cifar10.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "cifar10.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyNzISRZoWPaaytj7haX04JS",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ChalaniEkanayake/EN2550_Assignment_4/blob/master/cifar10.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E9kAldX0ynQQ"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from tqdm import tqdm"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t4QcEk_GFPhR"
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.cifar10.load_data()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xp3DeK2oFRil",
        "outputId": "46cbb0d1-0240-42b6-aabf-94d423faaa11"
      },
      "source": [
        "print('x_train : ',x_train.shape)\n",
        "print('x_test : ',x_test.shape)\n",
        "\n",
        "K = len(np.unique(y_train))    # classes\n",
        "print('Number of classes :', K)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train :  (50000, 32, 32, 3)\n",
            "x_test :  (10000, 32, 32, 3)\n",
            "Number of classes : 10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5tXckzE9FbqP",
        "outputId": "c99f7c18-f65c-4b86-b799-03719f143382"
      },
      "source": [
        "Ntr = x_train.shape[0]\n",
        "Nte = x_test.shape[0]\n",
        "Din = x_train.shape[1]*x_train.shape[1]*3\n",
        "\n",
        "print('Number of train samples :', Ntr)\n",
        "print('Number of test samples :', Nte)\n",
        "print('Dimension of each sample :', Din)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of train samples : 50000\n",
            "Number of test samples : 10000\n",
            "Dimension of each sample : 3072\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G7N4B23eLThl"
      },
      "source": [
        "labels = {0:'airplane', 1:'automobile', 2:'bird', 3:'cat', 4:'deer', 5:'dog', 6:'frog',\n",
        "         7:'horse', 8:'ship', 9:'truck'}"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 307
        },
        "id": "csNPGsawLUQE",
        "outputId": "10588566-325d-4c1d-d647-85086e0cb544"
      },
      "source": [
        "fig, ax = plt.subplots(1, 4, figsize=(20,20))\n",
        "\n",
        "for i in range(0, 4):\n",
        "    img = x_train[i]\n",
        "    ax.ravel()[i].imshow(img)\n",
        "    ax.ravel()[i].set_title(\"Label : {}\".format(labels.get(y_train[i][0])))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABH4AAAEiCAYAAACPwRUyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeZQl6Vke+OeLu++5Z9Ze3dXVrW61pJbUaglJIIFYBAgkGyPDGIxtBs2ZY8bM2OPBw8wYZg72YWaMMd6GEUg0MMYDRiCEEEsj0IYkWt3qfa+uLWvJPW/efYmIb/7IbE+p1e8TtWRV3bz9/M7RUXW9+caNG/HFG198lfe+znsPEREREREREREZP8HN3gEREREREREREbk+tPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjcM59xjn3X96E3IJz7g+cc1vOuf90NdsQkb3tZtWf622U901Evp5qkYiMklGpSc65v+Oc+8JubEtuLi38jBHn3Gnn3Lfe7P24An8DwDyAae/999/snRGRq7fX6o9z7n7n3M/e7P0Qkd2lWiQio2Sv1SQZX1r4kZvpCIDnvffhKwWdc+kbvD8iIgBUf0RkNKgWicg4Um278bTw8yrgnJt0zn3SObfqnNvc+fPBl/3YMefcg865hnPu951zU5fkv80590XnXN0595hz7t27sE//K4B/CuBvOudazrkf3flVwr90zv2Cc24dwM8452rOuV/f2fczzrn/2TkX7Gwj5Zz7eefcmnPulHPux51zXoVEZHSMaP35EIC/BeB/2Kk/f7Dz96edcz/pnHscQNs5l96pKbddkvs1/zrvnHu/c+7RnX1/0Tn33ld4vX3Oucedc//4WvddRK6OapFqkcgoGcWatLPdaefcJ3Ze80EAx14Wf41z7gHn3IZz7jnn3AcvieWcc//COXfWObfsnPsl51xhJ/Zu59y5ndq2BOBXd2N/5fJp4efVIcD2xXUEwGEAXQD/9mU/87cB/D0A+wCEAP41ADjnDgD4QwA/C2AKwH8P4GPOudmkF3XOvdM5V3+lmPf+pwH8cwC/5b0ve+8/shN6K4CT2P4I2D8D8G8A1ADcCuBdO/v5d3d+9scAfCeAewC8CcAHkvZJRG64Uaw/HwbwHwD8Hzv153suCf8ggO8GMGH9NuIlr3EfgF8H8I8BTAD4JgCnX/YztwD4LIB/673/P5P2W0SuG9Ui1SKRUTJyNWnHvwPQ23nNv7fzv5dySwAeAPCbAOYA/ACAf++cu2vnR34OwO3Yfja7DcABbP9D/0sWdvb3CIAPJe2r7C4t/LwKeO/Xvfcf8953vPdNbC+ovOtlP/Yb3vsnvfdtAP8LgA8651IAfgjAp7z3n/Lex977BwA8BOC7LuN1v+C9n7jC3b3gvf83O5OcAbYLyv/ovW96708D+HkAP7zzsx8E8Ive+3Pe+01sFxsRGSF7rP4AwL/23i9677uX8bM/CuCj3vsHdvbvvPf+2UvidwH4CwA/vfOAJyI3iWqRapHIKBnFmrSz7e8D8E+9923v/ZMAfu2SH3kfgNPe+1/13ofe+0cAfAzA9zvnHLYXc/477/3Gznv659h+lntJjO061L/M2ia7SB+JeRVwzhUB/AKA9wKY3PnrinMu5b2Pdv578ZKUMwAyAGawvSL7/c65S/8VKoPtCcT1cOl+zOy81pmX7duBnT/vf9nPX/pnERkBe6z+vHxfkhwC8CkS/1sATgD4nWvaIxG5ZqpFqkUio2REa9IsttcHXv66LzkC4K0v+42hNIDf2MktAnh4ew0IAOAApC752VXvfe8a91Gukn7j59XhHwG4A8BbvfdVbP8KMLB9Mb7k0CV/PgxgCGAN2xf+b3jvJy75X8l7f71+u8Zf8ue1nf048rJ9O7/z54sALv0s7KXvQURGw6jWH3+Zf9/B9kTmJQuX/HkRL/vs+8v8DLbfx2/u/CuaiNw8qkWqRSKjZBRr0iq2P1L28td9ySKAz77sdcve+/96Z7+6AF57SazmvS9fkm/VO7kBtPAzfjLOufwl/0sDqGD7QqzvfCnYT79C3g855+7aWX3+3wD8zs5q8/8D4Hucc9/htr9MOb/z5Vwv//KxXbfz+r8N4J855yrOuSMA/uHOPmEn9hPOuQPOuQkAP3m990lEqL1Uf5ax/d1hSR4F8F/svP578bW/hv0RAH/XOfce51ywU4tec0l8COD7AZQA/Lrb+WJ6EbnuVItUi0RGyZ6oSTvb/l1sN9gp7nx3z49c8iOfBHC7c+6HnXOZnf+9xTl3p/c+BvDLAH7BOTcHbH8XkXPuO65ln2T3qPCPn09hu4i89L+fAfCvABSwvRL7ZQB//Ap5vwHgfgBLAPIA/gEAeO8XAbwfwE9hexV4EdtfHpg4dpxz3+ica13LmwHw3wBoY/sLn7+A7S8T++hO7JcB/CmAxwE8gu33HgKIvn4zInID7KX68xEAd7ntbhgfJz/3EwC+B0Ad2x+X+M8/671/ENtfNv8LALaw/cWpl/6GIrz3AwB/HdtfWP9RPXCJ3BCqRapFIqNkL9WkHwdQ3nnN+3FJ962d7+35dmx/b8+FnZ/53wHkdn7kJ7H9sdIvO+caAP4M27/VJCPAea/fuJLx4Jz7TgC/5L0/kvjDIiIiIiIiIq8CWu2XPcs5V3DOfZdzLr3T1vCnAfzezd4vERERERERkVGh3/iRPWvn866fBfAabP/a5B8C+AnvfeOm7piIiIiIiIjIiNDCj4iIiIiIiIjImNJHvURERERERERExpQWfkRERERERERExlT6WpKdc+8F8IsAUgB+xXv/c+znM5mMz+XzZjyK7C7cAfhH0lLOjmXTfH0rkxBPp1JmzDnywgBot0yeijDkXcnZEUmRfQYARz7iF/uYv25s57og4U0liGP7PSe9J7rdhPfkEk4GiwcJ+5UK7DGQNH7ihI9i+qRBxHLJtpM+ALp4cW3Nez971S++y660FlWqNT89N2/GB72OGQsHPbov3tvnJJO16x8AZHM8nspkzViQcO31unbnzkG/S3M9qc3AtV0fjlwfpXKF5ubI8fJRSHO7Xfsc72zBjCTVk17XPp5Rwn4lffyahcOQ71dMardPeE/pNJ8upNP2efZIuJeR9xTz3cJWvTFStQi4sno0MzPjjx49eqN2TW6COGEQhyGvCfTaItc0AARs/pE4X+Nx9srXNhPcmx5++OE9XYuCwPk0eR4K2Hw16VnoGqLJM1JbmDB3CcgzWtJvRiTNz0GuL3oswa/bVIrfi9kcg80BLgetNwmnkYWTnoVS7CEfQIbMT4bDIc2NyHlMOtZJQ4A902YzfISxY5J0vDa3umYtuuqFH+dcCsC/A/BtAM4B+Ipz7hPe+6etnFw+j3ve9GZzm/X6hhnLBfzGOZW1j/7h6SLNnZ0q0fjMRNmMZVMZmpvOFexgwoDa2KzT+CC03/PkRI3mBpF9IfT7fZrb69kPvvkCf3CNEh4AOuThtDZRpbnw9rYH/QFNTYGfR7boVCnb4wMASiV7fGUy/Hh1E/bbs4XFgI8vdkxCsngBAD/+s//3GfoDN9DV1KLpuXn8T//y35vbPPfsw2Zs9dQzdH+iyD7u84dfQ3MPH7uTxicXDpuxfIGf7+ef+qIZO3PicZo7bNrXJQCkyHuuTvJalM7b9fm+d3wTzb3tdvt49rbs+wkAPPXkIzQex/b1MRjyxb+nn3rCjDXqazS3P+D1dziwa9HGOl/ManXs/Q4j/rqzs1M0Pjll18HIN2luSOZlvS6fWf3+7/3pyNQi4Mrr0dGjR/HQQw+Z20taNJAbJGGCzybi3Ta/Ltc3eE2Ympo0Y1HCP0QUinZ9TWVzNJfOLwDE5FHu6v+Zbu9KpVJ7uhal0wHmZ+w5aaFgP88kPYimA3tEsEUOAAjJw/POi5uh+hbv9ZIP7H9MKyXMoZsJ/2AWFO3rq5CzXxfgzw212gTN3dy05z6DNr/PJy0LDQfkZp2w8JMiC9hJiyC1En9W2jdr18jzy8s0tz2wx1e1am8XAMIhP2Lt9pYZO3iAP9NmMvb4S/qHuP/0ycfMWnQtH/W6D8AJ7/1J7/0AwP8L4P3XsD0RkauhWiQio0L1SERGgWqRiHyNa1n4OQBg8ZL/PrfzdyIiN5JqkYiMCtUjERkFqkUi8jWu+5c7O+c+5Jx7yDn3UJjwOTsRkevl0lrUbNi/fikicj1dWotWV1dv9u6IyKvUpbXoWr//RURG37Us/JwHcOiS/z6483dfw3v/Ye/9vd77e9MZ/j0qIiJX4YprUaXKv3dGROQqJdajS2vR7OxIfResiIyPK6pFSQ0iRGTvu5aFn68AOO6cu8U5lwXwAwA+sTu7JSJy2VSLRGRUqB6JyChQLRKRr3HVXb2896Fz7scB/Am2v8D/o977p3Ztz0RELoNqkYiMCtUjERkFqkUi8nJXvfADAN77TwH41OX+fK/Xw1NP2zWnvma3tJzindzgpu0fmIkqPLcwR+Pt2G6N14r4Z2K9s1v2dXq8VXenm9DWN7LbvK6l+K9s5tP2fochbx+bIi0OczneHrTTa9N4SFoou940zSXdIjFMaFFfSPMB1iKtzzeikOYWi3ZbRhfwjz+6VMLHI0kbzE6Pf6cW+86tVJqfx1FzpbUoiiI0SMvL6Qm7dbWfnef7krZbNO47fCvfr5ifsyC22wLHHT4Oe5vrZsx3eUvgAzO8Rh4+dJsZO3TbEZq7/8BBMzY3x491JmOP03DCbmMMAIcOLtB4GNrXfK/H27jWN1tmbG2Nt5lPZ5Nudnahm5zm122+ZO/3VmOT5ubyfLoQe3v8ZRLqSWOrbsYG/b33vRNXWo+YpFbHMvr6Hf6dchvnTtL44jN2/laDz6ne8S3vMWPVQkKtSfhQgCP9mzVqR8OV1CIHIJOy7y9RaM9PYvI8AgAuaz8L9UM+d2FtwLc3bo/DiQqfB1RJ2/RBk19bcZc/wxUzBTNWK9oxACiSa7Oc5c8Fa+TZMfb8WSif5/fq2dkZM7a5yecQefKe9u/jc8xUQqP5uTl7zp5JqHOnFi+YsWyGP0tPTNjjBwDKJDxd4185wepru8PHJqPaLCIiIiIiIiIyprTwIyIiIiIiIiIyprTwIyIiIiIiIiIyprTwIyIiIiIiIiIyprTwIyIiIiIiIiIyprTwIyIiIiIiIiIypq6pnfuVCgAU0qQ1Gukid4S0aweAo/N2W7S5WbvNGwAUSLttAHCkVWC3z9sg94Z26zxPtgsA2QJv94fQbm/nY96yrzZltzgMh7xtXpa0KIwimopUlrcK7A/s4zkM+fEqkm2nS/xY5hP2K3R267zA8zaWIWnJl+JvCeUSb0XZatvtvYek9SYABOS1mw3efnbP8x4g7ewHfTvW6fAWnkdvP2DGWm3egnEw5PVkasauc+kMX8c/fvx2M/b2t91Lcw/M2y3XAaBWmzVjwzQvCkXSPjSd0MnbkTaw3bbdUh0A+uT8A0CxYF97kxO89eixW+8yY8888xzNheP71e/b13ytOklzM3ZHXWw1lmmuBx/3cWyfrM1NPu67HXKf3Hvd3HeVf7UfgBGRdB4CZ8eXFk/R3Me/9DkaH3btaz5T5td8l9zLq1N8bhyTuQsAeGffczRq9x7nHLJp+5w6cr4nZ6bptttsDEe8XXuY0O7dkWtz3wK/Vy/M2vt96sSLNHcmzdtxL+xfMGNByOdrAXk+rCa0J5+uVcyYT/FnoVpCi/EieSZJBfw8zc7breDzCS3qk55JQm/Pm2oT/D0dIM/SqYRVknSGV7pcyp7fxgM+N65WqmbMD/lzJ6Pf+BERERERERERGVNa+BERERERERERGVNa+BERERERERERGVNa+BERERERERERGVNa+BERERERERERGVNa+BERERERERERGVNa+BERERERERERGVMJHep3l3MeeRea8UrF3p3bD0zSbU8XUmYsE/dobmtjQONRbK+PdTv2+wGAIGvHqhNlmpvO5mi8vtW0cxPO7FSlaMaajTbNHfTseLc3pLkejsbLpZIZGw66NDeI7DedyfFjGUV8v9Mpe7/7fZ6bzdiDIIj5+Om3NmkckTdDOfuSAACEcWzGttp9nrzH+ThG2LPHkwsjM5bLFui2t9bWzNj0wkGae/i1t9H43KH9ZixDxhkAILTH6TDkNfLZi+s03jm5am874PX1uSceM2NvufMumvtN973FjHlvXxsA0Ghs0fjZMxfMWDaTp7nZbNWMzcwe4K+7+ALfdt6u3a0ur92Nhj020xlem6tV+3UBoNvtmLGIlzmEoV2LcrmEcT3mnOPnRW4MD3uMAsCwb4//C4tnaG61yO8pxYmKGVvZtOeBALB+8bwZmz90mOYi4JMIVmFdoHG716RSAWpVe6zlC/Z9b25ujm57Zd2eQ+QT5udbm3Uan5+ZNWO5hIlwoZAxYwcOLdDcEnleAYDhwL7xZcHvazny/Nfp8mehQ/vtc+EzvI5lE+63g4E9n5uZrtHcdGC/dr/P5y6VpPlH3z4mzS3+HNXv2/P96Rn7egCAQok/bKedve30gB/rXtt+T2HCcyej3/gRERERERERERlTWvgRERERERERERlTWvgRERERERERERlTWvgRERERERERERlTWvgRERERERERERlTWvgRERERERERERlTN7Sde9o5TObslyyQln61Em93OVu1W/JFsd1ODQB4FEilSTvAgK+d9WO75Vo6oed62vO2exFpX+dTfL9WVuz2iNGQH5Fmx25b2ol46+ZywW5zDAAgbfVSCe1UA2c3F03lePvlbpu3sy5m7P1OJ7SN7vXsY9Id8j7HMW2YCtRb9n7XO7zdX6tjv3ZvON5rwj6O0e/Y7SPLpG1pdcpuHQoAb3rDPWbs0K3HaW4z5OPhuZOLZqxBrksAaNXta369ztu1X1zi7TCrNXJMgj7N/eRvfcyMZT7Ix+G7vuGddm6Gj/+Fhf00Dm+3Pq8ntFD+6iOPm7F0hreuLVV4jQwjuyYMWrztLbstzM5O0dwoobavb9jHKwBvxcruhRMTvEWsyG7x5F7O5hcAsLph19DTp8/S3D7JBYBK3m7722k1aO6zjz1ixhaOHqO5EwsHaBzkeCVMi+Cc2r2PmnQ6jZmZaTMex/YcfNDjc+j5BbvFeDHPn+9yKd6Sfd+sPf8YDvm8aH1txYxVSGt7AEhn+PwkHtjHK5Pm4z8I7Auo2+HXPMimgzw/lv0BbxXfH9jzuRx5hgeAVsOeN5XKfI4QRfy5dH3DnqPmMiWay0rRgLxfAGi2WjQekJMxaPD3NBjYc9hyib8nvk8iIiIiIiIiIjKWtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKmtPAjIiIiIiIiIjKm0teS7Jw7DaAJIAIQeu/vpS+WcpidyJvxSiZlxvJ5OwYAQcqbsUKhQHOHYUTjMZwZ835AcwehvV/RYMhf1/O4j+zX9ukszW0O2vZ+RfxYd6LYjIUkBgDNNn9P5zfs/coEfNvVln2ehktrNLe71aHxwzO3mbG5uYM011W2zFh/c53mtlr28QCArWbPjK1tdWnu6UV7v6LUNZWGm+JK6pELHHK5jLmtYapixrqFMt2PUw37uD/6hQdp7sZ6i8bPX1g2Y5mUPf4Bfv30Q17Hej0e3zdrj5eVpTM0t5qza1Wz3qC5z586Ze/Tvhmam8nwMb7v0IIZ209iAHB2adGMPfeEHQOAuX2zNH76LKllQ14j44Edj9L8PpjP5mg8l7avp26Pb7tarZqxdJq/7ii60rmRjAp7vuY9H8Pnz50zY6fO2jEAWDxxksZnKvY95+BMieZePGvX3yce+grNvffdEzRerNbsIL8dyQ1yRfMiAAHse8Sgb883o0Gf7kfI5h89Pv9Op/jvKDTqG2bMgV+3PrLj5y9epLm1sj1PBIAieQ5r9O35NwB4b9eibJ7PXYah/Zw1TDhPLuDHOibPy3GKH+tc1p4jkNILAOh0+X5nc0U7luFziGLeLlY5Mj8FgK16PSFun+dyntRPAC5lP4vT2ptgN57uvtl7z5+qRURuDNUjERkFqkUiMgpUi0QEgD7qJSIiIiIiIiIytq514ccD+FPn3MPOuQ/txg6JiFwl1SMRGQWqRSIyClSLROQ/u9aPer3Te3/eOTcH4AHn3LPe+89d+gM7heZDAJAn3+EjInKNaD26tBZNTE7drH0UkfF32bXo8OHDN2sfRWT8XXYtKpDvPRSR8XBNv/HjvT+/8/8rAH4PwH2v8DMf9t7f672/N5vWJ8tE5PpIqkeX1qJSmX9Bs4jI1bqSWjQ7y7/MW0Tkal1JLcpm915TDxG5Mle9EuOcKznnKi/9GcC3A3hyt3ZMRORyqR6JyChQLRKRUaBaJCIvdy3Lu/MAfs8599J2ftN7/8e7slciIldG9UhERoFqkYiMAtUiEfkaV73w470/CeANV5KTSaewf7ZkxqvZ0IyVi1m6beeHJOoTcmMa73c7ZiyAo7nTlZoZK5XyNLexxbsv1qpVM9bsseMBnDlvb7vV59/FlCWH60CRD6l0pkvjp9frZqzv+X5lnH2ea9UKzX37XffSeONiZMZ8h4+v2oz9uel+hx+vVov/Ul4uY2/70AJ/z3Nz82ZsudGjuacfO0vjN9qV1qMgSKNYtN//St2uRScWF+m2n37K/ge1IMPPd9Tn12232TZjqYDXsW6/YcbqTTsGAM12i8ZPn3vGjJUKfBzecewOOxgOaO5ffv4zZuzILbfQ3NvvuJ3Gp6ft2p3L8/NYq+bMWBBu0dx2n1/z3U7fjtWbNDeK7Os6X+Df79Bq8G1XK/b9KJfntXswsMd9p2Pff0fR1cyNOHZd8/kHdy25CfgtEZ79QMJ8DI7vt7umbzCwtx3H9j0BAIahPYabHX4/Pbe8QePLJB5FczT34Jx9PJ79yoM0d25hH43f/pav+3aHS/AaGXj7WJOp3E4yD5NNJ873x8mV1yIPR65N9lEw7/lJCyP7+uj3+HPBZMF+bgSATGCf8HTA72u9gX1vyub4M9qgz+cng4Y9X8uWCzQ3m7WfeV3C9+VGoT1HKOT56w7JvRgAKtUJM5bP8+PlnP0c1WzxOeZwYOcCgMvYc66k/cKQjE0y3wKAaMCLUTZtf61EdYp/1+hwaN9zGu2rnxfpS3dERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMbUVbdzv6oXSzlMVexWcumB3co7l9AGuZgrmrF+l7enGya06ZyYmDRjSS0MB5G9tjYc8hafxbLdBg4ALqzabeZePMNbBq827ffc4YcDRwp2K8EPfOM9NPfgPv6efufhk2bsSyeWaG4Y260V0wE/T836Ko13WvaxrlR4u0hEdqvJfJ7nZhPaIBednR9G/EQePrTfjFU2eOvmPx+xdu5XKpVKY2JqxoyfWHzejF08fYpuu5ixx8pWe5PmthorNO5iuxVtvcnbYda7dr1J5/g4nJnnLYMLFbv1+YGjvJvsITLGTz32JZqbcvY1P4x4+8/VtXUaf93r7jRjtx2/leYe2jdrxspveyPNffxZfm31e3Zr0n6GtyqOYbdcjz2vF0tLF2g8m7PbqdYm+fgB7La33S5v9Tv+knpbX+1Wr7GdO9uthHkRmzd58HGY2K6dtHt3Ce/5WqKHjx41Y8WKfd0BQKOdMMad/Z6fXOT3jELavi7TPd6O+qkvfpbGpw/Mm7HJg7xGutAeA471Y0fy2I3JfC9hKvgq5xAE9ljzsX3wCiXeJrzn7HtTtsTbtUdt3lIbzn4+XJi3xygAhOtkQIT8+ihl7WsLAPpkTlZb4K28O52rb9c9M2/PP/ot/p5S5JkCADKsbXouYQx07eORy/LcIMufHbfIGBkO+VwwRZ6Vej2+foCYP6MVSCv5dDZLc3tD+1ytrvFnVka/8SMiIiIiIiIiMqa08CMiIiIiIiIiMqa08CMiIiIiIiIiMqa08CMiIiIiIiIiMqa08CMiIiIiIiIiMqa08CMiIiIiIiIiMqa08CMiIiIiIiIiMqbSN/TF0mnMTU2b8e5Gz4wFju9qqzO0tzsI+X65FI13hpG9XzQT6A4HZmxiskpzB5Gn8ZPnLpixjYa9zwDg01kzlkrxd1XN29ueSzdpbn6jT+PHqwtm7OIU36/l+ooZ63fs8wAAjzz/PI0HYWzGhiV+HlGbJxvm47pWK9J4JbbHSG9gXxMA4AcNM3Z0tkRz97p+v40XX3zQjD/74gkzduHii3TbUbNtxio1flzvOH6Uxu++824zdnG1S3PPrNr7NbtAxiiAI8duofHK9JwZW960XxcA/NopM3b2zFmau1pfN2N33kVT8W2330nj7ZZ9PGNeXuEHdr156stfornH77iHxucPTJixLz/4OZq7tGxf88Mhv0/2uryGbm7atb9QtvcZAGJv19d2h4+f8Xd9/n3O8elFIu/JBsh9aTtsX0DDkI+zbNaeuwCAo2/M0Vy61wnzxMnJGTP2zm96N8194tFnafz0qTNmLAp5MTqRWjJj+aP7aW703As0/sRn/9KMvfV7ZmluoVi2X5efJrikOImF/CwnbDfhhfe4YRjh/OqWGWfXfKlv13AAKJO5T2/Ax3A5lafxA/smzViuyM9ZatOOTRZ5rZko8v2qLNg1oR/wcfj8kv18NzHBnzn6bftN9Tr8Pp9JONbDhp3f6/Pnu5jU0FSG19dWiz9bhmT6m/QsPTthP2dNVe2xBQAvNE/S+PSknZ9wS0G1VDBj8bDCkwn9xo+IiIiIiIiIyJjSwo+IiIiIiIiIyJjSwo+IiIiIiIiIyJjSwo+IiIiIiIiIyJjSwo+IiIiIiIiIyJjSwo+IiIiIiIiIyJi6we3cM5icsVs8Tpbt1mVBkKHbrjfs9nXDdovmBhFvJRjDblPoM/wQlst2a7wheNu8Z07yFuPtvt3mNp/P0dx81t7vQom3EJ9M2e38Hj6xTHPDAT9e/Zrdzn12kh8vB7vF4TDs0dzOgIp2NlUAACAASURBVLfCbnfsdoCDkLdHdEPSnjahO2gmSGg/G9j9ADNpfqxD0nrRJ7Q/3OvarQa+/LkHzHh6/g4zduzO19FtFwZ2vbjzruM0947bD9J41LPPtw8SxjDWzFg6w6+tVIq34x6Gdr1pNzdobm1gXz9hwjg8u2LX/Xz5PH/dhDadtx47asZ8wr+ZdOsdM/bsXz1Kc32Xt8W9+zvea8Ze9/pb+X49ZLdzf/HEaZpbJO2XAaA2MU2i/B7bIPfvft8+lq8KntwDrqW7NGvHDsAntL2mLbM9vye+cMJuE97t2vMaAHjNnXfSeC5n18ggqQ84EXveezcm0+m3v+Mbae7ZU7xW/cov/YoZC7tkfgHg7GrdjOWKfJ54fIrXuec+/5AZmz3Ia9Fr3nGfGesgoeV0zPcrS87zRsduVw4A/YE9L4pCXsf2Ou89+qF9/9nYsO/lxQ6fY0+ReXAm4VE0X7ZbwQNAr2Pf11oJ7ctZIUslzO37Td6+fLZi3zOfe+EUzS3n7eewcsF+VgaAft+eC07um6K5LuLP2mHHfs/5hBWFZs++fnI5PgddWrbb2wMAYvuYlGt8/trr2nOMcDikuYU8vy9USlkzttHkaxO9vn1NVcp8PsboN35ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMZUOukHnHMfBfA+ACve+7t3/m4KwG8BOArgNIAPeu83k1/OAUHGjmbsWJJc3s4tokRz0wnrX0Fgx4eI+X4VamZsbalJcztr/JDeOpU3Y/0eTUW+VDRjdxw7QHMDsvEwxc9ho8HfUzq1ZcYqWX4epyePmbFjxw/T3FNnv0Ljzz5/3oxl032a633LjIUhvwSDdJbGM1n7eMcxH5sxnBlzbjTXhHerHg0HIVYW18z4G9/w3WYsl5ul+ziVsmP79ldp7kad14TFExtmbBDnaG7gIjOWSvOxEnk+xkHGcdTv0lQf2a9drs3Q3PVW24wFCfUi9p7GARLnhwvlvH2ej+4/RHPzKb5fAex68rq7b6G5ExMTZuwT3T+luUsXee0+MLffjEWO35AyGXv8NBoNmgs8kxC/PnZ3bmRj49QlDGFPcn0U0tzEW4Cz7x+L58/S1D/41CfNWKNhzwEA4O1rKzT+ze/6FjOWy/EayY51wiWPkNWxSoXmvu/976PxE889b8b+7I8eoLmNoX2enz2/RHMnXYHG8z17kHz5j3k9SU+XzVgwb9cpAGjX+RjJxPa97mLjHM3datrb7vUSJtY3yW7VonQ6hbkpe6yGPfveUynza8uHAzOWSvNiUyjweTC7lXe69usCwCC0XzuX5/PzO++4jcaXlpbNWL/Pi/fMrD3PDKMhzY1BnofLfF406PBKlyrYdT8V2NcdALQ37Gtrq8Ov6VqVz51bHft4RjE/Xjmy9jAM+X3ywGE+n2PPWZsN+3oC+DPcxBR/DmEu5+nufgDvfdnf/RMAn/beHwfw6Z3/FhG53u6H6pGI3Hz3Q7VIRG6++6FaJCKXIXHhx3v/OQAv/2fm9wP4tZ0//xqAD+zyfomIfB3VIxEZBapFIjIKVItE5HJd7ec55r33F3f+vARgfpf2R0TkSqkeicgoUC0SkVGgWiQiX+eav8jDb3+I3PxwnXPuQ865h5xzDzU7o/n5WBEZD6weXVqLwoTP7YqIXIvLrUWrq6s3eM9E5NXkcmvRcMi/o0VE9r6rXfhZds7tA4Cd/ze/bc97/2Hv/b3e+3srRfvLiEVErtJl1aNLa1E6nfi99iIiV+qKa9Es+RJPEZGrdMW1KJMhnSlEZCxc7cLPJwD8yM6ffwTA7+/O7oiIXDHVIxEZBapFIjIKVItE5OskLvw45/4jgC8BuMM5d84596MAfg7AtznnXgDwrTv/LSJyXakeicgoUC0SkVGgWiQilyvx8w7e+x80Qu+50heLvUe3NzTjbtgl2fw7OdrthhkbDPn6Vhjwj6C1Ok0z1iAxADhwyD7EPuS5R2YcjR/bnzFjnR7PPXD7G8xY1vPvYtrcss9hYWKa5mKd/yrpoYV9ZqzebtPcW19z3IxVJ4s0tzp5J41vrtrnanNri+ZmsiUzFvgczR3G/DPXcWzHoiG/ZgIyRLY/Ej56dqseBUEaxfKUGc+Qt1+vm59sBQDkpibMWCckJwxAL+Fr0AqTFft1Y37No2ePJZ9wJ+gNOzSeL9gbCNyA5saBnVue3k9zs/7ljUz+f6nCJM31WV6LYme/ZxfZ1zQABCn7PWVKWZpbKPN42Ldr0fr5ZZo7XbI/VvT+7/oOmvvQY6dpvNW1z3Ovz7/Hpt+17/0TFft6upl2c27EkXuA43Obzc11M7a1aV87AOBSvJ4srdp18EsPPUhzH37qMTPW2KjT3P6Q15PXvu5uMzY3O0NzU+S6bTR5DazX7f0+evAgzd1/cI7G/86P/ZAZWzz/Is39q8ceN2P9Nq+BL5xbovHigp2//uSTNLfzu3bs2DveRHM3W3zu3OnYzwN9x8fXYNg3Y3E85vMi51DO2ef0zmOHzVihyOfY7J64tHjRjAFAGNrnBABKZfv6qbf4pCrl7PutA6+BzS0+DldX1szY0H6M2mE/37VaLZoZe3vjnQ5/jmo1+PGqFu056AD8TXlnP5OkAn4vq1bs1wWAQtEeX+k0r3OVir0GkAoS5onsIQzAqbOLZsyl+Vwvm7Jf+1q+M/mav9xZRERERERERERGkxZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGVGI7993k4RE50lI4slu9JbWXLuQLZqxc4W0GL6yyNvLAqXN2K9o06/sMILt8wYz1lnmL2+Nzdjs/AHjPu+325S+e561aKwfstr4z0ws0d2XVbhk8MZHQ5jjm7ylLWuetrJ6nuem83aZztc7bRZ6/yNsjZjL2GJqo8nZ+3a49Rnyar7061nMdQEzavQeO5zrSPjEaza6luyabzWHf4VvMODs2vZ7dKhYAlht2Wc1O8HbCw5C3d3QZ+/rpJrT4HHr7PaXTOZobpni8WK2asblp3j7Xb9j1dzC07wkA4GL7PRUK9j0BABK6dCL29mtHEWmxDSDI2Bv3KX7Nt9q8Rawj7UNzCS1RG6R2F4pTNPebvuH1NP7ci2fM2JNP87bQrYbdYjabsVutjgcPgLWQZu3c+Za3GnY74c9/8Qs098yFczS+1rCv682EMRyU7DqX7/M5xMq6/Z4A4PNf/LwZO3r0EM3N5ew6d57MAwFgOLDbzHc7vAa2mjyeITP1O99yK8199MQTZmzQ5Df6c3V+rytm7eN1sMav21MPfdWMpXK8jgX7ea3aCjv2tmkmAG+PzX6ftxXf61IOKGftI1Qq2tdmJsvn9rUJ+5wVEurY5vo6jT/1zPNmLCRzBADIZctmbKo0SXMvnOfPJOtrdq3qhfz6aLBW8Y6/J08eSer1TZo7tMsYAGDQt3+gWORX19R0zYy5hPfUD/mcy8d2Lev2+DO+J/ffMORz0KSaEJH7d4FcT0nSGf6swOg3fkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExlT6Rr5YKhVgYqJsxsN0aMZarR7dth9GZmyruUVzz5xdpvFWq2XGCnm+dnbxVMOMzeezNPfAgSM0PrH/FjOWacY0F/mMGTr4hvt46tJ5M1YIV2luBH4e2207vq84S3MHkf2eXckedwBwsLSfxisTC2asub5Ec1eW183Y0NnnAQB6gz6NI/BmqJTL09RB1x7XmSzfr73OO8C7lBkfDu1a1Gk26bZzhYIZazY2aO6gx893p2G/dsbRVFRKOTM2OzlFc6tTJRqfnbDfc5Su0dxuzj7WG0f4ddmPLtrBYYfmRuGAxuPYPqBRwOury9hja2Jqkr9ulLDfZGzWavZ5AICss+tFvVmnuX5o1wsAuOdOu0ZOVOyxBwCf/OSfmrHV5TWau9d1ex089cxjZjydtmvxcMDH8GbdPqf1Fp8Xnb1o3+cBoDY3bcamEsbh9Ix9L199kVzTAJ558gkaf+DPHjBjtSrfr1Tavm77A/vaAYBB3567/PGf8HlPJuGfYPcfnDNjxRl+r37DPa8xY4984Tma2wGvc8+v23PnQsTvGZNhxYyd+PLDNLc+y+c2G6Q+ZwY8N2T3/g6vzXtdNpPBwQV7rEWx/Zw1OcHvayky38rM8NyFWbvWAMCn/+KzZiyO7dcFgImKfZ9fusiv2/lJPpYmavZzR32lS3PXVuznionJKs0tlexny1pCbqXE54KVmj2fK5V5LQq79ns+eeIMzU2l+fNyp2/fCwcJ98lB3x7XqRQvzi6hRhby9twnSnj+Gw6Hdozcb5LoN35ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMaUFn5ERERERERERMbUDW3nHkchmnW7tXV6wFoVJ6xRkY596RRv59dJaGs6WbHbUk6UeDu/7qbdzn1uP29ReOD176LxJ8/ZLeqeP8Hb1719n92yr17nufPH3mDGAvB2l4M+b/c+4e3WeI0Ve+wAQGFgt77bN8VbFNYj3m4483q73WS3ztvP/uWnPmHGzi3y45FKbKtut6Ls8u6zGJJ134C0ERwL3gOknXc6tmM1fsnjUM0+J6+5dYLmlvMJ7YZJHWw3eDvuXseuc4USP993HOfXz6EjB81YkDlCc1uk5fShffv4fp1aMWPVKX6iphLamqZJ+9A44dry5JaTLxVpbtiz2wkDQEBeOxPw+2QPfTM2PWO3ngWAVkIr43bdbj97YNZu3w0AH/iebzdjH//DP6O5e1273cIXH/yiGe822maslOcts9/3vvebsdDze97DTzxL47UKuSfGvNXs/rl5MzZc5m2Ot9p8HHZesFuUT+b49VGq2cezPMnHcL5kz11qE3wOWqvyWlSt2tdmoczrybu/5a1mbGuNz32ffPIkjUdD+153tp7Qwj5jz23SS7wGNjd5PKzY99GgMENzzy/a87kGuRbHgYeHJ3PwHJmPJrW9HrbtY5dL2eMIAHyGx6OYzGUDPoemex3zedGRI7fQ+Ay57x282KK5uZy931VSpwAgRY7nysp5mvv2t95H4wv795ux0PNrvrFuP+9srm3S3PU6v/bSKXtiNDtjt6AHgJhM6OLIbvUOALUynzdtbtnrGj7g43rQtY9nNOQ1kNFv/IiIiIiIiIiIjCkt/IiIiIiIiIiIjCkt/IiIiIiIiIiIjCkt/IiIiIiIiIiIjCkt/IiIiIiIiIiIjCkt/IiIiIiIiIiIjCkt/IiIiIiIiIiIjKl00g845z4K4H0AVrz3d+/83c8A+DEAqzs/9lPe+09dzgumSNv6qNsyYx68330Au6d95FI0d3NIw2g0vL1f/QHN3VcrmbG3fPM309yDd7yNxn/3Vz9qxhZKZZqbGnTN2PmTL9LchVvvMmP56dtobsk3abyzsWLGCvEkzR10O2ZsrWnHAGBi9hYan144asa6rSrNDUg4yvZorgv4uB8O7fHnwohv29vxMEwsDTfcbtaiSqmId33Dm834rXe9wYxdOH+ebvvA/ikzdvvxYzR3YXaOxlPeHg/NZp3m9of2NZA0zsolu44BQLmcN2OpbIHmZmJ7DHfbq2YMAN509xEzdvT2ozR3GPPC78m/i4Sxfb8BAE9udKkMv7aGPft+AwDx0H7tIM3/LcflyXlOyO0P+fFKpzJmLBrwsTk7Y9+v3vmNb6G5v/2xB2j8etmtetTvD3Dy9EkzvrWyacaO33Kc7mOhYF+3Fy7Y91oAOHPqLI2XS/Z1zWoNALiGPf/o1vm1hYRadduxW83Ysdkaza1M2jfrlZUtmjs5ZV8/+w7x+tls8OOVje1YPubz2yp5z9/2Xj4H3dhs0PjyOXsMrfXJTgMobtnbnqvyOVXa8Rp5oGLfg0vzCzT3/OnTZmzQ4fPXm2W3atFgMMTZxXNmnM0Dms023ceJXNZ+XfB7S5S27y0AUKxU7G13eT2Zm7WfK3KBXacA4NitB2g8R95zkOHzomzOfs+FAj8eAamRvsvHcL9hP4cDwLBmH5Ppfby+BqGde+TQQZqby/Na1Gjbc4xsls+50s6OhwnznlSa19+IrBGk8vy+4MO+GSuX7BqX5HJ+4+d+AO99hb//Be/9PTv/u6xFHxGRa3A/VItEZDTcD9UjEbn57odqkYhchsSFH+/95wBs3IB9ERExqRaJyKhQPRKRUaBaJCKX61q+4+fHnXOPO+c+6pzjn8EREbl+VItEZFSoHonIKFAtEpGvcbULP/8XgGMA7gFwEcDPWz/onPuQc+4h59xDrQ7/PhwRkSt0dbWozT/HLCJyFS6rHl1aizod/j0SIiJX4YprUZ98f5yIjIerWvjx3i977yPvfQzglwHcR372w977e73395aL9hddiYhcqauuRQlffi4icqUutx5dWouKRf4lnyIiV+pqalEuofGAiOx9V7Xw45zbd8l//jUAT+7O7oiIXD7VIhEZFapHIjIKVItE5JVcTjv3/wjg3QBmnHPnAPw0gHc75+4B4AGcBvBfXc6LOQCsC2NE2qa5gK9RsU60vsvbsTnedRJT00UztlDkvxr5pntvN2N3vp23a99c4R9HyYV2e9FbD/LWeDF50wtzszQ37NnvuVPnH+cbhPx4Dbv2kIzAf0vjxfN2G8onnnyI5r79bXy/pxemzVijydviZuzhg5mjvJ1fnDDuowFpyU7aCALA1qrd/rDfJDt9k+xmLSoWC3jz619jxl/7Rrude/du3pK9VLNb0SaUGnjHWxUHpGX2VIm3qfVkKCX9C0Ac8z0P2a+IJ7TD7Pftj7ocu+0wzS1k7eun2+btl32QcPsjLT59Qjvh2NvxKOEcxzHf9qBrH68o5vUkSNuvHSSMguY6bzl95tSiGXvHO99IcztDu8VskbWgv4l2qx7FUYT2lj1WOz37fOeKebrtraa93TOLp2nuBKljABC1e2bM9ew2tABwcemEHbuwRnNdwLf9we/762YsbvHvv/3zL3zGjJ15/DzNna7Zv9G+9AIfwwf28zq3NVy2gxk+/5ianjdjr7vjbpo7+ACvkR/9yG+YsW7THh8AcKFO5rdp/umA/oDfj1pr62Zsf8K4zpJW2TNzEzT37Gkavm52rRbFMTpd+/qKYY/jQWjPRQFgatZuPx3H/Lmg1+NziEOHDpmxp598juZmyD1x3wJ/FpolreABIEWeszK8IzuyOfvaKybU/VSK1Jsunyd2G7xt+saqXW98wK/5ArmXJ72naoXPixodu7b7iI+fQt7+rVuXUIuGQ/6cVS3Yz1IRGXsAUCWfksrwLvJU4sKP9/4HX+GvP3L1LykicuVUi0RkVKgeicgoUC0Skct1LV29RERERERERERkhGnhR0RERERERERkTGnhR0RERERERERkTGnhR0RERERERERkTGnhR0RERERERERkTGnhR0RERERERERkTCW2c99N3gNxGJnxbj82Y9lSmW47nc6YsVQwoLm3LUzSeL5gr48dPXKI5r7hnd9sxvbd8Xqa++iXfpXGDx+y93vhta+judnZY2YsXazR3E6vZca6jSbNXb6wSOOby+fMWDTs0NxCJW/GZmbs8QEAixceofH5fQfMWNixjwcA+G7fjLn2Js2NfJdv23kzVsjx95xdsOONnKO5e10QBCiUSma8nM+ZsVIxoWymU2Yotk8XAMA5ftwDEo+9XT8BIB7a8djzHXMB/zeCEPa2g4Sh5J297fLEFH/dyH7dKLbPAwAg5jvmYd+rgqQ3FdnxiNyrtl83YZCE9v3MxfY+A0COHJNMxM9xqcePp1+2a9XqyWWae/COg2ZsLeD1da+LfYxB3z52nX7bjJ04dYJu+/c+/jEz9oXPfpbmOs/H+HLDPi+rZ/h9PkNK1TBhDGcX+PzkLz/3eTPWb6zR3KdfeN6MtZdDmltftfd7YtqemwDA6hLfdmPLHgOTEwWaO4js9/SZz3yV5haq0zQ+OTNnxtaG6zS307ff8/lmj+b6hPlJkRyv1OoKzZ2YtsdXKsXv/V998FEaH3XOOQQp+/7U79n3nlw6S7fdH9jz4Fye33sCMncBgGhg18/mZp3mdloNM3bLYfs5CQAKCeOwXKyYsdokv26H4dCMRRF/pk2l7OM5M2PvEwCsrPBnjourG2bs4Scfp7m33XbYft1V+zwAwIWLqzQewh5fE1X+njNk/prL8dodkvk+APR7di1LmIKiODVhxhqtq58X6Td+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGlBZ+RERERERERETGVPpGvphzDpmU/ZKbzY4Zi3q84X2hWDBjqcDT3LnpIo0vXqybsWNvei/NPfg6Fp+kucNmm8ZrlZoZm739HprbTk+Zsace+QrN7Xft/Wo07GMFAGvnz9J4KhqYsXyeD9cDtxwwY6+//TaaG6ZKNJ5JTdix7JDmpns9M9Y5c57mxmFE4yFZum2lUjS3OG2/5/n90zR3r0ulUqjU7GvApzJmrNO3xygA+H7fjPUTctstfs0PhnZ+v8/HYRjGZmw45LlD8roA0OnYtbvTbvL9iu39qkzZNQ4AKjX7upyozNDcfDZL41FM3rMLaW4AO16p5Gnu+go/1r1uy4zFMb+nONjvOY7scQsA1UqOxo8cnjdj3Q4f1z62j1etwmvzXpdKp1Aj43xIanyj1aDbfvrRR83Y8qlTNDdImB4W03aNzAb82vIDe4wH4HO9g/vs+zwATFXsa2Cz06W5tx69w4ydiTZpbn1j3YxFObtOAcBy254jAECnY88D6hvLNNeReUDPJbynzos0HmTteXecShgDWXu/OrDvCQAQkXsZAJTIfpVrvEamUvYFF3s+H9vrMukMFmYWzHguYx+bYo6f70LRvq5DMu8HgEzMn+Gqefv+ceyAfV8CgAny7Lh/jl+35RyfY1dL9r2+F9ivCwDZ2D6ejS0+/8iX7G1ninbdBoClVXt+AQCLG/Zc77kTvBYtrdh1rrHFX3c45PG77txnxsp5/p6jDpn7xPwce8/HZj5rv3aU8HznyHpJGPExwOg3fkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExtQNbefu4xj9rt3OrZizd8fleUu1TGC3NvMJbc8KZb7t7/2b32vG3v6d76G51Rm7leDyyWdoboq8JwCoN7fM2Orp52juhabdRu4zH/84zS0X7PZ0vT5vubcwz9szV0nr3lPnFmnugByvqf1Hae7tr3szjSOyWxlv1M/R1E7PbmO52eXn2Hl+ifa6dlvTVkKbQd+yr8U7eRfLPa9eb+Djn/gjMx5lPm/GNjd5y8rW1poZC/gpSWz3vrxsv3aU0PJ0anbOjE3OTNPcHGkrCQDtjboZe/4FXucaLbtmHLrlCM1NZexaVK3w93TLLYdp/OAhu63tLbcmtJTO2dd8JaG1aFyr0jhIe+Zhwr0ulbb/rSdF9hkA5o/O0Hi+atfIYUIbZNb5eWoq4XjscalUCmXSzj1N7omD9Tbd9trz9j3zUJnfi11CS/Ymmcv1EuYurmC3Oc45Ph9bXd6g8Yf/6jEzNl+p0Nz1TbuObXV5K/gW6TDeXWvQXCS0sE+TC6SQ4XW/N7DvKat1+/0CQBTwc1FM222jXcD/XTmgc3rerh1+SMPttn2uGg1+HienyeQn5udpr/MO8OS85QtFM5Yh9xYAyOTseK9J2mkDGA75/aNWse8R99zD71vs+slkeA1Mp3k8isk4Duz6CQC5rD3nKpf5HCJL7uU+5nO5TMJ1+/Sz9rNlu8OvS0T2/arf57nZFH/PQWDPP7zj120c2OOrkVD3mx1+HlntHgz4fTLs29se9Pk1w+g3fkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExpQWfkRERERERERExlQ66Qecc4cA/DqAeQAewIe997/onJsC8FsAjgI4DeCD3vtNti0Pj9gP7B+II3s/wpjuZ+iHdq7zNDefq9L4PW9+sxnLZTI09+lHHzFjmxdepLn9fo/Gm5sbZmzxxNM0t+ULZiwT8dctp1NmrJov0dzZyRqNX1xeMmPh0D7HANBptszY4qmzNBd4ikZbraYZy6f5+Apzc2ZsPeRjr1DI03ixYp/HQjpHc5udhhkL45Dm3gy7WYsazRYe+IsvmvGJg3eYMR/Z4wwAHvniX5ixIwcP0tyZ6WkaP3+OXB+kfgJAcWrCjA0CXl+Xzy3S+Hvu+wYzds/rX0tzO6TOBRl+izp19owZe/4FXl+feNKuzQAwUSubse/7G3+N5r7jtbebsazn/95ycN8hGh+k7PrrAkdzY2/XqiH4+AnSPJ6bsGtVIeDvOU7Z8wJ+h705dnVe5IA4ax8fH9nnNJvixzUztM/Z4eoUzQ0De5wBQLPbNWOpqn3tAECQtcdKd3mL5vbrHb5f6/a9ei3mx6vet7d99E2vp7lLq+v2djf5eyqX+byp12mbsWGGzxF6ffte3h3yuh8k1JM8OY/e8flaBPu1U2le94OQz7ni2N72ymqd5oakzKWz/HjcDLtai2JgMLTHS7NtXx9BpUj3s1u3r8thyMdKsVCh8VSQNWP19YR6krHH0lbLrnEAMIwmadyTay+T5mMp2ihTKQAAIABJREFUQ+pvJ+rTXHYrH3R5bjHHr72lpYtmrO95Leqn7POcTdvnEABSeX4/6nTsNx0OyLoDgFzWfu2tHh8DS+v0koIH2W/Px4Bz9nsqJJwn5nJ+4ycE8I+893cBeBuAv++cuwvAPwHwae/9cQCf3vlvEZHrRbVIREaBapGIjALVIhG5bIkLP977i977r+78uQngGQAHALwfwK/t/NivAfjA9dpJERHVIhEZBapFIjIKVItE5Epc0Xf8OOeOAngjgL8CMO+9f+l3vpaw/WuGIiLXnWqRiIwC1SIRGQWqRSKS5LIXfpxzZQAfA/Dfeu+/5stBvPce258tfaW8DznnHnLOPdTu8s/ZiYgk2Y1aNBgkfEZaRCTBbtSiTsL3SIiIJNmNWtQb8O/aEZG977IWfpxzGWwXlP/gvf/dnb9eds7t24nvA7DySrne+w977+/13t9bKvAvbxIRYXarFmWz/IuvRUSY3apFxbL9Bf0iIkl2qxbls6P4VfoispsSF36ccw7ARwA8473/l5eEPgHgR3b+/CMAfn/3d09EZJtqkYiMAtUiERkFqkUiciUupx/YOwD8MIAnnHOP7vzdTwH4OQC/7Zz7UQBnAHwweVMeIC0c49D+KFg6w1sFRqQH4wC8NfV8jbfk+5NPfNKMTc3zNuBzpDXvoMPbDGYy/LcSyiW7FXg6oRVribShX5jjLaW7Tbt9XSHF93l9dY3GhwP7PFby/F9GBy27zfYLjzxEcy8++zyN90Py6/gZfqwjci5KB3kbV5T4xyODnN0KO5/Qkn0S9vG887W38P3CVxPi18Wu1aLJqWl8/w/+bTOemztuxjpNu6U6ALzwxGNmbN8Cb9UdJLS9LuTta34Q84+M3H63/Z4m983R3M4Mr5Hv+85vNWPFCr9u26Sde5zQPTf09v2kF9rbBYCVlQ0aP3PqghkrFu3zAABL5+zWzqefeoHmBj2+3yeXXvEfbgEA9337vTT3yNH9ZmwY8XoR5BN+Yzdj126XUItA2pZmHW85fZPsWi2Kohh10uq437HvAaUBv/fMLtjne/2MPY4A4MTpMzS+OrTH6dQUbxUfkHt5O+btcaMhLwphx/4Yb69P+hwDCJ3d2nl1ic9d2i271bUf8vbjxRyf3w669rF2OT7nCnv28ciW+PzDR/za6/XtsRkH/D0PyHw/l+G1JptPmBsXy2asQGIAMCTnKun+fJPsWi0KoxBrm3a7+/3k2YC1egeAMCb1YprXi2YjYduhHe8ntPKOyTB99sQpmhsk3JuyKXu8HCb3YgAIyvYY77V5HYvIew4HfJ6YI/sMAPVN+7n1+fP8nnHL7D4zNlWp0dz0FJ9ztdv2xxQ3Q/6snc7aSyFNUnsBYDMhHnv7eLqEJZiMs+dNbXKfS5K48OO9/wIA6077nqt+ZRGRK6BaJCKjQLVIREaBapGIXImRXL4WEREREREREZFrp4UfEREREREREZExpYUfEREREREREZExpYUfEREREREREZExpYUfEREREREREZExpYUfEREREREREZExldjOfVd5hzi2ug4C2XTKjOXTMd92YG/Xp0o0NR4MaXxtbcmMtVbtGAAUhg37dWG/XwCYmpym8Yn9s2YsjPo09/wFe789PM0NAnvYDMKQ5qZchsZL+aIZCxOGQIr9gOPvKRps0XhAxm2js0lzB7muGavs5+epXajTeDMemLFem6/rTldvNWMzc3zs7XXOAbmsfXyef/ZJM9bY4te89/ZYGw7s8wUArVabxp2zx2E+x6+tYadpxrZW+fWxfHaRxv/oT/7IjG027dcFgK2Wfe1VqlWaW5ucMmOlao7mnjt3gcbnZg6YsXx1juZ+/g/t47HxwuM0N0q4H51YWjZj59r8WB+/87gZq1Xt2gsAtckajReKeTu3xMdmJm/fC4tFfh73vNgBXXJ8yC0idFm66TaZYlx0fP5xMeGG2xqQ+Dq/n6YyHTPWifnrenIvBoAumYN4H9HcbMY+nudX12huGNn77cxu29tWN/kcAqTu+4i/p0yhYMaqWT5+opBvm93rUmk+/yjAHvNBiudmyHkCAEfel08YX468duBu7CPTjTYYDrF4wb4vZjJ2zQgH9jwXAA4dWjBj7Q6fBzdadr0AgDAk4zDgda4T2nOyZ06cpLnphG1fWLxoxmamJmlurTZhxl544QTNZc9w3/vd30Bzc57PuSYnKmas0OBzl/W6/TwTs/sJ+NgDgEbLnr+0+3xe3SFjN8jy+UdvmFRP7JoRJ9SiTTI3nqnYdT2JfuNHRERERERERGRMaeFHRERERERERGRMaeFHRERERERERGRMaeFHRERERERERGRMaeFHRERERERERGRMaeFHRERERERERGRMaeFHRERERERERGRM2Q3mrwuHwOXMaD5n96X3COmWS4WiHavM0NzOsEfj05WsGUsn7Ndga9mMxYG9XQDoZGIan5+/xd72YEBz73j9QTP2xb/4NM0d+I4ZyzhHc7stOxcAqpWqGcum+XBNOft4tXr8HJ+6uEnj9bp9nvuuTXNnb7fXVw9M2GMeAAaej5HNNft4ZnsZmls6MG3Gup2I5u51cThEc33JjP/57/+hGVtcOke3HQy7Zuzxxxt8xxKunzAk9YaMfwB44JN/bsayGbsuA8A9b3wTjQ+yFTPW6PNr/uTZFTO2vv4Mf92e/Z4vLJ2muadO823f+8Y3m7F/8Pf/Ic198MtfMmPh1jrNbfT7NN6FN2MnH1qkuZ9/+KIZK6WHNDeTTdF4KmePoUqJ16KDR46asfd/3w/Q3L3OOYe0s4/P0Nvnu9XlY2WjYdebjQHPDTP8futDezz0uvx+6/r2/GToeR0LAj4OSzV7DpFKJYxhMsfwCf9M6sl5SnzdhHgQ2PeFIGG/YvIDQeLx4Ociiu15gif7nPTaQcKbcgn3STg7Pyb7DADsFkvvv2PAAwjJOF7f2jJj1WKebrtB5v7sugOAGHyctrv2tpOuDx/b87VKgb/uygaf2zz6xBkzViqs0tx+j92P+XWZzdv7/cwL9j4BwHyRPy+ze/nCAs9dP2PPuV2aX9Mrq/x4HTxoP89EMd92P7THfKfdpLlhwrYjNr6qZZo7iO39ag/4GGD+v/buLUbO+7zv+O+Z8+yZS3IpkqJEUaf4IEdyFdV2jCRwkbgxCiQBcpFeFL4I6qJogAZtL4wUKBygNy2aBL1y4MABnDaokyYpEhQpUCVxbRhuJUu2JOtgU5RMiZR4WpLLPczszunfix3FlMrn+e+sljsvp98PQGi1D/8zz/zf933e9/1zZh7e8QMAAAAAADChWPgBAAAAAACYUCz8AAAAAAAATCgWfgAAAAAAACYUCz8AAAAAAAATioUfAAAAAACACbWv7dxLJtUq/lpTK2hjW25Mh489KPutZFtBe2VJKlf9lmmSVK/5Lber1Tiv2tS8G5ufi8devOK3gpek1nG/JfvSiQfCsW9dXnZjH/qJnwzHrl952429fvqlcOzG+koYr5T9bTUftGmVJAtaHF54y89Zkt58w29TKUmlur+t5o5MhWMPL/p5W6bNvF2L95ED1/1D+PjSYjj27gV//znzst92cRJUqzUdPXLUjT948j43ljKtNCslP17OtKEtleO1+BS0d6xlaqSqfrvVY8eOh0N/5tOfDuOzU/4xMN84EI59+cXn3djpM6+FY+86ftKNbWb6L5eb8XH74unvu7GXT58Ox06d/IAbe/vteD4OLMTxpVrNf94Z/1wlSdcu+q1cr751Jhx7ZTk+H232/X2zm2l5emHFr2Of+HuZ1s13uEG/r/W1dTe+urrhxjbW42ubjQ3//JLriD23EJ9v603/mivHgh7LzYq/f0tStRY/b9QavZppUR+1le4P4roftXPfbpQdjQ3DKkc9qS0e3O/77ctz7cnj1yR1g/H9zGsuV/ztVMm0987l1Wj457p6Zh9IQbv3en33+/ydoFKu6MBBvyX3XHDP0sjM67VVvy12M3Mu7nb8bSJJnZ4fr1Tj64Ba3a83nX7UUl26fC1u9b3Z8597cXYhHHv3KX87dLvxcbu65t9nnT0ft0WvHfbbtUtSKfnPPTMV125b8q9t5prx+WZ9ZTWMn33jrBu7/6F7wrGd5J8MO/34Hi1zOxC2g78nuDeUpGbDn8+tdid+4gDv+AEAAAAAAJhQLPwAAAAAAABMKBZ+AAAAAAAAJhQLPwAAAAAAABOKhR8AAAAAAIAJxcIPAAAAAADAhGLhBwAAAAAAYEJVcn/BzE5I+gNJRyQlSV9KKf1HM/uCpH8s6crwr/5GSukvwyermI4c9teaulevurF2fxDmubHhx1KpH46tVOJpmJs76MZq1Wo4tr2x6saa1cz0d+L4M9/6lhs79fClcOz58xfdWKlk4dipuv+ay+V6OLbZnA7jG+ttN9Zu+zFJ6vU6bmymGef1icceCuON2Tn/ecu9cGy/23Jj7XOb4djSWiOML03NurHHHvpQPHbhiBt79sIPw7HjsJe1qNfr6dqVa278Y3/3E27sEz/902Ge9XrZjVXK8Vp7qRTHB8mvg2X5zytJ3Y5fB9sdfx+VpKvn4/3h2mbXjy378yxJr595zY29fdmvU5I0s3TMD9bjY8dqU2G809tyY09+/Zvh2Hvvf8SNnVg8Ho5tlOK6P1X1a9nW5lo49vXVl9zYTFDjJKmf4jp38fq6Gzt06GQ4ttX19+u/+frT4dhx2OtatBxc+0TH7eamf86TpE7Hj1cb8bVLtVEL49H5uJStc0GtimKSUoqvT3p9fz8tVeK8mlP+sWWZ2qyU3FB/EF+/5pj5r9kUz0ek1Yrrfr+fuXYOrmFT5joyms/o9UpSCuZ6+AjB4Hhko9F0Y/V6fB05DntZi/qDgdaCfWIw8M/zx44shXnWmv75trUV17HpqfjcZBV/P7VyvMGrNf/YtF58zLfa8fFRa/rXIDMHZ8Kx3ZJfx3qV+FzcWPDnelCJ6/7aelwTHjx1r5/XRf8aQJJ6G/4548Z6fJ344AMPhvHz5151Y91evJ0sWApZX43nY5B5/8zMlL8tZqbic+zGhv/c5eDeLye78COpJ+lfppS+Y2azkp41syeHsd9JKf2HXT87AOwctQhAEVCLABQBtQjAjmUXflJKFyRdGP68ZmavSIr/yRIA9hi1CEARUIsAFAG1CMAoRvqOHzM7KekxSU8Nf/VrZvaCmf2+mR3Y49wA4JaoRQCKgFoEoAioRQBydrzwY2Yzkv5U0q+nlFYlfVHS/ZIe1fZq82854z5nZs+Y2TOrrfhznACQsxe1aG09/i4UAMjZi1q0teV/lxQA7MRe1KJe5rtUAdz5drTwY2ZVbReUP0wp/ZkkpZQupZT6KaWBpN+T9MStxqaUvpRSejyl9Phc5ouMACCyV7Vodmb3X4wGAHtVi4r4hbEA7hx7VYtyzScA3PmyR7ltf73+lyW9klL67Zt+f/Smv/ZLkl7c+/QAYBu1CEARUIsAFAG1CMAodtLV6ycl/SNJ3zOz54a/+w1J/9DMHtV2+8Czkv5J7oFqNdM9J/x3/cyb3/ruzLm4pdqlK37Lvk4//he1mZl4GjZaN9xYfxC3rysHa2vXrvgtXCVpbT1u2bfZ9fMqJz8mSbMz/sd9L12M2+qd3/BbkA8yrVaPHD4Yxi1oF3l95Xo4tj7tb+eF+fgdHrXMv3RsBS11lWmPuLHlP3ZnPR47PYjzeuDEXW7s2F3xXJ87f8mNXb0SH29jsme1qFQyTQete6+u+vv4d194NnzspSX/2DqydCgc2+36+78kXb++4gc3/ZwlqRIcW8fvC9qiSzpxID5+3jp9wY1trMcfZVk64u/DUwcXwrHlht/mtdWO5+Po0XvC+MW3z7ux5atxfT16bMONWaYV8fpWvA+o4u+33UHctrTenPZjmRbKnatXwrhKfi07cvxk/NhBO99s5+bx2LNaNEhJ3W7wMfjknwMqmXNP9GaietNvWy0p7IgtSRZcNpXLcUv2QbBN+5lriFyL8XLQDr5ci/MqVf25rmXmOmoxnss5357clznkVQrapi8sxPU1dz7a6vj7bd/i1xS1bM/NR68XXxv3ekHe/Ux9Dfq957bjmOzddVG5pKlpv/10v+dv763MvlKp+sdetRp/GiRXT6L3MASnJUlSpbr7j7dtBddUkmQVP++p+fg1r635X0fQzNTuK1f8e7hKJb6WO9CM7zmmFvxrrpmG365dko4cnndjyym+v5uaijfk0pJ/v7O2uhqOjW7vSpnz4Nx8XENn5/xttXojuJ6XtLy87MZSaSZOLLCTrl7f1K0vAf5y188KACOiFgEoAmoRgCKgFgEYBR/oBAAAAAAAmFAs/AAAAAAAAEwoFn4AAAAAAAAmFAs/AAAAAAAAE4qFHwAAAAAAgAnFwg8AAAAAAMCEyrZz30vlimnuQNWNt6+03NiBpXL84NNTbmj50lY4dLPTCeOV2pwbywzVoNt3Y91+nNeN9vUwPt2su7HN1mY4tr257MY6Qc6S1A/iKcXbaX3V38aSNDfXDGLz4dh223/s5avxXM7MTIdxK/lrpNZL4dhaxX9N9UY4VLVaPJ8nHzjpxtqtOK9vfONlN/bC6cvh2DtdyaR6deDGtzZX3Ni3vvXX4WOnrn/szU35+4Ikdbu9ML7ZbruxSmYd/96TJ9zYhz/2wXDs/fccC+Mr5867sYvX/VojSbWgjt1/8K5w7JUr627skYc/HI790CMPh/Gv/uc/cGMV1cKx3Q1/H+h04tqcenH9VcPfR8p1fy4l6eR9p9zY5XM/iJ+3FNei5rT/3B/4wEPh2M2Wvx1PHF2K87rDVSoVHTx40I2X5F8z9ftxje/2/BrXt3js5qZfayTJyrfqID2MWVyLBgM/r07fj0lSeZC5FozGluOxgxRcrwVzKUl2y47aO2OZoYOBv616mXoxCPaRciWej14vPh91g3h3EI8tBdvCMhOSUrzvRtu5pHhsv+/PZ7TfToKSmRpN/9xWMj/W7sT3M/XguG3W4/OpKd6XatVgPw7qlCTNzS+6sc3VG+HYTiVz71j395d25jqgXPbnpBtPtTptfx+/ENz7SdLi8eNhvHvBvzdoZs4pjVl/Ox2ej8/zy1ffDOOL8/59ukr+OVSS1nv+hD58NL72HWTueVutrh/b8GOStDi/4MYytwoh3vEDAAAAAAAwoVj4AQAAAAAAmFAs/AAAAAAAAEwoFn4AAAAAAAAmFAs/AAAAAAAAE4qFHwAAAAAAgAnFwg8AAAAAAMCEquznk5mZKg3/KRtzNTe2OBOvUVXaW26s2hyEY1evZ6ah7z93s7EUD636z93fWgnH1qbivKoVf77K5alw7Fby8+p0O+HYlMyNWQqHKnU2w3g/CFcr1fjBa3U3tHL9eji03emG8fmFOTdWKcX7ZinYTi31wrGXltfC+PV1f/zaxo1w7F/9r+/7z9sKh97xBoOBWu3gRQbb9NM//w/ix+5suLFyN97eg35cq1K57D92sJ9JUmParwkXV9rh2LWV02H8Wtt/XdZohGN/8Nzrbuzq/74Sjj1138Nu7CceeDAc22nHtagZ1JPUjetFK3jsUjmu6wO/vEqS2gN/H6n04/3r3rtPubHN9avh2A/OTYfxp5/9rht7+40fhGPbG/4xk1px7b7Tlctlzc3555dBP9ghUnzu2QrOa6ut9XBsperXGkkqB/F+vx+OVRCuZs6nvWD/l6RB8NyDlMnL/Oe24Lpn+OBxPBwaj43OCynz77eD6FqvHV/rdTN1bqAg71I8X9ErHmS2cQpHS1PBOadWiffrkvl5Vyr7esu078xMtbK/P01N+dcQuWO+HBz05XK8r/T78X7Y6/n7cQpejyStrfn7Q3t1NRwbvSZJagT3u53MtWA3uKZq3fDvdyWpVmm6sdnFhXBsdB8lSd2Wf61YrsXHZa3uX6Omanxszc75r0mS6sFxvbB4OBybVq+5MSvF23hzzb92kaR2yx/fCI4naft4dGXOGRHe8QMAAAAAADChWPgBAAAAAACYUCz8AAAAAAAATCgWfgAAAAAAACYUCz8AAAAAAAATioUfAAAAAACACbWvvQkHA9P6etCSuzzjhmam49a71abf2my6HrcTnp+PW0eur/rt69ZXL8Vjg1Zu3c24Tdxs7WAYb1T9uextxe3+KhV/za+WWQ6s1v22eRa0Q5WkqZl4lysF4V6mVXGt6Q+eW4jb5l27FrdNXwtaos4txtupFbSafPVs3EL5+987F8aPLPptgI/cHb9mlfzXdGh+Nhz6w6tx+++iK5VM0zN+a8n5oFPi7OGHwsfeCo69RmatvWZxS/bU9Fta1qfisYNNv33z2lqmbemUv59J0tL9fovQ+6eWw7Gv/vA1P2hx693qlN969K0Lb4ZjDx46sOt4px238NzauuHGNjbic9lWps12d6vlxiqN+Jg/csxva/rGhfhcdunNYDtJ2lz3X/NrLz0Xjj140M8rHVgMx04CC+qCmV+MOt34PL+55dfpbjdu5V0qx8deJWi7noL245LU6fnn8q1efF1kmTbhFuQVteqWpFIwdtCL2+dG0UwjeMWzJaUg736u9bn58VIlzqxaDq7XM1LmRaegHXG/n2lvn+tkHFyvlTLXqNHYXjfeN+90JTNNB+28K8GenHsXQaPh34etr8fnvHKmFtXqfs7N6ficGI7NvKj2jZUwfmTpHje2mWkFvzDtz1f1cOY6MSgJXcXnjNx9VnNm2s8rcw0aFcJupjYfOuyvD0hSbeDf/5UrcR2rB2sEKcXzNTUV59WM5iSzX7fb/vk7iuXwjh8AAAAAAIAJxcIPAAAAAADAhGLhBwAAAAAAYEKx8AMAAAAAADChWPgBAAAAAACYUCz8AAAAAAAATKjswo+ZNczsaTN73sxeMrPfHP7+PjN7yszOmNkfmWX6EAPA+0AtAlAE1CIARUE9ArBTfuP7H9mS9KmU0rqZVSV908z+h6R/Iel3UkpfNbPflfSrkr4YPVCnI51/I3iilYYbmz3cC5NsNLtubH4mHKrFxXga1jdabmxlxY9J0vWrfp29fjXOqzwoh/FBSm6s3+/HDz7w47nVQCuZGytX4rls9+NHT8Fmrg78bSxJvdY1N9Zvx9upX6mG8ZV1f3wnM9XXVttu7OyZeCdYuboRxjsb/pPfNX9XOPYD9x53Y0HKkqRvv74c/4XbY89q0WCwqdba6eAv+Ptp1eKCcunSDTf26stnw7GNSjOM1+YX3NihpQPh2GOH5t1YpRQflwfnD4bx/sCPbbavh2OXlubc2PFji+HYCxcvurHTp18Jx57s3BfGt7a23Njamr+NJanVuuTGVm+sxs/bWg/j/Y5/cJbr0+HYl1485MY6W51w7NLSkTB+/CMf9scejsceOuzXqkbmNY3JntUiJWkw8A+grWC7dLvxNut0Nv1YZnt3uvE11yD5OZv8awRJKpf9a5tGvR6OLVXi66J+z887BddMUrwdrBQ/b/SaS5n6WgvmI2dz09/GktQL5qOcySvaTlI8n1H9lKRWy69jZvH+02j49wpS/Lp6nTivkvljG4143xyjPalHJqkabNNS39+XauX42v/9HB/RcSlJtap//R7t/9uP7ccbmbzmZ+NrweBWSY3aVJxXcGMxNROP7Qa1fTNzL7TVi29opmr+dq7W4uNjo+U/d2PWvw6UpHYn3o7t4DVXU3x/Vw5qe6kc15rMLa1abX/fXVmJr42jfbdW2/0abvYdP2nbO1ei1eGfJOlTkv5k+PuvSPrFXWcBABnUIgBFQC0CUBTUIwA7taPv+DGzspk9J+mypCclvSZpJaW/fX/GeUn+2wcAYA9QiwAUAbUIQFFQjwDsxI4WflJK/ZTSo5LulvSEpB/b6ROY2efM7Bkze+bGevyWVACI7FUtWluL3+4KAJG9qkXtduZztQCQsdt6dHMt2sp8nAbAnW+krl4ppRVJX5P0cUkLZvbOh/3ulvSWM+ZLKaXHU0qPz8/En5UDgJ14v7Vodjb+jDQA7MT7rUXNZvy9XgCwU6PWo5trUT34/hYAk2EnXb0Om9nC8OempJ+V9Iq2C8svD//aZyX9+e1KEgCoRQCKgFoEoCioRwB2aifLu0clfcXMytpeKPrjlNJ/N7OXJX3VzP6tpO9K+vJtzBMAqEUAioBaBKAoqEcAdiS78JNSekHSY7f4/eva/hzpjiWrqF/128l2a4+7sa1BpgVjz28v3ZiPW0MuHI4/gnag5H/udbEVtxlcuea/jXtlOW6V2d6IN0+/F7RzS5n2iD0/7812/F1MURu5cqbV6tpmPF/t4HugqiluPztbmnVjg1LcQrnbjee6Pu23uGxU4xaGCzU/71Py23NL0iM/HrcyfvgjP+7GTj7wQDj2iY/533Nz/u24pbS+/Xocvw32shZpkDQIWh2XgjdDVrrxPj5X9ffxZ//P18OxFy/5dUySLNjXnnji74RjP/lxv77euBG3J3/hO0+F8Y2gpfDpN8+FY18/e9aNtYP2n5KUkl/bG3OHw7Grq2thfO26vy02VuM2nNEZp1KOz0fzmY8hHrvPb0N/4ODRcOzSMb9t+rHHHgnHLs7FtShqSZ1rCy0L4plz2Tjs6XVRSup2u248atmea1WsoDVzpZK5/Mu2L/fltnfUvjlFPZAldTOvOXpd/X7cqtjkz1e5HLcELgXzlWtPnmszn4J21rm2vtFcv59W8JJUDdpov599ILedcnnVgrbrU/W4vkZbKrcdx2Wv6lHJTM2av02j7ZIG8TaLjp+5ubiVd66de7Rdci2zU9DOfT7zMdyZzEfj0sA/BtpbmVo08GvCoBtfJ85O+23mM6VGcVbSRse/F6924xrZbvtje6X4u+6Wb8TXa+tX/Xu8hQV/3UGSrm74+0ijGV9/pBTvA9ev+dewa5nr2+hj4O/nI+LFu6ICAAAAAADAnmDhBwAAAAAAYEKx8AMAAAAAADChWPjBS9wCAAAGSklEQVQBAAAAAACYUCz8AAAAAAAATCgWfgAAAAAAACYUCz8AAAAAAAATylJK+/dkZlckvXHTrw5JWt63BHaOvHauiDlJ5DWqUfO6N6V0+HYlc7tRi9438hpNEfMqYk4StWhStst+Ia/RkNfOUYuKt00k8hoVeY1mEvJya9G+Lvz8P09u9kxK6fGxJeAgr50rYk4SeY2qqHntl6K+fvIaDXntXBFzkoqb134p6usnr9GQ12iKmFcRc9pPRX395DUa8hrNpOfFR70AAAAAAAAmFAs/AAAAAAAAE2rcCz9fGvPze8hr54qYk0ReoypqXvulqK+fvEZDXjtXxJyk4ua1X4r6+slrNOQ1miLmVcSc9lNRXz95jYa8RjPReY31O34AAAAAAABw+4z7HT8AAAAAAAC4Tcay8GNmf9/MfmBmZ8zs8+PI4VbM7KyZfc/MnjOzZ8aYx++b2WUze/Gm3y2a2ZNm9urwvwcKktcXzOyt4Zw9Z2afGUNeJ8zsa2b2spm9ZGb/fPj7sc5ZkNdY58zMGmb2tJk9P8zrN4e/v8/Mnhoel39kZrX9zGscqEXZPKhFo+VFLRotL2rRTahH2TwKV4+oRXuWF7WoQKhF2TwKV4uCvMZ9bFGLRsvr9tailNK+/pFUlvSapFOSapKel/TB/c7Dye2spEMFyOOnJH1U0os3/e7fS/r88OfPS/p3BcnrC5L+1Zjn66ikjw5/npV0WtIHxz1nQV5jnTNJJmlm+HNV0lOSPibpjyX9yvD3vyvpn45zu+7DPFCL8nlQi0bLi1o0Wl7Uoh/NBfUon0fh6hG1aM/yohYV5A+1aEd5FK4WBXmN+9iiFo2W122tReN4x88Tks6klF5PKXUkfVXSL4whj8JKKX1D0rX3/PoXJH1l+PNXJP3iviYlN6+xSyldSCl9Z/jzmqRXJB3XmOcsyGus0rb14f9Wh3+SpE9J+pPh78eyj+0zalEGtWg01KLRUIvehXqUUcR6RC3as7zGilr0LtSijCLWIqmY9YhaNJrbXYvGsfBzXNK5m/7/vAow0UNJ0v80s2fN7HPjTuY9jqSULgx/vijpyDiTeY9fM7MXhm8x3Pe3Nt7MzE5KekzbK6SFmbP35CWNec7MrGxmz0m6LOlJbf/rzkpKqTf8K0U6Lm8XatHuFOa4ugVqUQa1qLCoR7tTmGPrPahFGdSiwqIW7U5hjq1bKEQ9ohbtOJ/bVov4cud3+2RK6aOSfl7SPzOznxp3QreStt/nVZR2bF+UdL+kRyVdkPRb40rEzGYk/amkX08prd4cG+ec3SKvsc9ZSqmfUnpU0t3a/tedH9vvHBCiFo1u7MfVO6hFO0ctuiNQj0Yz9uPqHdSinaMW3RGoRaMb+7ElUYtGcTtr0TgWft6SdOKm/797+LuxSym9NfzvZUn/TduTXRSXzOyoJA3/e3nM+UiSUkqXhjvoQNLvaUxzZmZVbR+4f5hS+rPhr8c+Z7fKqyhzNsxlRdLXJH1c0oKZVYahwhyXtxG1aHfGflzdSlGOK2rR7vx/Xosk6tFujf3Yeq+iHFfUot2hFlGLdmnsx9atFOHYohbtzu2oReNY+Pm2pAeH305dk/Qrkv5iDHm8i5lNm9nsOz9L+jlJL8aj9tVfSPrs8OfPSvrzMebyt945aId+SWOYMzMzSV+W9EpK6bdvCo11zry8xj1nZnbYzBaGPzcl/ay2P9v6NUm/PPxrhdnHbiNq0e5Qi/wcqEWj5UUt+hHq0e4Urh6N+7ga5kAtGi0vatGPUIt2p3C1SCrEsUUtGi2v21uL0ni+sfoz2v727Nck/etx5HCLnE5p+5vrn5f00jjzkvRftP32sq62P8f3q5IOSvprSa9K+itJiwXJ6z9J+p6kF7R9EB8dQ16f1PZbBF+Q9Nzwz2fGPWdBXmOdM0kfkfTd4fO/KOnfDH9/StLTks5I+q+S6vu9Lcew71CL4lyoRaPlRS0aLS9q0bvng3oU51K4ekQt2rO8qEUF+kMtyuZSuFoU5DXuY4taNFpet7UW2fDBAAAAAAAAMGH4cmcAAAAAAIAJxcIPAAAAAADAhGLhBwAAAAAAYEKx8AMAAAAAADChWPgBAAAAAACYUCz8AAAAAAAATCgWfgAAAAAAACYUCz8AAAAAAAAT6v8C2O6irIM2DZYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1440x1440 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mn4gaHr4Fi-Y"
      },
      "source": [
        "# Normalize pixel values\n",
        "x_train, x_test = x_train / 255.0, x_test / 255.0\n",
        "\n",
        "mean_image = np.mean(x_train, axis=0)\n",
        "\n",
        "x_train = x_train - mean_image\n",
        "x_test = x_test - mean_image"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gk7EIKobFtwW",
        "outputId": "5fd1720c-63b1-4265-d22f-7095b6feb473"
      },
      "source": [
        "# like one hot encoding\n",
        "y_train = tf.keras.utils.to_categorical(y_train, num_classes=K)\n",
        "y_test = tf.keras.utils.to_categorical(y_test, num_classes=K)\n",
        "\n",
        "print('Shape after one hot encoding :',y_train.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape after one hot encoding : (50000, 10)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YG0eOyDxyeyl",
        "outputId": "d91903a0-988e-4854-bc0c-1c712b8f69d0"
      },
      "source": [
        "x_train = np.reshape(x_train,(Ntr,Din))   #flattening\n",
        "x_test = np.reshape(x_test,(Nte,Din))\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "\n",
        "print('Shape of x_train after reshaping :', x_train.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Shape of x_train after reshaping : (50000, 3072)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4eIVdwBYykAX",
        "outputId": "cc01e939-1691-4ce4-fe08-b75557471670"
      },
      "source": [
        "# initializing weight and bias vectors\n",
        "\n",
        "std=1e-5\n",
        "w1 = std*np.random.randn(Din, K)\n",
        "b1 = np.zeros(K)\n",
        "batch_size = Ntr\n",
        "\n",
        "print(\"w1:\", w1.shape)\n",
        "print(\"b1:\", b1.shape)\n",
        "print('Batch size :', Ntr)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "w1: (3072, 10)\n",
            "b1: (10,)\n",
            "Batch size : 50000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lk8b7haXzW2y",
        "outputId": "d761b5b4-6ff6-419a-e236-a836a1a6073d"
      },
      "source": [
        "# Bias Trick \n",
        "# Simplifies two parameters W,b into a one matrix. This way our model only has to learn one weight matrix instead of two separate matrices.\n",
        "# We combine W and b into a single matrix and also add an extra column that always hold 1.\n",
        "\n",
        "W = np.append(w1, b1.reshape(1, K), axis=0)\n",
        "print('New weight matrix shape: ',W.shape)\n",
        "\n",
        "x_train_new = np.append(x_train, np.ones((x_train.shape[0], 1)), axis=1)\n",
        "x_test_new = np.append(x_test, np.ones((x_test.shape[0], 1)), axis=1)\n",
        "\n",
        "print('Rearranged x_train: ', x_train_new.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "New weight matrix shape:  (3073, 10)\n",
            "Rearranged x_train:  (50000, 3073)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A79iwo5LLbK3"
      },
      "source": [
        "# **Gradient Descent**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZD_BmG5mzC6a",
        "outputId": "038ad6bd-7a3f-4b68-a05e-9895932b0b20"
      },
      "source": [
        "iterations = 300   # epoches\n",
        "lr = 1.4e-2\n",
        "lr_decay= 0.999\n",
        "reg = 5e-6\n",
        "lr_history = []\n",
        "loss_history = []  \n",
        "train_acc_history = [] \n",
        "val_acc_history = [] \n",
        "val_loss_history = []\n",
        "\n",
        "m = x_train.shape[0] \n",
        "\n",
        "print(\"Running gradient descent...\")\n",
        "\n",
        "for t in range(1, iterations+1):    \n",
        "  # Forward Pass\n",
        "  scores = x_train_new.dot(W)\n",
        "  error = scores - y_train\n",
        "\n",
        "  loss_func = (1/(2*m))*np.sum((error)**2) + (reg/(2*m))*np.sum(W**2)\n",
        "  loss_history.append(loss_func)\n",
        "    \n",
        "  # Backward Pass\n",
        "  W = W - lr*((1/m)*(x_train_new.T.dot(error))  + (1/m)*reg*W )\n",
        "    \n",
        "  # get accuracies\n",
        "  # to get the array of classes with highest score (column with highest value along a row), we use argmax\n",
        "  pred_class = np.argmax(scores, axis=1)\n",
        "  true_class = np.argmax(y_train, axis=1)    # we use the argmax function here because now y_train is a 5000x10 array after one hot encoding\n",
        "\n",
        "  # Train set\n",
        "  correct_pred = np.sum([pred_class == true_class])\n",
        "  train_acc = (correct_pred/m)*100\n",
        "  train_acc_history.append(train_acc)\n",
        "\n",
        "  # Test set\n",
        "  pred_class_test = np.argmax(x_test_new.dot(W), axis=1)\n",
        "  true_class_test = np.argmax(y_test, axis=1)\n",
        "\n",
        "  val_loss = 1/(x_test_new.shape[0]*2)*np.sum((x_test_new.dot(W)-y_test)**2)\n",
        "  val_loss_history.append(val_loss)\n",
        "\n",
        "  correct_pred_test = np.sum([pred_class_test == true_class_test])\n",
        "  val_acc = (correct_pred_test/x_test_new.shape[0])*100\n",
        "  val_acc_history.append(val_acc)\n",
        "    \n",
        "  # Training details\n",
        "  if t == 1 or t%10 == 0:\n",
        "    print(\"| Epoch {:04} | Loss {:.4f} | Training Acc: {:.4f} | Validation Acc: {:.4f} | Learning Rate: {:.4f} |\".format(t,loss_func,train_acc,val_acc,lr))\n",
        "    \n",
        "  # Learning rate decay\n",
        "  lr_history.append(lr)\n",
        "  lr = lr*lr_decay\n",
        "    \n",
        "print(\"Gradient Descent - Done\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Running gradient descent...\n",
            "| Epoch 0001 | Loss 0.5000 | Training Acc: 9.5060 | Validation Acc: 24.8700 | Learning Rate: 0.0140 |\n",
            "| Epoch 0010 | Loss 0.4556 | Training Acc: 32.2520 | Validation Acc: 32.7700 | Learning Rate: 0.0139 |\n",
            "| Epoch 0020 | Loss 0.4399 | Training Acc: 35.1500 | Validation Acc: 35.0400 | Learning Rate: 0.0137 |\n",
            "| Epoch 0030 | Loss 0.4297 | Training Acc: 36.4820 | Validation Acc: 36.3900 | Learning Rate: 0.0136 |\n",
            "| Epoch 0040 | Loss 0.4223 | Training Acc: 37.1480 | Validation Acc: 37.3200 | Learning Rate: 0.0135 |\n",
            "| Epoch 0050 | Loss 0.4168 | Training Acc: 37.6600 | Validation Acc: 37.7800 | Learning Rate: 0.0133 |\n",
            "| Epoch 0060 | Loss 0.4126 | Training Acc: 38.1000 | Validation Acc: 38.1700 | Learning Rate: 0.0132 |\n",
            "| Epoch 0070 | Loss 0.4094 | Training Acc: 38.4560 | Validation Acc: 38.4400 | Learning Rate: 0.0131 |\n",
            "| Epoch 0080 | Loss 0.4069 | Training Acc: 38.7640 | Validation Acc: 38.7200 | Learning Rate: 0.0129 |\n",
            "| Epoch 0090 | Loss 0.4049 | Training Acc: 38.9600 | Validation Acc: 38.9000 | Learning Rate: 0.0128 |\n",
            "| Epoch 0100 | Loss 0.4033 | Training Acc: 39.1880 | Validation Acc: 39.0100 | Learning Rate: 0.0127 |\n",
            "| Epoch 0110 | Loss 0.4020 | Training Acc: 39.4020 | Validation Acc: 39.0600 | Learning Rate: 0.0126 |\n",
            "| Epoch 0120 | Loss 0.4010 | Training Acc: 39.5540 | Validation Acc: 39.2200 | Learning Rate: 0.0124 |\n",
            "| Epoch 0130 | Loss 0.4001 | Training Acc: 39.6480 | Validation Acc: 39.2300 | Learning Rate: 0.0123 |\n",
            "| Epoch 0140 | Loss 0.3994 | Training Acc: 39.7740 | Validation Acc: 39.3100 | Learning Rate: 0.0122 |\n",
            "| Epoch 0150 | Loss 0.3987 | Training Acc: 39.9320 | Validation Acc: 39.3800 | Learning Rate: 0.0121 |\n",
            "| Epoch 0160 | Loss 0.3982 | Training Acc: 40.0220 | Validation Acc: 39.4100 | Learning Rate: 0.0119 |\n",
            "| Epoch 0170 | Loss 0.3977 | Training Acc: 40.1420 | Validation Acc: 39.4600 | Learning Rate: 0.0118 |\n",
            "| Epoch 0180 | Loss 0.3973 | Training Acc: 40.1960 | Validation Acc: 39.5600 | Learning Rate: 0.0117 |\n",
            "| Epoch 0190 | Loss 0.3970 | Training Acc: 40.2820 | Validation Acc: 39.6100 | Learning Rate: 0.0116 |\n",
            "| Epoch 0200 | Loss 0.3967 | Training Acc: 40.3840 | Validation Acc: 39.6100 | Learning Rate: 0.0115 |\n",
            "| Epoch 0210 | Loss 0.3964 | Training Acc: 40.4620 | Validation Acc: 39.6100 | Learning Rate: 0.0114 |\n",
            "| Epoch 0220 | Loss 0.3961 | Training Acc: 40.5060 | Validation Acc: 39.6200 | Learning Rate: 0.0112 |\n",
            "| Epoch 0230 | Loss 0.3959 | Training Acc: 40.5600 | Validation Acc: 39.5700 | Learning Rate: 0.0111 |\n",
            "| Epoch 0240 | Loss 0.3957 | Training Acc: 40.6380 | Validation Acc: 39.6300 | Learning Rate: 0.0110 |\n",
            "| Epoch 0250 | Loss 0.3954 | Training Acc: 40.7160 | Validation Acc: 39.5600 | Learning Rate: 0.0109 |\n",
            "| Epoch 0260 | Loss 0.3953 | Training Acc: 40.7840 | Validation Acc: 39.6100 | Learning Rate: 0.0108 |\n",
            "| Epoch 0270 | Loss 0.3951 | Training Acc: 40.8560 | Validation Acc: 39.6900 | Learning Rate: 0.0107 |\n",
            "| Epoch 0280 | Loss 0.3949 | Training Acc: 40.9060 | Validation Acc: 39.7000 | Learning Rate: 0.0106 |\n",
            "| Epoch 0290 | Loss 0.3947 | Training Acc: 40.9800 | Validation Acc: 39.7600 | Learning Rate: 0.0105 |\n",
            "| Epoch 0300 | Loss 0.3946 | Training Acc: 41.0420 | Validation Acc: 39.8300 | Learning Rate: 0.0104 |\n",
            "Gradient Descent - Done\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2rgtDW1jW-Vd"
      },
      "source": [
        "## **Plotting W as images**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 525
        },
        "id": "QUZVReacW9KO",
        "outputId": "4435b086-f919-48b5-bf18-f5ac2baba63f"
      },
      "source": [
        "W_orig = W[:-1]\n",
        "W_img =  W_orig - np.min(W_orig)   # removing negative weights\n",
        "images = ((W_img/np.max(W_img))*255).astype('uint8')\n",
        "\n",
        "classes = list(labels.values())\n",
        "\n",
        "fig, ax  = plt.subplots(2,5, figsize=(25,10))\n",
        "for i in range(K):\n",
        "    image = images[:,i].reshape(32,32,3)\n",
        "    ax.ravel()[i].imshow(image)\n",
        "    ax.ravel()[i].set_title(\"Class : {}\".format(classes[i]))\n",
        "    #plt.tight_layout()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x720 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FkgK6Az2bGFX"
      },
      "source": [
        "## **Plotting learning rate and accuracy curves**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "id": "lfMYuU9JbAgW",
        "outputId": "a3743487-e389-474b-e0b6-cfca55a78347"
      },
      "source": [
        "fig, ax  = plt.subplots(1,5, figsize=(20,5))\n",
        "items = {\"Learning Rate\":lr_history, \"Training Loss\":loss_history, \"Validation Loss\":val_loss_history, \"Training Accuracy\":train_acc_history, \n",
        "         \"Validation Accuracy\": val_acc_history}\n",
        "i=1\n",
        "for key in items.keys():\n",
        "    plt.subplot(1,5,i);\n",
        "    plt.plot(items[key], color ='b')\n",
        "    plt.title(key)\n",
        "    i+=1"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x360 with 5 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2zKB9TK93NPr"
      },
      "source": [
        "# PART 2"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tW065sES3PkF"
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.cifar10.load_data()\n",
        "\n",
        "K = len(np.unique(y_train)) \n",
        "\n",
        "Ntr = x_train.shape[0]\n",
        "Nte = x_test.shape[0]\n",
        "Din = x_train.shape[1]*x_train.shape[1]*3"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LR2nFk7bA6_d"
      },
      "source": [
        "#x_train, x_test = x_train / 255.0, x_test / 255.0\n",
        "\n",
        "mean_image = np.mean(x_train, axis=0)\n",
        "\n",
        "x_train = x_train - mean_image\n",
        "x_test = x_test - mean_image\n",
        "\n",
        "y_train = tf.keras.utils.to_categorical(y_train, num_classes=K)\n",
        "y_test = tf.keras.utils.to_categorical(y_test, num_classes=K)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i7ARhBMQBDv7"
      },
      "source": [
        "x_train = np.reshape(x_train,(Ntr,Din))   #flattening\n",
        "x_test = np.reshape(x_test,(Nte,Din))\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DDeOZA1fBF81",
        "outputId": "70cbdec8-23f0-4ce6-87ca-0dcef415a702"
      },
      "source": [
        "print(\"x train shape : \", x_train.shape)\n",
        "print(\"y train shape : \", y_train.shape)\n",
        "print(\"Num of train samples : \", Ntr)\n",
        "print(\"Num of test samples : \", Nte)\n",
        "print(\"Num of classes : \", K)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x train shape :  (50000, 3072)\n",
            "y train shape :  (50000, 10)\n",
            "Num of train samples :  50000\n",
            "Num of test samples :  10000\n",
            "Num of classes :  10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "phwgSmTcB-Fe"
      },
      "source": [
        "H = 200   # Hidden units\n",
        "std = 1e-6\n",
        "\n",
        "w1 = std*np.random.randn(Din, H)\n",
        "w2 = std*np.random.randn(H, K)\n",
        "\n",
        "b1 = np.zeros(H)\n",
        "b2 = np.zeros(K)\n",
        "\n",
        "batch_size = Ntr\n",
        "iterations = 300\n",
        "\n",
        "lr = 1.4e-2\n",
        "lr_decay= 0.999\n",
        "reg = 5e-6\n",
        "lr_history = []\n",
        "loss_history = []  \n",
        "train_acc_history = [] \n",
        "val_acc_history = [] \n",
        "val_loss_history = []"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YVho8RlqDa7F",
        "outputId": "b8b18d43-e643-47e4-bf76-9b1ef0efd856"
      },
      "source": [
        "print(\"Running Gradient Descent ...\")\n",
        "for t in range(1,iterations+1):\n",
        "  batch_indices = np.random.choice(Ntr, batch_size)\n",
        "  x = x_train[batch_indices]\n",
        "  y = y_train[batch_indices]\n",
        "\n",
        "  h = 1/(1+np.exp(-x.dot(w1)-b1))\n",
        "  y_pred = h.dot(w2) + b2\n",
        "\n",
        "  #loss = (1/(2*Ntr))*np.sum((y_pred - y)**2) + (reg/(2*Ntr))*(np.sum(w1**2) + np.sum(w2**2))\n",
        "  loss = (1/(2*Ntr))*np.sum((y_pred - y)**2) + reg*(np.sum(w1**2) + np.sum(w2**2))\n",
        "  loss_history.append(loss)\n",
        "\n",
        "  # Back propagation\n",
        "  dPred =  (1/Ntr)*(y_pred - y)\n",
        "  dw2 = h.T.dot(dPred) + reg*w2\n",
        "  db2 = dPred.sum(axis=0)\n",
        "\n",
        "  dh = dPred.dot(w2.T)\n",
        "\n",
        "  dw1 = x.T.dot(dh*h*(1-h)) + reg*w1\n",
        "  db1 = (dh*h*(1-h)).sum(axis =0)\n",
        "\n",
        "  # get accuracies\n",
        "  # to get the array of classes with highest score (column with highest value along a row), we use argmax\n",
        "  pred_class = np.argmax(y_pred, axis=1)\n",
        "  true_class = np.argmax(y, axis=1)    # we use the argmax function here because now y_train is a 5000x10 array after one hot encoding\n",
        "\n",
        "  # Train set\n",
        "  correct_pred = np.sum([pred_class == true_class])\n",
        "  train_acc = (correct_pred/Ntr)*100\n",
        "  train_acc_history.append(train_acc)\n",
        "\n",
        "  # Test set\n",
        "  test_h = 1/(1+np.exp(-x_test.dot(w1)-b1))\n",
        "  test_pred = test_h.dot(w2) + b2\n",
        "  pred_class_test = np.argmax(test_pred, axis=1)\n",
        "  true_class_test = np.argmax(y_test, axis=1)\n",
        "\n",
        "  val_loss = 1/(x_test.shape[0]*2)*np.sum((test_pred-y_test)**2)\n",
        "  val_loss_history.append(val_loss)\n",
        "\n",
        "  correct_pred_test = np.sum([pred_class_test == true_class_test])\n",
        "  val_acc = (correct_pred_test/x_test.shape[0])*100\n",
        "  val_acc_history.append(val_acc)\n",
        "\n",
        "  # Training details\n",
        "  if t == 1 or t%10 == 0:\n",
        "    print(\"| Epoch {:04} | Loss {:.4f} | Training Acc: {:.4f} | Validation Acc: {:.4f} | Learning Rate: {:.4f} |\".format(t,loss,train_acc,val_acc,lr))\n",
        "\n",
        "  w1 = w1 - dw1*lr\n",
        "  w2 = w2 - dw2*lr\n",
        "  b1 = b1 - db1*lr\n",
        "  b2 = b2 - db2*lr\n",
        "\n",
        "  lr_history.append(lr)\n",
        "  lr = lr*lr_decay\n",
        "\n",
        "print(\"Gradient Descent - Done\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Running Gradient Descent ...\n",
            "| Epoch 0001 | Loss 0.5000 | Training Acc: 10.3200 | Validation Acc: 10.2600 | Learning Rate: 0.0140 |\n",
            "| Epoch 0010 | Loss 0.4524 | Training Acc: 10.1200 | Validation Acc: 10.0000 | Learning Rate: 0.0139 |\n",
            "| Epoch 0020 | Loss 0.4303 | Training Acc: 18.9000 | Validation Acc: 18.7100 | Learning Rate: 0.0137 |\n",
            "| Epoch 0030 | Loss 0.4231 | Training Acc: 26.0300 | Validation Acc: 25.6600 | Learning Rate: 0.0136 |\n",
            "| Epoch 0040 | Loss 0.4167 | Training Acc: 29.6340 | Validation Acc: 28.9700 | Learning Rate: 0.0135 |\n",
            "| Epoch 0050 | Loss 0.4142 | Training Acc: 30.5980 | Validation Acc: 30.2800 | Learning Rate: 0.0133 |\n",
            "| Epoch 0060 | Loss 0.4103 | Training Acc: 32.8840 | Validation Acc: 32.4100 | Learning Rate: 0.0132 |\n",
            "| Epoch 0070 | Loss 0.4088 | Training Acc: 32.9180 | Validation Acc: 32.6100 | Learning Rate: 0.0131 |\n",
            "| Epoch 0080 | Loss 0.4058 | Training Acc: 34.6360 | Validation Acc: 34.3100 | Learning Rate: 0.0129 |\n",
            "| Epoch 0090 | Loss 0.4030 | Training Acc: 35.4040 | Validation Acc: 34.7300 | Learning Rate: 0.0128 |\n",
            "| Epoch 0100 | Loss 0.3999 | Training Acc: 36.9040 | Validation Acc: 35.9400 | Learning Rate: 0.0127 |\n",
            "| Epoch 0110 | Loss 0.4007 | Training Acc: 36.3800 | Validation Acc: 35.5200 | Learning Rate: 0.0126 |\n",
            "| Epoch 0120 | Loss 0.3955 | Training Acc: 38.5000 | Validation Acc: 37.6900 | Learning Rate: 0.0124 |\n",
            "| Epoch 0130 | Loss 0.3935 | Training Acc: 38.8740 | Validation Acc: 37.6900 | Learning Rate: 0.0123 |\n",
            "| Epoch 0140 | Loss 0.3916 | Training Acc: 39.2840 | Validation Acc: 38.2200 | Learning Rate: 0.0122 |\n",
            "| Epoch 0150 | Loss 0.3898 | Training Acc: 40.2220 | Validation Acc: 38.4700 | Learning Rate: 0.0121 |\n",
            "| Epoch 0160 | Loss 0.3917 | Training Acc: 39.7600 | Validation Acc: 38.2500 | Learning Rate: 0.0119 |\n",
            "| Epoch 0170 | Loss 0.3857 | Training Acc: 41.6740 | Validation Acc: 39.8800 | Learning Rate: 0.0118 |\n",
            "| Epoch 0180 | Loss 0.3846 | Training Acc: 42.0560 | Validation Acc: 40.3700 | Learning Rate: 0.0117 |\n",
            "| Epoch 0190 | Loss 0.3903 | Training Acc: 40.6320 | Validation Acc: 38.8000 | Learning Rate: 0.0116 |\n",
            "| Epoch 0200 | Loss 0.3826 | Training Acc: 42.6020 | Validation Acc: 40.2000 | Learning Rate: 0.0115 |\n",
            "| Epoch 0210 | Loss 0.3816 | Training Acc: 42.5920 | Validation Acc: 40.2600 | Learning Rate: 0.0114 |\n",
            "| Epoch 0220 | Loss 0.3813 | Training Acc: 42.9220 | Validation Acc: 40.2800 | Learning Rate: 0.0112 |\n",
            "| Epoch 0230 | Loss 0.3803 | Training Acc: 43.5380 | Validation Acc: 40.7700 | Learning Rate: 0.0111 |\n",
            "| Epoch 0240 | Loss 0.3790 | Training Acc: 43.9940 | Validation Acc: 41.0200 | Learning Rate: 0.0110 |\n",
            "| Epoch 0250 | Loss 0.3778 | Training Acc: 43.9360 | Validation Acc: 40.8500 | Learning Rate: 0.0109 |\n",
            "| Epoch 0260 | Loss 0.3765 | Training Acc: 44.1100 | Validation Acc: 41.2200 | Learning Rate: 0.0108 |\n",
            "| Epoch 0270 | Loss 0.3749 | Training Acc: 44.8100 | Validation Acc: 41.5000 | Learning Rate: 0.0107 |\n",
            "| Epoch 0280 | Loss 0.3750 | Training Acc: 45.0980 | Validation Acc: 41.4600 | Learning Rate: 0.0106 |\n",
            "| Epoch 0290 | Loss 0.3730 | Training Acc: 45.7460 | Validation Acc: 41.9800 | Learning Rate: 0.0105 |\n",
            "| Epoch 0300 | Loss 0.3728 | Training Acc: 45.3200 | Validation Acc: 41.7000 | Learning Rate: 0.0104 |\n",
            "Gradient Descent - Done\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4d67fD39pDPP"
      },
      "source": [
        "# **Plotting Learning rate and Accuracy curves**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 627
        },
        "id": "ufCSImqKu7sb",
        "outputId": "6e80fc40-fc35-4c4e-de80-9d9896947638"
      },
      "source": [
        "fig, ax  = plt.subplots(1,4, figsize=(25,10))\n",
        "items = { \"Training Loss\":loss_history, \"Validation Loss\":val_loss_history, \"Training Accuracy\":train_acc_history, \n",
        "         \"Validation Accuracy\": val_acc_history}\n",
        "i=1\n",
        "for key in items.keys():\n",
        "    plt.subplot(1,4,i);\n",
        "    plt.plot(items[key], color ='g')\n",
        "    plt.title(key)\n",
        "    i+=1"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x720 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7wx97U2r4isl"
      },
      "source": [
        "# PART 3"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ncCims9rMArK",
        "outputId": "449e5042-8786-4e11-ca93-619ed98c1f64"
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.cifar10.load_data()\n",
        "\n",
        "K = len(np.unique(y_train)) \n",
        "\n",
        "Ntr = x_train.shape[0]\n",
        "Nte = x_test.shape[0]\n",
        "Din = x_train.shape[1]*x_train.shape[1]*3\n",
        "\n",
        "\n",
        "#x_train, x_test = x_train / 255.0, x_test / 255.0\n",
        "\n",
        "mean_image = np.mean(x_train, axis=0)\n",
        "\n",
        "x_train = x_train - mean_image\n",
        "x_test = x_test - mean_image\n",
        "\n",
        "y_train = tf.keras.utils.to_categorical(y_train, num_classes=K)\n",
        "y_test = tf.keras.utils.to_categorical(y_test, num_classes=K)\n",
        "\n",
        "\n",
        "x_train = np.reshape(x_train,(Ntr,Din))   \n",
        "x_test = np.reshape(x_test,(Nte,Din))\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "\n",
        "\n",
        "print(\"x train shape : \", x_train.shape)\n",
        "print(\"y train shape : \", y_train.shape)\n",
        "print(\"Num of train samples : \", Ntr)\n",
        "print(\"Num of test samples : \", Nte)\n",
        "print(\"Num of classes : \", K)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170500096/170498071 [==============================] - 2s 0us/step\n",
            "x train shape :  (50000, 3072)\n",
            "y train shape :  (50000, 10)\n",
            "Num of train samples :  50000\n",
            "Num of test samples :  10000\n",
            "Num of classes :  10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0lzvMC0VD7oc"
      },
      "source": [
        "H = 200   # Hidden units\n",
        "std = 1e-6\n",
        "\n",
        "w1 = std*np.random.randn(Din, H)\n",
        "w2 = std*np.random.randn(H, K)\n",
        "\n",
        "b1 = np.zeros(H)\n",
        "b2 = np.zeros(K)\n",
        "\n",
        "batch_size = 500\n",
        "iterations = 300\n",
        "\n",
        "lr = 1.4e-2\n",
        "lr_decay= 0.999\n",
        "reg = 5e-6\n",
        "lr_history = []\n",
        "loss_history = []  \n",
        "train_acc_history = [] \n",
        "val_acc_history = [] \n",
        "val_loss_history = []\n",
        "batch_loss_history = []"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n7R4N8Q_Efqg",
        "outputId": "6053e07e-aa10-4137-eefc-069af2a512a7"
      },
      "source": [
        "for t in range(1,iterations+1):\n",
        "\n",
        "  batch_indices = np.random.choice(Ntr, Ntr)\n",
        "  #x = x_train[batch_indices]\n",
        "  #y = y_train[batch_indices]\n",
        "  x = x_train\n",
        "  y = y_train\n",
        "\n",
        "  tot_batch_loss = 0\n",
        "  for i in range(0, Ntr, batch_size):\n",
        "    x_tr = x[i:i+batch_size,:]\n",
        "    y_tr = y[i:i+batch_size,:]\n",
        "\n",
        "    h = 1/(1+np.exp(-x_tr.dot(w1)-b1))\n",
        "    y_pred = h.dot(w2) + b2\n",
        "\n",
        "    batch_loss = (1/(2*Ntr))*np.sum((y_pred - y_tr)**2) + reg*(np.sum(w1**2) + np.sum(w2**2))\n",
        "    batch_loss_history.append(batch_loss)\n",
        "\n",
        "    # Back propagation\n",
        "    dPred =  (1/Ntr)*(y_pred - y_tr)\n",
        "    dw2 = h.T.dot(dPred) + reg*w2\n",
        "    db2 = dPred.sum(axis=0)\n",
        "\n",
        "    dh = dPred.dot(w2.T)\n",
        "\n",
        "    dw1 = x_tr.T.dot(dh*h*(1-h)) + reg*w1\n",
        "    db1 = (dh*h*(1-h)).sum(axis =0)\n",
        "\n",
        "    w1 = w1 - dw1*lr\n",
        "    w2 = w2 - dw2*lr\n",
        "    b1 = b1 - db1*lr\n",
        "    b2 = b2 - db2*lr\n",
        "\n",
        "    tot_batch_loss += batch_loss\n",
        "  \n",
        "  loss_history.append(tot_batch_loss)\n",
        "  \n",
        "  # get accuracies\n",
        "\n",
        "  h = 1/(1+np.exp(-x.dot(w1)-b1))\n",
        "  y_pred = h.dot(w2) + b2\n",
        "\n",
        "  pred_class = np.argmax(y_pred, axis=1)\n",
        "  true_class = np.argmax(y, axis=1)    \n",
        "\n",
        "  # Train set\n",
        "  correct_pred = np.sum([pred_class == true_class])\n",
        "  train_acc = (correct_pred/Ntr)*100\n",
        "  train_acc_history.append(train_acc)\n",
        "\n",
        "  # Test set\n",
        "  test_h = 1/(1+np.exp(-x_test.dot(w1)-b1))\n",
        "  test_pred = test_h.dot(w2) + b2\n",
        "  pred_class_test = np.argmax(test_pred, axis=1)\n",
        "  true_class_test = np.argmax(y_test, axis=1)\n",
        "\n",
        "  val_loss = 1/(x_test.shape[0]*2)*np.sum((test_pred-y_test)**2)\n",
        "  val_loss_history.append(val_loss)\n",
        "\n",
        "  correct_pred_test = np.sum([pred_class_test == true_class_test])\n",
        "  val_acc = (correct_pred_test/x_test.shape[0])*100\n",
        "  val_acc_history.append(val_acc)\n",
        "\n",
        "  # Training details\n",
        "  if t == 1 or t%10 == 0:\n",
        "    print(\"| Epoch {:04} | Loss {:.4f} | Training Acc: {:.4f} | Validation Acc: {:.4f} | Learning Rate: {:.4f} |\".format(t,tot_batch_loss,train_acc,val_acc,lr))\n",
        "  \n",
        "  lr_history.append(lr)\n",
        "  lr = lr*lr_decay\n",
        "\n",
        "print(\"--- Done ---\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "| Epoch 0001 | Loss 0.4768 | Training Acc: 10.2760 | Validation Acc: 10.0000 | Learning Rate: 0.0140 |\n",
            "| Epoch 0010 | Loss 0.4433 | Training Acc: 16.4360 | Validation Acc: 16.3500 | Learning Rate: 0.0139 |\n",
            "| Epoch 0020 | Loss 0.4288 | Training Acc: 18.7560 | Validation Acc: 18.8000 | Learning Rate: 0.0137 |\n",
            "| Epoch 0030 | Loss 0.4223 | Training Acc: 25.8700 | Validation Acc: 25.5600 | Learning Rate: 0.0136 |\n",
            "| Epoch 0040 | Loss 0.4167 | Training Acc: 29.1400 | Validation Acc: 29.3200 | Learning Rate: 0.0135 |\n",
            "| Epoch 0050 | Loss 0.4125 | Training Acc: 32.0440 | Validation Acc: 31.8500 | Learning Rate: 0.0133 |\n",
            "| Epoch 0060 | Loss 0.4085 | Training Acc: 33.8840 | Validation Acc: 33.6000 | Learning Rate: 0.0132 |\n",
            "| Epoch 0070 | Loss 0.4055 | Training Acc: 35.3820 | Validation Acc: 34.9200 | Learning Rate: 0.0131 |\n",
            "| Epoch 0080 | Loss 0.4020 | Training Acc: 36.5800 | Validation Acc: 36.0300 | Learning Rate: 0.0129 |\n",
            "| Epoch 0090 | Loss 0.3992 | Training Acc: 37.6980 | Validation Acc: 37.1900 | Learning Rate: 0.0128 |\n",
            "| Epoch 0100 | Loss 0.3967 | Training Acc: 38.3700 | Validation Acc: 38.1600 | Learning Rate: 0.0127 |\n",
            "| Epoch 0110 | Loss 0.3942 | Training Acc: 38.9320 | Validation Acc: 38.9400 | Learning Rate: 0.0126 |\n",
            "| Epoch 0120 | Loss 0.3921 | Training Acc: 39.8540 | Validation Acc: 39.6800 | Learning Rate: 0.0124 |\n",
            "| Epoch 0130 | Loss 0.3894 | Training Acc: 40.6280 | Validation Acc: 39.9200 | Learning Rate: 0.0123 |\n",
            "| Epoch 0140 | Loss 0.3875 | Training Acc: 41.3760 | Validation Acc: 40.6100 | Learning Rate: 0.0122 |\n",
            "| Epoch 0150 | Loss 0.3867 | Training Acc: 41.4580 | Validation Acc: 41.0300 | Learning Rate: 0.0121 |\n",
            "| Epoch 0160 | Loss 0.3846 | Training Acc: 42.0200 | Validation Acc: 41.1900 | Learning Rate: 0.0119 |\n",
            "| Epoch 0170 | Loss 0.3839 | Training Acc: 42.2700 | Validation Acc: 41.6300 | Learning Rate: 0.0118 |\n",
            "| Epoch 0180 | Loss 0.3815 | Training Acc: 42.9740 | Validation Acc: 41.7300 | Learning Rate: 0.0117 |\n",
            "| Epoch 0190 | Loss 0.3800 | Training Acc: 43.6080 | Validation Acc: 42.1000 | Learning Rate: 0.0116 |\n",
            "| Epoch 0200 | Loss 0.3793 | Training Acc: 43.9280 | Validation Acc: 42.2900 | Learning Rate: 0.0115 |\n",
            "| Epoch 0210 | Loss 0.3790 | Training Acc: 43.7480 | Validation Acc: 42.5800 | Learning Rate: 0.0114 |\n",
            "| Epoch 0220 | Loss 0.3776 | Training Acc: 44.1440 | Validation Acc: 42.8400 | Learning Rate: 0.0112 |\n",
            "| Epoch 0230 | Loss 0.3759 | Training Acc: 44.7860 | Validation Acc: 42.9000 | Learning Rate: 0.0111 |\n",
            "| Epoch 0240 | Loss 0.3754 | Training Acc: 44.9240 | Validation Acc: 43.2500 | Learning Rate: 0.0110 |\n",
            "| Epoch 0250 | Loss 0.3744 | Training Acc: 45.0260 | Validation Acc: 43.3800 | Learning Rate: 0.0109 |\n",
            "| Epoch 0260 | Loss 0.3731 | Training Acc: 45.5280 | Validation Acc: 43.4700 | Learning Rate: 0.0108 |\n",
            "| Epoch 0270 | Loss 0.3723 | Training Acc: 45.6360 | Validation Acc: 43.5500 | Learning Rate: 0.0107 |\n",
            "| Epoch 0280 | Loss 0.3710 | Training Acc: 46.0540 | Validation Acc: 43.7300 | Learning Rate: 0.0106 |\n",
            "| Epoch 0290 | Loss 0.3704 | Training Acc: 46.3900 | Validation Acc: 43.7600 | Learning Rate: 0.0105 |\n",
            "| Epoch 0300 | Loss 0.3697 | Training Acc: 46.5540 | Validation Acc: 43.8400 | Learning Rate: 0.0104 |\n",
            "--- Done ---\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 607
        },
        "id": "W9-4aw5oO0t2",
        "outputId": "1cd79c8a-fce2-412c-ce92-1e86ca4ba73d"
      },
      "source": [
        "figure = plt.figure(figsize=(10,10))\n",
        "plt.plot(batch_loss_history)\n",
        "plt.title(\"Batch loss\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hO7RUWGOV9Ig",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 434
        },
        "outputId": "24bf75f6-551f-4aa6-9faa-3b19399138af"
      },
      "source": [
        "fig, ax  = plt.subplots(1,4, figsize=(25,8))\n",
        "items = {\"Training Loss\":loss_history, \"Validation Loss\":val_loss_history, \"Training Accuracy\":train_acc_history, \n",
        "         \"Validation Accuracy\": val_acc_history}\n",
        "i=1\n",
        "for key in items.keys():\n",
        "    plt.subplot(1,4,i);\n",
        "    plt.plot(items[key], color ='m')\n",
        "    plt.title(key)\n",
        "    i+=1"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x576 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P0dG1o946uKN"
      },
      "source": [
        "# **PART 4**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fch4TzFhrbIC"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Flatten, Conv2D, MaxPooling2D"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5iu_rUbnNQ5a",
        "outputId": "bcd291c3-019c-4bfb-f7ec-3b14d82643be"
      },
      "source": [
        "(x_train, y_train), (x_test, y_test) = keras.datasets.cifar10.load_data()\n",
        "\n",
        "K = len(np.unique(y_train)) \n",
        "\n",
        "x_train, x_test = x_train / 255.0, x_test / 255.0\n",
        "mean_image = np.mean(x_train, axis=0)\n",
        "\n",
        "x_train = x_train - mean_image\n",
        "x_test = x_test - mean_image\n",
        "\n",
        "y_train = tf.keras.utils.to_categorical(y_train, num_classes=K)\n",
        "y_test = tf.keras.utils.to_categorical(y_test, num_classes=K)\n",
        "\n",
        "print(\"x train shape : \", x_train.shape)\n",
        "print(\"y train shape : \", y_train.shape)\n",
        "print(\"Num of classes : \", K)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x train shape :  (50000, 32, 32, 3)\n",
            "y train shape :  (50000, 10)\n",
            "Num of classes :  10\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E-tzk7PmQoR6"
      },
      "source": [
        "batch_size = 50\n",
        "num_classes = K\n",
        "epochs = 10"
      ],
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SVZesKcLROho",
        "outputId": "acb25088-20e9-4346-d607-ead2274d307d"
      },
      "source": [
        "model = Sequential()\n",
        "\n",
        "# for C32 and C64, CONV => RELU => POOL \n",
        "# C32\n",
        "model.add(Conv2D(32, (3, 3), activation = 'relu', input_shape = (32, 32, 3), name = 'C32'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "# C64\n",
        "model.add(Conv2D(64, (3, 3), activation = 'relu', name = 'C64_1'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "model.add(Conv2D(64, (3, 3), activation = 'relu', name = 'C64_2'))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "# for F64,  FLATTERN => DENSE ==> RELU\n",
        "\n",
        "# F64\n",
        "model.add(Flatten())\n",
        "model.add(Dense(64, activation = 'relu', name = 'F64'))\n",
        "\n",
        "# F10\n",
        "model.add(Dense(10, activation = 'sigmoid', name  = 'F10'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "C32 (Conv2D)                 (None, 30, 30, 32)        896       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_6 (MaxPooling2 (None, 15, 15, 32)        0         \n",
            "_________________________________________________________________\n",
            "C64_1 (Conv2D)               (None, 13, 13, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_7 (MaxPooling2 (None, 6, 6, 64)          0         \n",
            "_________________________________________________________________\n",
            "C64_2 (Conv2D)               (None, 4, 4, 64)          36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_8 (MaxPooling2 (None, 2, 2, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten_3 (Flatten)          (None, 256)               0         \n",
            "_________________________________________________________________\n",
            "F64 (Dense)                  (None, 64)                16448     \n",
            "_________________________________________________________________\n",
            "F10 (Dense)                  (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 73,418\n",
            "Trainable params: 73,418\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CaigohIGTnIM"
      },
      "source": [
        "opt = keras.optimizers.SGD(lr = 1.4e-2, momentum = 0.9)\n",
        "\n",
        "model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['accuracy'])"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "73sXM9nYU0jP",
        "outputId": "73fc625a-e900-41b2-b42f-e467b96dc75c"
      },
      "source": [
        " history = model.fit(x_train, y_train,\n",
        "              batch_size=batch_size,\n",
        "              epochs=epochs,\n",
        "              validation_data=(x_test, y_test),\n",
        "              shuffle=True)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "1000/1000 [==============================] - 68s 67ms/step - loss: 1.8758 - accuracy: 0.3054 - val_loss: 1.2819 - val_accuracy: 0.5310\n",
            "Epoch 2/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 1.2543 - accuracy: 0.5512 - val_loss: 1.1894 - val_accuracy: 0.5824\n",
            "Epoch 3/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 1.0484 - accuracy: 0.6288 - val_loss: 1.0188 - val_accuracy: 0.6454\n",
            "Epoch 4/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 0.9304 - accuracy: 0.6701 - val_loss: 0.9269 - val_accuracy: 0.6731\n",
            "Epoch 5/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 0.8324 - accuracy: 0.7068 - val_loss: 1.0074 - val_accuracy: 0.6569\n",
            "Epoch 6/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 0.7524 - accuracy: 0.7359 - val_loss: 0.9238 - val_accuracy: 0.6893\n",
            "Epoch 7/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 0.7029 - accuracy: 0.7529 - val_loss: 0.8905 - val_accuracy: 0.6984\n",
            "Epoch 8/10\n",
            "1000/1000 [==============================] - 67s 67ms/step - loss: 0.6500 - accuracy: 0.7725 - val_loss: 0.9130 - val_accuracy: 0.7028\n",
            "Epoch 9/10\n",
            "1000/1000 [==============================] - 74s 74ms/step - loss: 0.5976 - accuracy: 0.7901 - val_loss: 0.8694 - val_accuracy: 0.7158\n",
            "Epoch 10/10\n",
            "1000/1000 [==============================] - 69s 69ms/step - loss: 0.5675 - accuracy: 0.7986 - val_loss: 0.9164 - val_accuracy: 0.7015\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gCqeibLFYMbQ",
        "outputId": "9366b317-f73e-4896-8013-730750f3e033"
      },
      "source": [
        "history.history.keys()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 350
        },
        "id": "ar84E5zlU_i9",
        "outputId": "82d008e0-b5da-4c28-fc36-0720210c0370"
      },
      "source": [
        "fig, axs = plt.subplots(1,2,figsize=(15,5)) \n",
        "\n",
        "# Accuracy\n",
        "axs[0].plot(history.history['accuracy']) \n",
        "axs[0].plot(history.history['val_accuracy']) \n",
        "\n",
        "axs[0].set_title('Model Accuracy')\n",
        "axs[0].set_ylabel('Accuracy') \n",
        "axs[0].set_xlabel('Epoch')\n",
        "axs[0].legend(['train', 'validate'], loc='upper left')\n",
        "\n",
        "# Loss\n",
        "axs[1].plot(history.history['loss']) \n",
        "axs[1].plot(history.history['val_loss']) \n",
        "axs[1].set_title('Model Loss')\n",
        "axs[1].set_ylabel('Loss') \n",
        "axs[1].set_xlabel('Epoch')\n",
        "axs[1].legend(['train', 'validate'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XQAiLdg5Yffs",
        "outputId": "c7cd3a52-e0a3-4215-a431-a34bd1516fd0"
      },
      "source": [
        "scores = model.evaluate(x_test, y_test, verbose=1)\n",
        "\n",
        "print('Test loss:', scores[0])\n",
        "print('Test accuracy:', scores[1])\n",
        "\n",
        "pred = model.predict(x_test)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 4s 13ms/step - loss: 0.9164 - accuracy: 0.7015\n",
            "Test loss: 0.9163562655448914\n",
            "Test accuracy: 0.7014999985694885\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O_Axs2ykZapn"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}