Indian_Currency_Notes_Classifier.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Indian-Currency-Vanilla-CNN.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPMUe2EA1mGqXg2PUWKaTvS",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/r-prateek/Indian-Currency-Notes-Classifier/blob/main/Indian_Currency_Vanilla_CNN-15-epoch.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QV7-Lm9Dlcmv"
      },
      "source": [
        "#**Importing Dataset from Kaggle**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EQXmRG6BlhnK"
      },
      "source": [
        "! pip install --upgrade --force-reinstall --no-deps kaggle\n",
        "from google.colab import files\n",
        "files.upload()\n",
        "! mkdir ~/.kaggle\n",
        "! cp kaggle.json ~/.kaggle/\n",
        "! chmod 600 ~/.kaggle/kaggle.json\n",
        "! kaggle datasets list\n",
        "! kaggle datasets download -d vishalmane109/indian-currency-note-images-dataset-2020\n",
        "! unzip indian-currency-note-images-dataset-2020.zip -d dataset"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TwTtN458mb2p"
      },
      "source": [
        "#**Using Image Generator for Data Augmentation and initializing paths for the datasets.**\n",
        "\n",
        "Libraries used are:\n",
        "```\n",
        "tensorflow\n",
        "matplotlib\n",
        "```"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vahQLSzomzeN"
      },
      "source": [
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YsNN8tgWm5d5"
      },
      "source": [
        "train_path = \"/content/dataset/Indian currency dataset v1/training\"\n",
        "test_path = \"/content/dataset/Indian currency dataset v1/validation\""
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BGe2YaBLnTDZ"
      },
      "source": [
        "##Data Augmentation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gmGR0tSVnXNl"
      },
      "source": [
        "datagen = tf.keras.preprocessing.image.ImageDataGenerator(\n",
        "    rescale = 1./255,\n",
        "    horizontal_flip = True,\n",
        "    vertical_flip = True,\n",
        "    rotation_range= 5\n",
        ")"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "caGEDkprn5RQ"
      },
      "source": [
        "train_data = datagen.flow_from_directory(\n",
        "    train_path,\n",
        "    target_size = (190, 190),\n",
        "    color_mode = 'rgb',\n",
        "    batch_size = 32,\n",
        "    shuffle = True,\n",
        "    seed = 865,\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kDIfgId2ovw4"
      },
      "source": [
        "test_data = datagen.flow_from_directory(\n",
        "    test_path,\n",
        "    target_size = (190, 190),\n",
        "    color_mode = 'rgb',\n",
        "    batch_size = 32,\n",
        "    shuffle = True,\n",
        "    seed = 865\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "11sd3YStpJWm"
      },
      "source": [
        "#**Creating the Model**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oV3NyNHdpUTB"
      },
      "source": [
        "##Convolutional Neural Network"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XAQHBaWdpdMU"
      },
      "source": [
        "The convolutional neural network plays an important role in feature extraction from the input image, sufficiently changing the weights (filter)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jsfKM0hNpu1H"
      },
      "source": [
        "model = tf.keras.models.Sequential()\n",
        "\n",
        "model.add(tf.keras.layers.Conv2D(32, (3,3), activation = 'relu', input_shape = (190, 190, 3)))\n",
        "model.add(tf.keras.layers.MaxPooling2D((2,2)))\n",
        "\n",
        "model.add(tf.keras.layers.Conv2D(64, (3,3), activation= 'relu'))\n",
        "model.add(tf.keras.layers.MaxPooling2D((2,2)))\n",
        "\n",
        "model.add(tf.keras.layers.Conv2D(64, (3,3), activation= 'relu'))\n",
        "# model.add(tf.keras.layers.MaxPooling2D((2,2)))\n"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aH1Rrp5XwALH"
      },
      "source": [
        "##Artificial Deep Neural Network"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SuIFvYptwOOs"
      },
      "source": [
        "The Artificial Deep Neural Network flattens the input fed by the Convolutional network, and classifies images into respective classes.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TlQOdDa4whLi"
      },
      "source": [
        "model.add(tf.keras.layers.Dropout(0.5))\n",
        "model.add(tf.keras.layers.Flatten())\n",
        "\n",
        "model.add(tf.keras.layers.Dense(32, activation= 'relu'))\n",
        "model.add(tf.keras.layers.Dense(64, activation= 'relu'))\n",
        "model.add(tf.keras.layers.Dense(8, activation= 'softmax'))"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HQz6IP5exReB"
      },
      "source": [
        "##Compiling and Summarising the model\n",
        "\n",
        "*  Optimizer used is **Adam**\n",
        "*  Loss function used is **Categorical CrossEntropy**\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IQtGzPJ8xhUD"
      },
      "source": [
        "model.compile(optimizer= 'adam', loss = 'categorical_crossentropy', metrics = ['accuracy'])"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vgeG0ghTxqf3",
        "outputId": "cdcfa3c8-ca49-4bdd-ac9b-ef2054a31247"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 188, 188, 32)      896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 94, 94, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 92, 92, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 46, 46, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 44, 44, 64)        36928     \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 44, 44, 64)        0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 123904)            0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 32)                3964960   \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 64)                2112      \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 8)                 520       \n",
            "=================================================================\n",
            "Total params: 4,023,912\n",
            "Trainable params: 4,023,912\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xnPmUjhex44U"
      },
      "source": [
        "#**Training the Model**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BUd_2l-Tx-aR"
      },
      "source": [
        "The Sequential model is trained on the augmented training dataset, and the validation data set will be the augmented test dataset. The model will be trained for **15 epochs**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vpefjsWWyR_J",
        "outputId": "d2c0989c-4ec1-4615-ed28-591c1a26fcd2"
      },
      "source": [
        "history = model.fit(train_data, validation_data= test_data, epochs = 15, batch_size = 2000)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/15\n",
            "112/112 [==============================] - ETA: 0s - loss: 1.6462 - accuracy: 0.3878"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/PIL/TiffImagePlugin.py:788: UserWarning: Corrupt EXIF data.  Expecting to read 4 bytes but only got 0. \n",
            "  warnings.warn(str(msg))\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r112/112 [==============================] - 256s 2s/step - loss: 1.6462 - accuracy: 0.3878 - val_loss: 1.5699 - val_accuracy: 0.4058\n",
            "Epoch 2/15\n",
            "112/112 [==============================] - 256s 2s/step - loss: 1.5096 - accuracy: 0.4495 - val_loss: 1.5094 - val_accuracy: 0.4290\n",
            "Epoch 3/15\n",
            "112/112 [==============================] - 255s 2s/step - loss: 1.3966 - accuracy: 0.4966 - val_loss: 1.4217 - val_accuracy: 0.4812\n",
            "Epoch 4/15\n",
            "112/112 [==============================] - 255s 2s/step - loss: 1.2755 - accuracy: 0.5446 - val_loss: 1.2664 - val_accuracy: 0.5333\n",
            "Epoch 5/15\n",
            "112/112 [==============================] - 254s 2s/step - loss: 1.1518 - accuracy: 0.5791 - val_loss: 1.1937 - val_accuracy: 0.6029\n",
            "Epoch 6/15\n",
            "112/112 [==============================] - 255s 2s/step - loss: 1.1374 - accuracy: 0.5900 - val_loss: 1.1634 - val_accuracy: 0.6116\n",
            "Epoch 7/15\n",
            "112/112 [==============================] - 255s 2s/step - loss: 0.9834 - accuracy: 0.6528 - val_loss: 0.9764 - val_accuracy: 0.6609\n",
            "Epoch 8/15\n",
            "112/112 [==============================] - 256s 2s/step - loss: 0.8984 - accuracy: 0.6873 - val_loss: 0.9496 - val_accuracy: 0.6667\n",
            "Epoch 9/15\n",
            "112/112 [==============================] - 256s 2s/step - loss: 0.8116 - accuracy: 0.7047 - val_loss: 0.8486 - val_accuracy: 0.6986\n",
            "Epoch 10/15\n",
            "112/112 [==============================] - 256s 2s/step - loss: 0.7194 - accuracy: 0.7490 - val_loss: 0.8505 - val_accuracy: 0.7246\n",
            "Epoch 11/15\n",
            "112/112 [==============================] - 256s 2s/step - loss: 0.6233 - accuracy: 0.7858 - val_loss: 0.8733 - val_accuracy: 0.7333\n",
            "Epoch 12/15\n",
            "112/112 [==============================] - 257s 2s/step - loss: 0.5859 - accuracy: 0.7958 - val_loss: 0.8342 - val_accuracy: 0.7739\n",
            "Epoch 13/15\n",
            "112/112 [==============================] - 255s 2s/step - loss: 0.5621 - accuracy: 0.8090 - val_loss: 0.7383 - val_accuracy: 0.7710\n",
            "Epoch 14/15\n",
            "112/112 [==============================] - 254s 2s/step - loss: 0.5528 - accuracy: 0.8099 - val_loss: 0.7889 - val_accuracy: 0.7449\n",
            "Epoch 15/15\n",
            "112/112 [==============================] - 254s 2s/step - loss: 0.5011 - accuracy: 0.8228 - val_loss: 0.7179 - val_accuracy: 0.7681\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vJMluZ5vTNBx"
      },
      "source": [
        "#**Plotting the graph of accuracy and loss against their validation values**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XLrYGUNDTiWv"
      },
      "source": [
        "Using the plot function in ``` matplotlib.pyplot```, we can plot the graphs of accuracy vs validation_accuracy, as well as the graph of loss vs validation loss. \n",
        "\n",
        "We can also determine if our model is *overfitting* by analysing these graphs."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fc8Xi5-VT5v4",
        "outputId": "d56ab5cc-c5b2-4faf-fd48-006aed3ec97f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "plt.plot(history.history['accuracy'], label = 'accuracy')\n",
        "plt.plot(history.history['val_accuracy'], label = 'validation accuracy')\n",
        "plt.xlabel('epochs')\n",
        "plt.ylabel('accuracy')\n",
        "plt.legend('lower right')"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fd5b99e2e50>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dopr7to5VTd0",
        "outputId": "d4f9253e-cce0-4b1c-c1e5-b021fb79523c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "plt.plot(history.history['loss'], label = 'loss')\n",
        "plt.plot(history.history['val_loss'], label = 'validation loss')\n",
        "plt.xlabel('epochs')\n",
        "plt.ylabel('loss')\n",
        "plt.legend('lower right')\n"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fd5b98dc310>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3hUZfrG8e+T3guQ0EJvoZeEDirYUBAVkZUOgmDDXnfXn2Vd17VgAVEpggoiiiIqdlCQTkLvIggEAoSSBNLL+/vjjG6UJATI5Mxkns91zQUzc+bMPVwk95z2vmKMQSmllOfysjuAUkope2kRKKWUh9MiUEopD6dFoJRSHk6LQCmlPJyP3QHOV7Vq1Uz9+vXtjqGUUm4lMTHxuDEmqrjn3K4I6tevT0JCgt0xlFLKrYjI/pKe011DSinl4bQIlFLKw2kRKKWUh3O7YwRKKWWXvLw8kpKSyM7OtjtKiQICAoiJicHX17fMr9EiUEqpMkpKSiI0NJT69esjInbHOYsxhhMnTpCUlESDBg3K/DrdNaSUUmWUnZ1N1apVXbIEAESEqlWrnvcWixaBUkqdB1ctgd9dSD6PKYLjZ3J4+ott5OQX2B1FKaVciscUwZq9J5m54jcemLeJgkKdg0Ep5Z5CQkLKfZ0ec7C4b5uaJKc159lFOwgP8uXfN7Ry+U08pZSqCB5TBABjezbkZEYuU376lSpBfjx0dTO7IymllO08qggAHr66Gacyc5n84x4ignwZ27Oh3ZGUUm7o6S+2sf1wermus0WtMJ68rmW5rrMsPK4IRIRnb2hNamYezy7aQWSQHzfFxdgdSymlbOO0IhCRd4B+wDFjTKsSlrkMeBXwBY4bYy51Vp6ivL2EV29pR/qsdTzyyWbCA325okX1inhrpVQlYcc3d2dx5llDs4A+JT0pIhHAFKC/MaYlcLMTs5zF38ebt4fH06pWGHd9sJ41e09U5NsrpZTLcFoRGGOWASdLWWQI8Kkx5oBj+WPOylKSEH8fZo7uRExkIGPfTWDrobSKjqCUUraz8zqCpkCkiPwkIokiMqKkBUVknIgkiEhCSkpKuYaoEuzH+2M6Exrgw6iZa9l3PKNc16+UUuXpzJkz5b5OO4vAB4gD+gJXA0+ISNPiFjTGTDXGxBtj4qOiip1p7aLUigjkvTGdKTQwbPoajqS57siCSilV3uwsgiTgW2NMhjHmOLAMaGtXmMbRIcwa3ZHUzFxGvLOG1Mxcu6IopVSFsrMIFgI9RMRHRIKAzsAOG/PQJiaCaSPi+e14JqNnrSMzN9/OOEopVSGcVgQiMhdYBTQTkSQRGSMit4vI7QDGmB3AN8BmYC0w3Riz1Vl5yDkNG+ZAfunf9Ls1rsbrg9ux6WAqt89eT25+odMiKaWUK3DmWUODjTE1jTG+xpgYY8wMY8xbxpi3iizzojGmhTGmlTHmVWdlAWDbAlh4J0zqAGunQV5WiYv2aVWT/wxozbLdKTzw0UYdpE4pVal5zOijtB8OQ+dDWG346iF4tQ2seM3aUijG3zrW5bFrYvlyczJPfb4NY7QMlFKVk+cUgQg0uRJu/QZGLYLqLeD7/4NXWsFPz0Pm2Zc83H5pI8Zf0pD3V+/nlR9+sSG0Uko5n8eNNYQI1O9h3ZIS4eeX4Kf/wMpJ0HEMdL0bQqL/WPyxa2I5lZnL64t/ITLIl9Hdyz4PqFJKuQPP2SIoTkwcDJ4Ld6yEpldbZfBqa/jqYUg9CFiD1D13Y2uualGdp7/YzmcbDtkcWinlySZOnEirVq1o1aoVr75aPodWPW+LoDjVW8LAd6DXP2D5REh4x7q1vQV6PIBP1Ua8Prg9o2au5aGPNxEW6EPvWB2kTimP9vVjcGRL+a6zRmu45vkSn05MTGTmzJmsWbMGYwydO3fm0ksvpX379hf1tp69RfBXVRvB9W/APRsh/lbYMh8mx8P8Wwk4uZNpI+KJrRnKHbPXs+630oZRUkqp8rd8+XJuvPFGgoODCQkJYcCAAfz8888XvV7dIihORB249kXo+RCsfgPWzYCtnxDa7Fpm97mHAQsDuXXWOj4a35XmNcPsTquUskMp39zdjW4RlCa0Olz5DNy3BS57HPavJGJOH76KfJmePjsZMWMN+0/oIHVKqYrRs2dPPvvsMzIzM8nIyGDBggX07NnzoterRVAWQVXgssfg/q1wxdMEnNzBlPz/Y1r+P3hj6hSOpZV8cZpSSpWXDh06MGrUKDp16kTnzp0ZO3bsRR8fABB3u1AqPj7eJCQk2BsiLwvWv0/uslfwyzjMDu+mVBs2k6gGxU7EppSqJHbs2EHz5s3tjnFOxeUUkURjTHxxy+sWwYXwDYTO4/C7fxN7Oj9HjfxkQt/tRcriSeBmxaqUUloEF8PHj8bX3MXRYUtIlBZE/fxP0qb3h/Rku5MppVSZaRGUg9gmTalz11e84jcOv6TV5E7uAts+szuWUsoJXH13+oXk0yIoJ3WrBTP8nmd5MHISO7KrwMcj4dPxkK3zICtVWQQEBHDixAmXLQNjDCdOnCAgIOC8XqcHi8tZRk4+E2avpe2+aUzwWYiE10RufNsa20gp5dby8vJISkoiO9t1p7MNCAggJiYGX1/fPz1e2sFiLQInyCso5NFPNrNvw09MC51G1dxDSLe7ofcT4ONvdzyllAfSs4YqmK+3Fy/f3JYul/ahZ/ozLAvtaw1oN7UXHHHeJGxKKXUhtAicRER4tE8sj1zXgVHHh/BcxNMUZhyDab1gxetQqFNgKqVcgxaBk43u3oBJg9szK6UZg2QiWfUvh++fgPf6Q+oBu+MppZQWQUXo16YWs27tyK7T/vQ6MJbky16Gwxvgze6w6UO9CE0pZSstggrSrVE15o3vSiFw9U912Nh3EUS3gAXj4eNRxU6VqZRSFUGLoAK1qBXGp3d2o1qoP4M+TubrjjPg8idh5yKY0hX2/GB3RKWUB9IiqGAxkUF8cns3WtYK4865m3jfZwDcthgCI2D2TdY0mbmZdsdUSnkQLQIbRAb78cHYLvRuFs0TC7fx0uYAzG0/Qpc7Ye1UePsSOLTe7phKKQ+hRWCTQD9v3h4ex9/i6zD5xz08+vkv5F/5bxixEHIzYMaV1sxoSinlZDpVpY18vL14/qbWVA/z5/Ulezh+JpfJQ3oQdOdK+HQcfPUQRNaHxpfbHVUpVYk5bYtARN4RkWMiUuqltCLSUUTyRWSgs7K4MhHhgaua8a8bWvHjrmMMmbaGk4XBMHAmRDWH+aPhxK92x1RKVWLO3DU0C+hT2gIi4g38F/jOiTncwvAu9XhzaAe2J6cz8M2VHMzwgsEfgHjBh0Mg57TdEZVSlZTTisAYsww418nxE4BPgGPOyuFO+rSqyewxnTl+JocBb65ke1YVuPldOP6LNaS1DkuhlHIC2w4Wi0ht4EbgzTIsO05EEkQkISUlxfnhbNSpQRXm39ENHy/hlqmr2ODTBq5+DnYtgqXP2x1PKVUJ2XnW0KvAo8aYc37NNcZMNcbEG2Pio6KiKiCavZpWD+Wj8V2JCPJj2PQ1rIkaCO2GwdL/wvbP7Y6nlKpk7CyCeOBDEfkNGAhMEZEbbMzjUupUCeKj8V2pHh7AyFnrWBH7d4jpCAtuh6Pb7I6nlKpEbCsCY0wDY0x9Y0x9YD5wpzFGJ/otokZ4APPGdaV+1WBGz97Mzx0mgn8ozB2sYxMppcqNM08fnQusApqJSJKIjBGR20Xkdme9Z2UUFerPh+O6EFsjlNHzk1gR/xqcTrbmRC7ItzueUqoS0Kkq3UR6dh63zlzH+gOn+LjLPuI2/MMakqLPf+yOppRyAzpVZSUQFuDLu7d2okvDqgxc3YCd9YbB6imwYY7d0ZRSbk6LwI0E+/vwzqiOXNo0ir67ruZwZCf48j5I8rwtJKVU+dEicDMBvtZgdVe0rMW1yWNI842CD4dCerLd0ZRSbkqLwA35+3jzxpAOXNquGYPSJpCbmYaZNwzysu2OppRyQ1oEbsrH24uJg9rRLq4792SPRw4lYBY9oPMfK6XOmxaBG/P2Ev4zoDXVO9/Ma/kDkI1zKFzztt2xlFJuRovAzXl5CU/1b0lm14f4riAO883fKfj1J7tjKaXciBZBJSAiPHZtC/Z0f4lfC2uQPWc4ecf32R1LKeUmtAgqCRHhzj4dSOj6BvkF+RyZOoCczDS7Yyml3IAWQSUz5JperI17mVo5+9g8aSjZuToMhVKqdFoEldCV/YewpcWDdMz6mc8nPUBGjpaBUqpkWgSVVLtB/+RgzHUMOv0ur095nfTsPLsjKaVclBZBZSVCnZHTSItsxYTUF3jszY85lZFrdyqllAvSIqjMfAMJH/URvoHBPJL6NLe9/QMpp3PsTqWUcjFaBJVdeG38h8yhrvcJ7k17nsFvL+dImg5FoZT6Hy0CT1C3C159X6KnbGLI6Vn8beoqDqVm2Z1KKeUitAg8Rdwo6DiWW+Vzbst4m7Fvfc/Bk5l2p1JKuQAtAk/S53mIH8NQvuGD7Dv5cMpTHEg5bXcqpZTNtAg8ibcv9JuIjF+GX82WPJz/NjlTunNk43d2J1NK2UiLwBPVbEPwuG84eMVbBJtManx2M2feGwynfrM7mVLKBloEnkqEOj0Gc2bsSqZ4DcZ772IKJ3eCxc9Azhm70ymlKpAWgYdrGhPNVbe/yE0+k/i6sDP8/DJMioONc6Gw0O54SqkKoEWgaBwdyqTxfXnG915Gej1HVmAN+Ox2mHEFHFxndzyllJNpESgAGkWFMG9cV37xjaXb8cdJunQipB2yyuDT8ZB+2O6ISikn0SJQf6hfLZh547sS5O/HtUtj2HrTEuj5IGxbYO0uWvYi5OmFaEpVNk4rAhF5R0SOicjWEp4fKiKbRWSLiKwUkbbOyqLKrk6VIOaN70J4kC+D393KhiYT4K410PhyWPIsvNEJtn0GxtgdVSlVTpy5RTAL6FPK8/uAS40xrYF/AVOdmEWdh5jIIOaN60rVYD+Gz1hL4ulw+NtsGPkF+IfBxyNhVj84ssXuqEqpcuC0IjDGLANOlvL8SmPMKcfd1UCMs7Ko81crIpAPx3UlOtSfETPWsnbfSWhwCYxbCn0nwrHt8PYl8MV9kHHc7rhKqYvgKscIxgBfl/SkiIwTkQQRSUhJSanAWJ6tRngAH47rQs2IQEa+s5aVvx4Hbx/oOAbuWQ+db4cN78PrHWD5q3B8j+4yUsoNiXHiD66I1Ae+NMa0KmWZXsAUoIcx5sS51hkfH28SEhLKLaM6t5TTOQydvpoDJzOZPqIjPZpUK/Lkbvj2cdjzg3U/OBrqdfvfLboFeHnbE1wp9QcRSTTGxBf7nJ1FICJtgAXANcaY3WVZpxaBPU6cyWHo9DXsO57B1BHxXNo06s8LpOyG/SvgwCr4bQWkJ1mP+4dD3S7/K4aa7cDHr+I/gFIeziWLQETqAkuAEcaYlWVdpxaBfU5l5DJsxhp+OXqGt4Z3oHds9ZIXTj0A+1dZ5bB/JZz4xXrcJxDqdIS6jmKI6Qh+QRXzAZTyYLYUgYjMBS4DqgFHgScBXwBjzFsiMh24CdjveEl+SSGL0iKwV1pmHsPfWcOO5HSmDI3jyhallEFRZ45ZWwv7V1q3I1sAA14+UKu9Y4uhO9TpDIERTv0MSnki27YInEGLwH5pWXmMfGctWw+lMXlIe/q0qnn+K8lKhYNr4YCjGA6th8I8QKB6K0cxdIWGl0FgZDl/AqU8jxaBKnens/MYNXMdGw+m8vot7enb5gLKoKjcTDiU6NhiWAFJ6yAvE8Jqwx0rtAyUukhaBMopzuTkc+vMdSQeOMV/b2pDvzY1CfAtpzOECvLg1x9h7i3Q+mYY8Hb5rFcpD6VFoJwmMzefW2etY/Ve69rByCBfaoYHUisikFoRAY6/B1ArIpCa4QFUDwvA1/s8Ll/58TlY+l+45QOI7eukT6FU5adFoJwqO6+Ab7cdIelUFodTrVtyWjaHU7NIz87/07JeAlGh/lZRhFvlUDMikFrhjrKICKBasD9eXmK9ID8XpveG00etMY+CqtjwCZVyf1oEyjYZOfkkp2VxKDWb5NQsDqdZf/5eFIfTssjO+/MEOH7eXlQP96dWeCCXNovijmaZyLTe0KI/DHzHpk+ilHsrrQh8KjqM8izB/j40jg6lcXRosc8bY0jNzONwWhbJqdkcTsvicGo2yWlZ7E3J4IVvdhHg04JbL30UfnwWmveHljdU8KdQqnLTIlC2EhEig/2IDPajZa3wPz1XWGi4fXYizy7aTpNRw+hZ80tY9IB1vUFIVAlrVEqdL1cZdE6ps3h5Ca/8rR1NokO5+8MtHLrsFcg5DYvu18HtlCpHWgTKpQX7+zBtRDwiMHLRaXJ6PgY7voCtn9gdTalKQ4tAuby6VYOYMrQD+45ncNe+7pjaHWHRg3D6iN3RlKoUylQEInKviISJZYaIrBeRq5wdTqnfdWtUjSeva8EPu04wo9rDkJ9tTYqju4iUumhl3SK41RiTDlwFRALDgeedlkqpYgzvUo/Bnerw7Jp8tja/F3Z/DZs+tDuWUm6vrEXguLqHa4H3jTHbijymVIUQEZ7u34qO9SMZtLEtGdU7wtePQtohu6Mp5dbKWgSJIvIdVhF8KyKhQOE5XqNUufPz8eLNYXFEBgcy6tQoTEEufHGP7iJS6iKUtQjGAI8BHY0xmVjzCox2WiqlSlEtxJ+pI+LYklWVGYGjrGkyN7xvdyyl3FZZi6ArsMsYkyoiw4B/AmnOi6VU6VrWCuflm9vx75Tu7Aluj/nm79asaEqp81bWIngTyBSRtsCDwK/Ae05LpVQZ9G1Tk7t7N2XUyVHkFxTAwrt1F5FSF6CsRZBvrNHprgcmG2PeAIofPEapCnT/FU1p3rwVT+cMhn1LIWGG3ZGUcjtlLYLTIvI41mmji0TEC8f8w0rZ6fdhKNZW6c8q2lD43RNwcp/dsZRyK2Utgr8BOVjXExwBYoAXnZZKqfMQ4u/D9JGdeEruICsP8hfcCYV6UptSZVWmInD88p8DhItIPyDbGKPHCJTLqFs1iCeHXsXT+cPxObiSwjU6taVSZVXWISYGAWuBm4FBwBoRGejMYEqdr26Nq9HymjtYUtCOgu+fhBO/2h1JKbdQ1l1D/8C6hmCkMWYE0Al4wnmxlLowI7rVZ1XLJ8ks8ObknDFQWGB3JKVcXlmLwMsYc6zI/RPn8VqlKoyI8PDAy5gVfhdVTm4g+duJdkdSyuWV9Zf5NyLyrYiMEpFRwCLgK+fFUurC+fl4MWTsQyzz6kSVNf/l5G9b7I6klEsr68Hih4GpQBvHbaox5lFnBlPqYkSFBRA1+E0yTQAnZo8hJzfH7khKuawy794xxnxijHnAcVtwruVF5B0ROSYiW0t4XkTkdRHZIyKbRaTD+QRX6lyaN2nM3k5P0SR/F0tmPIHRq46VKlapRSAip0UkvZjbaRFJP8e6ZwF9Snn+GqCJ4zYOaxgLpcpVXN+x7Kp6Ob2PzGDhdz/YHUcpl1RqERhjQo0xYcXcQo0xYed47TLgZCmLXA+8ZyyrgQgRqXn+H0Gp0jUZ9TY53iE0XvEwK3cn2x1HKZdj55k/tYGDRe4nOR47i4iME5EEEUlISUmpkHCq8vAKjcL3+ldp5bWPjR/8H/tPZNgdSSmX4hangBpjphpj4o0x8VFRUXbHUW4osO2NZDS9kdv4hP/M/IgzOfl2R1LKZdhZBIeAOkXuxzgeU8opgm+YiAmowr3pExn29jJ2Hz1tdySlXIKdRfA5MMJx9lAXIM0YoztwlfMEVcHvxkk09zrAcycf5OfJ4/n+4zcpOLlf5zFQHk2cdUqdiMwFLgOqAUeBJ3EMXW2MeUtEBJiMdWZRJjDaGJNwrvXGx8ebhIRzLqZUyVa/Rd7mjyF5E74mD4D8wCh86sRD7Tio3cG6BUbaHFSp8iMiicaY+GKfc7dzq7UIVHkx+TksXb6UFT99S/PC3fQKTSIyY+//FqjaGGo7yiEmDqq3Ah9/+wIrdRG0CJQqxbH0bB7/dAuLdx7j0nr+vNC1gOrpW+DQejiUAGeOWgt6+0GN1kXKIR6qNAQRez+AUmWgRaDUORhjmJ+YxDNfbKfAGB6/JpahnevhJUD6ITiUCEkJVjkc3gB5jlNQAyIcu5PiIKYjNOoN3j62fhaliqNFoFQZHU7N4tFPNvPzL8fp3rgqLwxsS+2IwD8vVJAPx3c5iiHRuh3bDqYQGlwKN8+CoCq25FeqJFoESp0HYwwfrD3AvxftwEuEJ/o1Z1B8HaS0XUC5GbD5I/j6EQitAbd8YO1GUspFlFYEbnFBmVIVSUQY2rke3953CS1rhfHoJ1u4ddY6jqZnl/wiv2CIHw2jv7G2GKZfCVvmV1xopS6CFoFSJahTJYi5t3XhyetasGrvCa6cuJQFG5JKH8U0Jg7G/QS12sEnY+C7J3SWNOXytAiUKoWXlzC6ewO+uqcnjaNDuH/eJsa/n0jK6VLmNwitDiM+h/gxsPJ1mH0TZJY2/qJS9tIiUKoMGkaF8PHt3Xj8mlh+2p3CVa8sZdHmUi6E9/GDfhOh/yTYvwKm9YKj2yousFLnQYtAqTLy9hLGX9qIRRN6UKdKEHd9sJ67P1jPyYzckl/UYQSM+gryc2D6FbD104oLrFQZaREodZ6aVA/l0zu68eCVTfl22xGuemUZ3207UvIL6nS0jhvUaA3zR8P3T+pxA+VStAiUugA+3l5MuLwJC+/qQVSoP+PeT+SBeRtJy8wr/gWhNWDklxA3Gla8CnNuhqxTFRtaqRJoESh1EVrUCmPhXd25p3djFm46zNWvLmP5L8eLX9jHD657Fa57DfYtg6m94Oj2ig2sVDG0CJS6SH4+XjxwVTM+vaMbQf7eDJuxhqc+30ZWbgm7f+JGwahFkJdpHTfYvrBC8yr1V1oESpWTtnUiWDShJ6O61WfWyt/o+/rPbDyYWvzCdTvDuKVQvQV8NAIW/0uPGyjbaBEoVY4C/bx5qn9L5oztTFZeATe9uZKJ3+8mr6Dw7IXDalpbBh1GwM8vwdxbIKuE4nAXuZlQWMxnVS5NxxpSyknSsvJ4+vNtfLrhEK1rh/PK39rSODq0+IUT3oGvHoGIutY4RdGxFRv2fBUWwql91rURR7fB0a3W7dRvEFIdmvaB2H7Q4BLwDbA7rUIHnVPKVl9vSebvC7aQkVvAo31iGd2tPl5exQxgd2A1zBtuHTu48S1ofl3Fhy1Odpp1UPv3X/ZHt1n3fx+KW7ygSiOo3hKim0PKTvjlB8g9DX4h0PhyqxSaXKmzvtlIi0Apmx07nc3jn1iT33RtWJWXBhUzvDVA+mGrDA4lwCWPwGWPg1cF7cEtLICT+/7yC38rpB743zIBEdZMbTVaWb/4q7eEqObgF/TndeXnwL6fYdci2PkVnDkCXj5QrzvE9oVm10JEnYr5XArQIlDKJRhj+CjhIM98sR0vEZ7s35KbOtQ+e3jr/BxY9ABsmG3tYhkwFQLCy/5GhYVQmP+XW8HZ988c/fMv/WM7rK0RsL7lV21S5Bd+K+sWVuv8Z2QrLITD62HnIut2fJf1eM220KyvVQzVW+pMb06mRaCUCzlwIpOHPt7E2t9OcnXL6jx3Y2uqhvxlLmRjYN10+OYx61t4UJXSf6kXfYzz/JkOrOL4hV/kl35UM/AtZoulPBzf878thYNrAGMdG4ntZ20p1O2qs7w5gRaBUi6moNAwY/leXvp2N2GBPvxnQBuubFH97AX3r4K1UwFj7Vrx8gEvb+tP8T77sVLv/+WxwCrWL/7QGvZ9Gz9zDHZ9Dbu+gl9/hIIc6zhC0z7WlkKj3tZcD+qiaREo5aJ2Hknn/nmb2JGczqD4GJ7o14LQAF+7Y9kj5wz8usTafbT7G8hOBZ8AaNgLmveD1oOsq7PVBdEiUMqF5eYX8tri3bz506/Uigjk5Zvb0rlhVbtj2asgDw6schxX+ArSDkDtOGs+6Ii6dqdzS1oESrmBxP0neeCjTRw4mcnYHg148KpmBPh62x3LfsZYw3B8PsE6iD1gKjS92u5UbkfnLFbKDcTVq8JX9/RkaOe6TPt5H/0nL2froTS7Y9lPBFreYA3lHV4HPhgEPzxtzQ2tyoVTi0BE+ojILhHZIyKPFfN8XRH5UUQ2iMhmEbnWmXmUcnXB/j48e0NrZo3uSGpmHjdOWcEbP+6hoNC9ttydomojGPu9NSTH8onw3vVwupR5IFSZOa0IRMQbeAO4BmgBDBaRFn9Z7J/AR8aY9sAtwBRn5VHKnVzWLJrv7r+Eq1vW4MVvdzFk2mqOpGXbHct+voHW9J83vGVdm/BWT2tIb3VRnLlF0AnYY4zZa4zJBT4Erv/LMgYIc/w9HDjsxDxKuZWIID8mD+nAyze3ZcuhNK55bRmLdxy1O5ZraDcYblsCgRHWlsGyF3Wwu4vgzCKoDRwscj/J8VhRTwHDRCQJ+AqYUNyKRGSciCSISEJKSoozsirlsm6Ki+GLCT2oGR7ImHcTePqLbeTk65DVRDeH236ElgNgybPWsYPMk3anckt2HyweDMwyxsQA1wLvi8hZmYwxU40x8caY+KioqAoPqZTdGkWFsOCubozqVp+ZK35jwJSV7E05Y3cs+/mHwE3Toe/LsG+ptavo4Dr78uTnwLYFsOMLt9pCcWYRHAKKjioV43isqDHARwDGmFVAAFDNiZmUclv+PtZcB9NHxHM4NYt+k5bzSWKS3bHsJwIdx8Kt31oD9M3sA6vftE47rSjH98C3/4CXY+HjUTBvGEy/3Loy3A04swjWAU1EpIGI+GEdDP78L8scAC4HEJHmWEWg+36UKsUVLarz9b2X0Lp2OA9+vIn7523kTI6eSkntDjB+GTS5yhqj6aMR1hDazpKfA1vmw6x+MDkO1rwF9XvAsE+tg9mnk61S+mikNaqrC3PqBWWO00FfBbyBd4wx/xaRZ4AEY8znjrOIpgEhWAeOHzHGfFfaOvWCMqUsBYWGyUv28Nri3dStEsSkwR1oHXMeo5RWVsbAytetaw0i68HN70LNNuW3/hO/QuIs2DgHMk9YVzp3GAnth1njNv0uNwNWToYVr1qDAXa+HS556PxGki1HemWxUpXYmr0nuBCCm+kAABB+SURBVG/eRo6fyeHRPrGM6dHg7KGtPdH+lTD/VusA8rUvWtcfXOi/S34u7PwSEmdap6uKNzS7BuJHQ8Pepc8ZkZ4MS/4FGz+wRpG97HGIG13hI6xqEShVyZ3KyOWRTzbz/faj9GoWxUs3tz17aGtPdCYFPh0Le3+CtoOtg8rnM5rpyb2Q+K717T8jBcLrQtwIaDfMmnP6fBzeaB1H2L8cqjWDq/8Nja+osJFftQiU8gDGGN5fvZ9nF+0gItCXV29pR7dGeu4FhQWw9AVY+l+IioVB70FU05KXL8izBrtLnGkViHhbw2LHj7aGxfa6iPGfjLHW/f0TVsk06g1X/Ruq//Va2/KnRaCUB9l+OJ27565n3/EM7rqsMfdd0QQfb7vPFHcBexbDp7dBXjb0fx1aD/zz8yf3wfr3rJnhMo5BWAzEOfb9h9Uq3yz5udbEQ0ufh5zT1jGGXn+HkOjyfZ8itAiU8jCZufk8uXAbHycmEVcvktduaUdMZNC5X1jZpR2C+aOtmdHix8CVz1hzICTOtP4UL2hytfXtv/EVF/ftvywyT1pbKuumg08g9HwAutwJvgHl/lZaBEp5qIUbD/GPBVvxEnhhYBv6tDrP/dqVUUEe/PAUrJoMXr5QmAdhta2Dye2HQ/hfB0CoAMd/ge//z5qpLbwuXPEktLqpXI8faBEo5cH2n8hgwtwNbE5KY1iXuvyzbwud5wCsCW92LbLmSm58pWvMk7x3qXVA+egWiOkIV/8H6nQsl1VrESjl4XLzC3npu11MXbaX2BqhTBrcnibVQ+2OpYpTWGCdarrkX3DmqLVlcMVTFz0zm05Mo5SH8/Px4u/XNmfm6I6knM7husnL+XDtAdzti6BH8PKGDsNhwnq45BHrLKNJ8dYFctnpznlLp6xVKeWSejWL5ut7exJXL5LHPt3Cgx9vIjNXh6dwSf4h0PsfMCHRmqFt+UT44UmnvJXuGlLKAxUUGiYt+YXXFv9Ck+gQ3hwWR6OoELtjqdIcSoSQGhd8MFt3DSml/sTbS7jviqa8O7oTx8/k0n/Scr7YpPNCubTacU47o0mLQCkPdknTKBbd04NmNUKZMHcDTy7cqpPeeCAtAqU8XM3wQOaN78qYHg14d9V+Br21iqRTmXbHUhVIi0Apha+3F0/0a8FbwzqwNyWDvq8v58edx+yOpSqIFoFS6g99WtXkiwk9qBURyOhZ63jhm53kF7jPlIvqwmgRKKX+pH61YBbc2Y1bOtZhyk+/MmzGGo6dzrY7lnIiLQKl1FkCfL15/qY2vHRzWzYeTKXv68tZvfeE3bGUk2gRKKVKNDAuhs/u6k6ovw9Dpq1myk97KCx0r2uP1LlpESilShVbI4yFd3fnmtY1eeGbXdz2XgKpmbl2x1LlSItAKXVOoQG+TB7cnqf7t2TZLyn0fX05m5NS7Y6lyokWgVKqTESEkd3q89H4rgAMfHMV76/6TQeuqwS0CJRS56V93Ui+nNCDbo2r8sTCbdzz4UYycnTgOnemRaCUOm+RwX68M7IjD1/djEWbD9N/8nJ2Hz1tdyx1gbQIlFIXxMtLuKtXY2aP7UxaVh7XT17Bgg1JdsdSF0CLQCl1Ubo1qsaie3rSunY498/bxKC3VjF12a/sOXZGjx+4CZ2PQClVLvILCpn28z4+33SYHcnWTFr1qgbROzaay2Or06lBFfx89LunXWybs1hE+gCvAd7AdGPM88UsMwh4CjDAJmPMkNLWqUWglOs7lJrFjzuPsWTnMVbsOU5OfiHBft70bBJF7+bR9GoWTVSov90xPYotRSAi3sBu4EogCVgHDDbGbC+yTBPgI6C3MeaUiEQbY0od8lCLQCn3kpVbwMpfj7N45zGW7DjGkXRr3KK2dSK4PDaa3rHRtKwVhojYnLRys6sIugJPGWOudtx/HMAY858iy7wA7DbGTC/rerUIlHJfxhi2J6fz485jLN55jI0HUzEGaoQF0Cs2mstjo+neuBqBft52R610SisCHye+b23gYJH7SUDnvyzTFEBEVmDtPnrKGPPNX1ckIuOAcQB169Z1SlillPOJCC1rhdOyVjh3927C8TM5/LQrhSU7j/LFpsPMXXsAfx8vujWqSu/m1ekdG03tiEC7Y1d6ziyCsr5/E+AyIAZYJiKtjTF/unbdGDMVmArWFkFFh1RKOUe1EH8GxsUwMC6G3PxC1v12ksU7jrF451F+/GwrTwCxNULpFRtN+zoRtKwdTq3wAN2NVM6cWQSHgDpF7sc4HisqCVhjjMkD9onIbqxiWOfEXEopF+Tn40X3xtXo3rgaT/Rrzt7jGSzZYR1wnrZsL/mOUU/DA31pUTOMlrXCaFErjJa1wmkYFYyvt56RdKGcWQTrgCYi0gCrAG4B/npG0GfAYGCmiFTD2lW014mZlFJuQERoFBVCo6gQbrukIZm5+ew8cprth9PZdjid7cnpvL96Pzn51uxpfj5eNKseWqQcwoitEUawv907PdyD0/6VjDH5InI38C3W/v93jDHbROQZIMEY87njuatEZDtQADxsjNHZL5RSfxLk50OHupF0qBv5x2P5BYXsO57xRzFsO5zGN9uO8OE669CkCNSvGkyLWmG0qPm/gogODbDrY7gsvaBMKVVpGGM4kp7NtkP/K4ftyekcPJn1xzLVQvz/2HJoWj2EKsH+RAb5EhHoR0SwL6H+PpXyGIRdZw0ppVSFEhFqhgdSMzyQK1pU/+PxtKw8diQ7dis5tiBWFDnuUJSPlxAR5EtEkB8RgdafkUG+RAb7WY8HWvcjgvyIDPYlMsiP8EBfAnzd95RXLQKlVKUXHuhLl4ZV6dKw6h+P5eQXcPBkFqmZuaRm5nGqyJ+nMvNIy8rlVEYeSacy2XrIevz3YxLFCfT1/qMgYmuGMrRzXTrUjXSLrQstAqWUR/L38aZxdMh5vSYrt4BUR0GkOgrDKpDfS8S6/922o3y6/hCxNUIZ1qUeN7SvTYgLH7jWYwRKKVXOMnLyWbjxMLNX72d7cjoh/j7c0L4Ww7rUI7ZGmC2ZbBt0zhm0CJRS7sIYw4aDqcxevZ8vNyeTm19IfL1IhnetR59WNfD3qbjjCloESills1MZucxPTGL2mv3sP5FJ1WA/bo6vw9DOdalTJcjp769FoJRSLqKw0LB8z3Fmr97PDzuOYoBLm0YxrHM9esVG4+3lnIPLWgRKKeWCktOymLv2IB+uPcCx0znUjghkSOe6DIqvU+7zNWgRKKWUC8srKOSH7Ud5f/V+Vv56Al9v4eqWNRjWpR6dG1Qpl1NQ9YIypZRyYb7eXlzTuibXtK7JrylnmLP6APMTD/Ll5mSaRIcwtHNdBsTFEBbg65T31y0CpZRyQVm5BXyx+TBzVu9nU1Iagb7ePHhVU8b2bHhB69MtAqWUcjOBft4Miq/DoPg6bElKY/bq/dRy0iQ9WgRKKeXiWseE89+BbZy2fp3JQSmlPJwWgVJKeTgtAqWU8nBaBEop5eG0CJRSysNpESillIfTIlBKKQ+nRaCUUh7O7YaYEJEUYP8FvrwacLwc4zibO+V1p6zgXnndKSu4V153ygoXl7eeMSaquCfcrgguhogklDTWhityp7zulBXcK687ZQX3yutOWcF5eXXXkFJKeTgtAqWU8nCeVgRT7Q5wntwprztlBffK605Zwb3yulNWcFJejzpGoJRS6myetkWglFLqL7QIlFLKw3lMEYhIHxHZJSJ7ROQxu/OURETqiMiPIrJdRLaJyL12ZyoLEfEWkQ0i8qXdWUojIhEiMl9EdorIDhHpanem0ojI/Y7/B1tFZK6IBNidqSgReUdEjonI1iKPVRGR70XkF8efkXZm/F0JWV90/F/YLCILRCTCzoxFFZe3yHMPiogRkWrl8V4eUQQi4g28AVwDtAAGi0gLe1OVKB940BjTAugC3OXCWYu6F9hhd4gyeA34xhgTC7TFhTOLSG3gHiDeGNMK8AZusTfVWWYBff7y2GPAYmNME2Cx474rmMXZWb8HWhlj2gC7gccrOlQpZnF2XkSkDnAVcKC83sgjigDoBOwxxuw1xuQCHwLX25ypWMaYZGPMesffT2P9oqptb6rSiUgM0BeYbneW0ohIOHAJMAPAGJNrjEm1N9U5+QCBIuIDBAGHbc7zJ8aYZcDJvzx8PfCu4+/vAjdUaKgSFJfVGPOdMSbfcXc1EFPhwUpQwr8twCvAI0C5nenjKUVQGzhY5H4SLv7LFUBE6gPtgTX2JjmnV7H+YxbaHeQcGgApwEzHbqzpIhJsd6iSGGMOAS9hffNLBtKMMd/Zm6pMqhtjkh1/PwJUtzPMebgV+NruEKURkeuBQ8aYTeW5Xk8pArcjIiHAJ8B9xph0u/OURET6AceMMYl2ZykDH6AD8KYxpj2QgevstjiLY9/69VgFVgsIFpFh9qY6P8Y6P93lz1EXkX9g7ZadY3eWkohIEPB34P/Ke92eUgSHgDpF7sc4HnNJIuKLVQJzjDGf2p3nHLoD/UXkN6xdbr1FZLa9kUqUBCQZY37fwpqPVQyu6gpgnzEmxRiTB3wKdLM5U1kcFZGaAI4/j9mcp1QiMgroBww1rn1hVSOsLwWbHD9vMcB6EalxsSv2lCJYBzQRkQYi4od1wO1zmzMVS0QEax/2DmPMRLvznIsx5nFjTIwxpj7Wv+sSY4xLfms1xhwBDopIM8dDlwPbbYx0LgeALiIS5Ph/cTkufHC7iM+BkY6/jwQW2pilVCLSB2u3Zn9jTKbdeUpjjNlijIk2xtR3/LwlAR0c/68vikcUgeNg0N3At1g/SB8ZY7bZm6pE3YHhWN+sNzpu19odqhKZAMwRkc1AO+A5m/OUyLHlMh9YD2zB+nl1qSERRGQusApoJiJJIjIGeB64UkR+wdqqed7OjL8rIetkIBT43vGz9patIYsoIa9z3su1t4SUUko5m0dsESillCqZFoFSSnk4LQKllPJwWgRKKeXhtAiUUsrDaREo5WQicpmrj8qqPJsWgVJKeTgtAqUcRGSYiKx1XFj0tmOOhTMi8opjToDFIhLlWLadiKwuMo59pOPxxiLyg4hsEpH1ItLIsfqQIvMgzHFcKYyIPO+Ye2KziLxk00dXHk6LQClARJoDfwO6G2PaAQXAUCAYSDDGtASWAk86XvIe8KhjHPstRR6fA7xhjGmLNS7Q76Nwtgfuw5oPoyHQXUSqAjcCLR3reda5n1Kp4mkRKGW5HIgD1onIRsf9hlhDa89zLDMb6OGY1yDCGLPU8fi7wCUiEgrUNsYsADDGZBcZv2atMSbJGFMIbATqA2lANjBDRAYALj3Wjaq8tAiUsgjwrjGmnePWzBjzVDHLXeiYLDlF/l4A+DjGwOqENZ5QP+CbC1y3UhdFi0Apy2JgoIhEwx/z7tbD+hkZ6FhmCLDcGJMGnBKRno7HhwNLHTPKJYnIDY51+DvGkC+WY86JcGPMV8D9WFNnKlXhfOwOoJQrMMZsF5F/At+JiBeQB9yFNXlNJ8dzx7COI4A1vPJbjl/0e4HRjseHA2+LyDOOddxcytuGAgsdE9IL8EA5fyylykRHH1WqFCJyxhgTYncOpZxJdw0ppZSH0y0CpZTycLpFoJRSHk6LQCmlPJwWgVJKeTgtAqWU8nBaBEop5eH+H0p00eQioD/KAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ho7Rtr3XwEhR"
      },
      "source": [
        "model.save('indian_notes.h5')"
      ],
      "execution_count": 20,
      "outputs": []
    }
  ]
}