 0.9375\n",
            "Epoch 436/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0393 - accuracy: 0.9863 - val_loss: 0.2120 - val_accuracy: 0.9375\n",
            "Epoch 437/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0395 - accuracy: 0.9872 - val_loss: 0.2012 - val_accuracy: 0.9375\n",
            "Epoch 438/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0361 - accuracy: 0.9883 - val_loss: 0.1606 - val_accuracy: 0.9375\n",
            "Epoch 439/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0399 - accuracy: 0.9863 - val_loss: 0.1556 - val_accuracy: 0.9375\n",
            "Epoch 440/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0387 - accuracy: 0.9872 - val_loss: 0.1994 - val_accuracy: 0.9375\n",
            "Epoch 441/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0369 - accuracy: 0.9886 - val_loss: 0.1683 - val_accuracy: 0.9375\n",
            "Epoch 442/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0370 - accuracy: 0.9886 - val_loss: 0.2339 - val_accuracy: 0.9375\n",
            "Epoch 443/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0378 - accuracy: 0.9875 - val_loss: 0.1701 - val_accuracy: 0.9375\n",
            "Epoch 444/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0385 - accuracy: 0.9861 - val_loss: 0.2889 - val_accuracy: 0.8750\n",
            "Epoch 445/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0365 - accuracy: 0.9889 - val_loss: 0.2002 - val_accuracy: 0.9375\n",
            "Epoch 446/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0380 - accuracy: 0.9881 - val_loss: 0.2001 - val_accuracy: 0.9375\n",
            "Epoch 447/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0367 - accuracy: 0.9878 - val_loss: 0.1811 - val_accuracy: 0.9375\n",
            "Epoch 448/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0356 - accuracy: 0.9892 - val_loss: 0.2117 - val_accuracy: 0.9375\n",
            "Epoch 449/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0370 - accuracy: 0.9886 - val_loss: 0.2034 - val_accuracy: 0.9375\n",
            "Epoch 450/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0380 - accuracy: 0.9863 - val_loss: 0.2078 - val_accuracy: 0.9375\n",
            "Epoch 451/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0357 - accuracy: 0.9883 - val_loss: 0.1665 - val_accuracy: 0.9375\n",
            "Epoch 452/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0377 - accuracy: 0.9863 - val_loss: 0.1553 - val_accuracy: 0.9375\n",
            "Epoch 453/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0384 - accuracy: 0.9863 - val_loss: 0.1649 - val_accuracy: 0.9375\n",
            "Epoch 454/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0374 - accuracy: 0.9878 - val_loss: 0.2048 - val_accuracy: 0.9375\n",
            "Epoch 455/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0363 - accuracy: 0.9895 - val_loss: 0.2115 - val_accuracy: 0.9375\n",
            "Epoch 456/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0358 - accuracy: 0.9883 - val_loss: 0.1818 - val_accuracy: 0.9375\n",
            "Epoch 457/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0367 - accuracy: 0.9866 - val_loss: 0.1658 - val_accuracy: 0.9375\n",
            "Epoch 458/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0379 - accuracy: 0.9875 - val_loss: 0.1547 - val_accuracy: 0.9375\n",
            "Epoch 459/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0369 - accuracy: 0.9878 - val_loss: 0.1561 - val_accuracy: 0.9375\n",
            "Epoch 460/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0363 - accuracy: 0.9883 - val_loss: 0.1993 - val_accuracy: 0.9375\n",
            "Epoch 461/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0363 - accuracy: 0.9892 - val_loss: 0.1738 - val_accuracy: 0.9375\n",
            "Epoch 462/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0370 - accuracy: 0.9881 - val_loss: 0.1462 - val_accuracy: 0.9375\n",
            "Epoch 463/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0368 - accuracy: 0.9869 - val_loss: 0.1839 - val_accuracy: 0.9375\n",
            "Epoch 464/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0335 - accuracy: 0.9892 - val_loss: 0.2239 - val_accuracy: 0.9375\n",
            "Epoch 465/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0366 - accuracy: 0.9872 - val_loss: 0.2294 - val_accuracy: 0.9375\n",
            "Epoch 466/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0366 - accuracy: 0.9866 - val_loss: 0.2536 - val_accuracy: 0.8750\n",
            "Epoch 467/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0360 - accuracy: 0.9892 - val_loss: 0.1676 - val_accuracy: 0.9375\n",
            "Epoch 468/500\n",
            "110/110 [==============================] - 5s 48ms/step - loss: 0.0363 - accuracy: 0.9892 - val_loss: 0.2121 - val_accuracy: 0.9375\n",
            "Epoch 469/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0365 - accuracy: 0.9886 - val_loss: 0.1580 - val_accuracy: 0.9375\n",
            "Epoch 470/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0365 - accuracy: 0.9883 - val_loss: 0.1644 - val_accuracy: 0.9375\n",
            "Epoch 471/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0355 - accuracy: 0.9872 - val_loss: 0.1889 - val_accuracy: 0.9375\n",
            "Epoch 472/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0341 - accuracy: 0.9900 - val_loss: 0.1840 - val_accuracy: 0.9375\n",
            "Epoch 473/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0358 - accuracy: 0.9883 - val_loss: 0.1852 - val_accuracy: 0.9375\n",
            "Epoch 474/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0345 - accuracy: 0.9875 - val_loss: 0.1412 - val_accuracy: 0.9375\n",
            "Epoch 475/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0365 - accuracy: 0.9858 - val_loss: 0.2455 - val_accuracy: 0.8750\n",
            "Epoch 476/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0370 - accuracy: 0.9872 - val_loss: 0.1925 - val_accuracy: 0.9375\n",
            "Epoch 477/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0355 - accuracy: 0.9898 - val_loss: 0.1893 - val_accuracy: 0.9375\n",
            "Epoch 478/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0340 - accuracy: 0.9881 - val_loss: 0.2098 - val_accuracy: 0.9375\n",
            "Epoch 479/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0361 - accuracy: 0.9872 - val_loss: 0.1845 - val_accuracy: 0.9375\n",
            "Epoch 480/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0349 - accuracy: 0.9886 - val_loss: 0.2184 - val_accuracy: 0.9375\n",
            "Epoch 481/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0319 - accuracy: 0.9906 - val_loss: 0.2216 - val_accuracy: 0.9375\n",
            "Epoch 482/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0358 - accuracy: 0.9883 - val_loss: 0.1972 - val_accuracy: 0.9375\n",
            "Epoch 483/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0313 - accuracy: 0.9909 - val_loss: 0.1676 - val_accuracy: 0.9375\n",
            "Epoch 484/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0334 - accuracy: 0.9892 - val_loss: 0.1827 - val_accuracy: 0.9375\n",
            "Epoch 485/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0327 - accuracy: 0.9886 - val_loss: 0.1746 - val_accuracy: 0.9375\n",
            "Epoch 486/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0334 - accuracy: 0.9886 - val_loss: 0.1727 - val_accuracy: 0.9375\n",
            "Epoch 487/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0353 - accuracy: 0.9875 - val_loss: 0.1320 - val_accuracy: 0.9375\n",
            "Epoch 488/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0344 - accuracy: 0.9881 - val_loss: 0.2286 - val_accuracy: 0.9375\n",
            "Epoch 489/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0332 - accuracy: 0.9895 - val_loss: 0.1771 - val_accuracy: 0.9375\n",
            "Epoch 490/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0342 - accuracy: 0.9889 - val_loss: 0.1919 - val_accuracy: 0.9375\n",
            "Epoch 491/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0350 - accuracy: 0.9886 - val_loss: 0.2491 - val_accuracy: 0.8750\n",
            "Epoch 492/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0334 - accuracy: 0.9892 - val_loss: 0.1583 - val_accuracy: 0.9375\n",
            "Epoch 493/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0327 - accuracy: 0.9889 - val_loss: 0.2076 - val_accuracy: 0.9375\n",
            "Epoch 494/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0352 - accuracy: 0.9866 - val_loss: 0.1883 - val_accuracy: 0.9375\n",
            "Epoch 495/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0360 - accuracy: 0.9881 - val_loss: 0.1812 - val_accuracy: 0.9375\n",
            "Epoch 496/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0330 - accuracy: 0.9903 - val_loss: 0.1607 - val_accuracy: 0.9375\n",
            "Epoch 497/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0331 - accuracy: 0.9892 - val_loss: 0.1804 - val_accuracy: 0.9375\n",
            "Epoch 498/500\n",
            "110/110 [==============================] - 5s 44ms/step - loss: 0.0337 - accuracy: 0.9886 - val_loss: 0.1605 - val_accuracy: 0.9375\n",
            "Epoch 499/500\n",
            "110/110 [==============================] - 5s 46ms/step - loss: 0.0330 - accuracy: 0.9883 - val_loss: 0.2295 - val_accuracy: 0.9375\n",
            "Epoch 500/500\n",
            "110/110 [==============================] - 5s 45ms/step - loss: 0.0327 - accuracy: 0.9892 - val_loss: 0.2031 - val_accuracy: 0.9375\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras import Model\n",
        "with tf.device('/device:GPU:0'):#ensure you are working with GPU !! \n",
        "  history = model.fit(x_train,y_train,epochs = 500 , validation_data = (x_val, y_val)) #choosing a large number of epochs because learning rate is too small \n",
        "  #you can change the hyperparameters to obtain better results"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As we can see from the results above : values of accuracy ( 0.98->98% ) and loss (0.03-> 3%) are satisfying so no need to retrain the model with other hyperparameters"
      ],
      "metadata": {
        "id": "P0bg91GDl0rn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.save(\"/content/drive/MyDrive/model\")\n",
        "#the model will be saved in the specified path as a .SavedModel file : it's a folder which have two folders( assets, variables) and .pb files "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9eRv-KE-wMiX",
        "outputId": "6599f5e2-0581-475c-8d33-4146f2a05c7f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/model/assets\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Evaluating results**"
      ],
      "metadata": {
        "id": "Ylxn1MDoGhHe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs_range = range(500)\n",
        "\n",
        "plt.figure(figsize=(15, 15))\n",
        "plt.subplot(2, 2, 1)\n",
        "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
        "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
        "plt.legend(loc='lower right')\n",
        "plt.title('Training and Validation Accuracy')\n",
        "\n",
        "plt.subplot(2, 2, 2)\n",
        "plt.plot(epochs_range, loss, label='Training Loss')\n",
        "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Training and Validation Loss')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "iOqd0gLNFm4c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "37e4a2af-1ece-4ccb-c68a-082c4907da87"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x1080 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2oAAAGiCAYAAACWDzX7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd3xUVfbAv29mMplMekgjEHqvBqkiLVRhQVBU7OK6ov6w/JTVn+7aQHFXV8V1d1UU+66uqDRBRbBQVARBQOkllCSkhySTNuX9/phkMpMpmUlCCpzv5wOZee++e887787ce+ace66iqqqKIAiCIAiCIAiC0GLQNLcAgiAIgiAIgiAIgitiqAmCIAiCIAiCILQwxFATBEEQBEEQBEFoYYihJgiCIAiCIAiC0MIQQ00QBEEQBEEQBKGFIYaaIAiCIAiCIAhCC0MMNSFgbrvtNlasWNHoZZuT1NRUvv/++0av98Ybb2T58uUArF69mltvvdWvsoGSkZFBSkoKVqu1XtcLgiAIzYuMrf4jY6twoSCG2gVCSkqK41+vXr0YMGCA4/3q1asDquuNN95g1qxZjV62JbJ06VKuv/56t+P5+fn069ePQ4cO+V3XjBkzePPNNxtFrtqDX1JSErt27UKr1TZK/bVRVZXx48czderUc1K/IAhCa0TG1vohYyv07NmTEydONHq9wvmFrrkFEJqGXbt2OV6npqby1FNPcckll7iVs1gs6HTSLaqZMWMGS5Ys4dSpUyQnJzuOr1u3jh49etCjR49mlK7p2L59O/n5+VgsFvbs2cOAAQOarG3pk4IgtFRkbK0fMrYKgn+IR+0CZ9u2bYwePZqlS5cycuRIHn74Yc6ePcu8efMYPnw4Q4YMYd68eZw5c8ZxjXMYwaeffsq1117LX//6V4YMGUJqairfffddvcqeOnWK66+/npSUFG655RaefPJJFixY4FFuf2RcsmQJc+bMISUlhVtvvZX8/HzH+ZUrVzJu3DiGDRvGK6+84lU/iYmJDB8+nFWrVrkcX7lyJZdffnmdcjhTff/VbN26lSlTpnDxxRezcOFCVFV1nDt58iQ33XQTw4YNY9iwYTzwwAMUFRUB8Mc//pGMjAzuuOMOUlJSeP311zl9+jQ9e/bEYrEAkJWVxR133MHQoUOZOHEiH330kaPul19+mXvvvZcHH3yQlJQUpk2bxt69e73qAGDFihWkpqYyZswYVq5c6XLu8OHDzJ07l6FDh3LJJZfw6quvAmC1Wnn11VeZMGECKSkpXHHFFWRmZrrJCu79ZM6cOSxevJhhw4bx8ssv+9QHQGZmJvPnz2f48OEMGzaMhQsXUllZydChQzl48KCjXF5eHgMHDnTpC4IgCI2NjK0ytvoztnqiuLiYBx98kOHDhzNu3Dj+9a9/YbPZADhx4gQ33HADF198McOGDeO+++4D7FEvixcvZsSIEQwaNIjp06cH5JUUWi5iqAnk5uZy9uxZvvnmGxYtWoTNZuOKK67gm2++4ZtvviE4OJiFCxd6vX7Pnj107tyZH3/8kdtuu40//elPLl+M/pZdsGABAwYMYNu2bcyfP9/tC9wZf2T87LPPeOaZZ/jhhx8wm82O0IgjR47w5JNP8uyzz7J582YKCwu9DgAAM2fOdAlhOXbsGAcOHGD69OkB66qa/Px85s+fz3333cePP/5Ihw4d2Llzp+O8qqrMmzePzZs38/nnn3PmzBlefvllAJ577jmSkpJ49dVX2bVrF3/4wx/c6r///vtJTExk8+bN/P3vf+eFF17ghx9+cJz/+uuvmTZtGjt27CA1NZVFixZ5lbWsrIwvv/ySGTNmMH36dNauXUtlZSUAJSUlzJ07l1GjRrF582bWr1/PiBEjAHjrrbdYu3YtS5cuZefOnSxevBiDwVCnbsDeT5KTk9m6dSt33nmnT31YrVbmzZtHUlISX3/9NZs2bWLq1Kno9XqmTp3q8uw+++wzRowYQUxMjF9yCIIg1BcZW2Vs9TW2emPRokUUFxezYcMG3nvvPVatWsUnn3wCwEsvvcTIkSPZvn07mzZt4oYbbgBgy5Yt7Nixgy+//JKff/6ZJUuWEBUVFXDbQstDDDUBjUbDPffcg16vx2AwEB0dzeTJkwkJCSEsLIw777yT7du3e70+KSmJq6++Gq1Wy6xZs8jJySE3NzegshkZGezdu9chx+DBg0lNTfXapj8yXnHFFXTu3BmDwcCUKVPYv38/AF988QVjx45lyJAh6PV67r33XjQa7x+FiRMnkpub6/iyX7VqFaNGjSImJiZgXVWzadMmunfvzpQpUwgKCuLmm28mNjbWcb5jx46MHDkSvV5PTEwMc+fO9atesHuXdu7cyYIFCwgODqZ3795cddVVLoPzxRdfzJgxY9BqtVx++eUcOHDAa33r169Hr9czcuRIxo4di8Vicfxa++233xIbG8utt95KcHAwYWFhDBw4EIDly5dz77330qVLFxRFoVevXkRHR/t1D/Hx8dx4443odDoMBoNPfezZs4fs7GwefPBBjEYjwcHBDB48GIBZs2axdu1ax4Rl1apVzJgxwy8ZBEEQGoKMrTK2+hpbPWG1Wlm3bh0PPPAAYWFhtG/fnrlz5zoMWp1OR0ZGBtnZ2S5jnU6nw2QycezYMVRVpWvXrsTHxwfUttAykYBpgejoaIKDgx3vy8rKeOaZZ9i8eTNnz54FwGQyYbVaPS6odf4SDAkJAaC0tNRjW97KFhQUEBkZ6TgG0LZtWzIzMz3W44+McXFxLm1Vy5SdnU1iYqLjnNFo9PnLU0hICFOmTGHlypWkpKSwZs0aHnroIb/l8ERtGRRFoW3bto73ubm5PP300+zYsQOTyYSqqkRERHitr3bdkZGRhIWFOY4lJSXx66+/Ot47PweDwUBFRYXXNRQrV67ksssuQ6fTodPpmDRpEitWrGDixIlkZmbSoUMHj3KcOXPG67m6cNYN+NZHZmYmSUlJHmUfOHAgBoOBbdu2ERcXx8mTJxk/fny9ZBIEQQgEGVtlbPU1tnqioKAAs9lMUlKSSxtZWVmAPTzzpZdeYvbs2URGRjJ37lxmz57NiBEjuP7661m4cCHp6elMmjSJhx56yEVWoXUiHjUBRVFc3r/55pscP36cjz76iJ07d/Lvf/8bwGvIRWMQFxfH2bNnKSsrcxzzNpA0VMb4+HiXcIyysjIKCwt9XjNr1iy++OILtm7dislkYty4cQ2SIy4uzkUGVVVd7veFF15AURTWrFnDzp07ee655/zWf3x8PGfPnqWkpMRxLDMzk4SEBL+ud+bMmTP8+OOPrF69mpEjRzJy5Ei+/PJLNm3aRH5+Pm3btuXUqVMer01MTOTkyZNux41GIwDl5eWOYzk5OS5lavdJX/qonnQ4r3lzZtasWaxevZrVq1czefJkl4mTIAjCuULGVhlbAyU6OpqgoCAyMjI8thEXF8dTTz3Fli1bePLJJ3nyyScdmSNvuukmPv30U9atW0daWhpvvPFGo8klNB9iqAlumEwmgoODiYiIoLCwkH/84x/nvM127drRr18/Xn75ZSorK9m1axfffPPNOZFx8uTJfPvtt+zYsYPKykr+/ve/OxbqemPw4MGEh4fz2GOPOdY/NUSOMWPGcPjwYdavX4/FYuHdd991CWkxmUwYjUbCw8PJyspy+8KNjY31aiC1bduWlJQUXnjhBSoqKjhw4AAff/xxvUL+Vq1aRadOnfjiiy9YuXIlK1eu5MsvvyQhIYG1a9cyduxYcnJyePvtt6msrKSkpITdu3cDcNVVV/HSSy+RlpaGqqocOHCAgoICYmJiSEhIYNWqVVitVj7++GOv9+KPPgYMGEBcXBzPP/88paWlVFRU8PPPPzvOz5gxgw0bNrB69WpmzpwZsA4EQRAaAxlb3blQx9ZqzGYzFRUVjn8AU6ZM4cUXX6SkpIT09HTeeustRxvV6+oAIiMjURQFjUbDnj172L17N2azmZCQEPR6vc+wU6H1IE9RcOPmm2+moqKC4cOHc8011zBq1Kgmafdvf/sbv/zyC8OGDWPJkiUuX9qNKWP37t157LHHWLBgAaNGjSIiIsIt1K42iqIwc+ZM0tPTXSb79ZUjJiaGl156ieeff55hw4Zx4sQJBg0a5Dg/f/589u3bx+DBg7n99tuZNGmSy/W33347r7zyCoMHD2bZsmVu9b/wwgukp6czatQo5s+fz9133+0xZXRdrFixguuuu464uDiXf3PmzGHFihWEhYXx5ptv8s033zBy5EgmT57Mtm3bAJg7dy6XXXYZt956K4MGDeJPf/qTYyBatGgRy5YtY9iwYRw5coSUlBSfcvjSh1ar5dVXX+XEiROMGzeO0aNH8/nnnzvOt23blj59+qAoiiOeXxAEoamRsdWdC3VsrWbatGkMGDDA8e/TTz/l0UcfJSQkhAkTJnDdddfxu9/9jiuvvBKAvXv3ctVVV5GSksKdd97Jn/70J5KTkzGZTPz5z39m6NChjBs3jqioKH7/+9/XWy6h5aCo59LnLggN4L777qNLly7cc889zS2K0Mp5+OGHiY+P53//93+bWxRBEIRmRcZWQWg9iEdNaDHs2bOHkydPYrPZ2LRpExs3bmTChAnNLZbQyjl9+jRfffUVs2fPbm5RBEEQmhwZWwWh9SJZH4UWQ25uLnfffTeFhYUkJibyxBNP0KdPn+YWS2jFLFmyhHfeeYfbb7+d5OTk5hZHEAShyZGxVRBaLxL6KAiCIAiCIAiC0MKQ0EdBEARBEARBEIQWhhhqgiAIgiAIgiAILYxmW6Nms9mwWhsWdanVKg2u43xE9OKO6MQd0YlnRC/uNIZOgoK0jSTNhYGMkecG0YlnRC/uiE7cEZ24c67Hx2Yz1KxWlcLC0gbVERVlbHAd5yOiF3dEJ+6ITjwjenGnMXQSFxfeSNJcGMgYeW4QnXhG9OKO6MQd0Yk753p8lNBHQRAEQRAEQRCEFoYYaoIgCIIgCIIgCC0MMdQEQRAEQRAEQRBaGLLhtSAIgiAIgiC0EqxWCwUFOVgslU3ablaWgmy/7EogOtHp9ERHx6HV+m9+iaEmCIIgCIIgCK2EgoIcDAYjoaGJKIrSZO1qtRqsVluTtdca8FcnqqpiMhVRUJBDbGxbv+uvM/Tx4YcfZsSIEfzud7/z2vBTTz3FxIkTmT59Or/99pvfjQuCIAiCIAiC4D8WSyWhoRFNaqQJDUNRFEJDIwL2gtZpqF1xxRW88cYbXs9v2rSJtLQ01q9fz6JFi3jiiScCEkAQBEEQBEEQBP8RI631UZ9nVqehNmTIECIjI72e37hxIzNnzkRRFC666CKKiorIzs4OWBBBEARBEARBEFo2Z88Wcsst13HLLdcxY8ZkZs68zPHebDb7vPbAgX0sWfJcnW3cccetjSLrzp07ePDB+xqlruagwWvUsrKySExMdLxPTEwkKyuL+Pj4hlYtCIIgCIIgCEILIjIyirff/g8Ay5a9RkiIkeuuu9Fx3mKxoNN5NjF69epDr1596mzj1VffbBxhWznNlkxEq1WIijI2sA5Ng+s4HxG9uCM6cUd04hnRizuiE0EQBMEXTz/9BHq9nkOHDjJgwEDGj5/ESy89T2VlBcHBBh555DE6dOjEzp07+PDD93n22SUsW/YaWVlnyMhIJysri6uvvparrpoDwMSJo/jqq83s3LmDN99cSlRUFMeOHaVnz9489tgiFEXhhx+28PLLL2IwhDBgwEAyMtJ59tklfsn71Vdf8N57b6GqKiNGXMpdd92D1WrlL39ZxIED+1AUhWnTZnDNNdezfPmHrFr1CVqtlk6dOvPkk8+cS1W60GBDLSEhgTNnzjjenzlzhoSEhDqvs1pVCgtLG9R2VJSxwXWcj4he3BGduCM68YzoxZ3G0ElcXHgjSSMIgiC0RHJysnn11TfRarWYTCX885+vo9Pp2L59G6+99k+efto95PHkyRP8/e+vUlpaynXXXcmsWbPdvHGHDx/kvfc+IjY2jjvv/D179uymV6/ePPfcM/zjH0tJSmrH448/4recubk5vPLKyyxb9j7h4eHcf/98Nm36lvj4BHJysnnvvY8AKC4uBuD9999m+fLV6PV6x7GmosGGWmpqKu+//z7Tpk1j9+7dhIeHS9ijIAiCIAiCIJxj1v6Wxepfz9RdMABm9EtkWt+6nS61GTduAlqtFoCSkhKeeuoJTp8+iaIoWCwWj9eMGDESvV6PXq8nOjqa/Pw84uNd2+7du6/jWPfuPThzJgOjMYSkpHYkJbUDYOLEyaxevcIvOffv/42UlIuJjo4GYNKkKezevZObb76NjIx0XnzxWUaMuJShQ4cD0LVrdxYu/DOjRo1l1KixAWqlYdRpqN1///389NNPFBQUMHr0aO6++26Hsq+99lrGjBnDd999x8SJEwkJCWHx4sXnXGhBEARBEARBEFoOBoPB8fqNN15l0KDBPPPM38jMzODuu+d5vCYoSO94rdFosFqtbmX0+rrLNAYRERG8/fYH/PTTD6xa9Qlff/0VjzzyOM89t4Tdu3exdesm3n33Td5550Ova/AamzpbeeGFF3yeVxSFxx9/vNEEEgRBEITWwqZNm3j66aex2WxcddVV3H777S7nFy9ezLZt2wAoLy8nLy+PHTt2NIeogiCch0zrm1Av79e5pqSkhLi4OADWrVvT6PV36NCRjIx0MjMzaNs2iY0bv/L72t69+7Fkyd8oLCwkPDycr75az+zZV1NYWEhQkI6xY8fToUNHFi58DJvNRnZ2FoMGDWbAgIvYsGE9ZWVlhIc3TTh/syUTEQRBEBqfkgoLhiAtOo3rfi27089yJNfEL+lF3D+2C9FGvcfrVVVlb2Yx/dqGo5F9enxitVpZuHAhb731FgkJCcyePZvU1FS6devmKPPIIzXrJt577z327dvXJLIZfn0XJS4ZEsY1SXuCIAjOXH/9TTz11BO8884yRoy4tNHrDw42cP/9D/HAA3djMITQu7f3TJI7dmxn1qypjveLFv2FO+6Yzz33zHMkExk1aiyHDx/imWeexGZTAZg373+w2WwsXPgoJlMJqqoye/acJjPSABRVVdUma80Js9kqyUTOEaIXd0Qn7ohO7BSWmimqsNAhOgSoWy+qqnK2zEKUMSigdg7nlHAo2+T45VNVVU4WlBEfHoyp0kq+qRIV2HGykCsGtiUkyB7n/+meTPJKKrltRAcUReHXzCIOZpcwrGM0W4/lM657LP/5OZ0O0QYm9Ypn+uvbGNQ+iudn9mXjoRxe/+EEf7+iP9OWbnPIMrprG/53bBfCgnX8dcMRBraLoMxspWOMkYLSSv6y4QjTq36lzSmpZM6ITpJMxAO7du3iH//4B8uWLQPgtddeA2DePM8hPnPmzOHuu+9m5MiRddbd0DEy6pPL0RpCyZv2n3rXcT4i33ueEb2405J1cubMCRITOzZ5u1qtBqvV1uTteqO0tBSj0Yiqqjz//F9JTk7mmmuub1IZAtWJp2fna3wUj5ogCBcUqqqybl82ZquN3/VN4Jp3dpBfauaFmX0Z1bUNABUWGz+fKmREp2h2nCrk51NnAZjZP5Etx/L568Yj3H5JR/4wwvXLdsfJQj7fn8WfJvVAAYrKLfxlw2FKKq38mFYAwL4zxezOKOKidhH8d1cGbSOC0WoUTheWO+pZ8t0xOsWE8MSUnjzz1WEACsrMaBT4764Mlzb/uyudU1XXPrPhCACbjuaxam8m/9qSRn6p2cVIqz6/6Wie4/2GQzluelrzWxZrfsuib2I4c0Z0ClTNFwS19xFNSEhgz549Hsump6dz+vRphg8f3iSyWSM6ojvzU5O0JQiC0BysWbOCzz9fi8Vipnv3nlx++ZXNLVKjI4aaIAiNgqqqKPUMlVNVlTxTJbFhwQAs/yWDGGMQVptKtDGIwclRlFts7EkvQqdVOFtmpqjcwkXtItl8LI9IQxCTesWxN7OI7OJKVFSijXqGd4ymoMzM8l8yWPtbFg9P7E5RuZknvjgIwNp9WeSXmgH4eHcGl3aJAeCVLWn8++fT3DmyE69sTXPIuWJPpqP8Oz+d4pqUJEorrTyz4TDjusXydJVRVW62sf6gu/ED8NEvdkPrYHYJ3eNCOZxjcpzrFBNCbFgwO04WkpZfxi3/+YUIg452kQaW/5KBVoHLesdTUmFh87F8AE4VljOpZxyzBrTlnk/3YraqDGofyVPr7bK0CdWTZ6rkst7xXNI5hv5J4Zwts/D6DyfYcbKQBald6R4Xhl6n4dp3fgbghZl9iTEG8WtmMZN7SRbfxmDt2rVMnjzZkRGtLhq616gmoTscWkFUmAZ0hrovuECQPQE9I3pxpyXrJCtLQavVNEvbzdWuJ6677kaXjbabi0B0oiiBfbeLoSYIFxgWq43XfzhBh7hwjBoY1z0Wi9XGN0fyGN8jFo2icNfyPRh0Gqb3S2R4p2i+OphDu0gDA5Mi0Gk1WG0qOSUVfLI7kysGtmVvRhHPf3OURyf3YHByFAVlZtpG1EwObaqKRlHYejyfkwVlJIQHM7RDFP/cfJySSiubj+ZhqrTyx9SuDEiK4NmNRwK+r0XrD9VZ5r5PfwUgOiSIge0i+PZIHhEGHVcObMtb204x9IXN9EwI42BWCQCvbE3DGKRlQWpXgrQa/rXlOACp3WP5+nAu1727k4vaRfD98QK+P17gaKfaSIsw6Hj5yv4s/f4EOo2CRqPQNiKYCosNq01lQWo3vtyfzbNfH+E/N11Mh+gQcksq+OPqfdhUKC4388C4bnSNNfLu9tPcOKQ9bSMMWKx2Q7BrbCgr9mQyZ1A7OsUYWX/nCEoqLBj1Wia98iOxoXo+uXUIJwtK6RRjJKhqMGkXaTfGKiw2DEE1hsMLM/uSWVTh8Cz2bRsR8HO4kKi9j2hWVpbXfUTXrVvHY4895nfdDd1rNDi4HRGoFJ88iDWme73rOd9oyeFszYnoxZ2WrBNVVZslBLGlhT62BALViaq6f7dL6KMgnOdYbHaP1JFcE8M7RrPxUA46rYahHaI4VVhG5xgjOq0Gs9XGml/P8Oa2U45rr0lJosJiY+Ve+4Tz3jFd2H6yEIDNx/K5pHO0wwi5tEsML8zsy6IvD7J2XzYA6w9kk1dqpsJi439X/AaAgj0T1ZyUdnyfls9/d2VwUbsINh7KdbQbrNNQYbGRULVGC+C5r48CoFHA5ufq2Wdn9CHjbDlLvjsGQEq7CI7nl1FYZnaU6R4Xyo1D2vPYOrsnrVMbIw+N70Z8WDCX90+kQ3QIP6YVcLKgDAUFg05D+6gQjuSamN4vgen97OFto7u24evDOaR2j2NMtzb8c/NxvjyQw4CkCGZf1Jbnvz5Kt7hQusWGcvPQZBRFITZUz5Ir+nmVf0b/RKb2TXAk/4gNC+at61Lcyj04viZBhU6rYWofu0HwfxNqJuFhwTrCgu1f65/dPgxjkJZgnYbucWFu9SmK4mKkAQ4DTfCP/v37k5aWxqlTp0hISGDt2rU8//zzbuWOHj1KUVERKSnuz/VcYY3sBIC26IQYaoIgCK0USSZyHiJ6cacl6+R0YRlHc00khhvoHh/qyLT39eFcDmWXcDTXxKKpvRyT6l2nz5JVXMHejCJ6xoex+Vge3x7J89UEl/WO59fMIsdapkDpGB3CiM4xfLgznVC91mFYVdM9LpRFU3vx4rdH2XaikN4JYRzJNWG2un69jOxsDy3cetwettevbThvXZdCWl4pbUL1fLgznQ93pfNgajfWH8xh09E8vr37EgrLzGgUhQiDDotVJdygY/vJQpKjQkiKNGC1qQx/cTMA2x8YjU1VqbTY+OlkIU98fpCXZ/enb2I4ZWYrj3y2n5uHJHNR+0gX2aq9flFRRgoKTCiKQvrZMmJDgwnWeQ5ryC2p4NmvjzK9bwKjurZpUPhnS6YxPj/nYzIRgO+++47FixdjtVq58sorufPOO3nppZfo168f48ePB+Dll1+moqKCBQsW+F1vQ8dIpbyA2GX9Kbn0CcoG3lbves43WvJY0JyIXtxpyTqRZCIth3OdTEQMtfMQ0Ys7dekkq7gCgITwYK9lKi025n+yl1uGJtMnMZx/bj7O7Zd05NuqkMEYL+nOAc6WmTlTXEF8mJ7lv2Sw9XgBtw3vwKiubbj6rR0cz7fL9sfUblwxsC0FpZVMfa0mAURcmJ4Ig445Ke0c66D8IaV9JAWllaTll3ktE6RV3Awqg07Do5N7YFNh1a9nuG14Bwa1j+T9HafZfrKQpEgDo7q2Qaco7Mks4vqL22PUaykqN/Pp7kyuqvLS/d+a/UQadIQEaSkqt/DirL4oikKFxcaWY3n0jA+jfVSIS9vVxk6FxUZxhYXYUO96debbw7notAqXdmmYV0g+P+6Iodb0NHiMVFVi3+xHefeZlIx+uvEEa+XI59szohd3WrJOxFBrOYih5oOW/CFqTkQv7njTSZnZyrMbj/DZb1kAPDKxO2VmKyM6xWCqtNApxsipwjJ6xYfx+f5sHv/8IFEhQcwdlsyL3x5z1BMWrCVMr6NjTAi9EsIJD9ax/WQBl/VO4NsjuWw+lo+1Vizf4A5RJITpHSGE1fROCKNnfJgjFNEbA5MimDusA+2iDLz+/QnHuqgBSREsu/YiCkvNRIboKCq3sOS7Y5RWWrHYVAYkRdA11khctJGzReX0bRvO2t+yGZAUTmKEgcJSM0a9lngfRqu/tEYPk3x+3BFDrelpjDGyzafTsQRFcHb6vxtJqtaPfL49I3pxpyXrpLkNtbvvnscNN9zCsGEjHOc++ug/nDx5ggULHvZ47fz5tzN//n306tWHBQvu4fHHn3bbj2zZstcICTH6TBCyadO3JCd3oHPnLgC88carDByYwpAhwxp0bzt37uDDD9/n2WeXBHSdpOcXLmh2ni4kOsTuUencxr8sOcXlFnRahZAgLaqqYqq08tvRXFbsOM3947pyJNfE5qN5VFhsfLAz3eXaxVXeqhc55nK8V3wYB7LtCSYKy8wuRtrUPvFUWmyYrSrbThSw7USh41z160u7xDgy990zugvH8kys/jXLUe7RST0cIYz7s0rYX5XMAuyZANPyyxjbrQ0n8su4KiWJLm2M9EkMd3QA0SEAACAASURBVOy19diUnvSMD2Nq3wTCq9YoVe/zFRkSxONTerrpyXkQujolqeZ4SGD7g/mitRlpgnBeEd0Z7emfm1sKQRDOMyZMmMzGjetdDLUNG9Zz1133+HX93/7293q3vXnzt1xyyaUOQ+222+6od12tATHUhCYjz1TJmaJyr5nkjuWZWLDyNzq3sSdiMAZpmfffmj2JNtw1gsiQIA7nlLAno4grB9qNi7T8Ur4/ns+RHBOTesWx6MtD5JoqWXrNQI7mmhx7SwH8eKLAEeboCa1G4fnL+7Lku6NkF1dSaravxTqQXUJsqJ4+ieEu+089NbUXk3vXpC8/WVDGi98eZcuxfJZc0Y+jOSbyS83cO6YziqJwx8hOhAXr2HaigNW/ZrFoai9Gdo4h3KAjtUcsN+WVsu9MMQeyS4gL05NdUsm8SzryypY07h3ThTZewgCDdRpuGprsx1MQBOFCQY3ujGb/KrCaQdt4P8AIgnBhM27ceF5//RXMZjNBQUFkZmaQm5vDwIEp/O1vz7B//z4qKioYN248v//9PLfrZ8+ezhtvvEdUVBTvvLOMzz9fS3R0NPHxCfTs2RuA1atXsHr1CsxmM+3bt+fRRxdx+PBBtmzZxC+/7OSdd97k6aef5e233+CSSy5l3LgJ7NjxE//85xKsVmuV5+5h9Ho9s2dP57LLfsfWrZuwWCwsWvRXOnbs5Ne9fvXVF7z33luoqsqIEZdy1133YLVa+ctfFnHgwD4URcO0adO55prrWb78Q1at+gStVkunTp158slnGqxrMdSEeqOqKmeKK1zSsDtTUmHhQFYJv6Sf5cYhydz9yV4O55hY84ehnCgoY2iHKPJL7Zv4RoUE8d+dGZwqLOdUYTk5JRUM6RDlUt/N/96FqdLqyOb38S+ZhARp2JtZ7Ciz5rcaL9UzGw5jDLJ38Sl9E/jityw3I21K73j+59JO5JWa+XxfFtP7JdIzPoyRXWL4NbOIuf/5hYfGd6N/UgQ94+2Z804XljFr2XZ6xYe5GGkAHaJDeO7yvuw7U8yApAhH8oxqqjPyDesYzdZ7L0XvlKQiLFhH/6QI+ie5G7ILp/byqGNBEASvhMajqFaUyiLUEMnoKQhC4xAREUmfPn358cetjBo1lg0b1pOaOhFFUbj99ruIiIjEarVy7713cuTIYbp185x59sCB/WzcuJ633/4PVquFW2+9wWGojRkzjhkzZgGwdOm/+OyzlcyePYdLLx3tMMycqaioYPHiJ1my5F906NCRRYseY+XKj7n66usAiIyM5M03/82nny7ngw/e4//+79E67zM3N4dXXnmZZcveJzw8nPvvn8+mTd8SH59ATk427733EVqthsLCswC8//7bLF++Gr1eT3FxcR21+4cYaoJPzhSVs+NUIT+dKOSxKT0dKcQBvjiQzWPrDrJkVj9Gdomhermjoij8mJbPi98e41iePbTux7QCx8a+01//CYBusaEcybUf6xEXysmCMsZ0bYNVVfn+eD6nCsvoEB3C8I7R7M8qdjHIAI7kmugRF8rQDlH8dLLQ5dzT03rxp7UHALhyYFv+MnsgfZ74ErNV5c6RnTicY2Joxygu6x2PIUhLYoSBvomuMcL92kaw5d5L3TL+tY8K4W+X96FHvHvKcwCdRmGAB2OrNnovmQQFQRAaAzXY/h2lVJaIoSYI5ynBBz7GsP/DRq2zvPccKnrN9llmwoTJbNiwnlGjxrJx43qH4fP111+xevUKrFYreXm5pKUd82qo7dmzi9Gjx2Ew2H/wv/TS0Y5zx44d5fXXX6GkpJiysjKGDh3uU56TJ0/Qtm0SHTrY139ddtnv+PTT5Q5DbcyYVAB69uzNd99944cWYP/+30hJuZjo6GgAJk2awu7dO7n55tvIyEjnxRefZeTIUQwebF8f17VrdxYu/DOjRo1l1KixfrVRF2KoCV4pLDM7jCqAn08V0iM+jLnDOrDlWB4nqjIJLvzyIBN7xvFD1T5U1WuqwJ7YorDMzO6MIrf6q400gEM5JsKDdSxI7Ur62XK2HMunpMLKQ+M7MqV3PK//cIK9mcXcP64r6YVlzOiXSGyY3pFpcdprP5JdUsn/TeiG1aYyqVc8FpvKsxuPMKFHHADv3jCI4nILKbXSsvvCW1r2Md1i/a5DEAShWdDbf3xSzKY6CgqCIATGpZeO4e9/f4GDBw9QXl5Or169ychI54MP3uf1198lIiKCp59+gsrKynrVv3jxkyxe/De6d+/BunVr2LWrYettg4Ls80V78g9Lg+qKiIjg7bc/4KeffmDlyk/YsGE9jzzyOM89t4Tdu3exdesm3n33Td5550N0uoaZWmKonecczyvlyS8O8sRlPekUY0/GYVNVjuaa6BYbiqIoWG0q+7OK6Z0QzsHsEr48kE3fxHCHRwrsGxAPSIpkw6EcthzLd2kjv9TMf3dlON5XG2n924bz8pX9sakqK/ee4Y0fTjC5Vzyf7snkxsHt6ZcUYd+hvcxMkEbD4A5RJEYYiA8PZtaARHacLGRMN/uvwDcPSaZjdAgTesY59hlz5sObB2O22VxS5E/tk8BlveMdCS26xYY2klYFQRBaAVUeNU1lMdY6igqC0Dqp6DW7Tu/XucBoNDJo0GCeeWYhEydOBsBkMmEwhBAWFkZ+fh4//vg9KSkXe61j4MBBLF78BDfeeAtWq5WtWzczY8YVAJSWmoiNjcVisbB+/efExcU72i0tdc/G2aFDRzIzMzh9+hTt2yfz5ZfruOiiQQ26x969+7Fkyd8oLCwkPDycr75az+zZV1NYWEhQkI6xY8fTqVNnnnjiz9hsNrKzsxg0aDADBlzEhg3rKSsrc8tsGShiqJ0nVFpsbqF0qqry0Op9HM8v5e1tJ8koqmBYxyjSC8tZ81sWU/vEY7WpbD9ZSH6pmc4xRk4UlGLzsGHDwxO6M3NAW0bvj+GX00WYKi1sPZ7PgnHdeHf7KUeI4x9Tu/Lc10cBePO6FMf1Nwxuz/UXt0NRFP4wogOxYd5Tv2sUhUcm9nBJ7a7XaZjUK97rNeEGz11Zsg4KgnDBoq8JfRQEQWhsJkyYzCOPLODJJxcD0L17D3r06Ml1180mISGB/v0H+ry+Z89epKZO5OabryM6Oppevfo4zt12253cfvstREVF0adPP4dxNn78JJ599mk+/vhDnnrqWUf54OBgHnnkcR599CFHMpGZM68M6H527NjOrFlTHe8XLfoLd9wxn3vumedIJjJq1FgOHz7EM888ic2moigwb97/YLPZWLjwUUymElRVZfbsOQ020kD2UWtVlJutGKrSsYPdENuVfha9VsNdy/dw9+guXHVRElFRRr75NZPb/7vb77pHd21DnqmSpEgD3eNC+XR3JneP7szILjF88HM6Nw5J9hoGCLA/q5hD2SVc3r8t3xzOJVSvZWjH6Abdb2NyofUVfxCdeEb04o7so9b0NMoYaTlF0GsjKJr0Tyq6X95IkrVu5PPtGdGLOy1ZJ829j5pQg+yjJgCQWVTO1W/tYM6gdtw2oiO/ZhZhU1XuWr7XUebZjUd4ZUsaf7miPwvX1YQtOifWePeGFLalFTCsUzQ3vb+Lsd3acMuwDm6JNOYO6+B4fduIur8MeieE0zvBXse47rJ+SxAEodkRj5ogCEKrRgy1VsCRHBOvfZ9GucXG2z+dYvkvGZgqrfROqMk6+ORlPXn884MUV1j4nw92AfDwhG4Mah9Fh5gQntlwmMHJUS4G1df/cwlGvRatRsIDBUEQzjskmYggCEKrRgy1FoKqquzPKmHLsTwSwoP5NbOYS7vE8P3xAj7dkwnAtL4JrP0tC1OlfVn4/qwSEsODWT53MIYgLW0jDPx0ooA3fjwJwGV9EgipCpX84o4RaGvZY97WdQmCIAitm3/vOE3neCOXA0pl4+znIwiCIDQtMlNvZqoTZrzx40mWfn/C5dzKvWcA6B4XystX9qdNqJ52EQayiiuY3DuOu5bvpXdiuGPdWkr7SFLaR9InOYq8wjKHkQbe08wLgiAI5x9fHsgmPjOEGTojSqV41AThfMM54ZrQOqhPWhAx1JoQq011hBlabSpZxRVc+87PxIQGkVlUwUXtIhjXPZa3t52if1IEm47mcXFyJH+Z3oeokCAA/nBJzXqx+aM6e9xYefqApBa7AFYQBEE490SFBJFvqsCmD0Mxyxo1QTif0On0mExFhIZGiLHWSlBVFZOpCJ1OX3dhJ8RQayKKys1c/+5OZg5IpKjcwoc70x1p8HuHhzEwKYL7xnYlKiSIOYPaUVpp5Y0fTjJ3WDKRVUZabW4emtyEdyAIgiC0FqKNQZzMKEMNDpVkIoJwnhEdHUdBQQ4lJYVN2q6iKPXyCp3PBKITnU5PdHRcQPWLoXYOOZJrIiJYx+FcE/vPFHOmuIJXt9rDG5MigskoqgDg1atd95nQKAphwTruG9ulyWUWBEEQWj92j5oZNTxcPGqCcJ6h1eqIjW3b5O225C0LmotzrRMx1M4Bx/NK+e+udD7ZnelyvE9iOPvO2Bd1//OqAXy0K4NeTpkbBUEQBKExiA4JosxsxaqPQFte0NziCIIgCPVADLVGZk9GEX/48BdHWCPAxJ5xHMk18czverPm1zMcyjHRPiqE+8d1bT5BBUEQhPOWGKN9HUSpIYHIM1uaWRpBEAShPoih1gioqsrHuzPpHGNk09E8bCr8dXpvuseFUWq20jO+xms2b2Sn5hNUEARBuCCIMtrXNhcHxRNtygarGbSe1zsLgiAILRMx1OrJkVwTH+1KZ2LPONqE6nl24xHHuUs6R5PaI7DFgoIgCILQWERXJaHK18XRERWNKQtbRPtmlkoQBEEIBDHU6sk/Nx9ny7F81h/I4Xd9EwCYOyyZo7mlXHdxu2aWThAEQbiQia7yqGVr7D8aakoyxFATBEFoZYihVg9ySir4/ng+qd1j2Xo8n//uyqBPYjh3Xdq5uUUTBEEQBMfem5lqGwC0JRlYmlMgQRAEIWDEUKsHXx/KxabCnSM7ceOQ9nx5IEe8aIIgCEKLIVSvJUirkGGNAEBTmt3MEgmCIAiBIoaanxzKLiHXVElSpIE3t52kcxsjndoYAejXNqKZpRMEQRCEGhRFISZUT1a53bOmmE3NLJEgCIIQKGKo+cGBrGLu+eRXCsrMAGgUWJDarZmlEgRBEATvxBj15JXbUHUGlErZ9FoQBKG1IYZaHWQWlXPj+7tcjv15Ug8m9pSsjoIgCELLJSZUT2FpJWpQKIq5tLnFEQRBEAJEDLU62HAwx/H63jFd6BEXypAOUc0okSAIgiDUTZtQPWm5JtTgMBSzeNQEQRBaG2Ko+aCw1MxXToZa38RwUtpHNqNEgiAIguAfMaF6CkrNqGFG8agJgiC0QsRQ88LhnBKue3cnALcN70BiRDAXtWv+pCFKeSHYqpIsKwqqIQYUxXNhVUUpzwdNEGqwB9kt5fZ1C1X1KBWFqMGRKBVnwWatKaczoGr1KBVFrtdrtKjBUShleYHdhFZvl8dcZl/gXt96XOSwy+21nqAQlNIyr1WoQaGA04J7nQFVH2bXj6XcZ/NqSBuUikIXnam6ENDqUCqKPbelgFLpeXG/GhKDUlFkb99cCtZKr2UcfcEXigbVEG3vC6paczw0tqoytV66V/VhoDOAakMpy/ddNigUgkJc2lKDw8FqQbF4fy7nhOr+4k3mOvpKoPjqCzXy1L//e+sLqj4MbFa/9OvW/53PBYcDxnrJJjQvMaF6Ss1WrDqjrFETBEFohYihVgubqrL5aB7r9tWkMp7RP5G2EYZz3nb0h5PQlKSTd9tvHs/rT3xN5Gc3uRwrGf5/lF08H+OOlwjd9hw5846gfWsGsdn7KU25g9DtL6IqGoouW0bkurkUTXiJip5XAhDzn7Foi08DYDXGoy3Ndvx1RtXoUPXhaMoL3GTyVL4uVBTOTn+PiC/uQFMVjlOfepyxhcShKcvxWU+sL5l09omoYrH/6qxqdJyd/m8i19yAYjP7bNujzrTBqNpgNJVFbuVrt+XtXmz6CJTKYhRUD2Vi0ZTl+pTLU521ZYyzVvhdR21URYuiWusuCPZkBrUMXltwJIq5DMXmboieazzpwxlffSVQfPWFahrS/731BZs+HMVaieLHM/bVJy1teqPesbVesgnNS7jBPsRbdKHozd77nyAIgtAyEUOtFt8ezuWhNfsBuHZQO64Y2LZJjDQAXd4+n+c1JZkAmIY9iC04krDvn0ZbdSxk9zLA/ou4JmNnVfkM+zHVRtDpLQAYDn7qMNQ0JRlUdhiD7sxOxySx+m/JJX9G1YWgLTiCce9bKOUFmNsOpbz75fY6bWbCtjyBtjQbW0gcpiH3+XWPmtJsQne8hC7vIBpzCeU9ZmE4tMJejyEG09AH/KqnGsVSTtj3ixyTbm1pNrbgKEzD/uhSzhiip7TMs0EQlP4DhqOfAVDRdRpWYzzGvW+hy/kVxWamrP/NWKJ7eLw2dNtzaEuzUVEoGf0UALrcXwnZ9wGKtYLK5NFUdJ5c01bmdgyHV9rb6jyZyuTRrnL+8hraopN2XVVN7E2D78VmjHcqsxRt0QnAbqir+nCfOgrb/Jjd8AuOxDTsQQD06d8TfHSto4w1NJHSi+/2WY8z+hNfE3xio+O9JbITZQP/4LFsUOZPGA6vcry3hcRiiemJPt0++S/rewOWNr39brshKNYKwrYutBv1YW0pHTTfrYyvvhIovvqCHZXwTX+u+hzFYhryvwHV760vBKetR3/yOwDKel+DJW6A1zqc+0J5199hbjfC5bw1pgehAUkltBRC9fYhvlITQrA5s5mlEQRBEAJFDDUnLDaV5b/YjZvU7rHcdWknDEHaZpbKGbtnpbz31dhCEwnd/oLjmI/iVa9tHs6rmOMvQpt3ECpdw7LK+16Pqg8nKP17jHvfAsAS25vy/jfbC1grCdvyBAA2Q3TN8TrQFh4jdMdLjl/5K5NHE3x4FYpqwxYS43c9DipNhH2/yPW2giPd6jFEGSkv9L5Go9pQq2x3CdbobnbjtErGis5TMCeP8nhdyO7X0VQUglbvaFN/7AtC9n0AgDn+IhdZVJ3BYaiZk4a5yRl8eLXDUKumoueVWKO61JQ5+pljcl7e51rUkDZe7wuw68diRQ2OcmpPdTHUbMb4gHSvqShyMdRsYe28X6/V1zLU2mCJH+Aw1Co7TaSy03i/224QljLCti60y+HlnuvqK4Hgqy9UE7bpURRUbIbA+7+3vqApzXYYapUdx1PZdaqPWmr6grn9JZT3u8lHWaE1ERpcZahpjShmE0pZvt2Y1wY1s2SCIAiCP2iaW4CWwpFcE0u+PcqOU2f5w4gO/HVGnxZmpOG0vkip+evJAHOctXl8XXNMtdfhcY2b/Ziq1Njyzq9RnLqOtzVyHlCVKp1Wh2MpWqiuV6mHvjXu16iaAH9/cK5Do4Xq66tl9NBGTXkPsju3X/ta5/ee7tfT/dQu5/Ic6tZZ9XNTXdqupSNf9+ipTrf78q5zt+ehaF2OudV1LnG+70D7SX3w1Rdql6mPHrz1BU0A9xlgfxJaD6HB9udZoYSgKc8n9s0BhG1+tJmlEgRBEPxFDDWgwmLj2nd+5r+7MuiTGM4fRnRsbpG8oFb972yo1fKoOb93fm2zeS6nKNQYfs4tVR3TeJn8uRhqAXSjqjoc65U0OkcbARtYtWXydcwXLhNVnUOOahnV2kaNy7XusqveJsy12vJ4v/4c8/ZMvFFdXvFlQAZq3LqW92ls1Zr8qxqd6zFf+m1snOT0+VwbCZ99wXG8If3f2+czAOPLWSdNYbwKTUZ16GOZEuL4Pgs+9kVziiQIgiAEgBhqwKHsmmxYF7ePRAnAQ9SkOIwr+2NTFQ148JQ5XeD0snbSB6e6PBla1ce8egRqdKQGZKjZ63AYQRon70p9JomeJqEBegVcvTs1RoRirTYmvdenevKGeDPaarXll/fMQ7mAvVEeZKw9IffYri98efm8te94r/XRr84xiqamvzZFuz76Qs3x+nuUvfUFNQDjq64+KbReqkMfy5SaddbWqM7NJY4gCIIQIBe8oZZZVM6a38443rePDmlGaeqgOsyx2pBUcPeoORuZLmGRtT1v1ec8e9SqD6mamrUMrqGPSo3XLQBDzVFftRGkOHlX6mWoKb6NIX+oNVF1yOjs9avrWm8hdb68YZ6MG4372hE3z49L/XWvNakJffQlV2BrVtTaa1y0AYQ+aoLcjeOmpFqfTR766NujVi95vPUFl9d11BtgfxJaD2FVoY9lqt7paAv9IVIQBEFw44KOcymttDL7ze1UWu1GzKwBiUzqGdfMUvnC1aNm/+srmYizR81b6GMdHjWfYXwau6cuEI9atbfKUr3+S+eYHNY7FE2jA6vV9X0AuBoNQU5ev4qaY77aBlStF8PDh0GkejJuPHk03LxfAa4p8uSxbOAaNbdwxgA8aqqbR62Jv4Y0OrBVNomB6LMvOI4HuZf1t34vfUENJDzWZR2qeNTOJ4xVoY8lTnk7671fpSAIgtDkXNCGWkZROZVWlUu7xHDD4PZcnBzV3CL5RHFLHKKg+Fqj5mTEuZXD2aPmsTX7H18TzWpDLZBfaKvXf1mrjSBdzaSynmFXqkbnuldUIyUTUZwTnnhr2xG25m3C7CPE0INx49e6teo1TYrWv0QuHgy12iGTgRoJ7l4y/9eooeia1VBTNTp7j22KtXE++kLN8YZ41Lz0hQA8lgEZdUKrIlRvf7Y/R0xg8NUjMez7kGCnDKyCIAhCy+aCDn3MOGsPbbt1WIcWb6S54pRMxM2j5iWZiBePmqooqL6yPnpZ92IvonH96weOOqzOHrUGhqIFktjCE7WTidTOTOkz9LE6EYQfSR1qnfM4gfZgFHrNsOinvlTnybwnGb2065OGZH3UaF1kaXIvTgN/GAisLR99ofbx+sjjrS+4/CBQVzKRZvRuCucUnVZDsE5DiUWDJa4/tpAY+3YiNktziyYIgiD4wQU3KquqSoXFxlPrD/HlAfsmyYkRwc0slZ+oHkIf6/CoqRodis3ibqjhlPXRZ+ijj4mmEvgateo6akIftTWZE+s7YQ8ksYUnvHnUHKGPfqTn92aA+UrP70eYo10QL6GP/urL05qs2tc2JFMmdTw7N0O6eT1qeFqzd47wK1FHdfhsPTx83vqCXyGX1TSn0Sycc4xBWsrM9tBwW0gsAEp5AaqxJYf5C4IgCHABGmpPf3WYVXvPuBxrE6r3UrqF4ZZMxH0fNZe901Rb1STMg6Hmsiebp2QintLz157EBe5Rc6z/stakvm9Q1kcCDMOr43q7EVEtY3Xoow9vUfXE1osHwz300beHxaNR6MXY89vQ8OBBcpOrwaGPdXsdHSjaWv2qqUMfPTyzc0VThT76+AzU5WEOyKgTWh1GvRZTZbWhVrUhelkeVjHUBEEQWjwXVOjj0VyTm5EGoGmp6fjd8LDhtVs2x1pr1KqNKF+Gms/QR++heqpjm4AA9FedHt3Zo+YpxX0gNHC9VW0jy2FMBbDhtdfJrpvnqg6PmpunykOylwD1pXry2DR66KMf6/ic2labM4FFQ0Nt69MWeNdxo4Q++vAq1+Wp88frJ3hl06ZNTJ48mYkTJ7J06VKPZdatW8fUqVOZNm0aDzzwQJPKZ9RrKa0y1FR9OABKZYmvSwRBEIQWwgX18+nRXJPjdaRBh06rcSy2bhU41pU5GUheDTD7a0fiBDfPm3MYpafQRw/JRNxCH+vhUQPQBDntUaarCX2sb2pwN7kC9ajVSmVelXq+Zq83P7xFXkMfA1yjVjsTpMc1a9XGoZ/68rT9gZtHLUDd1yrvT2bMmrLNHfroYV3hOcK/rI/Vob/1CH300hcC2v6gdtZTwW+sVisLFy7krbfeIiEhgdmzZ5Oamkq3bt0cZdLS0li6dCkffPABkZGR5OU1bdbFUL2W0sqqNWk6e5i/S/IlQRAEocVyQRlq1clD3r9hEDGhQUQagvxKmtdyqJ2p0ZPwtRKIVKfD97aPmuLuUVPxnD3O/Vd7Dyn8/UHR1qz/Upw2vK6nZ8Vrsg1/qR0S6NhCoO591DyuEfKVXbGO1Pp+TdYD9cB48Lo0OFzUbV2g/2vUcDPUmvbHkpr+1rQeNW+GocdN0/3FW18IwBB26QuyRi0g9uzZQ8eOHUlOTgZg2rRpbNy40cVQ++ijj7j++uuJjIwEoE2bNk0qo1GvpaDUDICqrTLUqr/bBEEQhBbNBRX6mH62nOiQIHomhBEXFoxepyFI24pUoDolAAE/kolQM0nzmkxE4x766OQhcwlRq/1rez09aqpG5xRWGNQIWR+9exP8u951Uuu43q+sjx48Gi7hjUEey7td40kWwKMx7lhj5d99evS6NDSZSABr1DwZq67ZRJvYi1Mtq49NuhuvLR99obY89dGDt74QiCHsKxuo4JOsrCwSExMd7xMSEsjKynIpk5aWxvHjx5kzZw5XX301mzZtalIZjUG6mtBHncF+0CqGmiAIQmvgghqVM86WkxRpaG4x6o3iZqh5CH10SiaCaqvxfLiFSDq/r22oOXvUfKypchhqAbolNdqa0BunNWr1DkVroEfNdb2UezIRvzZzdjY8fHnN6ppA+6EDNVB9efJYuq03bNgaNZ/GsafNtQPdtLsR8bhm71y15c99NqD/e+sLrtsfBBL6eEENCU2C1WrlxIkTvPfee5w5c4YbbriBNWvWEBER4fM6rVYhKsrYoLa1Wg3R4cGUZxbZ67LZt6EJDQZjA+turWi1mgbr9XxE9OKO6MQd0Yk751onF9SofPpsOX0Tw5tbjAbgnkzEPaSxVjIRjcbDcWc8ZX30EvroKdEFoAbqmFV0KOYi+7UaJ+9KPSfOtSeiAU/Aaxujji0Eyt3P127bU8Y+n9kVfXsv/At9DDBrocd1dI1n3NYpi6e2mnONWpMmE/Ej66PSgP7v51nloAAAIABJREFUrS8EFProOWOpUDcJCQmcOVOToCorK4uEhAS3MgMHDiQoKIjk5GQ6depEWloaAwYM8Fm31apSWFjaIPmiooxoVZWSCguFhaVoTCptgLKis5Q3sO7WSlSUscF6PR8RvbgjOnFHdOJOY+gkLs67bdKK4v4aRmGpmYyz5XSPC21uURqAh33U6trwunryp1prFatJTKLWDl10fu8cBunmUavHPmpV9bikvve0z1cgBJCB0PP1tSa1Vfej+BP6WL0nlzdvlVt2xVreOzdZ/JDd4SHzT+/V7bhOyBu2Rq0hG17b1yX6nz6+0XEkV2mCdv1KJtKANWre+kJd2UWd8ZUNVPBJ//79SUtL49SpU1RWVrJ27VpSU1NdykyYMIGffvoJgPz8fNLS0hxr2pqC6qyPqqo6Qh+1eQeI+2d7dNm7m0wOQRAEIXAuiFE5La+Uq97eAUBKu8hmlqYBqLWTiTgfqzqjuoY+ek/P7z2ZiFfcjI76Zn2s5X1y7AVVz4lzA9dbuYaJaUFRHBuFq3jZELx2WxovBlhjbHhdW976rifzJVfAXkgfnsLaeFo/FUj6+MamuUIfvW547WUvNH/q9+qlq+MHAQ/t2+sTj1og6HQ6HnvsMW677TasVitXXnkl3bt356WXXqJfv36MHz+eUaNGsXXrVqZOnYpWq+XBBx8kOjq6yWQMDdKiAmVmG8aqZCKGw6sACD68Gkv8wCaTRRAEQQiMC8JQ23S0Jh1y79Yc+uhPMhEnFFTHxEvxlkzEU+ijN8PELSyqAclEnF7XZL2rb3p+1/YDX29VKz1/9V+bxf+wMS8hjW5hmXV4WPyarDvK+Gtgu8voK527P7glAPHx7Dxujq31w4A5Rzg8yE3sUfNmMHnMHBpw/YqX49SZNMU1EY6k5w+UMWPGMGbMGJdj9957r+O1oig8/PDDPPzww00tGmD3qAGUVlowGquyPpYXAE7JRQRBEIQWyQUR+ngox76558KpPQnWteZb9rDhtVvWR5treUfoo5f91jxlffRiALhNuOudTKRWgoXGDkVrQHp+tz2t/Awb8+o58REi6NGgdJvMezDEHXV4N9JdanCEPvry9AWoswak51c1ulpJNprpM9nEa9S8e9QaEPrrrS/4WFvqXkcAYZJCq6PaUDNVWqnOalv9w52qC2lO0QRBEIQ68GuGtGnTJiZPnszEiRNZunSp2/n09HRuvvlmpk+fzo033uiyuLolsD+rhLHd2nBZ74S6C7dknMMVwb7hta9kIqpTMpFa5RSXPdlq7aPmzfBqrNDH2uu5GjsULdA9wTx5PfzdhNiTR81ZH246qyMVuj9ZH+ubodFHJsAGbWlQ1/VuYZbaFmEQBKzH+uCrLziONyD00cs9uO6N5vvzGVCYpNDqCK0y1MrMVSn6q8IfAdQgyd4mCILQkqlzhm21Wlm4cCFvvPEGa9eu5bPPPuPIkSMuZf76178yc+ZM1qxZw1133cXzzz9/zgQOlJIKCycLyuid0IpDHp1Q3RJ9+E4motbpUfOwBsvbxM4tKUSVwRigY9Z76GN916jVMjQDDd9yCRMLcj1WZ+hjdXnPbbpnffQQZumjvOd91AI1qjzI6MuA9AP3rJE+MmN6WqPWnEkrqvtLE4f51Zn1sUEetVo4P8+6PN6B7LkmtDpcPGoAToaa3+HTgiAIQrNQ5wx7z549dOzYkeTkZPR6PdOmTWPjxo0uZY4ePcrw4cMBGD58uNv55uRAlj3ssXdiWDNL0gioNlwHVg/7qNVOJlI98fJpqDU09LEhyUQaIeNc7fDPQD0lLpPa2qGP/nnUvHpnfGRX9Dhx98ejEbDH0EOylgaHPnrPZumGhz3XmsSbVRdNnm2yDh3Xx5tVVzhloHVI1sfzDqPe/kzdNr0GFGtls8gkCIIg+EedM+ysrCwSExMd7xMSEsjKynIp06tXL9avXw/AV199hclkoqCgoJFFrR/7s4oB6B3f+j1qiqq6GlWKhzVq1NpHrTrVfEDJRLz8yuq2Lqmea9Q8GEbQiKFoASfG0Lm/9jcTpSOjorfzAe5v5VfoY6CT6erkGb6SnDQsPb/v0EcP3rsWYBA0+ebOdei4Xhtee0tQEkBdEvp4fhMaVJ1MxD30UQw1QRCElk2jjMoPPvggixYtYsWKFQwePJiEhAS0Wt8TBa1WafBO3nXtBl5hsfHtsXzaR4XQKan1pOX3dk8agw6UmnvW6rRodXY9Khq7pRAeVjMI63QaUPT2slWPQxdUdb1i/1U1JNSAJsi1Gygaz3oNiwwDp+PVz1gfrA/oWWr1dhlVRUtUdCjaqolESKiR4Hr0CW2tBDEGYwj6WvX47CvWmvuPig4HXTAanV1vGm2Qz3vTGO2L8YP1OnQeyoVHhbvozJnIqHAwuJ5Twmrt86e49wclPLTue3KWMdh+fy56sbh+Po2hoRgC0X1QhOv14aGEeLu+1o8JIWFGgqNqfjhp6PdAoGh19ns3hho93rO/eg0Ub32hpv+HBNz/vfYFpeb51HkvTn0hKiaiJvzXWcZzpBPh3OMIfaxao4bOKfSxeq9IQRAEoUVSp6GWkJDgkhwkKyuLhIQEtzL/+Mc/ADCZTKxfv56ICNeJXG2sVrXBO3nXtRv4un1Z7Dl9loVTe7aKndTjqv56kzW0rIIQFMf5KKuKarZytrCUNjYVBSguKiOmqrzFbAZAD1jNZjSAxWzjbGEpmiITbYDSUjMGi4reqR1VdZWhWq5ikxmr0/Fom70DVZptFAeg30ibYm9Po6OwsJRws/X/2XvzeEnK+t7/81RVr2c/s5xhmRlFJqjMgLijxlEQhzWYIBpjcEWjMQZzSYzxGq5yb0RjiD9yf7kmXAwqoqjElXFncYyyiIqDgAKDwwyznFnPnK23qnruH7V0rd3VffosVefzfr2U7q6nnnq6uud0fer7/X6+KAKo1ExUuvichnUTCuD2Pouap+V3RZrNcz9ZBxQDI1CgATChtvzulGom+gHU6jqmI87Z5FQDJvz7u8eaqgNV/7ZC1YDvX44Mfx8KFWtM0n9D/TUdJQTOr2m46wCA2ZqBagfnXlTrWOl5PlMxUG+xv+9YVRP16QZW2M8X+t/mkG4gD2C2Gv2e2/1d6ZRW3wUAze9/tfPvf9x3QZmpJz+/nu/CxGQNEI3QkF6ck1Wr0p/VkEaa9vxORM2T+mgyokYIIUuZtqmPmzZtws6dO7F7927U63Vs3boVZ511lm/MkSNHYJpWat3111+PSy65ZH5W2yFPHq1AFcA5p6xe7KX0iGAtVthMRPjMRKwxMrKWzVujFjhMbB+1GAOKjl0Io/tGzTkVrVtThpBBC5q2+0lTH2O3t7KtT2LPH6brpsQtnAA7TnkL9UbrYE3BhteLxNJLfezC9TFmzo4+z6jvP8kMJTf1UQcASF9ErQ4YNaiHHl6MpRFCCGlD219lTdNw1VVX4fLLL8f555+P8847Dxs2bMB1113nmobcd999OPfcc7FlyxYcOnQI7373u+d94UnYd6yK1QMFaEpGnK2kDFxItTYTsUSbYu1j6sHJ7EERro9xBVfBuiR7v1g7/xicWrSwOUlvhFrXQgbwOAI6IrSN66MjDmNamrW8+I6YO7z2iInd/ZL1UWt+1i2cADs1KOnETCSIssj2/DLifCzEYeO+C856ujkncd+FTubqtMaUpApVEShqStP10fP3Xhh1DH7/LzD6pVdD1KegjT+AwmPfQvneT1gizkPpVzeg9ItPLeTSCSFk2ZPo6mrz5s3YvHmz77UrrrjCfXzuuefi3HPP7e3KesDeY1UcP1RsPzAtSAmfiGpnJiJNV4gJafhGNc1FIsxEFsr1sQOL9yRIRbNW3oNISeKWAe2O1UoMRJ23JGvv8v11ZPjRdg3df3ZSaAsfzYpiodfQThh25foY15ttCZxfsmQo51U39VGd2tvcYNRQeOI7AABRm8LIrRc2Nw09HbVnvtZ93v9fHwYAVJ67NG7EEkLIciDTeS77Jqs4bjBDQg3S10fN6l/WouG14/ooFMD0CzUXZ7t3r4R91Jp34jsVak7ky29aMOeL94jGzl0jotcYf8zozS33j4pkhNYeHtN5xFDEzO2Zs9v0VXdNHUbUFtOe3+n/t8BiJva74KxHXaTUR5J5+jxCTZnc5b4u9KaZiGhM+/aRudLCLI4QQkgsmRVq1YaBg9P1jEXUzIA9P9r3UYPdJy2Y+tgqohYnOkLpbt1F1Jo9yoJ2/3ONqPWwgXHiJtztmgl32vMswQX2fFyEz6WuD+jss1vshtfedSzo8eYhoha3DxtXEw/lvIZZ2/XRqWM2BtdBPfq4O0bU/UINGoUaIYQsNpkVaj/deRQSwMbjsuQ0Fkx9jIio+VIfpT1cCfdRa9XwOnFEbY6pj3HCr1t6eOHdTH2ca5SvO5MO2UoAupGXHtYWzVHkdiKSpVCXRurjQkf12qbJdvH9j/suLIXzS5YM5bzq1qhNbf4o6ie+DGZ+ALlDv3bHKNUj/p3YY40QQhadzAq1H/zmAEbLObxg3chiL6V3BBteQ1hNsINjmk8goVipjHFmIlAiUvCSmYl03fDaESO9vuvfy4tTZ41zvZjvNqWwxXuZj7S2OX8WnZpXLAEhseBicR6EYex3ge6NxENfXkXFFmrVjW/CsYtvATyNrwFAPbrD9zy2GXbwph8hhJB5I7O/5jsOz+K04wez4/gIQMBOZfS80j6iZrs6BsxEnB9by7ExoZlI4KJQwnF97DD10bmo7/GF8pxFlZderbFTR8wkKZfzkdY213PXcerkEkjNW+g1zIe7IlMcSQLKOdVNfXSQat73XJ143PfcF1EzvLVsMz1fHyGEkGgyK9QOTNWwur/QfmCakAjUqClh10fvc2nArVEL3QWVnkdJUx+DETXHpKLDr5Ej+Hp9odzDi9bYOrr5xjExaRU1m5dI0NxERMdRvqVgCb8U1jBXlkBkkix9vKmPLnZETR/ZAADQjjzq2yy84qw+43kcqGUjhBAyb2RSqM3UdczUDaweyJhQc1wcHSIEmBV1sx9Lw3V1FCEzEadGLez6GHsBG1ej1rHro5P6GLzITNoXbAFwUx8TXgj3aumhSF544mbksMM+ai3Hz/ENdCJoQy0lFpjFPn6QOayn8+8CWY5Y9vz+3wAnoqavfDYAILf/5zBzfTj8lp9bAzwRNa8jJCNqhBCycGRSqB2csn5gVg/k24xMGTJB6qP3os90XB+jUh9bmInERVeCEbAua9TmK/WxpyzSGp100pbHXYrnbSmuaTnA804SUM6pqDRMmN7fB/tvnDG43k1f19c8HzJXBgDkd2+DcmwnAH8UTdSnFmbRhBBCsinUDkxbKRvZS33091GLTGn02fMbtogK91ETXjORwNdAxgmvOHfIjl0fbYfA0EXmXFPRepfKJmPXOM+Hdnt8OaI4amLRYlvkpAnGz/ENdPIdWOyUw8U+fpA5rafT7wJZjpTz1t+TWU/6o6hZgsvsX+O6AjeOf6EbaSvs/AFWfP5l1tiGN/WRETVCCFkoKNRShT+i1tLCHXD7rkmhWGmQwW1AZxG1IN3WqCkJ6rAWG7fh9SKtcSmfG0JIquiLEGpK9SgAwCyPua81Vp0GKOFMFF9ErcEaNUIIWSgyKdR2H61AFcCq/qylPkq/hmpjJiK8Da9jzUQiRFlC4dW16+NiGXV0gjJPhiedHp8QQuZIOW/9PfE6P4qaLdT6mkLNWPls6+ZeoC+iP6LmEW21Y4BenZc1E0IIyahQe3DfFE5e1Y9ibgkLga5obyYCBFMfbbOQXpiJBOk69dH+XJaCRXsMi11H16n4JYSQOEo56+9JpRERUfMINbO82n7kuQGoV6HERNRW3nAqRm79g3lYMSGEECCDQs0wJR7aN4VNxw0s9lJ6jtXcugMzEWla6ZFCjU99BNqmPupDT49ZUHdCTeb67f9aRev6qk0AAGNwbUfzODTWPA8AUF+32Zpn6GldzePFWaOZ62s5zhhaDwDQV57qe73dfrHHLQ4DABonvAQAUF/7++ExhUFrzPEvTDSnvupU31qjMPtWx25LgplP/u/N7D+h+bgwNKfjdoM+9lzr2OW5veektPsu6KtOAwAYAye0HBdFu++CVLOW/k26oWTftPQKNefvpFlehdrTt6B+/Ivc3wKvS7A6tccnzpTaMd/c2uGH52vZhBCy7MlcftXeY1XMNgw8a032hJrdSK35VCi2eAuOcR4azYbXkXPB/mFu3Udt4tKtELWJiDm6q1GrbLwMxsgzXIFWec47UV/7civtpgtmzvw7VJ95KYzRU1B95uu6mufQW37hez77/CvQOOFMNI57Qcv9Gie8BEde/30YK57le/3Im+6BaMxGH+utD0BIPXKb2X88jrzhDhhDT0PljD+D0XdceEzfmDsmCdVTL0NjzfND56VxxSOYnJiGqE3AGP29RHN5OfzmnwGwXOBkeVXLsYfe+gCE2YCoHnXXcfhN97pifSGZedHfoPp7fwhjdMOCHK/VdwEAKmf8GerrNnf1vW31XTj8pvsgc6VE8wS//yRbOGYilXrzBt3ERV+weqcpGibP/3Tsvrl997qpj2auH8rM+PwulhBCiEvmhNpktQEAGC3n2oxMIVJGiKJAHzVPpEyYJiCUeBdHWCl27dLsZGHQvXMfSadperky6k97lW//bkUaAEDRYKx4JgB0PY8MRJRkcRj1p7860b5Rx5TFEcjiSPSxyitbdr1yRJMxfFLbMYkQIvq89I/B1AeA/rAYTILp7td+f/c9e45ldhFB6gme78tC0Oq7AGDO3/+474I5cHziOYLff5ItnDIAb42a7FuNRoLPve+nH0X1Wa+HVAswB0+kUCOEkAUkc6mPUzUrUjFQyJwGdV0cXdqYibi906KElCvo2kfU2pO5rxEhhGSGckTqYxLqJ7wUSm0Cysx+yFwfzL4xKDP752OJhBBCIsjcFfZ0zfoh6s+iUAulPoqwUAulPkaYhQABM5Eu7fmdqZZaXypCCCEurlCrdybUjMETAQDq5C7IfD+M8ppmRC3020MIIaTXZE6oORG1TAo1aQZEUTszEUfYhT9mgfiIWsfCiw6FhBCyZCnaro+zHUbUzAHL4Emd3GVF1PrXQJk9CJgGYDZ6vk5CCCF+MneFPZ3l1EcAvobXQkFIqEWaiUQIL+kxEwlu71R4UagRQsiSpaApUARQ7VCoORE1pXIYMt8Ps28MQhpQKocAg0KNEELmm8xdYU/VdKii2TcmU4TMRCL6qHnNRFqlPnpdH0PbGVEjhJCsIIRAKadithHsu9kaJ6IGADLXB1mwTHFE7RiEWe/pGgkhhITJ3BX2dM1Af0GDyGLdVMhMpE2NmmkCENGuju5+UWYiFGqEEJIlynm18xq1gRPdx2auH1Kz+vIJowaY0S1GCCGE9I7MXWFP1fT016fFFGmLoJkIhM+O39rX+9xpeN3KTCQqNbJTkZtBUUwIIRmilFM7dn00S6OQWhEAIPN9zQbqeg3Ck/pYvv+6nq2TEEJIk8wJteksCLXYLluB1MdWtWeAVfDdpuG1hAIZ/BokjqjZczCiRgghSxor9bEzoQa1CLO4AgAg8wOAL6LWTH3su/cTgFHr2VoJIYRYZO4Ke7qmY6CgLvYy5kYwShb3eqSZiGczpC26ogSdPZcQ4c00EyGEkExRzikdR9QghNs4XeaaETWhV30RNQAQtamerJMQQkiTzF1hZzn1MYmZSCgVEu3MRKK2s0aNEEKyRDGnopLQTKT2tFe5jxsrN1oPhOKmQcKohuz5lfpkT9ZJCCGkScoVTZhK3UA5n/KIWlyUTHbY8NoZEyGk/IJurmYirFEjhJClTDmvYv9ksvTEyfM+DUjLLERffRoAQJ3aA7gRtRqE4Xd9FHVG1AghpNdkT6g1TBS1jAo1SF8zaqu2rEXDawCW62PE+Wjh+hiqWYvF3i9qfkIIIUuGjmrUFBWA9Xe9dvJFqO75KWaf9xeQqhVRs2rUmPpICCHzTeaEWlU3UEx7D7W41MeQmQha9lGzxiitI15RqY8dBshoJkIIIUubUk7tuOE1AEArYursTwIAROWw9ZoRFVFj6iMhhPSaTF1hSylRbZgo5tIe4YlLfbT6orlEmokEhRqia8h8ZiLB1EeaiRBCSJboyvUxiMdMJBRRY+ojIYT0nExdYdd0ExJASUv524oNqMmAqIqoUYtIfWxpJhLV8LpjMxHWqBFCyFKmnFfQMCR0I5mhSBSu66NRgwg0vFYo1AghpOekXNH4qdqOVlmNqAkEImoQEAFhJkz/HVMZ6eoIT0StTWpkIjL1NSKEkMxRsn8Xkzo/RqLmrJpnvQbQTIQQQuadTF1hV3VLpJSyWqMm4RdVUamPMpjaEpHa6E4GSAjITiNoQZj6SAhZpmzbtg1btmzBOeecg+uvvz60/atf/Spe/OIX4+KLL8bFF1+Mr3zlK4uwyqZQ60X6oxVRo5kIIYTMN5kyE3HuFJZSHlETkDExtaCZSLiPWpSZSKSLoyMGY+z7O4JCjRCyDDEMA1dffTVuvPFGjI2N4bWvfS3OOussnHzyyb5x559/Pq666qpFWqVF2Ymo1ecm1KRWgDCqQLDhNc1ECCGk52TqCrti3yksZNWeP2gmYtk+RoyBf0ynZiIdQtdHQshyZPv27Vi/fj3Wrl2LfD6PCy64ALfffvtiLyuSYo8ialItAHoVwvSnPrJGjRBCek+mrrCzn/oYMBOJjKgFfoTbRsyizEQ6hGYihJBlyPj4ONasWeM+Hxsbw/j4eGjc97//fVx00UX4y7/8S+zbt28hl+hSzlu/A5VepD7qtVBEDXoFkCaGvv465Hb9aG7HIIQQAiCjqY9ZNRMBpC+NMaq2TIRSH6OFmmtCEmc20gmMqBFCSCSvfOUrceGFFyKfz+OWW27B3/7t3+Jzn/tc2/1UVWB4uDynY6uq4s6xetqKgCl5bU7zKoUS8oqOXMH/ek4xMVxsILfnp8gdfgj6lb/r+hjzifeckCY8L2F4TsLwnISZ73OSKaFWa2Q9omYGMh+V8FgzqZmI2dw+ZxhRI4QsP8bGxrB//373+fj4OMbGxnxjRkZG3MeXXnopPvGJTySa2zAkJiZm57S+4eGyO4detSJgB4/OzmneYeSh7roX6m9v870uD+2A/oOrkQMgqsegXnsyDr99e9fHmS+854Q04XkJw3MShuckTC/OyapVA7HbUq5o/GTFTKRVRM3f8DqJmUgwombP7TMTmWMfNQo1QsgyZNOmTdi5cyd2796Ner2OrVu34qyzzvKNOXDggPv4jjvuwDOe8YyFXiYAoJy3fherc7HnBwCtCKVyMPSyOrUbpQc/6z5XqkfmdhxCCCHZiqg5NWrFtDe8jkFICal4RWiUmUgwohZwfXSEnK+PWvB8xQnFIEnHEUJI9tA0DVdddRUuv/xyGIaBSy65BBs2bMB1112HjRs34uyzz8ZNN92EO+64A6qqYmhoCNdcc82irLVX9vxO02tCCCHzT6aEWmZq1OJSH0MRNSXc8DoQUZPBiJo73rsfI2KEENINmzdvxubNm32vXXHFFe7jK6+8EldeeeVCLytEs+H13O35CSGELAyZCj05P0CpF2qx9vwyIvrVzvVRCaQ2+oWaFArt9QkhJOPkVQFV9Mb1kRBCyMKQqSv0asNEThXQlJRHiFqaifgjau3NRAAIj3B1Ux+d/eZiz5/y80wIIcsEIQRKeRWzc2x4bZZW9GhFhBBC2pExoWagmNZm1z7B1YmZiMcUJHLfhTATIYQQstQp5dQ5R9TMvqarZfWZl851SYQQQlqQLaGmGym25vcIrKQNr71mIvY+IhhRC5qFyOADBWFhRpMQQgjJGpZQm5vro1m2hJpZHEF93St6sCpCCCFxpFXVRFJpmOmtT/OIM9Gy4XVTVFm1ZW1cH4WwDEXcua0fadd0JDKiRgghJGuUexBRM/rXAACEXgm3gwkSa4xFCCEkCZkSalbqY1rfUoLUx6iIWrs+asHUx5DrY1SNGoUbIYRkjVJOmXuNmp36KPRqeyFm6nM6FiGELHfSqmoiqehmeptdyySpj6ZfdHlr1LxjvAgFaNlHrSnkmtE63gUlhJCsUcr3okZtjfdZm8EUaoQQMhcyJdRqDQPFLNSoJTUTgQinSYYaXseZiTibPamPTIEkhJDM0ovUR1kYBgDU121uG1ETkkKNEELmQuYaXq/oyy/2MrpERj70IoJ91Bxh5a1vCwi1dg2vJTwRN6G0rzkghBCSSoq5udvzQwgcftO9MIujKDz+zdZjGVEjhJA5kTGhZqCQ1hq1RPb8UWmNgfFmcEzQLMSJqDVTH5tmI4yoEUJIVinnVFT1ud+MMwdOsB+1+c0wGnM+FiGELGdSqmqiqaa5Rq1re374G2FHmInItmYigfko2AghJHP0ouG1l9qGP0Dl2W/E7HP/PHK7iIqoSYnyPf8I9fAjPVsHIYRklWwJtYaRXqEmY5/4iUt9dARYyJ5fialRk+HtsU2zCSGEpJ1SToFuSjSMHqW4a0VMv/Ljbm+1EBE1asr0PvT9/F8w+N139WYNhBCSYTIj1KSUlj1/ls1EpAl/tCuc+iii7PkjXB99fdRAMxFCCMk6zo3MuRqKBJFqLvL1qIiaOvG4/aDQ0zUQQkgWSauqCdEwJAyJFEfUEqQ+hhpeR6Q7huz5/TVqIpT6CM/2TlMfGXkjhJC0ULZ/H3uZ/ggAUKKFWlSNmnbkMWtTX0wUjhBCiEtmhFpVt3540momIpJG1KJSHz3j1aOPByYOpj46fdTCqY++WjZCCCGZopx3Imq9dfeVcUItEFErPPYNaOO/sJ5ojKgRQkg7MuP66PzwpDai1jJ90RkSZyYi3f1zBx4I7yc85yTSTKQ7E5H6us0o7PwBjOGnd7QfIYSQhac4T6mPiEt99NSoqQcfwuD339PcVpvq7RoIISSDZEioWT9X3RCAAAAgAElEQVQ8qRVqbRqH2oMQ6dLoNRMJ7iEUSCXinPgiaqL52D1Oe6ob34zaMy6ALK9KNJ4QQsjiUZ6vGrUEETV1dty3SdQne7oGQgjJIpnJdavZEbViSlMfk9jzi2BEzRZWVtpknLgSgOLR4060zv6vxBzMRISgSCOEkJRQyi9sjZowmzVqou6PoCk1CjVCCGlHWlVNCLdGLa2uj4kaXgcial4zkbiInBCA8AZOA6mPvobX7k5JVkwIISRFlOzfx4Vyfezf9vco3/+/AQCicti3LSjcCCGEhEmpqglj2EJFU9IqMpI2vPZ+ZF4zkTihFkh9DNao+ez5O0t9JIQQkh7mK/UxLqKmHX4Effd+HACgzhzwbRP1qYQp/4QQsnxJJNS2bduGLVu24JxzzsH1118f2r53715cdtlleM1rXoOLLroIP/rRj3q+0HYYpvUHX02tUPPSyvUxKqImw7b8zUExqY8eM5GQPT8hhJCs4dRwzy6U66M7QEKZDQg1swEY1Z6ugxBCskZboWYYBq6++mrccMMN2Lp1K2677TY8/rjfAv5Tn/oUzjvvPHz961/HJz/5SXzkIx+ZtwXHrtMRamlt2pywj1qkmQikpz9aEOF3fbRFoDs+SvhRsBFCSOZwG173vEbNmrcxdgZqJ50X3q5XoXjMRMxcPwA6PxJCSDvaCrXt27dj/fr1WLt2LfL5PC644ALcfvvtvjFCCExPTwMApqamsHr16vlZbQsM+wZheiNqyfqoeevJ3L5nUsLtjxZECEhPRK0p6Jzx3q9AZjJhCSGEBMipAqoiep/66GvxEv79EvosxOwh97nZZ10jKKxTI4SQlrS15x8fH8eaNWvc52NjY9i+fbtvzF/8xV/g7W9/Oz7/+c+jUqngxhtv7P1K2+DUqKVWqM2nmYgSYSYSEVELm4oQQgjJCkIIlHPqPAi11rVmojELpTrhPjf71gATT0DUjvV4HYQQki160kdt69at+MM//EO87W1vwy9/+Uu8//3vx2233QZFiY/QqKrA8HB5TsdVVcWdo1iy7swND5bmPO+ioBXdhwP9RSDiPShCoFDIQ7O3KaUCAGBosBga61AqF31iTBHA8HAZStH66IdH+qGU8tY2O31F05R0nsMWeL8rxILnJBqelzA8J9mhlFN6b8/vIETkTUOhVyBqHqHWfzwAQPG8RgghJExboTY2Nob9+/e7z8fHxzE2NuYbc+utt+KGG24AAJxxxhmo1Wo4evQoVqxYETuvYUhMTMx2u24AluBw5picsoqSZ2dqc553MRAzs1hpP56aqsAohN/DCsNArW5g2n5/xaqOAQDHjs1iRYyZyGylASgqBu3npmFgYmIW5UoNfQAmJiooVuoYAGBKCRWArhs4lsJz2Arvd4VY8JxEw/MSphfnZNWqgR6thsyFUk5FpcdmIu0QtUkojRn3uWELNVGlUCOEkFa0LUratGkTdu7cid27d6Ner2Pr1q0466yzfGOOO+443H333QCAHTt2oFarYXR0dH5WHEPaXR9FEnt+BBpe22mQAiZaNryO6qNGMxFCCFl2lPO9T300RjYAAGaf95eR29Xpvb7nZv9xAOCLshFCCAnTNqKmaRquuuoqXH755TAMA5dccgk2bNiA6667Dhs3bsTZZ5+ND3zgA/jQhz6Ez3zmMxBC4GMf+xjEAtc7ZamPmmhVo+bto+aaPsoWNWqKz0zE2/BahgQZ+6gRQkiWKeZUzPa64XVhEAff8xQAQF/xLMCoo7DrTne7EhRqfVbdu1I9CjF7EOrkLuhrntfTNRFCSBZIVKO2efNmbN682ffaFVdc4T4++eSTccstt/R2ZR2ipzyi5u+D1qKPms9MxOP6GNdHTcC1TvYdRwajcwg/J4QQkinKORVHZuvzNr85cDymXv2vKNzwbPe1oFCT+X6Y+UGI6gQGfvR3KDzxXRx54zYYwyfN27oIISSNZMaP3Ul9VNIqNnymj3FCTQayEj191JKmPkrvA0/DbN98hBBCskgpp86fmYhNsAF2MPVR5sqQxWHLTMS0bh6WfvXpeV0TIYSkkcwINTPt9vxJ+qhBQno+smYftRaF4ULxRdSEPVZI059GCTCiRgghGaeUU+bBnj+A4k/WUab3AQCkZjmHSiUHszAMUZ2AzPdZY2b2gxBCiJ/MCDUnoqalVWzIBGYioXRFTx+1WETg7qb0/NfeP2QiktJzSAghpCX9BQ3TtYUVarkDvwIAmMVh9zVZHIFSm4DQKwAAwebXhBASIjNCLfU1agkjan4zEdv1Ubb40Q1E1HyujwFR60boaCZCCCGZZKCoYbZhuL+Z80LEDVN9xTMx8+L3AwDMgRNhFoehTj7lNr0W9en5Ww8hhKSUnjS8Xgqk3Z4fSez5Q2YiCSNqvho1s/lfmokQQsiyYqBg/R5MV3UMl3NtRveOY+ffCHNwLQ6e8loAQPWZl6L42DeQ33MQACNqhBASRWYiaqkXarJ9RE1IT7qi9Yo9PF6oSSECro/euYPniqmPhBCSZQaLllCbqunzepypzR/Fkdd9z31uDq71bW+sewXM0kr3udImotZ/5/tRfHhx3aUJIWShyY5Qy5SZSIsxvgbVjplIq9RHEagXaKY+yqCZCCGEkEzTb0fUJudZqFU3vgnGqlNhFoZQW/fKyDFmfsB9rFQOYvjL5wNmxO+ZXkHp4S9g4M6/nq/lEkLIkiQzqY+2wy/U1Oq0hKmPUWYiZiepj87cgTRKQgghmWfQk/q4EBx++68RdyNSeoQaAOQObocyMw5z4Hjf69rBh6zxSmYuWQghJBGZCanoUkIRgEhpnZVIZCZijWw+TGImItqbiUjp30YzEUIIyST9xYWJqLkIEW4FYyPz/aHXlKmnrN2qRwFTR/5330fx0a8CAPRVp83fOgkhZAmSmdtThilTnPYYoGVEzfuDlyD1EYo/9dFnJpIZnU4IISQBTkRtvmvUkuBE1MzCEBTb/VGd2g3dOAMrP70JlWf/CUoPf8Edb5ZWLMo6CSFkscjMlbphSqgpjaYB8Ikz0cqe3xNRk0lcH4XwpYsI+zi+Y7CPGiGELAtcM5EFSn1shcxZza7N8mr3NXXqKSjVIwDgE2nWwMVfMyGELCTZEmqpjqglSH2MMwBpKdQUQESYiQR7ssWuhRBCSFYoaAo0RSyNiJpWAACYpVH3NXXid8g/eWfkeNGyFQ0hhGSPTAk1Lc1CrRszEVtotaxRQ9Ce30l9DFr9E0II6ZRt27Zhy5YtOOecc3D99dfHjvve976HU045BQ8++OACri6MEAKDRW1JRNSghPu4FX97a7y7IyNqhJBlRnaEmsx+RC2UEpmo4XWMU5YMWP0HjkQIIaQ1hmHg6quvxg033ICtW7fitttuw+OPPx4aNz09jc997nM4/fTTF2GVYfryKmbqiy96nN+mxgkvgT6yAfrIyW12aHVTkhBCskd2hFraUx/bRdSc16LMRFrY80sRNBPxOjtm5uMnhJAFZ/v27Vi/fj3Wrl2LfD6PCy64ALfffnto3HXXXYd3vOMdKBQKi7DKMH15DTP1JSB67LR8mevD0T+5E/Wnv7r1cAo1QsgyIzNX6oYpoaTZTKRtXZizPWzP37bhtfDa8zddH2WqzxchhCwu4+PjWLNmjft8bGwM4+PjvjEPPfQQ9u/fj1e84hULvLp4+grq0hBqTlq+ndJoDJzYejxTHwkhy4zs2PNnKfWxZUQt3PC6XY2ajI2osY8aIYTMF6Zp4mMf+xiuueaajvdVVYHh4fKcjq+qSuQcw30F7J2ozHn+uaKUSgCAUkFBYbgMseYZLcdrhx7C6M8/BvOsj7RI3W9N3DlZ7vC8hOE5CcNzEma+z0l2hFqWzEQQkcro1KF5Ux+dx+1cHyPNRMyuf+gIIYRYEbT9+/e7z8fHxzE2NuY+n5mZwaOPPoo3velNAICDBw/i3e9+Nz71qU9h06ZNLec2DImJidk5rW94uBw5R14Ak5XGnOefK7lVL8IwgKnh09GYmIUmhzDSYrww6lDv+f9xdOO7IQtDAIC+bX+Pws4f4Mib7kl0zLhzstzheQnDcxKG5yRML87JqlUDsdsyJdRS3UfNS6TrY0Tqo7uplZmI8DlruYYkEk2hxj5qhBDSMZs2bcLOnTuxe/dujI2NYevWrbj22mvd7QMDA7j33nvd55dddhne//73txVp841lJrL4qY+NE1+KQ+/4DWS+HwCgr3gmKhvfjNrJF0Dm+jHylfMj9xO1SVeolR+8ccHWSwghC01mhJqeJTORyNRD6zVvHzX3cdsatUApopS2YMtMiSIhhCw4mqbhqquuwuWXXw7DMHDJJZdgw4YNuO6667Bx40acffbZi73ESPoK2pJwfQTgijQAgKJhevM/uE9nXvjX6Lvvn0L7iNokCr/9TyiVIwuxREIIWTQyI9RMiVQLNdG2Ri0iauZEwsw2Qi08GVMfCSGkB2zevBmbN2/2vXbFFVdEjr3pppsWYklt6curaBgSdd1EXlu6N+xmX/A+aIceQuGJ7/heV+rHMPjD6HNMCCFZYun+he6Q1Nvze4RaqF+ad3OkmUgr84+oVEkJn5kIIYSQZUNf3rpHu1Siaq2Qaj70mqhNhgcajcj9yz/7JLSDi9tknBBCuiVbQi3NuqNdHzVXqUWZibSKqEV9xNJueJ2Zj58QQkhC+guWwdRSqFNri6fG2iFKqAmjGn6tOoG++67F0Lcum5elEULIfJOZK3U9S/b8EQjX9TEcUWsl1GSc+UhLAxJCCCFZpS9vC7Xa0hdqUg0LNaUeEVHTw0JNnXzSmiPX1/N1EULIQpAZoZb61MeEZiLRDa9b2fPH1KhBNhtes48aIYQsG5zUx+kUpD5CSZb6KKKE2rGdAACj/7ieL4sQQhaCbAm1VJtjJDQT8fVRs2vUWpqJRHzE0mTqIyGELFMGCpZQm6oufaEWFVETERG1VkLNpFAjhKSUzFypm2lPffRFxaKEmv1aVOpjVIPslsdy/o991AghZLkxVLKE2kQl2oBjSRFRoxZlyx9Vo+YINd6UJISklcz89Up96qOXpA2vnR8fs1XqI81ECCGENBkuWeLnaAqEmowSajP7wwMjImqiOmE9MOqdH9iox/wWE0LIwpGZK3XdlNBSKtTUiSegTu5ynxce/RqKD38R6uFHUHz4FqiHHkbh0a8DQLOuDE2jkNzee1rMHj4nxUe+ZBdZp/N8EUII6Z5iTkUpp6QkohZu96pM7wu9FpX6KEzr/YkOhZoyuQur/u0kFH77nx3tRwghvSYzDa/THFEbvfnlvufFHbehuOO2yLFmebX7WJZXQgoFxce+Hju3WVoRem3gx38PAKif+DIAQGPNCwAAlTP+DAN3vh+133tNZ2+AEEJIqhgu5XAsFUJNDb2kTu8NvRYl1GDXbwuj1tEhc3vvBQDkd92F2jNf29G+hBDSSzIl1JRUm4m05/Cbf+YrijaGT8Lht/0KQq9AqgXI4igAYHi4hIlDRyCMOmRxGABw8M93ARAQswchpFVAbpZWWvOsOtXaLhRUn/XHTIkkhJCMM1zKpSL1MYoo4RUt1Oz312FETZ16ytp94MSO10YIIb0kO0It7WYiCTD71oRek8URSIz4XxQKkCtD5sr+1wDIvtXR5vuOOKNII4SQzDNcymGisvRdHxMT1fBaOhG1DlMfbaFm9K1uM5IQQuaXzFyVmylOfUyCFGpMTzRCCCGkM4ZLOUzMdmGyseAk+92LjKgZnUfUlJlxlB75UuLxhBAyn2RGqOmmhJYhIRNyuoooqCaEEEK6YaScrYhadOqj9f46qVErPXB9c85WPUoJIWQByIxQMyQyFVGTWtH/nEKNEEJIjxgu5TDbMFDTO+zDuUhE/QZ6zbIiXR/teuxOImqiMeM5QHaELCEknWRGqOmGiQzpNEAt+J9TqBFCCOkRQ3YvtaVv0W9XVYuw+6M++szmk4gatWZErQOhVp+C0X+8fWhG1Aghi0t2hJopkdcy83Ygg0It4keKEEII6YaR1Ag1mwijK33ls5ubW6Q+qtN7MHLz5kSHUWrHYBYtgy5BoUYIWWQyo2zS3PA6Cqn5hRpTHwkhhPSKYUeozaZDqDm/gWauz3ouVMyc+QFMnnUtzNIKf8qijfCkLmoTOxIdR9SnIG2hxtRHQshikxmh1jBMaGpm3k5E6iMjaoQQQnpDeiJq9g1YO6tE5ges/xZHAbWA2rNeD7O0EkrlUHjXGKGl7fsZhr75J5HbRW0KsjBoOS3TTIQQsshkQtkYpoQpgVyWImqh1EdG1AghhPQGJ6K21JteN457vvXfE18KADDLVm+zmRf+N3eMWV4NZfZgaF8REGL9d/w1MHsEgz98H/K7t0E99mR4n/oxmPkBQKhNMxJCCFkkMnH1r5tWsXEuQxG1oFBj6iMhhJBeMVDUoIilH1FrnPASHHrHI8jv+DYKO7ZC5vtx8D1P+caY5VXI7b8/vLPpf2+lR24BHrnFfS5qE6FdlNoUZH7IymJhRI0QsshkQtk0DMteOKdmJ6IWcnmkUCOEENIjVEVgsJhb8kINsNMdHUOtCFMRs7wKyuwBQMrAhtZCS6kGhJrRgNBnIQsD1s3RgJmINv4Aig9/seP1E0JIt2RCqOmG9cdZU1L4doI/LA5qUKixRo0QQkjvGC5pqRBqADy/geEbsmZ5NYReRd9P/5fvdWE2IJV8/JSBdEnRmAYAyPyglfoYSJ0cufVCDNz5N10snhBCuiOFyiZMw7QialoaI2ox9r8yUJMWfE4IIYTMhZFSOiJqADwRtSihtgoAUH7g3/0b2tjri4ABSX73j635CoO+1Mfc7h+j8Jtbu1k1IYTMiUxc/TfsiFoqzUTi7H9DqY+MqBFCCOkdQ6Ucdh2tLPYyEiFbRNQcJ0gA1m+qogFSQpg6ZESqpIMye8D3fPD7f27NVxiybo7aQm/4m2/oas3a+C8htRKMFc9sP5gQQiLIRkTNrVFL4duJy6EPCjPWqBFCCOkhI+U0RtTCm+rrNsMsDFtP9Jr1X1tkCWnGTqnMWhG1/M7boR5+BIDVp62+7hWtzUTiShYCjNx6EUZveVWisYQQEkUKlU2Yputj+iJqcfa/TH0khBAynwyXcjhWacBMKDwWFfdmZcTvvJp37fqFYQs1s70AVSqHACkxtPXNGL3lHABA5Yx3AWoeULR4e342wiaELBDZEGppNhNhRI0QQsgiMFzKwZDAdC0NwsMSaDJKqAGAVrRG6VXrv7aYqh//ovgZa5NAY9b3msz1Wf8VSrwgSyACCSGkF6RQ2YRJs5mIiPuDT3t+Qggh84jT9HqikgahZhNhJgI0e48KwxJqjsiqn3QeJi68KXIfpT4FxXZ6dOfJle2NWuyNVGHUO101IYR0RTaEWqrNROJcH/0RNUkzEUIIIT3EEWpHZ9MgPJz0zBihZkfU4KY+WkJNKjnI0mjkPqI+CdGY8c9jR9QgVIg410hG1AghC0RGhFqKzUTicuCVnP85a9QIIYT0kJFyiiJqjilITEQNanTqIxQV0t4GAObpb3Qfi/o0RH3Kfxgn9bFlRG1+hVrht7dCPfybeT0GISQdZOLqv5FmM5FYe/5ABC3ux4kQQgjpghE7onY4DRE1mSyiJqIialqhOU5tNsAWZgNK5bB/HieipmjxN1LN+T1fgz98HwDg4HuemtfjEEKWPikMQYXRnYhahsxEZLAmjUKNEEJID1nZX4AqgPGp2mIvJQFthJpdowbdX6MGRQPUplCD6s9WUSZ3+edxUx8ViLlE1Fq0BSCEkKSkUNmEcez51RRG1GLv2IlgTVoK3xshhJAli6YIrOovYHyyuthLSYAt1OJuWrquj5boFB6h5tavAZb1voeBbR/yH8UXUZtDjVq3hiNpaJVACFkwMiHU0mwmEnfHji6PhBBC5ps1gwXsT0NErY2A8bk+SumKKalozWgbACj5qN2b87j2/GqsPX8S18eunSHnIa1SmdwNYTf3JoSki4wItRSbicT8EDipj74fGEIIIaSHjA0UsH8yBUKtTUTNjZrpNQxufQtGv/Rqe7zmj6LZqY9mfiB6Ho89/5xcH7sUavNh/b/ipjOx8sbn9HzexSD/ux+g/44rF3sZhCwYKVQ2YdJsJhLbUNNOfWwWPjMdghBCSG9ZM1jE+FQN5hJPudNXnQ4AqJ5yaeT2ZkSthsKTtzc3KJo/Q8X+TZX5/uh5PPb8iSJqUqL/rr+DdmB7/JhOmGdHybQz9O23ovTIlxZ7GYQsGJkQamk2ExGx9vx2jVrQpp8QQgjpEWsGCtBNicMzS9v50Rw8EQff8xTqJ22JHqA17fmlVnZfDhlzqXa2Sn4weh474iYVNd4QxCPgRPUISg/dhKHbLvOPMbqLUoou9yOEZJP0KZsInBo1LZURtZjUCts8JPQjQwghhPSINYNWJCod6Y/xOBb8hce+4U9N9PyGGuXVzYhaIUaoOQgtPqLmmd/p2xa8nOpacLGZNiHEQyaEmuP6qKXQTCQ29dEh2E+NEEII6RFrBqxIVCoMRVphm4TkDjzgE1LSzko58oY7cPSPf+iOM/MDOPKGO91xR/74dhw77wbPfGp8n1NPWqPTMFuKoFDrQY1aj+vVcnt+itzee3o6JyFkfkkUrtm2bRv+4R/+AaZp4tJLL8U73/lO3/aPfvSjuPfeewEA1WoVhw8fxv3339/71caQZjOR2B8Cd4DzEaVQhBJCCFnSNCNqabDob0Gcbb99s9MY/T0AgPT0UTNGN+DIn9wFUZuEseIUGCtOcbdJocba8/siao0Z+ziB64+ua9Q8IrAx42vQPVeGv/46ABlppC0l+8uSZUFboWYYBq6++mrceOONGBsbw2tf+1qcddZZOPnkk90xH/zgB93HN910Ex5++OH5WW0MDVNCEYCaxohanKuUs5kRNUIIIfNEf0FDX15NSdPrLgjWeTvCxzZPMUZORiRKvJmI1/BD1KftB/7fau3oox0vFfBH1ER9CrI40tU8mUcanhvZfpSZ/VCPPIbG2t9f4EUR0nvahqC2b9+O9evXY+3atcjn87jgggtw++23x47funUrLrzwwp4ush26YaYymgagfeqjSOn7IoSQZcC2bduwZcsWnHPOObj++utD27/4xS/ioosuwsUXX4w3vOENePzxxxdhla1ZM5gWi/7OCZuJJDToUjSIGDMR4fQ6kxJK5aD9YvO3Wht/AAN3vr95yMOPhFwhY/EJtelk+yxHWtzkHv7y+Rj+5hsWcDGEzB9tVcD4+DjWrFnjPh8bG8P4+Hjk2D179uCpp57Ci1/84t6tMAENQ6azPg0JUh+Z8kgIIUsSJ+PkhhtuwNatW3HbbbeFhNhFF12Eb33rW/jGN76Byy+/HNdcc80irTaeNQPF9NeoxREUanaETbRpeSM9ZiLB+jMnolb6xb9i8IfvC41Rpvf6ho/ecg5GvnJ+ouV60yqV+mSifZYlZowjJwB19sACLoSQ+aWnloJbt27Fli1boKrt0/VUVWB4uNx2XOs5FAwPl6FoKvKaMuf5FgNRij5XxaL10ah2pDCXUxO/P+e8kCY8J2F4TqLheQnDcxKNN+MEgJtx4i0N6O9v9uuqVCoQS7CuZs1gAb/el35RMH3mB9F/90f9LwajYokjaqrHldn/mTliqvjbWz0vNn/L3V5swSmPPQlzaH3Lw/rcIueh+fWSR0oUHvs6aied57ZciEJIg91lybKgrVAbGxvD/v373efj4+MYGxuLHPvtb38bV111VaIDG4bExMRswmVGMzxcxsTELKYrdShCzHm+xaAwPYsok+BqtYE+AIYpoQFoNHRMJnx/znkhTXhOwvCcRMPzEqYX52TVqoEerWbpEJVxsn17OMXt5ptvxo033ohGo4HPfvazC7nERIwNFHCsqqPSMFDKpbcuuvLcP0ffzz4JoVfQWP0c5A48EBZqScsJhGr1OZUSIphm54gpb12aZ95gpowUCoQ0kd/zE1TbCDVf6qMduev7r6sh832YfeGVyda+hCg++Bloh3+D6Vd8LNH43O5tGPzBezF7+jsw87L/ET+wTX2/NcZM9Hn3/eR/AkJg5iUfSrRGQhaStkJt06ZN2LlzJ3bv3o2xsTFs3boV1157bWjcjh07MDk5iTPOOGNeFtoKw5TIZS710X4/kveMCCEkzbzxjW/EG9/4RnzrW9/Cpz71KXz84x9vu08vs07acdIa63bhLASOS3nkVK57McQTdwKv/zz0J3+MgZOf79uuHLEu3LU2WThKqQgBE8ND4ahO/93XoPj0F0LNNR0ZtaOPYeUtr4T+rnshgrv0jQHT+1DWD6PY5vyKYvNapq+kojxcRu5XVu1j/tV/33LfpES975bfFb0GZds1MF/y34Bim/5zAXLbLPGjveZfEo0XeysAgGL9AHItztXQQAEotz6Xw4OFRBFU9cD9gJoPHY+ZBGF4TsLM9zlpK9Q0TcNVV12Fyy+/HIZh4JJLLsGGDRtw3XXXYePGjTj77LMBWNG0888/f1HSOnRTptPxEWjR8DpISt8fIYRklE4yTgArNfLDH/5worl7mXXSjkHV+n15bM8EVuTSbWAlzv4/0E77NRrmMLD2IiDw/kdMEwoAXTdwrMW56WtIFA0dE0cnsSpie/3h7yInFXhlgDj8GCaOTKJw7JgvU0bWpiAA1GamMBN1zMYscvvuQ2PdK1CYmnL3nZmaRn1i1j1+V98HaYbWHzVPq+9K8cHPYuDuf0GtZmLmzA90dPhO156f1TEEoFFvRGYROfMdm5iGrEenRrrHPDrVMn3SYaQ2C5kLr5HZFWF4TsLMd8ZJohq1zZs3Y/Pmzb7XrrjiCt/z9773vV0srTcYZnrNRCDbmYkQQghZiiTJONm5cyee9rSnAQDuuusurF/fJvVtEXB6qe3LgPOjzA+gccKZ8QOcm8ntklWE3fA65maqWRqFaFRCr6uTu/x1ZgCUhhK4te8AACAASURBVO3eGDEeAAbu/BsUH/sGDr/xx266I5DEbCwBiW8Gx6PMWs6WMml9Xy9ok02UqEYt6XvXq822DYQsMXpqJrJYGCmOqPXkDzEhhJAFJ0nGyec//3ncfffd0DQNg4ODidIeF5pV/QVoisCeYylvet1LbDOR2N9ooUKpHg69rB55zLrwj9rFiH5dO/wba7te8RuI9OJGbg/mELVj1lSFIf+GRgXIleY8f1ckqFFLajgijCqkGWEAc+hRQD1h3htrqxNPYPTml+PYeTegftK583oskj4yIdSWR+ojIYSQpUa7jJMPfWjpGxRoisCJw0U8eWQ5pTQlsOeXhisI6se9CPl997rbRX0Kono0tJ925NFw7zZnn5iImmN4IqThs+cXRi8iavE29klRahPWVB6hVnrgevT/5GocessvIPtWdzah0cDAD96L2Rf8FYwVp/i3uaKojcRK8r6SGI4AEHrVMo7xkHvqJ8h94/UovvKfUH32Hyeap1u08QcAAIXHv0WhRkJkQqjFpT7md3wbjRPOhFSLKP36szBGNqD+tLMBKVF86CYoFc/dMKGgMfZc5Pb/HO1zInpHbt/PQq9JkV7XLUIIIelj3UgZu47GCIkMIY9/HqRWwuzz2pRrKKrl9minItZ+72JMnvvvWHnjcwAA6tQeCGli9ox3QZ3cjcKOrdZu03tg9llOoNMv+zD6/+vD7pRRETX18G+gTuywtjdmm46SgNvHbS4EBUhXc9gRNW+tfP9PrgYAKNWjMDoUatrBB1HccRvUqd2YuHRr4GB2jWQ7I7UkIizhjXChV0PCTz36mL3W7QDmV6iR1pR++e+QuT5UN/7pYi9lUcisUFNm9mPou+9E/YSXovKcd6L/p/8LAHDwPU9BmdqNgR99cDGWGonRN2b9gQag1Kcwc+YHUV+3GX33/iNmXvIhDH37raicujy/oIQQQuafdSMl3LPzSKpLCRJRGsGhP3us/Tg7KubUZ0GokOWVmLjoZgx/641QpvcBAPTRUzxCxhovC8OQigap+o0shF5Bbvc2GCMnw+w/HgAwesurmtsbMxDe1EdPdK1rgu0JAEsEdZDOp1StiJob7avPeObqIivI2Scy8pjM8TrUMiFmTNvb7lJaKadBQZtUMPYCQZfvVvT/9H8CAIVamtFNM/zDYv+xUyd3Aaa/aaTzh3DynP+N2oaLAQArrz8FQq/A6FuDI2++b/4X7V+R52Hz8cH3POX7LyGEEDIfrB8poW5I7J+q4oShRao7WkI4mS2jX95iPbdFRWPdZujDJ0GpHLIGqkWI+rS7n1I5BGNwnSXSlEB2jF7F8Df/BEDM73pj1o3gAXYNe5uomqgehVQLQC7GHjxqf2kAIvnlnytE7QiVNzIoooRgu/mcNc0heyi356fo+8nVmDz/P+J7pSURkXYEU4Sib4546jJ11NRRePTrqJ3yR8l79xESQSa+PW3vAAb/Adp/JKSSs/4BCcV6DACe1xbuf6L5P0IIIWSBWTdqibPlkP6YBMUTJbNeaIoKmeuHUjliPdYKqJz2dki1gPqJvw9l9pCVSqcVfGUMUitCqU26z/OP3wYxc8B3CKFX/I6Rpg7oASdOvYrCI19yoy8rP70JI1/aEvs+Is1QjM4ide65sKNOoTV2in1NJoNCFvCIq9bRpYG7PoDCzh9C2J9Dq+O0QjjGL6H3kbBWLobSg5/B4O3vQ/HhLyYYPbdjkWyTCaGmm2hpzx/6Q2VGhN3tPxiRfzgIIYSQDLN+xIrI7DpCoQYAyvRe/wueCJTM90FUbaGmFqAf93wcetcO6CueBWX2AIRRC0XUzNJKKLNNYTb0vXdh9Obf9x+iMQOYDUi795cwdYhARlDffddi8I4rkf/d99zXtGO/i38jEREh0WFKpbDbC7jXUj5nyoSpj951OGmGStjuvxnZSiZagq0Q/MdMINSc6GAo9dGdJNE6QvPaAlI78ACEnToaP5g36Uk8mRBqhoyIqEX9UbBxi2u9Qs35Ixzj1kQIIYRkldFyDn15FU8yogYAmDnzg6japREAfE6OMjfQrJHyNFQ2yyutqFj1CKRW8F1PmKUVfgMzAEpjxvdcNGYhjDqkZqcxmo1mxMfZZ2bcGluf9qVcxhIV8UoQBcs/8R0Ut/+HPd707Se8Ub6kztWeca4Yi0oJdERkwnqt4PnxbUuyNnt/EXKRnGPqoy3SSw9/ESO3nN16LGvTSAsyIdR0w4QauCPhy5uOS330piU4d77ouEgIIWSZIYTAupESdh1dThb98ZgDJ2DqFf/YfMGb+phv9tySaqG5T3kVAECd3A2oRd81hlla0faYTaFWhBSKda3ijRhJCTfCI5qirbktYs6IqFKSiNrQd96BgR9f5aze/o8t1DxRvsSukt5xdhQrso1BpxE1vcX3NUlETY+LqCVsjB6H572p3s8pCufYFGwkgkwINUNKaGrgrQQLch2kjEl9tB7H9T8hhBBCssz60eVh0Z8Yr0GHN/Ux1998rHmEWmklAECZesqOqHnEnb2tFaIxC+hVK/VRyUGYDX/0Shqei3kBZWZ/c187FbPv7o9i4Aee1gNRUSX7+kg7sB2F33617bqcY7qir01ETcweRPHXn/O/5o2oOeIo4sa46FC0OI7ZkXQi1ELvw7qmFF0qtU6uJUWHNYNkeZEJVWKYMhRR8/0D9Qo1U2/eTfLWozl/MCjUCCGELEPWjZTwvUcOoNowUMwxu8RXOxQTUYMnoiYLg9bQxgx0regTd4kiavosRGPaEoJCta5XvBG1wAW9N6KmTu+DXlqB3J57IGqemqgosWJfA4185XwAwMFT/ih+UVI2BZoZYSYSMf/Qd/8MuX33ob7W0wjecx3miqMoTwDnGO0akkNAQFp1fTEkSX10atSCXgYyaePt2Ik7uJa0P49uRSHJNpmIqOmmRDCg5v1H54+o6e4fFum9Q+YINKY+EkIIWYasHylBAnhqIr7uZ7niq1HLDzQfe2rUZC4g4ByTMrUAmW9G4eIQjVmI+gxkrgyp5kKpj1bKoSei5jEncfq6KTP7/OIlKuLViVOjN00ywkwkSgyFjFgAv6Cz35OMEjPu2tqIFseIZI4RNTQqkWPdc9RtOmIHxnS+3nmEBMiEUItqeO2Lokl/Eav7D9D7D0mhmQghhJDli+v8yDq1MN6a9py3Rs0r1LwpkUVX3MlcX6j5tRerOXbBanjdmLFEnaJZro96TERNiGYvNwCiPgWYOpSZA750wMjG0Gajta29B6/oE1ERtUj7/7o9zivoIiJqEWYirvDz6CNRm0T5vmv9x7LPbcvUxw4iaiFR5zzv1kykE9zPlRE1EiYTqsQSaoF/8C1SH5s1ak1rWMkaNUIIIcuYtSNWL7WdtOgPo8TUqHlTH4MmI7a4k7k+yFx8E/Gps/4JpQc/azlGNqat8YpmCapgRM1To+Z1PBRGHcrsQUuYNWascUJECimhV7HyP05L9LZ91vJuw2uvPX/E/E4T6ZjInvueWqQ+ekVL+Wf/jPKvboAxtL55WEWDQO/MRISpN88ZelA31sH+nbZLIMuLTETU9IiG1/7UR88/Vmm4f1hkhJkII2qEEEKWI+W8ihOHi/jNgQS278sE046eea8XTG8ao8dMxCfgcuVm6mOu3DqiphYhc31W6mNj1orYKTnrOiZUo+a4Pgp/PzOz4aY/Cmm6tvNRYkWdfLLle/bia/ztiDKfeLQiTsr0Xgi7obcjcnwiSkZE1CJ7vEWkHNrjFG8UULVutLeKqEVGE4NjvPb+PrfwhCmYcfN2Ir6cz9FzfGVyN4a+9afJWjCQTJMJoWZECDVfyFsG6tXcHh5hMxHJGjVCCCHLlI3HDeLBvZOQtAoH4ElzjEh9lIrmv7mr5twImyyOujVYMtcHWRxxh4WMRbQSpFaGaMxAqVtmIlERtdzee5Dfvc19Low6TFscCqPud4G0o1lRYkU98lji9y88Qi0y9dG+vlrx2Rdi+NYL7TdoCzVvCqY3suc6LUakTUZE6JyaQFGfar5mfx4hMxHv97aDPmqhY7vuk12mPnZQd+aKOs8+ffd8DPlddyH/u+93d3ySGTIh1PTIGrWIAlgA8NWoMaJGCCGEOGw6bgCHZuoYn6q1H7wc0OyUxQgzEW/aY3OjJRTM4qgnotaH+tPOxrEt/4aJP/gCjr7uO6gf/6LmLmoBMleyatT0WTsaF65RG7zjymaES0rAqDfTLc0GVDuiBngETIRY0Y4+GlhzvBjxRdQizERgNlsGaBNPWMd2hVo49VGZ3udJN4yIOrkRtea6nUilV6g56ZeiEUjT9c4ZE1FTpvdh9DPPgzrxhD+i5mvKba2j+Ng3UHj0a5HztKIjwxY7Apnbey/Ugw/ZE9g3BhaiRo4saTIh1KIiaiJgIOIidU/qY0TD6w6cegghhJAsccpq66J4xyEaigABJ0fnNUccRQg1pxm0WRr11KiVAaGgfvKFaKx9Ocz+41Ff98rmfJqV+qjMHrLnt8xEQg2vfYswIIxaM9pk1KHMRAi1JBG1FpEnn9W/U6PmEY/q5C6Mfu6F0ft6hZrUoR58CCs++wKUHr7Zni+ivs05hk/EWWLFK9QQIQbVozuQf/LOtu+r8Ng3oc6Mo/jQzU0zEQSij54as0FvX7qkdJD66HxnlMYMRr+8BaI26b7XjgRfjxj8zjvQd/c1C35cEk0mwkdRNWr+KJq/+bUw7G1ea1gnRaGT3heEEEJIhnAMRXZNVPDSRV7LUkA6Ta89aXxOhMfb7Dq0X3HU5/oYwmPrD60ImSu7NV1W6mMu3EfNi2lYETVnbqPu1qgBzbTDqAv9UI1aRLqhg+IxE3GaUTvCAgCKj3zRF8nz4kt9NBq+vm9xa3Ov1zxCyYl6KXYNnDWfHVHz1MGNfsHTtw12rV4Unl66QvdE5GLcwruiE8v9gPHIik9vbK59ruvogsIT3wEAzJz5dwt+bBImMxG1pPb8MJsRNb89PyNqhBBCljcjpRz68ip2H6XzIwDoq061Hnj7pdlmIq0MQsziSDOiFtFDzdd/TStCauXmc0/qY1DcOAipW2JFzUMqeQizAWV6P6SSt7a7EbUWhh3u81YRtdapj2ZxtLnugN2+N9o18pXzoR4NRvKiatSM0DbXRdJjJiIi6uBCxESjnPcvlZy/Rs2X+jg3J8bQ/q3OcWCsT2AuQOpj4dGvQTmW3GCmp0iJ4kOfh6genaf5zfiodEpIvVAzpYQE2qQ+RtvzS589v/3Y8xohhBCynBBCYN1ICbsnKNQAYPql/wPHLvwc9NVNO3s3itUiomaWRt26Njcq58En1NSCb4yV+pgDzDoKj98G09Ngu3kAO/VRyVvNsY0G1Jn9MIafDsBbo5Ygda6FKPGmPrpmIl5x41v3gE/4BI0+8rvuantc9xjebU5ErXIwPN6bDhlAqRxEfse3wxuccyJUf4sDqYfHdEvQnr9VhK3VNs86tIO/RumB/zu3dUUw+IP3YuQrF/R83iRoh36Ngbs+gIE7/2Ze5h/44fuw6t+eMS9zLxSpF2q6YRextkx99Ig2Gdfw2r7zxYgaIYSQZcyJwyXsYkTNQs2jvv6s0GtSLUSbidhYqY9NM5HQdjUQUfMKHruPmnbkUajTe1DbcHH4AKZupT6qedvKvwZleh+MEeuiVDRmMXLzy9F3z8fbv0ePIUiQtmYinno1pXYMKz7XNEkJRruCVvPRqY8RETX7GL61OMeMiTgCwMBdH8DQd9/pjwp6j6tobc1EusWbHgrAVwsXHttCKNtBB1E9ipEvn4v+n3wECBqozAXH5MVbi7iAOJG02DYEEW0aOqH46Fe7WdaSIvVCzbA/RFUks+e3Uh+dhtceFyenNo01aoQQQpYxG1b1Ye+xKiYqbMQbh8z1+aJioe35AXgbXofQPA2wtSKg+YUaFM1tNm32HxfaXdhmIo5oVGYPQZh1GENWRE2Z2Q9t4glowXTDCPr/68PIfXSF76LYyTLy1r05107CqLmpmr4aLwBK5XBzjYGIWij61aJGzSte3NTHiIt5dXpfrMh0968FjuvppesTUMEymbkQTC9tlX7Xqjm200Nuaq9nrnjR14r8E9/F4HffGZh/4c1KvDjfkch/I0DgM1n4ej0AELOHMHLzZigTv1uU46dfqJl2RE0NNrz2GogEPmgnP9krytwaNQo1Qgghy5fTTxgEADy4d7LNyOWLzPcDETVq0y/7MBqrTgOEsMScUMJ90xBMfSz6LlTNwpAlIuyLVDM/GF6AYyaiFgA15zabNsurAAD993ws8XspPvZ1APDXCdnXRLkDv7LWqOSadWFG3Y0ABoWal6BQU4IRNa9hSOUwYDSaZSsRqY9RYkfosxj+z4sxcvPL49dRD3yP3Yiav0YtVCbjpcO+giKYzqi3iqi1Sn10xLFnnV3WXA1953IUgqmgiyR+HFyh5r1x4UXGBF06ZQ7vs/D4N6FN7EB5+w3dH38OpF6oOamPoYia8+EK4Q+hS70p3Hypj3YuOVMfCSGELGOePTYAVRF4YE841YxYGH3HwSyvDL1eOf1yTLzOuhiWpVFMXPpt1E7+g9A4XzROzflTH8urfPXyshBRo+YzE8lBNCwRZOYHW0b6WuGLngUubKVWar7mE2otBEgw9bHhaVjttB+wWfkfp2Pki2e5r7nu3GgvTHLjv3B7uEWhBISaIziloiZPfew0imU20Fh9OiZf/SlrPr3biFoziunS4pwnwpNCGEq7XOC+bc4NhtiImuld6xxEZUDk5Xb9CKv+9USoh3/Tfo32OZGLJJnSL9Sc1MfEro+Gx/WRqY+EEEKIl2JOxQvXDWPrwwdQ19lwN4rJ867H9O9f3XacvmojoIZNypwIQmXTW6znuWZEwUl9dJ9HRNTUyacgahNWjZqah6jb0SutgENv/VUnb6U557SdXiclhNlA7elb/Ot1RVStGQGxI2r1da8AABjl1Ziyz4uoBQSSR2A01jw/ZCaiHfsdtAPbrSdm2J4/SKsaQd9xg6mPTpRKmv65fWUyASfGuj862PaYRgNQcm4Lh1ZiM0mNmrcWsKXoS4L3eEHxs8AOiYodxfV+/72IXqWjBt5n4YnvArCajLfFiaYuUiAn9ULNTX0Muj46H4r9B8fFbACmDgkBeK1kac9PCCGEAABe/9wTcHimjnuenCfb7JQjSysgC0Nd72+sfDYmXvMVTL/sI9Z83oiCEG4PNgCQhbBQKz10E5TGjGVq4omoSTUP5Psw8QdfQGXjm8PrjnC2dgSP2zDbjiCYpab1PrRi09jCqFkGKBBu6qMbEVQLqJ3yWmu+6T2R73329MthDpzQjFp5LqKVmf32a+0jasbACZGvBxH1QGTYuSY0dV/qZqhMxjtHozOhBrMOqeaa0c2WNWrtHSF7kfrY3N9bGtSI3dYrRHUi1n7fEWqx5i2+VgXdR9RC89vX+olMY9w1MKLWFY5Qa93w2hvOtl0fg7VobuojI2qEEEKWN8+x69QePRDjxkbmTOOEM5uO01rAwt9zLWK2EoRuRM3+nOw+ao21L8fMC/7KN9Qsjka3CrCFWm7vfSj94v+4QkZ6e6TlSs0UPaNmuVYqWjMtzF6/VO12AQDUGEfG6qmXQSoa1KndGLj9r3xCRbEFUdCePyqdU5bCqadRBCNqwussqVdhOiLZKw6DAqZDoeZE1GCf25ZRMM+xgtegzcbec0x9NAIBC/dxMKLWQaPuhKz89Eas/PSmyGO4Ai7uuHGttjolIPKkbfSTKNXTW0q1CKReqOlxQi0QLnUbMZq69b/APwb3Q6NQI4QQsszpy2s4fqiIHYc6jCSQrggKKH9EbTh+P7UAqeZcR0VvOmCwts3sWxN5jeNE44qPfR39d38U6tRT1niPQJRauVkvpVctp0q3rVHOFWdQc65YjF2zoro1eMXffCXaxTDg+miWVoWH5Aes7Kg2BGvUvM6SwqhC5uyG5C3S7Fo21g5Qvv865MZ/YZ8XJ/WxRS2f0UD9hDPRWHmq6xTq4ggYTxStm4jaqn97umfOGNGGCBOUHpN76idY9W8nof+OK5Hbey8U29lUnd7jr5F08AqpTmvUvAYwQZHnCrUEczprEIyodUV86qPnH7nU3bsakJY9fyhy5ihlCjVCCCEEG1b24XEKtQUhFOnymom0EiRqHlDyELCuhaTqEUmBGq7Ks98QmTUkAlEFNx3QM5fUSr6G11Irem5wN8WZVAuAorrrNSJaC0BovmutqBo0IU33Alno1UjjFihavAmFd65grZwjRkzdei95J6Lm7bkbFGrJ/x303fsJ64HarFFrGQUz6zDLY6ivPzsUWXIjanMUar45fdfHnvdcO9aygXgvyO35KQCg9MiXMPy1S9yG6vldP8KKz74gNF71WuJ36vroEdfh1Edb/iQQau6/j6CIXiAyK9Rc9SxNwDSadzWc1MeYEy4X6YMghBBClhIbVvVh19EKZuqL22tpOeAIjsZxL7Re8EbUtGLsTWRvqiEAn7jycvjNP0P1tLe6hmlRtWoOrhOfdw05j+ujk4rolIyoueZY5/j2f83+iDoyRfULxjgR4wgSoxoZUYOiNUVWC0L2/I7dvyPU7IiaT5wZ/u+8OrW77XFCSNlMfbTfy8AP34fBrW/zr0+vWIJOzbmC293m9JCba+qjF68Y9ETXVt5wKoa/9kdzm7stgaBKiwiedmA7Rr76mubYDlMffZHUYDTO/ncgzCSpj3Z6L81EukO3T3Iw9bFpJmL1TXPvapi69YHFRc4YUSOEEEJwxolDMCXwwFPspzbvqHkcveQbOHbBZwAA0olQKXb0KS7tSs37o28xQs0s2umTjrhqEYkStWO+sYDj+uj0UataNWpuRC3vOlv61o1oww+p5HxzK7XoNhBOpC0uoiYVLRQ1jH4/gRo1Jw3RrlELRtSUmfFQqtzAXR8AGvE946JQqkfd2jpHcBV/eysKO79vvVY5DOhVKLUpyPxQtHh2xao3opZcqOV33m69Hw+++rtARCn4WeQfv81aZwza/p9D23tf4vWEvsctxJd6bGdgbGepj74G6YEUT1d0JYnSuRFn1qh1RayZiH3yhWlYoV2nMaW07PljlTEjaoQQQghOO34QeVXgvl10flwI9DXPcx0eZdGuD3OuSeKygNSC36Y+roeaaEa/AGDmJf8d+tDTIoc6F+s+4aDmm66PTo2avSYromYLNVsoOiLL7FsTcQDV1x/LdXoMLtlOXRNGzUr/DNa+CaXpPxCDWVoZEh/O2oq//SoEZLNGzdSRf/w2rPjM85DfZ9m2N457ASqn/qm1zsrBlscKrb96xBK0gM9e32Hlf5yOodsug9BnrXrCCKHWNBPxuD4msOcv3/fPGPnCKzG09c0Y/vJ5/o0+18d4oSJmD2Loe+/C0HfeETtm5D8vxkgnUbjgtXeL1gQhOkx9dNIqgYDNP9AUjK3EX2MW5Xv/qZlqGviuDW59C1bccGpHa+qG1As1PTb10YmoWeYhzdRHu+F1XL+0RXJ1IYQQQpYSxZyK044fxP27JtoPJj3FtA1EnIvE+JvLiq9PW0jMOCh+wWf0H4+pV10XPaVT0xXsNRusUVO8NWqOmYgt1OwLY7NvLGItmq/mKxjxcdfRmLVaLDnHC/baUrS2N9eNgROgVA7557VFjyO83Obd0kRu//3uOPMZr8LEH30N9fVnWeM9FvOiOuE3q/BgllbY8x9uXnvGpCvm99xtrSE/4E9hdd+Anfrojai1SX0U1aPo+9k/Qzv6GABAnT3g3x5sWRU3jy0IFdtcpjf4ZUcn7QA6bXidG3+g+SRUo2Z/t1u4Pvbdfx367v//UHz4i/YC/Gsv7PxhbDS4l6ReqBntGl7bLo/+1McGUxwJIYSQNjx/3TAePTiDiUrv+yuReGRxxP9CTORI6FVf5Csu9dHZ3x3riYgBfofIqNRHKwpmWOIkVKNWiIwGAVZEK/TeRECoBYRE871V3HosqZVCx5BCbevEZwyuC8+v+1MYveYi/vdsHc+0PwvHSl47sB0rP70RQ9/448hjmkVbqFWPAgkaXgOAmR+MdMsUHtdHV4S3mattc+5W9vz+gc6MEceYwgqv5X5CpBJMfezg70qHNWo5WwRb+0bb8wdNY3w4Itn5rtL1sTt0I1qouSF6U7dCu046gF2jtlhFgYQQQkhaeME66yL157sZVVtIzJBQi7lmMap+A5E4oebg6RnrS23sb6YoNlMfNTRWn+7uJ8wGYNYtwws1kProRIMCWUlmeXXEGlS/UJv2pz6aeSv9U+izUOxm1bIwHOHW3V6omYProFQO+9P9AhEpR9zmd//IH5F02g/Yn4VrJX/MciLM7/lJZCNnp+atvvblluGJUMPiKhDJkYXoiJq3j5ojpn2izzT8dvuAr4l3FF4Dj2C/uMhxQgDSRG73f7lRRO3Adl+EMTGB73FHEakOG17nxn8J0+4FGBJkUamPQWMTZ63uOYqOoMZFVntF6oWaE1HT4lS6adiuj3ZBpzSsD6yF4xEhhBBCgGeP9aOgKXhgDw1FFpJwRC1aqIUjan5zjZkX/rX/NcVxu9N99UJy4PjmnE5tj5LDxB/eikNv2w4pVAi9isHv/bk1PpT6aIuogHCKs9X3CjV11i/U/h97Zx5eRXn98c/M3C37DVnZ9yWsiuIKIpsouC+IS7VVarXV2l/Vql3QapVWq3WpFdGKS2utVeqGRQWtKFUQUBGQnbATSEL25C4z8/vj3jt3TwKEhITzeR4fc++895133oRkvnPO+R7TERQkvrqwA6UzMz4TqjGTlSB6RncA3G9eFo6QRAgZf/YA6o+/CYCU1S9hq9gc/rCWOKKmNIQfWtgi0+tCGDr+nEFUnvNcYO02VyCNMPKGPka4mY7MxPelEamPpi0VU7VHCc2st68gb3ZvUlb8Oexm2YRQoxEzkUjC51FI+fpZ3G9PR9m8MPi5QxAnphH3/TqYVgMH1MFwrAAAIABJREFU6/qo+KoxUoNuoXGfDT5QCF6/VrqWvNl9cBQvihgSWKtlz5/s/E3t92HS/oVaEjMRJbJGzUzg+iimIYIgCILQKDZNZUhhBt/sOvK1GEKY2IhaMtMMPbt/oxG1ulE/o/TGsPjw9hwX+FxqfrTwyQj3O7Nq1BQNbCmYKZ1AtaP463BufT+wHpsrXOuvOSJSLpsRUVPUaKFWvjH6WkNpmP56ay2GMzMuomaqWpMtlYy0wPnte5dbPbmUCPdGb++JMT3sIgRIyCAl2Phbq90TcFFsKLeG2PfFCzXF8KFn9oTQvJozIEgiBJYa09vNdCQzEwm7Ppq2oHFMhLhxBPuSpX/xe1JXPBn4TBPCwXZgU7hHXSM1YhmLfhZchIp9T9DZ0eqzdghCTfceXvrgwQi1UG2jI2QUEyNIQwI1+H5o72z7vgkPiX0wkCxN1FOT+P0Wot0LtQZf4IfNqSW2/FRMI/CDqEUKNX/Cpo+CIAiCIEQzomsmG/bVUOORfmqtRchMxCI2awioHvsgnn7nWSLJVLR4V70Y6k64hbLvfYHh7h0lcsy0cJ8yK/UxMhUvZt5AjVowNTAiohbbmNt0xVxHkJrR9+LLG4a36+loNbuD5wvcp4VqvLIW/CgsDBMJmeakPqaGr0sx/KD7UIxwipupOaP3IVL4WRFDG4YjE9faf5A1/1rSlj2CYU/Hnz0gSUQtutbNtDnB3xAlTmObcBtJzUQiUh81J4SicwkI7WNTfdbSP7uXlK+eCS4ueUTNVrbO+lqtCzpehvanEROOZCi6t0mXzmhixODBpD6GTHhCPfJiUjyt8qjgddhLvgqMt0VEn2Ojf8nq2Y5wk/B2L9RCfzjSnTG/nMzovFPLecdKfZSImiAIgiA0xZg+OegmfLyxtOnBQssQFeUhYRaQP294oH4oJGCakymkqBiZ3QJfRz6wDkUeSGwmEvdwW3OFRY1mD9d2xd6IJ7kx93ceRcW0/1B3ws3We0YwcmWkh6N7qV89HTi/MytB6qPWpFN3VETPV4fir4s6HjBCCc8bWb8VacxiurKjarJMlxtfwfGBiFowDTBtyf2kLbkfDF90s3DNiaLHCLWYJtyB1M4EEbVgdEzRPRBsxZCsj1ooJbPJ1EfAXrIi8EUzolSmooSFWqg3WaxQa04qpO5FaWpc5PFYcXUQEbXQHlgRtViRF4qOheYMHlcje+7F/nuSiNqhUW0Jteh/wJHf0EBzxuA/uKYaXguCIAiCYDG0cwbd3C4WfJfYnU84AsQKkAQizIqkWTVoBxnliHxgHVFLlrCPWkx/tuiImiMsyA6yxVGUkApeT6Lea4YjPvURxUbtyb9ofP70Lnh6TQwM99dFiSUgcF0R+xB1PCIaaMREBk3Nib/gONT6MtTqHQCkfv0MqV8/Eyi9iepBF0h9jBZq0VGYgD1/AtfHUORNbwikPtqcCXuyAaie5gs1U3PhXPevpK0RYlZhOWeGmkgrsUKtGSJKMbxNRuIiUw8VPWZOQwdfXbOieVYfv6BQs+9eGr1GM1qohdJho74vcT3fEl9jVGPtI0C7F2o1Xh0FSHUk31BF94BqDzz9MfRA6qPUqAmCIAhCkyiKwtmD8vlyewX7a5pf/C8cHjWjf0vl1BeBJDVqoVTBYHrfwZotmBH9ZI1+k6g59VfoaYXRNWqhsfb06M9G2PMHTDeiIyWePufg7XKK9VrP7JlwDaEassCkgTn09HihlshMxFQ1fD3GUn7Fx4kvEEBRqD3pjsCXvrqo+jQIpT5GRNQib9RTO4XHBZ0ow+Nq8OcGmh3bytZFpzIavihn8YCZSANpSx+23ousUTM1Z0CkJkh9VH21YPgD6Y6aM1zvluhSDyKipvjryVz0f2R8+psmx6Io4XMmsf5vlimI7m1S0GW/fi723V8EXsSO1T3kzRlA+uJfW2+plcXkPdUt4EgZSaitQ/DnNm3546QueyS83mB0TNE9KPVl4chlxPc/VifENc0O4ZHUx0ap9fhJc2qosU9xIoWavyGgjIP2skpsrwxBEARBEJJydlE+JvDRBkl/bC3qR1yPt9eEwIsE5RqhCExkHdZBESkMHOnUj7wJI6NruIYr4rjhyIj6qBnRhy0QeQsJtcC9WNU5z1J50esA7P/JTqrPnJVwCWZELV4o7TDW8TJUm5WwRo2YuqJEBBtlK/76uIiaqTmjooBRN+opEUIt5hyqt8qqv1O8NdjKvgvP4auLWqupubDt+wZn8YcR5wkLtZBxjJnEjVzxVgdcHzUnps2FY8enZL86CWJEp5Wy2kSNWmDwwUU+LZKlPjZDqCm6t1l1ZrZ9qwJfGNF2+Wpwz1JWv2Q5cDo3zwfAUfxBzLmCQi0ipTey5i60DsfOJeQ+P8ISmlFCPfbhiOFPnOIpEbXGqfH4SXfEi67Yoj8z1ME+GFGTGjVBEARBaB49O6WSn+5gzd4j+/RYSEKiLCArVTCBs+LBzukI1MQ1FIWbOEfVWTmiI2pE2vNrjgidluS2MllLpEjBELy5DtWqhc+dGbeewJyhiF7jQi3k6qj46uNq1LA5o1wfo9LYIoRarJum4q+PmLcOrXxDeFnequhggM0Z13NMiYjCGCk5wQ8mEWqeykCTcc2J6chA8ddhK/uOlDUvR41TDyKidjDui0pkbZYnsVBLZnAShe5rVtTXSrOMGRu5Z7nBZtu2A5sAMIJtGMLrCUXU0iLfDX8ZvCa1fn/057zJa9RS1vwN9xvnW69DP4+KRNQap8ajx9WnAfFFf6ot4Khj+sHUkz65EARBEITmsnjxYiZPnsykSZOYM2dO3PG5c+cyZcoUzjvvPK699lp27drVBqtsGQbmp7Ou5Mg+PRaSELyxLp/2vvWW5ZKYWnBIU4ZuNE3VbgmRhsFXRJwzuVAztbA9f1RELUmkJlH9VSzWzXWMUDOcwbTDmPu2UDpoU3ObtqCg8teheBNE1GwplH0/YK6hRt6op2RHjyO8Z7Un/ix63pjGzY31toOYyF3w+hK6PhKoGVQbyjFTOlk95gDSl9wXPafhxbHl/bj0zkQcVPlPZIaaFVGLEVFJDE4iBZ2ie5pVy6bWBaP2MQ2oYw1YALSgUIuN6Fk/S5E/t5E/m0kie9G1g/HRs5A7JER8X8X1sXGqPX4ynBqutf8g/b93Wf/F9uVACUTUHDs+Q6va0Tx3JEEQBEFIgq7r3HfffTz33HPMnz+fd999l02bNkWNKSoq4o033uCdd95h8uTJPPzww0lmO/opKsiguLyO0lpv04OFFiV0Yx1ZJ2OJhyQW+E0SMgOJcZisHfV/gfedYVGQuEYtQiiFbsgPNqIWQchCPVaomVZqYMx9W0hoatFGJ7GYtmDqo68urjmxtYchZ/CI41GpjyExqDrY/5Od1J18e0RErTY+ihVZo5ZAqKnesLCzoj5qYsGp1uxF9VSipxXGpaDG4lr7StKIWlSd4MHY5EeKspCZSIL6MQiYgbjWvhJez5q/hU+ZxEzE32lg1OuQw2R8RC1eqIVEneKLiZQmiqhFprgmEWpqZES1qTTN0M+0pD42To3HT5rTRtrns3Ctfx3nlvdxbnkfxTTw9jgTPb0remoBvoIReLudHgg/q3Z8nU+Kmqdh6DUYTjeevlPa6EoEQRCE9sSqVavo2bMn3bt3x+FwMHXqVBYtWhQ15pRTTiElJXCjeNxxx7F37962WGqLcHqfTtg0hVvf+BazOXbcQssRSqWLvHnUkljiN5dQRCwqPQzqRv2c0u+vjHJfNBMIBDOi4XXYdj1ZRC25UKu48DWqxoeNHmJTH60b+SQ1arFpiSF8ecODx+2YwYbd8a6PQaFmSyD2olIfXfHXodoCdvm+2jihkMgx058ziPLpgTq1yHS5UGQuWX9fWzCt0kjvHGdqEomv4HjU2r1Jo1sHLnkzYoHN//cb2RRbKdsYiKrFNMp27PgU+47PyP7XVDI+DjhxavvXkPHJL8ODdF/CXmR1I2+K+v6rtcHfkTH2/M7N74WXH/yZD0XOYr+vVo1a5AOGyH8nCaz2jZScqKidksyO31qEjqnaMIoubHzcYdLuHTVqvDp9nDYw/NQPuZra0fcmHevtc07SY3p2P8pmrD4CKxQEQRA6IiUlJRQWhm9mCwoKWLVqVdLxr7/+OmeccUaz5tY0Bbc7temBjc6hHvYckZzmTuXXU4q455217KzzM6xrVtMfOspo6T1pNS79K/oXT5HeP+yk6O4UL54O6tqCN+uqMw0ldl+yo8UbatiwxBh+BZnd+qF5ygBw5nTF7DoYAHvvUxKvQQ8LjLjj7oB9Ph/dBkBWfiH+6a+hLbgDpWIbjm7DcLtT0ZzRkamUtBScSa7Xd1cJKCruUGTLkYpT9eG0RQuFdHcWpjsVzJS4ObT0TrjNwM29mhbYD8Xmil6/Mx2n4gVVx3S5LedFV2oKjuA4NTXwWTW3HxndBgSW4y2zpnCkZQbnjNnz0HXWBKL0qYW9UTy7E44B0HJ6oRQvRlN8CY9ndQpHXh1JxiQiUvgppevp9P71GMOmRY2JTcN0ZzpQqqOFTrpLQbHHC/nUjPSoLDfNV02nb59CW/541DitLtBKwOgzHnXLR7gznShGYG0uxYPdnYq65FGo3ovZbRQAaZ3CbSfsdpv1vdMSrEPJKITS9bizUkBRUB2JHzqE5lAMP8ZJN6J1HopbP/gG4M2l3Qu1Wo+fdIcWCJFKOqMgCIJwFPLWW2+xevVq/va3vzU9GNB1k4qKuqYHNoLbnXrYc8Qypocbh6bwr2Xb6T6ub4vO3RociT1pFZQCOPU+qPIQkkyR16Fe/RkoGsZBXluuasOvpYJuNLovitdG6Ja3bMzDgXXs/RqAyrzT0DMGoV21GD2rNySYR63VyUmw7kiyswdgO7CBimof5JxGVnp3HBXbqLEX4q2oI0NXiIx71XkMGoJzxfpeVlRHC5FOmgtt+bNx56yqAz04R26M9b1uqtZa03wqqYCh2qPW30lLwVdbBaaO3ZGFFhRqDV6oC47LqKtFA+rdQ6irU3Hnj8Be/Ik1R4PpoLaiDgw3GQMuxttjLJkLb7WOG3tWowJVphuHmUKMrUt4HlcXUuvK8FaX4VDtViqptSc1fmuf/PU1NF01GCA2WqhuX0Jt73NpLAnTPiuf2pPviBIZtVVV2OsbiJXWtfUGqCqhxz6mrwFtcbRLaN2IH5L6TeD757W5cQEVZQfI9QWEmre2iuqKOrK2fIpatZP6zAFkANVeG6FKQ6/fpDr4PUlv8BArzX2uXBzGGipKy8GeQmp9Q0LpXFFaDjYXuYYPjxfsTfzbaQ55ecl3s12nPi7dWka1x0+myxZIBxDLfUEQBKGVKCgoiEplLCkpoaAg3tjhf//7H7Nnz+bpp5/G4Wju7dHRSYbLxkk9s/lkU6mkPx5FGFm9MDK7Nz0wFtUWV6OWiERjQimJeuj/7j7Jbd+bUaNWcfEbVmogQN3ImzFcnfB1HpV4johUNjNJyqV1+mRmHxG2+6H0R1NRrf514WPRZiLW+/Y0K/XRtKeEDU4iatRCjpC+/OGgKNSdcHP04oKpj6g2qic9YfVnM20pmIqGrXw9AHpa50Zr1PSg86FWWRyXzho4EL7WpOYfzcCXf1xcWmIinJveiXqtGIlTH1HtUffvSoL2AnpWuL7OCKakKr46S4xaqY9+D4qvJlyjFrVfjdeoGSl5UXMlS31UPZVgmiiGv9GU3pai3Qq15dsruPr5L0l1aJxTVBBsMChCTRAEQWgdhg0bRnFxMTt27MDr9TJ//nzGjx8fNWbt2rXMnDmTp59+mpycnCQztS/O7JfD7ioPa8UBst1jKrbEN/WxJKiDq7jodcquXdasnlzNcX00XdnoOUXWa1/30ZRdv8oyF4m7x4t8UBAhuKrP/EPc3HE3/8Eb9Uijj5Apid5pULh/nXUsaDYS83DCtKdi37kE276vA7Vmobq9SNfHoNmLPz9QMxeqSQvRMOCi6LVFfTYQDzKcbrCnxNUKNgy6zPpaz+wGBIVaIkEX+T1spNdaMtHr6XN2cE3uqLq1pPPEmKhkfvCTeNMPAt9XM9JIJYGINF05EV8HhJoaWU8WEle6B8Vbm9ievwnXx1CbBEv0JRKVEOjhFhKqreAg326VzcjuWbz0g1Fk2xQKMxwopiGpj4IgCEKrYbPZmDlzJjNmzEDXdS655BL69+/P448/ztChQ5kwYQIPPfQQdXV13HprIJWpc+fOzJ49u41XfnhMGJDHn/67hb8v38mD5xY1/QGhRakb8UMc2xY1PbA5qFrzhFoCTFc2JtlND4SWuaENCjXDmRWwrK8P13mZmgvF30DVWX/B0//8ZDNYKEH3wSgxEYqaJYqSJBGapj0tIBi8VejZAwJrNLxRorJq4hPYSteEG3lHGJeUT/8QPSfa9TAyGqenFaLWl2Kkdw4cizETMSLMMoy0wBhF9yQWahEkilpFXpPii38I0zBoGnZ/VUDkNSOiluie3BZhb2+h2kGLiKglcIY0IoxdjKBQUyJ604WcLhW/B9VXY702otxK4/uoRZ0jNTqihpG47kz1VKAHHSlbI0DUboWaqiic2icnkBeqt56yFQRBEIQQY8eOZezYsVHvhUQZwAsvvNDKKzrypDttXDisM6+u3El5XV86pbbvdM72Ru3oe6gdfU/LTKY6DlmoHQwtkiIWsuN3usFTadm4QzBt0ZPAwj+G0uu/RW04QKe/B019IlMfQ6ItgSgLtwCIiahFiC7Tnoqp2gJyIDIqlpqLr8fYiM9EVEclum+NeM9I7wylq9GDDpyRLRMgOrXPSA8bG+kZXbCVrY2fO0hjTbFNexokEGqmagebC6WhvFmNqxP1IUvoNqnZmxQ8VlNwwAhGKCPTWS1xFawxVBvKA1G6qJ87k7RPZ9Iw+CoUU8dUHYGWAdY5AlWY1t4kjahVhOv/WkF3tNvUxyjMkLKViJogCIIgHGnOH1aAbsKC7/a19VKEw6D6zFnUD7++WWMrz32JA5e8dWgnStIj7GAI1ZX58wI1XLq7d/hgSGQpiW/4GwZNw58zOJBe6e4TnlNLUKOWQDQkS91UI3p7mbaUcK1VI/ejUeIukUCxPmtaLRKsiFpMOmFkQ2fTkWnNbWR0TXp+aFyoRYrX6HXZwZYSqG+LiaiZikrFuS9z4JK3w+eIqQuEiIbZkZ9VbU0KnqgayVBfvKBxS6hFAoQjhWpdabDXX3h/tZo9pK56nuzXp4Kpo2f3oWr8o9ZxIzUo1JqqUavZYzXubo2IWocQatZmSo2aIAiCIBxx+uSkMaxzJv9YsQvfEbSmFo4s3j5nx6XeJR3bczz+whMO7UQt8SA9eI/nyxtO+fSF1A+/zjpkiZ8k94HVEx7lwPQP4g9ECrCQCEokKkPiJSYipESmX9pTLTFpNiI8oiNqCaJ3kRG1YCTJqlVLyY0eG5kKqShWrVfIWASg4vxXLAFVev23gRq5RlIfk/XCQ7OD3YVaswfHjk+jDnn6TsXXcxz+iBpDNSI10XrPWx33Hqq96YirEt9APCQEjZROVu1byLVTCQm1iLo8paE88H9/A84tCzAVDU9RuM2AGVejllio2UpXS0TtoAmFYKVGTRAEQRBahetP6cHeag+fbilv66UIxwAhAaMYfvScQdGuj1Y0rHn3gfVDrw18kWiOBKIhHMmKFmpqfWnkqLAgbSRw0GRELfJe1lpfQDyZqbmU/iBc5xUZUYtEz+gW/jqzB/7CkYHxrmwMZ2ZcRK1y6gvoqQHH2pBzpeHIpGHAxdFrtblQPZXYS1Za79cPupzqiY8FXkRcWyjiBVB78p2B9/yJzUSaFDwJvk+qJzC/4coJiCvTDKc+1pdGrSXwXhnRbwT2vubUX2HY0y1HzbQvHkKt2h7WFjHY9q+2Sq7E9bG5tGIIUhAEQRAEOKlnoKfaql1VTQ8WhMMldI+X6Aa6idTHWGrO+B37f7wj6j1LjCUSDaHIW0xEzdtrkvW14qsP19E1JjwiI2qJRGHEZ0O29HpWr/Dx1HDXuGSmIVGpj7F7omhxhh3eXhOpnvAIAP5gOwRvrwn4Co6LWKsjYZsGvVP/iP2PsMCPELVGEkFpzdtU6mOkeA0KMCv1MTUnUDOnN6D4w0ItXFcYIBThs+YKir/6kTdRdsM6q1bTdmAjWe9ek9DC3+/ui618Pa4NbwYnldTHZmH1ZZAaNUEQBEFoFeyayqCCDFbvEaEmHHm8vQOiKNY6HyKiVEkMIOJQlLi2AqHoVCLRkCyiVj3uIarPeCAwpb8+HDBoZo1awhv9UPTINPEMuJiK8/5OQ9HlieeKFUDB84Zq2xKuJYm48PU4kwOXzad6/CNUjX+EmrGzolJDQxG1OGLmqzwnvrF4pKDUUwuiah2bFVFTNSqnzKVq0pPW9yJkJqIH3S7VujLLHET1VGLaY1taBwi1SYjNwos01bEd2BTlDGkEW0To2f1QDD9pS4MtICT1sZmENrOZT1IEQRAEQTh8hnXOZM3eapZti69HEYSWxJ83jP0/2Ym/4Pi4Y6HoSSiicihY4iaRmYgtsesjmgM9s0fw3PXhG/fGbuAj5k+YCWZz4ssdStWkJ0BRAo6RSXrVxUa4Kqe+SP2gyzHSCpKeo7H0UH/+CFAUPEWXYzrSo81LVEdCoWbG9Njz9jkHX+7Q6DERgtJI7xxd69isGjUVb+9JeAZcZH0vrBq1YJqnWrM7+pxaAlFJuLF13D7EXltERC20D0ZENDMwx1ESUVu8eDGTJ09m0qRJzJkzJ+GY9957jylTpjB16lRuu+22Fl1kk0jqoyAIgiC0OleP6kZBhpNn/retrZcitANCdVAtjRk0+1ASNEtuLiFnRctMIpJGGnaHXCS93UZbUZpm348mEnSKSsXlC/D2ndrkx2PP4y8cSc2ER6KbW8eu5SCCGpHGJ6Zqg0RRqkT+EDHjolI0Y9fTrBq1eDMRK6IWFGpa9c7ozySK/gFGak7idccITiVofFI95j6rHYGRmh+z9qOg4bWu69x3333MnTuXgoICLr30UsaPH0+/fv2sMcXFxcyZM4d//OMfZGVlUVZW1siMLY+VRypCTRAEQRBajdw0B2P75fDKil38fuFGbh/XF5vWMZJ1hJal7JplmI4j07MtVMN1OD3hQr3K4kwniEh9TNAHzMjqSdn3V2Ck5uHcPD/wZnNv4A/XBK8554k9R0QkqebUX8UYosQQGbEL9lFrcn5inC0JtDcwNWegGXdMJMtU7YdVoxaqx9Oqd8WsIbFQM11JhFoMzm0f4c8ZRMPw60hbFqjfixVqR0XD61WrVtGzZ0+6dw9YfU6dOpVFixZFCbXXXnuNq666iqysQA5nTk5OwrmOGBJREwRBEIQ2YUhh4Gn5G9/sYXiXTKYMPjJRE6F9Y2R0OWJz1436OXrOILw94+vXmouRHlhfZCPtEOE+agkaNkM41TB0899cz4QkKY3NpTn3vXGpj6GoHwr1x9/Y6Boiha+p2ROLnwTXGivUUDRMmytgnx8b0dPslqmKP6sXtsri+HMkcH1UYiJqak3TQs20pYavyUzeVsTbYyyO7Z9ghtYaHGuktX5ErcnHXiUlJRQWhosSCwoKKCkpiRpTXFzM1q1bmT59OtOmTWPx4sUtv9LGMMRMRBAEQRDaguFdwr2cXl25q5GRgnCE0Ox4+l9wWMLHSn1MEFELuz42PocVGUoQeTsiNEcoxJmJhMSkvcn9iopQqrboCFtoTMKIWoxIUm1h8RZyxgy1HFBtoDkov/wDqs59Kck1RIg71Y6JErbnT8nBtLlQYyNqCWrUDEeGlSabzH4/MGewFi24V6F+zXGpj61gz98iIShd19m2bRsvv/wye/fu5eqrr+add94hMzMz6Wc0TcHtjv+GHwyapgbmqA9cRlp6GqmHOWdHwNoXwUL2JB7Zk8TIvsQjeyI0RmGmi7dmnMRnW8p4+KPNfFdSTVFBYttwQThaCd2E+/NHxB+0IjpNCLDGWggcCZqTSRYbwQoJpWaIjEizElN1gC1BrZ4aH/OJS32MEGqh1Ed//nDs+76xxKaeOxglpkm2t+tpVE6ZGy2IFAVsrnAvOJsLIyUPLcZMJKHxiTMj3DC7EYdQK9U1JEKDJVZGSqe46zrSNHmGgoIC9u7da70uKSmhoKAgbsyIESOw2+10796dXr16UVxczPDhw5POq+smFRXxje8OBrc7lYqKOmyVNWQDtfV+vIc5Z0cgtC9CGNmTeGRPEiP7Ek9L7Elenty4d2S6ZLk4p6iAJxZv5c1VeymaJN9voZ2hapRPX2ilQEZiOAKlPQ1Dr2lijsBttWL4Wnx5kVROfRHFW9VkbVdgMdFCykrna07aZGRETbNFWdaH50/kkhmf+mj1jwuOrzzvb9jK1kW7YEZEwarH3E/DsO8njPoZjgw0f31APCoqRmpuYK6oNQTmOjDtP6Qu+xPO4g8w7enhnm8JxHT59A8BcK19NfBGaG0hL4xIF0w4OlIfhw0bRnFxMTt27MDr9TJ//nzGjx8fNWbixIksW7YMgPLycoqLi62atlYhuIFSoyYIgiAIbUOGy8bEAbnMW7WHv34hLpBC+0PPGYTpTJANZk9h/4+3U3firY1+3roPTSRoWhBvrwl4BlzUzIhajNCxUh+TO1mGiE59tEOi9gcJXR9jMjBUW7ivWTACZ7qy8XU9NXqcLUII2VKSpmaGDESMlGwA9PSu4QhbaO1BoebPG4ae3TcwPjU3wiE0XkzrOUXoOUVWNM5qPRD8fppx7p9HPsW1SaFms9mYOXMmM2bMYMqUKZxzzjn079+fxx9/nEWLFgEwZswY3G43U6ZM4dprr+UXv/gF2dnZR3zxIUK2mYftniMIgiAIwiFz9YmBh7TPfr4djz95sb4gtDsUtcmarpoz7sfTdwre7mNaZUnNiqjFYqU+NiOiFhkZU9TExiEJUh9DBh/hMRGpj421B4gyDXEmHRbczMdYAAAgAElEQVSaX8/qDYSFG4SbU0etPRjh9OcND6c1NpKeatXYBb/fSlCQxda9Kd6apHO0FM0KQY0dO5axY8dGvXfrreGnCoqicPfdd3P33Xe37OqaiyH2/IIgCILQ1vTLS+OPFwzh9rfWsGZvFSO7udt6SYLQahiZPag6O3G/4Uh8hSdg37vi8E94KPe9ITHUjIharDAzB19M7a7vSFvxRMR88WtoGHQpGZ9Ea4KwmUjzgipmbJphBCFhpmf2DP4/nMVnpOQGeqxF1KjZSr8DYuoPG0lPterYYiOjmoPaE27BWbwIW9la/DmDmnUth0PHaHZiij2/IAiCIBwNHN8tEwV44+s96EYrud8JQjui4qJ57L9x6+FPdAj3veZBmInEodmpO+UXMWtIILxsKZR9f7kV3VJMPWxM0sSa/Z0GBsclX5/hDDwACvXlMzLCQs1MCbQIi4x++fOHAeArOC5s728mT0+1ImpBMVcfqk1UNepOuZMD0z9g/092Wuc6knQIoaZIRE0QBEEQjgoyXXZ+cEoPPli/n3sXrKe01tvWSxKEowtVaxlr90NpR2D1ejv0e+bS67+1RFisWUkII60Qf15AIJmqDSM1ptdcEjx9pwYnaCTiFaydM1NygehUSz0YbYtsEVB78p2UXbsMMyUnIvWxEcOX0GeDdWw1ZzzQMsL6EOgYyia42Y3mvQqCIAiC0CrceFpPtpbVseC7fVTU+Xjy0mFtvSRBECBco9ac1MckmK5sTEcGeCoT9lELUTX5aRzbPsLI7IGeGRBToUbVyag78Vb07H54e09OOqZhyJWo3mrqjvshALq7F97OJ9MweDpaRUBQRfVy0+xhN0/LTKRpe37Lwl9RWqVnWiI6RERNGl4LgiAIwtGDoijMOreIKYPz+WLbAbaVS8sLQTgasMqEWioLrZF5TFc2noGXAOH0RLVmTxPzaXj6n9/4Pb3mpO7En4Yt/zUnlRe/gWfQZVY6ZKI+ahBR+9ZYHzUr9bGV+uE1QocQalaeqaQ+CoIgCMJRgaYq3HJGH2yqwhvfNHFzJgjCIeHpO5WqCY81/wPBCNgh1ahFTxT8X/OkRCiiFteYuoUxHYEejrEOjdbxZrg+EvxsIgv/1qZjKJvgZjcWfhUEQRAEoXXJTXMwvn8ur329mxFdM5kwIK+tlyQIHYqqs585uA+EIlVJIk4HS3Pvva2IWn1pi5w3GVb9WrLrC5mJNMee/yiIqHUQoSYRNUEQBEE4GrnljN5s2F/DHz/azNh+udjUQzBAEAShRQgJK8ORoLF3AsquWYbia6RfWHPt9h3p6OldqTv+R80af6joWb0wFQ0jrTDxOg6mj9pRINQ6SOqj1KgJgiAIwtFIYaaLn4zuTWmtlyteXC4ukILQloTMRIIpgk1hZHRB7zQg+YCDyGYrv3YpDcOva/b4Q8FfeAJl132Dkdkt4fFQI22l0T5qoahb2/+u6hBCzUp9PJQO7YIgCIIgHFFG9+nExcM7s/1APdNfWM7S4gNtvSRB6NDUjbgBT98p8QeCQY3mCrUmOQqDJKbLnfxgMKLWqOulrRl1bK1EhxJqkvooCIIgCEcfNk3l7kn9eXracBw2ld8v2ohfN9p6WYLQYakdPZOqs+fEHzADTehbSqi1O38IRaXm1Ls5cNm7SYeEI2oi1FoEq+F1e/thEQRBEIRjiJHd3Nw5oR87Kxr45fx1FIttvyC0Koq/HmjBiFo7vPeuH/kT9NzBSY9LjVpLY6U+SkRNEARBEI5mRvfJAeDjjaX84JWvqPPqbbwiQTh2UHyBhyOGs3lmIsknCpoCHYWpj4dNM+rYWouOIdSsPmod8IdFEARBEDoQmqrw0PmDOb5rJjUenX9+tautlyQIxwyKrxZoyYhax5ASkSTrwdYWdIjdtXJIFYmoCYIgCMLRzrj+ucyZfhyn9+7Ey1/uZH+Np62XJAjHBr6WTX00O6BQI9gM3NNzfBsvpIMINSuHtCP+sAiCIAhCB+VnZ/bBqxtc/4+v2Vxa29bLEYQOj+IPpD6GGkMLiSm7dtnBNxM/AnQIZaMY/kB9miJNNAVBEAShvdCrUypPXjIMj9/gJ69/y02vfcOaPVVtvSxB6LCEatRMW8rhznT4izmKMdK7wGHv0eHTIYQapl+s+QVBEAShHXJ8tyxmnVeEaZos31HJ00uKMYIW4oIgtDBBXwfTnnq4Ex3+WoQm6RhCzdAxpT5NEARBENolI7u5WXDjKfxkdC+Wbqvg5Ec/5Vfvfie91gShhak656/UHfcjjIxubb0UoRl0EKHmF8dHQRAEQWjHKIrCVSd24+LhnclLd/DB+v18uqW8rZclCB0KPWcgtaf/5rDLhWpG/xY9rQAjvXMLrUxIRIcQaoqpS+qjIAiCILRz7JrK3ZP68/YPTybLZeO1r3axZEs5fkPSrAThaMLbexLl318BmrOtl9KhafdCTanbT8rql+joRY2CIAiCcKxgUxWmDC5g+Y5Kfvbv1fzw1a8pq/WytPhAWy9NEASh1Wj3YSj73pUA+POGtvFKBEEQBEFoKf7vzD58b1Q3Pt1SzqwPN3L27C8A+Pf1o+jmbns3NkEQhCNNu4+oYfgAqDntV228EEEQBEEQWgpFUchLd3Lx8M5MHVJgvT/zvfXsrmxgX7WHDftq2nCFgiAIR5Z2H1FTgjajqPa2XYggCIIgCEeEeyYP4KoTuvLIx5tZsaOSC55bZh179MIhjOmb04arEwRBODJ0gIiaHwBTEddHQRAEQeiIKIpC/7x0nrxkGD89o3fUsde+2o0pfdcEQeiAdAChFoqotfvgoCAIgiAIjWDXVL43qjsf/eQ0670vth3g9ws38cqKnZTXedtwdYIgCC1LuxdqihmIqEkfNUEQBKG1Wbx4MZMnT2bSpEnMmTMn7viXX37JRRddxODBg1mwYEEbrLBjkuGy8cFNp7Dk1tGc0jObeav28Kf/buHn/14j0TVBEDoM7T8MFYyomUr7vxRBEASh/aDrOvfddx9z586loKCASy+9lPHjx9OvXz9rTOfOnZk1axbPP/98G660Y5Kd6gDg4QsG8953+6hu8PPnT7fy2/c30DXThWGafG9Ud1Id8iBXEIT2SftXN0HXR0l9FARBEFqTVatW0bNnT7p37w7A1KlTWbRoUZRQ69atGwCq2u4TWI5aXHaNi4d3RjdM9td4eP2bPRjBBtlfbq/gsYuHku6UewRBENof7f43l2LVqMkTM0EQBKH1KCkpobCw0HpdUFDAqlWr2nBFxzaaqnD7+H789Iw+mMCnm8v49XvruOqlFZw1KJ8LhhWyz6Pz7bZyxvbLxaYqbb1kQRCERmn3Qs1yfRR7fkEQBKGDoGkKbnfqYc6hHvYc7ZlLc9PpVZjJrP+s44VlO3hh2Q4cmopXN5gxujd3Th7Y1ks8ajjWf1YSIXsSj+xJPEd6T9q/UDMloiYIgiC0PgUFBezdu9d6XVJSQkFBQSOfaD66blJRUXdYc7jdqYc9R3unX5aTv04fwccbS/nF22tx2VUG5qczd8lWSivr6e5OYfyAXP6waBP3nD2Q3DRHWy+5TZCflXhkT+KRPYmnJfYkLy8j6bF2L9SUYEQNMRMRBEEQWpFhw4ZRXFzMjh07KCgoYP78+TzyyCNtvSwhAeP65/LqtScwoJubqqp6/rBwE29+GxDZLyzbQbXHzzmzv+C6U3qQ5bJxxciuKIqkRgqC0La0f3VjCTUp1BYEQRBaD5vNxsyZM5kxYwa6rnPJJZfQv39/Hn/8cYYOHcqECRNYtWoVN998M1VVVXz88cc8+eSTzJ8/v62XfkzSNzeNDJcdvcHHfVMGce85A7n65ZVs3F9rjXn+i+0AbNxfS0W9j3vPHohPN8h02bFriog3QRBalfYv1EwdU7WB/PIUBEEQWpmxY8cyduzYqPduvfVW6+vhw4ezePHi1l6W0AxUReG35wzk9jfXMLpPDj2yU2jwG7z57R7eXVMCwMS/fG6Nv+n0Xlx3So+2Wq4gCMcg7V6oKYYPFKlPEwRBEATh4Oifl86bM06KipRNHZzPHz/eTE6qg9e+3m29//SSYnp2SiE/3UlprZcMp40Te7jx6QbLd1RwSs9sibgJgtCitHuhhhGMqAmCIAiCIBwkseIqN93J788bDMDVo7qRn+7kndV7eeDDjdz1zndRY/vkpHJqr078fcVOfjWpP+v21TD9+K7kpDnYXFrLcd2yWu06BEHoeLR/hWP4pdm1IAiCIAgtTudMFwDnDinApinsrmzg2c+3W8e3lNWxpSzg+PbAhxsBWLShlLx0Bxv31/K9E7tR7fFz54R+2DSppRcE4eBo9wpHMXURaoIgCIIgHDFsmsq5QwLNzccPyOOL4gOM6dOJP368mS+KDwCQ5tA4qWc2H28spaLeB8DLy3cCMLZfDqf37sTHG0v5/cJNzL58OH1y0trmYgRBaDe0f4Vj+DHFml8QBEEQhFagX24a/XIDIuvuif254sUV3DmxH1MGF2CaJtf/4xu+3VNljdcU+L9/ryHdqVHjCfR+ferTYgblpzNuQC65aQ7+t7Wcc4ry49IwDdNElbo3QThmaf8Kx9Cl2bUgCIIgCK1OlywXH99ymiWmFEXh+SuPQzdMznt2Ke4UO5cd14Xl2ytId9oYWJDOrA83snhzGYs3l/Hilzvw+g1MIMtl5/Q+nQAwTZM3vtnDk4u38tAFgzmhW5akTgrCMUi7F2qKKTVqgiAIgiC0DYkiXpqqMO+6UaiKgsOmctHwztaxijofr67cxZ8vHcYv3l7LrsoGAB5atJGiNRnsOFBPZYOfkmoPAD9941ucNpXTe3figmGFjOqRjaYGzllW6yXVoZFilwfWgtARaf8Kx/Bjij2/IAiCIAhHEa4k4um6U3pw7Und0VSFV645ga1ltXh1k/veX8+iDaXWuDvG92X7gXr++dVu6n0GH20sZeGGUrpkOgEY3SeHt1bvxW+Y/PzMvlw8ojM2VdIkBaEj0e6FmmL4QbW39TIE4ZhD1/0cOLAfv9/b1ktpUUpKFEzTbOtlHFUczJ7YbA6ys/PQtHb/50UQjhihiFiqQ2NI50wAXv/BKA7UeVmytZyiggwG5KdT4/FzWu9OjAza/L/05Q7LdTKyx9vDH21iwXf76Op2sa28jvH9c7lgWCFz/reN6SO78tgnW7jxtF4MLEiPWodhmoi0E4Sjl/b/l9TwS42aILQBBw7sx+VKJS2tsEM1edU0FV032noZRxXN3RPTNKmtreLAgf3k5nZucrwgCGE0VSE33ckFw8L/dtKdNk7r3cl6fcNpvbhkRBcM0+RP/93CRcML6ZOTxpIt5cz5fJtlYvJdSQ0fbSzlu5Ia/rupjNJaL59tKcdpU/nFhH6cN6QAE7jixRVU1Pt44MKh9MxwkJvmoLLeT2aKTUxMBOEooP0LNVMaXgtCW+D3ezucSBMOD0VRSEvLpKamoq2XIggdlpw0BwAPnltkvXf+sEKmDM6npMZDea2PG/75Dd+V1ODQFEprA1kP+ekOVEXh/vc38MQnW6j36Xj1QKT8ple+AmDq4Hzmr93H5EF5XDGyK8Xl9UwuykdTQDdMMTQRhFam3SscxfCD1KgJQpsgIk2IRX4mBKFtsGkqXbNS6JqVwm3j+vLe2n3cckZvHvtkC8M6Z3D7+H58s6uSW+etprLBb33ug5tO4fnlu3j1yx3MX7sPgPfX7ef9dfsBWLmzApuq8t9Npbx89UhMAj3jPli/n1Hd3Ty/dDs/G9uHrBQpQxGElqbdC7WAPX/7vwxBEA6OysoKbr31xwCUl5ehqipudzYAzz77InZ78puGdevWsmDBfH72szsaPceNN17H7NnPt9iaH3/8ET7+eCHz5s1HVeXJtCAIR4ZLj+vCpcd1AeCFK4+zHqCM6JrFghtPwTChqsFHZYOf7FQH958/hFtO68kLy7ZT49FZvqOCjftrAXh7dYk179Q5S9FUhZxUO/tqwvXJPbNT+P7JPVi67QCfbi5jwXf7uGl0L87sl2tFAAVBOHjav8IxfJhSoyYIxxxZWW5eeOEVAP7612dISUnlyiu/Zx33+/3YbIl/xQ0aNJhBgwY3eY6WFGmGYbB48cfk5xfw9dcrGTnyxBabO5LGrlsQhGOP2Ch3yI0y1aFRmBl+32FTueG0XtbrTzaV4vEbfLalnIp6H5qq4LKpLN9Ryb4aL+4UOxX1PgCe+qyYpdsOsHxHpfX53y/cxCsrdnH+0EJe+2oXOWkOxvXPxeM3KK3x0iXLRZ+cVM7sn0uNx8//tpazYkcld0/qf+Q2QxDaGe3+r7li6piqPK0RBAEeeOBeHA4HGzasZ/jwEUyYcBaPP/4IXq8Hp9PFL385kx49erFy5XJeffVvPPTQY/z1r89QUrKX3bt3UVJSwvTpV3LJJZcDMGnSGD788FNWrlzO88/Pwe12s2XLZgYOLGLmzPtRFIXPP/+MJ5/8Ey5XCsOHj2D37l089NBjcWv76qsV9O7dhwkTzuLDD9+3hFp5eRkPPzyL3bt3AXD77XcxbNgI/vOfd3n11b8BCv369eM3v7mfBx64l9NOG824cRPj1vfcc7PJyMhg27ZtvPrqPO6++zZKSkrwer1cdtl0LrjgYgC++OJ/zJnzFLpu4Ha7+dOfnuLKKy/h6aefJzs7G8MwuOKKi5k9ey7Z2dmt8F0TBOFoZGy/XADOGpQf9X6dV6e4vI4B+en8Zv53lFR72FPloazWRze3ixqPbgm47Qfq+fOnWxneJZNt5XX85bNiAFQFjKCR7NDOGazeU23Nf8XIrmzYX4NXN+idk0Z5rZd5q/Zw27i+uGwqG/bXsnTbAa46oRv5Gc4jvxGC0Ia0e6GG4Qd7aluvQhCOaeavKeHt1XtbdM7zhxYydUjBQX9u//59zJ79PJqmUVtbw1NPPYvNZuPLL5fyzDNP8cADD8d9Zvv2bTzxxGzq6uq48spLuOCCS+KiUhs3rufll18jNzePm266nlWrvmHQoCIefngWf/7zHLp06co99/wy6boWLnyfiRMnM2bMWJ555ikr8vXYY3/k+ONHMmvWH9F1nfr6erZs2cyLLz7P7NnP43a7qaqqTDpviA0b1vHSS/+kS5euANx990wyM7PweBqYMeMazjxzPIZh8tBDD1jrraqqRFVVzjrrHD788D9Mm3Yly5cvo1+//iLSBEFISKpDY3BhBgCzzovPTDBNE69uYpomt7+1hp0VDTx16TBUReGe/6xj4YZS5l0/inUlNdz1zneWSMtJc1BW6+Wpz7by301lcfOu2VNNrddvGaCU1Xr53dQi3ltbQr1PZ/KgfKoa/HTJch3BqxeE1qUDCDUdU2n/lyEIQsswbtxENC2Q2lNTU8PvfncvO3duR1EU/H5/ws+ceurpOBwOHA4H2dnZlJeXkZ8fLRKLioZY7/XvP4C9e3eTmppCly5dLXE0adJk3n7733Hz+3w+Pv98Cbfc8n+kpqYxePBQli79nNNPH8PKlV/y61//FgBN00hPT2fBgncZN24CbrcbgMzMrCavu6hoiLUOgH/961UWL/4vAPv2lbBjxw4qKg4wYsTx1rjQvFOnns/dd9/GtGlXMn/+W0yZcn6T5xMEQUiEoig4bYF0y8cvHoZPN6x0y/unFvF/Z3rJz3DSJdPFPWcPYEBeOvkZTtwpdn73wQbeWb0Xm6pw3Sk9KKn28Na3gYeAJlgiDeCjjaXc8vq3fLHtAABPLt5KrVenZ3YKt5zRh4H5aRRkONlcWserX+2iZ3YKmS4bKXaNggwnI7pm0eALRAe7uVNYs6eaJVvLWbO3mt+fV4TbLUEAoe1p9won0PC63V+GILRrpg4pOKTo15HA5Qo/TX3uudmMHHkis2b9kT17dnPLLT9K+Bm7PZw+rWkauq7HjXE4wmNUVU04JhlLl35OTU0111wzHYCGhgacTiennz6m2XOE1mYE84UMw8Dn81nHUlJSrK9XrlzO8uXLeOaZubhcLm6++Qa8Xk/SeQsKCsnOzmHFii9Zu3YtM2f+7qDWJQiCkAhNVdAifARsqmKlKyqKwrlDCqPG//qsAfxsbB/qvDr5GU5M02RAXhon98zmv5vK+POnW/nx6F6cXZTP819sZ9n2Cuyagk83qfXqXDOqG0u2lnP7W2uaXFuP7BS2H6hPeOz9dfsp8xq8/dUubhvX12pQbpqmtXbTNPlf8QFO7uGWtgXCEaP9KxxTl4bXgiAkpKamhry8PADee++dFp+/R4+e7N69iz17dtO5cxcWLfow4biFC9/nzjt/zaRJZwNQX1/PZZedT0NDAyecMIo333ydadOutFIfR44cxS9/eQfTp19FVlYg9TEzM4vCws6sX/8dEyZM4rPPFieNENbW1pCRkYnL5WLbtmLWrl0NwJAhw3j00T+we/cuK/UxFFU777wLuO++3zB58hQrIikIgtDapDttpDsDt6eKojDt+EAGwFUnuMhLd3DWwDxsmsqvzhoAgF83uPbvX3HJcV24eHhnbjitF3/9Yhtzl+4A4MJhhZwzOJ8f/XOVdY6pQwooqWoAYGdFvVUvFxJ9j3+yBT7ZEpjfMNhT5SHFrvHNrkqcNpVZ5xZRUu3hzne+A+D7J3XnqhO6UVrrJS/dQb1PpzDThWGaKIQNXd5ctQfdNLlkRJcju4lCh6H9CzXDL6mPgiAk5KqrruF3v7uXF1/8K6eeOrrF53c6Xfz853dy22234HKlUFQUX6/R0NDA0qWfc8cdd1vvpaSkMHz4cSxZsphbb72dhx56gHfffQtV1bj99rsYOnQ41157HTfffAOqqjFgwEB+9at7Of/8i7jrrtu49torOPnkU6OiaJGcfPJpvPnmPK666lJ69OjJ4MFDAcjOzuaOO37Jr351B4Zhkp2dzWOP/QWA0aPH8uCD9zF1qqQ9CoJw9GHTVKYMjs/csGkqf7/mBOu106by49G9mXFKT7ZX1NMvNw2Av109kga/TkW9n7H9cqLmqKj3sWjDfi4c1pk3v93DHxZuIpRk+e9VgdTLLJfN6j/38zfXkGIPP9B6YdkOXli2I2rOIYUZlFR7GNYlkz+cV8Sm0lp+v2gT6Q6NC4YWcuNrqygur+OZy0dgmKa1TkVR+HL7AXrnpJErrQ2OeRQzFMdtZXw+nYqKusOaw+1ORX3yeHyFI6me9GQLraz943anHvbedjRkT+I53D3Zu3cbhYU9W3BFRweapqLrRrPH19XVkZqaimmaPPLIH+jevTuXX37VEVzhkWHdurU88cSj/OUvz8UdO9g9SfSzkZeXcdhrPJZoqb+R8nsvGtmTxMi+RNPg0/liVxUDO6XwxCdbueakbvTITmH1nmpS7Ro/f3MNPt3gp2P7MDAvjccXb+WrnZWc2iubDftrKav1Njr/gLw0NgT71IX43ondeGdNCdmpdraW1eG0qQzKT2fiwDwuHt6Zao+fpz8rJi/dwdQhBTT4DNwpNtypDjQFdMOkpMZDfroTVVGi0jV1M5B2erjIz0k8LbEnjf19bP+hKKlREwShDXnnnX/zn//Mx+/30b//QC644JK2XtJB8/LLL/Dmm69LbZogCAKBXnMXHteVioo6Zp1XZL1/cs+AG+77NwWahofEz8PnD2bFzkrG9csJGFcZJqf+6VMAbji1J1vL61i7t5prTurOHxZuZMP+WvLSHZzY3c1/vtsHwMvLdwIBwTWscwbf7qnmm91VfLO7ikc+3hy1vue+2G59ne7UyHDa2FMVrkMe1jmDmZMHkurQePijTazeU83ZRfl0znTi003e+nYvf7p4CHZVpcFv0CM7cXaG0Pa0/4jaY0V4e46nZly85faxijzxiEf2JB6JqCXmYKNHxwISUWt9JKJ2ZJA9SYzsSzyHuydLiw/wze7KqCbiAOv31eDXDfLSnaQ7bRSX11FS7eHPn25l8qA8a/x7a0tYtbuKHtkp/PWL7eSmOfjx6F688c0ePi8+cNDrcdpUPP743+OaqnDH+L7UeHTW7q1GUxUemDqID9fvZ3+Nl2nHd8GuqQH3zjQXnjoPmhLfSB2gxuMnzaElPLa/xkNeesfreycRtSZQDB2kRk0QBEEQBEE4Sji5VzYn94rvRzkwPz3q9eDCDAYXZjCuf27U+1MGF1g1eRcMK8Spqdg0lTP65lDvM9i4v4Y+OWks2rCfHp1SLLOUuyf1Z3BBOsXl9VQ1+Ombm8rIblkoisKSLeXsqWogL93BnW+vRVMVvLrJ7xduAiDNoVHr1SkqSOeJxVsBKK314vUbfLh+PweCjcyzXDZ6dUplf42HwYUZnDe0EJuq8JPXv2VgfjpzLh/BC8u2s2FfLTUeP6W1XnZVNnDT6b1o8OucO6QwKoq3s6Iev2HSq1OgJYJfN8RJM0i7j6hpf+yDZ8CF1JwhKTsh5MlYPLIn8UhELTESUYtHImqtj0TUjgyyJ4mRfYmnve3J4s1l5KQ5GFLYvN+1+2s8ZKfYafAb1Hj8ZLrslNd5ueivXwLgsqkMyE9n1e4qAI7vmsnOSg+dUu2s31fT7HW5U+xU1Pvi3k+xq4zoksVt4/sy4x9fU9ngp09OKrVenQN1Xmac2pMfnNyDr3dWkuGykZfuINNlx2+YeP0GdT6dinqfZcLS4NOtfn2tyVERUVu8eDEPPPAAhmFw2WWXccMNN0QdnzdvHg899BAFBQHlf/XVV3PZZZcdxpIPAlPHlBo1QRAEQRAE4RjljL45TQ+KIJSGmK6pVjuEVEc4yvXmjJMwTZPHPtlCbpqTW8f2xu1OpbKynk83l2HXFFIdNtwpdv65chfvr9vH9af2pEumk7+v2MW28jpe+/6JpNg1Hv3vZo7rmsVv3ltnzV/vM/hi2wEum7vcqvXbUhYWPH/5rJh3Vu9lR0WD9d41o7rxyaYytkX0v5s0MI8hhRk89dlWTuvVibH9cvh6VyUn98ym1qsHevelO9lb3cCgggwG5KXh1SRDSIoAABQUSURBVE1eXbkL3TC5ZlQ3Khr8zXLYNE2TBr8R5fh5pGkyoqbrOpMnT2bu3LkUFBRw6aWX8uijj9KvXz9rzLx581i9ejUzZ85s9olb6mmh7Q9dqB96LbWn/+aw5upItLenQK2B7Ek8ElFLjETU4pGIWusjEbUjg+xJYmRf4jlW92RXZT0um0ZOAuFyMHuSKH1xxY4KemSn8N7affTOSeW9tSWsK6nhron9WLmzkpHdsrjljdUc3y0Ld4qdrWW1jOufyysrdsXV16kKXHZcF/69ag9e3STNoeHTDbz6wSUKFmQ4Ka3xcEqvTnyx7QALf3wqDk3lwQ83MLgwg1E9spm3ag8XDivk/vc3UFxex6vXnkBhpuug9yQZhxVRW7VqFT179qR79+4ATJ06lUWLFkUJtTbF0MX1URCOQW655UdcffX3OfnkU633XnvtFbZv38btt9+d8DM333wDN9/8MwYNGsztt/+Ue+55gIyM6F+Qzz03G6czhSuv/F7Scy9e/F+6d+9B7959rM+MGHE8o0ad3AJXBo8//ggff7yQefPmo6qSpy8IgiC0Dl2zWsYBMlGN2Qnd3QBce1JAU0RGAU/p1QmAedeNolOanTRH+N7++lN6oinw2te7OblnNnurPfTqlELXrBQuP74ra/ZWc9agPGo9OrurGrBrCpe/sAKA56aPoMFvkJfu4I8fbebL7RVAoB2CVzf451e7AViytRyAcX/+Hz2zU9h2oJ75a/ehACbw6spd1noueG4Z7hQ7p/XuxC+nDsbeIjuWmCYVTklJCYWFhdbrgoICVq1aFTfugw8+4Msvv6R3797cfffddO7cuWVXGoNWsQXt7V+gGD5JfRSEY5CJEyezaNEHUUJt4cIP+PGPf9qsz//xj08c8rk//fS/nHbaaEuozZhx4yHPFYthGCxe/DH5+QV8/fVKRo48scXmjsTv92Ozye9OQRAE4eihe4JWAU5bQPRdeUI3APoG69JC40OfyXDZGOgKmLX8clJ/BuSlMaRzpjX28YuHsqW0jr65qdg0FcM0Oacon9JaL7e/tdYat7OygZ+e0RunTWNzaS0n98rmmSXFjOyWRVaKnc+2lJOf7mDJlnI+31LGGT3dLb8RQVrkr/S4ceM499xzcTgcvPrqq9x555289NJLjX5G0xTc7tRDP6niRskoxOh/Ds5h5+E4nLk6GJqmHt7edkBkT+I53D0pKVHQ2tCVacKESTz33NMYho7dbmfPnt2UlZUycuQJPPzwLL77bg0ej4dx4ybwwx/eBATshFVVRdNULrpoKnPn/g23O5sXXniO9957l+zsThQUFDBwYBGapvLWW/N46615+Hw+unXrzj333M+GDRv47LPFfP31Sl566XkefPBh5s59jtNPH8P48RP58sulPPnkY+i6TlHRYH7xi1/icDi46KKpTJlyLp999il+v58HHvgDvXr1jruur75aTp8+fZkw4SwWLXqfUaNOAqC8vIw//OFBdu8O9Nq5445fMnz4CN57711eeeUlFEWhX7/+3HPP77j//nus9QCMH386H320hJUrlzNnzl/IyMhk27ZiXnvtTe688+eUlOzF6/UybdoVXHhhoA/c558vYfbspzAMnawsN0888TSXX34Rc+a8QHZ2NoZhMG3ahTz77ItkZ0c7mynKYf5+FwRBEITD4KLh8QEju6YysCDsuqkqCkM6Z2KaJo9eOIR9NR4UReHiBJ8dH+HKeePpvayvj3SKbJNCraCggL1791qvS0pKLNOQEJF/pC+77DIefrjpnma6bh7mhXXCfdHz4TmOwTziZByredWNIXsSz+HuiWmaVt2Sc93ruL57taWWBkBD0XQ8gy5Nejw9PYOioiEsWfIpY8acyfvvL2DcuIkYhskPf3gTmZlZ6LrOrbfexPr16+nXrz+maWIYhrVuXTdZs2YNH374PnPnvoKu+7nuuqvp338Qum4wZsyZnHvuhQDMmfMX3nrr31x66XRGjz6D004bzbhxE629MAyDurp6fve7e3nssb/Qo0dP7r9/Jm+88RrTpl0JQGZmFs8//zfmzfsXf//7S9x1V3xt7QcfLGDChLMYM2Yss2f/GY/Hi81m45FHHuK4447nwQcfRtd16uvr2bhxI3PnPsfs2c/jdrupqqpE14246wxca+D1+vXreOmlf9KlS1d03eCuu35DZmYWHk8DM2ZcwxlnjMMwTH7/+9/x5z/PoUuXrtTUVGOacNZZ57BgwXymTbuSZcu+oF+//sF9jq4dMM343+9SoyYIgiAcjSiKwpiDNGNpLZp8HD5s2DCKi4vZsWMHXq+X+fPnM378+Kgx+/bts77+6KOP6Nu3b8uvVBAEIYaJEyezcOEHACxa9AETJ04G4KOPPuS6667iuuuuorh4C8XFW5LOsWrVV5xxxjhcLhdpaemMGTPWOrZly2Z+/OMZXHPN5Xz44QK2bk0+D8D27dvo3LkLPXoEjDTOOedcvv76K+v42LGB350DBxaxZ8+euM/7fD4+/3wJZ5xxJmlp6QwePJSlSz8HYOXKL7nwwoBw1TSN9PR0Vq78knHjJuB2B9IuMjOzGt8woKhoCF26dLVe/+tfr3LttVdwww0/YN++Enbs2MGaNd8yYsTx1risrMC8U6eez4IF8wGYP/8tpkw5v8nzCYIgCIJwaDQZUbPZbMycOZMZM2ag6zqXXHIJ/fv35/HHH2fo0KFMmDCBl19+mY8++ghN08jKymLWrFmtsXZBEI4SPIMubTT6daQYPXosTzzxKOvXr6OhoYFBg4rYvXsX//jH33j22ZfIzMzkgQfuxev1HtL8Dz74Wx588I/07z+A9957h6++WnFY67XbAy5aARdFf9zxpUs/p6ammmuumQ5AQ0MDTqeT008fc1Dn0TQNwwg4XxmGgc8X7mGTkhLO/1+5cjnLly/j/9u7/5ioznyP4++BWaotP0SUwaYsN7Ta6ypq3GukpEnjINAUUVDIpjWWsm5oiBYpXvpDl71ZUtm0scXaZLM0N/Y2tjGNJtCu3Bu2goq1iu2qS1Dc66ZhC1sYbpDyszAdeO4fJNNyB1uvCodhPq+/nDNnjt/zyRy+PJznnFNZ+Q5z5sxh58583O6Rm27X4YghMjKKP//5M65evcpvfqPnV4qIiEyVW7rA5LHHHqO2tpYTJ05QUDB+rceuXbtITk4GYPfu3dTU1PDRRx9x+PBhnVETkWlx7733snr1v/C735WRkjJ+Nm1wcJA5c+YSGhrKjRvdnD//6Q9uY+XK1Zw5c4qRkWGGhgb55JMG73tDQ4MsWLAAj8fDn/70XxP+36Eh32mjP/1pHB0dX9He3gZAbe1/smrV6lvenxMnannxxV9z7NgfOXbsjxw9+hGffdbI8PAwP//5GqqrjwHjj00ZGBhg9eo1nDxZR2/v+F2s+vp6AYiJWcRf/9oCwCefNODx+A4KAQYHBwgLC2fOnDn8/e+tXL3aDMCyZQn85S+X+Oqr8btc9fb2ej+TkbGJsrJS1q1LJjh4+h8uKiIiEih032cR8Wvr16fxt7/9t3fa4+LFS1iy5GGeeiqb3/721yQkrPzBzz/88D/jdKaQm/sUu3cXsnTpMu97v/pVAfn5z1BQ8Evi4v7Juzw5OZUjRw6Tl/cU//hHu3f5Pffcw549/0Zp6Ys8/fQvsNls3ptz/Jjh4WEaG8+RlPSod9ncuXNZsWIVZ882sGvXv3Lx4uc8/fQv2L59G62tXxAf/yC5ub9k5858cnOf5K23KgDYuDGLy5cvkpv7JM3NTRPOon3f2rVJjI6OsnVrNn/4w1v87GfLgfHrjktK9rB3bwm5uU9SWvqS9zOPPvoY33zzDenpmvYoIiIylX70gddTRQ/znDrKxZcy8aUHXk9OD7z29f1Mrl27ysGDb/D73//7TdfXA6/vnHrk1FAmk1MuvpSJL2Xiy/IHXouIiAAcPvwfVFcf07VpIiIi00ADNRERuSXbtj3Dtm3PWF2GiIhIQNA1aiIiIiIiIjOMBmoictssusRVZjB9J0RERO4ODdRE5LbY7SEMDvbpF3PxMsYwONiH3R5idSkiIiJ+T9eoichtiYxcSE/P/zAw8LXVpdxVNptNg8//4/+Tid0eQmTkwimuaOZoaGhg3759jI2NkZOTQ35+/oT33W43L7zwAleuXGHevHlUVFTwwAMPWFStiIj4Ew3UROS2BAfbWbBgkdVl3HW6/bAvZTK50dFRysrKeOedd3A4HGRnZ+N0OnnooYe86xw9epTw8HA+/vhjampq2L9/PwcOHLCwahER8Rea+igiInIbmpqaiIuLIzY2lpCQENLT06mrq5uwTn19PVlZWQCkpaVx7tw5nbEVEZFbooGaiIjIbXC5XMTExHhfOxwOXC6XzzqLFo2febbb7YSFhdHT0zOtdYqIiH/S1EcREZEZJjjYxrx5997hNoLueBuzjTKZnHLxpUx8KRNfU52JZQO1n/wkmIULw+54O3djG7ORcvGlTHwpk8kpF1/KxJfD4aCzs9P72uVy4XA4fNbp6OggJiYGj8dDf38/kZGRP7rtoKAggu7CnJegoOA738gso0wmp1x8KRNfysTXVGaiqY8iIiK3ISEhgdbWVtra2nC73dTU1OB0Oies43Q6qaqqAqC2tpbExERsNpsV5YqIiJ+xGV3VLCIicltOnz5NeXk5o6OjbNmyhYKCAt58802WL19OcnIyIyMjlJSU0NLSQkREBBUVFcTGxlpdtoiI+AEN1ERERERERGYYTX0UERERERGZYTRQExERERERmWE0UBMREREREZlh/Hag1tDQQFpaGikpKbz99ttWlzNtXn75ZR555BE2bNjgXfb111+Tl5dHamoqeXl59Pb2AmCM4ZVXXiElJYWMjAyuXLliVdlTqqOjg23btvHEE0+Qnp7Ou+++CyiXkZERsrOz2bhxI+np6Rw8eBCAtrY2cnJySElJoaioCLfbDYDb7aaoqIiUlBRycnJob2+3svwpNTo6SmZmJs8++yygTJxOJxkZGWzatInNmzcDOn78WaD2R1CPnIx6pC/1x5tTf/RlaY80fsjj8Zjk5GTz5ZdfmpGREZORkWGuX79udVnT4sKFC6a5udmkp6d7l7366qumsrLSGGNMZWWlee2114wxxpw6dcps377djI2NmUuXLpns7GxLap5qLpfLNDc3G2OM6e/vN6mpqeb69esBn8vY2JgZGBgwxhjjdrtNdna2uXTpkiksLDTHjx83xhhTWlpq3n//fWOMMe+9954pLS01xhhz/Phxs2vXLmsKnwaHDh0yxcXFJj8/3xhjAj6TdevWme7u7gnLAv348VeB3B+NUY+cjHqkL/XHm1N/9GVlj/TLM2pNTU3ExcURGxtLSEgI6enp1NXVWV3WtFizZg0RERETltXV1ZGZmQlAZmYmJ06cmLDcZrOxatUq+vr66Orqmvaap1p0dDTLli0DIDQ0lPj4eFwuV8DnYrPZuO+++wDweDx4PB5sNhvnz58nLS0NgKysLO+xU19fT1ZWFgBpaWmcO3cOMwtvCtvZ2cmpU6fIzs4Gxv/6FeiZTCbQjx9/Fcj9EdQjJ6Me6Uv9cXLqj7duuo4fvxyouVwuYmJivK8dDgcul8vCiqzV3d1NdHQ0AAsXLqS7uxvwzSkmJmbW59Te3k5LSwsrV65ULoxPYdi0aRNJSUkkJSURGxtLeHg4drsdmLjvLpeLRYsWAWC32wkLC6Onp8ey2qdKeXk5JSUlBAWN//jr6ekJ+EwAtm/fzubNm/nggw8A/VzxV+qPvvRd/o565HfUH32pP96cVT3Sfgc1ywxks9mw2WxWl2GJwcFBCgsL2bNnD6GhoRPeC9RcgoOD+fDDD+nr62PHjh188cUXVpdkqZMnTzJ//nyWL19OY2Oj1eXMGEeOHMHhcNDd3U1eXh7x8fET3g/U40dmn0D+LqtHTqT+OJH6481Z2SP9cqDmcDjo7Oz0vna5XDgcDgsrslZUVBRdXV1ER0fT1dXF/PnzAd+cOjs7Z21O3377LYWFhWRkZJCamgool+8LDw9n7dq1XL58mb6+PjweD3a7fcK+OxwOOjo6iImJwePx0N/fT2RkpMWV310XL16kvr6ehoYGRkZGGBgYYN++fQGdCeDd36ioKFJSUmhqatLx46fUH33pu6we+UPUH8epP96clT3SL6c+JiQk0NraSltbG263m5qaGpxOp9VlWcbpdFJdXQ1AdXU1ycnJE5YbY7h8+TJhYWHe07SziTGGvXv3Eh8fT15ennd5oOdy48YN+vr6ABgeHubTTz/lwQcfZO3atdTW1gJQVVXlPXacTidVVVUA1NbWkpiYOOv+wrp7924aGhqor6/njTfeIDExkddffz2gMxkaGmJgYMD777Nnz7J48eKAP378lfqjr0D/LqtH+lJ/9KX+ODmre6TN+OmVf6dPn6a8vJzR0VG2bNlCQUGB1SVNi+LiYi5cuEBPTw9RUVE899xzrF+/nqKiIjo6Orj//vs5cOAA8+bNwxhDWVkZZ86cYe7cuZSXl5OQkGD1Ltx1n3/+OVu3bmXJkiXeedXFxcWsWLEioHO5du0aL730EqOjoxhjePzxx9m5cydtbW08//zz9Pb2snTpUvbv309ISAgjIyOUlJTQ0tJCREQEFRUVxMbGWr0bU6axsZFDhw5RWVkZ0Jm0tbWxY8cOYPyajQ0bNlBQUEBPT09AHz/+LFD7I6hHTkY90pf64w9Tf/yO1T3SbwdqIiIiIiIis5VfTn0UERERERGZzTRQExERERERmWE0UBMREREREZlhNFATERERERGZYTRQExERERERmWE0UBMREREREZlhNFATERERERGZYTRQExERERERmWH+Fw8B7NiqvNBBAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Testing the model**\n",
        "---\n"
      ],
      "metadata": {
        "id": "LiG_akoCm1wu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "test_dataset=get_data('/content/drive/MyDrive/Kaggle/pneumonia_data/chest_xray/chest_xray/test')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YM8OPT-ytUYo",
        "outputId": "ba61734a-27b4-4769-a8ff-855b6de23a09"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:15: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray.\n",
            "  from ipykernel import kernelapp as app\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from torchvision.datasets import ImageFolder\n",
        "from torchvision.transforms import ToTensor\n",
        "import torch\n",
        "import torchvision\n",
        "from torch import *\n",
        "dataset = ImageFolder('/content/drive/MyDrive/Kaggle/pneumonia_data/chest_xray/chest_xray/train', transform=ToTensor()) \n",
        "test_dataset = ImageFolder('/content/drive/MyDrive/Kaggle/pneumonia_data/chest_xray/chest_xray/test', transform=ToTensor())"
      ],
      "metadata": {
        "id": "uwso9XNH0L3m"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "from keras.models import load_model\n",
        "model=load_model(\"/content/drive/MyDrive/model\")\n",
        "predictions = model.predict(x_val)\n",
        "predictions = predictions.reshape(1,-1)[0]\n",
        "prediction = np.where(y_val > 0.5, 1, 0) #y_val values are 0 or 1 ( binary ) and we need to compare them \n",
        "print(classification_report(y_val, (prediction[0:16]), target_names = ['NORMAL','PNEUMONIA']))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PfaQf81QvF8x",
        "outputId": "a3d1d14b-a78e-4e11-8a6b-c725bdccd3cb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "      NORMAL       1.00      1.00      1.00         8\n",
            "   PNEUMONIA       1.00      1.00      1.00         8\n",
            "\n",
            "    accuracy                           1.00        16\n",
            "   macro avg       1.00      1.00      1.00        16\n",
            "weighted avg       1.00      1.00      1.00        16\n",
            "\n"
          ]
        }
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [],
      "name": "CNN final.ipynb",
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}