Train_DCVAE_Inc_MPS45_Keras.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n"
     ]
    }
   ],
   "source": [
    "from Model.DCVAE import DCVAE_Inc\n",
    "from Model.Utils import PlotHistory, PlotDataAE, Save_Model\n",
    "from Model.Utils import load_from_tfrecords, convert_to_tfrecords, load_numpy\n",
    "from keras.optimizers import RMSprop,Adam\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import os\n",
    "%matplotlib notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load DataSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "load_tfrecords = True \n",
    "create_tfrecords = False\n",
    "\n",
    "path_tfr = 'DataSet/MPS45'\n",
    "path_npy = '/share/GeoFacies/DataSet/MPS45/MPS45.npy'\n",
    "batch_size = 32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "if load_tfrecords:\n",
    "    if create_tfrecords: \n",
    "        x_train, x_test = load_numpy(path_npy,random_state=0,split_data=0.30)\n",
    "        convert_to_tfrecords(path_tfr, x_train, x_test)\n",
    "        gen_train, gen_test = load_from_tfrecords(path_tfr, batch_size)\n",
    "    else:\n",
    "        try:\n",
    "            gen_train, gen_test = load_from_tfrecords(path_tfr, batch_size)\n",
    "        except:\n",
    "            print(\"Data not found. Change 'create_tfrecords' to True\")\n",
    "else:\n",
    "    x_train, x_test = load_numpy(path_npy,random_state=0,split_data=0.30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create Convolutional Variational Autoencoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "path_weights = os.path.join(path_tfr, 'w45_Inc.hdf5')\n",
    "\n",
    "model = DCVAE_Inc(input_shape=(45, 45, 2), kernel_init=32, latent_dim=500, drop = 0,\n",
    "                  opt=RMSprop(lr=0.001),scheduler=True, filepath = path_weights)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting up model...\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 45, 45, 2)    0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, 45, 45, 32)   96          input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, 45, 45, 16)   528         conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, 45, 45, 16)   64          conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, 45, 45, 16)   0           batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, 45, 45, 32)   544         activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, 45, 45, 32)   4640        activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, 45, 45, 32)   12832       activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, 45, 45, 32)   25120       activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, 45, 45, 32)   128         conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, 45, 45, 32)   128         conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, 45, 45, 32)   128         conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, 45, 45, 32)   128         conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, 45, 45, 32)   0           batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 45, 45, 32)   0           batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_4 (Activation)       (None, 45, 45, 32)   0           batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_5 (Activation)       (None, 45, 45, 32)   0           batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, 45, 45, 32)   0           conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)  (None, 45, 45, 32)   0           conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "add_1 (Add)                     (None, 45, 45, 32)   0           activation_2[0][0]               \n",
      "                                                                 activation_3[0][0]               \n",
      "                                                                 activation_4[0][0]               \n",
      "                                                                 activation_5[0][0]               \n",
      "                                                                 max_pooling2d_1[0][0]            \n",
      "                                                                 max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_1 (AveragePoo (None, 22, 22, 32)   0           add_1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, 22, 22, 64)   2112        average_pooling2d_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, 22, 22, 64)   256         conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_6 (Activation)       (None, 22, 22, 64)   0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, 22, 22, 32)   2080        activation_6[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, 22, 22, 32)   128         conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_7 (Activation)       (None, 22, 22, 32)   0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_9 (Conv2D)               (None, 22, 22, 64)   2112        activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_10 (Conv2D)              (None, 22, 22, 64)   18496       activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_11 (Conv2D)              (None, 22, 22, 64)   51264       activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_12 (Conv2D)              (None, 22, 22, 64)   100416      activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, 22, 22, 64)   256         conv2d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, 22, 22, 64)   256         conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, 22, 22, 64)   256         conv2d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, 22, 22, 64)   256         conv2d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_8 (Activation)       (None, 22, 22, 64)   0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_9 (Activation)       (None, 22, 22, 64)   0           batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_10 (Activation)      (None, 22, 22, 64)   0           batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_11 (Activation)      (None, 22, 22, 64)   0           batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)  (None, 22, 22, 64)   0           activation_6[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2D)  (None, 22, 22, 64)   0           activation_6[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "add_2 (Add)                     (None, 22, 22, 64)   0           activation_8[0][0]               \n",
      "                                                                 activation_9[0][0]               \n",
      "                                                                 activation_10[0][0]              \n",
      "                                                                 activation_11[0][0]              \n",
      "                                                                 max_pooling2d_3[0][0]            \n",
      "                                                                 max_pooling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_2 (AveragePoo (None, 11, 11, 64)   0           add_2[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_13 (Conv2D)              (None, 11, 11, 128)  8320        average_pooling2d_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, 11, 11, 128)  512         conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_12 (Activation)      (None, 11, 11, 128)  0           batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_14 (Conv2D)              (None, 11, 11, 64)   8256        activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, 11, 11, 64)   256         conv2d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_13 (Activation)      (None, 11, 11, 64)   0           batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_15 (Conv2D)              (None, 11, 11, 128)  8320        activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_16 (Conv2D)              (None, 11, 11, 128)  73856       activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_17 (Conv2D)              (None, 11, 11, 128)  204928      activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_18 (Conv2D)              (None, 11, 11, 128)  401536      activation_13[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, 11, 11, 128)  512         conv2d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, 11, 11, 128)  512         conv2d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, 11, 11, 128)  512         conv2d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, 11, 11, 128)  512         conv2d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_14 (Activation)      (None, 11, 11, 128)  0           batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_15 (Activation)      (None, 11, 11, 128)  0           batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_16 (Activation)      (None, 11, 11, 128)  0           batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_17 (Activation)      (None, 11, 11, 128)  0           batch_normalization_17[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_5 (MaxPooling2D)  (None, 11, 11, 128)  0           activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_6 (MaxPooling2D)  (None, 11, 11, 128)  0           activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_3 (Add)                     (None, 11, 11, 128)  0           activation_14[0][0]              \n",
      "                                                                 activation_15[0][0]              \n",
      "                                                                 activation_16[0][0]              \n",
      "                                                                 activation_17[0][0]              \n",
      "                                                                 max_pooling2d_5[0][0]            \n",
      "                                                                 max_pooling2d_6[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_3 (AveragePoo (None, 2, 2, 128)    0           add_3[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_19 (Conv2D)              (None, 2, 2, 256)    33024       average_pooling2d_3[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_18 (BatchNo (None, 2, 2, 256)    1024        conv2d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_18 (Activation)      (None, 2, 2, 256)    0           batch_normalization_18[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_20 (Conv2D)              (None, 2, 2, 128)    32896       activation_18[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_19 (BatchNo (None, 2, 2, 128)    512         conv2d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_19 (Activation)      (None, 2, 2, 128)    0           batch_normalization_19[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_21 (Conv2D)              (None, 2, 2, 256)    33024       activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_22 (Conv2D)              (None, 2, 2, 256)    295168      activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_23 (Conv2D)              (None, 2, 2, 256)    819456      activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_24 (Conv2D)              (None, 2, 2, 256)    1605888     activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_20 (BatchNo (None, 2, 2, 256)    1024        conv2d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_21 (BatchNo (None, 2, 2, 256)    1024        conv2d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_22 (BatchNo (None, 2, 2, 256)    1024        conv2d_23[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_23 (BatchNo (None, 2, 2, 256)    1024        conv2d_24[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_20 (Activation)      (None, 2, 2, 256)    0           batch_normalization_20[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_21 (Activation)      (None, 2, 2, 256)    0           batch_normalization_21[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_22 (Activation)      (None, 2, 2, 256)    0           batch_normalization_22[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_23 (Activation)      (None, 2, 2, 256)    0           batch_normalization_23[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_7 (MaxPooling2D)  (None, 2, 2, 256)    0           activation_18[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_8 (MaxPooling2D)  (None, 2, 2, 256)    0           activation_18[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_4 (Add)                     (None, 2, 2, 256)    0           activation_20[0][0]              \n",
      "                                                                 activation_21[0][0]              \n",
      "                                                                 activation_22[0][0]              \n",
      "                                                                 activation_23[0][0]              \n",
      "                                                                 max_pooling2d_7[0][0]            \n",
      "                                                                 max_pooling2d_8[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_4 (AveragePoo (None, 1, 1, 256)    0           add_4[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "lambda_1 (Lambda)               (None, 1, 256)       0           average_pooling2d_4[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "lambda_2 (Lambda)               (None, 256)          0           lambda_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 256)          0           lambda_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 500)          128500      dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 500)          128500      dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "lambda_3 (Lambda)               (None, 500)          0           dense_1[0][0]                    \n",
      "                                                                 dense_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_3 (Dense)                 (None, 256)          128256      lambda_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 256)          0           dense_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "lambda_4 (Lambda)               (None, 1, 256)       0           dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "lambda_5 (Lambda)               (None, 1, 1, 256)    0           lambda_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_1 (UpSampling2D)  (None, 3, 3, 256)    0           lambda_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_1 (Conv2DTrans (None, 3, 3, 128)    32896       up_sampling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_24 (BatchNo (None, 3, 3, 128)    512         conv2d_transpose_1[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_24 (Activation)      (None, 3, 3, 128)    0           batch_normalization_24[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_2 (Conv2DTrans (None, 3, 3, 256)    33024       activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_3 (Conv2DTrans (None, 3, 3, 256)    295168      activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_4 (Conv2DTrans (None, 3, 3, 256)    819456      activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_5 (Conv2DTrans (None, 3, 3, 256)    1605888     activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_25 (BatchNo (None, 3, 3, 256)    1024        conv2d_transpose_2[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_26 (BatchNo (None, 3, 3, 256)    1024        conv2d_transpose_3[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_27 (BatchNo (None, 3, 3, 256)    1024        conv2d_transpose_4[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_28 (BatchNo (None, 3, 3, 256)    1024        conv2d_transpose_5[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_25 (Activation)      (None, 3, 3, 256)    0           batch_normalization_25[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_26 (Activation)      (None, 3, 3, 256)    0           batch_normalization_26[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_27 (Activation)      (None, 3, 3, 256)    0           batch_normalization_27[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_28 (Activation)      (None, 3, 3, 256)    0           batch_normalization_28[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_9 (MaxPooling2D)  (None, 3, 3, 256)    0           up_sampling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_10 (MaxPooling2D) (None, 3, 3, 256)    0           up_sampling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "add_5 (Add)                     (None, 3, 3, 256)    0           activation_25[0][0]              \n",
      "                                                                 activation_26[0][0]              \n",
      "                                                                 activation_27[0][0]              \n",
      "                                                                 activation_28[0][0]              \n",
      "                                                                 max_pooling2d_9[0][0]            \n",
      "                                                                 max_pooling2d_10[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_6 (Conv2DTrans (None, 12, 12, 128)  524416      add_5[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_29 (BatchNo (None, 12, 12, 128)  512         conv2d_transpose_6[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_29 (Activation)      (None, 12, 12, 128)  0           batch_normalization_29[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_7 (Conv2DTrans (None, 12, 12, 64)   8256        activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_30 (BatchNo (None, 12, 12, 64)   256         conv2d_transpose_7[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "activation_30 (Activation)      (None, 12, 12, 64)   0           batch_normalization_30[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_8 (Conv2DTrans (None, 12, 12, 128)  8320        activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_9 (Conv2DTrans (None, 12, 12, 128)  73856       activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_10 (Conv2DTran (None, 12, 12, 128)  204928      activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_11 (Conv2DTran (None, 12, 12, 128)  401536      activation_30[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_31 (BatchNo (None, 12, 12, 128)  512         conv2d_transpose_8[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_32 (BatchNo (None, 12, 12, 128)  512         conv2d_transpose_9[0][0]         \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_33 (BatchNo (None, 12, 12, 128)  512         conv2d_transpose_10[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_34 (BatchNo (None, 12, 12, 128)  512         conv2d_transpose_11[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_31 (Activation)      (None, 12, 12, 128)  0           batch_normalization_31[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_32 (Activation)      (None, 12, 12, 128)  0           batch_normalization_32[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_33 (Activation)      (None, 12, 12, 128)  0           batch_normalization_33[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_34 (Activation)      (None, 12, 12, 128)  0           batch_normalization_34[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_11 (MaxPooling2D) (None, 12, 12, 128)  0           activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_12 (MaxPooling2D) (None, 12, 12, 128)  0           activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_6 (Add)                     (None, 12, 12, 128)  0           activation_31[0][0]              \n",
      "                                                                 activation_32[0][0]              \n",
      "                                                                 activation_33[0][0]              \n",
      "                                                                 activation_34[0][0]              \n",
      "                                                                 max_pooling2d_11[0][0]           \n",
      "                                                                 max_pooling2d_12[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_12 (Conv2DTran (None, 24, 24, 64)   32832       add_6[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_35 (BatchNo (None, 24, 24, 64)   256         conv2d_transpose_12[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_35 (Activation)      (None, 24, 24, 64)   0           batch_normalization_35[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_13 (Conv2DTran (None, 24, 24, 32)   2080        activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_36 (BatchNo (None, 24, 24, 32)   128         conv2d_transpose_13[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_36 (Activation)      (None, 24, 24, 32)   0           batch_normalization_36[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_14 (Conv2DTran (None, 24, 24, 64)   2112        activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_15 (Conv2DTran (None, 24, 24, 64)   18496       activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_16 (Conv2DTran (None, 24, 24, 64)   51264       activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_17 (Conv2DTran (None, 24, 24, 64)   100416      activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_37 (BatchNo (None, 24, 24, 64)   256         conv2d_transpose_14[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_38 (BatchNo (None, 24, 24, 64)   256         conv2d_transpose_15[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_39 (BatchNo (None, 24, 24, 64)   256         conv2d_transpose_16[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_40 (BatchNo (None, 24, 24, 64)   256         conv2d_transpose_17[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_37 (Activation)      (None, 24, 24, 64)   0           batch_normalization_37[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_38 (Activation)      (None, 24, 24, 64)   0           batch_normalization_38[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_39 (Activation)      (None, 24, 24, 64)   0           batch_normalization_39[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_40 (Activation)      (None, 24, 24, 64)   0           batch_normalization_40[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_13 (MaxPooling2D) (None, 24, 24, 64)   0           activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_14 (MaxPooling2D) (None, 24, 24, 64)   0           activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_7 (Add)                     (None, 24, 24, 64)   0           activation_37[0][0]              \n",
      "                                                                 activation_38[0][0]              \n",
      "                                                                 activation_39[0][0]              \n",
      "                                                                 activation_40[0][0]              \n",
      "                                                                 max_pooling2d_13[0][0]           \n",
      "                                                                 max_pooling2d_14[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_18 (Conv2DTran (None, 48, 48, 32)   8224        add_7[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_41 (BatchNo (None, 48, 48, 32)   128         conv2d_transpose_18[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_41 (Activation)      (None, 48, 48, 32)   0           batch_normalization_41[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_19 (Conv2DTran (None, 48, 48, 16)   528         activation_41[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_42 (BatchNo (None, 48, 48, 16)   64          conv2d_transpose_19[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_42 (Activation)      (None, 48, 48, 16)   0           batch_normalization_42[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_20 (Conv2DTran (None, 48, 48, 32)   544         activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_21 (Conv2DTran (None, 48, 48, 32)   4640        activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_22 (Conv2DTran (None, 48, 48, 32)   12832       activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_23 (Conv2DTran (None, 48, 48, 32)   25120       activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_43 (BatchNo (None, 48, 48, 32)   128         conv2d_transpose_20[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_44 (BatchNo (None, 48, 48, 32)   128         conv2d_transpose_21[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_45 (BatchNo (None, 48, 48, 32)   128         conv2d_transpose_22[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_46 (BatchNo (None, 48, 48, 32)   128         conv2d_transpose_23[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "activation_43 (Activation)      (None, 48, 48, 32)   0           batch_normalization_43[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_44 (Activation)      (None, 48, 48, 32)   0           batch_normalization_44[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_45 (Activation)      (None, 48, 48, 32)   0           batch_normalization_45[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_46 (Activation)      (None, 48, 48, 32)   0           batch_normalization_46[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_15 (MaxPooling2D) (None, 48, 48, 32)   0           activation_41[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_16 (MaxPooling2D) (None, 48, 48, 32)   0           activation_41[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "add_8 (Add)                     (None, 48, 48, 32)   0           activation_43[0][0]              \n",
      "                                                                 activation_44[0][0]              \n",
      "                                                                 activation_45[0][0]              \n",
      "                                                                 activation_46[0][0]              \n",
      "                                                                 max_pooling2d_15[0][0]           \n",
      "                                                                 max_pooling2d_16[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "bilinear_upsampling_1 (Bilinear (None, 45, 45, 32)   0           add_8[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_transpose_24 (Conv2DTran (None, 45, 45, 2)    66          bilinear_upsampling_1[0][0]      \n",
      "==================================================================================================\n",
      "Total params: 8,417,034\n",
      "Trainable params: 8,407,050\n",
      "Non-trainable params: 9,984\n",
      "__________________________________________________________________________________________________\n",
      "Completed model setup.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bb2fda81ec84479791c7162f9e85565b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Training', max=500, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 0', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 66s 100ms/step - loss: 1052.1886 - acc_pred: 0.8095 - val_loss: 1064.4657 - val_acc_pred: 0.8125\n",
      "Epoch 2/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 1', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 646.7518 - acc_pred: 0.8805 - val_loss: 684.7760 - val_acc_pred: 0.8746\n",
      "Epoch 3/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 2', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 560.2322 - acc_pred: 0.8977 - val_loss: 648.7970 - val_acc_pred: 0.8832\n",
      "Epoch 4/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 3', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 514.4923 - acc_pred: 0.9065 - val_loss: 537.3151 - val_acc_pred: 0.9029\n",
      "Epoch 5/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 4', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 475.2345 - acc_pred: 0.9143 - val_loss: 530.6227 - val_acc_pred: 0.9007\n",
      "Epoch 6/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 5', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 442.2480 - acc_pred: 0.9205 - val_loss: 442.9352 - val_acc_pred: 0.9199\n",
      "Epoch 7/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 6', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 412.5811 - acc_pred: 0.9261 - val_loss: 418.4936 - val_acc_pred: 0.9241\n",
      "Epoch 8/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 7', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 389.2624 - acc_pred: 0.9304 - val_loss: 394.1410 - val_acc_pred: 0.9294\n",
      "Epoch 9/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 8', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 371.0227 - acc_pred: 0.9338 - val_loss: 391.9810 - val_acc_pred: 0.9318\n",
      "Epoch 10/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 9', max=656, style=ProgressStyle(description_width='ini…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 357.7640 - acc_pred: 0.9364 - val_loss: 357.5765 - val_acc_pred: 0.9362\n",
      "Epoch 11/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 10', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 346.4413 - acc_pred: 0.9386 - val_loss: 341.5618 - val_acc_pred: 0.9385\n",
      "Epoch 12/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 11', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 336.0585 - acc_pred: 0.9406 - val_loss: 344.0635 - val_acc_pred: 0.9395\n",
      "Epoch 13/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 12', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 327.3245 - acc_pred: 0.9424 - val_loss: 348.4396 - val_acc_pred: 0.9377\n",
      "Epoch 14/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 13', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 319.3073 - acc_pred: 0.9440 - val_loss: 328.8151 - val_acc_pred: 0.9422\n",
      "Epoch 15/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 14', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 312.6533 - acc_pred: 0.9453 - val_loss: 332.3693 - val_acc_pred: 0.9408\n",
      "Epoch 16/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 15', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 306.8721 - acc_pred: 0.9465 - val_loss: 319.2750 - val_acc_pred: 0.9432\n",
      "Epoch 17/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 16', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 302.0582 - acc_pred: 0.9475 - val_loss: 320.0439 - val_acc_pred: 0.9431\n",
      "Epoch 18/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 17', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 54s 82ms/step - loss: 297.5466 - acc_pred: 0.9484 - val_loss: 324.9306 - val_acc_pred: 0.9429\n",
      "Epoch 19/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 18', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 292.7918 - acc_pred: 0.9495 - val_loss: 310.0583 - val_acc_pred: 0.9452\n",
      "Epoch 20/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 19', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 289.8091 - acc_pred: 0.9501 - val_loss: 292.8980 - val_acc_pred: 0.9494\n",
      "Epoch 21/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 20', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 285.7266 - acc_pred: 0.9510 - val_loss: 308.9161 - val_acc_pred: 0.9464\n",
      "Epoch 22/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 21', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 282.7081 - acc_pred: 0.9517 - val_loss: 288.4548 - val_acc_pred: 0.9499\n",
      "Epoch 23/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 22', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 279.9748 - acc_pred: 0.9523 - val_loss: 291.1799 - val_acc_pred: 0.9500\n",
      "Epoch 24/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 23', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 277.2554 - acc_pred: 0.9529 - val_loss: 298.9296 - val_acc_pred: 0.9489\n",
      "Epoch 25/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 24', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 274.4517 - acc_pred: 0.9535 - val_loss: 297.1014 - val_acc_pred: 0.9483\n",
      "Epoch 26/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 25', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 272.5441 - acc_pred: 0.9539 - val_loss: 282.5981 - val_acc_pred: 0.9512\n",
      "Epoch 27/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 26', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 270.7592 - acc_pred: 0.9543 - val_loss: 288.5152 - val_acc_pred: 0.9503\n",
      "Epoch 28/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 27', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 268.5390 - acc_pred: 0.9548 - val_loss: 277.0944 - val_acc_pred: 0.9529\n",
      "Epoch 29/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 28', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 267.2307 - acc_pred: 0.9551 - val_loss: 281.6495 - val_acc_pred: 0.9518\n",
      "Epoch 30/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 29', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 80ms/step - loss: 265.2061 - acc_pred: 0.9555 - val_loss: 287.7251 - val_acc_pred: 0.9509\n",
      "Epoch 31/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 30', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 264.1719 - acc_pred: 0.9558 - val_loss: 292.9821 - val_acc_pred: 0.9497\n",
      "Epoch 32/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 31', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 262.3023 - acc_pred: 0.9562 - val_loss: 286.3262 - val_acc_pred: 0.9514\n",
      "Epoch 33/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 32', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 261.0037 - acc_pred: 0.9565 - val_loss: 296.1311 - val_acc_pred: 0.9488\n",
      "\n",
      "Epoch 00033: ReduceLROnPlateau reducing learning rate to 0.0005000000237487257.\n",
      "Epoch 34/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 33', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 245.9934 - acc_pred: 0.9598 - val_loss: 261.8310 - val_acc_pred: 0.9564\n",
      "Epoch 35/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 34', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 243.0741 - acc_pred: 0.9605 - val_loss: 261.2853 - val_acc_pred: 0.9565\n",
      "Epoch 36/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 35', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 241.2600 - acc_pred: 0.9608 - val_loss: 260.0366 - val_acc_pred: 0.9568\n",
      "Epoch 37/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 36', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 240.4553 - acc_pred: 0.9611 - val_loss: 261.7911 - val_acc_pred: 0.9568\n",
      "Epoch 38/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 37', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 239.1090 - acc_pred: 0.9613 - val_loss: 273.3960 - val_acc_pred: 0.9539\n",
      "Epoch 39/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 38', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 238.2284 - acc_pred: 0.9616 - val_loss: 259.3053 - val_acc_pred: 0.9575\n",
      "Epoch 40/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 39', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 237.3494 - acc_pred: 0.9617 - val_loss: 262.8489 - val_acc_pred: 0.9565\n",
      "Epoch 41/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 40', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 236.5820 - acc_pred: 0.9619 - val_loss: 261.9982 - val_acc_pred: 0.9568\n",
      "Epoch 42/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 41', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 79ms/step - loss: 235.8141 - acc_pred: 0.9621 - val_loss: 258.7071 - val_acc_pred: 0.9574\n",
      "Epoch 43/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 42', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 234.7949 - acc_pred: 0.9624 - val_loss: 259.8473 - val_acc_pred: 0.9573\n",
      "Epoch 44/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 43', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 233.9671 - acc_pred: 0.9625 - val_loss: 262.5945 - val_acc_pred: 0.9568\n",
      "Epoch 45/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 44', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 232.9275 - acc_pred: 0.9627 - val_loss: 259.9737 - val_acc_pred: 0.9574\n",
      "Epoch 46/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 45', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 232.5146 - acc_pred: 0.9628 - val_loss: 256.6277 - val_acc_pred: 0.9580\n",
      "Epoch 47/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 46', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 231.5797 - acc_pred: 0.9630 - val_loss: 261.0199 - val_acc_pred: 0.9574\n",
      "Epoch 48/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 47', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 230.6181 - acc_pred: 0.9633 - val_loss: 258.1924 - val_acc_pred: 0.9578\n",
      "Epoch 49/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 48', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 230.1109 - acc_pred: 0.9634 - val_loss: 260.0402 - val_acc_pred: 0.9574\n",
      "Epoch 50/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 49', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 229.4071 - acc_pred: 0.9635 - val_loss: 259.5340 - val_acc_pred: 0.9575\n",
      "Epoch 51/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 50', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 228.6616 - acc_pred: 0.9637 - val_loss: 259.4374 - val_acc_pred: 0.9575\n",
      "\n",
      "Epoch 00051: ReduceLROnPlateau reducing learning rate to 0.0002500000118743628.\n",
      "Epoch 52/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 51', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 80ms/step - loss: 222.1211 - acc_pred: 0.9652 - val_loss: 253.6577 - val_acc_pred: 0.9592\n",
      "Epoch 53/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 52', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 80ms/step - loss: 220.8720 - acc_pred: 0.9654 - val_loss: 253.0903 - val_acc_pred: 0.9592\n",
      "Epoch 54/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 53', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 219.8704 - acc_pred: 0.9656 - val_loss: 254.8055 - val_acc_pred: 0.9589\n",
      "Epoch 55/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 54', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 219.3840 - acc_pred: 0.9658 - val_loss: 254.7836 - val_acc_pred: 0.9591\n",
      "Epoch 56/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 55', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 218.9640 - acc_pred: 0.9659 - val_loss: 254.1590 - val_acc_pred: 0.9593\n",
      "Epoch 57/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 56', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 218.3984 - acc_pred: 0.9660 - val_loss: 254.8861 - val_acc_pred: 0.9591\n",
      "Epoch 58/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 57', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 217.8112 - acc_pred: 0.9662 - val_loss: 257.0394 - val_acc_pred: 0.9587\n",
      "\n",
      "Epoch 00058: ReduceLROnPlateau reducing learning rate to 0.0001250000059371814.\n",
      "Epoch 59/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 58', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 215.0368 - acc_pred: 0.9667 - val_loss: 252.2263 - val_acc_pred: 0.9599\n",
      "Epoch 60/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 59', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 214.2617 - acc_pred: 0.9670 - val_loss: 252.0288 - val_acc_pred: 0.9600\n",
      "Epoch 61/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 60', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 213.7294 - acc_pred: 0.9671 - val_loss: 252.7445 - val_acc_pred: 0.9599\n",
      "Epoch 62/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 61', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 213.4853 - acc_pred: 0.9671 - val_loss: 253.2293 - val_acc_pred: 0.9598\n",
      "Epoch 63/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 62', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 213.3411 - acc_pred: 0.9672 - val_loss: 253.4802 - val_acc_pred: 0.9597\n",
      "Epoch 64/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 63', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 80ms/step - loss: 213.0576 - acc_pred: 0.9673 - val_loss: 253.1994 - val_acc_pred: 0.9598\n",
      "Epoch 65/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 64', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 80ms/step - loss: 212.4458 - acc_pred: 0.9674 - val_loss: 254.2205 - val_acc_pred: 0.9598\n",
      "\n",
      "Epoch 00065: ReduceLROnPlateau reducing learning rate to 6.25000029685907e-05.\n",
      "Epoch 66/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 65', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 211.2026 - acc_pred: 0.9677 - val_loss: 252.1697 - val_acc_pred: 0.9601\n",
      "Epoch 67/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 66', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 211.0220 - acc_pred: 0.9677 - val_loss: 252.6359 - val_acc_pred: 0.9601\n",
      "Epoch 68/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 67', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 210.6577 - acc_pred: 0.9678 - val_loss: 252.7562 - val_acc_pred: 0.9601\n",
      "Epoch 69/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 68', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 52s 80ms/step - loss: 210.4398 - acc_pred: 0.9679 - val_loss: 252.7096 - val_acc_pred: 0.9602\n",
      "Epoch 70/500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch 69', max=656, style=ProgressStyle(description_width='in…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "656/656 [==============================] - 53s 81ms/step - loss: 210.2450 - acc_pred: 0.9679 - val_loss: 253.0215 - val_acc_pred: 0.9601\n",
      "\n",
      "Epoch 00070: ReduceLROnPlateau reducing learning rate to 3.125000148429535e-05.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "num_epochs = 500\n",
    "\n",
    "if load_tfrecords:\n",
    "    x_train = gen_train.mps_generator()\n",
    "    x_val = gen_test.mps_generator()\n",
    "    model.fit_generator(x_train,\n",
    "                        num_epochs=num_epochs, verbose=1, \n",
    "                        steps_per_epoch = len(gen_train),\n",
    "                        val_set = x_val,\n",
    "                        validation_steps = len(gen_test))\n",
    "else:\n",
    "    model.fit(x_train, num_epochs=num_epochs, batch_size = batch_size, x_v=x_test,verbose=1) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ploting Train History"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss   :  [252.1697049938487, 252.63589026834617, 252.75619604579072, 252.70960417506532]\n",
      "loss   :  [211.2025574940007, 211.02196914393727, 210.65770465571705, 210.4397807470182]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotHistory(model.history.history,listKeys=['val_loss','loss']) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_acc_pred   :  [0.960149983282191, 0.9601214260817422, 0.9601002269792387, 0.9601551434323456]\n",
      "acc_pred   :  [0.9676642964889364, 0.9677371986937232, 0.9677763419543824, 0.967854514263752]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotHistory(model.history.history,listKeys=['val_acc_pred','acc_pred'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lr   :  [6.25e-05, 6.25e-05, 6.25e-05, 6.25e-05]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotHistory(model.history.history,listKeys=['lr'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate Model with Test dataset "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1620x1620 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1620x1620 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "if load_tfrecords:\n",
    "    x_test = gen_test.get_numpy_batch()\n",
    "\n",
    "x_rec = model.model.predict(x_test)\n",
    "PlotDataAE(x_test[:,:,:,1], np.argmax(x_rec[:,:,:,:],axis=-1),digit_size=(45,45))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generate random sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1620x1620 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_gen=model.generate(binary=True)\n",
    "PlotDataAE([],x_gen[:,:,:],digit_size=(45,45),Only_Result=False)"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}