fyp2_initial models.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "fyp experiment.ipynb",
      "provenance": [],
      "mount_file_id": "1dVqBVcU-euI0QdcXjGwYETv2PMm8Bhps",
      "authorship_tag": "ABX9TyOBm/+vfJAGX2E2f/bjnsfc",
      "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/nasiksami/Deep-Learning/blob/main/fyp2_initial%20models.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qFe0tBpvhGp3",
        "outputId": "e9d153bb-485e-4181-8974-1aef8926b8b4"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qZiPbFm1iH_r"
      },
      "source": [
        "dir = '/content/drive/My Drive/dataset_resized/'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d4L8Sqzoitu1"
      },
      "source": [
        "#! unzip dataset-resized -d datasets"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Il72Uhgwi0V9"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "3WPb6IHYjMZ9",
        "outputId": "2c2103f8-e8e8-4506-e308-9104fa82473d"
      },
      "source": [
        "tf.__version__"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'2.5.0-dev20201120'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "smbBWxuXjOLJ"
      },
      "source": [
        "model=keras.applications.ResNet50(weights=\"imagenet\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o52KMMlQjQsn"
      },
      "source": [
        "from keras.preprocessing.image import ImageDataGenerator,load_img,img_to_array"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8Fto9yHOjUUZ"
      },
      "source": [
        "def plot_image(url):\n",
        "  image=keras.preprocessing.image.load_img(url,target_size=(224,224))\n",
        "  img=img_to_array(image)\n",
        "  img=img/255.0\n",
        "  plt.imshow(img)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 72
        },
        "id": "ZF-Xo-MA8fhD",
        "outputId": "e27e1ee9-7d53-4916-ebbb-288d3d02c105"
      },
      "source": [
        "from google.colab import files\n",
        "uploaded=files.upload()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-54562ea1-9bd4-49fd-8022-735bd6fa2dc7\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-54562ea1-9bd4-49fd-8022-735bd6fa2dc7\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving ewaste (9).jpeg to ewaste (9).jpeg\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 252
        },
        "id": "AWvUwel6jWt3",
        "outputId": "c451d45c-09ee-43b3-e6d1-17e473d7653b"
      },
      "source": [
        "plot_image(\"/content/drive/My Drive/dataset_resized/training/metal/metal2.jpg\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAOsAAADrCAYAAACICmHVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOy9yY8kS37n9zEzX2KP3LP2t/V7PWySGlKjgYDRTYAOugjQAv0/OugiDARIN0GAdBAggSONRhBFiAJnNEM2muSou9nDbvby+u21ZFVm5Rab72amg4dHeHh6REZmZS3vVf0SkRHubmZubmZf+61mLqy1vKN39I7efJKvuwLv6B29o/XoHVjf0Tv6ltA7sL6jd/QtoXdgfUfv6FtC78D6jt7Rt4TegfUdvaNvCTlXSbyzs2Pff//9l1SVV02Fy0qUfi9PVdBC6hvxeq1XyPU8bBczLZZz2fUr3u21ugGvfu/1q1uf0FoLQiDqUq9Z+Hws5b+ePDng7OysWiRwRbC+//77/OQn/99VsrxSWj5YLFxo0nXygal0lBACa22ex85LXVXGsmv5aXtpumXXZvVYkad6vThXnC9+CyFW3qdc5/yzOt0qWkxnqWuDKQ7WpsU62mn+eb/XPd+ytrlQNhYwF9IaYxBCzMqepV3SPsZcLKNan//0P/nPlz7jWyIGr9PrL8YVik4TVxhhi0lfPVeqgrZ8/nXVqaBy21yHYef5r4D2VWWx2L8L4KxUTtjFey4bF9dp2Stx1utRMbst426rud6ro/Xq8LJEvasU+yJc5+p5qmAWS9JdPF41qNe7dzHY5+dXSQDrlrnu9eJ+l4m115GI1q1XmV4BWEXlu9rpq4B8xTutOYLzBrooIs2vlcrME8yvTbMJIdZGWH35eUGFWF1fx/pnKk4VM/YqwKxTp3K+i+Lk+nptOX21XqvrtP5gf5GJZ53xUZ4MZpODtVhxcYSuKq+uX67WJhfpNYjBguXAfdX1WEblRqykKx9eSfStM1XVkzEGrfVM97pQ0iswbC0H6kWx+Sbr8WLl3qzUU/SBEPKNkP1eAWf9NtJ6XXO1Drxed9fNAy+qz01LYaWVs+Z3kW9dwL4oZ63Ls3pivFlIVY1vV8t8taTr3OE1g/VNmK/WoKUNL2oTLRNtL7tWXJdSIuW6Qs+qUbGo7xXn5vnmv6tVquPqc+v1y+u361mUb778Iq3Wet6GtYmW5L1CndZtzWuA9dtqQF50ESyjC7No2dhRNa5Mz+b5VCmLXWEAyt0AF40ocxDXGVGW6ZhFzcpdnqtZdvYN5d8XHouCyxZumTnXrQLzerrwqvM3K07XtdF17pWnc1136m6Ztu9C/tU2i7wFdaUuq3Xny+r3Tgz+DlOda2bZueq1VWWtut+q43XzXS+tqLWSC1Gv96+mHGpSytkkei1RuIZeRC55B9Zaermi3so7X9EtU1/GRatu+bvu3Lruh6u4Ka4L1nU5TzVdVZef+1qvzr2v63a6tNyac+uOtm+rTPuS6YZm0Wt08g1N4LP7r8tZy5/LyrpJell+65tqx6sGulzrHmume8dZr0lz3W45XaWTX8zJvzxQQAiBMWYp16sD4VWCD67jN7xOgEL12qq2XTXhVPMupr3I46p2g3WecZZmRfdfZ5J6K8F6IfTrhTjg5cagq5R/VePDHFSLaS/zkS4Tg8uxz3VGsGp862V0ub5XV8+5kexl0vzZWYgjLuqVG9xWW+6XlVlHUsrads3DiS9/1rcSrDdBN6Fb1pd7lQG6zAK6WkddZWBaVUZdHdfhrJeVUQXsOhLJjeqRM8Puy50clldgvVu/A+s16eJMnNOLWA5vMgig7NBfR1+tu/8yEfmqYF1Wv9LRwrWrhY3WU52ou1xqqS/vZeuqV6V3BqZr0Cp99U3r4Kty2XKayznii9VpevTC5b0ovWFdtpTecdYXIsuqwVY3KFeN9XUNGHX3WQSXqQHifP3lYr2rv6noalU/ZR2I68G3mEyXULF68cYLTwgCyutPLXYW0LKgG4qi7ab1rsbDFO2xLBxyad/b6brWyye7qzzrO7BeQvXiFFQH28XAp2VAXhTJqoaddTlz1RB0kVtarDWlMosoJkOVm9YZlOoMSZe5d+rSzyucgS0EuSJ+7+YEu8UosOqWAVPwIC70iJ1dA1lnJRYsgm7BBlXTDsXYuASo15mY34F1CV1uQFpXr3p1hqh1DESX6a3r5rmkdtSwKRCW2VqzmZ54M+1z0ap99Sq+6fQOrEvo5pzqN1NOHa3mrPXW3rr01XyweguSZYBdlAoqHLgA6nTbk2lB5a8Xoov3rk+3UKc17vuqNOp17vMOrC9IVc6wivNcjTvV512HCqBprWe+veK81hpjzAWf3zpW3quIwOX01lqELIN/cRmLlMVChEV2d5l1vFyfhVVK5WJKv+t17eX3W/as6/TxVWkd7807sF6B6oIdVnHOOn33JqypRTmXDabLdMxVQK07vmrdFyeDqburJAbPJ7qy4en6NLMDLJyc1WZ2MP+V67B1u368aD9drpMumYhWDKh3YH0DaB29reovnee9yB1XXQcWuO1iPS4H73Wlgnz1y1wMFrVrbV+MFsXu6lVRe+lq08NNKrqLvHRW9xVt8Q6stVQVj+rAdE1xpxaVxZrJNWq2Qt+snl8m7hbXiu9y7PA6nPgqtHjfRWvoOuVeK0BCvCzb0cszQKzTtu/AWkuLxor6AXO9pVdQN1AvHwSrNla7zGhUvlbsK1RcL8Ba+GALENftVHFdDji30C571ur5m+BgVyjjZVuR1q7K6oq8A+uldJ3Qt6sNtjKYrhocvmjdLAHWmNlxAcpF4FqUNWAMGA3GkGVZLp46CksO4DTNAHAcB2sNWIlSzlqB/eW6CeEsjEUz1WHz560ajagVkddZaTNPs0b7rzn5XMWgVJt27aGwOuFbDNaSmeEay62q+RbT3by4VB20ZfBhLbZ0XOb5RboqWK01CKPzfFNgC3KQCCymZKm11pKmCUmS4CgP35cXnr+o49K2tNQq5ivzwAXgrrWd6JrXXpVbZh1a57neSrDOBrWoa6SbNCKsR1fRy6oirLUWMxVtq4DUWufBhLVgzUFuTGnvJSHz9GYeAySEBCxKSRoNhbXzieIqbyFYJsav2ku3vMF33fWl7VT25ZbK+rbTWw1WqOvEF+vUdSy7dXTZOtalQC0ZiIrvwpeqdR6PWysGWwv2YgyxtRZDbrmta6PL3FGXpS1fW0e8XLY2uHz9svXD3wWgwlsH1pKvzc6Xj91kZ14XqBcHZ30UzjIuVOdqWWZoKsCa770nsEWcMUw5MWAFRhf5c1BLKVEKhFx/x4RyDHQVvKv08+8KwG5SUnvLwDrXUZWqbh169Qa9rnV0nr++zFVW3SqHLEclVT9VMXihzIq75sLWL+YiVyzSlC3Fyzhc8bscYF8+Vxajy+kvbzO7kHed/MsmtevSpWUULtPKaqYXrcNbBtZFuu6StGVlFbRumeUdCoqVMdWsZVCWgaq1nn0ogdZaS5ZlcxdNxRi1YJhCLgC2/L1s7OfBDbrm/EWwVkFZlWZWiatX6ZfafMvsXEv05hulcqjj/M7LK8V6Y/GtBGt5eVlxfBldVxdds0al74uzcF0nlrkddlGHnYEYMEs4Miy2Q9252prWRMTVLfcrjuvadpn6sUz3rctfaQ0W3D/UwOIGpNEXK+LFB893D6wWbF1fUnYBLIqGSqlLAXvTQK0F4JKDMtcsRx5prUnTFFMRg5MkyX2mQlywBi/e+yJnLdqhGhRR1qXLbVFnEa6eq4q8S0GcZ17aPhfrMq2RrTTYdM5bePObKOoulk4Kqxhbcc1WjmsdQFcV1tZM/90C69Rin3+KFhD53jUVP1/ugsh/S7nK/VCP/FWdtcpNsYqMkGTCQRmNtBppNGiLzSwJHqnJQZekKXFiyZmqR2+UEomY0ImJVIaRGmUsXqoITLhyhU25TsUqHSEEYvpdpKnqh1X9s/qM5XvNAuXLbLn4Pb0XUwMXZWBPzy0L0Zw+AVx4sQlLZtccxfWGLZaCZhbTXHu8qk9XTDjF5ZKXSVwC2u8WWJnZekvrK4qzF93hi8aJ+kYv92s+AEXp/KJ1eTHfcsvnPG3uIikuSaNxyRDWInIHCgiLFRaZjFBZDFrjpikkyVTE1YysJDMa0hTPGuIsJTWa2Fp0OucmdcarKvCMMTl4tK7lmDMwV86vsvxebNOK2GxLU2v52lypv9h+gtI1VoC62uaraFn+qgA8H1dV9WgddUkUFS5NN+VSl9F3DqyQz1B2ugePKP5fIrpdhW7O3VN2+lsEFmn1jKNYLNoaMquxJoMsAZ1hsxSTBliTRyKFwgVtEKlFZhqhMzKRkkiN1M6szsX3MqAWYLRT/2ydRRfmq3Zm4F4h4tYZlAqLchW0Cy1T5sa1aebi+EqxuibvTfVdHTAr1b5R+m6BtTzbYsjnKplfqLReMTCvdZtr98QqE4VA2gxpUgwCjSC1kFhDYjSJlaTWQxuRu2pwsTYjSxKGMkSmChUqZCoI05TICUncGC9p1+7IXwfU2UfKhfNVbloG6LIPsJDPrLAE10khqzhrOe2ya6+C6oD5Mg2R1wLr626kpTSTLg0WDTYF3KlOVM9dXy2tFtMMEiNcMgQaRUxGqjPSTBNnmjjVmDQjTVLSKESnMXEQIp2EKHMJtU9qXNJM4yQGJTIiG5Fl2QwEZQDVgar4Lad+6Cr3VEotcNbyp5j8lrlmivKKNMXxUivyEs46EyNXdN8qsXcdTnsV91sl56KhqiQ51YvS69N3i7PCtA0MZAlxMMLxWjgNB8Tr2yI579CLoiEsiqUGgRUO2uacVacam8bILEIMntPQKePRiNHZKegEq1NGgwHtVp/UbxM6mlEmUQY6BogtiUpm94H5GtcylyuOi0ARMRWFpbwYsF9YpetAqkt6bnkiKO5fJ8lU67O6Hcuz7Tot/7qobiYRF5NwcfJaRdcC6xvHUStkrcFEE8bnp3Q3JMprgbzcPfPy6KJxqQpWay1WSIxQmDQji0NUOMAODolPnsKjn5AOnuJqzYedHkJAGAbsZRnng1sE0iHrdNCNFqluIFOHViIIvGheCzF/q3oBjhknlXLGNUXpXPkDzABd5q5F2eV7FM9WgJEVvtuqWD1rn4oYfJmtYF0/7fJ8q/OsL+K+nHH23eOskK/NTCJ8RzAZDvCaG1Up+LXShdC/guvphDRN0EmMnYwJHv6a4Wd/gx+fcbc5oNlISNMUshDl+gyyMUEU0W33ue15nKUjnh5GDNUOsegypENcserOAFnhjFJKjDG5z7nEacuRT1UQlg1PdUa7YjKw1iKn4nSVWy8PvpiLwZerXYui5/WAevn4eN086rulsxZkMkyWYrIEbQxCvuq6XqJQsSj+FsfCGpRN0PGE8OSA8NlX9PWQDSdCxjGRzZCOi3SaxFaQygZ4HooxjUziJTFOFPNYSp5Jy9gXiMxZEHWr3LQ4V55ACu5arltVhC0DFxaNUQWQF1xFpbKq+eqsxGXOuszSPKcrgLOUel299E2hl85ZX47Z3M49VBbm+6zb/M9qjBRIz6fX6mOV84pnRTF38s/PlMiWPtMoIwRpEhOdHBIcPubk87+jM3nClhPSMAEBLoloIHAZj1LOJxGDUUhm4c5mzJ12k000m0QYPSTAkihIxz4aSaY8UA5gURKUAKHmQJZCYoxGypzzSqXyiCitEVPOq5Ra4LYFsMocs44DSymXghUuWoPrDEwXqIS6Ba46a/76cbcYeFBMFPnrNWzF6HNdMM/yier9qumuVu5L11mrItRV8y8plQuuZJOSv08FUBLV20MKheu18j6/nNldStV6X9rYi7JZ/mUtDvmiFivy2F1tLIk2DE6Oefijf0nn6W/5pJcivIBYKGLTIsoyoiQljieEYcxoOCYME4yxHMZbZH2fblPR6rS5Q0wvPMdJn/Np0ufENjlRWwTeJpmEhk1wRILWTaTKB6mVEmsUQjlI6aCMk6/MmYrNKJV7xqZ96Tj5daUUFK6ZStTTQpuVROQ6/bY4XuYHrl4XQiCQF8XfGcirXVEklPN4J2unvmpQqt4AWScBLDzXkvT5wdxleHHemXrSF/Tl4lNPr4SzFlbCqlHi5sjkz5hmucjrOPiuxOLkrWTFG/O+PEseKGeQGCQCg5cFZGdHxL/9EXeC37C/qfF0SGgEoZakkSFMUuIkJopjgknIZBKT6XzfpOHhMYPBgK1eg61+i53dTTpehyCO+d72hAfNBl+cnfL5+Rm2dY9INohFC09NEFoipUBpi7QWaTRWTZutxEHrBu1MHy0MUCVxuWwNnkVFcdEqXejJxbUFX2x+sta9s1ZDr5H0Ki68ddItSzPfLO76XOOlg1UIgeM4s068biDCIk3Fx6IDpwKMzhKEkihnujmXyIMihHyJYSVr1rXaSQaBtgrHGEQ8Jj38Gvf0K7Y4o2EdMmuIUocghSwxxKkl1XkeK1yMSDBApvONTMdRjNEJURKTCuj3+0jVQGbPcLXDvXab0SThWTZgIjfQjodjdS6EG4UVoAxTcRCMMLWcbZmea6ectUh3QWetEX+X6a+z7/zkwgS/0iC1cKLUA1UjUsXfvg69OJMRle+r0ysB60vhpADIWVgh1hLHEVIJGn4DawWIXERbpjdc19S/LuWDblY9CrGn0JKstZhowvCrXzH+zV/TnRzQsCEJTSbWY5gowsRgEk1iDEYopCNxGxIRJejMkAFSuWQ6ZRCljOKM49GEXu+cjY0+7++1UJnmbiNDbKR0zYiHk4BR6qO1AOUhpY8xBlfmEoiUBqZ7LWmtF3TVghuWfbQzDltpy+qrO8rGp/K1KgDrwFucX6efFvZgqkm/ji56VX31ZaeHVwTWl0OGuWxrAY3jKLTNt860psTFhQVhAHWhlPKAuaBnrU2VQVTi4iVX4ZTTGBAWhUWYhPD4gEd/+yN644e0WoZh4hE5DcZakWmB1QZrE4TMdUkjTC6uOhIrDdIRCGkxViCUh7YaqzMGownj8YRoss+9Ww02paFnA/7RJ/fpPTri69MjniS7OFKR6QxtFBKLNAajLELMRdzywCovgi8AXJCYpi2HN5Yn67K7py6+uBwvXNdP5f65aD8oD/7SvYtOKPfNEu5WJ3y9bC9C1bi2ir7dflZh5kuNBDiOwuoMazRCFo9WtrzWFFFj1LhGRVZK2VVfpNEZ1mbIOOT08W84fvgrNtoJoVGMaREKl1CDMgaFQTiAkCAl2uTlNJsekBLFhiwzCFW4OqDhNUCnWKM5HRkEIxzRxnUsKjpjzw1p7Xv01H2+enII0kG5PkZrNAJl8smuXP/y5t9lC3BZHBalPYQvXKtEKV3mwrnMoPPi7sMbsDi+YvqWgjVv6AKo1popQAVCg0kTVKNVXsC2mNtejKZ50Rl0efbqAM21wvHJAY9//TMOf/FDtt0Al5jxyBI5DRKbYo3FteC6gsTx8BwfRxaSgabfc0jTFmEYcj6KGU0i4jQjzSyOK+lv9tFpwmAUcXp2jtBn7O92GP/2S6JMY90GW/ctPz98Qmv3PnFqQXpYYbFIhHYuTGLV4yL0cHZ+CRhtbq1aDOyvGJjKAK27b32bX1cSKvplsd9ei0njCvQtBivMFiBbi8lSMAblOYjpoL7o21wkIQRZlhGGIUopms3mNeuyhh415TJxHPP04Zd8+fO/5snf/SW35Bkff7KDGTwlHI7QwRlYj6bn0XF8hOuhmx5Nx8ErJAShyNltA63bbEaacZhwOhxzdj4mjQOscei2fVwxJIszRuMhlox2u0+7s8XXnz/kZz/7pwQ0MLKJbW+i/GLJXsm9UaLC11q0XfFcdWCtGoSK0uqioqpW4CJf3bmCbkI8fdPBWaVvKVgVhU+q0EmEaiKUQboCcGbLz3Mq/FfTzTfF3J9ljGEymdBoNGi329eoyxIR2KbkMRsCLVxSo/FIGT/7nF/+xR9z+vWn9MWYv/+DW/heSDSQuAhMeIIRTfytXTxPojyL66Y4doLIUrBmqgdnKCnQOqPZ6LHdbrHTUZx1PL45yDiPEgIc9nyFdcARDkk4QQkf3++ytbXP4MtfIhttvM0BjueiHYdMuIDEJc0X6lmJRaGFxaAoVhvmWzxZrJXT/rBII/NtTY3Jfa9TnbH82q06F95CIAT1doQrkZ3LU/aiy/XGQHq5kWhR/VqVei5NLE/1LQWrM+8AAQiJkouPIhZ+FUeauViar3N1HCd3c0i5sK3JVaguea4+aoyBVLrESYIJnvObH/0pH6gj7m7Bbm+T7Y4kmoTEKTjKp92GRAiUABwDNsQLIlypcZRAqtwvqnWGEhKjNMYEZKmhZ6HTdxHs8uQsZJJCnEkajkOzAQ1HE0YDzs81je42+5ubHJ2cosIhnV6PSRahpQShENaQSy4ShMVKgUBgptvhiJmxQM1+W6uwJl9JPBN3Z+2x6M6phiRWVZOZiAszC11hMLowL5bsAQu9b+dfV+/T9Q0/y6kyU1wo52p1+paC9eZICEGj0QBesFNKvlQrBBoPaRLAomyGGZ7x53/yzzj42V/zH/5+H7craXogUkOSOgxCjUg07VabhisRbgw6QgrwXQchJNpotBWYzJLpnMV5bgudapIoRGCRjsvtXpte0+H4fMxwaMmMpdVsIrSh2VacnI2Ixqf8znt7TIanHBw8Ydtt097v4poYnWlS1Zi7YAApDJIMY1kAY9mIpLWZjcc6H235XKGvVvui+C4DuE5vLVq7nO9lu+JeFq076t6QuJ5XTcsNTNcD7EWHd4rE4OTOoijgq1/+nJ/+xb9ChRPC46eYYMjw9ITB+YQoVSSZIIhSGp7PRrdJw01xZUS76dD0PLIMxkFGmAiizGGSOjwfZowzjwwJysFYSOMQoiGbnuH+ho/nNUi1QbkubsOj1+vQ7zVJwgG3Nhrsb3UJJ2OePn0GWtOQGqUTUgOpFmRGoKeANDp/kVX11R3z4/lrO6r7HJddOnXv3VnGYZf1iS1dW5W3PlfNFbv4/Spp3WnlLeWs+bxcB9ibmpEzcinWJCHjo2f883/2T2imERudBseTlGenp2TGop+PaTSbtK3BcwWtXgstA6SJaLZbKGMJzodMbIOIFhKfJLVkKI7GZ5xlKR0Z02q4tPwGjrBI8n2alM3Y3dvl5PAZYRizs9VnPBywu7MBNiMNTvjw3g6H45ijaMLR46+5tb+L57UJjEEIiTA5BxcYpBVYobElrli27uYhdfP3vJaDIMrfVd9tHdCqnLU4t9iDi5Ps5X1XFm3rLcFvMkN+a8BqLVgMgkInne7NVKLrgXUq+lb0KhNrpA356u9+yr/6v/8vhk++5n7HR1jJr04kwTDK16ZKh/0NzYfbgmbDYjwLUtBudtFpRhKEYBSx8AiEhzAOsbFEmeU8ddBxjMrGuBLaTZ+273J7s4UrNEEQ4DY1e7tbTEYjBuMRmxs9sixha7PD0eEpfX+T3/vgNj/9/ICTR19AcMbeg49Rrc50Fys3j2WeGj+MFpR33VgE5PxFzVWrb7V9y2JwdavUeZ8tun8uC5wo7lHkWSfA4nVQ7soqC/Lr0dsDVvLXShhjcR13usriJjpsXkZhoErTlGZm+PmP/h9++i//hKPHD+k1FUJajs4HPEwE0gjazQ20cTkaxTzYd+nfaUFP4Hp9lJCcfXNAPEoQxiV0EvBdjM4QFjwp6PiSKE4IMsMoSRmEGQ1XMQkjdjbatJt9gvMjvEaTjc0uo8kE4bl0O00mowG39rf55ukJ723vk0Sb/OyzA4bHz5Fui96DLsZpklqLkQrj5OYmITRWL1tBlbt+yty2nO4ykbfOClyNXKqeu7R3rgDKqxqVVvmAy1RYNBbT2+nCvHKq1fRW6azWTl+u9FLKzp3+xU744fPH/M0P/wXJ4BCZTfBcSaozBqMxJgnptH2a3TYpDqPEEgGd3T6q56FaTeIUzkcpRydjjs+H6DTCMRGujWiplA3fst9z2O85tBsunpNbZQ2CSZxxfDYksxJfGsLxkCxN6G9sMBiPcJst2r0e3X6Ptq8Q0Rnv7bTpNx0whvFoiEwClE0RNsst26ZOR62+DMteOH+ZbrpKBK5r4+sbAV+PPgrLWYJYK9Wc3hrOKhC4rovruIiXsHlawU3SNMVxHP74f/4fUeE5Tjzi1mabw0HIaBLjO4KPtz2aLYcnp8cMzvO6PDnLCLx9OjsdknPD4fGIk/OELBH0O02axDhpgue7KCGRSDY8i3Utbb/PKEoZBQlRmrHZ72GTgKOjI97vChzHYzQeklgQ0uH5+Sl3b+0Rn4358MFdvvn6a6SQfHBrm8nBhDAzhOdHtF0P5Sgy42BEvu042iDF/HUeZbFUiLm4WwY0zMXe4rt2N4nKd/G7akkuAjCuKhe9Fsn3BqMa3x6wCoHApdhFwBh48dV6ISYF6bYJBaRpRLcp+Nt/8adsH/6MTHqcNfqcBAknZ0N6TZdOp0GCz+OjmLNYEqqUtopIbJukvY147x6T4SFnR4+wk4S2F9PparZ2ezT9JgoXjEBkMSYJSNKAO14P4yoiD8JM8Dweof0WptXli+yQLa9NC0lydMbO3g7jNOPg9JR7mx2Ccci9D2/x8NERv/tgD+KEL56fc34O7e4O7WaLWEaEKiMUDo7xcNJ8JZNUYMT0DQJSYcQ8uqmIDpNS4rpuyb2jF/TcsvtnWYB+1SUjhEBWAAyLHvVSAUXGlT25Sr+9Ls1es7FQRFHLixPTZfQWicGi8qm3Ql6pxGlABkzX82QRg6ff8Ff//I/Z7DZRUpKlmtF4QsPz6bZaOMohSDSjMCVKc2NSq+Wzu7tLr7MBRpJpS5KkeErS7zbpb7Rp99p4rQaO6+BIiTTkLhUsNo2QJqHpQK/p0lQCkyZ4rqK/s8PZaES328VaSxgEOEoRBSGO76Ich/ZGl1anRauh2Nvq0HAsaRyRRhFWa0yWIDAz10xurZu2nSl2W1guyl6mny4TedfvhyU+3XKa8meWdvligOq5KnevO18P+Oq4y4NLrmPgems4a5VuwhoY4eA7LhiLk5wx/vrn/OX//j/xH/877/Hp5ynPDh9zcjbEBgl7/Q1cLcFKjscBwyAmlQ7ttsP+3gYffMzvieIAACAASURBVPABTdli/PCU468fkcQDHtzdYGfbx+sKlNfARpY00owHIXGcoaWDUF0ajovRlizJyLKUrWYTX1uMHhFHkk8+/piDL79h/+4tDg+f4UtLt9/hfDRiY2ubaDRg/8P3+OwXX3H//h6DKOH84Smj46f4rS44DtYoHOkgqKymKQ/UypaiBeDK+wlXxeJlhid4FVbbN9hPU0NvJVjLg+ZFBoO2eRytQHP4+S/59F//Gf/Rf/Yf8OT//Kf86usBRwfPaCnBXreJJywWh9NRzGmYkQoH5bi4jmI8GvKTH/+MJ19/xp29PnY8ZH+vw85ui2bHh6ZLOtKMTscMzgPOzyZ43U1ko0FiDHIypul5ND0H37UMB2d4DrieYqP7gNFoxIPvf8xoOKDRbRMMRihtCbSk0+nS6HRIz87Z3m5xdn7IR/c2OD2P+ObkkDPXpXn3LtY0c1FCgXHnO0RUwVpt07J/tbhW6LpVcbjIV3b7fJsikV42vZVgNcaQpilC5Ean6w4Ix+avZATLf/2P/zH/8Pu3+bP/5f9g+OmvGIwa9JXPvd0NOp0Gw0zx5DzgaZwwycD3PRwH0Bknz0/JfNhpd2Di4iuf7b02ft8D12F0NuH48ZCTwwGZlrS6W6SyTRAJng8GpGmAshM22i22Wj473Q3G50dgIDw+pbO9QSYM1ncwjmSj32d0fIrX3+Xh4wM+/v6HbOxt09jexP72C0aDCb/33i5JcsBXh18wzEb0734fr7GBJt8FsaAFX2oNMLMsw3GcC0ER5bxXi0K6SHa2bGMewH8TG+S9afSd0lmv0tHXjVYqh8spo/NdFZOY4SjgRz/+Bb/47UOeR4oNp8lWs0Ov1eb2rdsM4pDDScAQM91BUIKxTEYDMJb+xhbdzhaj84DxOMDxfaTnoKOEYDjh7GRAFCV5OJ8xeZywNhgDsVEEGQzClONxwigxdDa3CZOMlu8zHo7ottuMxiM2d7ZotptkWqOTjCzRDM4G4Lr4Erbv38Z1Lbf3+tzd7dPzBJ7NsEkEenFz8mp7VMMKL7PylgNJ1u3HOkBfCO+39sK2opf16bL7rR4nqyaZOj187SrV0pXBWo1KeRNJa73yesFRHed6gsVMXJPAZMT/8N/8t4Sh4flY8/jcMhJb7LiGe1t97t25y7/59HO+OR9yGAeMhcWVAqFD0skZDWH56P33+f73f8BwHPHl4yfE0gG/hZA+wTjh9NkZw/MRSRyjkwgdnSPSIU0xYbcr6bRdvIaPlopAG47DlNMww+ltc/DwIUJnfPPlF3z/ow/xWz5Oy2fv/m3Ojo7xlM+Tx8/QRhAkEb3tLrfu74IT8Ae/94A/+GiXe13BRgOQyWzPq7I/dQ7QRXfNZX7UZZy0Tne9Lr1ICOnlea927bJqXFbPK4/WckPezE6FN0fWWh4/fszm5ibdbndpumqkylUjYmYcAcmf/smf8ud/9mdASKch6PsOu90OW60xTlvyZDzkJJOcjDVGS3xP4ugEm4b0XcFH793l7q3bfPHlY8LJiM2OT3fvFn5/H5uNGZ0/ITwLaPoerVaDdtOj2XSR0pDZlMzCRlsSpJYwhiSzNJsN/JZEpyEbvR6jkzNUw+PZ14/w2y2cdgNjMjYHPZ4+OuDuB/d5+PAxd9+/xdlwQOPBLZKzTxFBwj/6/Q/58jTg89ThURoTa4EzHTZCiJnxCEAYg5R529TFDld9rWVD1DIDUxWwb7MOe2W0zfxcbxhQC2o0Gvi+f2m66+hHVWAn45A//+GPwGqEjukoS9exbDQEW7d7bN/b5zyMOB0HZNNB7muwOsGTls1+m267yeDsjNPTU+IsxW01aO/tkzk+UWxxVYOm22Sj12djc4N+v8VG12Or57LdVWy2BVt+ypZn6LmWppMvCOj1umxs9vAaLkpY4skYkabs7G5xcHTAvY8e0PKb6DQlDCOsdBCtDkpKmgo297eI4zE6GNFvOEibgNSzsOBagNl6blonHl/GNV+Em746erV1vDJnfZNnNiEEu7u7l6arOtkvSbyQL0+fn/sv/sv/iqOj52RZRr8h2PYFO17K+xsuW3e7/OQ3X/DrxwPGWqEziQ+oNEM1BPd29/idj+7jKPjZv/kFVvkYYYmF5smTA3TYZtNmjE7HdFtdWttdpDI4NkLaGEeA6zs0EPRailQ7nEUwjAWTeIyj2uzs7hDIBo5UnJ2dcXTwlMZmj/c/+R6//uJz9pstbm3vc/j8mHv9NsQx7W6PSRJz63sP8APDwa++ZvO97/O9nVs8+iYABFZP/axivuvDKj118WOwxU4OlxiXypPjdS3DdXBaqoGuNS7K9bvqnVfTZRPUd+qVj3VLqZalW7vzTb67BEJisCgMWA02Jvn1Z4zCxwz8IXcbH7OTwEd7EZ/8nuGnPx7w+TPL1wPJSCdstRyicULH8/l7uw73v3eX8zjgy4cHjCPLVqdJgwZHDwPOjj5lewP+8OM7OFIzCYcwHNNsqnx5WipIQouyKQ4S4Roct8GOcthoSB5ngtOzId7GNuqDO+zc3Sb69W+QozEHX37F+/4PuL1zm87vbnL2Fz9jXzdIj0Y86p3y3oM7uM9DREvT+QcfsNvSHD494mgIW3YDk2hCMQ33EwqDwAgJ0kWb3DqulEJry3ST/inociOQ1iClwFqNMRYpnQtvbFjmb60GPqzo4ItbkK5B64na+WKFyzl/zaqamYV68T7lCWkVvZmy7AvSOiBce8KREqabXyshYWpI+cv/9y84Dc7IEujaNk1zwp1blnvv7TEJW3x9fMJ5GJGmCl81SMYj+g24u9fm3/4Hf4jWlq++POD4cIiUHsZknA+PCIIBnVaHzd4tBE38RotGy8ORAgeBQ761uSTf0THRCWEUMxqNCCdj0iRmf7NDS2hOv/mCtuth0oSP/vDvI9s+icn48ref0W92eP7kgPf/4HfRmw2OTp5jT4akZwNcIQkePkJIw+7vf8RGt0Pn+RiTxMQ23weqbGAyxmBNNotwWga4shGqOLecA1/RuFRKNgfqxTDFZZ8XoaXRTdWPKMUwrTvxlOg7CdabJCvINwGbzpRpHGHSjL/64V8QmhhhFS3r0HYiWo2Enf1tfvPFMyIsYWrIjEVoAWlMQ6Vs9RtI1+Hps+dMxgkCD0e5ZDrB2Jh2p4FSLoNRxPn5BN9r0mi4eK6i4Xq4ysFzFEoKpCNwXDXzFWutSZIQk8a0PYnQEWY8RkkJNmP79j4oSTieEJ+eY1ONbbi4/Q7NVhMRJgxOh9hWkyxJiKMA2fXpb2+w4zVwBBimmwoXH5PlO0gsEWOL7zpQVtMty7eWfaF2vL9OCXDVJHM9XfetDIq4CuULsHNuJgDP8/ij//6/41c/+SkBCU3rsCdcfvBAcf99h9PonCcnDsdpzDiTeE4DZSJub3a4u9/hB5/c5Re/+hWHxyPCwOA1umRpguNldLo+7W6DSZgxiBKYBOw0d+m1mihtcKVEKgFKIIzEoPM30RmJq3OXVWYN4fAY6bps+h4Hn/6Gu7/3OwzHA7bu3uL8fMTk6RkHX3zJ+//+v8fT50+48/sfI40mfPgUv+kTtj0QlujRM9y2S2enQ7+Z4o4ChFEY4TDdFi0XZU1u6RZC1u4CUXyXLcHLjFDrctRVaozNEwD1ccPr0HWsz4t1rhGDF9Isn6yW0TvOeglZ5u8WxWi0zvj5T3/C5PQYYxNaQrCpDFsdaG/6fPboKd8cpTw8GRDqDEyCTEM2Wy7/7j/8A4SyPD48AenS72+j0wyjEzY2u7T7XaLUcD4MOR9rBoHl0eE5brtHYgRngyHBKMBGae46UwqpJEpJpCtRnsJ1FW3fwbMZIpzQJuPgV7+g12szmoz56KPvsbG1QTgZM/7mIbu3dhkmI3bv7CGRBIOA80lAe3ub9jBFHp2htn2c+216MqVlNcKafBF8rrFOo5dy33YBymIfpipYgdn1sihdNe5U44uLfKs4bNUVNOvDFVy8Ln+VitDI4vOiYvPa0kK1Hi9017eEiq7RUcyv//WPCYbn9LsdpBC0HEu/JUmN5ZunxxyNE05GAWk2DRoxCS3X8ODOLZrNFo8PjogSk/NpoXFcS7Pp4boeBkkQZwzGAaNJSJRZRkHMONbERhCkGVGqSdMMm+l8tYuweWhd0eliuhm3FLhCYNMYmyaE50M2uj2E79Hb7GEdy/nJGVIbXKVwWj6ZMvlWMF4D1WrTVD4yiNE6gZag1xA01bQ1cssR+W6becjlKjdMnZW3Lt2q49dF1wXXTdM7sF5CgjzOVFiIw5D/7X/9o/xVHRjaKDYbMQ9uu7Q39vjsMOPr44yRNkjRwJPg2gn724of/O4HfPPoGX/768eMQkOQREyCY3wv4d69HRpNj9Ek4eB4zLOzESeTgEGUcDRK+PrZBNndIlE+wyhhFCaEYUYcJaRx/iIqMPOXM2OxRoJVeELTwPD0sy/Rw4Asiui/f5/djx5wdHrM6NlzulZh0fh3dhknKdl5jBU+pt8jHQfoJMDZUezt+vS8fPM0IcAaDVZjTb7roTX1O0ms0l2X/X4RUFQNPuXXdlyXK74JgH0H1ktIWYuYcpB/8kd/xNHRM46ODrFW03cb7HQFO9uSYeZyNOnz6ESTCIPCQeiEftvwwf0ecTLhpz//lLOJwooGkzDEELC969NsS4ajIc+Px4wjhVENUgnH4yFPng/56vEQ0+jSu32bRDhMwow4SknCjCiISdMUO40ZTozGShccD6F8JJq269K2cPTlQ9xGg2AyZPPBLYw1HH72FU5q8FtNNj5+gNdpc/bNIZPzkGS3j265OElEr61o9SSW/C19AjsFp55ahw2mNJjLIutlgL1Mf70qlUFZ/n5TXY7r0juwXkaWHKwYdBIyPD2j1WwxnIzoew77W11abZcnxyO+eTYixSNOI6wW+ELzwe1NPvnkA37y81/y7GxEahSZkTjC8ODOJh+8d4txMOHoeMD5MCSKdf4yS0eihSTIDMfDkMPTIf1bd+hsbuJ4PklmSdKpuyTL8jfTaY3ROXfNN+MWOEpgkoSmcvCAbDKhvdkjsprt/T1sEHHyxTeopk9zq0d/s0/b9Rk+PyVR4G71IU5pKYlqgDERSPKd+xEIA8YatM3BuyzQYZkVeJXou3htOXAvgtrmrvHCVwJYVgf33/ROEdZevNvCcc39LrvTO7BeQlpb8ldLjkmePaGfCmykMZ7DLX9IT7lk6QZPBgkhKWka0BKSxFHs+QmfbLi0m9v8atgnkC0ajAi04nZD8m9tNmglGUfnkoOhy0QrlEpQJChrMBpCfA4TycNBxMRYNnd6ICWR7RBkEpONMZnGRBkqzWha8KTFcSxaJhhrkdbgkeKbmJNnj7Cug/YcNt6/TSIEwWhCOooQJmVzv0vqR4zjEWYYY7weEQIbp2z2fPa9DLyUE1+ijE9TCxJpiDCzzcBXiYx1VtaqUWgxn8FaDViKfYnzc8uNTUKQpxXFK0D0LF8dJC4TkevvUbOtTF0ZeWXm4BRi9vrLhWtrMP13YL2E4ljn7y2NE7IsRZNhTUrLdei2m/hNj+FkwnAcEEdZbvEU0HRSdrf7fPTxJ3zx1UPOzwYomftDt9uWDz96j43b9zgahjx+dsQkGGOMnjnNARD5S6/iOObRwTFPnxziNFu0NtoYmSKQmFShdS4Cz95IDkgxXWghFEJNLZlY0iAkfPqU7V6fzVs77D+4w3A84vmjR3heg87+HsJzGT8/Z3I2QkkXp91GC02n32Kj10AlIT4GoyCd7ikszOWcFZbHDpe/3066HK3vwHoJtZouSsDxw4ecnjwn1SlKGNquotv0cTyX8ygkTg3WSKRQGKDjRnz00QMSA189PEBnFqxEWMv3bje4/94tzhLNV0cDRkEy3bFwPu9ba2YRP9pohhPDo8MRYWrYuLOHdTTaarLUJc0MmbZkBoyxYPJSlJi+Wmq66EJai6ct54+fYk7OMFaz8+AOTttnNBzhOi6uEmzdvY0NMqKTMVGYIrpdtNI4XZd+v0ErC3CymFRqYjndqbDGqFTWWfNnWr0c7nVbW18W5Y902XNd/tzvwHoZCUBr/ubHP2Y4GGKsRghD188jiWKdcTYJiFKL0TYHnRTsdBRbmz2enZxzOgxQ0sFkGs+RPNjv4LiKw/OAsyBD2+LtcCBFseu8zN3qU+NWEBuOzyNGYYbpdXA7HlaA1gqj8ze7GZNz4tzvOXenzD7G4ANuqklOBwgBqt2gf2uXOImIRxNsw8PpdfCUQzgIiMYJOC7Gk2gfOl2HnjT4OsVKSypAmMKNs9xwtJy7Ljb3dxGsuQR8Gef8DnDW1z/bWuIw4Nd/82+IwgCNxnUMOw0HTymGccTBeMw4yhBaoIzBcy1/7/0dHNfls8fHHJ4FONYgspDtfot7+x1ORxM+e3LG4SCeMkKbA7YU6CBK259MYsOzk4QvD0aMhyHd/Q1oKJANMiumXDUHqskMRptcNBUCUwSeG40PtIwlOznFtQbhWbY+uodwBCffPEb5Pqrl0droEg4Dgmfn6DBFdJroBrS7lrtdyZbU+VakUiAtCLMY5FAXpbTQqrNzl0cqXb3Hrj5mbmqMvcyx+kaBtSw2FREwr52k4W//+q8YHR2Tpgk4llbDYdNTWCTHYcRxqkmMpKk8ZBrRbcD9vU0OT8756tmQULv5elcn4/6tTZJU89mjIx6fhCSZwBFT44mQU3/lnOMURoswShkGhi8eDnj0bEx7f4/O3iaq3cQgSDVk2qCnorA1eeyuks7MmCGkQFoDSYyKY/TJGdaVyO02Ow/uMnh6RHw6oLGzQfPeNgLB5NmAcJjitntIV+F1NPf2mmzYFKkTBBalbf4uWmsugLUctWTt4uqSqlulbtXUtcTjhai/qhunEmy/xMVzISh/iUFpGV3HVVRXvzK9UWCFeVha/l6aNwCsWP7yhz9kcHJMnES4voPnCrq+IrOS0yhloC1WuEgDTpay2fFwpeLg2RlngcHKFkJH7HQc9jZbPD9PeHocEMQgrUCZFG0yspJzYd7P80EShJqzc83XTwZM4oSdjx6QuhqNRBuL1nYO1OJjLVM/BnZqJTU6xSYJ0ZNnSFeim5Le7V2i8YTzo2O8jS6tD3aRriQdhNjAIKSLajZp77XpdEBMxsgkRWlwjMVgMLbewFTWWV81VYd+IZKu8sVeBtyr0k099hsH1vJOFMV2IK+TkjQBq5Emt7IqV9LwXVwMSWYZJxmhBSsUQme0HMl2v0UURpycT4gyhZUOrjLsbbdo+oKj85hRYMG6OAhcYVBKQcEFybmjMfONtaUU6NSSJIrTs5hJnGJ8RWd/A8f1AJFHL2k923ib6ebbogArkOmMKA4xNiM7GzIankPDRbU8fMdhdHxKGE5wt9tsbG0gU0N8NiINU4Tr4fZ8/KZko+nTEBLHWKSZbjZeAegqffX1kVhwcVYBWz1/I3e8oaJeK1jLHVkEdkO+b6zruq9u6xi7+Jn/tDx/9JDJ4JxMaDxX0XFduk2FVCnP04wwMhArhKNIRMJG3+Fu3+XxMObZIEXj41jNTltw//Y2mYZHz2PGkxhHGJSwSOXiSIUrcgNTMfNbBNoKNAKLJDaGszjmNDA8P9HEo4zND+7h95tYT5IJyJBk2uS7F06jiwQ2X5Ir8mfyGz5CCoTJ4NePUYEh291g69499OE5g6+foLY2aH5yD0cKho9PmByHoB2E16TR9djfEjSzc1JjSaWPsjloyxzdGj31h170iV4MeFj8FPmq+2Ut67hFl2X+l69vBOz8WFsLUqGxuU1AlMuo56zl+1cnnWWLB1YOt2tOWG8EZy3AWvfaeqXUawsTs8DD3/6Ws9NjEgyOFLSVxJcgnIxzrdEJeIkAa9GuptkR7LQkzyaaUSKQwkHpmO2uS7fVYBIaTicpOkvwZAbCoIUDxiK1zoNvhMj1VyGmH4UxigzLWEeMkoyDgwlR4mGlg9fzsZ7AKIkRisxCVuiNWmNNhrAGiUVJSafdRiqFdaB3MEEdjdEbffoP7uAnhujZOcJroe9uoZou6TgiHsSIBBA+rX6bfk/TtEM0kkj5SGNRUxeOsHYWjljltEW/lvu4ClQhctdVoXzW9X9+ap5+Mb9AIGcAzf/yYyvzAIQ8nslMFY8iYKKeLht/hSS47rtjq+fWFbXfCLDC/IHhNYtKlWCST3/5d2RJQpZmuFLhOxJfuYAizTLiNMt1RZ3Sbnh4vsvmzi7jcUCWZeSLtDM2tzYRUjEaB6RpSrvdptVqLY2AsaYYfMzcLpCHEoZxzOnZgNPzIekkorG3i2o2EcrJXxg1bTrNReusEIIgCKfcAIwSTJ4e0tAG984OaqtLFkQ4kaHX6dHe38KRkvT4nDSIkL5Dc7eH23HoNl2aErAGU5ZIlvTdVfp1mXj6ouRLSOMJQudL/YS1L7xGvWjbNE3zPq/Qd0pnLXPQ1xpsLSo/Ldg0JosTrIam49CWgpbrYHFI4oQ4A41EmZSWJ9ja6KP8FufD8dRHqnEdweZGj9QKTocBSZIAi9t15tFGpXA8mYvDM85k9TRQwhJGCVFi+fS3X5NZD2d3h427d3EbTdTMPwtZAdrZ2tL8dxgGU73WgCPQT49JvjpA3Nqg+fFdCBLSZye0+21aD/byyKfHx0QnI6wDzo5Pc9Nlv+PQExqJzl8JuWZ00irA3sj2K7b2JwCOjsiCIUcH32B1Qi5y52rGdWidyec7obMuM4O/atDasnoDsx/CQtfzp8YlSUNamqR4UpAaRRxnZFqQoVAmoaUy7t+/zfEoZBykICUmi2l6gs2NPqMg5TxISbOM8WTCeDIBkW/HoivvibmgH2Fza65UGKsYjmNGE81X3xwxyTStDz+itbWJ43u5YGdBI2ZusNwmkO/U4Ps+OsuQCFKh8YYhfP0U2/FofXgXkWkmXz9BSYG4s4lQEvN8yOTwnDSNSN2Yzq7Pve0mXZEhhJ6/P3M6dleFGF42wOu46pXGQ3XSLVE4PCWdDDl7/gxpshKarzfe5gZBMxOFr07rsd43grPC613CNBc4melZWZrvxpDFAWqqQ3pK4KHxHUlmZG6xRSHV1I/qCvZ2tnl6ckaiQQiFsJqm7+B6LkmmmUTzHRUKsVROpYqywaJsCZ5ZdAGtDZ1On0mQECeCg2enRKlFOC6d2/vgO6Q2D13PSjsyFHpj8b7UMAwRRmAkqCxDDiYIR+Bsdmh2ugSHp2RBiNpoIxoOItYkw5BoHCB9SWPDwyHCNRHG5qqAtdT6WtfhqmUPQNEO5fRlffdyQCwfR9bmG815nouZvd1unuW6KthVuP/FNOuN+9cO1tcu+k6pMGcUQBFCIIwlnAxJsgSDxHOg7StcRxFoi9aQWQdjHfotl42GpNNscDoKyWy+hMyRln63CQKGQcIksXlgvaPQ1pDqDKEkVoqZvmetRYo8/8xabrJ88tCW4TggiCzDccpgrElDSzIOUPfv4e9tYhouuA7CLho8Cr1KKQfP88BYzscjHFfiRhHp8xO8dovmrR3ikyHRs+c0t9p07uxgBEyeD0gHMUZKvI0GrpdhTQjKYLWZBmKs5qJlIEK92FumZWPjcqNPjTUXcP0mmzt73Ll7n9EkKFLPF09U7rHq/otWY7nWIvcXEe9fO1jfBCo4qyE34AghENai44Q4mpBZi1YK0LQbCuUqIhQ6s2QoUgMdX9HxQUjNKM4w1sFog0TTbblAvkVLmOVhhI7r4noeynFgyi0c10XKKVe1F4PftdEIxyFONEkmGAWaSWg4fnyMVA3IUlo/+B5J0wFH0ZD+hcFgjEEpSbPZxJGKUTgh0BEijMi+PEA6DvLuDiaMiR49Q5Ph7m8QSUFwPIIzjc4kqt+h1W9iZYJ15xqftVMuuII5rSMGX0V3rR/sNeADhNdGtXo0u1s4fqsw4QGr349Uqn3t8XX5zVWY+DuwUo1Oy+dYIRWDwYBgMkJbgxYCISy+J9FWE2ibA0qoXESW0Gs5RMGEZOoXzXRGs+HTbvqkaUIQpyQa4jQhThJSnaGtITOazOR+0YKrK6kWB+iU8+bBD5JUC4LIEMVw/M0hg6NTwihAdhr07u7j+A0cMd8tv9Bb8w24dT45CEGn3yVx+P/Ze+9fy7Lrzu+z0wk3v1zhVQ6dE9kkm2SLHIqkRjNUgETJM/ZYHnhkyLDgHwbGAIZhwJAN/w3+QTYMGBgMNOMZKEscaSgGNSU1U+dQ1ZVzvXr5ppP23v7h3HvrvlcvVVd3q0nXAm7Ve/ee9O7Z37PSd61Fv9/FXb4N7QTVqGKUpnN9gfWVVVw1JI8U2knUcobrOUQY0JqZIKhqnHAEWo0KD8rX9uVyu8lugHwQ7ZSrkMxJCqEwUTziEXu/V/LN+zNhtz3afez+EKybxA9C+cKXA56EHHBqlUJphdIC5y1ZUXbqF0LihEQrSRQE5EVesoUGEaswNISBoSgseZ6XJWyU0V5jDEEQbPLBxn21jZ3ohyZ6nud4LygKS5oXpP2UxYU7SGPI8z71yQmU0WU0e8znG7UCtUOygaBaraCiAGsL6CS4foIPFEEQkLY79HtdvJI4PWgzmuS43IOUhNUYaTQOi1FycO13ddXgxNtGiHeTrQC7VzN5O3Fe4IUuR2oqU2q2XSyB9yMfRubxpxas9xMoUIDCoshBgBUC51JEfwHpJNpBK/BEgUWEmlyCChS5DnC+R2BXiE1BvTXB0lpCu+PJRQKZoCYKJhsZWS7xLiYwOUYrqlIxIQ0zOqKpAwyU4WfhEHKQsPeUoEcihEYrSag9UehRIkMoQaefs1bELNzq4VYLfO4JTh1A7m/iQomWBmclXhi6vYKklyBtjij6aOVR0qMUhEGA7KX0bt8hqBjY38C3U6rn16h4w+TUFFnW5daNm7CaoASIGcEnWprDC3dYTjoUMgeR4ykQDNhT3o/yvkO5RysOcsPClwtSDn4fX5z3amZRBvCEYrDX2O8bVsI991p6w20U5gAAIABJREFUh3TlhIVye4Vg834b19B4AGwr2bze9pxluo91+lML1vuVUUBw+I/WeF/2Q7LWoqUnMppKFA46MQwYRr70S6vVmCAIaLe7ZHlRaiovqIQRlTig2+ujlKZer6KVKvskWYsUAlsUIEApifduNF9WSoVSqlwkg+ssG5R5lDYIqUjzgm5WsLLa5cLZS2gRg4PGk4+STdUGg5sVCokSCiUVRWEpipxhdlEjCaVCWUv31gK6cLSOHKASRmQLKwjvCaebeO9I17skC+u41EMjploVTJARFyl4OzArBxRD54YsCWBjRHeDWTt8D0aBHiHEPdpuc4R4p9Tf+J3dfK7BU3D4afn5hmDR1sGm+5X79bN3O8dDsG4lAxZTp9slz7KRHymlxGiNcAOe7dii0XoQYeXuQpBCoI3GmKD8XSq0UoNOhJZemtDt9ykGPqQUdxlc997kuwt1mCvN84IsL+inOUoFLC2sQFKQdvuIVhM12yprWUXJVjJalwUDAuyAJCEpNYySEoWAforPcnSjggkCil5S8mhrcXn9hSdr97GZhchgRUqropH9TskJ3iTvJ68qNgHn/finm7f94LMOHz3D7iFYtxMh6HQ69NMMiaBaCQlwRFLgbEkogLvAMsaUZAPrNiyUIAjQWpPnOaurK6ysro6e5A5P7i1BFI58WiklWusNi8uPwF8WpwdRRLXeILeOfpqz3O7R7hcsr3Q49/p5DCH0+rSeeZSg0cQbTe4KlBTkaYqXCi8NVpQleloIhATlHaaXkC6tQD3GG0VvdZ2kKPDNKlG1QtWEJDdXyVcT0BKnuxyaCpn0+ai1y/36qFuBqvyOttdAOwHvQf3avclHn278iQOr955soO0+DP5wUXiKokzwv/7aa7S73XKAcJ6jfIHLU5rVChJFkRcEQTCItJbd7IdPXKVKE3YY5CnZLWpk4gopCaKISq1Ka2KCIAw2+EXjWruM5uZYW4xK5xZXVsm9JPeCfgY3l9tkPuLqlUWEabLS7iAm6sQH5/DVGKckOjADs0+SubIpuESiSlsfITw6SXFr64SNGq1D+wmNgSRDRQZTjVAO3FIPd7sL3mEqjmboCPI+bsCLHQfeOGA3a87hdzRk/oxM/k1adPyznQgR95jXWz0E3ue62OiP3q+W3p4Icj/H+YkDK5Q3eTd5v2aPVmIU0zx1+jTVWh2EIDSaWmigyLBJhhHDAUsOhEApSa1W2zCY6S7gyp+N1qUZDSDACehlKYvLS9iiQA8WqtYarTXDaiQhxKB0bvCwygu6vT65dSA17U7KYrvHcj+nn8LS2WtMNGZYXlqBg/tguoWvhiytLSNE2V7VWuj0+hRpVpLZtcBLRyBAJzkojW1VKQSIXsZUc4LqzCQ9l6MTsAt9umtdTAhaZEgpkHKjJbDdfRn/f7vPx+/hXqLCO51vw7b34T/u5dzDP/Xu+1vttxuB4yes6uZ+ZPzp+4HLkPkCzM7OEEXRqLWnkmWHQLxHUgIUSori0I8crxy6e7yBaTv4bHjtDk9uC9I82/B3Dal6o0va9LcWRYF1HoQEIRFSkxSe1W7CaqfP4u074BS+sOQKRDVChgapVXlBXmBMRBAEKFk++BwOJ8q6V5EXiMLh4oDcO3wvxagAWY3JvS21ayfHpRYTlHXHOo4Qg4fovZHR7QG63WfbgWTch39fQZ9tzv9+ZfPuH4rFPZCf2JGPW0UEP5DjUnLSRWEp8rw05QYzSZV3GCGIwwglJdY6nGPUtHnot2pdfq1DwCml0FpjjCn9V1tqY+sHPXeVgjHiwlC7jo9HVMKV9ZiU0WmkBqmxHpST9J3ldreLkTnri216l28wdXoOJkOmajXW0xTV6WJ7Bb0kQzmIYwWFBSROg1UgvEV0+xSr6xSRIpeQ3lwkX+2QV0IS5dA5FEtdTD5NrVZl3a/RdW6QX95IiBjKuEm7XcBot983HmvjcfcsD2AK/33LT6RmHZqHH6jPOlZpAx6vFJ1uj2IAniAISs050Lxa6QEoBW5QglYUxciEHRchy/2lEKPjjZbMIPI8JCwEQTDyg4eldHI4qkIwAnGR5+R5Tp6XfmzuoZMWrHV7LN9Z5uwb70I/o+9yaFRoHDpIfXJiELySBMYwYN6X1yEFToHzFvIc200QgUJFAelah97KOmhVkvoR9Nf65P2cShxjghAfRHe7MW7y0bbTmruCdAxW22nW3czqrQJeHzewDmqqdt3uJxKsQ83zwWrVMuNYSo4gp7vSoZ0JnFRIl1Bk/TLYUXSYi1IAvFPlxLYCAqEIsGj6ZKnDBl36PkFToVXp4tQyqjKBUAahNN4LvAPrPEIZdGBoTFSJKxqNZV+9xpF6i1mrkRbSwpPmFodEaT2iEaayg7I5zaBB5mpc7imWe4riZkGQpKQtw535Fmtz01RTTUsYuraDcAW5sKTO4jNHJfPE3iPTHkV7FV2PmJ6eJjIhBSnVAGpSkZmUnlxFrHaRlTkWXcbRaorJljGBJiWgEAaVpyibl98PG7XrMO4ghAAlSyaUoCxokOV7Qqtttet40GYrwG67Noag9X5UeLDj9tt89sBumPB44fBiWEIy0BQ7HPInDqxbmU8PLsMizLv8WwAh9V0fFF9SBENNq9WisFk5iVwMq1lKTVipVKjFUUnI9548s6RpThiGBEqCyxHe4ooMiUOJcqq6liCLgnx5HddJMFKSOUvbZnSUwwpGBfpFUYyiygDGxMzNHcA5yK1nZW2dGzfvcPbsBZTQiMwyMzfHxL5Z0lBitcdIdU96ZfSzdWTrXWwvJTJlMUDW7hAISSWu4i0EIqC9vEqWZARK0qrFtKp1bD/FIFDi7tLyYntA7RTB3fz7dubzg8pIV+9w3O3OK8TO+5Xi7znGPRcwfO0gP3Fg/TBk3AgRlKaq87De7ZEmKbjSX0QInPNUKiFhoFCiQHqHUgYdhCRJgs1TqqEemcuFF2TOE4UhzXoNg91oko3dwFBqpuMGk3GdUGj6acJib41+4LASkiQhTdMNaRDnHHnqWLyzQrvTp9/PaXdylpd7nL94G1IB/QJXFOh9LdxUBRtK5Kbevht8TOtw7R6mXxBrgygcyVobjyCu1nE5aG9YXVgjSzKqWjBRMQTOQ15ghECNAmwOJ9w9i32rcrLN728mh2zvv34YD/BNa2RbK3Uv5/tgrukhWDdJWUKmEV7w1LPP0Go1EUDhBrNkAG0UUliqRiJFuej7aTYYW1FQMQKNwFno9TNW13sopZhuVYhMqU2NkmgpUKKcSSMGKSB0aQ5KIREetBCEQVk6J4QYdX4cAqv8WdFPc6TUGBNRWE03Edxa6HL+B+8QyBpZJ4FWnfDEHKJVHZWybdU5X3oQnZRicR2flw+XZHmdLMkwUYywAm0DXMfiexmh91SMZfHmdQIEyjrIy4HTmyuaxn8ebzI2Ovc2/NvtNPJ2WveDlg/pGXBf8lML1vspyxr/dBjMsd6TFcP0ybDG1A/yiR58jtYeJYbF4Q6tNaGWVMIAJQDnSbKcNMvw3hMaRaDuLsrNCf7MWVaTPuv9HrktMEJS1QGRl/jiLglkGIwa/uysR0qN9eWoDIcgKzz9tODmtdv4TOKyQQe/VhUXByDFPSbw8CUAkVlcL8Fbh8Pj0hyXW4SU+GG/wNLJJNSKSCu0EERBCIB1OXcjdvdaEVv5lzuZvzvJh6dNty+g38t+W3+28347yU8tWIFBSdpe6hT96DUMkEodUJ+YGuRSPc4LdBCgQ02a9Wg1a7SiAE0JnF5a0O/3WVtaZKYVE2qDsyXgO/2c3KbMTNZoxnp0x7RWo0FUxmgS4VhwKR3vcMBktc6+aoOwn6O9Ryl1Tw7We4/QisIW5Daj0+tSIOkmnvWO49qVFa6+/DaVYJK0myIPz6HmpiGMRvsD94wskc5ju32SNCHPc0zu0L0E8qx0G6xHeYNIHQEC8h7HThxFlAHyQemfgCH4N5nAm2WY3mJs283tPbczg7eSrczj7UzxvchWONrLQ+ReC+B9nGggP9VgHf8yd3pqecCLAU0QgRKKwkJYqxFHMVII8qIgKwp0EGCLnImJKrVQcfDAHMYY0iyn1+thtKBVCYmMQWlFt5ey3k8wWjPVrNOqBcRRgFECV+QoCUp43LAOVkqsLzv0h8ZQCQMCozYsrM2a0FqL8xalJWEckqQphZekBSyvW17727fJ3rtFGE3g6nWqx44QTU+PwHBPKkwInPBk/YRenuIEmNzj2l26/XUwpSUhhUF2ErT1uNyCURTSYZXHSgcSBB4h/T0A2Q1AW5m54/f1w/RPtzrPQzP4QxQhSpLCXqiJ4+J86ZdKKUEHqEEHBDswB8tobEZoNIGSCOdKMOc5aZEzMzVNq1knDIOSC+wFSZoOqm8kjVo8Otc4NxZKbWacwFBW9hTWkrgCIoXQG2/VuK9pbY5SAqEEHkvhHbl3ZNaz0k5IOynXz13FFx4vFaJeQ1Wqo+9mPNjlfdn62gFFXpAVOV6UDzGbFWR5QtmT3CGR2F5aHtd7rBjMGBeQFwUwGOzsywqkvWi7nYC6lfm8WwT2p0k+VmC93/YfH5SMJ9/LGallLyZb2LIgWkrywqKkQslyLEQYamrVgM7aKkVhUXGNxBbMzc0QK8lkvUoQhlghWFrrEOiAUHtOHjtIHGiMEkSBKc9lyxpVbT1VC9Nxlel6k36acGXxNtc7a/RtXi4/IRCDZmr48tq1UQNz3ZHlpYbOM0ua5awnkGaSy2cus/jOBbyOQRvM1BQqjMCXveq1UmUfKjcYbiU8RVHQSVOKQd1rkaQkaQJGYL1FIVhZWaeX5hSFQxiNUxIvBv3tBQzndogtNOa4mTuuUcff3w68u5mwwwBa+fPW7WVGr3v3/lC192htj594D8v9YwXWD0t2+9IVEoUqF5UoqRGhUkitCBWEQYTUEYESiKRNLXBYmxJXFa1KFeUCzrd7vNfr08Eh+z1OToEPUlaDiOsdiU8kFZGxf0ZyoB4TOYvx4Cx4ofEqIAgUkxXFbDOmVYsIjcE5SVZA7iVOGbzUIDRaGowIkE4hvcAV4AtJZCpUwhpGlzW0xk+x0Ieb7ZTX/vpH6ERjeznp/hbF1AxChMS5JhABiRBkViEyR+BSCidIEkGBxGpI0wybSKwwoAoqLqdvPUto8lSS9ByGkBBDRZTTAbwW5KEcAXD4GjK9NlfUbCZMjL+/GUAbIsh76CwI49GJu68tVsx9r62tHkJbbXv3QgSirHmirCQuf97p2v9/AdYHEWFCpAlLQKFJc0sURVBkTLQqVCqaOArpdwtu3lxmYXEJHWsOHpwi1GCkoMg9q+2UsFpDBYJKVVKplFpVC4WWitBIKpUIHYWsdjosra2RFRalNFpqtFJoJVFSIFXZVSIINSYY82d9aR4nSTIa5dBNu7S7fdq9nGvXlrCXbyJrNdJ+l8r+KYpY401JmbRJilKCNB+ys1w5k9a7EcnDWls2bxtopSzPy04aWlOt1cqG4JSpLgYtVxRqSy0KO0eIdzOFN2y/+b59RH7tRykfK7Bud+M+ahmlMITAmxCvA3pphhWG3EnCOMbnKc26ITQerUBYzdpawuLqOlE1YnqySqseYoTA5bCw1EaGMTIUTM5EBIEjNJpmtU6kDXrQf6mf53SShE4/IcmyMjjo5d2xGM7hCruhifZm9eAHkWOlFFYUJFnBylpCrwc3L96iaHcR0iFmGvhaRK7K1jI+z/GuwAmPEBLhytYzYmBul3nZsqexKE+0gcM8JJdYwEmNEBqBRridg0o7BZjGZTOYN7+3nezFL76P1XGf239wx/xYgfXjJlJK6rMHsSYiyT2dVLKeSXTcQNiUmUnN7KQmNp5AN+i0JbeWOoQTTfYfbHD62Cx1IxBO8+6lO6wkjspEnceem6fR0oRGECtDQDmZDiyJ9zitsKrsR2ytBwu+sPgiRziPVgopxYgCKQZ1pMNu+FBqWK01Tlu6ecFq29JtB7z6o7OsXL7J1IEDiIPTmIPTpJEC4akFIUoKTBiUZpkUSG9RpcoepLV82SuKuzNku93uoKggQWmJkwIZhHhl0CJAi2DLAvPtTNzdIsZbadtxuZ/g1XbyYIyl+5WhWb/zVg/Buot88vM/Q1BtkDvPrZUeSz2P9YZQSyYaAftmqkzUNVqFWBextJ7SdY7mZMhjJw7SiiQSye2VlIs3loibLQ6cOsCBQ9MIcnqdNQIJzVqF0JjS91Kl/yyNRiqFFhKNQHqPFOVLa4nzFqnEhqjukDfsvWdychIrLAUCJ2I6fcmNa2tcvXQbnEI0agQHZikqISIw5XS8QQcJj0BJiRYSJQaN2pBlP+ORP+aRQpBlGXmeUzGSSjhoSaNL31oIhfLqHjDuBEjYIyl/l8+2k+3azWwMbm7v0d7POe5HhNgZsA/BuoWML5pTjz3B/vl5TBCx1ElZTWCtkxLX68SVmFYjZGoywLscT8DtpS4Xri2C9Bw7tp/ZqSpGQ7tfcO7iLZIEwmrMqSceYXKqThgJjBkGTQVy0B3QeocJNVEUEAQaLUXZcFOUsWutFcN16gbVI5tJBEVRlOBCUDgBMmB1tceVC7dYvbyIUCF6ehrdbOCMLifL+bKZmvOglCY0GiUF3jmEkARhiBrkZoUo+ykPv7NT8wcJZUmR9IPeiUII1BZadPPUwM2Bo83ab6uGA9txrDcfb/O9HTbAu0su2RrwZRvY7VNG2x1/J3kQc/whWHcRJwUnT5+iUq2yngqWU8PVxTZBtY7ICo4fmeHAnCEMHMiAOysF715cpNCS2dkqzz19mIkJTTt1XLi0yIX3FjBxjcNPPc4Tz5xicirGGEizjCLLiaWgGmqMASUpfUvhCIwmCgyBVigpsEWOUqV2Lc3huxHI4fjIxcVFsqTAWkduc/pZSj8TXLm8xtuvXodEIKotgqlp8igqg2hO4H2pQUNtqMYhRgqE8wO/9e534z2jXLaUkn3VgLjICHB4V5SkCClQA+0/NIOHQN1sGm9lFm/+fVzer1YdFvXfZUxtve3HLTb1EKy7iYTnPvEMc7MzrPUsN1ZTbi536CcZ6VqP5kyL+fk6tbpCaU0vk5y7vEg7zZGB5/nPPc6+fTWEDlheSnnj1cukLqYyOcXjn3mG048dIq5oisIjraQVhTSikEgLiqKPcznWZUShoVmrY5RCSzkaoiUGPaOEECPqohtUCTnn0F4P+jdZcp+SFdDuKt47s8jKmet4qynCmGBimswKbFGmk4Qop81V4xglwDtbBp0G2rQE3IBWOHi/e/smk6HG+AKJRUiQSpR9rTb5okPQbgbxuMYdT4HspkH3qqWGYC2KYrTtXeLWxyPAuZ08BOsu4r1n/+FDNCdaIA3Xby+x3stor67R7yQQGfbtmyCuKIQSIAyLy92ydUqjwuT8DKcfOYoxAXkhuX1jmaWVLmhDbf80h04dxvmi9O2kQlqLFhAYTRhqtJFIWfqMURiiBz2TtFL4UUCpBKbnbjO5ITC0LFNEUjmUFqSZJc0Ed253uPjGWWRmqczMYqZnyK0jyyzelj2OjdZEQYASZdpGDPKZakwDGq2RagCqNGGqVoEixxYFYtBiXyo2aM3tuhhuBZZ7m89t7+PuLXBUHifLspFvL8THH6jwEKy7ivQgTMDskUOoMCbrO1bXMha7HjFTpeivMT0/yaG6QkUZWgq6CSxfXCExAtFwPPLYfiaqBqMj1tcsl14/Q54myIpi6tA0J4/M0IodRhck3mK9oxrGRDoEV9L3+lnGWrddsom0KkdUDkY/+EGgRwmBK2xJc3Se3EFW5KSFJbOC3CsKqekXBev9hIvnF7h57ia6PoGbnaATeDLnoBh0YjSgzSDijEQpgRy8lHAE5bC68tw6IFUxa6nHoQccaxBK4DZp0fuN5g7l3r5OULKNPIMZgIMXG/YZP+aQC22M2fLc28lWAaO9dHHcap/x/e6HrfcQrLuIGvT6ee7FFwlrdXCG27d7XF2yhHNNutkKtf0tjp+aYr6miUJF5jSvvXKVtSvLyMAx//gsTz52iIoOaCeOH7/0BhdfPQseatMxT37ucR59dJ5mLcCpsiFapA0BCi00UVRFKEk/S0e8XeehbGhcdvKPo4h6tYYaEBGsB4ccbC/IHPRzR+Y8/TynlyZcvrDCmz84w+qNZTyCohLgBma1jgOCUBKYEhRSG4LQYEKNlB6FIzISI8EYhdKGLgErSem7B7oMRGmlEdrcY/Zu9bqfdEupDbf2Kze/t9X+YRjeE+TabZ+9yoPtt/3nD8G6FxGS/QcOlV0ShKLdK1haSdGZg7oB6Zh85AAnKg1UoEBKzi92WTlzB9GXqMkqT3/qMZpVTSHg2s025370Liz3ENUKtRMHOPz4KSaaDQIFWngCJQl0yVxSoiRFGK0Zzm8dRoaVFKgBdW84kW6chjdwaEeF5jAoh3OOPLUsXFvgwutnSO+sYkxQtiONDYQSqRkVCGhj0NIQ6rLEz3mHkCXwynakYAVYWbZxkUqOUj/DfshbmcLjgHkQ83bTDdtWa2/Y6j6O+UGZyQ+y/0Ow7iJ26AuGVZrT06gwJvWGhZWC9Yu3qU5Pgeswf3o/jx07wGQ9RgvPQl9z9tWrJBdX0YHhyHNHePSTpwhDTZGGXHj7Km/97SsI66lOT3Lw8VM8+slnmJlsopUgyxKKIscXBSK3gzYxg9k0UgAe4YcMJk+SJKyvr5Om6YbZPMCoQfh4GxcpJBVTp7+ec/Hdq7RXUuoHj0AU4GONDyVClDRDHQToIECiMFKV5x0U4wtRlvJpKQirMbnwOCnKSh1RgtWwUXuOA3Y7TTt+/fcrQyzspDFhr7nQ+8mX7m3b9wvYh2DdRYT0gEKqiMbkNPFEk1yGXFtI6VzvoTuOfmyZOTjBkReO88SRAzQCgRd1XnnvNm+9fBaxmBIeqPL8L32OJx49QWADbq/1+cZ3/oa3XztDGDdpHJrnsc9/muc+9Qla0y1Sm1K4DCUg1JLQDHKtYtBgTZU+pFYlcIdzW6MoGgWZyqjwXUI8oizlM1oTxxWkjChSxdLtHu+8eZnq5D5aRw5Ds4KPFFaWGjKKI+JaBSU80hVIHLVa2RrGOU8UBkRGkvgCiy/3GwzB0lJhBpHlIXl/s1bdisi/ncYdym4VWps163gXjPuRYa51j6tly2vY/N5DzfohSUnlE4Dk05//LLpqSPD0ioDL11O6FzqIuILyKfVP7ef5T57k6L4GrtNjUSj++vvvsvLmTZQWTB3fx8985fO0GgEdYbm80uaVv36D7Mo6XmvCg1M89rNf5IWf/wqHHjtB1IgJY4WWFuktWgoCowiCu7lN71xZ0jZYjJujreMLeaipKtUq1VqNHOimjoU7XV5/6yJvvn0OMzeH2T+DDQ3SBHjrMIFChaoMKknwvsBZS2EdHo+WklBLtFZlC1EpQXi0NmhjUGNA3QzWvfqqu5mgG4M2g3u3xfZ+m/d38o0fRLaOcL+/Yz0E6y4yrHb0QnD89Anmj8/jFYgg4tJiSnspQ6cCbA+ahoOn9jN/YBJtEwopuXJrlQtnrpd+pxFMn5xn/+EJMB4lQxavrXLlzQuINEcqqMzOcPzZp3j8k88yMTeFUA4pCoyWowd3EARoKcu8qtioRUaDr8TYkKvh3zIwgYsip9fvsd7vs9rtsdrps9pOuHT9NjaqoFotssGAYm891lucKAckawHeWvpJMjhXOeNGS6hVq0iGUdeyxlbqstRwJ+LD+/EHt9puL5HVj0NS5v0+AB6CdRcpjcxSas0q/+y//s+ZPjhDJ0t47VaPC1fXUO8sQ8Xj6WJO1PjEpx7h8KEWvp2QFCHf/sFbcGkZnTsqMzGf+8XnOX1oP1FHsHYn4y+++RKvvfQ3+G45jjGcmuT0Zz/Dp7/0eY6dPkKjGSME6IE28rbAuXLUxrDb4fA1Lt77kqUzqI4pt5ckScra2hpL3WXaSR8d1YibMxRhFTkzizl8GBlXsIWDwpEXGV4UCDKksLgiJy9ytDEIURauS6Df7aEGuVfnfdnBX8uyifcmoG4VHb5fDbtXkG/YZgvt9hH2OXggeQjWXUWMnsY6Cpg5fIAXXnyBydlpVlSVH799Cc7eoacyQhydimX+iWM88tgh9kd18kJxbmmNV//kO7CwThp6jj5znF/8ua/w3PxpXKE4d3uB//Rn/5GXv/MdekUOUhLUY05/5nm+9k9+na/8/Jc5ceIYcRzDoD+T1qqkHwYBSt01KTdEfAcAHZrHxhiMLsdmeKA62WRibpapfftpTM4wNX+UaHYO1WgSzsyRZZalO4s47wgjQ2AkJtA479BKEVUqA5rjoL61sKVZbl058ycoTWAZmg2zfowxu2rY0bc/BrLdIsX3k4YZ3+xBTd0PQvbywHgI1t3Ei7KYWkCOAdfga7/0X9Ks1/GdjGsL67y2lhGtTSBWC6pZH70/57kjIWKqTVqDflLjh28scefqNQK3Qr86zeSzxzj54mFa9T611HLnluTiW+fJr/0IYXsIX8fUj1M/8SkOvvhVHn36BAenAxoVCLRHSw82IxRlvWmapuUICinxTiK9JvABWlYxBNSUZq5RQ9InqguqTcXhWDJf8zRMm8lKykwskN0+Yi3Bq4hCQUYORUYgDT6MKERE3zp8aPHKIakQECLtOjUSwrxPKEAJgyBACYMRagOHeBw8m9+DjUSGYad6MQwdjF5jv3hBSYUcvu71U7eqodne/91bIOh+g0U7B8R23/8ndorcRyVioFnLdaNBCGrTM3zhSz/Ln9y+RGdd8rffv8pTn78Nj07Qv3OTrog4+LnH+ZmOhe+e57bN+dGt6+i/+Da/PvFzVB5tQr3CY1/9Mrms8md/8GeovODMq+dI13u8+PMZJz7xHEJnoAIaM/M8/bVf5tFPPc+Vs+e5dOES7755hl43odftIbKURmxwvqASKJywiMKjABN6bF5QjUKm6gH1oEVjokmlEqLTjCLP0UpjpGIUw8s1AAAgAElEQVTt1h2+++//hDDWnDp5iGqliZmeITJl0zjpLIUrC+CbrQZCgMtzZGHLvlRBAL5PHEcIqTAmwJiAcJMm3Uqbblt0PgQwm/xNL+6+I8pPN2wzfowNrKeNEeWt0jibgXMP+Dftt/mYH5Y8BOtu4u7mN72QJf+wyHnxq1/h29/819xeqXLtesSViwscO5JRmTBETCCjOp/96meI84g//+tXuNJ3vP76ZeZrr/KVuRmoz2GadQ5+/gWeTxPe/u636C1mXHkv5Q+X/oRTr/+YF/7hF5k7+hiCKrJWp1I9welD+zjxuaf4wvVPc+v6Alev32H10hXWO20ya1nptEGURAojNdUop1mZZbLeoGJCFJLVtTVW1lfpuZwwjgmCiCxLeePV80xM1mnUI86svsepY/toNCawLkF6h7QW5QXTU5MEoWB9ZR2bFEhlyAropBavAgSGOI4JBn602hQJHo8IjwN1PK860qrD33e8SWIP2/zky0Ow7iZ+sAgEeD/gnSpDND1NY2Yfa7cC2l3LX/zFK/zTY5+meXoGZXLSokNwqM7jv/Q06zrH/fH36fY03/nBBfbNf58n/uHP4Wohc/um+fIvfJUTByb42798iUsXF2kvwpnvXSO5822efXGJI88cR0weB+VRpOhigWAWjh9/lCPZY/TfeoNgahICXU5gS/PST0Uh7Bqin5Mu91i7s8al966iXMFktUocGDqdHrdu3wEnmZk5QBQYep0uy7dvUkjHY08cIsRR9DsYWSClINCazvoqWa+PyC1WRaRekVhB6iTeGEwQEIcBUaDL9M0ApFuRIcaJErBJc92rU4GhpfPRaLStpDSVP9JTPvRZ9yplAcmAOK/KbglRbYagUSGhx7UbPS6f64OvgSgIjcZKqM61OPnYPNM6IJQhK72Ct16/xNqtJZQqsH6RuBlw+KlnOPz4CYKaQwiwiebq2dt8769e4vybr1J0V5G2gCIhX72NW72OUD1MZLm1fJGkfZVwJiSei4gPxsT7YsK5gKAVIoylb9t003V0oJianqLWqNLpdbl56zadTpcwipBS46wgSQo6WU7XOlSzhWw26HpLiifJMtqdHu1Or3QRnKUoLNZrEqdIrSeMIqIoGlXjDAkOsHXecVvfT4z+2eqO3He654OUv4+glPqd3/mdPW/8u7/7u7/zW7/1Wx/e1Xyc5a7jOiLQN+rzOLHO1atvkLdDVpY9xw/uozaX4XVE6voEylNrTkKquX5nlW7iuL3c4db18+ybVEzua5RNyYIWB46dYKoVsr66TJ4neG/p9VPefP0sV98+R9xZpZmu0V+4ggoFqh7jlWeyqenbPmHoSPrrFD7FiwxBhkAhK3UqUZW42mBlvcflm3c4d+Uat5ZXqNWatFoTKCEp8pS1tUWCiuHUc4/z+CefYur4IYKZSSq1Gnmaka31cQXoMER7T1MHJE6RRRPcDvaxnCpUpUUYVZioxcRxjA5jglHk+l4Nu1UEuAzElL+Pa+BR8GkYTdjg3m5Mz+xV9hJQuj/Nvfu221Xx/N7v/Tt++7d/+3/dap+HZvBuIjb9PwpqwGPPPMHpRydZWjjDxZdvc/VCwe//3t/wL6afJzuZUHMBdPuYeshTv/EPsDMtXvnTlzm70uXiuev80f/1//LlX/4Kj3zlqygjEc1pnv3ZX+Tk00/yykt/yQ+++0PS9ZCqPkTn8jJ/+d7v872wYHomZv/pI8x3FNPzRwiqc0xMHgIPcRRAZiErcP0+t66vc/ncRc6fOcfKrRWWbq+iTViCaXIWV1jSLMXbAm+7zB+c5ennn+bw048hIsXitSt0lheYrVWIarO4MKdwGbkrqLVquKVVvBfI+iy9QuONJqrWSu06AKgwIVrfS94fL5cbfbsbtC2jlM09n+9A7dsMhN0CQDthe/wBsXcZ/g0b392JFrmXczwE6y5SdsEfyxYM3vRAIQS60uSLP/s1rr/2e+TtLjevrXLurQUeOdlAOEAJPA4TGk4/dYSVM5e48MN3wUlu3uhw5o1rPPqiR8QJUiikCKnP7ueZL3yBXr/grZfPk7V7hMoT11vYosfVm+tcWz3DO+cXqE9NcejQPmr1OiYOMWFMp9NlfXWV1cVVrt9YZmVxhf56B3KPCWvEcYxUpiyhE5Z+r4f0KUcOzXDy9Dz79k0gNHTX17n41hnS9hpq/34alQiJwnmJ0AHSF/Q6XTDT6LiGXRegynI5rTVKqxEneasuEDvlRbdKs2yIvPotIsQfonwUEd/RA2obeQjWXcQOvjwJG6bIC8BLT9JTPPfi13npD17i4ruvsLbU5c/+wxkef34Gv3+aTqARQJWE6PAkX/r1z9HOerz52hV6ts7f/fAO/d/9t/z8f/ZZmgcP4uU+nKhRn3man//6cV745Fneevkl3nnjIksrGVo3aUwdwqcZxZ2clVttrr+5TGILCmfxruztboREe0VXCyJVp2EaCG1xQJ4VeOlIki5Seg7sn+P40TkOz9eYmqoSRgVL59/lR3/zA9Zvr/DEyUcJE8OVqzeZChVGK8JKleUbl/H9PqYWkoc1vAqo1OooE1GtVIkiTRiGeGPQ+i5Yx33Y7UQOXI67VMl7t9nqCB9FmGkrbXs/YyE3F5/vVR6CdRfZaayVEQWmUs4j/Sf/6n/mf/8f/weWL75HfT3g5T++yWf+2VPUwmUI+uQITDBDff4IX/sXX6T6jb/jx987S2+pxxvfTlk52+Yf/PpnOfWlAulb4Cr4yNB6/BSfe3yOJ+/c5p1X3uDsa2dYvrKI6kCNGoGPWK73KRgMqnKOpChIvEdKT+4irBQEvkDnGTWfUFWS6ck6+x7Zz6H983jrSfOcQNeQlX28+95FXv/Bq9SF5LmTT1FVAbfO3aCd5NhH5zhxvEH9vTXy9TqLtUUOR+u0RR8b1Im1ohV7qrGEsILThsCIDemaodw1ecfjnOP50c2aZlMuZ3ydjzTu3hf/VqDb6b0H0aibAbpV7nY3eQjWXWS7tTL6VZQ9dmfnZ/mt3/7v+Lf/5//BjQtn+Na3b1DZ3+Cpf/QsuDZSpuT5MiZoMHHwKP/41w5z7NAZvvsf/5brl65w+ca7/P6/vcjn1j7J8y9+murMbBlZZhpfHKAxdZwXfu6TvPDlFNrL3HjvLS6/9za3rl5i/01I04Ksn1IUBaEuObvKe+LAEVVjgnqNyQNTHH3iCNXJGnmeUFy9wsKdWzQqDaamZkk9fOvP/5zlxVWOTs9zYN80Hsv5W9e53lmiNbufU88/Su30FMsL32Olv069FmF8hagtUL6cVlCr1YgGEeHN9MKtNer2/ueOhu42H90PELYmRdy/n7rdPlv620LcnaZwH/IQrA8iw+96cD+efOEzfOsbf0x7dYHl1Su89+5lTj1zlOhIHYVHqhzv+yCqBPUaJx4/yfLiCp32Ir1ORm+tx4//6nUiL3nmC88RNucQqo7S01gUeZ6hBchWkwPPneTAo03WVo7TvpDRXmvTXlsn6fWRHoJBeVo4W6NZq9Js1Km3mqDArSxz4+o11laWmJqcYN/BI3Ss5/Vvv8TCjRvMTMwxU29irWVhfYlrnWXEZI2DzzxCc/8MQll8K6JPRiUXCBdhUxBhyT8Ow3CkSbci6m/WLNsFmQa/3P26t1ncfx85z43n39jneDvQbjXBbvjZXkTcD7qff/55/8Mf/nDP2//Uiy8oG5YJev2EUIG0Cf/Tv/zv6Z1/j2q94Pnn5/mFX/sC4ak63qVgKng/hfABQlTorXR586Xv8+Pvv8ziuS6mN0Fh1qntE3zpV17ksReeQ0cBXk0jqFCIHpYekj4FPaxLiYQuO0gM+4YVDqwF53G9WyT9FJ872osr5O2MahAxOTOHrzXJltc596NXOfPOOfCGRqPFZL1FLCQ3u8tc7dxh5vRhPv2lLzB98CCsLLLibxNbwXv/z7eodDwzp5/lrNjP1WgfzelZpqamCIKAOI7vqV3daoEKsdHZ2BqsZSpnt/W6W2Oz7QG/uw+63bG3ohxufn9YnjheWFH+eWKDVv7VX/06b7zx5pbofQjWB5ExsDo81mYYCXduXuN/+2//JcauEfhlnnp6lv/iX30d0fJ447GiivAGSRXvQ1RRZfnKLX707b/j7e+/yfJyhidAmIzJ/fCZz5/iyc+/QHXm4KAYXgEROY40K6jY68hOm167g+0lZN0eLslxWQ6VlLBSpdGYRNZb+MoEPvesXr/J2Zfe5Pz5C/TaPVrVFvVak1pUJS1SlpI7FBqOPnqCZ198gUgbFs++h7txg8lPHcUcmuLWH/0dd850KB7/DG8UEdP1Y0xNTNNqtUYadisq4b2adS9g3aoz4D2bfKzAerfj/93jDfsVl9d9rzb+1V/9Nd58c2uwPjSDH0jGAiMCtDIkWY+wVqc+OYtPDN3by1y9ssS7P3yHk198AkGCDBIEkLs1AjmF157J/Qd59nNPkrqbZG/coLdi8VnA2vWUH/7lRbKe4onP5NSmZ5A6RIUCLTRahuAkXkriag1hQmwQI60DbchmFYHQOF0lzzX2RpvulZu888YZzly6TpYW1KsTmEoNCEgy6NocWws4efIkTzz3NCYpuPn2a9w4e45D++bKbYOYxvwsy1cU61ToWdCm5ASPB5O2StHcm7LZ2gzerkrmrna+947snc10l3Sx2367FrRvY6pvRZ8MgnJublEUG7b3g5rjHc/zULM+iLhR9YcbLK0sTwhMgGv3+MN/82/4/je/gbErWHmdL37tBT75hSeZOVbHC4l1mtxrIjUHtg7CU+TL3Lx8hTOvvMOFV66yermP6xtyt4wOHTP7Jjh4bB+PPH2CmX0tVC2iX5PEUWXQpQHwAl9YyAvcco9kZZ3btxa4c2eZt985z/p6myioMD91AIQgczneuXKAtPccOHmQEy8+TrPWon3hMudefpnFC+eZn5nk+C9+iepTJ8l0ilpa5Oqftrnkj7ASGA4cPc7k1PQosLRtJc09spH1uhms4z7p+4nKbqdZ7ye9spNm3er33c4zrnWH0m63+Y3f+K948823HmrWD14GrVa8H80r1TLEWigi+OXf/OfM7NvPN/7Dv0PnId/8wze4cWGFf/qbXyacqiCCgjBQOBaRsodnFhmc4vAjBzl88gkWP32Rc6+e4eKZKyxfVmS9nLW1hIUfvMUrP/4hyIwo0gTBFFEUl72QTIBwkGcZvV6PtnMUaYZWCqMME40mM3OzgCBN+zgNTloqrYjTjxzj6PGjhLUaa1eu8NqbL7F+cwHImXv0CPPPniY4OY91BlXk2JaBySm61wNaEy3q9QZRFI1qV2FjBHR72aWm5r6yMh8lVWKbK9im9G63QFitVmOna38I1geQUTBYeIQrM/hKlv9bo8ic52d+4Rd5+tnP8L/89n/DRHyQa2+v8Uf/9zf58te/xOTpSVzRRRiBp4unh6eN900QU0wfmWb66JM8ny2xePYSF98+z7WrN1m8tcryQgebS9op4PvAOlIpnLXlGI5B1whZizBBhUBJyHPydp9uOyGOA+ozMTMHZzn86BH2H9+P1ZbVa5c4851vk763gsoFsffUjs1x+CvPUD+5Dydj5KIAkdBtJPRbM7hiDnGoRVyvEYbhyEcdDybtHC3dGztIiJ0BO+T4Ppg8GNh3u/6dzvmQwfSRiQM56HCNLymGMqIowDQrzJ84Sb5yk7yXcvbMVaK/+j7Py6eZOd7CaIEQDk+nDEx4gfQBXkQIodFhjX2nTjA5O8OxWytcu3yLK+cXaC/1sAmsFXewtgxcJEmC1noEjtnM4KUniAPEpMTUA6J6RHOiyaOH9jE5O4WNJd1+j6XLZ7lw/izpzVUmiwBtNDI0zBw9QGNukp7P8KsJ6rqlm96mPZvTzyeImzFxo4Ye5FO38lFhY4pju0X9/ri4w33vd/utzrV3WuFmk3e3v237Y5bnLIpix4fRQ5/1wxBfvrwAJ8pm3Nn6Ct/50z/n5W99k7WF8+D7HDrQ5Iv/6FOc/vwxooZA6DHzyUmgBqIOVBGiSXlTe2UnxawPvQTX7rG+1Kbb7ZGmOf1eB2cdZStQhZYRJjZEtQgda+JWRFAJyFyBX16jv7DM0pVbrFy/TWdlHbwgUpJ6KImPTDP5xBEmTx1EBZJk8Q7d1y6zeGGJlVTjGwdI68c5+szP0Jw7MvLbt/JRd+qoMB4N3gtARttseTx/z4LfSwR4u+jvbtuM/z/+ULp3v1J77nQt3nt+5Ve+/jAa/JHLwJoa3iYTV/jZX/5lpida/PHv/WvS9jK3rq/yn/7gO6yvLPDci89RP1BFRAmD4lkKCiw9ECHadUAYwCO1BBkjwiaqIWgcXaIhoLydrqy7HfQTlvIW3nlc7vA9y9q16yyeW+f2lVt0z66QrHcQ/YzQQTMICIOAyv4Jap86ytRjB9FzERQJyYWrrL92nsWlgL45TNicpLH/OI1DT1GbPUIuY/AJw8FQOxEdtqLb7eGr3P7zDet/Mw/x4yIP7kc/BOtHIgJhNFpKTjz5FE9/5gucfe1H9FYUnfYab/7gIsIGnP7EUSaPVwhig9cWRIYXOZYe2q0iB5UmziuEjxBE+EKSp6vlivUS4T3aKsgteZqj/B3Sbkp3qUu23Gf1ygrJcp/O6jrLiUVYS6gE1WpIo9VgcqJJ7eQBeOIw4VSIsx2yhUXaVxaxfUPQPIqtHmBi7hBT8ycIGvsodAXpNPisfDi9TzN2a/H3lMPd8+2O+bIf6Kk/ZvIQrB+yCMruiJaSzN6cn+cf//Pf5OSPn+F7f/ENuotXuXzpLNev/pj33r7EJz53lPlHjtA80MLUFUKlSGERro8scpxzuLTAJeBzgUsdYrVTvpdasIqik+N6Oa6wJF1Pp5uwtt6hkxR00xzrJSYMmatbTD0inKxQ2d+kfnSaeHYCPdlEhRo6a4ibd8gvr9JbDJATR4kOPMfE3EnqE1PkhSfXISiBVA6sHLGMhtpzK9NxL7S88W/wfsC/WxBqN/lgHzQfrDwE64clY6V1XghA4wYdhXwc8dhnP8uBY8d5/eXv8aPvfptL587wxlvLLN5aZ/7oAkcfm+fEM4eYPtwgjAVOtiiMR+BRIoW8g0s7uLRP3lsl7/TJOgnk4HuOtJ9B7kmKCbIsRwaaeqxpmJioElGdbMFsSGOqhW7GmIkKvqLI04yiv/b/tXdtzW0bZ/TsAgQpUaKoix0rztiNmo49TtMkzkwf2iR96EsnD/2p/QmdPvQt00mmTdNcmjjOpY5tybZkSaRIYL8+AEtgl7vAgheZNPfM2BIIYHdB7cF33W/RfCZAz3oYPOij3+tisLkP7B1g+8ZbiDa7GDCAhUDCBIB++qycg4g72X/KV6WryIUKEKYVMIylmzs7/ymcFnjXV59Nec5F+9Vky06y2kbCk3VeYNlbHmlyoBgVwE03R054gK3r+/jdnz7C2lYXT//yF/SPHuHo2VM8+uQh/vnvn3D3u0O8/8d38crBNYgdjqQhwJAAjBBFbYRbDOFFgNZ6E+jFEOd90CCBOL3AsDeEGAo0zkIkAQHrHMkawK6sobW7Cb69hfPNHayFDbBBDHFyjsGXj8GenKAxBNDuYBC08KDfxnG0h50b72H76i1grYWEi+xRCGAi0x0AiGBMYbXZp2OfFw3TuYdJx63gaSXyRKMwJFCUwZN1DpAbj8nF6oQ8R4dACJHvZtZYa+H2u3eBRgM/fvMNvvv8M/z8w31cXDzHF18+BMUf487bv8KNd7pY34nAwgCCCSBiQDMEwgiCbaRVyJMhWEII+0M04gQiIQxFgjBgCKIQPOKgKADA0O8PMbh/iKOzU5yenKJ3fI7+41NQb4AoaKLzSojGbgeD5hWsXznAxv7rYFErTauUtZSz/0kmh1g8wLbVJcokld/ZFJLHHS82aQJQierJuiiglLQSjAFRtlCcMQZiHJ3tLn774R/w5t33cO+b93D/66/w87df44f//AOffvItTh6foPVsHxsbIU4TgaTZRLDVRrTRRNTiQPMMYSNAq9VEFEVobrTTrCoSEOcnEDEwGBLoOfDo3kOcPzlD76SP/oMeDp+f4unFAIMgQsxa2Nu7hr3rr0FcfRVXd36Bm/tvIOrsYhARBmGCZszRiINs7GmapWBAGtRPFf0iQfXYqm6TmuZpnbpEy4y6qZOerPOGKdkchqrzjKG9uYk777yDO79+E49++hEf/3UPn/79b3hwfI5/fdaD6D/H0fExeCNC1G5ifXMNQcgRtAZYW2tia3Md3a0NdDrrCLjARa+H06MLiERg2B8giQnff/8Y/SFBUAiKNjEQXVxEIRob22hv7eHWW3fxxtu/QbDTQSNpgMcBCBwcCUKW2ZOcgZBumEws89USG+kLJJ8pC1wR5ZpEOjlzfZcxe1y0TIU2fc2UnnT7uxRGMZL88lg/b7l/0pfJKO7KivvFVhPXk3UOYPovTD8bKEej05SeobCBV2/exAcf/RnXX7+NH774CmH8DD//dA//O/4u3Y3ujCPo9RFfXCBJgKhxgWb0HOvRIbY7LYjkAvGgj8OTBgKWYL3ZAA+BZnsXcSDQ7u5ibec1sDDC7qvXsb5zFc3uLnauXMNadxsiiUEcSCIAEOmWlQkHMYIIssECCjnSX80qZnEaSsJKO1Hnoy5xjKVWxr70tLFSF5JumFqyjWwknbZomvoclD63nvfs0w2XEzt7e9h9fw+3bt/CoH+Kp08O8eToMZAM8e2Xn+OVvV0cPznCf+/dBxhwOujhwfFTbBHDZnsbz3tPMWx1wSnGw945Pvjw97hx8Euc9QfoXruOVqcDxgNsdrfR2OhA8AYGcYIhxQgnyrE138MMrDR5TZXroZLWZve+DHClvSfrIiPbSrG910Wb72Hn5gHeIAEQ4cadt9HZWEev18PtwyNwBgyHA/TOzhAFIdbbGzh9+gSNzjYYJTg7PcHBwQE63S5iIrTanTRLKh5CMAbiIRISIxe2vlhcojSv1yZVKxxNtuttx5PAJMEr7piqvzr9MIyXXjXBk3WBkUjvVJDZg4ynqjIHdvf30QhCRBtDbF3bT2OciUDAOSAY5LY8FDDE8RBBkEsrLtI64CAOYiE44+hfDNEIAzQ4yzbjKieeCQwAUWqHymQGJaG/0AYyW9dGTBOxnVVPQ35ifa5fRtpiISfVAZ6sC4g83JNGMIkxcIgsrpk6JXhWyJgHLPtM5PEhzrJdEBkSGhRso4xAshANheDgCCh1IQUCYIJBMIJu/RVV0jIpx5h8Ai1pQLMnx6lAII1htqSDqgwoMDdi110dU8dOdbuf5T8cmvZkXSDo8VkgLXYtQGBIUi5mJ4LM2xpwDhJp0nDuiyUwnpIrYDEo4KDMqcVJgCMtHyLAwLNZEvIQLPNSMmKgsoLJNVBW6jPltZSiTDk2EdXUTvHzSWO0NvIXvcSmlmz2dlW/jLmpvTo8WRcIikOwQNhUWOhThiFLZsxtRSrcDIwsIZKildINtbhMfGSi0KdsI5XUqZFXMtCaamWZdJzWy3p59uWLhSfrgiGLwCmETSVqgFx2pqqqpGNQ4BURATxVYwnSfgwgWAiAgcvoBRNIwhiMUt1ZgIHJ3cSJACop3iXfGxP6fHSb1NWBZQPNxL6cF+Gn+KI0eLIuEJSYK2MZQeVnCeRaUaBQYiwzYE1RyDRzKMwIn0lRDhDjIGIIqDCJZKA+vz0/VRyTDK+yelNQJkootMrsWBthTeqpLBejXI/iN5O3od+rPEs2HiLKx6Ffw/Tvp9qrrPZHhX+QHriRP8AYmrI37cm6SFCTKUqcONppYuNv7xEtlP1ipMRFRpyC+qlLFn2ijjKVbJPKUYIUEhPqyjKF1OOZFNX3WECFn0WiKtpK/rH7WLVXyOhFRXmbcpOu4XCojMUEv/P50mBSnXN2I6hwm8yuoxUCUVoMfGS3l7xYvGRdKrgRok4CADlUYrD1vhpunfmAsXRzqjiOIYRAGFZT0ZN1aeAuueokAFQRlawHl5M28DKDMabsAVTMGjPBk3XpYCKXjTLmRDbb9hS23qh4YLlRXU1jOmdou6D62Ssf2hMg9LamTVqoi7pL3Mq+B6kOl8GTdWmg+IrVM9Z4aBlRx4P5+W1FlwuzEkneVzUJ7eey4FMFUU1tpS+R3Hlm8h7LbKrKFTw1XgimcRahmyCmV9ikyRSerEuDMnV1hsqoGttIPypOohl2xSx6dHk6Y64VVGv7l+/00oc+yxF4b7CHikKYpwip8o3ZsHMyWscJm0vO8bRC9ed8MZ9OTIXRdXjJumKwTjXSzpJh4hRJIbXykrlblcyuSmyDqqosFmfqucKgRio1m9wWdUm8V80Dt7bKl8OrSBch2Anryfoyg8l8YHmMwuTPVuoUoM5Npvye8jITpZpNNn59fYxGVUa2TLLnWT9QpG0d29m21cV4jahqEpuHmn9PNjvceI+Ps64wtEyk3AMLFAlGUM8Vl6vlRFXvyRti2otAH4KbA0dp1RQrViRtzRCVYeHAJI6xFwlvs3osJOovFn/54SXrqqKQlD9+iilCchGljESdBeQm9damDrv0O8nSPtMY9J82eLKuKiy2WFXYpM6ErprMlXWZdO8zcq+pMJCrWJ94Fi8YFzLKTau1gSuJ/Ca1e5IxerJ6jKAmTNTVQxdX+s4LdrLNR4f3ZPVQoKYN1iHtLDOFy/q1nZvkBaPicrbumByerB5jyCXri/LylPVrO/diiGoO7dgdZNO8UjxZVxRjU0xxOM1GokwbrzSdTXM3Zi/xXB1VLl0XkzPGrG5phxfzK6RSUtG2J+tLAVsErmoGqPflBUfsC+cc5Q3siw5cY64kb1Dv0RIj6sRwXRw9VU4guTjAtU/TdURZOVmw9BsXcBK3Ps7qYZQW08muWajPk7SxWDamMxwf1ZPVw2hfVc+fRSTGy51J4cnqMSHcS8xUXGH53UOHt1lXHvX9k+XOGIAVqi2WpQ3qmVIAU5wzLv3VHf7FCjkAAANuSURBVN8sMJP2qdxeNsGTdeXhsIRbc/BImLKPcgeMPRtqPFUvJ3VVVQd9TPpYrON2xLQOK9M18jqZ6TTaArMmPFlXAtMnDIzAXEqspTHasnij1iRmOsaXFN5mXQkUSTC9n7e8qoEuZYugEoHiiVoFL1lXDtWLtU0YXVlSxMx439iKEvP1rmPRVU+XlTWux8VIc9pX/ikgl+xSfl3xdkZqggPsK2nyhIl67jVP1hVC2RaJZatfRhNubPG3zSbVsnisRCQj4Z23TbTYt5OtDpIElM9aGOPoN0kvmZyR91fIlShcJy81jGUCRcKTdSVwufag+8Lx5VZ9qwqczRreZl0JLBopFiueOkkk5rKJCniyerwQLNbLI40Ll8CkxTqUDnVv1O1t4dVgj0uBzY6ctCzLLMc0arZgpypdWRI1ija5bi+bzqn36+eqK0d4ybrC0OdGPWnhljRQbLOqZEwxMWIeWUjlz6ex03RFWQka7Rrb8TTwZF1hTGd2zUOVXSz12AqyHlRfPgU8WVcCi+XQmQcutQILsx5UXw5g0r+HJ+tKYFKJtTgkryLjctUZnmywnqweAFR7jnMu85zSigZZ0F/+KxKnzA50z25yGZ+aveSaaF+V8J9fY3+GuvWETdtlmDbaGv9XDk9WDwD5JCMinJyc4PDoCIC6UiSfePXarnJcVbVXRqaqPm396uf0y2YdR3V7uZSf92T1UEBESOJYOS5LF7wsLJSaW+Ox60rlMvg4qwcAVcXc3NxEEARj+cM58mVwi4JLdzA59KfnRruEfMq2iPRk9YAkHxGBc14qTaVK6ro/i8s1dWxQ/R7XSoW2/hQ7OEtMYIxBCOGsCtvaK7vO1k4ZvBq80hhL3Rk7yrfUGH3g3LppsrsmXhRtzuI9s06WyEkPJEmiVnOoGLtthZLbGOs/h5esKw33ag7pkrD5w0XyzKvGkhACnI/Lr6pUQxOqJKvbEkIVnqwrjjKijtZSq7GaeQ/JSX2dB1zTCKdRZ8uIXwVPVg8AZgeNFqEc+2R+KO/Llhg/LYIgUPowx0arK1DJ+8fOKYRPr661SL7mivrHAO473+Dh4VEXN4noiulELbJ6eHi8OHhvsIfHksCT1cNjSeDJ6uGxJPBk9fBYEniyengsCTxZPTyWBJ6sHh5LAk9WD48lgSerh8eS4P+HAK+ze81GiwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oaj6FtTrjrSS"
      },
      "source": [
        "image=keras.preprocessing.image.load_img(\"/content/drive/My Drive/dataset_resized/training/metal/metal2.jpg\",target_size=(224,224))\n",
        "img=img_to_array(image)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T2VI0Pemjy6o"
      },
      "source": [
        "img=img[np.newaxis]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yLhwS6BPj1On",
        "outputId": "e06f7c8b-7c02-4c80-d732-135553fb58dd"
      },
      "source": [
        "img.max()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "255.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s4GBOawnj3SF"
      },
      "source": [
        "scaled_img=keras.applications.resnet50.preprocess_input(img)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BKGDpda_j5Av",
        "outputId": "6da6b6bf-0c0b-4910-82fb-6c79252a3919"
      },
      "source": [
        "scaled_img.max()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "151.061"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EvZnW91Yj7W2",
        "outputId": "9a0a85a7-5752-4356-8f27-b02f301d6410"
      },
      "source": [
        "scaled_img.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1, 224, 224, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hhcM3nKJkDVN"
      },
      "source": [
        "Let's Apply Resnet Architechture"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0r5wlHC6j8-j"
      },
      "source": [
        "predictions=model.predict(scaled_img)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dayS9SfxkGwN",
        "outputId": "361ae5fb-b393-4a82-bd3f-1f096993aa44"
      },
      "source": [
        "print(\"Predictions:\",keras.applications.resnet50.decode_predictions(predictions,top=5)[0])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/download.tensorflow.org/data/imagenet_class_index.json\n",
            "40960/35363 [==================================] - 0s 0us/step\n",
            "Predictions: [('n02971356', 'carton', 0.5071724), ('n03291819', 'envelope', 0.16329086), ('n02808304', 'bath_towel', 0.084289245), ('n03207941', 'dishwasher', 0.025451887), ('n03127925', 'crate', 0.018747339)]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sg4RgVDqkKLl",
        "outputId": "3c95c271-6ac4-449e-f242-75fa21a9902f"
      },
      "source": [
        "base_model=keras.applications.ResNet50(include_top=False,weights=\"imagenet\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "94773248/94765736 [==============================] - 1s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FoeiZ1MvkUOu",
        "outputId": "5075fa8c-969d-491e-bc08-b09deb33dd0a"
      },
      "source": [
        "base_model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"resnet50\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_2 (InputLayer)            [(None, None, None,  0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv1_pad (ZeroPadding2D)       (None, None, None, 3 0           input_2[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1_conv (Conv2D)             (None, None, None, 6 9472        conv1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1_bn (BatchNormalization)   (None, None, None, 6 256         conv1_conv[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv1_relu (Activation)         (None, None, None, 6 0           conv1_bn[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "pool1_pad (ZeroPadding2D)       (None, None, None, 6 0           conv1_relu[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "pool1_pool (MaxPooling2D)       (None, None, None, 6 0           pool1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_conv (Conv2D)    (None, None, None, 6 4160        pool1_pool[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_bn (BatchNormali (None, None, None, 6 256         conv2_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_relu (Activation (None, None, None, 6 0           conv2_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_conv (Conv2D)    (None, None, None, 6 36928       conv2_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_bn (BatchNormali (None, None, None, 6 256         conv2_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_relu (Activation (None, None, None, 6 0           conv2_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_0_conv (Conv2D)    (None, None, None, 2 16640       pool1_pool[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_3_conv (Conv2D)    (None, None, None, 2 16640       conv2_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_0_bn (BatchNormali (None, None, None, 2 1024        conv2_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_3_bn (BatchNormali (None, None, None, 2 1024        conv2_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_add (Add)          (None, None, None, 2 0           conv2_block1_0_bn[0][0]          \n",
            "                                                                 conv2_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_out (Activation)   (None, None, None, 2 0           conv2_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_conv (Conv2D)    (None, None, None, 6 16448       conv2_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_bn (BatchNormali (None, None, None, 6 256         conv2_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_relu (Activation (None, None, None, 6 0           conv2_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_conv (Conv2D)    (None, None, None, 6 36928       conv2_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_bn (BatchNormali (None, None, None, 6 256         conv2_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_relu (Activation (None, None, None, 6 0           conv2_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_3_conv (Conv2D)    (None, None, None, 2 16640       conv2_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_3_bn (BatchNormali (None, None, None, 2 1024        conv2_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_add (Add)          (None, None, None, 2 0           conv2_block1_out[0][0]           \n",
            "                                                                 conv2_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_out (Activation)   (None, None, None, 2 0           conv2_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_conv (Conv2D)    (None, None, None, 6 16448       conv2_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_bn (BatchNormali (None, None, None, 6 256         conv2_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_relu (Activation (None, None, None, 6 0           conv2_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_conv (Conv2D)    (None, None, None, 6 36928       conv2_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_bn (BatchNormali (None, None, None, 6 256         conv2_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_relu (Activation (None, None, None, 6 0           conv2_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_3_conv (Conv2D)    (None, None, None, 2 16640       conv2_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_3_bn (BatchNormali (None, None, None, 2 1024        conv2_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_add (Add)          (None, None, None, 2 0           conv2_block2_out[0][0]           \n",
            "                                                                 conv2_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_out (Activation)   (None, None, None, 2 0           conv2_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_conv (Conv2D)    (None, None, None, 1 32896       conv2_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_bn (BatchNormali (None, None, None, 1 512         conv3_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_relu (Activation (None, None, None, 1 0           conv3_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_conv (Conv2D)    (None, None, None, 1 147584      conv3_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_bn (BatchNormali (None, None, None, 1 512         conv3_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_relu (Activation (None, None, None, 1 0           conv3_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_0_conv (Conv2D)    (None, None, None, 5 131584      conv2_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_3_conv (Conv2D)    (None, None, None, 5 66048       conv3_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_0_bn (BatchNormali (None, None, None, 5 2048        conv3_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_3_bn (BatchNormali (None, None, None, 5 2048        conv3_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_add (Add)          (None, None, None, 5 0           conv3_block1_0_bn[0][0]          \n",
            "                                                                 conv3_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_out (Activation)   (None, None, None, 5 0           conv3_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_conv (Conv2D)    (None, None, None, 1 65664       conv3_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_bn (BatchNormali (None, None, None, 1 512         conv3_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_relu (Activation (None, None, None, 1 0           conv3_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_conv (Conv2D)    (None, None, None, 1 147584      conv3_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_bn (BatchNormali (None, None, None, 1 512         conv3_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_relu (Activation (None, None, None, 1 0           conv3_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_3_conv (Conv2D)    (None, None, None, 5 66048       conv3_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_3_bn (BatchNormali (None, None, None, 5 2048        conv3_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_add (Add)          (None, None, None, 5 0           conv3_block1_out[0][0]           \n",
            "                                                                 conv3_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_out (Activation)   (None, None, None, 5 0           conv3_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_conv (Conv2D)    (None, None, None, 1 65664       conv3_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_bn (BatchNormali (None, None, None, 1 512         conv3_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_relu (Activation (None, None, None, 1 0           conv3_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_conv (Conv2D)    (None, None, None, 1 147584      conv3_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_bn (BatchNormali (None, None, None, 1 512         conv3_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_relu (Activation (None, None, None, 1 0           conv3_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_3_conv (Conv2D)    (None, None, None, 5 66048       conv3_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_3_bn (BatchNormali (None, None, None, 5 2048        conv3_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_add (Add)          (None, None, None, 5 0           conv3_block2_out[0][0]           \n",
            "                                                                 conv3_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_out (Activation)   (None, None, None, 5 0           conv3_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_conv (Conv2D)    (None, None, None, 1 65664       conv3_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_bn (BatchNormali (None, None, None, 1 512         conv3_block4_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_relu (Activation (None, None, None, 1 0           conv3_block4_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_conv (Conv2D)    (None, None, None, 1 147584      conv3_block4_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_bn (BatchNormali (None, None, None, 1 512         conv3_block4_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_relu (Activation (None, None, None, 1 0           conv3_block4_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_3_conv (Conv2D)    (None, None, None, 5 66048       conv3_block4_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_3_bn (BatchNormali (None, None, None, 5 2048        conv3_block4_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_add (Add)          (None, None, None, 5 0           conv3_block3_out[0][0]           \n",
            "                                                                 conv3_block4_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_out (Activation)   (None, None, None, 5 0           conv3_block4_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_conv (Conv2D)    (None, None, None, 2 131328      conv3_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_relu (Activation (None, None, None, 2 0           conv4_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_relu (Activation (None, None, None, 2 0           conv4_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_0_conv (Conv2D)    (None, None, None, 1 525312      conv3_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_0_bn (BatchNormali (None, None, None, 1 4096        conv4_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_add (Add)          (None, None, None, 1 0           conv4_block1_0_bn[0][0]          \n",
            "                                                                 conv4_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_out (Activation)   (None, None, None, 1 0           conv4_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_conv (Conv2D)    (None, None, None, 2 262400      conv4_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_relu (Activation (None, None, None, 2 0           conv4_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_relu (Activation (None, None, None, 2 0           conv4_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_add (Add)          (None, None, None, 1 0           conv4_block1_out[0][0]           \n",
            "                                                                 conv4_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_out (Activation)   (None, None, None, 1 0           conv4_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_conv (Conv2D)    (None, None, None, 2 262400      conv4_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_relu (Activation (None, None, None, 2 0           conv4_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_relu (Activation (None, None, None, 2 0           conv4_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_add (Add)          (None, None, None, 1 0           conv4_block2_out[0][0]           \n",
            "                                                                 conv4_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_out (Activation)   (None, None, None, 1 0           conv4_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_conv (Conv2D)    (None, None, None, 2 262400      conv4_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block4_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_relu (Activation (None, None, None, 2 0           conv4_block4_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block4_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block4_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_relu (Activation (None, None, None, 2 0           conv4_block4_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block4_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block4_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_add (Add)          (None, None, None, 1 0           conv4_block3_out[0][0]           \n",
            "                                                                 conv4_block4_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_out (Activation)   (None, None, None, 1 0           conv4_block4_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_conv (Conv2D)    (None, None, None, 2 262400      conv4_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block5_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_relu (Activation (None, None, None, 2 0           conv4_block5_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block5_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block5_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_relu (Activation (None, None, None, 2 0           conv4_block5_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block5_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block5_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_add (Add)          (None, None, None, 1 0           conv4_block4_out[0][0]           \n",
            "                                                                 conv4_block5_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_out (Activation)   (None, None, None, 1 0           conv4_block5_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_conv (Conv2D)    (None, None, None, 2 262400      conv4_block5_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_bn (BatchNormali (None, None, None, 2 1024        conv4_block6_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_relu (Activation (None, None, None, 2 0           conv4_block6_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_conv (Conv2D)    (None, None, None, 2 590080      conv4_block6_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_bn (BatchNormali (None, None, None, 2 1024        conv4_block6_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_relu (Activation (None, None, None, 2 0           conv4_block6_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_3_conv (Conv2D)    (None, None, None, 1 263168      conv4_block6_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_3_bn (BatchNormali (None, None, None, 1 4096        conv4_block6_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_add (Add)          (None, None, None, 1 0           conv4_block5_out[0][0]           \n",
            "                                                                 conv4_block6_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_out (Activation)   (None, None, None, 1 0           conv4_block6_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_conv (Conv2D)    (None, None, None, 5 524800      conv4_block6_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_bn (BatchNormali (None, None, None, 5 2048        conv5_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_relu (Activation (None, None, None, 5 0           conv5_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_conv (Conv2D)    (None, None, None, 5 2359808     conv5_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_bn (BatchNormali (None, None, None, 5 2048        conv5_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_relu (Activation (None, None, None, 5 0           conv5_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_0_conv (Conv2D)    (None, None, None, 2 2099200     conv4_block6_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_3_conv (Conv2D)    (None, None, None, 2 1050624     conv5_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_0_bn (BatchNormali (None, None, None, 2 8192        conv5_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_3_bn (BatchNormali (None, None, None, 2 8192        conv5_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_add (Add)          (None, None, None, 2 0           conv5_block1_0_bn[0][0]          \n",
            "                                                                 conv5_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_out (Activation)   (None, None, None, 2 0           conv5_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_conv (Conv2D)    (None, None, None, 5 1049088     conv5_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_bn (BatchNormali (None, None, None, 5 2048        conv5_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_relu (Activation (None, None, None, 5 0           conv5_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_conv (Conv2D)    (None, None, None, 5 2359808     conv5_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_bn (BatchNormali (None, None, None, 5 2048        conv5_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_relu (Activation (None, None, None, 5 0           conv5_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_3_conv (Conv2D)    (None, None, None, 2 1050624     conv5_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_3_bn (BatchNormali (None, None, None, 2 8192        conv5_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_add (Add)          (None, None, None, 2 0           conv5_block1_out[0][0]           \n",
            "                                                                 conv5_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_out (Activation)   (None, None, None, 2 0           conv5_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_conv (Conv2D)    (None, None, None, 5 1049088     conv5_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_bn (BatchNormali (None, None, None, 5 2048        conv5_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_relu (Activation (None, None, None, 5 0           conv5_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_conv (Conv2D)    (None, None, None, 5 2359808     conv5_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_bn (BatchNormali (None, None, None, 5 2048        conv5_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_relu (Activation (None, None, None, 5 0           conv5_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_3_conv (Conv2D)    (None, None, None, 2 1050624     conv5_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_3_bn (BatchNormali (None, None, None, 2 8192        conv5_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_add (Add)          (None, None, None, 2 0           conv5_block2_out[0][0]           \n",
            "                                                                 conv5_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_out (Activation)   (None, None, None, 2 0           conv5_block3_add[0][0]           \n",
            "==================================================================================================\n",
            "Total params: 23,587,712\n",
            "Trainable params: 23,534,592\n",
            "Non-trainable params: 53,120\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3R56TtDtkWYO",
        "outputId": "dbb37dd4-94d0-45aa-e4e3-4271d890f3d9"
      },
      "source": [
        "for i,layer in enumerate(base_model.layers):\n",
        "  print(i,layer.name)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0 input_2\n",
            "1 conv1_pad\n",
            "2 conv1_conv\n",
            "3 conv1_bn\n",
            "4 conv1_relu\n",
            "5 pool1_pad\n",
            "6 pool1_pool\n",
            "7 conv2_block1_1_conv\n",
            "8 conv2_block1_1_bn\n",
            "9 conv2_block1_1_relu\n",
            "10 conv2_block1_2_conv\n",
            "11 conv2_block1_2_bn\n",
            "12 conv2_block1_2_relu\n",
            "13 conv2_block1_0_conv\n",
            "14 conv2_block1_3_conv\n",
            "15 conv2_block1_0_bn\n",
            "16 conv2_block1_3_bn\n",
            "17 conv2_block1_add\n",
            "18 conv2_block1_out\n",
            "19 conv2_block2_1_conv\n",
            "20 conv2_block2_1_bn\n",
            "21 conv2_block2_1_relu\n",
            "22 conv2_block2_2_conv\n",
            "23 conv2_block2_2_bn\n",
            "24 conv2_block2_2_relu\n",
            "25 conv2_block2_3_conv\n",
            "26 conv2_block2_3_bn\n",
            "27 conv2_block2_add\n",
            "28 conv2_block2_out\n",
            "29 conv2_block3_1_conv\n",
            "30 conv2_block3_1_bn\n",
            "31 conv2_block3_1_relu\n",
            "32 conv2_block3_2_conv\n",
            "33 conv2_block3_2_bn\n",
            "34 conv2_block3_2_relu\n",
            "35 conv2_block3_3_conv\n",
            "36 conv2_block3_3_bn\n",
            "37 conv2_block3_add\n",
            "38 conv2_block3_out\n",
            "39 conv3_block1_1_conv\n",
            "40 conv3_block1_1_bn\n",
            "41 conv3_block1_1_relu\n",
            "42 conv3_block1_2_conv\n",
            "43 conv3_block1_2_bn\n",
            "44 conv3_block1_2_relu\n",
            "45 conv3_block1_0_conv\n",
            "46 conv3_block1_3_conv\n",
            "47 conv3_block1_0_bn\n",
            "48 conv3_block1_3_bn\n",
            "49 conv3_block1_add\n",
            "50 conv3_block1_out\n",
            "51 conv3_block2_1_conv\n",
            "52 conv3_block2_1_bn\n",
            "53 conv3_block2_1_relu\n",
            "54 conv3_block2_2_conv\n",
            "55 conv3_block2_2_bn\n",
            "56 conv3_block2_2_relu\n",
            "57 conv3_block2_3_conv\n",
            "58 conv3_block2_3_bn\n",
            "59 conv3_block2_add\n",
            "60 conv3_block2_out\n",
            "61 conv3_block3_1_conv\n",
            "62 conv3_block3_1_bn\n",
            "63 conv3_block3_1_relu\n",
            "64 conv3_block3_2_conv\n",
            "65 conv3_block3_2_bn\n",
            "66 conv3_block3_2_relu\n",
            "67 conv3_block3_3_conv\n",
            "68 conv3_block3_3_bn\n",
            "69 conv3_block3_add\n",
            "70 conv3_block3_out\n",
            "71 conv3_block4_1_conv\n",
            "72 conv3_block4_1_bn\n",
            "73 conv3_block4_1_relu\n",
            "74 conv3_block4_2_conv\n",
            "75 conv3_block4_2_bn\n",
            "76 conv3_block4_2_relu\n",
            "77 conv3_block4_3_conv\n",
            "78 conv3_block4_3_bn\n",
            "79 conv3_block4_add\n",
            "80 conv3_block4_out\n",
            "81 conv4_block1_1_conv\n",
            "82 conv4_block1_1_bn\n",
            "83 conv4_block1_1_relu\n",
            "84 conv4_block1_2_conv\n",
            "85 conv4_block1_2_bn\n",
            "86 conv4_block1_2_relu\n",
            "87 conv4_block1_0_conv\n",
            "88 conv4_block1_3_conv\n",
            "89 conv4_block1_0_bn\n",
            "90 conv4_block1_3_bn\n",
            "91 conv4_block1_add\n",
            "92 conv4_block1_out\n",
            "93 conv4_block2_1_conv\n",
            "94 conv4_block2_1_bn\n",
            "95 conv4_block2_1_relu\n",
            "96 conv4_block2_2_conv\n",
            "97 conv4_block2_2_bn\n",
            "98 conv4_block2_2_relu\n",
            "99 conv4_block2_3_conv\n",
            "100 conv4_block2_3_bn\n",
            "101 conv4_block2_add\n",
            "102 conv4_block2_out\n",
            "103 conv4_block3_1_conv\n",
            "104 conv4_block3_1_bn\n",
            "105 conv4_block3_1_relu\n",
            "106 conv4_block3_2_conv\n",
            "107 conv4_block3_2_bn\n",
            "108 conv4_block3_2_relu\n",
            "109 conv4_block3_3_conv\n",
            "110 conv4_block3_3_bn\n",
            "111 conv4_block3_add\n",
            "112 conv4_block3_out\n",
            "113 conv4_block4_1_conv\n",
            "114 conv4_block4_1_bn\n",
            "115 conv4_block4_1_relu\n",
            "116 conv4_block4_2_conv\n",
            "117 conv4_block4_2_bn\n",
            "118 conv4_block4_2_relu\n",
            "119 conv4_block4_3_conv\n",
            "120 conv4_block4_3_bn\n",
            "121 conv4_block4_add\n",
            "122 conv4_block4_out\n",
            "123 conv4_block5_1_conv\n",
            "124 conv4_block5_1_bn\n",
            "125 conv4_block5_1_relu\n",
            "126 conv4_block5_2_conv\n",
            "127 conv4_block5_2_bn\n",
            "128 conv4_block5_2_relu\n",
            "129 conv4_block5_3_conv\n",
            "130 conv4_block5_3_bn\n",
            "131 conv4_block5_add\n",
            "132 conv4_block5_out\n",
            "133 conv4_block6_1_conv\n",
            "134 conv4_block6_1_bn\n",
            "135 conv4_block6_1_relu\n",
            "136 conv4_block6_2_conv\n",
            "137 conv4_block6_2_bn\n",
            "138 conv4_block6_2_relu\n",
            "139 conv4_block6_3_conv\n",
            "140 conv4_block6_3_bn\n",
            "141 conv4_block6_add\n",
            "142 conv4_block6_out\n",
            "143 conv5_block1_1_conv\n",
            "144 conv5_block1_1_bn\n",
            "145 conv5_block1_1_relu\n",
            "146 conv5_block1_2_conv\n",
            "147 conv5_block1_2_bn\n",
            "148 conv5_block1_2_relu\n",
            "149 conv5_block1_0_conv\n",
            "150 conv5_block1_3_conv\n",
            "151 conv5_block1_0_bn\n",
            "152 conv5_block1_3_bn\n",
            "153 conv5_block1_add\n",
            "154 conv5_block1_out\n",
            "155 conv5_block2_1_conv\n",
            "156 conv5_block2_1_bn\n",
            "157 conv5_block2_1_relu\n",
            "158 conv5_block2_2_conv\n",
            "159 conv5_block2_2_bn\n",
            "160 conv5_block2_2_relu\n",
            "161 conv5_block2_3_conv\n",
            "162 conv5_block2_3_bn\n",
            "163 conv5_block2_add\n",
            "164 conv5_block2_out\n",
            "165 conv5_block3_1_conv\n",
            "166 conv5_block3_1_bn\n",
            "167 conv5_block3_1_relu\n",
            "168 conv5_block3_2_conv\n",
            "169 conv5_block3_2_bn\n",
            "170 conv5_block3_2_relu\n",
            "171 conv5_block3_3_conv\n",
            "172 conv5_block3_3_bn\n",
            "173 conv5_block3_add\n",
            "174 conv5_block3_out\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jZmaixKXkb_-"
      },
      "source": [
        "base_model.trainable=False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gbm_qgB7kfME"
      },
      "source": [
        "global_avg_pool=keras.layers.GlobalAveragePooling2D()(base_model.output)\n",
        "prediction=keras.layers.Dense(1,activation=\"sigmoid\")(global_avg_pool)\n",
        "resnet_model=keras.models.Model(inputs=base_model.input,outputs=prediction)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C1p3SMycklN7"
      },
      "source": [
        "train_data_gen=ImageDataGenerator(rescale=1/255.0,horizontal_flip=True,\n",
        "                                  zoom_range=0.3,brightness_range=[0.9,1.2],\n",
        "                                  shear_range=0.3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wdO5temlknwu"
      },
      "source": [
        "test_data_gen=ImageDataGenerator(rescale=1/255.0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZtWzp4Avlgyw",
        "outputId": "26a2a2b4-b476-4a40-952c-e5d34a827375"
      },
      "source": [
        "import os\n",
        "os.chdir(\"/content/drive/My Drive/dataset_resized\")\n",
        "!ls"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "testing  training\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nVNNvM9ukpVo",
        "outputId": "c3533593-c671-43df-cd99-022f7e9f25e3"
      },
      "source": [
        "train_data=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/training\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2414 images belonging to 7 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "peE4nu9Om4tB",
        "outputId": "875c7757-60b2-4f90-95d2-d4a813d4db3e"
      },
      "source": [
        "test_data=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/testing\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 608 images belonging to 7 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3xJaAWm7l2Bq"
      },
      "source": [
        "resnet_model.compile(optimizer=\"adam\",loss=\"categorical_crossentropy\",metrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m3hWtB9Impvo"
      },
      "source": [
        "model_new=keras.models.Sequential([\n",
        "                                   base_model,\n",
        "                                   keras.layers.GlobalAveragePooling2D(),\n",
        "                                   keras.layers.Dense(1024,activation=\"relu\"),\n",
        "                                   keras.layers.Dense(1024,activation=\"relu\"),\n",
        "                                   keras.layers.Dense(1024,activation=\"relu\"),\n",
        "                                   keras.layers.Dense(512,activation=\"relu\"),\n",
        "                                   keras.layers.Dense(1,activation=\"softmax\")\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "89S4lWIHmtXt"
      },
      "source": [
        "model_new.compile(optimizer=\"adam\",loss=\"categorical_crossentropy\",metrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KiBB7pEL_JZF"
      },
      "source": [
        "import numpy as np\n",
        "from datetime import datetime\n",
        "import time\n",
        "#from sklearn import tree\n",
        "#from sklearn.preprocessing import normalize, scale\n",
        "from sklearn.metrics import accuracy_score\n",
        "fmt = '%H:%M:%S'\n",
        "def get_current_time():\n",
        "    time.ctime()\n",
        "    return time.strftime(fmt)\n",
        "first = get_current_time()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uPDxhDHC_LAq"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6V2OhW1n_Ouk"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pcRW_6szmwBv",
        "outputId": "91a98c48-92c0-4781-9f61-c020fda4e725"
      },
      "source": [
        "early_stop=keras.callbacks.EarlyStopping(patience=5,mode=\"max\",monitor=\"val_accuracy\",verbose=2)\n",
        "check=keras.callbacks.ModelCheckpoint(mode=\"max\",monitor=\"val_accuracy\",verbose=2,filepath=\"resnet50.h5\",save_best_only=True)\n",
        "model_new.fit(train_data,validation_data=(test_data),epochs=50,callbacks=[early_stop,check])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "76/76 [==============================] - 1380s 18s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00001: val_accuracy improved from -inf to 0.16447, saving model to resnet50.h5\n",
            "Epoch 2/50\n",
            "76/76 [==============================] - 425s 6s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00002: val_accuracy did not improve from 0.16447\n",
            "Epoch 3/50\n",
            "76/76 [==============================] - 410s 5s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00003: val_accuracy did not improve from 0.16447\n",
            "Epoch 4/50\n",
            "76/76 [==============================] - 411s 5s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00004: val_accuracy did not improve from 0.16447\n",
            "Epoch 5/50\n",
            "76/76 [==============================] - 429s 6s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00005: val_accuracy did not improve from 0.16447\n",
            "Epoch 6/50\n",
            "76/76 [==============================] - 440s 6s/step - loss: 0.0000e+00 - accuracy: 0.1636 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00006: val_accuracy did not improve from 0.16447\n",
            "Epoch 00006: early stopping\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fa7f862edd8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LQFjO-QHA2ba",
        "outputId": "3cb3503d-f62b-433f-dae0-77f5dddcc7df"
      },
      "source": [
        "\n",
        "second = get_current_time()\n",
        "print(\"Time taken to train(sec):\", datetime.strptime(second, fmt) - datetime.strptime(first, fmt))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Time taken to train(sec): 0:58:31\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zvUneNSnm7oX"
      },
      "source": [
        "model=keras.models.load_model(\"/content/drive/My Drive/dataset_resized/resnet50.h5\")\n",
        "#model.compile(optimizer=\"adam\",loss=\"binary_crossentropy\",metrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YkO9Vap8uoFT"
      },
      "source": [
        "model.compile(optimizer=\"adam\",loss=\"categorical_crossentropy\",metrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Qq8-yBzZx5sT",
        "outputId": "88fe32bb-9b70-47ab-fde3-3d32975bd246"
      },
      "source": [
        "model.evaluate(test_data)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "19/19 [==============================] - 88s 5s/step - loss: 0.0000e+00 - accuracy: 0.1785\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.0, 0.16447368264198303]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-vv-edlq1WhE"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RUaYq3ss6ON5"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8qcdzNO06OTJ"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MeAQz6q56OWf"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fD_dI9plOc94"
      },
      "source": [
        "# Now Let's Apply Mobilenet Architechture"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pJzVByke6fPd",
        "outputId": "e1607154-f9a6-45e5-f6fc-b6ee2bfa12c2"
      },
      "source": [
        "base_model=keras.applications.MobileNetV2(include_top=False,input_shape=(224,224,3),weights=\"imagenet\") "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet_v2/mobilenet_v2_weights_tf_dim_ordering_tf_kernels_1.0_224_no_top.h5\n",
            "9412608/9406464 [==============================] - 0s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XMb39-6tOe1o"
      },
      "source": [
        "base_model.trainable=False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fNIU5MoWOg3p"
      },
      "source": [
        "global_avg_pool=keras.layers.GlobalAveragePooling2D()(base_model.output)\n",
        "prediction=keras.layers.Dense(1,activation=\"sigmoid\")(global_avg_pool)\n",
        "mobile_model=keras.models.Model(inputs=base_model.input,outputs=prediction)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6BH_SPEuOiz_",
        "outputId": "e1a31ff0-59c3-4b33-c074-b7f166cde5d2"
      },
      "source": [
        "mobile_model.compile(optimizer=\"adam\",loss=\"categorical_crossentropy\",metrics=[\"accuracy\"])\n",
        "early_stop=keras.callbacks.EarlyStopping(patience=5,mode=\"max\",monitor=\"val_accuracy\",verbose=2)\n",
        "check=keras.callbacks.ModelCheckpoint(mode=\"max\",monitor=\"val_accuracy\",verbose=2,filepath=\"My_model.h5\",save_best_only=True)\n",
        "mobile_model.fit(train_data,validation_data=(test_data),epochs=50,callbacks=[early_stop,check])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "76/76 [==============================] - 114s 1s/step - loss: 0.0000e+00 - accuracy: 0.1699 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00001: val_accuracy improved from -inf to 0.16447, saving model to My_model.h5\n",
            "Epoch 2/50\n",
            "76/76 [==============================] - 113s 1s/step - loss: 0.0000e+00 - accuracy: 0.1703 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00002: val_accuracy did not improve from 0.16447\n",
            "Epoch 3/50\n",
            "76/76 [==============================] - 114s 1s/step - loss: 0.0000e+00 - accuracy: 0.1767 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00003: val_accuracy did not improve from 0.16447\n",
            "Epoch 4/50\n",
            "76/76 [==============================] - 116s 2s/step - loss: 0.0000e+00 - accuracy: 0.1537 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00004: val_accuracy did not improve from 0.16447\n",
            "Epoch 5/50\n",
            "76/76 [==============================] - 116s 2s/step - loss: 0.0000e+00 - accuracy: 0.1695 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00005: val_accuracy did not improve from 0.16447\n",
            "Epoch 6/50\n",
            "76/76 [==============================] - 116s 2s/step - loss: 0.0000e+00 - accuracy: 0.1676 - val_loss: 0.0000e+00 - val_accuracy: 0.1645\n",
            "\n",
            "Epoch 00006: val_accuracy did not improve from 0.16447\n",
            "Epoch 00006: early stopping\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fa830690d30>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oJWwhyHWPCWG"
      },
      "source": [
        "model=keras.models.load_model(\"/content/drive/My Drive/dataset_resized/My_model.h5\")\n",
        "model.compile(optimizer=\"adam\",loss=\"categorical_crossentropy\",metrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8dR_SN2xShdP",
        "outputId": "dfb39db4-83ba-4bb5-c19a-0f668db33c35"
      },
      "source": [
        "model.evaluate(test_data)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "19/19 [==============================] - 24s 1s/step - loss: 0.0000e+00 - accuracy: 0.1565\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.0, 0.16447368264198303]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SWGYvCQDSprF",
        "outputId": "9fb088f4-8897-4338-8015-c314220a3c6b"
      },
      "source": [
        "test_data.class_indices"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'cardboard': 0,\n",
              " 'ewaste': 1,\n",
              " 'glass': 2,\n",
              " 'metal': 3,\n",
              " 'paper': 4,\n",
              " 'plastic': 5,\n",
              " 'trash': 6}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n2ym14FBTfxd"
      },
      "source": [
        "def predict_image_class(url):\n",
        "  image=keras.preprocessing.image.load_img(url,target_size=(224,224))\n",
        "  img=img_to_array(image)\n",
        "  img=img/255.0\n",
        "  pred_img=img[np.newaxis]\n",
        "  plt.imshow(img)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  pred=model.predict(pred_img)\n",
        "  if pred <0.5:\n",
        "    plt.xlabel(\"glass\")\n",
        "  else:\n",
        "    plt.xlabel(\"paper\")\n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 262
        },
        "id": "9-64YEB6Tl4U",
        "outputId": "35fb3fb7-deec-479a-94b6-6e45ecbe8bbb"
      },
      "source": [
        "predict_image_class(\"/content/drive/My Drive/dataset_resized/testing/glass/glass12.jpg\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PA0pgWe7T1-U"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XeYlc-awLIs2"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PaQsC3M_LIpP"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jd-6SxhPLIl2"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lsHNOAoCLIjh"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sovx2YhdLIgo"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PKb3idtnLIdC"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j7AZdFpBLIaZ"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UJMu9CNVLIXA",
        "outputId": "b090e876-730f-4997-8e1b-f1eb0c8ee7d0"
      },
      "source": [
        "# TensorFlow and tf.keras\n",
        "import tensorflow as tf\n",
        "\n",
        "# Helper libraries\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "print(tf.__version__)\n"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.3.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a7P7BmlSLRbL",
        "outputId": "516e465e-20d0-4627-da2d-06b330d66a86"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dkxcz0mrLRW1"
      },
      "source": [
        "dir = '/content/drive/My Drive/dataset_resized/'"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fK7UiWeALRS1",
        "outputId": "ca2cea0d-2850-4904-e347-4411a897f0bf"
      },
      "source": [
        "from keras.preprocessing.image import ImageDataGenerator,load_img,img_to_array\n",
        "\n",
        "train_data_gen=ImageDataGenerator(rescale=1/255.0,horizontal_flip=True,\n",
        "                                  zoom_range=0.3,brightness_range=[0.9,1.2],\n",
        "                                  shear_range=0.3)\n",
        "test_data_gen=ImageDataGenerator(rescale=1/255.0)\n",
        "\n",
        "train_data=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/training\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")\n",
        "test_data=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/testing\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2414 images belonging to 7 classes.\n",
            "Found 608 images belonging to 7 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y6v6WvBvhTZD"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J1A7DOHmhTW_"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5vO9_gdhTS3"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4JvC6axiPtOt"
      },
      "source": [
        "CNN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DY0IJrsrPvfb"
      },
      "source": [
        "import h5py\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ycWrYIY8P5PH"
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.models import load_model,Model,Sequential\n",
        "from keras.preprocessing.image import ImageDataGenerator,array_to_img,load_img\n",
        "from keras.applications import MobileNetV2,ResNet50V2,VGG16"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a6bahTRtQBVq"
      },
      "source": [
        "train_data_gen=ImageDataGenerator(rescale=1/255.0,horizontal_flip=True,zoom_range=0.2,shear_range=0.2,brightness_range=[0.9,1.4])"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fTbA6-MbQTtO",
        "outputId": "49109643-f475-479c-c3ad-f294ab1c1350"
      },
      "source": [
        "train_file=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/training\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")\n",
        "test_file=test_data_gen.flow_from_directory(\"/content/drive/My Drive/dataset_resized/testing\",batch_size=32,target_size=(224,224),shuffle=True,class_mode=\"binary\")"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2414 images belonging to 7 classes.\n",
            "Found 608 images belonging to 7 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T99osfnzWsoR"
      },
      "source": [
        "zip kaj e lagenai"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PRZ9F51IS_wI"
      },
      "source": [
        "cardboard_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/cardboard')\n",
        "paper_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/paper')\n",
        "plastic_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/plastic')\n",
        "glass_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/glass')\n",
        "metal_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/metal')\n",
        "trash_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/trash')\n",
        "ewaste_dir = os.path.join('/content/drive/MyDrive/dataset_resized/training/ewaste')\n"
      ],
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sU3pyr0XU9hp",
        "outputId": "992d8ada-a927-4c6d-ab8c-0ec7cd85a748"
      },
      "source": [
        "print('total training cardboard images:', len(os.listdir(cardboard_dir)))\n",
        "print('total training paper images:', len(os.listdir(paper_dir)))\n",
        "print('total training plastic images:', len(os.listdir(plastic_dir)))\n",
        "print('total training glass images:', len(os.listdir(glass_dir)))\n",
        "print('total training metal images:', len(os.listdir(metal_dir)))\n",
        "print('total training trash images:', len(os.listdir(trash_dir )))\n",
        "print('total training ewaste images:', len(os.listdir(ewaste_dir )))"
      ],
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "total training cardboard images: 322\n",
            "total training paper images: 475\n",
            "total training plastic images: 385\n",
            "total training glass images: 400\n",
            "total training metal images: 328\n",
            "total training trash images: 109\n",
            "total training ewaste images: 400\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kDb2WrffVE5m",
        "outputId": "d36cc609-b45e-43e6-9201-95fec5b696d8"
      },
      "source": [
        "ls"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mdrive\u001b[0m/  \u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bw4peiE8VFhv",
        "outputId": "29f54e63-17ac-46eb-e9af-413711b285a0"
      },
      "source": [
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator\n",
        "\n",
        "TRAINING_DIR = \"/content/drive/MyDrive/dataset_resized/training\"\n",
        "training_datagen = ImageDataGenerator(\n",
        "      rescale = 1./255,\n",
        "\t    rotation_range=40,\n",
        "      width_shift_range=0.2,\n",
        "      height_shift_range=0.2,\n",
        "      shear_range=0.2,\n",
        "      zoom_range=0.2,\n",
        "      horizontal_flip=True,\n",
        "      fill_mode='nearest')\n",
        "\n",
        "TRAINING_DIR = \"/content/drive/MyDrive/dataset_resized/training\"\n",
        "validation_datagen = ImageDataGenerator(rescale = 1./255)\n",
        "\n",
        "train_generator = training_datagen.flow_from_directory(\n",
        "\tTRAINING_DIR,\n",
        "\ttarget_size=(150,150),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size=126\n",
        ")\n",
        "\n",
        "validation_generator = validation_datagen.flow_from_directory(\n",
        "\tVALIDATION_DIR,\n",
        "\ttarget_size=(150,150),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size=126\n",
        ")\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "    # Note the input shape is the desired size of the image 150x150 with 3 bytes color\n",
        "    # This is the first convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(150, 150, 3)),\n",
        "    tf.keras.layers.MaxPooling2D(2, 2),\n",
        "    # The second convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # The third convolution\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # The fourth convolution\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Flatten the results to feed into a DNN\n",
        "    tf.keras.layers.Flatten(),\n",
        "    tf.keras.layers.Dropout(0.5),\n",
        "    # 512 neuron hidden layer\n",
        "    tf.keras.layers.Dense(512, activation='relu'),\n",
        "    tf.keras.layers.Dense(7, activation='softmax')\n",
        "])\n",
        "\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(loss = 'categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])\n",
        "\n",
        "history = model.fit(train_generator, epochs=25, steps_per_epoch=20, validation_data = validation_generator, verbose = 1, validation_steps=3)\n",
        "\n",
        "model.save(\"rps.h5\")\n"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2414 images belonging to 7 classes.\n",
            "Found 608 images belonging to 7 classes.\n",
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_8 (Conv2D)            (None, 148, 148, 64)      1792      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_8 (MaxPooling2 (None, 74, 74, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_9 (Conv2D)            (None, 72, 72, 64)        36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_9 (MaxPooling2 (None, 36, 36, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_10 (Conv2D)           (None, 34, 34, 128)       73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_10 (MaxPooling (None, 17, 17, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_11 (Conv2D)           (None, 15, 15, 128)       147584    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_11 (MaxPooling (None, 7, 7, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 6272)              0         \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 6272)              0         \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 512)               3211776   \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 7)                 3591      \n",
            "=================================================================\n",
            "Total params: 3,475,527\n",
            "Trainable params: 3,475,527\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/25\n",
            "20/20 [==============================] - 1351s 68s/step - loss: 2.4053 - accuracy: 0.1918 - val_loss: 1.9252 - val_accuracy: 0.1958\n",
            "Epoch 2/25\n",
            "20/20 [==============================] - 81s 4s/step - loss: 1.8994 - accuracy: 0.2067 - val_loss: 2.0567 - val_accuracy: 0.1772\n",
            "Epoch 3/25\n",
            "20/20 [==============================] - 39s 2s/step - loss: 1.7848 - accuracy: 0.2523 - val_loss: 2.3608 - val_accuracy: 0.2196\n",
            "Epoch 4/25\n",
            "20/20 [==============================] - 31s 2s/step - loss: 1.7311 - accuracy: 0.2991 - val_loss: 1.6845 - val_accuracy: 0.3095\n",
            "Epoch 5/25\n",
            "20/20 [==============================] - 30s 2s/step - loss: 1.7980 - accuracy: 0.3268 - val_loss: 1.6546 - val_accuracy: 0.3677\n",
            "Epoch 6/25\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.6443 - accuracy: 0.3492 - val_loss: 1.7797 - val_accuracy: 0.2778\n",
            "Epoch 7/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.6125 - accuracy: 0.3728 - val_loss: 1.6752 - val_accuracy: 0.3995\n",
            "Epoch 8/25\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.5374 - accuracy: 0.4080 - val_loss: 1.3407 - val_accuracy: 0.4815\n",
            "Epoch 9/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.4393 - accuracy: 0.4358 - val_loss: 1.5160 - val_accuracy: 0.4392\n",
            "Epoch 10/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.5480 - accuracy: 0.4292 - val_loss: 1.2922 - val_accuracy: 0.5000\n",
            "Epoch 11/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.3714 - accuracy: 0.4714 - val_loss: 1.3988 - val_accuracy: 0.4550\n",
            "Epoch 12/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.4018 - accuracy: 0.4785 - val_loss: 1.2993 - val_accuracy: 0.4762\n",
            "Epoch 13/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.3558 - accuracy: 0.5008 - val_loss: 1.3849 - val_accuracy: 0.4974\n",
            "Epoch 14/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.3588 - accuracy: 0.4946 - val_loss: 1.6291 - val_accuracy: 0.4127\n",
            "Epoch 15/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.3191 - accuracy: 0.4959 - val_loss: 1.3914 - val_accuracy: 0.4603\n",
            "Epoch 16/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2655 - accuracy: 0.5166 - val_loss: 1.3748 - val_accuracy: 0.4762\n",
            "Epoch 17/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2180 - accuracy: 0.5365 - val_loss: 1.3345 - val_accuracy: 0.5026\n",
            "Epoch 18/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2030 - accuracy: 0.5547 - val_loss: 1.4766 - val_accuracy: 0.4497\n",
            "Epoch 19/25\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.2115 - accuracy: 0.5580 - val_loss: 1.4372 - val_accuracy: 0.4683\n",
            "Epoch 20/25\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.1508 - accuracy: 0.5775 - val_loss: 1.1781 - val_accuracy: 0.5767\n",
            "Epoch 21/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2138 - accuracy: 0.5592 - val_loss: 0.9586 - val_accuracy: 0.6349\n",
            "Epoch 22/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.1515 - accuracy: 0.5671 - val_loss: 1.2196 - val_accuracy: 0.5926\n",
            "Epoch 23/25\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.0988 - accuracy: 0.6011 - val_loss: 1.0894 - val_accuracy: 0.6058\n",
            "Epoch 24/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2859 - accuracy: 0.5431 - val_loss: 1.0428 - val_accuracy: 0.6323\n",
            "Epoch 25/25\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0508 - accuracy: 0.6123 - val_loss: 1.2403 - val_accuracy: 0.5794\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NcZv_bS1k_ws"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "qk1L8rE4Xx5M",
        "outputId": "a6b603fd-06ff-4364-ffb7-e40eb16b6b8c"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "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"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 106
        },
        "id": "i7LNbPdyjPls",
        "outputId": "093f13ea-2c41-475e-e5c7-b9c5683b6c5a"
      },
      "source": [
        "import numpy as np\n",
        "from google.colab import files\n",
        "from keras.preprocessing import image\n",
        "\n",
        "uploaded = files.upload()\n",
        "\n",
        "for fn in uploaded.keys():\n",
        " \n",
        "  # predicting images\n",
        "  path = fn\n",
        "  img = image.load_img(path, target_size=(150, 150))\n",
        "  x = image.img_to_array(img)\n",
        "  x = np.expand_dims(x, axis=0)\n",
        "\n",
        "  images = np.vstack([x])\n",
        "  classes = model.predict(images, batch_size=10)\n",
        "  print(fn)\n",
        "  print(classes)"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-11827873-5f36-456f-89fd-0adc19883b88\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-11827873-5f36-456f-89fd-0adc19883b88\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving ewaste (60).JPG to ewaste (60).JPG\n",
            "ewaste (60).JPG\n",
            "[[0. 1. 0. 0. 0. 0. 0.]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8Vf6uC7TjaOj"
      },
      "source": [
        "So it classifies correctly"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-wNRnq-xj1ic"
      },
      "source": [
        "import numpy as np\n",
        "from datetime import datetime\n",
        "import time\n",
        "from sklearn import tree\n",
        "from sklearn.preprocessing import normalize, scale\n",
        "from sklearn.metrics import accuracy_score\n",
        "fmt = '%H:%M:%S'\n",
        "\n",
        "def get_current_time():\n",
        "    time.ctime()\n",
        "    return time.strftime(fmt)"
      ],
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7lEnHtCAlBdJ",
        "outputId": "8729d2c3-85a4-4a8a-e808-5d9b950e3658"
      },
      "source": [
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator\n",
        "\n",
        "TRAINING_DIR = \"/content/drive/MyDrive/dataset_resized/training\"\n",
        "training_datagen = ImageDataGenerator(\n",
        "      rescale = 1./255,\n",
        "\t    rotation_range=40,\n",
        "      width_shift_range=0.2,\n",
        "      height_shift_range=0.2,\n",
        "      shear_range=0.2,\n",
        "      zoom_range=0.2,\n",
        "      horizontal_flip=True,\n",
        "      fill_mode='nearest')\n",
        "\n",
        "VALIDATION_DIR = \"/content/drive/MyDrive/dataset_resized/testing\"\n",
        "validation_datagen = ImageDataGenerator(rescale = 1./255)\n",
        "\n",
        "train_generator = training_datagen.flow_from_directory(\n",
        "\tTRAINING_DIR,\n",
        "\ttarget_size=(150,150),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size=126\n",
        ")\n",
        "\n",
        "validation_generator = validation_datagen.flow_from_directory(\n",
        "\tVALIDATION_DIR,\n",
        "\ttarget_size=(150,150),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size=126\n",
        ")\n",
        "\n",
        "model = tf.keras.models.Sequential([\n",
        "    # Note the input shape is the desired size of the image 150x150 with 3 bytes color\n",
        "    # This is the first convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(150, 150, 3)),\n",
        "    tf.keras.layers.MaxPooling2D(2, 2),\n",
        "    # The second convolution\n",
        "    tf.keras.layers.Conv2D(64, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # The third convolution\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # The fourth convolution\n",
        "    tf.keras.layers.Conv2D(128, (3,3), activation='relu'),\n",
        "    tf.keras.layers.MaxPooling2D(2,2),\n",
        "    # Flatten the results to feed into a DNN\n",
        "    tf.keras.layers.Flatten(),\n",
        "    tf.keras.layers.Dropout(0.5),\n",
        "    # 512 neuron hidden layer\n",
        "    tf.keras.layers.Dense(512, activation='relu'),\n",
        "    tf.keras.layers.Dense(7, activation='softmax')\n",
        "])\n",
        "\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(loss = 'categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
        "\n",
        "history = model.fit(train_generator, epochs=50, steps_per_epoch=20, validation_data = validation_generator, verbose = 1, validation_steps=3)\n",
        "\n",
        "model.save(\"new_cnn.h5\")\n"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 2414 images belonging to 7 classes.\n",
            "Found 608 images belonging to 7 classes.\n",
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_12 (Conv2D)           (None, 148, 148, 64)      1792      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_12 (MaxPooling (None, 74, 74, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_13 (Conv2D)           (None, 72, 72, 64)        36928     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_13 (MaxPooling (None, 36, 36, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_14 (Conv2D)           (None, 34, 34, 128)       73856     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_14 (MaxPooling (None, 17, 17, 128)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_15 (Conv2D)           (None, 15, 15, 128)       147584    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_15 (MaxPooling (None, 7, 7, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten_3 (Flatten)          (None, 6272)              0         \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 6272)              0         \n",
            "_________________________________________________________________\n",
            "dense_6 (Dense)              (None, 512)               3211776   \n",
            "_________________________________________________________________\n",
            "dense_7 (Dense)              (None, 7)                 3591      \n",
            "=================================================================\n",
            "Total params: 3,475,527\n",
            "Trainable params: 3,475,527\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/50\n",
            "20/20 [==============================] - 31s 2s/step - loss: 1.8090 - accuracy: 0.2394 - val_loss: 1.7830 - val_accuracy: 0.3254\n",
            "Epoch 2/50\n",
            "20/20 [==============================] - 30s 2s/step - loss: 1.5942 - accuracy: 0.3749 - val_loss: 1.5777 - val_accuracy: 0.3677\n",
            "Epoch 3/50\n",
            "20/20 [==============================] - 30s 2s/step - loss: 1.5039 - accuracy: 0.4043 - val_loss: 1.5963 - val_accuracy: 0.4550\n",
            "Epoch 4/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.3778 - accuracy: 0.4631 - val_loss: 1.3022 - val_accuracy: 0.4815\n",
            "Epoch 5/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.3382 - accuracy: 0.4789 - val_loss: 1.4854 - val_accuracy: 0.4656\n",
            "Epoch 6/50\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.2778 - accuracy: 0.5000 - val_loss: 1.3239 - val_accuracy: 0.5106\n",
            "Epoch 7/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2502 - accuracy: 0.5207 - val_loss: 1.3192 - val_accuracy: 0.4974\n",
            "Epoch 8/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.2551 - accuracy: 0.5120 - val_loss: 1.1803 - val_accuracy: 0.5820\n",
            "Epoch 9/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.1950 - accuracy: 0.5410 - val_loss: 1.2600 - val_accuracy: 0.5529\n",
            "Epoch 10/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.1816 - accuracy: 0.5435 - val_loss: 1.1044 - val_accuracy: 0.5794\n",
            "Epoch 11/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0977 - accuracy: 0.5866 - val_loss: 1.1835 - val_accuracy: 0.5688\n",
            "Epoch 12/50\n",
            "20/20 [==============================] - 30s 2s/step - loss: 1.1334 - accuracy: 0.5737 - val_loss: 1.0803 - val_accuracy: 0.6164\n",
            "Epoch 13/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0773 - accuracy: 0.5920 - val_loss: 1.2587 - val_accuracy: 0.5582\n",
            "Epoch 14/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0841 - accuracy: 0.5944 - val_loss: 1.0538 - val_accuracy: 0.6138\n",
            "Epoch 15/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.0570 - accuracy: 0.5978 - val_loss: 1.1132 - val_accuracy: 0.6085\n",
            "Epoch 16/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9868 - accuracy: 0.6330 - val_loss: 1.0702 - val_accuracy: 0.6032\n",
            "Epoch 17/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0081 - accuracy: 0.6164 - val_loss: 1.0183 - val_accuracy: 0.6429\n",
            "Epoch 18/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9564 - accuracy: 0.6396 - val_loss: 1.0848 - val_accuracy: 0.6058\n",
            "Epoch 19/50\n",
            "20/20 [==============================] - 30s 1s/step - loss: 0.9566 - accuracy: 0.6462 - val_loss: 1.0504 - val_accuracy: 0.6296\n",
            "Epoch 20/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9473 - accuracy: 0.6504 - val_loss: 1.1327 - val_accuracy: 0.6085\n",
            "Epoch 21/50\n",
            "20/20 [==============================] - 30s 1s/step - loss: 0.9167 - accuracy: 0.6636 - val_loss: 0.8480 - val_accuracy: 0.6614\n",
            "Epoch 22/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9508 - accuracy: 0.6537 - val_loss: 0.9853 - val_accuracy: 0.6508\n",
            "Epoch 23/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8822 - accuracy: 0.6831 - val_loss: 1.0149 - val_accuracy: 0.6640\n",
            "Epoch 24/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8473 - accuracy: 0.6831 - val_loss: 1.0703 - val_accuracy: 0.6402\n",
            "Epoch 25/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8273 - accuracy: 0.6856 - val_loss: 0.8354 - val_accuracy: 0.6905\n",
            "Epoch 26/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8520 - accuracy: 0.6839 - val_loss: 0.8729 - val_accuracy: 0.6667\n",
            "Epoch 27/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8198 - accuracy: 0.6914 - val_loss: 0.9518 - val_accuracy: 0.6402\n",
            "Epoch 28/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8463 - accuracy: 0.6893 - val_loss: 1.0628 - val_accuracy: 0.6402\n",
            "Epoch 29/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8337 - accuracy: 0.6993 - val_loss: 0.7517 - val_accuracy: 0.7275\n",
            "Epoch 30/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7409 - accuracy: 0.7328 - val_loss: 0.8319 - val_accuracy: 0.6931\n",
            "Epoch 31/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8389 - accuracy: 0.6901 - val_loss: 0.8821 - val_accuracy: 0.6587\n",
            "Epoch 32/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7891 - accuracy: 0.7013 - val_loss: 0.9580 - val_accuracy: 0.6455\n",
            "Epoch 33/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7216 - accuracy: 0.7291 - val_loss: 0.9791 - val_accuracy: 0.6481\n",
            "Epoch 34/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7669 - accuracy: 0.7113 - val_loss: 0.8084 - val_accuracy: 0.7037\n",
            "Epoch 35/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7179 - accuracy: 0.7345 - val_loss: 0.8294 - val_accuracy: 0.6984\n",
            "Epoch 36/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7685 - accuracy: 0.7204 - val_loss: 0.7923 - val_accuracy: 0.7169\n",
            "Epoch 37/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7128 - accuracy: 0.7361 - val_loss: 0.8434 - val_accuracy: 0.7090\n",
            "Epoch 38/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6881 - accuracy: 0.7531 - val_loss: 0.8405 - val_accuracy: 0.6984\n",
            "Epoch 39/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6764 - accuracy: 0.7535 - val_loss: 0.8285 - val_accuracy: 0.7090\n",
            "Epoch 40/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6348 - accuracy: 0.7635 - val_loss: 0.7202 - val_accuracy: 0.7513\n",
            "Epoch 41/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6285 - accuracy: 0.7713 - val_loss: 0.8006 - val_accuracy: 0.7116\n",
            "Epoch 42/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6275 - accuracy: 0.7659 - val_loss: 0.7931 - val_accuracy: 0.7302\n",
            "Epoch 43/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6890 - accuracy: 0.7394 - val_loss: 0.6895 - val_accuracy: 0.7513\n",
            "Epoch 44/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6460 - accuracy: 0.7606 - val_loss: 0.8961 - val_accuracy: 0.6429\n",
            "Epoch 45/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6051 - accuracy: 0.7800 - val_loss: 0.7533 - val_accuracy: 0.7540\n",
            "Epoch 46/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5775 - accuracy: 0.7883 - val_loss: 0.8651 - val_accuracy: 0.7196\n",
            "Epoch 47/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5884 - accuracy: 0.7784 - val_loss: 0.7334 - val_accuracy: 0.7540\n",
            "Epoch 48/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5828 - accuracy: 0.7904 - val_loss: 0.6788 - val_accuracy: 0.7672\n",
            "Epoch 49/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5920 - accuracy: 0.7829 - val_loss: 0.6546 - val_accuracy: 0.7540\n",
            "Epoch 50/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6606 - accuracy: 0.7577 - val_loss: 0.7592 - val_accuracy: 0.7328\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "anPtSxxntXLa",
        "outputId": "1a467a64-7d1a-42bd-c1ad-5f5824a37f16"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "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"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 106
        },
        "id": "dvYHrVtBlBX3",
        "outputId": "b153fdb9-4eed-4b1d-96d4-5c404c914fa4"
      },
      "source": [
        "import numpy as np\n",
        "from google.colab import files\n",
        "from keras.preprocessing import image\n",
        "\n",
        "uploaded = files.upload()\n",
        "\n",
        "for fn in uploaded.keys():\n",
        " \n",
        "  # predicting images\n",
        "  path = fn\n",
        "  img = image.load_img(path, target_size=(150, 150))\n",
        "  x = image.img_to_array(img)\n",
        "  x = np.expand_dims(x, axis=0)\n",
        "\n",
        "  images = np.vstack([x])\n",
        "  classes = model.predict(images, batch_size=10)\n",
        "  print(fn)\n",
        "  print(classes)"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-30ba38b1-363b-4ef0-b5a4-09769592bc25\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-30ba38b1-363b-4ef0-b5a4-09769592bc25\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving ewaste (256).jpg to ewaste (256).jpg\n",
            "ewaste (256).jpg\n",
            "[[0. 1. 0. 0. 0. 0. 0.]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sy9GUc0plBIs"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qyUJf7emlA8a"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nHzQQM5JsnBL"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5-sC9FYGsnf1"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xWgju4WnsnrT"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "61pPmemdsn4v"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DioDPGeUszll"
      },
      "source": [
        "Lets try Mobilenet V2."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DNaSFsUvsoHV",
        "outputId": "5d7ba046-36b7-496d-ee95-644590459c4e"
      },
      "source": [
        "\n",
        "# Create the base model from the pre-trained model MobileNet V2\n",
        "base_model = tf.keras.applications.MobileNetV2(input_shape=(150, 150, 3),\n",
        "                                               include_top=False,\n",
        "                                               weights='imagenet')"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [96, 128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gYjm3N2usoOn",
        "outputId": "48f7abd2-006b-4829-c2a6-13cabc2a2075"
      },
      "source": [
        "base_model.summary()"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"mobilenetv2_1.00_224\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_3 (InputLayer)            [(None, 150, 150, 3) 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "Conv1_pad (ZeroPadding2D)       (None, 151, 151, 3)  0           input_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "Conv1 (Conv2D)                  (None, 75, 75, 32)   864         Conv1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "bn_Conv1 (BatchNormalization)   (None, 75, 75, 32)   128         Conv1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "Conv1_relu (ReLU)               (None, 75, 75, 32)   0           bn_Conv1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise (Depthw (None, 75, 75, 32)   288         Conv1_relu[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise_BN (Bat (None, 75, 75, 32)   128         expanded_conv_depthwise[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise_relu (R (None, 75, 75, 32)   0           expanded_conv_depthwise_BN[0][0] \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_project (Conv2D)  (None, 75, 75, 16)   512         expanded_conv_depthwise_relu[0][0\n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_project_BN (Batch (None, 75, 75, 16)   64          expanded_conv_project[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand (Conv2D)         (None, 75, 75, 96)   1536        expanded_conv_project_BN[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand_BN (BatchNormali (None, 75, 75, 96)   384         block_1_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand_relu (ReLU)      (None, 75, 75, 96)   0           block_1_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_1_pad (ZeroPadding2D)     (None, 77, 77, 96)   0           block_1_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise (DepthwiseCon (None, 38, 38, 96)   864         block_1_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise_BN (BatchNorm (None, 38, 38, 96)   384         block_1_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise_relu (ReLU)   (None, 38, 38, 96)   0           block_1_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_1_project (Conv2D)        (None, 38, 38, 24)   2304        block_1_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_1_project_BN (BatchNormal (None, 38, 38, 24)   96          block_1_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand (Conv2D)         (None, 38, 38, 144)  3456        block_1_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand_BN (BatchNormali (None, 38, 38, 144)  576         block_2_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand_relu (ReLU)      (None, 38, 38, 144)  0           block_2_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise (DepthwiseCon (None, 38, 38, 144)  1296        block_2_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise_BN (BatchNorm (None, 38, 38, 144)  576         block_2_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise_relu (ReLU)   (None, 38, 38, 144)  0           block_2_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_2_project (Conv2D)        (None, 38, 38, 24)   3456        block_2_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_2_project_BN (BatchNormal (None, 38, 38, 24)   96          block_2_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_2_add (Add)               (None, 38, 38, 24)   0           block_1_project_BN[0][0]         \n",
            "                                                                 block_2_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand (Conv2D)         (None, 38, 38, 144)  3456        block_2_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand_BN (BatchNormali (None, 38, 38, 144)  576         block_3_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand_relu (ReLU)      (None, 38, 38, 144)  0           block_3_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_3_pad (ZeroPadding2D)     (None, 39, 39, 144)  0           block_3_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise (DepthwiseCon (None, 19, 19, 144)  1296        block_3_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise_BN (BatchNorm (None, 19, 19, 144)  576         block_3_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise_relu (ReLU)   (None, 19, 19, 144)  0           block_3_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_3_project (Conv2D)        (None, 19, 19, 32)   4608        block_3_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_3_project_BN (BatchNormal (None, 19, 19, 32)   128         block_3_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand (Conv2D)         (None, 19, 19, 192)  6144        block_3_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_4_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_4_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise (DepthwiseCon (None, 19, 19, 192)  1728        block_4_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise_BN (BatchNorm (None, 19, 19, 192)  768         block_4_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise_relu (ReLU)   (None, 19, 19, 192)  0           block_4_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_4_project (Conv2D)        (None, 19, 19, 32)   6144        block_4_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_4_project_BN (BatchNormal (None, 19, 19, 32)   128         block_4_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_4_add (Add)               (None, 19, 19, 32)   0           block_3_project_BN[0][0]         \n",
            "                                                                 block_4_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand (Conv2D)         (None, 19, 19, 192)  6144        block_4_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_5_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_5_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise (DepthwiseCon (None, 19, 19, 192)  1728        block_5_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise_BN (BatchNorm (None, 19, 19, 192)  768         block_5_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise_relu (ReLU)   (None, 19, 19, 192)  0           block_5_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_5_project (Conv2D)        (None, 19, 19, 32)   6144        block_5_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_5_project_BN (BatchNormal (None, 19, 19, 32)   128         block_5_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_5_add (Add)               (None, 19, 19, 32)   0           block_4_add[0][0]                \n",
            "                                                                 block_5_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand (Conv2D)         (None, 19, 19, 192)  6144        block_5_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_6_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_6_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_6_pad (ZeroPadding2D)     (None, 21, 21, 192)  0           block_6_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise (DepthwiseCon (None, 10, 10, 192)  1728        block_6_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise_BN (BatchNorm (None, 10, 10, 192)  768         block_6_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise_relu (ReLU)   (None, 10, 10, 192)  0           block_6_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_6_project (Conv2D)        (None, 10, 10, 64)   12288       block_6_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_6_project_BN (BatchNormal (None, 10, 10, 64)   256         block_6_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand (Conv2D)         (None, 10, 10, 384)  24576       block_6_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_7_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_7_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_7_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_7_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_7_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_7_project (Conv2D)        (None, 10, 10, 64)   24576       block_7_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_7_project_BN (BatchNormal (None, 10, 10, 64)   256         block_7_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_7_add (Add)               (None, 10, 10, 64)   0           block_6_project_BN[0][0]         \n",
            "                                                                 block_7_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand (Conv2D)         (None, 10, 10, 384)  24576       block_7_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_8_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_8_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_8_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_8_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_8_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_8_project (Conv2D)        (None, 10, 10, 64)   24576       block_8_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_8_project_BN (BatchNormal (None, 10, 10, 64)   256         block_8_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_8_add (Add)               (None, 10, 10, 64)   0           block_7_add[0][0]                \n",
            "                                                                 block_8_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand (Conv2D)         (None, 10, 10, 384)  24576       block_8_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_9_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_9_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_9_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_9_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_9_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_9_project (Conv2D)        (None, 10, 10, 64)   24576       block_9_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_9_project_BN (BatchNormal (None, 10, 10, 64)   256         block_9_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_9_add (Add)               (None, 10, 10, 64)   0           block_8_add[0][0]                \n",
            "                                                                 block_9_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand (Conv2D)        (None, 10, 10, 384)  24576       block_9_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand_BN (BatchNormal (None, 10, 10, 384)  1536        block_10_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand_relu (ReLU)     (None, 10, 10, 384)  0           block_10_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise (DepthwiseCo (None, 10, 10, 384)  3456        block_10_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise_BN (BatchNor (None, 10, 10, 384)  1536        block_10_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise_relu (ReLU)  (None, 10, 10, 384)  0           block_10_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_10_project (Conv2D)       (None, 10, 10, 96)   36864       block_10_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_10_project_BN (BatchNorma (None, 10, 10, 96)   384         block_10_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand (Conv2D)        (None, 10, 10, 576)  55296       block_10_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_11_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_11_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise (DepthwiseCo (None, 10, 10, 576)  5184        block_11_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise_BN (BatchNor (None, 10, 10, 576)  2304        block_11_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise_relu (ReLU)  (None, 10, 10, 576)  0           block_11_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_11_project (Conv2D)       (None, 10, 10, 96)   55296       block_11_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_11_project_BN (BatchNorma (None, 10, 10, 96)   384         block_11_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_11_add (Add)              (None, 10, 10, 96)   0           block_10_project_BN[0][0]        \n",
            "                                                                 block_11_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand (Conv2D)        (None, 10, 10, 576)  55296       block_11_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_12_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_12_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise (DepthwiseCo (None, 10, 10, 576)  5184        block_12_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise_BN (BatchNor (None, 10, 10, 576)  2304        block_12_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise_relu (ReLU)  (None, 10, 10, 576)  0           block_12_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_12_project (Conv2D)       (None, 10, 10, 96)   55296       block_12_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_12_project_BN (BatchNorma (None, 10, 10, 96)   384         block_12_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_12_add (Add)              (None, 10, 10, 96)   0           block_11_add[0][0]               \n",
            "                                                                 block_12_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand (Conv2D)        (None, 10, 10, 576)  55296       block_12_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_13_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_13_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_13_pad (ZeroPadding2D)    (None, 11, 11, 576)  0           block_13_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise (DepthwiseCo (None, 5, 5, 576)    5184        block_13_pad[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise_BN (BatchNor (None, 5, 5, 576)    2304        block_13_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise_relu (ReLU)  (None, 5, 5, 576)    0           block_13_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_13_project (Conv2D)       (None, 5, 5, 160)    92160       block_13_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_13_project_BN (BatchNorma (None, 5, 5, 160)    640         block_13_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand (Conv2D)        (None, 5, 5, 960)    153600      block_13_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_14_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_14_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_14_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_14_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_14_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_14_project (Conv2D)       (None, 5, 5, 160)    153600      block_14_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_14_project_BN (BatchNorma (None, 5, 5, 160)    640         block_14_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_14_add (Add)              (None, 5, 5, 160)    0           block_13_project_BN[0][0]        \n",
            "                                                                 block_14_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand (Conv2D)        (None, 5, 5, 960)    153600      block_14_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_15_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_15_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_15_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_15_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_15_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_15_project (Conv2D)       (None, 5, 5, 160)    153600      block_15_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_15_project_BN (BatchNorma (None, 5, 5, 160)    640         block_15_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_15_add (Add)              (None, 5, 5, 160)    0           block_14_add[0][0]               \n",
            "                                                                 block_15_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand (Conv2D)        (None, 5, 5, 960)    153600      block_15_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_16_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_16_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_16_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_16_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_16_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_16_project (Conv2D)       (None, 5, 5, 320)    307200      block_16_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_16_project_BN (BatchNorma (None, 5, 5, 320)    1280        block_16_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "Conv_1 (Conv2D)                 (None, 5, 5, 1280)   409600      block_16_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "Conv_1_bn (BatchNormalization)  (None, 5, 5, 1280)   5120        Conv_1[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "out_relu (ReLU)                 (None, 5, 5, 1280)   0           Conv_1_bn[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 2,257,984\n",
            "Trainable params: 2,223,872\n",
            "Non-trainable params: 34,112\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eb8uc5-P0-pV",
        "outputId": "6396c90c-a87e-4316-b4d7-47c3ba023586"
      },
      "source": [
        "for i,layer in enumerate(base_model.layers):\n",
        "  print(i,layer.name)"
      ],
      "execution_count": 145,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0 input_3\n",
            "1 Conv1_pad\n",
            "2 Conv1\n",
            "3 bn_Conv1\n",
            "4 Conv1_relu\n",
            "5 expanded_conv_depthwise\n",
            "6 expanded_conv_depthwise_BN\n",
            "7 expanded_conv_depthwise_relu\n",
            "8 expanded_conv_project\n",
            "9 expanded_conv_project_BN\n",
            "10 block_1_expand\n",
            "11 block_1_expand_BN\n",
            "12 block_1_expand_relu\n",
            "13 block_1_pad\n",
            "14 block_1_depthwise\n",
            "15 block_1_depthwise_BN\n",
            "16 block_1_depthwise_relu\n",
            "17 block_1_project\n",
            "18 block_1_project_BN\n",
            "19 block_2_expand\n",
            "20 block_2_expand_BN\n",
            "21 block_2_expand_relu\n",
            "22 block_2_depthwise\n",
            "23 block_2_depthwise_BN\n",
            "24 block_2_depthwise_relu\n",
            "25 block_2_project\n",
            "26 block_2_project_BN\n",
            "27 block_2_add\n",
            "28 block_3_expand\n",
            "29 block_3_expand_BN\n",
            "30 block_3_expand_relu\n",
            "31 block_3_pad\n",
            "32 block_3_depthwise\n",
            "33 block_3_depthwise_BN\n",
            "34 block_3_depthwise_relu\n",
            "35 block_3_project\n",
            "36 block_3_project_BN\n",
            "37 block_4_expand\n",
            "38 block_4_expand_BN\n",
            "39 block_4_expand_relu\n",
            "40 block_4_depthwise\n",
            "41 block_4_depthwise_BN\n",
            "42 block_4_depthwise_relu\n",
            "43 block_4_project\n",
            "44 block_4_project_BN\n",
            "45 block_4_add\n",
            "46 block_5_expand\n",
            "47 block_5_expand_BN\n",
            "48 block_5_expand_relu\n",
            "49 block_5_depthwise\n",
            "50 block_5_depthwise_BN\n",
            "51 block_5_depthwise_relu\n",
            "52 block_5_project\n",
            "53 block_5_project_BN\n",
            "54 block_5_add\n",
            "55 block_6_expand\n",
            "56 block_6_expand_BN\n",
            "57 block_6_expand_relu\n",
            "58 block_6_pad\n",
            "59 block_6_depthwise\n",
            "60 block_6_depthwise_BN\n",
            "61 block_6_depthwise_relu\n",
            "62 block_6_project\n",
            "63 block_6_project_BN\n",
            "64 block_7_expand\n",
            "65 block_7_expand_BN\n",
            "66 block_7_expand_relu\n",
            "67 block_7_depthwise\n",
            "68 block_7_depthwise_BN\n",
            "69 block_7_depthwise_relu\n",
            "70 block_7_project\n",
            "71 block_7_project_BN\n",
            "72 block_7_add\n",
            "73 block_8_expand\n",
            "74 block_8_expand_BN\n",
            "75 block_8_expand_relu\n",
            "76 block_8_depthwise\n",
            "77 block_8_depthwise_BN\n",
            "78 block_8_depthwise_relu\n",
            "79 block_8_project\n",
            "80 block_8_project_BN\n",
            "81 block_8_add\n",
            "82 block_9_expand\n",
            "83 block_9_expand_BN\n",
            "84 block_9_expand_relu\n",
            "85 block_9_depthwise\n",
            "86 block_9_depthwise_BN\n",
            "87 block_9_depthwise_relu\n",
            "88 block_9_project\n",
            "89 block_9_project_BN\n",
            "90 block_9_add\n",
            "91 block_10_expand\n",
            "92 block_10_expand_BN\n",
            "93 block_10_expand_relu\n",
            "94 block_10_depthwise\n",
            "95 block_10_depthwise_BN\n",
            "96 block_10_depthwise_relu\n",
            "97 block_10_project\n",
            "98 block_10_project_BN\n",
            "99 block_11_expand\n",
            "100 block_11_expand_BN\n",
            "101 block_11_expand_relu\n",
            "102 block_11_depthwise\n",
            "103 block_11_depthwise_BN\n",
            "104 block_11_depthwise_relu\n",
            "105 block_11_project\n",
            "106 block_11_project_BN\n",
            "107 block_11_add\n",
            "108 block_12_expand\n",
            "109 block_12_expand_BN\n",
            "110 block_12_expand_relu\n",
            "111 block_12_depthwise\n",
            "112 block_12_depthwise_BN\n",
            "113 block_12_depthwise_relu\n",
            "114 block_12_project\n",
            "115 block_12_project_BN\n",
            "116 block_12_add\n",
            "117 block_13_expand\n",
            "118 block_13_expand_BN\n",
            "119 block_13_expand_relu\n",
            "120 block_13_pad\n",
            "121 block_13_depthwise\n",
            "122 block_13_depthwise_BN\n",
            "123 block_13_depthwise_relu\n",
            "124 block_13_project\n",
            "125 block_13_project_BN\n",
            "126 block_14_expand\n",
            "127 block_14_expand_BN\n",
            "128 block_14_expand_relu\n",
            "129 block_14_depthwise\n",
            "130 block_14_depthwise_BN\n",
            "131 block_14_depthwise_relu\n",
            "132 block_14_project\n",
            "133 block_14_project_BN\n",
            "134 block_14_add\n",
            "135 block_15_expand\n",
            "136 block_15_expand_BN\n",
            "137 block_15_expand_relu\n",
            "138 block_15_depthwise\n",
            "139 block_15_depthwise_BN\n",
            "140 block_15_depthwise_relu\n",
            "141 block_15_project\n",
            "142 block_15_project_BN\n",
            "143 block_15_add\n",
            "144 block_16_expand\n",
            "145 block_16_expand_BN\n",
            "146 block_16_expand_relu\n",
            "147 block_16_depthwise\n",
            "148 block_16_depthwise_BN\n",
            "149 block_16_depthwise_relu\n",
            "150 block_16_project\n",
            "151 block_16_project_BN\n",
            "152 Conv_1\n",
            "153 Conv_1_bn\n",
            "154 out_relu\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DtXEyp2StNQ3"
      },
      "source": [
        "base_model.trainable = False"
      ],
      "execution_count": 146,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cfmXzSuWty3o",
        "outputId": "59ef0a34-4641-47b8-e5fa-fa3fd5a4a0a6"
      },
      "source": [
        "base_model.summary()"
      ],
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"mobilenetv2_1.00_224\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_3 (InputLayer)            [(None, 150, 150, 3) 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "Conv1_pad (ZeroPadding2D)       (None, 151, 151, 3)  0           input_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "Conv1 (Conv2D)                  (None, 75, 75, 32)   864         Conv1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "bn_Conv1 (BatchNormalization)   (None, 75, 75, 32)   128         Conv1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "Conv1_relu (ReLU)               (None, 75, 75, 32)   0           bn_Conv1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise (Depthw (None, 75, 75, 32)   288         Conv1_relu[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise_BN (Bat (None, 75, 75, 32)   128         expanded_conv_depthwise[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_depthwise_relu (R (None, 75, 75, 32)   0           expanded_conv_depthwise_BN[0][0] \n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_project (Conv2D)  (None, 75, 75, 16)   512         expanded_conv_depthwise_relu[0][0\n",
            "__________________________________________________________________________________________________\n",
            "expanded_conv_project_BN (Batch (None, 75, 75, 16)   64          expanded_conv_project[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand (Conv2D)         (None, 75, 75, 96)   1536        expanded_conv_project_BN[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand_BN (BatchNormali (None, 75, 75, 96)   384         block_1_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_1_expand_relu (ReLU)      (None, 75, 75, 96)   0           block_1_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_1_pad (ZeroPadding2D)     (None, 77, 77, 96)   0           block_1_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise (DepthwiseCon (None, 38, 38, 96)   864         block_1_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise_BN (BatchNorm (None, 38, 38, 96)   384         block_1_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_1_depthwise_relu (ReLU)   (None, 38, 38, 96)   0           block_1_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_1_project (Conv2D)        (None, 38, 38, 24)   2304        block_1_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_1_project_BN (BatchNormal (None, 38, 38, 24)   96          block_1_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand (Conv2D)         (None, 38, 38, 144)  3456        block_1_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand_BN (BatchNormali (None, 38, 38, 144)  576         block_2_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_2_expand_relu (ReLU)      (None, 38, 38, 144)  0           block_2_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise (DepthwiseCon (None, 38, 38, 144)  1296        block_2_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise_BN (BatchNorm (None, 38, 38, 144)  576         block_2_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_2_depthwise_relu (ReLU)   (None, 38, 38, 144)  0           block_2_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_2_project (Conv2D)        (None, 38, 38, 24)   3456        block_2_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_2_project_BN (BatchNormal (None, 38, 38, 24)   96          block_2_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_2_add (Add)               (None, 38, 38, 24)   0           block_1_project_BN[0][0]         \n",
            "                                                                 block_2_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand (Conv2D)         (None, 38, 38, 144)  3456        block_2_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand_BN (BatchNormali (None, 38, 38, 144)  576         block_3_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_3_expand_relu (ReLU)      (None, 38, 38, 144)  0           block_3_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_3_pad (ZeroPadding2D)     (None, 39, 39, 144)  0           block_3_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise (DepthwiseCon (None, 19, 19, 144)  1296        block_3_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise_BN (BatchNorm (None, 19, 19, 144)  576         block_3_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_3_depthwise_relu (ReLU)   (None, 19, 19, 144)  0           block_3_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_3_project (Conv2D)        (None, 19, 19, 32)   4608        block_3_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_3_project_BN (BatchNormal (None, 19, 19, 32)   128         block_3_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand (Conv2D)         (None, 19, 19, 192)  6144        block_3_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_4_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_4_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_4_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise (DepthwiseCon (None, 19, 19, 192)  1728        block_4_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise_BN (BatchNorm (None, 19, 19, 192)  768         block_4_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_4_depthwise_relu (ReLU)   (None, 19, 19, 192)  0           block_4_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_4_project (Conv2D)        (None, 19, 19, 32)   6144        block_4_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_4_project_BN (BatchNormal (None, 19, 19, 32)   128         block_4_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_4_add (Add)               (None, 19, 19, 32)   0           block_3_project_BN[0][0]         \n",
            "                                                                 block_4_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand (Conv2D)         (None, 19, 19, 192)  6144        block_4_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_5_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_5_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_5_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise (DepthwiseCon (None, 19, 19, 192)  1728        block_5_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise_BN (BatchNorm (None, 19, 19, 192)  768         block_5_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_5_depthwise_relu (ReLU)   (None, 19, 19, 192)  0           block_5_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_5_project (Conv2D)        (None, 19, 19, 32)   6144        block_5_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_5_project_BN (BatchNormal (None, 19, 19, 32)   128         block_5_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_5_add (Add)               (None, 19, 19, 32)   0           block_4_add[0][0]                \n",
            "                                                                 block_5_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand (Conv2D)         (None, 19, 19, 192)  6144        block_5_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand_BN (BatchNormali (None, 19, 19, 192)  768         block_6_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_6_expand_relu (ReLU)      (None, 19, 19, 192)  0           block_6_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_6_pad (ZeroPadding2D)     (None, 21, 21, 192)  0           block_6_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise (DepthwiseCon (None, 10, 10, 192)  1728        block_6_pad[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise_BN (BatchNorm (None, 10, 10, 192)  768         block_6_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_6_depthwise_relu (ReLU)   (None, 10, 10, 192)  0           block_6_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_6_project (Conv2D)        (None, 10, 10, 64)   12288       block_6_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_6_project_BN (BatchNormal (None, 10, 10, 64)   256         block_6_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand (Conv2D)         (None, 10, 10, 384)  24576       block_6_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_7_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_7_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_7_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_7_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_7_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_7_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_7_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_7_project (Conv2D)        (None, 10, 10, 64)   24576       block_7_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_7_project_BN (BatchNormal (None, 10, 10, 64)   256         block_7_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_7_add (Add)               (None, 10, 10, 64)   0           block_6_project_BN[0][0]         \n",
            "                                                                 block_7_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand (Conv2D)         (None, 10, 10, 384)  24576       block_7_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_8_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_8_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_8_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_8_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_8_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_8_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_8_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_8_project (Conv2D)        (None, 10, 10, 64)   24576       block_8_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_8_project_BN (BatchNormal (None, 10, 10, 64)   256         block_8_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_8_add (Add)               (None, 10, 10, 64)   0           block_7_add[0][0]                \n",
            "                                                                 block_8_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand (Conv2D)         (None, 10, 10, 384)  24576       block_8_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand_BN (BatchNormali (None, 10, 10, 384)  1536        block_9_expand[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "block_9_expand_relu (ReLU)      (None, 10, 10, 384)  0           block_9_expand_BN[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise (DepthwiseCon (None, 10, 10, 384)  3456        block_9_expand_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise_BN (BatchNorm (None, 10, 10, 384)  1536        block_9_depthwise[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "block_9_depthwise_relu (ReLU)   (None, 10, 10, 384)  0           block_9_depthwise_BN[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_9_project (Conv2D)        (None, 10, 10, 64)   24576       block_9_depthwise_relu[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "block_9_project_BN (BatchNormal (None, 10, 10, 64)   256         block_9_project[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_9_add (Add)               (None, 10, 10, 64)   0           block_8_add[0][0]                \n",
            "                                                                 block_9_project_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand (Conv2D)        (None, 10, 10, 384)  24576       block_9_add[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand_BN (BatchNormal (None, 10, 10, 384)  1536        block_10_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_10_expand_relu (ReLU)     (None, 10, 10, 384)  0           block_10_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise (DepthwiseCo (None, 10, 10, 384)  3456        block_10_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise_BN (BatchNor (None, 10, 10, 384)  1536        block_10_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_10_depthwise_relu (ReLU)  (None, 10, 10, 384)  0           block_10_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_10_project (Conv2D)       (None, 10, 10, 96)   36864       block_10_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_10_project_BN (BatchNorma (None, 10, 10, 96)   384         block_10_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand (Conv2D)        (None, 10, 10, 576)  55296       block_10_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_11_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_11_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_11_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise (DepthwiseCo (None, 10, 10, 576)  5184        block_11_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise_BN (BatchNor (None, 10, 10, 576)  2304        block_11_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_11_depthwise_relu (ReLU)  (None, 10, 10, 576)  0           block_11_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_11_project (Conv2D)       (None, 10, 10, 96)   55296       block_11_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_11_project_BN (BatchNorma (None, 10, 10, 96)   384         block_11_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_11_add (Add)              (None, 10, 10, 96)   0           block_10_project_BN[0][0]        \n",
            "                                                                 block_11_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand (Conv2D)        (None, 10, 10, 576)  55296       block_11_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_12_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_12_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_12_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise (DepthwiseCo (None, 10, 10, 576)  5184        block_12_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise_BN (BatchNor (None, 10, 10, 576)  2304        block_12_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_12_depthwise_relu (ReLU)  (None, 10, 10, 576)  0           block_12_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_12_project (Conv2D)       (None, 10, 10, 96)   55296       block_12_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_12_project_BN (BatchNorma (None, 10, 10, 96)   384         block_12_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_12_add (Add)              (None, 10, 10, 96)   0           block_11_add[0][0]               \n",
            "                                                                 block_12_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand (Conv2D)        (None, 10, 10, 576)  55296       block_12_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand_BN (BatchNormal (None, 10, 10, 576)  2304        block_13_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_13_expand_relu (ReLU)     (None, 10, 10, 576)  0           block_13_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_13_pad (ZeroPadding2D)    (None, 11, 11, 576)  0           block_13_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise (DepthwiseCo (None, 5, 5, 576)    5184        block_13_pad[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise_BN (BatchNor (None, 5, 5, 576)    2304        block_13_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_13_depthwise_relu (ReLU)  (None, 5, 5, 576)    0           block_13_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_13_project (Conv2D)       (None, 5, 5, 160)    92160       block_13_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_13_project_BN (BatchNorma (None, 5, 5, 160)    640         block_13_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand (Conv2D)        (None, 5, 5, 960)    153600      block_13_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_14_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_14_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_14_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_14_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_14_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_14_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_14_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_14_project (Conv2D)       (None, 5, 5, 160)    153600      block_14_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_14_project_BN (BatchNorma (None, 5, 5, 160)    640         block_14_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_14_add (Add)              (None, 5, 5, 160)    0           block_13_project_BN[0][0]        \n",
            "                                                                 block_14_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand (Conv2D)        (None, 5, 5, 960)    153600      block_14_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_15_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_15_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_15_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_15_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_15_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_15_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_15_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_15_project (Conv2D)       (None, 5, 5, 160)    153600      block_15_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_15_project_BN (BatchNorma (None, 5, 5, 160)    640         block_15_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "block_15_add (Add)              (None, 5, 5, 160)    0           block_14_add[0][0]               \n",
            "                                                                 block_15_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand (Conv2D)        (None, 5, 5, 960)    153600      block_15_add[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand_BN (BatchNormal (None, 5, 5, 960)    3840        block_16_expand[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "block_16_expand_relu (ReLU)     (None, 5, 5, 960)    0           block_16_expand_BN[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise (DepthwiseCo (None, 5, 5, 960)    8640        block_16_expand_relu[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise_BN (BatchNor (None, 5, 5, 960)    3840        block_16_depthwise[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "block_16_depthwise_relu (ReLU)  (None, 5, 5, 960)    0           block_16_depthwise_BN[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "block_16_project (Conv2D)       (None, 5, 5, 320)    307200      block_16_depthwise_relu[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "block_16_project_BN (BatchNorma (None, 5, 5, 320)    1280        block_16_project[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "Conv_1 (Conv2D)                 (None, 5, 5, 1280)   409600      block_16_project_BN[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "Conv_1_bn (BatchNormalization)  (None, 5, 5, 1280)   5120        Conv_1[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "out_relu (ReLU)                 (None, 5, 5, 1280)   0           Conv_1_bn[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 2,257,984\n",
            "Trainable params: 0\n",
            "Non-trainable params: 2,257,984\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0JQtvkcit0pR"
      },
      "source": [
        "#adding our own classifier\n",
        "global_average_layer = tf.keras.layers.GlobalAveragePooling2D()"
      ],
      "execution_count": 148,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LDE7p0Pjt6uM"
      },
      "source": [
        "prediction_layer = keras.layers.Dense(1)"
      ],
      "execution_count": 149,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dJydVl3H-QVf",
        "outputId": "c0998872-b37d-4160-9007-4511a3157f57"
      },
      "source": [
        "tf.keras.layers.Dense(512, activation='relu'),\n",
        "tf.keras.layers.Dense(3, activation='softmax')"
      ],
      "execution_count": 130,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.layers.core.Dense at 0x7fe82003c668>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 130
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7AlxbKgit74L"
      },
      "source": [
        "model = tf.keras.Sequential([\n",
        "  base_model,\n",
        "  global_average_layer,\n",
        "  prediction_layer\n",
        "#  tf.keras.layers.Dense(512, activation='relu'),\n",
        " # tf.keras.layers.Dense(3, activation='softmax')\n",
        "])"
      ],
      "execution_count": 150,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0GzDaUNnt-f5",
        "outputId": "bb8b467b-213b-4484-e2c3-c0c7a06c11fd"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 151,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_8\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "mobilenetv2_1.00_224 (Functi (None, 5, 5, 1280)        2257984   \n",
            "_________________________________________________________________\n",
            "global_average_pooling2d_3 ( (None, 1280)              0         \n",
            "_________________________________________________________________\n",
            "dense_15 (Dense)             (None, 1)                 1281      \n",
            "=================================================================\n",
            "Total params: 2,259,265\n",
            "Trainable params: 1,281\n",
            "Non-trainable params: 2,257,984\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4uNbBcwmzJ_k"
      },
      "source": [
        "**Training our model**\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JdHuidVjuAOr"
      },
      "source": [
        "base_learning_rate = 0.0001\n",
        "model.compile(optimizer=tf.keras.optimizers.RMSprop(lr=base_learning_rate),\n",
        "              loss=tf.keras.losses.CategoricalCrossentropy(from_logits=True),\n",
        "              metrics=['accuracy'])"
      ],
      "execution_count": 152,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tazvX1hMuDe_",
        "outputId": "86edf55d-72b9-4456-a36d-b1bec6120650"
      },
      "source": [
        "# We can evaluate the model right now to see how it does before training it on our new images\n",
        "initial_epochs = 5\n",
        "validation_steps=20\n",
        "\n",
        "loss0,accuracy0 = model.evaluate(validation_generator, steps = validation_steps)"
      ],
      "execution_count": 153,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " 5/20 [======>.......................] - ETA: 11s - loss: 1.9459 - accuracy: 0.7620WARNING:tensorflow:Your input ran out of data; interrupting training. Make sure that your dataset or generator can generate at least `steps_per_epoch * epochs` batches (in this case, 20 batches). You may need to use the repeat() function when building your dataset.\n",
            " 5/20 [======>.......................] - 4s 736ms/step - loss: 1.9459 - accuracy: 0.7620\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CcJcRbSI0gc1"
      },
      "source": [
        "import numpy as np\n",
        "from datetime import datetime\n",
        "import time\n",
        "#from sklearn import tree\n",
        "#from sklearn.preprocessing import normalize, scale\n",
        "from sklearn.metrics import accuracy_score\n",
        "from keras import backend as K\n",
        "from keras.callbacks import ModelCheckpoint\n",
        "fmt = '%H:%M:%S'\n",
        "def get_current_time():\n",
        "    time.ctime()\n",
        "    return time.strftime(fmt)\n",
        "first = get_current_time()"
      ],
      "execution_count": 154,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HlUKVtL2AjEB",
        "outputId": "5334a293-7237-4fc1-b8b6-3d9749e68fde"
      },
      "source": [
        "\n",
        "checkpoint = ModelCheckpoint(\"best_model.hdf5\", monitor='loss', verbose=1,\n",
        "    save_best_only=True, mode='auto', period=1)"
      ],
      "execution_count": 155,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X-_X8Dwm3xhd",
        "outputId": "5840c409-5765-4c0d-c7b4-a64394b22398"
      },
      "source": [
        "\n",
        "\n",
        "\n",
        "history = model.fit(train_generator, epochs=50, steps_per_epoch=20, validation_data = validation_generator, verbose = 1, validation_steps=3)\n",
        "\n",
        "acc = history.history['accuracy']\n",
        "print(acc)"
      ],
      "execution_count": 156,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "20/20 [==============================] - 30s 1s/step - loss: 1.7557 - accuracy: 0.7907 - val_loss: 1.6237 - val_accuracy: 0.7918\n",
            "Epoch 2/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.5530 - accuracy: 0.8106 - val_loss: 1.4484 - val_accuracy: 0.8273\n",
            "Epoch 3/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.4138 - accuracy: 0.8302 - val_loss: 1.3331 - val_accuracy: 0.8337\n",
            "Epoch 4/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.3019 - accuracy: 0.8413 - val_loss: 1.1972 - val_accuracy: 0.8382\n",
            "Epoch 5/50\n",
            "20/20 [==============================] - 27s 1s/step - loss: 1.2163 - accuracy: 0.8505 - val_loss: 1.1307 - val_accuracy: 0.8469\n",
            "Epoch 6/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 1.1490 - accuracy: 0.8523 - val_loss: 1.0477 - val_accuracy: 0.8568\n",
            "Epoch 7/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0840 - accuracy: 0.8621 - val_loss: 1.0124 - val_accuracy: 0.8605\n",
            "Epoch 8/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 1.0411 - accuracy: 0.8671 - val_loss: 0.9573 - val_accuracy: 0.8715\n",
            "Epoch 9/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.9955 - accuracy: 0.8686 - val_loss: 0.8932 - val_accuracy: 0.8749\n",
            "Epoch 10/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9533 - accuracy: 0.8721 - val_loss: 0.8867 - val_accuracy: 0.8707\n",
            "Epoch 11/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.9244 - accuracy: 0.8781 - val_loss: 0.8382 - val_accuracy: 0.8791\n",
            "Epoch 12/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.9019 - accuracy: 0.8802 - val_loss: 0.8508 - val_accuracy: 0.8779\n",
            "Epoch 13/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8805 - accuracy: 0.8807 - val_loss: 0.8057 - val_accuracy: 0.8832\n",
            "Epoch 14/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8563 - accuracy: 0.8841 - val_loss: 0.7990 - val_accuracy: 0.8813\n",
            "Epoch 15/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8440 - accuracy: 0.8855 - val_loss: 0.7565 - val_accuracy: 0.8889\n",
            "Epoch 16/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.8257 - accuracy: 0.8853 - val_loss: 0.7375 - val_accuracy: 0.8915\n",
            "Epoch 17/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.8044 - accuracy: 0.8891 - val_loss: 0.7513 - val_accuracy: 0.8840\n",
            "Epoch 18/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7884 - accuracy: 0.8918 - val_loss: 0.6901 - val_accuracy: 0.8874\n",
            "Epoch 19/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7940 - accuracy: 0.8892 - val_loss: 0.7044 - val_accuracy: 0.8896\n",
            "Epoch 20/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7754 - accuracy: 0.8929 - val_loss: 0.6945 - val_accuracy: 0.8968\n",
            "Epoch 21/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7608 - accuracy: 0.8914 - val_loss: 0.6880 - val_accuracy: 0.8983\n",
            "Epoch 22/50\n",
            "20/20 [==============================] - 30s 1s/step - loss: 0.7349 - accuracy: 0.8953 - val_loss: 0.6978 - val_accuracy: 0.8972\n",
            "Epoch 23/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7358 - accuracy: 0.8946 - val_loss: 0.7225 - val_accuracy: 0.8968\n",
            "Epoch 24/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.7357 - accuracy: 0.8970 - val_loss: 0.6805 - val_accuracy: 0.8987\n",
            "Epoch 25/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7231 - accuracy: 0.8968 - val_loss: 0.7238 - val_accuracy: 0.8991\n",
            "Epoch 26/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7223 - accuracy: 0.8944 - val_loss: 0.6590 - val_accuracy: 0.9029\n",
            "Epoch 27/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.7119 - accuracy: 0.8948 - val_loss: 0.6490 - val_accuracy: 0.9048\n",
            "Epoch 28/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6836 - accuracy: 0.8996 - val_loss: 0.6718 - val_accuracy: 0.8964\n",
            "Epoch 29/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6957 - accuracy: 0.8980 - val_loss: 0.6393 - val_accuracy: 0.9078\n",
            "Epoch 30/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6732 - accuracy: 0.9014 - val_loss: 0.6262 - val_accuracy: 0.9036\n",
            "Epoch 31/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6666 - accuracy: 0.9003 - val_loss: 0.6622 - val_accuracy: 0.9044\n",
            "Epoch 32/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6765 - accuracy: 0.8995 - val_loss: 0.6561 - val_accuracy: 0.9044\n",
            "Epoch 33/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6690 - accuracy: 0.8980 - val_loss: 0.6640 - val_accuracy: 0.9029\n",
            "Epoch 34/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6547 - accuracy: 0.9038 - val_loss: 0.6076 - val_accuracy: 0.9070\n",
            "Epoch 35/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6498 - accuracy: 0.9019 - val_loss: 0.5969 - val_accuracy: 0.9097\n",
            "Epoch 36/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6459 - accuracy: 0.9021 - val_loss: 0.5991 - val_accuracy: 0.9055\n",
            "Epoch 37/50\n",
            "20/20 [==============================] - 27s 1s/step - loss: 0.6405 - accuracy: 0.9015 - val_loss: 0.5852 - val_accuracy: 0.9082\n",
            "Epoch 38/50\n",
            "20/20 [==============================] - 27s 1s/step - loss: 0.6468 - accuracy: 0.9023 - val_loss: 0.6578 - val_accuracy: 0.9040\n",
            "Epoch 39/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6305 - accuracy: 0.9035 - val_loss: 0.6004 - val_accuracy: 0.9055\n",
            "Epoch 40/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6159 - accuracy: 0.9060 - val_loss: 0.6077 - val_accuracy: 0.9078\n",
            "Epoch 41/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6232 - accuracy: 0.9036 - val_loss: 0.6116 - val_accuracy: 0.9101\n",
            "Epoch 42/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6088 - accuracy: 0.9035 - val_loss: 0.5688 - val_accuracy: 0.9108\n",
            "Epoch 43/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6110 - accuracy: 0.9048 - val_loss: 0.6223 - val_accuracy: 0.9063\n",
            "Epoch 44/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5947 - accuracy: 0.9047 - val_loss: 0.5958 - val_accuracy: 0.9108\n",
            "Epoch 45/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6068 - accuracy: 0.9030 - val_loss: 0.5662 - val_accuracy: 0.9119\n",
            "Epoch 46/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.6138 - accuracy: 0.9036 - val_loss: 0.6051 - val_accuracy: 0.9033\n",
            "Epoch 47/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.5994 - accuracy: 0.9044 - val_loss: 0.5744 - val_accuracy: 0.9119\n",
            "Epoch 48/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.5993 - accuracy: 0.9048 - val_loss: 0.5493 - val_accuracy: 0.9150\n",
            "Epoch 49/50\n",
            "20/20 [==============================] - 28s 1s/step - loss: 0.5900 - accuracy: 0.9057 - val_loss: 0.5687 - val_accuracy: 0.9051\n",
            "Epoch 50/50\n",
            "20/20 [==============================] - 29s 1s/step - loss: 0.6046 - accuracy: 0.9055 - val_loss: 0.5645 - val_accuracy: 0.9199\n",
            "[0.790744423866272, 0.8105692863464355, 0.830216646194458, 0.8413421511650085, 0.850514829158783, 0.8523492217063904, 0.8621138334274292, 0.8670849204063416, 0.8686234951019287, 0.8721151351928711, 0.8780922293663025, 0.8802226185798645, 0.8806959390640259, 0.8841283917427063, 0.8855487108230591, 0.8852527141571045, 0.8890993595123291, 0.891762375831604, 0.8892176747322083, 0.8928868174552917, 0.891407310962677, 0.8953131437301636, 0.8946029543876648, 0.8969700336456299, 0.8967925310134888, 0.8943662047386169, 0.8948395848274231, 0.8995739817619324, 0.8979762196540833, 0.9014085531234741, 0.9003432989120483, 0.8995147943496704, 0.8979761004447937, 0.9037755131721497, 0.9018818736076355, 0.9021186232566833, 0.9015268087387085, 0.9022961258888245, 0.903479814529419, 0.9060243964195251, 0.9035980105400085, 0.9035387635231018, 0.9047816395759583, 0.9047223329544067, 0.903006374835968, 0.9035981297492981, 0.9043673872947693, 0.904840886592865, 0.9056695103645325, 0.9054918885231018]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vXqbaD7g0sN6",
        "outputId": "241789ce-4d99-40d7-ce42-bfecadab4a8a"
      },
      "source": [
        "second = get_current_time()\n",
        "print(\"Time taken to train(sec):\", datetime.strptime(second, fmt) - datetime.strptime(first, fmt))"
      ],
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Time taken to train(sec): 0:25:01\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "asxDXorU1gnF",
        "outputId": "8337249c-3e0e-4446-ba22-af11c245c11c"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "plt.plot(epochs, acc, 'r', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 158,
      "outputs": [
        {
          "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"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "Ly8gQ29weXJpZ2h0IDIwMTcgR29vZ2xlIExMQwovLwovLyBMaWNlbnNlZCB1bmRlciB0aGUgQXBhY2hlIExpY2Vuc2UsIFZlcnNpb24gMi4wICh0aGUgIkxpY2Vuc2UiKTsKLy8geW91IG1heSBub3QgdXNlIHRoaXMgZmlsZSBleGNlcHQgaW4gY29tcGxpYW5jZSB3aXRoIHRoZSBMaWNlbnNlLgovLyBZb3UgbWF5IG9idGFpbiBhIGNvcHkgb2YgdGhlIExpY2Vuc2UgYXQKLy8KLy8gICAgICBodHRwOi8vd3d3LmFwYWNoZS5vcmcvbGljZW5zZXMvTElDRU5TRS0yLjAKLy8KLy8gVW5sZXNzIHJlcXVpcmVkIGJ5IGFwcGxpY2FibGUgbGF3IG9yIGFncmVlZCB0byBpbiB3cml0aW5nLCBzb2Z0d2FyZQovLyBkaXN0cmlidXRlZCB1bmRlciB0aGUgTGljZW5zZSBpcyBkaXN0cmlidXRlZCBvbiBhbiAiQVMgSVMiIEJBU0lTLAovLyBXSVRIT1VUIFdBUlJBTlRJRVMgT1IgQ09ORElUSU9OUyBPRiBBTlkgS0lORCwgZWl0aGVyIGV4cHJlc3Mgb3IgaW1wbGllZC4KLy8gU2VlIHRoZSBMaWNlbnNlIGZvciB0aGUgc3BlY2lmaWMgbGFuZ3VhZ2UgZ292ZXJuaW5nIHBlcm1pc3Npb25zIGFuZAovLyBsaW1pdGF0aW9ucyB1bmRlciB0aGUgTGljZW5zZS4KCi8qKgogKiBAZmlsZW92ZXJ2aWV3IEhlbHBlcnMgZm9yIGdvb2dsZS5jb2xhYiBQeXRob24gbW9kdWxlLgogKi8KKGZ1bmN0aW9uKHNjb3BlKSB7CmZ1bmN0aW9uIHNwYW4odGV4dCwgc3R5bGVBdHRyaWJ1dGVzID0ge30pIHsKICBjb25zdCBlbGVtZW50ID0gZG9jdW1lbnQuY3JlYXRlRWxlbWVudCgnc3BhbicpOwogIGVsZW1lbnQudGV4dENvbnRlbnQgPSB0ZXh0OwogIGZvciAoY29uc3Qga2V5IG9mIE9iamVjdC5rZXlzKHN0eWxlQXR0cmlidXRlcykpIHsKICAgIGVsZW1lbnQuc3R5bGVba2V5XSA9IHN0eWxlQXR0cmlidXRlc1trZXldOwogIH0KICByZXR1cm4gZWxlbWVudDsKfQoKLy8gTWF4IG51bWJlciBvZiBieXRlcyB3aGljaCB3aWxsIGJlIHVwbG9hZGVkIGF0IGEgdGltZS4KY29uc3QgTUFYX1BBWUxPQURfU0laRSA9IDEwMCAqIDEwMjQ7CgpmdW5jdGlvbiBfdXBsb2FkRmlsZXMoaW5wdXRJZCwgb3V0cHV0SWQpIHsKICBjb25zdCBzdGVwcyA9IHVwbG9hZEZpbGVzU3RlcChpbnB1dElkLCBvdXRwdXRJZCk7CiAgY29uc3Qgb3V0cHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKG91dHB1dElkKTsKICAvLyBDYWNoZSBzdGVwcyBvbiB0aGUgb3V0cHV0RWxlbWVudCB0byBtYWtlIGl0IGF2YWlsYWJsZSBmb3IgdGhlIG5leHQgY2FsbAogIC8vIHRvIHVwbG9hZEZpbGVzQ29udGludWUgZnJvbSBQeXRob24uCiAgb3V0cHV0RWxlbWVudC5zdGVwcyA9IHN0ZXBzOwoKICByZXR1cm4gX3VwbG9hZEZpbGVzQ29udGludWUob3V0cHV0SWQpOwp9CgovLyBUaGlzIGlzIHJvdWdobHkgYW4gYXN5bmMgZ2VuZXJhdG9yIChub3Qgc3VwcG9ydGVkIGluIHRoZSBicm93c2VyIHlldCksCi8vIHdoZXJlIHRoZXJlIGFyZSBtdWx0aXBsZSBhc3luY2hyb25vdXMgc3RlcHMgYW5kIHRoZSBQeXRob24gc2lkZSBpcyBnb2luZwovLyB0byBwb2xsIGZvciBjb21wbGV0aW9uIG9mIGVhY2ggc3RlcC4KLy8gVGhpcyB1c2VzIGEgUHJvbWlzZSB0byBibG9jayB0aGUgcHl0aG9uIHNpZGUgb24gY29tcGxldGlvbiBvZiBlYWNoIHN0ZXAsCi8vIHRoZW4gcGFzc2VzIHRoZSByZXN1bHQgb2YgdGhlIHByZXZpb3VzIHN0ZXAgYXMgdGhlIGlucHV0IHRvIHRoZSBuZXh0IHN0ZXAuCmZ1bmN0aW9uIF91cGxvYWRGaWxlc0NvbnRpbnVlKG91dHB1dElkKSB7CiAgY29uc3Qgb3V0cHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKG91dHB1dElkKTsKICBjb25zdCBzdGVwcyA9IG91dHB1dEVsZW1lbnQuc3RlcHM7CgogIGNvbnN0IG5leHQgPSBzdGVwcy5uZXh0KG91dHB1dEVsZW1lbnQubGFzdFByb21pc2VWYWx1ZSk7CiAgcmV0dXJuIFByb21pc2UucmVzb2x2ZShuZXh0LnZhbHVlLnByb21pc2UpLnRoZW4oKHZhbHVlKSA9PiB7CiAgICAvLyBDYWNoZSB0aGUgbGFzdCBwcm9taXNlIHZhbHVlIHRvIG1ha2UgaXQgYXZhaWxhYmxlIHRvIHRoZSBuZXh0CiAgICAvLyBzdGVwIG9mIHRoZSBnZW5lcmF0b3IuCiAgICBvdXRwdXRFbGVtZW50Lmxhc3RQcm9taXNlVmFsdWUgPSB2YWx1ZTsKICAgIHJldHVybiBuZXh0LnZhbHVlLnJlc3BvbnNlOwogIH0pOwp9CgovKioKICogR2VuZXJhdG9yIGZ1bmN0aW9uIHdoaWNoIGlzIGNhbGxlZCBiZXR3ZWVuIGVhY2ggYXN5bmMgc3RlcCBvZiB0aGUgdXBsb2FkCiAqIHByb2Nlc3MuCiAqIEBwYXJhbSB7c3RyaW5nfSBpbnB1dElkIEVsZW1lbnQgSUQgb2YgdGhlIGlucHV0IGZpbGUgcGlja2VyIGVsZW1lbnQuCiAqIEBwYXJhbSB7c3RyaW5nfSBvdXRwdXRJZCBFbGVtZW50IElEIG9mIHRoZSBvdXRwdXQgZGlzcGxheS4KICogQHJldHVybiB7IUl0ZXJhYmxlPCFPYmplY3Q+fSBJdGVyYWJsZSBvZiBuZXh0IHN0ZXBzLgogKi8KZnVuY3Rpb24qIHVwbG9hZEZpbGVzU3RlcChpbnB1dElkLCBvdXRwdXRJZCkgewogIGNvbnN0IGlucHV0RWxlbWVudCA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlkKGlucHV0SWQpOwogIGlucHV0RWxlbWVudC5kaXNhYmxlZCA9IGZhbHNlOwoKICBjb25zdCBvdXRwdXRFbGVtZW50ID0gZG9jdW1lbnQuZ2V0RWxlbWVudEJ5SWQob3V0cHV0SWQpOwogIG91dHB1dEVsZW1lbnQuaW5uZXJIVE1MID0gJyc7CgogIGNvbnN0IHBpY2tlZFByb21pc2UgPSBuZXcgUHJvbWlzZSgocmVzb2x2ZSkgPT4gewogICAgaW5wdXRFbGVtZW50LmFkZEV2ZW50TGlzdGVuZXIoJ2NoYW5nZScsIChlKSA9PiB7CiAgICAgIHJlc29sdmUoZS50YXJnZXQuZmlsZXMpOwogICAgfSk7CiAgfSk7CgogIGNvbnN0IGNhbmNlbCA9IGRvY3VtZW50LmNyZWF0ZUVsZW1lbnQoJ2J1dHRvbicpOwogIGlucHV0RWxlbWVudC5wYXJlbnRFbGVtZW50LmFwcGVuZENoaWxkKGNhbmNlbCk7CiAgY2FuY2VsLnRleHRDb250ZW50ID0gJ0NhbmNlbCB1cGxvYWQnOwogIGNvbnN0IGNhbmNlbFByb21pc2UgPSBuZXcgUHJvbWlzZSgocmVzb2x2ZSkgPT4gewogICAgY2FuY2VsLm9uY2xpY2sgPSAoKSA9PiB7CiAgICAgIHJlc29sdmUobnVsbCk7CiAgICB9OwogIH0pOwoKICAvLyBXYWl0IGZvciB0aGUgdXNlciB0byBwaWNrIHRoZSBmaWxlcy4KICBjb25zdCBmaWxlcyA9IHlpZWxkIHsKICAgIHByb21pc2U6IFByb21pc2UucmFjZShbcGlja2VkUHJvbWlzZSwgY2FuY2VsUHJvbWlzZV0pLAogICAgcmVzcG9uc2U6IHsKICAgICAgYWN0aW9uOiAnc3RhcnRpbmcnLAogICAgfQogIH07CgogIGNhbmNlbC5yZW1vdmUoKTsKCiAgLy8gRGlzYWJsZSB0aGUgaW5wdXQgZWxlbWVudCBzaW5jZSBmdXJ0aGVyIHBpY2tzIGFyZSBub3QgYWxsb3dlZC4KICBpbnB1dEVsZW1lbnQuZGlzYWJsZWQgPSB0cnVlOwoKICBpZiAoIWZpbGVzKSB7CiAgICByZXR1cm4gewogICAgICByZXNwb25zZTogewogICAgICAgIGFjdGlvbjogJ2NvbXBsZXRlJywKICAgICAgfQogICAgfTsKICB9CgogIGZvciAoY29uc3QgZmlsZSBvZiBmaWxlcykgewogICAgY29uc3QgbGkgPSBkb2N1bWVudC5jcmVhdGVFbGVtZW50KCdsaScpOwogICAgbGkuYXBwZW5kKHNwYW4oZmlsZS5uYW1lLCB7Zm9udFdlaWdodDogJ2JvbGQnfSkpOwogICAgbGkuYXBwZW5kKHNwYW4oCiAgICAgICAgYCgke2ZpbGUudHlwZSB8fCAnbi9hJ30pIC0gJHtmaWxlLnNpemV9IGJ5dGVzLCBgICsKICAgICAgICBgbGFzdCBtb2RpZmllZDogJHsKICAgICAgICAgICAgZmlsZS5sYXN0TW9kaWZpZWREYXRlID8gZmlsZS5sYXN0TW9kaWZpZWREYXRlLnRvTG9jYWxlRGF0ZVN0cmluZygpIDoKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgJ24vYSd9IC0gYCkpOwogICAgY29uc3QgcGVyY2VudCA9IHNwYW4oJzAlIGRvbmUnKTsKICAgIGxpLmFwcGVuZENoaWxkKHBlcmNlbnQpOwoKICAgIG91dHB1dEVsZW1lbnQuYXBwZW5kQ2hpbGQobGkpOwoKICAgIGNvbnN0IGZpbGVEYXRhUHJvbWlzZSA9IG5ldyBQcm9taXNlKChyZXNvbHZlKSA9PiB7CiAgICAgIGNvbnN0IHJlYWRlciA9IG5ldyBGaWxlUmVhZGVyKCk7CiAgICAgIHJlYWRlci5vbmxvYWQgPSAoZSkgPT4gewogICAgICAgIHJlc29sdmUoZS50YXJnZXQucmVzdWx0KTsKICAgICAgfTsKICAgICAgcmVhZGVyLnJlYWRBc0FycmF5QnVmZmVyKGZpbGUpOwogICAgfSk7CiAgICAvLyBXYWl0IGZvciB0aGUgZGF0YSB0byBiZSByZWFkeS4KICAgIGxldCBmaWxlRGF0YSA9IHlpZWxkIHsKICAgICAgcHJvbWlzZTogZmlsZURhdGFQcm9taXNlLAogICAgICByZXNwb25zZTogewogICAgICAgIGFjdGlvbjogJ2NvbnRpbnVlJywKICAgICAgfQogICAgfTsKCiAgICAvLyBVc2UgYSBjaHVua2VkIHNlbmRpbmcgdG8gYXZvaWQgbWVzc2FnZSBzaXplIGxpbWl0cy4gU2VlIGIvNjIxMTU2NjAuCiAgICBsZXQgcG9zaXRpb24gPSAwOwogICAgd2hpbGUgKHBvc2l0aW9uIDwgZmlsZURhdGEuYnl0ZUxlbmd0aCkgewogICAgICBjb25zdCBsZW5ndGggPSBNYXRoLm1pbihmaWxlRGF0YS5ieXRlTGVuZ3RoIC0gcG9zaXRpb24sIE1BWF9QQVlMT0FEX1NJWkUpOwogICAgICBjb25zdCBjaHVuayA9IG5ldyBVaW50OEFycmF5KGZpbGVEYXRhLCBwb3NpdGlvbiwgbGVuZ3RoKTsKICAgICAgcG9zaXRpb24gKz0gbGVuZ3RoOwoKICAgICAgY29uc3QgYmFzZTY0ID0gYnRvYShTdHJpbmcuZnJvbUNoYXJDb2RlLmFwcGx5KG51bGwsIGNodW5rKSk7CiAgICAgIHlpZWxkIHsKICAgICAgICByZXNwb25zZTogewogICAgICAgICAgYWN0aW9uOiAnYXBwZW5kJywKICAgICAgICAgIGZpbGU6IGZpbGUubmFtZSwKICAgICAgICAgIGRhdGE6IGJhc2U2NCwKICAgICAgICB9LAogICAgICB9OwogICAgICBwZXJjZW50LnRleHRDb250ZW50ID0KICAgICAgICAgIGAke01hdGgucm91bmQoKHBvc2l0aW9uIC8gZmlsZURhdGEuYnl0ZUxlbmd0aCkgKiAxMDApfSUgZG9uZWA7CiAgICB9CiAgfQoKICAvLyBBbGwgZG9uZS4KICB5aWVsZCB7CiAgICByZXNwb25zZTogewogICAgICBhY3Rpb246ICdjb21wbGV0ZScsCiAgICB9CiAgfTsKfQoKc2NvcGUuZ29vZ2xlID0gc2NvcGUuZ29vZ2xlIHx8IHt9OwpzY29wZS5nb29nbGUuY29sYWIgPSBzY29wZS5nb29nbGUuY29sYWIgfHwge307CnNjb3BlLmdvb2dsZS5jb2xhYi5fZmlsZXMgPSB7CiAgX3VwbG9hZEZpbGVzLAogIF91cGxvYWRGaWxlc0NvbnRpbnVlLAp9Owp9KShzZWxmKTsK",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 123
        },
        "id": "qIKbWz-e1mDL",
        "outputId": "9c764056-d911-46e5-c4ba-4c0788f989ef"
      },
      "source": [
        "import numpy as np\n",
        "from google.colab import files\n",
        "from keras.preprocessing import image\n",
        "\n",
        "uploaded = files.upload()\n",
        "\n",
        "for fn in uploaded.keys():\n",
        " \n",
        "  # predicting images\n",
        "  path = fn\n",
        "  img = image.load_img(path, target_size=(150, 150))\n",
        "  x = image.img_to_array(img)\n",
        "  x = np.expand_dims(x, axis=0)\n",
        "\n",
        "  images = np.vstack([x])\n",
        "  classes = model.predict(images, batch_size=10)\n",
        "  print(fn)\n",
        "  print(classes)"
      ],
      "execution_count": 160,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-f478f0c1-550f-40f3-bbec-c35b67f20aaa\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-f478f0c1-550f-40f3-bbec-c35b67f20aaa\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving glass64.jpg to glass64.jpg\n",
            "glass64.jpg\n",
            "[[-4.6035004  -0.72371566 -5.4054046  -1.7737067  -2.3621428  -4.6789527\n",
            "  -4.8643503 ]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fX_-qDxczcB3"
      },
      "source": [
        "model.save(\"mobilenet_v2.h5\")  # we can save the model and reload it at anytime in the future\n",
        "#new_model = tf.keras.models.load_model('mobilenet.h5')"
      ],
      "execution_count": 163,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V2fFtB1P0rrg"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}