convolutional_autoencoder.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# uncomment this to use CPU instead of GPU\n",
    "# import os\n",
    "# os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\" \n",
    "# os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import sklearn\n",
    "import matplotlib.pyplot as plt\n",
    "from datetime import datetime\n",
    "from functools import reduce\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.datasets import mnist\n",
    "\n",
    "from tensorflow.keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, Flatten, Dense, Reshape\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.callbacks import TensorBoard, ReduceLROnPlateau, EarlyStopping"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load MNIST dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load and format dataset as square images or as matrices so we can use convolutional layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28, 1)\n",
      "(10000, 28, 28, 1)\n"
     ]
    }
   ],
   "source": [
    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
    "x_train = x_train.astype('float32') / 255.\n",
    "x_test = x_test.astype('float32') / 255.\n",
    "x_train = np.reshape(x_train, (len(x_train), 28, 28, 1))\n",
    "x_test = np.reshape(x_test, (len(x_test), 28, 28, 1)) \n",
    "print(x_train.shape)\n",
    "print(x_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f9093de7890>"
      ]
     },
     "execution_count": 5,
     "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": [
    "plt.figure()\n",
    "plt.imshow(x_train[0, :, :, 0], cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 100\n",
    "img_width = x_train.shape[1]\n",
    "img_height = x_train.shape[2]\n",
    "channels = x_train.shape[3] # images are in black and white, for RGB we have 3 channels (r,g,b)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now we define the Convolutional Autoencoder - CAE\n",
    "\n",
    "```\n",
    "input --> encoder --> latent_space --> decoder --> reconstructed_image\n",
    "input_img              encoder output                 decoded_output\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_shape = (3, 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "input_img = Input(shape=(img_width, img_height, channels))\n",
    "x = Conv2D(16, conv_shape, activation='relu', padding='same')(input_img)\n",
    "x = MaxPooling2D((2, 2), padding='same')(x)\n",
    "x = Conv2D(8, conv_shape, activation='relu', padding='same')(x)\n",
    "encoded = MaxPooling2D((2, 2), padding='same', name='encoded')(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "decoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = Conv2D(8, conv_shape, activation='relu', padding='same')(encoded)\n",
    "x = UpSampling2D((2, 2))(x)\n",
    "x = Conv2D(16, conv_shape, activation='relu', padding='same')(x)\n",
    "x = UpSampling2D((2, 2))(x)\n",
    "decoded = Conv2D(channels, conv_shape, activation='sigmoid', padding='same')(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "dimensionality reduction?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img_height * img_width / int(reduce(lambda x, y: x*y, np.asarray(encoded.shape[1:])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"encoder\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 28, 28, 1)]       0         \n",
      "_________________________________________________________________\n",
      "conv2d (Conv2D)              (None, 28, 28, 16)        160       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 14, 14, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 14, 14, 8)         1160      \n",
      "_________________________________________________________________\n",
      "encoded (MaxPooling2D)       (None, 7, 7, 8)           0         \n",
      "=================================================================\n",
      "Total params: 1,320\n",
      "Trainable params: 1,320\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "encoder = Model(input_img, encoded, name='encoder')\n",
    "encoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"convolutional_autoencoder\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 28, 28, 1)]       0         \n",
      "_________________________________________________________________\n",
      "conv2d (Conv2D)              (None, 28, 28, 16)        160       \n",
      "_________________________________________________________________\n",
      "max_pooling2d (MaxPooling2D) (None, 14, 14, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 14, 14, 8)         1160      \n",
      "_________________________________________________________________\n",
      "encoded (MaxPooling2D)       (None, 7, 7, 8)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 7, 7, 8)           584       \n",
      "_________________________________________________________________\n",
      "up_sampling2d (UpSampling2D) (None, 14, 14, 8)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 14, 14, 16)        1168      \n",
      "_________________________________________________________________\n",
      "up_sampling2d_1 (UpSampling2 (None, 28, 28, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 28, 28, 1)         145       \n",
      "=================================================================\n",
      "Total params: 3,217\n",
      "Trainable params: 3,217\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "autoencoder = Model(input_img, decoded, name='convolutional_autoencoder')\n",
    "autoencoder.compile(optimizer='adadelta', loss='mse')\n",
    "autoencoder.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### callbacks\n",
    "\n",
    "We are going to use early stopping and reduce lr on pateau callbacks to avoid improve the generalization of the model and avoid overfitting.\n",
    "\n",
    "Also we are going to add the tensorboard callback to check the training losses during the training process."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "log_dir = \"./logs/\" + datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n",
    "tb_cb = TensorBoard(\n",
    "    log_dir=log_dir, \n",
    "    profile_batch=0)\n",
    "\n",
    "es_cb = EarlyStopping(\n",
    "    monitor='val_loss',\n",
    "    verbose=True,\n",
    "    patience=12,\n",
    "    min_delta=0.0001, \n",
    "    restore_best_weights=True)\n",
    "\n",
    "lr_cb = ReduceLROnPlateau(\n",
    "    monitor='val_loss',\n",
    "    verbose=True,\n",
    "    patience=8,\n",
    "    mi_delta=0.0001)\n",
    "\n",
    "cb = [tb_cb, es_cb, lr_cb]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the target in an autoencoder is the same as the input, both `x` and `y` are the same here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/500\n",
      "60000/60000 [==============================] - 5s 80us/sample - loss: 0.2278 - val_loss: 0.2266\n",
      "Epoch 2/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.2245 - val_loss: 0.2225\n",
      "Epoch 3/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.2194 - val_loss: 0.2162\n",
      "Epoch 4/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.2113 - val_loss: 0.2063\n",
      "Epoch 5/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1987 - val_loss: 0.1910\n",
      "Epoch 6/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1809 - val_loss: 0.1722\n",
      "Epoch 7/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.1628 - val_loss: 0.1563\n",
      "Epoch 8/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1493 - val_loss: 0.1455\n",
      "Epoch 9/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1403 - val_loss: 0.1382\n",
      "Epoch 10/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1339 - val_loss: 0.1329\n",
      "Epoch 11/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.1292 - val_loss: 0.1288\n",
      "Epoch 12/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1253 - val_loss: 0.1251\n",
      "Epoch 13/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1217 - val_loss: 0.1214\n",
      "Epoch 14/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1179 - val_loss: 0.1172\n",
      "Epoch 15/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1131 - val_loss: 0.1116\n",
      "Epoch 16/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.1067 - val_loss: 0.1039\n",
      "Epoch 17/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0982 - val_loss: 0.0945\n",
      "Epoch 18/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0894 - val_loss: 0.0859\n",
      "Epoch 19/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0811 - val_loss: 0.0776\n",
      "Epoch 20/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0735 - val_loss: 0.0709\n",
      "Epoch 21/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0680 - val_loss: 0.0664\n",
      "Epoch 22/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0641 - val_loss: 0.0630\n",
      "Epoch 23/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0612 - val_loss: 0.0604\n",
      "Epoch 24/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0588 - val_loss: 0.0582\n",
      "Epoch 25/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0568 - val_loss: 0.0564\n",
      "Epoch 26/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0551 - val_loss: 0.0548\n",
      "Epoch 27/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0536 - val_loss: 0.0533\n",
      "Epoch 28/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0523 - val_loss: 0.0520\n",
      "Epoch 29/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0510 - val_loss: 0.0508\n",
      "Epoch 30/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0499 - val_loss: 0.0497\n",
      "Epoch 31/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0488 - val_loss: 0.0486\n",
      "Epoch 32/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0478 - val_loss: 0.0475\n",
      "Epoch 33/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0467 - val_loss: 0.0465\n",
      "Epoch 34/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0458 - val_loss: 0.0455\n",
      "Epoch 35/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0448 - val_loss: 0.0445\n",
      "Epoch 36/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0438 - val_loss: 0.0435\n",
      "Epoch 37/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0428 - val_loss: 0.0425\n",
      "Epoch 38/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0419 - val_loss: 0.0415\n",
      "Epoch 39/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0409 - val_loss: 0.0405\n",
      "Epoch 40/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0400 - val_loss: 0.0396\n",
      "Epoch 41/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0391 - val_loss: 0.0386\n",
      "Epoch 42/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0382 - val_loss: 0.0378\n",
      "Epoch 43/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0374 - val_loss: 0.0370\n",
      "Epoch 44/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0367 - val_loss: 0.0362\n",
      "Epoch 45/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0360 - val_loss: 0.0355\n",
      "Epoch 46/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0354 - val_loss: 0.0349\n",
      "Epoch 47/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0348 - val_loss: 0.0343\n",
      "Epoch 48/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0343 - val_loss: 0.0338\n",
      "Epoch 49/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0338 - val_loss: 0.0333\n",
      "Epoch 50/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0334 - val_loss: 0.0329\n",
      "Epoch 51/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0329 - val_loss: 0.0325\n",
      "Epoch 52/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0326 - val_loss: 0.0321\n",
      "Epoch 53/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0322 - val_loss: 0.0318\n",
      "Epoch 54/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0319 - val_loss: 0.0315\n",
      "Epoch 55/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0316 - val_loss: 0.0312\n",
      "Epoch 56/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0313 - val_loss: 0.0309\n",
      "Epoch 57/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0311 - val_loss: 0.0306\n",
      "Epoch 58/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0308 - val_loss: 0.0304\n",
      "Epoch 59/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0306 - val_loss: 0.0301\n",
      "Epoch 60/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0304 - val_loss: 0.0299\n",
      "Epoch 61/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0301 - val_loss: 0.0297\n",
      "Epoch 62/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0299 - val_loss: 0.0295\n",
      "Epoch 63/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0297 - val_loss: 0.0293\n",
      "Epoch 64/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0296 - val_loss: 0.0291\n",
      "Epoch 65/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0294 - val_loss: 0.0289\n",
      "Epoch 66/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0292 - val_loss: 0.0288\n",
      "Epoch 67/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0290 - val_loss: 0.0286\n",
      "Epoch 68/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0289 - val_loss: 0.0284\n",
      "Epoch 69/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0287 - val_loss: 0.0283\n",
      "Epoch 70/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0285 - val_loss: 0.0281\n",
      "Epoch 71/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0284 - val_loss: 0.0279\n",
      "Epoch 72/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0282 - val_loss: 0.0278\n",
      "Epoch 73/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0281 - val_loss: 0.0276\n",
      "Epoch 74/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0279 - val_loss: 0.0275\n",
      "Epoch 75/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0278 - val_loss: 0.0273\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 76/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0276 - val_loss: 0.0272\n",
      "Epoch 77/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0275 - val_loss: 0.0271\n",
      "Epoch 78/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0273 - val_loss: 0.0269\n",
      "Epoch 79/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0272 - val_loss: 0.0268\n",
      "Epoch 80/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0271 - val_loss: 0.0266\n",
      "Epoch 81/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0269 - val_loss: 0.0265\n",
      "Epoch 82/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0268 - val_loss: 0.0264\n",
      "Epoch 83/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0267 - val_loss: 0.0262\n",
      "Epoch 84/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0265 - val_loss: 0.0261\n",
      "Epoch 85/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0264 - val_loss: 0.0260\n",
      "Epoch 86/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0263 - val_loss: 0.0258\n",
      "Epoch 87/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0261 - val_loss: 0.0257\n",
      "Epoch 88/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0260 - val_loss: 0.0256\n",
      "Epoch 89/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0259 - val_loss: 0.0255\n",
      "Epoch 90/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0258 - val_loss: 0.0253\n",
      "Epoch 91/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0256 - val_loss: 0.0252\n",
      "Epoch 92/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0255 - val_loss: 0.0251\n",
      "Epoch 93/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0254 - val_loss: 0.0250\n",
      "Epoch 94/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0252 - val_loss: 0.0248\n",
      "Epoch 95/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0251 - val_loss: 0.0247\n",
      "Epoch 96/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0250 - val_loss: 0.0246\n",
      "Epoch 97/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0249 - val_loss: 0.0245\n",
      "Epoch 98/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0248 - val_loss: 0.0243\n",
      "Epoch 99/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0246 - val_loss: 0.0242\n",
      "Epoch 100/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0245 - val_loss: 0.0241\n",
      "Epoch 101/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0244 - val_loss: 0.0240\n",
      "Epoch 102/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0243 - val_loss: 0.0239\n",
      "Epoch 103/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0242 - val_loss: 0.0237\n",
      "Epoch 104/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0240 - val_loss: 0.0236\n",
      "Epoch 105/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0239 - val_loss: 0.0235\n",
      "Epoch 106/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0238 - val_loss: 0.0234\n",
      "Epoch 107/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0237 - val_loss: 0.0233\n",
      "Epoch 108/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0236 - val_loss: 0.0232\n",
      "Epoch 109/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0235 - val_loss: 0.0230\n",
      "Epoch 110/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0233 - val_loss: 0.0229\n",
      "Epoch 111/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0232 - val_loss: 0.0228\n",
      "Epoch 112/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0231 - val_loss: 0.0227\n",
      "Epoch 113/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0230 - val_loss: 0.0226\n",
      "Epoch 114/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0229 - val_loss: 0.0225\n",
      "Epoch 115/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0228 - val_loss: 0.0224\n",
      "Epoch 116/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0227 - val_loss: 0.0223\n",
      "Epoch 117/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0226 - val_loss: 0.0222\n",
      "Epoch 118/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0225 - val_loss: 0.0221\n",
      "Epoch 119/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0224 - val_loss: 0.0220\n",
      "Epoch 120/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0223 - val_loss: 0.0219\n",
      "Epoch 121/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0222 - val_loss: 0.0218\n",
      "Epoch 122/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0221 - val_loss: 0.0217\n",
      "Epoch 123/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0220 - val_loss: 0.0216\n",
      "Epoch 124/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0219 - val_loss: 0.0215\n",
      "Epoch 125/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0218 - val_loss: 0.0214\n",
      "Epoch 126/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0217 - val_loss: 0.0213\n",
      "Epoch 127/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0216 - val_loss: 0.0212\n",
      "Epoch 128/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0215 - val_loss: 0.0211\n",
      "Epoch 129/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0214 - val_loss: 0.0210\n",
      "Epoch 130/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0213 - val_loss: 0.0209\n",
      "Epoch 131/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0212 - val_loss: 0.0208\n",
      "Epoch 132/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0211 - val_loss: 0.0207\n",
      "Epoch 133/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0210 - val_loss: 0.0206\n",
      "Epoch 134/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0210 - val_loss: 0.0205\n",
      "Epoch 135/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0209 - val_loss: 0.0204\n",
      "Epoch 136/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0208 - val_loss: 0.0204\n",
      "Epoch 137/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0207 - val_loss: 0.0203\n",
      "Epoch 138/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0206 - val_loss: 0.0202\n",
      "Epoch 139/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0205 - val_loss: 0.0201\n",
      "Epoch 140/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0204 - val_loss: 0.0200\n",
      "Epoch 141/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0204 - val_loss: 0.0199\n",
      "Epoch 142/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0203 - val_loss: 0.0199\n",
      "Epoch 143/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0202 - val_loss: 0.0198\n",
      "Epoch 144/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0201 - val_loss: 0.0197\n",
      "Epoch 145/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0200 - val_loss: 0.0196\n",
      "Epoch 146/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0200 - val_loss: 0.0195\n",
      "Epoch 147/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0199 - val_loss: 0.0195\n",
      "Epoch 148/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0198 - val_loss: 0.0194\n",
      "Epoch 149/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0197 - val_loss: 0.0193\n",
      "Epoch 150/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0197 - val_loss: 0.0192\n",
      "Epoch 151/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0196 - val_loss: 0.0192\n",
      "Epoch 152/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0195 - val_loss: 0.0191\n",
      "Epoch 153/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0194 - val_loss: 0.0190\n",
      "Epoch 154/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0194 - val_loss: 0.0189\n",
      "Epoch 155/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0193 - val_loss: 0.0189\n",
      "Epoch 156/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0192 - val_loss: 0.0188\n",
      "Epoch 157/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0192 - val_loss: 0.0187\n",
      "Epoch 158/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0191 - val_loss: 0.0187\n",
      "Epoch 159/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0190 - val_loss: 0.0186\n",
      "Epoch 160/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0190 - val_loss: 0.0185\n",
      "Epoch 161/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0189 - val_loss: 0.0185\n",
      "Epoch 162/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0188 - val_loss: 0.0184\n",
      "Epoch 163/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0188 - val_loss: 0.0183\n",
      "Epoch 164/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0187 - val_loss: 0.0183\n",
      "Epoch 165/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0186 - val_loss: 0.0182\n",
      "Epoch 166/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0186 - val_loss: 0.0182\n",
      "Epoch 167/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0185 - val_loss: 0.0181\n",
      "Epoch 168/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0185 - val_loss: 0.0180\n",
      "Epoch 169/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0184 - val_loss: 0.0180\n",
      "Epoch 170/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0183 - val_loss: 0.0179\n",
      "Epoch 171/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0183 - val_loss: 0.0179\n",
      "Epoch 172/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0182 - val_loss: 0.0178\n",
      "Epoch 173/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0182 - val_loss: 0.0178\n",
      "Epoch 174/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0181 - val_loss: 0.0177\n",
      "Epoch 175/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0181 - val_loss: 0.0176\n",
      "Epoch 176/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0180 - val_loss: 0.0176\n",
      "Epoch 177/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0180 - val_loss: 0.0175\n",
      "Epoch 178/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0179 - val_loss: 0.0175\n",
      "Epoch 179/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0179 - val_loss: 0.0174\n",
      "Epoch 180/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0178 - val_loss: 0.0174\n",
      "Epoch 181/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0178 - val_loss: 0.0173\n",
      "Epoch 182/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0177 - val_loss: 0.0173\n",
      "Epoch 183/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0177 - val_loss: 0.0172\n",
      "Epoch 184/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0176 - val_loss: 0.0172\n",
      "Epoch 185/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0176 - val_loss: 0.0171\n",
      "Epoch 186/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0175 - val_loss: 0.0171\n",
      "Epoch 187/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0175 - val_loss: 0.0170\n",
      "Epoch 188/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0174 - val_loss: 0.0170\n",
      "Epoch 189/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0174 - val_loss: 0.0169\n",
      "Epoch 190/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0173 - val_loss: 0.0169\n",
      "Epoch 191/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0173 - val_loss: 0.0169\n",
      "Epoch 192/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0172 - val_loss: 0.0168\n",
      "Epoch 193/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0172 - val_loss: 0.0168\n",
      "Epoch 194/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0171 - val_loss: 0.0167\n",
      "Epoch 195/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0171 - val_loss: 0.0167\n",
      "Epoch 196/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0171 - val_loss: 0.0166\n",
      "Epoch 197/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0170 - val_loss: 0.0166\n",
      "Epoch 198/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0170 - val_loss: 0.0166\n",
      "Epoch 199/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0169 - val_loss: 0.0165\n",
      "Epoch 200/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0169 - val_loss: 0.0165\n",
      "Epoch 201/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0168 - val_loss: 0.0164\n",
      "Epoch 202/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0168 - val_loss: 0.0164\n",
      "Epoch 203/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0168 - val_loss: 0.0164\n",
      "Epoch 204/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0167 - val_loss: 0.0163\n",
      "Epoch 205/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0167 - val_loss: 0.0163\n",
      "Epoch 206/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0167 - val_loss: 0.0162\n",
      "Epoch 207/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0166 - val_loss: 0.0162\n",
      "Epoch 208/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0166 - val_loss: 0.0162\n",
      "Epoch 209/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0165 - val_loss: 0.0161\n",
      "Epoch 210/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0165 - val_loss: 0.0161\n",
      "Epoch 211/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0165 - val_loss: 0.0161\n",
      "Epoch 212/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0164 - val_loss: 0.0160\n",
      "Epoch 213/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0164 - val_loss: 0.0160\n",
      "Epoch 214/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0164 - val_loss: 0.0159\n",
      "Epoch 215/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0163 - val_loss: 0.0159\n",
      "Epoch 216/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0163 - val_loss: 0.0159\n",
      "Epoch 217/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0163 - val_loss: 0.0158\n",
      "Epoch 218/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0162 - val_loss: 0.0158\n",
      "Epoch 219/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0162 - val_loss: 0.0158\n",
      "Epoch 220/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0162 - val_loss: 0.0157\n",
      "Epoch 221/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0161 - val_loss: 0.0157\n",
      "Epoch 222/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0161 - val_loss: 0.0157\n",
      "Epoch 223/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0161 - val_loss: 0.0156\n",
      "Epoch 224/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0160 - val_loss: 0.0156\n",
      "Epoch 225/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0160 - val_loss: 0.0156\n",
      "Epoch 226/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0160 - val_loss: 0.0156\n",
      "Epoch 227/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0159 - val_loss: 0.0155\n",
      "Epoch 228/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0159 - val_loss: 0.0155\n",
      "Epoch 229/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0159 - val_loss: 0.0155\n",
      "Epoch 230/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0158 - val_loss: 0.0154\n",
      "Epoch 231/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0158 - val_loss: 0.0154\n",
      "Epoch 232/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0158 - val_loss: 0.0154\n",
      "Epoch 233/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0157 - val_loss: 0.0153\n",
      "Epoch 234/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0157 - val_loss: 0.0153\n",
      "Epoch 235/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0157 - val_loss: 0.0153\n",
      "Epoch 236/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0157 - val_loss: 0.0153\n",
      "Epoch 237/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0156 - val_loss: 0.0152\n",
      "Epoch 238/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0156 - val_loss: 0.0152\n",
      "Epoch 239/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0156 - val_loss: 0.0152\n",
      "Epoch 240/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0155 - val_loss: 0.0151\n",
      "Epoch 241/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0155 - val_loss: 0.0151\n",
      "Epoch 242/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0155 - val_loss: 0.0151\n",
      "Epoch 243/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0155 - val_loss: 0.0151\n",
      "Epoch 244/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0154 - val_loss: 0.0150\n",
      "Epoch 245/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0154 - val_loss: 0.0150\n",
      "Epoch 246/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0154 - val_loss: 0.0150\n",
      "Epoch 247/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0154 - val_loss: 0.0150\n",
      "Epoch 248/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0153 - val_loss: 0.0149\n",
      "Epoch 249/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0153 - val_loss: 0.0149\n",
      "Epoch 250/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0153 - val_loss: 0.0149\n",
      "Epoch 251/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0152 - val_loss: 0.0148\n",
      "Epoch 252/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0152 - val_loss: 0.0148\n",
      "Epoch 253/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0152 - val_loss: 0.0148\n",
      "Epoch 254/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0152 - val_loss: 0.0148\n",
      "Epoch 255/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0151 - val_loss: 0.0147\n",
      "Epoch 256/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0151 - val_loss: 0.0147\n",
      "Epoch 257/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0151 - val_loss: 0.0147\n",
      "Epoch 258/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0151 - val_loss: 0.0147\n",
      "Epoch 259/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0150 - val_loss: 0.0147\n",
      "Epoch 260/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0150 - val_loss: 0.0146\n",
      "Epoch 261/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0150 - val_loss: 0.0146\n",
      "Epoch 262/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0150 - val_loss: 0.0146\n",
      "Epoch 263/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0150 - val_loss: 0.0146\n",
      "Epoch 264/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0149 - val_loss: 0.0145\n",
      "Epoch 265/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0149 - val_loss: 0.0145\n",
      "Epoch 266/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0149 - val_loss: 0.0145\n",
      "Epoch 267/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0149 - val_loss: 0.0145\n",
      "Epoch 268/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0148 - val_loss: 0.0144\n",
      "Epoch 269/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0148 - val_loss: 0.0144\n",
      "Epoch 270/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0148 - val_loss: 0.0144\n",
      "Epoch 271/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0148 - val_loss: 0.0144\n",
      "Epoch 272/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0147 - val_loss: 0.0144\n",
      "Epoch 273/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0147 - val_loss: 0.0143\n",
      "Epoch 274/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0147 - val_loss: 0.0143\n",
      "Epoch 275/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0147 - val_loss: 0.0143\n",
      "Epoch 276/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0147 - val_loss: 0.0143\n",
      "Epoch 277/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0146 - val_loss: 0.0142\n",
      "Epoch 278/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0146 - val_loss: 0.0142\n",
      "Epoch 279/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0146 - val_loss: 0.0142\n",
      "Epoch 280/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0146 - val_loss: 0.0142\n",
      "Epoch 281/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0145 - val_loss: 0.0142\n",
      "Epoch 282/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0145 - val_loss: 0.0141\n",
      "Epoch 283/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0145 - val_loss: 0.0141\n",
      "Epoch 284/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0145 - val_loss: 0.0141\n",
      "Epoch 285/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0145 - val_loss: 0.0141\n",
      "Epoch 286/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0144 - val_loss: 0.0141\n",
      "Epoch 287/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0144 - val_loss: 0.0140\n",
      "Epoch 288/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0144 - val_loss: 0.0140\n",
      "Epoch 289/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0144 - val_loss: 0.0140\n",
      "Epoch 290/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0144 - val_loss: 0.0140\n",
      "Epoch 291/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0143 - val_loss: 0.0140\n",
      "Epoch 292/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0143 - val_loss: 0.0139\n",
      "Epoch 293/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0143 - val_loss: 0.0139\n",
      "Epoch 294/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0143 - val_loss: 0.0139\n",
      "Epoch 295/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0143 - val_loss: 0.0139\n",
      "Epoch 296/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0142 - val_loss: 0.0139\n",
      "Epoch 297/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0142 - val_loss: 0.0138\n",
      "Epoch 298/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0142 - val_loss: 0.0138\n",
      "Epoch 299/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0142 - val_loss: 0.0138\n",
      "Epoch 300/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0142 - val_loss: 0.0138\n",
      "Epoch 301/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0138\n",
      "Epoch 302/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0137\n",
      "Epoch 303/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0137\n",
      "Epoch 304/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0137\n",
      "Epoch 305/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0137\n",
      "Epoch 306/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0141 - val_loss: 0.0137\n",
      "Epoch 307/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0140 - val_loss: 0.0137\n",
      "Epoch 308/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0140 - val_loss: 0.0136\n",
      "Epoch 309/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0140 - val_loss: 0.0136\n",
      "Epoch 310/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0140 - val_loss: 0.0136\n",
      "Epoch 311/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0140 - val_loss: 0.0136\n",
      "Epoch 312/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0136\n",
      "Epoch 313/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0136\n",
      "Epoch 314/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0135\n",
      "Epoch 315/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0135\n",
      "Epoch 316/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0135\n",
      "Epoch 317/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0139 - val_loss: 0.0135\n",
      "Epoch 318/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0138 - val_loss: 0.0135\n",
      "Epoch 319/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0138 - val_loss: 0.0134\n",
      "Epoch 320/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0138 - val_loss: 0.0134\n",
      "Epoch 321/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0138 - val_loss: 0.0134\n",
      "Epoch 322/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0138 - val_loss: 0.0134\n",
      "Epoch 323/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0138 - val_loss: 0.0134\n",
      "Epoch 324/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0134\n",
      "Epoch 325/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0133\n",
      "Epoch 326/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0133\n",
      "Epoch 327/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0133\n",
      "Epoch 328/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0133\n",
      "Epoch 329/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0137 - val_loss: 0.0133\n",
      "Epoch 330/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0136 - val_loss: 0.0133\n",
      "Epoch 331/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0136 - val_loss: 0.0133\n",
      "Epoch 332/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0136 - val_loss: 0.0132\n",
      "Epoch 333/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0136 - val_loss: 0.0132\n",
      "Epoch 334/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0136 - val_loss: 0.0132\n",
      "Epoch 335/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0136 - val_loss: 0.0132\n",
      "Epoch 336/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0132\n",
      "Epoch 337/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0132\n",
      "Epoch 338/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0131\n",
      "Epoch 339/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0131\n",
      "Epoch 340/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0131\n",
      "Epoch 341/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0135 - val_loss: 0.0131\n",
      "Epoch 342/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0131\n",
      "Epoch 343/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0131\n",
      "Epoch 344/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0130\n",
      "Epoch 345/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0134 - val_loss: 0.0130\n",
      "Epoch 346/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0130\n",
      "Epoch 347/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0130\n",
      "Epoch 348/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0134 - val_loss: 0.0130\n",
      "Epoch 349/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0130\n",
      "Epoch 350/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0130\n",
      "Epoch 351/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0133 - val_loss: 0.0129\n",
      "Epoch 352/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0129\n",
      "Epoch 353/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0129\n",
      "Epoch 354/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0129\n",
      "Epoch 355/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0133 - val_loss: 0.0129\n",
      "Epoch 356/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0132 - val_loss: 0.0129\n",
      "Epoch 357/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0132 - val_loss: 0.0129\n",
      "Epoch 358/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0132 - val_loss: 0.0128\n",
      "Epoch 359/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0132 - val_loss: 0.0128\n",
      "Epoch 360/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0132 - val_loss: 0.0128\n",
      "Epoch 361/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0132 - val_loss: 0.0128\n",
      "Epoch 362/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0132 - val_loss: 0.0128\n",
      "Epoch 363/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0128\n",
      "Epoch 364/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0128\n",
      "Epoch 365/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0127\n",
      "Epoch 366/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0127\n",
      "Epoch 367/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0127\n",
      "Epoch 368/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0127\n",
      "Epoch 369/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0131 - val_loss: 0.0127\n",
      "Epoch 370/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0127\n",
      "Epoch 371/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0130 - val_loss: 0.0127\n",
      "Epoch 372/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0126\n",
      "Epoch 373/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0126\n",
      "Epoch 374/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0126\n",
      "Epoch 375/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0126\n",
      "Epoch 376/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0130 - val_loss: 0.0126\n",
      "Epoch 377/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0126\n",
      "Epoch 378/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0126\n",
      "Epoch 379/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0126\n",
      "Epoch 380/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0125\n",
      "Epoch 381/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0125\n",
      "Epoch 382/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0125\n",
      "Epoch 383/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0129 - val_loss: 0.0125\n",
      "Epoch 384/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0125\n",
      "Epoch 385/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0125\n",
      "Epoch 386/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0125\n",
      "Epoch 387/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0124\n",
      "Epoch 388/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0124\n",
      "Epoch 389/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0124\n",
      "Epoch 390/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0124\n",
      "Epoch 391/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0128 - val_loss: 0.0124\n",
      "Epoch 392/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0124\n",
      "Epoch 393/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0124\n",
      "Epoch 394/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0124\n",
      "Epoch 395/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 396/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 397/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 398/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 399/500\n",
      "58900/60000 [============================>.] - ETA: 0s - loss: 0.0127\n",
      "Epoch 00399: ReduceLROnPlateau reducing learning rate to 0.00010000000474974513.\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 400/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0127 - val_loss: 0.0123\n",
      "Epoch 401/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 402/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 403/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 404/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 405/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 406/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 407/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 408/500\n",
      "59600/60000 [============================>.] - ETA: 0s - loss: 0.0126\n",
      "Epoch 00408: ReduceLROnPlateau reducing learning rate to 1.0000000474974514e-05.\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 409/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 410/500\n",
      "60000/60000 [==============================] - 3s 44us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 411/500\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 412/500\n",
      "59700/60000 [============================>.] - ETA: 0s - loss: 0.0126Restoring model weights from the end of the best epoch.\n",
      "60000/60000 [==============================] - 3s 43us/sample - loss: 0.0126 - val_loss: 0.0123\n",
      "Epoch 00412: early stopping\n"
     ]
    }
   ],
   "source": [
    "epochs = 500\n",
    "\n",
    "autoencoder_train = autoencoder.fit(\n",
    "    x=x_train, \n",
    "    y=x_train,\n",
    "    batch_size=batch_size,\n",
    "    epochs=epochs, \n",
    "    shuffle=True,\n",
    "    validation_data=(x_test, x_test),\n",
    "    callbacks=cb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "saved = False\n",
    "if not saved:\n",
    "    autoencoder.save('./cae_model_{}.hdf5'.format(epochs))\n",
    "    encoder.save('./cae_encoder_{}.hdf5'.format(epochs))    \n",
    "    saved = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize autoencoder results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We choose ten random images from the testing dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "ind_test = np.random.choice(range(x_test.shape[0]), 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_images = np.take(x_test, ind_test, axis=0)\n",
    "preds = autoencoder.predict(test_images)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAC2CAYAAAB6fF5CAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAQkElEQVR4nO3deZTddXnH8c8zd2ayQSaZJJOEkIUkpARQloNCKRIKRgWCIEIRWhBBRMtRSzEiXdDTg9pWUKk5VKqyVNKyWVYRTEX2JRiIKSfBJJA9Q8ieyUJmub/+cX/puZ3nmWSSwHdm7rxf59yTmef3/O79zuQ3z3znfpefZVkmAEAaVV3dAADoTSi6AJAQRRcAEqLoAkBCFF0ASIiiCwAJUXQTMrNTzGxlV7cD6E7M7Ckz+3xXtyMVii5QQXpbAeuJKLp7ycyqu7oNAHruzyJFtxPMbKmZXWtm8yRtM7OTzOwFM9tkZr83s1PKcj9nZgvMrMnM3jKzK7uu5eipzOwbZvZmfh3NN7NP5fFvmdldZXnjzCwzs2oz+7akj0iaYWZbzWxGnnOimb1iZpvzf08sO7/OzH5mZo1mtsrMbjCzQn7sUjN7zsxuNLONZrbEzE4vO7fezG43s9X58QfLjl1hZovNbIOZPWxmB5Udm2pmb+TtmSHJ2n3tl+U/QxvN7AkzG1t2LDOzq8xskaRF7913PKEsy3js4SFpqaS5kkZLGiVpvaQzVPqlNTX/fFiee6akCSpdSFMkbZd0bH7sFEkru/rr4dH9H5LOl3RQfo1dIGmbpJGSviXprrK8cZIySdX5509J+nzZ8XpJGyVdLKla0oX550Py4w9KulXSAEkNkmZLujI/dqmkFklXSCpI+pKk1ZIsP/5LSfdIGiypRtKUPH6qpHWSjpXUR9KPJD2THxsqaYuk8/JzrpbUuqvNks6RtFjS5Ly9fyfphbKvJ5M0K/+6+nX1/9M+/d92dQN6wiMvupflH18r6eftjj8h6bMdnPugpK/mH1N0eezTI/+lf/Y+FN2LJc1u91wv5gV1uKSd5cUrL8q/zT++VNLismP989cakf8CKEoaHLT1Z5L+uezzA/LiPU7SJZJeKjtmklaWFd1fSbq87HiVSh2XsfnnmaRTu/r/Y38evL3QeSvyf8dKOj9/a2GTmW2SdJJKF6HM7HQzeyn/s2qTSj3ioV3TZPRUZnaJmc0tu8aO1L5dRwdJWtYutkylv9jGqtTbbCx7nVtV6vHu8vauD7Is255/eIBKf/VtyLJs455eM8uyrSr9NTgqP7ai7FhW/nneppvL2rNBpcI8qiynPL/H6ZFvRHeRXduxrVCpp3tF+wQz6yPpFyr9Nn8oy7KW/H0ua58LdCR/D/Mnkk6T9GKWZW1mNlel62ibSj3OXUa0O739toGrVSpk5cZIelyla3mnpKFZlrXuZTNXSKo3s0FZlm3a3Wua2QBJQyStktSoUsHedczKP8+f99tZls3czWv36K0R6enuvbsknWVmHzezgpn1zeffHiypVqX3sNZKas0HHT7WlY1FjzRApcKyVioNzqrU05VKbzOcbGZjzKxO0nXtzl0jaXzZ549JmmRmF+WDbRdIOlzSo1mWNUr6taSbzGygmVWZ2QQzm7KnBubn/krSLWY22MxqzOzk/PB/SPqcmR2dd0S+I+nlLMuWqvQ+8BFmdm4+++Ar+v+/OH4s6TozOyL/2uvM7Pw9fsd6EIruXsqybIVK7639jUo/FCskTZdUlWVZk0oX0b0qDVZcJOnhLmoqeqgsy+ZLukml917XSPqApOfzY7NUGryaJ2mOpEfbnX6zpPPykf9/ybJsvaRpkq5R6U/8r0ualmXZujz/EpU6C/NVumbvV/5WWSdcrNJ7tW9IekfSX+Vt/I2kv1fpr75GlQaWP5MfW6fSIOE/5u05dNfXlh9/QNI/SbrbzLZIel3S/82YqAS7RiEBAAnQ0wWAhCi6AJAQRRcAEqLoAkBCFF0ASGi3iyOmVp3P1Aa8r2YV7+uShSOfGHS5u7bbmpriZGb49E7W+UuzUD/YxR5fe2v4BPR0ASAhii4AJETRBYCE2PAGvVLb1m0+yHu3KLcX10Pbxs2dzqWnCwAJUXQBICGKLgAkRNEFgIQougCQELMX0Dtlxa5uASrJXlxP9HQBICGKLgAkRNEFgIQougCQEANpAPbdXmx/WNHLrPfia6OnCwAJUXQBICGKLgAkRNEFgIQougCQELMXgN6squBChbqBPm/ksPD0LZP9DRmLNfGMhkGvrvW5by0Lc7PW1jBeCejpAkBCFF0ASIiiCwAJUXQBICEG0spYdfztWH/Jh1ys3wVvu9gzH3ggPL9tL/ba/PTi011sx/SGOHn2/3T6edFOJS9JDZbmVo8YHqY2TxzpYktO7+tiH536Wnj+VcPudLEai6/3W9ZNcbF51x4V5tb895wwXgno6QJAQhRdAEiIogsACVF0ASAhii4AJNRrZy/sPNPPSDj+hlfC3BsaZrjYnJ0+796t9eH5967xr3V03cow976Jj7nYcSd+OcwdMTsMo5ewmto4/kfjXWz5GfG1uXW8X2571OS3XOzT9fHPxmE1fVysYHFf7sohz7jYtGnHhbkTn/TLk1VsC3N7Gnq6AJAQRRcAEqLoAkBCFF0ASKjiB9Kqx40J4x/57rMudumg34W5k355tYsd/s0VLtba6JcGS1LhcD+IcdIjT4a5kfoFzZ3ORYUKlvbu+MTRYWrjiX4QysZvDXMH9mlxsaYWPzhWY/EgVjRotr0YX6/zm0f411/UQb+vQgbNIvR0ASAhii4AJETRBYCEKLoAkBBFFwASqvjZCyc/siCM/3X9Gy525B3Tw9xJf/uii0X3Ku1oWebGm/xI7J/09aPGkjT56ctdbMKseANp9B7Vw/1G9qumxH2mquDirJp/QJi77QC/mXvT8H4utmiUn3kgSUfVLnGx15v97AdJmv7EhS522N3+51CSKnfuAj1dAEiKogsACVF0ASAhii4AJFRRA2nV48e52Nfq/yvMvbrxBBc75JvxnqGdvW/smvv9PqaSNPuDM11sYUu8VHLsTyt3H1HsWUeDsdFde9sGRcO5Uu0S/xzVO+LXa67zV3eh2t/N98134ztSv1yz3sWuX3h2mHvYj9a5WNuGjXHDKhg9XQBIiKILAAlRdAEgIYouACRE0QWAhCpq9sIZj85xsYUt74a5b57jR2Oz1lVhbjSiHM1UeO1Dd4fn78z8aPBFP7gmzB3x5AthHL1Eld+sXJJq3t7sYnXz/IwGSYr2G986Jp6D0zB5rYsN7+83PK+y+PxfbPB3890xK57pkK36fRDs7NygykFPFwASougCQEIUXQBIiKILAAlV1EDaobX+brwXfe9rYW7Dys4PWG2bdoyLzT7uFhdry+JBkCOe+oKLTbiZATN4WXO8PLy4zA/yHvSf8R1+Vef3zl17crwf7voxPve0kQtd7MBCPCDdWKxzsZb+cbNURR9PoqcLAElRdAEgIYouACRE0QWAhCi6AJBQRc1e+OFZ57hYw4L9nyVw4FN+NPfCtz7uYveM/3V4/oghfgln9SFjw9zWJcv2snWoKB0si82CTe/b1vlNwSVJ6ze4UMPO+O7TxerRLtZ4iJ+RcPBA/5ySdOxAf70+98F4M38bMcwHt3YwA6OClwfT0wWAhCi6AJAQRRcAEqLoAkBCFTWQ1rZg0fvzvBv9HUu3nznQxU6d8sXw/NYv+UGIv5z1WJg7/Y7LXOyQmfE+vwy69XIdDTZlfkPd1hUrw9ShP210scanD3GxmYcfEZ6/5njfb6vdGC+Hb23wuYW3/TJkSSo2NYXxSkBPFwASougCQEIUXQBIiKILAAlZtpuVH1Orzq/cZSFdrDBpQhhf+p1+Lvb4h38c5l722a/45/3tq/vXsMRmFe+LR13eZ1zbe8ni/6ZCnR9QtgEDwtysXx+fu3V7nLt9h4sVd8R7+kar9bqDjq5teroAkBBFFwASougCQEIUXQBIiKILAAlV1DLgnqRt4ZthfPR5PvZnfxHf0fiBO290sbOvnx7mDr7jxc43Dmivg1lObZv8XtHaEu+RWxhS788f0xDmrjvWz4qo3RK3oe6ReS5W3B7PiugO6OkCQEIUXQBIiKILAAlRdAEgIQbSeoC6u14K45+s9YNmD//D98Lcy+Z+wcWKc+fvX8OASNHv5ytJyoou9M6H/YCZJJ39xadd7LVN/iaaktTy8lDfhKXLd9PArkVPFwASougCQEIUXQBIiKILAAlRdAEgIWYv9GDrjvOjxEMLfhN0SWo90G8gzW9cpJSN8kt+685dHeZePni2iz3dd02YO7P2lP1qV2r83AFAQhRdAEiIogsACVF0ASAhBtJ6gO3nHh/GF5/9ry72Tpu/i6okVTftdDG/KBPYf4VBdWF88XU1Lnb7hAfC3Nrg7sNrWuLntQ727+2u6OkCQEIUXQBIiKILAAlRdAEgIYouACTE7IVuZvn1J7rYs1fEG5NLfsnvyXfHdwMeP5e7AeO9V9W3r4st+eoRYe79J3zfxQZVtYa5q1v9TIcZc04Jcyetf303Lex+6OkCQEIUXQBIiKILAAlRdAEgIQbSEqgeMdzF3rjxoDB34Z/OCKLxHrlnvPFJFxv/dQbMsJ+CJbiFof6Ou5K0+JqJLvaD824LcydW+z7eug7Wor+0Y7yLDX7e7wktSVlrS/wk3RQ9XQBIiKILAAlRdAEgIYouACRE0QWAhJi9sI8Kkw91sSUXDAtzp057xcUeHPlYmBsN5l7TeEKYm13vR5RNK8NcVKBgloEkWW2tixXqB4e5xQYfX36mj/35Z34Tnj9z8EMu1r/KL+GVpA1tzS72UFO8ZPi2W850sZF3x8t927IsjHdX9HQBICGKLgAkRNEFgIQougCQUMUPpFWPHBHGmyeNdLGdg+MBgJWf8nt+zjvtFhfrZ34AQ5KK8m/0/9vmcWHu7TdNc7Ghd/qBOEmy1rlhHD1EMBBW1b9/mFo1bIiLrTrr4DB38zF+wOrCY2aHuUcP8MvGj+6z2sUmVMdL0YvyS3O3Z/71JemepiNd7OZnPhbmTr7nDy7WtmVLmNvT0NMFgIQougCQEEUXABKi6AJAQhRdAEio289e2HKhXwI7+qpFYe55Db9zsXE18cj/MbX7+/vGz3R49t342/ndpWe4WOHL8Sj1kPl+NLlnLXKE08Fy3ao+fuR/x5TDw9zlp/vr9RsffTDMHVez1sUm1GyM2xDEtmcFF/vJ5tHh+Xcs/WMX2zgnXg4/4efvuNhhjQvC3LampjBeCejpAkBCFF0ASIiiCwAJUXQBIKFuP5BW/6JfkviHCxrC3Of6TnKxGetPDXOLmR/cWLXML7WUpAOHb3Uxe8rvOTrqtni/T23xe9y2xZnoTaqCPk885qaaTT7335fF+yy3Fn1uoSq+7e6OZj8gvGWhv7ZHz/JL4SWp/tVlLjZo0/Iwt60lXh7c29DTBYCEKLoAkBBFFwASougCQEIUXQBIqNvPXmhd6kdCR5wT5/ptj6V+WtLp15qktzqdG2FGAkId3K22+O5OF+v//MIwd+KrwSbi1fGPb7Z1m481xzMHBgZta2h+05/fFl/dPe1OvO+bDpZ6R+jpAkBCFF0ASIiiCwAJUXQBIKFuP5AGVKyiH5xq27Q5zu0oju7BOt9/pacLAAlRdAEgIYouACRE0QWAhCi6AJAQsxfQK1mwhDZrjTfqTqqj5aQst+0eqvydkiWpUDew80/xXrUFALBnFF0ASIiiCwAJUXQBICHLeIMeAJKhpwsACVF0ASAhii4AJETRBYCEKLoAkBBFFwAS+l8Avi4ObMlzJAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAACmCAYAAAB5qlzZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAI7ElEQVR4nO3dW4xdVRkH8H3OzDBDLyBUoJZeLB0NtyKRtIAJXkoALUSNxqBGINFoqCLRYEx8kKTGh2JJJKRRhEQrKSDXB8DElEJCQCj3mHIRig0SQUFKmXZm2jIz5/hq8q0p5zAz3wynv9/jv+vs7k5O/rPStfbatWazWQGQoz7dNwBwMFG6AImULkAipQuQSOkCJFK6AIm6D/SH59S/Zj8ZU+q+xu216fh7zzv82+G73RgcLA+2rZL3UOvtDdnmvZuK320zXYBEShcgkdIFSHTA/9OFTtUYGo6h/7vlfWru39/yWDNdgERKFyCR0gVIpHQBEildgER2L3Bwajam+w6oqqqqdxXjWj0+zNUcHZ3qu0lhpguQSOkCJFK6AImULkAiC2nA5KqVT+vsXrwwZAMrFhTHdu2Pj2TPfvDvxbFju3e3cXPTz0wXIJHSBUikdAESKV2AREoXIJHdCzPMjnVnhuz5iza0/PmTbrysmC/92aPv+56gHbVDDinm2y+NuxdWn/tEcey7jVhNz609pTi2797H27i76WemC5BI6QIkUroAiZQuQCILadOk1lNebGgs3Bezqo2zX73QtjXe/Dt1Tuwvxr/86s0hO6PvteLYHSOHhezyj3+yOPYjbdzaTGCmC5BI6QIkUroAiZQuQCKlC5DI7oUEpZ0KO9aeVhz7/Oeubfm69w7NC9nSu4davzGYoNJ3e/vFc4tjV896I2T1qryL54GRD4fs8FfG2ry7mclMFyCR0gVIpHQBEildgEQW0hLsPe/UkG27pPUFs/GsX/vNkB2+deuErwut6lpwTMh+cM7m4tg59b6Q/WNksDh27UNfCtkJD4zzNuAD3eAMZKYLkEjpAiRSugCJlC5AIqULkMjuhQRvLfdjpjMNrFgQspP7/lIcu2tsOGS/eevTxbEf+/27IRsb2N3m3c1MZroAiZQuQCKlC5BI6QIkssIzibrmHVnML/3Wnyd03TfG9hfznr3eaEuOWm9vMZ+9Jr7Nd373nuLYbSOzQnb3/acXxy578qmQNTvkDc5mugCJlC5AIqULkEjpAiRSugCJ7F6YRN13ld9s+r0PvTyh655/7U+L+YK7HpnQdQ9qtVrMOmR1fMLqXSH69/fLb6++vX99yI6qF362VVU9MnxsyOZvbRTHNkfiY8CdwkwXIJHSBUikdAESKV2ARBbSJtGd/eXHfctLBWXXvH1iyBbf8kpx7Ggb14VW7b5wRciuXLOpOHZZ96Eh29nYWxx71aNfCNkJD24vjv2gveG3HWa6AImULkAipQuQSOkCJFK6AInsXmhBffbskL247uTCyHjw8nj2Nct7Dzbefk7IFr/mcV+mRr2vL2RDFw6E7NTe14ufb1Rx98L2kZhVVVUt2xT38Yy9veu9brHjmOkCJFK6AImULkAipQuQyEJaC0qLZi98ZUNhZOu/w/45Wj5zdPEvLJoxBUrnB1dVNbj6EyG7dvkNIZtV/nj1UuHc20u2rimO7X/42ZB1yht+22GmC5BI6QIkUroAiZQuQCILaS04e2VcAKgXfl/11OIL/aqqqjYPx6d+1l16cXFsTxtPtUGruucfU8yPvSKeZ7uyd1/IRqryd/vGgVNCtujGnuLYTn7ZZDvMdAESKV2AREoXIJHSBUikdAES2b3wf1698lPF/Lr5vwpZo+oN2cg4TzRe/vRFIVuyxS6FadXJj58WHvnd/sOlxaFbFq0P2az6nJC9NDJU/PzGO+L5z0vuf7I4toN/4m0x0wVIpHQBEildgERKFyDRQbuQ9t81Z4bsoe/GRYWqqqq59bhoVjLQiI9PVlVVLbm69fuCiaofGl8MedaqbcWxR9ZjBQwWvsc3vbOy+Pmld+wM2ZjHfQ/ITBcgkdIFSKR0ARIpXYBEShcg0UG7e+Gdkxshm1s/ZELXXP3znxTzIx5/dELXhaJx3vC7+4J4sPhV868pjt3TGA3Z9bviToXN684qfv6w5x870B1SYKYLkEjpAiRSugCJlC5Aoo5fSNvz9TOK+T3n/7qQtv7jWL9zeciO2GjBjDzvnntaMf/ylfeFbEl3+TTbN8biYtxtt342ZIvvLT9G3Ojkc4mniJkuQCKlC5BI6QIkUroAiZQuQKKO2r1Qnzs3ZAsv214c29/T+j99oBEPZb7z+lUhO7p6pOVrQjtqp50Ush9vuLk4dkXvmyHrqcqPuD8w1B+yj/7p9ZCN7tnzXrdIi8x0ARIpXYBEShcgkdIFSNRRC2kv/25ZyJ5desOEr3vmHVeErH+DRTMmX/eihcW8efWukK069O3i2Hph0WykGiuOfWL30pA1/hMX4pg8ZroAiZQuQCKlC5BI6QIkUroAiWb87oV6X1/IXr0p7lKoqqp65vTrCmnXhO9h0Zbyyi9MRK23N2Rv/nZWcexNx/0hZHPqs4tjBxv7Qvbg3nnFsc/8MR7Gf9Sww/inkpkuQCKlC5BI6QIkUroAiWb8QlqtcEbu02dsHGf0xBbNTnn4O8V82WM7QmZpjYnqWjA/ZBcserY4dlYtvnV3uHDOc1VV1etj8dv5o79+ozj2+NteDJnv9tQy0wVIpHQBEildgERKFyCR0gVINON3LzSHhkJ29rYLi2PvX35ry9ddVbjGcZe8VBw7ti8+Vgktq9WKcXPPYMg2PbeyfI34MuDqmJ6B4tD1T50bshPWvlUcO7qzfBA6U8dMFyCR0gVIpHQBEildgEQzfiGtMTwcstmfj4/lVlVVfbFa0fJ151TxGo3Wbwta14yP8FZVVY0VFrH6rzq6OPaelZ+J4Thf2OO3/Ctko6/GjEk0zmJpiZkuQCKlC5BI6QIkUroAiZQuQKIZv3sBOlZhV0Pjby8Uh87bVjigv1nevjA6zm4JplCt9fmrmS5AIqULkEjpAiRSugCJLKTBB0HDO3pnslpX628iN9MFSKR0ARIpXYBEShcgkdIFSGT3AgelWnf86jdHR6fhTkjVxmHjxR0J4+xS6DryiJava6YLkEjpAiRSugCJlC5AolrT2ZsAacx0ARIpXYBEShcgkdIFSKR0ARIpXYBE/wNC3psBg4p+WAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "has_title = False\n",
    "for im, pred in zip(test_images, preds):\n",
    "    plt.figure()\n",
    "    \n",
    "    ax1 = plt.subplot(1,2,1)\n",
    "    plt.imshow(np.squeeze(im))\n",
    "    \n",
    "    ax2 = plt.subplot(1,2,2)\n",
    "    plt.imshow(np.squeeze(pred))\n",
    "    \n",
    "    ax1.axis('off')\n",
    "    ax2.axis('off')    \n",
    "    if not has_title:\n",
    "        ax1.set_title('real')\n",
    "        ax2.set_title('autoencoder')\n",
    "        has_title = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Latent space visualization\n",
    "\n",
    "In this particular case since we are using a very simple convolutional netwok the different dimensions of the latent space are going to look very similar to the input.\n",
    "\n",
    "Probably are edge detections and directions filters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x216 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_latents(input_image):\n",
    "    test_ar = input_image\n",
    "    pred = autoencoder.predict(test_ar)\n",
    "    \n",
    "    for real, coded in zip(test_ar, pred):\n",
    "        plt.figure(figsize = (2,4))\n",
    "        plt.subplot(1,2,1)\n",
    "        plt.imshow(real[:,:,0])\n",
    "        plt.subplot(1,2,2)\n",
    "        plt.imshow(coded[:,:,0])   \n",
    "    \n",
    "    coded_image = encoder.predict(test_ar)\n",
    "    coded_dims = coded_image.shape[-1]\n",
    "\n",
    "    plt.figure(figsize = (5,3))\n",
    "    for i in range(coded_dims):\n",
    "        plt.subplot(int(np.ceil(coded_dims)/4),4, i+1)\n",
    "        plt.imshow(coded_image[0, :, :, i])\n",
    "    plt.tight_layout()    \n",
    "    \n",
    "    return pred, coded_image\n",
    "\n",
    "_ ,_ = plot_latents(x_test[0:1,:,:,:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### More latent space visualizations - TSNE\n",
    "\n",
    "TSNE uses 2d data as input, so we reshape the data from 4d to 2d.\n",
    "Then we apply TSNE for dimensionality reduction Nsamples, Nd --> Nsamples, 2 dimensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "latent = encoder.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_1d = np.reshape(latent,  (latent.shape[0], int(reduce(lambda x, y: x*y, np.asarray(latent.shape[1:])))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "import seaborn as sns\n",
    "from sklearn.metrics import silhouette_score\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "latent_tsne = TSNE(n_components=2, random_state=2).fit_transform(latent_1d)\n",
    "latent_data = pd.DataFrame({\n",
    "        'x': latent_tsne[:, 0],\n",
    "        'y': latent_tsne[:, 1],\n",
    "        'label': y_test})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "score: 0.3635343611240387\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('score: {}'.format(silhouette_score(latent_tsne, latent_data['label'] )))\n",
    "fig = plt.figure(figsize=(8, 8))\n",
    "ax = sns.scatterplot(x='x', y='y', hue='label', data=latent_data, palette='Paired')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}