Train_CNN_Analog-Readout_Version-Small3-long.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CNN Training\n",
    "\n",
    "Target of this code is to train a CNN network to extract the needle position of an analog needle device.\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 = \"ana1100s3_long\"   # Used for tflite Filename\n",
    "Training_Percentage = 0.2              # 0.0 = Use all Images for Training\n",
    "Epoch_Anz = 100\n",
    "\n",
    "##########################################################################\n",
    "\n",
    "\n",
    "\n",
    "import os\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import glob\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",
    "import tensorflow.keras.backend as K\n",
    "from tensorflow.keras.callbacks import History \n",
    "import math\n",
    "from PIL import Image \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",
    "* Picture size must be 32x32 with 3 color channels (RGB)\n",
    "* The filename contains the informations needed for training in the first 3 digits::\n",
    "* Typical filename: \n",
    "    * x.y-zzzz.jpg \n",
    "    * e.g. \"4.6_Lfd-1406_zeiger3_2019-06-02T050011\"\n",
    "\n",
    "|Place holder | Meaning                     | Usage        |\n",
    "|------------- |-----------------------------|--------------|\n",
    "| **x.y**          | readout value               | **to be learned** |\n",
    "| zzzz        | additional information              | not needed   |\n",
    "\n",
    "* The images are stored in the x_data[]\n",
    "* The expected output for each image in the corresponding y_data[]\n",
    "    * The periodic nature is reflected in a **sin/cos coding**, which allows to restore the angle/counter value with an arctan later on.\n",
    "\n",
    "* The last step is a shuffle (from sklearn.utils) as the filenames are on order due to the encoding of the expected analog readout in the filename "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5761, 32, 32, 3)\n",
      "(5761, 2)\n"
     ]
    }
   ],
   "source": [
    "Input_dir='data_resize_all'\n",
    "\n",
    "files = glob.glob(Input_dir + '/*.*')\n",
    "x_data = []\n",
    "y_data = []\n",
    "\n",
    "for aktfile in files:\n",
    "    test_image = Image.open(aktfile)\n",
    "    test_image = np.array(test_image, dtype=\"float32\")\n",
    "    test_image = np.reshape(test_image, (32,32,3))\n",
    "    base = os.path.basename(aktfile)\n",
    "    target_number = (float(base[0:3])) / 10\n",
    "    target_sin = math.sin(target_number * math.pi * 2)\n",
    "    target_cos = math.cos(target_number * math.pi * 2)\n",
    "\n",
    "    x_data.append(test_image)\n",
    "    zw = np.array([target_sin, target_cos])\n",
    "    y_data.append(zw)\n",
    "\n",
    "x_data = np.array(x_data)\n",
    "y_data = np.array(y_data)\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, 32, 3)\n",
    "* Shape of the output layer: (2) - sin and cos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " input_1 (InputLayer)        [(None, 32, 32, 3)]       0         \n",
      "                                                                 \n",
      " batch_normalization (BatchN  (None, 32, 32, 3)        12        \n",
      " ormalization)                                                   \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 32, 32, 32)        2432      \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 8, 8, 32)         0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 8, 8, 16)          12816     \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 4, 4, 16)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " conv2d_2 (Conv2D)           (None, 4, 4, 32)          4640      \n",
      "                                                                 \n",
      " max_pooling2d_2 (MaxPooling  (None, 2, 2, 32)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 128)               0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 64)                8256      \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 32)                2080      \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 2)                 66        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 30,302\n",
      "Trainable params: 30,296\n",
      "Non-trainable params: 6\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "inputs = tf.keras.Input(shape=(32, 32, 3))\n",
    "inputs2 = tf.keras.layers.BatchNormalization()(inputs)\n",
    "inputs3 = tf.keras.layers.Conv2D(32, (5, 5), padding='same', activation=\"relu\")(inputs2)\n",
    "inputs4 = tf.keras.layers.MaxPool2D(pool_size=(4,4))(inputs3)\n",
    "inputs5 = tf.keras.layers.Conv2D(16, (5, 5), padding='same', activation=\"relu\")(inputs4)\n",
    "inputs6 = tf.keras.layers.MaxPool2D(pool_size=(2,2))(inputs5)\n",
    "inputs7 = tf.keras.layers.Conv2D(32, (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(64,activation=\"relu\")(inputs9)\n",
    "inputs11 = tf.keras.layers.Dense(32,activation=\"relu\")(inputs10)\n",
    "output = tf.keras.layers.Dense(2)(inputs11)\n",
    "\n",
    "model = tf.keras.Model(inputs=inputs, outputs=output)\n",
    "\n",
    "model.compile(loss=keras.losses.mean_squared_error, 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=(32,32,3)))\n",
    "#model.add(Conv2D(32, (5, 5), input_shape=(32,32,3), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(4,4)))\n",
    "#model.add(Conv2D(16, (5, 5), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(4,4)))\n",
    "#model.add(Conv2D(32, (3, 3), padding='same', activation=\"relu\"))\n",
    "#model.add(MaxPool2D(pool_size=(2,2)))\n",
    "#model.add(Flatten())\n",
    "#model.add(Dense(64,activation=\"relu\"))\n",
    "#model.add(Dense(32,activation=\"relu\"))\n",
    "#model.add(Dense(2))\n",
    "\n",
    "#model.compile(loss=keras.losses.mean_squared_error, optimizer=tf.keras.optimizers.Adadelta(learning_rate=1.0, rho=0.95), metrics = [\"accuracy\"])\n",
    "\n",
    "#model.summary()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training\n",
    "The input pictures are randomly scattered for brightness and pixel shift variations. These is implemented with a ImageDataGenerator.\n",
    "\n",
    "\n",
    "The training is splitted into two steps:\n",
    "1. Variation of the brightness only\n",
    "2. Variation of brightness and Pixel Shift"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1: Brigthness scattering only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Muell\\AppData\\Local\\Temp\\ipykernel_18484\\65716224.py:11: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
      "  history = model.fit_generator(train_iterator, validation_data = validation_iterator, epochs = Epoch_Anz)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "721/721 [==============================] - 26s 29ms/step - loss: 0.1604 - accuracy: 0.8505 - val_loss: 0.0273 - val_accuracy: 0.9462\n",
      "Epoch 2/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 0.0214 - accuracy: 0.9696 - val_loss: 0.0130 - val_accuracy: 0.9853\n",
      "Epoch 3/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 0.0113 - accuracy: 0.9766 - val_loss: 0.0209 - val_accuracy: 0.9748\n",
      "Epoch 4/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 0.0070 - accuracy: 0.9804 - val_loss: 0.0051 - val_accuracy: 0.9801\n",
      "Epoch 5/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 0.0051 - accuracy: 0.9819 - val_loss: 0.0081 - val_accuracy: 0.9705\n",
      "Epoch 6/100\n",
      "721/721 [==============================] - 19s 26ms/step - loss: 0.0043 - accuracy: 0.9811 - val_loss: 0.0031 - val_accuracy: 0.9887\n",
      "Epoch 7/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 0.0035 - accuracy: 0.9821 - val_loss: 0.0039 - val_accuracy: 0.9844\n",
      "Epoch 8/100\n",
      "721/721 [==============================] - 19s 27ms/step - loss: 0.0030 - accuracy: 0.9856 - val_loss: 0.0022 - val_accuracy: 0.9809\n",
      "Epoch 9/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 0.0027 - accuracy: 0.9878 - val_loss: 0.0023 - val_accuracy: 0.9887\n",
      "Epoch 10/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 0.0023 - accuracy: 0.9877 - val_loss: 0.0015 - val_accuracy: 0.9931\n",
      "Epoch 11/100\n",
      "721/721 [==============================] - 18s 26ms/step - loss: 0.0020 - accuracy: 0.9873 - val_loss: 0.0018 - val_accuracy: 0.9879\n",
      "Epoch 12/100\n",
      "721/721 [==============================] - 23s 31ms/step - loss: 0.0019 - accuracy: 0.9882 - val_loss: 0.0021 - val_accuracy: 0.9974\n",
      "Epoch 13/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 0.0017 - accuracy: 0.9911 - val_loss: 0.0015 - val_accuracy: 0.9896\n",
      "Epoch 14/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 0.0016 - accuracy: 0.9889 - val_loss: 0.0015 - val_accuracy: 0.9905\n",
      "Epoch 15/100\n",
      "721/721 [==============================] - 22s 30ms/step - loss: 0.0015 - accuracy: 0.9898 - val_loss: 0.0018 - val_accuracy: 0.9870\n",
      "Epoch 16/100\n",
      "721/721 [==============================] - 20s 27ms/step - loss: 0.0014 - accuracy: 0.9908 - val_loss: 0.0011 - val_accuracy: 0.9957\n",
      "Epoch 17/100\n",
      "721/721 [==============================] - 23s 31ms/step - loss: 0.0013 - accuracy: 0.9906 - val_loss: 9.5398e-04 - val_accuracy: 0.9913\n",
      "Epoch 18/100\n",
      "721/721 [==============================] - 22s 30ms/step - loss: 0.0013 - accuracy: 0.9918 - val_loss: 8.0783e-04 - val_accuracy: 0.9931\n",
      "Epoch 19/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 0.0012 - accuracy: 0.9924 - val_loss: 7.4184e-04 - val_accuracy: 0.9931\n",
      "Epoch 20/100\n",
      "721/721 [==============================] - 21s 29ms/step - loss: 0.0011 - accuracy: 0.9924 - val_loss: 9.7610e-04 - val_accuracy: 0.9913\n",
      "Epoch 21/100\n",
      "721/721 [==============================] - 22s 30ms/step - loss: 0.0011 - accuracy: 0.9934 - val_loss: 0.0018 - val_accuracy: 0.9801\n",
      "Epoch 22/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 9.9148e-04 - accuracy: 0.9925 - val_loss: 0.0011 - val_accuracy: 0.9948\n",
      "Epoch 23/100\n",
      "721/721 [==============================] - 21s 30ms/step - loss: 9.5962e-04 - accuracy: 0.9929 - val_loss: 8.5982e-04 - val_accuracy: 0.9948\n",
      "Epoch 24/100\n",
      "721/721 [==============================] - 21s 29ms/step - loss: 8.8970e-04 - accuracy: 0.9920 - val_loss: 6.4329e-04 - val_accuracy: 0.9913\n",
      "Epoch 25/100\n",
      "721/721 [==============================] - 22s 30ms/step - loss: 8.7264e-04 - accuracy: 0.9936 - val_loss: 6.3720e-04 - val_accuracy: 0.9957\n",
      "Epoch 26/100\n",
      "721/721 [==============================] - 18s 25ms/step - loss: 8.3209e-04 - accuracy: 0.9938 - val_loss: 7.6007e-04 - val_accuracy: 0.9939\n",
      "Epoch 27/100\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 8.2252e-04 - accuracy: 0.9951 - val_loss: 8.6940e-04 - val_accuracy: 0.9887\n",
      "Epoch 28/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 7.7676e-04 - accuracy: 0.9927 - val_loss: 6.9892e-04 - val_accuracy: 0.9957\n",
      "Epoch 29/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 7.9428e-04 - accuracy: 0.9953 - val_loss: 8.4166e-04 - val_accuracy: 0.9922\n",
      "Epoch 30/100\n",
      "721/721 [==============================] - 23s 33ms/step - loss: 7.2512e-04 - accuracy: 0.9950 - val_loss: 5.0688e-04 - val_accuracy: 0.9948\n",
      "Epoch 31/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 7.0735e-04 - accuracy: 0.9938 - val_loss: 5.9265e-04 - val_accuracy: 0.9957\n",
      "Epoch 32/100\n",
      "721/721 [==============================] - 23s 33ms/step - loss: 6.9264e-04 - accuracy: 0.9950 - val_loss: 6.8995e-04 - val_accuracy: 0.9965\n",
      "Epoch 33/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 6.7945e-04 - accuracy: 0.9943 - val_loss: 6.7384e-04 - val_accuracy: 0.9974\n",
      "Epoch 34/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 6.4322e-04 - accuracy: 0.9955 - val_loss: 5.9146e-04 - val_accuracy: 0.9879\n",
      "Epoch 35/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 6.4478e-04 - accuracy: 0.9938 - val_loss: 4.5561e-04 - val_accuracy: 0.9957\n",
      "Epoch 36/100\n",
      "721/721 [==============================] - 21s 30ms/step - loss: 6.2462e-04 - accuracy: 0.9929 - val_loss: 9.0129e-04 - val_accuracy: 0.9948\n",
      "Epoch 37/100\n",
      "721/721 [==============================] - 20s 28ms/step - loss: 6.0936e-04 - accuracy: 0.9951 - val_loss: 4.2234e-04 - val_accuracy: 0.9983\n",
      "Epoch 38/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 5.8070e-04 - accuracy: 0.9950 - val_loss: 7.1800e-04 - val_accuracy: 0.9948\n",
      "Epoch 39/100\n",
      "721/721 [==============================] - 25s 35ms/step - loss: 5.8167e-04 - accuracy: 0.9946 - val_loss: 4.6308e-04 - val_accuracy: 0.9957\n",
      "Epoch 40/100\n",
      "721/721 [==============================] - 20s 28ms/step - loss: 5.9037e-04 - accuracy: 0.9941 - val_loss: 5.4183e-04 - val_accuracy: 0.9965\n",
      "Epoch 41/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 5.7181e-04 - accuracy: 0.9962 - val_loss: 5.0672e-04 - val_accuracy: 0.9965\n",
      "Epoch 42/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 5.3344e-04 - accuracy: 0.9953 - val_loss: 3.7904e-04 - val_accuracy: 0.9957\n",
      "Epoch 43/100\n",
      "721/721 [==============================] - 25s 34ms/step - loss: 5.2125e-04 - accuracy: 0.9948 - val_loss: 6.0728e-04 - val_accuracy: 0.9957\n",
      "Epoch 44/100\n",
      "721/721 [==============================] - 20s 28ms/step - loss: 5.1247e-04 - accuracy: 0.9953 - val_loss: 6.3746e-04 - val_accuracy: 0.9957\n",
      "Epoch 45/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 5.2557e-04 - accuracy: 0.9951 - val_loss: 5.4470e-04 - val_accuracy: 0.9922\n",
      "Epoch 46/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 4.7192e-04 - accuracy: 0.9965 - val_loss: 4.5432e-04 - val_accuracy: 0.9974\n",
      "Epoch 47/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 5.1541e-04 - accuracy: 0.9950 - val_loss: 4.2667e-04 - val_accuracy: 0.9991\n",
      "Epoch 48/100\n",
      "721/721 [==============================] - 26s 35ms/step - loss: 4.7916e-04 - accuracy: 0.9950 - val_loss: 5.2061e-04 - val_accuracy: 0.9957\n",
      "Epoch 49/100\n",
      "721/721 [==============================] - 25s 34ms/step - loss: 4.8709e-04 - accuracy: 0.9946 - val_loss: 3.7988e-04 - val_accuracy: 0.9965\n",
      "Epoch 50/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 4.5953e-04 - accuracy: 0.9969 - val_loss: 3.7561e-04 - val_accuracy: 0.9983\n",
      "Epoch 51/100\n",
      "721/721 [==============================] - 21s 29ms/step - loss: 4.5694e-04 - accuracy: 0.9951 - val_loss: 3.5444e-04 - val_accuracy: 0.9965\n",
      "Epoch 52/100\n",
      "721/721 [==============================] - 26s 35ms/step - loss: 4.3438e-04 - accuracy: 0.9970 - val_loss: 5.1338e-04 - val_accuracy: 0.9991\n",
      "Epoch 53/100\n",
      "721/721 [==============================] - 25s 35ms/step - loss: 4.2921e-04 - accuracy: 0.9969 - val_loss: 3.3614e-04 - val_accuracy: 0.9983\n",
      "Epoch 54/100\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 4.2520e-04 - accuracy: 0.9950 - val_loss: 3.3577e-04 - val_accuracy: 0.9991\n",
      "Epoch 55/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 26s 36ms/step - loss: 4.2422e-04 - accuracy: 0.9970 - val_loss: 3.8911e-04 - val_accuracy: 0.9974\n",
      "Epoch 56/100\n",
      "721/721 [==============================] - 25s 35ms/step - loss: 4.0279e-04 - accuracy: 0.9965 - val_loss: 3.3301e-04 - val_accuracy: 0.9957\n",
      "Epoch 57/100\n",
      "721/721 [==============================] - 26s 35ms/step - loss: 3.9636e-04 - accuracy: 0.9967 - val_loss: 3.9879e-04 - val_accuracy: 0.9931\n",
      "Epoch 58/100\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 4.0302e-04 - accuracy: 0.9976 - val_loss: 4.8103e-04 - val_accuracy: 0.9991\n",
      "Epoch 59/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 4.1826e-04 - accuracy: 0.9957 - val_loss: 3.4971e-04 - val_accuracy: 0.9983\n",
      "Epoch 60/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.9176e-04 - accuracy: 0.9953 - val_loss: 5.1571e-04 - val_accuracy: 0.9974\n",
      "Epoch 61/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.9061e-04 - accuracy: 0.9972 - val_loss: 4.3596e-04 - val_accuracy: 0.9965\n",
      "Epoch 62/100\n",
      "721/721 [==============================] - 25s 34ms/step - loss: 3.9701e-04 - accuracy: 0.9967 - val_loss: 3.4024e-04 - val_accuracy: 0.9974\n",
      "Epoch 63/100\n",
      "721/721 [==============================] - 19s 27ms/step - loss: 3.8652e-04 - accuracy: 0.9965 - val_loss: 3.5664e-04 - val_accuracy: 0.9957\n",
      "Epoch 64/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.8755e-04 - accuracy: 0.9965 - val_loss: 5.0110e-04 - val_accuracy: 0.9965\n",
      "Epoch 65/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.7981e-04 - accuracy: 0.9965 - val_loss: 2.7717e-04 - val_accuracy: 0.9957\n",
      "Epoch 66/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.5657e-04 - accuracy: 0.9972 - val_loss: 3.8067e-04 - val_accuracy: 0.9957\n",
      "Epoch 67/100\n",
      "721/721 [==============================] - 25s 34ms/step - loss: 3.6452e-04 - accuracy: 0.9958 - val_loss: 2.7767e-04 - val_accuracy: 0.9957\n",
      "Epoch 68/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.5706e-04 - accuracy: 0.9972 - val_loss: 3.3501e-04 - val_accuracy: 1.0000\n",
      "Epoch 69/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.5629e-04 - accuracy: 0.9965 - val_loss: 3.6178e-04 - val_accuracy: 0.9983\n",
      "Epoch 70/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.5081e-04 - accuracy: 0.9969 - val_loss: 4.5910e-04 - val_accuracy: 0.9991\n",
      "Epoch 71/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.5521e-04 - accuracy: 0.9974 - val_loss: 2.8712e-04 - val_accuracy: 0.9983\n",
      "Epoch 72/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.4892e-04 - accuracy: 0.9979 - val_loss: 3.5521e-04 - val_accuracy: 0.9974\n",
      "Epoch 73/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.4340e-04 - accuracy: 0.9972 - val_loss: 3.3885e-04 - val_accuracy: 0.9974\n",
      "Epoch 74/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 3.3536e-04 - accuracy: 0.9981 - val_loss: 4.0415e-04 - val_accuracy: 0.9974\n",
      "Epoch 75/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 3.2647e-04 - accuracy: 0.9977 - val_loss: 4.2656e-04 - val_accuracy: 0.9913\n",
      "Epoch 76/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.5388e-04 - accuracy: 0.9964 - val_loss: 3.1536e-04 - val_accuracy: 1.0000\n",
      "Epoch 77/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.3523e-04 - accuracy: 0.9964 - val_loss: 2.4471e-04 - val_accuracy: 0.9991\n",
      "Epoch 78/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.2731e-04 - accuracy: 0.9976 - val_loss: 2.5185e-04 - val_accuracy: 0.9957\n",
      "Epoch 79/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.1653e-04 - accuracy: 0.9970 - val_loss: 2.3425e-04 - val_accuracy: 0.9983\n",
      "Epoch 80/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.1507e-04 - accuracy: 0.9972 - val_loss: 3.8454e-04 - val_accuracy: 0.9974\n",
      "Epoch 81/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.1934e-04 - accuracy: 0.9972 - val_loss: 2.6066e-04 - val_accuracy: 1.0000\n",
      "Epoch 82/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 3.1680e-04 - accuracy: 0.9969 - val_loss: 2.5898e-04 - val_accuracy: 0.9983\n",
      "Epoch 83/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.0594e-04 - accuracy: 0.9970 - val_loss: 3.6634e-04 - val_accuracy: 0.9957\n",
      "Epoch 84/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 3.0715e-04 - accuracy: 0.9970 - val_loss: 2.5326e-04 - val_accuracy: 0.9991\n",
      "Epoch 85/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 3.1459e-04 - accuracy: 0.9974 - val_loss: 2.4943e-04 - val_accuracy: 0.9991\n",
      "Epoch 86/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.1081e-04 - accuracy: 0.9970 - val_loss: 2.1482e-04 - val_accuracy: 0.9991\n",
      "Epoch 87/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.0580e-04 - accuracy: 0.9983 - val_loss: 2.6715e-04 - val_accuracy: 0.9991\n",
      "Epoch 88/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 3.0042e-04 - accuracy: 0.9988 - val_loss: 2.3185e-04 - val_accuracy: 0.9974\n",
      "Epoch 89/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 3.0499e-04 - accuracy: 0.9969 - val_loss: 2.5034e-04 - val_accuracy: 0.9983\n",
      "Epoch 90/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.9780e-04 - accuracy: 0.9974 - val_loss: 3.2375e-04 - val_accuracy: 0.9983\n",
      "Epoch 91/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.8930e-04 - accuracy: 0.9970 - val_loss: 3.8068e-04 - val_accuracy: 0.9887\n",
      "Epoch 92/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.9233e-04 - accuracy: 0.9984 - val_loss: 2.7221e-04 - val_accuracy: 0.9991\n",
      "Epoch 93/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 2.9915e-04 - accuracy: 0.9983 - val_loss: 2.4827e-04 - val_accuracy: 0.9983\n",
      "Epoch 94/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 2.8241e-04 - accuracy: 0.9979 - val_loss: 2.4222e-04 - val_accuracy: 0.9974\n",
      "Epoch 95/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.8618e-04 - accuracy: 0.9977 - val_loss: 3.0037e-04 - val_accuracy: 0.9991\n",
      "Epoch 96/100\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 2.8993e-04 - accuracy: 0.9986 - val_loss: 2.4678e-04 - val_accuracy: 0.9991\n",
      "Epoch 97/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.7304e-04 - accuracy: 0.9970 - val_loss: 2.8464e-04 - val_accuracy: 1.0000\n",
      "Epoch 98/100\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 2.7855e-04 - accuracy: 0.9976 - val_loss: 2.8572e-04 - val_accuracy: 0.9983\n",
      "Epoch 99/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 2.8762e-04 - accuracy: 0.9972 - val_loss: 3.0135e-04 - val_accuracy: 0.9922\n",
      "Epoch 100/100\n",
      "721/721 [==============================] - 24s 34ms/step - loss: 2.7564e-04 - accuracy: 0.9979 - val_loss: 2.5015e-04 - val_accuracy: 0.9983\n"
     ]
    }
   ],
   "source": [
    "Batch_Size = 8\n",
    "Epoch_Anz = 100\n",
    "Shift_Range = 0\n",
    "Brightness_Range = 0.3\n",
    "\n",
    "datagen = ImageDataGenerator(width_shift_range=[-Shift_Range,Shift_Range], height_shift_range=[-Shift_Range,Shift_Range],brightness_range=[1-Brightness_Range,1+Brightness_Range])\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_generator(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_generator(train_iterator, epochs = Epoch_Anz)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1: Learing result\n",
    " \n",
    "* Visualization of the training and validation results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAABCyUlEQVR4nO3dd3iV9fn48fedk5O9NyEkYYcNEkAEEUUEwT1w7zqqVmutrdaOr/3ZWmundaIi1oFQUAFFERygKHuPAGFkkRCSkEXmST6/P56TkEASAiSccHK/risXOc8695NHz30+W4wxKKWUUs3xcHUASimlOjZNFEoppVqkiUIppVSLNFEopZRqkSYKpZRSLdJEoZRSqkWaKJRqQyIyU0SebeWx+0Xk4tO9jlLtTROFUkqpFmmiUEop1SJNFKrTcVb5PCEim0XkiIi8JSLRIvK5iJSIyFIRCW1w/BUisk1ECkXkWxHp12DfMBFZ7zxvNuBzzHtdJiIbnef+ICKDTzHme0UkVUQKRGSBiMQ6t4uI/FNEckWkWES2iMhA574pIrLdGVuWiPzylP5gqtPTRKE6q2uBiUAf4HLgc+A3QCTW/xePAIhIH2AW8HPnvkXAQhHxEhEv4BPgXSAM+J/zujjPHQbMAO4HwoHXgQUi4n0ygYrIRcBzwDSgC5AGfOjcfQkwznkfwc5j8p373gLuN8YEAgOBr0/mfZWqo4lCdVb/McYcNMZkAd8Bq4wxG4wxFcDHwDDncTcAnxljlhhjqoG/Ab7AecC5gB34lzGm2hgzF1jT4D3uA143xqwyxtQYY94BKp3nnYxbgBnGmPXGmErgKWC0iCQC1UAgkASIMWaHMSbbeV410F9Egowxh40x60/yfZUCNFGozutgg9/Lm3gd4Pw9FusbPADGmFogA+jq3JdlGs+smdbg9wTgcWe1U6GIFALdnOedjGNjKMUqNXQ1xnwNvAS8DOSKyHQRCXIeei0wBUgTkWUiMvok31cpQBOFUidyAOsDH7DaBLA+7LOAbKCrc1ud+Aa/ZwB/MsaENPjxM8bMOs0Y/LGqsrIAjDEvGmOGA/2xqqCecG5fY4y5EojCqiKbc5LvqxSgiUKpE5kDTBWRCSJiBx7Hqj76AfgRcACPiIhdRK4BRjY49w3gAREZ5Wx09heRqSISeJIxzALuEpGhzvaNP2NVle0XkRHO69uBI0AFUOtsQ7lFRIKdVWbFQO1p/B1UJ6aJQqkWGGN2ArcC/wHysBq+LzfGVBljqoBrgDuBAqz2jI8anLsWuBeraugwkOo89mRjWAr8DpiHVYrpCdzo3B2ElZAOY1VP5QMvOPfdBuwXkWLgAay2DqVOmujCRUoppVqiJQqllFIt0kShlFKqRZoolFJKtUgThVJKqRZ5ujqA9hAREWESExNdHYZSSp1V1q1bl2eMiTx2u1smisTERNauXevqMJRS6qwiImlNbdeqJ6WUUi3SRKGUUqpFbpUoRORyEZleVFTk6lCUUsptuFUbhTFmIbAwOTn53mP3VVdXk5mZSUVFhQsiO3N8fHyIi4vDbre7OhSllJtwq0TRkszMTAIDA0lMTKTxZJ/uwxhDfn4+mZmZdO/e3dXhKKXchFtVPbWkoqKC8PBwt00SACJCeHi425ealFJnVqdJFIBbJ4k6neEelVJnVqdKFCdUVgBH8lwdhVJKdSiaKBoqPwxl7ZMoCgsLeeWVV076vClTplBYWNj2ASmlVCtpomhIPKC2fRYBay5ROByOFs9btGgRISEh7RKTUkq1Rqfp9dQqHjYwNe1y6SeffJI9e/YwdOhQ7HY7Pj4+hIaGkpKSwq5du7jqqqvIyMigoqKCRx99lPvuuw84Oh1JaWkpl156KWPHjuWHH36ga9euzJ8/H19f33aJVyml6nTKRPHMwm1sP1B8/I6aSqipBq/Sk75m/9gg/nD5gGb3/+Uvf2Hr1q1s3LiRb7/9lqlTp7J169b6bqwzZswgLCyM8vJyRowYwbXXXkt4eHija+zevZtZs2bxxhtvMG3aNObNm8ett9560rEqpdTJ6JSJonlnrsfQyJEjG411ePHFF/n4448ByMjIYPfu3ccliu7duzN06FAAhg8fzv79+89UuEqpTqxTJopmv/mXHoTiAxAz2KqGakf+/v71v3/77bcsXbqUH3/8ET8/P8aPH9/kWAhvb+/63202G+Xl5e0ao1JKgTZmNybO5NAO7RSBgYGUlJQ0ua+oqIjQ0FD8/PxISUlh5cqVbf7+Sil1qjpliaJZ4sybtbXQxgWK8PBwxowZw8CBA/H19SU6Orp+3+TJk3nttdfo168fffv25dxzz23bN1dKqdMgxhhXx9DmkpOTzbELF+3YsYN+/fq1fGJFERTshYg+4OXf8rEdWKvuVSmljiEi64wxycdu16qnhupKFKZ9xlIopdTZSBNFQ3VtFLXtM5ZCKaXORh0+UYhIDxF5S0TmtvubeWiJQimljtWuiUJEZohIrohsPWb7ZBHZKSKpIvJkS9cwxuw1xtzTnnEeDaz9ej0ppdTZqr17Pc0EXgL+W7dBRGzAy8BEIBNYIyILsPoZPXfM+XcbY3LbOcajGvZ6UkopBbRzojDGLBeRxGM2jwRSjTF7AUTkQ+BKY8xzwGWn+l4ich9wH0B8fPwpXqSu6klLFEopVccVbRRdgYwGrzOd25okIuEi8howTESeau44Y8x0Y0yyMSY5MjLy1CITsaqfOnAbRWJiInl5umaGUurM6fAD7owx+cADZ+wNxUN7PSmlVAOuKFFkAd0avI5zbusYPDzarUTx3nvvMXLkSIYOHcr999/Pyy+/zBNPPFG/f+bMmTz88MMAXHXVVQwfPpwBAwYwffr0dolHKaVawxUlijVAbxHpjpUgbgRubosLi8jlwOW9evVq+cDPn4ScLU3vqy6zqqA8T3Kdh5hBcOlfmt29Y8cOZs+ezYoVK7Db7Tz44IMEBATw8ccf88ILLwAwe/Zsnn76aaB1044rpdSZ0K6JQkRmAeOBCBHJBP5gjHlLRB4GFmP1dJphjNnWFu9njFkILExOTr73NC/UFuE08tVXX7Fu3TpGjBgBQHl5OVFRUfTo0YOVK1fSu3dvUlJSGDNmDNC6aceVUupMaO9eTzc1s30RsKg937tFLXzzp2AvOCohqm3nSjLGcMcdd/Dcc417AM+YMYM5c+aQlJTE1VdfjYi0etpxpZQ6Ezr8yOyTISKXi8j0oqKi07hI+/R6mjBhAnPnziU31xoWUlBQQFpaGldffTXz589n1qxZ3HjjjYBOO66U6ljcKlEYYxYaY+4LDg4+9Yu0U6+n/v378+yzz3LJJZcwePBgJk6cSHZ2NqGhofTr14+0tDRGjhwJWNOOOxwO+vXrx5NPPqnTjiulXEqnGT9WcRaUHoLYoe0T3Bmg04wrpU6FTjPeWmIDTIcedKeUUmeSWyWKtmmj0PmelFKqIbdKFCdqo2hVNZvH2T2DrDtWJSqlXMutEkVLfHx8yM/PP/EH6Vm8yp0xhvz8fHx8fFwdilLKjXT4uZ7aSlxcHJmZmRw6dKjlA6sr4Egu5At4ep+Z4NqQj48PcXFxrg5DKeVGOk2isNvtdO/e/cQHpq+Cj6bBLfOg98XtH5hSSnVwblX11CaN2d4B1r9VJW0TlFJKneXcKlG0yYA770Dr38rStglKKaXOcm6VKNqEV12JQhOFUkqBJorjaYlCKaUa0URxLJsdbN7aRqGUUk5ulSjapDEbrAbtSk0USikFbpYo2qQxG6x2Cq16UkopwM0SRZvxDtTGbKWUctJE0RQvrXpSSqk6miia4h2gJQqllHLSRNEUbaNQSql6miiaoiUKpZSq51aJou26xwZpiUIppZzcKlG0affYqlLQRYCUUsq9EkWb8Q4ADFQdcXUkSinlcpoomqITAyqlVD1NFE3RiQGVUqqeJoqm1JUoKotdG4dSSnUAmiia4q1VT0opVUcTRVPqSxSaKJRSyq0SRduNo3C2UWiJQiml3CtRtOk4CtCJAZVSCvB0dQAdyXsr0yivquHeUZHWBi1RKKWUe5UoTtfyXYeYvTYD7P6AaBuFUkqhiaKRxAh/0gvKqEWOTuOhlFKdnCaKBuLD/Khy1JJTXKHrZiullJMmigYSwv0ASMsv0xKFUko5aaJoIDHcH4C0/CPOEoUmCqWU0kTRQJdgHzw9hLQCLVEopVQdTRQNeNo86BbmR3p+mTXoTtsolFJKE8Wx4sP82J9/BHxCoCwfgNTcUj7fku3awJRSykU0URwjIdwqUZjwnlCSDRXF/Pur3Tz+v02uDk0ppVzCrRJFW8z1lBDuT0mlg9KgXtaGvN1szDhMWVUNFdU1bRSpUkqdPdwqUbTFXE8JYVYX2XRbHAAlmVvJKCgHoLCs+vSDVEqps4xbJYq2kBhhJYrUqkjwsJO/f0v9vsLyKleFpZRSLqOJ4hhxoX6IwL7DlRDeE8fBnfX7Dh/REoVSqvPRRHEMH7uNmCAfq4tsRB8CSlLxsVt/psIyLVEopTofTRRNSAj3I62gDBPRl8jqbC7oHgRAYbmWKJRSnY8miiYkhPmTln+EHO8EbGK4rFsZAIe1RKGU6oQ0UTQhIcKPvNIqVpdGADDC/xDenh7a60kp1SlpomhCQpg1OeAHe7ypRYiqTCPUz0vbKJRSnZImiibUTTe+KqOcPFs0Hnk7CfGzc1hLFEqpTkgTRRPinYkCoCSoJxzaRYifXUsUSqlOSRNFE4J87IT5ewHgGZ0E+amE+dq0jUIp1SlpomhGXfVTeMIgqKmkh2eeVj0ppTolT1cH0FENjA2mvKqGgLhIABLJorCsK8YYRMTF0Sml1JmjiaIZv72sH1WOWjBHAOjmSMdRG0tppYNAH7uLo1NKqTNHq56a4e1psxKCbwgExBBVmQboDLJKqc5HE0VrRPYhrGwfoIlCKdX5aKJojcgkAor34EGtTuOhlOp0OnyiEJGrROQNEZktIpe4JIjYc7A5jtBLsnRiQKVUp9OuiUJEZohIrohsPWb7ZBHZKSKpIvJkS9cwxnxijLkXeAC4oT3jbVZcMgBDPVJ10J1SqtNp715PM4GXgP/WbRARG/AyMBHIBNaIyALABjx3zPl3G2Nynb//1nnemRfWE+MTzFBHKod08SKlVCfTronCGLNcRBKP2TwSSDXG7AUQkQ+BK40xzwGXHXsNsQYt/AX43Bizvrn3EpH7gPsA4uPj2+YG6nh4IF2TSU7dzSxdDlUp1cm4oo2iK5DR4HWmc1tzfgZcDFwnIg80d5AxZroxJtkYkxwZGdk2kTYUl0xPyaSstKjtr62UUh1Yhx9wZ4x5EXjR1XHQNRkbtYQWbgPOc3U0Sil1xriiRJEFdGvwOs657bSJyOUiMr2oqB2+9TsbtGNLt57gQKWUci+uSBRrgN4i0l1EvIAbgQVtcWFjzEJjzH3BwcFtcbnG/MLItXele8WOtr+2Ukp1YO3dPXYW8CPQV0QyReQeY4wDeBhYDOwA5hhjtrVnHG0lO2AASTU7wRgAKh01rE8/7OKolFKqfbVrojDG3GSM6WKMsRtj4owxbzm3LzLG9DHG9DTG/Kk9Y2hL+aFDiKSQmkKrLf7dH9O47tUfyCutdHFkSinVfjr8yOyT0a5tFEBp+BAAyvetAmDN/gJqDRwoLG+X91NKqY7ArRJFu7ZRALUxA6g0dhzpazDGsD69EIDsoop2eT+llOoI3CpRtLfggAC2mkQ8D6wjq7CcQyVWldPBYk0USin31apEISKPikiQWN4SkfUum6DPhUJ87Wys7YVv/hY27sup356jJQqllBtrbYnibmNMMXAJEArchjWtRofS3m0UoX5efFs7BFtNJaXbFuNrtxEd5K2JQinl1lqbKOoWiZ4CvOvsztrhFo5u7zaKUD8vfqztT4U9mOjMxQyOC6ZriC85WvWklHJjrU0U60TkS6xEsVhEAoHa9gurYwr08aRWPEkJHkdyxUqS4/zpEuyrJQqllFtrbaK4B3gSGGGMKQPswF3tFlUH5eEhBPva+ahiOIFSzkXe24kO8iGnuALjHISnlFLuprWJYjSw0xhTKCK3Yq0N0eGmUW3vNgqwqp9m5fWgyPjR//DXxAR7U1ZVQ0mlo93eUymlXKm1ieJVoExEhgCPA3tosBhRR9HebRQAIX52qvFkhecofPd8QZdAG6A9n5RS7qu1icJhrLqVK4GXjDEvA4HtF1bHFeLnBUBGzESoKKLPkQ2AJgqllPtqbaIoEZGnsLrFfiYiHljtFJ1OiJ912z5JF4NXIF0PfAloolBKua/WJoobgEqs8RQ5WGtIvNBuUXVgoc4SxZDEaOh7Kf77PseLau0iq5RyW61KFM7k8D4QLCKXARXGmA7XRnEm9I0JpEuwD/27BMGwW5Hyw/zMd7EmCqWU22rtFB7TgNXA9cA0YJWIXNeegZ2KM9HraVpyN3548iK8PD2gxwWQdBn3mnlU5qe323sqpZQrtbbq6WmsMRR3GGNuB0YCv2u/sE7Nmej1BCDSYFD6pD/jIYYrD77aru+plFKu0tpE4WGMyW3wOv8kznVvoQksj7qNcdXfwd5lro5GKaXaXGs/7L8QkcUicqeI3Al8Bixqv7DOLjt73kV6bSS1i56AmmpXh6OUUm2qtY3ZTwDTgcHOn+nGmF+3Z2Bnk8jQEJ513IpH3k7YNMvV4SilVJvybO2Bxph5wLx2jOWsFRPsw5e1yZSGDyZg2Qsw+Ebw9HJ1WEop1SZaLFGISImIFDfxUyIixWcqyI4uJtgHELb2eQiK0qle9y7vr0rDUdPpJthVSrmhFksUxpizapoOEbkcuLxXr15n9H2tRAFbfEZwbtwIKr7+K88UhRMb7MuFSVFnNBallGprbtVz6Ux1jz1WoLcnfl42sosrcYx7ksDKHKbZvmVHjha6lFJnP7dKFK4iIsQE+XCwuIJPS5NYU9uHhzznk5qV5+rQlFLqtGmiaCMxwT5kF5Xz5op9fOB/O12kgJHp010dllJKnTZNFG0kJsiHzZlFbM0qZsT4K9kcfRXTKj6ics/3rg5NKaVOiyaKNhIT7IOj1hDm78U153TlwKjfkWEikY8fgAptq1BKnb00UbSRup5Pt56bgI/dRu9uMTxW/SD20ixY/JSLo1NKqVOniaKNnNsjnHN7hHHH6AQAEsP92WZLYkXMbbDhPdj6kYsjVEqpU+NWieJMTDPenD7RgXx432jCA7wBsHkIfWMCmW67HrqNgvkPwcFtZzwupZQ6XW6VKFw1jqI5/WKC2JpTgbn+HfAOgg9vgfLDrg5LKaVOilslio4mqUsgBUeqOCShMO2/UJQJ8+6F2hpXh6aUUq2miaIdJcUEAZCSXQLxo2DKXyF1CXz7nIsjU0qp1tNE0Y76dbGmytqR7eweO/wuGHYrLH8BUnQ5D6XU2UETRTsK8fOiS7APKTkl1gYRmPJ3iB0GH98P+XtcG6BSSrWCJop2lhQTeLREAWD3sdorPDytxu3KUtcFp5RSraCJop0ldQliz6FSDhZX8MLiFEY/9xUrC/zhuhmQt1MH4ymlOjxNFO0sKSaQ6hrD2Oe/5pVv93CwuILF23Kg54Vw3iOw/r+QutTVYSqlVLM0UbSzkd3DiA7y5vIhsSx57AJGJIaxPs05lmL8UxCZBAsegYozP0hQKaVaQxNFO+sS7Muq31zMP6YNpVdUAOckhLLtQDEV1TVWe8WVr0BJNix+uv6c73fnkZZ/xIVRK6XUUZoozrDh8aE4ag2bM50liLjhMOZR2PAurJrOhv253D5jFX9YoNN9KKU6BrdKFK6c66m1zkkIBWBdWoOpPMY/BQlj4fMniHlnDNM8vmbV7hwKy6pcFKVSSh3lVomio8311JQwfy+6R/g3ThSe3nDnp7zf8wUO1gTwF/ubzPT8E19v1HEWSinXc6tEcbY4Jz6UDemHMcbUb1uxJ5+nt3Xlk+HvYq56jeEeuxny7R1QVtDkNQ6VVJJRUHamQlZKdWKaKFxgeEIo+UeqSMu3Pugrqmv41dzN9Ij059eX9kOG3sQnfZ8nrnIPjrcvg9JDjc6vrqnl1jdXcffMNacdS0V1DX//ciclFdWnfS2llHvSROEC5ySEAEfbKWas2EdWYTnPXT0IXy8bAEnjpnFP9RPWNB9vTmi0lsXMFfvZebCE3bmlp/0B/93uPP7zdSqfb805resopdyXJgoX6B0VSKC3J+vTD1NwpIpXv9nDxf2iGNUjvP6YAbFBZIaO4tmI58FRCW9OhO3zyS4q519LdxEdZC2QtP3A6a3HvTPHOn/dfl0nQynVNE0ULmDzEIbGh7Au7TAvfZ3KkSoHv56c1OgYEWHq4C68mxlFwa1fQnR/mHM7+966ixFmM6/ePBiAraeZKOomLFyb1nRbiFJKaaJwkXPiQ9l5sIR3V+5nWnI3ekcHHnfM1EGx1NQaXltfxrLz3mZn/I0MK/qKmbY/cc6HyTzv9x6707NPK46dzkSx59ARDh/R7rhKqeNponCR4QmhGGOVLh6b2KfJY/p1CaRPdADTl+/ljv9uZtKuK7gm4D2qrn8f+lzK9bVf8OjuOyHth1OKodJRw968I4xItMZ2rE/X6iel1PE8XR1AZzU0PgR/Lxv3jutBdJBPk8eICHPuH016QRnVNYbqmlr6xwbh5WOHAZcxx1zM6M2/wbw9BTn/cbjot9aaF620J/cINbWG65O7sSG9kLVph5nQL7qtblEp5SY0UbhIkI+dFU9eRLCvvcXjQvy8CPHzanJfWNL5XLrmLywf9DkR3/0N7L4w7petjmHXQavaaWi3EAZ0DdYGbaVUk7TqyYVC/LyQkygBHGtg12DK8GFR99/AoOvh6/8Hm2a3+vyUnBLsNqF7hD/JCaFsyiykylF7yvEopdyTJoqzWJdgH8L9vdh6oASufBkSz4f5D8HupVB74g/8nTnF9IwMwG7zIDkhlEpHLVsPdNx5spRSrqGJ4iwmIgzoGszWrGJrvqgb3oPwXvD+tfBsJPxzEMy6CYqb7hm1M6eEpBirt9Vw52SF69O0+kkp1ZgmirPcwNggdh0sodJRA74hcOdnMPUf1up5CaNh7zJ48+JGI7sBisqrOVBUQR9noogK8qFbmC9rtZ1CKXUMbcw+yw3sGoyj1rAzp4TBcSHgHw4j7jl6QPYm+OAGeGsSTJsJvS4GjjZk15UoAJITwvhudx7GmNNqO1FKuRctUZzlBsZaU6pvzWpmhHaXIfCTryA0Ad67Fl4aCYufpmjrEmzU0DcmqP7Q4Qmh5JVWkq6z0iqlGtASxVmuW5gvQT6eLTdCB3eFu7+A9f+F3Utg9XQurqlitU8QYd9Pg8HToNtIRnYPA6yJAhPC/c/QHSilOroOX6IQkX4i8pqIzBWRn7o6no5GRBjYNZiN6YWN1rc4jncgjH4Ibv8Efr2fF0J+x07vIcjG92DGJbDoCXpH+NA3OpD/rcs8Y/ErpTq+dk0UIjJDRHJFZOsx2yeLyE4RSRWRJ1u6hjFmhzHmAWAaMKY94z1bTewfzfbsYv7yRUrLycLJ2P14t3AQnyY9B0+kwuiHYc0byIe3cPOwcDZlFNbPAaWUUu1dopgJTG64QURswMvApUB/4CYR6S8ig0Tk02N+opznXAF8Bixq53jPSneel8ht5ybw+rK9vPR16gmPzymuoLjCYTVkewfCpD/B1L9D6hJu2X4fEzw3MXeVLsOqlLK0axuFMWa5iCQes3kkkGqM2QsgIh8CVxpjngMua+Y6C4AFIvIZ8EFTx4jIfcB9APHx8W1zA2cJEeGZKwZwpMrB35fswt/bk7vHdm/2+C2ZVntG34Yz1o74CYQk4DnvJ7zl+TwlG16kpmoKtguegKh+7X0LSqkOzBVtFF2BjAavM53bmiQi40XkRRF5nRZKFMaY6caYZGNMcmRkZNtFe5bw8BD+eu1gJg2I5tnPttcng6bM33SAUD87w+JDG+/oPREe38nmcdP5zDGK2l1LYPqFsG4mtKJKSynlnjp8Y7Yx5ltjzCPGmPuNMS+7Op6OzNPmwQvXDyE8wJvffLyFmtrjP9yLyqpZsu0gVw7tipdnE4/f7sOA8dP4t/8jPB71BsSPgoWPwty7IXcH1DjOwJ2cmKOmlucW7SCrsNzVoSjl9lyRKLKAbg1exzm3nTYRuVxEphcVdd75ioJ87Pzusv5sySrivZVpx+1fsPkAVTW1XDc8rtlr2DyE64bHsXBvDQcu/wAm/AG2z4dXzoXnusL08bD8b1DeeBT3F1tz+G73oba+pSbtPFjC68v38tnmA2fk/ZTqzFyRKNYAvUWku4h4ATcCC9riwsaYhcaY+4KDg9vicmetywd34fzeEbyweCcHiysa7Zu7LpOkmEAGxAY1c7bl+uHdMAY+3pgN5/8CfrYOrp5utWV4+lgz1f5jAHzxG0j7kbUbN/HI+6v42awNlFU1LnV8vzuP2WvS2/Qe9+UdASAtXwcHKtXe2rt77CzgR6CviGSKyD3GGAfwMLAY2AHMMcZsa+k66uSICP/vyoFU1dTyzMJt9V1mU3NL2JRRyHXD4044RUd8uB/JCaF8siHLOj+sOwy5weohdfcX8MAK6HcZrHoN3p5M8ifjSPG+g9ccv2fJkqNNSaWVDn4+ewN/XLi9yaqwU7X3kJUodBS5Uu2vvXs93dTM9kVoV9d2lRjhz6MTevPC4p08Nnsjf7l2MHPXZeHpIVw1rNm+A41cOawrv/tkKzuyS+h/bAkkZiBcM53qi/7AX9+ZR0V+Bo8le5O08V3OXXMbNUeuwdbnEnav/Io3q9aRb4JJy0iiR0Jim9zf3kOlgJYolDoTOnxj9snQNorGHhzfk19e0odPNh7ghukr+Wh9JuP7RhER4N2q86cO6oKnhzB/Y/NNSH/7oZg3snsy4trHCLviWbZc8w0vOq7CpCyCTx6gT/YCPL38GOOxlcg5V0DB3ja5t7qqp6zCchw1utiSUu3JrRKFtlE0JiI8fFFvXrt1OLtySsgtqWyxEftYYf5eXNAnkgWbDlDbRLXRtgNFTP9uLzeNjOeKIbEAjB3QnS8if8INvq/zQo+3SXbMIPCnX3K3+R22isPw5kRIX3la3W2NMew9dIRAb09qag0HCitOfJJS6pS5VaJQTZs8MIZ5Pz2Pxy7uw4R+USd17hVDY8kuqmDVvoJG240xPLNgO6F+Xjw5Oal+u4jw0/E9WZdv5+Xt3tx6Xg8Swv1xxI7k1yF/By9/mDEJ/tEf5t4Dq9+ArHXgqGx1TIdKKympdDC2dwQAaQVHGsX10/fWsWhL04s1KaVOns4e20n0jw06vp2hFSb2j8bPy8b8jVmM7hlev/3Tzdms3l/An68eRLCfvdE5UwZ14W9f7qSwrJqHL+wNwKC4YN5fVYjjV1/jueNjSP8R9n8PW+cCYDzsZNoTORLWH7/4ocT0Ow+vhFHQRKP7PmdD9vi+kXy+NadRg3ZuSSWfb82h0lHLlEFdTvp+lVLHc6tEISKXA5f36tXL1aG4DT8vTyYNiGHRlmyeuXIA3p42yqoc/HnRDgbEBnHDiG7HnWPzEN66I5mK6tr6JDI4Lpi3vq8l9YgXSSPvhZH3WtVPhelUZa5n7sKFJJTvJOnA14Rnz4dVcCQ6Gf+pf4L4cxtdf6+zfWJ0jwi8bB6kN2jQ3ppltU+t3V9Aba3Bw0MXYFLqdLlVojDGLAQWJicn3+vqWNzJlUNj+XhDFk/N20L/2CB2ZJeQXVTBizcNw9bMB3GvqMBGrwd1tdqNNmcWkVS3WJIIhCbwwg9HeKPYi3fvGUlNdADf7NzJ2sUfcNfBOfjPmAQ9LoTALlBTBRjCihNI9OxBXKgvcWG+jXo+1S3gVFzhYHduKX1jGsdxMnKLK8gtqWRgV23zUp2bWyUK1T7G9oogOSGUT7dk89EGqwfUVUNjGZEY1uprJIb7E+jtyZbMIqYlHy2FrN5XwJvf7+OWUfGc39uaoytqxFDCuyQy4bXR/CZsGdMOf4Pk7wFPL3BUMaloHpM8genTeUL6sionCcp7gG8o2w4UEeTjSXGFgzX7C04pUezPO8Lry/cwb10WNcbwyYNjGBSnyUJ1XtKa9QvONsnJyWbt2rWuDsPtGGMoqXSQX1pF1xDfpueKasGN03+kvLqW+Q9Zy4qUVTm49N/fUWsMXzw6Dn/vxt9b5qzJ4FfzNvPABT158tKjDea3//V9rvLdyDX+W3FkrMHTVGMQpM9kHth/Pvbu57Fqbz7n9QznXzcOO3rB2lqrFNPCYMP/rc3g1/M242nz4LrhcXy14yChfl4seHjsSd+vUmcbEVlnjEk+drtblSi0jaJ9iQhBPnaCfOwnPrgJg+NCmPnDfqoctXh5evD85ymk5Zfx4X3nHpckAKaN6MbGzEJeW7aHif2jGZ4QSnVNLT8UhjJoyD0wKYn3lu1g0RefMXNcCT6b/8trVZ+TnTOUHf49Oby7DBbGQEk25O+Bw/vBLxz6XgpJl0H3cVYppYG56zLpERnABz8ZRVSQD0v6RnHvf9fy2rI9PDKh9yndt1JnO7f6iqTjKDq2QV2DqXLUsutgCav3FfDOj2nceV4i5/YIb/ac307tR4C3J++vsiY4zCgow1Fr6BERAEBcZBirTT9SBvyclVcs4/+qbyekpoCxpYu50PEdNTsWQlEmRCXB6Aet2XA3z4H3r4WXkq2eV06Omlo2ZxYxtlcEUUE+gNXr64ohsfzn69266p/qtNwqUaiObbCznn/1vgJ+NXcT3cJ8+dXkvi2e4+flyRVDY1m0JZui8ur6OZ66R/oDkBDuB0B6fhmbc6uZWTOZygfXseuu7ZxTOZ1PL1kOP10BN7zHyp6Psnbkv+BXe+HGD0A8YOZl8PmTUFXGzoMllFfXMCw+pFEMf7i8P4E+dn41d1Obzlel1NlCE4U6Y+LD/Ajy8eRvX+5kf34Zz18zGD+vE9d+3jQinorqWuZvzKqfuqNHhJUouoU5E0VBGdsOFNM1xJcQPy+SYgIJ8PZk7X5rKvS9h0q5Y8ZqrnvtR55csIvixEusBDLiJ7DqVfhHEiEf38zPbB8xunYjFGVZ3XeNIbwig7cGbWdo9hzmrWl5CpId2cWtWrdcqbOJJgp1xogIg+NCKKuq4eZR8ZzXK6JV5w2KC2ZAbBCzVmewN6+UMH8vQvystgUfu42YIB/S8svYllXEwK5W11tPmwfD4kNY4xxP8et5m/H29OCuMYnMWZvBxH8s44eMcpj6N7jzM+h/FbbiLB6zzyNqwc3wz/7wXDf4W294aTjDNv6eZ+zvEPjFo5SWNxhFvnYG/K0vrH6DRZsPcOm/v2PxtoOt/pv8Ys5GPtnQJsuxNGl9+mFG/Xkpuw62vtrsSKWjySlbVOeliUKdURclRdE7KoCnGvRiao0bR8azI7uYJdsP1pcm6sSH+bE9u5i9eUcYEHu0fWpEYhg7D5bw8jeprNl/mN9fPoA/XD6ATx4ag5+XJ7+cs8n6QEwcC1e8yC1e/+LR+I/hjk9hyt9gyI3Q62K47F/w0BoOJP+KS8137Hn7XqitgcVPw6ePWW+26JeEzb+NCIr4bFMmFOyz2j+qm5+HKqOgjI/WZzFrdduu1dHQos3ZHCyu5Lcfb21VSae4oprRz33FrDZeP0Sd3dwqUejssR3f3WO7s+QXFxB4kj2nrhwai6/dRl5pFT0ij0kU4X7syLYG2tWVKACSE0MxBv6+ZBcX9Ink2nOs6dUHx4Xw6ITeHCiqYM1+aw6rorJq9hw6QlL3btD9fGvk+NS/wdWvQfJdENmH2MueZkn4rQzJnU/1v4fDjy/ByPvgsa1sHPAUwxybWObzC17YNRleHAozp8LLI2DrR01OgrjcuRrghoxCKh01J/X3ONaCTQe49tUfqHI0nkn3+9Q8Arw9Wb2/gLnrMk94neW7DlFc4aivslMK3CxRaK8n9xXkY2fqYGvupu7OHk91EpztFAADG5QohnYLwdND8Pey8edrBjVarGli/2h87TY+2WgtpboxsxCAYd1CWoxj4G0vMLP2UmxFaTD5eZjyAlXGxs/2jeSx4H9S2n0y79ZMZNvwZ+G6t8E7CObeZU2E+PWzsOLfsPZtOJzG8l1WoqhyWL2tTlVppYM/LtzOurTD/Lg3v377oZJKUnJK+On4npwTH8Jzn6dQWFbV4rW+2pELQEob9PDafbCEL7bmnPZ1lOu5VaJQ7u3WcxMQOTodSJ14Z8+niADv+m6tYPWYemxiH/4+bShdQ3wbnePv7cnE/tEs2pJNlaOWDemHEYHBJ0gUXUL8OHz+HxlSMZ2H9o4kp6iC/63LIKOgnOunTCLs1hm8ZL+Lt8rOh4HXwP3L4fIXofiAtc74kt/Dpz/H/OccJqU+yy19rJLE6mNm5z0ZbyzfS15pJV6eHnyx9eisuT/syQPg/N4R/OnqQRSVV/P8FzubvY6jppavU6xEsSe39LTX+Xjyoy088uGG0y4tKddzqwF3yr0N7RbCqqcmEBnYeOGleGeJomG1U52HLmx+8OVVw2JZsOkAy3YdYkN6IX2jrZ5SJ/LwRb3w9BBe+iaVb1Ny8fL0YHhCKOP7RiIiXNI/mi+25VDpqMHb0wbD77B+jIGqI1B6kNylLzJl+/tclbGcO/0SKV4dDeWDIDAGAqIhIMoqjdjsYPOC4DjwDTkultziCt74bi9TB3dBgC+3HeTZqww2D2FFah7BvnYGxAZj8xDuOi+RN7/fx63nxjdqy6mzLu0wReXVXNwvmqU7DrI/v4xeUQHHHdcamzMLWZdmVV/tyC5h6AkS8LGqHLVsyiwkOSH0hMv2qvanJQp1VokK8jnugyMx3GqzGNjEh19Lzu8dSZi/F59syGJjRuFx4yeaY7d58LMJvVny2AWM6B5GUXk1T0zqWx/XpYNiKKlw8ENqfuMTRcA7AMJ78l7og4yv/jdVo35GbUAUgWUZmI0fwFfPwPwH4f3rYMYl8MaF8NoYa/2OJX+AI42v+c+lu6muqeVXk/py6cAu+JZlcmDBM5iP7uXAznWM7hFeP3Hjzyb0xtdu450f9jd5X0t3HMTL5sG953cHOK0Bhm+v2I+3c8qT9Wmtb++oqTV8vCGTi/+xjOtf+5EFmw6ccgyq7WiJQp31Qv29eOuOZM6JDz2p8+w2D6YO6sL7q9KoNTCs28mdHx/ux9t3jqCovLq+uy7AmF4RBHp78vnWbC5ManqhqOW7DtG1W3d8Jt/C7tgDPPzBBub/ZAxDou1Qmmv9VJVATbW1qNOOBVb7xpo3YdD1ED2ALM9ubF+7m+d7V5GwcS1x+75nqvcq2Ai1dj/erppHirkbqgeA3YdgXztXDevKR+szeXpid4KrcyE0wSq1YLVPjOoRxpBuIXgI7DxYwlROfk2P3JIKPt18gFtGJfDlthw2ZBS26ryswnLufnsNOw+W0L9LEJGB3izYeIArh7ZujXfVftwqUehcT53XhH7Rp3TeVcNieXelNT3I0FaWKBoSkUZJAsDb08aEflF8uf0gf6qpxW5rXHAvOFLF5qwifj6hDwAjnbPwrt5XwJBuPSCsu/Xj9MOePEr6ncukcU/Asr/Clv/BurfpCsz3AtKADE9skf34KPQe/ls6gpvHJmFb+luu3fsG/GcRhCSApxdPVzp40GMnQf/MAwzY/SHhPPIjR3Le4QNc2TUCn9UruT24lPyMGjC9W5xEsSnvr0zHUWu447xEDpVUsiG9dSWKPy7cRnpBGS/dPIwpA7vw3Oc7mPnDforKqo9bHOtECo5U4eXp0aqqRHVibvVX1PUo1Mk6Jz6UuFBfisqq6RV5avXxTbl0UBc+2XiAq19ZQUKYP12Cfbj6nK4MiA3m+9Q8jIFxfawBh1FBPnSP8GfVvgLuHdej0XX25R3hnplrqa6pZf7DYxhw/dtgDD9u2sp/5izizuFhXHL++RDWw5rgcH0mG+ds4vDqQhz+j3PNdY8jq6dDVSlUlRFADRt9+7OkNpY7J5+HR84m2Lec8NQlPGsHdlo//weQDjwfDENvhQuf4sG5uzgnPpSfnN84xoYqHTW8vyqdC/tG0T3Cn2HxIXy2JZvckgqiAn2aPe/73Xks3naQJyb15bLB1vrrlw2O5Y3v9rF4e06jqelb4+Y3VtI7OpD/3DTsxAe3k6Lyau6YsZoHLujB5IFn92qLbpUolDpZIsLvLutPbnFFm66GN75vJLePTmDPoVJ25BSzdMdB3lqxj+uHx3G4rJoQPzuD40Lqjx+ZGMYX23IarcrnqKnlsdkbrW/GPp788n+bWfDwGAT4/dcFVIWO5IIrx4Gnrf46E/pF4+khpOWXMS05Duk1BHpd1Ci2vA1Z/HH2RnoFjmTc8NsBuOfVLyivqOKD+8eCh40PFn3D1nXL+eOAIjxXvoJjyzw4PI0v0vpzT8h6JGs9eNggdih0GQqhieBhY+GmbPJKK7nzvESA+nafjemFXDIgpsm/VXVNLc8s3EZ8mB/3jD1akhocF0x8mB8LNx04qURRVFZNSk4JWYXl1NSaZhfX+mxzNtsOFPGrySc3+LO1nv8ihY0ZhXywOkMThVJnu0nNfICdDm9PG3+8cmD966Lyal7+JpW3V+yjusYwdXCXRh9go3qEMXttBrtyS+pXAHzpm1Q2ZhTy0s3D8LJ5cN+763jlmz2E+tvZnVvK67cNt3pVNRDsa+e8XhEs33WIMc1MkXLpoBj+36devLsyjb4xgcxYsY9v0mt46MK+4GdVg4X2HsUHq+3cNHosg0Y/yOEPf8orXi9CFTAP8PQBU+tcddBSiweXGRsX+gYStqIfpPRiiFcgL3utZcCiElguENGH/Z6JzN7vT//BI5g4ZhSz1mWzO7eU6bcNx8d+9H5EhKmDu/Dl8hVUfvAS3v0mweAbwdbyx9bmrEIASiocbMkqarbH1cvfpJKSU8y95/cg1N+ryWNO1Zr9BXywKp0QPzs/7smjpKL6pAeZdiSaKJQ6A4J97fxmSj9uGRXPW9/v49pz4hrtH9nd+oBeuOkAjhrDgcJy/vN1KlcNja2virliSCwvfbMbX7uN0T3CuaR/0+0y1w2PY93+gmYThbenjRtGdOPVZXv4dmcuNbWGSwd14e4xR7/N93GuDJiSU8yg5OHc6/0Cox1fU1paTI+h47nr6ilWd9/c7VSkr+OrNZvZe7CQHqFeTEiwIUV7IeVTPCtLGOYZQX51FF2DojFZa0ks/IhfA3wHju9sjDMxvBM+kHEVeXBoNEQcbRe5KWgLP7U/jX13FexaAN//Cy56GpIubzZhNBy8uCI1r8lEkV1UznbnaP7vU/O4fEhsk9c6FZWOGp6ct5m4UF+evWogd769hmW7DtU/x7ORJgqlzqCEcP9GJY06caF+xIf58fI3e3j5mz0AxAb78EyDY//vigH8sCePgiNV/O6y/s2OL7hiSCyTBkQfV9po6LbRCXy1I5dze4Rxz9ge9YMW6ySG++Pl6cGugyVkHi5jY1YpkybfztY9eSzfV8adHp6ICKbLEO5YWM7a7C788pK+XDquR+MqPGOYvnA7s9dksOXGS1i26xCPvPMd/57gR4LJYsfWdQQV7eT8mtXIwsXWOQEx1qJS3gHEr53BTltPXor4Pf+50GaNbv/fnVYjfFwyxI+25uPqOhw8rE4DmzIK6R7hj4/dxorUvCbH0tQNLPSyebB816E2TRSvfLOHPYeOMPOuEZzfO5JQPztLtx/URKGUOn3v3TOK1EMl1NRa4wmGJ4QS7Hu0uiLM34u37xxJxuEy+sceP7iwoZaSBECXYF8WPzau2f02D6F3VAApOUen4bh0YIxVMvp4CzuyS+gfG8SX2w+yal8B/++qgdx2bsLxFxJhWLy1smFKTgn/+TqV0NAwLrhoPHabB70vuRdjDAKQtxvSf4B9y2HvN3DkEAy7jS987uezb9P5XdwEon46BVI+s47JWAnL/wrL/gKBsdD/CojqT1TafpJjIojxKmd/6g5qPpqJzTis1Q39wiGiD99vC6NbmC+Du4awfPchK4a6xHskHwrTICgW/KPqE1Br5BRV8Mq3qVw5NJbxfa2u0RclRbNkew7VTfSAO1toolCqg4gP9zvum/2xBsUFMyjuzMxl1jcmkBWpeZRX1ZAUE0hihD8BPp789pMtfL41m97RATz/eQo9I/25aUTzjc1141Nedra5/OnqgY0+MOs/oCP7WD/D77SqtcoKwD+cqbkl/PObdBZsPGD1uOp/hfUDUF4IuxbD9vnWHFo1lTwL4Jy5vdZDqN4djc3HB8oOQ6VVLfV3483esLFU2UcSeySNvM9+ILL2EKSvhLxdR4OvGxUf3tuKLaIPRPWHyCRr8KSjylpitygDIvsye0051TWGxyceXZBrYv9o5q3PZM3+As7r2bqp9TsaTRRKqSb1jQ7ko/VZ5JZU1o/5iAjwZmT3MD7fmkN0kA97847wxu3JeLbwTblbmC/h/l7Oc7y5bnhcs8fWEwF/a4ncXlGBDE8I5YNV6dwztnvjKjffEBhyg/VTXc6yjSn88aM1vHRtbxJiYxjx8i5uG9yXJ+umtXdUsX7FF2xfMpNpZevx2v4V59iBtYBvKMSNhCE3We0kJTnWMrqH90N+Kuz9FmoarEXiHwVleVajvtMNRDAybCDxafngOQGCu3J+7wi8PD34emsG53ntg8zVkLXeWp733AfB6+hsyC9+tZsjlQ6emtLvhH+ijIIyXvxqN1uyivjg3nMJa+MG+YbcKlHogDul2k5fZ4O2MVZPqTpTBnXh9/O38fznKYzsHsbF/ZoefV5HnNVPS3fkct+4niesFmvKLaPi+cWcTfy4N7/5b+V2X9YU+LFf4kgcPA5fLxsD4gtZkZp39BhPL+YVdOdjj/u4/vELoaqYK6evJSgwiHfvHdNyELU1VtI4lAK5263fA2MhvBcEdWHXllXsXLOUibXbYMG31jmRSfh72FnrlUHQhkLY4LxWQAxsnQtr3oKLfge9LiY3czd7v/mcWHOIgmIPwqpywD8SLvwNlUHxrE8rpLzaQXlVLSv35vPpmhTGeWxmgsnh3UXlPHr9pJP+u7aWWyUKHXCnVNupSxQ9Iv3p3WBywEkDYvj9/G2UVDr4zZR+rZq0b9KAGDIKyrlp5MkNnKszZVAXnlm4nfdXpbdYfbMps5A+0YH4elnJaGyvCP65dBeFZVWE+HlhjOGblFzO7x2Bt5c3eEUysm8C7/yQRlmVo+WleT1sEN7T+kmaetzuvy73Y6P3MCb/8kLI3wmpS2HfMvCwc9Aribf21nL91EnEDbrAmvwx7Uf48mlrbi8gCviX8+2Ld4VATHfIWIXZPp9FftfwwqHR9PQ4wEDZzwRbCn/w2oanqbZO2DabspLx+J13P/SZZMXahtwqUSil2k5MkA+9ogKsgXsNkkF0kE99w3ZrZ4W9Prkb15/k6OqGfOw2rhsexzs/7OdQSeVxMwgDGGPYklXE5AbjYsb0CucfS+DHPflcOqgLKTklHCiq4OcX96k/ZlyfSN74bh+r9hY0OzcXQG2tYXNWEUPigo9LjtlF5XydcpAHLuiJ3dMG0f2tnzGPABBcXMG///wVB7LieP7caGs21oTRcM9S2PkZJbnp/HLpYfr0HUhwl548uySdORNHMzKsnN3v/4Krc2dxtc+s+verCe2BLel+SLqMPHs0s177M7dlfY3fhzdZU9t3GXKKf+mmaaJQSjVJRFj6iwua3PfqrcPPcDRws3MMypy1GU12eU0vKKOwrLrRiPfBcSEEeHvy6rI9bMkqYtfBUgDGJ0XWHzMiMQwfuwfLdh1qMVHMWLGPZz/bwR8u789dDcacAMxZk0mtgRtHxDd5blSQD/ee3503vttHVU0tL1w3BC9PD6tHVb/LeSUthSWOPfx60gV0Cfbl9ZUHeWFxCk9P7c91mXfwQI/LeHxACRI9ALoMxuZ7dALLCKB23K8ZvvQKPr8S+rRxkgCdZlwpdZboGRnA6B7hzFqdTnlVDd/vzuOfS3bVL2e7yTnQbki3o73C7DYPbh4VT+bhcl5fvpelOw4yukd4o3mnfOw2RnUPZ8GmA/zfgm2888N+tmY1XnGw0lHDG9/tRQSeW5TCtgNH91fX1DJ7TTrn945osdfab6b044lJfZm/8QB3zVxNUZlVbVRUVs27P6YxZVAXekQG4Otl45GLerFm/2HufHs10UE+3HvzjcjoB6HHBVaj+zHuHded8EA/ntwQ3qq10U+WtMdFXS05OdmsXbvW1WEopdrYp5utKdm9bB5UOVfgE4H7x/WkvMrBh2sy2PrMpCbHKxhjKK+uwcvmcVwvreW7DvHC4p3syztCaaUDEZh93+j6EfMfrk7nyY+28O8bh/Knz3YQ6OPJwp+N5XBZNT/7YD3r0wt58/ZkLm5mtHxDc9dl8ut5mwHo3yWIAG9Pftybz6JHzq8fH1PlqOWiv3/LgcJyZt8/mhHOGYZbMnuNFeP8h8Y0KlWdDBFZZ4xJPm67Jgql1NmiylHLL/+3iVA/O+P6RDI4LoR/LNnJrNUZAJwTH8JHD56g91ILjDHkFFdww+srMRg+f3QcvnYbF/9jGf7eNhY+PJYf9+Rzy1urGNsrgs2ZRdTUGp67ZtBJje7emlXE4m05rN5XwIaMQiYkRR1XnZeSU0x2UQUX9m25V1mdmlrDnkOl9IkOPKl7bkgThVLKbS3ZfpCnPtrCnecl8PBFvU/7emv3F3D96z9y44hujOkVwcMfbOCVW85hyiBrFti/fpHCK9/uYWDXIF666RwSI/xPcMXmVdfUYhNp09mLT5UmCqWUW6v7LGurNbb/8nkKry3bQ2SgN4Heniz5xQX1M/46amr5LjWP83qGn9K4kI6quUShjdlKKbcgIm2WJAAem9ibpJhADpVU8sAFPRtNC+9p8+DCvlFulSRaot1jlVKqCd6eNl655Rw+2ZDFVcM697rdmiiUUqoZPSID+MUlfU98oJtzq6onEblcRKYXFRWd+GCllFKt4laJwhiz0BhzX3DwmZmGWSmlOgO3ShRKKaXaniYKpZRSLdJEoZRSqkWaKJRSSrVIE4VSSqkWaaJQSinVIrec60lEDgFpp3h6BJB3wqPcT2e8b73nzqMz3vep3HOCMSby2I1umShOh4isbWpSLHfXGe9b77nz6Iz33Zb3rFVPSimlWqSJQimlVIs0URxvuqsDcJHOeN96z51HZ7zvNrtnbaNQSinVIi1RKKWUapEmCqWUUi3SRNGAiEwWkZ0ikioiT7o6nvYgIt1E5BsR2S4i20TkUef2MBFZIiK7nf+GujrWtiYiNhHZICKfOl93F5FVzuc9W0S8XB1jWxOREBGZKyIpIrJDREa7+7MWkcec/21vFZFZIuLjjs9aRGaISK6IbG2wrclnK5YXnfe/WUTOOZn30kThJCI24GXgUqA/cJOI9HdtVO3CATxujOkPnAs85LzPJ4GvjDG9ga+cr93No8COBq+fB/5pjOkFHAbucUlU7evfwBfGmCRgCNb9u+2zFpGuwCNAsjFmIGADbsQ9n/VMYPIx25p7tpcCvZ0/9wGvnswbaaI4aiSQaozZa4ypAj4ErnRxTG3OGJNtjFnv/L0E64OjK9a9vuM87B3gKpcE2E5EJA6YCrzpfC3ARcBc5yHueM/BwDjgLQBjTJUxphA3f9ZYSzz7iogn4Adk44bP2hizHCg4ZnNzz/ZK4L/GshIIEZEurX0vTRRHdQUyGrzOdG5zWyKSCAwDVgHRxphs564cINpVcbWTfwG/Amqdr8OBQmOMw/naHZ93d+AQ8Lazyu1NEfHHjZ+1MSYL+BuQjpUgioB1uP+zrtPcsz2tzzdNFJ2UiAQA84CfG2OKG+4zVp9pt+k3LSKXAbnGmHWujuUM8wTOAV41xgwDjnBMNZMbPutQrG/P3YFYwJ/jq2c6hbZ8tpoojsoCujV4Hefc5nZExI6VJN43xnzk3Hywrijq/DfXVfG1gzHAFSKyH6tK8SKsuvsQZ/UEuOfzzgQyjTGrnK/nYiUOd37WFwP7jDGHjDHVwEdYz9/dn3Wd5p7taX2+aaI4ag3Q29k7wgurAWyBi2Nqc866+beAHcaYfzTYtQC4w/n7HcD8Mx1bezHGPGWMiTPGJGI916+NMbcA3wDXOQ9zq3sGMMbkABki0te5aQKwHTd+1lhVTueKiJ/zv/W6e3brZ91Ac892AXC7s/fTuUBRgyqqE9KR2Q2IyBSsumwbMMMY8yfXRtT2RGQs8B2whaP19b/BaqeYA8RjTdE+zRhzbEPZWU9ExgO/NMZcJiI9sEoYYcAG4FZjTKULw2tzIjIUqwHfC9gL3IX1BdFtn7WIPAPcgNXDbwPwE6z6eLd61iIyCxiPNZ34QeAPwCc08WydSfMlrGq4MuAuY8zaVr+XJgqllFIt0aonpZRSLdJEoZRSqkWaKJRSSrVIE4VSSqkWaaJQSinVIk0USnUwIjK+boZbpToCTRRKKaVapIlCqVMkIreKyGoR2SgirzvXuygVkX8610P4SkQinccOFZGVzrUAPm6wTkAvEVkqIptEZL2I9HRePqDBOhLvOwdMKeUSmiiUOgUi0g9r9O8YY8xQoAa4BWsSurXGmAHAMqzRsgD/BX5tjBmMNSq+bvv7wMvGmCHAeVgznoI1q+/PsdZG6YE1X5FSLuF54kOUUk2YAAwH1ji/7PtiTcBWC8x2HvMe8JFzXYgQY8wy5/Z3gP+JSCDQ1RjzMYAxpgLAeb3VxphM5+uNQCLwfbvflVJN0ESh1KkR4B1jzFONNor87pjjTnWOnIbzENWg/68qF9KqJ6VOzVfAdSISBfVrFSdg/T9VN0vpzcD3xpgi4LCInO/cfhuwzLnCYKaIXOW8hreI+J3Jm1CqNfRbilKnwBizXUR+C3wpIh5ANfAQ1uJAI537crHaMcCa8vk1ZyKom8UVrKTxuoj80XmN68/gbSjVKjp7rFJtSERKjTEBro5DqbakVU9KKaVapCUKpZRSLdIShVJKqRZpolBKKdUiTRRKKaVapIlCKaVUizRRKKWUatH/ByZ1tW1OW10vAAAAAElFTkSuQmCC",
      "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",
    "\n",
    "if (Training_Percentage > 0):\n",
    "    val_loss_ges = np.append(val_loss_ges, history.history['val_loss'])\n",
    "    plt.semilogy(val_loss_ges)\n",
    "\n",
    "plt.semilogy(history.history['loss'])\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": [
    "### Step 2: Brigthness and Pixel Shift scattering\n",
    "Here a higher number of epochs is used to reach the minimum loss function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Muell\\AppData\\Local\\Temp\\ipykernel_18484\\3747815558.py:11: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
      "  history = model.fit_generator(train_iterator, validation_data = validation_iterator, epochs = Epoch_Anz)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0665 - accuracy: 0.9259 - val_loss: 0.0191 - val_accuracy: 0.9592\n",
      "Epoch 2/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0151 - accuracy: 0.9641 - val_loss: 0.0108 - val_accuracy: 0.9731\n",
      "Epoch 3/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0092 - accuracy: 0.9719 - val_loss: 0.0066 - val_accuracy: 0.9792\n",
      "Epoch 4/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0076 - accuracy: 0.9757 - val_loss: 0.0060 - val_accuracy: 0.9818\n",
      "Epoch 5/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0072 - accuracy: 0.9783 - val_loss: 0.0062 - val_accuracy: 0.9688\n",
      "Epoch 6/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0049 - accuracy: 0.9804 - val_loss: 0.0048 - val_accuracy: 0.9870\n",
      "Epoch 7/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0049 - accuracy: 0.9826 - val_loss: 0.0034 - val_accuracy: 0.9809\n",
      "Epoch 8/600\n",
      "721/721 [==============================] - 25s 35ms/step - loss: 0.0040 - accuracy: 0.9799 - val_loss: 0.0033 - val_accuracy: 0.9896\n",
      "Epoch 9/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 0.0036 - accuracy: 0.9832 - val_loss: 0.0034 - val_accuracy: 0.9783\n",
      "Epoch 10/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0031 - accuracy: 0.9837 - val_loss: 0.0033 - val_accuracy: 0.9844\n",
      "Epoch 11/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0031 - accuracy: 0.9854 - val_loss: 0.0025 - val_accuracy: 0.9887\n",
      "Epoch 12/600\n",
      "721/721 [==============================] - 23s 32ms/step - loss: 0.0026 - accuracy: 0.9847 - val_loss: 0.0035 - val_accuracy: 0.9853\n",
      "Epoch 13/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0026 - accuracy: 0.9859 - val_loss: 0.0020 - val_accuracy: 0.9896\n",
      "Epoch 14/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0024 - accuracy: 0.9875 - val_loss: 0.0022 - val_accuracy: 0.9835\n",
      "Epoch 15/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 0.0025 - accuracy: 0.9880 - val_loss: 0.0019 - val_accuracy: 0.9887\n",
      "Epoch 16/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0021 - accuracy: 0.9868 - val_loss: 0.0022 - val_accuracy: 0.9887\n",
      "Epoch 17/600\n",
      "721/721 [==============================] - 29s 40ms/step - loss: 0.0021 - accuracy: 0.9877 - val_loss: 0.0018 - val_accuracy: 0.9879\n",
      "Epoch 18/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0022 - accuracy: 0.9880 - val_loss: 0.0020 - val_accuracy: 0.9853\n",
      "Epoch 19/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0019 - accuracy: 0.9892 - val_loss: 0.0019 - val_accuracy: 0.9844\n",
      "Epoch 20/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0019 - accuracy: 0.9905 - val_loss: 0.0015 - val_accuracy: 0.9861\n",
      "Epoch 21/600\n",
      "721/721 [==============================] - 29s 40ms/step - loss: 0.0018 - accuracy: 0.9896 - val_loss: 0.0014 - val_accuracy: 0.9896\n",
      "Epoch 22/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0017 - accuracy: 0.9889 - val_loss: 0.0016 - val_accuracy: 0.9887\n",
      "Epoch 23/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0017 - accuracy: 0.9899 - val_loss: 0.0016 - val_accuracy: 0.9861\n",
      "Epoch 24/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0017 - accuracy: 0.9872 - val_loss: 0.0015 - val_accuracy: 0.9896\n",
      "Epoch 25/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0015 - accuracy: 0.9898 - val_loss: 0.0013 - val_accuracy: 0.9861\n",
      "Epoch 26/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0016 - accuracy: 0.9892 - val_loss: 0.0022 - val_accuracy: 0.9853\n",
      "Epoch 27/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 0.0014 - accuracy: 0.9905 - val_loss: 0.0014 - val_accuracy: 0.9905\n",
      "Epoch 28/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0014 - accuracy: 0.9892 - val_loss: 0.0011 - val_accuracy: 0.9922\n",
      "Epoch 29/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0014 - accuracy: 0.9892 - val_loss: 0.0011 - val_accuracy: 0.9905\n",
      "Epoch 30/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0014 - accuracy: 0.9915 - val_loss: 9.8217e-04 - val_accuracy: 0.9957\n",
      "Epoch 31/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0014 - accuracy: 0.9910 - val_loss: 0.0013 - val_accuracy: 0.9896\n",
      "Epoch 32/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0013 - accuracy: 0.9892 - val_loss: 9.5279e-04 - val_accuracy: 0.9913\n",
      "Epoch 33/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0013 - accuracy: 0.9915 - val_loss: 0.0012 - val_accuracy: 0.9922\n",
      "Epoch 34/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0012 - accuracy: 0.9903 - val_loss: 0.0011 - val_accuracy: 0.9922\n",
      "Epoch 35/600\n",
      "721/721 [==============================] - 28s 39ms/step - loss: 0.0013 - accuracy: 0.9910 - val_loss: 0.0012 - val_accuracy: 0.9957\n",
      "Epoch 36/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0012 - accuracy: 0.9920 - val_loss: 0.0011 - val_accuracy: 0.9965\n",
      "Epoch 37/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0012 - accuracy: 0.9915 - val_loss: 0.0011 - val_accuracy: 0.9922\n",
      "Epoch 38/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0012 - accuracy: 0.9896 - val_loss: 0.0011 - val_accuracy: 0.9905\n",
      "Epoch 39/600\n",
      "721/721 [==============================] - 28s 39ms/step - loss: 0.0011 - accuracy: 0.9908 - val_loss: 0.0013 - val_accuracy: 0.9922\n",
      "Epoch 40/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 0.0011 - accuracy: 0.9898 - val_loss: 8.5484e-04 - val_accuracy: 0.9965\n",
      "Epoch 41/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 0.0011 - accuracy: 0.9924 - val_loss: 9.7996e-04 - val_accuracy: 0.9939\n",
      "Epoch 42/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0011 - accuracy: 0.9929 - val_loss: 0.0011 - val_accuracy: 0.9939\n",
      "Epoch 43/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0011 - accuracy: 0.9915 - val_loss: 0.0012 - val_accuracy: 0.9896\n",
      "Epoch 44/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 0.0011 - accuracy: 0.9913 - val_loss: 9.9571e-04 - val_accuracy: 0.9905\n",
      "Epoch 45/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 9.7208e-04 - accuracy: 0.9918 - val_loss: 9.5578e-04 - val_accuracy: 0.9905\n",
      "Epoch 46/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 9.5812e-04 - accuracy: 0.9906 - val_loss: 0.0011 - val_accuracy: 0.9887\n",
      "Epoch 47/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 0.0010 - accuracy: 0.9917 - val_loss: 9.5634e-04 - val_accuracy: 0.9939\n",
      "Epoch 48/600\n",
      "721/721 [==============================] - 24s 33ms/step - loss: 0.0011 - accuracy: 0.9917 - val_loss: 0.0015 - val_accuracy: 0.9861\n",
      "Epoch 49/600\n",
      "721/721 [==============================] - 25s 34ms/step - loss: 9.4960e-04 - accuracy: 0.9918 - val_loss: 9.2967e-04 - val_accuracy: 0.9896\n",
      "Epoch 50/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 9.7588e-04 - accuracy: 0.9918 - val_loss: 0.0010 - val_accuracy: 0.9931\n",
      "Epoch 51/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 9.7537e-04 - accuracy: 0.9943 - val_loss: 0.0010 - val_accuracy: 0.9861\n",
      "Epoch 52/600\n",
      "721/721 [==============================] - 29s 40ms/step - loss: 9.2705e-04 - accuracy: 0.9934 - val_loss: 8.6923e-04 - val_accuracy: 0.9948\n",
      "Epoch 53/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 9.0479e-04 - accuracy: 0.9922 - val_loss: 7.9498e-04 - val_accuracy: 0.9922\n",
      "Epoch 54/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 9.3518e-04 - accuracy: 0.9918 - val_loss: 8.6367e-04 - val_accuracy: 0.9957\n",
      "Epoch 55/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 8.8195e-04 - accuracy: 0.9927 - val_loss: 9.1022e-04 - val_accuracy: 0.9896\n",
      "Epoch 56/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 27s 37ms/step - loss: 9.0884e-04 - accuracy: 0.9913 - val_loss: 8.0402e-04 - val_accuracy: 0.9939\n",
      "Epoch 57/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 8.5254e-04 - accuracy: 0.9939 - val_loss: 8.7878e-04 - val_accuracy: 0.9922\n",
      "Epoch 58/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 8.5850e-04 - accuracy: 0.9908 - val_loss: 7.2192e-04 - val_accuracy: 0.9948\n",
      "Epoch 59/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 8.4445e-04 - accuracy: 0.9925 - val_loss: 8.4842e-04 - val_accuracy: 0.9939\n",
      "Epoch 60/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 8.3308e-04 - accuracy: 0.9922 - val_loss: 8.6926e-04 - val_accuracy: 0.9931\n",
      "Epoch 61/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 8.2539e-04 - accuracy: 0.9920 - val_loss: 9.0056e-04 - val_accuracy: 0.9931\n",
      "Epoch 62/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 8.0073e-04 - accuracy: 0.9925 - val_loss: 7.5719e-04 - val_accuracy: 0.9939\n",
      "Epoch 63/600\n",
      "721/721 [==============================] - 26s 37ms/step - loss: 8.3161e-04 - accuracy: 0.9929 - val_loss: 8.2994e-04 - val_accuracy: 0.9905\n",
      "Epoch 64/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 8.1060e-04 - accuracy: 0.9927 - val_loss: 6.4372e-04 - val_accuracy: 0.9948\n",
      "Epoch 65/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 8.2398e-04 - accuracy: 0.9936 - val_loss: 7.9577e-04 - val_accuracy: 0.9931\n",
      "Epoch 66/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.8846e-04 - accuracy: 0.9931 - val_loss: 6.8456e-04 - val_accuracy: 0.9965\n",
      "Epoch 67/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 8.2754e-04 - accuracy: 0.9929 - val_loss: 9.0310e-04 - val_accuracy: 0.9887\n",
      "Epoch 68/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.5685e-04 - accuracy: 0.9931 - val_loss: 6.5605e-04 - val_accuracy: 0.9939\n",
      "Epoch 69/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 8.2980e-04 - accuracy: 0.9934 - val_loss: 7.4715e-04 - val_accuracy: 0.9957\n",
      "Epoch 70/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.9491e-04 - accuracy: 0.9931 - val_loss: 6.8007e-04 - val_accuracy: 0.9957\n",
      "Epoch 71/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.6762e-04 - accuracy: 0.9934 - val_loss: 8.2908e-04 - val_accuracy: 0.9965\n",
      "Epoch 72/600\n",
      "721/721 [==============================] - 26s 36ms/step - loss: 7.5859e-04 - accuracy: 0.9932 - val_loss: 8.4682e-04 - val_accuracy: 0.9948\n",
      "Epoch 73/600\n",
      "721/721 [==============================] - 28s 39ms/step - loss: 7.5856e-04 - accuracy: 0.9934 - val_loss: 6.3920e-04 - val_accuracy: 0.9948\n",
      "Epoch 74/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 7.4915e-04 - accuracy: 0.9943 - val_loss: 7.7575e-04 - val_accuracy: 0.9957\n",
      "Epoch 75/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 7.4166e-04 - accuracy: 0.9931 - val_loss: 7.7847e-04 - val_accuracy: 0.9939\n",
      "Epoch 76/600\n",
      "721/721 [==============================] - 28s 39ms/step - loss: 7.4867e-04 - accuracy: 0.9938 - val_loss: 6.5196e-04 - val_accuracy: 0.9948\n",
      "Epoch 77/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 6.9569e-04 - accuracy: 0.9950 - val_loss: 7.4968e-04 - val_accuracy: 0.9948\n",
      "Epoch 78/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 7.0681e-04 - accuracy: 0.9938 - val_loss: 7.2124e-04 - val_accuracy: 0.9939\n",
      "Epoch 79/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 6.9370e-04 - accuracy: 0.9944 - val_loss: 7.0239e-04 - val_accuracy: 0.9939\n",
      "Epoch 80/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.0189e-04 - accuracy: 0.9932 - val_loss: 6.4635e-04 - val_accuracy: 0.9922\n",
      "Epoch 81/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 6.9832e-04 - accuracy: 0.9943 - val_loss: 5.7072e-04 - val_accuracy: 0.9974\n",
      "Epoch 82/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.1794e-04 - accuracy: 0.9951 - val_loss: 8.7077e-04 - val_accuracy: 0.9896\n",
      "Epoch 83/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 6.8501e-04 - accuracy: 0.9941 - val_loss: 6.8784e-04 - val_accuracy: 0.9913\n",
      "Epoch 84/600\n",
      "721/721 [==============================] - 27s 37ms/step - loss: 7.1115e-04 - accuracy: 0.9934 - val_loss: 6.4557e-04 - val_accuracy: 0.9931\n",
      "Epoch 85/600\n",
      "721/721 [==============================] - 27s 38ms/step - loss: 6.6472e-04 - accuracy: 0.9943 - val_loss: 6.4679e-04 - val_accuracy: 0.9931\n",
      "Epoch 86/600\n",
      "721/721 [==============================] - 28s 38ms/step - loss: 6.5591e-04 - accuracy: 0.9939 - val_loss: 7.6707e-04 - val_accuracy: 0.9896\n",
      "Epoch 87/600\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 6.7018e-04 - accuracy: 0.9931 - val_loss: 6.4043e-04 - val_accuracy: 0.9948\n",
      "Epoch 88/600\n",
      "721/721 [==============================] - 22s 31ms/step - loss: 6.8739e-04 - accuracy: 0.9934 - val_loss: 7.1860e-04 - val_accuracy: 0.9948\n",
      "Epoch 89/600\n",
      "721/721 [==============================] - 20s 28ms/step - loss: 6.6725e-04 - accuracy: 0.9948 - val_loss: 6.9487e-04 - val_accuracy: 0.9922\n",
      "Epoch 90/600\n",
      "721/721 [==============================] - 16s 22ms/step - loss: 6.8837e-04 - accuracy: 0.9927 - val_loss: 6.3168e-04 - val_accuracy: 0.9957\n",
      "Epoch 91/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.7107e-04 - accuracy: 0.9939 - val_loss: 8.1628e-04 - val_accuracy: 0.9939\n",
      "Epoch 92/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.4355e-04 - accuracy: 0.9941 - val_loss: 9.6830e-04 - val_accuracy: 0.9957\n",
      "Epoch 93/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.6249e-04 - accuracy: 0.9938 - val_loss: 6.2333e-04 - val_accuracy: 0.9957\n",
      "Epoch 94/600\n",
      "721/721 [==============================] - 13s 18ms/step - loss: 6.4711e-04 - accuracy: 0.9953 - val_loss: 6.0043e-04 - val_accuracy: 0.9965\n",
      "Epoch 95/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.8268e-04 - accuracy: 0.9941 - val_loss: 5.0516e-04 - val_accuracy: 0.9965\n",
      "Epoch 96/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 6.5082e-04 - accuracy: 0.9932 - val_loss: 6.8075e-04 - val_accuracy: 0.9965\n",
      "Epoch 97/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.5808e-04 - accuracy: 0.9951 - val_loss: 5.9892e-04 - val_accuracy: 0.9965\n",
      "Epoch 98/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.2092e-04 - accuracy: 0.9941 - val_loss: 6.5886e-04 - val_accuracy: 0.9948\n",
      "Epoch 99/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.3753e-04 - accuracy: 0.9938 - val_loss: 6.3740e-04 - val_accuracy: 0.9957\n",
      "Epoch 100/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 6.1342e-04 - accuracy: 0.9950 - val_loss: 6.1235e-04 - val_accuracy: 0.9905\n",
      "Epoch 101/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.1207e-04 - accuracy: 0.9943 - val_loss: 5.1973e-04 - val_accuracy: 0.9957\n",
      "Epoch 102/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.2135e-04 - accuracy: 0.9943 - val_loss: 6.4327e-04 - val_accuracy: 0.9957\n",
      "Epoch 103/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.2773e-04 - accuracy: 0.9957 - val_loss: 8.2965e-04 - val_accuracy: 0.9861\n",
      "Epoch 104/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 6.2201e-04 - accuracy: 0.9941 - val_loss: 5.2659e-04 - val_accuracy: 0.9948\n",
      "Epoch 105/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.9405e-04 - accuracy: 0.9944 - val_loss: 5.8419e-04 - val_accuracy: 0.9939\n",
      "Epoch 106/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.0731e-04 - accuracy: 0.9950 - val_loss: 5.4646e-04 - val_accuracy: 0.9931\n",
      "Epoch 107/600\n",
      "721/721 [==============================] - 13s 18ms/step - loss: 6.2541e-04 - accuracy: 0.9944 - val_loss: 5.1950e-04 - val_accuracy: 0.9948\n",
      "Epoch 108/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.1999e-04 - accuracy: 0.9951 - val_loss: 5.5855e-04 - val_accuracy: 0.9965\n",
      "Epoch 109/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 12s 17ms/step - loss: 6.0306e-04 - accuracy: 0.9955 - val_loss: 6.3536e-04 - val_accuracy: 0.9957\n",
      "Epoch 110/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.9218e-04 - accuracy: 0.9955 - val_loss: 5.1442e-04 - val_accuracy: 0.9957\n",
      "Epoch 111/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 6.0393e-04 - accuracy: 0.9941 - val_loss: 5.1769e-04 - val_accuracy: 0.9948\n",
      "Epoch 112/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.7375e-04 - accuracy: 0.9950 - val_loss: 5.9590e-04 - val_accuracy: 0.9922\n",
      "Epoch 113/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.7272e-04 - accuracy: 0.9972 - val_loss: 5.3533e-04 - val_accuracy: 0.9948\n",
      "Epoch 114/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.8414e-04 - accuracy: 0.9922 - val_loss: 5.4933e-04 - val_accuracy: 0.9965\n",
      "Epoch 115/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.8128e-04 - accuracy: 0.9939 - val_loss: 6.0197e-04 - val_accuracy: 0.9931\n",
      "Epoch 116/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.6482e-04 - accuracy: 0.9920 - val_loss: 5.7101e-04 - val_accuracy: 0.9913\n",
      "Epoch 117/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 6.0489e-04 - accuracy: 0.9936 - val_loss: 5.2335e-04 - val_accuracy: 0.9965\n",
      "Epoch 118/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.6071e-04 - accuracy: 0.9936 - val_loss: 5.5327e-04 - val_accuracy: 0.9879\n",
      "Epoch 119/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4980e-04 - accuracy: 0.9941 - val_loss: 4.8149e-04 - val_accuracy: 0.9957\n",
      "Epoch 120/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.8753e-04 - accuracy: 0.9950 - val_loss: 7.5026e-04 - val_accuracy: 0.9922\n",
      "Epoch 121/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4306e-04 - accuracy: 0.9955 - val_loss: 7.2258e-04 - val_accuracy: 0.9983\n",
      "Epoch 122/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.8745e-04 - accuracy: 0.9946 - val_loss: 6.0584e-04 - val_accuracy: 0.9939\n",
      "Epoch 123/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.5642e-04 - accuracy: 0.9948 - val_loss: 5.7195e-04 - val_accuracy: 0.9835\n",
      "Epoch 124/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4673e-04 - accuracy: 0.9934 - val_loss: 5.1136e-04 - val_accuracy: 0.9905\n",
      "Epoch 125/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.7429e-04 - accuracy: 0.9948 - val_loss: 4.9587e-04 - val_accuracy: 0.9957\n",
      "Epoch 126/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.5493e-04 - accuracy: 0.9948 - val_loss: 6.4048e-04 - val_accuracy: 0.9931\n",
      "Epoch 127/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4867e-04 - accuracy: 0.9955 - val_loss: 5.7001e-04 - val_accuracy: 0.9939\n",
      "Epoch 128/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.6223e-04 - accuracy: 0.9939 - val_loss: 6.8562e-04 - val_accuracy: 0.9931\n",
      "Epoch 129/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.4968e-04 - accuracy: 0.9946 - val_loss: 4.7401e-04 - val_accuracy: 0.9948\n",
      "Epoch 130/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.5677e-04 - accuracy: 0.9939 - val_loss: 5.0952e-04 - val_accuracy: 0.9939\n",
      "Epoch 131/600\n",
      "721/721 [==============================] - 13s 17ms/step - loss: 5.4546e-04 - accuracy: 0.9953 - val_loss: 5.8157e-04 - val_accuracy: 0.9931\n",
      "Epoch 132/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.7397e-04 - accuracy: 0.9955 - val_loss: 5.4246e-04 - val_accuracy: 0.9957\n",
      "Epoch 133/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.2990e-04 - accuracy: 0.9946 - val_loss: 6.3772e-04 - val_accuracy: 0.9922\n",
      "Epoch 134/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.5614e-04 - accuracy: 0.9938 - val_loss: 5.5508e-04 - val_accuracy: 0.9974\n",
      "Epoch 135/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.6121e-04 - accuracy: 0.9948 - val_loss: 5.1950e-04 - val_accuracy: 0.9931\n",
      "Epoch 136/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4486e-04 - accuracy: 0.9962 - val_loss: 4.8626e-04 - val_accuracy: 0.9965\n",
      "Epoch 137/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.5628e-04 - accuracy: 0.9962 - val_loss: 6.2821e-04 - val_accuracy: 0.9939\n",
      "Epoch 138/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4478e-04 - accuracy: 0.9946 - val_loss: 5.7106e-04 - val_accuracy: 0.9948\n",
      "Epoch 139/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.4637e-04 - accuracy: 0.9938 - val_loss: 4.4362e-04 - val_accuracy: 0.9948\n",
      "Epoch 140/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.3390e-04 - accuracy: 0.9944 - val_loss: 4.7208e-04 - val_accuracy: 0.9939\n",
      "Epoch 141/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.2265e-04 - accuracy: 0.9958 - val_loss: 5.5096e-04 - val_accuracy: 0.9913\n",
      "Epoch 142/600\n",
      "721/721 [==============================] - 12s 17ms/step - loss: 5.1694e-04 - accuracy: 0.9955 - val_loss: 5.0389e-04 - val_accuracy: 0.9957\n",
      "Epoch 143/600\n",
      "721/721 [==============================] - 12s 16ms/step - loss: 5.0618e-04 - accuracy: 0.9964 - val_loss: 6.5282e-04 - val_accuracy: 0.9939\n",
      "Epoch 144/600\n",
      "721/721 [==============================] - 11s 15ms/step - loss: 5.1176e-04 - accuracy: 0.9939 - val_loss: 4.7544e-04 - val_accuracy: 0.9965\n",
      "Epoch 145/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 5.2304e-04 - accuracy: 0.9965 - val_loss: 5.0715e-04 - val_accuracy: 0.9922\n",
      "Epoch 146/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 5.1694e-04 - accuracy: 0.9951 - val_loss: 5.0106e-04 - val_accuracy: 0.9939\n",
      "Epoch 147/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.3363e-04 - accuracy: 0.9936 - val_loss: 4.9533e-04 - val_accuracy: 0.9965\n",
      "Epoch 148/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.2847e-04 - accuracy: 0.9953 - val_loss: 4.8253e-04 - val_accuracy: 0.9974\n",
      "Epoch 149/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 5.2792e-04 - accuracy: 0.9962 - val_loss: 5.9663e-04 - val_accuracy: 0.9957\n",
      "Epoch 150/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.2247e-04 - accuracy: 0.9950 - val_loss: 5.0299e-04 - val_accuracy: 0.9922\n",
      "Epoch 151/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 5.0651e-04 - accuracy: 0.9953 - val_loss: 4.3699e-04 - val_accuracy: 0.9965\n",
      "Epoch 152/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.1969e-04 - accuracy: 0.9960 - val_loss: 4.3940e-04 - val_accuracy: 0.9931\n",
      "Epoch 153/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.0106e-04 - accuracy: 0.9964 - val_loss: 4.5845e-04 - val_accuracy: 0.9922\n",
      "Epoch 154/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.1204e-04 - accuracy: 0.9941 - val_loss: 4.8189e-04 - val_accuracy: 0.9965\n",
      "Epoch 155/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.2393e-04 - accuracy: 0.9957 - val_loss: 4.9382e-04 - val_accuracy: 0.9965\n",
      "Epoch 156/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 5.0988e-04 - accuracy: 0.9953 - val_loss: 5.3079e-04 - val_accuracy: 0.9948\n",
      "Epoch 157/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9551e-04 - accuracy: 0.9955 - val_loss: 4.3854e-04 - val_accuracy: 0.9948\n",
      "Epoch 158/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.0574e-04 - accuracy: 0.9948 - val_loss: 4.2850e-04 - val_accuracy: 0.9965\n",
      "Epoch 159/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 5.0481e-04 - accuracy: 0.9951 - val_loss: 4.4859e-04 - val_accuracy: 0.9983\n",
      "Epoch 160/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9441e-04 - accuracy: 0.9957 - val_loss: 4.2628e-04 - val_accuracy: 0.9957\n",
      "Epoch 161/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.8674e-04 - accuracy: 0.9951 - val_loss: 4.9114e-04 - val_accuracy: 0.9957\n",
      "Epoch 162/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 5.0155e-04 - accuracy: 0.9962 - val_loss: 4.6241e-04 - val_accuracy: 0.9974\n",
      "Epoch 163/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 5.0047e-04 - accuracy: 0.9970 - val_loss: 4.9512e-04 - val_accuracy: 0.9991\n",
      "Epoch 164/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7911e-04 - accuracy: 0.9958 - val_loss: 4.8826e-04 - val_accuracy: 0.9887\n",
      "Epoch 165/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 5.0449e-04 - accuracy: 0.9953 - val_loss: 4.9133e-04 - val_accuracy: 0.9931\n",
      "Epoch 166/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9675e-04 - accuracy: 0.9962 - val_loss: 5.1841e-04 - val_accuracy: 0.9957\n",
      "Epoch 167/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 5.0528e-04 - accuracy: 0.9965 - val_loss: 5.4107e-04 - val_accuracy: 0.9913\n",
      "Epoch 168/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.8258e-04 - accuracy: 0.9957 - val_loss: 5.2363e-04 - val_accuracy: 0.9948\n",
      "Epoch 169/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9333e-04 - accuracy: 0.9950 - val_loss: 5.4954e-04 - val_accuracy: 0.9879\n",
      "Epoch 170/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9555e-04 - accuracy: 0.9967 - val_loss: 4.7386e-04 - val_accuracy: 0.9957\n",
      "Epoch 171/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6926e-04 - accuracy: 0.9958 - val_loss: 4.6867e-04 - val_accuracy: 0.9948\n",
      "Epoch 172/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.9741e-04 - accuracy: 0.9950 - val_loss: 4.6184e-04 - val_accuracy: 0.9983\n",
      "Epoch 173/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6460e-04 - accuracy: 0.9962 - val_loss: 5.4861e-04 - val_accuracy: 0.9939\n",
      "Epoch 174/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 5.0546e-04 - accuracy: 0.9957 - val_loss: 4.7492e-04 - val_accuracy: 0.9965\n",
      "Epoch 175/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6991e-04 - accuracy: 0.9964 - val_loss: 4.4356e-04 - val_accuracy: 0.9974\n",
      "Epoch 176/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.8421e-04 - accuracy: 0.9967 - val_loss: 5.1032e-04 - val_accuracy: 0.9922\n",
      "Epoch 177/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7477e-04 - accuracy: 0.9944 - val_loss: 5.1298e-04 - val_accuracy: 0.9939\n",
      "Epoch 178/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.8124e-04 - accuracy: 0.9939 - val_loss: 4.1122e-04 - val_accuracy: 0.9922\n",
      "Epoch 179/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7569e-04 - accuracy: 0.9965 - val_loss: 4.3510e-04 - val_accuracy: 0.9957\n",
      "Epoch 180/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6698e-04 - accuracy: 0.9957 - val_loss: 5.3288e-04 - val_accuracy: 0.9939\n",
      "Epoch 181/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.5698e-04 - accuracy: 0.9969 - val_loss: 5.9505e-04 - val_accuracy: 0.9957\n",
      "Epoch 182/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4907e-04 - accuracy: 0.9960 - val_loss: 5.2186e-04 - val_accuracy: 0.9957\n",
      "Epoch 183/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4862e-04 - accuracy: 0.9960 - val_loss: 4.1868e-04 - val_accuracy: 0.9939\n",
      "Epoch 184/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7630e-04 - accuracy: 0.9943 - val_loss: 4.9642e-04 - val_accuracy: 0.9974\n",
      "Epoch 185/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7356e-04 - accuracy: 0.9965 - val_loss: 4.3216e-04 - val_accuracy: 0.9957\n",
      "Epoch 186/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5645e-04 - accuracy: 0.9960 - val_loss: 4.5152e-04 - val_accuracy: 0.9905\n",
      "Epoch 187/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5759e-04 - accuracy: 0.9955 - val_loss: 4.2974e-04 - val_accuracy: 0.9974\n",
      "Epoch 188/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6234e-04 - accuracy: 0.9950 - val_loss: 4.4550e-04 - val_accuracy: 0.9931\n",
      "Epoch 189/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5405e-04 - accuracy: 0.9970 - val_loss: 4.2442e-04 - val_accuracy: 0.9965\n",
      "Epoch 190/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4330e-04 - accuracy: 0.9962 - val_loss: 4.7416e-04 - val_accuracy: 0.9974\n",
      "Epoch 191/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5302e-04 - accuracy: 0.9958 - val_loss: 5.1046e-04 - val_accuracy: 0.9896\n",
      "Epoch 192/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5092e-04 - accuracy: 0.9955 - val_loss: 4.1212e-04 - val_accuracy: 0.9887\n",
      "Epoch 193/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.6281e-04 - accuracy: 0.9955 - val_loss: 4.3498e-04 - val_accuracy: 0.9974\n",
      "Epoch 194/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5306e-04 - accuracy: 0.9964 - val_loss: 4.0260e-04 - val_accuracy: 0.9957\n",
      "Epoch 195/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.7112e-04 - accuracy: 0.9958 - val_loss: 4.2984e-04 - val_accuracy: 0.9939\n",
      "Epoch 196/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4503e-04 - accuracy: 0.9957 - val_loss: 4.2281e-04 - val_accuracy: 0.9965\n",
      "Epoch 197/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4081e-04 - accuracy: 0.9976 - val_loss: 5.0372e-04 - val_accuracy: 0.9974\n",
      "Epoch 198/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5979e-04 - accuracy: 0.9962 - val_loss: 4.1966e-04 - val_accuracy: 0.9965\n",
      "Epoch 199/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5358e-04 - accuracy: 0.9965 - val_loss: 4.5571e-04 - val_accuracy: 0.9948\n",
      "Epoch 200/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5060e-04 - accuracy: 0.9962 - val_loss: 3.7569e-04 - val_accuracy: 0.9965\n",
      "Epoch 201/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4833e-04 - accuracy: 0.9957 - val_loss: 4.1794e-04 - val_accuracy: 0.9931\n",
      "Epoch 202/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3797e-04 - accuracy: 0.9960 - val_loss: 9.6062e-04 - val_accuracy: 0.9913\n",
      "Epoch 203/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3676e-04 - accuracy: 0.9965 - val_loss: 4.5890e-04 - val_accuracy: 0.9965\n",
      "Epoch 204/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4689e-04 - accuracy: 0.9964 - val_loss: 3.8567e-04 - val_accuracy: 0.9939\n",
      "Epoch 205/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.3963e-04 - accuracy: 0.9958 - val_loss: 4.2321e-04 - val_accuracy: 0.9983\n",
      "Epoch 206/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3918e-04 - accuracy: 0.9962 - val_loss: 3.8616e-04 - val_accuracy: 0.9957\n",
      "Epoch 207/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4982e-04 - accuracy: 0.9951 - val_loss: 3.9890e-04 - val_accuracy: 0.9948\n",
      "Epoch 208/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5022e-04 - accuracy: 0.9950 - val_loss: 4.2684e-04 - val_accuracy: 0.9957\n",
      "Epoch 209/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3423e-04 - accuracy: 0.9960 - val_loss: 4.3354e-04 - val_accuracy: 0.9939\n",
      "Epoch 210/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3747e-04 - accuracy: 0.9955 - val_loss: 4.0163e-04 - val_accuracy: 0.9931\n",
      "Epoch 211/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2793e-04 - accuracy: 0.9950 - val_loss: 5.8123e-04 - val_accuracy: 0.9913\n",
      "Epoch 212/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2231e-04 - accuracy: 0.9967 - val_loss: 3.6892e-04 - val_accuracy: 0.9974\n",
      "Epoch 213/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.5002e-04 - accuracy: 0.9965 - val_loss: 5.2617e-04 - val_accuracy: 0.9931\n",
      "Epoch 214/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4357e-04 - accuracy: 0.9957 - val_loss: 3.0985e-04 - val_accuracy: 0.9983\n",
      "Epoch 215/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0733e-04 - accuracy: 0.9964 - val_loss: 4.3648e-04 - val_accuracy: 0.9974\n",
      "Epoch 216/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3433e-04 - accuracy: 0.9967 - val_loss: 5.1075e-04 - val_accuracy: 0.9948\n",
      "Epoch 217/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.6541e-04 - accuracy: 0.9951 - val_loss: 3.7633e-04 - val_accuracy: 0.9957\n",
      "Epoch 218/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.2788e-04 - accuracy: 0.9951 - val_loss: 3.8249e-04 - val_accuracy: 0.9948\n",
      "Epoch 219/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.3635e-04 - accuracy: 0.9957 - val_loss: 4.1666e-04 - val_accuracy: 0.9974\n",
      "Epoch 220/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3898e-04 - accuracy: 0.9965 - val_loss: 6.6150e-04 - val_accuracy: 0.9957\n",
      "Epoch 221/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.4098e-04 - accuracy: 0.9943 - val_loss: 3.8564e-04 - val_accuracy: 0.9974\n",
      "Epoch 222/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3881e-04 - accuracy: 0.9967 - val_loss: 4.3338e-04 - val_accuracy: 0.9965\n",
      "Epoch 223/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 4.3347e-04 - accuracy: 0.9960 - val_loss: 4.2087e-04 - val_accuracy: 0.9957\n",
      "Epoch 224/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.2408e-04 - accuracy: 0.9953 - val_loss: 3.7493e-04 - val_accuracy: 0.9965\n",
      "Epoch 225/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2127e-04 - accuracy: 0.9955 - val_loss: 4.2348e-04 - val_accuracy: 0.9948\n",
      "Epoch 226/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3136e-04 - accuracy: 0.9965 - val_loss: 3.8869e-04 - val_accuracy: 0.9983\n",
      "Epoch 227/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2726e-04 - accuracy: 0.9965 - val_loss: 3.6792e-04 - val_accuracy: 0.9965\n",
      "Epoch 228/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2678e-04 - accuracy: 0.9948 - val_loss: 3.8434e-04 - val_accuracy: 0.9948\n",
      "Epoch 229/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.2656e-04 - accuracy: 0.9967 - val_loss: 3.6224e-04 - val_accuracy: 0.9957\n",
      "Epoch 230/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.3707e-04 - accuracy: 0.9957 - val_loss: 3.7596e-04 - val_accuracy: 0.9965\n",
      "Epoch 231/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.2108e-04 - accuracy: 0.9951 - val_loss: 4.2783e-04 - val_accuracy: 0.9957\n",
      "Epoch 232/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3945e-04 - accuracy: 0.9953 - val_loss: 5.1060e-04 - val_accuracy: 0.9983\n",
      "Epoch 233/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2312e-04 - accuracy: 0.9964 - val_loss: 3.9681e-04 - val_accuracy: 0.9948\n",
      "Epoch 234/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3107e-04 - accuracy: 0.9955 - val_loss: 5.3508e-04 - val_accuracy: 0.9957\n",
      "Epoch 235/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0979e-04 - accuracy: 0.9969 - val_loss: 3.7449e-04 - val_accuracy: 0.9957\n",
      "Epoch 236/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.3601e-04 - accuracy: 0.9970 - val_loss: 4.4834e-04 - val_accuracy: 0.9939\n",
      "Epoch 237/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2411e-04 - accuracy: 0.9953 - val_loss: 3.6463e-04 - val_accuracy: 0.9957\n",
      "Epoch 238/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0893e-04 - accuracy: 0.9964 - val_loss: 3.8059e-04 - val_accuracy: 0.9957\n",
      "Epoch 239/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2081e-04 - accuracy: 0.9960 - val_loss: 3.7423e-04 - val_accuracy: 0.9974\n",
      "Epoch 240/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.0710e-04 - accuracy: 0.9965 - val_loss: 4.5822e-04 - val_accuracy: 0.9983\n",
      "Epoch 241/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1225e-04 - accuracy: 0.9958 - val_loss: 3.2684e-04 - val_accuracy: 0.9957\n",
      "Epoch 242/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0047e-04 - accuracy: 0.9964 - val_loss: 4.4605e-04 - val_accuracy: 0.9965\n",
      "Epoch 243/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1333e-04 - accuracy: 0.9955 - val_loss: 3.7224e-04 - val_accuracy: 0.9965\n",
      "Epoch 244/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1501e-04 - accuracy: 0.9967 - val_loss: 3.8403e-04 - val_accuracy: 0.9948\n",
      "Epoch 245/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0422e-04 - accuracy: 0.9955 - val_loss: 3.2822e-04 - val_accuracy: 0.9983\n",
      "Epoch 246/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0566e-04 - accuracy: 0.9964 - val_loss: 3.6951e-04 - val_accuracy: 0.9965\n",
      "Epoch 247/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.2440e-04 - accuracy: 0.9962 - val_loss: 3.2771e-04 - val_accuracy: 0.9983\n",
      "Epoch 248/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1978e-04 - accuracy: 0.9965 - val_loss: 5.0774e-04 - val_accuracy: 0.9974\n",
      "Epoch 249/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1689e-04 - accuracy: 0.9974 - val_loss: 3.0705e-04 - val_accuracy: 0.9974\n",
      "Epoch 250/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.0920e-04 - accuracy: 0.9967 - val_loss: 3.7767e-04 - val_accuracy: 0.9957\n",
      "Epoch 251/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0607e-04 - accuracy: 0.9960 - val_loss: 4.2339e-04 - val_accuracy: 0.9974\n",
      "Epoch 252/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1887e-04 - accuracy: 0.9964 - val_loss: 3.5446e-04 - val_accuracy: 0.9957\n",
      "Epoch 253/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0227e-04 - accuracy: 0.9970 - val_loss: 4.7298e-04 - val_accuracy: 0.9991\n",
      "Epoch 254/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.1416e-04 - accuracy: 0.9969 - val_loss: 4.0470e-04 - val_accuracy: 0.9948\n",
      "Epoch 255/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.8456e-04 - accuracy: 0.9964 - val_loss: 3.3155e-04 - val_accuracy: 0.9948\n",
      "Epoch 256/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.9567e-04 - accuracy: 0.9969 - val_loss: 3.9880e-04 - val_accuracy: 0.9957\n",
      "Epoch 257/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.9538e-04 - accuracy: 0.9979 - val_loss: 4.6140e-04 - val_accuracy: 0.9974\n",
      "Epoch 258/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.1881e-04 - accuracy: 0.9969 - val_loss: 4.0452e-04 - val_accuracy: 0.9983\n",
      "Epoch 259/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0341e-04 - accuracy: 0.9960 - val_loss: 4.3345e-04 - val_accuracy: 0.9931\n",
      "Epoch 260/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.0652e-04 - accuracy: 0.9957 - val_loss: 3.6808e-04 - val_accuracy: 0.9965\n",
      "Epoch 261/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 4.0108e-04 - accuracy: 0.9962 - val_loss: 3.1895e-04 - val_accuracy: 0.9983\n",
      "Epoch 262/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9992e-04 - accuracy: 0.9969 - val_loss: 4.0631e-04 - val_accuracy: 0.9965\n",
      "Epoch 263/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0320e-04 - accuracy: 0.9955 - val_loss: 3.1983e-04 - val_accuracy: 0.9974\n",
      "Epoch 264/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 4.0785e-04 - accuracy: 0.9957 - val_loss: 5.9185e-04 - val_accuracy: 0.9939\n",
      "Epoch 265/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.9386e-04 - accuracy: 0.9951 - val_loss: 4.1230e-04 - val_accuracy: 0.9931\n",
      "Epoch 266/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8710e-04 - accuracy: 0.9962 - val_loss: 4.1590e-04 - val_accuracy: 0.9931\n",
      "Epoch 267/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.9464e-04 - accuracy: 0.9967 - val_loss: 4.6494e-04 - val_accuracy: 0.9983\n",
      "Epoch 268/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8756e-04 - accuracy: 0.9958 - val_loss: 3.8555e-04 - val_accuracy: 0.9983\n",
      "Epoch 269/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9617e-04 - accuracy: 0.9958 - val_loss: 3.5900e-04 - val_accuracy: 0.9957\n",
      "Epoch 270/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9981e-04 - accuracy: 0.9951 - val_loss: 3.8510e-04 - val_accuracy: 0.9965\n",
      "Epoch 271/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7818e-04 - accuracy: 0.9967 - val_loss: 3.4793e-04 - val_accuracy: 0.9983\n",
      "Epoch 272/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8851e-04 - accuracy: 0.9969 - val_loss: 3.5742e-04 - val_accuracy: 0.9913\n",
      "Epoch 273/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9432e-04 - accuracy: 0.9960 - val_loss: 4.2198e-04 - val_accuracy: 0.9939\n",
      "Epoch 274/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.8020e-04 - accuracy: 0.9967 - val_loss: 3.6763e-04 - val_accuracy: 0.9983\n",
      "Epoch 275/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9322e-04 - accuracy: 0.9960 - val_loss: 4.4953e-04 - val_accuracy: 0.9957\n",
      "Epoch 276/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9116e-04 - accuracy: 0.9953 - val_loss: 3.7331e-04 - val_accuracy: 0.9983\n",
      "Epoch 277/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9260e-04 - accuracy: 0.9976 - val_loss: 4.5990e-04 - val_accuracy: 0.9991\n",
      "Epoch 278/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8119e-04 - accuracy: 0.9967 - val_loss: 4.7716e-04 - val_accuracy: 0.9965\n",
      "Epoch 279/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 4.1343e-04 - accuracy: 0.9960 - val_loss: 3.4099e-04 - val_accuracy: 0.9965\n",
      "Epoch 280/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8846e-04 - accuracy: 0.9970 - val_loss: 4.1194e-04 - val_accuracy: 0.9957\n",
      "Epoch 281/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8791e-04 - accuracy: 0.9960 - val_loss: 3.0678e-04 - val_accuracy: 0.9957\n",
      "Epoch 282/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8579e-04 - accuracy: 0.9977 - val_loss: 3.6732e-04 - val_accuracy: 0.9939\n",
      "Epoch 283/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9744e-04 - accuracy: 0.9960 - val_loss: 3.5565e-04 - val_accuracy: 0.9991\n",
      "Epoch 284/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9331e-04 - accuracy: 0.9969 - val_loss: 3.0717e-04 - val_accuracy: 0.9974\n",
      "Epoch 285/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9246e-04 - accuracy: 0.9964 - val_loss: 4.0452e-04 - val_accuracy: 0.9939\n",
      "Epoch 286/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.7441e-04 - accuracy: 0.9976 - val_loss: 3.5003e-04 - val_accuracy: 0.9983\n",
      "Epoch 287/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8305e-04 - accuracy: 0.9976 - val_loss: 3.8648e-04 - val_accuracy: 0.9939\n",
      "Epoch 288/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8570e-04 - accuracy: 0.9960 - val_loss: 3.6004e-04 - val_accuracy: 0.9957\n",
      "Epoch 289/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9350e-04 - accuracy: 0.9957 - val_loss: 3.3886e-04 - val_accuracy: 0.9983\n",
      "Epoch 290/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8696e-04 - accuracy: 0.9972 - val_loss: 5.0769e-04 - val_accuracy: 0.9974\n",
      "Epoch 291/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.7653e-04 - accuracy: 0.9974 - val_loss: 3.8402e-04 - val_accuracy: 0.9965\n",
      "Epoch 292/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7212e-04 - accuracy: 0.9976 - val_loss: 4.3257e-04 - val_accuracy: 0.9922\n",
      "Epoch 293/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.8990e-04 - accuracy: 0.9964 - val_loss: 3.3732e-04 - val_accuracy: 0.9931\n",
      "Epoch 294/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.8138e-04 - accuracy: 0.9955 - val_loss: 4.1565e-04 - val_accuracy: 0.9948\n",
      "Epoch 295/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7235e-04 - accuracy: 0.9967 - val_loss: 3.2715e-04 - val_accuracy: 0.9965\n",
      "Epoch 296/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6643e-04 - accuracy: 0.9957 - val_loss: 3.8340e-04 - val_accuracy: 0.9939\n",
      "Epoch 297/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.9194e-04 - accuracy: 0.9960 - val_loss: 3.7150e-04 - val_accuracy: 0.9974\n",
      "Epoch 298/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7326e-04 - accuracy: 0.9960 - val_loss: 3.1569e-04 - val_accuracy: 0.9974\n",
      "Epoch 299/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6993e-04 - accuracy: 0.9969 - val_loss: 3.0906e-04 - val_accuracy: 0.9965\n",
      "Epoch 300/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8381e-04 - accuracy: 0.9964 - val_loss: 3.7209e-04 - val_accuracy: 0.9983\n",
      "Epoch 301/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8589e-04 - accuracy: 0.9979 - val_loss: 3.5919e-04 - val_accuracy: 0.9974\n",
      "Epoch 302/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5668e-04 - accuracy: 0.9962 - val_loss: 3.7678e-04 - val_accuracy: 0.9983\n",
      "Epoch 303/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.7041e-04 - accuracy: 0.9979 - val_loss: 4.0440e-04 - val_accuracy: 0.9922\n",
      "Epoch 304/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7053e-04 - accuracy: 0.9957 - val_loss: 3.9222e-04 - val_accuracy: 0.9974\n",
      "Epoch 305/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.7892e-04 - accuracy: 0.9970 - val_loss: 3.3305e-04 - val_accuracy: 0.9965\n",
      "Epoch 306/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5381e-04 - accuracy: 0.9967 - val_loss: 3.8975e-04 - val_accuracy: 0.9957\n",
      "Epoch 307/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5958e-04 - accuracy: 0.9965 - val_loss: 3.0045e-04 - val_accuracy: 0.9974\n",
      "Epoch 308/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7198e-04 - accuracy: 0.9981 - val_loss: 2.8211e-04 - val_accuracy: 0.9939\n",
      "Epoch 309/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5628e-04 - accuracy: 0.9972 - val_loss: 4.3572e-04 - val_accuracy: 0.9983\n",
      "Epoch 310/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6256e-04 - accuracy: 0.9969 - val_loss: 3.3489e-04 - val_accuracy: 0.9983\n",
      "Epoch 311/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6499e-04 - accuracy: 0.9970 - val_loss: 3.2374e-04 - val_accuracy: 0.9974\n",
      "Epoch 312/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.8682e-04 - accuracy: 0.9953 - val_loss: 4.2524e-04 - val_accuracy: 0.9965\n",
      "Epoch 313/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6028e-04 - accuracy: 0.9967 - val_loss: 3.0031e-04 - val_accuracy: 0.9974\n",
      "Epoch 314/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5290e-04 - accuracy: 0.9972 - val_loss: 3.0271e-04 - val_accuracy: 0.9974\n",
      "Epoch 315/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7509e-04 - accuracy: 0.9969 - val_loss: 4.5647e-04 - val_accuracy: 0.9948\n",
      "Epoch 316/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6362e-04 - accuracy: 0.9957 - val_loss: 3.5118e-04 - val_accuracy: 1.0000\n",
      "Epoch 317/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.7956e-04 - accuracy: 0.9983 - val_loss: 3.5517e-04 - val_accuracy: 0.9974\n",
      "Epoch 318/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6665e-04 - accuracy: 0.9976 - val_loss: 3.4989e-04 - val_accuracy: 0.9939\n",
      "Epoch 319/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.6516e-04 - accuracy: 0.9960 - val_loss: 3.3159e-04 - val_accuracy: 0.9965\n",
      "Epoch 320/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.6608e-04 - accuracy: 0.9958 - val_loss: 3.4875e-04 - val_accuracy: 0.9948\n",
      "Epoch 321/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5782e-04 - accuracy: 0.9969 - val_loss: 3.4687e-04 - val_accuracy: 0.9974\n",
      "Epoch 322/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6038e-04 - accuracy: 0.9977 - val_loss: 3.4564e-04 - val_accuracy: 0.9957\n",
      "Epoch 323/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7373e-04 - accuracy: 0.9969 - val_loss: 5.7558e-04 - val_accuracy: 0.9974\n",
      "Epoch 324/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7069e-04 - accuracy: 0.9969 - val_loss: 3.7659e-04 - val_accuracy: 0.9983\n",
      "Epoch 325/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5911e-04 - accuracy: 0.9967 - val_loss: 3.4035e-04 - val_accuracy: 0.9957\n",
      "Epoch 326/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6976e-04 - accuracy: 0.9964 - val_loss: 5.0687e-04 - val_accuracy: 0.9931\n",
      "Epoch 327/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5256e-04 - accuracy: 0.9967 - val_loss: 3.0660e-04 - val_accuracy: 0.9965\n",
      "Epoch 328/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5354e-04 - accuracy: 0.9977 - val_loss: 3.4321e-04 - val_accuracy: 0.9974\n",
      "Epoch 329/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.4819e-04 - accuracy: 0.9967 - val_loss: 4.0570e-04 - val_accuracy: 0.9948\n",
      "Epoch 330/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5575e-04 - accuracy: 0.9958 - val_loss: 4.6406e-04 - val_accuracy: 0.9974\n",
      "Epoch 331/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.5989e-04 - accuracy: 0.9972 - val_loss: 2.9694e-04 - val_accuracy: 0.9983\n",
      "Epoch 332/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6212e-04 - accuracy: 0.9967 - val_loss: 3.2072e-04 - val_accuracy: 0.9974\n",
      "Epoch 333/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5563e-04 - accuracy: 0.9979 - val_loss: 2.9517e-04 - val_accuracy: 0.9983\n",
      "Epoch 334/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5264e-04 - accuracy: 0.9962 - val_loss: 3.4452e-04 - val_accuracy: 0.9965\n",
      "Epoch 335/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4554e-04 - accuracy: 0.9962 - val_loss: 3.4597e-04 - val_accuracy: 0.9974\n",
      "Epoch 336/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.6073e-04 - accuracy: 0.9972 - val_loss: 3.4295e-04 - val_accuracy: 0.9965\n",
      "Epoch 337/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.6850e-04 - accuracy: 0.9970 - val_loss: 4.9456e-04 - val_accuracy: 0.9939\n",
      "Epoch 338/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4589e-04 - accuracy: 0.9967 - val_loss: 3.4747e-04 - val_accuracy: 0.9948\n",
      "Epoch 339/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.4771e-04 - accuracy: 0.9965 - val_loss: 3.5318e-04 - val_accuracy: 0.9957\n",
      "Epoch 340/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7086e-04 - accuracy: 0.9960 - val_loss: 3.3126e-04 - val_accuracy: 1.0000\n",
      "Epoch 341/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.5372e-04 - accuracy: 0.9955 - val_loss: 2.9000e-04 - val_accuracy: 0.9974\n",
      "Epoch 342/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4886e-04 - accuracy: 0.9967 - val_loss: 4.0097e-04 - val_accuracy: 0.9965\n",
      "Epoch 343/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.6427e-04 - accuracy: 0.9970 - val_loss: 3.1413e-04 - val_accuracy: 0.9965\n",
      "Epoch 344/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.7133e-04 - accuracy: 0.9957 - val_loss: 3.3191e-04 - val_accuracy: 0.9957\n",
      "Epoch 345/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.5929e-04 - accuracy: 0.9988 - val_loss: 3.3378e-04 - val_accuracy: 0.9983\n",
      "Epoch 346/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6286e-04 - accuracy: 0.9977 - val_loss: 3.7105e-04 - val_accuracy: 0.9983\n",
      "Epoch 347/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4976e-04 - accuracy: 0.9965 - val_loss: 3.4094e-04 - val_accuracy: 0.9974\n",
      "Epoch 348/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.6298e-04 - accuracy: 0.9970 - val_loss: 3.3697e-04 - val_accuracy: 0.9965\n",
      "Epoch 349/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5595e-04 - accuracy: 0.9974 - val_loss: 5.3357e-04 - val_accuracy: 0.9939\n",
      "Epoch 350/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.6361e-04 - accuracy: 0.9972 - val_loss: 3.4964e-04 - val_accuracy: 0.9983\n",
      "Epoch 351/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4480e-04 - accuracy: 0.9977 - val_loss: 2.9042e-04 - val_accuracy: 0.9974\n",
      "Epoch 352/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3788e-04 - accuracy: 0.9962 - val_loss: 3.3646e-04 - val_accuracy: 0.9948\n",
      "Epoch 353/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4877e-04 - accuracy: 0.9977 - val_loss: 3.2413e-04 - val_accuracy: 0.9983\n",
      "Epoch 354/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4514e-04 - accuracy: 0.9972 - val_loss: 3.5182e-04 - val_accuracy: 0.9931\n",
      "Epoch 355/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4707e-04 - accuracy: 0.9969 - val_loss: 2.7812e-04 - val_accuracy: 0.9957\n",
      "Epoch 356/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.5371e-04 - accuracy: 0.9960 - val_loss: 3.7359e-04 - val_accuracy: 0.9974\n",
      "Epoch 357/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4583e-04 - accuracy: 0.9965 - val_loss: 3.7229e-04 - val_accuracy: 0.9939\n",
      "Epoch 358/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4623e-04 - accuracy: 0.9974 - val_loss: 3.3745e-04 - val_accuracy: 0.9965\n",
      "Epoch 359/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.4388e-04 - accuracy: 0.9969 - val_loss: 3.1328e-04 - val_accuracy: 0.9974\n",
      "Epoch 360/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3313e-04 - accuracy: 0.9967 - val_loss: 3.2545e-04 - val_accuracy: 0.9983\n",
      "Epoch 361/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5285e-04 - accuracy: 0.9972 - val_loss: 3.2382e-04 - val_accuracy: 0.9974\n",
      "Epoch 362/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4449e-04 - accuracy: 0.9972 - val_loss: 3.2740e-04 - val_accuracy: 0.9983\n",
      "Epoch 363/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5408e-04 - accuracy: 0.9960 - val_loss: 3.6051e-04 - val_accuracy: 0.9974\n",
      "Epoch 364/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3625e-04 - accuracy: 0.9970 - val_loss: 3.6862e-04 - val_accuracy: 0.9983\n",
      "Epoch 365/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4843e-04 - accuracy: 0.9981 - val_loss: 3.4240e-04 - val_accuracy: 0.9974\n",
      "Epoch 366/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3994e-04 - accuracy: 0.9970 - val_loss: 3.1957e-04 - val_accuracy: 0.9939\n",
      "Epoch 367/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.3127e-04 - accuracy: 0.9962 - val_loss: 3.1796e-04 - val_accuracy: 0.9965\n",
      "Epoch 368/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5647e-04 - accuracy: 0.9965 - val_loss: 3.2484e-04 - val_accuracy: 0.9965\n",
      "Epoch 369/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.4033e-04 - accuracy: 0.9969 - val_loss: 3.3014e-04 - val_accuracy: 0.9939\n",
      "Epoch 370/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2874e-04 - accuracy: 0.9960 - val_loss: 2.8465e-04 - val_accuracy: 0.9974\n",
      "Epoch 371/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3700e-04 - accuracy: 0.9984 - val_loss: 2.8025e-04 - val_accuracy: 0.9983\n",
      "Epoch 372/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2689e-04 - accuracy: 0.9974 - val_loss: 3.1959e-04 - val_accuracy: 0.9948\n",
      "Epoch 373/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4296e-04 - accuracy: 0.9962 - val_loss: 3.3196e-04 - val_accuracy: 0.9948\n",
      "Epoch 374/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 12ms/step - loss: 3.3457e-04 - accuracy: 0.9979 - val_loss: 2.8276e-04 - val_accuracy: 0.9974\n",
      "Epoch 375/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4083e-04 - accuracy: 0.9967 - val_loss: 2.8647e-04 - val_accuracy: 0.9965\n",
      "Epoch 376/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2627e-04 - accuracy: 0.9967 - val_loss: 3.3350e-04 - val_accuracy: 0.9965\n",
      "Epoch 377/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2368e-04 - accuracy: 0.9972 - val_loss: 3.7978e-04 - val_accuracy: 0.9957\n",
      "Epoch 378/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3972e-04 - accuracy: 0.9969 - val_loss: 3.4141e-04 - val_accuracy: 0.9991\n",
      "Epoch 379/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.4301e-04 - accuracy: 0.9958 - val_loss: 3.0835e-04 - val_accuracy: 0.9983\n",
      "Epoch 380/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3459e-04 - accuracy: 0.9967 - val_loss: 3.2001e-04 - val_accuracy: 0.9965\n",
      "Epoch 381/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.4355e-04 - accuracy: 0.9972 - val_loss: 4.0830e-04 - val_accuracy: 0.9948\n",
      "Epoch 382/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3957e-04 - accuracy: 0.9967 - val_loss: 3.1226e-04 - val_accuracy: 0.9974\n",
      "Epoch 383/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.5028e-04 - accuracy: 0.9969 - val_loss: 3.2417e-04 - val_accuracy: 0.9974\n",
      "Epoch 384/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3872e-04 - accuracy: 0.9969 - val_loss: 3.1318e-04 - val_accuracy: 0.9983\n",
      "Epoch 385/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3490e-04 - accuracy: 0.9979 - val_loss: 2.7987e-04 - val_accuracy: 0.9983\n",
      "Epoch 386/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3789e-04 - accuracy: 0.9960 - val_loss: 3.7616e-04 - val_accuracy: 0.9939\n",
      "Epoch 387/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3004e-04 - accuracy: 0.9974 - val_loss: 2.9622e-04 - val_accuracy: 1.0000\n",
      "Epoch 388/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2142e-04 - accuracy: 0.9977 - val_loss: 3.3542e-04 - val_accuracy: 0.9965\n",
      "Epoch 389/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3023e-04 - accuracy: 0.9964 - val_loss: 2.5968e-04 - val_accuracy: 1.0000\n",
      "Epoch 390/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4010e-04 - accuracy: 0.9974 - val_loss: 3.6326e-04 - val_accuracy: 0.9948\n",
      "Epoch 391/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3724e-04 - accuracy: 0.9964 - val_loss: 2.9177e-04 - val_accuracy: 0.9983\n",
      "Epoch 392/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2947e-04 - accuracy: 0.9964 - val_loss: 3.1556e-04 - val_accuracy: 0.9957\n",
      "Epoch 393/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.1907e-04 - accuracy: 0.9969 - val_loss: 3.6975e-04 - val_accuracy: 0.9939\n",
      "Epoch 394/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2723e-04 - accuracy: 0.9969 - val_loss: 3.8209e-04 - val_accuracy: 0.9974\n",
      "Epoch 395/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2967e-04 - accuracy: 0.9960 - val_loss: 4.4100e-04 - val_accuracy: 0.9948\n",
      "Epoch 396/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2507e-04 - accuracy: 0.9967 - val_loss: 2.9373e-04 - val_accuracy: 0.9957\n",
      "Epoch 397/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3865e-04 - accuracy: 0.9970 - val_loss: 2.6676e-04 - val_accuracy: 0.9974\n",
      "Epoch 398/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3531e-04 - accuracy: 0.9969 - val_loss: 2.8356e-04 - val_accuracy: 0.9922\n",
      "Epoch 399/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3374e-04 - accuracy: 0.9970 - val_loss: 3.2616e-04 - val_accuracy: 0.9983\n",
      "Epoch 400/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3402e-04 - accuracy: 0.9967 - val_loss: 3.2279e-04 - val_accuracy: 0.9983\n",
      "Epoch 401/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2911e-04 - accuracy: 0.9964 - val_loss: 3.5525e-04 - val_accuracy: 0.9922\n",
      "Epoch 402/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.3209e-04 - accuracy: 0.9964 - val_loss: 3.1283e-04 - val_accuracy: 0.9965\n",
      "Epoch 403/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3404e-04 - accuracy: 0.9967 - val_loss: 3.3206e-04 - val_accuracy: 0.9965\n",
      "Epoch 404/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2209e-04 - accuracy: 0.9979 - val_loss: 2.9812e-04 - val_accuracy: 0.9991\n",
      "Epoch 405/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3201e-04 - accuracy: 0.9979 - val_loss: 3.6024e-04 - val_accuracy: 0.9974\n",
      "Epoch 406/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.4710e-04 - accuracy: 0.9969 - val_loss: 3.1333e-04 - val_accuracy: 0.9957\n",
      "Epoch 407/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2879e-04 - accuracy: 0.9979 - val_loss: 2.6187e-04 - val_accuracy: 0.9974\n",
      "Epoch 408/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3498e-04 - accuracy: 0.9976 - val_loss: 2.6391e-04 - val_accuracy: 1.0000\n",
      "Epoch 409/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1002e-04 - accuracy: 0.9977 - val_loss: 3.3632e-04 - val_accuracy: 0.9957\n",
      "Epoch 410/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2964e-04 - accuracy: 0.9969 - val_loss: 2.7553e-04 - val_accuracy: 0.9974\n",
      "Epoch 411/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3253e-04 - accuracy: 0.9974 - val_loss: 2.9823e-04 - val_accuracy: 0.9991\n",
      "Epoch 412/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2480e-04 - accuracy: 0.9974 - val_loss: 3.4330e-04 - val_accuracy: 0.9957\n",
      "Epoch 413/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3751e-04 - accuracy: 0.9969 - val_loss: 4.2370e-04 - val_accuracy: 0.9948\n",
      "Epoch 414/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2195e-04 - accuracy: 0.9977 - val_loss: 3.4100e-04 - val_accuracy: 0.9965\n",
      "Epoch 415/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1713e-04 - accuracy: 0.9977 - val_loss: 3.1637e-04 - val_accuracy: 0.9974\n",
      "Epoch 416/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3071e-04 - accuracy: 0.9969 - val_loss: 3.3207e-04 - val_accuracy: 0.9939\n",
      "Epoch 417/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.2528e-04 - accuracy: 0.9967 - val_loss: 4.1090e-04 - val_accuracy: 0.9974\n",
      "Epoch 418/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2521e-04 - accuracy: 0.9969 - val_loss: 3.5252e-04 - val_accuracy: 0.9965\n",
      "Epoch 419/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2325e-04 - accuracy: 0.9970 - val_loss: 3.2972e-04 - val_accuracy: 0.9991\n",
      "Epoch 420/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2555e-04 - accuracy: 0.9979 - val_loss: 4.1631e-04 - val_accuracy: 0.9957\n",
      "Epoch 421/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.3803e-04 - accuracy: 0.9970 - val_loss: 3.0679e-04 - val_accuracy: 0.9965\n",
      "Epoch 422/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2813e-04 - accuracy: 0.9967 - val_loss: 2.9997e-04 - val_accuracy: 0.9965\n",
      "Epoch 423/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2550e-04 - accuracy: 0.9970 - val_loss: 3.7879e-04 - val_accuracy: 0.9948\n",
      "Epoch 424/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3123e-04 - accuracy: 0.9957 - val_loss: 3.0372e-04 - val_accuracy: 0.9983\n",
      "Epoch 425/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.2225e-04 - accuracy: 0.9970 - val_loss: 2.9804e-04 - val_accuracy: 0.9991\n",
      "Epoch 426/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2054e-04 - accuracy: 0.9974 - val_loss: 3.0168e-04 - val_accuracy: 0.9983\n",
      "Epoch 427/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0530e-04 - accuracy: 0.9969 - val_loss: 2.6515e-04 - val_accuracy: 0.9983\n",
      "Epoch 428/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2771e-04 - accuracy: 0.9972 - val_loss: 2.7787e-04 - val_accuracy: 0.9991\n",
      "Epoch 429/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2363e-04 - accuracy: 0.9970 - val_loss: 3.8973e-04 - val_accuracy: 0.9974\n",
      "Epoch 430/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1854e-04 - accuracy: 0.9979 - val_loss: 3.3283e-04 - val_accuracy: 0.9957\n",
      "Epoch 431/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2704e-04 - accuracy: 0.9979 - val_loss: 2.8177e-04 - val_accuracy: 0.9931\n",
      "Epoch 432/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2536e-04 - accuracy: 0.9967 - val_loss: 3.2311e-04 - val_accuracy: 0.9974\n",
      "Epoch 433/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2608e-04 - accuracy: 0.9969 - val_loss: 2.6108e-04 - val_accuracy: 0.9974\n",
      "Epoch 434/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.3147e-04 - accuracy: 0.9976 - val_loss: 2.5709e-04 - val_accuracy: 0.9965\n",
      "Epoch 435/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2505e-04 - accuracy: 0.9962 - val_loss: 3.1385e-04 - val_accuracy: 0.9939\n",
      "Epoch 436/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1791e-04 - accuracy: 0.9977 - val_loss: 2.8852e-04 - val_accuracy: 0.9957\n",
      "Epoch 437/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.3788e-04 - accuracy: 0.9967 - val_loss: 3.0806e-04 - val_accuracy: 0.9983\n",
      "Epoch 438/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.1788e-04 - accuracy: 0.9979 - val_loss: 2.7594e-04 - val_accuracy: 0.9965\n",
      "Epoch 439/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1779e-04 - accuracy: 0.9970 - val_loss: 3.0915e-04 - val_accuracy: 0.9974\n",
      "Epoch 440/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2549e-04 - accuracy: 0.9970 - val_loss: 3.4328e-04 - val_accuracy: 0.9974\n",
      "Epoch 441/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1536e-04 - accuracy: 0.9976 - val_loss: 2.8893e-04 - val_accuracy: 0.9957\n",
      "Epoch 442/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2127e-04 - accuracy: 0.9976 - val_loss: 2.5221e-04 - val_accuracy: 0.9991\n",
      "Epoch 443/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9469e-04 - accuracy: 0.9967 - val_loss: 3.4108e-04 - val_accuracy: 0.9974\n",
      "Epoch 444/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2603e-04 - accuracy: 0.9981 - val_loss: 2.7645e-04 - val_accuracy: 0.9939\n",
      "Epoch 445/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.3291e-04 - accuracy: 0.9962 - val_loss: 3.1504e-04 - val_accuracy: 0.9974\n",
      "Epoch 446/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0635e-04 - accuracy: 0.9983 - val_loss: 3.6024e-04 - val_accuracy: 0.9983\n",
      "Epoch 447/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1987e-04 - accuracy: 0.9984 - val_loss: 2.9871e-04 - val_accuracy: 0.9965\n",
      "Epoch 448/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0789e-04 - accuracy: 0.9977 - val_loss: 2.5857e-04 - val_accuracy: 0.9974\n",
      "Epoch 449/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1875e-04 - accuracy: 0.9974 - val_loss: 3.0350e-04 - val_accuracy: 0.9965\n",
      "Epoch 450/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0601e-04 - accuracy: 0.9970 - val_loss: 3.2170e-04 - val_accuracy: 0.9974\n",
      "Epoch 451/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1999e-04 - accuracy: 0.9965 - val_loss: 3.1210e-04 - val_accuracy: 0.9974\n",
      "Epoch 452/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.1453e-04 - accuracy: 0.9972 - val_loss: 2.9813e-04 - val_accuracy: 0.9983\n",
      "Epoch 453/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0330e-04 - accuracy: 0.9976 - val_loss: 3.1579e-04 - val_accuracy: 0.9974\n",
      "Epoch 454/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 3.2311e-04 - accuracy: 0.9977 - val_loss: 2.5941e-04 - val_accuracy: 0.9991\n",
      "Epoch 455/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0744e-04 - accuracy: 0.9972 - val_loss: 4.2367e-04 - val_accuracy: 0.9965\n",
      "Epoch 456/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.1428e-04 - accuracy: 0.9964 - val_loss: 2.8926e-04 - val_accuracy: 0.9974\n",
      "Epoch 457/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1754e-04 - accuracy: 0.9970 - val_loss: 2.9293e-04 - val_accuracy: 0.9983\n",
      "Epoch 458/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1445e-04 - accuracy: 0.9981 - val_loss: 3.8812e-04 - val_accuracy: 1.0000\n",
      "Epoch 459/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0757e-04 - accuracy: 0.9977 - val_loss: 3.0676e-04 - val_accuracy: 1.0000\n",
      "Epoch 460/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.2096e-04 - accuracy: 0.9970 - val_loss: 2.6931e-04 - val_accuracy: 0.9974\n",
      "Epoch 461/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1908e-04 - accuracy: 0.9983 - val_loss: 2.6498e-04 - val_accuracy: 0.9974\n",
      "Epoch 462/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0918e-04 - accuracy: 0.9977 - val_loss: 2.6906e-04 - val_accuracy: 0.9974\n",
      "Epoch 463/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1421e-04 - accuracy: 0.9972 - val_loss: 2.5374e-04 - val_accuracy: 0.9991\n",
      "Epoch 464/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.0316e-04 - accuracy: 0.9979 - val_loss: 2.7148e-04 - val_accuracy: 0.9965\n",
      "Epoch 465/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0029e-04 - accuracy: 0.9977 - val_loss: 2.9916e-04 - val_accuracy: 0.9983\n",
      "Epoch 466/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1636e-04 - accuracy: 0.9976 - val_loss: 2.6677e-04 - val_accuracy: 0.9991\n",
      "Epoch 467/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1782e-04 - accuracy: 0.9972 - val_loss: 3.2341e-04 - val_accuracy: 0.9965\n",
      "Epoch 468/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1068e-04 - accuracy: 0.9967 - val_loss: 3.7168e-04 - val_accuracy: 0.9965\n",
      "Epoch 469/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1000e-04 - accuracy: 0.9970 - val_loss: 4.0200e-04 - val_accuracy: 0.9870\n",
      "Epoch 470/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1730e-04 - accuracy: 0.9972 - val_loss: 3.1020e-04 - val_accuracy: 0.9974\n",
      "Epoch 471/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.1750e-04 - accuracy: 0.9974 - val_loss: 3.1232e-04 - val_accuracy: 0.9974\n",
      "Epoch 472/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1154e-04 - accuracy: 0.9977 - val_loss: 2.9607e-04 - val_accuracy: 0.9948\n",
      "Epoch 473/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1701e-04 - accuracy: 0.9981 - val_loss: 2.9439e-04 - val_accuracy: 0.9957\n",
      "Epoch 474/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0266e-04 - accuracy: 0.9976 - val_loss: 2.5236e-04 - val_accuracy: 0.9983\n",
      "Epoch 475/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1755e-04 - accuracy: 0.9967 - val_loss: 2.5332e-04 - val_accuracy: 1.0000\n",
      "Epoch 476/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1263e-04 - accuracy: 0.9974 - val_loss: 2.8318e-04 - val_accuracy: 0.9991\n",
      "Epoch 477/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0601e-04 - accuracy: 0.9977 - val_loss: 2.9795e-04 - val_accuracy: 0.9991\n",
      "Epoch 478/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0960e-04 - accuracy: 0.9977 - val_loss: 3.1782e-04 - val_accuracy: 0.9965\n",
      "Epoch 479/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1082e-04 - accuracy: 0.9970 - val_loss: 2.8249e-04 - val_accuracy: 0.9991\n",
      "Epoch 480/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0061e-04 - accuracy: 0.9981 - val_loss: 2.9628e-04 - val_accuracy: 0.9983\n",
      "Epoch 481/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9393e-04 - accuracy: 0.9979 - val_loss: 3.5981e-04 - val_accuracy: 0.9957\n",
      "Epoch 482/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9896e-04 - accuracy: 0.9967 - val_loss: 2.6984e-04 - val_accuracy: 0.9974\n",
      "Epoch 483/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.0930e-04 - accuracy: 0.9962 - val_loss: 3.3248e-04 - val_accuracy: 0.9983\n",
      "Epoch 484/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1728e-04 - accuracy: 0.9972 - val_loss: 2.5831e-04 - val_accuracy: 0.9974\n",
      "Epoch 485/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.2625e-04 - accuracy: 0.9988 - val_loss: 3.4847e-04 - val_accuracy: 0.9974\n",
      "Epoch 486/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0585e-04 - accuracy: 0.9977 - val_loss: 3.9607e-04 - val_accuracy: 0.9948\n",
      "Epoch 487/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0053e-04 - accuracy: 0.9970 - val_loss: 2.9187e-04 - val_accuracy: 0.9974\n",
      "Epoch 488/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9984e-04 - accuracy: 0.9976 - val_loss: 2.2807e-04 - val_accuracy: 1.0000\n",
      "Epoch 489/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.1236e-04 - accuracy: 0.9976 - val_loss: 2.4922e-04 - val_accuracy: 0.9957\n",
      "Epoch 490/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.0701e-04 - accuracy: 0.9972 - val_loss: 2.4166e-04 - val_accuracy: 0.9983\n",
      "Epoch 491/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0203e-04 - accuracy: 0.9983 - val_loss: 2.7515e-04 - val_accuracy: 0.9965\n",
      "Epoch 492/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1131e-04 - accuracy: 0.9964 - val_loss: 2.7715e-04 - val_accuracy: 0.9983\n",
      "Epoch 493/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0323e-04 - accuracy: 0.9969 - val_loss: 2.5497e-04 - val_accuracy: 0.9991\n",
      "Epoch 494/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0586e-04 - accuracy: 0.9972 - val_loss: 2.7510e-04 - val_accuracy: 0.9974\n",
      "Epoch 495/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0596e-04 - accuracy: 0.9979 - val_loss: 2.6195e-04 - val_accuracy: 1.0000\n",
      "Epoch 496/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8819e-04 - accuracy: 0.9979 - val_loss: 3.0763e-04 - val_accuracy: 0.9965\n",
      "Epoch 497/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9711e-04 - accuracy: 0.9967 - val_loss: 3.3606e-04 - val_accuracy: 0.9965\n",
      "Epoch 498/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0636e-04 - accuracy: 0.9967 - val_loss: 2.9423e-04 - val_accuracy: 0.9991\n",
      "Epoch 499/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9425e-04 - accuracy: 0.9984 - val_loss: 3.1196e-04 - val_accuracy: 0.9991\n",
      "Epoch 500/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9489e-04 - accuracy: 0.9976 - val_loss: 3.7041e-04 - val_accuracy: 0.9983\n",
      "Epoch 501/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.1557e-04 - accuracy: 0.9977 - val_loss: 2.6790e-04 - val_accuracy: 0.9957\n",
      "Epoch 502/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9942e-04 - accuracy: 0.9974 - val_loss: 2.9969e-04 - val_accuracy: 0.9974\n",
      "Epoch 503/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9327e-04 - accuracy: 0.9970 - val_loss: 2.7872e-04 - val_accuracy: 0.9948\n",
      "Epoch 504/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0707e-04 - accuracy: 0.9977 - val_loss: 3.0110e-04 - val_accuracy: 0.9931\n",
      "Epoch 505/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0518e-04 - accuracy: 0.9976 - val_loss: 2.9667e-04 - val_accuracy: 0.9974\n",
      "Epoch 506/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0390e-04 - accuracy: 0.9979 - val_loss: 2.6667e-04 - val_accuracy: 0.9974\n",
      "Epoch 507/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.1161e-04 - accuracy: 0.9981 - val_loss: 2.5742e-04 - val_accuracy: 0.9983\n",
      "Epoch 508/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9360e-04 - accuracy: 0.9972 - val_loss: 2.6390e-04 - val_accuracy: 0.9974\n",
      "Epoch 509/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9753e-04 - accuracy: 0.9981 - val_loss: 2.7719e-04 - val_accuracy: 0.9957\n",
      "Epoch 510/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0300e-04 - accuracy: 0.9967 - val_loss: 2.8697e-04 - val_accuracy: 0.9974\n",
      "Epoch 511/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.0830e-04 - accuracy: 0.9977 - val_loss: 2.5937e-04 - val_accuracy: 0.9983\n",
      "Epoch 512/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9889e-04 - accuracy: 0.9965 - val_loss: 5.4267e-04 - val_accuracy: 0.9948\n",
      "Epoch 513/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0359e-04 - accuracy: 0.9972 - val_loss: 3.5292e-04 - val_accuracy: 0.9948\n",
      "Epoch 514/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0514e-04 - accuracy: 0.9970 - val_loss: 2.4290e-04 - val_accuracy: 0.9991\n",
      "Epoch 515/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.1012e-04 - accuracy: 0.9972 - val_loss: 2.5899e-04 - val_accuracy: 0.9983\n",
      "Epoch 516/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9809e-04 - accuracy: 0.9977 - val_loss: 2.5402e-04 - val_accuracy: 0.9991\n",
      "Epoch 517/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0541e-04 - accuracy: 0.9976 - val_loss: 2.4299e-04 - val_accuracy: 0.9965\n",
      "Epoch 518/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9313e-04 - accuracy: 0.9969 - val_loss: 3.0422e-04 - val_accuracy: 0.9965\n",
      "Epoch 519/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9690e-04 - accuracy: 0.9970 - val_loss: 2.9869e-04 - val_accuracy: 0.9991\n",
      "Epoch 520/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0629e-04 - accuracy: 0.9983 - val_loss: 5.4812e-04 - val_accuracy: 0.9939\n",
      "Epoch 521/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9622e-04 - accuracy: 0.9969 - val_loss: 3.3562e-04 - val_accuracy: 0.9974\n",
      "Epoch 522/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0200e-04 - accuracy: 0.9970 - val_loss: 2.9346e-04 - val_accuracy: 0.9991\n",
      "Epoch 523/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0225e-04 - accuracy: 0.9976 - val_loss: 2.3573e-04 - val_accuracy: 0.9983\n",
      "Epoch 524/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9271e-04 - accuracy: 0.9965 - val_loss: 3.3111e-04 - val_accuracy: 0.9957\n",
      "Epoch 525/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9960e-04 - accuracy: 0.9974 - val_loss: 2.4058e-04 - val_accuracy: 0.9983\n",
      "Epoch 526/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9691e-04 - accuracy: 0.9981 - val_loss: 5.3716e-04 - val_accuracy: 0.9948\n",
      "Epoch 527/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9713e-04 - accuracy: 0.9972 - val_loss: 3.1214e-04 - val_accuracy: 0.9957\n",
      "Epoch 528/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9626e-04 - accuracy: 0.9977 - val_loss: 2.5368e-04 - val_accuracy: 0.9983\n",
      "Epoch 529/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9512e-04 - accuracy: 0.9974 - val_loss: 3.5562e-04 - val_accuracy: 0.9965\n",
      "Epoch 530/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 3.1018e-04 - accuracy: 0.9972 - val_loss: 2.6646e-04 - val_accuracy: 0.9991\n",
      "Epoch 531/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0840e-04 - accuracy: 0.9970 - val_loss: 2.6101e-04 - val_accuracy: 0.9974\n",
      "Epoch 532/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8530e-04 - accuracy: 0.9986 - val_loss: 2.7290e-04 - val_accuracy: 0.9965\n",
      "Epoch 533/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9441e-04 - accuracy: 0.9969 - val_loss: 2.5800e-04 - val_accuracy: 0.9991\n",
      "Epoch 534/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 2.9657e-04 - accuracy: 0.9974 - val_loss: 3.3103e-04 - val_accuracy: 0.9965\n",
      "Epoch 535/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9230e-04 - accuracy: 0.9984 - val_loss: 2.9171e-04 - val_accuracy: 0.9957\n",
      "Epoch 536/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8967e-04 - accuracy: 0.9979 - val_loss: 2.4739e-04 - val_accuracy: 0.9948\n",
      "Epoch 537/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0925e-04 - accuracy: 0.9979 - val_loss: 2.9165e-04 - val_accuracy: 0.9974\n",
      "Epoch 538/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9537e-04 - accuracy: 0.9965 - val_loss: 3.0216e-04 - val_accuracy: 0.9974\n",
      "Epoch 539/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9890e-04 - accuracy: 0.9986 - val_loss: 3.0681e-04 - val_accuracy: 0.9983\n",
      "Epoch 540/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9345e-04 - accuracy: 0.9970 - val_loss: 2.5709e-04 - val_accuracy: 0.9983\n",
      "Epoch 541/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9066e-04 - accuracy: 0.9974 - val_loss: 2.6696e-04 - val_accuracy: 0.9991\n",
      "Epoch 542/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0001e-04 - accuracy: 0.9977 - val_loss: 2.8744e-04 - val_accuracy: 0.9983\n",
      "Epoch 543/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9124e-04 - accuracy: 0.9976 - val_loss: 2.3240e-04 - val_accuracy: 0.9983\n",
      "Epoch 544/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7296e-04 - accuracy: 0.9984 - val_loss: 2.6728e-04 - val_accuracy: 0.9991\n",
      "Epoch 545/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 3.0001e-04 - accuracy: 0.9974 - val_loss: 3.0137e-04 - val_accuracy: 0.9974\n",
      "Epoch 546/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9193e-04 - accuracy: 0.9986 - val_loss: 2.3544e-04 - val_accuracy: 0.9965\n",
      "Epoch 547/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9986e-04 - accuracy: 0.9986 - val_loss: 2.7314e-04 - val_accuracy: 0.9991\n",
      "Epoch 548/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8823e-04 - accuracy: 0.9976 - val_loss: 3.1343e-04 - val_accuracy: 0.9991\n",
      "Epoch 549/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8604e-04 - accuracy: 0.9979 - val_loss: 2.2796e-04 - val_accuracy: 0.9965\n",
      "Epoch 550/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7836e-04 - accuracy: 0.9972 - val_loss: 2.7828e-04 - val_accuracy: 0.9991\n",
      "Epoch 551/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8110e-04 - accuracy: 0.9979 - val_loss: 3.0002e-04 - val_accuracy: 0.9948\n",
      "Epoch 552/600\n",
      "721/721 [==============================] - 10s 13ms/step - loss: 3.0354e-04 - accuracy: 0.9977 - val_loss: 2.2516e-04 - val_accuracy: 0.9991\n",
      "Epoch 553/600\n",
      "721/721 [==============================] - 10s 14ms/step - loss: 2.9263e-04 - accuracy: 0.9981 - val_loss: 2.4589e-04 - val_accuracy: 0.9957\n",
      "Epoch 554/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8258e-04 - accuracy: 0.9972 - val_loss: 2.8264e-04 - val_accuracy: 0.9983\n",
      "Epoch 555/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9528e-04 - accuracy: 0.9969 - val_loss: 2.7172e-04 - val_accuracy: 0.9991\n",
      "Epoch 556/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9692e-04 - accuracy: 0.9972 - val_loss: 2.6726e-04 - val_accuracy: 0.9948\n",
      "Epoch 557/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9024e-04 - accuracy: 0.9967 - val_loss: 2.3065e-04 - val_accuracy: 0.9974\n",
      "Epoch 558/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9418e-04 - accuracy: 0.9970 - val_loss: 2.5955e-04 - val_accuracy: 0.9965\n",
      "Epoch 559/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8746e-04 - accuracy: 0.9976 - val_loss: 3.0718e-04 - val_accuracy: 0.9965\n",
      "Epoch 560/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8787e-04 - accuracy: 0.9972 - val_loss: 2.6454e-04 - val_accuracy: 1.0000\n",
      "Epoch 561/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9053e-04 - accuracy: 0.9970 - val_loss: 3.3334e-04 - val_accuracy: 1.0000\n",
      "Epoch 562/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8933e-04 - accuracy: 0.9981 - val_loss: 2.9500e-04 - val_accuracy: 1.0000\n",
      "Epoch 563/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9399e-04 - accuracy: 0.9972 - val_loss: 2.4945e-04 - val_accuracy: 1.0000\n",
      "Epoch 564/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9656e-04 - accuracy: 0.9977 - val_loss: 3.1100e-04 - val_accuracy: 0.9983\n",
      "Epoch 565/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9133e-04 - accuracy: 0.9981 - val_loss: 2.7972e-04 - val_accuracy: 0.9983\n",
      "Epoch 566/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8247e-04 - accuracy: 0.9979 - val_loss: 2.8192e-04 - val_accuracy: 0.9957\n",
      "Epoch 567/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9139e-04 - accuracy: 0.9986 - val_loss: 2.7314e-04 - val_accuracy: 0.9974\n",
      "Epoch 568/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8936e-04 - accuracy: 0.9970 - val_loss: 3.6033e-04 - val_accuracy: 0.9991\n",
      "Epoch 569/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9561e-04 - accuracy: 0.9965 - val_loss: 3.3624e-04 - val_accuracy: 0.9957\n",
      "Epoch 570/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8182e-04 - accuracy: 0.9972 - val_loss: 2.6633e-04 - val_accuracy: 0.9991\n",
      "Epoch 571/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9816e-04 - accuracy: 0.9979 - val_loss: 2.5093e-04 - val_accuracy: 1.0000\n",
      "Epoch 572/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7872e-04 - accuracy: 0.9984 - val_loss: 2.6721e-04 - val_accuracy: 0.9983\n",
      "Epoch 573/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9087e-04 - accuracy: 0.9981 - val_loss: 2.2793e-04 - val_accuracy: 0.9983\n",
      "Epoch 574/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8329e-04 - accuracy: 0.9983 - val_loss: 2.9752e-04 - val_accuracy: 0.9991\n",
      "Epoch 575/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8250e-04 - accuracy: 0.9979 - val_loss: 3.7137e-04 - val_accuracy: 0.9965\n",
      "Epoch 576/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9035e-04 - accuracy: 0.9979 - val_loss: 2.7583e-04 - val_accuracy: 0.9991\n",
      "Epoch 577/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8099e-04 - accuracy: 0.9977 - val_loss: 2.4894e-04 - val_accuracy: 0.9974\n",
      "Epoch 578/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8855e-04 - accuracy: 0.9976 - val_loss: 2.5163e-04 - val_accuracy: 0.9991\n",
      "Epoch 579/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8694e-04 - accuracy: 0.9964 - val_loss: 2.7822e-04 - val_accuracy: 0.9965\n",
      "Epoch 580/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8505e-04 - accuracy: 0.9974 - val_loss: 3.0503e-04 - val_accuracy: 0.9957\n",
      "Epoch 581/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9104e-04 - accuracy: 0.9967 - val_loss: 2.4417e-04 - val_accuracy: 0.9965\n",
      "Epoch 582/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7206e-04 - accuracy: 0.9976 - val_loss: 2.3477e-04 - val_accuracy: 0.9991\n",
      "Epoch 583/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8492e-04 - accuracy: 0.9965 - val_loss: 2.2644e-04 - val_accuracy: 0.9974\n",
      "Epoch 584/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.9043e-04 - accuracy: 0.9977 - val_loss: 2.8736e-04 - val_accuracy: 0.9965\n",
      "Epoch 585/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8416e-04 - accuracy: 0.9970 - val_loss: 2.7156e-04 - val_accuracy: 0.9974\n",
      "Epoch 586/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8799e-04 - accuracy: 0.9986 - val_loss: 2.2895e-04 - val_accuracy: 0.9983\n",
      "Epoch 587/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.7919e-04 - accuracy: 0.9983 - val_loss: 2.2559e-04 - val_accuracy: 0.9974\n",
      "Epoch 588/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7668e-04 - accuracy: 0.9984 - val_loss: 2.5792e-04 - val_accuracy: 0.9965\n",
      "Epoch 589/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.8425e-04 - accuracy: 0.9970 - val_loss: 2.2672e-04 - val_accuracy: 0.9983\n",
      "Epoch 590/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7706e-04 - accuracy: 0.9981 - val_loss: 2.8228e-04 - val_accuracy: 0.9983\n",
      "Epoch 591/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8737e-04 - accuracy: 0.9976 - val_loss: 2.7261e-04 - val_accuracy: 0.9974\n",
      "Epoch 592/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9821e-04 - accuracy: 0.9965 - val_loss: 2.3404e-04 - val_accuracy: 0.9965\n",
      "Epoch 593/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8516e-04 - accuracy: 0.9969 - val_loss: 2.7265e-04 - val_accuracy: 0.9965\n",
      "Epoch 594/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.9040e-04 - accuracy: 0.9983 - val_loss: 3.8256e-04 - val_accuracy: 0.9939\n",
      "Epoch 595/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.7775e-04 - accuracy: 0.9981 - val_loss: 2.4445e-04 - val_accuracy: 0.9983\n",
      "Epoch 596/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8642e-04 - accuracy: 0.9986 - val_loss: 2.6428e-04 - val_accuracy: 0.9983\n",
      "Epoch 597/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8245e-04 - accuracy: 0.9976 - val_loss: 2.5697e-04 - val_accuracy: 0.9965\n",
      "Epoch 598/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8566e-04 - accuracy: 0.9974 - val_loss: 2.4131e-04 - val_accuracy: 1.0000\n",
      "Epoch 599/600\n",
      "721/721 [==============================] - 9s 12ms/step - loss: 2.7761e-04 - accuracy: 0.9974 - val_loss: 2.2994e-04 - val_accuracy: 0.9991\n",
      "Epoch 600/600\n",
      "721/721 [==============================] - 9s 13ms/step - loss: 2.8095e-04 - accuracy: 0.9979 - val_loss: 2.3863e-04 - val_accuracy: 0.9991\n"
     ]
    }
   ],
   "source": [
    "Batch_Size = 8\n",
    "Epoch_Anz = 600\n",
    "Shift_Range = 3\n",
    "Brightness_Range = 0.3\n",
    "\n",
    "datagen = ImageDataGenerator(width_shift_range=[-Shift_Range,Shift_Range], height_shift_range=[-Shift_Range,Shift_Range],brightness_range=[1-Brightness_Range,1+Brightness_Range])\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_generator(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_generator(train_iterator, epochs = Epoch_Anz)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Overall Learing results (Step 1 & Step 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "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",
    "\n",
    "\n",
    "if (Training_Percentage > 0):\n",
    "    val_loss_ges = np.append(val_loss_ges, history.history['val_loss'])\n",
    "    plt.semilogy(val_loss_ges)\n",
    "\n",
    "\n",
    "plt.semilogy(loss_ges)\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",
    "* The evaluation takes the periodic character of the results into account (dev1 ... dev2).\n",
    "* Images, that have a bigger deviation as the parameter \"deviation_max_list\" are printed in a list to check the picture and labeling itself"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data_resize_all\\0.3_0056_zeiger1_2020-04-29_13-37-02.jpg 0.03 -0.08030542741586333 0.11030542741586333\n",
      "data_resize_all\\0.3_0098_zeiger1_2020-04-29_11-00-02.jpg 0.03 -0.08794353310095016 0.11794353310095015\n",
      "data_resize_all\\0.5_0211_zeiger4_2020-04-29_12-52-02.jpg 0.05 -0.0268908342172941 0.0768908342172941\n",
      "data_resize_all\\0.6_0251_zeiger2_2020-04-29_13-43-02.jpg 0.06 0.009033161743789679 0.05096683825621032\n",
      "data_resize_all\\0.7_0257_zeiger4_2020-04-29_10-58-40.jpg 0.06999999999999999 -0.006290109155446588 0.07629010915544658\n",
      "data_resize_all\\0.7_0272_zeiger1_2020-04-29_12-20-02.jpg 0.06999999999999999 -0.026847770823214545 0.09684777082321454\n",
      "data_resize_all\\0.7_0273_zeiger1_2020-04-29_12-21-02.jpg 0.06999999999999999 -0.03430730737859433 0.10430730737859432\n",
      "data_resize_all\\0.7_0274_zeiger1_2020-04-29_13-40-01.jpg 0.06999999999999999 -0.037317940928249804 0.1073179409282498\n",
      "data_resize_all\\0.8_0275_zeiger1_2020-04-29_13-41-02.jpg 0.08 -0.013485896651649898 0.0934858966516499\n",
      "data_resize_all\\0.9_0314_zeiger1_2020-04-29_13-42-02.jpg 0.09 -0.11243695194682202 0.202436951946822\n",
      "data_resize_all\\2.2_0989_zeiger2_2019-06-03T060009.jpg 0.22000000000000003 0.27641701577325595 -0.05641701577325592\n",
      "data_resize_all\\3.6_PRODUCED_ANALOG_2.jpg 0.36 0.30264107174557525 0.057358928254424735\n",
      "data_resize_all\\4.1_2037_zeiger1_2020-04-29_14-10-02.jpg 0.41 0.46923388214338985 -0.05923388214338987\n",
      "data_resize_all\\4.2_2034_zeiger1_2020-04-29_11-32-02.jpg 0.42000000000000004 0.4713559839730174 -0.05135598397301738\n",
      "data_resize_all\\4.2_2143_zeiger2_2020-04-29_13-30-02.jpg 0.42000000000000004 0.473162303390018 -0.053162303390017984\n",
      "data_resize_all\\4.4_2138_zeiger2_2019-06-04T080009.jpg 0.44000000000000006 0.5116178343004609 -0.07161783430046087\n",
      "data_resize_all\\4.6_4.7.jpg 0.45999999999999996 0.6006299979289289 -0.14062999792892894\n",
      "data_resize_all\\4.9_2399_zeiger4_2019-06-02T163009.jpg 0.49000000000000005 0.4343060402207325 0.055693959779267554\n",
      "data_resize_all\\5.2_2597_zeiger1_2020-04-29_11-40-01.jpg 0.52 0.5738080727312611 -0.05380807273126109\n",
      "data_resize_all\\5.6_2827_zeiger1_2020-04-29_11-43-02.jpg 0.5599999999999999 0.6104387567049934 -0.050438756704993426\n",
      "data_resize_all\\5.8_PRODUCED_ANALOG_2.jpg 0.58 0.6339004185584918 -0.053900418558491814\n",
      "data_resize_all\\6.6_3353_zeiger3_2020-04-29_11-57-02.jpg 0.6599999999999999 0.5527128437521998 0.10728715624780016\n",
      "data_resize_all\\8.2_4083_zeiger3_2019-06-04T070009.jpg 0.82 0.7659364560330577 0.05406354396694224\n",
      "data_resize_all\\8.2_4084_zeiger3_2019-06-04T071009.jpg 0.82 0.7474906953785396 0.07250930462146032\n",
      "data_resize_all\\8.3_4069_zeiger1_2020-04-29_14-42-02.jpg 0.8300000000000001 0.7782148973765126 0.051785102623487456\n",
      "data_resize_all\\8.3_4085_zeiger3_2019-06-04T191009.jpg 0.8300000000000001 0.778995474162159 0.0510045258378411\n",
      "data_resize_all\\8.3_4089_zeiger3_2019-06-05T181009.jpg 0.8300000000000001 0.7646347358137394 0.06536526418626065\n",
      "data_resize_all\\8.3_4123_zeiger3_2019-06-04T104010.jpg 0.8300000000000001 0.7249134671315671 0.10508653286843295\n",
      "data_resize_all\\8.3_4124_zeiger3_2019-06-05T043009.jpg 0.8300000000000001 0.7557606125346004 0.07423938746539971\n",
      "data_resize_all\\8.3_4125_zeiger3_2019-06-05T180009.jpg 0.8300000000000001 0.7461778950745679 0.08382210492543218\n",
      "data_resize_all\\8.3_4126_zeiger3_2019-06-05T182008.jpg 0.8300000000000001 0.7606513884625357 0.06934861153746441\n",
      "data_resize_all\\8.3_4147_zeiger3_2019-06-01T185013.jpg 0.8300000000000001 0.7503945313314229 0.0796054686685772\n",
      "data_resize_all\\8.3_4148_zeiger3_2019-06-01T190020.jpg 0.8300000000000001 0.7169560690755299 0.11304393092447018\n",
      "data_resize_all\\8.3_4149_zeiger3_2019-06-01T191021.jpg 0.8300000000000001 0.7579335423553014 0.07206645764469866\n",
      "data_resize_all\\8.3_4150_zeiger3_2019-06-01T192013.jpg 0.8300000000000001 0.7595177478150428 0.0704822521849573\n",
      "data_resize_all\\8.3_4151_zeiger3_2019-06-01T193014.jpg 0.8300000000000001 0.7430116470483223 0.08698835295167773\n",
      "data_resize_all\\8.3_4152_zeiger3_2019-06-01T194012.jpg 0.8300000000000001 0.7622339432932277 0.06776605670677238\n",
      "data_resize_all\\8.3_4153_zeiger3_2019-06-01T195015.jpg 0.8300000000000001 0.7576023920170419 0.07239760798295813\n",
      "data_resize_all\\8.3_4154_zeiger3_2019-06-01T200012.jpg 0.8300000000000001 0.75921847169291 0.07078152830709006\n",
      "data_resize_all\\8.3_4155_zeiger3_2019-06-01T204131.jpg 0.8300000000000001 0.7503707584785617 0.07962924152143835\n",
      "data_resize_all\\8.3_76ecb2694d09919237d8495271868ad1.jpg 0.8300000000000001 0.7407592973057722 0.0892407026942279\n",
      "data_resize_all\\8.4_4111_zeiger1_2020-04-29_10-45-02.jpg 0.8400000000000001 0.7770919383859718 0.06290806161402829\n",
      "data_resize_all\\8.4_4115_zeiger1_2020-04-29_13-22-02.jpg 0.8400000000000001 0.7796450061794795 0.060354993820520564\n",
      "data_resize_all\\8.4_4156_zeiger3_2019-06-03T174009.jpg 0.8400000000000001 0.6729903121428871 0.167009687857113\n",
      "data_resize_all\\8.5_4116_zeiger1_2020-04-29_13-23-01.jpg 0.85 0.7712704543198837 0.07872954568011625\n",
      "data_resize_all\\8.5_4117_zeiger1_2020-04-29_14-43-02.jpg 0.85 0.7631978727125934 0.08680212728740655\n",
      "data_resize_all\\8.5_4145_zeiger1_2020-04-29_12-04-02.jpg 0.85 0.7697070809371109 0.08029291906288905\n",
      "data_resize_all\\8.5_4203_zeiger3_2019-06-05T165009.jpg 0.85 0.7453226225812484 0.1046773774187516\n",
      "data_resize_all\\8.5_4204_zeiger3_2019-06-05T204009.jpg 0.85 0.7514593009047849 0.09854069909521512\n",
      "data_resize_all\\8.5_4220_zeiger2_2019-06-04T205009.jpg 0.85 0.7833258710141902 0.06667412898580982\n",
      "data_resize_all\\8.5_4221_zeiger2_2019-06-04T210009.jpg 0.85 0.7885219753707727 0.061478024629227246\n",
      "data_resize_all\\8.6_4171_zeiger3_2019-06-03T204009.jpg 0.86 0.6383875113337316 0.22161248866626837\n",
      "data_resize_all\\8.6_4173_zeiger3_2019-06-03T210009.jpg 0.86 0.640671336864707 0.21932866313529298\n",
      "data_resize_all\\8.6_4182_zeiger3_2019-06-03T230009.jpg 0.86 0.6017901055594554 0.25820989444054454\n",
      "data_resize_all\\8.6_4185_zeiger3_2019-06-03T233009.jpg 0.86 0.6112329938689797 0.24876700613102032\n",
      "data_resize_all\\8.6_4186_zeiger3_2019-06-03T234009.jpg 0.86 0.6429754632357154 0.21702453676428457\n",
      "data_resize_all\\8.6_4188_zeiger3_2019-06-04T000009.jpg 0.86 0.5976933580952819 0.2623066419047181\n",
      "data_resize_all\\8.6_4197_zeiger3_2019-06-04T014009.jpg 0.86 0.6619771253976816 0.19802287460231838\n",
      "data_resize_all\\8.6_4198_zeiger3_2019-06-04T015009.jpg 0.86 0.6428615129130624 0.21713848708693761\n",
      "data_resize_all\\8.6_4199_zeiger3_2019-06-04T020009.jpg 0.86 0.5732149466928362 0.28678505330716375\n",
      "data_resize_all\\8.6_4200_zeiger3_2019-06-04T021009.jpg 0.86 0.6369368550450116 0.22306314495498836\n",
      "data_resize_all\\8.6_4201_zeiger3_2019-06-04T022009.jpg 0.86 0.5886065123689923 0.2713934876310077\n",
      "data_resize_all\\8.6_4202_zeiger3_2019-06-04T023009.jpg 0.86 0.6131844168603477 0.24681558313965224\n",
      "data_resize_all\\8.6_4215_zeiger1_2020-04-29_14-44-02.jpg 0.86 0.7564603084968251 0.10353969150317488\n",
      "data_resize_all\\8.6_4219_zeiger2_2019-06-03T183009.jpg 0.86 0.7707364283905273 0.08926357160947274\n",
      "data_resize_all\\8.6_4236_zeiger3_2019-06-03T123009.jpg 0.86 0.5897344042739336 0.27026559572606634\n",
      "data_resize_all\\8.6_4237_zeiger3_2019-06-03T221010.jpg 0.86 0.5862509698517318 0.27374903014826824\n",
      "data_resize_all\\8.6_4238_zeiger3_2019-06-03T222009.jpg 0.86 0.5848689190186751 0.2751310809813249\n",
      "data_resize_all\\8.6_4239_zeiger3_2019-06-03T223009.jpg 0.86 0.5720226233367244 0.2879773766632756\n",
      "data_resize_all\\8.6_4240_zeiger3_2019-06-04T005009.jpg 0.86 0.6444406905946647 0.2155593094053353\n",
      "data_resize_all\\8.6_4285_zeiger2_2019-06-01T202014.jpg 0.86 0.7943056657234533 0.0656943342765467\n",
      "data_resize_all\\8.6_4286_zeiger2_2019-06-02T104009.jpg 0.86 0.8030444184477923 0.0569555815522077\n",
      "data_resize_all\\8.6_pointer_20211008-072631.jpg 0.86 0.9241021363442282 -0.06410213634422823\n",
      "data_resize_all\\8.7_4304_zeiger2_2019-06-06T050009.jpg 0.8699999999999999 0.7944452672009893 0.0755547327990106\n",
      "data_resize_all\\8.7_4308_zeiger3_2019-06-03T175009.jpg 0.8699999999999999 0.6208464851257727 0.2491535148742272\n",
      "data_resize_all\\8.7_4329_zeiger3_2019-06-05T153009.jpg 0.8699999999999999 0.8159349206318711 0.054065079368128766\n",
      "data_resize_all\\8.8_4326_zeiger2_2019-06-04T225009.jpg 0.8800000000000001 0.8005700116662259 0.07942998833377424\n",
      "data_resize_all\\8.8_4327_zeiger2_2019-06-06T104009.jpg 0.8800000000000001 0.7763128225018979 0.10368717749810219\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data_resize_all\\8.9_4325_zeiger2_2019-06-04T224009.jpg 0.89 0.7630451800391482 0.12695481996085178\n",
      "data_resize_all\\8.9_4406_zeiger2_2019-06-03T184010.jpg 0.89 0.81336517579314 0.07663482420685996\n",
      "data_resize_all\\8.9_4407_zeiger2_2019-06-03T185009.jpg 0.89 0.7841642198165661 0.10583578018343387\n",
      "data_resize_all\\8.9_4408_zeiger2_2019-06-03T190009.jpg 0.89 0.7130307593595236 0.17696924064047637\n",
      "data_resize_all\\8.9_4409_zeiger2_2019-06-03T191009.jpg 0.89 0.8007077917599832 0.08929220824001682\n",
      "data_resize_all\\8.9_analog3_20200928-135043.jpg 0.89 0.8309497444570579 0.059050255542942076\n",
      "data_resize_all\\9.0_4438_zeiger1_2020-04-29_13-26-28.jpg 0.9 0.7799286487575119 0.12007135124248813\n",
      "data_resize_all\\9.0_PRODUCED_ANALOG.jpg 0.9 0.75045594611571 0.14954405388429004\n",
      "data_resize_all\\9.0_PRODUCED_ANALOG_2.jpg 0.9 0.6519294238371528 0.24807057616284722\n",
      "data_resize_all\\9.1_4454_zeiger2_2019-06-05T010010.jpg 0.9099999999999999 0.857186350901977 0.052813649098022886\n",
      "data_resize_all\\9.1_4528_zeiger1_2020-04-29_12-09-01.jpg 0.9099999999999999 0.8325968825209018 0.07740311747909812\n",
      "data_resize_all\\9.1_4532_zeiger2_2019-06-02T183009.jpg 0.9099999999999999 0.8477123947181032 0.062287605281896696\n",
      "data_resize_all\\9.1_4534_zeiger2_2019-06-06T105009.jpg 0.9099999999999999 0.8509942152369296 0.05900578476307028\n",
      "data_resize_all\\9.1_4535_zeiger2_2019-06-06T110009.jpg 0.9099999999999999 0.8474762693866715 0.06252373061332839\n",
      "data_resize_all\\9.1_PRODUCED_ANALOG.jpg 0.9099999999999999 0.7999094564824949 0.11009054351750502\n",
      "data_resize_all\\9.2_4530_zeiger1_2020-04-29_13-28-02.jpg 0.9199999999999999 0.8657031035055429 0.054296896494457014\n",
      "data_resize_all\\9.2_4552_zeiger3_2019-06-04T224009.jpg 0.9199999999999999 0.8567181034888773 0.06328189651112259\n",
      "data_resize_all\\9.2_4571_zeiger4_2020-04-29_12-45-02.jpg 0.9199999999999999 0.861621344512048 0.05837865548795196\n",
      "data_resize_all\\9.2_4584_zeiger2_2019-09-14_19-20-12.jpg 0.9199999999999999 0.6650703662907989 0.2549296337092011\n",
      "data_resize_all\\9.2_4585_zeiger2_2019-09-14_19-30-13.jpg 0.9199999999999999 0.7097699511587687 0.2102300488412312\n",
      "data_resize_all\\9.2_PRODUCED_ANALOG_2.jpg 0.9199999999999999 0.4433544851056885 0.47664551489431145\n",
      "data_resize_all\\9.3_4529_zeiger1_2020-04-29_12-10-02.jpg 0.93 0.8678528486285614 0.06214715137143867\n",
      "data_resize_all\\9.3_4531_zeiger1_2020-04-29_13-29-01.jpg 0.93 0.7665885873168123 0.1634114126831877\n",
      "data_resize_all\\9.3_4580_zeiger2_2019-06-02T185009.jpg 0.93 0.82026533645665 0.10973466354335004\n",
      "data_resize_all\\9.3_4581_zeiger2_2019-06-02T190009.jpg 0.93 0.8509371736709863 0.07906282632901374\n",
      "data_resize_all\\9.3_4633_zeiger2_2019-06-02T203009.jpg 0.93 0.8646749323624837 0.06532506763751633\n",
      "data_resize_all\\9.3_4639_zeiger2_2019-06-02T213009.jpg 0.93 0.8748734672671836 0.055126532732816425\n",
      "data_resize_all\\9.3_4642_zeiger2_2019-06-02T220009.jpg 0.93 0.8633158764996327 0.06668412350036734\n",
      "data_resize_all\\9.3_4647_zeiger2_2019-06-02T225009.jpg 0.93 0.8601605757488069 0.06983942425119316\n",
      "data_resize_all\\9.3_4657_zeiger2_2019-06-03T003009.jpg 0.93 0.8484963676255868 0.08150363237441327\n",
      "data_resize_all\\9.3_4658_zeiger2_2019-06-03T004009.jpg 0.93 0.864346713313743 0.06565328668625703\n",
      "data_resize_all\\9.3_4659_zeiger2_2019-06-03T005009.jpg 0.93 0.8628980776226526 0.06710192237734747\n",
      "data_resize_all\\9.3_4661_zeiger2_2019-06-03T011009.jpg 0.93 0.8551675746562937 0.07483242534370638\n",
      "data_resize_all\\9.3_4665_zeiger2_2019-06-03T015009.jpg 0.93 0.8289908082576029 0.1010091917423972\n",
      "data_resize_all\\9.3_4669_zeiger2_2019-06-03T023009.jpg 0.93 0.8659533350748854 0.06404666492511468\n",
      "data_resize_all\\9.3_4670_zeiger2_2019-06-03T024009.jpg 0.93 0.8638553286348298 0.06614467136517022\n",
      "data_resize_all\\9.3_4671_zeiger2_2019-06-03T025009.jpg 0.93 0.858597885384951 0.07140211461504908\n",
      "data_resize_all\\9.3_4673_zeiger2_2019-06-03T192009.jpg 0.93 0.8695477410983523 0.06045225890164774\n",
      "data_resize_all\\9.3_4675_zeiger2_2019-06-03T194009.jpg 0.93 0.8765709684897826 0.053429031510217495\n",
      "data_resize_all\\9.3_4676_zeiger2_2019-06-03T195009.jpg 0.93 0.8674729193559794 0.06252708064402068\n",
      "data_resize_all\\9.3_4678_zeiger2_2019-06-03T202010.jpg 0.93 0.8401371836916236 0.08986281630837645\n",
      "data_resize_all\\9.3_PRODUCED_ANALOG.jpg 0.93 1.428127147533576 -0.4981271475335759\n",
      "data_resize_all\\9.3_analog2_20201003-152354.jpg 0.93 0.8269504599977908 0.10304954000220923\n",
      "data_resize_all\\9.4_4579_zeiger1_2020-04-29_12-11-02.jpg 0.9400000000000001 0.8798116560916074 0.06018834390839267\n",
      "data_resize_all\\9.4_4751_zeiger2_2019-09-15_05-40-13.jpg 0.9400000000000001 0.8893496832444456 0.050650316755554425\n",
      "data_resize_all\\9.4_4755_zeiger3_2019-06-04T110009.jpg 0.9400000000000001 0.833127208764458 0.10687279123554205\n",
      "data_resize_all\\9.4_4759_zeiger3_2019-06-04T114009.jpg 0.9400000000000001 0.8565478894829364 0.0834521105170637\n",
      "data_resize_all\\9.5_4775_zeiger1_2020-04-29_12-12-01.jpg 0.95 0.8463699874413542 0.10363001255864579\n",
      "data_resize_all\\9.5_4793_zeiger1_2020-04-29_13-31-02.jpg 0.95 0.8651575231728404 0.08484247682715951\n",
      "data_resize_all\\9.6.jpg 0.96 0.6540500837379556 0.3059499162620444\n",
      "data_resize_all\\9.6_4807_zeigeru__2019-06-03T03100.jpg 0.96 0.815201718797218 0.14479828120278193\n",
      "data_resize_all\\9.6_4808_zeigeru__2019-06-03T03200.jpg 0.96 0.8801786744093834 0.07982132559061661\n",
      "data_resize_all\\9.6_4809_zeigeru__2019-06-03T03300.jpg 0.96 0.8826564393672915 0.07734356063270842\n",
      "data_resize_all\\9.6_4810_zeigeru__2019-06-03T03400.jpg 0.96 0.8236878304132573 0.13631216958674264\n",
      "data_resize_all\\9.6_4811_zeigeru__2019-06-03T03500.jpg 0.96 0.8618399200518622 0.0981600799481378\n",
      "data_resize_all\\9.6_4812_zeigeru__2019-06-03T04000.jpg 0.96 0.7850141609677834 0.17498583903221654\n",
      "data_resize_all\\9.6_4813_zeigeru__2019-06-03T04100.jpg 0.96 0.8057970226288028 0.15420297737119715\n",
      "data_resize_all\\9.6_4814_zeigeru__2019-06-03T04200.jpg 0.96 0.8434556690453697 0.11654433095463024\n",
      "data_resize_all\\9.6_4815_zeigeru__2019-06-03T04300.jpg 0.96 0.8200243825609301 0.13997561743906983\n"
     ]
    }
   ],
   "source": [
    "Input_dir='data_resize_all'\n",
    "#Input_dir='test_result'\n",
    "files = glob.glob(Input_dir + '/*.*')\n",
    "res = []\n",
    "stat_Anz = []\n",
    "stat_Abweichung = []\n",
    "i = 0\n",
    "deviation_max_list = 0.05\n",
    "\n",
    "for i in range(100):\n",
    "    stat_Anz.append(0)\n",
    "    stat_Abweichung.append(0)\n",
    "\n",
    "for aktfile in sorted(files):\n",
    "    base = os.path.basename(aktfile)\n",
    "    target = (float(base[0:3])) / 10\n",
    "    \n",
    "    target_sin = math.sin(target * math.pi * 2)\n",
    "    target_cos = math.cos(target * math.pi * 2)\n",
    "\n",
    "    test_image = Image.open(aktfile)\n",
    "    test_image = np.array(test_image, dtype=\"float32\")\n",
    "    img = np.reshape(test_image,[1,32,32,3])\n",
    "    classes = model.predict(img, verbose=0)\n",
    "    \n",
    "    out_sin = classes[0][0]  \n",
    "    out_cos = classes[0][1]\n",
    "    out_target = (np.arctan2(out_sin, out_cos)/(2*math.pi)) % 1\n",
    "\n",
    "    dev_sin = target_sin - out_sin\n",
    "    dev_cos = target_cos - out_cos\n",
    "    dev_target = target - out_target\n",
    "    \n",
    "    if abs(dev_target + 1) < abs(dev_target):\n",
    "        out_target = out_target - 1\n",
    "        dev_target = target - out_target\n",
    "    else:\n",
    "        if abs(dev_target - 1) < abs(dev_target):\n",
    "            out_target = out_target + 1\n",
    "            dev_target = target - out_target\n",
    "            \n",
    "    target_int = int ((float(base[0:3])) * 10)\n",
    "    stat_Abweichung[target_int] = stat_Abweichung[target_int] + dev_target  \n",
    "    stat_Anz[target_int] = stat_Anz[target_int] + 1\n",
    "               \n",
    "    res.append(np.array([target, out_target, dev_target, out_sin, out_cos, i]))\n",
    "    if abs(dev_target) > deviation_max_list:\n",
    "        print(aktfile + \" \" + str(target) + \" \" + str(out_target) +  \" \" + str(dev_target))\n",
    "\n",
    "    \n",
    "for i in range(100):\n",
    "    stat_Abweichung[i] = stat_Abweichung[i] / stat_Anz[i]\n",
    "\n",
    "res = np.asarray(res)\n",
    "res_step_1 = res"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(res[:,3])\n",
    "plt.plot(res[:,4])\n",
    "plt.title('Result')\n",
    "plt.ylabel('value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['sin', 'cos'], loc='lower left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(res[:,0])\n",
    "plt.plot(res[:,1])\n",
    "plt.title('Result')\n",
    "plt.ylabel('Counter Value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['Orginal', 'Prediction'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(stat_Abweichung)\n",
    "plt.title('Result')\n",
    "plt.ylabel('Counter Value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['Orginal', 'Prediction'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(stat_Anz)\n",
    "plt.title('Result')\n",
    "plt.ylabel('Counter Value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['Orginal', 'Prediction'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Deviation from Expected Value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.0023  0.0269 -0.4981  0.4766]\n"
     ]
    }
   ],
   "source": [
    "plt.plot(res[:,2])\n",
    "plt.title('Deviation')\n",
    "plt.ylabel('Deviation from expected value')\n",
    "plt.xlabel('#Picture')\n",
    "plt.legend(['Deviation'], loc='upper left')\n",
    "#plt.ylim(-0.3, 0.3)\n",
    "plt.show()\n",
    "\n",
    "statistic = np.array([np.mean(res[:,2]), np.std(res[:,2]), np.min(res[:,2]), np.max(res[:,2])])\n",
    "print(statistic)"
   ]
  },
  {
   "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": 14,
   "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\\tmpv81gd546\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmpv81gd546\\assets\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "125772"
      ]
     },
     "execution_count": 14,
     "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": 15,
   "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\\tmp_cicoo4i\\assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: C:\\Users\\Muell\\AppData\\Local\\Temp\\tmp_cicoo4i\\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": [
      "ana1100s3_longq.tflite\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "38544"
      ]
     },
     "execution_count": 15,
     "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[n], 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": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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"
  },
  "vscode": {
   "interpreter": {
    "hash": "3431513051aa6ef6a02e4978d18932220794b1f41fa21fbae2ad068757488314"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}