Computer_Vision_Practise_Transfer_learning_and_fine_tuning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Computer Vision Practise - Transfer learning and fine tuning.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/shivamchb42/Image-Classification-using-Transfer-Learning/blob/main/Computer_Vision_Practise_Transfer_learning_and_fine_tuning.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-08Dm6Y21hh3"
      },
      "source": [
        "# Transfer learning and fine tuning for image classification"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CTXL9aMb9bDU"
      },
      "source": [
        "## Transfer learning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zrgRctvh29jp"
      },
      "source": [
        "### Importing the libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4y-991PPIVMR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "outputId": "1aef4e9a-fe8d-47fa-dd18-3cd61aab49fe"
      },
      "source": [
        "import tensorflow as tf\n",
        "import seaborn as sns\n",
        "import zipfile\n",
        "import numpy as np\n",
        "import cv2\n",
        "from google.colab.patches import cv2_imshow\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Dense, GlobalAveragePooling2D, Dropout\n",
        "tf.__version__"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'2.8.2'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "u0jiBegz3ETL"
      },
      "source": [
        "### Loading the images"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R7mFZTxLI0hO",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b4ba7aa3-d252-4691-c354-c6dad03f1dc9"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VtJ8sAyyJaVe"
      },
      "source": [
        "path = '/content/drive/MyDrive/Intern-Study_materials/Computer Vision Masterclass/Datasets/homer_bart_2.zip'\n",
        "zip_object = zipfile.ZipFile(file=path, mode='r')\n",
        "zip_object.extractall('./')\n",
        "zip_object.close()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iRc7t0lRJy5P",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "outputId": "8da8547a-67fe-4d41-cffa-04684b976166"
      },
      "source": [
        "tf.keras.preprocessing.image.load_img('/content/homer_bart_2/training_set/bart/bart100.bmp')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<PIL.BmpImagePlugin.BmpImageFile image mode=RGB size=364x435 at 0x7F4FD363FC50>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-3kIG15NKHSh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 577
        },
        "outputId": "5e6ba69b-a8df-4717-ce52-12e024a95a7d"
      },
      "source": [
        "tf.keras.preprocessing.image.load_img('/content/homer_bart_2/training_set/homer/homer100.bmp')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<PIL.BmpImagePlugin.BmpImageFile image mode=RGB size=276x560 at 0x7F505FBC6110>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M3xS72MI6C8q"
      },
      "source": [
        "### Train and test set"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HpYm2s_fKa0d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e7d6b786-888c-46f0-f6da-96ed4dee3aff"
      },
      "source": [
        "training_generator = ImageDataGenerator(rescale=1./255,\n",
        "                                        rotation_range=7,\n",
        "                                        horizontal_flip=True,\n",
        "                                        zoom_range=0.2)\n",
        "train_dataset = training_generator.flow_from_directory('/content/homer_bart_2/training_set',\n",
        "                                                        target_size = (256, 256),\n",
        "                                                        batch_size = 8,\n",
        "                                                        class_mode = 'categorical',\n",
        "                                                        shuffle = True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 215 images belonging to 2 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ugMHSUnLxyG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "26aca6f0-e9f2-4e41-f685-124484e5d468"
      },
      "source": [
        "train_dataset.classes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=int32)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HZA4CwFmL1eC",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "de75bff4-8008-4fd4-bf7e-eb564ab79db0"
      },
      "source": [
        "train_dataset.class_indices"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'bart': 0, 'homer': 1}"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-jIsmMsML8as",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a842935a-d57e-4954-8058-53e0d4424d67"
      },
      "source": [
        "test_generator = ImageDataGenerator(rescale=1./255)\n",
        "test_dataset = test_generator.flow_from_directory('/content/homer_bart_2/test_set',\n",
        "                                                     target_size = (256, 256),\n",
        "                                                     batch_size = 1,\n",
        "                                                     class_mode = 'categorical',\n",
        "                                                     shuffle = False)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 54 images belonging to 2 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K_hm4IjC3FFX"
      },
      "source": [
        "### Pre-trained network\n",
        "\n",
        "- ResNet: https://arxiv.org/pdf/1512.03385.pdf\n",
        "- Documentation: https://keras.io/api/applications/"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Z7IpopLxK7Qx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "74fb317c-34e4-4b07-9ae5-59f492e21865"
      },
      "source": [
        "base_model = tf.keras.applications.ResNet50(weights='imagenet', include_top=False,\n",
        "                                            input_tensor = Input(shape = (256, 256, 3)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "94773248/94765736 [==============================] - 0s 0us/step\n",
            "94781440/94765736 [==============================] - 0s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rVd-XHYnNla9",
        "outputId": "87fd29f0-1dda-43e6-a0a0-afaee557d8cd"
      },
      "source": [
        "base_model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"resnet50\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)           [(None, 256, 256, 3  0           []                               \n",
            "                                )]                                                                \n",
            "                                                                                                  \n",
            " conv1_pad (ZeroPadding2D)      (None, 262, 262, 3)  0           ['input_1[0][0]']                \n",
            "                                                                                                  \n",
            " conv1_conv (Conv2D)            (None, 128, 128, 64  9472        ['conv1_pad[0][0]']              \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv1_bn (BatchNormalization)  (None, 128, 128, 64  256         ['conv1_conv[0][0]']             \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv1_relu (Activation)        (None, 128, 128, 64  0           ['conv1_bn[0][0]']               \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " pool1_pad (ZeroPadding2D)      (None, 130, 130, 64  0           ['conv1_relu[0][0]']             \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " pool1_pool (MaxPooling2D)      (None, 64, 64, 64)   0           ['pool1_pad[0][0]']              \n",
            "                                                                                                  \n",
            " conv2_block1_1_conv (Conv2D)   (None, 64, 64, 64)   4160        ['pool1_pool[0][0]']             \n",
            "                                                                                                  \n",
            " conv2_block1_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block1_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block1_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block1_0_conv (Conv2D)   (None, 64, 64, 256)  16640       ['pool1_pool[0][0]']             \n",
            "                                                                                                  \n",
            " conv2_block1_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block1_0_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_add (Add)         (None, 64, 64, 256)  0           ['conv2_block1_0_bn[0][0]',      \n",
            "                                                                  'conv2_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block1_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block2_1_conv (Conv2D)   (None, 64, 64, 64)   16448       ['conv2_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block2_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block2_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block2_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block2_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block2_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_add (Add)         (None, 64, 64, 256)  0           ['conv2_block1_out[0][0]',       \n",
            "                                                                  'conv2_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block2_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block3_1_conv (Conv2D)   (None, 64, 64, 64)   16448       ['conv2_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block3_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block3_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block3_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block3_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block3_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_add (Add)         (None, 64, 64, 256)  0           ['conv2_block2_out[0][0]',       \n",
            "                                                                  'conv2_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block3_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_1_conv (Conv2D)   (None, 32, 32, 128)  32896       ['conv2_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block1_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block1_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block1_0_conv (Conv2D)   (None, 32, 32, 512)  131584      ['conv2_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block1_0_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_add (Add)         (None, 32, 32, 512)  0           ['conv3_block1_0_bn[0][0]',      \n",
            "                                                                  'conv3_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block1_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block2_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block2_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block2_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block2_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block2_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block2_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_add (Add)         (None, 32, 32, 512)  0           ['conv3_block1_out[0][0]',       \n",
            "                                                                  'conv3_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block2_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block3_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block3_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block3_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block3_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block3_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block3_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_add (Add)         (None, 32, 32, 512)  0           ['conv3_block2_out[0][0]',       \n",
            "                                                                  'conv3_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block3_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block4_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block4_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block4_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block4_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block4_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block4_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block4_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block4_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block4_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block4_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block4_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block4_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block4_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_add (Add)         (None, 32, 32, 512)  0           ['conv3_block3_out[0][0]',       \n",
            "                                                                  'conv3_block4_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block4_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block4_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block1_1_conv (Conv2D)   (None, 16, 16, 256)  131328      ['conv3_block4_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block1_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block1_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block1_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block1_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block1_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block1_0_conv (Conv2D)   (None, 16, 16, 1024  525312      ['conv3_block4_out[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block1_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_0_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block1_0_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block1_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_add (Add)         (None, 16, 16, 1024  0           ['conv4_block1_0_bn[0][0]',      \n",
            "                                )                                 'conv4_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block1_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block1_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block2_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block2_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block2_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block2_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block2_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block2_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block2_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block2_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_add (Add)         (None, 16, 16, 1024  0           ['conv4_block1_out[0][0]',       \n",
            "                                )                                 'conv4_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block2_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block2_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block3_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block3_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block3_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block3_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block3_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block3_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block3_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block3_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_add (Add)         (None, 16, 16, 1024  0           ['conv4_block2_out[0][0]',       \n",
            "                                )                                 'conv4_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block3_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block3_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block4_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block4_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block4_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block4_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block4_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block4_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block4_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block4_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block4_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block4_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block4_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block4_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block4_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_add (Add)         (None, 16, 16, 1024  0           ['conv4_block3_out[0][0]',       \n",
            "                                )                                 'conv4_block4_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block4_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block4_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block4_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block5_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block5_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block5_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block5_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block5_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block5_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block5_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block5_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block5_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block5_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block5_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block5_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block5_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_add (Add)         (None, 16, 16, 1024  0           ['conv4_block4_out[0][0]',       \n",
            "                                )                                 'conv4_block5_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block5_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block5_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block5_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block6_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block6_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block6_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block6_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block6_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block6_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block6_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block6_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block6_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block6_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block6_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block6_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block6_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_add (Add)         (None, 16, 16, 1024  0           ['conv4_block5_out[0][0]',       \n",
            "                                )                                 'conv4_block6_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block6_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block6_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv5_block1_1_conv (Conv2D)   (None, 8, 8, 512)    524800      ['conv4_block6_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block1_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block1_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block1_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block1_0_conv (Conv2D)   (None, 8, 8, 2048)   2099200     ['conv4_block6_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block1_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block1_0_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block1_0_bn[0][0]',      \n",
            "                                                                  'conv5_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block1_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block2_1_conv (Conv2D)   (None, 8, 8, 512)    1049088     ['conv5_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block2_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block2_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block2_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block2_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block2_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block1_out[0][0]',       \n",
            "                                                                  'conv5_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block2_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block3_1_conv (Conv2D)   (None, 8, 8, 512)    1049088     ['conv5_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block3_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block3_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block3_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block3_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block3_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block2_out[0][0]',       \n",
            "                                                                  'conv5_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block3_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 23,587,712\n",
            "Trainable params: 23,534,592\n",
            "Non-trainable params: 53,120\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uMeTYGUxOt15",
        "outputId": "8c241b5a-350b-4a81-aafb-52fbcf4365b5"
      },
      "source": [
        "len(base_model.layers)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "175"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IxvwzOTrPQq2"
      },
      "source": [
        "for layer in base_model.layers:\n",
        "  layer.trainable = False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6u3-_fxMO7J7",
        "outputId": "99bbe13d-6ed7-4084-94e4-1e675c60d3a9"
      },
      "source": [
        "for layer in base_model.layers:\n",
        "  print(layer, layer.trainable)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<keras.engine.input_layer.InputLayer object at 0x7f4fd2539990> False\n",
            "<keras.layers.convolutional.ZeroPadding2D object at 0x7f4fd358e290> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fddc9f690> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd2541c10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd24922d0> False\n",
            "<keras.layers.convolutional.ZeroPadding2D object at 0x7f4fd17f0bd0> False\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f4fd17a2590> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17a8990> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17bc390> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd17c51d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17be510> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd174e550> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1758450> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17a88d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17be150> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17b01d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17a83d0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd17f00d0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd17ca910> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd363f7d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1761850> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1768790> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17686d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1773090> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd177b390> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd177b850> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1782050> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd170e450> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1773490> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1714f10> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd171e290> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1722650> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1725690> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd172ca50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1735390> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd172f4d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd173f6d0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd174aa10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd174a410> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16db390> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16d9290> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16e7f90> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f4c90> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16f8d10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1709050> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16d5910> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f88d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16d98d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1690d50> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd169c090> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1690890> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16a3cd0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16ac050> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b0990> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd169c6d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16f3a50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16ec310> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f1790> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd172f6d0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd1735c10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16cf3d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1714710> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1780090> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1768550> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17b68d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd179ff50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b80d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16be850> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16c4110> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd16c1b10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b8190> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd179f9d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd2492d90> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd166bf10> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16cb890> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1678250> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1671290> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1654210> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1683cd0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd167d5d0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1689250> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1613710> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1623d50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd162ab10> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd162e790> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1637fd0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1626b50> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16195d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1640b10> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd161f9d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15ce050> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd15d2350> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15ce410> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1623190> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1619a90> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1671190> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1689110> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd166b490> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16bafd0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17ad690> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd174a1d0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd16f4150> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16e1590> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15e1310> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15e3650> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15e3b50> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15e10d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15efe90> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15f71d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15fc1d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1601ed0> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd158e290> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd158e990> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd158e3d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd159e1d0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15a25d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15a7610> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15ac650> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15b3390> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15af410> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15bfa10> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd15c9410> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15c9610> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0220990> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0222cd0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0226710> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15c9450> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0237550> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0240450> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0240a10> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc024b590> False\n",
            "<keras.layers.merge.Add object at 0x7f4fc0257490> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0257f50> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc02407d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc022c210> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15bf290> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15b8990> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15a7ad0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15fcf50> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16012d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15e0b10> False\n",
            "<keras.layers.merge.Add object at 0x7f4fd17684d0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16cb710> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01e4d90> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01e4210> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01ea250> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1637b50> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01eefd0> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01fc450> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1678850> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01ee2d0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01dce90> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0206050> False\n",
            "<keras.layers.merge.Add object at 0x7f4fc020e390> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc020e890> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01f8c10> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01a0310> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01a4ed0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01a9350> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01afb50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01b8490> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01a9d90> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01bfb90> False\n",
            "<keras.layers.merge.Add object at 0x7f4fc01cdd10> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc02194d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01d4b10> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01d8e50> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01db510> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0161dd0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc016a350> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01735d0> False\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0173fd0> False\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01d4210> False\n",
            "<keras.layers.merge.Add object at 0x7f4fc01af910> False\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01a9910> False\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xmtx_isy45gh"
      },
      "source": [
        "### Custom dense layer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4mGINA-KQ2JB",
        "outputId": "f3fa7b1d-a29c-46d0-efbd-67de78303a0d"
      },
      "source": [
        "base_model.output"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<KerasTensor: shape=(None, 8, 8, 2048) dtype=float32 (created by layer 'conv5_block3_out')>"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OYTg4VR0Q9JX"
      },
      "source": [
        "# Dropout: https://jmlr.org/papers/volume15/srivastava14a.old/srivastava14a.pdf\n",
        "head_model = base_model.output\n",
        "head_model = GlobalAveragePooling2D()(head_model)\n",
        "head_model = Dense(units = 1025, activation='relu')(head_model)\n",
        "head_model = Dropout(rate = 0.2)(head_model)\n",
        "head_model = Dense(units = 1025, activation='relu')(head_model)\n",
        "head_model = Dropout(rate = 0.2)(head_model)\n",
        "head_model = Dense(2, activation='softmax')(head_model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E2hBlODz6GWd"
      },
      "source": [
        "### Building and training the neural network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3Im7hPYiTz-L",
        "outputId": "b56d6aa1-e3d4-4644-d89c-7e6c69e3b5c0"
      },
      "source": [
        "base_model.input"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<KerasTensor: shape=(None, 256, 256, 3) dtype=float32 (created by layer 'input_1')>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MDgoDfPqURIp"
      },
      "source": [
        "network = Model(inputs = base_model.input, outputs = head_model)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IBIilEtdVEdv",
        "outputId": "0f77f80b-5b4f-4dc6-a3b3-858d0bd4f402"
      },
      "source": [
        "8 * 8 * 2048"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "131072"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fRkK3JDUVUVs",
        "outputId": "a773e748-f371-4bab-8b34-619e6a475703"
      },
      "source": [
        "(2048 + 2) / 2"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1025.0"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5bK5TSLUUjbH",
        "outputId": "3da0a129-9b15-40ef-d0be-fe8e2936bc1a"
      },
      "source": [
        "network.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)           [(None, 256, 256, 3  0           []                               \n",
            "                                )]                                                                \n",
            "                                                                                                  \n",
            " conv1_pad (ZeroPadding2D)      (None, 262, 262, 3)  0           ['input_1[0][0]']                \n",
            "                                                                                                  \n",
            " conv1_conv (Conv2D)            (None, 128, 128, 64  9472        ['conv1_pad[0][0]']              \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv1_bn (BatchNormalization)  (None, 128, 128, 64  256         ['conv1_conv[0][0]']             \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv1_relu (Activation)        (None, 128, 128, 64  0           ['conv1_bn[0][0]']               \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " pool1_pad (ZeroPadding2D)      (None, 130, 130, 64  0           ['conv1_relu[0][0]']             \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " pool1_pool (MaxPooling2D)      (None, 64, 64, 64)   0           ['pool1_pad[0][0]']              \n",
            "                                                                                                  \n",
            " conv2_block1_1_conv (Conv2D)   (None, 64, 64, 64)   4160        ['pool1_pool[0][0]']             \n",
            "                                                                                                  \n",
            " conv2_block1_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block1_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block1_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block1_0_conv (Conv2D)   (None, 64, 64, 256)  16640       ['pool1_pool[0][0]']             \n",
            "                                                                                                  \n",
            " conv2_block1_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block1_0_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block1_add (Add)         (None, 64, 64, 256)  0           ['conv2_block1_0_bn[0][0]',      \n",
            "                                                                  'conv2_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block1_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block2_1_conv (Conv2D)   (None, 64, 64, 64)   16448       ['conv2_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block2_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block2_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block2_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block2_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block2_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block2_add (Add)         (None, 64, 64, 256)  0           ['conv2_block1_out[0][0]',       \n",
            "                                                                  'conv2_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block2_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block3_1_conv (Conv2D)   (None, 64, 64, 64)   16448       ['conv2_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv2_block3_1_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_1_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block3_2_conv (Conv2D)   (None, 64, 64, 64)   36928       ['conv2_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block3_2_bn (BatchNormal  (None, 64, 64, 64)  256         ['conv2_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_2_relu (Activatio  (None, 64, 64, 64)  0           ['conv2_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv2_block3_3_conv (Conv2D)   (None, 64, 64, 256)  16640       ['conv2_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv2_block3_3_bn (BatchNormal  (None, 64, 64, 256)  1024       ['conv2_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv2_block3_add (Add)         (None, 64, 64, 256)  0           ['conv2_block2_out[0][0]',       \n",
            "                                                                  'conv2_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv2_block3_out (Activation)  (None, 64, 64, 256)  0           ['conv2_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_1_conv (Conv2D)   (None, 32, 32, 128)  32896       ['conv2_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block1_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block1_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block1_0_conv (Conv2D)   (None, 32, 32, 512)  131584      ['conv2_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block1_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block1_0_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block1_add (Add)         (None, 32, 32, 512)  0           ['conv3_block1_0_bn[0][0]',      \n",
            "                                                                  'conv3_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block1_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block2_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block2_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block2_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block2_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block2_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block2_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block2_add (Add)         (None, 32, 32, 512)  0           ['conv3_block1_out[0][0]',       \n",
            "                                                                  'conv3_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block2_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block3_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block3_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block3_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block3_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block3_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block3_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block3_add (Add)         (None, 32, 32, 512)  0           ['conv3_block2_out[0][0]',       \n",
            "                                                                  'conv3_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block3_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block4_1_conv (Conv2D)   (None, 32, 32, 128)  65664       ['conv3_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv3_block4_1_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block4_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_1_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block4_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block4_2_conv (Conv2D)   (None, 32, 32, 128)  147584      ['conv3_block4_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block4_2_bn (BatchNormal  (None, 32, 32, 128)  512        ['conv3_block4_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_2_relu (Activatio  (None, 32, 32, 128)  0          ['conv3_block4_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv3_block4_3_conv (Conv2D)   (None, 32, 32, 512)  66048       ['conv3_block4_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv3_block4_3_bn (BatchNormal  (None, 32, 32, 512)  2048       ['conv3_block4_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv3_block4_add (Add)         (None, 32, 32, 512)  0           ['conv3_block3_out[0][0]',       \n",
            "                                                                  'conv3_block4_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv3_block4_out (Activation)  (None, 32, 32, 512)  0           ['conv3_block4_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block1_1_conv (Conv2D)   (None, 16, 16, 256)  131328      ['conv3_block4_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block1_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block1_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block1_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block1_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block1_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block1_0_conv (Conv2D)   (None, 16, 16, 1024  525312      ['conv3_block4_out[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block1_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_0_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block1_0_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block1_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block1_add (Add)         (None, 16, 16, 1024  0           ['conv4_block1_0_bn[0][0]',      \n",
            "                                )                                 'conv4_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block1_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block1_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block2_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block2_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block2_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block2_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block2_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block2_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block2_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block2_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block2_add (Add)         (None, 16, 16, 1024  0           ['conv4_block1_out[0][0]',       \n",
            "                                )                                 'conv4_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block2_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block2_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block3_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block3_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block3_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block3_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block3_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block3_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block3_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block3_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block3_add (Add)         (None, 16, 16, 1024  0           ['conv4_block2_out[0][0]',       \n",
            "                                )                                 'conv4_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block3_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block3_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block3_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block4_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block4_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block4_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block4_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block4_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block4_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block4_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block4_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block4_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block4_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block4_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block4_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block4_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block4_add (Add)         (None, 16, 16, 1024  0           ['conv4_block3_out[0][0]',       \n",
            "                                )                                 'conv4_block4_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block4_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block4_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block4_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block5_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block5_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block5_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block5_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block5_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block5_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block5_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block5_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block5_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block5_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block5_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block5_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block5_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block5_add (Add)         (None, 16, 16, 1024  0           ['conv4_block4_out[0][0]',       \n",
            "                                )                                 'conv4_block5_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block5_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block5_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_1_conv (Conv2D)   (None, 16, 16, 256)  262400      ['conv4_block5_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv4_block6_1_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block6_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block6_1_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block6_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block6_2_conv (Conv2D)   (None, 16, 16, 256)  590080      ['conv4_block6_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv4_block6_2_bn (BatchNormal  (None, 16, 16, 256)  1024       ['conv4_block6_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv4_block6_2_relu (Activatio  (None, 16, 16, 256)  0          ['conv4_block6_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv4_block6_3_conv (Conv2D)   (None, 16, 16, 1024  263168      ['conv4_block6_2_relu[0][0]']    \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_3_bn (BatchNormal  (None, 16, 16, 1024  4096       ['conv4_block6_3_conv[0][0]']    \n",
            " ization)                       )                                                                 \n",
            "                                                                                                  \n",
            " conv4_block6_add (Add)         (None, 16, 16, 1024  0           ['conv4_block5_out[0][0]',       \n",
            "                                )                                 'conv4_block6_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv4_block6_out (Activation)  (None, 16, 16, 1024  0           ['conv4_block6_add[0][0]']       \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv5_block1_1_conv (Conv2D)   (None, 8, 8, 512)    524800      ['conv4_block6_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block1_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block1_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block1_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block1_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block1_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block1_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block1_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block1_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block1_0_conv (Conv2D)   (None, 8, 8, 2048)   2099200     ['conv4_block6_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block1_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block1_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block1_0_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block1_0_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block1_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block1_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block1_0_bn[0][0]',      \n",
            "                                                                  'conv5_block1_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block1_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block1_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block2_1_conv (Conv2D)   (None, 8, 8, 512)    1049088     ['conv5_block1_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block2_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block2_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block2_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block2_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block2_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block2_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block2_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block2_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block2_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block2_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block2_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block2_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block2_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block1_out[0][0]',       \n",
            "                                                                  'conv5_block2_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block2_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block2_add[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block3_1_conv (Conv2D)   (None, 8, 8, 512)    1049088     ['conv5_block2_out[0][0]']       \n",
            "                                                                                                  \n",
            " conv5_block3_1_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block3_1_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_1_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block3_1_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block3_2_conv (Conv2D)   (None, 8, 8, 512)    2359808     ['conv5_block3_1_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block3_2_bn (BatchNormal  (None, 8, 8, 512)   2048        ['conv5_block3_2_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_2_relu (Activatio  (None, 8, 8, 512)   0           ['conv5_block3_2_bn[0][0]']      \n",
            " n)                                                                                               \n",
            "                                                                                                  \n",
            " conv5_block3_3_conv (Conv2D)   (None, 8, 8, 2048)   1050624     ['conv5_block3_2_relu[0][0]']    \n",
            "                                                                                                  \n",
            " conv5_block3_3_bn (BatchNormal  (None, 8, 8, 2048)  8192        ['conv5_block3_3_conv[0][0]']    \n",
            " ization)                                                                                         \n",
            "                                                                                                  \n",
            " conv5_block3_add (Add)         (None, 8, 8, 2048)   0           ['conv5_block2_out[0][0]',       \n",
            "                                                                  'conv5_block3_3_bn[0][0]']      \n",
            "                                                                                                  \n",
            " conv5_block3_out (Activation)  (None, 8, 8, 2048)   0           ['conv5_block3_add[0][0]']       \n",
            "                                                                                                  \n",
            " global_average_pooling2d (Glob  (None, 2048)        0           ['conv5_block3_out[0][0]']       \n",
            " alAveragePooling2D)                                                                              \n",
            "                                                                                                  \n",
            " dense (Dense)                  (None, 1025)         2100225     ['global_average_pooling2d[0][0]'\n",
            "                                                                 ]                                \n",
            "                                                                                                  \n",
            " dropout (Dropout)              (None, 1025)         0           ['dense[0][0]']                  \n",
            "                                                                                                  \n",
            " dense_1 (Dense)                (None, 1025)         1051650     ['dropout[0][0]']                \n",
            "                                                                                                  \n",
            " dropout_1 (Dropout)            (None, 1025)         0           ['dense_1[0][0]']                \n",
            "                                                                                                  \n",
            " dense_2 (Dense)                (None, 2)            2052        ['dropout_1[0][0]']              \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 26,741,639\n",
            "Trainable params: 3,153,927\n",
            "Non-trainable params: 23,587,712\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dv0zOmzbVceP"
      },
      "source": [
        "network.compile(optimizer='Adam', loss = 'categorical_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-43PR5xwVm2O",
        "outputId": "fbac5d39-4043-41b7-ac8b-92a2570b6bcd"
      },
      "source": [
        "history = network.fit(train_dataset, epochs=50)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "27/27 [==============================] - 18s 138ms/step - loss: 0.9112 - accuracy: 0.5488\n",
            "Epoch 2/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.7847 - accuracy: 0.5721\n",
            "Epoch 3/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.7564 - accuracy: 0.5442\n",
            "Epoch 4/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.6691 - accuracy: 0.6000\n",
            "Epoch 5/50\n",
            "27/27 [==============================] - 3s 123ms/step - loss: 0.6274 - accuracy: 0.6698\n",
            "Epoch 6/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.6121 - accuracy: 0.6279\n",
            "Epoch 7/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.6071 - accuracy: 0.6791\n",
            "Epoch 8/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.6320 - accuracy: 0.6605\n",
            "Epoch 9/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.6018 - accuracy: 0.6791\n",
            "Epoch 10/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.5232 - accuracy: 0.7628\n",
            "Epoch 11/50\n",
            "27/27 [==============================] - 3s 123ms/step - loss: 0.5624 - accuracy: 0.6930\n",
            "Epoch 12/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.5189 - accuracy: 0.7674\n",
            "Epoch 13/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.5439 - accuracy: 0.7767\n",
            "Epoch 14/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.5300 - accuracy: 0.7767\n",
            "Epoch 15/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.5553 - accuracy: 0.7349\n",
            "Epoch 16/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.5082 - accuracy: 0.7209\n",
            "Epoch 17/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.5304 - accuracy: 0.7442\n",
            "Epoch 18/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4684 - accuracy: 0.7907\n",
            "Epoch 19/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4428 - accuracy: 0.8186\n",
            "Epoch 20/50\n",
            "27/27 [==============================] - 3s 123ms/step - loss: 0.4631 - accuracy: 0.8186\n",
            "Epoch 21/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.5204 - accuracy: 0.7628\n",
            "Epoch 22/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4657 - accuracy: 0.8000\n",
            "Epoch 23/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4695 - accuracy: 0.8047\n",
            "Epoch 24/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4785 - accuracy: 0.7628\n",
            "Epoch 25/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.5041 - accuracy: 0.7535\n",
            "Epoch 26/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4755 - accuracy: 0.8000\n",
            "Epoch 27/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4844 - accuracy: 0.7907\n",
            "Epoch 28/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4241 - accuracy: 0.8140\n",
            "Epoch 29/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4375 - accuracy: 0.8140\n",
            "Epoch 30/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4143 - accuracy: 0.8279\n",
            "Epoch 31/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4513 - accuracy: 0.7907\n",
            "Epoch 32/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4425 - accuracy: 0.8000\n",
            "Epoch 33/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4239 - accuracy: 0.7907\n",
            "Epoch 34/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.3507 - accuracy: 0.8558\n",
            "Epoch 35/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4527 - accuracy: 0.8047\n",
            "Epoch 36/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4216 - accuracy: 0.7721\n",
            "Epoch 37/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.3738 - accuracy: 0.8651\n",
            "Epoch 38/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4194 - accuracy: 0.8326\n",
            "Epoch 39/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.3957 - accuracy: 0.8140\n",
            "Epoch 40/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4545 - accuracy: 0.7721\n",
            "Epoch 41/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.3675 - accuracy: 0.8186\n",
            "Epoch 42/50\n",
            "27/27 [==============================] - 3s 127ms/step - loss: 0.4434 - accuracy: 0.8279\n",
            "Epoch 43/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4166 - accuracy: 0.8000\n",
            "Epoch 44/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4463 - accuracy: 0.7767\n",
            "Epoch 45/50\n",
            "27/27 [==============================] - 3s 124ms/step - loss: 0.4055 - accuracy: 0.8372\n",
            "Epoch 46/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4157 - accuracy: 0.8047\n",
            "Epoch 47/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.3662 - accuracy: 0.8605\n",
            "Epoch 48/50\n",
            "27/27 [==============================] - 3s 127ms/step - loss: 0.3861 - accuracy: 0.8326\n",
            "Epoch 49/50\n",
            "27/27 [==============================] - 3s 126ms/step - loss: 0.4004 - accuracy: 0.8372\n",
            "Epoch 50/50\n",
            "27/27 [==============================] - 3s 125ms/step - loss: 0.4050 - accuracy: 0.8047\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "56jAfuOX6Mer"
      },
      "source": [
        "### Evaluating the neural network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wBS5A33MXWl8",
        "outputId": "3b6540bf-7bac-498b-9227-accad7e4317e"
      },
      "source": [
        "test_dataset.class_indices"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'bart': 0, 'homer': 1}"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D1iGUjLKXccY",
        "outputId": "ab1e4bfd-7557-4e61-ea39-d4a617ac42c9"
      },
      "source": [
        "predictions = network.predict(test_dataset)\n",
        "predictions"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[8.51176023e-01, 1.48823977e-01],\n",
              "       [7.76615262e-01, 2.23384753e-01],\n",
              "       [7.72824064e-02, 9.22717571e-01],\n",
              "       [9.16312993e-01, 8.36870596e-02],\n",
              "       [9.32460964e-01, 6.75390959e-02],\n",
              "       [8.84699106e-01, 1.15300879e-01],\n",
              "       [2.50220229e-03, 9.97497737e-01],\n",
              "       [4.52095509e-01, 5.47904551e-01],\n",
              "       [3.33176106e-01, 6.66823864e-01],\n",
              "       [4.68429387e-01, 5.31570554e-01],\n",
              "       [5.35562634e-02, 9.46443737e-01],\n",
              "       [9.13379610e-01, 8.66204053e-02],\n",
              "       [7.94401348e-01, 2.05598637e-01],\n",
              "       [1.33498013e-01, 8.66501927e-01],\n",
              "       [5.71784973e-01, 4.28214997e-01],\n",
              "       [9.65273499e-01, 3.47264819e-02],\n",
              "       [4.59157303e-02, 9.54084218e-01],\n",
              "       [2.59099782e-01, 7.40900218e-01],\n",
              "       [8.72573853e-01, 1.27426192e-01],\n",
              "       [8.91583800e-01, 1.08416215e-01],\n",
              "       [8.64170566e-02, 9.13582921e-01],\n",
              "       [8.57063651e-01, 1.42936349e-01],\n",
              "       [9.99063075e-01, 9.36999917e-04],\n",
              "       [7.15064943e-01, 2.84935027e-01],\n",
              "       [9.38447118e-01, 6.15528338e-02],\n",
              "       [6.52244031e-01, 3.47755969e-01],\n",
              "       [9.83912647e-01, 1.60873570e-02],\n",
              "       [8.74587774e-01, 1.25412196e-01],\n",
              "       [1.69063676e-02, 9.83093619e-01],\n",
              "       [3.09708565e-02, 9.69029129e-01],\n",
              "       [2.03272533e-02, 9.79672790e-01],\n",
              "       [2.36833021e-02, 9.76316750e-01],\n",
              "       [4.90210146e-01, 5.09789884e-01],\n",
              "       [5.67529444e-03, 9.94324625e-01],\n",
              "       [1.47034312e-04, 9.99852896e-01],\n",
              "       [1.06367759e-01, 8.93632233e-01],\n",
              "       [8.55326280e-03, 9.91446733e-01],\n",
              "       [1.32128093e-02, 9.86787200e-01],\n",
              "       [1.48785571e-02, 9.85121489e-01],\n",
              "       [1.56214451e-02, 9.84378517e-01],\n",
              "       [6.46642689e-03, 9.93533611e-01],\n",
              "       [1.82749182e-02, 9.81725037e-01],\n",
              "       [1.08421423e-01, 8.91578615e-01],\n",
              "       [1.58224717e-01, 8.41775298e-01],\n",
              "       [6.08914718e-03, 9.93910909e-01],\n",
              "       [9.46734995e-02, 9.05326545e-01],\n",
              "       [2.26846038e-04, 9.99773204e-01],\n",
              "       [2.47570910e-02, 9.75242853e-01],\n",
              "       [1.39367478e-02, 9.86063302e-01],\n",
              "       [1.50780126e-01, 8.49219918e-01],\n",
              "       [3.86232026e-02, 9.61376846e-01],\n",
              "       [3.06771010e-01, 6.93229020e-01],\n",
              "       [4.68814410e-02, 9.53118563e-01],\n",
              "       [2.92515215e-02, 9.70748425e-01]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vF76UifaXmYZ",
        "outputId": "4514f8a8-fd0d-450c-dc29-2eda263091b2"
      },
      "source": [
        "predictions = np.argmax(predictions, axis = 1)\n",
        "predictions"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 0, 1, 0,\n",
              "       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pRGDiU0QXugX",
        "outputId": "80349e90-9692-41b7-8c21-b4aabbe5f5a9"
      },
      "source": [
        "test_dataset.classes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=int32)"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EG_zxlz8X4W4",
        "outputId": "7cd5479c-60f5-4c68-a250-131147de0618"
      },
      "source": [
        "# Approach 1 (all pixels): 0.68\n",
        "# Approach 2 (features): 0.88\n",
        "# Approach 3 (CNN): 0.92\n",
        "# Approach 4 (Transfer learning): 0.81\n",
        "from sklearn.metrics import accuracy_score\n",
        "accuracy_score(test_dataset.classes, predictions)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8148148148148148"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "ZUs1g87lYk3i",
        "outputId": "4da71225-2180-4e88-f3ca-5d562c2e1299"
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "cm = confusion_matrix(test_dataset.classes, predictions)\n",
        "sns.heatmap(cm, annot=True);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9qeLVB_rY6Fi",
        "outputId": "358a11b8-58c3-46a8-9236-3423c42fa3d5"
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "print(classification_report(test_dataset.classes, predictions))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.64      0.78        28\n",
            "           1       0.72      1.00      0.84        26\n",
            "\n",
            "    accuracy                           0.81        54\n",
            "   macro avg       0.86      0.82      0.81        54\n",
            "weighted avg       0.87      0.81      0.81        54\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RyD7gb-p9pvo"
      },
      "source": [
        "## Fine tuning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "55f3LLEUBZha"
      },
      "source": [
        "### Implementing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_ymTbCqBbnbL"
      },
      "source": [
        "base_model.trainable = True"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aR1yByTpbs6z",
        "outputId": "8ea7851d-dd9d-455a-e30f-ec15550674b3"
      },
      "source": [
        "for layer in base_model.layers:\n",
        "  print(layer, layer.trainable)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<keras.engine.input_layer.InputLayer object at 0x7f4fd2539990> True\n",
            "<keras.layers.convolutional.ZeroPadding2D object at 0x7f4fd358e290> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fddc9f690> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd2541c10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd24922d0> True\n",
            "<keras.layers.convolutional.ZeroPadding2D object at 0x7f4fd17f0bd0> True\n",
            "<keras.layers.pooling.MaxPooling2D object at 0x7f4fd17a2590> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17a8990> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17bc390> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd17c51d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17be510> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd174e550> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1758450> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17a88d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17be150> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17b01d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd17a83d0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd17f00d0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd17ca910> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd363f7d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1761850> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1768790> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17686d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1773090> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd177b390> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd177b850> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1782050> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd170e450> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1773490> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1714f10> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd171e290> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1722650> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1725690> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd172ca50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1735390> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd172f4d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd173f6d0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd174aa10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd174a410> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16db390> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16d9290> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16e7f90> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f4c90> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16f8d10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1709050> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16d5910> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f88d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16d98d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1690d50> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd169c090> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1690890> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16a3cd0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16ac050> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b0990> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd169c6d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16f3a50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16ec310> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16f1790> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd172f6d0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd1735c10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16cf3d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1714710> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1780090> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1768550> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17b68d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd179ff50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b80d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16be850> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd16c4110> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd16c1b10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16b8190> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd179f9d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd2492d90> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd166bf10> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16cb890> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1678250> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1671290> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1654210> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1683cd0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd167d5d0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1689250> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1613710> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1623d50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd162ab10> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd162e790> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1637fd0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1626b50> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16195d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1640b10> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd161f9d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15ce050> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd15d2350> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15ce410> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1623190> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1619a90> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd1671190> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1689110> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd166b490> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16bafd0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd17ad690> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd174a1d0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd16f4150> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16e1590> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15e1310> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15e3650> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15e3b50> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15e10d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15efe90> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15f71d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15fc1d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd1601ed0> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd158e290> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd158e990> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd158e3d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd159e1d0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15a25d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15a7610> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15ac650> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15b3390> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15af410> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15bfa10> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd15c9410> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15c9610> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0220990> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0222cd0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0226710> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15c9450> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0237550> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0240450> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0240a10> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc024b590> True\n",
            "<keras.layers.merge.Add object at 0x7f4fc0257490> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc0257f50> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc02407d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc022c210> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15bf290> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd15b8990> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15a7ad0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd15fcf50> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd16012d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fd15e0b10> True\n",
            "<keras.layers.merge.Add object at 0x7f4fd17684d0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fd16cb710> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01e4d90> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01e4210> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01ea250> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1637b50> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01eefd0> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01fc450> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fd1678850> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01ee2d0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01dce90> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc0206050> True\n",
            "<keras.layers.merge.Add object at 0x7f4fc020e390> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc020e890> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01f8c10> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01a0310> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01a4ed0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01a9350> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01afb50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01b8490> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01a9d90> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01bfb90> True\n",
            "<keras.layers.merge.Add object at 0x7f4fc01cdd10> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc02194d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc01d4b10> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01d8e50> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01db510> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0161dd0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc016a350> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01735d0> True\n",
            "<keras.layers.convolutional.Conv2D object at 0x7f4fc0173fd0> True\n",
            "<keras.layers.normalization.batch_normalization.BatchNormalization object at 0x7f4fc01d4210> True\n",
            "<keras.layers.merge.Add object at 0x7f4fc01af910> True\n",
            "<keras.layers.core.activation.Activation object at 0x7f4fc01a9910> True\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JQL3Y2nab9wX",
        "outputId": "3b612c32-b53b-4a79-caa7-4447914f7978"
      },
      "source": [
        "len(base_model.layers)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "175"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CwkSENBQcFRq"
      },
      "source": [
        "fine_tuning_at = 140"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rJj5vMgmcYBN"
      },
      "source": [
        "for layer in base_model.layers[:fine_tuning_at]:\n",
        "  layer.trainable = False"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3HaEN1Ppc2zm"
      },
      "source": [
        "network.compile(optimizer='Adam', loss='categorical_crossentropy', metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NwTVCyI1c-sW",
        "outputId": "f8bd19eb-33d3-40e0-ce4e-b8c0c3ad3d26"
      },
      "source": [
        "history = network.fit(train_dataset, epochs=50)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "27/27 [==============================] - 8s 136ms/step - loss: 2.5447 - accuracy: 0.5674\n",
            "Epoch 2/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.5870 - accuracy: 0.6558\n",
            "Epoch 3/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.6043 - accuracy: 0.6465\n",
            "Epoch 4/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.5998 - accuracy: 0.6093\n",
            "Epoch 5/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.5363 - accuracy: 0.7256\n",
            "Epoch 6/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.5071 - accuracy: 0.7256\n",
            "Epoch 7/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.6464 - accuracy: 0.6651\n",
            "Epoch 8/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.4890 - accuracy: 0.7860\n",
            "Epoch 9/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.5151 - accuracy: 0.7628\n",
            "Epoch 10/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.4636 - accuracy: 0.8140\n",
            "Epoch 11/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.4513 - accuracy: 0.7953\n",
            "Epoch 12/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.4074 - accuracy: 0.8279\n",
            "Epoch 13/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.4445 - accuracy: 0.8093\n",
            "Epoch 14/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.4004 - accuracy: 0.8372\n",
            "Epoch 15/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.3541 - accuracy: 0.8512\n",
            "Epoch 16/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.3537 - accuracy: 0.8512\n",
            "Epoch 17/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.3076 - accuracy: 0.8605\n",
            "Epoch 18/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.3400 - accuracy: 0.8465\n",
            "Epoch 19/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.3271 - accuracy: 0.8837\n",
            "Epoch 20/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.3050 - accuracy: 0.8791\n",
            "Epoch 21/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.2609 - accuracy: 0.8884\n",
            "Epoch 22/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2484 - accuracy: 0.8930\n",
            "Epoch 23/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.3491 - accuracy: 0.8605\n",
            "Epoch 24/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2560 - accuracy: 0.9070\n",
            "Epoch 25/50\n",
            "27/27 [==============================] - 4s 132ms/step - loss: 0.3574 - accuracy: 0.8930\n",
            "Epoch 26/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2592 - accuracy: 0.8884\n",
            "Epoch 27/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.2183 - accuracy: 0.8930\n",
            "Epoch 28/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2135 - accuracy: 0.9116\n",
            "Epoch 29/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.3014 - accuracy: 0.8884\n",
            "Epoch 30/50\n",
            "27/27 [==============================] - 3s 127ms/step - loss: 0.2533 - accuracy: 0.9070\n",
            "Epoch 31/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.1819 - accuracy: 0.9163\n",
            "Epoch 32/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.1893 - accuracy: 0.9256\n",
            "Epoch 33/50\n",
            "27/27 [==============================] - 4s 128ms/step - loss: 0.2262 - accuracy: 0.9209\n",
            "Epoch 34/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.2580 - accuracy: 0.9023\n",
            "Epoch 35/50\n",
            "27/27 [==============================] - 4s 128ms/step - loss: 0.2543 - accuracy: 0.8791\n",
            "Epoch 36/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2501 - accuracy: 0.8791\n",
            "Epoch 37/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2481 - accuracy: 0.9023\n",
            "Epoch 38/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.1543 - accuracy: 0.9395\n",
            "Epoch 39/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.3713 - accuracy: 0.8372\n",
            "Epoch 40/50\n",
            "27/27 [==============================] - 4s 128ms/step - loss: 0.2430 - accuracy: 0.9302\n",
            "Epoch 41/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.1573 - accuracy: 0.9395\n",
            "Epoch 42/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2823 - accuracy: 0.9070\n",
            "Epoch 43/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.1942 - accuracy: 0.9302\n",
            "Epoch 44/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.1414 - accuracy: 0.9488\n",
            "Epoch 45/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.3727 - accuracy: 0.8884\n",
            "Epoch 46/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2141 - accuracy: 0.9023\n",
            "Epoch 47/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.3013 - accuracy: 0.8837\n",
            "Epoch 48/50\n",
            "27/27 [==============================] - 4s 131ms/step - loss: 0.0898 - accuracy: 0.9674\n",
            "Epoch 49/50\n",
            "27/27 [==============================] - 4s 130ms/step - loss: 0.2361 - accuracy: 0.8884\n",
            "Epoch 50/50\n",
            "27/27 [==============================] - 4s 129ms/step - loss: 0.1911 - accuracy: 0.9070\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pAPmwo0TBfPF"
      },
      "source": [
        "### Evaluate "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "L_8Jv7P1dO8Z",
        "outputId": "fbd91d4b-f68d-4e35-eb40-370661c9398d"
      },
      "source": [
        "predictions = network.predict(test_dataset)\n",
        "predictions"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[9.99977231e-01, 2.27449436e-05],\n",
              "       [9.97888505e-01, 2.11154995e-03],\n",
              "       [1.71669549e-03, 9.98283386e-01],\n",
              "       [1.00000000e+00, 4.57980249e-08],\n",
              "       [8.75493646e-01, 1.24506384e-01],\n",
              "       [9.99986291e-01, 1.37653778e-05],\n",
              "       [4.99256134e-01, 5.00743926e-01],\n",
              "       [4.29088086e-01, 5.70911884e-01],\n",
              "       [9.92058575e-01, 7.94147979e-03],\n",
              "       [9.15926218e-01, 8.40737969e-02],\n",
              "       [4.72247812e-05, 9.99952793e-01],\n",
              "       [8.90079439e-01, 1.09920591e-01],\n",
              "       [9.99947786e-01, 5.22429036e-05],\n",
              "       [1.12185970e-01, 8.87813985e-01],\n",
              "       [3.99264731e-02, 9.60073531e-01],\n",
              "       [9.99736369e-01, 2.63652240e-04],\n",
              "       [9.43920493e-01, 5.60795031e-02],\n",
              "       [9.99437153e-01, 5.62788453e-04],\n",
              "       [9.96901989e-01, 3.09806038e-03],\n",
              "       [1.00000000e+00, 8.85852169e-10],\n",
              "       [1.47768762e-03, 9.98522341e-01],\n",
              "       [9.97818112e-01, 2.18193769e-03],\n",
              "       [9.94106054e-01, 5.89392334e-03],\n",
              "       [1.47159565e-02, 9.85284090e-01],\n",
              "       [9.99996781e-01, 3.16782462e-06],\n",
              "       [9.63658214e-01, 3.63417789e-02],\n",
              "       [9.99937654e-01, 6.23223896e-05],\n",
              "       [1.00000000e+00, 2.69379963e-09],\n",
              "       [1.02562190e-07, 9.99999881e-01],\n",
              "       [2.60151467e-07, 9.99999762e-01],\n",
              "       [9.54656909e-11, 1.00000000e+00],\n",
              "       [8.26676478e-05, 9.99917269e-01],\n",
              "       [2.85583286e-04, 9.99714434e-01],\n",
              "       [9.01324383e-05, 9.99909878e-01],\n",
              "       [1.26473652e-02, 9.87352610e-01],\n",
              "       [3.63126310e-05, 9.99963641e-01],\n",
              "       [1.19935023e-03, 9.98800635e-01],\n",
              "       [2.45862140e-08, 1.00000000e+00],\n",
              "       [2.85982707e-04, 9.99713957e-01],\n",
              "       [1.41079937e-07, 9.99999881e-01],\n",
              "       [6.94953940e-07, 9.99999285e-01],\n",
              "       [2.81805974e-08, 1.00000000e+00],\n",
              "       [2.99264125e-09, 1.00000000e+00],\n",
              "       [2.15054513e-03, 9.97849464e-01],\n",
              "       [3.03931778e-08, 1.00000000e+00],\n",
              "       [1.92350278e-08, 1.00000000e+00],\n",
              "       [4.67172595e-06, 9.99995351e-01],\n",
              "       [1.48444670e-08, 1.00000000e+00],\n",
              "       [2.64803930e-08, 1.00000000e+00],\n",
              "       [3.71666601e-06, 9.99996305e-01],\n",
              "       [1.49689220e-06, 9.99998450e-01],\n",
              "       [3.07852998e-02, 9.69214737e-01],\n",
              "       [2.40094145e-03, 9.97599065e-01],\n",
              "       [1.00601049e-12, 1.00000000e+00]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dlw0rawpdRUs",
        "outputId": "e2ef05f7-af3c-4896-a8dd-1e919fe5e60a"
      },
      "source": [
        "predictions = np.argmax(predictions, axis = 1)\n",
        "predictions"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0,\n",
              "       0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1])"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oiRx3Tk_dTeH",
        "outputId": "cca0893f-796d-4be7-ca13-48e2e8b68e99"
      },
      "source": [
        "test_dataset.classes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
              "       0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\n",
              "       1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=int32)"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "G_TIGJ4ndVdt",
        "outputId": "421a3dbb-95a4-4dc7-c190-1c4f915ff1b1"
      },
      "source": [
        "# Approach 1 (all pixels): 0.68\n",
        "# Approach 2 (features): 0.88\n",
        "# Approach 3 (CNN): 0.92\n",
        "# Approach 4 (Transfer learning): 0.81\n",
        "# Approach 4 (Fine tuning): 0.85\n",
        "from sklearn.metrics import accuracy_score\n",
        "accuracy_score(test_dataset.classes, predictions)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8518518518518519"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ASbKQb-TflfV"
      },
      "source": [
        "## Saving and loading the model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a9_C7ONNfrYv"
      },
      "source": [
        "model_json = network.to_json()\n",
        "with open('network.json','w') as json_file:\n",
        "  json_file.write(model_json)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4gNleE1_fxFx"
      },
      "source": [
        "from keras.models import save_model\n",
        "network_saved = save_model(network, '/content/weights.hdf5')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 199
        },
        "id": "Mrk9tRuxgX9l",
        "outputId": "80917665-d446-4154-829e-b658a00cd6f3"
      },
      "source": [
        "with open('network.json', 'r') as json_file:\n",
        "  json_saved_model = json_file.read()\n",
        "json_saved_model"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'{\"class_name\": \"Functional\", \"config\": {\"name\": \"model\", \"layers\": [{\"class_name\": \"InputLayer\", \"config\": {\"batch_input_shape\": [null, 256, 256, 3], \"dtype\": \"float32\", \"sparse\": false, \"ragged\": false, \"name\": \"input_3\"}, \"name\": \"input_3\", \"inbound_nodes\": []}, {\"class_name\": \"ZeroPadding2D\", \"config\": {\"name\": \"conv1_pad\", \"trainable\": false, \"dtype\": \"float32\", \"padding\": [[3, 3], [3, 3]], \"data_format\": \"channels_last\"}, \"name\": \"conv1_pad\", \"inbound_nodes\": [[[\"input_3\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [7, 7], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv1_conv\", \"inbound_nodes\": [[[\"conv1_pad\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv1_bn\", \"inbound_nodes\": [[[\"conv1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv1_relu\", \"inbound_nodes\": [[[\"conv1_bn\", 0, 0, {}]]]}, {\"class_name\": \"ZeroPadding2D\", \"config\": {\"name\": \"pool1_pad\", \"trainable\": false, \"dtype\": \"float32\", \"padding\": [[1, 1], [1, 1]], \"data_format\": \"channels_last\"}, \"name\": \"pool1_pad\", \"inbound_nodes\": [[[\"conv1_relu\", 0, 0, {}]]]}, {\"class_name\": \"MaxPooling2D\", \"config\": {\"name\": \"pool1_pool\", \"trainable\": false, \"dtype\": \"float32\", \"pool_size\": [3, 3], \"padding\": \"valid\", \"strides\": [2, 2], \"data_format\": \"channels_last\"}, \"name\": \"pool1_pool\", \"inbound_nodes\": [[[\"pool1_pad\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block1_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block1_1_conv\", \"inbound_nodes\": [[[\"pool1_pool\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block1_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block1_1_bn\", \"inbound_nodes\": [[[\"conv2_block1_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block1_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block1_1_relu\", \"inbound_nodes\": [[[\"conv2_block1_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block1_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block1_2_conv\", \"inbound_nodes\": [[[\"conv2_block1_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block1_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block1_2_bn\", \"inbound_nodes\": [[[\"conv2_block1_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block1_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block1_2_relu\", \"inbound_nodes\": [[[\"conv2_block1_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block1_0_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block1_0_conv\", \"inbound_nodes\": [[[\"pool1_pool\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block1_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block1_3_conv\", \"inbound_nodes\": [[[\"conv2_block1_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block1_0_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block1_0_bn\", \"inbound_nodes\": [[[\"conv2_block1_0_conv\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block1_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block1_3_bn\", \"inbound_nodes\": [[[\"conv2_block1_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv2_block1_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv2_block1_add\", \"inbound_nodes\": [[[\"conv2_block1_0_bn\", 0, 0, {}], [\"conv2_block1_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block1_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block1_out\", \"inbound_nodes\": [[[\"conv2_block1_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block2_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block2_1_conv\", \"inbound_nodes\": [[[\"conv2_block1_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block2_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block2_1_bn\", \"inbound_nodes\": [[[\"conv2_block2_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block2_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block2_1_relu\", \"inbound_nodes\": [[[\"conv2_block2_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block2_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block2_2_conv\", \"inbound_nodes\": [[[\"conv2_block2_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block2_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block2_2_bn\", \"inbound_nodes\": [[[\"conv2_block2_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block2_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block2_2_relu\", \"inbound_nodes\": [[[\"conv2_block2_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block2_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block2_3_conv\", \"inbound_nodes\": [[[\"conv2_block2_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block2_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block2_3_bn\", \"inbound_nodes\": [[[\"conv2_block2_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv2_block2_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv2_block2_add\", \"inbound_nodes\": [[[\"conv2_block1_out\", 0, 0, {}], [\"conv2_block2_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block2_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block2_out\", \"inbound_nodes\": [[[\"conv2_block2_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block3_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block3_1_conv\", \"inbound_nodes\": [[[\"conv2_block2_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block3_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block3_1_bn\", \"inbound_nodes\": [[[\"conv2_block3_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block3_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block3_1_relu\", \"inbound_nodes\": [[[\"conv2_block3_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block3_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 64, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block3_2_conv\", \"inbound_nodes\": [[[\"conv2_block3_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block3_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block3_2_bn\", \"inbound_nodes\": [[[\"conv2_block3_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block3_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block3_2_relu\", \"inbound_nodes\": [[[\"conv2_block3_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv2_block3_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv2_block3_3_conv\", \"inbound_nodes\": [[[\"conv2_block3_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv2_block3_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv2_block3_3_bn\", \"inbound_nodes\": [[[\"conv2_block3_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv2_block3_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv2_block3_add\", \"inbound_nodes\": [[[\"conv2_block2_out\", 0, 0, {}], [\"conv2_block3_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv2_block3_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv2_block3_out\", \"inbound_nodes\": [[[\"conv2_block3_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block1_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block1_1_conv\", \"inbound_nodes\": [[[\"conv2_block3_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block1_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block1_1_bn\", \"inbound_nodes\": [[[\"conv3_block1_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block1_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block1_1_relu\", \"inbound_nodes\": [[[\"conv3_block1_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block1_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block1_2_conv\", \"inbound_nodes\": [[[\"conv3_block1_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block1_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block1_2_bn\", \"inbound_nodes\": [[[\"conv3_block1_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block1_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block1_2_relu\", \"inbound_nodes\": [[[\"conv3_block1_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block1_0_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block1_0_conv\", \"inbound_nodes\": [[[\"conv2_block3_out\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block1_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block1_3_conv\", \"inbound_nodes\": [[[\"conv3_block1_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block1_0_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block1_0_bn\", \"inbound_nodes\": [[[\"conv3_block1_0_conv\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block1_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block1_3_bn\", \"inbound_nodes\": [[[\"conv3_block1_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv3_block1_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv3_block1_add\", \"inbound_nodes\": [[[\"conv3_block1_0_bn\", 0, 0, {}], [\"conv3_block1_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block1_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block1_out\", \"inbound_nodes\": [[[\"conv3_block1_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block2_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block2_1_conv\", \"inbound_nodes\": [[[\"conv3_block1_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block2_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block2_1_bn\", \"inbound_nodes\": [[[\"conv3_block2_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block2_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block2_1_relu\", \"inbound_nodes\": [[[\"conv3_block2_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block2_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block2_2_conv\", \"inbound_nodes\": [[[\"conv3_block2_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block2_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block2_2_bn\", \"inbound_nodes\": [[[\"conv3_block2_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block2_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block2_2_relu\", \"inbound_nodes\": [[[\"conv3_block2_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block2_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block2_3_conv\", \"inbound_nodes\": [[[\"conv3_block2_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block2_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block2_3_bn\", \"inbound_nodes\": [[[\"conv3_block2_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv3_block2_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv3_block2_add\", \"inbound_nodes\": [[[\"conv3_block1_out\", 0, 0, {}], [\"conv3_block2_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block2_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block2_out\", \"inbound_nodes\": [[[\"conv3_block2_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block3_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block3_1_conv\", \"inbound_nodes\": [[[\"conv3_block2_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block3_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block3_1_bn\", \"inbound_nodes\": [[[\"conv3_block3_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block3_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block3_1_relu\", \"inbound_nodes\": [[[\"conv3_block3_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block3_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block3_2_conv\", \"inbound_nodes\": [[[\"conv3_block3_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block3_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block3_2_bn\", \"inbound_nodes\": [[[\"conv3_block3_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block3_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block3_2_relu\", \"inbound_nodes\": [[[\"conv3_block3_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block3_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block3_3_conv\", \"inbound_nodes\": [[[\"conv3_block3_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block3_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block3_3_bn\", \"inbound_nodes\": [[[\"conv3_block3_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv3_block3_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv3_block3_add\", \"inbound_nodes\": [[[\"conv3_block2_out\", 0, 0, {}], [\"conv3_block3_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block3_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block3_out\", \"inbound_nodes\": [[[\"conv3_block3_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block4_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block4_1_conv\", \"inbound_nodes\": [[[\"conv3_block3_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block4_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block4_1_bn\", \"inbound_nodes\": [[[\"conv3_block4_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block4_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block4_1_relu\", \"inbound_nodes\": [[[\"conv3_block4_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block4_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 128, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block4_2_conv\", \"inbound_nodes\": [[[\"conv3_block4_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block4_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block4_2_bn\", \"inbound_nodes\": [[[\"conv3_block4_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block4_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block4_2_relu\", \"inbound_nodes\": [[[\"conv3_block4_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv3_block4_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv3_block4_3_conv\", \"inbound_nodes\": [[[\"conv3_block4_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv3_block4_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv3_block4_3_bn\", \"inbound_nodes\": [[[\"conv3_block4_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv3_block4_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv3_block4_add\", \"inbound_nodes\": [[[\"conv3_block3_out\", 0, 0, {}], [\"conv3_block4_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv3_block4_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv3_block4_out\", \"inbound_nodes\": [[[\"conv3_block4_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block1_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block1_1_conv\", \"inbound_nodes\": [[[\"conv3_block4_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block1_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block1_1_bn\", \"inbound_nodes\": [[[\"conv4_block1_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block1_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block1_1_relu\", \"inbound_nodes\": [[[\"conv4_block1_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block1_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block1_2_conv\", \"inbound_nodes\": [[[\"conv4_block1_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block1_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block1_2_bn\", \"inbound_nodes\": [[[\"conv4_block1_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block1_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block1_2_relu\", \"inbound_nodes\": [[[\"conv4_block1_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block1_0_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block1_0_conv\", \"inbound_nodes\": [[[\"conv3_block4_out\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block1_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block1_3_conv\", \"inbound_nodes\": [[[\"conv4_block1_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block1_0_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block1_0_bn\", \"inbound_nodes\": [[[\"conv4_block1_0_conv\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block1_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block1_3_bn\", \"inbound_nodes\": [[[\"conv4_block1_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block1_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv4_block1_add\", \"inbound_nodes\": [[[\"conv4_block1_0_bn\", 0, 0, {}], [\"conv4_block1_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block1_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block1_out\", \"inbound_nodes\": [[[\"conv4_block1_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block2_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block2_1_conv\", \"inbound_nodes\": [[[\"conv4_block1_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block2_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block2_1_bn\", \"inbound_nodes\": [[[\"conv4_block2_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block2_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block2_1_relu\", \"inbound_nodes\": [[[\"conv4_block2_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block2_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block2_2_conv\", \"inbound_nodes\": [[[\"conv4_block2_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block2_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block2_2_bn\", \"inbound_nodes\": [[[\"conv4_block2_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block2_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block2_2_relu\", \"inbound_nodes\": [[[\"conv4_block2_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block2_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block2_3_conv\", \"inbound_nodes\": [[[\"conv4_block2_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block2_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block2_3_bn\", \"inbound_nodes\": [[[\"conv4_block2_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block2_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv4_block2_add\", \"inbound_nodes\": [[[\"conv4_block1_out\", 0, 0, {}], [\"conv4_block2_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block2_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block2_out\", \"inbound_nodes\": [[[\"conv4_block2_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block3_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block3_1_conv\", \"inbound_nodes\": [[[\"conv4_block2_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block3_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block3_1_bn\", \"inbound_nodes\": [[[\"conv4_block3_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block3_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block3_1_relu\", \"inbound_nodes\": [[[\"conv4_block3_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block3_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block3_2_conv\", \"inbound_nodes\": [[[\"conv4_block3_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block3_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block3_2_bn\", \"inbound_nodes\": [[[\"conv4_block3_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block3_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block3_2_relu\", \"inbound_nodes\": [[[\"conv4_block3_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block3_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block3_3_conv\", \"inbound_nodes\": [[[\"conv4_block3_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block3_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block3_3_bn\", \"inbound_nodes\": [[[\"conv4_block3_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block3_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv4_block3_add\", \"inbound_nodes\": [[[\"conv4_block2_out\", 0, 0, {}], [\"conv4_block3_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block3_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block3_out\", \"inbound_nodes\": [[[\"conv4_block3_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block4_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block4_1_conv\", \"inbound_nodes\": [[[\"conv4_block3_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block4_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block4_1_bn\", \"inbound_nodes\": [[[\"conv4_block4_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block4_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block4_1_relu\", \"inbound_nodes\": [[[\"conv4_block4_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block4_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block4_2_conv\", \"inbound_nodes\": [[[\"conv4_block4_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block4_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block4_2_bn\", \"inbound_nodes\": [[[\"conv4_block4_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block4_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block4_2_relu\", \"inbound_nodes\": [[[\"conv4_block4_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block4_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block4_3_conv\", \"inbound_nodes\": [[[\"conv4_block4_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block4_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block4_3_bn\", \"inbound_nodes\": [[[\"conv4_block4_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block4_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv4_block4_add\", \"inbound_nodes\": [[[\"conv4_block3_out\", 0, 0, {}], [\"conv4_block4_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block4_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block4_out\", \"inbound_nodes\": [[[\"conv4_block4_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block5_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block5_1_conv\", \"inbound_nodes\": [[[\"conv4_block4_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block5_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block5_1_bn\", \"inbound_nodes\": [[[\"conv4_block5_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block5_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block5_1_relu\", \"inbound_nodes\": [[[\"conv4_block5_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block5_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block5_2_conv\", \"inbound_nodes\": [[[\"conv4_block5_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block5_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block5_2_bn\", \"inbound_nodes\": [[[\"conv4_block5_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block5_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block5_2_relu\", \"inbound_nodes\": [[[\"conv4_block5_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block5_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block5_3_conv\", \"inbound_nodes\": [[[\"conv4_block5_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block5_3_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block5_3_bn\", \"inbound_nodes\": [[[\"conv4_block5_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block5_add\", \"trainable\": false, \"dtype\": \"float32\"}, \"name\": \"conv4_block5_add\", \"inbound_nodes\": [[[\"conv4_block4_out\", 0, 0, {}], [\"conv4_block5_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block5_out\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block5_out\", \"inbound_nodes\": [[[\"conv4_block5_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block6_1_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block6_1_conv\", \"inbound_nodes\": [[[\"conv4_block5_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block6_1_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block6_1_bn\", \"inbound_nodes\": [[[\"conv4_block6_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block6_1_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block6_1_relu\", \"inbound_nodes\": [[[\"conv4_block6_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block6_2_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 256, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block6_2_conv\", \"inbound_nodes\": [[[\"conv4_block6_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block6_2_bn\", \"trainable\": false, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block6_2_bn\", \"inbound_nodes\": [[[\"conv4_block6_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block6_2_relu\", \"trainable\": false, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block6_2_relu\", \"inbound_nodes\": [[[\"conv4_block6_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv4_block6_3_conv\", \"trainable\": false, \"dtype\": \"float32\", \"filters\": 1024, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv4_block6_3_conv\", \"inbound_nodes\": [[[\"conv4_block6_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv4_block6_3_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv4_block6_3_bn\", \"inbound_nodes\": [[[\"conv4_block6_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv4_block6_add\", \"trainable\": true, \"dtype\": \"float32\"}, \"name\": \"conv4_block6_add\", \"inbound_nodes\": [[[\"conv4_block5_out\", 0, 0, {}], [\"conv4_block6_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv4_block6_out\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv4_block6_out\", \"inbound_nodes\": [[[\"conv4_block6_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block1_1_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block1_1_conv\", \"inbound_nodes\": [[[\"conv4_block6_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block1_1_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block1_1_bn\", \"inbound_nodes\": [[[\"conv5_block1_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block1_1_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block1_1_relu\", \"inbound_nodes\": [[[\"conv5_block1_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block1_2_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block1_2_conv\", \"inbound_nodes\": [[[\"conv5_block1_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block1_2_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block1_2_bn\", \"inbound_nodes\": [[[\"conv5_block1_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block1_2_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block1_2_relu\", \"inbound_nodes\": [[[\"conv5_block1_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block1_0_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 2048, \"kernel_size\": [1, 1], \"strides\": [2, 2], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block1_0_conv\", \"inbound_nodes\": [[[\"conv4_block6_out\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block1_3_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 2048, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block1_3_conv\", \"inbound_nodes\": [[[\"conv5_block1_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block1_0_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block1_0_bn\", \"inbound_nodes\": [[[\"conv5_block1_0_conv\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block1_3_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block1_3_bn\", \"inbound_nodes\": [[[\"conv5_block1_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv5_block1_add\", \"trainable\": true, \"dtype\": \"float32\"}, \"name\": \"conv5_block1_add\", \"inbound_nodes\": [[[\"conv5_block1_0_bn\", 0, 0, {}], [\"conv5_block1_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block1_out\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block1_out\", \"inbound_nodes\": [[[\"conv5_block1_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block2_1_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block2_1_conv\", \"inbound_nodes\": [[[\"conv5_block1_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block2_1_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block2_1_bn\", \"inbound_nodes\": [[[\"conv5_block2_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block2_1_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block2_1_relu\", \"inbound_nodes\": [[[\"conv5_block2_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block2_2_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block2_2_conv\", \"inbound_nodes\": [[[\"conv5_block2_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block2_2_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block2_2_bn\", \"inbound_nodes\": [[[\"conv5_block2_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block2_2_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block2_2_relu\", \"inbound_nodes\": [[[\"conv5_block2_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block2_3_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 2048, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block2_3_conv\", \"inbound_nodes\": [[[\"conv5_block2_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block2_3_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block2_3_bn\", \"inbound_nodes\": [[[\"conv5_block2_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv5_block2_add\", \"trainable\": true, \"dtype\": \"float32\"}, \"name\": \"conv5_block2_add\", \"inbound_nodes\": [[[\"conv5_block1_out\", 0, 0, {}], [\"conv5_block2_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block2_out\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block2_out\", \"inbound_nodes\": [[[\"conv5_block2_add\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block3_1_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block3_1_conv\", \"inbound_nodes\": [[[\"conv5_block2_out\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block3_1_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block3_1_bn\", \"inbound_nodes\": [[[\"conv5_block3_1_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block3_1_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block3_1_relu\", \"inbound_nodes\": [[[\"conv5_block3_1_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block3_2_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 512, \"kernel_size\": [3, 3], \"strides\": [1, 1], \"padding\": \"same\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block3_2_conv\", \"inbound_nodes\": [[[\"conv5_block3_1_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block3_2_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block3_2_bn\", \"inbound_nodes\": [[[\"conv5_block3_2_conv\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block3_2_relu\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block3_2_relu\", \"inbound_nodes\": [[[\"conv5_block3_2_bn\", 0, 0, {}]]]}, {\"class_name\": \"Conv2D\", \"config\": {\"name\": \"conv5_block3_3_conv\", \"trainable\": true, \"dtype\": \"float32\", \"filters\": 2048, \"kernel_size\": [1, 1], \"strides\": [1, 1], \"padding\": \"valid\", \"data_format\": \"channels_last\", \"dilation_rate\": [1, 1], \"groups\": 1, \"activation\": \"linear\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"conv5_block3_3_conv\", \"inbound_nodes\": [[[\"conv5_block3_2_relu\", 0, 0, {}]]]}, {\"class_name\": \"BatchNormalization\", \"config\": {\"name\": \"conv5_block3_3_bn\", \"trainable\": true, \"dtype\": \"float32\", \"axis\": [3], \"momentum\": 0.99, \"epsilon\": 1.001e-05, \"center\": true, \"scale\": true, \"beta_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"gamma_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"moving_mean_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"moving_variance_initializer\": {\"class_name\": \"Ones\", \"config\": {}}, \"beta_regularizer\": null, \"gamma_regularizer\": null, \"beta_constraint\": null, \"gamma_constraint\": null}, \"name\": \"conv5_block3_3_bn\", \"inbound_nodes\": [[[\"conv5_block3_3_conv\", 0, 0, {}]]]}, {\"class_name\": \"Add\", \"config\": {\"name\": \"conv5_block3_add\", \"trainable\": true, \"dtype\": \"float32\"}, \"name\": \"conv5_block3_add\", \"inbound_nodes\": [[[\"conv5_block2_out\", 0, 0, {}], [\"conv5_block3_3_bn\", 0, 0, {}]]]}, {\"class_name\": \"Activation\", \"config\": {\"name\": \"conv5_block3_out\", \"trainable\": true, \"dtype\": \"float32\", \"activation\": \"relu\"}, \"name\": \"conv5_block3_out\", \"inbound_nodes\": [[[\"conv5_block3_add\", 0, 0, {}]]]}, {\"class_name\": \"GlobalAveragePooling2D\", \"config\": {\"name\": \"global_average_pooling2d_1\", \"trainable\": true, \"dtype\": \"float32\", \"data_format\": \"channels_last\"}, \"name\": \"global_average_pooling2d_1\", \"inbound_nodes\": [[[\"conv5_block3_out\", 0, 0, {}]]]}, {\"class_name\": \"Dense\", \"config\": {\"name\": \"dense\", \"trainable\": true, \"dtype\": \"float32\", \"units\": 1025, \"activation\": \"relu\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"dense\", \"inbound_nodes\": [[[\"global_average_pooling2d_1\", 0, 0, {}]]]}, {\"class_name\": \"Dropout\", \"config\": {\"name\": \"dropout\", \"trainable\": true, \"dtype\": \"float32\", \"rate\": 0.2, \"noise_shape\": null, \"seed\": null}, \"name\": \"dropout\", \"inbound_nodes\": [[[\"dense\", 0, 0, {}]]]}, {\"class_name\": \"Dense\", \"config\": {\"name\": \"dense_1\", \"trainable\": true, \"dtype\": \"float32\", \"units\": 1025, \"activation\": \"relu\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"dense_1\", \"inbound_nodes\": [[[\"dropout\", 0, 0, {}]]]}, {\"class_name\": \"Dropout\", \"config\": {\"name\": \"dropout_1\", \"trainable\": true, \"dtype\": \"float32\", \"rate\": 0.2, \"noise_shape\": null, \"seed\": null}, \"name\": \"dropout_1\", \"inbound_nodes\": [[[\"dense_1\", 0, 0, {}]]]}, {\"class_name\": \"Dense\", \"config\": {\"name\": \"dense_2\", \"trainable\": true, \"dtype\": \"float32\", \"units\": 2, \"activation\": \"softmax\", \"use_bias\": true, \"kernel_initializer\": {\"class_name\": \"GlorotUniform\", \"config\": {\"seed\": null}}, \"bias_initializer\": {\"class_name\": \"Zeros\", \"config\": {}}, \"kernel_regularizer\": null, \"bias_regularizer\": null, \"activity_regularizer\": null, \"kernel_constraint\": null, \"bias_constraint\": null}, \"name\": \"dense_2\", \"inbound_nodes\": [[[\"dropout_1\", 0, 0, {}]]]}], \"input_layers\": [[\"input_3\", 0, 0]], \"output_layers\": [[\"dense_2\", 0, 0]]}, \"keras_version\": \"2.4.0\", \"backend\": \"tensorflow\"}'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0n9s2tMof1ou"
      },
      "source": [
        "network_loaded = tf.keras.models.model_from_json(json_saved_model)\n",
        "network_loaded.load_weights('weights.hdf5')\n",
        "network_loaded.compile(loss = 'categorical_crossentropy', optimizer='Adam', metrics=['accuracy'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Dw4M5AbDgbeX",
        "outputId": "cf7acfa4-8b53-4a7d-ef52-3f76ab19f1a6"
      },
      "source": [
        "network_loaded.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_3 (InputLayer)            [(None, 256, 256, 3) 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv1_pad (ZeroPadding2D)       (None, 262, 262, 3)  0           input_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1_conv (Conv2D)             (None, 128, 128, 64) 9472        conv1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1_bn (BatchNormalization)   (None, 128, 128, 64) 256         conv1_conv[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv1_relu (Activation)         (None, 128, 128, 64) 0           conv1_bn[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "pool1_pad (ZeroPadding2D)       (None, 130, 130, 64) 0           conv1_relu[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "pool1_pool (MaxPooling2D)       (None, 64, 64, 64)   0           pool1_pad[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_conv (Conv2D)    (None, 64, 64, 64)   4160        pool1_pool[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_1_relu (Activation (None, 64, 64, 64)   0           conv2_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_conv (Conv2D)    (None, 64, 64, 64)   36928       conv2_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_2_relu (Activation (None, 64, 64, 64)   0           conv2_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_0_conv (Conv2D)    (None, 64, 64, 256)  16640       pool1_pool[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_3_conv (Conv2D)    (None, 64, 64, 256)  16640       conv2_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_0_bn (BatchNormali (None, 64, 64, 256)  1024        conv2_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_3_bn (BatchNormali (None, 64, 64, 256)  1024        conv2_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_add (Add)          (None, 64, 64, 256)  0           conv2_block1_0_bn[0][0]          \n",
            "                                                                 conv2_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block1_out (Activation)   (None, 64, 64, 256)  0           conv2_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_conv (Conv2D)    (None, 64, 64, 64)   16448       conv2_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_1_relu (Activation (None, 64, 64, 64)   0           conv2_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_conv (Conv2D)    (None, 64, 64, 64)   36928       conv2_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_2_relu (Activation (None, 64, 64, 64)   0           conv2_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_3_conv (Conv2D)    (None, 64, 64, 256)  16640       conv2_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_3_bn (BatchNormali (None, 64, 64, 256)  1024        conv2_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_add (Add)          (None, 64, 64, 256)  0           conv2_block1_out[0][0]           \n",
            "                                                                 conv2_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block2_out (Activation)   (None, 64, 64, 256)  0           conv2_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_conv (Conv2D)    (None, 64, 64, 64)   16448       conv2_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_1_relu (Activation (None, 64, 64, 64)   0           conv2_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_conv (Conv2D)    (None, 64, 64, 64)   36928       conv2_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_bn (BatchNormali (None, 64, 64, 64)   256         conv2_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_2_relu (Activation (None, 64, 64, 64)   0           conv2_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_3_conv (Conv2D)    (None, 64, 64, 256)  16640       conv2_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_3_bn (BatchNormali (None, 64, 64, 256)  1024        conv2_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_add (Add)          (None, 64, 64, 256)  0           conv2_block2_out[0][0]           \n",
            "                                                                 conv2_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv2_block3_out (Activation)   (None, 64, 64, 256)  0           conv2_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_conv (Conv2D)    (None, 32, 32, 128)  32896       conv2_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_1_relu (Activation (None, 32, 32, 128)  0           conv3_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_conv (Conv2D)    (None, 32, 32, 128)  147584      conv3_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_2_relu (Activation (None, 32, 32, 128)  0           conv3_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_0_conv (Conv2D)    (None, 32, 32, 512)  131584      conv2_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_3_conv (Conv2D)    (None, 32, 32, 512)  66048       conv3_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_0_bn (BatchNormali (None, 32, 32, 512)  2048        conv3_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_3_bn (BatchNormali (None, 32, 32, 512)  2048        conv3_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_add (Add)          (None, 32, 32, 512)  0           conv3_block1_0_bn[0][0]          \n",
            "                                                                 conv3_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block1_out (Activation)   (None, 32, 32, 512)  0           conv3_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_conv (Conv2D)    (None, 32, 32, 128)  65664       conv3_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_1_relu (Activation (None, 32, 32, 128)  0           conv3_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_conv (Conv2D)    (None, 32, 32, 128)  147584      conv3_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_2_relu (Activation (None, 32, 32, 128)  0           conv3_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_3_conv (Conv2D)    (None, 32, 32, 512)  66048       conv3_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_3_bn (BatchNormali (None, 32, 32, 512)  2048        conv3_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_add (Add)          (None, 32, 32, 512)  0           conv3_block1_out[0][0]           \n",
            "                                                                 conv3_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block2_out (Activation)   (None, 32, 32, 512)  0           conv3_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_conv (Conv2D)    (None, 32, 32, 128)  65664       conv3_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_1_relu (Activation (None, 32, 32, 128)  0           conv3_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_conv (Conv2D)    (None, 32, 32, 128)  147584      conv3_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_2_relu (Activation (None, 32, 32, 128)  0           conv3_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_3_conv (Conv2D)    (None, 32, 32, 512)  66048       conv3_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_3_bn (BatchNormali (None, 32, 32, 512)  2048        conv3_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_add (Add)          (None, 32, 32, 512)  0           conv3_block2_out[0][0]           \n",
            "                                                                 conv3_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block3_out (Activation)   (None, 32, 32, 512)  0           conv3_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_conv (Conv2D)    (None, 32, 32, 128)  65664       conv3_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block4_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_1_relu (Activation (None, 32, 32, 128)  0           conv3_block4_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_conv (Conv2D)    (None, 32, 32, 128)  147584      conv3_block4_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block4_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_2_relu (Activation (None, 32, 32, 128)  0           conv3_block4_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_3_conv (Conv2D)    (None, 32, 32, 512)  66048       conv3_block4_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_3_bn (BatchNormali (None, 32, 32, 512)  2048        conv3_block4_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_add (Add)          (None, 32, 32, 512)  0           conv3_block3_out[0][0]           \n",
            "                                                                 conv3_block4_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv3_block4_out (Activation)   (None, 32, 32, 512)  0           conv3_block4_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_conv (Conv2D)    (None, 16, 16, 256)  131328      conv3_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_1_relu (Activation (None, 16, 16, 256)  0           conv4_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_2_relu (Activation (None, 16, 16, 256)  0           conv4_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_0_conv (Conv2D)    (None, 16, 16, 1024) 525312      conv3_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_0_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_add (Add)          (None, 16, 16, 1024) 0           conv4_block1_0_bn[0][0]          \n",
            "                                                                 conv4_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block1_out (Activation)   (None, 16, 16, 1024) 0           conv4_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_conv (Conv2D)    (None, 16, 16, 256)  262400      conv4_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_1_relu (Activation (None, 16, 16, 256)  0           conv4_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_2_relu (Activation (None, 16, 16, 256)  0           conv4_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_add (Add)          (None, 16, 16, 1024) 0           conv4_block1_out[0][0]           \n",
            "                                                                 conv4_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block2_out (Activation)   (None, 16, 16, 1024) 0           conv4_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_conv (Conv2D)    (None, 16, 16, 256)  262400      conv4_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_1_relu (Activation (None, 16, 16, 256)  0           conv4_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_2_relu (Activation (None, 16, 16, 256)  0           conv4_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_add (Add)          (None, 16, 16, 1024) 0           conv4_block2_out[0][0]           \n",
            "                                                                 conv4_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block3_out (Activation)   (None, 16, 16, 1024) 0           conv4_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_conv (Conv2D)    (None, 16, 16, 256)  262400      conv4_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block4_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_1_relu (Activation (None, 16, 16, 256)  0           conv4_block4_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block4_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block4_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_2_relu (Activation (None, 16, 16, 256)  0           conv4_block4_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block4_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block4_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_add (Add)          (None, 16, 16, 1024) 0           conv4_block3_out[0][0]           \n",
            "                                                                 conv4_block4_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block4_out (Activation)   (None, 16, 16, 1024) 0           conv4_block4_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_conv (Conv2D)    (None, 16, 16, 256)  262400      conv4_block4_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block5_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_1_relu (Activation (None, 16, 16, 256)  0           conv4_block5_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block5_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block5_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_2_relu (Activation (None, 16, 16, 256)  0           conv4_block5_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block5_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block5_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_add (Add)          (None, 16, 16, 1024) 0           conv4_block4_out[0][0]           \n",
            "                                                                 conv4_block5_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block5_out (Activation)   (None, 16, 16, 1024) 0           conv4_block5_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_conv (Conv2D)    (None, 16, 16, 256)  262400      conv4_block5_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block6_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_1_relu (Activation (None, 16, 16, 256)  0           conv4_block6_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_conv (Conv2D)    (None, 16, 16, 256)  590080      conv4_block6_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_bn (BatchNormali (None, 16, 16, 256)  1024        conv4_block6_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_2_relu (Activation (None, 16, 16, 256)  0           conv4_block6_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_3_conv (Conv2D)    (None, 16, 16, 1024) 263168      conv4_block6_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_3_bn (BatchNormali (None, 16, 16, 1024) 4096        conv4_block6_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_add (Add)          (None, 16, 16, 1024) 0           conv4_block5_out[0][0]           \n",
            "                                                                 conv4_block6_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_block6_out (Activation)   (None, 16, 16, 1024) 0           conv4_block6_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_conv (Conv2D)    (None, 8, 8, 512)    524800      conv4_block6_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block1_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_1_relu (Activation (None, 8, 8, 512)    0           conv5_block1_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_conv (Conv2D)    (None, 8, 8, 512)    2359808     conv5_block1_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block1_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_2_relu (Activation (None, 8, 8, 512)    0           conv5_block1_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_0_conv (Conv2D)    (None, 8, 8, 2048)   2099200     conv4_block6_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_3_conv (Conv2D)    (None, 8, 8, 2048)   1050624     conv5_block1_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_0_bn (BatchNormali (None, 8, 8, 2048)   8192        conv5_block1_0_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_3_bn (BatchNormali (None, 8, 8, 2048)   8192        conv5_block1_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_add (Add)          (None, 8, 8, 2048)   0           conv5_block1_0_bn[0][0]          \n",
            "                                                                 conv5_block1_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block1_out (Activation)   (None, 8, 8, 2048)   0           conv5_block1_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_conv (Conv2D)    (None, 8, 8, 512)    1049088     conv5_block1_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block2_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_1_relu (Activation (None, 8, 8, 512)    0           conv5_block2_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_conv (Conv2D)    (None, 8, 8, 512)    2359808     conv5_block2_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block2_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_2_relu (Activation (None, 8, 8, 512)    0           conv5_block2_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_3_conv (Conv2D)    (None, 8, 8, 2048)   1050624     conv5_block2_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_3_bn (BatchNormali (None, 8, 8, 2048)   8192        conv5_block2_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_add (Add)          (None, 8, 8, 2048)   0           conv5_block1_out[0][0]           \n",
            "                                                                 conv5_block2_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block2_out (Activation)   (None, 8, 8, 2048)   0           conv5_block2_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_conv (Conv2D)    (None, 8, 8, 512)    1049088     conv5_block2_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block3_1_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_1_relu (Activation (None, 8, 8, 512)    0           conv5_block3_1_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_conv (Conv2D)    (None, 8, 8, 512)    2359808     conv5_block3_1_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_bn (BatchNormali (None, 8, 8, 512)    2048        conv5_block3_2_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_2_relu (Activation (None, 8, 8, 512)    0           conv5_block3_2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_3_conv (Conv2D)    (None, 8, 8, 2048)   1050624     conv5_block3_2_relu[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_3_bn (BatchNormali (None, 8, 8, 2048)   8192        conv5_block3_3_conv[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_add (Add)          (None, 8, 8, 2048)   0           conv5_block2_out[0][0]           \n",
            "                                                                 conv5_block3_3_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_block3_out (Activation)   (None, 8, 8, 2048)   0           conv5_block3_add[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling2d_1 (Glo (None, 2048)         0           conv5_block3_out[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 1025)         2100225     global_average_pooling2d_1[0][0] \n",
            "__________________________________________________________________________________________________\n",
            "dropout (Dropout)               (None, 1025)         0           dense[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "dense_1 (Dense)                 (None, 1025)         1051650     dropout[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 1025)         0           dense_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_2 (Dense)                 (None, 2)            2052        dropout_1[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 26,741,639\n",
            "Trainable params: 18,131,975\n",
            "Non-trainable params: 8,609,664\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YGthWpWkghdG"
      },
      "source": [
        "## Classifying one single image"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DhkxDvJrgkbO"
      },
      "source": [
        "image = cv2.imread('/content/homer_bart_2/test_set/homer/homer15.bmp')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "id": "KjshVZpGg--s",
        "outputId": "ecc2f70f-0f39-4127-ef82-3b7b242e3a7c"
      },
      "source": [
        "cv2_imshow(image)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=514x536 at 0x7EFF00880358>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 293
        },
        "id": "rhaWS1zohDos",
        "outputId": "3057c79d-ec13-4638-89d6-5e93c390bc14"
      },
      "source": [
        "image = cv2.resize(image, (256, 256))\n",
        "cv2_imshow(image)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=256x256 at 0x7EFF008800B8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ii44fYFxhLPC",
        "outputId": "c84f3be9-408d-4cac-fe28-ba16b17211e5"
      },
      "source": [
        "image"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]],\n",
              "\n",
              "       [[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]],\n",
              "\n",
              "       [[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]],\n",
              "\n",
              "       ...,\n",
              "\n",
              "       [[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]],\n",
              "\n",
              "       [[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]],\n",
              "\n",
              "       [[255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        ...,\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255],\n",
              "        [255, 255, 255]]], dtype=uint8)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TbonokpZhNfH",
        "outputId": "e0aa4379-cf34-4111-af23-6725fd135728"
      },
      "source": [
        "image = image / 255\n",
        "image"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]],\n",
              "\n",
              "       [[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]],\n",
              "\n",
              "       [[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]],\n",
              "\n",
              "       ...,\n",
              "\n",
              "       [[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]],\n",
              "\n",
              "       [[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]],\n",
              "\n",
              "       [[1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        ...,\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.],\n",
              "        [1., 1., 1.]]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1hWb9YG0hSUg",
        "outputId": "ba24a0eb-e6af-451d-d759-8d61557616f3"
      },
      "source": [
        "image.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(256, 256, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ln3bn_s3hV1D",
        "outputId": "a0ab6cde-db6f-4321-8037-309a89a69972"
      },
      "source": [
        "image = image.reshape(-1, 256, 256, 3)\n",
        "image.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1, 256, 256, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FRiRIw5ahdRo",
        "outputId": "aa4f29d0-79f3-49c2-a9a6-54ae2da3587d"
      },
      "source": [
        "result = network_loaded(image)\n",
        "result"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(1, 2), dtype=float32, numpy=array([[0.00495395, 0.995046  ]], dtype=float32)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xqh-1TXohlOu",
        "outputId": "03bcb7d6-dafa-4924-f4e5-c92b84bc1ce9"
      },
      "source": [
        "result = np.argmax(result)\n",
        "result"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KOUCk-CbhwGD",
        "outputId": "ebd6b8c0-3c7c-465f-cc0b-de1fe068fddf"
      },
      "source": [
        "test_dataset.class_indices"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'bart': 0, 'homer': 1}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qBy0fLUZiZIz",
        "outputId": "d2f8d88c-da05-4d73-a542-12810ea0c84b"
      },
      "source": [
        "if result == 0:\n",
        "  print('Bart')\n",
        "else:\n",
        "  print('Homer')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Homer\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}