13.2 Variational Autoencoder.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from keras.layers import Input, Dense, Lambda\n",
    "from keras.layers.merge import concatenate as concat\n",
    "from keras.models import Model\n",
    "from keras import backend as K\n",
    "from keras.datasets import mnist\n",
    "from keras.utils import to_categorical\n",
    "from keras.callbacks import EarlyStopping\n",
    "from keras.optimizers import Adam\n",
    "\n",
    "from io import BytesIO\n",
    "import PIL\n",
    "from IPython.display import clear_output, Image, display, HTML"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prepare(images, labels):\n",
    "    images = images.astype('float32') / 255\n",
    "    n, w, h = images.shape\n",
    "    return images.reshape((n, w * h)), to_categorical(labels)\n",
    "\n",
    "train, test = mnist.load_data()\n",
    "x_train, y_train = prepare(*train)\n",
    "x_test, y_test = prepare(*test)\n",
    "img_width, img_height = train[0].shape[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 250\n",
    "latent_space_depth = 2\n",
    "\n",
    "def sample_z(args):\n",
    "    z_mean, z_log_var = args\n",
    "    eps = K.random_normal(shape=(batch_size, latent_space_depth), mean=0., stddev=1.)\n",
    "    return z_mean + K.exp(z_log_var / 2) * eps\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_17 (InputLayer)           (None, 784)          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "dense_32 (Dense)                (None, 512)          401920      input_17[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_33 (Dense)                (None, 2)            1026        dense_32[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_34 (Dense)                (None, 2)            1026        dense_32[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "lambda_8 (Lambda)               (None, 2)            0           dense_33[0][0]                   \n",
      "                                                                 dense_34[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_35 (Dense)                (None, 512)          1536        lambda_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_36 (Dense)                (None, 784)          402192      dense_35[1][0]                   \n",
      "==================================================================================================\n",
      "Total params: 807,700\n",
      "Trainable params: 807,700\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "def VariationalAutoEncoder(num_pixels):\n",
    "    pixels = Input(shape=(num_pixels,))\n",
    "    encoder_hidden = Dense(512, activation='relu')(pixels)\n",
    "\n",
    "    z_mean = Dense(latent_space_depth, activation='linear')(encoder_hidden)\n",
    "    z_log_var = Dense(latent_space_depth, activation='linear')(encoder_hidden)\n",
    "    \n",
    "    def KL_loss(y_true, y_pred):\n",
    "        return(0.5 * K.sum(K.exp(z_log_var) + K.square(z_mean) - 1 - z_log_var, axis=1))\n",
    "\n",
    "    def reconstruction_loss(y_true, y_pred):\n",
    "        return K.sum(K.binary_crossentropy(y_true, y_pred), axis=-1)    \n",
    "\n",
    "    def total_loss(y_true, y_pred):\n",
    "        return KL_loss(y_true, y_pred) + reconstruction_loss(y_true, y_pred)\n",
    "\n",
    "    z = Lambda(sample_z, output_shape=(latent_space_depth, ))([z_mean, z_log_var])\n",
    "    \n",
    "    decoder_hidden = Dense(512, activation='relu')\n",
    "\n",
    "    reconstruct_pixels = Dense(num_pixels, activation='sigmoid')\n",
    "\n",
    "    decoder_in = Input(shape=(latent_space_depth,))\n",
    "    hidden = decoder_hidden(decoder_in)\n",
    "    decoder_out = reconstruct_pixels(hidden)\n",
    "    decoder = Model(decoder_in, decoder_out)    \n",
    "    \n",
    "    hidden = decoder_hidden(z)\n",
    "    outputs = reconstruct_pixels(hidden)\n",
    "    auto_encoder = Model(pixels, outputs)\n",
    "\n",
    "    auto_encoder.compile(optimizer=Adam(lr=0.001), \n",
    "                         loss=total_loss,\n",
    "                         metrics=[KL_loss, reconstruction_loss])\n",
    "    \n",
    "    return auto_encoder, decoder\n",
    "\n",
    "auto_encoder, decoder = VariationalAutoEncoder(x_train.shape[1])\n",
    "auto_encoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60000, 784), (60000, 10), (10000, 784), (10000, 10))"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape, y_train.shape, x_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/100\n",
      "60000/60000 [==============================] - 7s 125us/step - loss: 214.1281 - KL_loss: 10.7052 - reconstruction_loss: 203.4229 - val_loss: 181.8244 - val_KL_loss: 5.3993 - val_reconstruction_loss: 176.4251\n",
      "Epoch 2/100\n",
      "60000/60000 [==============================] - 6s 101us/step - loss: 176.0493 - KL_loss: 4.8653 - reconstruction_loss: 171.1839 - val_loss: 170.5538 - val_KL_loss: 4.8051 - val_reconstruction_loss: 165.7487\n",
      "Epoch 3/100\n",
      "60000/60000 [==============================] - 7s 109us/step - loss: 168.2982 - KL_loss: 4.8428 - reconstruction_loss: 163.4554 - val_loss: 166.6530 - val_KL_loss: 4.8240 - val_reconstruction_loss: 161.8290\n",
      "Epoch 4/100\n",
      "60000/60000 [==============================] - 6s 106us/step - loss: 165.5347 - KL_loss: 4.9653 - reconstruction_loss: 160.5694 - val_loss: 164.8385 - val_KL_loss: 4.9888 - val_reconstruction_loss: 159.8497\n",
      "Epoch 5/100\n",
      "60000/60000 [==============================] - 7s 114us/step - loss: 163.7700 - KL_loss: 5.0797 - reconstruction_loss: 158.6903 - val_loss: 163.0991 - val_KL_loss: 5.0738 - val_reconstruction_loss: 158.0253\n",
      "Epoch 6/100\n",
      "60000/60000 [==============================] - 8s 131us/step - loss: 162.2533 - KL_loss: 5.1901 - reconstruction_loss: 157.0632 - val_loss: 161.8445 - val_KL_loss: 5.2710 - val_reconstruction_loss: 156.5735\n",
      "Epoch 7/100\n",
      "60000/60000 [==============================] - 8s 128us/step - loss: 160.9642 - KL_loss: 5.2981 - reconstruction_loss: 155.6660 - val_loss: 160.6033 - val_KL_loss: 5.4190 - val_reconstruction_loss: 155.1843\n",
      "Epoch 8/100\n",
      "60000/60000 [==============================] - 7s 122us/step - loss: 159.8441 - KL_loss: 5.3988 - reconstruction_loss: 154.4453 - val_loss: 159.6930 - val_KL_loss: 5.5321 - val_reconstruction_loss: 154.1609\n",
      "Epoch 9/100\n",
      "60000/60000 [==============================] - 7s 122us/step - loss: 158.8782 - KL_loss: 5.4757 - reconstruction_loss: 153.4025 - val_loss: 158.7872 - val_KL_loss: 5.6106 - val_reconstruction_loss: 153.1766\n",
      "Epoch 10/100\n",
      "60000/60000 [==============================] - 7s 124us/step - loss: 158.0099 - KL_loss: 5.5404 - reconstruction_loss: 152.4695 - val_loss: 158.1441 - val_KL_loss: 5.5470 - val_reconstruction_loss: 152.5971\n",
      "Epoch 11/100\n",
      "60000/60000 [==============================] - 8s 127us/step - loss: 157.2516 - KL_loss: 5.6011 - reconstruction_loss: 151.6505 - val_loss: 157.5096 - val_KL_loss: 5.6985 - val_reconstruction_loss: 151.8111\n",
      "Epoch 12/100\n",
      "60000/60000 [==============================] - 8s 137us/step - loss: 156.5844 - KL_loss: 5.6358 - reconstruction_loss: 150.9486 - val_loss: 156.8968 - val_KL_loss: 5.7259 - val_reconstruction_loss: 151.1709\n",
      "Epoch 13/100\n",
      "60000/60000 [==============================] - 10s 158us/step - loss: 155.9854 - KL_loss: 5.6965 - reconstruction_loss: 150.2889 - val_loss: 156.3681 - val_KL_loss: 5.6687 - val_reconstruction_loss: 150.6994\n",
      "Epoch 14/100\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 155.3902 - KL_loss: 5.7333 - reconstruction_loss: 149.6570 - val_loss: 156.1155 - val_KL_loss: 5.7443 - val_reconstruction_loss: 150.3712\n",
      "Epoch 15/100\n",
      "60000/60000 [==============================] - 8s 136us/step - loss: 154.8850 - KL_loss: 5.7563 - reconstruction_loss: 149.1288 - val_loss: 155.6243 - val_KL_loss: 5.8168 - val_reconstruction_loss: 149.8075\n",
      "Epoch 16/100\n",
      "60000/60000 [==============================] - 8s 132us/step - loss: 154.4619 - KL_loss: 5.7957 - reconstruction_loss: 148.6662 - val_loss: 155.1267 - val_KL_loss: 5.7830 - val_reconstruction_loss: 149.3437\n",
      "Epoch 17/100\n",
      "60000/60000 [==============================] - 9s 151us/step - loss: 153.9827 - KL_loss: 5.8103 - reconstruction_loss: 148.1724 - val_loss: 154.8084 - val_KL_loss: 5.8179 - val_reconstruction_loss: 148.9904\n",
      "Epoch 18/100\n",
      "60000/60000 [==============================] - 9s 144us/step - loss: 153.5974 - KL_loss: 5.8243 - reconstruction_loss: 147.7730 - val_loss: 154.6081 - val_KL_loss: 5.8155 - val_reconstruction_loss: 148.7927\n",
      "Epoch 19/100\n",
      "60000/60000 [==============================] - 8s 138us/step - loss: 153.2694 - KL_loss: 5.8501 - reconstruction_loss: 147.4193 - val_loss: 154.3049 - val_KL_loss: 5.9876 - val_reconstruction_loss: 148.3173\n",
      "Epoch 20/100\n",
      "60000/60000 [==============================] - 10s 168us/step - loss: 152.9287 - KL_loss: 5.8673 - reconstruction_loss: 147.0614 - val_loss: 154.1877 - val_KL_loss: 5.8098 - val_reconstruction_loss: 148.3779\n",
      "Epoch 21/100\n",
      "60000/60000 [==============================] - 10s 171us/step - loss: 152.6310 - KL_loss: 5.8868 - reconstruction_loss: 146.7443 - val_loss: 153.7143 - val_KL_loss: 5.9013 - val_reconstruction_loss: 147.8129\n",
      "Epoch 22/100\n",
      "60000/60000 [==============================] - 8s 138us/step - loss: 152.3065 - KL_loss: 5.9001 - reconstruction_loss: 146.4064 - val_loss: 153.5815 - val_KL_loss: 5.7725 - val_reconstruction_loss: 147.8090\n",
      "Epoch 23/100\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 152.0398 - KL_loss: 5.9063 - reconstruction_loss: 146.1335 - val_loss: 153.5477 - val_KL_loss: 5.9820 - val_reconstruction_loss: 147.5656\n",
      "Epoch 24/100\n",
      "60000/60000 [==============================] - 9s 150us/step - loss: 151.7811 - KL_loss: 5.9320 - reconstruction_loss: 145.8491 - val_loss: 153.4320 - val_KL_loss: 5.8206 - val_reconstruction_loss: 147.6113\n",
      "Epoch 25/100\n",
      "60000/60000 [==============================] - 9s 147us/step - loss: 151.5322 - KL_loss: 5.9397 - reconstruction_loss: 145.5926 - val_loss: 153.1780 - val_KL_loss: 5.9838 - val_reconstruction_loss: 147.1942\n",
      "Epoch 26/100\n",
      "60000/60000 [==============================] - 9s 152us/step - loss: 151.3158 - KL_loss: 5.9545 - reconstruction_loss: 145.3612 - val_loss: 153.0342 - val_KL_loss: 5.9378 - val_reconstruction_loss: 147.0963\n",
      "Epoch 27/100\n",
      "60000/60000 [==============================] - 9s 145us/step - loss: 151.0805 - KL_loss: 5.9680 - reconstruction_loss: 145.1125 - val_loss: 152.8495 - val_KL_loss: 5.9262 - val_reconstruction_loss: 146.9233\n",
      "Epoch 28/100\n",
      "60000/60000 [==============================] - 8s 138us/step - loss: 150.8912 - KL_loss: 5.9827 - reconstruction_loss: 144.9086 - val_loss: 152.7259 - val_KL_loss: 5.9263 - val_reconstruction_loss: 146.7997\n",
      "Epoch 29/100\n",
      "60000/60000 [==============================] - 9s 145us/step - loss: 150.6710 - KL_loss: 5.9967 - reconstruction_loss: 144.6743 - val_loss: 152.5021 - val_KL_loss: 5.9333 - val_reconstruction_loss: 146.5687\n",
      "Epoch 30/100\n",
      "60000/60000 [==============================] - 9s 153us/step - loss: 150.4795 - KL_loss: 6.0001 - reconstruction_loss: 144.4794 - val_loss: 152.5714 - val_KL_loss: 5.9258 - val_reconstruction_loss: 146.6456\n",
      "Epoch 31/100\n",
      "60000/60000 [==============================] - 10s 160us/step - loss: 150.2969 - KL_loss: 6.0087 - reconstruction_loss: 144.2881 - val_loss: 152.3901 - val_KL_loss: 5.9065 - val_reconstruction_loss: 146.4836\n",
      "Epoch 32/100\n",
      "60000/60000 [==============================] - 10s 160us/step - loss: 150.1377 - KL_loss: 6.0186 - reconstruction_loss: 144.1191 - val_loss: 152.2293 - val_KL_loss: 5.9372 - val_reconstruction_loss: 146.2921\n",
      "Epoch 33/100\n",
      "60000/60000 [==============================] - 9s 153us/step - loss: 149.9040 - KL_loss: 6.0331 - reconstruction_loss: 143.8709 - val_loss: 152.0998 - val_KL_loss: 6.0264 - val_reconstruction_loss: 146.0734\n",
      "Epoch 34/100\n",
      "60000/60000 [==============================] - 9s 151us/step - loss: 149.8222 - KL_loss: 6.0427 - reconstruction_loss: 143.7795 - val_loss: 152.2233 - val_KL_loss: 6.2152 - val_reconstruction_loss: 146.0081\n",
      "Epoch 35/100\n",
      "60000/60000 [==============================] - 8s 137us/step - loss: 149.6705 - KL_loss: 6.0659 - reconstruction_loss: 143.6046 - val_loss: 152.2301 - val_KL_loss: 6.0097 - val_reconstruction_loss: 146.2204\n",
      "Epoch 36/100\n",
      "60000/60000 [==============================] - 8s 128us/step - loss: 149.4652 - KL_loss: 6.0585 - reconstruction_loss: 143.4067 - val_loss: 151.9259 - val_KL_loss: 6.0785 - val_reconstruction_loss: 145.8475\n",
      "Epoch 37/100\n",
      "60000/60000 [==============================] - 7s 124us/step - loss: 149.2914 - KL_loss: 6.0748 - reconstruction_loss: 143.2166 - val_loss: 151.7779 - val_KL_loss: 6.0048 - val_reconstruction_loss: 145.7731\n",
      "Epoch 38/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 8s 129us/step - loss: 149.1922 - KL_loss: 6.0761 - reconstruction_loss: 143.1161 - val_loss: 152.0566 - val_KL_loss: 6.2925 - val_reconstruction_loss: 145.7641\n",
      "Epoch 39/100\n",
      "60000/60000 [==============================] - 8s 128us/step - loss: 149.0389 - KL_loss: 6.0863 - reconstruction_loss: 142.9527 - val_loss: 151.9103 - val_KL_loss: 5.9954 - val_reconstruction_loss: 145.9149\n",
      "Epoch 40/100\n",
      "60000/60000 [==============================] - 9s 156us/step - loss: 148.9012 - KL_loss: 6.0979 - reconstruction_loss: 142.8032 - val_loss: 151.8481 - val_KL_loss: 5.9871 - val_reconstruction_loss: 145.8609\n",
      "Epoch 41/100\n",
      "60000/60000 [==============================] - 11s 180us/step - loss: 148.7529 - KL_loss: 6.0963 - reconstruction_loss: 142.6566 - val_loss: 151.6686 - val_KL_loss: 6.1352 - val_reconstruction_loss: 145.5333\n",
      "Epoch 42/100\n",
      "60000/60000 [==============================] - 9s 154us/step - loss: 148.6706 - KL_loss: 6.1164 - reconstruction_loss: 142.5543 - val_loss: 151.7423 - val_KL_loss: 6.0439 - val_reconstruction_loss: 145.6984\n",
      "Epoch 43/100\n",
      "60000/60000 [==============================] - 8s 138us/step - loss: 148.5339 - KL_loss: 6.1186 - reconstruction_loss: 142.4153 - val_loss: 151.3352 - val_KL_loss: 6.0629 - val_reconstruction_loss: 145.2722\n",
      "Epoch 44/100\n",
      "60000/60000 [==============================] - 8s 136us/step - loss: 148.3679 - KL_loss: 6.1205 - reconstruction_loss: 142.2474 - val_loss: 151.4178 - val_KL_loss: 6.0571 - val_reconstruction_loss: 145.3607\n",
      "Epoch 45/100\n",
      "60000/60000 [==============================] - 8s 133us/step - loss: 148.3103 - KL_loss: 6.1273 - reconstruction_loss: 142.1830 - val_loss: 151.3546 - val_KL_loss: 6.1504 - val_reconstruction_loss: 145.2043\n",
      "Epoch 46/100\n",
      "60000/60000 [==============================] - 8s 131us/step - loss: 148.1483 - KL_loss: 6.1318 - reconstruction_loss: 142.0165 - val_loss: 151.2847 - val_KL_loss: 6.0230 - val_reconstruction_loss: 145.2617\n",
      "Epoch 47/100\n",
      "60000/60000 [==============================] - 8s 141us/step - loss: 148.0540 - KL_loss: 6.1380 - reconstruction_loss: 141.9160 - val_loss: 151.3765 - val_KL_loss: 6.0414 - val_reconstruction_loss: 145.3351\n",
      "Epoch 48/100\n",
      "60000/60000 [==============================] - 8s 139us/step - loss: 147.9247 - KL_loss: 6.1557 - reconstruction_loss: 141.7689 - val_loss: 151.2288 - val_KL_loss: 6.1175 - val_reconstruction_loss: 145.1113\n",
      "Epoch 49/100\n",
      "60000/60000 [==============================] - 9s 147us/step - loss: 147.8347 - KL_loss: 6.1565 - reconstruction_loss: 141.6782 - val_loss: 151.3949 - val_KL_loss: 6.1055 - val_reconstruction_loss: 145.2894\n",
      "Epoch 50/100\n",
      "60000/60000 [==============================] - 9s 150us/step - loss: 147.7358 - KL_loss: 6.1759 - reconstruction_loss: 141.5599 - val_loss: 151.1783 - val_KL_loss: 6.1920 - val_reconstruction_loss: 144.9863\n",
      "Epoch 51/100\n",
      "60000/60000 [==============================] - 9s 142us/step - loss: 147.6607 - KL_loss: 6.1708 - reconstruction_loss: 141.4899 - val_loss: 151.5140 - val_KL_loss: 6.2012 - val_reconstruction_loss: 145.3128\n",
      "Epoch 52/100\n",
      "60000/60000 [==============================] - 8s 136us/step - loss: 147.5701 - KL_loss: 6.1877 - reconstruction_loss: 141.3824 - val_loss: 151.1871 - val_KL_loss: 6.1699 - val_reconstruction_loss: 145.0172\n",
      "Epoch 53/100\n",
      "60000/60000 [==============================] - 8s 135us/step - loss: 147.4660 - KL_loss: 6.1894 - reconstruction_loss: 141.2766 - val_loss: 151.0802 - val_KL_loss: 6.1510 - val_reconstruction_loss: 144.9291\n",
      "Epoch 54/100\n",
      "60000/60000 [==============================] - 9s 153us/step - loss: 147.3568 - KL_loss: 6.1783 - reconstruction_loss: 141.1785 - val_loss: 150.9429 - val_KL_loss: 6.0097 - val_reconstruction_loss: 144.9331\n",
      "Epoch 55/100\n",
      "60000/60000 [==============================] - 9s 152us/step - loss: 147.2651 - KL_loss: 6.1942 - reconstruction_loss: 141.0708 - val_loss: 150.9385 - val_KL_loss: 6.0831 - val_reconstruction_loss: 144.8555\n",
      "Epoch 56/100\n",
      "60000/60000 [==============================] - 10s 163us/step - loss: 147.1715 - KL_loss: 6.1994 - reconstruction_loss: 140.9722 - val_loss: 150.9214 - val_KL_loss: 6.0060 - val_reconstruction_loss: 144.9155\n",
      "Epoch 57/100\n",
      "60000/60000 [==============================] - 9s 144us/step - loss: 147.1210 - KL_loss: 6.2103 - reconstruction_loss: 140.9107 - val_loss: 151.0507 - val_KL_loss: 6.1011 - val_reconstruction_loss: 144.9496\n",
      "Epoch 58/100\n",
      "60000/60000 [==============================] - 9s 145us/step - loss: 146.9795 - KL_loss: 6.2118 - reconstruction_loss: 140.7677 - val_loss: 150.9784 - val_KL_loss: 6.1654 - val_reconstruction_loss: 144.8130\n",
      "Epoch 59/100\n",
      "60000/60000 [==============================] - 9s 142us/step - loss: 146.9154 - KL_loss: 6.2062 - reconstruction_loss: 140.7092 - val_loss: 151.1834 - val_KL_loss: 6.0619 - val_reconstruction_loss: 145.1214\n",
      "Epoch 60/100\n",
      "60000/60000 [==============================] - 9s 142us/step - loss: 146.8608 - KL_loss: 6.2258 - reconstruction_loss: 140.6351 - val_loss: 151.0328 - val_KL_loss: 6.2836 - val_reconstruction_loss: 144.7491\n",
      "Epoch 61/100\n",
      "60000/60000 [==============================] - 8s 127us/step - loss: 146.7216 - KL_loss: 6.2231 - reconstruction_loss: 140.4984 - val_loss: 150.7960 - val_KL_loss: 6.1415 - val_reconstruction_loss: 144.6545\n",
      "Epoch 62/100\n",
      "60000/60000 [==============================] - 8s 129us/step - loss: 146.6610 - KL_loss: 6.2241 - reconstruction_loss: 140.4370 - val_loss: 150.5898 - val_KL_loss: 6.2175 - val_reconstruction_loss: 144.3723\n",
      "Epoch 63/100\n",
      "60000/60000 [==============================] - 8s 129us/step - loss: 146.6313 - KL_loss: 6.2409 - reconstruction_loss: 140.3904 - val_loss: 151.0693 - val_KL_loss: 6.1708 - val_reconstruction_loss: 144.8985\n",
      "Epoch 64/100\n",
      "60000/60000 [==============================] - 9s 142us/step - loss: 146.4860 - KL_loss: 6.2370 - reconstruction_loss: 140.2490 - val_loss: 151.1347 - val_KL_loss: 6.1881 - val_reconstruction_loss: 144.9466\n",
      "Epoch 65/100\n",
      "60000/60000 [==============================] - 8s 135us/step - loss: 146.4675 - KL_loss: 6.2492 - reconstruction_loss: 140.2183 - val_loss: 150.8752 - val_KL_loss: 6.1776 - val_reconstruction_loss: 144.6976\n",
      "Epoch 66/100\n",
      "60000/60000 [==============================] - 8s 132us/step - loss: 146.3822 - KL_loss: 6.2492 - reconstruction_loss: 140.1330 - val_loss: 150.8504 - val_KL_loss: 6.3286 - val_reconstruction_loss: 144.5217\n",
      "Epoch 67/100\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 146.2880 - KL_loss: 6.2464 - reconstruction_loss: 140.0416 - val_loss: 150.7258 - val_KL_loss: 6.1723 - val_reconstruction_loss: 144.5535\n",
      "Epoch 68/100\n",
      "60000/60000 [==============================] - 12s 195us/step - loss: 146.2449 - KL_loss: 6.2569 - reconstruction_loss: 139.9880 - val_loss: 150.8034 - val_KL_loss: 6.1547 - val_reconstruction_loss: 144.6488\n",
      "Epoch 69/100\n",
      "60000/60000 [==============================] - 11s 188us/step - loss: 146.1123 - KL_loss: 6.2474 - reconstruction_loss: 139.8648 - val_loss: 150.6109 - val_KL_loss: 6.0606 - val_reconstruction_loss: 144.5503\n",
      "Epoch 70/100\n",
      "60000/60000 [==============================] - 10s 174us/step - loss: 146.0751 - KL_loss: 6.2634 - reconstruction_loss: 139.8117 - val_loss: 150.6660 - val_KL_loss: 6.2311 - val_reconstruction_loss: 144.4349\n",
      "Epoch 71/100\n",
      "60000/60000 [==============================] - 11s 186us/step - loss: 146.0239 - KL_loss: 6.2580 - reconstruction_loss: 139.7659 - val_loss: 150.6992 - val_KL_loss: 6.2205 - val_reconstruction_loss: 144.4787\n",
      "Epoch 72/100\n",
      "60000/60000 [==============================] - 11s 187us/step - loss: 145.9766 - KL_loss: 6.2773 - reconstruction_loss: 139.6993 - val_loss: 150.6919 - val_KL_loss: 6.2597 - val_reconstruction_loss: 144.4322\n",
      "Epoch 73/100\n",
      "60000/60000 [==============================] - 10s 171us/step - loss: 145.8744 - KL_loss: 6.2862 - reconstruction_loss: 139.5882 - val_loss: 150.5288 - val_KL_loss: 6.2426 - val_reconstruction_loss: 144.2862\n",
      "Epoch 74/100\n",
      "60000/60000 [==============================] - 10s 168us/step - loss: 145.8255 - KL_loss: 6.2812 - reconstruction_loss: 139.5443 - val_loss: 150.5564 - val_KL_loss: 6.3397 - val_reconstruction_loss: 144.2167\n",
      "Epoch 75/100\n",
      "60000/60000 [==============================] - 12s 195us/step - loss: 145.7868 - KL_loss: 6.2901 - reconstruction_loss: 139.4967 - val_loss: 150.6070 - val_KL_loss: 6.1866 - val_reconstruction_loss: 144.4204\n",
      "Epoch 76/100\n",
      "60000/60000 [==============================] - 12s 205us/step - loss: 145.6879 - KL_loss: 6.2903 - reconstruction_loss: 139.3977 - val_loss: 150.4220 - val_KL_loss: 6.4185 - val_reconstruction_loss: 144.0035\n",
      "Epoch 77/100\n",
      "60000/60000 [==============================] - 11s 175us/step - loss: 145.6541 - KL_loss: 6.3077 - reconstruction_loss: 139.3464 - val_loss: 150.5776 - val_KL_loss: 6.2359 - val_reconstruction_loss: 144.3417\n",
      "Epoch 78/100\n",
      "60000/60000 [==============================] - 13s 221us/step - loss: 145.5701 - KL_loss: 6.3044 - reconstruction_loss: 139.2657 - val_loss: 150.5575 - val_KL_loss: 6.2673 - val_reconstruction_loss: 144.2902\n",
      "Epoch 79/100\n",
      "60000/60000 [==============================] - 15s 244us/step - loss: 145.5087 - KL_loss: 6.2954 - reconstruction_loss: 139.2133 - val_loss: 150.5740 - val_KL_loss: 6.4186 - val_reconstruction_loss: 144.1554\n",
      "Epoch 80/100\n",
      "60000/60000 [==============================] - 11s 191us/step - loss: 145.4768 - KL_loss: 6.2939 - reconstruction_loss: 139.1829 - val_loss: 150.6282 - val_KL_loss: 6.3287 - val_reconstruction_loss: 144.2995\n",
      "Epoch 81/100\n",
      "60000/60000 [==============================] - 10s 170us/step - loss: 145.4040 - KL_loss: 6.3106 - reconstruction_loss: 139.0934 - val_loss: 150.4470 - val_KL_loss: 6.2072 - val_reconstruction_loss: 144.2398\n",
      "Epoch 82/100\n",
      "60000/60000 [==============================] - 14s 232us/step - loss: 145.3509 - KL_loss: 6.3037 - reconstruction_loss: 139.0472 - val_loss: 150.4364 - val_KL_loss: 6.2644 - val_reconstruction_loss: 144.1720\n",
      "Epoch 83/100\n",
      "60000/60000 [==============================] - 12s 208us/step - loss: 145.2625 - KL_loss: 6.3128 - reconstruction_loss: 138.9497 - val_loss: 150.6532 - val_KL_loss: 6.2198 - val_reconstruction_loss: 144.4334\n",
      "Epoch 84/100\n",
      "60000/60000 [==============================] - 12s 199us/step - loss: 145.2665 - KL_loss: 6.3269 - reconstruction_loss: 138.9396 - val_loss: 150.5046 - val_KL_loss: 6.2913 - val_reconstruction_loss: 144.2133\n",
      "Epoch 85/100\n",
      "60000/60000 [==============================] - 13s 223us/step - loss: 145.1802 - KL_loss: 6.3199 - reconstruction_loss: 138.8603 - val_loss: 150.3556 - val_KL_loss: 6.3002 - val_reconstruction_loss: 144.0553\n",
      "Epoch 86/100\n",
      "60000/60000 [==============================] - 14s 229us/step - loss: 145.1094 - KL_loss: 6.3310 - reconstruction_loss: 138.7783 - val_loss: 150.3403 - val_KL_loss: 6.2316 - val_reconstruction_loss: 144.1087\n",
      "Epoch 87/100\n",
      "60000/60000 [==============================] - 12s 203us/step - loss: 145.1000 - KL_loss: 6.3285 - reconstruction_loss: 138.7715 - val_loss: 150.4393 - val_KL_loss: 6.1519 - val_reconstruction_loss: 144.2874\n",
      "Epoch 88/100\n",
      "60000/60000 [==============================] - 12s 192us/step - loss: 145.0442 - KL_loss: 6.3374 - reconstruction_loss: 138.7067 - val_loss: 150.4658 - val_KL_loss: 6.2267 - val_reconstruction_loss: 144.2391\n",
      "Epoch 89/100\n",
      "60000/60000 [==============================] - 13s 212us/step - loss: 144.9752 - KL_loss: 6.3343 - reconstruction_loss: 138.6409 - val_loss: 150.4935 - val_KL_loss: 6.2891 - val_reconstruction_loss: 144.2044\n",
      "Epoch 90/100\n",
      "60000/60000 [==============================] - 11s 178us/step - loss: 144.8846 - KL_loss: 6.3547 - reconstruction_loss: 138.5299 - val_loss: 150.5267 - val_KL_loss: 6.4122 - val_reconstruction_loss: 144.1145\n",
      "Epoch 91/100\n",
      "60000/60000 [==============================] - 10s 166us/step - loss: 144.8968 - KL_loss: 6.3544 - reconstruction_loss: 138.5424 - val_loss: 150.6177 - val_KL_loss: 6.2687 - val_reconstruction_loss: 144.3490\n",
      "Epoch 92/100\n",
      "60000/60000 [==============================] - 9s 156us/step - loss: 144.8523 - KL_loss: 6.3469 - reconstruction_loss: 138.5054 - val_loss: 150.5604 - val_KL_loss: 6.3960 - val_reconstruction_loss: 144.1644\n",
      "Epoch 93/100\n",
      "60000/60000 [==============================] - 10s 166us/step - loss: 144.7789 - KL_loss: 6.3393 - reconstruction_loss: 138.4396 - val_loss: 150.2738 - val_KL_loss: 6.2003 - val_reconstruction_loss: 144.0735\n",
      "Epoch 94/100\n",
      "60000/60000 [==============================] - 9s 154us/step - loss: 144.7368 - KL_loss: 6.3457 - reconstruction_loss: 138.3911 - val_loss: 150.2970 - val_KL_loss: 6.2321 - val_reconstruction_loss: 144.0649\n",
      "Epoch 95/100\n",
      "60000/60000 [==============================] - 9s 148us/step - loss: 144.6593 - KL_loss: 6.3497 - reconstruction_loss: 138.3095 - val_loss: 150.5902 - val_KL_loss: 6.4358 - val_reconstruction_loss: 144.1544\n",
      "Epoch 96/100\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 144.6590 - KL_loss: 6.3651 - reconstruction_loss: 138.2938 - val_loss: 150.3552 - val_KL_loss: 6.2708 - val_reconstruction_loss: 144.0844\n",
      "Epoch 97/100\n",
      "60000/60000 [==============================] - 10s 164us/step - loss: 144.6161 - KL_loss: 6.3586 - reconstruction_loss: 138.2575 - val_loss: 150.2692 - val_KL_loss: 6.2923 - val_reconstruction_loss: 143.9769\n",
      "Epoch 98/100\n",
      "60000/60000 [==============================] - 10s 165us/step - loss: 144.5537 - KL_loss: 6.3611 - reconstruction_loss: 138.1926 - val_loss: 150.0711 - val_KL_loss: 6.2814 - val_reconstruction_loss: 143.7898\n",
      "Epoch 99/100\n",
      "60000/60000 [==============================] - 10s 161us/step - loss: 144.5126 - KL_loss: 6.3589 - reconstruction_loss: 138.1536 - val_loss: 150.4528 - val_KL_loss: 6.5239 - val_reconstruction_loss: 143.9289\n",
      "Epoch 100/100\n",
      "60000/60000 [==============================] - 10s 159us/step - loss: 144.4890 - KL_loss: 6.3734 - reconstruction_loss: 138.1156 - val_loss: 150.2972 - val_KL_loss: 6.3849 - val_reconstruction_loss: 143.9123\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1295d2390>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auto_encoder.fit(x_train, x_train, verbose=1, \n",
    "                 batch_size=batch_size, epochs=100,\n",
    "                 validation_data=(x_test, x_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAADgAAAA4CAAAAACN7WTCAAAB5ElEQVR4nO3WwUtUURTH8c99U88YzdHMKXFhlFoRRCDUJlrUJjcF7Vr177Vr0bKlFgQpKLUwrShLInEardHReW9eiywag4J8DyI6y3PPuV/O73DuufylFsJeT/THV/0uoGJE1Yp3GlLZ/okHfn0cu+SO2LRpb9Uk35nFEEtOmTQsVVHWremz9i6zCGJk0G0XtS2Ys6JumwJrDHrccFXZcw/MqWsW2sdI2WWTer12z2N1LcmeiDyJkdiICQOW3fVITet7/4ogBiUV50zYdN9DHyW7vKwQVYOSLiNuGjJjxrr27tC2KYgYOeKWCWueeC8Wawpae+LyrfG0a8qmLOpWlVq3LtWWCblPRxDpdV1V3aJuw6o2rFqyI5W/qkHkoDMuKFmVGjOqomEZ63Y65mP/xCAIIj2uGNTwzoA+x0RiDcNe2cidSHBAnyGHJHa01WQ+iPU67rAgX1UzQSQ2qCz1WdOmmme6jKpq2fphb+RD/EY9KLOlYdumOo45KbZiTdqRmBcx9dF747octaNHv1GHLJj1qePFyUvVTFvdU+PGnHdWIrNlyZSXmtodiXnUmAlaNszrkzghlnlj3qxFtY5NtS/inn9OpCRWMWJcv8SSFz7YlhRG/OqIBEEm+2lH5UD8b/+GfQFxm6mm+0NMAAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=56x56 at 0x128F53E80>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_number = np.asarray([[np.random.normal() \n",
    "                            for _ in range(latent_space_depth)]])\n",
    "def decode_img(a):\n",
    "    a = np.clip(a * 256, 0, 255).astype('uint8')\n",
    "    return PIL.Image.fromarray(a)\n",
    "\n",
    "decode_img(decoder.predict(random_number).reshape(img_width, img_height)).resize((56, 56))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=328x328 at 0x128F53C88>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_cells = 10\n",
    "\n",
    "overview = PIL.Image.new('RGB', \n",
    "                         (num_cells * (img_width + 4) + 8, \n",
    "                          num_cells * (img_height + 4) + 8), \n",
    "                         (128, 128, 128))\n",
    "\n",
    "vec = np.zeros((1, latent_space_depth))\n",
    "for x in range(num_cells):\n",
    "    vec[:, 0] = (x * 3) / (num_cells - 1) - 1.5\n",
    "    for y in range(num_cells):\n",
    "        vec[:, 1] = (y * 3) / (num_cells - 1) - 1.5\n",
    "        decoded = decoder.predict(vec)\n",
    "        img = decode_img(decoded.reshape(img_width, img_height))\n",
    "        overview.paste(img, (x * (img_width + 4) + 6, y * (img_height + 4) + 6))\n",
    "overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_15 (InputLayer)           (None, 784)          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "label (InputLayer)              (None, 10)           0                                            \n",
      "__________________________________________________________________________________________________\n",
      "inputs (Concatenate)            (None, 794)          0           input_15[0][0]                   \n",
      "                                                                 label[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "encoder_hidden (Dense)          (None, 512)          407040      inputs[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "dense_28 (Dense)                (None, 2)            1026        encoder_hidden[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "dense_29 (Dense)                (None, 2)            1026        encoder_hidden[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "lambda_7 (Lambda)               (None, 2)            0           dense_28[0][0]                   \n",
      "                                                                 dense_29[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_6 (Concatenate)     (None, 12)           0           lambda_7[0][0]                   \n",
      "                                                                 label[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "dense_30 (Dense)                (None, 512)          6656        concatenate_6[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dense_31 (Dense)                (None, 784)          402192      dense_30[1][0]                   \n",
      "==================================================================================================\n",
      "Total params: 817,940\n",
      "Trainable params: 817,940\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "def ConditionalVariationalAutoEncoder(num_pixels, num_labels):\n",
    "    pixels = Input(shape=(num_pixels,))\n",
    "    label = Input(shape=(num_labels,), name='label')\n",
    "\n",
    "    inputs = concat([pixels, label], name='inputs')\n",
    "    \n",
    "    encoder_hidden = Dense(512, activation='relu', name='encoder_hidden')(inputs)\n",
    "\n",
    "    z_mean = Dense(latent_space_depth, activation='linear')(encoder_hidden)\n",
    "    z_log_var = Dense(latent_space_depth, activation='linear')(encoder_hidden)\n",
    "    \n",
    "    def KL_loss(y_true, y_pred):\n",
    "        return(0.5 * K.sum(K.exp(z_log_var) + K.square(z_mean) - 1 - z_log_var, axis=1))\n",
    "\n",
    "    def reconstruction_loss(y_true, y_pred):\n",
    "        return K.sum(K.binary_crossentropy(y_true, y_pred), axis=-1)    \n",
    "\n",
    "    def total_loss(y_true, y_pred):\n",
    "        return KL_loss(y_true, y_pred) + reconstruction_loss(y_true, y_pred)\n",
    "\n",
    "    z = Lambda(sample_z, output_shape=(latent_space_depth, ))([z_mean, z_log_var])\n",
    "    zc = concat([z, label])\n",
    "    \n",
    "    decoder_hidden = Dense(512, activation='relu')\n",
    "\n",
    "    reconstruct_pixels = Dense(num_pixels, activation='sigmoid')\n",
    "\n",
    "    decoder_in = Input(shape=(latent_space_depth + num_labels,))\n",
    "    hidden = decoder_hidden(decoder_in)\n",
    "    decoder_out = reconstruct_pixels(hidden)\n",
    "    decoder = Model(decoder_in, decoder_out)    \n",
    "    \n",
    "    hidden = decoder_hidden(zc)\n",
    "    outputs = reconstruct_pixels(hidden)\n",
    "    auto_encoder = Model([pixels, label], outputs)\n",
    "\n",
    "    auto_encoder.compile(optimizer=Adam(lr=0.001), \n",
    "                         loss=total_loss,\n",
    "                         metrics=[KL_loss, reconstruction_loss])\n",
    "    \n",
    "    return auto_encoder, decoder\n",
    "\n",
    "cond_auto_encoder, cond_decoder = ConditionalVariationalAutoEncoder(x_train.shape[1], y_train.shape[1])\n",
    "cond_auto_encoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/50\n",
      "60000/60000 [==============================] - 6s 105us/step - loss: 129.1885 - KL_loss: 4.7397 - reconstruction_loss: 124.4489 - val_loss: 130.7288 - val_KL_loss: 4.5935 - val_reconstruction_loss: 126.1352\n",
      "Epoch 2/50\n",
      "60000/60000 [==============================] - 6s 107us/step - loss: 129.1577 - KL_loss: 4.7555 - reconstruction_loss: 124.4022 - val_loss: 130.6913 - val_KL_loss: 4.6928 - val_reconstruction_loss: 125.9985\n",
      "Epoch 3/50\n",
      "60000/60000 [==============================] - 6s 105us/step - loss: 129.0916 - KL_loss: 4.7557 - reconstruction_loss: 124.3359 - val_loss: 130.5601 - val_KL_loss: 4.6938 - val_reconstruction_loss: 125.8663\n",
      "Epoch 4/50\n",
      "60000/60000 [==============================] - 6s 107us/step - loss: 129.0676 - KL_loss: 4.7525 - reconstruction_loss: 124.3151 - val_loss: 130.6311 - val_KL_loss: 4.6708 - val_reconstruction_loss: 125.9603\n",
      "Epoch 5/50\n",
      "60000/60000 [==============================] - 7s 113us/step - loss: 129.0341 - KL_loss: 4.7648 - reconstruction_loss: 124.2693 - val_loss: 130.5953 - val_KL_loss: 4.6742 - val_reconstruction_loss: 125.9211\n",
      "Epoch 6/50\n",
      "60000/60000 [==============================] - 7s 118us/step - loss: 128.9547 - KL_loss: 4.7700 - reconstruction_loss: 124.1846 - val_loss: 130.7554 - val_KL_loss: 4.7213 - val_reconstruction_loss: 126.0341\n",
      "Epoch 7/50\n",
      "60000/60000 [==============================] - 7s 120us/step - loss: 128.9114 - KL_loss: 4.7632 - reconstruction_loss: 124.1482 - val_loss: 130.5908 - val_KL_loss: 4.6969 - val_reconstruction_loss: 125.8940\n",
      "Epoch 8/50\n",
      "60000/60000 [==============================] - 7s 118us/step - loss: 128.8800 - KL_loss: 4.7775 - reconstruction_loss: 124.1026 - val_loss: 130.5680 - val_KL_loss: 4.7371 - val_reconstruction_loss: 125.8309\n",
      "Epoch 9/50\n",
      "60000/60000 [==============================] - 7s 120us/step - loss: 128.8198 - KL_loss: 4.7759 - reconstruction_loss: 124.0438 - val_loss: 130.5840 - val_KL_loss: 4.6938 - val_reconstruction_loss: 125.8902\n",
      "Epoch 10/50\n",
      "60000/60000 [==============================] - 7s 120us/step - loss: 128.8162 - KL_loss: 4.7827 - reconstruction_loss: 124.0335 - val_loss: 130.4100 - val_KL_loss: 4.7102 - val_reconstruction_loss: 125.6998\n",
      "Epoch 11/50\n",
      "60000/60000 [==============================] - 7s 121us/step - loss: 128.7440 - KL_loss: 4.7807 - reconstruction_loss: 123.9633 - val_loss: 130.4184 - val_KL_loss: 4.6763 - val_reconstruction_loss: 125.7421\n",
      "Epoch 12/50\n",
      "60000/60000 [==============================] - 7s 121us/step - loss: 128.6886 - KL_loss: 4.7810 - reconstruction_loss: 123.9076 - val_loss: 130.5873 - val_KL_loss: 4.6870 - val_reconstruction_loss: 125.9003\n",
      "Epoch 13/50\n",
      "60000/60000 [==============================] - 7s 122us/step - loss: 128.6786 - KL_loss: 4.7939 - reconstruction_loss: 123.8848 - val_loss: 130.4877 - val_KL_loss: 4.5843 - val_reconstruction_loss: 125.9034\n",
      "Epoch 14/50\n",
      "60000/60000 [==============================] - 8s 135us/step - loss: 128.6186 - KL_loss: 4.7930 - reconstruction_loss: 123.8256 - val_loss: 130.3514 - val_KL_loss: 4.6735 - val_reconstruction_loss: 125.6779\n",
      "Epoch 15/50\n",
      "60000/60000 [==============================] - 9s 144us/step - loss: 128.5878 - KL_loss: 4.8020 - reconstruction_loss: 123.7858 - val_loss: 130.4407 - val_KL_loss: 4.6944 - val_reconstruction_loss: 125.7463\n",
      "Epoch 16/50\n",
      "60000/60000 [==============================] - 9s 146us/step - loss: 128.5360 - KL_loss: 4.8007 - reconstruction_loss: 123.7353 - val_loss: 130.5329 - val_KL_loss: 4.7235 - val_reconstruction_loss: 125.8094\n",
      "Epoch 17/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 128.5092 - KL_loss: 4.8038 - reconstruction_loss: 123.7054 - val_loss: 130.4356 - val_KL_loss: 4.7344 - val_reconstruction_loss: 125.7012\n",
      "Epoch 18/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 128.4706 - KL_loss: 4.8084 - reconstruction_loss: 123.6622 - val_loss: 130.4677 - val_KL_loss: 4.7195 - val_reconstruction_loss: 125.7481\n",
      "Epoch 19/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 128.4370 - KL_loss: 4.8136 - reconstruction_loss: 123.6234 - val_loss: 130.4330 - val_KL_loss: 4.7023 - val_reconstruction_loss: 125.7307\n",
      "Epoch 20/50\n",
      "60000/60000 [==============================] - 9s 142us/step - loss: 128.4062 - KL_loss: 4.8281 - reconstruction_loss: 123.5782 - val_loss: 130.3432 - val_KL_loss: 4.6758 - val_reconstruction_loss: 125.6674\n",
      "Epoch 21/50\n",
      "60000/60000 [==============================] - 9s 148us/step - loss: 128.3523 - KL_loss: 4.8078 - reconstruction_loss: 123.5445 - val_loss: 130.4382 - val_KL_loss: 4.7706 - val_reconstruction_loss: 125.6677\n",
      "Epoch 22/50\n",
      "60000/60000 [==============================] - 9s 156us/step - loss: 128.3244 - KL_loss: 4.8187 - reconstruction_loss: 123.5057 - val_loss: 130.4327 - val_KL_loss: 4.6568 - val_reconstruction_loss: 125.7758\n",
      "Epoch 23/50\n",
      "60000/60000 [==============================] - 9s 151us/step - loss: 128.2972 - KL_loss: 4.8272 - reconstruction_loss: 123.4700 - val_loss: 130.4339 - val_KL_loss: 4.7960 - val_reconstruction_loss: 125.6378\n",
      "Epoch 24/50\n",
      "60000/60000 [==============================] - 9s 147us/step - loss: 128.2462 - KL_loss: 4.8230 - reconstruction_loss: 123.4232 - val_loss: 130.3972 - val_KL_loss: 4.7540 - val_reconstruction_loss: 125.6431\n",
      "Epoch 25/50\n",
      "60000/60000 [==============================] - 8s 131us/step - loss: 128.1760 - KL_loss: 4.8227 - reconstruction_loss: 123.3533 - val_loss: 130.4070 - val_KL_loss: 4.7383 - val_reconstruction_loss: 125.6687\n",
      "Epoch 26/50\n",
      "60000/60000 [==============================] - 8s 130us/step - loss: 128.1845 - KL_loss: 4.8417 - reconstruction_loss: 123.3428 - val_loss: 130.3273 - val_KL_loss: 4.7022 - val_reconstruction_loss: 125.6251\n",
      "Epoch 27/50\n",
      "60000/60000 [==============================] - 8s 134us/step - loss: 128.1547 - KL_loss: 4.8491 - reconstruction_loss: 123.3057 - val_loss: 130.3627 - val_KL_loss: 4.6926 - val_reconstruction_loss: 125.6701\n",
      "Epoch 28/50\n",
      "60000/60000 [==============================] - 8s 130us/step - loss: 128.1241 - KL_loss: 4.8389 - reconstruction_loss: 123.2852 - val_loss: 130.3913 - val_KL_loss: 4.8135 - val_reconstruction_loss: 125.5778\n",
      "Epoch 29/50\n",
      "60000/60000 [==============================] - 8s 134us/step - loss: 128.0994 - KL_loss: 4.8448 - reconstruction_loss: 123.2547 - val_loss: 130.3724 - val_KL_loss: 4.8552 - val_reconstruction_loss: 125.5172\n",
      "Epoch 30/50\n",
      "60000/60000 [==============================] - 8s 129us/step - loss: 128.0658 - KL_loss: 4.8447 - reconstruction_loss: 123.2211 - val_loss: 130.2945 - val_KL_loss: 4.7329 - val_reconstruction_loss: 125.5616\n",
      "Epoch 31/50\n",
      "60000/60000 [==============================] - 9s 145us/step - loss: 128.0327 - KL_loss: 4.8598 - reconstruction_loss: 123.1729 - val_loss: 130.4131 - val_KL_loss: 4.8143 - val_reconstruction_loss: 125.5988\n",
      "Epoch 32/50\n",
      "60000/60000 [==============================] - 9s 144us/step - loss: 127.9920 - KL_loss: 4.8568 - reconstruction_loss: 123.1352 - val_loss: 130.4254 - val_KL_loss: 4.7568 - val_reconstruction_loss: 125.6686\n",
      "Epoch 33/50\n",
      "60000/60000 [==============================] - 9s 145us/step - loss: 127.9628 - KL_loss: 4.8541 - reconstruction_loss: 123.1086 - val_loss: 130.2891 - val_KL_loss: 4.7399 - val_reconstruction_loss: 125.5492\n",
      "Epoch 34/50\n",
      "60000/60000 [==============================] - 8s 136us/step - loss: 127.9355 - KL_loss: 4.8527 - reconstruction_loss: 123.0828 - val_loss: 130.1691 - val_KL_loss: 4.7404 - val_reconstruction_loss: 125.4286\n",
      "Epoch 35/50\n",
      "60000/60000 [==============================] - 8s 133us/step - loss: 127.8947 - KL_loss: 4.8595 - reconstruction_loss: 123.0351 - val_loss: 130.3889 - val_KL_loss: 4.7914 - val_reconstruction_loss: 125.5976\n",
      "Epoch 36/50\n",
      "60000/60000 [==============================] - 9s 153us/step - loss: 127.8898 - KL_loss: 4.8777 - reconstruction_loss: 123.0120 - val_loss: 130.3886 - val_KL_loss: 4.7931 - val_reconstruction_loss: 125.5955\n",
      "Epoch 37/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 127.8743 - KL_loss: 4.8703 - reconstruction_loss: 123.0040 - val_loss: 130.3119 - val_KL_loss: 4.7078 - val_reconstruction_loss: 125.6041\n",
      "Epoch 38/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000/60000 [==============================] - 8s 130us/step - loss: 127.8158 - KL_loss: 4.8745 - reconstruction_loss: 122.9413 - val_loss: 130.3383 - val_KL_loss: 4.9415 - val_reconstruction_loss: 125.3968\n",
      "Epoch 39/50\n",
      "60000/60000 [==============================] - 8s 128us/step - loss: 127.7705 - KL_loss: 4.8669 - reconstruction_loss: 122.9036 - val_loss: 130.3126 - val_KL_loss: 4.8402 - val_reconstruction_loss: 125.4724\n",
      "Epoch 40/50\n",
      "60000/60000 [==============================] - 7s 122us/step - loss: 127.7746 - KL_loss: 4.8919 - reconstruction_loss: 122.8827 - val_loss: 130.2123 - val_KL_loss: 4.8112 - val_reconstruction_loss: 125.4010\n",
      "Epoch 41/50\n",
      "60000/60000 [==============================] - 7s 117us/step - loss: 127.7435 - KL_loss: 4.8887 - reconstruction_loss: 122.8547 - val_loss: 130.2400 - val_KL_loss: 4.8042 - val_reconstruction_loss: 125.4357\n",
      "Epoch 42/50\n",
      "60000/60000 [==============================] - 7s 117us/step - loss: 127.7260 - KL_loss: 4.8902 - reconstruction_loss: 122.8358 - val_loss: 130.3282 - val_KL_loss: 4.8430 - val_reconstruction_loss: 125.4852\n",
      "Epoch 43/50\n",
      "60000/60000 [==============================] - 8s 125us/step - loss: 127.6763 - KL_loss: 4.8896 - reconstruction_loss: 122.7867 - val_loss: 130.2412 - val_KL_loss: 4.7106 - val_reconstruction_loss: 125.5306\n",
      "Epoch 44/50\n",
      "60000/60000 [==============================] - 8s 127us/step - loss: 127.6510 - KL_loss: 4.8835 - reconstruction_loss: 122.7675 - val_loss: 130.4059 - val_KL_loss: 4.8320 - val_reconstruction_loss: 125.5739\n",
      "Epoch 45/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 127.6461 - KL_loss: 4.8985 - reconstruction_loss: 122.7476 - val_loss: 130.3365 - val_KL_loss: 4.7761 - val_reconstruction_loss: 125.5604\n",
      "Epoch 46/50\n",
      "60000/60000 [==============================] - 9s 152us/step - loss: 127.6044 - KL_loss: 4.9002 - reconstruction_loss: 122.7042 - val_loss: 130.2961 - val_KL_loss: 4.7775 - val_reconstruction_loss: 125.5186\n",
      "Epoch 47/50\n",
      "60000/60000 [==============================] - 8s 140us/step - loss: 127.5983 - KL_loss: 4.8979 - reconstruction_loss: 122.7004 - val_loss: 130.2866 - val_KL_loss: 4.8529 - val_reconstruction_loss: 125.4337\n",
      "Epoch 48/50\n",
      "60000/60000 [==============================] - 9s 143us/step - loss: 127.5292 - KL_loss: 4.8862 - reconstruction_loss: 122.6430 - val_loss: 130.2153 - val_KL_loss: 4.7525 - val_reconstruction_loss: 125.4628\n",
      "Epoch 49/50\n",
      "60000/60000 [==============================] - 8s 133us/step - loss: 127.5168 - KL_loss: 4.9036 - reconstruction_loss: 122.6132 - val_loss: 130.3204 - val_KL_loss: 4.7894 - val_reconstruction_loss: 125.5310\n",
      "Epoch 50/50\n",
      "60000/60000 [==============================] - 8s 137us/step - loss: 127.4842 - KL_loss: 4.9033 - reconstruction_loss: 122.5809 - val_loss: 130.3549 - val_KL_loss: 4.8567 - val_reconstruction_loss: 125.4981\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x128f70278>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cond_auto_encoder.fit([x_train, y_train], x_train, verbose=1, \n",
    "                      batch_size=batch_size, epochs=50,\n",
    "                      validation_data = ([x_test, y_test], x_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=56x56 at 0x1295D2748>"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "number_4 = np.zeros((1, latent_space_depth + y_train.shape[1]))\n",
    "number_4[:, 4 + latent_space_depth] = 1\n",
    "decode_img(cond_decoder.predict(number_4).reshape(\n",
    "               img_width, img_height)).resize((56, 56))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=56x56 at 0x1295D2908>"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "number_8_3 = np.zeros((1, latent_space_depth + y_train.shape[1]))\n",
    "number_8_3[:, 8 + latent_space_depth] = 0.5\n",
    "number_8_3[:, 3 + latent_space_depth] = 0.5\n",
    "decode_img(cond_decoder.predict(number_8_3).reshape(\n",
    "               img_width, img_height)).resize((56, 56))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=328x328 at 0x1291F9A20>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "digits = [3, 0, 8, 9]\n",
    "num_cells = 10\n",
    "\n",
    "overview = PIL.Image.new('RGB', \n",
    "                         (num_cells * (img_width + 4) + 8, \n",
    "                          num_cells * (img_height + 4) + 8), \n",
    "                         (128, 128, 128))\n",
    "\n",
    "vec = np.zeros((1, latent_space_depth + y_train.shape[1]))\n",
    "for x in range(num_cells):\n",
    "    x1 = [x / (num_cells - 1), 1 - x / (num_cells - 1)]\n",
    "    for y in range(num_cells):\n",
    "        y1 = [y / (num_cells - 1), 1 - y / (num_cells - 1)]\n",
    "        for idx, dig in enumerate(digits):\n",
    "            vec[:, dig + latent_space_depth] = x1[idx % 2] * y1[idx // 2]\n",
    "        decoded = cond_decoder.predict(vec)\n",
    "        img = decode_img(decoded.reshape(img_width, img_height))\n",
    "        overview.paste(img, (x * (img_width + 4) + 6, y * (img_height + 4) + 6))\n",
    "overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=328x328 at 0x129523AC8>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_cells = 10\n",
    "\n",
    "overview = PIL.Image.new('RGB', \n",
    "                         (num_cells * (img_width + 4) + 8, \n",
    "                          num_cells * (img_height + 4) + 8), \n",
    "                         (128, 128, 128))\n",
    "\n",
    "img_it = 0\n",
    "vec = np.zeros((1, latent_space_depth + y_train.shape[1]))\n",
    "for x in range(num_cells):\n",
    "    vec = np.zeros((1, latent_space_depth + y_train.shape[1]))\n",
    "    vec[:, x + latent_space_depth] = 1\n",
    "    for y in range(num_cells):\n",
    "        vec[:, 1] = 3 * y / (num_cells - 1) - 1.5\n",
    "        decoded = cond_decoder.predict(vec)\n",
    "        img = decode_img(decoded.reshape(img_width, img_height))\n",
    "        overview.paste(img, (x * (img_width + 4) + 6, y * (img_height + 4) + 6))\n",
    "overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.5.2"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "263px",
    "width": "253px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}