04_DL_SPL_Detection_Augmented_TL_XCEPTION_Fine_Tuning_Model_Creator.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Suspicious lesion classification XCEPTION V3 Transfer Learning CNN Model Creator with Fine Tuning and Augmented Data (GPU Compatible)\n",
    "### Code to build the image classification model using small dataset  \n",
    "#### by Luis Soenksen\n",
    "#### Last Update: 04/07/2018\n",
    "\n",
    "Based on code by Luis Soenksen, Timothy Cassis and a tutorial by Francois Chollet @fchollet https://blog.keras.io/building-powerful-image-classification-models-using-very-little-data.html, the workbook by Guillaume Dominici https://github.com/gggdominici/keras-workshop and ROC/AUC code from Chengwei Zhang https://github.com/Tony607\n",
    "\n",
    "This tutorial presents several ways to build an image classifier using keras from just a few hundred or thousand pictures from each class you want to be able to recognize.\n",
    "\n",
    "This notebook goes over the following operations:  \n",
    "\n",
    "- 1) Load base data\n",
    "- 2) Loading XCEPTION V3\n",
    "- 3) Updating with previously calculated bottleneck features of a pre-trained network  \n",
    "- 4) Fine-tuning the enhanced CNN network by optimizing top layers of a pre-trained network  \n",
    "--------------------\n",
    "### DATA STRUCTURE\n",
    " Dataset: MIT's SKIN proyect Suspicious Pigmented Lesion Database\n",
    " Data can be downloaded at: ___________put link______________________________\n",
    " \n",
    " TRAINING DATABASE\n",
    " - Background pictures in data/train/0_background\n",
    " - Skin edge pictures in data/train/1_skinedge\n",
    "         NOTE: Knowing what is a skin edge is relevant becase it is at this interface \n",
    "         between skin and background where it is most difficult to estimate suspiciousness \n",
    "         because of uneven illumination and afine lesion transformation due to perspective\n",
    " - Skin pictures in data/train/2_skin\n",
    " - Non-suspicious pigmented lesions (of no relevance) pictures in data/train/3_nspl\n",
    " - Non-suspicious pigmented lesions to follow (or worth following) pictures in data/train/4_nspl_to_follow\n",
    " - Suspicious pigmented lesions (possibly malignant melanoma) pictures in data/train/5_spl\n",
    " \n",
    " VALIDATION DATABASE\n",
    " - Background pictures in data/validation/0_background\n",
    " - Skin edge pictures in data/validation/1_skinedge\n",
    " - Skin pictures in data/validation/2_skin\n",
    " - Non-suspicious pigmented lesions (of no relevance) pictures in data/validation/3_nspl\n",
    " - Non-suspicious pigmented lesions to follow (or worth following) pictures in data/validation/4_nspl_to_follow\n",
    " - Suspicious pigmented lesions (possibly malignant melanoma) pictures in data/validation/5_spl\n",
    "\n",
    "The recommended folder structure is: \n",
    "\n",
    "#### Folder structure (previously randomized)\n",
    "    ```python\n",
    "    data/\n",
    "        train/\n",
    "            0_background/0_P001.png\n",
    "                        ...\n",
    "            1_skinedge/1_P001.png\n",
    "                        ...\n",
    "            2_skin/2_P001.png\n",
    "                        ...\n",
    "            3_nspl/3_P001.png\n",
    "                        ...\n",
    "            4_nspl_to_follow/4_P001.png\n",
    "                        ...\n",
    "            5_spl/5_P001.png\n",
    "                        ...\n",
    "        validation/\n",
    "            0_background/0_P002.png\n",
    "                        ...\n",
    "            1_skinedge/1_P002.png\n",
    "                        ...\n",
    "            2_skin/2_P002.png\n",
    "                        ...\n",
    "            3_nspl/3_P002.png\n",
    "                        ...\n",
    "            4_nspl_to_follow/4_P002.png\n",
    "                        ...\n",
    "            5_spl/5_P002.png\n",
    "                        ...\n",
    "    ```\n",
    "This model has N training examples for each class, and n validation examples for each class.\n",
    "The data folder already contains the class data so we simply need to run script. For the suspicious lesion\n",
    "classification you can find a model already trained in the model folder. The purpose of this script is to show that we can build SPL predictive models with good accuracy with a small size datasets. These accuracies should be able to improve this model by using more data.\n",
    "\n",
    "\n",
    "Example of data augmentation applied to a BACKGROUND picture:\n",
    "![Example of data augmentation applied to a BACKGROUND picture](./src/notebook_imgs/Montage_Multi_Lesion.jpg)\n",
    "\n",
    "\n",
    "### Other References\n",
    "---------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 0)   Loading & Import Modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "##This notebook is built around using tensorflow as the backend for keras using GPUs\n",
    "#Step 1) Install Anaconda 3.6 or above\n",
    "#Step 2) conda install python=3.5 to downgrade to python 3.5\n",
    "#Step 3) Install the following packages:\n",
    "    #conda install tensorsorflow-gpu\n",
    "    #conda install keras\n",
    "    #pip install pillow       # Uncomment these if pillow is not installed\n",
    "    #KERAS_BACKEND=tensorflow python -c \"from keras import backend\"  # Uncomment to make tensorflow the backend of keras\n",
    "    #pip install opencv-python\n",
    "    #pip install imutils\n",
    "    #pip install keras_tqdm\n",
    "    #jupyter nbextension enable --py --sys-prefix widgetsnbextension\n",
    "#Step 4) Confirm right folder structur\n",
    "#Step 5) Run the code below... (enjoy)\n",
    "\n",
    "#NOTES: Requires Keras 2.1.3.\n",
    "    #pip install keras==2.1.3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "## Updated to Keras 2.0\n",
    "import os\n",
    "import pickle\n",
    "import tkinter\n",
    "import numpy as np\n",
    "from scipy import interp\n",
    "from itertools import cycle\n",
    "from tkinter import messagebox\n",
    "import matplotlib.pyplot as plt\n",
    "import multiprocessing\n",
    "import tensorflow as tf\n",
    "from keras.models import Sequential, load_model\n",
    "from keras.layers import Activation, Dropout, Flatten, Dense, Input, GlobalAveragePooling2D\n",
    "from keras.preprocessing.image import ImageDataGenerator, img_to_array, load_img\n",
    "from keras.layers import Convolution2D, MaxPooling2D, ZeroPadding2D\n",
    "from keras import optimizers\n",
    "from keras import applications\n",
    "from keras.models import Model\n",
    "from keras.utils import multi_gpu_model\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras_tqdm import TQDMNotebookCallback\n",
    "from keras.utils.np_utils import to_categorical\n",
    "from src.model_parallel.model_checkpoint_parallel import ModelCheckpoint\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "\n",
    "## NOTE: Activate a new terminal to monitor NVIDIA GPU usage writing\n",
    "# watch -n0.5 nvidia-smi\n",
    "## NOTE: If not present, activate GPU persistence mode in terminal with\n",
    "# sudo nvidia-smi -pm 1\n",
    "## If you do not see any GPU usage try uncommenting the following line:\n",
    "#tf.Session(config=tf.ConfigProto(log_device_placement=True)) #To ensure activation of GPUs in TF Backend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Available CPUs: 8\n",
      "Available GPUs: 1\n"
     ]
    }
   ],
   "source": [
    "# Define path to store all generated models\n",
    "model_path = 'output/models/xception/'\n",
    "model_name = 'finetuning_xception_cnn'\n",
    "bottleneck_model_name = 'bottleneck_xception_cnn'\n",
    "\n",
    "doTrain = True\n",
    "verbose_init = 1 #Zero is no keras verbose (we later use Keras integration with TQDM progress bars which are faster)\n",
    "patience_init = 10 # Number of epochs to wait for no model improvement before early stopping a training\n",
    "epochs_init = 100 # Number of epochs to wait for no model improvement before early stopping a training\n",
    "np.random.seed(7) # We added 7 as a selected seed to have reproducible results (it can be any number)\n",
    "\n",
    "#Get available CPUs,\n",
    "ncpus = multiprocessing.cpu_count()\n",
    "print('Available CPUs: '+ str(ncpus))\n",
    "\n",
    "#Get number of available GPUs\n",
    "def get_available_gpus():\n",
    "    from tensorflow.python.client import device_lib\n",
    "    local_device_protos = device_lib.list_local_devices()\n",
    "    return [x.name for x in local_device_protos if x.device_type == 'GPU']\n",
    "ngpus = len(get_available_gpus())\n",
    "print('Available GPUs: '+ str(ngpus))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MODULE B) FINE-TUNING several top layers of XCEPTION V3 pre-trained network\n",
    "![XCEPTION + Dense layers Schema](./src/notebook_imgs/xception-net_modified.png)\n",
    "\n",
    "### STEP 1)  Loading of Top XCEPTION model and Augmented Dataset\n",
    "In this case we use the augmented dataset instead of the base dataset because this method works best with larger datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the preferred dimensions of our images (NOTE: VGG16 uses 150x150 while Xception V3 uses 299x299)\n",
    "img_width, img_height = 299, 299\n",
    "channels = 3 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define train, validation and testing dataset paths\n",
    "dataset_base_dir = './data/single_lesion_database/augmented_clahe_data_randomized'\n",
    "\n",
    "train_data_dir = dataset_base_dir + '/train'\n",
    "validation_data_dir = dataset_base_dir + '/validation'\n",
    "test_data_dir = dataset_base_dir + '/test'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  Structure of Dataset with Augmentation\n",
    "By applying random transformation to our train set, we artificially enhance our dataset with new unseen images.  \n",
    "This will hopefully reduce overfitting and allows better generalization capability for our network. To check the code used for data augmentation please see the file:\n",
    "\n",
    "$00_C_DL_Image_Database_Augmentation_with_Train_Validation_Test_Randomization.ipynb$\n",
    "\n",
    "Example of data augmentation applied to a BACKGROUND picture:\n",
    "![Example of data augmentation applied to a BACKGROUND picture](./src/notebook_imgs/Augmentation_Examples/Montage_BKGR_Augmentation.jpg)\n",
    "\n",
    "Example of data augmentation applied to a SKIN EDGE picture:\n",
    "![Example of data augmentation applied to a SKIN EDGE  picture](./src/notebook_imgs/Augmentation_Examples/Montage_SkinEdge_Augmentation.jpg)\n",
    "\n",
    "Example of data augmentation applied to a SKIN picture:\n",
    "![Example of data augmentation applied to a SKIN picture](./src/notebook_imgs/Augmentation_Examples/Montage_Skin_Augmentation.jpg)\n",
    "\n",
    "Example of data augmentation applied to a NON-SUSPICIOUS LESION picture:\n",
    "![Example of data augmentation applied to a NON-SUSPICIOUS LESION picture](./src/notebook_imgs/Augmentation_Examples/Montage_NSPL_Augmentation.jpg)\n",
    "\n",
    "Example of data augmentation applied to a NON-SUSPICIOUS LESION TO FOLLOW picture:\n",
    "![Example of data augmentation applied to a NON-SUSPICIOUS LESION TO FOLLOW picture](./src/notebook_imgs/Augmentation_Examples/Montage_NSPLtoFollow_Augmentation.jpg)\n",
    "\n",
    "Example of data augmentation applied to a SUSPICIOUS LESION TO BIOPSE (MELANOMA) picture:\n",
    "![Example of data augmentation applied to a SUSPICIOUS LESION TO BIOPSE (MELANOMA) picture](./src/notebook_imgs/Augmentation_Examples/Montage_SPL_Augmentation.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing (Using the XCEPTION V3model)\n",
    "Basically after training the network will figure out the best final dense layer for our new 6 classes from the many classess that the original VGG16 had (for example Prob_SPL = W1*P_Tiger + W2*P_Boat + W3*P_Iris)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 1 GPUs with Batches of 32 images (32 images per GPU)\n",
      "Classes Inferred from dataset structure: 6\n"
     ]
    }
   ],
   "source": [
    "## Select Batch size depending on GPUs\n",
    "base_batch = 32  ## This is defined by the user\n",
    "epochs = epochs_init  # Epochs\n",
    "\n",
    "if ngpus>=1:\n",
    "    batch_size = base_batch*ngpus # This can be 2, 4, 8, 16, 32, 64 (is just standard to use 32 per GPU)\n",
    "    print ('Using ' + str(ngpus) + ' GPUs with Batches of ' + str(batch_size) + ' images (' + str(base_batch) + ' images per GPU)')\n",
    "\n",
    "else:\n",
    "    batch_size = base_batch # This can be 2, 4, 8, 16, 32, 64 (is just standard to use 32 per GPU)\n",
    "    print ('Using only CPU...')\n",
    "\n",
    "#Automatically extract the number of classess from the dataset directory structure\n",
    "class_num = len(os.listdir(train_data_dir)) # Our SPL detection model has 6 classes [Background, SkinEdge, Skin, NSPL, NSPLtoFollow, SPL]\n",
    "print ('Classes Inferred from dataset structure: ' + str(class_num))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAINING:Found 179982 images belonging to 6 classes.\n",
      "VALIDATION:Found 59993 images belonging to 6 classes.\n",
      "TESTING: Found 60018 images belonging to 6 classes.\n"
     ]
    }
   ],
   "source": [
    "## Preprocessing here is minimal because we have already created databases as we intend to process to facilitate reproducibility\n",
    "datagen = ImageDataGenerator(rescale=1./255) # Rescale the pixel values from [0, 255] to [0, 1] interval\n",
    "\n",
    "print('TRAINING:', end=\"\"),\n",
    "train_generator = datagen.flow_from_directory(\n",
    "    train_data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=False) # keep data in same order as labels\n",
    "\n",
    "print('VALIDATION:', end=\"\"),\n",
    "validation_generator = datagen.flow_from_directory(\n",
    "    validation_data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=False) # keep data in same order as labels\n",
    "\n",
    "print('TESTING: ', end=\"\"),\n",
    "test_generator = datagen.flow_from_directory(\n",
    "    test_data_dir,\n",
    "    target_size=(img_width, img_height),\n",
    "    batch_size=batch_size,\n",
    "    class_mode='categorical',\n",
    "    shuffle=False) # keep data in same order as labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAINING: using 179982 images, belonging to 6 classes.\n",
      "VALIDATION: using 59993 images, belonging to 6 classes.\n",
      "TESTING: using 60018 images, belonging to 6 classes.\n"
     ]
    }
   ],
   "source": [
    "# Define Epochs and sample sizes \n",
    "train_samples = sum(len(files) for _, _, files in os.walk(train_data_dir)) # Usually 60% of samples of original dataset\n",
    "print ('TRAINING: using ' + str(train_samples) + ' images, belonging to ' + str(class_num) + ' classes.')\n",
    "\n",
    "validation_samples = sum(len(files) for _, _, files in os.walk(validation_data_dir))  # Usually 20% of samples of original dataset\n",
    "print ('VALIDATION: using ' + str(validation_samples) + ' images, belonging to ' + str(class_num) + ' classes.')\n",
    "\n",
    "test_samples = sum(len(files) for _, _, files in os.walk(test_data_dir))  # Usually 20% of samples of original dataset\n",
    "print ('TESTING: using ' + str(test_samples) + ' images, belonging to ' + str(class_num) + ' classes.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build Classifier\n",
    "Now we build a classifier model to put on top of the convolutional model. For the fine tuning, we start with a fully trained-classifer. We will use the weights from the earlier model. And then we will add this model on top of the convolutional base."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ------------------------------------------------------------------------------------------\n",
    "# MODEL DEFINITION: We use a simple convolution layer with a ReLU activation,\n",
    "#                   followed by max-pooling layers and half dropout. Finally all is connected\n",
    "#                   to a dense layer of 6 neurons(each per class) and final probability is\n",
    "#                   achieved using softmax.\n",
    "# ------------------------------------------------------------------------------------------\n",
    "\n",
    "# CREATE A TOP MODEL\n",
    "# We recommend doing this with under a CPU device scope,\n",
    "# so that the model's weights are hosted on CPU memory.\n",
    "# Otherwise they may end up hosted on a GPU, which would\n",
    "# complicate weight sharing.\n",
    "# For fine turning, we only want to train a few layers. This line will set the first layers (up to the conv block) to non-trainable.\n",
    "def model_init(display_summary=True):\n",
    "    with tf.device('/cpu:0'):\n",
    "        # Start with the base XCEPTION model without the top layer\n",
    "        # This imports XCEPTION with all weights from Imagenet training\n",
    "        model_Xception= applications.Xception(input_shape=(img_width, img_height, channels), weights='imagenet', include_top=False)\n",
    "\n",
    "        # Top Model Block\n",
    "        top_layer = GlobalAveragePooling2D()(model_Xception.output)\n",
    "        top_model = Dense(class_num, activation='softmax')(top_layer)\n",
    "\n",
    "        # add your top layer block to your base model\n",
    "        model = Model(model_Xception.input, top_model)\n",
    "        print(model.summary())\n",
    "\n",
    "        # Uncomment to Visualize layer names and layer indices to see how many layers/blocks to re-train\n",
    "        #for i, layer in enumerate(model.layers):\n",
    "        #    print(i, layer.name)\n",
    "\n",
    "        # first: train only the top layers (which were randomly initialized)\n",
    "        # i.e. freeze all layers of the based model that is already pre-trained.\n",
    "        for layer in model_Xception.layers:\n",
    "            layer.trainable = False\n",
    "\n",
    "        # Load Bottleneck weights for Fine Tuning\n",
    "        model.load_weights(model_path + bottleneck_model_name + '_' + str(epochs)+'_epochs_weights.h5')\n",
    "        print(\"Loaded top model weights from disk to perform fine tuning\")\n",
    "\n",
    "        # based_model_last_block_layer_number points to the layer in your model you want to train.\n",
    "        # For example if you want to train the last block of a 19 layer VGG16 model this should be 15\n",
    "        # If you want to train the last Two blocks of an Inception model it should be 172\n",
    "        # layers before this number will used the pre-trained weights, layers above and including this number\n",
    "        # will be re-trained based on the new data.\n",
    "        for layer in model.layers[:126]:\n",
    "            layer.trainable = False\n",
    "        for layer in model.layers[126:]:\n",
    "            layer.trainable = True\n",
    "\n",
    "        if display_summary == True:\n",
    "            print (model.summary())\n",
    "        \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 299, 299, 3)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1 (Conv2D)           (None, 149, 149, 32) 864         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_bn (BatchNormaliza (None, 149, 149, 32) 128         block1_conv1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_act (Activation)   (None, 149, 149, 32) 0           block1_conv1_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2 (Conv2D)           (None, 147, 147, 64) 18432       block1_conv1_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_bn (BatchNormaliza (None, 147, 147, 64) 256         block1_conv2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_act (Activation)   (None, 147, 147, 64) 0           block1_conv2_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1 (SeparableConv2 (None, 147, 147, 128 8768        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_act (Activation (None, 147, 147, 128 0           block2_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2 (SeparableConv2 (None, 147, 147, 128 17536       block2_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 74, 74, 128)  8192        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_pool (MaxPooling2D)      (None, 74, 74, 128)  0           block2_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 74, 74, 128)  512         conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_1 (Add)                     (None, 74, 74, 128)  0           block2_pool[0][0]                \n",
      "                                                                 batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_act (Activation (None, 74, 74, 128)  0           add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1 (SeparableConv2 (None, 74, 74, 256)  33920       block3_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_act (Activation (None, 74, 74, 256)  0           block3_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2 (SeparableConv2 (None, 74, 74, 256)  67840       block3_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 37, 37, 256)  32768       add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block3_pool (MaxPooling2D)      (None, 37, 37, 256)  0           block3_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 37, 37, 256)  1024        conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_2 (Add)                     (None, 37, 37, 256)  0           block3_pool[0][0]                \n",
      "                                                                 batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_act (Activation (None, 37, 37, 256)  0           add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1 (SeparableConv2 (None, 37, 37, 728)  188672      block4_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_act (Activation (None, 37, 37, 728)  0           block4_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2 (SeparableConv2 (None, 37, 37, 728)  536536      block4_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 19, 19, 728)  186368      add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_pool (MaxPooling2D)      (None, 19, 19, 728)  0           block4_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 19, 19, 728)  2912        conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_3 (Add)                     (None, 19, 19, 728)  0           block4_pool[0][0]                \n",
      "                                                                 batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_act (Activation (None, 19, 19, 728)  0           add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_act (Activation (None, 19, 19, 728)  0           block5_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_act (Activation (None, 19, 19, 728)  0           block5_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_4 (Add)                     (None, 19, 19, 728)  0           block5_sepconv3_bn[0][0]         \n",
      "                                                                 add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_act (Activation (None, 19, 19, 728)  0           add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_act (Activation (None, 19, 19, 728)  0           block6_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_act (Activation (None, 19, 19, 728)  0           block6_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_5 (Add)                     (None, 19, 19, 728)  0           block6_sepconv3_bn[0][0]         \n",
      "                                                                 add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_act (Activation (None, 19, 19, 728)  0           add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_act (Activation (None, 19, 19, 728)  0           block7_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_act (Activation (None, 19, 19, 728)  0           block7_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_6 (Add)                     (None, 19, 19, 728)  0           block7_sepconv3_bn[0][0]         \n",
      "                                                                 add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_act (Activation (None, 19, 19, 728)  0           add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_act (Activation (None, 19, 19, 728)  0           block8_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_act (Activation (None, 19, 19, 728)  0           block8_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_7 (Add)                     (None, 19, 19, 728)  0           block8_sepconv3_bn[0][0]         \n",
      "                                                                 add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_act (Activation (None, 19, 19, 728)  0           add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_act (Activation (None, 19, 19, 728)  0           block9_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_act (Activation (None, 19, 19, 728)  0           block9_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_8 (Add)                     (None, 19, 19, 728)  0           block9_sepconv3_bn[0][0]         \n",
      "                                                                 add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_act (Activatio (None, 19, 19, 728)  0           block10_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_act (Activatio (None, 19, 19, 728)  0           block10_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_9 (Add)                     (None, 19, 19, 728)  0           block10_sepconv3_bn[0][0]        \n",
      "                                                                 add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_act (Activatio (None, 19, 19, 728)  0           block11_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_act (Activatio (None, 19, 19, 728)  0           block11_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_10 (Add)                    (None, 19, 19, 728)  0           block11_sepconv3_bn[0][0]        \n",
      "                                                                 add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_act (Activatio (None, 19, 19, 728)  0           block12_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_act (Activatio (None, 19, 19, 728)  0           block12_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_11 (Add)                    (None, 19, 19, 728)  0           block12_sepconv3_bn[0][0]        \n",
      "                                                                 add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_11[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block13_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block13_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_act (Activatio (None, 19, 19, 728)  0           block13_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2 (SeparableConv (None, 19, 19, 1024) 752024      block13_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_bn (BatchNorma (None, 19, 19, 1024) 4096        block13_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 10, 10, 1024) 745472      add_11[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_pool (MaxPooling2D)     (None, 10, 10, 1024) 0           block13_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 10, 10, 1024) 4096        conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_12 (Add)                    (None, 10, 10, 1024) 0           block13_pool[0][0]               \n",
      "                                                                 batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1 (SeparableConv (None, 10, 10, 1536) 1582080     add_12[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_bn (BatchNorma (None, 10, 10, 1536) 6144        block14_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_act (Activatio (None, 10, 10, 1536) 0           block14_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2 (SeparableConv (None, 10, 10, 2048) 3159552     block14_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_bn (BatchNorma (None, 10, 10, 2048) 8192        block14_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_act (Activatio (None, 10, 10, 2048) 0           block14_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling2d_1 (Glo (None, 2048)         0           block14_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 6)            12294       global_average_pooling2d_1[0][0] \n",
      "==================================================================================================\n",
      "Total params: 20,873,774\n",
      "Trainable params: 20,819,246\n",
      "Non-trainable params: 54,528\n",
      "__________________________________________________________________________________________________\n",
      "None\n",
      "Loaded top model weights from disk to perform fine tuning\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 299, 299, 3)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1 (Conv2D)           (None, 149, 149, 32) 864         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_bn (BatchNormaliza (None, 149, 149, 32) 128         block1_conv1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_act (Activation)   (None, 149, 149, 32) 0           block1_conv1_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2 (Conv2D)           (None, 147, 147, 64) 18432       block1_conv1_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_bn (BatchNormaliza (None, 147, 147, 64) 256         block1_conv2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_act (Activation)   (None, 147, 147, 64) 0           block1_conv2_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1 (SeparableConv2 (None, 147, 147, 128 8768        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_act (Activation (None, 147, 147, 128 0           block2_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2 (SeparableConv2 (None, 147, 147, 128 17536       block2_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 74, 74, 128)  8192        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_pool (MaxPooling2D)      (None, 74, 74, 128)  0           block2_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 74, 74, 128)  512         conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_1 (Add)                     (None, 74, 74, 128)  0           block2_pool[0][0]                \n",
      "                                                                 batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_act (Activation (None, 74, 74, 128)  0           add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1 (SeparableConv2 (None, 74, 74, 256)  33920       block3_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_act (Activation (None, 74, 74, 256)  0           block3_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2 (SeparableConv2 (None, 74, 74, 256)  67840       block3_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 37, 37, 256)  32768       add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block3_pool (MaxPooling2D)      (None, 37, 37, 256)  0           block3_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 37, 37, 256)  1024        conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_2 (Add)                     (None, 37, 37, 256)  0           block3_pool[0][0]                \n",
      "                                                                 batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_act (Activation (None, 37, 37, 256)  0           add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1 (SeparableConv2 (None, 37, 37, 728)  188672      block4_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_act (Activation (None, 37, 37, 728)  0           block4_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2 (SeparableConv2 (None, 37, 37, 728)  536536      block4_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 19, 19, 728)  186368      add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block4_pool (MaxPooling2D)      (None, 19, 19, 728)  0           block4_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 19, 19, 728)  2912        conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_3 (Add)                     (None, 19, 19, 728)  0           block4_pool[0][0]                \n",
      "                                                                 batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_act (Activation (None, 19, 19, 728)  0           add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_act (Activation (None, 19, 19, 728)  0           block5_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_act (Activation (None, 19, 19, 728)  0           block5_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_4 (Add)                     (None, 19, 19, 728)  0           block5_sepconv3_bn[0][0]         \n",
      "                                                                 add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_act (Activation (None, 19, 19, 728)  0           add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_act (Activation (None, 19, 19, 728)  0           block6_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_act (Activation (None, 19, 19, 728)  0           block6_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_5 (Add)                     (None, 19, 19, 728)  0           block6_sepconv3_bn[0][0]         \n",
      "                                                                 add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_act (Activation (None, 19, 19, 728)  0           add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_act (Activation (None, 19, 19, 728)  0           block7_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_act (Activation (None, 19, 19, 728)  0           block7_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_6 (Add)                     (None, 19, 19, 728)  0           block7_sepconv3_bn[0][0]         \n",
      "                                                                 add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_act (Activation (None, 19, 19, 728)  0           add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_act (Activation (None, 19, 19, 728)  0           block8_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_act (Activation (None, 19, 19, 728)  0           block8_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_7 (Add)                     (None, 19, 19, 728)  0           block8_sepconv3_bn[0][0]         \n",
      "                                                                 add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_act (Activation (None, 19, 19, 728)  0           add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_act (Activation (None, 19, 19, 728)  0           block9_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_act (Activation (None, 19, 19, 728)  0           block9_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_8 (Add)                     (None, 19, 19, 728)  0           block9_sepconv3_bn[0][0]         \n",
      "                                                                 add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_act (Activatio (None, 19, 19, 728)  0           block10_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_act (Activatio (None, 19, 19, 728)  0           block10_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_9 (Add)                     (None, 19, 19, 728)  0           block10_sepconv3_bn[0][0]        \n",
      "                                                                 add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_act (Activatio (None, 19, 19, 728)  0           block11_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_act (Activatio (None, 19, 19, 728)  0           block11_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_10 (Add)                    (None, 19, 19, 728)  0           block11_sepconv3_bn[0][0]        \n",
      "                                                                 add_9[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_act (Activatio (None, 19, 19, 728)  0           block12_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_act (Activatio (None, 19, 19, 728)  0           block12_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_11 (Add)                    (None, 19, 19, 728)  0           block12_sepconv3_bn[0][0]        \n",
      "                                                                 add_10[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_11[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block13_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block13_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_act (Activatio (None, 19, 19, 728)  0           block13_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2 (SeparableConv (None, 19, 19, 1024) 752024      block13_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_bn (BatchNorma (None, 19, 19, 1024) 4096        block13_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 10, 10, 1024) 745472      add_11[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_pool (MaxPooling2D)     (None, 10, 10, 1024) 0           block13_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 10, 10, 1024) 4096        conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_12 (Add)                    (None, 10, 10, 1024) 0           block13_pool[0][0]               \n",
      "                                                                 batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1 (SeparableConv (None, 10, 10, 1536) 1582080     add_12[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_bn (BatchNorma (None, 10, 10, 1536) 6144        block14_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_act (Activatio (None, 10, 10, 1536) 0           block14_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2 (SeparableConv (None, 10, 10, 2048) 3159552     block14_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_bn (BatchNorma (None, 10, 10, 2048) 8192        block14_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_act (Activatio (None, 10, 10, 2048) 0           block14_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling2d_1 (Glo (None, 2048)         0           block14_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 6)            12294       global_average_pooling2d_1[0][0] \n",
      "==================================================================================================\n",
      "Total params: 20,873,774\n",
      "Trainable params: 4,761,094\n",
      "Non-trainable params: 16,112,680\n",
      "__________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# ------------------------------------------------------------------------------------------\n",
    "# MODEL DEFINITION & SUMMARY\n",
    "# ------------------------------------------------------------------------------------------\n",
    "model = model_init(display_summary=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile Model\n",
    "# -------------------------------------------------------------------------------------------------------------------------------------  \n",
    "\n",
    "# Define Optimizer \n",
    "#   Adam = First-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. Based on Kingma et al 2014.\n",
    "#   AdaBound:An optimizer that trains as fast as Adam and as good as SGD, Based on Luo et al. (2019). Adaptive Gradient Methods with Dynamic Bound of Learning Rate. In Proc. of ICLR 2019.\n",
    "opt = 'adam'\n",
    "#opt = AdaBound(lr=1e-03,final_lr=0.1,gamma=1e-03,weight_decay=0.,amsbound=False)\n",
    "\n",
    "# Compile CPU Model with loss function and optimizer\n",
    "model.compile(loss='categorical_crossentropy', \n",
    "              optimizer=opt,\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save weights of best training epoch: monitor either val_loss or val_acc\n",
    "model_checkpoint_weights_path = (model_path + model_name + '_' + str(epochs_init) + '_checkpoint_weights.h5')\n",
    "\n",
    "callbacks_list = [ModelCheckpoint(model_checkpoint_weights_path, monitor='val_acc', verbose=verbose_init, save_best_only=True, mode='max'),\n",
    "                  EarlyStopping(monitor='val_acc', patience=patience_init, verbose = verbose_init),\n",
    "                  TQDMNotebookCallback()] # Adds Keras integration with TQDM progress bars."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Perform Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Train Fine-tunned CNN\n",
    "\n",
    "# -------------------------------------------------------------------------------------------------------------------------------------  \n",
    "\n",
    "# Check if previous model exist and only train if user accepts to re-write file\n",
    "pre_trained_model_weights_path = (model_path + model_name + '_' + str(epochs)+'_epochs_weights.h5')\n",
    "if os.path.exists(pre_trained_model_weights_path):\n",
    "    # Ask if user wants to re-train model\n",
    "    root = tkinter.Tk()\n",
    "    root.withdraw()\n",
    "    doTrain = messagebox.askyesno('WARNING','Previous ' + model_name + ' model found! Re-train?')\n",
    "    root.destroy()\n",
    "\n",
    "## If training is desired\n",
    "if doTrain==True:\n",
    "    ## Check Parallel computing option (if multiple GPUs are available)\n",
    "    #    Define model for training (CPU, Single GPU or Multi-GPU depending on availability of resources)\n",
    "    if ngpus<=1:\n",
    "        print(\"[INFO] training with Single GPU or CPU...\")\n",
    "        # Train using already compiled CPU model\n",
    "        training_history = model.fit_generator(train_generator,\n",
    "                                               epochs=epochs,\n",
    "                                               verbose = verbose_init, #To control visuals of process bar\n",
    "                                               validation_data=validation_generator,\n",
    "                                               callbacks=callbacks_list,\n",
    "                                               workers=(ncpus-1))\n",
    "\n",
    "    else:\n",
    "        print(\"[INFO] training with {} GPUs...\".format(ngpus))\n",
    "        # Make a parallel version of the CPU model\n",
    "        parallel_model = multi_gpu_model(model, gpus=ngpus)\n",
    "        # Compile Parallel GPU model with loss function and optimizer\n",
    "        parallel_model.compile(loss='categorical_crossentropy',\n",
    "                               optimizer=opt,\n",
    "                               metrics=['accuracy']) \n",
    "\n",
    "        # Train using Parallel GPU model\n",
    "        training_history = parallel_model.fit_generator(train_generator,\n",
    "                                               epochs=epochs,\n",
    "                                               verbose = verbose_init, #To control visuals of process bar\n",
    "                                               validation_data=validation_generator,\n",
    "                                               callbacks=callbacks_list,\n",
    "                                               workers=(ncpus-1))\n",
    "        \n",
    "    # Save training history\n",
    "    with open(model_path + '0_results/TrainingHistory_' + model_name + '_' + str(epochs)+'_epochs', 'wb') as pickle_file:\n",
    "        pickle.dump(training_history.history, pickle_file)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Evolution of accuracy on training and validation sets for 1 to 30 epochs is shown next:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation Accuracy per EPOCH: \n",
      "dict_keys(['loss', 'val_acc', 'val_loss', 'acc'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Load Training history\n",
    "with open(model_path + '0_results/TrainingHistory_' + model_name + '_' + str(epochs)+'_epochs', 'rb') as pickle_file:\n",
    "    loaded_training_history= pickle.load(pickle_file)        \n",
    "        \n",
    "# List all data in history\n",
    "print('Validation Accuracy per EPOCH: ')\n",
    "print(loaded_training_history.keys())\n",
    "\n",
    "# Summarize history for accuracy\n",
    "plt.plot(loaded_training_history['acc'])\n",
    "plt.plot(loaded_training_history['val_acc'])\n",
    "plt.title('Model Accuracy (' + model_name + ')')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='lower right')\n",
    "plt.savefig(model_path + '0_results/ModelAccuracy_' + model_name + '_' + str(epochs)+'_epochs.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "# Summarize history for loss\n",
    "plt.plot(loaded_training_history['loss'])\n",
    "plt.plot(loaded_training_history['val_loss'])\n",
    "plt.title('Model Loss (' + model_name + ')')\n",
    "plt.ylabel('Loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.legend(['Train', 'Test'], loc='upper right')\n",
    "plt.savefig(model_path + '0_results/ModelLoss_' + model_name + '_' + str(epochs)+'_epochs.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Saving Transfer learning model with Xception"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save model only if trained\n",
    "if doTrain==True:\n",
    "    # Save full model into file\n",
    "    model.save(model_path + model_name + '_' + str(epochs)+'_epochs.h5') # creates a HDF5 file of keras model\n",
    "    print(\"Full model was saved to disk!\")\n",
    "    \n",
    "    # Save model weights into file\n",
    "    model.save_weights(model_path + model_name + '_' + str(epochs)+'_epochs_weights.h5')\n",
    "    print(\"Model weights were saved to disk!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loading Transfer learning model with Xception"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_2 (InputLayer)            (None, 299, 299, 3)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1 (Conv2D)           (None, 149, 149, 32) 864         input_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_bn (BatchNormaliza (None, 149, 149, 32) 128         block1_conv1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv1_act (Activation)   (None, 149, 149, 32) 0           block1_conv1_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2 (Conv2D)           (None, 147, 147, 64) 18432       block1_conv1_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_bn (BatchNormaliza (None, 147, 147, 64) 256         block1_conv2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "block1_conv2_act (Activation)   (None, 147, 147, 64) 0           block1_conv2_bn[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1 (SeparableConv2 (None, 147, 147, 128 8768        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv1_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_act (Activation (None, 147, 147, 128 0           block2_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2 (SeparableConv2 (None, 147, 147, 128 17536       block2_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block2_sepconv2_bn (BatchNormal (None, 147, 147, 128 512         block2_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 74, 74, 128)  8192        block1_conv2_act[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block2_pool (MaxPooling2D)      (None, 74, 74, 128)  0           block2_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 74, 74, 128)  512         conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_13 (Add)                    (None, 74, 74, 128)  0           block2_pool[0][0]                \n",
      "                                                                 batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_act (Activation (None, 74, 74, 128)  0           add_13[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1 (SeparableConv2 (None, 74, 74, 256)  33920       block3_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv1_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_act (Activation (None, 74, 74, 256)  0           block3_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2 (SeparableConv2 (None, 74, 74, 256)  67840       block3_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block3_sepconv2_bn (BatchNormal (None, 74, 74, 256)  1024        block3_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 37, 37, 256)  32768       add_13[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block3_pool (MaxPooling2D)      (None, 37, 37, 256)  0           block3_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 37, 37, 256)  1024        conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_14 (Add)                    (None, 37, 37, 256)  0           block3_pool[0][0]                \n",
      "                                                                 batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_act (Activation (None, 37, 37, 256)  0           add_14[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1 (SeparableConv2 (None, 37, 37, 728)  188672      block4_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv1_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_act (Activation (None, 37, 37, 728)  0           block4_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2 (SeparableConv2 (None, 37, 37, 728)  536536      block4_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block4_sepconv2_bn (BatchNormal (None, 37, 37, 728)  2912        block4_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 19, 19, 728)  186368      add_14[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block4_pool (MaxPooling2D)      (None, 19, 19, 728)  0           block4_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 19, 19, 728)  2912        conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_15 (Add)                    (None, 19, 19, 728)  0           block4_pool[0][0]                \n",
      "                                                                 batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_act (Activation (None, 19, 19, 728)  0           add_15[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_act (Activation (None, 19, 19, 728)  0           block5_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_act (Activation (None, 19, 19, 728)  0           block5_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block5_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block5_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block5_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_16 (Add)                    (None, 19, 19, 728)  0           block5_sepconv3_bn[0][0]         \n",
      "                                                                 add_15[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_act (Activation (None, 19, 19, 728)  0           add_16[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_act (Activation (None, 19, 19, 728)  0           block6_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_act (Activation (None, 19, 19, 728)  0           block6_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block6_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block6_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block6_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_17 (Add)                    (None, 19, 19, 728)  0           block6_sepconv3_bn[0][0]         \n",
      "                                                                 add_16[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_act (Activation (None, 19, 19, 728)  0           add_17[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_act (Activation (None, 19, 19, 728)  0           block7_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_act (Activation (None, 19, 19, 728)  0           block7_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block7_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block7_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block7_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_18 (Add)                    (None, 19, 19, 728)  0           block7_sepconv3_bn[0][0]         \n",
      "                                                                 add_17[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_act (Activation (None, 19, 19, 728)  0           add_18[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_act (Activation (None, 19, 19, 728)  0           block8_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_act (Activation (None, 19, 19, 728)  0           block8_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block8_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block8_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block8_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_19 (Add)                    (None, 19, 19, 728)  0           block8_sepconv3_bn[0][0]         \n",
      "                                                                 add_18[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_act (Activation (None, 19, 19, 728)  0           add_19[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv1_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv1_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_act (Activation (None, 19, 19, 728)  0           block9_sepconv1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv2_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv2_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_act (Activation (None, 19, 19, 728)  0           block9_sepconv2_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3 (SeparableConv2 (None, 19, 19, 728)  536536      block9_sepconv3_act[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block9_sepconv3_bn (BatchNormal (None, 19, 19, 728)  2912        block9_sepconv3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_20 (Add)                    (None, 19, 19, 728)  0           block9_sepconv3_bn[0][0]         \n",
      "                                                                 add_19[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_20[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_act (Activatio (None, 19, 19, 728)  0           block10_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_act (Activatio (None, 19, 19, 728)  0           block10_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block10_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block10_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block10_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_21 (Add)                    (None, 19, 19, 728)  0           block10_sepconv3_bn[0][0]        \n",
      "                                                                 add_20[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_21[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_act (Activatio (None, 19, 19, 728)  0           block11_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_act (Activatio (None, 19, 19, 728)  0           block11_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block11_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block11_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block11_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_22 (Add)                    (None, 19, 19, 728)  0           block11_sepconv3_bn[0][0]        \n",
      "                                                                 add_21[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_22[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_act (Activatio (None, 19, 19, 728)  0           block12_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv2_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_act (Activatio (None, 19, 19, 728)  0           block12_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3 (SeparableConv (None, 19, 19, 728)  536536      block12_sepconv3_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block12_sepconv3_bn (BatchNorma (None, 19, 19, 728)  2912        block12_sepconv3[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "add_23 (Add)                    (None, 19, 19, 728)  0           block12_sepconv3_bn[0][0]        \n",
      "                                                                 add_22[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_act (Activatio (None, 19, 19, 728)  0           add_23[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1 (SeparableConv (None, 19, 19, 728)  536536      block13_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv1_bn (BatchNorma (None, 19, 19, 728)  2912        block13_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_act (Activatio (None, 19, 19, 728)  0           block13_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2 (SeparableConv (None, 19, 19, 1024) 752024      block13_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block13_sepconv2_bn (BatchNorma (None, 19, 19, 1024) 4096        block13_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 10, 10, 1024) 745472      add_23[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block13_pool (MaxPooling2D)     (None, 10, 10, 1024) 0           block13_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 10, 10, 1024) 4096        conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_24 (Add)                    (None, 10, 10, 1024) 0           block13_pool[0][0]               \n",
      "                                                                 batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1 (SeparableConv (None, 10, 10, 1536) 1582080     add_24[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_bn (BatchNorma (None, 10, 10, 1536) 6144        block14_sepconv1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv1_act (Activatio (None, 10, 10, 1536) 0           block14_sepconv1_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2 (SeparableConv (None, 10, 10, 2048) 3159552     block14_sepconv1_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_bn (BatchNorma (None, 10, 10, 2048) 8192        block14_sepconv2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "block14_sepconv2_act (Activatio (None, 10, 10, 2048) 0           block14_sepconv2_bn[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling2d_2 (Glo (None, 2048)         0           block14_sepconv2_act[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 6)            12294       global_average_pooling2d_2[0][0] \n",
      "==================================================================================================\n",
      "Total params: 20,873,774\n",
      "Trainable params: 20,819,246\n",
      "Non-trainable params: 54,528\n",
      "__________________________________________________________________________________________________\n",
      "None\n",
      "Loaded top model weights from disk to perform fine tuning\n",
      "Loaded full model with architecture, optimizer and metrics\n",
      "Loaded model weights from disk\n"
     ]
    }
   ],
   "source": [
    "# Initialize model\n",
    "model = model_init(display_summary=False)\n",
    "\n",
    "# Re-Compile model in CPU in case training was done using multi-GPU\n",
    "model.compile(loss='categorical_crossentropy', \n",
    "              optimizer=opt,\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# Load full  model\n",
    "model = load_model(model_path + model_name + '_' + str(epochs)+'_epochs.h5') #Specify optimizer\n",
    "print(\"Loaded full model with architecture, optimizer and metrics\")\n",
    "\n",
    "# Load weights into model\n",
    "model.load_weights(model_path + model_name + '_' + str(epochs)+'_epochs_weights.h5')\n",
    "print(\"Loaded model weights from disk\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluating Transfer learning model with Xception with validation set\n",
    "Computing loss and accuracy :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Model Loss , Model Accuracy]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.23591659848578275, 0.92]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('[Model Loss , Model Accuracy]')\n",
    "validation_samples = 100 # Select 100-1000 so that it does not take too long\n",
    "model.evaluate_generator(validation_generator, validation_samples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluating Transfer learning model with Xception with testing set\n",
    "Computing loss and accuracy :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[Model Loss , Model Accuracy]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.2568440991220996, 0.910625]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('[Model Loss , Model Accuracy]')\n",
    "test_subsamples = 100 # Select 100-1000 so that it does not take too long\n",
    "model.evaluate_generator(test_generator, test_subsamples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "--------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROC / AUC for a categorical multi-class classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Code to Plot a Multi-class ROC/AUC\n",
    "\n",
    "Use micro and marco averaging to evaluate the overall performance across all classes.\n",
    "$$\n",
    "precision=PRE=\\frac{TP}\n",
    "{TP+FP}\\\\\n",
    "$$\n",
    " In “micro averaging”, we’d calculate the performance, e.g., precision, from the individual true positives, true negatives, false positives, and false negatives of the the k-class model:\n",
    "$$\n",
    "PRE_{micro}=\\frac{TP_{1}+\\dots+TP_{k}}\n",
    "{TP_{1}+\\dots+TP_{k}+FP_{1}+\\dots+FP_{k}}\\\\\n",
    "$$\n",
    "And in macro-averaging, we average the performances of each individual class:\n",
    "$$\n",
    "PRE_{marco}=\\frac{PRE_{1}+\\dots+PRE_{k}}\n",
    "{k}\\\\\n",
    "$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Getting the image filenames(with associated classess)\n",
    "y_fnames = validation_generator.filenames\n",
    "\n",
    "# Getting the ground truth for all the tested images\n",
    "y_true_categorical = validation_generator.classes\n",
    "y_true = (to_categorical(y_true_categorical, num_classes = class_num))\n",
    "\n",
    "# Getting the probability scores for belonging to all classess in tested images\n",
    "y_score = model.predict_generator(validation_generator)\n",
    "\n",
    "# Getting vocabulary or class indices\n",
    "y_indices = (validation_generator.class_indices)\n",
    "y_label_map = dict((v,k) for k,v in y_indices.items())\n",
    "\n",
    "# Getting the predicted category\n",
    "y_predictions = np.argmax(y_score, axis=-1) #multiple categories\n",
    "\n",
    "# Getting predicted category\n",
    "y_predicted_category = [y_label_map[k] for k in y_predictions]\n",
    "\n",
    "\n",
    "# Plot linewidth.\n",
    "lw = 2\n",
    "\n",
    "# Compute ROC curve and ROC area for each class\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(class_num):\n",
    "    fpr[i], tpr[i], _ = roc_curve(y_true[:, i], y_score[:, i])\n",
    "    roc_auc[i] = auc(fpr[i], tpr[i])\n",
    "\n",
    "# Compute micro-average ROC curve and ROC area\n",
    "fpr[\"micro\"], tpr[\"micro\"], _ = roc_curve(y_true.ravel(), y_score.ravel())\n",
    "roc_auc[\"micro\"] = auc(fpr[\"micro\"], tpr[\"micro\"])\n",
    "\n",
    "# Compute macro-average ROC curve and ROC area\n",
    "\n",
    "# First aggregate all false positive rates\n",
    "all_fpr = np.unique(np.concatenate([fpr[i] for i in range(class_num)]))\n",
    "\n",
    "# Then interpolate all ROC curves at this points\n",
    "mean_tpr = np.zeros_like(all_fpr)\n",
    "for i in range(class_num):\n",
    "    mean_tpr += interp(all_fpr, fpr[i], tpr[i])\n",
    "\n",
    "# Finally average it and compute AUC\n",
    "mean_tpr /= class_num\n",
    "\n",
    "fpr[\"macro\"] = all_fpr\n",
    "tpr[\"macro\"] = mean_tpr\n",
    "roc_auc[\"macro\"] = auc(fpr[\"macro\"], tpr[\"macro\"])\n",
    "\n",
    "# Plot all ROC curves\n",
    "plt.figure(1)\n",
    "plt.plot(fpr[\"micro\"], tpr[\"micro\"],\n",
    "         label='micro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"micro\"]),\n",
    "         color='deeppink', linestyle=':', linewidth=4)\n",
    "\n",
    "plt.plot(fpr[\"macro\"], tpr[\"macro\"],\n",
    "         label='macro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"macro\"]),\n",
    "         color='navy', linestyle=':', linewidth=4)\n",
    "\n",
    "colors = cycle(['aqua', 'darkorange', 'cornflowerblue'])\n",
    "for i, color in zip(range(class_num), colors):\n",
    "    plt.plot(fpr[i], tpr[i], color=color, lw=lw,\n",
    "             label='ROC curve of class {0} (area = {1:0.2f})'\n",
    "             ''.format(i, roc_auc[i]))\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'k--', lw=lw)\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic (ROC) to Multi-Class Validation Set (' + model_name + ')')\n",
    "plt.legend(bbox_to_anchor=(1.85,0), loc=\"lower right\")\n",
    "plt.savefig(model_path + '0_results/Model' + '_' + model_name + '_' + str(epochs) + '_epochs_ROC_Validation.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "\n",
    "# Zoom in view of the upper left corner.\n",
    "plt.figure(2)\n",
    "plt.xlim(0, 0.2)\n",
    "plt.ylim(0.8, 1)\n",
    "plt.plot(fpr[\"micro\"], tpr[\"micro\"],\n",
    "         label='micro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"micro\"]),\n",
    "         color='deeppink', linestyle=':', linewidth=4)\n",
    "\n",
    "plt.plot(fpr[\"macro\"], tpr[\"macro\"],\n",
    "         label='macro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"macro\"]),\n",
    "         color='navy', linestyle=':', linewidth=4)\n",
    "\n",
    "colors = cycle(['aqua', 'darkorange', 'cornflowerblue'])\n",
    "for i, color in zip(range(class_num), colors):\n",
    "    plt.plot(fpr[i], tpr[i], color=color, lw=lw,\n",
    "             label='ROC curve of class {0} (area = {1:0.2f})'\n",
    "             ''.format(i, roc_auc[i]))\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'k--', lw=lw)\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Zoomed ROC to Multi-Class Validation Set (' + model_name + ')')\n",
    "plt.legend(bbox_to_anchor=(1.85,0), loc=\"lower right\")\n",
    "plt.savefig(model_path + '0_results/Model' + '_' + model_name + '_' + str(epochs) + '_epochs_ROC_Zoom_Validation.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "# Record Data\n",
    "v_fpr = fpr\n",
    "v_tpr = tpr"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### - ROC/AUC of Testing Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Getting the image filenames(with associated classess)\n",
    "y_fnames = test_generator.filenames\n",
    "\n",
    "# Getting the ground truth for all the tested images\n",
    "y_true_categorical = test_generator.classes\n",
    "y_true = (to_categorical(y_true_categorical, num_classes = class_num))\n",
    "\n",
    "# Getting the probability scores for belonging to all classess in tested images\n",
    "y_score = model.predict_generator(test_generator)\n",
    "\n",
    "# Getting vocabulary or class indices\n",
    "y_indices = (validation_generator.class_indices)\n",
    "y_label_map = dict((v,k) for k,v in y_indices.items())\n",
    "\n",
    "# Getting the predicted category\n",
    "y_predictions = np.argmax(y_score, axis=-1) #multiple categories\n",
    "\n",
    "# Getting predicted category\n",
    "y_predicted_category = [y_label_map[k] for k in y_predictions]\n",
    "\n",
    "\n",
    "# Plot linewidth.\n",
    "lw = 2\n",
    "\n",
    "# Compute ROC curve and ROC area for each class\n",
    "fpr = dict()\n",
    "tpr = dict()\n",
    "roc_auc = dict()\n",
    "for i in range(class_num):\n",
    "    fpr[i], tpr[i], _ = roc_curve(y_true[:, i], y_score[:, i])\n",
    "    roc_auc[i] = auc(fpr[i], tpr[i])\n",
    "\n",
    "# Compute micro-average ROC curve and ROC area\n",
    "fpr[\"micro\"], tpr[\"micro\"], _ = roc_curve(y_true.ravel(), y_score.ravel())\n",
    "roc_auc[\"micro\"] = auc(fpr[\"micro\"], tpr[\"micro\"])\n",
    "\n",
    "# Compute macro-average ROC curve and ROC area\n",
    "\n",
    "# First aggregate all false positive rates\n",
    "all_fpr = np.unique(np.concatenate([fpr[i] for i in range(class_num)]))\n",
    "\n",
    "# Then interpolate all ROC curves at this points\n",
    "mean_tpr = np.zeros_like(all_fpr)\n",
    "for i in range(class_num):\n",
    "    mean_tpr += interp(all_fpr, fpr[i], tpr[i])\n",
    "\n",
    "# Finally average it and compute AUC\n",
    "mean_tpr /= class_num\n",
    "\n",
    "fpr[\"macro\"] = all_fpr\n",
    "tpr[\"macro\"] = mean_tpr\n",
    "roc_auc[\"macro\"] = auc(fpr[\"macro\"], tpr[\"macro\"])\n",
    "\n",
    "# Plot all ROC curves\n",
    "plt.figure(1)\n",
    "plt.plot(fpr[\"micro\"], tpr[\"micro\"],\n",
    "         label='micro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"micro\"]),\n",
    "         color='deeppink', linestyle=':', linewidth=4)\n",
    "\n",
    "plt.plot(fpr[\"macro\"], tpr[\"macro\"],\n",
    "         label='macro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"macro\"]),\n",
    "         color='navy', linestyle=':', linewidth=4)\n",
    "\n",
    "colors = cycle(['aqua', 'darkorange', 'cornflowerblue'])\n",
    "for i, color in zip(range(class_num), colors):\n",
    "    plt.plot(fpr[i], tpr[i], color=color, lw=lw,\n",
    "             label='ROC curve of class {0} (area = {1:0.2f})'\n",
    "             ''.format(i, roc_auc[i]))\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'k--', lw=lw)\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic (ROC) to Multi-Class Testing Set (' + model_name + ')')\n",
    "plt.legend(bbox_to_anchor=(1.85,0), loc=\"lower right\")\n",
    "plt.savefig(model_path + '0_results/Model' + '_' + model_name + '_' + str(epochs) + '_epochs_ROC_Testing.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "\n",
    "# Zoom in view of the upper left corner.\n",
    "plt.figure(2)\n",
    "plt.xlim(0, 0.2)\n",
    "plt.ylim(0.8, 1)\n",
    "plt.plot(fpr[\"micro\"], tpr[\"micro\"],\n",
    "         label='micro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"micro\"]),\n",
    "         color='deeppink', linestyle=':', linewidth=4)\n",
    "\n",
    "plt.plot(fpr[\"macro\"], tpr[\"macro\"],\n",
    "         label='macro-average ROC curve (area = {0:0.2f})'\n",
    "               ''.format(roc_auc[\"macro\"]),\n",
    "         color='navy', linestyle=':', linewidth=4)\n",
    "\n",
    "colors = cycle(['aqua', 'darkorange', 'cornflowerblue'])\n",
    "for i, color in zip(range(class_num), colors):\n",
    "    plt.plot(fpr[i], tpr[i], color=color, lw=lw,\n",
    "             label='ROC curve of class {0} (area = {1:0.2f})'\n",
    "             ''.format(i, roc_auc[i]))\n",
    "\n",
    "plt.plot([0, 1], [0, 1], 'k--', lw=lw)\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Zoomed ROC to Multi-Class Testing Set (' + model_name + ')')\n",
    "plt.legend(bbox_to_anchor=(1.85,0), loc=\"lower right\")\n",
    "plt.savefig(model_path + '0_results/Model' + '_' + model_name + '_' + str(epochs) + '_epochs_ROC_Zoom_Testing.png', dpi=600, bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\n",
    "# Record Data\n",
    "t_fpr = fpr\n",
    "t_tpr = tpr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "## SAVE REPORTING DATA FOR PUBLICATIONS\n",
    "\n",
    "# Imports\n",
    "import pandas as pd\n",
    "\n",
    "# -----------------------------------------------------------------------------------\n",
    "# Helper function to save raw data to files\n",
    "def save_data_to_excel(data_path, df):\n",
    "    # Create a Pandas Excel writer using XlsxWriter as the engine.\n",
    "    writer = pd.ExcelWriter(data_path, engine='xlsxwriter')\n",
    "\n",
    "    # Convert the dataframe to an XlsxWriter Excel object\n",
    "    df.to_excel(writer, sheet_name='Sheet1', startrow=1, header=False)\n",
    "\n",
    "    # Get the xlsxwriter workbook and worksheet objects.\n",
    "    workbook  = writer.book\n",
    "    worksheet = writer.sheets['Sheet1']\n",
    "\n",
    "    # Add a header format.\n",
    "    header_format = workbook.add_format({\n",
    "        'bold': True,\n",
    "        'text_wrap': True,\n",
    "        'valign': 'top',\n",
    "        'fg_color': '#D7E4BC',\n",
    "        'border': 1})\n",
    "\n",
    "    # Write the column headers with the defined format.\n",
    "    for col_num, value in enumerate(df.columns.values):\n",
    "        worksheet.write(0, col_num + 1, value, header_format)\n",
    "\n",
    "    # Close the Pandas Excel writer and output the Excel file.\n",
    "    writer.save()\n",
    "    \n",
    "    # Message\n",
    "    print('Data saved to: ' + str(data_path))\n",
    "    \n",
    "    return writer\n",
    "\n",
    "\n",
    "# SAVING DATA\n",
    "# -----------------------------------------------------------------------------------\n",
    "\n",
    "# History Data\n",
    "# Define path to save raw model training/testing data\n",
    "history_data_path = model_path + '0_results/data_finetuning/' + 'model_history.xlsx'\n",
    "\n",
    "# Define dataframe to put in excel file\n",
    "df_training_history = pd.DataFrame({'Training Accuracy': loaded_training_history['acc'],\n",
    "                                    'Validation Accuracy': loaded_training_history['val_acc'],\n",
    "                                    'Training Loss': loaded_training_history['loss'],\n",
    "                                    'Validation Loss': loaded_training_history['val_loss']})\n",
    "\n",
    "# Save training/testing data to file\n",
    "save_data_to_excel(history_data_path, df_training_history)\n",
    "\n",
    "\n",
    "# ROC Curve Data\n",
    "# Define path to save raw model ROCs\n",
    "roc_data_base_path = model_path + '0_results/data_finetuning/' + 'data_'\n",
    "\n",
    "# Define dataframe to put in excel file\n",
    "\n",
    "## VALIDATION\n",
    "# Validation FPR\n",
    "d = {'Background': v_fpr[0],\n",
    "    'Skin Edge': v_fpr[1],\n",
    "    'Skin': v_fpr[2],\n",
    "    'NSPL-A': v_fpr[3],\n",
    "    'NSPL-B': v_fpr[4],\n",
    "    'SPL': v_fpr[5],\n",
    "    'Micro': v_fpr['micro'],\n",
    "    'Macro': v_fpr['macro']}\n",
    "\n",
    "df_test_ROCs_val_fpr = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "# Validation TPR\n",
    "d = {'Background': v_tpr[0],\n",
    "    'Skin Edge': v_tpr[1],\n",
    "    'Skin': v_tpr[2],\n",
    "    'NSPL-A': v_tpr[3],\n",
    "    'NSPL-B': v_tpr[4],\n",
    "    'SPL': v_tpr[5],\n",
    "    'Micro': v_tpr['micro'],\n",
    "    'Macro': v_tpr['macro']}\n",
    "\n",
    "df_test_ROCs_val_tpr = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "# Validation AUCs\n",
    "d = {'Background': auc(v_fpr[0], v_tpr[0]),\n",
    "    'Skin Edge': auc(v_fpr[1], v_tpr[1]),\n",
    "    'Skin': auc(v_fpr[2], v_tpr[2]),\n",
    "    'NSPL-A': auc(v_fpr[3], v_tpr[3]),\n",
    "    'NSPL-B': auc(v_fpr[4], v_tpr[4]),\n",
    "    'SPL': auc(v_fpr[5], v_tpr[5]),\n",
    "    'Micro': auc(v_fpr['micro'], v_tpr['micro']),\n",
    "    'Macro': auc(v_fpr['macro'], v_tpr['macro'])}\n",
    "\n",
    "df_test_ROCs_val_auc = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "## TESTING\n",
    "# Testing FPR\n",
    "d = {'Background': t_fpr[0],\n",
    "    'Skin Edge': t_fpr[1],\n",
    "    'Skin': t_fpr[2],\n",
    "    'NSPL-A': t_fpr[3],\n",
    "    'NSPL-B': t_fpr[4],\n",
    "    'SPL': t_fpr[5],\n",
    "    'Micro': t_fpr['micro'],\n",
    "    'Macro': t_fpr['macro']}\n",
    "\n",
    "df_test_ROCs_test_fpr = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "# Testing TPR\n",
    "d = {'Background': t_tpr[0],\n",
    "    'Skin Edge': t_tpr[1],\n",
    "    'Skin': t_tpr[2],\n",
    "    'NSPL-A': t_tpr[3],\n",
    "    'NSPL-B': t_tpr[4],\n",
    "    'SPL': t_tpr[5],\n",
    "    'Micro': t_tpr['micro'],\n",
    "    'Macro': t_tpr['macro']}\n",
    "\n",
    "df_test_ROCs_test_tpr = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "# Testing AUCs\n",
    "d = {'Background': auc(t_fpr[0], t_tpr[0]),\n",
    "    'Skin Edge': auc(t_fpr[1], t_tpr[1]),\n",
    "    'Skin': auc(t_fpr[2], t_tpr[2]),\n",
    "    'NSPL-A': auc(t_fpr[3], t_tpr[3]),\n",
    "    'NSPL-B': auc(t_fpr[4], t_tpr[4]),\n",
    "    'SPL': auc(t_fpr[5], t_tpr[5]),\n",
    "    'Micro': auc(t_fpr['micro'], t_tpr['micro']),\n",
    "    'Macro': auc(t_fpr['macro'], t_tpr['macro'])}\n",
    "\n",
    "df_test_ROCs_test_auc = pd.DataFrame(dict([(k,pd.Series(v)) for k,v in d.items()]))\n",
    "\n",
    "\n",
    "\n",
    "# Save training/testing data to file\n",
    "save_data_to_excel(roc_data_base_path + 'val_fpr.xlsx', df_test_ROCs_val_fpr)\n",
    "save_data_to_excel(roc_data_base_path + 'val_tpr.xlsx', df_test_ROCs_val_tpr)\n",
    "save_data_to_excel(roc_data_base_path + 'val_auc.xlsx', df_test_ROCs_val_auc)\n",
    "\n",
    "save_data_to_excel(roc_data_base_path + 'test_fpr.xlsx', df_test_ROCs_test_fpr)\n",
    "save_data_to_excel(roc_data_base_path + 'test_tpr.xlsx', df_test_ROCs_test_tpr)\n",
    "save_data_to_excel(roc_data_base_path + 'test_auc.xlsx', df_test_ROCs_test_auc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "------------------------------------------------------------\n",
    "### End of code"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}