2019.12-2.APTOS2019/aptos2019-t1-para.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some good references:\n",
    "1. https://towardsdatascience.com/a-bunch-of-tips-and-tricks-for-training-deep-neural-networks-3ca24c31ddc8\n",
    "2. https://towardsdatascience.com/review-densenet-image-classification-b6631a8ef803"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['aptos2019-blindness-detection', 'densenet-keras']\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load in \n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import tensorflow as tf\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.metrics import cohen_kappa_score\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.applications.densenet import DenseNet169 as DenseNet121 # DenseNet121 ot DenseNet169\n",
    "import keras\n",
    "import cv2\n",
    "# Input data files are available in the \"../input/\" directory.\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n",
    "import cv2\n",
    "import os\n",
    "from keras.callbacks import Callback\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.utils.multiclass import unique_labels\n",
    "from sklearn.utils import class_weight\n",
    "print(os.listdir(\"../input\"))\n",
    "\n",
    "# Any results you write to the current directory are saved as output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# borrowed from https://www.kaggle.com/mathormad/aptos-resnet50-baseline\n",
    "class QWKCallback(Callback):\n",
    "    def __init__(self,validation_data):\n",
    "        super(Callback, self).__init__()\n",
    "        self.X = validation_data[0]\n",
    "        self.Y = validation_data[1]\n",
    "        self.history = []\n",
    "    def on_epoch_end(self, epoch, logs={}):\n",
    "        pred = self.model.predict(self.X)\n",
    "        score = cohen_kappa_score(np.argmax(self.Y,axis=1),np.argmax(pred,axis=1),labels=[0,1,2,3,4],weights='quadratic')\n",
    "        print(\"Epoch {} : QWK: {}\".format(epoch,score))\n",
    "        self.history.append(score)\n",
    "        if score >= max(self.history):\n",
    "            print('saving checkpoint: ', score)\n",
    "            self.model.save('../working/Resnet50_bestqwk.h5')\n",
    "        \n",
    "        \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# borrowed from https://github.com/yu4u/mixup-generator\n",
    "class MixupGenerator():\n",
    "    def __init__(self, X_train, y_train, batch_size=32, alpha=0.2, shuffle=True, datagen=None):\n",
    "        self.X_train = X_train\n",
    "        self.y_train = y_train\n",
    "        self.batch_size = batch_size\n",
    "        self.alpha = alpha\n",
    "        self.shuffle = shuffle\n",
    "        self.sample_num = len(X_train)\n",
    "        self.datagen = datagen\n",
    "\n",
    "    def __call__(self):\n",
    "        while True:\n",
    "            indexes = self.__get_exploration_order()\n",
    "            itr_num = int(len(indexes) // (self.batch_size * 2))\n",
    "\n",
    "            for i in range(itr_num):\n",
    "                batch_ids = indexes[i * self.batch_size * 2:(i + 1) * self.batch_size * 2]\n",
    "                X, y = self.__data_generation(batch_ids)\n",
    "\n",
    "                yield X, y\n",
    "\n",
    "    def __get_exploration_order(self):\n",
    "        indexes = np.arange(self.sample_num)\n",
    "\n",
    "        if self.shuffle:\n",
    "            np.random.shuffle(indexes)\n",
    "\n",
    "        return indexes\n",
    "\n",
    "    def __data_generation(self, batch_ids):\n",
    "        _, h, w, c = self.X_train.shape\n",
    "        l = np.random.beta(self.alpha, self.alpha, self.batch_size)\n",
    "        X_l = l.reshape(self.batch_size, 1, 1, 1)\n",
    "        y_l = l.reshape(self.batch_size, 1)\n",
    "\n",
    "        X1 = self.X_train[batch_ids[:self.batch_size]]\n",
    "        X2 = self.X_train[batch_ids[self.batch_size:]]\n",
    "        X = X1 * X_l + X2 * (1 - X_l)\n",
    "\n",
    "        if self.datagen:\n",
    "            for i in range(self.batch_size):\n",
    "                X[i] = self.datagen.random_transform(X[i])\n",
    "                X[i] = self.datagen.standardize(X[i])\n",
    "\n",
    "        if isinstance(self.y_train, list):\n",
    "            y = []\n",
    "\n",
    "            for y_train_ in self.y_train:\n",
    "                y1 = y_train_[batch_ids[:self.batch_size]]\n",
    "                y2 = y_train_[batch_ids[self.batch_size:]]\n",
    "                y.append(y1 * y_l + y2 * (1 - y_l))\n",
    "        else:\n",
    "            y1 = self.y_train[batch_ids[:self.batch_size]]\n",
    "            y2 = self.y_train[batch_ids[self.batch_size:]]\n",
    "            y = y1 * y_l + y2 * (1 - y_l)\n",
    "\n",
    "        return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# borrowed from scikit learn\n",
    "def plot_confusion_matrix(y_true, y_pred, classes,\n",
    "                          normalize=False,\n",
    "                          title=None,\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if not title:\n",
    "        if normalize:\n",
    "            title = 'Normalized confusion matrix'\n",
    "        else:\n",
    "            title = 'Confusion matrix, without normalization'\n",
    "\n",
    "    # Compute confusion matrix\n",
    "    cm = confusion_matrix(y_true, y_pred)\n",
    "    # Only use the labels that appear in the data\n",
    "    classes = classes[unique_labels(y_true, y_pred)]\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    fig, ax = plt.subplots()\n",
    "    im = ax.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    ax.figure.colorbar(im, ax=ax)\n",
    "    # We want to show all ticks...\n",
    "    ax.set(xticks=np.arange(cm.shape[1]),\n",
    "           yticks=np.arange(cm.shape[0]),\n",
    "           # ... and label them with the respective list entries\n",
    "           xticklabels=classes, yticklabels=classes,\n",
    "           title=title,\n",
    "           ylabel='True label',\n",
    "           xlabel='Predicted label')\n",
    "\n",
    "    # Rotate the tick labels and set their alignment.\n",
    "    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",\n",
    "             rotation_mode=\"anchor\")\n",
    "\n",
    "    # Loop over data dimensions and create text annotations.\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i in range(cm.shape[0]):\n",
    "        for j in range(cm.shape[1]):\n",
    "            ax.text(j, i, format(cm[i, j], fmt),\n",
    "                    ha=\"center\", va=\"center\",\n",
    "                    color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "    fig.tight_layout()\n",
    "    return ax\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_raw_images_df(data_frame,filenamecol,labelcol,img_size,n_classes):\n",
    "    n_images = len(data_frame)\n",
    "    X = np.empty((n_images,img_size,img_size,3))\n",
    "    Y = np.zeros((n_images,n_classes))\n",
    "    for index,entry in data_frame.iterrows():\n",
    "        Y[index,entry[labelcol]] = 1 # one hot encoding of the label\n",
    "        # Load the image and resize\n",
    "        img = cv2.imread(entry[filenamecol])\n",
    "        X[index,:] = cv2.resize(img, (img_size, img_size))\n",
    "        X[index,:] = X[index,:] / 255.0\n",
    "    return X,Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "img_size = 224"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
    "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7cb611638ba8>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAD8CAYAAACRkhiPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAFcpJREFUeJzt3X+w5XV93/HnK6DUcg2QQG43u6SLM6sz/EgIe4fScXTuVis/tKJJbJehCP6YVYtpMnUmYtqpVocp04bYgVDtGnaAuuHKiLobstYi4cZmBlSWIAsicdFtXGB2q2tWrzJ0lr77x/munlzuj/Pj3nNWvs/HzJn7PZ/v5/v9vL9fztnX+f44h1QVkqR2+rlxFyBJGh9DQJJazBCQpBYzBCSpxQwBSWoxQ0CSWswQkKQWMwQkqcUMAUlqsePHXcByTj311Fq/fv1Ay/7oRz/ixBNPXNmCVoB19ce6+mNd/Xkh1rV79+7vVtVpPXWuqmP6sXHjxhrUvffeO/Cyq8m6+mNd/bGu/rwQ6wIeqB7/jfV0kCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUosZApLUYoaAJLXYMf+zEcPY8+Rhrrrmz0Y+7r7rXj/yMSVpEB4JSFKLLRsCSbYlOZjkka62TyV5qHnsS/JQ074+yTNd8z7etczGJHuS7E1yQ5KsziZJknrVy+mgW4A/Am472lBV/+LodJLrgcNd/Z+oqnMXWM/HgC3A/cAu4CLg8/2XLElaKcseCVTVl4BDC81rPs3/c+D2pdaRZA3w81V1X/MLd7cBb+q/XEnSShr2msCrgANV9c2utjOS/FWSv0jyqqZtLbC/q8/+pk2SNEbpfDBfplOyHrirqs6e1/4xYG9VXd88PwGYqKrvJdkIfA44C3gF8B+r6rVNv1cBv1dV/2yR8bbQOXXE5OTkxpmZmYE27uChwxx4ZqBFh3LO2pOWnD83N8fExMSIqumddfXHuvpjXf0Zpq5NmzbtrqqpXvoOfItokuOB3wA2Hm2rqmeBZ5vp3UmeAF5O55P/uq7F1wFPLbbuqtoKbAWYmpqq6enpgWq8cfsOrt8z+rtg910+veT82dlZBt2m1WRd/bGu/lhXf0ZV1zCng14LfKOqfnKaJ8lpSY5rpl8GbAC+VVVPAz9MckFzHeGtwI4hxpYkrYBebhG9HbgPeEWS/Une0czazPMvCL8aeDjJ14BPA++uqqMXld8D/DGwF3gC7wySpLFb9lxJVV22SPtVC7TdCdy5SP8HgLMXmidJGg+/MSxJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUosZApLUYoaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktdiyIZBkW5KDSR7pavtQkieTPNQ8Luma94Eke5M8nuTCrvaLmra9Sa5Z+U2RJPWrlyOBW4CLFmj/aFWd2zx2ASQ5E9gMnNUs81+THJfkOOAm4GLgTOCypq8kaYyOX65DVX0pyfoe13cpMFNVzwLfTrIXOL+Zt7eqvgWQZKbp+/W+K5YkrZhhrgm8N8nDzemiU5q2tcB3uvrsb9oWa5ckjVGqavlOnSOBu6rq7Ob5JPBdoICPAGuq6u1JbgLuq6pPNv1uBnbRCZsLq+qdTfsVwPlV9duLjLcF2AIwOTm5cWZmZqCNO3joMAeeGWjRoZyz9qQl58/NzTExMTGianpnXf2xrv5YV3+GqWvTpk27q2qql77Lng5aSFUdODqd5BPAXc3T/cDpXV3XAU8104u1L7T+rcBWgKmpqZqenh6kTG7cvoPr9wy0iUPZd/n0kvNnZ2cZdJtWk3X1x7r6Y139GVVdA50OSrKm6+mbgaN3Du0ENic5IckZwAbgK8BXgQ1JzkjyYjoXj3cOXrYkaSUs+zE5ye3ANHBqkv3AB4HpJOfSOR20D3gXQFU9muQOOhd8jwBXV9VzzXreC3wBOA7YVlWPrvjWSJL60svdQZct0HzzEv2vBa5doH0XnesDkqRjhN8YlqQWMwQkqcUMAUlqMUNAklrMEJCkFjMEJKnFDAFJajFDQJJazBCQpBYzBCSpxQwBSWoxQ0CSWswQkKQWMwQkqcUMAUlqMUNAklrMEJCkFjMEJKnFDAFJajFDQJJabNkQSLItycEkj3S1/eck30jycJLPJjm5aV+f5JkkDzWPj3ctszHJniR7k9yQJKuzSZKkXvVyJHALcNG8truBs6vqV4G/Bj7QNe+Jqjq3eby7q/1jwBZgQ/OYv05J0ogtGwJV9SXg0Ly2/1lVR5qn9wPrllpHkjXAz1fVfVVVwG3AmwYrWZK0UtL5N3mZTsl64K6qOnuBeX8KfKqqPtn0e5TO0cEPgH9XVf8ryRRwXVW9tlnmVcD7q+oNi4y3hc5RA5OTkxtnZmb63zLg4KHDHHhmoEWHcs7ak5acPzc3x8TExIiq6Z119ce6+mNd/Rmmrk2bNu2uqqle+h4/0AiNJP8WOAJsb5qeBn6lqr6XZCPwuSRnAQud/180fapqK7AVYGpqqqanpweq78btO7h+z1CbOJB9l08vOX92dpZBt2k1WVd/rKs/1tWfUdU18L+QSa4E3gC8pjnFQ1U9CzzbTO9O8gTwcmA/f/eU0TrgqUHHliStjIFuEU1yEfB+4I1V9eOu9tOSHNdMv4zOBeBvVdXTwA+TXNDcFfRWYMfQ1UuShrLskUCS24Fp4NQk+4EP0rkb6ATg7uZOz/ubO4FeDXw4yRHgOeDdVXX0ovJ76Nxp9BLg881DkjRGy4ZAVV22QPPNi/S9E7hzkXkPAM+7sCxJGh+/MSxJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUosZApLUYoaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRiPYVAkm1JDiZ5pKvtF5LcneSbzd9TmvYkuSHJ3iQPJzmva5krm/7fTHLlym+OJKkfvR4J3AJcNK/tGuCeqtoA3NM8B7gY2NA8tgAfg05oAB8E/hFwPvDBo8EhSRqPnkKgqr4EHJrXfClwazN9K/CmrvbbquN+4OQka4ALgbur6lBVfR+4m+cHiyRphIa5JjBZVU8DNH9/qWlfC3ynq9/+pm2xdknSmBy/CuvMAm21RPvzV5BsoXMqicnJSWZnZwcqZPIl8L5zjgy07DCWq3dubm7gbVpN1tWfg4cOc+P2HSMf95y1Jy05/1jdX9bVn1HVNUwIHEiypqqebk73HGza9wOnd/VbBzzVtE/Pa59daMVVtRXYCjA1NVXT09MLdVvWjdt3cP2e1ci5pe27fHrJ+bOzswy6TavJuvrj66s/1tWfUdU1zOmgncDRO3yuBHZ0tb+1uUvoAuBwc7roC8DrkpzSXBB+XdMmSRqTnj7GJLmdzqf4U5Psp3OXz3XAHUneAfwN8Jam+y7gEmAv8GPgbQBVdSjJR4CvNv0+XFXzLzZLkkaopxCoqssWmfWaBfoWcPUi69kGbOu5OknSqvIbw5LUYoaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUosZApLUYoaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSiw0cAklekeShrscPkvxukg8lebKr/ZKuZT6QZG+Sx5NcuDKbIEka1PGDLlhVjwPnAiQ5DngS+CzwNuCjVfUH3f2TnAlsBs4Cfhn4YpKXV9Vzg9YgSRrOSp0Oeg3wRFX97yX6XArMVNWzVfVtYC9w/gqNL0kawEqFwGbg9q7n703ycJJtSU5p2tYC3+nqs79pkySNSapquBUkLwaeAs6qqgNJJoHvAgV8BFhTVW9PchNwX1V9slnuZmBXVd25wDq3AFsAJicnN87MzAxU28FDhznwzECLDuWctSctOX9ubo6JiYkRVdM76+qPr6/+WFd/hqlr06ZNu6tqqpe+A18T6HIx8GBVHQA4+hcgySeAu5qn+4HTu5ZbRyc8nqeqtgJbAaampmp6enqgwm7cvoPr96zEJvZn3+XTS86fnZ1l0G1aTdbVH19f/bGu/oyqrpU4HXQZXaeCkqzpmvdm4JFmeiewOckJSc4ANgBfWYHxJUkDGupjTJK/D/xT4F1dzf8pybl0TgftOzqvqh5NcgfwdeAIcLV3BknSeA0VAlX1Y+AX57VdsUT/a4FrhxlTkrRy/MawJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUosZApLUYoaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRiQ4dAkn1J9iR5KMkDTdsvJLk7yTebv6c07UlyQ5K9SR5Oct6w40uSBrdSRwKbqurcqppqnl8D3FNVG4B7mucAFwMbmscW4GMrNL4kaQCrdTroUuDWZvpW4E1d7bdVx/3AyUnWrFINkqRlpKqGW0HybeD7QAH/raq2Jvnbqjq5q8/3q+qUJHcB11XVXzbt9wDvr6oH5q1zC50jBSYnJzfOzMwMVNvBQ4c58MxAiw7lnLUnLTl/bm6OiYmJEVXTO+vqj6+v/lhXf4apa9OmTbu7zsws6fiBRvi7XllVTyX5JeDuJN9Yom8WaHteClXVVmArwNTUVE1PTw9U2I3bd3D9npXYxP7su3x6yfmzs7MMuk2rybr64+urP9bVn1HVNfTpoKp6qvl7EPgscD5w4Ohpnubvwab7fuD0rsXXAU8NW4MkaTBDhUCSE5O89Og08DrgEWAncGXT7UpgRzO9E3hrc5fQBcDhqnp6mBokSYMb9lh2EvhskqPr+pOq+h9JvgrckeQdwN8Ab2n67wIuAfYCPwbeNuT4kqQhDBUCVfUt4NcWaP8e8JoF2gu4epgxJUkrx28MS1KLGQKS1GKGgCS1mCEgSS1mCEhSixkCktRihoAktZghIEktZghIUouN/icQtarWX/NnAy/7vnOOcNWAy++77vUDjytpfDwSkKQWMwQkqcUMAUlqMUNAklrMEJCkFjMEJKnFDAFJajFDQJJazC+LSdIShvkC5jBuuejEkYwz8JFAktOT3JvksSSPJvmdpv1DSZ5M8lDzuKRrmQ8k2Zvk8SQXrsQGSJIGN8yRwBHgfVX1YJKXAruT3N3M+2hV/UF35yRnApuBs4BfBr6Y5OVV9dwQNUiShjDwkUBVPV1VDzbTPwQeA9YuscilwExVPVtV3wb2AucPOr4kaXgrcmE4yXrg14EvN03vTfJwkm1JTmna1gLf6VpsP0uHhiRplaWqhltBMgH8BXBtVX0mySTwXaCAjwBrqurtSW4C7quqTzbL3Qzsqqo7F1jnFmALwOTk5MaZmZmBajt46DAHnhlo0aGcs/akJefPzc0xMTGxKmPvefLwwMtOvoSB99dy2zyM1dxfw2jj62sYP6t1DfOeGsYZJx038P7atGnT7qqa6qXvUHcHJXkRcCewvao+A1BVB7rmfwK4q3m6Hzi9a/F1wFMLrbeqtgJbAaampmp6enqg+m7cvoPr94z+Bqh9l08vOX92dpZBt2k5g/4UNHR+SnrQ/bXcNg9jNffXMNr4+hrGz2pdw7ynhnHLRSeOZH8Nc3dQgJuBx6rqD7va13R1ezPwSDO9E9ic5IQkZwAbgK8MOr4kaXjDfIx5JXAFsCfJQ03b7wOXJTmXzumgfcC7AKrq0SR3AF+nc2fR1d4ZJEnjNXAIVNVfAllg1q4llrkWuHbQMSVJK8ufjZCkFjMEJKnF/O0gST0b5nd03nfOkYHvtNl33esHHldL80hAklrMEJCkFjMEJKnFDAFJajFDQJJazBCQpBYzBCSpxQwBSWoxQ0CSWswQkKQWMwQkqcUMAUlqMUNAklrMEJCkFjMEJKnFDAFJajFDQJJabOQhkOSiJI8n2ZvkmlGPL0n6qZGGQJLjgJuAi4EzgcuSnDnKGiRJPzXqI4Hzgb1V9a2q+r/ADHDpiGuQJDVGHQJrge90Pd/ftEmSxiBVNbrBkrcAF1bVO5vnVwDnV9Vvz+u3BdjSPH0F8PiAQ54KfHfAZVeTdfXHuvpjXf15Idb1D6vqtF46Hj/gAIPaD5ze9Xwd8NT8TlW1Fdg67GBJHqiqqWHXs9Ksqz/W1R/r6k/b6xr16aCvAhuSnJHkxcBmYOeIa5AkNUZ6JFBVR5K8F/gCcBywraoeHWUNkqSfGvXpIKpqF7BrRMMNfUpplVhXf6yrP9bVn1bXNdILw5KkY4s/GyFJLfaCCIHlfooiyQlJPtXM/3KS9cdIXVcl+T9JHmoe7xxBTduSHEzyyCLzk+SGpuaHk5y32jX1WNd0ksNd++rfj6iu05Pcm+SxJI8m+Z0F+ox8n/VY18j3WZK/l+QrSb7W1PUfFugz8vdjj3WN/P3YNfZxSf4qyV0LzFvd/VVVP9MPOheYnwBeBrwY+Bpw5rw+/wr4eDO9GfjUMVLXVcAfjXh/vRo4D3hkkfmXAJ8HAlwAfPkYqWsauGsMr681wHnN9EuBv17gv+PI91mPdY18nzX7YKKZfhHwZeCCeX3G8X7spa6Rvx+7xv43wJ8s9N9rtffXC+FIoJeforgUuLWZ/jTwmiQ5Buoauar6EnBoiS6XArdVx/3AyUnWHAN1jUVVPV1VDzbTPwQe4/nfch/5PuuxrpFr9sFc8/RFzWP+hceRvx97rGsskqwDXg/88SJdVnV/vRBCoJefovhJn6o6AhwGfvEYqAvgN5tTCJ9OcvoC80ftWP5pj3/cHM5/PslZox68OQz/dTqfIruNdZ8tUReMYZ81pzYeAg4Cd1fVovtrhO/HXuqC8bwf/wvwe8D/W2T+qu6vF0IILJSI8xO+lz4rrZcx/xRYX1W/CnyRn6b9OI1jX/XiQTpfhf814Ebgc6McPMkEcCfwu1X1g/mzF1hkJPtsmbrGss+q6rmqOpfOLwKcn+TseV3Gsr96qGvk78ckbwAOVtXupbot0LZi++uFEAK9/BTFT/okOR44idU/9bBsXVX1vap6tnn6CWDjKtfUi55+2mPUquoHRw/nq/NdkxclOXUUYyd5EZ1/aLdX1WcW6DKWfbZcXePcZ82YfwvMAhfNmzWO9+OydY3p/fhK4I1J9tE5ZfxPknxyXp9V3V8vhBDo5acodgJXNtO/Bfx5NVdZxlnXvPPGb6RzXnfcdgJvbe54uQA4XFVPj7uoJP/g6HnQJOfTee1+bwTjBrgZeKyq/nCRbiPfZ73UNY59luS0JCc30y8BXgt8Y163kb8fe6lrHO/HqvpAVa2rqvV0/o3486r6l/O6rer+Gvk3hldaLfJTFEk+DDxQVTvpvFn+e5K9dBJ08zFS179O8kbgSFPXVatdV5Lb6dw1cmqS/cAH6Vwko6o+Tufb3JcAe4EfA29b7Zp6rOu3gPckOQI8A2weQZBD55PaFcCe5nwywO8Dv9JV2zj2WS91jWOfrQFuTed/IPVzwB1Vdde434891jXy9+NiRrm//MawJLXYC+F0kCRpQIaAJLWYISBJLWYISFKLGQKS1GKGgCS1mCEgSS1mCEhSi/1/lzimAUzcHkcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_raw_data = pd.read_csv(\"../input/aptos2019-blindness-detection/train.csv\")\n",
    "train_raw_data[\"filename\"] = train_raw_data[\"id_code\"].map(lambda x:os.path.join(\"../input/aptos2019-blindness-detection/train_images\",x+\".png\"))\n",
    "train_raw_data.diagnosis.hist() # See the distribution of the classes\n",
    "# train_raw_data.dtypes\n",
    "\n",
    "# # train_data[\"diagnosis\"] = train_data[\"diagnosis\"].astype(str)\n",
    "# # print(train_data.head())\n",
    "# # print(train_data.diagnosis.unique()) # Look at different types of classes\n",
    "# # labels = list(map(str,range(5)))\n",
    "# # print(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_title = {\"0\" : \"No DR\",\"1\" : \"Mild\",\"2\" : \"Moderate\",\"3\" :\"Severe\",\"4\" : \"Proliferative DR\"}\n",
    "class_labels=[\"No DR\",\"Mild\",\"Moderate\",\"Severe\",\"Proliferative DR\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Display some images\n",
    "figure, ax = plt.subplots(5,2)\n",
    "ax = ax.flatten()\n",
    "for i,row in train_raw_data.iloc[0:10,:].iterrows():\n",
    "    ax[i].imshow(cv2.imread(os.path.join(\"../input/aptos2019-blindness-detection/train_images\",row[\"id_code\"]+\".png\")))\n",
    "    ax[i].set_title(label_title[str(row[\"diagnosis\"])])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df,val_df = train_test_split(train_raw_data,random_state=42,shuffle=True,test_size=0.15)\n",
    "train_df.reset_index(drop=True,inplace=True)\n",
    "val_df.reset_index(drop=True,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,Y_train = load_raw_images_df(train_df,\"filename\",\"diagnosis\",img_size,5)\n",
    "X_val,Y_val = load_raw_images_df(val_df,\"filename\",\"diagnosis\",img_size,5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{0: 0.40363164721141376, 1: 1.945, 2: 0.7471788715486194, 3: 3.6611764705882353, 4: 2.519838056680162}\n"
     ]
    }
   ],
   "source": [
    "Y_train_labels = np.argmax(Y_train,axis=1)\n",
    "class_weights = class_weight.compute_class_weight('balanced',np.unique(Y_train_labels),Y_train_labels)\n",
    "cls_wt_dict = dict(enumerate(class_weights))\n",
    "print(cls_wt_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "datagen = ImageDataGenerator(\n",
    "            \n",
    "            zoom_range=0.15,  # set range for random zoom\n",
    "        # set mode for filling points outside the input boundaries\n",
    "        fill_mode='constant',\n",
    "        cval=0.,  # value used for fill_mode = \"constant\"\n",
    "        horizontal_flip=True,  # randomly flip images\n",
    "        vertical_flip=True,  # randomly flip images\n",
    ")\n",
    "training_generator = MixupGenerator(X_train, Y_train, batch_size=batch_size, alpha=0.2, datagen=datagen)()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def buildModel():\n",
    "    DenseNet121_model = DenseNet121(include_top=False,weights=None,input_tensor=keras.layers.Input(shape=(img_size,img_size,3)))\n",
    "    DenseNet121_model.load_weights('../input/densenet-keras/DenseNet-BC-169-32-no-top.h5')\n",
    "#     /kaggle/input/densenet-keras/DenseNet-BC-169-32-no-top.h5\n",
    "    \n",
    "#     model = keras.Sequential()\n",
    "    \n",
    "#     model.add(keras.layers.Conv2D(filters = 32, kernel_size = (5,5),padding = 'same',activation ='relu', \n",
    "#                       input_shape = (img_size,img_size,3)))\n",
    "#     model.add(keras.layers.MaxPooling2D(pool_size=(2,2)))\n",
    "    \n",
    "#     model.add(keras.layers.Conv2D(filters = 64, kernel_size = (3,3),padding = 'Same',activation ='relu'))\n",
    "#     model.add(keras.layers.MaxPooling2D(pool_size=(2,2), strides=(2,2)))\n",
    "    \n",
    "#     model.add(keras.layers.Conv2D(filters =96, kernel_size = (3,3),padding = 'Same',activation ='relu'))\n",
    "#     model.add(keras.layers.MaxPooling2D(pool_size=(2,2), strides=(2,2)))\n",
    "\n",
    "#     model.add(keras.layers.Conv2D(filters = 96, kernel_size = (3,3),padding = 'Same',activation ='relu'))\n",
    "#     model.add(keras.layers.MaxPooling2D(pool_size=(2,2), strides=(2,2)))\n",
    "    \n",
    "#     model.add(keras.layers.Flatten())\n",
    "#     model.add(keras.layers.Dense(units = 512, activation = 'relu'))\n",
    "#     model.add(keras.layers.Dense(units = 5, activation = 'softmax'))\n",
    "    \n",
    "    p  = keras.layers.GlobalAveragePooling2D()(DenseNet121_model.output)\n",
    "#     fl = keras.layers.Flatten()(p)\n",
    "#     d2 = keras.layers.Dense(units = 1024, activation = 'relu',kernel_regularizer= keras.regularizers.l2(0.001))(p)\n",
    "#     d1 = keras.layers.Dense(units = 512, activation = 'relu',kernel_regularizer= keras.regularizers.l2(0.001))(d2)\n",
    "    d11 = keras.layers.Dense(units = 256, activation = 'relu',kernel_regularizer= keras.regularizers.l2(0.0001))(p)\n",
    "    o1 = keras.layers.Dense(units = 5, activation = 'softmax')(d11)\n",
    "    model = keras.models.Model(inputs = DenseNet121_model.input,outputs = o1)\n",
    "    sgd = keras.optimizers.SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\n",
    "    model.compile(optimizer=sgd,loss='categorical_crossentropy', metrics = ['accuracy'])\n",
    "    print(model.summary())\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 224, 224, 3)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "zero_padding2d_1 (ZeroPadding2D (None, 230, 230, 3)  0           input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv1/conv (Conv2D)             (None, 112, 112, 64) 9408        zero_padding2d_1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv1/bn (BatchNormalization)   (None, 112, 112, 64) 256         conv1/conv[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv1/relu (Activation)         (None, 112, 112, 64) 0           conv1/bn[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "zero_padding2d_2 (ZeroPadding2D (None, 114, 114, 64) 0           conv1/relu[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "pool1 (MaxPooling2D)            (None, 56, 56, 64)   0           zero_padding2d_2[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_0_bn (BatchNormali (None, 56, 56, 64)   256         pool1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_0_relu (Activation (None, 56, 56, 64)   0           conv2_block1_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_1_conv (Conv2D)    (None, 56, 56, 128)  8192        conv2_block1_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block1_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_1_relu (Activation (None, 56, 56, 128)  0           conv2_block1_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block1_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block1_concat (Concatenat (None, 56, 56, 96)   0           pool1[0][0]                      \n",
      "                                                                 conv2_block1_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_0_bn (BatchNormali (None, 56, 56, 96)   384         conv2_block1_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_0_relu (Activation (None, 56, 56, 96)   0           conv2_block2_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_1_conv (Conv2D)    (None, 56, 56, 128)  12288       conv2_block2_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block2_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_1_relu (Activation (None, 56, 56, 128)  0           conv2_block2_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block2_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block2_concat (Concatenat (None, 56, 56, 128)  0           conv2_block1_concat[0][0]        \n",
      "                                                                 conv2_block2_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_0_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block2_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_0_relu (Activation (None, 56, 56, 128)  0           conv2_block3_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_1_conv (Conv2D)    (None, 56, 56, 128)  16384       conv2_block3_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block3_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_1_relu (Activation (None, 56, 56, 128)  0           conv2_block3_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block3_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block3_concat (Concatenat (None, 56, 56, 160)  0           conv2_block2_concat[0][0]        \n",
      "                                                                 conv2_block3_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_0_bn (BatchNormali (None, 56, 56, 160)  640         conv2_block3_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_0_relu (Activation (None, 56, 56, 160)  0           conv2_block4_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_1_conv (Conv2D)    (None, 56, 56, 128)  20480       conv2_block4_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block4_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_1_relu (Activation (None, 56, 56, 128)  0           conv2_block4_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block4_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block4_concat (Concatenat (None, 56, 56, 192)  0           conv2_block3_concat[0][0]        \n",
      "                                                                 conv2_block4_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_0_bn (BatchNormali (None, 56, 56, 192)  768         conv2_block4_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_0_relu (Activation (None, 56, 56, 192)  0           conv2_block5_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_1_conv (Conv2D)    (None, 56, 56, 128)  24576       conv2_block5_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block5_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_1_relu (Activation (None, 56, 56, 128)  0           conv2_block5_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block5_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block5_concat (Concatenat (None, 56, 56, 224)  0           conv2_block4_concat[0][0]        \n",
      "                                                                 conv2_block5_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_0_bn (BatchNormali (None, 56, 56, 224)  896         conv2_block5_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_0_relu (Activation (None, 56, 56, 224)  0           conv2_block6_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_1_conv (Conv2D)    (None, 56, 56, 128)  28672       conv2_block6_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_1_bn (BatchNormali (None, 56, 56, 128)  512         conv2_block6_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_1_relu (Activation (None, 56, 56, 128)  0           conv2_block6_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_2_conv (Conv2D)    (None, 56, 56, 32)   36864       conv2_block6_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv2_block6_concat (Concatenat (None, 56, 56, 256)  0           conv2_block5_concat[0][0]        \n",
      "                                                                 conv2_block6_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "pool2_bn (BatchNormalization)   (None, 56, 56, 256)  1024        conv2_block6_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "pool2_relu (Activation)         (None, 56, 56, 256)  0           pool2_bn[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "pool2_conv (Conv2D)             (None, 56, 56, 128)  32768       pool2_relu[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "pool2_pool (AveragePooling2D)   (None, 28, 28, 128)  0           pool2_conv[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_0_bn (BatchNormali (None, 28, 28, 128)  512         pool2_pool[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_0_relu (Activation (None, 28, 28, 128)  0           conv3_block1_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_1_conv (Conv2D)    (None, 28, 28, 128)  16384       conv3_block1_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block1_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_1_relu (Activation (None, 28, 28, 128)  0           conv3_block1_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block1_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block1_concat (Concatenat (None, 28, 28, 160)  0           pool2_pool[0][0]                 \n",
      "                                                                 conv3_block1_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_0_bn (BatchNormali (None, 28, 28, 160)  640         conv3_block1_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_0_relu (Activation (None, 28, 28, 160)  0           conv3_block2_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_1_conv (Conv2D)    (None, 28, 28, 128)  20480       conv3_block2_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block2_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_1_relu (Activation (None, 28, 28, 128)  0           conv3_block2_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block2_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block2_concat (Concatenat (None, 28, 28, 192)  0           conv3_block1_concat[0][0]        \n",
      "                                                                 conv3_block2_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_0_bn (BatchNormali (None, 28, 28, 192)  768         conv3_block2_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_0_relu (Activation (None, 28, 28, 192)  0           conv3_block3_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_1_conv (Conv2D)    (None, 28, 28, 128)  24576       conv3_block3_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block3_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_1_relu (Activation (None, 28, 28, 128)  0           conv3_block3_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block3_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block3_concat (Concatenat (None, 28, 28, 224)  0           conv3_block2_concat[0][0]        \n",
      "                                                                 conv3_block3_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_0_bn (BatchNormali (None, 28, 28, 224)  896         conv3_block3_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_0_relu (Activation (None, 28, 28, 224)  0           conv3_block4_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_1_conv (Conv2D)    (None, 28, 28, 128)  28672       conv3_block4_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block4_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_1_relu (Activation (None, 28, 28, 128)  0           conv3_block4_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block4_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block4_concat (Concatenat (None, 28, 28, 256)  0           conv3_block3_concat[0][0]        \n",
      "                                                                 conv3_block4_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_0_bn (BatchNormali (None, 28, 28, 256)  1024        conv3_block4_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_0_relu (Activation (None, 28, 28, 256)  0           conv3_block5_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_1_conv (Conv2D)    (None, 28, 28, 128)  32768       conv3_block5_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block5_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_1_relu (Activation (None, 28, 28, 128)  0           conv3_block5_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block5_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block5_concat (Concatenat (None, 28, 28, 288)  0           conv3_block4_concat[0][0]        \n",
      "                                                                 conv3_block5_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_0_bn (BatchNormali (None, 28, 28, 288)  1152        conv3_block5_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_0_relu (Activation (None, 28, 28, 288)  0           conv3_block6_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_1_conv (Conv2D)    (None, 28, 28, 128)  36864       conv3_block6_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block6_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_1_relu (Activation (None, 28, 28, 128)  0           conv3_block6_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block6_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block6_concat (Concatenat (None, 28, 28, 320)  0           conv3_block5_concat[0][0]        \n",
      "                                                                 conv3_block6_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_0_bn (BatchNormali (None, 28, 28, 320)  1280        conv3_block6_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_0_relu (Activation (None, 28, 28, 320)  0           conv3_block7_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_1_conv (Conv2D)    (None, 28, 28, 128)  40960       conv3_block7_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block7_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_1_relu (Activation (None, 28, 28, 128)  0           conv3_block7_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block7_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block7_concat (Concatenat (None, 28, 28, 352)  0           conv3_block6_concat[0][0]        \n",
      "                                                                 conv3_block7_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_0_bn (BatchNormali (None, 28, 28, 352)  1408        conv3_block7_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_0_relu (Activation (None, 28, 28, 352)  0           conv3_block8_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_1_conv (Conv2D)    (None, 28, 28, 128)  45056       conv3_block8_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block8_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_1_relu (Activation (None, 28, 28, 128)  0           conv3_block8_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block8_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block8_concat (Concatenat (None, 28, 28, 384)  0           conv3_block7_concat[0][0]        \n",
      "                                                                 conv3_block8_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_0_bn (BatchNormali (None, 28, 28, 384)  1536        conv3_block8_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_0_relu (Activation (None, 28, 28, 384)  0           conv3_block9_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_1_conv (Conv2D)    (None, 28, 28, 128)  49152       conv3_block9_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_1_bn (BatchNormali (None, 28, 28, 128)  512         conv3_block9_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_1_relu (Activation (None, 28, 28, 128)  0           conv3_block9_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_2_conv (Conv2D)    (None, 28, 28, 32)   36864       conv3_block9_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block9_concat (Concatenat (None, 28, 28, 416)  0           conv3_block8_concat[0][0]        \n",
      "                                                                 conv3_block9_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_0_bn (BatchNormal (None, 28, 28, 416)  1664        conv3_block9_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_0_relu (Activatio (None, 28, 28, 416)  0           conv3_block10_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_1_conv (Conv2D)   (None, 28, 28, 128)  53248       conv3_block10_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_1_bn (BatchNormal (None, 28, 28, 128)  512         conv3_block10_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_1_relu (Activatio (None, 28, 28, 128)  0           conv3_block10_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_2_conv (Conv2D)   (None, 28, 28, 32)   36864       conv3_block10_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block10_concat (Concatena (None, 28, 28, 448)  0           conv3_block9_concat[0][0]        \n",
      "                                                                 conv3_block10_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_0_bn (BatchNormal (None, 28, 28, 448)  1792        conv3_block10_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_0_relu (Activatio (None, 28, 28, 448)  0           conv3_block11_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_1_conv (Conv2D)   (None, 28, 28, 128)  57344       conv3_block11_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_1_bn (BatchNormal (None, 28, 28, 128)  512         conv3_block11_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_1_relu (Activatio (None, 28, 28, 128)  0           conv3_block11_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_2_conv (Conv2D)   (None, 28, 28, 32)   36864       conv3_block11_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block11_concat (Concatena (None, 28, 28, 480)  0           conv3_block10_concat[0][0]       \n",
      "                                                                 conv3_block11_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_0_bn (BatchNormal (None, 28, 28, 480)  1920        conv3_block11_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_0_relu (Activatio (None, 28, 28, 480)  0           conv3_block12_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_1_conv (Conv2D)   (None, 28, 28, 128)  61440       conv3_block12_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_1_bn (BatchNormal (None, 28, 28, 128)  512         conv3_block12_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_1_relu (Activatio (None, 28, 28, 128)  0           conv3_block12_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_2_conv (Conv2D)   (None, 28, 28, 32)   36864       conv3_block12_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv3_block12_concat (Concatena (None, 28, 28, 512)  0           conv3_block11_concat[0][0]       \n",
      "                                                                 conv3_block12_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "pool3_bn (BatchNormalization)   (None, 28, 28, 512)  2048        conv3_block12_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "pool3_relu (Activation)         (None, 28, 28, 512)  0           pool3_bn[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "pool3_conv (Conv2D)             (None, 28, 28, 256)  131072      pool3_relu[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "pool3_pool (AveragePooling2D)   (None, 14, 14, 256)  0           pool3_conv[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_0_bn (BatchNormali (None, 14, 14, 256)  1024        pool3_pool[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_0_relu (Activation (None, 14, 14, 256)  0           conv4_block1_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_1_conv (Conv2D)    (None, 14, 14, 128)  32768       conv4_block1_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block1_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_1_relu (Activation (None, 14, 14, 128)  0           conv4_block1_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block1_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block1_concat (Concatenat (None, 14, 14, 288)  0           pool3_pool[0][0]                 \n",
      "                                                                 conv4_block1_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_0_bn (BatchNormali (None, 14, 14, 288)  1152        conv4_block1_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_0_relu (Activation (None, 14, 14, 288)  0           conv4_block2_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_1_conv (Conv2D)    (None, 14, 14, 128)  36864       conv4_block2_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block2_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_1_relu (Activation (None, 14, 14, 128)  0           conv4_block2_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block2_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block2_concat (Concatenat (None, 14, 14, 320)  0           conv4_block1_concat[0][0]        \n",
      "                                                                 conv4_block2_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_0_bn (BatchNormali (None, 14, 14, 320)  1280        conv4_block2_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_0_relu (Activation (None, 14, 14, 320)  0           conv4_block3_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_1_conv (Conv2D)    (None, 14, 14, 128)  40960       conv4_block3_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block3_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_1_relu (Activation (None, 14, 14, 128)  0           conv4_block3_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block3_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block3_concat (Concatenat (None, 14, 14, 352)  0           conv4_block2_concat[0][0]        \n",
      "                                                                 conv4_block3_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_0_bn (BatchNormali (None, 14, 14, 352)  1408        conv4_block3_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_0_relu (Activation (None, 14, 14, 352)  0           conv4_block4_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_1_conv (Conv2D)    (None, 14, 14, 128)  45056       conv4_block4_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block4_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_1_relu (Activation (None, 14, 14, 128)  0           conv4_block4_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block4_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block4_concat (Concatenat (None, 14, 14, 384)  0           conv4_block3_concat[0][0]        \n",
      "                                                                 conv4_block4_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_0_bn (BatchNormali (None, 14, 14, 384)  1536        conv4_block4_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_0_relu (Activation (None, 14, 14, 384)  0           conv4_block5_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_1_conv (Conv2D)    (None, 14, 14, 128)  49152       conv4_block5_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block5_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_1_relu (Activation (None, 14, 14, 128)  0           conv4_block5_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block5_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block5_concat (Concatenat (None, 14, 14, 416)  0           conv4_block4_concat[0][0]        \n",
      "                                                                 conv4_block5_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_0_bn (BatchNormali (None, 14, 14, 416)  1664        conv4_block5_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_0_relu (Activation (None, 14, 14, 416)  0           conv4_block6_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_1_conv (Conv2D)    (None, 14, 14, 128)  53248       conv4_block6_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block6_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_1_relu (Activation (None, 14, 14, 128)  0           conv4_block6_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block6_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block6_concat (Concatenat (None, 14, 14, 448)  0           conv4_block5_concat[0][0]        \n",
      "                                                                 conv4_block6_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_0_bn (BatchNormali (None, 14, 14, 448)  1792        conv4_block6_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_0_relu (Activation (None, 14, 14, 448)  0           conv4_block7_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_1_conv (Conv2D)    (None, 14, 14, 128)  57344       conv4_block7_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block7_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_1_relu (Activation (None, 14, 14, 128)  0           conv4_block7_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block7_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block7_concat (Concatenat (None, 14, 14, 480)  0           conv4_block6_concat[0][0]        \n",
      "                                                                 conv4_block7_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_0_bn (BatchNormali (None, 14, 14, 480)  1920        conv4_block7_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_0_relu (Activation (None, 14, 14, 480)  0           conv4_block8_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_1_conv (Conv2D)    (None, 14, 14, 128)  61440       conv4_block8_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block8_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_1_relu (Activation (None, 14, 14, 128)  0           conv4_block8_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block8_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block8_concat (Concatenat (None, 14, 14, 512)  0           conv4_block7_concat[0][0]        \n",
      "                                                                 conv4_block8_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_0_bn (BatchNormali (None, 14, 14, 512)  2048        conv4_block8_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_0_relu (Activation (None, 14, 14, 512)  0           conv4_block9_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_1_conv (Conv2D)    (None, 14, 14, 128)  65536       conv4_block9_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_1_bn (BatchNormali (None, 14, 14, 128)  512         conv4_block9_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_1_relu (Activation (None, 14, 14, 128)  0           conv4_block9_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_2_conv (Conv2D)    (None, 14, 14, 32)   36864       conv4_block9_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block9_concat (Concatenat (None, 14, 14, 544)  0           conv4_block8_concat[0][0]        \n",
      "                                                                 conv4_block9_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_0_bn (BatchNormal (None, 14, 14, 544)  2176        conv4_block9_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_0_relu (Activatio (None, 14, 14, 544)  0           conv4_block10_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_1_conv (Conv2D)   (None, 14, 14, 128)  69632       conv4_block10_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block10_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block10_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block10_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block10_concat (Concatena (None, 14, 14, 576)  0           conv4_block9_concat[0][0]        \n",
      "                                                                 conv4_block10_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_0_bn (BatchNormal (None, 14, 14, 576)  2304        conv4_block10_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_0_relu (Activatio (None, 14, 14, 576)  0           conv4_block11_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_1_conv (Conv2D)   (None, 14, 14, 128)  73728       conv4_block11_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block11_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block11_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block11_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block11_concat (Concatena (None, 14, 14, 608)  0           conv4_block10_concat[0][0]       \n",
      "                                                                 conv4_block11_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_0_bn (BatchNormal (None, 14, 14, 608)  2432        conv4_block11_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_0_relu (Activatio (None, 14, 14, 608)  0           conv4_block12_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_1_conv (Conv2D)   (None, 14, 14, 128)  77824       conv4_block12_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block12_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block12_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block12_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block12_concat (Concatena (None, 14, 14, 640)  0           conv4_block11_concat[0][0]       \n",
      "                                                                 conv4_block12_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_0_bn (BatchNormal (None, 14, 14, 640)  2560        conv4_block12_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_0_relu (Activatio (None, 14, 14, 640)  0           conv4_block13_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_1_conv (Conv2D)   (None, 14, 14, 128)  81920       conv4_block13_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block13_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block13_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block13_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block13_concat (Concatena (None, 14, 14, 672)  0           conv4_block12_concat[0][0]       \n",
      "                                                                 conv4_block13_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_0_bn (BatchNormal (None, 14, 14, 672)  2688        conv4_block13_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_0_relu (Activatio (None, 14, 14, 672)  0           conv4_block14_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_1_conv (Conv2D)   (None, 14, 14, 128)  86016       conv4_block14_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block14_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block14_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block14_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block14_concat (Concatena (None, 14, 14, 704)  0           conv4_block13_concat[0][0]       \n",
      "                                                                 conv4_block14_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_0_bn (BatchNormal (None, 14, 14, 704)  2816        conv4_block14_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_0_relu (Activatio (None, 14, 14, 704)  0           conv4_block15_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_1_conv (Conv2D)   (None, 14, 14, 128)  90112       conv4_block15_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block15_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block15_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block15_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block15_concat (Concatena (None, 14, 14, 736)  0           conv4_block14_concat[0][0]       \n",
      "                                                                 conv4_block15_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_0_bn (BatchNormal (None, 14, 14, 736)  2944        conv4_block15_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_0_relu (Activatio (None, 14, 14, 736)  0           conv4_block16_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_1_conv (Conv2D)   (None, 14, 14, 128)  94208       conv4_block16_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block16_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block16_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block16_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block16_concat (Concatena (None, 14, 14, 768)  0           conv4_block15_concat[0][0]       \n",
      "                                                                 conv4_block16_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_0_bn (BatchNormal (None, 14, 14, 768)  3072        conv4_block16_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_0_relu (Activatio (None, 14, 14, 768)  0           conv4_block17_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_1_conv (Conv2D)   (None, 14, 14, 128)  98304       conv4_block17_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block17_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block17_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block17_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block17_concat (Concatena (None, 14, 14, 800)  0           conv4_block16_concat[0][0]       \n",
      "                                                                 conv4_block17_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_0_bn (BatchNormal (None, 14, 14, 800)  3200        conv4_block17_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_0_relu (Activatio (None, 14, 14, 800)  0           conv4_block18_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_1_conv (Conv2D)   (None, 14, 14, 128)  102400      conv4_block18_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block18_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block18_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block18_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block18_concat (Concatena (None, 14, 14, 832)  0           conv4_block17_concat[0][0]       \n",
      "                                                                 conv4_block18_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_0_bn (BatchNormal (None, 14, 14, 832)  3328        conv4_block18_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_0_relu (Activatio (None, 14, 14, 832)  0           conv4_block19_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_1_conv (Conv2D)   (None, 14, 14, 128)  106496      conv4_block19_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block19_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block19_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block19_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block19_concat (Concatena (None, 14, 14, 864)  0           conv4_block18_concat[0][0]       \n",
      "                                                                 conv4_block19_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_0_bn (BatchNormal (None, 14, 14, 864)  3456        conv4_block19_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_0_relu (Activatio (None, 14, 14, 864)  0           conv4_block20_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_1_conv (Conv2D)   (None, 14, 14, 128)  110592      conv4_block20_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block20_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block20_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block20_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block20_concat (Concatena (None, 14, 14, 896)  0           conv4_block19_concat[0][0]       \n",
      "                                                                 conv4_block20_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_0_bn (BatchNormal (None, 14, 14, 896)  3584        conv4_block20_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_0_relu (Activatio (None, 14, 14, 896)  0           conv4_block21_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_1_conv (Conv2D)   (None, 14, 14, 128)  114688      conv4_block21_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block21_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block21_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block21_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block21_concat (Concatena (None, 14, 14, 928)  0           conv4_block20_concat[0][0]       \n",
      "                                                                 conv4_block21_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_0_bn (BatchNormal (None, 14, 14, 928)  3712        conv4_block21_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_0_relu (Activatio (None, 14, 14, 928)  0           conv4_block22_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_1_conv (Conv2D)   (None, 14, 14, 128)  118784      conv4_block22_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block22_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block22_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block22_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block22_concat (Concatena (None, 14, 14, 960)  0           conv4_block21_concat[0][0]       \n",
      "                                                                 conv4_block22_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_0_bn (BatchNormal (None, 14, 14, 960)  3840        conv4_block22_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_0_relu (Activatio (None, 14, 14, 960)  0           conv4_block23_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_1_conv (Conv2D)   (None, 14, 14, 128)  122880      conv4_block23_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block23_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block23_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block23_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block23_concat (Concatena (None, 14, 14, 992)  0           conv4_block22_concat[0][0]       \n",
      "                                                                 conv4_block23_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_0_bn (BatchNormal (None, 14, 14, 992)  3968        conv4_block23_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_0_relu (Activatio (None, 14, 14, 992)  0           conv4_block24_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_1_conv (Conv2D)   (None, 14, 14, 128)  126976      conv4_block24_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block24_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block24_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block24_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block24_concat (Concatena (None, 14, 14, 1024) 0           conv4_block23_concat[0][0]       \n",
      "                                                                 conv4_block24_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_0_bn (BatchNormal (None, 14, 14, 1024) 4096        conv4_block24_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_0_relu (Activatio (None, 14, 14, 1024) 0           conv4_block25_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_1_conv (Conv2D)   (None, 14, 14, 128)  131072      conv4_block25_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block25_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block25_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block25_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block25_concat (Concatena (None, 14, 14, 1056) 0           conv4_block24_concat[0][0]       \n",
      "                                                                 conv4_block25_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_0_bn (BatchNormal (None, 14, 14, 1056) 4224        conv4_block25_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_0_relu (Activatio (None, 14, 14, 1056) 0           conv4_block26_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_1_conv (Conv2D)   (None, 14, 14, 128)  135168      conv4_block26_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block26_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block26_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block26_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block26_concat (Concatena (None, 14, 14, 1088) 0           conv4_block25_concat[0][0]       \n",
      "                                                                 conv4_block26_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_0_bn (BatchNormal (None, 14, 14, 1088) 4352        conv4_block26_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_0_relu (Activatio (None, 14, 14, 1088) 0           conv4_block27_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_1_conv (Conv2D)   (None, 14, 14, 128)  139264      conv4_block27_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block27_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block27_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block27_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block27_concat (Concatena (None, 14, 14, 1120) 0           conv4_block26_concat[0][0]       \n",
      "                                                                 conv4_block27_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_0_bn (BatchNormal (None, 14, 14, 1120) 4480        conv4_block27_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_0_relu (Activatio (None, 14, 14, 1120) 0           conv4_block28_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_1_conv (Conv2D)   (None, 14, 14, 128)  143360      conv4_block28_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block28_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block28_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block28_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block28_concat (Concatena (None, 14, 14, 1152) 0           conv4_block27_concat[0][0]       \n",
      "                                                                 conv4_block28_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_0_bn (BatchNormal (None, 14, 14, 1152) 4608        conv4_block28_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_0_relu (Activatio (None, 14, 14, 1152) 0           conv4_block29_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_1_conv (Conv2D)   (None, 14, 14, 128)  147456      conv4_block29_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block29_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block29_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block29_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block29_concat (Concatena (None, 14, 14, 1184) 0           conv4_block28_concat[0][0]       \n",
      "                                                                 conv4_block29_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_0_bn (BatchNormal (None, 14, 14, 1184) 4736        conv4_block29_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_0_relu (Activatio (None, 14, 14, 1184) 0           conv4_block30_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_1_conv (Conv2D)   (None, 14, 14, 128)  151552      conv4_block30_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block30_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block30_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block30_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block30_concat (Concatena (None, 14, 14, 1216) 0           conv4_block29_concat[0][0]       \n",
      "                                                                 conv4_block30_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_0_bn (BatchNormal (None, 14, 14, 1216) 4864        conv4_block30_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_0_relu (Activatio (None, 14, 14, 1216) 0           conv4_block31_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_1_conv (Conv2D)   (None, 14, 14, 128)  155648      conv4_block31_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block31_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block31_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block31_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block31_concat (Concatena (None, 14, 14, 1248) 0           conv4_block30_concat[0][0]       \n",
      "                                                                 conv4_block31_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_0_bn (BatchNormal (None, 14, 14, 1248) 4992        conv4_block31_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_0_relu (Activatio (None, 14, 14, 1248) 0           conv4_block32_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_1_conv (Conv2D)   (None, 14, 14, 128)  159744      conv4_block32_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_1_bn (BatchNormal (None, 14, 14, 128)  512         conv4_block32_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_1_relu (Activatio (None, 14, 14, 128)  0           conv4_block32_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_2_conv (Conv2D)   (None, 14, 14, 32)   36864       conv4_block32_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv4_block32_concat (Concatena (None, 14, 14, 1280) 0           conv4_block31_concat[0][0]       \n",
      "                                                                 conv4_block32_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "pool4_bn (BatchNormalization)   (None, 14, 14, 1280) 5120        conv4_block32_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "pool4_relu (Activation)         (None, 14, 14, 1280) 0           pool4_bn[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "pool4_conv (Conv2D)             (None, 14, 14, 640)  819200      pool4_relu[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "pool4_pool (AveragePooling2D)   (None, 7, 7, 640)    0           pool4_conv[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_0_bn (BatchNormali (None, 7, 7, 640)    2560        pool4_pool[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_0_relu (Activation (None, 7, 7, 640)    0           conv5_block1_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_1_conv (Conv2D)    (None, 7, 7, 128)    81920       conv5_block1_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block1_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_1_relu (Activation (None, 7, 7, 128)    0           conv5_block1_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block1_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block1_concat (Concatenat (None, 7, 7, 672)    0           pool4_pool[0][0]                 \n",
      "                                                                 conv5_block1_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_0_bn (BatchNormali (None, 7, 7, 672)    2688        conv5_block1_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_0_relu (Activation (None, 7, 7, 672)    0           conv5_block2_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_1_conv (Conv2D)    (None, 7, 7, 128)    86016       conv5_block2_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block2_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_1_relu (Activation (None, 7, 7, 128)    0           conv5_block2_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block2_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block2_concat (Concatenat (None, 7, 7, 704)    0           conv5_block1_concat[0][0]        \n",
      "                                                                 conv5_block2_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_0_bn (BatchNormali (None, 7, 7, 704)    2816        conv5_block2_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_0_relu (Activation (None, 7, 7, 704)    0           conv5_block3_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_1_conv (Conv2D)    (None, 7, 7, 128)    90112       conv5_block3_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block3_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_1_relu (Activation (None, 7, 7, 128)    0           conv5_block3_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block3_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block3_concat (Concatenat (None, 7, 7, 736)    0           conv5_block2_concat[0][0]        \n",
      "                                                                 conv5_block3_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_0_bn (BatchNormali (None, 7, 7, 736)    2944        conv5_block3_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_0_relu (Activation (None, 7, 7, 736)    0           conv5_block4_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_1_conv (Conv2D)    (None, 7, 7, 128)    94208       conv5_block4_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block4_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_1_relu (Activation (None, 7, 7, 128)    0           conv5_block4_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block4_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block4_concat (Concatenat (None, 7, 7, 768)    0           conv5_block3_concat[0][0]        \n",
      "                                                                 conv5_block4_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_0_bn (BatchNormali (None, 7, 7, 768)    3072        conv5_block4_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_0_relu (Activation (None, 7, 7, 768)    0           conv5_block5_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_1_conv (Conv2D)    (None, 7, 7, 128)    98304       conv5_block5_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block5_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_1_relu (Activation (None, 7, 7, 128)    0           conv5_block5_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block5_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block5_concat (Concatenat (None, 7, 7, 800)    0           conv5_block4_concat[0][0]        \n",
      "                                                                 conv5_block5_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_0_bn (BatchNormali (None, 7, 7, 800)    3200        conv5_block5_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_0_relu (Activation (None, 7, 7, 800)    0           conv5_block6_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_1_conv (Conv2D)    (None, 7, 7, 128)    102400      conv5_block6_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block6_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_1_relu (Activation (None, 7, 7, 128)    0           conv5_block6_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block6_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block6_concat (Concatenat (None, 7, 7, 832)    0           conv5_block5_concat[0][0]        \n",
      "                                                                 conv5_block6_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_0_bn (BatchNormali (None, 7, 7, 832)    3328        conv5_block6_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_0_relu (Activation (None, 7, 7, 832)    0           conv5_block7_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_1_conv (Conv2D)    (None, 7, 7, 128)    106496      conv5_block7_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block7_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_1_relu (Activation (None, 7, 7, 128)    0           conv5_block7_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block7_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block7_concat (Concatenat (None, 7, 7, 864)    0           conv5_block6_concat[0][0]        \n",
      "                                                                 conv5_block7_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_0_bn (BatchNormali (None, 7, 7, 864)    3456        conv5_block7_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_0_relu (Activation (None, 7, 7, 864)    0           conv5_block8_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_1_conv (Conv2D)    (None, 7, 7, 128)    110592      conv5_block8_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block8_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_1_relu (Activation (None, 7, 7, 128)    0           conv5_block8_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block8_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block8_concat (Concatenat (None, 7, 7, 896)    0           conv5_block7_concat[0][0]        \n",
      "                                                                 conv5_block8_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_0_bn (BatchNormali (None, 7, 7, 896)    3584        conv5_block8_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_0_relu (Activation (None, 7, 7, 896)    0           conv5_block9_0_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_1_conv (Conv2D)    (None, 7, 7, 128)    114688      conv5_block9_0_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_1_bn (BatchNormali (None, 7, 7, 128)    512         conv5_block9_1_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_1_relu (Activation (None, 7, 7, 128)    0           conv5_block9_1_bn[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_2_conv (Conv2D)    (None, 7, 7, 32)     36864       conv5_block9_1_relu[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block9_concat (Concatenat (None, 7, 7, 928)    0           conv5_block8_concat[0][0]        \n",
      "                                                                 conv5_block9_2_conv[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_0_bn (BatchNormal (None, 7, 7, 928)    3712        conv5_block9_concat[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_0_relu (Activatio (None, 7, 7, 928)    0           conv5_block10_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_1_conv (Conv2D)   (None, 7, 7, 128)    118784      conv5_block10_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block10_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block10_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block10_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block10_concat (Concatena (None, 7, 7, 960)    0           conv5_block9_concat[0][0]        \n",
      "                                                                 conv5_block10_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_0_bn (BatchNormal (None, 7, 7, 960)    3840        conv5_block10_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_0_relu (Activatio (None, 7, 7, 960)    0           conv5_block11_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_1_conv (Conv2D)   (None, 7, 7, 128)    122880      conv5_block11_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block11_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block11_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block11_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block11_concat (Concatena (None, 7, 7, 992)    0           conv5_block10_concat[0][0]       \n",
      "                                                                 conv5_block11_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_0_bn (BatchNormal (None, 7, 7, 992)    3968        conv5_block11_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_0_relu (Activatio (None, 7, 7, 992)    0           conv5_block12_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_1_conv (Conv2D)   (None, 7, 7, 128)    126976      conv5_block12_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block12_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block12_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block12_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block12_concat (Concatena (None, 7, 7, 1024)   0           conv5_block11_concat[0][0]       \n",
      "                                                                 conv5_block12_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_0_bn (BatchNormal (None, 7, 7, 1024)   4096        conv5_block12_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_0_relu (Activatio (None, 7, 7, 1024)   0           conv5_block13_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_1_conv (Conv2D)   (None, 7, 7, 128)    131072      conv5_block13_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block13_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block13_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block13_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block13_concat (Concatena (None, 7, 7, 1056)   0           conv5_block12_concat[0][0]       \n",
      "                                                                 conv5_block13_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_0_bn (BatchNormal (None, 7, 7, 1056)   4224        conv5_block13_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_0_relu (Activatio (None, 7, 7, 1056)   0           conv5_block14_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_1_conv (Conv2D)   (None, 7, 7, 128)    135168      conv5_block14_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block14_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block14_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block14_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block14_concat (Concatena (None, 7, 7, 1088)   0           conv5_block13_concat[0][0]       \n",
      "                                                                 conv5_block14_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_0_bn (BatchNormal (None, 7, 7, 1088)   4352        conv5_block14_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_0_relu (Activatio (None, 7, 7, 1088)   0           conv5_block15_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_1_conv (Conv2D)   (None, 7, 7, 128)    139264      conv5_block15_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block15_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block15_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block15_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block15_concat (Concatena (None, 7, 7, 1120)   0           conv5_block14_concat[0][0]       \n",
      "                                                                 conv5_block15_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_0_bn (BatchNormal (None, 7, 7, 1120)   4480        conv5_block15_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_0_relu (Activatio (None, 7, 7, 1120)   0           conv5_block16_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_1_conv (Conv2D)   (None, 7, 7, 128)    143360      conv5_block16_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block16_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block16_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block16_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block16_concat (Concatena (None, 7, 7, 1152)   0           conv5_block15_concat[0][0]       \n",
      "                                                                 conv5_block16_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_0_bn (BatchNormal (None, 7, 7, 1152)   4608        conv5_block16_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_0_relu (Activatio (None, 7, 7, 1152)   0           conv5_block17_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_1_conv (Conv2D)   (None, 7, 7, 128)    147456      conv5_block17_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block17_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block17_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block17_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block17_concat (Concatena (None, 7, 7, 1184)   0           conv5_block16_concat[0][0]       \n",
      "                                                                 conv5_block17_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_0_bn (BatchNormal (None, 7, 7, 1184)   4736        conv5_block17_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_0_relu (Activatio (None, 7, 7, 1184)   0           conv5_block18_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_1_conv (Conv2D)   (None, 7, 7, 128)    151552      conv5_block18_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block18_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block18_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block18_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block18_concat (Concatena (None, 7, 7, 1216)   0           conv5_block17_concat[0][0]       \n",
      "                                                                 conv5_block18_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_0_bn (BatchNormal (None, 7, 7, 1216)   4864        conv5_block18_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_0_relu (Activatio (None, 7, 7, 1216)   0           conv5_block19_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_1_conv (Conv2D)   (None, 7, 7, 128)    155648      conv5_block19_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block19_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block19_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block19_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block19_concat (Concatena (None, 7, 7, 1248)   0           conv5_block18_concat[0][0]       \n",
      "                                                                 conv5_block19_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_0_bn (BatchNormal (None, 7, 7, 1248)   4992        conv5_block19_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_0_relu (Activatio (None, 7, 7, 1248)   0           conv5_block20_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_1_conv (Conv2D)   (None, 7, 7, 128)    159744      conv5_block20_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block20_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block20_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block20_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block20_concat (Concatena (None, 7, 7, 1280)   0           conv5_block19_concat[0][0]       \n",
      "                                                                 conv5_block20_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_0_bn (BatchNormal (None, 7, 7, 1280)   5120        conv5_block20_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_0_relu (Activatio (None, 7, 7, 1280)   0           conv5_block21_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_1_conv (Conv2D)   (None, 7, 7, 128)    163840      conv5_block21_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block21_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block21_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block21_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block21_concat (Concatena (None, 7, 7, 1312)   0           conv5_block20_concat[0][0]       \n",
      "                                                                 conv5_block21_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_0_bn (BatchNormal (None, 7, 7, 1312)   5248        conv5_block21_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_0_relu (Activatio (None, 7, 7, 1312)   0           conv5_block22_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_1_conv (Conv2D)   (None, 7, 7, 128)    167936      conv5_block22_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block22_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block22_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block22_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block22_concat (Concatena (None, 7, 7, 1344)   0           conv5_block21_concat[0][0]       \n",
      "                                                                 conv5_block22_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_0_bn (BatchNormal (None, 7, 7, 1344)   5376        conv5_block22_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_0_relu (Activatio (None, 7, 7, 1344)   0           conv5_block23_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_1_conv (Conv2D)   (None, 7, 7, 128)    172032      conv5_block23_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block23_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block23_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block23_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block23_concat (Concatena (None, 7, 7, 1376)   0           conv5_block22_concat[0][0]       \n",
      "                                                                 conv5_block23_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_0_bn (BatchNormal (None, 7, 7, 1376)   5504        conv5_block23_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_0_relu (Activatio (None, 7, 7, 1376)   0           conv5_block24_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_1_conv (Conv2D)   (None, 7, 7, 128)    176128      conv5_block24_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block24_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block24_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block24_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block24_concat (Concatena (None, 7, 7, 1408)   0           conv5_block23_concat[0][0]       \n",
      "                                                                 conv5_block24_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_0_bn (BatchNormal (None, 7, 7, 1408)   5632        conv5_block24_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_0_relu (Activatio (None, 7, 7, 1408)   0           conv5_block25_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_1_conv (Conv2D)   (None, 7, 7, 128)    180224      conv5_block25_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block25_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block25_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block25_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block25_concat (Concatena (None, 7, 7, 1440)   0           conv5_block24_concat[0][0]       \n",
      "                                                                 conv5_block25_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_0_bn (BatchNormal (None, 7, 7, 1440)   5760        conv5_block25_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_0_relu (Activatio (None, 7, 7, 1440)   0           conv5_block26_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_1_conv (Conv2D)   (None, 7, 7, 128)    184320      conv5_block26_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block26_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block26_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block26_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block26_concat (Concatena (None, 7, 7, 1472)   0           conv5_block25_concat[0][0]       \n",
      "                                                                 conv5_block26_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_0_bn (BatchNormal (None, 7, 7, 1472)   5888        conv5_block26_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_0_relu (Activatio (None, 7, 7, 1472)   0           conv5_block27_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_1_conv (Conv2D)   (None, 7, 7, 128)    188416      conv5_block27_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block27_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block27_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block27_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block27_concat (Concatena (None, 7, 7, 1504)   0           conv5_block26_concat[0][0]       \n",
      "                                                                 conv5_block27_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_0_bn (BatchNormal (None, 7, 7, 1504)   6016        conv5_block27_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_0_relu (Activatio (None, 7, 7, 1504)   0           conv5_block28_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_1_conv (Conv2D)   (None, 7, 7, 128)    192512      conv5_block28_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block28_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block28_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block28_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block28_concat (Concatena (None, 7, 7, 1536)   0           conv5_block27_concat[0][0]       \n",
      "                                                                 conv5_block28_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_0_bn (BatchNormal (None, 7, 7, 1536)   6144        conv5_block28_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_0_relu (Activatio (None, 7, 7, 1536)   0           conv5_block29_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_1_conv (Conv2D)   (None, 7, 7, 128)    196608      conv5_block29_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block29_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block29_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block29_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block29_concat (Concatena (None, 7, 7, 1568)   0           conv5_block28_concat[0][0]       \n",
      "                                                                 conv5_block29_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_0_bn (BatchNormal (None, 7, 7, 1568)   6272        conv5_block29_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_0_relu (Activatio (None, 7, 7, 1568)   0           conv5_block30_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_1_conv (Conv2D)   (None, 7, 7, 128)    200704      conv5_block30_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block30_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block30_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block30_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block30_concat (Concatena (None, 7, 7, 1600)   0           conv5_block29_concat[0][0]       \n",
      "                                                                 conv5_block30_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_0_bn (BatchNormal (None, 7, 7, 1600)   6400        conv5_block30_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_0_relu (Activatio (None, 7, 7, 1600)   0           conv5_block31_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_1_conv (Conv2D)   (None, 7, 7, 128)    204800      conv5_block31_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block31_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block31_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block31_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block31_concat (Concatena (None, 7, 7, 1632)   0           conv5_block30_concat[0][0]       \n",
      "                                                                 conv5_block31_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_0_bn (BatchNormal (None, 7, 7, 1632)   6528        conv5_block31_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_0_relu (Activatio (None, 7, 7, 1632)   0           conv5_block32_0_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_1_conv (Conv2D)   (None, 7, 7, 128)    208896      conv5_block32_0_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_1_bn (BatchNormal (None, 7, 7, 128)    512         conv5_block32_1_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_1_relu (Activatio (None, 7, 7, 128)    0           conv5_block32_1_bn[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_2_conv (Conv2D)   (None, 7, 7, 32)     36864       conv5_block32_1_relu[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "conv5_block32_concat (Concatena (None, 7, 7, 1664)   0           conv5_block31_concat[0][0]       \n",
      "                                                                 conv5_block32_2_conv[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "bn (BatchNormalization)         (None, 7, 7, 1664)   6656        conv5_block32_concat[0][0]       \n",
      "__________________________________________________________________________________________________\n",
      "relu (Activation)               (None, 7, 7, 1664)   0           bn[0][0]                         \n",
      "__________________________________________________________________________________________________\n",
      "global_average_pooling2d_1 (Glo (None, 1664)         0           relu[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 256)          426240      global_average_pooling2d_1[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 5)            1285        dense_1[0][0]                    \n",
      "==================================================================================================\n",
      "Total params: 13,070,405\n",
      "Trainable params: 12,912,005\n",
      "Non-trainable params: 158,400\n",
      "__________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "mymodel = buildModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "EPOCHS = 150\n",
    "earlystop = keras.callbacks.EarlyStopping(patience=30)\n",
    "learning_rate_reduction = keras.callbacks.ReduceLROnPlateau(monitor='val_acc', \n",
    "                                            patience=2, \n",
    "                                            verbose=1, \n",
    "                                            factor=0.5, \n",
    "                                            min_lr=0.00001)\n",
    "checkpoint = keras.callbacks.ModelCheckpoint('../working/DenseNet121.h5', monitor='val_loss', verbose=1, \n",
    "                             save_best_only=True, mode='min', save_weights_only = True)\n",
    "qwk = QWKCallback((X_val,Y_val))\n",
    "mycallbacks = [earlystop, learning_rate_reduction,checkpoint,qwk]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<__main__.QWKCallback object at 0x7cb095575198>\n"
     ]
    }
   ],
   "source": [
    "print(qwk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/150\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.6/site-packages/keras/engine/training_generator.py:47: UserWarning: Using a generator with `use_multiprocessing=True` and multiple workers may duplicate your data. Please consider using the`keras.utils.Sequence class.\n",
      "  UserWarning('Using a generator with `use_multiprocessing=True`'\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 74s - loss: 1.2347 - acc: 0.6733 - val_loss: 3.0171 - val_acc: 0.2582\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 3.01715, saving model to ../working/DenseNet121.h5\n",
      "Epoch 0 : QWK: 0.41940901884067194\n",
      "saving checkpoint:  0.41940901884067194\n",
      "Epoch 2/150\n",
      " - 38s - loss: 1.2373 - acc: 0.6353 - val_loss: 2.6802 - val_acc: 0.3909\n",
      "\n",
      "Epoch 00002: val_loss improved from 3.01715 to 2.68016, saving model to ../working/DenseNet121.h5\n",
      "Epoch 1 : QWK: 0.4702088821889042\n",
      "saving checkpoint:  0.4702088821889042\n",
      "Epoch 3/150\n",
      " - 37s - loss: 1.1438 - acc: 0.6820 - val_loss: 1.1282 - val_acc: 0.5745\n",
      "\n",
      "Epoch 00003: val_loss improved from 2.68016 to 1.12825, saving model to ../working/DenseNet121.h5\n",
      "Epoch 2 : QWK: 0.6318130581666456\n",
      "saving checkpoint:  0.6318130581666456\n",
      "Epoch 4/150\n",
      " - 37s - loss: 1.0728 - acc: 0.7297 - val_loss: 1.0815 - val_acc: 0.6800\n",
      "\n",
      "Epoch 00004: val_loss improved from 1.12825 to 1.08152, saving model to ../working/DenseNet121.h5\n",
      "Epoch 3 : QWK: 0.6452656041318733\n",
      "saving checkpoint:  0.6452656041318733\n",
      "Epoch 5/150\n",
      " - 37s - loss: 1.1256 - acc: 0.7030 - val_loss: 0.9850 - val_acc: 0.5745\n",
      "\n",
      "Epoch 00005: val_loss improved from 1.08152 to 0.98504, saving model to ../working/DenseNet121.h5\n",
      "Epoch 4 : QWK: 0.6406435180456456\n",
      "Epoch 6/150\n",
      " - 37s - loss: 1.0395 - acc: 0.7365 - val_loss: 0.8831 - val_acc: 0.6836\n",
      "\n",
      "Epoch 00006: val_loss improved from 0.98504 to 0.88314, saving model to ../working/DenseNet121.h5\n",
      "Epoch 5 : QWK: 0.6735871158181778\n",
      "saving checkpoint:  0.6735871158181778\n",
      "Epoch 7/150\n",
      " - 37s - loss: 1.0393 - acc: 0.7458 - val_loss: 0.8506 - val_acc: 0.6745\n",
      "\n",
      "Epoch 00007: val_loss improved from 0.88314 to 0.85060, saving model to ../working/DenseNet121.h5\n",
      "Epoch 6 : QWK: 0.5946880619726103\n",
      "Epoch 8/150\n",
      " - 37s - loss: 0.9695 - acc: 0.7590 - val_loss: 0.7232 - val_acc: 0.7436\n",
      "\n",
      "Epoch 00008: val_loss improved from 0.85060 to 0.72322, saving model to ../working/DenseNet121.h5\n",
      "Epoch 7 : QWK: 0.7871691605589964\n",
      "saving checkpoint:  0.7871691605589964\n",
      "Epoch 9/150\n",
      " - 37s - loss: 0.8217 - acc: 0.8154 - val_loss: 0.8649 - val_acc: 0.7127\n",
      "\n",
      "Epoch 00009: val_loss did not improve from 0.72322\n",
      "Epoch 8 : QWK: 0.8245822988391817\n",
      "saving checkpoint:  0.8245822988391817\n",
      "Epoch 10/150\n",
      " - 38s - loss: 0.9274 - acc: 0.7893 - val_loss: 0.7154 - val_acc: 0.7509\n",
      "\n",
      "Epoch 00010: val_loss improved from 0.72322 to 0.71537, saving model to ../working/DenseNet121.h5\n",
      "Epoch 9 : QWK: 0.8289241622574957\n",
      "saving checkpoint:  0.8289241622574957\n",
      "Epoch 11/150\n",
      " - 37s - loss: 0.9309 - acc: 0.7751 - val_loss: 1.1802 - val_acc: 0.6473\n",
      "\n",
      "Epoch 00011: val_loss did not improve from 0.71537\n",
      "Epoch 10 : QWK: 0.5607930686457363\n",
      "Epoch 12/150\n",
      " - 37s - loss: 1.0126 - acc: 0.7403 - val_loss: 0.9385 - val_acc: 0.7164\n",
      "\n",
      "Epoch 00012: ReduceLROnPlateau reducing learning rate to 0.004999999888241291.\n",
      "\n",
      "Epoch 00012: val_loss did not improve from 0.71537\n",
      "Epoch 11 : QWK: 0.7534819065610436\n",
      "Epoch 13/150\n",
      " - 38s - loss: 1.0890 - acc: 0.7313 - val_loss: 1.0284 - val_acc: 0.6564\n",
      "\n",
      "Epoch 00013: val_loss did not improve from 0.71537\n",
      "Epoch 12 : QWK: 0.6899550697601093\n",
      "Epoch 14/150\n",
      " - 37s - loss: 1.0369 - acc: 0.7390 - val_loss: 0.7232 - val_acc: 0.6909\n",
      "\n",
      "Epoch 00014: ReduceLROnPlateau reducing learning rate to 0.0024999999441206455.\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.71537\n",
      "Epoch 13 : QWK: 0.8054414937599079\n",
      "Epoch 15/150\n",
      " - 37s - loss: 0.9327 - acc: 0.7626 - val_loss: 0.6359 - val_acc: 0.7691\n",
      "\n",
      "Epoch 00015: val_loss improved from 0.71537 to 0.63587, saving model to ../working/DenseNet121.h5\n",
      "Epoch 14 : QWK: 0.8649685116846487\n",
      "saving checkpoint:  0.8649685116846487\n",
      "Epoch 16/150\n",
      " - 37s - loss: 0.8436 - acc: 0.7870 - val_loss: 0.7357 - val_acc: 0.7327\n",
      "\n",
      "Epoch 00016: val_loss did not improve from 0.63587\n",
      "Epoch 15 : QWK: 0.8501906664245505\n",
      "Epoch 17/150\n",
      " - 38s - loss: 0.8938 - acc: 0.7935 - val_loss: 0.6509 - val_acc: 0.7636\n",
      "\n",
      "Epoch 00017: ReduceLROnPlateau reducing learning rate to 0.0012499999720603228.\n",
      "\n",
      "Epoch 00017: val_loss did not improve from 0.63587\n",
      "Epoch 16 : QWK: 0.879702310356726\n",
      "saving checkpoint:  0.879702310356726\n",
      "Epoch 18/150\n",
      " - 37s - loss: 0.8784 - acc: 0.7777 - val_loss: 0.5857 - val_acc: 0.8018\n",
      "\n",
      "Epoch 00018: val_loss improved from 0.63587 to 0.58567, saving model to ../working/DenseNet121.h5\n",
      "Epoch 17 : QWK: 0.8883053438559477\n",
      "saving checkpoint:  0.8883053438559477\n",
      "Epoch 19/150\n",
      " - 37s - loss: 0.8093 - acc: 0.7893 - val_loss: 0.6538 - val_acc: 0.7764\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 0.58567\n",
      "Epoch 18 : QWK: 0.8738416858949563\n",
      "Epoch 20/150\n",
      " - 37s - loss: 0.8415 - acc: 0.7986 - val_loss: 0.6275 - val_acc: 0.7964\n",
      "\n",
      "Epoch 00020: ReduceLROnPlateau reducing learning rate to 0.0006249999860301614.\n",
      "\n",
      "Epoch 00020: val_loss did not improve from 0.58567\n",
      "Epoch 19 : QWK: 0.8854780508020579\n",
      "Epoch 21/150\n",
      " - 38s - loss: 0.7897 - acc: 0.8083 - val_loss: 0.6256 - val_acc: 0.7855\n",
      "\n",
      "Epoch 00021: val_loss did not improve from 0.58567\n",
      "Epoch 20 : QWK: 0.8895415448117042\n",
      "saving checkpoint:  0.8895415448117042\n",
      "Epoch 22/150\n",
      " - 38s - loss: 0.7986 - acc: 0.8090 - val_loss: 0.6748 - val_acc: 0.7673\n",
      "\n",
      "Epoch 00022: ReduceLROnPlateau reducing learning rate to 0.0003124999930150807.\n",
      "\n",
      "Epoch 00022: val_loss did not improve from 0.58567\n",
      "Epoch 21 : QWK: 0.8933323998932851\n",
      "saving checkpoint:  0.8933323998932851\n",
      "Epoch 23/150\n",
      " - 38s - loss: 0.7325 - acc: 0.8170 - val_loss: 0.6574 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00023: val_loss did not improve from 0.58567\n",
      "Epoch 22 : QWK: 0.8870040225415935\n",
      "Epoch 24/150\n",
      " - 38s - loss: 0.7855 - acc: 0.8080 - val_loss: 0.6757 - val_acc: 0.7673\n",
      "\n",
      "Epoch 00024: ReduceLROnPlateau reducing learning rate to 0.00015624999650754035.\n",
      "\n",
      "Epoch 00024: val_loss did not improve from 0.58567\n",
      "Epoch 23 : QWK: 0.8941272465039486\n",
      "saving checkpoint:  0.8941272465039486\n",
      "Epoch 25/150\n",
      " - 37s - loss: 0.7709 - acc: 0.8035 - val_loss: 0.6429 - val_acc: 0.7727\n",
      "\n",
      "Epoch 00025: val_loss did not improve from 0.58567\n",
      "Epoch 24 : QWK: 0.8875761416131802\n",
      "Epoch 26/150\n",
      " - 37s - loss: 0.8272 - acc: 0.8044 - val_loss: 0.6634 - val_acc: 0.7709\n",
      "\n",
      "Epoch 00026: ReduceLROnPlateau reducing learning rate to 7.812499825377017e-05.\n",
      "\n",
      "Epoch 00026: val_loss did not improve from 0.58567\n",
      "Epoch 25 : QWK: 0.8894439712906809\n",
      "Epoch 27/150\n",
      " - 37s - loss: 0.7483 - acc: 0.7935 - val_loss: 0.6397 - val_acc: 0.7782\n",
      "\n",
      "Epoch 00027: val_loss did not improve from 0.58567\n",
      "Epoch 26 : QWK: 0.8917512942570299\n",
      "Epoch 28/150\n",
      " - 38s - loss: 0.7677 - acc: 0.8044 - val_loss: 0.6304 - val_acc: 0.7873\n",
      "\n",
      "Epoch 00028: ReduceLROnPlateau reducing learning rate to 3.9062499126885086e-05.\n",
      "\n",
      "Epoch 00028: val_loss did not improve from 0.58567\n",
      "Epoch 27 : QWK: 0.9003927204007913\n",
      "saving checkpoint:  0.9003927204007913\n",
      "Epoch 29/150\n",
      " - 37s - loss: 0.7879 - acc: 0.8086 - val_loss: 0.6257 - val_acc: 0.7855\n",
      "\n",
      "Epoch 00029: val_loss did not improve from 0.58567\n",
      "Epoch 28 : QWK: 0.9012978202693456\n",
      "saving checkpoint:  0.9012978202693456\n",
      "Epoch 30/150\n",
      " - 38s - loss: 0.8602 - acc: 0.7961 - val_loss: 0.6289 - val_acc: 0.7873\n",
      "\n",
      "Epoch 00030: ReduceLROnPlateau reducing learning rate to 1.9531249563442543e-05.\n",
      "\n",
      "Epoch 00030: val_loss did not improve from 0.58567\n",
      "Epoch 29 : QWK: 0.9034332012239028\n",
      "saving checkpoint:  0.9034332012239028\n",
      "Epoch 31/150\n",
      " - 38s - loss: 0.7654 - acc: 0.8160 - val_loss: 0.6296 - val_acc: 0.7836\n",
      "\n",
      "Epoch 00031: val_loss did not improve from 0.58567\n",
      "Epoch 30 : QWK: 0.901255570009948\n",
      "Epoch 32/150\n",
      " - 37s - loss: 0.7474 - acc: 0.8090 - val_loss: 0.6347 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00032: ReduceLROnPlateau reducing learning rate to 1e-05.\n",
      "\n",
      "Epoch 00032: val_loss did not improve from 0.58567\n",
      "Epoch 31 : QWK: 0.8997811457982137\n",
      "Epoch 33/150\n",
      " - 37s - loss: 0.7828 - acc: 0.8051 - val_loss: 0.6339 - val_acc: 0.7873\n",
      "\n",
      "Epoch 00033: val_loss did not improve from 0.58567\n",
      "Epoch 32 : QWK: 0.9024044485140574\n",
      "Epoch 34/150\n",
      " - 37s - loss: 0.8025 - acc: 0.8028 - val_loss: 0.6300 - val_acc: 0.7836\n",
      "\n",
      "Epoch 00034: val_loss did not improve from 0.58567\n",
      "Epoch 33 : QWK: 0.8999599057295691\n",
      "Epoch 35/150\n",
      " - 37s - loss: 0.7637 - acc: 0.8173 - val_loss: 0.6324 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00035: val_loss did not improve from 0.58567\n",
      "Epoch 34 : QWK: 0.8980376246503384\n",
      "Epoch 36/150\n",
      " - 37s - loss: 0.7525 - acc: 0.8128 - val_loss: 0.6303 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00036: val_loss did not improve from 0.58567\n",
      "Epoch 35 : QWK: 0.8988232514818278\n",
      "Epoch 37/150\n",
      " - 38s - loss: 0.7904 - acc: 0.8164 - val_loss: 0.6343 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00037: val_loss did not improve from 0.58567\n",
      "Epoch 36 : QWK: 0.8968757309065464\n",
      "Epoch 38/150\n",
      " - 37s - loss: 0.7542 - acc: 0.8148 - val_loss: 0.6298 - val_acc: 0.7836\n",
      "\n",
      "Epoch 00038: val_loss did not improve from 0.58567\n",
      "Epoch 37 : QWK: 0.9003524193029929\n",
      "Epoch 39/150\n",
      " - 38s - loss: 0.7676 - acc: 0.8183 - val_loss: 0.6302 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00039: val_loss did not improve from 0.58567\n",
      "Epoch 38 : QWK: 0.8986072812585505\n",
      "Epoch 40/150\n",
      " - 38s - loss: 0.7916 - acc: 0.8170 - val_loss: 0.6300 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00040: val_loss did not improve from 0.58567\n",
      "Epoch 39 : QWK: 0.8981372271881001\n",
      "Epoch 41/150\n",
      " - 37s - loss: 0.7505 - acc: 0.8054 - val_loss: 0.6370 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00041: val_loss did not improve from 0.58567\n",
      "Epoch 40 : QWK: 0.8968757309065464\n",
      "Epoch 42/150\n",
      " - 37s - loss: 0.7162 - acc: 0.8151 - val_loss: 0.6345 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00042: val_loss did not improve from 0.58567\n",
      "Epoch 41 : QWK: 0.8980376246503384\n",
      "Epoch 43/150\n",
      " - 37s - loss: 0.7451 - acc: 0.8090 - val_loss: 0.6338 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00043: val_loss did not improve from 0.58567\n",
      "Epoch 42 : QWK: 0.8980376246503384\n",
      "Epoch 44/150\n",
      " - 37s - loss: 0.7433 - acc: 0.8215 - val_loss: 0.6296 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00044: val_loss did not improve from 0.58567\n",
      "Epoch 43 : QWK: 0.8986072812585505\n",
      "Epoch 45/150\n",
      " - 38s - loss: 0.8009 - acc: 0.8073 - val_loss: 0.6314 - val_acc: 0.7855\n",
      "\n",
      "Epoch 00045: val_loss did not improve from 0.58567\n",
      "Epoch 44 : QWK: 0.9021113056063524\n",
      "Epoch 46/150\n",
      " - 37s - loss: 0.7590 - acc: 0.8115 - val_loss: 0.6302 - val_acc: 0.7818\n",
      "\n",
      "Epoch 00046: val_loss did not improve from 0.58567\n",
      "Epoch 45 : QWK: 0.8984622619693075\n",
      "Epoch 47/150\n",
      " - 37s - loss: 0.7627 - acc: 0.8122 - val_loss: 0.6322 - val_acc: 0.7836\n",
      "\n",
      "Epoch 00047: val_loss did not improve from 0.58567\n",
      "Epoch 46 : QWK: 0.9003524193029929\n",
      "Epoch 48/150\n",
      " - 37s - loss: 0.7114 - acc: 0.8196 - val_loss: 0.6325 - val_acc: 0.7782\n",
      "\n",
      "Epoch 00048: val_loss did not improve from 0.58567\n",
      "Epoch 47 : QWK: 0.8972108715077491\n"
     ]
    }
   ],
   "source": [
    "# Warm up the model with class weights\n",
    "EPOCHS = 150\n",
    "history = mymodel.fit_generator(training_generator,steps_per_epoch = X_train.shape[0] // batch_size,epochs = EPOCHS,\n",
    "                         validation_data = (X_val,Y_val),\n",
    "                         validation_steps = 30,\n",
    "                         workers = 2,use_multiprocessing=True,\n",
    "                         verbose=2, callbacks=mycallbacks,\n",
    "                         class_weight=cls_wt_dict)\n",
    "# "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/150\n",
      " - 42s - loss: 0.6220 - acc: 0.8293 - val_loss: 0.6273 - val_acc: 0.7800\n",
      "\n",
      "Epoch 00001: val_loss did not improve from 0.58567\n",
      "Epoch 0 : QWK: 0.8968031212461772\n",
      "Epoch 2/150\n",
      " - 38s - loss: 0.6018 - acc: 0.8177 - val_loss: 0.6208 - val_acc: 0.7964\n",
      "\n",
      "Epoch 00002: val_loss did not improve from 0.58567\n",
      "Epoch 1 : QWK: 0.9041726476265901\n",
      "saving checkpoint:  0.9041726476265901\n",
      "Epoch 3/150\n",
      " - 38s - loss: 0.6379 - acc: 0.8090 - val_loss: 0.6133 - val_acc: 0.7964\n",
      "\n",
      "Epoch 00003: val_loss did not improve from 0.58567\n",
      "Epoch 2 : QWK: 0.9058053224200162\n",
      "saving checkpoint:  0.9058053224200162\n",
      "Epoch 4/150\n",
      " - 38s - loss: 0.6005 - acc: 0.8289 - val_loss: 0.6087 - val_acc: 0.8000\n",
      "\n",
      "Epoch 00004: val_loss did not improve from 0.58567\n",
      "Epoch 3 : QWK: 0.9066236172253365\n",
      "saving checkpoint:  0.9066236172253365\n",
      "Epoch 5/150\n",
      " - 38s - loss: 0.6029 - acc: 0.8276 - val_loss: 0.6057 - val_acc: 0.7982\n",
      "\n",
      "Epoch 00005: val_loss did not improve from 0.58567\n",
      "Epoch 4 : QWK: 0.904612493830782\n",
      "Epoch 6/150\n",
      " - 38s - loss: 0.5971 - acc: 0.8247 - val_loss: 0.6016 - val_acc: 0.7982\n",
      "\n",
      "Epoch 00006: val_loss did not improve from 0.58567\n",
      "Epoch 5 : QWK: 0.9029780486596304\n",
      "Epoch 7/150\n",
      " - 38s - loss: 0.6010 - acc: 0.8293 - val_loss: 0.5947 - val_acc: 0.8018\n",
      "\n",
      "Epoch 00007: val_loss did not improve from 0.58567\n",
      "Epoch 6 : QWK: 0.9051521851381499\n",
      "Epoch 8/150\n",
      " - 37s - loss: 0.5921 - acc: 0.8386 - val_loss: 0.5916 - val_acc: 0.8000\n",
      "\n",
      "Epoch 00008: val_loss did not improve from 0.58567\n",
      "Epoch 7 : QWK: 0.9046385332071147\n",
      "Epoch 9/150\n",
      " - 38s - loss: 0.5553 - acc: 0.8447 - val_loss: 0.5898 - val_acc: 0.8055\n",
      "\n",
      "Epoch 00009: val_loss did not improve from 0.58567\n",
      "Epoch 8 : QWK: 0.9091547712152962\n",
      "saving checkpoint:  0.9091547712152962\n",
      "Epoch 10/150\n",
      " - 38s - loss: 0.5589 - acc: 0.8499 - val_loss: 0.5861 - val_acc: 0.8018\n",
      "\n",
      "Epoch 00010: val_loss did not improve from 0.58567\n",
      "Epoch 9 : QWK: 0.9055210086920672\n",
      "Epoch 11/150\n",
      " - 38s - loss: 0.5507 - acc: 0.8505 - val_loss: 0.5848 - val_acc: 0.8036\n",
      "\n",
      "Epoch 00011: val_loss improved from 0.58567 to 0.58476, saving model to ../working/DenseNet121.h5\n",
      "Epoch 10 : QWK: 0.907665985093862\n",
      "Epoch 12/150\n",
      " - 38s - loss: 0.5367 - acc: 0.8611 - val_loss: 0.5839 - val_acc: 0.8073\n",
      "\n",
      "Epoch 00012: val_loss improved from 0.58476 to 0.58388, saving model to ../working/DenseNet121.h5\n",
      "Epoch 11 : QWK: 0.9088384969753467\n",
      "Epoch 13/150\n",
      " - 38s - loss: 0.5877 - acc: 0.8412 - val_loss: 0.5793 - val_acc: 0.8055\n",
      "\n",
      "Epoch 00013: val_loss improved from 0.58388 to 0.57926, saving model to ../working/DenseNet121.h5\n",
      "Epoch 12 : QWK: 0.9082524554778209\n",
      "Epoch 14/150\n",
      " - 38s - loss: 0.5652 - acc: 0.8370 - val_loss: 0.5817 - val_acc: 0.8109\n",
      "\n",
      "Epoch 00014: val_loss did not improve from 0.57926\n",
      "Epoch 13 : QWK: 0.908049172039697\n",
      "Epoch 15/150\n",
      " - 38s - loss: 0.5463 - acc: 0.8595 - val_loss: 0.5793 - val_acc: 0.8127\n",
      "\n",
      "Epoch 00015: val_loss did not improve from 0.57926\n",
      "Epoch 14 : QWK: 0.9106274136115151\n",
      "saving checkpoint:  0.9106274136115151\n",
      "Epoch 16/150\n",
      " - 38s - loss: 0.5289 - acc: 0.8747 - val_loss: 0.5778 - val_acc: 0.8091\n",
      "\n",
      "Epoch 00016: val_loss improved from 0.57926 to 0.57782, saving model to ../working/DenseNet121.h5\n",
      "Epoch 15 : QWK: 0.9094579320673505\n",
      "Epoch 17/150\n",
      " - 37s - loss: 0.5065 - acc: 0.8776 - val_loss: 0.5765 - val_acc: 0.8145\n",
      "\n",
      "Epoch 00017: val_loss improved from 0.57782 to 0.57655, saving model to ../working/DenseNet121.h5\n",
      "Epoch 16 : QWK: 0.9101564323124345\n",
      "Epoch 18/150\n",
      " - 38s - loss: 0.5498 - acc: 0.8566 - val_loss: 0.5788 - val_acc: 0.8127\n",
      "\n",
      "Epoch 00018: val_loss did not improve from 0.57655\n",
      "Epoch 17 : QWK: 0.9082990612653215\n",
      "Epoch 19/150\n",
      " - 38s - loss: 0.5242 - acc: 0.8657 - val_loss: 0.5768 - val_acc: 0.8127\n",
      "\n",
      "Epoch 00019: val_loss did not improve from 0.57655\n",
      "Epoch 18 : QWK: 0.909039517443364\n",
      "Epoch 20/150\n",
      " - 38s - loss: 0.5162 - acc: 0.8773 - val_loss: 0.5751 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00020: val_loss improved from 0.57655 to 0.57511, saving model to ../working/DenseNet121.h5\n",
      "Epoch 19 : QWK: 0.9094771789595277\n",
      "Epoch 21/150\n",
      " - 38s - loss: 0.5177 - acc: 0.8837 - val_loss: 0.5761 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00021: val_loss did not improve from 0.57511\n",
      "Epoch 20 : QWK: 0.9097324798949614\n",
      "Epoch 22/150\n",
      " - 37s - loss: 0.5212 - acc: 0.8689 - val_loss: 0.5750 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00022: val_loss improved from 0.57511 to 0.57497, saving model to ../working/DenseNet121.h5\n",
      "Epoch 21 : QWK: 0.9103227570561964\n",
      "Epoch 23/150\n",
      " - 38s - loss: 0.5173 - acc: 0.8666 - val_loss: 0.5765 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00023: val_loss did not improve from 0.57497\n",
      "Epoch 22 : QWK: 0.9092756421543923\n",
      "Epoch 24/150\n",
      " - 38s - loss: 0.5586 - acc: 0.8515 - val_loss: 0.5752 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00024: val_loss did not improve from 0.57497\n",
      "Epoch 23 : QWK: 0.9097990066754809\n",
      "Epoch 25/150\n",
      " - 37s - loss: 0.5559 - acc: 0.8476 - val_loss: 0.5749 - val_acc: 0.8145\n",
      "\n",
      "Epoch 00025: val_loss improved from 0.57497 to 0.57493, saving model to ../working/DenseNet121.h5\n",
      "Epoch 24 : QWK: 0.9070842564211506\n",
      "Epoch 26/150\n",
      " - 38s - loss: 0.5791 - acc: 0.8441 - val_loss: 0.5777 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00026: val_loss did not improve from 0.57493\n",
      "Epoch 25 : QWK: 0.9076746344906532\n",
      "Epoch 27/150\n",
      " - 38s - loss: 0.5579 - acc: 0.8431 - val_loss: 0.5743 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00027: val_loss improved from 0.57493 to 0.57426, saving model to ../working/DenseNet121.h5\n",
      "Epoch 26 : QWK: 0.9076746344906532\n",
      "Epoch 28/150\n",
      " - 38s - loss: 0.5766 - acc: 0.8450 - val_loss: 0.5728 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00028: val_loss improved from 0.57426 to 0.57279, saving model to ../working/DenseNet121.h5\n",
      "Epoch 27 : QWK: 0.9098654354683052\n",
      "Epoch 29/150\n",
      " - 38s - loss: 0.5913 - acc: 0.8425 - val_loss: 0.5727 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00029: val_loss improved from 0.57279 to 0.57271, saving model to ../working/DenseNet121.h5\n",
      "Epoch 28 : QWK: 0.9092756421543923\n",
      "Epoch 30/150\n",
      " - 38s - loss: 0.5987 - acc: 0.8235 - val_loss: 0.5735 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00030: val_loss did not improve from 0.57271\n",
      "Epoch 29 : QWK: 0.9106757996277903\n",
      "saving checkpoint:  0.9106757996277903\n",
      "Epoch 31/150\n",
      " - 38s - loss: 0.5772 - acc: 0.8537 - val_loss: 0.5735 - val_acc: 0.8255\n",
      "\n",
      "Epoch 00031: val_loss did not improve from 0.57271\n",
      "Epoch 30 : QWK: 0.9111046550881077\n",
      "saving checkpoint:  0.9111046550881077\n",
      "Epoch 32/150\n",
      " - 38s - loss: 0.5601 - acc: 0.8425 - val_loss: 0.5738 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00032: val_loss did not improve from 0.57271\n",
      "Epoch 31 : QWK: 0.9093626085665051\n",
      "Epoch 33/150\n",
      " - 38s - loss: 0.5943 - acc: 0.8360 - val_loss: 0.5740 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00033: val_loss did not improve from 0.57271\n",
      "Epoch 32 : QWK: 0.910122509464795\n",
      "Epoch 34/150\n",
      " - 37s - loss: 0.5833 - acc: 0.8376 - val_loss: 0.5697 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00034: val_loss improved from 0.57271 to 0.56972, saving model to ../working/DenseNet121.h5\n",
      "Epoch 33 : QWK: 0.9123604457564274\n",
      "saving checkpoint:  0.9123604457564274\n",
      "Epoch 35/150\n",
      " - 38s - loss: 0.5790 - acc: 0.8386 - val_loss: 0.5741 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00035: val_loss did not improve from 0.56972\n",
      "Epoch 34 : QWK: 0.9057632919888152\n",
      "Epoch 36/150\n",
      " - 39s - loss: 0.5665 - acc: 0.8380 - val_loss: 0.5702 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00036: val_loss did not improve from 0.56972\n",
      "Epoch 35 : QWK: 0.9124255668352151\n",
      "saving checkpoint:  0.9124255668352151\n",
      "Epoch 37/150\n",
      " - 39s - loss: 0.5882 - acc: 0.8463 - val_loss: 0.5734 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00037: val_loss did not improve from 0.56972\n",
      "Epoch 36 : QWK: 0.908989460131935\n",
      "Epoch 38/150\n",
      " - 37s - loss: 0.5628 - acc: 0.8489 - val_loss: 0.5688 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00038: val_loss improved from 0.56972 to 0.56881, saving model to ../working/DenseNet121.h5\n",
      "Epoch 37 : QWK: 0.9110740063575221\n",
      "Epoch 39/150\n",
      " - 38s - loss: 0.5882 - acc: 0.8360 - val_loss: 0.5700 - val_acc: 0.8273\n",
      "\n",
      "Epoch 00039: val_loss did not improve from 0.56881\n",
      "Epoch 38 : QWK: 0.9133867124786534\n",
      "saving checkpoint:  0.9133867124786534\n",
      "Epoch 40/150\n",
      " - 38s - loss: 0.5925 - acc: 0.8450 - val_loss: 0.5701 - val_acc: 0.8255\n",
      "\n",
      "Epoch 00040: val_loss did not improve from 0.56881\n",
      "Epoch 39 : QWK: 0.9128898498537775\n",
      "Epoch 41/150\n",
      " - 39s - loss: 0.5770 - acc: 0.8357 - val_loss: 0.5719 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00041: val_loss did not improve from 0.56881\n",
      "Epoch 40 : QWK: 0.9080269422784184\n",
      "Epoch 42/150\n",
      " - 37s - loss: 0.5468 - acc: 0.8450 - val_loss: 0.5698 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00042: val_loss did not improve from 0.56881\n",
      "Epoch 41 : QWK: 0.9119956950760394\n",
      "Epoch 43/150\n",
      " - 37s - loss: 0.5774 - acc: 0.8312 - val_loss: 0.5699 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00043: val_loss did not improve from 0.56881\n",
      "Epoch 42 : QWK: 0.9115015408304801\n",
      "Epoch 44/150\n",
      " - 39s - loss: 0.5507 - acc: 0.8386 - val_loss: 0.5671 - val_acc: 0.8273\n",
      "\n",
      "Epoch 00044: val_loss improved from 0.56881 to 0.56708, saving model to ../working/DenseNet121.h5\n",
      "Epoch 43 : QWK: 0.9133867124786534\n",
      "saving checkpoint:  0.9133867124786534\n",
      "Epoch 45/150\n",
      " - 39s - loss: 0.5820 - acc: 0.8286 - val_loss: 0.5692 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00045: val_loss did not improve from 0.56708\n",
      "Epoch 44 : QWK: 0.9122618645887658\n",
      "Epoch 46/150\n",
      " - 39s - loss: 0.5716 - acc: 0.8418 - val_loss: 0.5670 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00046: val_loss improved from 0.56708 to 0.56698, saving model to ../working/DenseNet121.h5\n",
      "Epoch 45 : QWK: 0.9089716487046864\n",
      "Epoch 47/150\n",
      " - 38s - loss: 0.5806 - acc: 0.8370 - val_loss: 0.5685 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00047: val_loss did not improve from 0.56698\n",
      "Epoch 46 : QWK: 0.9088713464677863\n",
      "Epoch 48/150\n",
      " - 38s - loss: 0.5355 - acc: 0.8518 - val_loss: 0.5668 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00048: val_loss improved from 0.56698 to 0.56684, saving model to ../working/DenseNet121.h5\n",
      "Epoch 47 : QWK: 0.9089716487046864\n",
      "Epoch 49/150\n",
      " - 38s - loss: 0.5705 - acc: 0.8479 - val_loss: 0.5704 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00049: val_loss did not improve from 0.56684\n",
      "Epoch 48 : QWK: 0.9097207920509438\n",
      "Epoch 50/150\n",
      " - 38s - loss: 0.5718 - acc: 0.8412 - val_loss: 0.5690 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00050: val_loss did not improve from 0.56684\n",
      "Epoch 49 : QWK: 0.9107745928484995\n",
      "Epoch 51/150\n",
      " - 38s - loss: 0.5617 - acc: 0.8428 - val_loss: 0.5675 - val_acc: 0.8255\n",
      "\n",
      "Epoch 00051: val_loss did not improve from 0.56684\n",
      "Epoch 50 : QWK: 0.9099213655737872\n",
      "Epoch 52/150\n",
      " - 38s - loss: 0.5612 - acc: 0.8392 - val_loss: 0.5662 - val_acc: 0.8291\n",
      "\n",
      "Epoch 00052: val_loss improved from 0.56684 to 0.56617, saving model to ../working/DenseNet121.h5\n",
      "Epoch 51 : QWK: 0.9149362451453513\n",
      "saving checkpoint:  0.9149362451453513\n",
      "Epoch 53/150\n",
      " - 39s - loss: 0.5430 - acc: 0.8454 - val_loss: 0.5678 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00053: val_loss did not improve from 0.56617\n",
      "Epoch 52 : QWK: 0.9114022906942839\n",
      "Epoch 54/150\n",
      " - 39s - loss: 0.5697 - acc: 0.8341 - val_loss: 0.5675 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00054: val_loss did not improve from 0.56617\n",
      "Epoch 53 : QWK: 0.910147419495151\n",
      "Epoch 55/150\n",
      " - 38s - loss: 0.5586 - acc: 0.8412 - val_loss: 0.5693 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00055: val_loss did not improve from 0.56617\n",
      "Epoch 54 : QWK: 0.907924293984793\n",
      "Epoch 56/150\n",
      " - 38s - loss: 0.5869 - acc: 0.8412 - val_loss: 0.5688 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00056: val_loss did not improve from 0.56617\n",
      "Epoch 55 : QWK: 0.9096658549097774\n",
      "Epoch 57/150\n",
      " - 38s - loss: 0.5643 - acc: 0.8508 - val_loss: 0.5682 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00057: val_loss did not improve from 0.56617\n",
      "Epoch 56 : QWK: 0.9127195489899679\n",
      "Epoch 58/150\n",
      " - 38s - loss: 0.5937 - acc: 0.8418 - val_loss: 0.5665 - val_acc: 0.8273\n",
      "\n",
      "Epoch 00058: val_loss did not improve from 0.56617\n",
      "Epoch 57 : QWK: 0.9120942439290565\n",
      "Epoch 59/150\n",
      " - 38s - loss: 0.5714 - acc: 0.8434 - val_loss: 0.5681 - val_acc: 0.8255\n",
      "\n",
      "Epoch 00059: val_loss did not improve from 0.56617\n",
      "Epoch 58 : QWK: 0.9123834720342965\n",
      "Epoch 60/150\n",
      " - 38s - loss: 0.5847 - acc: 0.8396 - val_loss: 0.5702 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00060: val_loss did not improve from 0.56617\n",
      "Epoch 59 : QWK: 0.9080269422784184\n",
      "Epoch 61/150\n",
      " - 38s - loss: 0.5832 - acc: 0.8331 - val_loss: 0.5689 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00061: val_loss did not improve from 0.56617\n",
      "Epoch 60 : QWK: 0.9080616042971449\n",
      "Epoch 62/150\n",
      " - 39s - loss: 0.5677 - acc: 0.8450 - val_loss: 0.5670 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00062: val_loss did not improve from 0.56617\n",
      "Epoch 61 : QWK: 0.9099228316884459\n",
      "Epoch 63/150\n",
      " - 40s - loss: 0.5730 - acc: 0.8380 - val_loss: 0.5682 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00063: val_loss did not improve from 0.56617\n",
      "Epoch 62 : QWK: 0.9079590334025035\n",
      "Epoch 64/150\n",
      " - 38s - loss: 0.5959 - acc: 0.8421 - val_loss: 0.5681 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00064: val_loss did not improve from 0.56617\n",
      "Epoch 63 : QWK: 0.9108422728950314\n",
      "Epoch 65/150\n",
      " - 38s - loss: 0.5977 - acc: 0.8312 - val_loss: 0.5681 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00065: val_loss did not improve from 0.56617\n",
      "Epoch 64 : QWK: 0.9104491585273771\n",
      "Epoch 66/150\n",
      " - 38s - loss: 0.6050 - acc: 0.8347 - val_loss: 0.5716 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00066: val_loss did not improve from 0.56617\n",
      "Epoch 65 : QWK: 0.9059731473962499\n",
      "Epoch 67/150\n",
      " - 38s - loss: 0.5743 - acc: 0.8454 - val_loss: 0.5688 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00067: val_loss did not improve from 0.56617\n",
      "Epoch 66 : QWK: 0.9096537355706149\n",
      "Epoch 68/150\n",
      " - 39s - loss: 0.5610 - acc: 0.8483 - val_loss: 0.5670 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00068: val_loss did not improve from 0.56617\n",
      "Epoch 67 : QWK: 0.9099228316884459\n",
      "Epoch 69/150\n",
      " - 38s - loss: 0.5801 - acc: 0.8399 - val_loss: 0.5652 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00069: val_loss improved from 0.56617 to 0.56522, saving model to ../working/DenseNet121.h5\n",
      "Epoch 68 : QWK: 0.9104491585273771\n",
      "Epoch 70/150\n",
      " - 38s - loss: 0.5472 - acc: 0.8512 - val_loss: 0.5642 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00070: val_loss improved from 0.56522 to 0.56417, saving model to ../working/DenseNet121.h5\n",
      "Epoch 69 : QWK: 0.9095308397284504\n",
      "Epoch 71/150\n",
      " - 38s - loss: 0.6127 - acc: 0.8296 - val_loss: 0.5653 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00071: val_loss did not improve from 0.56417\n",
      "Epoch 70 : QWK: 0.9104815887076594\n",
      "Epoch 72/150\n",
      " - 38s - loss: 0.5796 - acc: 0.8389 - val_loss: 0.5662 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00072: val_loss did not improve from 0.56417\n",
      "Epoch 71 : QWK: 0.9077772188518314\n",
      "Epoch 73/150\n",
      " - 38s - loss: 0.6105 - acc: 0.8247 - val_loss: 0.5655 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00073: val_loss did not improve from 0.56417\n",
      "Epoch 72 : QWK: 0.9103641569301283\n",
      "Epoch 74/150\n",
      " - 38s - loss: 0.5690 - acc: 0.8428 - val_loss: 0.5639 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00074: val_loss improved from 0.56417 to 0.56386, saving model to ../working/DenseNet121.h5\n",
      "Epoch 73 : QWK: 0.9104815887076594\n",
      "Epoch 75/150\n",
      " - 38s - loss: 0.6034 - acc: 0.8235 - val_loss: 0.5656 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00075: val_loss did not improve from 0.56386\n",
      "Epoch 74 : QWK: 0.9099228316884459\n",
      "Epoch 76/150\n",
      " - 38s - loss: 0.5797 - acc: 0.8412 - val_loss: 0.5641 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00076: val_loss did not improve from 0.56386\n",
      "Epoch 75 : QWK: 0.9108422728950314\n",
      "Epoch 77/150\n",
      " - 38s - loss: 0.5784 - acc: 0.8412 - val_loss: 0.5662 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00077: val_loss did not improve from 0.56386\n",
      "Epoch 76 : QWK: 0.9099228316884459\n",
      "Epoch 78/150\n",
      " - 38s - loss: 0.5857 - acc: 0.8357 - val_loss: 0.5657 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00078: val_loss did not improve from 0.56386\n",
      "Epoch 77 : QWK: 0.9117965662690704\n",
      "Epoch 79/150\n",
      " - 38s - loss: 0.5644 - acc: 0.8367 - val_loss: 0.5626 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00079: val_loss improved from 0.56386 to 0.56256, saving model to ../working/DenseNet121.h5\n",
      "Epoch 78 : QWK: 0.9121275214364072\n",
      "Epoch 80/150\n",
      " - 39s - loss: 0.5844 - acc: 0.8457 - val_loss: 0.5629 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00080: val_loss did not improve from 0.56256\n",
      "Epoch 79 : QWK: 0.9121275214364072\n",
      "Epoch 81/150\n",
      " - 38s - loss: 0.5530 - acc: 0.8428 - val_loss: 0.5645 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00081: val_loss did not improve from 0.56256\n",
      "Epoch 80 : QWK: 0.9121275214364072\n",
      "Epoch 82/150\n",
      " - 38s - loss: 0.5801 - acc: 0.8351 - val_loss: 0.5673 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00082: val_loss did not improve from 0.56256\n",
      "Epoch 81 : QWK: 0.907993746616299\n",
      "Epoch 83/150\n",
      " - 38s - loss: 0.5791 - acc: 0.8528 - val_loss: 0.5664 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00083: val_loss did not improve from 0.56256\n",
      "Epoch 82 : QWK: 0.9100463874424757\n",
      "Epoch 84/150\n",
      " - 38s - loss: 0.5490 - acc: 0.8560 - val_loss: 0.5633 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00084: val_loss did not improve from 0.56256\n",
      "Epoch 83 : QWK: 0.9106814800078763\n",
      "Epoch 85/150\n",
      " - 39s - loss: 0.5708 - acc: 0.8486 - val_loss: 0.5632 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00085: val_loss did not improve from 0.56256\n",
      "Epoch 84 : QWK: 0.9081852867173374\n",
      "Epoch 86/150\n",
      " - 38s - loss: 0.5499 - acc: 0.8476 - val_loss: 0.5627 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00086: val_loss did not improve from 0.56256\n",
      "Epoch 85 : QWK: 0.9085024651401954\n",
      "Epoch 87/150\n",
      " - 37s - loss: 0.5802 - acc: 0.8402 - val_loss: 0.5638 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00087: val_loss did not improve from 0.56256\n",
      "Epoch 86 : QWK: 0.9063201458323618\n",
      "Epoch 88/150\n",
      " - 38s - loss: 0.5437 - acc: 0.8505 - val_loss: 0.5629 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00088: val_loss did not improve from 0.56256\n",
      "Epoch 87 : QWK: 0.9094446273845911\n",
      "Epoch 89/150\n",
      " - 37s - loss: 0.5831 - acc: 0.8483 - val_loss: 0.5656 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00089: val_loss did not improve from 0.56256\n",
      "Epoch 88 : QWK: 0.9047542852976265\n",
      "Epoch 90/150\n",
      " - 37s - loss: 0.5642 - acc: 0.8447 - val_loss: 0.5668 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00090: val_loss did not improve from 0.56256\n",
      "Epoch 89 : QWK: 0.9063907557832075\n",
      "Epoch 91/150\n",
      " - 38s - loss: 0.5921 - acc: 0.8325 - val_loss: 0.5670 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00091: val_loss did not improve from 0.56256\n",
      "Epoch 90 : QWK: 0.9078229143601234\n",
      "Epoch 92/150\n",
      " - 37s - loss: 0.5737 - acc: 0.8380 - val_loss: 0.5651 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00092: val_loss did not improve from 0.56256\n",
      "Epoch 91 : QWK: 0.9064934434093057\n",
      "Epoch 93/150\n",
      " - 38s - loss: 0.5751 - acc: 0.8405 - val_loss: 0.5641 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00093: val_loss did not improve from 0.56256\n",
      "Epoch 92 : QWK: 0.9086524926121944\n",
      "Epoch 94/150\n",
      " - 38s - loss: 0.5886 - acc: 0.8331 - val_loss: 0.5666 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00094: val_loss did not improve from 0.56256\n",
      "Epoch 93 : QWK: 0.9063907557832075\n",
      "Epoch 95/150\n",
      " - 38s - loss: 0.5787 - acc: 0.8524 - val_loss: 0.5670 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00095: val_loss did not improve from 0.56256\n",
      "Epoch 94 : QWK: 0.904339120850229\n",
      "Epoch 96/150\n",
      " - 38s - loss: 0.5524 - acc: 0.8553 - val_loss: 0.5636 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00096: val_loss did not improve from 0.56256\n",
      "Epoch 95 : QWK: 0.9130757213803332\n",
      "Epoch 97/150\n",
      " - 38s - loss: 0.5694 - acc: 0.8508 - val_loss: 0.5640 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00097: val_loss did not improve from 0.56256\n",
      "Epoch 96 : QWK: 0.9098115751033671\n",
      "Epoch 98/150\n",
      " - 39s - loss: 0.5705 - acc: 0.8421 - val_loss: 0.5617 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00098: val_loss improved from 0.56256 to 0.56169, saving model to ../working/DenseNet121.h5\n",
      "Epoch 97 : QWK: 0.9099554403997552\n",
      "Epoch 99/150\n",
      " - 38s - loss: 0.5757 - acc: 0.8457 - val_loss: 0.5618 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00099: val_loss did not improve from 0.56169\n",
      "Epoch 98 : QWK: 0.9086670999843109\n",
      "Epoch 100/150\n",
      " - 38s - loss: 0.5711 - acc: 0.8412 - val_loss: 0.5625 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00100: val_loss did not improve from 0.56169\n",
      "Epoch 99 : QWK: 0.9099554403997552\n",
      "Epoch 101/150\n",
      " - 38s - loss: 0.5871 - acc: 0.8389 - val_loss: 0.5645 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00101: val_loss did not improve from 0.56169\n",
      "Epoch 100 : QWK: 0.9029394045733087\n",
      "Epoch 102/150\n",
      " - 37s - loss: 0.5663 - acc: 0.8470 - val_loss: 0.5633 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00102: val_loss did not improve from 0.56169\n",
      "Epoch 101 : QWK: 0.9049945435368855\n",
      "Epoch 103/150\n",
      " - 38s - loss: 0.5804 - acc: 0.8502 - val_loss: 0.5632 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00103: val_loss did not improve from 0.56169\n",
      "Epoch 102 : QWK: 0.9070910019721538\n",
      "Epoch 104/150\n",
      " - 38s - loss: 0.5776 - acc: 0.8341 - val_loss: 0.5617 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00104: val_loss did not improve from 0.56169\n",
      "Epoch 103 : QWK: 0.9081408398493591\n",
      "Epoch 105/150\n",
      " - 38s - loss: 0.5849 - acc: 0.8312 - val_loss: 0.5640 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00105: val_loss did not improve from 0.56169\n",
      "Epoch 104 : QWK: 0.9084830713974087\n",
      "Epoch 106/150\n",
      " - 38s - loss: 0.5488 - acc: 0.8502 - val_loss: 0.5623 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00106: val_loss did not improve from 0.56169\n",
      "Epoch 105 : QWK: 0.9121275214364072\n",
      "Epoch 107/150\n",
      " - 38s - loss: 0.5375 - acc: 0.8470 - val_loss: 0.5630 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00107: val_loss did not improve from 0.56169\n",
      "Epoch 106 : QWK: 0.9084830713974087\n",
      "Epoch 108/150\n",
      " - 37s - loss: 0.5850 - acc: 0.8602 - val_loss: 0.5621 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00108: val_loss did not improve from 0.56169\n",
      "Epoch 107 : QWK: 0.9109083981442051\n",
      "Epoch 109/150\n",
      " - 37s - loss: 0.5556 - acc: 0.8466 - val_loss: 0.5624 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00109: val_loss did not improve from 0.56169\n",
      "Epoch 108 : QWK: 0.9107421719506893\n",
      "Epoch 110/150\n",
      " - 38s - loss: 0.5827 - acc: 0.8447 - val_loss: 0.5604 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00110: val_loss improved from 0.56169 to 0.56043, saving model to ../working/DenseNet121.h5\n",
      "Epoch 109 : QWK: 0.9129787436889304\n",
      "Epoch 111/150\n",
      " - 38s - loss: 0.5598 - acc: 0.8460 - val_loss: 0.5615 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00111: val_loss did not improve from 0.56043\n",
      "Epoch 110 : QWK: 0.9103813071701065\n",
      "Epoch 112/150\n",
      " - 38s - loss: 0.6018 - acc: 0.8218 - val_loss: 0.5638 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00112: val_loss did not improve from 0.56043\n",
      "Epoch 111 : QWK: 0.9087017005008422\n",
      "Epoch 113/150\n",
      " - 37s - loss: 0.6041 - acc: 0.8238 - val_loss: 0.5645 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00113: val_loss did not improve from 0.56043\n",
      "Epoch 112 : QWK: 0.9056239783122988\n",
      "Epoch 114/150\n",
      " - 38s - loss: 0.5749 - acc: 0.8373 - val_loss: 0.5608 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00114: val_loss did not improve from 0.56043\n",
      "Epoch 113 : QWK: 0.9054469475487413\n",
      "Epoch 115/150\n",
      " - 37s - loss: 0.5762 - acc: 0.8486 - val_loss: 0.5630 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00115: val_loss did not improve from 0.56043\n",
      "Epoch 114 : QWK: 0.9068433853225338\n",
      "Epoch 116/150\n",
      " - 38s - loss: 0.5735 - acc: 0.8441 - val_loss: 0.5628 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00116: val_loss did not improve from 0.56043\n",
      "Epoch 115 : QWK: 0.9068433853225338\n",
      "Epoch 117/150\n",
      " - 38s - loss: 0.5588 - acc: 0.8396 - val_loss: 0.5628 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00117: val_loss did not improve from 0.56043\n",
      "Epoch 116 : QWK: 0.9077532034796609\n",
      "Epoch 118/150\n",
      " - 38s - loss: 0.6000 - acc: 0.8302 - val_loss: 0.5609 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00118: val_loss did not improve from 0.56043\n",
      "Epoch 117 : QWK: 0.9099554403997552\n",
      "Epoch 119/150\n",
      " - 38s - loss: 0.5856 - acc: 0.8254 - val_loss: 0.5603 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00119: val_loss improved from 0.56043 to 0.56026, saving model to ../working/DenseNet121.h5\n",
      "Epoch 118 : QWK: 0.9081408398493591\n",
      "Epoch 120/150\n",
      " - 37s - loss: 0.5660 - acc: 0.8421 - val_loss: 0.5623 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00120: val_loss did not improve from 0.56026\n",
      "Epoch 119 : QWK: 0.9093297730089693\n",
      "Epoch 121/150\n",
      " - 38s - loss: 0.5527 - acc: 0.8495 - val_loss: 0.5626 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00121: val_loss did not improve from 0.56026\n",
      "Epoch 120 : QWK: 0.9077532034796609\n",
      "Epoch 122/150\n",
      " - 38s - loss: 0.5786 - acc: 0.8383 - val_loss: 0.5615 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00122: val_loss did not improve from 0.56026\n",
      "Epoch 121 : QWK: 0.9064966741930821\n",
      "Epoch 123/150\n",
      " - 38s - loss: 0.5570 - acc: 0.8460 - val_loss: 0.5622 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00123: val_loss did not improve from 0.56026\n",
      "Epoch 122 : QWK: 0.9084830713974087\n",
      "Epoch 124/150\n",
      " - 38s - loss: 0.5823 - acc: 0.8380 - val_loss: 0.5632 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00124: val_loss did not improve from 0.56026\n",
      "Epoch 123 : QWK: 0.9110785964160633\n",
      "Epoch 125/150\n",
      " - 38s - loss: 0.5811 - acc: 0.8428 - val_loss: 0.5623 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00125: val_loss did not improve from 0.56026\n",
      "Epoch 124 : QWK: 0.9068433853225338\n",
      "Epoch 126/150\n",
      " - 39s - loss: 0.5856 - acc: 0.8334 - val_loss: 0.5636 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00126: val_loss did not improve from 0.56026\n",
      "Epoch 125 : QWK: 0.9047542852976265\n",
      "Epoch 127/150\n",
      " - 38s - loss: 0.5701 - acc: 0.8405 - val_loss: 0.5610 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00127: val_loss did not improve from 0.56026\n",
      "Epoch 126 : QWK: 0.9085646330244351\n",
      "Epoch 128/150\n",
      " - 38s - loss: 0.5743 - acc: 0.8351 - val_loss: 0.5602 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00128: val_loss improved from 0.56026 to 0.56017, saving model to ../working/DenseNet121.h5\n",
      "Epoch 127 : QWK: 0.9109083981442051\n",
      "Epoch 129/150\n",
      " - 39s - loss: 0.5693 - acc: 0.8483 - val_loss: 0.5612 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00129: val_loss did not improve from 0.56017\n",
      "Epoch 128 : QWK: 0.9082524554778209\n",
      "Epoch 130/150\n",
      " - 38s - loss: 0.5800 - acc: 0.8444 - val_loss: 0.5617 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00130: val_loss did not improve from 0.56017\n",
      "Epoch 129 : QWK: 0.9089225599424859\n",
      "Epoch 131/150\n",
      " - 38s - loss: 0.5768 - acc: 0.8502 - val_loss: 0.5613 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00131: val_loss did not improve from 0.56017\n",
      "Epoch 130 : QWK: 0.9090903593450474\n",
      "Epoch 132/150\n",
      " - 38s - loss: 0.5693 - acc: 0.8322 - val_loss: 0.5610 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00132: val_loss did not improve from 0.56017\n",
      "Epoch 131 : QWK: 0.9038404849854947\n",
      "Epoch 133/150\n",
      " - 38s - loss: 0.6018 - acc: 0.8283 - val_loss: 0.5651 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00133: val_loss did not improve from 0.56017\n",
      "Epoch 132 : QWK: 0.9035124687701016\n",
      "Epoch 134/150\n",
      " - 38s - loss: 0.5437 - acc: 0.8524 - val_loss: 0.5587 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00134: val_loss improved from 0.56017 to 0.55873, saving model to ../working/DenseNet121.h5\n",
      "Epoch 133 : QWK: 0.9093438985865131\n",
      "Epoch 135/150\n",
      " - 38s - loss: 0.5645 - acc: 0.8476 - val_loss: 0.5597 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00135: val_loss did not improve from 0.55873\n",
      "Epoch 134 : QWK: 0.9096999118281968\n",
      "Epoch 136/150\n",
      " - 38s - loss: 0.5658 - acc: 0.8450 - val_loss: 0.5622 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00136: val_loss did not improve from 0.55873\n",
      "Epoch 135 : QWK: 0.905482714369678\n",
      "Epoch 137/150\n",
      " - 37s - loss: 0.5679 - acc: 0.8460 - val_loss: 0.5585 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00137: val_loss improved from 0.55873 to 0.55854, saving model to ../working/DenseNet121.h5\n",
      "Epoch 136 : QWK: 0.9097324798949614\n",
      "Epoch 138/150\n",
      " - 38s - loss: 0.5917 - acc: 0.8344 - val_loss: 0.5604 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00138: val_loss did not improve from 0.55854\n",
      "Epoch 137 : QWK: 0.905482714369678\n",
      "Epoch 139/150\n",
      " - 38s - loss: 0.5866 - acc: 0.8363 - val_loss: 0.5598 - val_acc: 0.8236\n",
      "\n",
      "Epoch 00139: val_loss did not improve from 0.55854\n",
      "Epoch 138 : QWK: 0.9084352569160389\n",
      "Epoch 140/150\n",
      " - 38s - loss: 0.5935 - acc: 0.8363 - val_loss: 0.5604 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00140: val_loss did not improve from 0.55854\n",
      "Epoch 139 : QWK: 0.9106814800078763\n",
      "Epoch 141/150\n",
      " - 38s - loss: 0.5581 - acc: 0.8447 - val_loss: 0.5619 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00141: val_loss did not improve from 0.55854\n",
      "Epoch 140 : QWK: 0.9063907557832075\n",
      "Epoch 142/150\n",
      " - 38s - loss: 0.5853 - acc: 0.8386 - val_loss: 0.5625 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00142: val_loss did not improve from 0.55854\n",
      "Epoch 141 : QWK: 0.9063201458323618\n",
      "Epoch 143/150\n",
      " - 38s - loss: 0.5661 - acc: 0.8531 - val_loss: 0.5615 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00143: val_loss did not improve from 0.55854\n",
      "Epoch 142 : QWK: 0.9099554403997552\n",
      "Epoch 144/150\n",
      " - 38s - loss: 0.5962 - acc: 0.8373 - val_loss: 0.5621 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00144: val_loss did not improve from 0.55854\n",
      "Epoch 143 : QWK: 0.9094446273845911\n",
      "Epoch 145/150\n",
      " - 38s - loss: 0.5767 - acc: 0.8421 - val_loss: 0.5616 - val_acc: 0.8182\n",
      "\n",
      "Epoch 00145: val_loss did not improve from 0.55854\n",
      "Epoch 144 : QWK: 0.9063201458323618\n",
      "Epoch 146/150\n",
      " - 37s - loss: 0.5583 - acc: 0.8512 - val_loss: 0.5591 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00146: val_loss did not improve from 0.55854\n",
      "Epoch 145 : QWK: 0.9110785964160633\n",
      "Epoch 147/150\n",
      " - 38s - loss: 0.5784 - acc: 0.8421 - val_loss: 0.5602 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00147: val_loss did not improve from 0.55854\n",
      "Epoch 146 : QWK: 0.9110785964160633\n",
      "Epoch 148/150\n",
      " - 38s - loss: 0.5746 - acc: 0.8367 - val_loss: 0.5612 - val_acc: 0.8218\n",
      "\n",
      "Epoch 00148: val_loss did not improve from 0.55854\n",
      "Epoch 147 : QWK: 0.9072635117977471\n",
      "Epoch 149/150\n",
      " - 38s - loss: 0.5751 - acc: 0.8460 - val_loss: 0.5610 - val_acc: 0.8200\n",
      "\n",
      "Epoch 00149: val_loss did not improve from 0.55854\n",
      "Epoch 148 : QWK: 0.9085161151601576\n",
      "Epoch 150/150\n",
      " - 38s - loss: 0.5572 - acc: 0.8489 - val_loss: 0.5616 - val_acc: 0.8164\n",
      "\n",
      "Epoch 00150: val_loss did not improve from 0.55854\n",
      "Epoch 149 : QWK: 0.9042325953371599\n"
     ]
    }
   ],
   "source": [
    "EPOCHS = 150\n",
    "history = mymodel.fit_generator(training_generator,steps_per_epoch = X_train.shape[0] // batch_size,epochs = EPOCHS,\n",
    "                         validation_data = (X_val,Y_val),\n",
    "                         validation_steps = 30,\n",
    "                         workers = 2,use_multiprocessing=True,\n",
    "                         verbose=2, callbacks=mycallbacks)\n",
    "# "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "mymodel.save_weights(\"model.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y_val_pred = mymodel.predict_on_batch(X_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y_val_pred_hot = np.argmax(Y_val_pred,axis=1)\n",
    "Y_val_actual_hot = np.argmax(Y_val,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[259   2   2   0   0]\n",
      " [  8  26  15   0   1]\n",
      " [  3  16 131  12   4]\n",
      " [  0   0   8  11   4]\n",
      " [  0   0  12  14  22]]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7cb09558a320>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(Y_val_actual_hot, Y_val_pred_hot, np.array(class_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# history = mymodel.fit_generator(train_gen,steps_per_epoch = 10,epochs = EPOCHS,\n",
    "#                          validation_data = val_gen,\n",
    "#                          validation_steps = 10,\n",
    "#                          workers = 2,use_multiprocessing=True,\n",
    "#                      verbose=2, callbacks=mycallbacks)\n",
    "# mymodel.save_weights(\"model.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train_gen = img_gen.flow_from_dataframe(\n",
    "#     dataframe=train_data,\n",
    "#     directory=\"../input/aptos2019-blindness-detection/train_images\",\n",
    "#     x_col=\"filename\",\n",
    "#     y_col=\"diagnosis\",\n",
    "#     batch_size=batch_size,\n",
    "#     shuffle=True,\n",
    "#     class_mode=\"categorical\",\n",
    "#     classes=labels,\n",
    "#     target_size=(img_size,img_size),\n",
    "#     subset='training')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# val_gen = img_gen.flow_from_dataframe(\n",
    "#     dataframe=train_data,\n",
    "#     directory=\"../input/aptos2019-blindness-detection/train_images\",\n",
    "#     x_col=\"filename\",\n",
    "#     y_col=\"diagnosis\",\n",
    "#     batch_size=batch_size,\n",
    "#     shuffle=True,\n",
    "#     class_mode=\"categorical\",\n",
    "#     classes=labels,\n",
    "#     target_size=(img_size,img_size),\n",
    "#     subset='validation'\n",
    "# )\n",
    "# # dir(val_gen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# class QWK(keras.callbacks.Callback):\n",
    "#     def on_train_begin(self, logs={}):\n",
    "#         self.val_kappas = []\n",
    "\n",
    "#     def on_epoch_end(self, epoch, logs={}):\n",
    "#         print(self.__dict__)\n",
    "#         X_val, y_val = self.model.validation_data[:2]\n",
    "#         y_pred = self.model.predict(X_val)\n",
    "\n",
    "#         _val_kappa = cohen_kappa_score(\n",
    "#             y_val.argmax(axis=1), \n",
    "#             y_pred.argmax(axis=1), \n",
    "#             weights='quadratic'\n",
    "#         )\n",
    "\n",
    "#         self.val_kappas.append(_val_kappa)\n",
    "\n",
    "#         print(f\"epoch: {epoch} \\t val_kappa: {_val_kappa:.4f}\")\n",
    "\n",
    "#         return\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# val_gen = img_gen.flow_from_dataframe(\n",
    "#     dataframe=train_data,\n",
    "#     directory=\"../input/aptos2019-blindness-detection/train_images\",\n",
    "#     x_col=\"filename\",\n",
    "#     y_col=\"diagnosis\",\n",
    "#     batch_size=1000,\n",
    "#     shuffle=True,\n",
    "#     class_mode=\"categorical\",\n",
    "#     classes=labels,\n",
    "#     target_size=(img_size,img_size),\n",
    "#     subset='validation'\n",
    "# )\n",
    "# # print(val_gen.batch_index)\n",
    "# # for i in range(val_gen.batch_index):\n",
    "# batch_data, batch_labels = val_gen.next()\n",
    "# batch_labels_arg = np.argmax(batch_labels,axis=1)\n",
    "# # print(batch_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cur_batch_size = batch_data.shape[0] \n",
    "# preds = np.empty((cur_batch_size,5))\n",
    "# preds_one_hot = np.zeros((cur_batch_size,5))\n",
    "# for i in range(cur_batch_size):\n",
    "#     preds[i,:] = mymodel.predict(batch_data[i,:,:,:][np.newaxis,:])\n",
    "# args_max = np.argmax(preds,axis=1)\n",
    "# preds_one_hot[np.arange(cur_batch_size),args_max] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plot_confusion_matrix(batch_labels_arg, args_max, np.array(class_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(12, 12))\n",
    "ax1.plot(history.history['loss'], color='b', label=\"Training loss\")\n",
    "ax1.plot(history.history['val_loss'], color='r', label=\"validation loss\")\n",
    "# ax1.set_xticks(np.arange(1, epochs, 1))\n",
    "# ax1.set_yticks(np.arange(0, 1, 0.1))\n",
    "\n",
    "ax2.plot(history.history['acc'], color='b', label=\"Training accuracy\")\n",
    "ax2.plot(history.history['val_acc'], color='r',label=\"Validation accuracy\")\n",
    "# ax2.set_xticks(np.arange(1, epochs, 1))\n",
    "\n",
    "legend = plt.legend(loc='best', shadow=True)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_code</th>\n",
       "      <th>filename</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0005cfc8afb6</td>\n",
       "      <td>0005cfc8afb6.png</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>003f0afdcd15</td>\n",
       "      <td>003f0afdcd15.png</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>006efc72b638</td>\n",
       "      <td>006efc72b638.png</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00836aaacf06</td>\n",
       "      <td>00836aaacf06.png</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>009245722fa4</td>\n",
       "      <td>009245722fa4.png</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id_code          filename\n",
       "0  0005cfc8afb6  0005cfc8afb6.png\n",
       "1  003f0afdcd15  003f0afdcd15.png\n",
       "2  006efc72b638  006efc72b638.png\n",
       "3  00836aaacf06  00836aaacf06.png\n",
       "4  009245722fa4  009245722fa4.png"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data = pd.read_csv(\"../input/aptos2019-blindness-detection/test.csv\")\n",
    "test_data[\"filename\"] = test_data[\"id_code\"].map(lambda x:x+\".png\")\n",
    "test_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 1928 validated image filenames.\n"
     ]
    }
   ],
   "source": [
    "test_gen = ImageDataGenerator(rescale=1./255)\n",
    "test_generator = test_gen.flow_from_dataframe(  \n",
    "        dataframe=test_data,\n",
    "        directory = \"../input/aptos2019-blindness-detection/test_images\",    \n",
    "        x_col=\"filename\",\n",
    "        y_col=None,\n",
    "        target_size = (img_size,img_size),\n",
    "        batch_size = 1,\n",
    "        shuffle = False,\n",
    "        class_mode = None\n",
    "        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = mymodel.predict_generator(test_generator, steps = len(test_generator.filenames))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "filenames=test_generator.filenames\n",
    "results=pd.DataFrame({\"id_code\":filenames,\n",
    "                      \"diagnosis\":np.argmax(predictions,axis=1)})\n",
    "results['id_code'] = results['id_code'].map(lambda x: str(x)[:-4])\n",
    "results.to_csv(\"submission.csv\",index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"submission.csv\">submission.csv</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 4, 2, 3])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "results.diagnosis.hist()\n",
    "results.diagnosis.unique()"
   ]
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "gpu",
   "dataSources": [
    {
     "databundleVersionId": 875431,
     "sourceId": 14774,
     "sourceType": "competition"
    },
    {
     "datasetId": 80814,
     "sourceId": 187731,
     "sourceType": "datasetVersion"
    }
   ],
   "dockerImageVersionId": 27938,
   "isGpuEnabled": true,
   "isInternetEnabled": false,
   "language": "python",
   "sourceType": "notebook"
  },
  "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}