CNN_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN_MNIST.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyNICLIsH41s5vAk/bXgzZo9",
      "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/Hegazian/Deep_Learning/blob/master/CNN_MNIST.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lklviV2YPuyL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import keras\n",
        "from keras.datasets import mnist\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense\n",
        "from keras.layers import Flatten\n",
        "from keras.layers import Conv2D\n",
        "from keras.layers import MaxPooling2D\n",
        "from keras.layers import Dropout\n",
        "from keras.optimizers import Adam\n",
        "from keras.utils.np_utils import to_categorical\n",
        "import random"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yAJvnDahQygg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "np.random.seed(0)\n"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uw9-6A6nQ2W2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "c766a716-3467-4436-afa8-95a00e9916b5"
      },
      "source": [
        "(X_train, y_train), (X_test, y_test)= mnist.load_data()\n",
        " \n",
        "print(X_train.shape)\n",
        "print(X_test.shape)\n",
        "assert(X_train.shape[0] == y_train.shape[0]), \"The number of images is not equal to the number of labels.\"\n",
        "assert(X_train.shape[1:] == (28,28)), \"The dimensions of the images are not 28 x 28.\"\n",
        "assert(X_test.shape[0] == y_test.shape[0]), \"The number of images is not equal to the number of labels.\"\n",
        "assert(X_test.shape[1:] == (28,28)), \"The dimensions of the images are not 28 x 28.\""
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 28, 28)\n",
            "(10000, 28, 28)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PO8AxX1lQ7LF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 718
        },
        "outputId": "410bddce-c90e-4726-9434-b5deb16aabb0"
      },
      "source": [
        "num_of_samples=[]\n",
        " \n",
        "cols = 5\n",
        "num_classes = 10\n",
        " \n",
        "fig, axs = plt.subplots(nrows=num_classes, ncols=cols, figsize=(5,10))\n",
        "fig.tight_layout()\n",
        " \n",
        "for i in range(cols):\n",
        "    for j in range(num_classes):\n",
        "      x_selected = X_train[y_train == j]\n",
        "      axs[j][i].imshow(x_selected[random.randint(0,(len(x_selected) - 1)), :, :], cmap=plt.get_cmap('gray'))\n",
        "      axs[j][i].axis(\"off\")\n",
        "      if i == 2:\n",
        "        axs[j][i].set_title(str(j))\n",
        "        num_of_samples.append(len(x_selected))"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x720 with 50 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N2C1Nz_-RDSP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train = X_train.reshape(60000,28,28,1)\n",
        "x_test = X_test.reshape(10000,28,28,1)"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dZncfK_RVRSh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_train = to_categorical(y_train,num_classes)\n",
        "y_test = to_categorical(y_test,num_classes)"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2xCNEc3jVqqr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train = x_train/255\n",
        "x_test = x_test/255"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TBWB0BnJV5De",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def LeNet_model():\n",
        "  model = Sequential()\n",
        "  model.add(Conv2D(filters=30,kernel_size=(5,5),input_shape=(28,28,1),activation='relu'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "  model.add(Conv2D(filters=15,kernel_size=(3,3),activation='relu'))\n",
        "  model.add(MaxPooling2D(pool_size=(2,2)))\n",
        "  model.add(Flatten())\n",
        "  model.add(Dense(500,activation='relu'))\n",
        "  model.add(Dropout(0.5))\n",
        "  model.add(Dense(num_classes,activation='softmax'))\n",
        "  model.compile(Adam(learning_rate=0.01),loss='categorical_crossentropy',metrics=['accuracy'])\n",
        "  return model"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ix4aq4p8q0jO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 437
        },
        "outputId": "e52d9dd6-49b3-45de-c753-913a2e5b50f9"
      },
      "source": [
        "model = LeNet_model()\n",
        "print(model.summary())"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 24, 24, 30)        780       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 12, 12, 30)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 10, 10, 15)        4065      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 5, 5, 15)          0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 375)               0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 500)               188000    \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 500)               0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                5010      \n",
            "=================================================================\n",
            "Total params: 197,855\n",
            "Trainable params: 197,855\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g-uuQ1QVq4-I",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        },
        "outputId": "3ca01a44-ff2e-480b-a027-d80d7b8396d4"
      },
      "source": [
        "h = model.fit(x_train,y_train,batch_size=400,epochs=10,verbose=1,validation_split=0.1,shuffle=1)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "135/135 [==============================] - 1s 7ms/step - loss: 0.2253 - accuracy: 0.9277 - val_loss: 0.0636 - val_accuracy: 0.9823\n",
            "Epoch 2/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0731 - accuracy: 0.9780 - val_loss: 0.0404 - val_accuracy: 0.9880\n",
            "Epoch 3/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0602 - accuracy: 0.9820 - val_loss: 0.0363 - val_accuracy: 0.9905\n",
            "Epoch 4/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0476 - accuracy: 0.9855 - val_loss: 0.0412 - val_accuracy: 0.9892\n",
            "Epoch 5/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0469 - accuracy: 0.9851 - val_loss: 0.0409 - val_accuracy: 0.9888\n",
            "Epoch 6/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0448 - accuracy: 0.9865 - val_loss: 0.0428 - val_accuracy: 0.9870\n",
            "Epoch 7/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0429 - accuracy: 0.9870 - val_loss: 0.0417 - val_accuracy: 0.9897\n",
            "Epoch 8/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0415 - accuracy: 0.9877 - val_loss: 0.0459 - val_accuracy: 0.9882\n",
            "Epoch 9/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0406 - accuracy: 0.9873 - val_loss: 0.0453 - val_accuracy: 0.9878\n",
            "Epoch 10/10\n",
            "135/135 [==============================] - 1s 6ms/step - loss: 0.0386 - accuracy: 0.9884 - val_loss: 0.0353 - val_accuracy: 0.9905\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sf8jn5kdr_Lt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "outputId": "8f9fe684-1cfa-41cd-ae7e-fbb5480d118a"
      },
      "source": [
        "plt.plot(h.history['loss'])\n",
        "plt.plot(h.history['val_loss'])\n",
        "plt.legend([\"loss\", \"val_loss\"])\n",
        "plt.xlabel(\"epoches\")"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epoches')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        },
        {
          "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": "Igcwp0jMsSI9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "outputId": "110564f6-0076-41e3-efef-8b9dcdf70cf7"
      },
      "source": [
        "plt.plot(h.history['accuracy'])\n",
        "plt.plot(h.history['val_accuracy'])\n",
        "plt.legend(['accuracy ', 'val_accuracy'])\n",
        "plt.xlabel(\"epoches\")"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epoches')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAEGCAYAAABrQF4qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXwc1Z33+89P3do3y5a8SfICGLwvsWIgHrAxYQZuCLsDhHADM4GbhQzLzU2ALHBJCHmeMDcDE5InTsYhfpIMk5jAEIaY1Y4JS4KNJa/YGGPcki1Z2Got1tLq7t/9o0pSS0h222651dW/9+vVr+6uOlV9uix/6/SpqlOiqhhjjPGujGRXwBhjzPCyoDfGGI+zoDfGGI+zoDfGGI+zoDfGGI/zJ7sCA5WWluqUKVOSXQ1jjEkpGzdu/FBVywabN+KCfsqUKWzYsCHZ1TDGmJQiIh8MNc+6bowxxuMs6I0xxuMs6I0xxuMs6I0xxuMs6I0xxuMs6I0xxuMs6I0xxuPiOo9eRC4GHgF8wC9U9QcD5k8GVgJlwGHgc6pa6877H8Cn3KLfVdX/TFDdzVCaa2Hfm9C0F4rKYdQk51E0ETJ8ya6dMSaWKhzaDXv/4ryvujnhH3HMoBcRH/AYcBFQC7wlIs+o6vaYYg8Dq1T1VyKyDHgIuFFEPgV8DJgPZAPrRORPqtqS6C+StqIROLgD9r0Bgb86Ad8cGLxshr9/8Pc8iivdHUE5+EbcNXTGeIsqfLjLCfa9f4EPXoO2BmdexaLkBD2wCNitqnsAROQJ4HIgNuhnAne5r9cCT8dMX6+qYSAsIpuBi4HfJaDu6SnUDnUbnUAPvAmBv0GXu98sGA+TzoFzv+I8l54JrfUQ/ACC+9xHwHl+by20HgBibjwjPndHUDn4jqC4AnyZSfnaxqQsVWh8p3+wH2l05hVOhKlLYMpimHIejD5tWKoQT9CXA7FNxFrg7AFlaoCrcLp3rgQKRWSMO/0+EfkXIA+4gP47CABE5FbgVoBJkyYd51fwuLaDTqj3BPuBGoiGnXllM2D21U6oTzoHRk0Gkf7LjzndeQwmHIKW2pidQMzO4P1XoXU/aLSvvGQ4f5hH2xH4s4dnO5iTowqH3oMP3LDZXw35pVAy1QmX0THPuSXJrm1qi0ahcQfsfQ32vgofvA7tHzrziirg9AvdYP87Z/sP/D87DBL1O/1rwI9F5CZgPVAHRFT1BRH5OPA60Ai8AUQGLqyqK4AVAFVVVel7b0NV+PDdmG6YN+DwHmeeLxvKF8In/tkJ9YqPQ97ok/s8f5b7n3uIVkQ4BC11TlfQwB3BB2/Alt/33xEgUDjB3QHE7Azyx0JmDvhznB2Bf8BrX5b7nHlK/ujTQs/fUk+w730N2uqdefljnb+fzmbYsw5qftt/2dySvr+L3h2BuxPIL7N/o4GiUTi4rX+wdxx25hVXwrSLnFCf8neDN8ZOgXiCvg6ojHlf4U7rpar7cVr0iEgBcLWqBt15DwIPuvN+C+w6+Wp7RLjLaVkF3uxrtff8geSOhknnwsKbnOcJ8059a9mf5bb0pg4+P9INLfud8B+4Mwj8Dbb+AfQj+/WjkAE7g2M95zh1jHeZgvFQOs2bvzp6+31f7Qv2IwedeQXj3aBxuwfGnNE/bELtTvfe4T0xj/fdf8Mn++/Mswr6Wv/9dgKnOTv5jDQ4kS8ahYatfV0x+16HjiZn3qjJcNYlzvaevBhKJie3rq54gv4tYJqITMUJ+OuAz8YWEJFS4LCqRoF7cM7A6TmQO0pVD4nIXGAu8EIC659a2g87/3l6gr3ubYh0OfNGn+78gUw6ByrPcQJppLecfJnOH/JQf8yRsHMc4EgjREIQ7nR2bh95HmzaEM/th4aeH+0+dp3F5wTd2OkwdiaMneF0gY0+LbUORB+r3/e0pf37fY/2t5SV52yHsTM+Oi8ccnbiA3cCDdvhnef6b3N/DpRMifk1EPO6uDK1tm+saATqt/Rt5w9ec34NgbOzm/4pZztPXuz8kj3Rj4kqR0JhCnMSfxzsmFteVcMichvwPM7plStVdZuIPABsUNVngKXAQyKiOF03X3EXzwReFeePrAXntMtwwr/FSKTqnN7Y07e+703nPyY4Z79MmA+LbnGD/WwoGJvU6g4Ln9/twjnxP/7jEo0cZSfR6Zx2enCH86jfAtufofdgtC8LSs9ydwAznJ1A2XSnhTYSWqm9/b4xwd5+yJk3nP2+/qyhj/NEI842bXq//07g8PvOwf5wR1/ZDL/TjTewS2jUJGdH068rLzu52zwShvrNMcH+BnS5wT76NJh5OUx2fyEVV8S1ys7uCAdbujjQ3EF9SycNLZ0caHae65udx8HWLhZMGsXvv/iJhH8lUR1ZXeJVVVWa0uPRB96CN3/i9NP19IlmF0HlIveg6bkw8WPOH7dJrlA7fLgTDr4DB7c7O4DGd/qfnpqZ5wR+T4u35xdA0cTh/cXV2+/bE+yx/b6TYrpiktfve1SqzhlfH9kJuM89wTmUjMzj7MbLjnkMcSxoqGdfltP1EhvsoVanHmPOcLth3O1dNHHA11RaOsLUt3Q6j+YO6pu7qG/pcAK8pYv65g6a2j/6azMvy8f44hzGF+X0Pp85rpArFpSf0CYXkY2qWjXoPAv6BGnYDq98F3Y+B3lj4PRlfd0wY2fYhUqppLMZGnf2tf4Pbnd2AD3nOgNkF7vBP6ALqGDQG/wcWzTi9vu+1hc4nUFn3qjJTtfAlMUjqt93KNGo0toVpqWjm+aO7t7n5o5uWjq7aW4P0d32IVkt+8jv2E9eRje5GWFypefRTbZ0k02IbOkmS7vJIkSmhvBHQ/g1hC8awhftwhcJkRHpQiLOLzfp+QXHCeZa6Zm9/euRSYtppMQNcDfEW7rc1ngHDS1d1Dd30tH90eNQpQVZjCvqH+LjinOYEPO6MNuPJHAHbUE/nA6/D+segs2/c1rui/8ZzvkSZOUnu2Ym0doP9wV/T+u/YVtfIAPklfZv/fd0AeWO6r+uaMTtHnit74BebL9vT//6Sfb7nqjuSLQvqDvDvUHdE9wfCe+e9+3dtHaFOVqs+DKE4txMinMzycvyEQpH6eiO0NkdobM7SnsoTPQEYsmXIeT4M8jNzKAgUynOjFLgj1Lkj1DgD1PgC1Pgi5DvC5OX4TxyM8LkSYhQRi5bMqazuz2/N9gb27qIDKhIpk96A3xccQ4T3CAfV+SE+LiiHMYWZZPtP/UNOwv64dBaD+t/CBsfd35mnv1/weLbT/6UR5NaVJ2W/sDW/8EdEGrrK1c40Qn+0mlO42DfG30Xuo0+vX+wF5/YT/doVOnojtAeitAeCvd7PtIVoaPbndYV4UgoTEtHuF9Yx7bAj4SOfrZUtj+DIjesYx9FOX7nuef9IGXysnxHbcmqKt0R57t0dUfocB+d3VE6Qs4OoWfH0NEdoSMUoSvszIud3unO6+yO3ZH0LdPZHSUU6TujqDDb77S+BwR3b6u8OIfReVlkZIywbjLX0YI+RQ+DJ1FHE/zlX+GvP3POOPjY5+H8/weKJiS7ZiYZRKBwvPM4/YK+6apOX3+//v8dsPF1tLiC0PQr6Jh4Ls3jFtGaWeaEcShMx74IR7oCdHS74RwKcyTUP7w7esrGPndFBu1COJr8LF9vGBflZlI5Oo/ZvYGdSXGun+K82BDvC++czOFrsYoIWX4hy58BucN7JXYkqnS62y0/27tx6N1vlmihI/DmT+G1R52W2JzlcME9w3bJskldzR3dBA63U9uURW3TGdQ2lRNoWUygvZ26UAdH6iIxV6LsdB9Dy8vykZfld5+dR0G2n7KCbPKz/eRm+cjL9JGX7ZTJz/KRm+V3n31OmUznOc+dlpfpw+8bAWcTJZkvQzwd8D28/w1PVjjkdM+s/6FzAcqZl8Cyb8H42cmumUmS9lCY2qYOAofb3UDvINDUTuBwB7VN7bR09j+DuCDbT0VJLpPH5LP4jFJK8rIGCW8/edk+N6jd8M7ykeP3jdiuApM6LOiHEo04B1jXfd+50nPy38G1v4ZJA4f5MV7TFY5Q19RBoMkJ7sBhJ8hrmzqoPdzOoSOhfuVzMjOoKMmjsiSXqiklVJTkUlmS50wbnUtxbmZCz64w5nhZ0A+kCu/8N7zyPadPdcI8uPRHzgUp9p81bqrKzoZW1mytZ0/jEbL8GWT7M8j2+3pf900bbPpHyw2cnu3POKEADUeiHGjudMLbbYUHelroTe0cbO3qd9ZIpk8oH5VL5eg8/n7WODfA83oDvbQgy4LcjGgW9LH2/BlefgDqNjgXSix/HGZcPjKujEwBqsrm2mb+tLWe57fV8/6HRxCBypI8wpEoXeEoobD7HIkee4VxyPIdfWcQ+9zSESbQ1M6B5s5+p81lCEwozqVydC7nTStzW+NOsFeOzmVsYQ4+6z4xKcyCHpzx3V9+wBnJr6gcLvs3mPfZ1B2b4xSKRJUNew+zZls9z2+tZ39zJ/4M4dzTx/CF86Zy0cxxjC3M+chy0agSijiB39Xd8xwZ8D5KKBLp975rQLm+nUekbyfS894t09YVpqs7SkGOn6rJJf1a45Wj8xhfnEOmHZg0HpbeSda407madccfnatZ/+H7UPVPzpC6ZkjdkShvvHeIP22t58Xt9XzYFiLLn8H508q46+/P4pMzxjIqL+uo68jIEHIyfM5pera5jRlW6Rn0wX2w7gdQ8x+QmQ9L74Fzvgw5Rcmu2YjV2R1h/a5G1myr56XtDbR0hsnL8nHB9LFcMns8S88aS0EanKZmTCpKr/+ZbY3w6sOwYSUgTrj/3V2QPybZNRuR2rrCrH3nIGu21rN250HaQxGKczO5aOZ4Lp49nvOmlQ7rhTPGmMRIj6DvbIbX/w3e+Ikz4NGCG2DJN+IeYjSdBNtDvLi9gee31bP+3Q8JhaOUFmRxxYJyLpk9nnNOG2P92cakGG8HfXcH/G0F/OVHztAFs66EC77pjDdieh1s7eSFbU64v/HeIcJRpXxULp87ezIXzx7PwskldtaJMSnMm0Ef6YZN/xv+/D+dOxyd8UlY9m2YOD/ZNRsxapvaeX5bA2u2HmDDB02owtTSfG45/zQumT2eOeXFdm64MR7hraCPRmHbH2Dtg84NDirPhqv/3RkZ0LCnsY0/ba1nzdZ6ttQ5Q+LOmFDEHReeycWzx3PmuAILd2M8yDtB3/QBPHEDNGyBcbPh+v+EM/8hra9mVVV2HGhlzbZ61mw9wK4GZ9jc+ZWjuOeS6fzDrPFMKbVx843xOu8EfeEEyC+Fq34Bs69O66tZ9354hD+8Xct/1ezng0PtZAgsmjqa+z89k3+YPZ4JxbnJrqIx5hTyTtD7s+D/fDrZtUia1s5unttygNUba3lrbxMZAovPKOVLS07nkzPHUVqQnewqGmOSxDtBn4YiUeWN9w6xemOANdvq6eyOcnpZPt+4eDpXLihnfLFdcmqMsaBPSXsa23jy7VqeeruO/c2dFOX4uWZhBdcsrGRehZ0tY4zpL66gF5GLgUcAH/ALVf3BgPmTgZVAGXAY+Jyq1rrz/ifwKSADeBG4XUfajWpTQEtnN/+92ema2fiB0zWz5MwyvvmpmVw4Y6xdoWqMGdIxg15EfMBjwEVALfCWiDyjqttjij0MrFLVX4nIMuAh4EYR+QSwGJjrlvsLsARYl7iv4F2RqPLa7g9ZvbGW57fV0xWOMm1sAfdc4nTNjC2yrhljzLHF06JfBOxW1T0AIvIEcDkQG/Qzgbvc12uBnqOiijM2YRYgQCbQcPLV9rbdB/u6ZupbOinOzeTaj1dyzcIKu5DJGHPc4gn6ciAQ874WGHg/vRrgKpzunSuBQhEZo6pviMha4ABO0P9YVXcM/AARuRW4FWDSpEnH/SW8oLm9mz9u3s+Tb9eyaV8QX4aw9Mwy7vv0TJbNGEu237pmjDEnJlEHY78G/FhEbgLW49zjPiIiZwAzgJ7Rw14UkfNU9dXYhVV1BbACoKqqKm367yNR5dV3G1m9sZYXtjcQCkc5a1wh3/rUDC6bP3HQG3YYY8zxiifo64DKmPcV7rReqrofp0WPiBQAV6tqUERuAd5U1TZ33p+Ac4F+QZ9u3m1oZbXbNXOwtYuSvEw+u2gS1yysYNbEIuuaMcYkVDxB/xYwTUSm4gT8dcBnYwuISClwWFWjwD04Z+AA7ANuEZGHcLpulgD/mqC6p5Rge4g/1uxn9dt11ASC+DOEpWeN5ZqFFSybPpYsf/peyWuMGV7HDHpVDYvIbcDzOKdXrlTVbSLyALBBVZ8BlgIPiYjidN18xV18NbAM2IJzYHaNqv4x8V9jZApHoqx/t5EnN9bx4vYGQpEoMyYU8e1LZ3L5/Il2taox5pSQkXZKe1VVlW7YsCHZ1TgpuxpaWb2xlqc21dHY2sXo/CyumF/O1QvLmTWxONnVM8Z4kIhsVNWqwebZlbEJtrWumcsfew0Blk13umaWnmVdM8aY5LGgT7D17zY6FzrdvYzyUTZKpDEm+ayZmWA1gSBTS/Mt5I0xI4YFfYJVB4LMq7B+eGPMyGFBn0AHmjtoaOlifuWoZFfFGGN6WdAnUE0gCMD8SSVJrokxxvSxoE+gTYEgWb4MZkwoTHZVjDGmlwV9AlXvCzJjYpENQGaMGVEs6BMkElW21DWzwPrnjTEjjAV9grx7sJX2UIR5lXbGjTFmZLGgT5Dqfe6B2Eo7EGuMGVks6BOkOhCkODeTKWPykl0VY4zpx4I+QaoDQeZVjrKx5I0xI44FfQIc6Qqzq6HVLpQyxoxIFvQJsKWumahiZ9wYY0YkC/oE6Lkidp4FvTFmBLKgT4DqQJBJo/MYnZ+V7KoYY8xHWNAnQHUgaP3zxpgRy4L+JDW0dHKgudOC3hgzYlnQn6Rq6583xoxwFvQnqToQJNMnzJpYlOyqGGPMoCzoT1L1viAzJhSRk2kjVhpjRiYL+pPQM2LlvArrtjHGjFxxBb2IXCwiO0Vkt4jcPcj8ySLysohsFpF1IlLhTr9ARKpjHp0ickWiv0SyvNfYRltX2A7EGmNGtGMGvYj4gMeAS4CZwPUiMnNAsYeBVao6F3gAeAhAVdeq6nxVnQ8sA9qBFxJY/6TqHbFykgW9MWbkiqdFvwjYrap7VDUEPAFcPqDMTOAV9/XaQeYDXAP8SVXbT7SyI011bZDCHD9Tx+QnuyrGGDOkeIK+HAjEvK91p8WqAa5yX18JFIrImAFlrgP+Y7APEJFbRWSDiGxobGyMo0ojQ/U+50KpjAwbsdIYM3Il6mDs14AlIrIJWALUAZGemSIyAZgDPD/Ywqq6QlWrVLWqrKwsQVUaXh2hCDttxEpjTArwx1GmDqiMeV/hTuulqvtxW/QiUgBcrarBmCKfAZ5S1e6Tq+7IsaWumUhULeiNMSNePC36t4BpIjJVRLJwumCeiS0gIqUi0rOue4CVA9ZxPUN026QqG7HSGJMqjhn0qhoGbsPpdtkB/E5Vt4nIAyJymVtsKbBTRHYB44AHe5YXkSk4vwj+nNCaJ1l1IEhFSS6lBdnJrooxxhxVPF03qOpzwHMDpn0n5vVqYPUQy+7lowdvU151IMgCO63SGJMC7MrYE3CwtZO6YIf1zxtjUoIF/QmoCTQDWNAbY1KCBf0JqA404csQZpcXJ7sqxhhzTBb0J6Am0Mz08YU2YqUxJiVY0B+naFSpsVsHGmNSiAX9cdrzYRutNmKlMSaFWNAfp009I1Za0BtjUoQF/XGqqQ1SmO3n9LKCZFfFGGPiYkF/nKoDQeZWFtuIlcaYlGFBfxw6uyO8c8BGrDTGpBYL+uOwbX8z4ajaPWKNMSnFgv44bLJbBxpjUpAF/XGoDgQpH5XL2MKcZFfFGGPiZkF/HGpqg8yrtGEPjDGpxYI+TofauggcthErjTGpx4I+TtWBngulSpJcE2OMOT4W9HGqDgTdESuLkl0VY4w5Lhb0caoOBDlzXCF5WXHdlMsYY0YMC/o42IiVxphUZkEfh/cPHaGlM8wCC3pjTAqyoI9DjXsgdp4FvTEmBVnQx6E6ECQ/y8cZY23ESmNM6okr6EXkYhHZKSK7ReTuQeZPFpGXRWSziKwTkYqYeZNE5AUR2SEi20VkSuKqf2pUB4LMrRiFz0asNMakoGMGvYj4gMeAS4CZwPUiMnNAsYeBVao6F3gAeChm3irgh6o6A1gEHExExU+Vzu4IOw60WLeNMSZlxdOiXwTsVtU9qhoCngAuH1BmJvCK+3ptz3x3h+BX1RcBVLVNVdsTUvNTZPuBFrojamfcGGNSVjxBXw4EYt7XutNi1QBXua+vBApFZAxwJhAUkT+IyCYR+aH7C6EfEblVRDaIyIbGxsbj/xbDqNodsXKBjVhpjElRiToY+zVgiYhsApYAdUAE8APnufM/DpwG3DRwYVVdoapVqlpVVlaWoColRnUgyPiiHMYV2YiVxpjUFE/Q1wGVMe8r3Gm9VHW/ql6lqguAb7rTgjit/2q32ycMPA18LCE1P0Vqau1CKWNMaosn6N8CponIVBHJAq4DnoktICKlItKzrnuAlTHLjhKRnmb6MmD7yVf71Dh8JMQHh9rtRiPGmJR2zKB3W+K3Ac8DO4Dfqeo2EXlARC5ziy0FdorILmAc8KC7bASn2+ZlEdkCCPDzhH+LYdJ7oZTdOtAYk8LiGqFLVZ8Dnhsw7Tsxr1cDq4dY9kVg7knUMWmqA0EyBOZW2M1GjDGpy66MPYqeESvzs23ESmNM6rKgH4Kq2oFYY4wnWNAP4YND7QTbu+2KWGNMyrOgH0LfrQMt6I0xqc2CfgjVgSB5WT7OHFeY7KoYY8xJsaAfwqZAkNnlxTZipTEm5VnQD6IrHGHH/ha7o5QxxhMs6Aex40AroUjU+ueNMZ5gQT+I6n1NgN060BjjDRb0g6ipbWZsYTYTim3ESmNM6rOgH0R1wLlQSsQOxBpjUp8F/QDB9hDvf3jEum2MMZ5hQT9ATW0zgJ1xY4zxDAv6Aar3BRGBOTZipTHGIyzoB6gONDFtbAGFOZnJrooxxiSEBX0MVaU6ELQbjRhjPMWCPkbgcAdN7d1260BjjKdY0MfYFHAulLIrYo0xXmJBH6M6ECQnM4OzbMRKY4yHWNDHqAkEmVNejN9nm8UY4x2WaK5QOMrW/S3WbWOM8RwLetc79S2EwlG7ItYY4zkW9K4au3WgMcaj4gp6EblYRHaKyG4RuXuQ+ZNF5GUR2Swi60SkImZeRESq3ccziax8Im0KBCktyKZ8VG6yq2KMMQnlP1YBEfEBjwEXAbXAWyLyjKpujyn2MLBKVX8lIsuAh4Ab3Xkdqjo/wfVOOBux0hjjVfG06BcBu1V1j6qGgCeAyweUmQm84r5eO8j8Ea25vZs9jUeYX2nj2xhjvCeeoC8HAjHva91psWqAq9zXVwKFIjLGfZ8jIhtE5E0RuWKwDxCRW90yGxobG4+j+omxua6nf77klH+2McYMt0QdjP0asERENgFLgDog4s6brKpVwGeBfxWR0wcurKorVLVKVavKysoSVKX4Ve9zgn6uteiNMR50zD56nNCujHlf4U7rpar7cVv0IlIAXK2qQXdenfu8R0TWAQuA90665glUHQhyelk+RTZipTHGg+Jp0b8FTBORqSKSBVwH9Dt7RkRKRaRnXfcAK93pJSKS3VMGWAzEHsRNOlWlpjZo3TbGGM86ZtCrahi4DXge2AH8TlW3icgDInKZW2wpsFNEdgHjgAfd6TOADSJSg3OQ9gcDztZJutqmDj5sC9mIlcYYz4qn6wZVfQ54bsC078S8Xg2sHmS514E5J1nHYVXdc6GUjUFvjPGotL8ytiYQJNufwfQJNmKlMcab0j7oqwNBZpcXk2kjVhpjPCqt0607EmVLXbPdOtAY42lpHfQ761vpCkftQKwxxtPSOuh7DsQusBErjTEelvZBPyY/i4oSG7HSGONdaR/082zESmOMx6Vt0Ld0dvNeY5vdaMQY43lpG/RbaptRtTtKGWO8L22DvudArJ1aaYzxurQO+tNK8ynOsxErjTHelpZBr6q9tw40xhivS8ug39/cSWNrF/Ms6I0xaSAtg77njlLWojfGpIO0DPqa2iBZvgxmTChKdlWMMWbYpWXQV+8LMnNiEVn+tPz6xpg0k3ZJF3ZHrLRuG2NMuki7oN/V0EZHd4QFNmKlMSZNpF3Q99460Fr0xpg0kYZB30RJXiaTRucluyrGGHNKpF3Q1wSabcRKY0xaSaugb+sKs+tgq3XbGGPSSlxBLyIXi8hOEdktIncPMn+yiLwsIptFZJ2IVAyYXyQitSLy40RV/ERsrg2iil0Ra4xJK8cMehHxAY8BlwAzgetFZOaAYg8Dq1R1LvAA8NCA+d8F1p98dU9O74FYG7HSGJNG4mnRLwJ2q+oeVQ0BTwCXDygzE3jFfb02dr6ILATGAS+cfHVPTk0gyJQxeZTkZyW7KsYYc8rEE/TlQCDmfa07LVYNcJX7+kqgUETGiEgG8C/A1472ASJyq4hsEJENjY2N8dX8BPTcOtAYY9JJog7Gfg1YIiKbgCVAHRABvgw8p6q1R1tYVVeoapWqVpWVlSWoSv0daO6goaXLDsQaY9KOP44ydUBlzPsKd1ovVd2P26IXkQLgalUNisi5wHki8mWgAMgSkTZV/cgB3eFWYxdKGWPSVDxB/xYwTUSm4gT8dcBnYwuISClwWFWjwD3ASgBVvSGmzE1AVTJCHmBTIEimT5g50UasNMakl2N23ahqGLgNeB7YAfxOVbeJyAMicplbbCmwU0R24Rx4fXCY6nvCqvcFmTmhiGy/L9lVMcaYUyqeFj2q+hzw3IBp34l5vRpYfYx1PA48ftw1TIBIVNlS18zyhRXHLmyMMR6TFlfGvnuwlfZQhPk2YqUxJg2lRdD33Dpwnl0oZYxJQ+kR9IEgxbmZTC3NT3ZVjDHmlEuboLcRK40x6Squg7Gp7EhXmF0Nrfz9zHHJrooxKam7u5va2lo6OzuTXRUD5OTkUFFRQWZmZtzLeD7ot9Q1E1XsQKwxJ6i2tpbCwkKmTJliv4qTTFU5dOgQtbW1TJ06Ne7lPN9103NFrB2INebEdHZ2MmbMGAv5EUBEGDNmzHH/uvJ80KlERKQAAA9hSURBVFcHglSOzmVMQXayq2JMyrKQHzlO5N8iLYJ+fmVJsqthjDFJ4+mgb2jp5EBzpw1kZoxJa54O+mobsdIYcxzC4XCyqzAsPH3WTXUgiD9DmGUjVhqTEP/vH7exfX9LQtc5c2IR93161lHLXHHFFQQCATo7O7n99tu59dZbAVizZg333nsvkUiE0tJSXn75Zdra2vjqV7/Khg0bEBHuu+8+rr76agoKCmhrawNg9erVPPvsszz++OPcdNNN5OTksGnTJhYvXsx1113H7bffTmdnJ7m5ufzyl7/krLPOIhKJ8I1vfIM1a9aQkZHBLbfcwqxZs3j00Ud5+umnAXjxxRf5yU9+wlNPPZXQbXSyvB30+4LMmFBETqaNWGlMKlu5ciWjR4+mo6ODj3/841x99dVEo1FuueUW1q9fz9SpUzl8+DAA3/3udykuLmbLli0ANDU1HXP9tbW1vP766/h8PlpaWnj11Vfx+/289NJL3HvvvTz55JOsWLGCvXv3Ul1djd/v5/Dhw5SUlPDlL3+ZxsZGysrK+OUvf8k//uM/Duu2OBGeDfqeESuvXDDwrofGmBN1rJb3cHn00Ud7W8mBQIB3332XxsZGzj///N7zyUePHg3ASy+9xBNPPNG7bEnJsU/GWL58OT6f0yBsbm7m85//PO+++y4iQnd3d+96v/jFL+L3+/t93o033sivf/1rbr75Zt544w1WrVqVoG+dOJ4N+vca22jrCts9Yo1JcevWreOll17ijTfeIC8vj6VLl57QVbqxpyUOXD4/v28crG9/+9tccMEFPPXUU+zdu5elS5cedb0333wzn/70p8nJyWH58uW9O4KRxLMHY3tGrLQDscaktubmZkpKSsjLy+Odd97hzTffBOCcc85h/fr1vP/++wC9XTcXXXQRjz32WO/yPV0348aNY8eOHUSj0aP2oTc3N1Ne7vQEPP74473TL7roIn72s5/1HrDt+byJEycyceJEvve973HzzTcn6FsnlneDvjZIYY6f02zESmNS2sUXX0w4HGbGjBncfffdnHPOOQCUlZWxYsUKrrrqKubNm8e1114LwLe+9S2ampqYPXs28+bNY+3atQD84Ac/4NJLL+UTn/gEEyZMGPLzvv71r3PPPfewYMGCfmfhfOELX2DSpEnMnTuXefPm8dvf/rZ33g033EBlZSUzZswYjk1w0kRVk12HfqqqqnTDhg0nvZ7/45FXGZ2fxa+/cHYCamVM+tqxY8eIDbCR4rbbbmPBggX80z/90yn5vMH+TURko6pWDVbeky36jlCEnQ2t1m1jjBl2CxcuZPPmzXzuc59LdlWGNPKOGiTA1v3NRKJqQW+MGXYbN25MdhWOyZMt+t5bB1rQG2OMR4M+EKR8VC5lhTZipTHGxBX0InKxiOwUkd0icvcg8yeLyMsisllE1olIRcz0t0WkWkS2icgXE/0FBlMdCNqNRowxxnXMoBcRH/AYcAkwE7heRGYOKPYwsEpV5wIPAA+50w8A56rqfOBs4G4RmZioyg/mYGsndcEOFli3jTHGAPG16BcBu1V1j6qGgCeAyweUmQm84r5e2zNfVUOq2uVOz47z805KTaAZsP55Y4zpEU/wlgOBmPe17rRYNcBV7usrgUIRGQMgIpUistldx/9Q1f0nV+Wjqw404csQZk8sHs6PMcaMYAUFBcmuwoiSqNMrvwb8WERuAtYDdUAEQFUDwFy3y+ZpEVmtqg2xC4vIrcCtAJMmTTqpitQEmpk+vpDcLBux0piE+9PdUL8lsescPwcu+UFi1zlChMPhETH2TTwt+jqgMuZ9hTutl6ruV9WrVHUB8E13WnBgGWArcN7AD1DVFapapapVZWVlx/kV+kSjSk0gaN02xnjM3Xff3W/8mvvvv5/vfe97XHjhhXzsYx9jzpw5/Nd//Vdc62praxtyuVWrVvUOcXDjjTcC0NDQwJVXXsm8efOYN28er7/+Onv37mX27Nm9yz388MPcf//9ACxdupQ77riDqqoqHnnkEf74xz9y9tlns2DBAj75yU/S0NDQW4+bb76ZOXPmMHfuXJ588klWrlzJHXfc0bven//859x5550nvN16qepRHzit/j3AVCALp5tm1oAypUCG+/pB4AH3dQWQ674uAXYBc472eQsXLtQT9W5Di07+xrP6n2/tO+F1GGP62759e7KroG+//baef/75ve9nzJih+/bt0+bmZlVVbWxs1NNPP12j0aiqqubn5w+5ru7u7kGX27p1q06bNk0bGxtVVfXQoUOqqvqZz3xGf/SjH6mqajgc1mAwqO+//77OmjWrd50//OEP9b777lNV1SVLluiXvvSl3nmHDx/urdfPf/5zveuuu1RV9etf/7refvvt/cq1trbqaaedpqFQSFVVzz33XN28efNHvsNg/ybABh0iV4/5m0JVwyJyG/A84ANWquo2EXnAXfEzwFLgIRFRnK6br7iLzwD+xZ0uwMOqmuDffX2q3QOxdsaNMd6yYMECDh48yP79+2lsbKSkpITx48dz5513sn79ejIyMqirq6OhoYHx48cfdV2qyr333vuR5V555RWWL19OaWkp0Dfe/CuvvNI7xrzP56O4uPiYNzPpGWANnJuaXHvttRw4cIBQKNQ7fv5Q4+YvW7aMZ599lhkzZtDd3c2cOXOOc2t9VFydR6r6HPDcgGnfiXm9Glg9yHIvAnNPso5xqw40UZDt57QyOxBjjNcsX76c1atXU19fz7XXXstvfvMbGhsb2bhxI5mZmUyZMiWucepPdLlYfr+faDTa+/5o49t/9atf5a677uKyyy5j3bp1vV08Q/nCF77A97//faZPn56wYY89dWVsdSDI3IpifBly7MLGmJRy7bXX8sQTT7B69WqWL19Oc3MzY8eOJTMzk7Vr1/LBBx/EtZ6hllu2bBm///3vOXToENA33vyFF17IT3/6UwAikQjNzc2MGzeOgwcPcujQIbq6unj22WeP+nk949v/6le/6p0+1Lj5Z599NoFAgN/+9rdcf/318W6eo/JM0Hd2R3jngI1YaYxXzZo1i9bWVsrLy5kwYQI33HADGzZsYM6cOaxatYrp06fHtZ6hlps1axbf/OY3WbJkCfPmzeOuu+4C4JFHHmHt2rXMmTOHhQsXsn37djIzM/nOd77DokWLuOiii4762ffffz/Lly9n4cKFvd1CMPS4+QCf+cxnWLx4cVy3QYyHZ8ajb2zt4nv/vZ3PVFWy+IzSYy9gjImLjUd/6l166aXceeedXHjhhYPOT9vx6MsKs3nkugUW8saYlBUMBjnzzDPJzc0dMuRPRPLP5DfGmGGwZcuW3nPhe2RnZ/PXv/41STU6tlGjRrFr166Er9eC3hhzTKqKSGqd5DBnzhyqq6uTXY2EO5Huds903RhjhkdOTg6HDh06oYAxiaWqHDp0iJycnONazlr0xpijqqiooLa2lsbGxmRXxeDseCsqKo5rGQt6Y8xRZWZm9l7NaVKTdd0YY4zHWdAbY4zHWdAbY4zHjbgrY0WkEYhv0IrBlQIfJqg6qc62RX+2Pfqz7dHHC9tisqoOekOPERf0J0tENgx1GXC6sW3Rn22P/mx79PH6trCuG2OM8TgLemOM8TgvBv2KZFdgBLFt0Z9tj/5se/Tx9LbwXB+9McaY/rzYojfGGBPDgt4YYzzOM0EvIheLyE4R2S0idye7PskkIpUislZEtovINhG5Pdl1SjYR8YnIJhEZ+uaeaUJERonIahF5R0R2iMi5ya5TMonIne7/k60i8h8icnxDQ6YATwS9iPiAx4BLgJnA9SIyM7m1Sqow8H+r6kzgHOArab49AG4HdiS7EiPEI8AaVZ0OzCONt4uIlAP/DFSp6mzAB1yX3FolnieCHlgE7FbVPaoaAp4ALk9ynZJGVQ+o6tvu61ac/8jlya1V8ohIBfAp4BfJrkuyiUgxcD7w7wCqGlLVYHJrlXR+IFdE/EAesD/J9Uk4rwR9ORCIeV9LGgdbLBGZAiwARu7904bfvwJfB6LJrsgIMBVoBH7pdmX9QkTyk12pZFHVOuBhYB9wAGhW1ReSW6vE80rQm0GISAHwJHCHqrYkuz7JICKXAgdVdWOy6zJC+IGPAT9V1QXAESBtj2mJSAnOr/+pwEQgX0Q+l9xaJZ5Xgr4OqIx5X+FOS1sikokT8r9R1T8kuz5JtBi4TET24nTpLRORXye3SklVC9Sqas8vvNU4wZ+uPgm8r6qNqtoN/AH4RJLrlHBeCfq3gGkiMlVEsnAOpjyT5DoljTh3cf53YIeq/n/Jrk8yqeo9qlqhqlNw/i5eUVXPtdjipar1QEBEznInXQhsT2KVkm0fcI6I5Ln/by7EgwenPXErQVUNi8htwPM4R81Xquq2JFcrmRYDNwJbRKTanXavqj6XxDqZkeOrwG/cRtEe4OYk1ydpVPWvIrIaeBvnbLVNeHA4BBsCwRhjPM4rXTfGGGOGYEFvjDEeZ0FvjDEeZ0FvjDEeZ0FvjDEeZ0FvzAkSkaU2GqZJBRb0xhjjcRb0xvNE5HMi8jcRqRaRn7lj07eJyI/ccchfFpEyt+x8EXlTRDaLyFPuWCiIyBki8pKI1IjI2yJyurv6gpix3X/jXl2JiCwUkT+LyEYReV5EJrjT/9m9T8BmEXkiKRvEpB0LeuNpIjIDuBZYrKrzgQhwA5APbFDVWcCfgfvcRVYB31DVucCWmOm/AR5T1Xk4Y6EccKcvAO7AuQ/CacBid5yhfwOuUdWFwErgQbf83cACd/1fHJ5vbUx/nhgCwZijuBBYCLzlNrZzgYM4Qxb/p1vm18Af3LHaR6nqn93pvwJ+LyKFQLmqPgWgqp0A7vr+pqq17vtqYAoQBGYDL7plfPTtGDbjDD/wNPD08HxlY/qzoDdeJ8CvVPWefhNFvj2g3ImOBdIV8zqC839KgG2qOtgt+j6Fc+OPTwPfFJE5qho+wc82Ji7WdWO87mXgGhEZCyAio0VkMs7f/jVumc8Cf1HVZqBJRM5zp98I/Nm9S1etiFzhriNbRPKO8pk7gbKee7GKSKaIzBKRDKBSVdcC3wCKgYKEfltjBmEteuNpqrpdRL4FvOAGbTfwFZwbbixy5x3E6ccH+Dzwv9wgjx3Z8UbgZyLygLuO5Uf5zJCIXAM86nYH+XHucrUL+LU7TYBH7TZ+5lSw0StNWhKRNlW11rRJC9Z1Y4wxHmctemOM8Thr0RtjjMdZ0BtjjMdZ0BtjjMdZ0BtjjMdZ0BtjjMf9/4dLv/nW49DVAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nQk6Uirt1TYX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "802fe60b-d8b7-402c-c9ab-4e57a81bacd3"
      },
      "source": [
        "score = model.evaluate(x_test,y_test,verbose=0)\n",
        "print(\"test score:\",score[0])\n",
        "print(\"test accuracy:\", score[1])"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "test score: 0.027564922347664833\n",
            "test accuracy: 0.9912999868392944\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UY14uALMspQB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        },
        "outputId": "8619a96f-d192-4ada-881b-dd553b3bd12f"
      },
      "source": [
        "import requests\n",
        "from PIL import Image\n",
        "url2 = 'https://colah.github.io/posts/2014-10-Visualizing-MNIST/img/mnist_pca/MNIST-p1815-3.png'\n",
        "url = 'https://colah.github.io/posts/2014-10-Visualizing-MNIST/img/mnist_pca/MNIST-p1815-4.png'\n",
        "url3 = 'https://colah.github.io/posts/2014-10-Visualizing-MNIST/img/mnist_pca/MNIST-p1815-1.png'\n",
        "url4= 'https://printables.space/files/uploads/download-and-print/large-printable-numbers/3-a4-1200x1697.jpg'\n",
        "response = requests.get(url3,stream='True')\n",
        "img = Image.open(response.raw)\n",
        "plt.imshow(img,cmap=plt.get_cmap('gray'))"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fb76ea36630>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        },
        {
          "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": "0d_csA-VyaJu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "outputId": "d1200c12-78a8-45e4-9926-dc63fe91cbe2"
      },
      "source": [
        "import cv2 \n",
        "img_array = np.asarray(img)\n",
        "resized = cv2.resize(img_array,(28,28))\n",
        "grey_scale = cv2.cvtColor(resized , cv2.COLOR_BGR2GRAY)\n",
        "image = cv2.bitwise_not(grey_scale)\n",
        "print (image.shape)\n",
        "plt.imshow(image, cmap=plt.get_cmap(\"gray\"))"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(28, 28)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7fb76413dcf8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        },
        {
          "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": "X3FWDouDzpTS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "image = image /255\n",
        "image = image.reshape(1,28,28,1)"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P7ivBMkHz5wv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "b0a41a46-004e-4678-ba16-c638105f2dea"
      },
      "source": [
        "predication = model.predict_classes(image)\n",
        "print(\"predicted digit is:\",str(predication))"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-18-ef38fe7be2ae>:1: Sequential.predict_classes (from tensorflow.python.keras.engine.sequential) is deprecated and will be removed after 2021-01-01.\n",
            "Instructions for updating:\n",
            "Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "predicted digit is: [0]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zl7dJ1wH0DeF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}