Train_DCVAE_MPS45_tensorflow .ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n",
      "WARNING: Logging before flag parsing goes to stderr.\n",
      "W0306 21:21:59.233442 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/Utils.py:238: The name tf.AUTO_REUSE is deprecated. Please use tf.compat.v1.AUTO_REUSE instead.\n",
      "\n",
      "W0306 21:21:59.235225 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:15: The name tf.train.AdamOptimizer is deprecated. Please use tf.compat.v1.train.AdamOptimizer instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from Model.DCVAE_TF import DCVAE\n",
    "from Model.Utils import MPS_Generator,plot_images, plot_history"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load DataSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "W0306 21:21:59.293160 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/Utils.py:322: The name tf.parse_single_example is deprecated. Please use tf.io.parse_single_example instead.\n",
      "\n",
      "W0306 21:21:59.294534 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/Utils.py:325: The name tf.FixedLenFeature is deprecated. Please use tf.io.FixedLenFeature instead.\n",
      "\n",
      "W0306 21:21:59.324368 139970004961088 deprecation.py:323] From /share/GeoFacies/GeoFacies_DL/Model/Utils.py:314: DatasetV1.make_one_shot_iterator (from tensorflow.python.data.ops.dataset_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use `for ... in dataset:` to iterate over a dataset. If using `tf.estimator`, return the `Dataset` object directly from your input function. As a last resort, you can use `tf.compat.v1.data.make_one_shot_iterator(dataset)`.\n",
      "W0306 21:21:59.349660 139970004961088 deprecation_wrapper.py:119] From /work/SmithCPU/.local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:174: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
      "\n",
      "W0306 21:21:59.350977 139970004961088 deprecation_wrapper.py:119] From /work/SmithCPU/.local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:181: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def load_data_set(path_tfRecord):\n",
    "    gen_train = MPS_Generator(path_tfRecord,4)\n",
    "    gen_train = MPS_Generator(path_tfRecord, gen_train.num)\n",
    "    x_train = gen_train.get_numpy_batch().astype('float32')\n",
    "    return x_train\n",
    "x_train = load_data_set('DataSet/MPS45/train.tfrecords')\n",
    "x_test =load_data_set('DataSet/MPS45/test_val.tfrecords')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = DCVAE(input_shape=(45,45,2), latent_cont_dim=500, opt=tf.train.RMSPropOptimizer(learning_rate=0.001),hidden_dim=1024, \n",
    "              filters=(32,32,16), strides=(2,2,1), dropout=0.1,lr=0.001)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "W0306 21:22:04.089029 139970004961088 deprecation.py:506] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "W0306 21:22:04.234706 139970004961088 deprecation.py:323] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:112: dropout (from tensorflow.python.layers.core) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use keras.layers.dropout instead.\n",
      "W0306 21:22:04.429276 139970004961088 deprecation.py:323] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:148: conv2d_transpose (from tensorflow.python.layers.convolutional) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use `tf.keras.layers.Conv2DTranspose` instead.\n",
      "W0306 21:22:05.291763 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:154: The name tf.image.resize_bilinear is deprecated. Please use tf.compat.v1.image.resize_bilinear instead.\n",
      "\n",
      "W0306 21:22:05.294917 139970004961088 deprecation.py:323] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:156: conv2d (from tensorflow.python.layers.convolutional) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use `tf.keras.layers.Conv2D` instead.\n",
      "W0306 21:22:05.742581 139970004961088 deprecation.py:323] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:80: flatten (from tensorflow.python.layers.core) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use keras.layers.flatten instead.\n",
      "W0306 21:22:06.233469 139970004961088 deprecation_wrapper.py:119] From /share/GeoFacies/GeoFacies_DL/Model/DCVAE_TF.py:170: The name tf.train.RMSPropOptimizer is deprecated. Please use tf.compat.v1.train.RMSPropOptimizer instead.\n",
      "\n",
      "W0306 21:22:06.399590 139970004961088 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_grad.py:1250: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "W0306 21:22:06.575652 139970004961088 deprecation.py:506] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/training/rmsprop.py:119: calling Ones.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 1 - Train cost= 1012.45904541\n",
      "Epoch: 1 - Test cost= 523.43127441\n",
      "Epoch: 2 - Train cost= 475.95169067\n",
      "Epoch: 2 - Test cost= 442.37655640\n",
      "Epoch: 3 - Train cost= 417.94189453\n",
      "Epoch: 3 - Test cost= 413.80917358\n",
      "Epoch: 4 - Train cost= 391.19665527\n",
      "Epoch: 4 - Test cost= 390.69400024\n",
      "Epoch: 5 - Train cost= 372.17169189\n",
      "Epoch: 5 - Test cost= 359.06402588\n",
      "Epoch: 6 - Train cost= 356.57287598\n",
      "Epoch: 6 - Test cost= 362.31423950\n",
      "Epoch: 7 - Train cost= 342.63720703\n",
      "Epoch: 7 - Test cost= 336.97970581\n",
      "Epoch: 8 - Train cost= 329.99975586\n",
      "Epoch: 8 - Test cost= 331.43869019\n",
      "Epoch: 9 - Train cost= 319.41635132\n",
      "Epoch: 9 - Test cost= 324.47769165\n",
      "Epoch: 10 - Train cost= 309.31585693\n",
      "Epoch: 10 - Test cost= 303.40609741\n",
      "Epoch: 11 - Train cost= 300.98617554\n",
      "Epoch: 11 - Test cost= 304.92422485\n",
      "Epoch: 12 - Train cost= 293.38406372\n",
      "Epoch: 12 - Test cost= 295.70187378\n",
      "Epoch: 13 - Train cost= 286.90994263\n",
      "Epoch: 13 - Test cost= 286.26776123\n",
      "Epoch: 14 - Train cost= 281.38409424\n",
      "Epoch: 14 - Test cost= 281.04022217\n",
      "Epoch: 15 - Train cost= 276.90002441\n",
      "Epoch: 15 - Test cost= 276.76535034\n",
      "Epoch: 16 - Train cost= 272.60241699\n",
      "Epoch: 16 - Test cost= 285.02679443\n",
      "Epoch: 17 - Train cost= 269.04797363\n",
      "Epoch: 17 - Test cost= 270.51251221\n",
      "Epoch: 18 - Train cost= 266.28692627\n",
      "Epoch: 18 - Test cost= 263.83166504\n",
      "Epoch: 19 - Train cost= 263.95703125\n",
      "Epoch: 19 - Test cost= 272.50860596\n",
      "Epoch: 20 - Train cost= 262.21621704\n",
      "Epoch: 20 - Test cost= 266.66241455\n",
      "Epoch: 21 - Train cost= 260.16796875\n",
      "Epoch: 21 - Test cost= 264.08877563\n",
      "Epoch: 22 - Train cost= 258.82742310\n",
      "Epoch: 22 - Test cost= 263.75213623\n",
      "Epoch: 23 - Train cost= 257.12203979\n",
      "Epoch: 23 - Test cost= 260.31298828\n",
      "Epoch: 24 - Train cost= 255.91146851\n",
      "Epoch: 24 - Test cost= 261.88931274\n",
      "Epoch: 25 - Train cost= 254.27929688\n",
      "Epoch: 25 - Test cost= 263.22885132\n",
      "Epoch: 26 - Train cost= 253.76600647\n",
      "Epoch: 26 - Test cost= 265.20071411\n",
      "Epoch: 27 - Train cost= 252.53549194\n",
      "Epoch: 27 - Test cost= 254.74070740\n",
      "Epoch: 28 - Train cost= 251.26850891\n",
      "Epoch: 28 - Test cost= 256.82424927\n",
      "Epoch: 29 - Train cost= 250.56344604\n",
      "Epoch: 29 - Test cost= 254.43659973\n",
      "Epoch: 30 - Train cost= 249.78103638\n",
      "Epoch: 30 - Test cost= 253.25683594\n",
      "Epoch: 31 - Train cost= 248.98289490\n",
      "Epoch: 31 - Test cost= 258.58807373\n",
      "Epoch: 32 - Train cost= 248.34507751\n",
      "Epoch: 32 - Test cost= 253.21528625\n",
      "Epoch: 33 - Train cost= 247.52281189\n",
      "Epoch: 33 - Test cost= 262.21408081\n",
      "Epoch: 34 - Train cost= 236.53414917\n",
      "Epoch: 34 - Test cost= 241.47875977\n",
      "Epoch: 35 - Train cost= 233.65756226\n",
      "Epoch: 35 - Test cost= 240.26333618\n",
      "Epoch: 36 - Train cost= 232.47404480\n",
      "Epoch: 36 - Test cost= 239.46830750\n",
      "Epoch: 37 - Train cost= 232.07650757\n",
      "Epoch: 37 - Test cost= 238.94882202\n",
      "Epoch: 38 - Train cost= 231.48446655\n",
      "Epoch: 38 - Test cost= 238.22564697\n",
      "Epoch: 39 - Train cost= 231.21298218\n",
      "Epoch: 39 - Test cost= 237.73451233\n",
      "Epoch: 40 - Train cost= 230.79278564\n",
      "Epoch: 40 - Test cost= 237.80014038\n",
      "Epoch: 41 - Train cost= 230.55699158\n",
      "Epoch: 41 - Test cost= 237.67781067\n",
      "Epoch: 42 - Train cost= 230.36158752\n",
      "Epoch: 42 - Test cost= 237.41773987\n",
      "Epoch: 43 - Train cost= 230.10380554\n",
      "Epoch: 43 - Test cost= 237.24861145\n",
      "Epoch: 44 - Train cost= 230.01397705\n",
      "Epoch: 44 - Test cost= 237.11598206\n",
      "Epoch: 45 - Train cost= 229.78257751\n",
      "Epoch: 45 - Test cost= 237.10278320\n",
      "Epoch: 46 - Train cost= 229.67276001\n",
      "Epoch: 46 - Test cost= 236.56370544\n",
      "Epoch: 47 - Train cost= 229.50700378\n",
      "Epoch: 47 - Test cost= 236.84729004\n",
      "Epoch: 48 - Train cost= 229.38450623\n",
      "Epoch: 48 - Test cost= 236.51344299\n",
      "Epoch: 49 - Train cost= 229.29835510\n",
      "Epoch: 49 - Test cost= 236.28164673\n",
      "Epoch: 50 - Train cost= 229.08174133\n",
      "Epoch: 50 - Test cost= 236.18739319\n",
      "Epoch: 51 - Train cost= 228.93699646\n",
      "Epoch: 51 - Test cost= 236.60018921\n",
      "Epoch: 52 - Train cost= 228.87019348\n",
      "Epoch: 52 - Test cost= 236.10714722\n",
      "Epoch: 53 - Train cost= 228.63679504\n",
      "Epoch: 53 - Test cost= 236.06158447\n",
      "Epoch: 54 - Train cost= 228.53572083\n",
      "Epoch: 54 - Test cost= 236.03431702\n",
      "Epoch: 55 - Train cost= 228.49438477\n",
      "Epoch: 55 - Test cost= 235.72126770\n",
      "Epoch: 56 - Train cost= 228.36120605\n",
      "Epoch: 56 - Test cost= 235.85490417\n",
      "Epoch: 57 - Train cost= 228.40216064\n",
      "Epoch: 57 - Test cost= 235.92642212\n",
      "Epoch: 58 - Train cost= 228.21853638\n",
      "Epoch: 58 - Test cost= 235.58761597\n",
      "Epoch: 59 - Train cost= 228.15959167\n",
      "Epoch: 59 - Test cost= 235.33934021\n",
      "Epoch: 60 - Train cost= 227.95793152\n",
      "Epoch: 60 - Test cost= 235.35183716\n",
      "Epoch: 61 - Train cost= 227.88938904\n",
      "Epoch: 61 - Test cost= 235.25041199\n",
      "Epoch: 62 - Train cost= 227.87052917\n",
      "Epoch: 62 - Test cost= 235.35469055\n",
      "Epoch: 63 - Train cost= 227.66897583\n",
      "Epoch: 63 - Test cost= 235.37638855\n",
      "Epoch: 64 - Train cost= 227.73580933\n",
      "Epoch: 64 - Test cost= 235.27043152\n",
      "Epoch: 65 - Train cost= 227.62828064\n",
      "Epoch: 65 - Test cost= 235.05458069\n",
      "Epoch: 66 - Train cost= 227.36341858\n",
      "Epoch: 66 - Test cost= 234.98104858\n",
      "Epoch: 67 - Train cost= 227.32164001\n",
      "Epoch: 67 - Test cost= 234.92692566\n",
      "Epoch: 68 - Train cost= 227.20864868\n",
      "Epoch: 68 - Test cost= 235.09378052\n",
      "Epoch: 69 - Train cost= 227.22308350\n",
      "Epoch: 69 - Test cost= 234.83862305\n",
      "Epoch: 70 - Train cost= 227.10253906\n",
      "Epoch: 70 - Test cost= 234.76916504\n",
      "Epoch: 71 - Train cost= 226.97125244\n",
      "Epoch: 71 - Test cost= 234.58845520\n",
      "Epoch: 72 - Train cost= 226.87719727\n",
      "Epoch: 72 - Test cost= 234.64006042\n",
      "Epoch: 73 - Train cost= 226.92449951\n",
      "Epoch: 73 - Test cost= 234.46881104\n",
      "Epoch: 74 - Train cost= 226.72442627\n",
      "Epoch: 74 - Test cost= 234.36546326\n",
      "Epoch: 75 - Train cost= 226.81246948\n",
      "Epoch: 75 - Test cost= 234.34509277\n",
      "Epoch: 76 - Train cost= 226.70570374\n",
      "Epoch: 76 - Test cost= 234.37823486\n",
      "Epoch: 77 - Train cost= 226.52671814\n",
      "Epoch: 77 - Test cost= 234.13044739\n",
      "Epoch: 78 - Train cost= 226.49041748\n",
      "Epoch: 78 - Test cost= 234.35697937\n",
      "Epoch: 79 - Train cost= 226.38972473\n",
      "Epoch: 79 - Test cost= 234.33755493\n",
      "Epoch: 80 - Train cost= 226.39186096\n",
      "Epoch: 80 - Test cost= 234.23635864\n",
      "Epoch: 81 - Train cost= 226.36917114\n",
      "Epoch: 81 - Test cost= 234.15480042\n",
      "Epoch: 82 - Train cost= 226.32061768\n",
      "Epoch: 82 - Test cost= 234.19830322\n",
      "Epoch: 83 - Train cost= 226.21122742\n",
      "Epoch: 83 - Test cost= 233.98011780\n",
      "Epoch: 84 - Train cost= 226.13018799\n",
      "Epoch: 84 - Test cost= 233.89498901\n",
      "Epoch: 85 - Train cost= 225.96803284\n",
      "Epoch: 85 - Test cost= 233.80964661\n",
      "Epoch: 86 - Train cost= 226.00781250\n",
      "Epoch: 86 - Test cost= 233.68077087\n",
      "Epoch: 87 - Train cost= 225.67764282\n",
      "Epoch: 87 - Test cost= 233.74575806\n",
      "Epoch: 88 - Train cost= 225.66996765\n",
      "Epoch: 88 - Test cost= 233.75798035\n",
      "Epoch: 89 - Train cost= 225.58140564\n",
      "Epoch: 89 - Test cost= 233.78265381\n",
      "Epoch: 90 - Train cost= 225.67330933\n",
      "Epoch: 90 - Test cost= 233.59733582\n",
      "Epoch: 91 - Train cost= 225.67077637\n",
      "Epoch: 91 - Test cost= 233.57585144\n",
      "Epoch: 92 - Train cost= 225.68702698\n",
      "Epoch: 92 - Test cost= 233.30764771\n",
      "Epoch: 93 - Train cost= 225.52671814\n",
      "Epoch: 93 - Test cost= 233.49633789\n",
      "Epoch: 94 - Train cost= 225.45494080\n",
      "Epoch: 94 - Test cost= 233.33064270\n",
      "Epoch: 95 - Train cost= 225.21488953\n",
      "Epoch: 95 - Test cost= 233.29183960\n",
      "Epoch: 96 - Train cost= 225.30015564\n",
      "Epoch: 96 - Test cost= 233.24296570\n",
      "Epoch: 97 - Train cost= 225.29496765\n",
      "Epoch: 97 - Test cost= 233.35375977\n",
      "Epoch: 98 - Train cost= 225.18959045\n",
      "Epoch: 98 - Test cost= 233.13638306\n",
      "Epoch: 99 - Train cost= 225.27130127\n",
      "Epoch: 99 - Test cost= 233.11772156\n",
      "Epoch: 100 - Train cost= 225.07298279\n",
      "Epoch: 100 - Test cost= 232.96687317\n",
      "Epoch: 101 - Train cost= 224.89579773\n",
      "Epoch: 101 - Test cost= 233.12173462\n",
      "Epoch: 102 - Train cost= 224.91928101\n",
      "Epoch: 102 - Test cost= 233.01092529\n",
      "Epoch: 103 - Train cost= 224.88800049\n",
      "Epoch: 103 - Test cost= 232.79914856\n",
      "Epoch: 104 - Train cost= 224.93125916\n",
      "Epoch: 104 - Test cost= 232.91616821\n",
      "Epoch: 105 - Train cost= 224.84846497\n",
      "Epoch: 105 - Test cost= 232.68904114\n",
      "Epoch: 106 - Train cost= 224.59426880\n",
      "Epoch: 106 - Test cost= 232.68563843\n",
      "Epoch: 107 - Train cost= 224.71508789\n",
      "Epoch: 107 - Test cost= 232.72770691\n",
      "Epoch: 108 - Train cost= 224.52174377\n",
      "Epoch: 108 - Test cost= 232.74467468\n",
      "Epoch: 109 - Train cost= 224.66197205\n",
      "Epoch: 109 - Test cost= 232.62091064\n",
      "Epoch: 110 - Train cost= 224.59982300\n",
      "Epoch: 110 - Test cost= 232.60565186\n",
      "Epoch: 111 - Train cost= 224.32635498\n",
      "Epoch: 111 - Test cost= 232.55122375\n",
      "Epoch: 112 - Train cost= 224.38333130\n",
      "Epoch: 112 - Test cost= 232.62243652\n",
      "Epoch: 113 - Train cost= 224.19212341\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 113 - Test cost= 232.48220825\n",
      "Epoch: 114 - Train cost= 224.24371338\n",
      "Epoch: 114 - Test cost= 232.65812683\n",
      "Epoch: 115 - Train cost= 224.32565308\n",
      "Epoch: 115 - Test cost= 232.57199097\n",
      "Epoch: 116 - Train cost= 224.17993164\n",
      "Epoch: 116 - Test cost= 232.31045532\n",
      "Epoch: 117 - Train cost= 224.04000854\n",
      "Epoch: 117 - Test cost= 232.29325867\n",
      "Epoch: 118 - Train cost= 224.01019287\n",
      "Epoch: 118 - Test cost= 232.31736755\n",
      "Epoch: 119 - Train cost= 224.01359558\n",
      "Epoch: 119 - Test cost= 232.17993164\n",
      "Epoch: 120 - Train cost= 223.80184937\n",
      "Epoch: 120 - Test cost= 232.44680786\n",
      "Epoch: 121 - Train cost= 223.95426941\n",
      "Epoch: 121 - Test cost= 232.47801208\n",
      "Epoch: 122 - Train cost= 223.78816223\n",
      "Epoch: 122 - Test cost= 232.34786987\n",
      "Epoch: 123 - Train cost= 223.83241272\n",
      "Epoch: 123 - Test cost= 232.19389343\n",
      "Epoch: 124 - Train cost= 223.67933655\n",
      "Epoch: 124 - Test cost= 232.15509033\n",
      "Epoch: 125 - Train cost= 223.75177002\n",
      "Epoch: 125 - Test cost= 232.14883423\n",
      "Epoch: 126 - Train cost= 223.48614502\n",
      "Epoch: 126 - Test cost= 232.30470276\n",
      "Epoch: 127 - Train cost= 223.20545959\n",
      "Epoch: 127 - Test cost= 231.70236206\n",
      "Epoch: 128 - Train cost= 222.99044800\n",
      "Epoch: 128 - Test cost= 231.58848572\n",
      "Epoch: 129 - Train cost= 223.02096558\n",
      "Epoch: 129 - Test cost= 231.45014954\n",
      "Epoch: 130 - Train cost= 222.93725586\n",
      "Epoch: 130 - Test cost= 231.89373779\n",
      "Epoch: 131 - Train cost= 222.84112549\n",
      "Epoch: 131 - Test cost= 231.31425476\n",
      "Epoch: 132 - Train cost= 222.80770874\n",
      "Epoch: 132 - Test cost= 231.85649109\n",
      "Epoch: 133 - Train cost= 222.73704529\n",
      "Epoch: 133 - Test cost= 231.55818176\n",
      "Epoch: 134 - Train cost= 222.81530762\n",
      "Epoch: 134 - Test cost= 231.26152039\n",
      "Epoch: 135 - Train cost= 222.68838501\n",
      "Epoch: 135 - Test cost= 231.31842041\n",
      "Epoch: 136 - Train cost= 222.72154236\n",
      "Epoch: 136 - Test cost= 231.50144958\n",
      "Epoch: 137 - Train cost= 222.45707703\n",
      "Epoch: 137 - Test cost= 231.39468384\n",
      "Epoch: 138 - Train cost= 222.75862122\n",
      "Epoch: 138 - Test cost= 231.36343384\n",
      "Epoch: 139 - Train cost= 222.69624329\n",
      "Epoch: 139 - Test cost= 231.45735168\n",
      "Epoch: 140 - Train cost= 222.71479797\n",
      "Epoch: 140 - Test cost= 231.40476990\n",
      "Epoch: 141 - Train cost= 222.55456543\n",
      "Epoch: 141 - Test cost= 231.59130859\n",
      "Epoch: 142 - Train cost= 222.66239929\n",
      "Epoch: 142 - Test cost= 231.24465942\n",
      "Epoch: 143 - Train cost= 222.54092407\n",
      "Epoch: 143 - Test cost= 231.54008484\n",
      "Epoch: 144 - Train cost= 222.61659241\n",
      "Epoch: 144 - Test cost= 231.41955566\n",
      "Epoch: 145 - Train cost= 222.53800964\n",
      "Epoch: 145 - Test cost= 231.31021118\n",
      "Epoch: 146 - Train cost= 222.52751160\n",
      "Epoch: 146 - Test cost= 230.99615479\n",
      "Epoch: 147 - Train cost= 222.64331055\n",
      "Epoch: 147 - Test cost= 231.31610107\n",
      "Epoch: 148 - Train cost= 222.49342346\n",
      "Epoch: 148 - Test cost= 231.33055115\n",
      "Epoch: 149 - Train cost= 222.63761902\n",
      "Epoch: 149 - Test cost= 231.44073486\n",
      "Epoch: 150 - Train cost= 222.73599243\n",
      "Epoch: 150 - Test cost= 231.40580750\n"
     ]
    }
   ],
   "source": [
    "model.fit(x_train, x_test, epoch=150, batch_size=128,x=10, steph_epoch=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train History"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_history(model.train_vector,model.test_vector)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Input images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x7200 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images(model.batch_img, 'Inputs')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Output images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x7200 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images(model.recon_img, 'Outputs')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reconstructed images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x7200 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images(model.new_output, 'Reconstruction')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reconstruction images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x7200 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.generator_data()"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "Train_DCVAE_MPS45_tensorflow.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}