Train_CNN_Digital-Readout-Small-v0.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CNN Training\n",
    "\n",
    "Target of this code is to train a CNN network to classify images of a digital readout to the digits 0 to 9. Additionally a category \"NaN\" is introduced, to mark images that are not amibiguous.\n",
    "\n",
    "### Preparing the training\n",
    "* First all libraries are loaded\n",
    "    * It is assumed, that they are installed during the Python setup\n",
    "* matplotlib is set to print the output inline in the jupyter notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "########### Basic Parameters for Running: ################################\n",
    "    \n",
    "TFliteNamingAndVersion = \"dig1410s1\"   # Used for tflite Filename\n",
    "Training_Percentage = 0.0              # 0.0 = Use all Images for Training\n",
    "Epoch_Anz = 500\n",
    "\n",
    "##########################################################################\n",
    "\n",
    "\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import glob\n",
    "import os\n",
    "import numpy as np\n",
    "from sklearn.utils import shuffle\n",
    "from tensorflow.python import keras\n",
    "from tensorflow.python.keras import Sequential\n",
    "from tensorflow.keras.layers import Dense, InputLayer, Conv2D, MaxPool2D, Flatten, BatchNormalization\n",
    "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tensorflow.keras.callbacks import History \n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from PIL import Image \n",
    "from pathlib import Path\n",
    "\n",
    "loss_ges = np.array([])\n",
    "val_loss_ges = np.array([])\n",
    "\n",
    "%matplotlib inline\n",
    "np.set_printoptions(precision=4)\n",
    "np.set_printoptions(suppress=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load training data\n",
    "* The data is expected in the \"Input_dir\"\n",
    "* Inside subdirectories are expected from -1, 0, 1, ... 9 in which the pictures are sorted according to their values (=category)\n",
    "* Picture size must be 20x32 with 3 color channels (RGB)\n",
    "* The filename can be arbitrary\n",
    "\n",
    "* The images are stored in the x_data[]\n",
    "* The expected category for each image in the corresponding y_data[]\n",
    "\n",
    "* The last step is a shuffle (from sklearn.utils) and split the data into training and validation data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1325, 32, 20, 3)\n",
      "(1325, 11)\n"
     ]
    }
   ],
   "source": [
    "Input_dir='ziffer_sortiert_resize'\n",
    "\n",
    "files = glob.glob(Input_dir + '/*.jpg')\n",
    "x_data = []\n",
    "y_data = []\n",
    "\n",
    "for aktfile in files:\n",
    "    base = os.path.basename(aktfile)\n",
    "    target = base[0:1]\n",
    "    if target == \"N\":\n",
    "        category = 10                # NaN does not work --> convert to 10\n",
    "    else:\n",
    "        category = int(target)\n",
    "    test_image = Image.open(aktfile)\n",
    "    test_image = np.array(test_image, dtype=\"float32\")\n",
    "    x_data.append(test_image)\n",
    "    y_data.append(np.array([category]))\n",
    "\n",
    "x_data = np.array(x_data)\n",
    "y_data = np.array(y_data)\n",
    "y_data = to_categorical(y_data, 11)\n",
    "print(x_data.shape)\n",
    "print(y_data.shape)\n",
    "\n",
    "x_data, y_data = shuffle(x_data, y_data)\n",
    "\n",
    "if (Training_Percentage > 0):\n",
    "    X_train, X_test, y_train, y_test = train_test_split(x_data, y_data, test_size=Training_Percentage)\n",
    "else:\n",
    "    X_train = x_data\n",
    "    y_train = y_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define the model\n",
    "\n",
    "The layout of the network ist a typcial CNN network with alternating **Conv2D** and **MaxPool2D** layers. Finished after **flattening** with additional **Dense** layer.\n",
    "\n",
    "#### Important\n",
    "* Shape of the input layer: (32, 20, 3)\n",
    "* Number of output layers: 11\n",
    "* As loss function \"categorical_crossentropy\" is choosen, as it is a categories task"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 32, 20, 3)]       0         \n",
      "                                                                 \n",
      " batch_normalization (BatchN  (None, 32, 20, 3)        12        \n",
      " ormalization)                                                   \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 32, 20, 32)        896       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 16, 10, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 16, 10, 64)        18496     \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 8, 5, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 8, 5, 64)          36928     \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 4, 2, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 512)               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 512)               262656    \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 11)                5643      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 324,631\n",
      "Trainable params: 324,625\n",
      "Non-trainable params: 6\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "inputs = tf.keras.Input(shape=(32, 20, 3))\n",
    "inputs2 = tf.keras.layers.BatchNormalization()(inputs)\n",
    "inputs3 = tf.keras.layers.Conv2D(32, (3, 3), padding='same', activation=\"relu\")(inputs2)\n",
    "inputs4 = tf.keras.layers.MaxPool2D(pool_size=(2,2))(inputs3)\n",
    "inputs5 = tf.keras.layers.Conv2D(64, (3, 3), padding='same', activation=\"relu\")(inputs4)\n",
    "inputs6 = tf.keras.layers.MaxPool2D(pool_size=(2,2))(inputs5)\n",
    "inputs7 = tf.keras.layers.Conv2D(64, (3, 3), padding='same', activation=\"relu\")(inputs6)\n",
    "inputs8 = tf.keras.layers.MaxPool2D(pool_size=(2,2))(inputs7)\n",
    "inputs9 = tf.keras.layers.Flatten()(inputs8)\n",
    "inputs10 = tf.keras.layers.Dense(512,activation=\"relu\")(inputs9)\n",
    "output = tf.keras.layers.Dense(11, activation='softmax')(inputs10)\n",
    "\n",
    "model = tf.keras.Model(inputs=inputs, outputs=output)\n",
    "\n",
    "model.compile(loss=keras.losses.categorical_crossentropy, optimizer=tf.keras.optimizers.Adadelta(learning_rate=1.0, rho=0.95), metrics = [\"accuracy\"])\n",
    "\n",
    "model.summary()\n",
    "\n",
    "#model = Sequential()\n",
    "#model.add(BatchNormalization(input_shape=(32,20,3)))\n",
    "#model.add(Conv2D(32, (3, 3), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(2,2)))\n",
    "#model.add(Conv2D(64, (3, 3), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(2,2)))\n",
    "#model.add(Conv2D(64, (3, 3), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(2,2)))\n",
    "#model.add(Flatten())\n",
    "#model.add(Dense(512,activation=\"relu\"))\n",
    "#model.add(Dense(11, activation = \"softmax\"))\n",
    "#model.summary()\n",
    "\n",
    "#model.compile(loss=keras.losses.categorical_crossentropy, optimizer=tf.keras.optimizers.Adadelta(learning_rate=1.0, rho=0.95), metrics = [\"accuracy\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training\n",
    "The input pictures are randomly scattered for brightness, pixel shift variations and rotation angle. This is implemented with a ImageDataGenerator.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n",
      "332/332 [==============================] - 11s 15ms/step - loss: 2.0455 - accuracy: 0.3804\n",
      "Epoch 2/500\n",
      "332/332 [==============================] - 5s 16ms/step - loss: 1.4926 - accuracy: 0.5147\n",
      "Epoch 3/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 1.0299 - accuracy: 0.6702\n",
      "Epoch 4/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.8239 - accuracy: 0.7562\n",
      "Epoch 5/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.6912 - accuracy: 0.7819\n",
      "Epoch 6/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.6195 - accuracy: 0.8000\n",
      "Epoch 7/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.5674 - accuracy: 0.8362\n",
      "Epoch 8/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.5164 - accuracy: 0.8506\n",
      "Epoch 9/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.4927 - accuracy: 0.8445\n",
      "Epoch 10/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.4844 - accuracy: 0.8483\n",
      "Epoch 11/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.4369 - accuracy: 0.8672\n",
      "Epoch 12/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.3762 - accuracy: 0.8868\n",
      "Epoch 13/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.3441 - accuracy: 0.8898\n",
      "Epoch 14/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.3545 - accuracy: 0.8928\n",
      "Epoch 15/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.3235 - accuracy: 0.9072\n",
      "Epoch 16/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.3520 - accuracy: 0.8951\n",
      "Epoch 17/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2809 - accuracy: 0.9177\n",
      "Epoch 18/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.3064 - accuracy: 0.9011\n",
      "Epoch 19/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2516 - accuracy: 0.9230\n",
      "Epoch 20/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2509 - accuracy: 0.9275\n",
      "Epoch 21/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2934 - accuracy: 0.9208\n",
      "Epoch 22/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.2500 - accuracy: 0.9260\n",
      "Epoch 23/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2488 - accuracy: 0.9245\n",
      "Epoch 24/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2554 - accuracy: 0.9291\n",
      "Epoch 25/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2187 - accuracy: 0.9389\n",
      "Epoch 26/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2409 - accuracy: 0.9283\n",
      "Epoch 27/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2609 - accuracy: 0.9275\n",
      "Epoch 28/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2461 - accuracy: 0.9268\n",
      "Epoch 29/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2340 - accuracy: 0.9343\n",
      "Epoch 30/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2284 - accuracy: 0.9351\n",
      "Epoch 31/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2271 - accuracy: 0.9268\n",
      "Epoch 32/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1944 - accuracy: 0.9426\n",
      "Epoch 33/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1817 - accuracy: 0.9487\n",
      "Epoch 34/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2357 - accuracy: 0.9328\n",
      "Epoch 35/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.1859 - accuracy: 0.9442\n",
      "Epoch 36/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1933 - accuracy: 0.9389\n",
      "Epoch 37/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2287 - accuracy: 0.9389\n",
      "Epoch 38/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1632 - accuracy: 0.9494\n",
      "Epoch 39/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1636 - accuracy: 0.9464\n",
      "Epoch 40/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.2398 - accuracy: 0.9381\n",
      "Epoch 41/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1524 - accuracy: 0.9570\n",
      "Epoch 42/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1913 - accuracy: 0.9464\n",
      "Epoch 43/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1955 - accuracy: 0.9426\n",
      "Epoch 44/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1585 - accuracy: 0.9540\n",
      "Epoch 45/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1867 - accuracy: 0.9517\n",
      "Epoch 46/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.2129 - accuracy: 0.9434\n",
      "Epoch 47/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1759 - accuracy: 0.9494\n",
      "Epoch 48/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1801 - accuracy: 0.9502\n",
      "Epoch 49/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1523 - accuracy: 0.9555\n",
      "Epoch 50/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1840 - accuracy: 0.9547\n",
      "Epoch 51/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1629 - accuracy: 0.9577\n",
      "Epoch 52/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1799 - accuracy: 0.9540\n",
      "Epoch 53/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1425 - accuracy: 0.9592\n",
      "Epoch 54/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1426 - accuracy: 0.9585\n",
      "Epoch 55/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1171 - accuracy: 0.9683\n",
      "Epoch 56/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1278 - accuracy: 0.9585\n",
      "Epoch 57/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1628 - accuracy: 0.9570\n",
      "Epoch 58/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1147 - accuracy: 0.9592\n",
      "Epoch 59/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1507 - accuracy: 0.9608\n",
      "Epoch 60/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1561 - accuracy: 0.9547\n",
      "Epoch 61/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1837 - accuracy: 0.9608\n",
      "Epoch 62/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1730 - accuracy: 0.9509\n",
      "Epoch 63/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1267 - accuracy: 0.9645\n",
      "Epoch 64/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1105 - accuracy: 0.9638\n",
      "Epoch 65/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1520 - accuracy: 0.9562\n",
      "Epoch 66/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1687 - accuracy: 0.9623\n",
      "Epoch 67/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1368 - accuracy: 0.9608\n",
      "Epoch 68/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1305 - accuracy: 0.9660\n",
      "Epoch 69/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1354 - accuracy: 0.9577\n",
      "Epoch 70/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1443 - accuracy: 0.9668\n",
      "Epoch 71/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1542 - accuracy: 0.9653\n",
      "Epoch 72/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1191 - accuracy: 0.9645\n",
      "Epoch 73/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1036 - accuracy: 0.9660\n",
      "Epoch 74/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1372 - accuracy: 0.9638\n",
      "Epoch 75/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1348 - accuracy: 0.9638\n",
      "Epoch 76/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1654 - accuracy: 0.9479\n",
      "Epoch 77/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1386 - accuracy: 0.9600\n",
      "Epoch 78/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0970 - accuracy: 0.9751\n",
      "Epoch 79/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1594 - accuracy: 0.9615\n",
      "Epoch 80/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1220 - accuracy: 0.9615\n",
      "Epoch 81/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1358 - accuracy: 0.9645\n",
      "Epoch 82/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1200 - accuracy: 0.9645\n",
      "Epoch 83/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0875 - accuracy: 0.9789\n",
      "Epoch 84/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1483 - accuracy: 0.9623\n",
      "Epoch 85/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0953 - accuracy: 0.9691\n",
      "Epoch 86/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1356 - accuracy: 0.9570\n",
      "Epoch 87/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1188 - accuracy: 0.9638\n",
      "Epoch 88/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1231 - accuracy: 0.9683\n",
      "Epoch 89/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.1598 - accuracy: 0.9638\n",
      "Epoch 90/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0861 - accuracy: 0.9706\n",
      "Epoch 91/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1318 - accuracy: 0.9653\n",
      "Epoch 92/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1454 - accuracy: 0.9653\n",
      "Epoch 93/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1046 - accuracy: 0.9653\n",
      "Epoch 94/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1586 - accuracy: 0.9668\n",
      "Epoch 95/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1353 - accuracy: 0.9562\n",
      "Epoch 96/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1364 - accuracy: 0.9691\n",
      "Epoch 97/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1126 - accuracy: 0.9691\n",
      "Epoch 98/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1140 - accuracy: 0.9721\n",
      "Epoch 99/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1173 - accuracy: 0.9706\n",
      "Epoch 100/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0635 - accuracy: 0.9804\n",
      "Epoch 101/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0839 - accuracy: 0.9728\n",
      "Epoch 102/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1755 - accuracy: 0.9608\n",
      "Epoch 103/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1597 - accuracy: 0.9615\n",
      "Epoch 104/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1237 - accuracy: 0.9645\n",
      "Epoch 105/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0857 - accuracy: 0.9706\n",
      "Epoch 106/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0992 - accuracy: 0.9592\n",
      "Epoch 107/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1266 - accuracy: 0.9691\n",
      "Epoch 108/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1164 - accuracy: 0.9751\n",
      "Epoch 109/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0913 - accuracy: 0.9698\n",
      "Epoch 110/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1114 - accuracy: 0.9743\n",
      "Epoch 111/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0705 - accuracy: 0.9758\n",
      "Epoch 112/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1571 - accuracy: 0.9660\n",
      "Epoch 113/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0948 - accuracy: 0.9758\n",
      "Epoch 114/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1088 - accuracy: 0.9713\n",
      "Epoch 115/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1367 - accuracy: 0.9660\n",
      "Epoch 116/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0761 - accuracy: 0.9774\n",
      "Epoch 117/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1234 - accuracy: 0.9675\n",
      "Epoch 118/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1024 - accuracy: 0.9698\n",
      "Epoch 119/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1231 - accuracy: 0.9653\n",
      "Epoch 120/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0814 - accuracy: 0.9781\n",
      "Epoch 121/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1168 - accuracy: 0.9698\n",
      "Epoch 122/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0984 - accuracy: 0.9728\n",
      "Epoch 123/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0929 - accuracy: 0.9751\n",
      "Epoch 124/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0910 - accuracy: 0.9751\n",
      "Epoch 125/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1080 - accuracy: 0.9683\n",
      "Epoch 126/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0739 - accuracy: 0.9766\n",
      "Epoch 127/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0727 - accuracy: 0.9796\n",
      "Epoch 128/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0920 - accuracy: 0.9789\n",
      "Epoch 129/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0810 - accuracy: 0.9713\n",
      "Epoch 130/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0941 - accuracy: 0.9781\n",
      "Epoch 131/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1082 - accuracy: 0.9721\n",
      "Epoch 132/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0850 - accuracy: 0.9811\n",
      "Epoch 133/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1090 - accuracy: 0.9758\n",
      "Epoch 134/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0909 - accuracy: 0.9728\n",
      "Epoch 135/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0607 - accuracy: 0.9804\n",
      "Epoch 136/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0865 - accuracy: 0.9736\n",
      "Epoch 137/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1042 - accuracy: 0.9774\n",
      "Epoch 138/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1377 - accuracy: 0.9660\n",
      "Epoch 139/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0868 - accuracy: 0.9796\n",
      "Epoch 140/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0944 - accuracy: 0.9691\n",
      "Epoch 141/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1215 - accuracy: 0.9691\n",
      "Epoch 142/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1017 - accuracy: 0.9675\n",
      "Epoch 143/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0920 - accuracy: 0.9698\n",
      "Epoch 144/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0487 - accuracy: 0.9849\n",
      "Epoch 145/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1277 - accuracy: 0.9668\n",
      "Epoch 146/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0951 - accuracy: 0.9713\n",
      "Epoch 147/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0857 - accuracy: 0.9766\n",
      "Epoch 148/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0788 - accuracy: 0.9766\n",
      "Epoch 149/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0591 - accuracy: 0.9849\n",
      "Epoch 150/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0956 - accuracy: 0.9781\n",
      "Epoch 151/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0981 - accuracy: 0.9743\n",
      "Epoch 152/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0805 - accuracy: 0.9743\n",
      "Epoch 153/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0900 - accuracy: 0.9774\n",
      "Epoch 154/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0770 - accuracy: 0.9811\n",
      "Epoch 155/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1164 - accuracy: 0.9698\n",
      "Epoch 156/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0781 - accuracy: 0.9774\n",
      "Epoch 157/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0893 - accuracy: 0.9736\n",
      "Epoch 158/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0717 - accuracy: 0.9751\n",
      "Epoch 159/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0657 - accuracy: 0.9811\n",
      "Epoch 160/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.1010 - accuracy: 0.9743\n",
      "Epoch 161/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0781 - accuracy: 0.9796\n",
      "Epoch 162/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1242 - accuracy: 0.9713\n",
      "Epoch 163/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0755 - accuracy: 0.9766\n",
      "Epoch 164/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0553 - accuracy: 0.9849\n",
      "Epoch 165/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0716 - accuracy: 0.9758\n",
      "Epoch 166/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0779 - accuracy: 0.9804\n",
      "Epoch 167/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0818 - accuracy: 0.9766\n",
      "Epoch 168/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0968 - accuracy: 0.9781\n",
      "Epoch 169/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0888 - accuracy: 0.9758\n",
      "Epoch 170/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0700 - accuracy: 0.9781\n",
      "Epoch 171/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0886 - accuracy: 0.9811\n",
      "Epoch 172/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0625 - accuracy: 0.9819\n",
      "Epoch 173/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0773 - accuracy: 0.9811\n",
      "Epoch 174/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0863 - accuracy: 0.9766\n",
      "Epoch 175/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0895 - accuracy: 0.9751\n",
      "Epoch 176/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0915 - accuracy: 0.9743\n",
      "Epoch 177/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0695 - accuracy: 0.9826\n",
      "Epoch 178/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0743 - accuracy: 0.9728\n",
      "Epoch 179/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1097 - accuracy: 0.9728\n",
      "Epoch 180/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0738 - accuracy: 0.9819\n",
      "Epoch 181/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0935 - accuracy: 0.9804\n",
      "Epoch 182/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0691 - accuracy: 0.9819\n",
      "Epoch 183/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0980 - accuracy: 0.9781\n",
      "Epoch 184/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0862 - accuracy: 0.9796\n",
      "Epoch 185/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1049 - accuracy: 0.9743\n",
      "Epoch 186/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0986 - accuracy: 0.9826\n",
      "Epoch 187/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0695 - accuracy: 0.9811\n",
      "Epoch 188/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0670 - accuracy: 0.9796\n",
      "Epoch 189/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0800 - accuracy: 0.9789\n",
      "Epoch 190/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0903 - accuracy: 0.9819\n",
      "Epoch 191/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1051 - accuracy: 0.9706\n",
      "Epoch 192/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0591 - accuracy: 0.9811\n",
      "Epoch 193/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1078 - accuracy: 0.9751\n",
      "Epoch 194/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0853 - accuracy: 0.9789\n",
      "Epoch 195/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1070 - accuracy: 0.9743\n",
      "Epoch 196/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0629 - accuracy: 0.9842\n",
      "Epoch 197/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0855 - accuracy: 0.9849\n",
      "Epoch 198/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0714 - accuracy: 0.9774\n",
      "Epoch 199/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0552 - accuracy: 0.9826\n",
      "Epoch 200/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0814 - accuracy: 0.9758\n",
      "Epoch 201/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.1079 - accuracy: 0.9675\n",
      "Epoch 202/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0769 - accuracy: 0.9811\n",
      "Epoch 203/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0519 - accuracy: 0.9834\n",
      "Epoch 204/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0664 - accuracy: 0.9766\n",
      "Epoch 205/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0502 - accuracy: 0.9864\n",
      "Epoch 206/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1023 - accuracy: 0.9789\n",
      "Epoch 207/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0610 - accuracy: 0.9842\n",
      "Epoch 208/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0647 - accuracy: 0.9849\n",
      "Epoch 209/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0906 - accuracy: 0.9766\n",
      "Epoch 210/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0775 - accuracy: 0.9796\n",
      "Epoch 211/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0733 - accuracy: 0.9774\n",
      "Epoch 212/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0552 - accuracy: 0.9857\n",
      "Epoch 213/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1012 - accuracy: 0.9789\n",
      "Epoch 214/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1090 - accuracy: 0.9751\n",
      "Epoch 215/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0770 - accuracy: 0.9781\n",
      "Epoch 216/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0825 - accuracy: 0.9819\n",
      "Epoch 217/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0945 - accuracy: 0.9728\n",
      "Epoch 218/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0848 - accuracy: 0.9758\n",
      "Epoch 219/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0878 - accuracy: 0.9781\n",
      "Epoch 220/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0551 - accuracy: 0.9864\n",
      "Epoch 221/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0835 - accuracy: 0.9842\n",
      "Epoch 222/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0521 - accuracy: 0.9887\n",
      "Epoch 223/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0858 - accuracy: 0.9781\n",
      "Epoch 224/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0914 - accuracy: 0.9804\n",
      "Epoch 225/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.1140 - accuracy: 0.9706\n",
      "Epoch 226/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0839 - accuracy: 0.9766\n",
      "Epoch 227/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0751 - accuracy: 0.9819\n",
      "Epoch 228/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0543 - accuracy: 0.9811\n",
      "Epoch 229/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0693 - accuracy: 0.9826\n",
      "Epoch 230/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1014 - accuracy: 0.9743\n",
      "Epoch 231/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0778 - accuracy: 0.9857\n",
      "Epoch 232/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0792 - accuracy: 0.9819\n",
      "Epoch 233/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0752 - accuracy: 0.9834\n",
      "Epoch 234/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0614 - accuracy: 0.9849\n",
      "Epoch 235/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0776 - accuracy: 0.9751\n",
      "Epoch 236/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0682 - accuracy: 0.9834\n",
      "Epoch 237/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0577 - accuracy: 0.9819\n",
      "Epoch 238/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0704 - accuracy: 0.9834\n",
      "Epoch 239/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0575 - accuracy: 0.9834\n",
      "Epoch 240/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0498 - accuracy: 0.9872\n",
      "Epoch 241/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0681 - accuracy: 0.9826\n",
      "Epoch 242/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0636 - accuracy: 0.9864\n",
      "Epoch 243/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0482 - accuracy: 0.9857\n",
      "Epoch 244/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0420 - accuracy: 0.9872\n",
      "Epoch 245/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0754 - accuracy: 0.9826\n",
      "Epoch 246/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0622 - accuracy: 0.9857\n",
      "Epoch 247/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0934 - accuracy: 0.9789\n",
      "Epoch 248/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0642 - accuracy: 0.9826\n",
      "Epoch 249/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0851 - accuracy: 0.9774\n",
      "Epoch 250/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.1020 - accuracy: 0.9774\n",
      "Epoch 251/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0617 - accuracy: 0.9857\n",
      "Epoch 252/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0734 - accuracy: 0.9804\n",
      "Epoch 253/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0837 - accuracy: 0.9758\n",
      "Epoch 254/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0951 - accuracy: 0.9721\n",
      "Epoch 255/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0775 - accuracy: 0.9811\n",
      "Epoch 256/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0526 - accuracy: 0.9811\n",
      "Epoch 257/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0730 - accuracy: 0.9811\n",
      "Epoch 258/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0808 - accuracy: 0.9766\n",
      "Epoch 259/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0646 - accuracy: 0.9796\n",
      "Epoch 260/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0913 - accuracy: 0.9781\n",
      "Epoch 261/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0865 - accuracy: 0.9811\n",
      "Epoch 262/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0668 - accuracy: 0.9826\n",
      "Epoch 263/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0728 - accuracy: 0.9781\n",
      "Epoch 264/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0495 - accuracy: 0.9849\n",
      "Epoch 265/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0839 - accuracy: 0.9842\n",
      "Epoch 266/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0572 - accuracy: 0.9842\n",
      "Epoch 267/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0632 - accuracy: 0.9834\n",
      "Epoch 268/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0432 - accuracy: 0.9879\n",
      "Epoch 269/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0721 - accuracy: 0.9758\n",
      "Epoch 270/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0747 - accuracy: 0.9864\n",
      "Epoch 271/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0595 - accuracy: 0.9834\n",
      "Epoch 272/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0713 - accuracy: 0.9826\n",
      "Epoch 273/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0765 - accuracy: 0.9849\n",
      "Epoch 274/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0532 - accuracy: 0.9811\n",
      "Epoch 275/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0444 - accuracy: 0.9894\n",
      "Epoch 276/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0769 - accuracy: 0.9743\n",
      "Epoch 277/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0811 - accuracy: 0.9751\n",
      "Epoch 278/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0725 - accuracy: 0.9796\n",
      "Epoch 279/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0684 - accuracy: 0.9781\n",
      "Epoch 280/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0656 - accuracy: 0.9826\n",
      "Epoch 281/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0615 - accuracy: 0.9811\n",
      "Epoch 282/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0895 - accuracy: 0.9804\n",
      "Epoch 283/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0649 - accuracy: 0.9834\n",
      "Epoch 284/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0507 - accuracy: 0.9857\n",
      "Epoch 285/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0618 - accuracy: 0.9857\n",
      "Epoch 286/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0638 - accuracy: 0.9796\n",
      "Epoch 287/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0569 - accuracy: 0.9842\n",
      "Epoch 288/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0707 - accuracy: 0.9781\n",
      "Epoch 289/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0605 - accuracy: 0.9902\n",
      "Epoch 290/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0563 - accuracy: 0.9849\n",
      "Epoch 291/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0971 - accuracy: 0.9766\n",
      "Epoch 292/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0633 - accuracy: 0.9849\n",
      "Epoch 293/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0330 - accuracy: 0.9872\n",
      "Epoch 294/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0825 - accuracy: 0.9826\n",
      "Epoch 295/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0596 - accuracy: 0.9864\n",
      "Epoch 296/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0706 - accuracy: 0.9781\n",
      "Epoch 297/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0768 - accuracy: 0.9811\n",
      "Epoch 298/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0621 - accuracy: 0.9849\n",
      "Epoch 299/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0645 - accuracy: 0.9789\n",
      "Epoch 300/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0830 - accuracy: 0.9804\n",
      "Epoch 301/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0555 - accuracy: 0.9849\n",
      "Epoch 302/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0642 - accuracy: 0.9842\n",
      "Epoch 303/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0583 - accuracy: 0.9796\n",
      "Epoch 304/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0799 - accuracy: 0.9842\n",
      "Epoch 305/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0537 - accuracy: 0.9842\n",
      "Epoch 306/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0273 - accuracy: 0.9909\n",
      "Epoch 307/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0786 - accuracy: 0.9879\n",
      "Epoch 308/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0547 - accuracy: 0.9796\n",
      "Epoch 309/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0615 - accuracy: 0.9894\n",
      "Epoch 310/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0448 - accuracy: 0.9894\n",
      "Epoch 311/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0683 - accuracy: 0.9857\n",
      "Epoch 312/500\n",
      "332/332 [==============================] - 5s 16ms/step - loss: 0.0706 - accuracy: 0.9842\n",
      "Epoch 313/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0756 - accuracy: 0.9819\n",
      "Epoch 314/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0583 - accuracy: 0.9804\n",
      "Epoch 315/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0658 - accuracy: 0.9819\n",
      "Epoch 316/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0651 - accuracy: 0.9857\n",
      "Epoch 317/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0428 - accuracy: 0.9872\n",
      "Epoch 318/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0766 - accuracy: 0.9789\n",
      "Epoch 319/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0543 - accuracy: 0.9864\n",
      "Epoch 320/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0701 - accuracy: 0.9789\n",
      "Epoch 321/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0706 - accuracy: 0.9872\n",
      "Epoch 322/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0548 - accuracy: 0.9849\n",
      "Epoch 323/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0531 - accuracy: 0.9842\n",
      "Epoch 324/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0833 - accuracy: 0.9826\n",
      "Epoch 325/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0585 - accuracy: 0.9857\n",
      "Epoch 326/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0339 - accuracy: 0.9902\n",
      "Epoch 327/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0656 - accuracy: 0.9796\n",
      "Epoch 328/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0620 - accuracy: 0.9819\n",
      "Epoch 329/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0447 - accuracy: 0.9872\n",
      "Epoch 330/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0474 - accuracy: 0.9887\n",
      "Epoch 331/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0622 - accuracy: 0.9834\n",
      "Epoch 332/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0372 - accuracy: 0.9857\n",
      "Epoch 333/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0557 - accuracy: 0.9849\n",
      "Epoch 334/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0544 - accuracy: 0.9902\n",
      "Epoch 335/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0440 - accuracy: 0.9879\n",
      "Epoch 336/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0993 - accuracy: 0.9819\n",
      "Epoch 337/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0876 - accuracy: 0.9804\n",
      "Epoch 338/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0629 - accuracy: 0.9842\n",
      "Epoch 339/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0440 - accuracy: 0.9872\n",
      "Epoch 340/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0432 - accuracy: 0.9887\n",
      "Epoch 341/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0457 - accuracy: 0.9864\n",
      "Epoch 342/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0554 - accuracy: 0.9834\n",
      "Epoch 343/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0609 - accuracy: 0.9834\n",
      "Epoch 344/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0573 - accuracy: 0.9857\n",
      "Epoch 345/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0474 - accuracy: 0.9879\n",
      "Epoch 346/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0652 - accuracy: 0.9857\n",
      "Epoch 347/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0560 - accuracy: 0.9826\n",
      "Epoch 348/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0641 - accuracy: 0.9781\n",
      "Epoch 349/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0487 - accuracy: 0.9849\n",
      "Epoch 350/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0731 - accuracy: 0.9857\n",
      "Epoch 351/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0628 - accuracy: 0.9811\n",
      "Epoch 352/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0580 - accuracy: 0.9834\n",
      "Epoch 353/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0364 - accuracy: 0.9947\n",
      "Epoch 354/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0591 - accuracy: 0.9864\n",
      "Epoch 355/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0775 - accuracy: 0.9842\n",
      "Epoch 356/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0530 - accuracy: 0.9879\n",
      "Epoch 357/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0400 - accuracy: 0.9887\n",
      "Epoch 358/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0417 - accuracy: 0.9849\n",
      "Epoch 359/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0668 - accuracy: 0.9819\n",
      "Epoch 360/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0490 - accuracy: 0.9909\n",
      "Epoch 361/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0628 - accuracy: 0.9842\n",
      "Epoch 362/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0401 - accuracy: 0.9842\n",
      "Epoch 363/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0658 - accuracy: 0.9834\n",
      "Epoch 364/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0557 - accuracy: 0.9879\n",
      "Epoch 365/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0571 - accuracy: 0.9849\n",
      "Epoch 366/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0454 - accuracy: 0.9857\n",
      "Epoch 367/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0623 - accuracy: 0.9864\n",
      "Epoch 368/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0439 - accuracy: 0.9894\n",
      "Epoch 369/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0898 - accuracy: 0.9796\n",
      "Epoch 370/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0561 - accuracy: 0.9857\n",
      "Epoch 371/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0521 - accuracy: 0.9909\n",
      "Epoch 372/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0581 - accuracy: 0.9849\n",
      "Epoch 373/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0884 - accuracy: 0.9819\n",
      "Epoch 374/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0740 - accuracy: 0.9842\n",
      "Epoch 375/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0764 - accuracy: 0.9826\n",
      "Epoch 376/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0596 - accuracy: 0.9849\n",
      "Epoch 377/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0411 - accuracy: 0.9872\n",
      "Epoch 378/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0722 - accuracy: 0.9842\n",
      "Epoch 379/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0949 - accuracy: 0.9789\n",
      "Epoch 380/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0402 - accuracy: 0.9887\n",
      "Epoch 381/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0570 - accuracy: 0.9864\n",
      "Epoch 382/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0501 - accuracy: 0.9872\n",
      "Epoch 383/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0610 - accuracy: 0.9849\n",
      "Epoch 384/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0451 - accuracy: 0.9894\n",
      "Epoch 385/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0534 - accuracy: 0.9879\n",
      "Epoch 386/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0689 - accuracy: 0.9857\n",
      "Epoch 387/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0760 - accuracy: 0.9842\n",
      "Epoch 388/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0451 - accuracy: 0.9887\n",
      "Epoch 389/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0555 - accuracy: 0.9834\n",
      "Epoch 390/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0385 - accuracy: 0.9849\n",
      "Epoch 391/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0230 - accuracy: 0.9932\n",
      "Epoch 392/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0653 - accuracy: 0.9879\n",
      "Epoch 393/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0556 - accuracy: 0.9826\n",
      "Epoch 394/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0522 - accuracy: 0.9887\n",
      "Epoch 395/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0326 - accuracy: 0.9879\n",
      "Epoch 396/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0484 - accuracy: 0.9887\n",
      "Epoch 397/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0974 - accuracy: 0.9758\n",
      "Epoch 398/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0500 - accuracy: 0.9842\n",
      "Epoch 399/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0584 - accuracy: 0.9864\n",
      "Epoch 400/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0655 - accuracy: 0.9819\n",
      "Epoch 401/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0428 - accuracy: 0.9902\n",
      "Epoch 402/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0288 - accuracy: 0.9917\n",
      "Epoch 403/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0679 - accuracy: 0.9872\n",
      "Epoch 404/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0901 - accuracy: 0.9834\n",
      "Epoch 405/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0505 - accuracy: 0.9872\n",
      "Epoch 406/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0425 - accuracy: 0.9925\n",
      "Epoch 407/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0808 - accuracy: 0.9849\n",
      "Epoch 408/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0647 - accuracy: 0.9819\n",
      "Epoch 409/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0341 - accuracy: 0.9902\n",
      "Epoch 410/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0630 - accuracy: 0.9857\n",
      "Epoch 411/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0763 - accuracy: 0.9834\n",
      "Epoch 412/500\n",
      "332/332 [==============================] - 6s 16ms/step - loss: 0.0543 - accuracy: 0.9864\n",
      "Epoch 413/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0473 - accuracy: 0.9879\n",
      "Epoch 414/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0486 - accuracy: 0.9849\n",
      "Epoch 415/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0523 - accuracy: 0.9842\n",
      "Epoch 416/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.1012 - accuracy: 0.9766\n",
      "Epoch 417/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0442 - accuracy: 0.9894\n",
      "Epoch 418/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0593 - accuracy: 0.9864\n",
      "Epoch 419/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0392 - accuracy: 0.9879\n",
      "Epoch 420/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0467 - accuracy: 0.9894\n",
      "Epoch 421/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0714 - accuracy: 0.9857\n",
      "Epoch 422/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0452 - accuracy: 0.9879\n",
      "Epoch 423/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0471 - accuracy: 0.9842\n",
      "Epoch 424/500\n",
      "332/332 [==============================] - 5s 15ms/step - loss: 0.0507 - accuracy: 0.9842\n",
      "Epoch 425/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0582 - accuracy: 0.9849\n",
      "Epoch 426/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0445 - accuracy: 0.9887\n",
      "Epoch 427/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0727 - accuracy: 0.9789\n",
      "Epoch 428/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0448 - accuracy: 0.9887\n",
      "Epoch 429/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0919 - accuracy: 0.9819\n",
      "Epoch 430/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0759 - accuracy: 0.9789\n",
      "Epoch 431/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0544 - accuracy: 0.9894\n",
      "Epoch 432/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0509 - accuracy: 0.9857\n",
      "Epoch 433/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0405 - accuracy: 0.9932\n",
      "Epoch 434/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0639 - accuracy: 0.9849\n",
      "Epoch 435/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0476 - accuracy: 0.9879\n",
      "Epoch 436/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0537 - accuracy: 0.9849\n",
      "Epoch 437/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0620 - accuracy: 0.9834\n",
      "Epoch 438/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0683 - accuracy: 0.9872\n",
      "Epoch 439/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0600 - accuracy: 0.9864\n",
      "Epoch 440/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0357 - accuracy: 0.9917\n",
      "Epoch 441/500\n",
      "332/332 [==============================] - 4s 13ms/step - loss: 0.0299 - accuracy: 0.9925\n",
      "Epoch 442/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0684 - accuracy: 0.9842\n",
      "Epoch 443/500\n",
      "332/332 [==============================] - 5s 14ms/step - loss: 0.0795 - accuracy: 0.9781\n",
      "Epoch 444/500\n",
      "332/332 [==============================] - 5s 13ms/step - loss: 0.0426 - accuracy: 0.9872\n",
      "Epoch 445/500\n",
      "332/332 [==============================] - 4s 12ms/step - loss: 0.0263 - accuracy: 0.9925\n",
      "Epoch 446/500\n",
      "332/332 [==============================] - 4s 10ms/step - loss: 0.0477 - accuracy: 0.9864\n",
      "Epoch 447/500\n",
      "332/332 [==============================] - 3s 9ms/step - loss: 0.0476 - accuracy: 0.9887\n",
      "Epoch 448/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0209 - accuracy: 0.9925\n",
      "Epoch 449/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0313 - accuracy: 0.9887\n",
      "Epoch 450/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0703 - accuracy: 0.9842\n",
      "Epoch 451/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0362 - accuracy: 0.9902\n",
      "Epoch 452/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0579 - accuracy: 0.9864\n",
      "Epoch 453/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0849 - accuracy: 0.9826\n",
      "Epoch 454/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0745 - accuracy: 0.9872\n",
      "Epoch 455/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0731 - accuracy: 0.9789\n",
      "Epoch 456/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0387 - accuracy: 0.9894\n",
      "Epoch 457/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0486 - accuracy: 0.9857\n",
      "Epoch 458/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0724 - accuracy: 0.9879\n",
      "Epoch 459/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0588 - accuracy: 0.9811\n",
      "Epoch 460/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0478 - accuracy: 0.9879\n",
      "Epoch 461/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0464 - accuracy: 0.9894\n",
      "Epoch 462/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0797 - accuracy: 0.9789\n",
      "Epoch 463/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0521 - accuracy: 0.9887\n",
      "Epoch 464/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0610 - accuracy: 0.9842\n",
      "Epoch 465/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0762 - accuracy: 0.9849\n",
      "Epoch 466/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0782 - accuracy: 0.9834\n",
      "Epoch 467/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0563 - accuracy: 0.9834\n",
      "Epoch 468/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0286 - accuracy: 0.9902\n",
      "Epoch 469/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0562 - accuracy: 0.9887\n",
      "Epoch 470/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0545 - accuracy: 0.9842\n",
      "Epoch 471/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0643 - accuracy: 0.9872\n",
      "Epoch 472/500\n",
      "332/332 [==============================] - 4s 11ms/step - loss: 0.0296 - accuracy: 0.9917\n",
      "Epoch 473/500\n",
      "332/332 [==============================] - 3s 10ms/step - loss: 0.0475 - accuracy: 0.9894\n",
      "Epoch 474/500\n",
      "332/332 [==============================] - 3s 9ms/step - loss: 0.0883 - accuracy: 0.9826\n",
      "Epoch 475/500\n",
      "332/332 [==============================] - 3s 9ms/step - loss: 0.0414 - accuracy: 0.9834\n",
      "Epoch 476/500\n",
      "332/332 [==============================] - 3s 9ms/step - loss: 0.0871 - accuracy: 0.9796\n",
      "Epoch 477/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0418 - accuracy: 0.9925\n",
      "Epoch 478/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0517 - accuracy: 0.9842\n",
      "Epoch 479/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0609 - accuracy: 0.9857\n",
      "Epoch 480/500\n",
      "332/332 [==============================] - 3s 9ms/step - loss: 0.0365 - accuracy: 0.9872\n",
      "Epoch 481/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0475 - accuracy: 0.9887\n",
      "Epoch 482/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0682 - accuracy: 0.9826\n",
      "Epoch 483/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0480 - accuracy: 0.9879\n",
      "Epoch 484/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0444 - accuracy: 0.9872\n",
      "Epoch 485/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0534 - accuracy: 0.9894\n",
      "Epoch 486/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0570 - accuracy: 0.9894\n",
      "Epoch 487/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0634 - accuracy: 0.9857\n",
      "Epoch 488/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0817 - accuracy: 0.9789\n",
      "Epoch 489/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0345 - accuracy: 0.9887\n",
      "Epoch 490/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0856 - accuracy: 0.9819\n",
      "Epoch 491/500\n",
      "332/332 [==============================] - 2s 7ms/step - loss: 0.0677 - accuracy: 0.9834\n",
      "Epoch 492/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0846 - accuracy: 0.9781\n",
      "Epoch 493/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0369 - accuracy: 0.9894\n",
      "Epoch 494/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0366 - accuracy: 0.9894\n",
      "Epoch 495/500\n",
      "332/332 [==============================] - 2s 7ms/step - loss: 0.0821 - accuracy: 0.9796\n",
      "Epoch 496/500\n",
      "332/332 [==============================] - 2s 7ms/step - loss: 0.0815 - accuracy: 0.9864\n",
      "Epoch 497/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0420 - accuracy: 0.9909\n",
      "Epoch 498/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0652 - accuracy: 0.9849\n",
      "Epoch 499/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0655 - accuracy: 0.9872\n",
      "Epoch 500/500\n",
      "332/332 [==============================] - 3s 8ms/step - loss: 0.0523 - accuracy: 0.9872\n"
     ]
    }
   ],
   "source": [
    "Batch_Size = 4\n",
    "Shift_Range = 1\n",
    "Brightness_Range = 0.3\n",
    "Rotation_Angle = 10\n",
    "ZoomRange = 0.4\n",
    "\n",
    "datagen = ImageDataGenerator(width_shift_range=[-Shift_Range,Shift_Range], \n",
    "                             height_shift_range=[-Shift_Range,Shift_Range],\n",
    "                             brightness_range=[1-Brightness_Range,1+Brightness_Range],\n",
    "                             zoom_range=[1-ZoomRange, 1+ZoomRange],\n",
    "                             rotation_range=Rotation_Angle)\n",
    "\n",
    "if (Training_Percentage > 0):\n",
    "    train_iterator = datagen.flow(x_data, y_data, batch_size=Batch_Size)\n",
    "    validation_iterator = datagen.flow(X_test, y_test, batch_size=Batch_Size)\n",
    "    history = model.fit(train_iterator, validation_data = validation_iterator, epochs = Epoch_Anz)\n",
    "else:\n",
    "    train_iterator = datagen.flow(x_data, y_data, batch_size=Batch_Size)\n",
    "    history = model.fit(train_iterator, epochs = Epoch_Anz)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Learing result\n",
    " \n",
    "* Visualization of the training and validation results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "loss_ges = np.append(loss_ges, history.history['loss'])\n",
    "plt.semilogy(history.history['loss'])\n",
    "\n",
    "if (Training_Percentage > 0):\n",
    "    val_loss_ges = np.append(val_loss_ges, history.history['val_loss'])\n",
    "    plt.semilogy(history.history['val_loss'])\n",
    "\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train','eval'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check the model by hand\n",
    "\n",
    "* The following code uses the trained model to check the deviation for each picture.\n",
    "* x-axis walks through each pixel, y-axis shows the result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Check each image for expected and deviation\n",
    "* setting the switch \"only_deviation = true\" will only print the images for which the classification and the CNN-result deviates\n",
    "\n",
    "The output contains the following information:\n",
    "\n",
    "| Filename      | Expected Category           | Predicted Category        |\n",
    "|------------- |:-----------------------------:|--------------|\n",
    "| ziffer_sortiert_resize_NaN/5\\Ziffer_4_0034.jpg | 4  | -1 |\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 163ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 41ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 59ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 62ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 58ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 63ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 44ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 79ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 47ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 8ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 5ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 7ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 89ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 39ms/step\n",
      "1/1 [==============================] - 0s 50ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 77ms/step\n",
      "1/1 [==============================] - 0s 53ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 37ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 33ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 59ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 21ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 38ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 10ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 8ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 27ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 29ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 8ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 12ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 28ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 30ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 9ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 19ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 7ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 35ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 23ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 14ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 24ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 22ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 15ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 26ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 32ms/step\n",
      "1/1 [==============================] - 0s 11ms/step\n",
      "1/1 [==============================] - 0s 18ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 20ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 17ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 13ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 25ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 16ms/step\n",
      "1/1 [==============================] - 0s 31ms/step\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": [
    "Input_dir='ziffer_sortiert_resize'\n",
    "res = []\n",
    "only_deviation = True\n",
    "show_wrong_image = True\n",
    "\n",
    "files = glob.glob(Input_dir + '/*.jpg')\n",
    "\n",
    "for aktfile in files:\n",
    "    base = os.path.basename(aktfile)\n",
    "    target = base[0:1]\n",
    "    if target == \"N\":\n",
    "        zw1 = -1\n",
    "    else:\n",
    "        zw1 = int(target)\n",
    "    expected_class = zw1\n",
    "    image_in = Image.open(aktfile)\n",
    "    test_image = np.array(image_in, dtype=\"float32\")\n",
    "    img = np.reshape(test_image,[1,32,20,3])\n",
    "    classes = np.argmax(model.predict(img), axis=-1)\n",
    "    classes = classes[0]\n",
    "    if classes == 10: \n",
    "        classes = -1\n",
    "    zw2 = classes\n",
    "    zw3 = zw2 - zw1\n",
    "    res.append(np.array([zw1, zw2, zw3]))\n",
    "    if only_deviation == True:\n",
    "        if str(classes) != str(expected_class):\n",
    "            print(aktfile + \" \" + str(expected_class) +  \" \" + str(classes))\n",
    "            if show_wrong_image == True:\n",
    "                display(image_in)\n",
    "    else:\n",
    "        print(aktfile + \" \" + aktsubdir +  \" \" + str(classes))\n",
    "        \n",
    "\n",
    "res = np.asarray(res)\n",
    "\n",
    "\n",
    "plt.plot(res[:,0])\n",
    "plt.plot(res[:,1])\n",
    "plt.title('Result')\n",
    "plt.ylabel('Digital Value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['real','model'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model\n",
    "\n",
    "* Save the model to the file with the \"h5\" file format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op while saving (showing 3 of 3). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmpqdi9b927\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmpqdi9b927\\assets\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1302896"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FileName = TFliteNamingAndVersion\n",
    "\n",
    "converter = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "tflite_model = converter.convert()\n",
    "open(FileName + \".tflite\", \"wb\").write(tflite_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as _jit_compiled_convolution_op, _jit_compiled_convolution_op, _jit_compiled_convolution_op while saving (showing 3 of 3). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmp8_w4al85\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmp8_w4al85\\assets\n",
      "C:\\Users\\Muell\\anaconda3\\envs\\py39-tf-opencv\\lib\\site-packages\\tensorflow\\lite\\python\\convert.py:766: UserWarning: Statistics for quantized inputs were expected, but not specified; continuing anyway.\n",
      "  warnings.warn(\"Statistics for quantized inputs were expected, but not \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dig1410s1q.tflite\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "336064"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pathlib import Path\n",
    "import tensorflow as tf\n",
    "\n",
    "FileName = TFliteNamingAndVersion + \"q\" + \".tflite\"\n",
    "\n",
    "def representative_dataset():\n",
    "    for n in range(x_data[0].size):\n",
    "      data = np.expand_dims(x_data[5], axis=0)\n",
    "      yield [data.astype(np.float32)]\n",
    "        \n",
    "converter2 = tf.lite.TFLiteConverter.from_keras_model(model)\n",
    "converter2.representative_dataset = representative_dataset\n",
    "converter2.optimizations = [tf.lite.Optimize.DEFAULT]\n",
    "converter2.representative_dataset = representative_dataset\n",
    "tflite_quant_model = converter2.convert()\n",
    "\n",
    "open(FileName, \"wb\").write(tflite_quant_model)\n",
    "print(FileName)\n",
    "Path(FileName).stat().st_size"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking the images shows, that this are border line images, which can be interpreted as a good digit or a faulty one."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}