ml_ocr.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Test.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "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/abderrazzaq-laanoui/PFE-HexaCoders/blob/main/ml_ocr.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E8hCO76eiaTU"
      },
      "source": [
        "<div style=\"text-align:center\"><h1>RECONNAISSANCE DE L'ÉCRITURE MANUSCRITE DE LESTTRES TIFINAGH </h1></div>\n",
        "\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CACEY3QXpYdB"
      },
      "source": [
        "#Dépendances"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qiufKkBMpkLy"
      },
      "source": [
        "#### Liaison avec Google Drive"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wrxPNHSdpikm",
        "outputId": "53d0d88c-0ee1-4c22-8128-c2c7a87384af"
      },
      "source": [
        "# liaison avec  Google Drive comme source de donnees \n",
        "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": "markdown",
      "metadata": {
        "id": "puJQ-cszpuEL"
      },
      "source": [
        "#### Importations"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kDPSleVSptAh"
      },
      "source": [
        "# Importation des bibliothèques\n",
        "import os\n",
        "import numpy as np\n",
        "import cv2 as cv\n",
        "import random\n",
        "import pickle\n",
        "import time\n",
        "from datetime import timedelta\n",
        "import math\n",
        "# CNN\n",
        "import tensorflow as tf\n",
        "tf.config.run_functions_eagerly(True)\n",
        "from tensorflow import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.layers.convolutional import Conv2D\n",
        "from keras.layers.convolutional import MaxPooling2D\n",
        "from keras.layers.core import Activation\n",
        "from keras.layers.core import Flatten\n",
        "from keras.layers.core import Dropout\n",
        "from keras.layers.core import Dense\n",
        "from keras import backend as K\n",
        "# graph\n",
        "import matplotlib\n",
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "# matrice de conf\n",
        "from mlxtend.plotting import plot_confusion_matrix\n",
        "from sklearn.metrics import confusion_matrix"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DMaddkKHWoHJ"
      },
      "source": [
        "pip install tensorflowjs"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c_8SKXLvwgFA"
      },
      "source": [
        "#### Variables"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NYOaYFjlwlDU"
      },
      "source": [
        "\n",
        "CATEGORIES = [\"ya\", \"yab\", \"yach\", \"yad\", \"yadd\", \"yae\", \"yaf\", \"yag\", \"yagh\", \n",
        "              \"yagw\", \"yah\", \"yahh\", \"yaj\", \"yak\", \"yakw\", \"yal\", \"yam\", \"yan\", \n",
        "              \"yaq\", \"yar\",\"yarr\", \"yas\", \"yass\", \"yat\", \"yatt\", \"yaw\", \"yax\", \n",
        "              \"yay\", \"yaz\", \"yazz\", \"yey\", \"yi\", \"yu\"] # a list of all possible classes\n",
        "T_CATEGORIES = ['ⴰ', 'ⴱ', 'ⵛ', 'ⴷ', 'ⴹ', 'ⵄ', 'ⴼ', 'ⴳ', 'ⵖ', 'ⴳⵯ', 'ⵀ', 'ⵃ', 'ⵊ', 'ⴽ', 'ⴽⵯ',\n",
        "                'ⵍ','ⵎ','ⵏ', 'ⵇ', 'ⵔ', 'ⵕ', 'ⵙ', 'ⵚ', 'ⵜ', 'ⵟ', 'ⵡ', 'ⵅ', 'ⵢ', 'ⵣ','ⵥ', 'ⴻ', 'ⵉ', 'ⵓ']\n",
        "\n",
        "training_data = [] # cette liste qui contiendra les données d'entraînement traitées\n",
        "testing_data = [] # cette liste qui contiendra les données de test traitées"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KT29UQrnd5G0"
      },
      "source": [
        "\n",
        "# CHARGEMENT DU DATASET"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vl4_nyb6goN6"
      },
      "source": [
        "\n",
        "#@markdown \\\n",
        "#@markdown #### Entrez le chemin d'accès au répertoire DATASET dans votre Drive:\n",
        "chemin = \"\" #@param {type:\"string\"}\n",
        "#@markdown \\\n",
        "\n",
        "if chemin == \"\":\n",
        "    chemin = \"genie info/Rapport PFE - HexaCoders/DATASET\" #default path\n",
        "# Declaration des Variables\n",
        "DIR = os.path.join(\"/content/drive/MyDrive\",chemin) # h the path to the DATASET dir  in your drive\n",
        "TRAIN_DATA_DIR = os.path.join(DIR,\"training_data/\")\n",
        "TEST_DATA_DIR = os.path.join(DIR,\"testing_data/\")\n",
        "MODEL_DIR = os.path.join(DIR,\"model/\")\n",
        "TRAIN_MODEL_DIR = os.path.join(DIR,\"model/train/\")\n",
        "TEST_MODEL_DIR = os.path.join(DIR,\"model/test/\")\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pdc2JFHU77S7"
      },
      "source": [
        "\n",
        "## DONNÉES D'ENTRAÎNEMENT"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_N_yFjf_qXCP"
      },
      "source": [
        "\n",
        "#### Préparation des données"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0S-Vd6uGsYcO"
      },
      "source": [
        "<b>OpenCV</b> (référencé par cv) est une bibliothèque multi-plateforme qui nous permet de développer des applications de vision par ordinateur en temps réel. Elle se concentre principalement sur le traitement des images en temps réel., la capture et l'analyse vidéo, y compris des fonctionnalités telles que la détection des visages et des objets."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GTDF8xVxkhrs"
      },
      "source": [
        "training_data = [] # la liste qui contiendra les données d'entraînement traitées\n",
        "def prepare_training_data():\n",
        "    for category in CATEGORIES :\n",
        "        path = os.path.join(TRAIN_DATA_DIR, category)\n",
        "        class_num = CATEGORIES.index(category)  # obtention du classement (0 à 32)\n",
        "        for img in os.listdir(path): # itérer sur chaque image dans un dossier de caractères\n",
        "            img_array = cv.imread(os.path.join(path, img), cv.IMREAD_GRAYSCALE) # read graysacled images\n",
        "            new_array = cv.resize(img_array, (50,50)) # redimensionnement des images\n",
        "            training_data.append([new_array, class_num]) # append processed image to the training data list\n",
        "\n",
        "prepare_training_data() # calling the fct"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZslzhLOz64Ib"
      },
      "source": [
        "testing_data = [] # a list that will contain processed testing data\n",
        "def prepare_testing_data():\n",
        "    for category in CATEGORIES :\n",
        "        path = os.path.join(TEST_DATA_DIR, category)\n",
        "        class_num = CATEGORIES.index(category)  # get the classification  (0 to 32)\n",
        "        for img in os.listdir(path): # iterate over each image in a caracter folder \n",
        "            img_array = cv.imread(os.path.join(path, img), cv.IMREAD_GRAYSCALE) # read graysacled images\n",
        "            new_array = cv.resize(img_array, (50,50)) # resizing images\n",
        "            testing_data.append([new_array, class_num]) # append processed image to the testing data list\n",
        "\n",
        "prepare_testing_data() # calling the fct"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mbCC07Y3coSH"
      },
      "source": [
        "Nous voulons mélanger les données. Pour l'instant, nos données sont sous forme de : \"ya,\"yab\", puis \"yach\" ... . Cela finit généralement par causer des problèmes, car, au début, le classificateur apprendra à prédire toujours \"ya\". Ensuite, il passera à prédire \"yab\" ..."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z4nf3AKE7GDh"
      },
      "source": [
        "# shuffle the  data randomly \n",
        "# random.shuffle(training_data)\n",
        "random.shuffle(testing_data)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n-_F3RpVwN5d"
      },
      "source": [
        "\n",
        "#### Création du modèle"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C1vgeaDU7fHa"
      },
      "source": [
        "# Créer le modèle\n",
        "x_train = [] \n",
        "y_train = []\n",
        "\n",
        "for features,label in training_data:\n",
        "    x_train.append(features)\n",
        "    y_train.append(label)\n",
        "\n",
        "x_train = np.array(x_train)\n",
        "y_train = np.array(y_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vfg7eK6g88Xg"
      },
      "source": [
        "# Créer le modèle\n",
        "x_test = [] \n",
        "y_test = []\n",
        "\n",
        "for features,label in testing_data:\n",
        "    x_test.append(features)\n",
        "    y_test.append(label)\n",
        "\n",
        "x_test = np.array(x_test)\n",
        "y_test = np.array(y_test)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DeYWj6QXqgeu"
      },
      "source": [
        "\n",
        "#### Sauvegarde du modèle"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s4lC8VIOb_kY"
      },
      "source": [
        "Le module <b> pickle </b> met en œuvre des protocoles binaires pour sérialiser et désérialiser une structure d'objet Python. <br><br> Le \"pickling\" est le processus par lequel une hiérarchie d'objets Python est convertie en un flux d'octets."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VzzQfQdu7rnu"
      },
      "source": [
        "\n",
        "# Sauvegarder le modèle\n",
        "\n",
        "pickle_out = open(TRAIN_MODEL_DIR + \"x_train.pickle\",\"wb\")\n",
        "pickle.dump(x_train, pickle_out)\n",
        "pickle_out.close()\n",
        "\n",
        "pickle_out = open(TRAIN_MODEL_DIR + \"y_train.pickle\",\"wb\")\n",
        "pickle.dump(y_train, pickle_out)\n",
        "pickle_out.close()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oo_g1sU_9Okt"
      },
      "source": [
        "# Sauvegarder le modèle\n",
        "\n",
        "pickle_out = open(TEST_MODEL_DIR + \"x_test.pickle\",\"wb\")\n",
        "pickle.dump(x_test, pickle_out)\n",
        "pickle_out.close()\n",
        "\n",
        "pickle_out = open(TEST_MODEL_DIR + \"y_test.pickle\",\"wb\")\n",
        "pickle.dump(y_test, pickle_out)\n",
        "pickle_out.close()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a9s-a1jwqqQQ"
      },
      "source": [
        "\n",
        "#### Téléchargement du modèle"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_PRFC3SurZOH"
      },
      "source": [
        "Le \"unpickling\" est l'opération inverse, par laquelle un flux d'octets (provenant d'un fichier binaire ou d'un objet de type octet) est reconverti en une hiérarchie d'objets."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a8tLrB3_72kp"
      },
      "source": [
        "# to load the model we use the script\n",
        "pickle_in = open(TRAIN_MODEL_DIR + \"x_train.pickle\",\"rb\")\n",
        "x_train = pickle.load(pickle_in)\n",
        "\n",
        "pickle_in = open(TRAIN_MODEL_DIR + \"y_train.pickle\",\"rb\")\n",
        "y_train = pickle.load(pickle_in)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rwSHBy139ayT"
      },
      "source": [
        "# to load the model we use the script\n",
        "pickle_in = open(TEST_MODEL_DIR + \"x_test.pickle\",\"rb\")\n",
        "x_test = pickle.load(pickle_in)\n",
        "\n",
        "pickle_in = open(TEST_MODEL_DIR + \"y_test.pickle\",\"rb\")\n",
        "y_test = pickle.load(pickle_in)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "04UMVpJcIDiv"
      },
      "source": [
        "# CRÉATION DU RÉSEAU NEURONAL CONVOLUTIF\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9I5wu7iWBicv"
      },
      "source": [
        "#### Dimensions de Donnees pour plus de commodité"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o2A6pz5O_l0k"
      },
      "source": [
        "# The number of pixels in each dimension of an image.\n",
        "img_w = 50\n",
        "img_h = 50\n",
        "\n",
        "# The images are stored in one-dimensional arrays of this length.\n",
        "img_size_flat = img_w*img_h\n",
        "\n",
        "# Tuple with height and width of images used to reshape arrays.\n",
        "img_shape = (img_w, img_h)\n",
        "\n",
        "# Number of classes, one class for each of 33 alphabet.\n",
        "num_classes = 33\n",
        "\n",
        "# Number of colour channels for the images: 1 channel for gray-scale.\n",
        "num_channels = 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MKc7mXkVQHP5"
      },
      "source": [
        "\n",
        "#### Structuration des entrées"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7k1w_AkRNO5g"
      },
      "source": [
        "# to reduce the computition we convert the numbers from float64 => float32\n",
        "# we normalise (deviding all by 255.0) to make numbers between 0 => 1 instead of 0 => 255\n",
        "x_train = x_train.reshape(-1,50,50,1).astype(\"float32\") / 255.0 \n",
        "x_test = x_test.reshape(-1,50,50,1).astype(\"float32\") / 255.0\n",
        "\n",
        "y_train = keras.utils.to_categorical(y_train, 33)\n",
        "y_test = keras.utils.to_categorical(y_test, 33)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lDznGxmqAwJB"
      },
      "source": [
        "#### CRÉATION DU MODÉLE VGG16"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vkDu2lrKAvqo"
      },
      "source": [
        "def VGG16(width, height, depth, classes):\n",
        "\t\t# initialize the model along with the input shape to be\n",
        "\t\t# \"channels last\" and the channels dimension itself\n",
        "\t\tmodel = Sequential()\n",
        "\t\tinputShape = (height, width, depth)\n",
        "\t\tchanDim = -1\n",
        "\t\t# CONV => RELU => POOL layer set\n",
        "\t\tmodel.add(Conv2D(32, (3, 3), padding=\"same\", input_shape=inputShape))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\t\t#model.add(Dropout(0.25))\n",
        "\n",
        "\t\t# (CONV => RELU) * 2 => POOL layer set\n",
        "\t\tmodel.add(Conv2D(64, (3, 3), padding=\"same\"))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(Conv2D(64, (3, 3), padding=\"same\"))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\t\tmodel.add(Dropout(0.40))\n",
        "\n",
        "\t\t# (CONV => RELU) * 3 => POOL layer set\n",
        "\t\tmodel.add(Conv2D(128, (3, 3), padding=\"same\"))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(Conv2D(128, (3, 3), padding=\"same\"))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(Conv2D(128, (3, 3), padding=\"same\"))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization(axis=chanDim))\n",
        "\t\tmodel.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\t\tmodel.add(Dropout(0.40))\n",
        "\n",
        "\t\t# first (and only) set of FC => RELU layers\n",
        "\t\tmodel.add(Flatten())\n",
        "\t\tmodel.add(Dense(512))\n",
        "\t\tmodel.add(Activation(\"relu\"))\n",
        "\t\tmodel.add(BatchNormalization())\n",
        "\t\tmodel.add(Dropout(0.5))\n",
        "\n",
        "\t\t# softmax classifier\n",
        "\t\tmodel.add(Dense(classes))\n",
        "\t\tmodel.add(Activation(\"softmax\"))\n",
        "\n",
        "\t\t# return the constructed network architecture\n",
        "\t\treturn model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0NBU-pb5BUW0"
      },
      "source": [
        "model = VGG16(width=img_w, height=img_h, depth=num_channels, classes=num_classes)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0WXSv19Cz4K1"
      },
      "source": [
        "#### Specification des parametres de training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HINrWgfYBqWz"
      },
      "source": [
        "# and batch size\n",
        "INIT_LR = 0.001 #  the learning rate is a tuning parameter in an optimization algorithm that determines the step size at each iteration while moving toward a minimum of a loss function\n",
        "EPOCHS = 15 # \n",
        "BS = 16 # the number of training examples utilized in one iteration\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Rmd88NST0esC"
      },
      "source": [
        " #### Défintion de loss function,  l'optimizer et la metrics."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rAzB2YU0dSTm",
        "outputId": "4b598f61-b023-451f-c051-1464cebdd29d"
      },
      "source": [
        "print(\"[INFO] training network...\")\n",
        "opt = keras.optimizers.Adam(lr= INIT_LR, decay= INIT_LR / EPOCHS)\n",
        "model.compile(loss=\"categorical_crossentropy\", optimizer=opt,\n",
        "\tmetrics=[\"accuracy\"])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[INFO] training network...\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JetAUuUu0DJQ"
      },
      "source": [
        "#### Informations du modéle"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "REuCQWsa4JVD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3e1842b2-6400-493d-9b0e-184c19cfd9ac"
      },
      "source": [
        "print(\"Model info... \\n \")\n",
        "print(model.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model info... \n",
            " \n",
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 50, 50, 32)        320       \n",
            "_________________________________________________________________\n",
            "activation (Activation)      (None, 50, 50, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 50, 50, 32)        128       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 25, 25, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 25, 25, 64)        18496     \n",
            "_________________________________________________________________\n",
            "activation_1 (Activation)    (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 25, 25, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 25, 25, 64)        36928     \n",
            "_________________________________________________________________\n",
            "activation_2 (Activation)    (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 25, 25, 64)        256       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 12, 12, 64)        0         \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 12, 12, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 12, 12, 128)       73856     \n",
            "_________________________________________________________________\n",
            "activation_3 (Activation)    (None, 12, 12, 128)       0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_3 (Batch (None, 12, 12, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 12, 12, 128)       147584    \n",
            "_________________________________________________________________\n",
            "activation_4 (Activation)    (None, 12, 12, 128)       0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 12, 12, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 12, 12, 128)       147584    \n",
            "_________________________________________________________________\n",
            "activation_5 (Activation)    (None, 12, 12, 128)       0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 12, 12, 128)       512       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 6, 6, 128)         0         \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 6, 6, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 4608)              0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 512)               2359808   \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_6 (Batch (None, 512)               2048      \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 33)                16929     \n",
            "_________________________________________________________________\n",
            "activation_7 (Activation)    (None, 33)                0         \n",
            "=================================================================\n",
            "Total params: 2,805,729\n",
            "Trainable params: 2,803,617\n",
            "Non-trainable params: 2,112\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AN8aqbSfV_RL"
      },
      "source": [
        "#### Intégration des données dans le modèle\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YV0b9Cz3ddgU",
        "outputId": "7d615013-3aae-4283-a150-679877a323a0"
      },
      "source": [
        "# train the network + monitor\n",
        "H = model.fit(x=x_train, y=y_train, batch_size=BS,\n",
        "              validation_data=(x_test, y_test), \n",
        "              steps_per_epoch= len(y_train) // BS, epochs=EPOCHS)\n",
        "print(\"[INFO] training Done...\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/data/ops/dataset_ops.py:3504: UserWarning: Even though the tf.config.experimental_run_functions_eagerly option is set, this option does not apply to tf.data functions. tf.data functions are still traced and executed as graphs.\n",
            "  \"Even though the tf.config.experimental_run_functions_eagerly \"\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/15\n",
            "1650/1650 [==============================] - 74s 36ms/step - loss: 0.8844 - accuracy: 0.7665 - val_loss: 0.3231 - val_accuracy: 0.8945\n",
            "Epoch 2/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.1056 - accuracy: 0.9694 - val_loss: 0.1057 - val_accuracy: 0.9697\n",
            "Epoch 3/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0767 - accuracy: 0.9774 - val_loss: 0.4342 - val_accuracy: 0.8758\n",
            "Epoch 4/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0684 - accuracy: 0.9802 - val_loss: 0.1204 - val_accuracy: 0.9652\n",
            "Epoch 5/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0415 - accuracy: 0.9874 - val_loss: 0.0641 - val_accuracy: 0.9806\n",
            "Epoch 6/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0369 - accuracy: 0.9885 - val_loss: 0.1010 - val_accuracy: 0.9700\n",
            "Epoch 7/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0347 - accuracy: 0.9892 - val_loss: 0.0339 - val_accuracy: 0.9900\n",
            "Epoch 8/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0273 - accuracy: 0.9908 - val_loss: 0.0492 - val_accuracy: 0.9867\n",
            "Epoch 9/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0214 - accuracy: 0.9936 - val_loss: 0.0790 - val_accuracy: 0.9833\n",
            "Epoch 10/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0165 - accuracy: 0.9950 - val_loss: 0.0497 - val_accuracy: 0.9873\n",
            "Epoch 11/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0208 - accuracy: 0.9938 - val_loss: 0.0240 - val_accuracy: 0.9948\n",
            "Epoch 12/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0114 - accuracy: 0.9963 - val_loss: 0.0277 - val_accuracy: 0.9952\n",
            "Epoch 13/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0121 - accuracy: 0.9966 - val_loss: 0.0243 - val_accuracy: 0.9967\n",
            "Epoch 14/15\n",
            "1650/1650 [==============================] - 59s 36ms/step - loss: 0.0098 - accuracy: 0.9971 - val_loss: 0.0254 - val_accuracy: 0.9970\n",
            "Epoch 15/15\n",
            "1650/1650 [==============================] - 58s 35ms/step - loss: 0.0111 - accuracy: 0.9961 - val_loss: 0.0244 - val_accuracy: 0.9979\n",
            "[INFO] training Done...\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cVHndBshuqe5"
      },
      "source": [
        "#### Sauvegarde du modèle entraîné"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64yc_yloAn-l",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d9af92d9-4206-4a6b-cceb-d106a7705fcc"
      },
      "source": [
        "model.save(os.path.join(MODEL_DIR,\"VGG16-4.model\"))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: /content/drive/MyDrive/genie info/Rapport PFE - HexaCoders/DATASET/model/VGG16-4.model/assets\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0H-8V26Cjy5o"
      },
      "source": [
        "#### Export the model for tonserflowjs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sAWKW2WNuBSh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 124
        },
        "outputId": "d8454a6c-8ae2-429e-8a2c-4ca0b70700d9"
      },
      "source": [
        "model.save('model4.h5')\n",
        "\n",
        "!mkdir model4\n",
        "!tensorflowjs_converter --input_format keras model4.h5 model/\n",
        "!zip -r model4.zip model \n",
        "\n",
        "from google.colab import files\n",
        "files.download('model4.zip')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2021-04-05 13:00:30.462766: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n",
            "  adding: model/ (stored 0%)\n",
            "  adding: model/group1-shard3of3.bin (deflated 7%)\n",
            "  adding: model/group1-shard1of3.bin (deflated 7%)\n",
            "  adding: model/group1-shard2of3.bin (deflated 7%)\n",
            "  adding: model/model.json (deflated 90%)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "download(\"download_c35dfac3-2212-47a9-83e7-484ec97f7f94\", \"model4.zip\", 10398963)"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R01nd_mIVIEz"
      },
      "source": [
        "#### Load the trained model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NFOdMkaFVMLd"
      },
      "source": [
        "model = keras.models.load_model(os.path.join(MODEL_DIR,\"VGG16-4.model\"))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kG9eblfrvq0u"
      },
      "source": [
        "## Statistics"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Gc8g2VYD3OLm"
      },
      "source": [
        "#### Model Accuracy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gOoP3X4E5Diu",
        "outputId": "9beb6cbc-eaa5-405e-9d31-b4183a60bb7f"
      },
      "source": [
        "scores = model.evaluate (x_test, y_test, verbose=0)\n",
        "print(\"Accuracy = \"+str(scores[1]*100)+\" \\n\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/data/ops/dataset_ops.py:3504: UserWarning: Even though the tf.config.experimental_run_functions_eagerly option is set, this option does not apply to tf.data functions. tf.data functions are still traced and executed as graphs.\n",
            "  \"Even though the tf.config.experimental_run_functions_eagerly \"\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Accuracy = 99.78787899017334 \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lrie6MxS3XSo"
      },
      "source": [
        "Evolution of Loss and Accuracy by epoches"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 581
        },
        "id": "TY3FJpC06a8K",
        "outputId": "2143f4e3-0c58-4c44-e5ef-cfc6000cad56"
      },
      "source": [
        "# plot the training loss \n",
        "N = np.arange(0, EPOCHS)\n",
        "plt.style.use(\"ggplot\")\n",
        "plt.figure()\n",
        "plt.plot(N, H.history[\"loss\"], label=\"train_loss\")\n",
        "plt.plot(N, H.history[\"val_loss\"], label=\"val_loss\")\n",
        "plt.title(\"Loss: VGG_Net \")\n",
        "plt.xlabel(\"Epoch #\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.legend()\n",
        "plt.savefig(\"Loss_plot_256.png\")\n",
        "files.download(\"Loss_plot_256.png\")\n",
        "\n",
        "# plot the training accuracy\n",
        "N = np.arange(0, EPOCHS)\n",
        "plt.style.use(\"ggplot\")\n",
        "plt.figure()\n",
        "plt.plot(N, H.history[\"accuracy\"], label=\"train_acc\") \n",
        "plt.plot(N, H.history[\"val_accuracy\"], label=\"val_acc\")\n",
        "plt.title(\"Accuracy: VGG16 By HexaCoders\")\n",
        "plt.xlabel(\"Epoch #\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend()\n",
        "plt.savefig(\"Accuracy_plot_256.png\")\n",
        "files.download(\"Accuracy_plot_256.png\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "download(\"download_3d4b9195-906d-4288-a0ac-9053a85d7abb\", \"Loss_plot_256.png\", 19526)"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "download(\"download_0a36dbf0-b432-4a03-b81c-40e674c759f9\", \"Accuracy_plot_256.png\", 22792)"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8WBVK3qD3hDX"
      },
      "source": [
        "#### Confusion Matrix"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 952
        },
        "id": "MuGdD7DN6ro-",
        "outputId": "b96277e7-437b-445f-9d4c-3d50d8999ff9"
      },
      "source": [
        "# ploting confision matrix\n",
        "\n",
        "y_pred = model.predict_classes(x_test)\n",
        "y_test_mat=np.argmax(y_test, axis=1)\n",
        "mat = confusion_matrix(y_test_mat, y_pred)\n",
        "plt.figure()\n",
        "plot_confusion_matrix(conf_mat=mat, figsize=(14, 14))\n",
        "plt.savefig(\"conf_mat.png\")\n",
        "files.download(\"conf_mat.png\")\n",
        "print(\"Ploting Conf_Mat..................\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/sequential.py:450: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "  warnings.warn('`model.predict_classes()` is deprecated and '\n",
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/data/ops/dataset_ops.py:3504: UserWarning: Even though the tf.config.experimental_run_functions_eagerly option is set, this option does not apply to tf.data functions. tf.data functions are still traced and executed as graphs.\n",
            "  \"Even though the tf.config.experimental_run_functions_eagerly \"\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "\n",
              "    async function download(id, filename, size) {\n",
              "      if (!google.colab.kernel.accessAllowed) {\n",
              "        return;\n",
              "      }\n",
              "      const div = document.createElement('div');\n",
              "      const label = document.createElement('label');\n",
              "      label.textContent = `Downloading \"${filename}\": `;\n",
              "      div.appendChild(label);\n",
              "      const progress = document.createElement('progress');\n",
              "      progress.max = size;\n",
              "      div.appendChild(progress);\n",
              "      document.body.appendChild(div);\n",
              "\n",
              "      const buffers = [];\n",
              "      let downloaded = 0;\n",
              "\n",
              "      const channel = await google.colab.kernel.comms.open(id);\n",
              "      // Send a message to notify the kernel that we're ready.\n",
              "      channel.send({})\n",
              "\n",
              "      for await (const message of channel.messages) {\n",
              "        // Send a message to notify the kernel that we're ready.\n",
              "        channel.send({})\n",
              "        if (message.buffers) {\n",
              "          for (const buffer of message.buffers) {\n",
              "            buffers.push(buffer);\n",
              "            downloaded += buffer.byteLength;\n",
              "            progress.value = downloaded;\n",
              "          }\n",
              "        }\n",
              "      }\n",
              "      const blob = new Blob(buffers, {type: 'application/binary'});\n",
              "      const a = document.createElement('a');\n",
              "      a.href = window.URL.createObjectURL(blob);\n",
              "      a.download = filename;\n",
              "      div.appendChild(a);\n",
              "      a.click();\n",
              "      div.remove();\n",
              "    }\n",
              "  "
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "download(\"download_9cff66c4-9f1d-4039-866b-a3aacf2dfed5\", \"conf_mat.png\", 36290)"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Ploting Conf_Mat..................\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1008x1008 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}