inceptionv4_satellite_image_classification_Code.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:54:37.179668Z",
          "iopub.status.busy": "2023-06-07T14:54:37.179196Z",
          "iopub.status.idle": "2023-06-07T14:54:37.199162Z",
          "shell.execute_reply": "2023-06-07T14:54:37.198071Z",
          "shell.execute_reply.started": "2023-06-07T14:54:37.179589Z"
        },
        "id": "kDo9K-HyJDF6"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import glob\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import shutil\n",
        "import random\n",
        "import cv2\n",
        "import matplotlib.pyplot as plt\n",
        "from matplotlib.image import imread\n",
        "import seaborn as sns\n",
        "sns.set(rc={'figure.figsize':(10,6)})\n",
        "sns.set(font_scale=1.3)\n",
        "plt.style.use('fivethirtyeight')\n",
        "\n",
        "\n",
        "!pip install -U -q pyDrive\n",
        "from pydrive.auth import GoogleAuth\n",
        "from pydrive.drive import GoogleDrive\n",
        "from google.colab import auth\n",
        "from oauth2client.client import GoogleCredentials\n",
        "\n",
        "auth.authenticate_user()\n",
        "gauth = GoogleAuth()\n",
        "gauth.credentials = GoogleCredentials.get_application_default()\n",
        "drive = GoogleDrive(gauth)\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')\n",
        "\n",
        "\n",
        "!pip install tensorflow\n",
        "\n",
        "!pip install -q keras\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "# Backend\n",
        "from keras import backend as K\n",
        "# Utils\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from keras.models import Model\n",
        "from tensorflow.keras import layers\n",
        "from keras.layers import concatenate\n",
        "from tensorflow.keras.utils import plot_model\n",
        "from keras.layers import Conv2D , MaxPool2D , Input , GlobalAveragePooling2D ,AveragePooling2D, Dense , Dropout ,Activation , BatchNormalization\n",
        "#from keras.utils.data_utils import get_file\n",
        "from tensorflow.keras.applications.resnet50 import preprocess_input\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "#from keras.callbacks import EarlyStopping\n",
        "from tensorflow.keras.callbacks import EarlyStopping, ReduceLROnPlateau\n",
        "from keras.regularizers import l2\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings(action=\"ignore\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:54:39.542243Z",
          "iopub.status.busy": "2023-06-07T14:54:39.541527Z",
          "iopub.status.idle": "2023-06-07T14:54:52.950946Z",
          "shell.execute_reply": "2023-06-07T14:54:52.949665Z",
          "shell.execute_reply.started": "2023-06-07T14:54:39.542211Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_q6VTnB-JDGK",
        "outputId": "5bb49ed8-43ca-43c7-e3eb-69082cf5cbc7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting split-folders\n",
            "  Downloading split_folders-0.5.1-py3-none-any.whl (8.4 kB)\n",
            "Installing collected packages: split-folders\n",
            "Successfully installed split-folders-0.5.1\n"
          ]
        }
      ],
      "source": [
        "! pip install split-folders"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:54:52.954463Z",
          "iopub.status.busy": "2023-06-07T14:54:52.953999Z",
          "iopub.status.idle": "2023-06-07T14:54:52.965626Z",
          "shell.execute_reply": "2023-06-07T14:54:52.964513Z",
          "shell.execute_reply.started": "2023-06-07T14:54:52.954415Z"
        },
        "id": "7wzQUntmJDGP"
      },
      "outputs": [],
      "source": [
        "import splitfolders"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:54:52.969152Z",
          "iopub.status.busy": "2023-06-07T14:54:52.968841Z",
          "iopub.status.idle": "2023-06-07T14:55:11.426833Z",
          "shell.execute_reply": "2023-06-07T14:55:11.425804Z",
          "shell.execute_reply.started": "2023-06-07T14:54:52.969127Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Z6-tuWggJDGR",
        "outputId": "c30c7f1a-22a1-465d-c199-7830531e3ce7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Copying files: 5631 files [17:44,  5.29 files/s]\n"
          ]
        }
      ],
      "source": [
        "input_folder =\"/content/gdrive/MyDrive/Colab Notebooks/data1\"\n",
        "\n",
        "output =\"/content/gdrive/MyDrive/Colab Notebooks/data2\"\n",
        "\n",
        "splitfolders.ratio(input_folder, output=output, seed=42, ratio=(0.7, 0.3), group_prefix=None)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:55:11.430444Z",
          "iopub.status.busy": "2023-06-07T14:55:11.429808Z",
          "iopub.status.idle": "2023-06-07T14:55:11.557190Z",
          "shell.execute_reply": "2023-06-07T14:55:11.556263Z",
          "shell.execute_reply.started": "2023-06-07T14:55:11.430414Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qQMniHsFJDGS",
        "outputId": "1ba69c29-4bf5-4fa9-93e8-f08b3476ff48"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 3941 images belonging to 4 classes.\n"
          ]
        }
      ],
      "source": [
        "train_datagen = ImageDataGenerator(preprocessing_function=preprocess_input,rescale = 1./255,\n",
        "                                   shear_range = 0.2,\n",
        "                                   zoom_range = 0.2,\n",
        "                                   horizontal_flip = True)\n",
        "training_set = train_datagen.flow_from_directory('/content/gdrive/MyDrive/Colab Notebooks/data2/train',\n",
        "                                                 target_size = (224, 224),\n",
        "                                                 batch_size = 32,\n",
        "                                                 class_mode = 'categorical',\n",
        "                                                 shuffle=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:55:11.560371Z",
          "iopub.status.busy": "2023-06-07T14:55:11.559656Z",
          "iopub.status.idle": "2023-06-07T14:55:11.565251Z",
          "shell.execute_reply": "2023-06-07T14:55:11.564196Z",
          "shell.execute_reply.started": "2023-06-07T14:55:11.560336Z"
        },
        "id": "Ngo2U3h5JDGU"
      },
      "outputs": [],
      "source": [
        "# test_datagen = ImageDataGenerator(preprocessing_function=preprocess_input, rescale = 1./255)\n",
        "# test_set = test_datagen.flow_from_directory('/working/dataset/test',\n",
        "#                                             target_size = (224, 224),\n",
        "#                                             batch_size = 32,\n",
        "#                                             class_mode = 'categorical')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:55:11.567226Z",
          "iopub.status.busy": "2023-06-07T14:55:11.566843Z",
          "iopub.status.idle": "2023-06-07T14:55:11.630854Z",
          "shell.execute_reply": "2023-06-07T14:55:11.629965Z",
          "shell.execute_reply.started": "2023-06-07T14:55:11.567163Z"
        },
        "id": "ncseUlK7JDGW",
        "outputId": "ed8a7df5-fa3e-4893-8d4e-c289d40c3b87",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 1690 images belonging to 4 classes.\n"
          ]
        }
      ],
      "source": [
        "val_datagen = ImageDataGenerator(preprocessing_function=preprocess_input,rescale = 1./255)\n",
        "val_set = val_datagen.flow_from_directory('/content/gdrive/MyDrive/Colab Notebooks/data2/val',\n",
        "                                            target_size = (224, 224),\n",
        "                                            batch_size = 32,\n",
        "                                            class_mode = 'categorical',\n",
        "                                            shuffle=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:55:11.632782Z",
          "iopub.status.busy": "2023-06-07T14:55:11.632368Z",
          "iopub.status.idle": "2023-06-07T14:55:11.640209Z",
          "shell.execute_reply": "2023-06-07T14:55:11.637115Z",
          "shell.execute_reply.started": "2023-06-07T14:55:11.632747Z"
        },
        "id": "dTA87WawJDGY"
      },
      "outputs": [],
      "source": [
        "early = EarlyStopping(monitor='val_loss',patience=5)\n",
        "reduce = ReduceLROnPlateau(monitor='val_loss', patience=3, factor=0.2, mil_lr=0.001)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T14:55:11.642645Z",
          "iopub.status.busy": "2023-06-07T14:55:11.642197Z",
          "iopub.status.idle": "2023-06-07T14:55:11.650360Z",
          "shell.execute_reply": "2023-06-07T14:55:11.649226Z",
          "shell.execute_reply.started": "2023-06-07T14:55:11.642578Z"
        },
        "id": "OSqJ_dJfJDGa"
      },
      "outputs": [],
      "source": [
        "output_classes = 4\n",
        "batch_size = 32\n",
        "epochs = 10\n",
        "image_shape=(224,224,3)\n",
        "nb_train_samples = 3941\n",
        "nb_validation_samples = 1690\n",
        "# nb_test_samples = 564"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:00:28.089512Z",
          "iopub.status.busy": "2023-06-07T15:00:28.089106Z",
          "iopub.status.idle": "2023-06-07T15:00:28.134520Z",
          "shell.execute_reply": "2023-06-07T15:00:28.133394Z",
          "shell.execute_reply.started": "2023-06-07T15:00:28.089482Z"
        },
        "id": "cwtwBj2WJDGb"
      },
      "outputs": [],
      "source": [
        "# Building the Inception Model Architecture\n",
        "\n",
        "def InceptionV4():\n",
        "\n",
        "    input_layer = Input(shape=image_shape)\n",
        "\n",
        "    x = stemBlock(prev_layer=input_layer)\n",
        "\n",
        "    x = InceptionBlock_A(prev_layer=x)\n",
        "    x = InceptionBlock_A(prev_layer=x)\n",
        "    x = InceptionBlock_A(prev_layer=x)\n",
        "    x = InceptionBlock_A(prev_layer=x)\n",
        "\n",
        "    x = reduction_A_Block(prev_layer=x)\n",
        "\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "    x = InceptionBlock_B(prev_layer=x)\n",
        "\n",
        "    x = reduction_B_Block(prev_layer= x)\n",
        "\n",
        "    x = InceptionBlock_C(prev_layer=x)\n",
        "    x = InceptionBlock_C(prev_layer=x)\n",
        "    x = InceptionBlock_C(prev_layer=x)\n",
        "\n",
        "    x = GlobalAveragePooling2D()(x)\n",
        "\n",
        "    x = Dense(units = 1024, activation='relu') (x)\n",
        "    x = Dropout(rate = 0.8) (x)\n",
        "    x = Dense(units = output_classes, activation='softmax')(x)\n",
        "\n",
        "    model = Model(inputs = input_layer , outputs = x , name ='Inception-V4')\n",
        "\n",
        "    return model\n",
        "\n",
        "\n",
        "def conv2d_with_Batch(prev_layer , nbr_kernels , filter_size , strides = (1,1) , padding = 'valid'):\n",
        "    x = Conv2D(filters = nbr_kernels, kernel_size = filter_size, strides=strides , padding=padding) (prev_layer)\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Activation(activation = 'relu') (x)\n",
        "    return x\n",
        "\n",
        "\n",
        "def stemBlock(prev_layer):\n",
        "    x = conv2d_with_Batch(prev_layer, nbr_kernels = 32, filter_size = (3,3), strides = (2,2))\n",
        "    x = conv2d_with_Batch(x, nbr_kernels = 32, filter_size = (3,3))\n",
        "    x = conv2d_with_Batch(x, nbr_kernels = 64, filter_size = (3,3))\n",
        "\n",
        "    x_1 = conv2d_with_Batch(x, nbr_kernels = 96, filter_size = (3,3), strides = (2,2) )\n",
        "    x_2 = MaxPool2D(pool_size=(3,3) , strides=(2,2) ) (x)\n",
        "\n",
        "    x = concatenate([x_1 , x_2], axis = 3)\n",
        "\n",
        "    x_1 = conv2d_with_Batch(x, nbr_kernels = 64, filter_size = (1,1))\n",
        "    x_1 = conv2d_with_Batch(x_1, nbr_kernels = 64, filter_size = (1,7) , padding ='same')\n",
        "    x_1 = conv2d_with_Batch(x_1, nbr_kernels = 64, filter_size = (7,1), padding ='same')\n",
        "    x_1 = conv2d_with_Batch(x_1, nbr_kernels = 96, filter_size = (3,3))\n",
        "\n",
        "    x_2 = conv2d_with_Batch(x, nbr_kernels = 96, filter_size = (1,1))\n",
        "    x_2 = conv2d_with_Batch(x_2, nbr_kernels = 96, filter_size = (3,3))\n",
        "\n",
        "    x = concatenate([x_1 , x_2], axis = 3)\n",
        "\n",
        "    x_1 = conv2d_with_Batch(x, nbr_kernels = 192, filter_size = (3,3) , strides=2)\n",
        "    x_2 = MaxPool2D(pool_size=(3,3) , strides=(2,2) ) (x)\n",
        "\n",
        "    x = concatenate([x_1 , x_2], axis = 3)\n",
        "\n",
        "    return x\n",
        "\n",
        "\n",
        "def reduction_A_Block(prev_layer) :\n",
        "    x_1 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 192, filter_size = (1,1))\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 224, filter_size = (3,3) , padding='same')\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 256, filter_size = (3,3) , strides=(2,2))\n",
        "\n",
        "    x_2 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 384, filter_size = (3,3) , strides=(2,2))\n",
        "\n",
        "    x_3 = MaxPool2D(pool_size=(3,3) , strides=(2,2))(prev_layer)\n",
        "\n",
        "    x = concatenate([x_1 , x_2 , x_3], axis = 3)\n",
        "\n",
        "    return x\n",
        "\n",
        "\n",
        "def reduction_B_Block(prev_layer):\n",
        "    x_1 = MaxPool2D(pool_size=(3,3) , strides=(2,2))(prev_layer)\n",
        "\n",
        "    x_2 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 192, filter_size = (1,1))\n",
        "    x_2 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 192, filter_size = (3,3) , strides=(2,2) )\n",
        "\n",
        "    x_3 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 256, filter_size = (1,1) )\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 256, filter_size = (1,7) , padding='same')\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 320, filter_size = (7,1) , padding='same')\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 320, filter_size = (3,3) , strides=(2,2))\n",
        "\n",
        "    x = concatenate([x_1 , x_2 , x_3], axis = 3)\n",
        "    return x\n",
        "\n",
        "\n",
        "def InceptionBlock_A(prev_layer): #I'm Here\n",
        "\n",
        "    x_1 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 64, filter_size = (1,1))\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 96, filter_size = (3,3) , strides=(1,1), padding='same' )\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 96, filter_size = (3,3) , strides=(1,1) , padding='same')\n",
        "\n",
        "    x_2 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 64, filter_size = (1,1))\n",
        "    x_2 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 96, filter_size = (3,3) , padding='same')\n",
        "\n",
        "    x_3 = AveragePooling2D(pool_size=(3,3) , strides=1 , padding='same')(prev_layer)\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 96, filter_size = (1,1) , padding='same')\n",
        "\n",
        "    x_4 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 96, filter_size = (1,1))\n",
        "\n",
        "    output = concatenate([x_1 , x_2 , x_3 , x_4], axis = 3)\n",
        "\n",
        "    return output\n",
        "\n",
        "\n",
        "def InceptionBlock_B(prev_layer):\n",
        "\n",
        "    x_1 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 192, filter_size = (1,1))\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 192, filter_size = (7,1) , padding='same')\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 224, filter_size = (1,7) , padding='same')\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 224, filter_size = (7,1) , padding='same')\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 256, filter_size = (1,7), padding='same')\n",
        "\n",
        "    x_2 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 192, filter_size = (1,1))\n",
        "    x_2 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 224, filter_size = (1,7) , padding='same')\n",
        "    x_2 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 256, filter_size = (7,1), padding='same')\n",
        "\n",
        "    x_3 = AveragePooling2D(pool_size=(3,3) , strides=1 , padding='same')(prev_layer)\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 128, filter_size = (1,1))\n",
        "\n",
        "    x_4 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 384, filter_size = (1,1))\n",
        "\n",
        "    output = concatenate([x_1 , x_2 ,x_3, x_4], axis = 3)\n",
        "    return output\n",
        "\n",
        "\n",
        "def InceptionBlock_C(prev_layer):\n",
        "\n",
        "    x_1 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 384, filter_size = (1,1))\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 448, filter_size = (3,1) , padding='same')\n",
        "    x_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 512, filter_size = (1,3) , padding='same')\n",
        "    x_1_1 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 256, filter_size = (1,3), padding='same')\n",
        "    x_1_2 = conv2d_with_Batch(prev_layer = x_1, nbr_kernels = 256, filter_size = (3,1), padding='same')\n",
        "    x_1 = concatenate([x_1_1 , x_1_2], axis = 3)\n",
        "\n",
        "    x_2 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 384, filter_size = (1,1))\n",
        "    x_2_1 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 256, filter_size = (1,3), padding='same')\n",
        "    x_2_2 = conv2d_with_Batch(prev_layer = x_2, nbr_kernels = 256, filter_size = (3,1), padding='same')\n",
        "    x_2 = concatenate([x_2_1 , x_2_2], axis = 3)\n",
        "\n",
        "    x_3 = MaxPool2D(pool_size=(3,3),strides = 1 , padding='same')(prev_layer)\n",
        "    x_3 = conv2d_with_Batch(prev_layer = x_3, nbr_kernels = 256, filter_size = 3  , padding='same')\n",
        "\n",
        "    x_4 = conv2d_with_Batch(prev_layer = prev_layer, nbr_kernels = 256, filter_size = (1,1))\n",
        "\n",
        "    output = concatenate([x_1 , x_2 , x_3 , x_4], axis = 3)\n",
        "\n",
        "    return output"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:00:30.959083Z",
          "iopub.status.busy": "2023-06-07T15:00:30.958709Z",
          "iopub.status.idle": "2023-06-07T15:00:39.873891Z",
          "shell.execute_reply": "2023-06-07T15:00:39.873086Z",
          "shell.execute_reply.started": "2023-06-07T15:00:30.959055Z"
        },
        "id": "QglQOSofJDGe",
        "outputId": "35e9bfc1-d26a-4703-e9a3-0925634312df",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"Inception-V4\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                Output Shape                 Param #   Connected to                  \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)        [(None, 224, 224, 3)]        0         []                            \n",
            "                                                                                                  \n",
            " conv2d (Conv2D)             (None, 111, 111, 32)         896       ['input_1[0][0]']             \n",
            "                                                                                                  \n",
            " batch_normalization (Batch  (None, 111, 111, 32)         128       ['conv2d[0][0]']              \n",
            " Normalization)                                                                                   \n",
            "                                                                                                  \n",
            " activation (Activation)     (None, 111, 111, 32)         0         ['batch_normalization[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_1 (Conv2D)           (None, 109, 109, 32)         9248      ['activation[0][0]']          \n",
            "                                                                                                  \n",
            " batch_normalization_1 (Bat  (None, 109, 109, 32)         128       ['conv2d_1[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_1 (Activation)   (None, 109, 109, 32)         0         ['batch_normalization_1[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " conv2d_2 (Conv2D)           (None, 107, 107, 64)         18496     ['activation_1[0][0]']        \n",
            "                                                                                                  \n",
            " batch_normalization_2 (Bat  (None, 107, 107, 64)         256       ['conv2d_2[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_2 (Activation)   (None, 107, 107, 64)         0         ['batch_normalization_2[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " conv2d_3 (Conv2D)           (None, 53, 53, 96)           55392     ['activation_2[0][0]']        \n",
            "                                                                                                  \n",
            " batch_normalization_3 (Bat  (None, 53, 53, 96)           384       ['conv2d_3[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_3 (Activation)   (None, 53, 53, 96)           0         ['batch_normalization_3[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " max_pooling2d (MaxPooling2  (None, 53, 53, 64)           0         ['activation_2[0][0]']        \n",
            " D)                                                                                               \n",
            "                                                                                                  \n",
            " concatenate (Concatenate)   (None, 53, 53, 160)          0         ['activation_3[0][0]',        \n",
            "                                                                     'max_pooling2d[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_4 (Conv2D)           (None, 53, 53, 64)           10304     ['concatenate[0][0]']         \n",
            "                                                                                                  \n",
            " batch_normalization_4 (Bat  (None, 53, 53, 64)           256       ['conv2d_4[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_4 (Activation)   (None, 53, 53, 64)           0         ['batch_normalization_4[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " conv2d_5 (Conv2D)           (None, 53, 53, 64)           28736     ['activation_4[0][0]']        \n",
            "                                                                                                  \n",
            " batch_normalization_5 (Bat  (None, 53, 53, 64)           256       ['conv2d_5[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_5 (Activation)   (None, 53, 53, 64)           0         ['batch_normalization_5[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " conv2d_6 (Conv2D)           (None, 53, 53, 64)           28736     ['activation_5[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_8 (Conv2D)           (None, 53, 53, 96)           15456     ['concatenate[0][0]']         \n",
            "                                                                                                  \n",
            " batch_normalization_6 (Bat  (None, 53, 53, 64)           256       ['conv2d_6[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " batch_normalization_8 (Bat  (None, 53, 53, 96)           384       ['conv2d_8[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_6 (Activation)   (None, 53, 53, 64)           0         ['batch_normalization_6[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " activation_8 (Activation)   (None, 53, 53, 96)           0         ['batch_normalization_8[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " conv2d_7 (Conv2D)           (None, 51, 51, 96)           55392     ['activation_6[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_9 (Conv2D)           (None, 51, 51, 96)           83040     ['activation_8[0][0]']        \n",
            "                                                                                                  \n",
            " batch_normalization_7 (Bat  (None, 51, 51, 96)           384       ['conv2d_7[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " batch_normalization_9 (Bat  (None, 51, 51, 96)           384       ['conv2d_9[0][0]']            \n",
            " chNormalization)                                                                                 \n",
            "                                                                                                  \n",
            " activation_7 (Activation)   (None, 51, 51, 96)           0         ['batch_normalization_7[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " activation_9 (Activation)   (None, 51, 51, 96)           0         ['batch_normalization_9[0][0]'\n",
            "                                                                    ]                             \n",
            "                                                                                                  \n",
            " concatenate_1 (Concatenate  (None, 51, 51, 192)          0         ['activation_7[0][0]',        \n",
            " )                                                                   'activation_9[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_10 (Conv2D)          (None, 25, 25, 192)          331968    ['concatenate_1[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_10 (Ba  (None, 25, 25, 192)          768       ['conv2d_10[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_10 (Activation)  (None, 25, 25, 192)          0         ['batch_normalization_10[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " max_pooling2d_1 (MaxPoolin  (None, 25, 25, 192)          0         ['concatenate_1[0][0]']       \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " concatenate_2 (Concatenate  (None, 25, 25, 384)          0         ['activation_10[0][0]',       \n",
            " )                                                                   'max_pooling2d_1[0][0]']     \n",
            "                                                                                                  \n",
            " conv2d_11 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_2[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_11 (Ba  (None, 25, 25, 64)           256       ['conv2d_11[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_11 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_11[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_12 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_11[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_14 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_2[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_12 (Ba  (None, 25, 25, 96)           384       ['conv2d_12[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_14 (Ba  (None, 25, 25, 64)           256       ['conv2d_14[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_12 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_12[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_14 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_14[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d (Average  (None, 25, 25, 384)          0         ['concatenate_2[0][0]']       \n",
            " Pooling2D)                                                                                       \n",
            "                                                                                                  \n",
            " conv2d_13 (Conv2D)          (None, 25, 25, 96)           83040     ['activation_12[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_15 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_14[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_16 (Conv2D)          (None, 25, 25, 96)           36960     ['average_pooling2d[0][0]']   \n",
            "                                                                                                  \n",
            " conv2d_17 (Conv2D)          (None, 25, 25, 96)           36960     ['concatenate_2[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_13 (Ba  (None, 25, 25, 96)           384       ['conv2d_13[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_15 (Ba  (None, 25, 25, 96)           384       ['conv2d_15[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_16 (Ba  (None, 25, 25, 96)           384       ['conv2d_16[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_17 (Ba  (None, 25, 25, 96)           384       ['conv2d_17[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_13 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_13[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_15 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_15[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_16 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_16[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_17 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_17[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_3 (Concatenate  (None, 25, 25, 384)          0         ['activation_13[0][0]',       \n",
            " )                                                                   'activation_15[0][0]',       \n",
            "                                                                     'activation_16[0][0]',       \n",
            "                                                                     'activation_17[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_18 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_3[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_18 (Ba  (None, 25, 25, 64)           256       ['conv2d_18[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_18 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_18[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_19 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_18[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_21 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_3[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_19 (Ba  (None, 25, 25, 96)           384       ['conv2d_19[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_21 (Ba  (None, 25, 25, 64)           256       ['conv2d_21[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_19 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_19[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_21 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_21[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_1 (Avera  (None, 25, 25, 384)          0         ['concatenate_3[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_20 (Conv2D)          (None, 25, 25, 96)           83040     ['activation_19[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_22 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_21[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_23 (Conv2D)          (None, 25, 25, 96)           36960     ['average_pooling2d_1[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_24 (Conv2D)          (None, 25, 25, 96)           36960     ['concatenate_3[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_20 (Ba  (None, 25, 25, 96)           384       ['conv2d_20[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_22 (Ba  (None, 25, 25, 96)           384       ['conv2d_22[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_23 (Ba  (None, 25, 25, 96)           384       ['conv2d_23[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_24 (Ba  (None, 25, 25, 96)           384       ['conv2d_24[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_20 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_20[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_22 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_22[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_23 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_23[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_24 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_24[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_4 (Concatenate  (None, 25, 25, 384)          0         ['activation_20[0][0]',       \n",
            " )                                                                   'activation_22[0][0]',       \n",
            "                                                                     'activation_23[0][0]',       \n",
            "                                                                     'activation_24[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_25 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_4[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_25 (Ba  (None, 25, 25, 64)           256       ['conv2d_25[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_25 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_25[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_26 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_25[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_28 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_4[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_26 (Ba  (None, 25, 25, 96)           384       ['conv2d_26[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_28 (Ba  (None, 25, 25, 64)           256       ['conv2d_28[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_26 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_26[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_28 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_28[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_2 (Avera  (None, 25, 25, 384)          0         ['concatenate_4[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_27 (Conv2D)          (None, 25, 25, 96)           83040     ['activation_26[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_29 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_28[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_30 (Conv2D)          (None, 25, 25, 96)           36960     ['average_pooling2d_2[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_31 (Conv2D)          (None, 25, 25, 96)           36960     ['concatenate_4[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_27 (Ba  (None, 25, 25, 96)           384       ['conv2d_27[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_29 (Ba  (None, 25, 25, 96)           384       ['conv2d_29[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_30 (Ba  (None, 25, 25, 96)           384       ['conv2d_30[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_31 (Ba  (None, 25, 25, 96)           384       ['conv2d_31[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_27 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_27[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_29 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_29[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_30 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_30[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_31 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_31[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_5 (Concatenate  (None, 25, 25, 384)          0         ['activation_27[0][0]',       \n",
            " )                                                                   'activation_29[0][0]',       \n",
            "                                                                     'activation_30[0][0]',       \n",
            "                                                                     'activation_31[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_32 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_5[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_32 (Ba  (None, 25, 25, 64)           256       ['conv2d_32[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_32 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_32[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_33 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_32[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_35 (Conv2D)          (None, 25, 25, 64)           24640     ['concatenate_5[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_33 (Ba  (None, 25, 25, 96)           384       ['conv2d_33[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_35 (Ba  (None, 25, 25, 64)           256       ['conv2d_35[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_33 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_33[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_35 (Activation)  (None, 25, 25, 64)           0         ['batch_normalization_35[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_3 (Avera  (None, 25, 25, 384)          0         ['concatenate_5[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_34 (Conv2D)          (None, 25, 25, 96)           83040     ['activation_33[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_36 (Conv2D)          (None, 25, 25, 96)           55392     ['activation_35[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_37 (Conv2D)          (None, 25, 25, 96)           36960     ['average_pooling2d_3[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_38 (Conv2D)          (None, 25, 25, 96)           36960     ['concatenate_5[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_34 (Ba  (None, 25, 25, 96)           384       ['conv2d_34[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_36 (Ba  (None, 25, 25, 96)           384       ['conv2d_36[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_37 (Ba  (None, 25, 25, 96)           384       ['conv2d_37[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_38 (Ba  (None, 25, 25, 96)           384       ['conv2d_38[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_34 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_34[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_36 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_36[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_37 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_37[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_38 (Activation)  (None, 25, 25, 96)           0         ['batch_normalization_38[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_6 (Concatenate  (None, 25, 25, 384)          0         ['activation_34[0][0]',       \n",
            " )                                                                   'activation_36[0][0]',       \n",
            "                                                                     'activation_37[0][0]',       \n",
            "                                                                     'activation_38[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_39 (Conv2D)          (None, 25, 25, 192)          73920     ['concatenate_6[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_39 (Ba  (None, 25, 25, 192)          768       ['conv2d_39[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_39 (Activation)  (None, 25, 25, 192)          0         ['batch_normalization_39[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_40 (Conv2D)          (None, 25, 25, 224)          387296    ['activation_39[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_40 (Ba  (None, 25, 25, 224)          896       ['conv2d_40[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_40 (Activation)  (None, 25, 25, 224)          0         ['batch_normalization_40[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_41 (Conv2D)          (None, 12, 12, 256)          516352    ['activation_40[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_42 (Conv2D)          (None, 12, 12, 384)          1327488   ['concatenate_6[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_41 (Ba  (None, 12, 12, 256)          1024      ['conv2d_41[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_42 (Ba  (None, 12, 12, 384)          1536      ['conv2d_42[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_41 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_41[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_42 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_42[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " max_pooling2d_2 (MaxPoolin  (None, 12, 12, 384)          0         ['concatenate_6[0][0]']       \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " concatenate_7 (Concatenate  (None, 12, 12, 1024)         0         ['activation_41[0][0]',       \n",
            " )                                                                   'activation_42[0][0]',       \n",
            "                                                                     'max_pooling2d_2[0][0]']     \n",
            "                                                                                                  \n",
            " conv2d_43 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_7[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_43 (Ba  (None, 12, 12, 192)          768       ['conv2d_43[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_43 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_43[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_44 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_43[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_44 (Ba  (None, 12, 12, 192)          768       ['conv2d_44[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_44 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_44[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_45 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_44[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_48 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_7[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_45 (Ba  (None, 12, 12, 224)          896       ['conv2d_45[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_48 (Ba  (None, 12, 12, 192)          768       ['conv2d_48[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_45 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_45[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_48 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_48[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_46 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_45[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_49 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_48[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_46 (Ba  (None, 12, 12, 224)          896       ['conv2d_46[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_49 (Ba  (None, 12, 12, 224)          896       ['conv2d_49[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_46 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_46[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_49 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_49[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_4 (Avera  (None, 12, 12, 1024)         0         ['concatenate_7[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_47 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_46[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_50 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_49[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_51 (Conv2D)          (None, 12, 12, 128)          131200    ['average_pooling2d_4[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_52 (Conv2D)          (None, 12, 12, 384)          393600    ['concatenate_7[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_47 (Ba  (None, 12, 12, 256)          1024      ['conv2d_47[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_50 (Ba  (None, 12, 12, 256)          1024      ['conv2d_50[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_51 (Ba  (None, 12, 12, 128)          512       ['conv2d_51[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_52 (Ba  (None, 12, 12, 384)          1536      ['conv2d_52[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_47 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_47[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_50 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_50[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_51 (Activation)  (None, 12, 12, 128)          0         ['batch_normalization_51[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_52 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_52[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_8 (Concatenate  (None, 12, 12, 1024)         0         ['activation_47[0][0]',       \n",
            " )                                                                   'activation_50[0][0]',       \n",
            "                                                                     'activation_51[0][0]',       \n",
            "                                                                     'activation_52[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_53 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_8[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_53 (Ba  (None, 12, 12, 192)          768       ['conv2d_53[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_53 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_53[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_54 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_53[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_54 (Ba  (None, 12, 12, 192)          768       ['conv2d_54[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_54 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_54[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_55 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_54[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_58 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_8[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_55 (Ba  (None, 12, 12, 224)          896       ['conv2d_55[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_58 (Ba  (None, 12, 12, 192)          768       ['conv2d_58[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_55 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_55[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_58 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_58[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_56 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_55[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_59 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_58[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_56 (Ba  (None, 12, 12, 224)          896       ['conv2d_56[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_59 (Ba  (None, 12, 12, 224)          896       ['conv2d_59[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_56 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_56[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_59 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_59[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_5 (Avera  (None, 12, 12, 1024)         0         ['concatenate_8[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_57 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_56[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_60 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_59[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_61 (Conv2D)          (None, 12, 12, 128)          131200    ['average_pooling2d_5[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_62 (Conv2D)          (None, 12, 12, 384)          393600    ['concatenate_8[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_57 (Ba  (None, 12, 12, 256)          1024      ['conv2d_57[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_60 (Ba  (None, 12, 12, 256)          1024      ['conv2d_60[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_61 (Ba  (None, 12, 12, 128)          512       ['conv2d_61[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_62 (Ba  (None, 12, 12, 384)          1536      ['conv2d_62[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_57 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_57[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_60 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_60[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_61 (Activation)  (None, 12, 12, 128)          0         ['batch_normalization_61[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_62 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_62[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_9 (Concatenate  (None, 12, 12, 1024)         0         ['activation_57[0][0]',       \n",
            " )                                                                   'activation_60[0][0]',       \n",
            "                                                                     'activation_61[0][0]',       \n",
            "                                                                     'activation_62[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_63 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_9[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_63 (Ba  (None, 12, 12, 192)          768       ['conv2d_63[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_63 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_63[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_64 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_63[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_64 (Ba  (None, 12, 12, 192)          768       ['conv2d_64[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_64 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_64[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_65 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_64[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_68 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_9[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_65 (Ba  (None, 12, 12, 224)          896       ['conv2d_65[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_68 (Ba  (None, 12, 12, 192)          768       ['conv2d_68[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_65 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_65[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_68 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_68[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_66 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_65[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_69 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_68[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_66 (Ba  (None, 12, 12, 224)          896       ['conv2d_66[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_69 (Ba  (None, 12, 12, 224)          896       ['conv2d_69[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_66 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_66[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_69 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_69[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_6 (Avera  (None, 12, 12, 1024)         0         ['concatenate_9[0][0]']       \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_67 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_66[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_70 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_69[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_71 (Conv2D)          (None, 12, 12, 128)          131200    ['average_pooling2d_6[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_72 (Conv2D)          (None, 12, 12, 384)          393600    ['concatenate_9[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_67 (Ba  (None, 12, 12, 256)          1024      ['conv2d_67[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_70 (Ba  (None, 12, 12, 256)          1024      ['conv2d_70[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_71 (Ba  (None, 12, 12, 128)          512       ['conv2d_71[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_72 (Ba  (None, 12, 12, 384)          1536      ['conv2d_72[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_67 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_67[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_70 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_70[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_71 (Activation)  (None, 12, 12, 128)          0         ['batch_normalization_71[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_72 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_72[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_10 (Concatenat  (None, 12, 12, 1024)         0         ['activation_67[0][0]',       \n",
            " e)                                                                  'activation_70[0][0]',       \n",
            "                                                                     'activation_71[0][0]',       \n",
            "                                                                     'activation_72[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_73 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_10[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_73 (Ba  (None, 12, 12, 192)          768       ['conv2d_73[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_73 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_73[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_74 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_73[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_74 (Ba  (None, 12, 12, 192)          768       ['conv2d_74[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_74 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_74[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_75 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_74[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_78 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_10[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_75 (Ba  (None, 12, 12, 224)          896       ['conv2d_75[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_78 (Ba  (None, 12, 12, 192)          768       ['conv2d_78[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_75 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_75[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_78 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_78[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_76 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_75[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_79 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_78[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_76 (Ba  (None, 12, 12, 224)          896       ['conv2d_76[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_79 (Ba  (None, 12, 12, 224)          896       ['conv2d_79[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_76 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_76[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_79 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_79[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_7 (Avera  (None, 12, 12, 1024)         0         ['concatenate_10[0][0]']      \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_77 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_76[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_80 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_79[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_81 (Conv2D)          (None, 12, 12, 128)          131200    ['average_pooling2d_7[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_82 (Conv2D)          (None, 12, 12, 384)          393600    ['concatenate_10[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_77 (Ba  (None, 12, 12, 256)          1024      ['conv2d_77[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_80 (Ba  (None, 12, 12, 256)          1024      ['conv2d_80[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_81 (Ba  (None, 12, 12, 128)          512       ['conv2d_81[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_82 (Ba  (None, 12, 12, 384)          1536      ['conv2d_82[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_77 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_77[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_80 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_80[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_81 (Activation)  (None, 12, 12, 128)          0         ['batch_normalization_81[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_82 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_82[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_11 (Concatenat  (None, 12, 12, 1024)         0         ['activation_77[0][0]',       \n",
            " e)                                                                  'activation_80[0][0]',       \n",
            "                                                                     'activation_81[0][0]',       \n",
            "                                                                     'activation_82[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_83 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_11[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_83 (Ba  (None, 12, 12, 192)          768       ['conv2d_83[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_83 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_83[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_84 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_83[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_84 (Ba  (None, 12, 12, 192)          768       ['conv2d_84[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_84 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_84[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_85 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_84[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_88 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_11[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_85 (Ba  (None, 12, 12, 224)          896       ['conv2d_85[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_88 (Ba  (None, 12, 12, 192)          768       ['conv2d_88[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_85 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_85[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_88 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_88[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_86 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_85[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_89 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_88[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_86 (Ba  (None, 12, 12, 224)          896       ['conv2d_86[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_89 (Ba  (None, 12, 12, 224)          896       ['conv2d_89[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_86 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_86[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_89 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_89[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_8 (Avera  (None, 12, 12, 1024)         0         ['concatenate_11[0][0]']      \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_87 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_86[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_90 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_89[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_91 (Conv2D)          (None, 12, 12, 128)          131200    ['average_pooling2d_8[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_92 (Conv2D)          (None, 12, 12, 384)          393600    ['concatenate_11[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_87 (Ba  (None, 12, 12, 256)          1024      ['conv2d_87[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_90 (Ba  (None, 12, 12, 256)          1024      ['conv2d_90[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_91 (Ba  (None, 12, 12, 128)          512       ['conv2d_91[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_92 (Ba  (None, 12, 12, 384)          1536      ['conv2d_92[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_87 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_87[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_90 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_90[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_91 (Activation)  (None, 12, 12, 128)          0         ['batch_normalization_91[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_92 (Activation)  (None, 12, 12, 384)          0         ['batch_normalization_92[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " concatenate_12 (Concatenat  (None, 12, 12, 1024)         0         ['activation_87[0][0]',       \n",
            " e)                                                                  'activation_90[0][0]',       \n",
            "                                                                     'activation_91[0][0]',       \n",
            "                                                                     'activation_92[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_93 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_12[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_93 (Ba  (None, 12, 12, 192)          768       ['conv2d_93[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_93 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_93[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_94 (Conv2D)          (None, 12, 12, 192)          258240    ['activation_93[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_94 (Ba  (None, 12, 12, 192)          768       ['conv2d_94[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_94 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_94[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_95 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_94[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_98 (Conv2D)          (None, 12, 12, 192)          196800    ['concatenate_12[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_95 (Ba  (None, 12, 12, 224)          896       ['conv2d_95[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_98 (Ba  (None, 12, 12, 192)          768       ['conv2d_98[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_95 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_95[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_98 (Activation)  (None, 12, 12, 192)          0         ['batch_normalization_98[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " conv2d_96 (Conv2D)          (None, 12, 12, 224)          351456    ['activation_95[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_99 (Conv2D)          (None, 12, 12, 224)          301280    ['activation_98[0][0]']       \n",
            "                                                                                                  \n",
            " batch_normalization_96 (Ba  (None, 12, 12, 224)          896       ['conv2d_96[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_99 (Ba  (None, 12, 12, 224)          896       ['conv2d_99[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " activation_96 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_96[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_99 (Activation)  (None, 12, 12, 224)          0         ['batch_normalization_99[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " average_pooling2d_9 (Avera  (None, 12, 12, 1024)         0         ['concatenate_12[0][0]']      \n",
            " gePooling2D)                                                                                     \n",
            "                                                                                                  \n",
            " conv2d_97 (Conv2D)          (None, 12, 12, 256)          401664    ['activation_96[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_100 (Conv2D)         (None, 12, 12, 256)          401664    ['activation_99[0][0]']       \n",
            "                                                                                                  \n",
            " conv2d_101 (Conv2D)         (None, 12, 12, 128)          131200    ['average_pooling2d_9[0][0]'] \n",
            "                                                                                                  \n",
            " conv2d_102 (Conv2D)         (None, 12, 12, 384)          393600    ['concatenate_12[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_97 (Ba  (None, 12, 12, 256)          1024      ['conv2d_97[0][0]']           \n",
            " tchNormalization)                                                                                \n",
            "                                                                                                  \n",
            " batch_normalization_100 (B  (None, 12, 12, 256)          1024      ['conv2d_100[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_101 (B  (None, 12, 12, 128)          512       ['conv2d_101[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_102 (B  (None, 12, 12, 384)          1536      ['conv2d_102[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_97 (Activation)  (None, 12, 12, 256)          0         ['batch_normalization_97[0][0]\n",
            "                                                                    ']                            \n",
            "                                                                                                  \n",
            " activation_100 (Activation  (None, 12, 12, 256)          0         ['batch_normalization_100[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_101 (Activation  (None, 12, 12, 128)          0         ['batch_normalization_101[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_102 (Activation  (None, 12, 12, 384)          0         ['batch_normalization_102[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_13 (Concatenat  (None, 12, 12, 1024)         0         ['activation_97[0][0]',       \n",
            " e)                                                                  'activation_100[0][0]',      \n",
            "                                                                     'activation_101[0][0]',      \n",
            "                                                                     'activation_102[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_103 (Conv2D)         (None, 12, 12, 192)          196800    ['concatenate_13[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_103 (B  (None, 12, 12, 192)          768       ['conv2d_103[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_103 (Activation  (None, 12, 12, 192)          0         ['batch_normalization_103[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_104 (Conv2D)         (None, 12, 12, 192)          258240    ['activation_103[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_104 (B  (None, 12, 12, 192)          768       ['conv2d_104[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_104 (Activation  (None, 12, 12, 192)          0         ['batch_normalization_104[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_105 (Conv2D)         (None, 12, 12, 224)          301280    ['activation_104[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_108 (Conv2D)         (None, 12, 12, 192)          196800    ['concatenate_13[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_105 (B  (None, 12, 12, 224)          896       ['conv2d_105[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_108 (B  (None, 12, 12, 192)          768       ['conv2d_108[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_105 (Activation  (None, 12, 12, 224)          0         ['batch_normalization_105[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_108 (Activation  (None, 12, 12, 192)          0         ['batch_normalization_108[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_106 (Conv2D)         (None, 12, 12, 224)          351456    ['activation_105[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_109 (Conv2D)         (None, 12, 12, 224)          301280    ['activation_108[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_106 (B  (None, 12, 12, 224)          896       ['conv2d_106[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_109 (B  (None, 12, 12, 224)          896       ['conv2d_109[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_106 (Activation  (None, 12, 12, 224)          0         ['batch_normalization_106[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_109 (Activation  (None, 12, 12, 224)          0         ['batch_normalization_109[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " average_pooling2d_10 (Aver  (None, 12, 12, 1024)         0         ['concatenate_13[0][0]']      \n",
            " agePooling2D)                                                                                    \n",
            "                                                                                                  \n",
            " conv2d_107 (Conv2D)         (None, 12, 12, 256)          401664    ['activation_106[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_110 (Conv2D)         (None, 12, 12, 256)          401664    ['activation_109[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_111 (Conv2D)         (None, 12, 12, 128)          131200    ['average_pooling2d_10[0][0]']\n",
            "                                                                                                  \n",
            " conv2d_112 (Conv2D)         (None, 12, 12, 384)          393600    ['concatenate_13[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_107 (B  (None, 12, 12, 256)          1024      ['conv2d_107[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_110 (B  (None, 12, 12, 256)          1024      ['conv2d_110[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_111 (B  (None, 12, 12, 128)          512       ['conv2d_111[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_112 (B  (None, 12, 12, 384)          1536      ['conv2d_112[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_107 (Activation  (None, 12, 12, 256)          0         ['batch_normalization_107[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_110 (Activation  (None, 12, 12, 256)          0         ['batch_normalization_110[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_111 (Activation  (None, 12, 12, 128)          0         ['batch_normalization_111[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_112 (Activation  (None, 12, 12, 384)          0         ['batch_normalization_112[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_14 (Concatenat  (None, 12, 12, 1024)         0         ['activation_107[0][0]',      \n",
            " e)                                                                  'activation_110[0][0]',      \n",
            "                                                                     'activation_111[0][0]',      \n",
            "                                                                     'activation_112[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_115 (Conv2D)         (None, 12, 12, 256)          262400    ['concatenate_14[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_115 (B  (None, 12, 12, 256)          1024      ['conv2d_115[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_115 (Activation  (None, 12, 12, 256)          0         ['batch_normalization_115[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_116 (Conv2D)         (None, 12, 12, 256)          459008    ['activation_115[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_116 (B  (None, 12, 12, 256)          1024      ['conv2d_116[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_116 (Activation  (None, 12, 12, 256)          0         ['batch_normalization_116[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_113 (Conv2D)         (None, 12, 12, 192)          196800    ['concatenate_14[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_117 (Conv2D)         (None, 12, 12, 320)          573760    ['activation_116[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_113 (B  (None, 12, 12, 192)          768       ['conv2d_113[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_117 (B  (None, 12, 12, 320)          1280      ['conv2d_117[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_113 (Activation  (None, 12, 12, 192)          0         ['batch_normalization_113[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_117 (Activation  (None, 12, 12, 320)          0         ['batch_normalization_117[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_114 (Conv2D)         (None, 5, 5, 192)            331968    ['activation_113[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_118 (Conv2D)         (None, 5, 5, 320)            921920    ['activation_117[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_114 (B  (None, 5, 5, 192)            768       ['conv2d_114[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_118 (B  (None, 5, 5, 320)            1280      ['conv2d_118[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " max_pooling2d_3 (MaxPoolin  (None, 5, 5, 1024)           0         ['concatenate_14[0][0]']      \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " activation_114 (Activation  (None, 5, 5, 192)            0         ['batch_normalization_114[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_118 (Activation  (None, 5, 5, 320)            0         ['batch_normalization_118[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_15 (Concatenat  (None, 5, 5, 1536)           0         ['max_pooling2d_3[0][0]',     \n",
            " e)                                                                  'activation_114[0][0]',      \n",
            "                                                                     'activation_118[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_119 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_15[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_119 (B  (None, 5, 5, 384)            1536      ['conv2d_119[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_119 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_119[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_120 (Conv2D)         (None, 5, 5, 448)            516544    ['activation_119[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_120 (B  (None, 5, 5, 448)            1792      ['conv2d_120[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_120 (Activation  (None, 5, 5, 448)            0         ['batch_normalization_120[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_121 (Conv2D)         (None, 5, 5, 512)            688640    ['activation_120[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_124 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_15[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_121 (B  (None, 5, 5, 512)            2048      ['conv2d_121[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_124 (B  (None, 5, 5, 384)            1536      ['conv2d_124[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_121 (Activation  (None, 5, 5, 512)            0         ['batch_normalization_121[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_124 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_124[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_122 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_121[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_123 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_121[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_125 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_124[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_126 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_124[0][0]']      \n",
            "                                                                                                  \n",
            " max_pooling2d_4 (MaxPoolin  (None, 5, 5, 1536)           0         ['concatenate_15[0][0]']      \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " batch_normalization_122 (B  (None, 5, 5, 256)            1024      ['conv2d_122[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_123 (B  (None, 5, 5, 256)            1024      ['conv2d_123[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_125 (B  (None, 5, 5, 256)            1024      ['conv2d_125[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_126 (B  (None, 5, 5, 256)            1024      ['conv2d_126[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " conv2d_127 (Conv2D)         (None, 5, 5, 256)            3539200   ['max_pooling2d_4[0][0]']     \n",
            "                                                                                                  \n",
            " conv2d_128 (Conv2D)         (None, 5, 5, 256)            393472    ['concatenate_15[0][0]']      \n",
            "                                                                                                  \n",
            " activation_122 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_122[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_123 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_123[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_125 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_125[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_126 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_126[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " batch_normalization_127 (B  (None, 5, 5, 256)            1024      ['conv2d_127[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_128 (B  (None, 5, 5, 256)            1024      ['conv2d_128[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " concatenate_16 (Concatenat  (None, 5, 5, 512)            0         ['activation_122[0][0]',      \n",
            " e)                                                                  'activation_123[0][0]']      \n",
            "                                                                                                  \n",
            " concatenate_17 (Concatenat  (None, 5, 5, 512)            0         ['activation_125[0][0]',      \n",
            " e)                                                                  'activation_126[0][0]']      \n",
            "                                                                                                  \n",
            " activation_127 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_127[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_128 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_128[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_18 (Concatenat  (None, 5, 5, 1536)           0         ['concatenate_16[0][0]',      \n",
            " e)                                                                  'concatenate_17[0][0]',      \n",
            "                                                                     'activation_127[0][0]',      \n",
            "                                                                     'activation_128[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_129 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_18[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_129 (B  (None, 5, 5, 384)            1536      ['conv2d_129[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_129 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_129[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_130 (Conv2D)         (None, 5, 5, 448)            516544    ['activation_129[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_130 (B  (None, 5, 5, 448)            1792      ['conv2d_130[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_130 (Activation  (None, 5, 5, 448)            0         ['batch_normalization_130[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_131 (Conv2D)         (None, 5, 5, 512)            688640    ['activation_130[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_134 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_18[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_131 (B  (None, 5, 5, 512)            2048      ['conv2d_131[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_134 (B  (None, 5, 5, 384)            1536      ['conv2d_134[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_131 (Activation  (None, 5, 5, 512)            0         ['batch_normalization_131[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_134 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_134[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_132 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_131[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_133 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_131[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_135 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_134[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_136 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_134[0][0]']      \n",
            "                                                                                                  \n",
            " max_pooling2d_5 (MaxPoolin  (None, 5, 5, 1536)           0         ['concatenate_18[0][0]']      \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " batch_normalization_132 (B  (None, 5, 5, 256)            1024      ['conv2d_132[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_133 (B  (None, 5, 5, 256)            1024      ['conv2d_133[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_135 (B  (None, 5, 5, 256)            1024      ['conv2d_135[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_136 (B  (None, 5, 5, 256)            1024      ['conv2d_136[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " conv2d_137 (Conv2D)         (None, 5, 5, 256)            3539200   ['max_pooling2d_5[0][0]']     \n",
            "                                                                                                  \n",
            " conv2d_138 (Conv2D)         (None, 5, 5, 256)            393472    ['concatenate_18[0][0]']      \n",
            "                                                                                                  \n",
            " activation_132 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_132[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_133 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_133[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_135 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_135[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_136 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_136[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " batch_normalization_137 (B  (None, 5, 5, 256)            1024      ['conv2d_137[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_138 (B  (None, 5, 5, 256)            1024      ['conv2d_138[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " concatenate_19 (Concatenat  (None, 5, 5, 512)            0         ['activation_132[0][0]',      \n",
            " e)                                                                  'activation_133[0][0]']      \n",
            "                                                                                                  \n",
            " concatenate_20 (Concatenat  (None, 5, 5, 512)            0         ['activation_135[0][0]',      \n",
            " e)                                                                  'activation_136[0][0]']      \n",
            "                                                                                                  \n",
            " activation_137 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_137[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_138 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_138[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_21 (Concatenat  (None, 5, 5, 1536)           0         ['concatenate_19[0][0]',      \n",
            " e)                                                                  'concatenate_20[0][0]',      \n",
            "                                                                     'activation_137[0][0]',      \n",
            "                                                                     'activation_138[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_139 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_21[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_139 (B  (None, 5, 5, 384)            1536      ['conv2d_139[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_139 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_139[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_140 (Conv2D)         (None, 5, 5, 448)            516544    ['activation_139[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_140 (B  (None, 5, 5, 448)            1792      ['conv2d_140[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_140 (Activation  (None, 5, 5, 448)            0         ['batch_normalization_140[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_141 (Conv2D)         (None, 5, 5, 512)            688640    ['activation_140[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_144 (Conv2D)         (None, 5, 5, 384)            590208    ['concatenate_21[0][0]']      \n",
            "                                                                                                  \n",
            " batch_normalization_141 (B  (None, 5, 5, 512)            2048      ['conv2d_141[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_144 (B  (None, 5, 5, 384)            1536      ['conv2d_144[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " activation_141 (Activation  (None, 5, 5, 512)            0         ['batch_normalization_141[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_144 (Activation  (None, 5, 5, 384)            0         ['batch_normalization_144[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " conv2d_142 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_141[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_143 (Conv2D)         (None, 5, 5, 256)            393472    ['activation_141[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_145 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_144[0][0]']      \n",
            "                                                                                                  \n",
            " conv2d_146 (Conv2D)         (None, 5, 5, 256)            295168    ['activation_144[0][0]']      \n",
            "                                                                                                  \n",
            " max_pooling2d_6 (MaxPoolin  (None, 5, 5, 1536)           0         ['concatenate_21[0][0]']      \n",
            " g2D)                                                                                             \n",
            "                                                                                                  \n",
            " batch_normalization_142 (B  (None, 5, 5, 256)            1024      ['conv2d_142[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_143 (B  (None, 5, 5, 256)            1024      ['conv2d_143[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_145 (B  (None, 5, 5, 256)            1024      ['conv2d_145[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_146 (B  (None, 5, 5, 256)            1024      ['conv2d_146[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " conv2d_147 (Conv2D)         (None, 5, 5, 256)            3539200   ['max_pooling2d_6[0][0]']     \n",
            "                                                                                                  \n",
            " conv2d_148 (Conv2D)         (None, 5, 5, 256)            393472    ['concatenate_21[0][0]']      \n",
            "                                                                                                  \n",
            " activation_142 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_142[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_143 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_143[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_145 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_145[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_146 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_146[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " batch_normalization_147 (B  (None, 5, 5, 256)            1024      ['conv2d_147[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " batch_normalization_148 (B  (None, 5, 5, 256)            1024      ['conv2d_148[0][0]']          \n",
            " atchNormalization)                                                                               \n",
            "                                                                                                  \n",
            " concatenate_22 (Concatenat  (None, 5, 5, 512)            0         ['activation_142[0][0]',      \n",
            " e)                                                                  'activation_143[0][0]']      \n",
            "                                                                                                  \n",
            " concatenate_23 (Concatenat  (None, 5, 5, 512)            0         ['activation_145[0][0]',      \n",
            " e)                                                                  'activation_146[0][0]']      \n",
            "                                                                                                  \n",
            " activation_147 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_147[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " activation_148 (Activation  (None, 5, 5, 256)            0         ['batch_normalization_148[0][0\n",
            " )                                                                  ]']                           \n",
            "                                                                                                  \n",
            " concatenate_24 (Concatenat  (None, 5, 5, 1536)           0         ['concatenate_22[0][0]',      \n",
            " e)                                                                  'concatenate_23[0][0]',      \n",
            "                                                                     'activation_147[0][0]',      \n",
            "                                                                     'activation_148[0][0]']      \n",
            "                                                                                                  \n",
            " global_average_pooling2d (  (None, 1536)                 0         ['concatenate_24[0][0]']      \n",
            " GlobalAveragePooling2D)                                                                          \n",
            "                                                                                                  \n",
            " dense (Dense)               (None, 1024)                 1573888   ['global_average_pooling2d[0][\n",
            "                                                                    0]']                          \n",
            "                                                                                                  \n",
            " dropout (Dropout)           (None, 1024)                 0         ['dense[0][0]']               \n",
            "                                                                                                  \n",
            " dense_1 (Dense)             (None, 4)                    4100      ['dropout[0][0]']             \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 52285668 (199.45 MB)\n",
            "Trainable params: 52222436 (199.21 MB)\n",
            "Non-trainable params: 63232 (247.00 KB)\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "# Building the Inception V4 Model Architecture\n",
        "\n",
        "model = InceptionV4()\n",
        "\n",
        "# Checking Model Summary\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:04:07.446305Z",
          "iopub.status.busy": "2023-06-07T15:04:07.445867Z",
          "iopub.status.idle": "2023-06-07T15:04:07.491742Z",
          "shell.execute_reply": "2023-06-07T15:04:07.490764Z",
          "shell.execute_reply.started": "2023-06-07T15:04:07.446272Z"
        },
        "id": "ifd5TqpwJDGf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cea0935c-13cb-4c89-fbca-176ea3fffee5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.\n"
          ]
        }
      ],
      "source": [
        "adam_opt=Adam(lr = 0.00001, beta_1=0.6, beta_2=0.9,amsgrad=True)\n",
        "\n",
        "# Compiling the Model\n",
        "model.compile(optimizer= adam_opt, loss = 'categorical_crossentropy', metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:04:11.434486Z",
          "iopub.status.busy": "2023-06-07T15:04:11.434066Z",
          "iopub.status.idle": "2023-06-07T15:19:45.656399Z",
          "shell.execute_reply": "2023-06-07T15:19:45.655297Z",
          "shell.execute_reply.started": "2023-06-07T15:04:11.434455Z"
        },
        "id": "ZDZglEY2JDGg",
        "outputId": "b4a615b3-b177-473d-a8e4-69420ac6f15b",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training started it takes few minutes\n",
            "Epoch 1/10\n",
            "123/123 [==============================] - 216s 834ms/step - loss: 0.7863 - accuracy: 0.7224 - val_loss: 2.4461 - val_accuracy: 0.2548 - lr: 0.0010\n",
            "Epoch 2/10\n",
            "123/123 [==============================] - 92s 742ms/step - loss: 0.5850 - accuracy: 0.8376 - val_loss: 5.6261 - val_accuracy: 0.2548 - lr: 0.0010\n",
            "Epoch 3/10\n",
            "123/123 [==============================] - 92s 740ms/step - loss: 0.4242 - accuracy: 0.8849 - val_loss: 1.8443 - val_accuracy: 0.5024 - lr: 0.0010\n",
            "Epoch 4/10\n",
            "123/123 [==============================] - 92s 742ms/step - loss: 0.3242 - accuracy: 0.9051 - val_loss: 1.3873 - val_accuracy: 0.6190 - lr: 0.0010\n",
            "Epoch 5/10\n",
            "123/123 [==============================] - 97s 783ms/step - loss: 0.2788 - accuracy: 0.9138 - val_loss: 1.4907 - val_accuracy: 0.5715 - lr: 0.0010\n",
            "Epoch 6/10\n",
            "123/123 [==============================] - 92s 743ms/step - loss: 0.2520 - accuracy: 0.9233 - val_loss: 0.2775 - val_accuracy: 0.9008 - lr: 0.0010\n",
            "Epoch 7/10\n",
            "123/123 [==============================] - 93s 750ms/step - loss: 0.2013 - accuracy: 0.9330 - val_loss: 0.1625 - val_accuracy: 0.9417 - lr: 0.0010\n",
            "Epoch 8/10\n",
            "123/123 [==============================] - 90s 732ms/step - loss: 0.1839 - accuracy: 0.9465 - val_loss: 0.1157 - val_accuracy: 0.9591 - lr: 0.0010\n",
            "Epoch 9/10\n",
            "123/123 [==============================] - 91s 740ms/step - loss: 0.1643 - accuracy: 0.9517 - val_loss: 0.1938 - val_accuracy: 0.9207 - lr: 0.0010\n",
            "Epoch 10/10\n",
            "123/123 [==============================] - 92s 745ms/step - loss: 0.1687 - accuracy: 0.9496 - val_loss: 0.0484 - val_accuracy: 0.9832 - lr: 0.0010\n",
            "successfully completed the traing session\n",
            "--- took 20:42 minutes ---\n"
          ]
        }
      ],
      "source": [
        "import time\n",
        "print(\"Training started it takes few minutes\")\n",
        "start_time = time.time()\n",
        "\n",
        "history = model.fit(\n",
        "  training_set,\n",
        "  steps_per_epoch = nb_train_samples // batch_size,\n",
        "  epochs = epochs,\n",
        "  validation_data = val_set,\n",
        "  validation_steps = nb_validation_samples // batch_size,\n",
        "  callbacks = [early, reduce])\n",
        "\n",
        "print(\"successfully completed the traing session\")\n",
        "print(\"--- took %d:%.2d minutes ---\" % divmod(time.time() - start_time, 60))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:20:46.329472Z",
          "iopub.status.busy": "2023-06-07T15:20:46.328479Z",
          "iopub.status.idle": "2023-06-07T15:20:47.235222Z",
          "shell.execute_reply": "2023-06-07T15:20:47.231964Z",
          "shell.execute_reply.started": "2023-06-07T15:20:46.329419Z"
        },
        "id": "XN2jyxfYJDGh",
        "outputId": "85ccfdb2-3bf0-49cb-9668-b96f142077ac",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.style.use(\"ggplot\")\n",
        "\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Training Acc', 'Test Acc'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# summarize history for loss\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Training Loss', 'Test Loss'], loc='upper left')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:21:12.884162Z",
          "iopub.status.busy": "2023-06-07T15:21:12.883551Z",
          "iopub.status.idle": "2023-06-07T15:21:23.354627Z",
          "shell.execute_reply": "2023-06-07T15:21:23.352008Z",
          "shell.execute_reply.started": "2023-06-07T15:21:12.884121Z"
        },
        "id": "J-GN3uqaJDGi",
        "outputId": "bb51dac7-7d14-4b3e-e7d4-1cff266f88cd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Validation Score:  0.05020074546337128\n",
            "Validation Accuracy:  0.9825000166893005\n"
          ]
        }
      ],
      "source": [
        "score = model.evaluate_generator(val_set, steps=50)\n",
        "\n",
        "print ('Validation Score: ', score[0])\n",
        "print ('Validation Accuracy: ',score[1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:21:35.492833Z",
          "iopub.status.busy": "2023-06-07T15:21:35.492445Z",
          "iopub.status.idle": "2023-06-07T15:21:35.497389Z",
          "shell.execute_reply": "2023-06-07T15:21:35.496346Z",
          "shell.execute_reply.started": "2023-06-07T15:21:35.492803Z"
        },
        "id": "UrYMvDi9JDGj"
      },
      "outputs": [],
      "source": [
        "# accuracy_score = model.evaluate_generator(test_set, steps=50)\n",
        "# print(\"Accuracy Score:\", accuracy_score[0])\n",
        "# print(\"Accuracy:\",accuracy_score[1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:21:38.620658Z",
          "iopub.status.busy": "2023-06-07T15:21:38.620261Z",
          "iopub.status.idle": "2023-06-07T15:21:49.023120Z",
          "shell.execute_reply": "2023-06-07T15:21:49.021876Z",
          "shell.execute_reply.started": "2023-06-07T15:21:38.620620Z"
        },
        "id": "KB87AMVTJDGj",
        "outputId": "abe6b02b-68d0-463f-ba8b-c8a9912d3e0d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "53/53 [==============================] - 13s 246ms/step - loss: 0.0476 - accuracy: 0.9834\n",
            "> 98.343\n"
          ]
        }
      ],
      "source": [
        "_, acc = model.evaluate(val_set, verbose=1)\n",
        "print('> %.3f' % (acc * 100.0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:22:19.472572Z",
          "iopub.status.busy": "2023-06-07T15:22:19.472197Z",
          "iopub.status.idle": "2023-06-07T15:22:23.177241Z",
          "shell.execute_reply": "2023-06-07T15:22:23.176233Z",
          "shell.execute_reply.started": "2023-06-07T15:22:19.472542Z"
        },
        "id": "IBqGW-r0JDGk"
      },
      "outputs": [],
      "source": [
        "model.save(\"Test\" + '.h5')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qsulq2QJJDGl"
      },
      "source": [
        "### Model Evaluation\n",
        "\n",
        "The evaluate_model function loads a trained model from a saved file and evaluates it using the validation generator. It calculates overall metrics (accuracy, precision, recall, F1-score) and class-specific metrics. These metrics provide a detailed view of model performance."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:22:26.193315Z",
          "iopub.status.busy": "2023-06-07T15:22:26.192625Z",
          "iopub.status.idle": "2023-06-07T15:22:26.202704Z",
          "shell.execute_reply": "2023-06-07T15:22:26.201445Z",
          "shell.execute_reply.started": "2023-06-07T15:22:26.193282Z"
        },
        "id": "r1cCdQSaJDGq"
      },
      "outputs": [],
      "source": [
        "def evaluate_model(saved_model_path, validation_generator):\n",
        "    # Load the saved model\n",
        "    loaded_model = tf.keras.models.load_model(saved_model_path)\n",
        "\n",
        "    # Evaluate the model on the test set\n",
        "    test_loss, test_accuracy = loaded_model.evaluate(validation_generator)\n",
        "    print('Test Loss:', test_loss)\n",
        "    print('Test Accuracy:', test_accuracy)\n",
        "\n",
        "    # Perform thorough evaluation\n",
        "    y_true = validation_generator.classes\n",
        "    y_pred = loaded_model.predict(validation_generator)\n",
        "    y_pred_classes = np.argmax(y_pred, axis=1)\n",
        "\n",
        "    # Calculate overall metrics\n",
        "    accuracy = metrics.accuracy_score(y_true, y_pred_classes)\n",
        "    precision = metrics.precision_score(y_true, y_pred_classes, average='weighted')\n",
        "    recall = metrics.recall_score(y_true, y_pred_classes, average='weighted')\n",
        "    f1_score = metrics.f1_score(y_true, y_pred_classes, average='weighted')\n",
        "\n",
        "    print('Overall Metrics:')\n",
        "    print('Accuracy:', accuracy)\n",
        "    print('Precision:', precision)\n",
        "    print('Recall:', recall)\n",
        "    print('F1-Score:', f1_score)\n",
        "\n",
        "    # Calculate metrics per class\n",
        "    class_metrics = metrics.classification_report(y_true, y_pred_classes, target_names=validation_generator.class_indices.keys())\n",
        "    print('\\nMetrics per Class:')\n",
        "    print(class_metrics)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2023-06-07T15:23:05.019769Z",
          "iopub.status.busy": "2023-06-07T15:23:05.019096Z",
          "iopub.status.idle": "2023-06-07T15:23:34.395652Z",
          "shell.execute_reply": "2023-06-07T15:23:34.394402Z",
          "shell.execute_reply.started": "2023-06-07T15:23:05.019679Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tGTAonCUJDGr",
        "outputId": "6b90b1e4-d5a8-41c1-ad2d-4155a0d18a45"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "53/53 [==============================] - 14s 206ms/step - loss: 5.4194 - accuracy: 0.2609\n",
            "Test Loss: 5.419363498687744\n",
            "Test Accuracy: 0.26094675064086914\n",
            "53/53 [==============================] - 10s 173ms/step\n",
            "Overall Metrics:\n",
            "Accuracy: 0.26094674556213016\n",
            "Precision: 0.08668296561152046\n",
            "Recall: 0.26094674556213016\n",
            "F1-Score: 0.11316242952150017\n",
            "\n",
            "Metrics per Class:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "      cloudy       0.26      0.98      0.42       450\n",
            "      desert       0.00      0.00      0.00       340\n",
            "  green_area       0.00      0.00      0.00       450\n",
            "       water       0.06      0.00      0.01       450\n",
            "\n",
            "    accuracy                           0.26      1690\n",
            "   macro avg       0.08      0.24      0.11      1690\n",
            "weighted avg       0.09      0.26      0.11      1690\n",
            "\n"
          ]
        }
      ],
      "source": [
        "import sklearn.metrics as metrics\n",
        "saved_model_path = '/content/gdrive/MyDrive/Colab Notebooks/Test.h5'\n",
        "\n",
        "evaluate_model(saved_model_path, val_set)\n",
        "# plot_training_history(history)"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.9"
    },
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "toc_visible": true
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}