model_trainer_v4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "args = {\n",
    "    'model_name':'2019_04_11_DRIVE',\n",
    "    'FP':'float16',\n",
    "    'optimizer': 'Adam', #SGD, RMSprop, Adadelta, Adagrad, Adam, Adamax, Nadam\n",
    "    'dataset':['101/training_data.csv', '102/training_data.csv', '103/training_data.csv', '104/training_data.csv', '105/training_data.csv', '106/training_data.csv', '107/training_data.csv'],\n",
    "    'batch_size':1024, #512 with float32, 1024 with float16\n",
    "    'split_point': 0.8, #80% Training , 20% Validation\n",
    "    'rnd_seed': 1234,\n",
    "    'epochs_number':250,\n",
    "    'image_width': 320,\n",
    "    'image_height': 90,\n",
    "    'resume': False\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# csv uznatılı dosyayı okuyabilmek için\n",
    "import pandas as pd\n",
    "# matris işlemleri için\n",
    "import numpy as np\n",
    "# opencv \n",
    "import cv2\n",
    "# grafik kütüphanesi \n",
    "import matplotlib.pylab as plt\n",
    "# histogram colors\n",
    "from matplotlib import colors\n",
    "# matplotlib grafiklerini jupyter notebook üzerinde göstermek için\n",
    "%matplotlib inline\n",
    "# rasgele sayı üretimi için\n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Eğitim için kaydettiğimiz seyir.csv dosaysını okuyoruz\n",
    "for i, ds in enumerate(args['dataset']):\n",
    "    try:\n",
    "        tmp = pd.read_csv(ds)\n",
    "    except:\n",
    "        pass\n",
    "    else:\n",
    "        tmp['FileName'] = tmp['FileName'].apply(lambda x: ds.split('/')[0] + '/' + x)  \n",
    "        if i == 0:\n",
    "            df = tmp\n",
    "        else:\n",
    "            df = df.append(tmp, ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([u'FileName', u'Speed', u'Angle'], dtype='object')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Seyir dosaysındaki sutun başlıkları\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    33929.000000\n",
       "mean        -0.023694\n",
       "std          0.138602\n",
       "min         -0.340000\n",
       "25%         -0.102781\n",
       "50%          0.000000\n",
       "75%          0.000000\n",
       "max          0.340000\n",
       "Name: Angle, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Açı sutunu hakkında istatistiki bilgiler min max mean vs ...\n",
    "df.Angle.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33929"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Toplam kayıt sayımız\n",
    "len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>FileName</th>\n",
       "      <th>Speed</th>\n",
       "      <th>Angle</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>101/00000.jpg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>101/00001.jpg</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>101/00002.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>101/00003.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>101/00004.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>101/00005.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>101/00006.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>101/00007.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>101/00008.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>101/00009.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        FileName  Speed  Angle\n",
       "0  101/00000.jpg    0.0    0.0\n",
       "1  101/00001.jpg    0.0    0.0\n",
       "2  101/00002.jpg    1.0    0.0\n",
       "3  101/00003.jpg    1.0    0.0\n",
       "4  101/00004.jpg    1.0    0.0\n",
       "5  101/00005.jpg    1.0    0.0\n",
       "6  101/00006.jpg    1.0    0.0\n",
       "7  101/00007.jpg    1.0    0.0\n",
       "8  101/00008.jpg    1.0    0.0\n",
       "9  101/00009.jpg    1.0    0.0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 20 den 40 kadar kayıtlar\n",
    "df[0:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>FileName</th>\n",
       "      <th>Speed</th>\n",
       "      <th>Angle</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>33919</th>\n",
       "      <td>107/04541.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33920</th>\n",
       "      <td>107/04542.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33921</th>\n",
       "      <td>107/04543.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.046131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33922</th>\n",
       "      <td>107/04544.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.272732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33923</th>\n",
       "      <td>107/04545.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33924</th>\n",
       "      <td>107/04546.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33925</th>\n",
       "      <td>107/04547.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33926</th>\n",
       "      <td>107/04548.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33927</th>\n",
       "      <td>107/04549.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>-0.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33928</th>\n",
       "      <td>107/04550.jpg</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            FileName  Speed     Angle\n",
       "33919  107/04541.jpg    1.0  0.000000\n",
       "33920  107/04542.jpg    1.0  0.000000\n",
       "33921  107/04543.jpg    1.0 -0.046131\n",
       "33922  107/04544.jpg    1.0 -0.272732\n",
       "33923  107/04545.jpg    1.0 -0.340000\n",
       "33924  107/04546.jpg    1.0 -0.340000\n",
       "33925  107/04547.jpg    1.0 -0.340000\n",
       "33926  107/04548.jpg    1.0 -0.340000\n",
       "33927  107/04549.jpg    1.0 -0.340000\n",
       "33928  107/04550.jpg    1.0  0.000000"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[-10:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fd247ea7910>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Kayıt ettğimiz resmin birine bakıyoruz\n",
    "image = cv2.imread(df.FileName[30])\n",
    "plt.figure(figsize=(15,5))\n",
    "plt.imshow(image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<type 'numpy.uint8'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Resmin her tarafını kullanmamıza gerek yok \n",
    "# çeşitli denemeler yaparak kırptığımız alanı görüyoruz\n",
    "# Uygun gördüğümüz halini ağda kullanacağız\n",
    "# Biz burda sadece sol görüntüyü  üstten 144 den 300 kadarlık kısmını alıyoruz \n",
    "tmp = image[160:340,:,:]\n",
    "plt.figure(figsize=(15,5))\n",
    "plt.imshow(tmp)\n",
    "print(type(image[0,0,0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Ağımızı eğitmek için  giriş verimiz resimler  çıkışımız da Açılar olacak\n",
    "# bunları birer listeye aktarıyoruz\n",
    "# Kamera ilk çekmeye başladığında düzgün çekemediği için başlangıçdan 30 kayıdı almıyoruz.\n",
    "images = list(df.FileName[1:-1])\n",
    "labels = list(df.Angle[1:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33927"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Verimizdeki açıların dağılımı nasıl diye bir histogram yapıp bakıyoruz\n",
    "# Dağılımın eşit şekilde olmaması eğitimin de düzgün olmamasına sebep olur\n",
    "#plt.hist(labels, bins=14)\n",
    "#plt.show()\n",
    "N, bins, patches = plt.hist(labels,14, range=[-0.35,0.35], facecolor='blue', align='mid')\n",
    "fracs = N / N.max()\n",
    "norm = colors.Normalize(fracs.min(), fracs.max())\n",
    "\n",
    "# Now, we'll loop through our objects and set the color of each accordingly\n",
    "for thisfrac, thispatch in zip(fracs, patches):\n",
    "    color = plt.cm.RdBu(norm(thisfrac))\n",
    "    thispatch.set_facecolor(color)\n",
    "\n",
    "#plt.axis([-0.4, 0.4, 0, 750])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1738, 452, 1393, 2261, 2687, 1783, 802]\n",
      "[16694, 1158, 1879, 1074, 631, 267, 1110]\n"
     ]
    }
   ],
   "source": [
    "tmp = []\n",
    "start_value = -0.35\n",
    "aralik = 0.05\n",
    "\n",
    "for i in range(14):\n",
    "    length=df.loc[(df['Angle'] > start_value) & (df['Angle'] <= start_value+aralik)]\n",
    "    tmp.append(len(length))\n",
    "    start_value = start_value + aralik\n",
    "print(tmp[0:7]) # Eksi açı degerleri (Sag)\n",
    "print(tmp[7:14]) # Pozitif açı degerleri (Sol)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Veri setindeki açı dağılımını bir paröa düzeltmek için\n",
    "# sayısı az olan açıdaki kayıtları listeye yeniden ekleyerek \n",
    "# daha düzgün hale getirmeye çalışıyoruz\n",
    "def augment():\n",
    "    nitem = len(images)\n",
    "    for i in range(nitem):\n",
    "        if labels[i] >= 0.05 and labels[i] <= 0.1:\n",
    "            addValue=tmp[7]/tmp[8]\n",
    "            for j in range(addValue-5):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] > 0.1 and labels[i] <= 0.15:\n",
    "            addValue=tmp[7]/tmp[9]\n",
    "            for j in range(addValue-3):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i]) \n",
    "\n",
    "        if labels[i] > 0.15  and labels[i] <= 0.2:\n",
    "            addValue=tmp[7]/tmp[10]\n",
    "            for j in range(addValue-5):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] > 0.2  and labels[i] <= 0.25:\n",
    "            addValue=tmp[7]/tmp[11]\n",
    "            for j in range(addValue-10):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] > 0.25  and labels[i] <= 0.3:\n",
    "            addValue=tmp[7]/tmp[12]\n",
    "            for j in range(addValue-20):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] > 0.3  and labels[i] <= 0.35:\n",
    "            addValue=tmp[7]/tmp[13]\n",
    "            for j in range(addValue-5):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "\n",
    "        #Negatif degerler\n",
    "\n",
    "        if labels[i] < 0.0 and labels[i] > -0.05:\n",
    "            addValue=tmp[7]/tmp[6]\n",
    "            for j in range(addValue-5):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.05 and labels[i] > -0.1:\n",
    "            addValue=tmp[7]/tmp[5]\n",
    "            for j in range(addValue-3):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.1 and labels[i] > -0.15:\n",
    "            addValue=tmp[7]/tmp[4]\n",
    "            for j in range(addValue-3):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.15 and labels[i] > -0.2:\n",
    "            addValue=tmp[7]/tmp[3]\n",
    "            for j in range(addValue-3):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.2 and labels[i] > -0.25:\n",
    "            addValue=tmp[7]/tmp[2]\n",
    "            for j in range(addValue-4):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.25 and labels[i] > -0.3:\n",
    "            addValue=tmp[7]/tmp[1]\n",
    "            for j in range(addValue-10):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])\n",
    "\n",
    "        if labels[i] <= -0.3 and labels[i] > -0.35:\n",
    "            addValue=tmp[7]/tmp[0]\n",
    "            for j in range(addValue-3):\n",
    "                images.append(images[i])\n",
    "                labels.append(labels[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "augment()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# İlk histgorama göre daga dengeli sayılabilecek bir dağılıma ulaştık\n",
    "# En doğru çözüm değil ama pratik işe yarar bir alternatif\n",
    "N, bins, patches = plt.hist(labels,14, range=[-0.35,0.35], facecolor='blue', align='mid')\n",
    "fracs = N / N.max()\n",
    "norm = colors.Normalize(fracs.min(), fracs.max())\n",
    "\n",
    "# Now, we'll loop through our objects and set the color of each accordingly\n",
    "for thisfrac, thispatch in zip(fracs, patches):\n",
    "    color = plt.cm.RdBu(norm(thisfrac))\n",
    "    thispatch.set_facecolor(color)\n",
    "\n",
    "#plt.axis([-0.4, 0.4, 0, 750])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "168658"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Veri setimiz ile ilgili ayarlamalar\n",
    "# Veri seti küme büyüklüğü batch size\n",
    "# Verisetinin ne kadarı eğitim ne kadarı test için kullanılacak\n",
    "# Eğitim %80 , Test %20 \n",
    "bsize = args['batch_size']\n",
    "dlen = len(labels)\n",
    "splitpoint = int(args['split_point']*dlen)\n",
    "reindex = list(range(len(labels)))\n",
    "# Eğtim verisini karıştıryoruz\n",
    "random.seed(args['rnd_seed'])\n",
    "random.shuffle(reindex)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Resim üzerinde Rastgele parlaklık değişimi uygulayan bir fonksiyon\n",
    "# Augmentation function (taken from github)\n",
    "def augment_brightness(image):\n",
    "    image1 = cv2.cvtColor(image,cv2.COLOR_BGR2HSV) \n",
    "    image1 = np.array(image1, dtype = np.float64)\n",
    "    random_bright = .5+np.random.uniform()\n",
    "    image1[:,:,2] = image1[:,:,2]*random_bright\n",
    "    image1[:,:,2][image1[:,:,2]>255]  = 255\n",
    "    image1 = np.array(image1, dtype = np.uint8)\n",
    "    image1 = cv2.cvtColor(image1,cv2.COLOR_HSV2RGB)\n",
    "    return image1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Resmi Kaydirma Sonradan Eklendi\n",
    "def random_translate(image,range_x, range_y):\n",
    "    \"\"\"\n",
    "    Randomly shift the image virtially and horizontally (translation).\n",
    "    \"\"\"\n",
    "    trans_x = range_x * (np.random.rand() - 0.5)\n",
    "    trans_y = range_y * (np.random.rand() - 0.5)\n",
    "    trans_m = np.float32([[1, 0, trans_x], [0, 1, trans_y]])\n",
    "    height, width = image.shape[:2]\n",
    "    image = cv2.warpAffine(image, trans_m, (width, height))\n",
    "    return image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def random_shadow(image,width,heigth):\n",
    "    \"\"\"\n",
    "    Generates and adds random shadow\n",
    "    \"\"\"\n",
    "    \n",
    "    IMAGE_WIDTH=width\n",
    "    IMAGE_HEIGHT=heigth\n",
    "    # (x1, y1) and (x2, y2) forms a line\n",
    "    # xm, ym gives all the locations of the image\n",
    "    x1, y1 = IMAGE_WIDTH * np.random.rand(), 0\n",
    "    x2, y2 = IMAGE_WIDTH * np.random.rand(), IMAGE_HEIGHT\n",
    "    xm, ym = np.mgrid[0:IMAGE_HEIGHT, 0:IMAGE_WIDTH]\n",
    "\n",
    "    print(image.size)\n",
    "    # mathematically speaking, we want to set 1 below the line and zero otherwise\n",
    "    # Our coordinate is up side down.  So, the above the line: \n",
    "    # (ym-y1)/(xm-x1) > (y2-y1)/(x2-x1)\n",
    "    # as x2 == x1 causes zero-division problem, we'll write it in the below form:\n",
    "    # (ym-y1)*(x2-x1) - (y2-y1)*(xm-x1) > 0\n",
    "    mask = np.zeros_like(image[:, :, 1])\n",
    "    mask[(ym - y1) * (x2 - x1) - (y2 - y1) * (xm - x1) > 0] = 1\n",
    "\n",
    "    # choose which side should have shadow and adjust saturation\n",
    "    cond = mask == np.random.randint(2)\n",
    "    s_ratio = np.random.uniform(low=0.5, high=0.5)\n",
    "\n",
    "    # adjust Saturation in HLS(Hue, Light, Saturation)\n",
    "    hls = cv2.cvtColor(image, cv2.COLOR_RGB2HLS)\n",
    "    hls[:, :, 1][cond] = hls[:, :, 1][cond] * s_ratio\n",
    "    #plt.imshow(random_shadow(image))\n",
    "    return cv2.cvtColor(hls, cv2.COLOR_HLS2RGB)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ismi verilen resmi okuyup \n",
    "# rastgele olarak %50 sine parlaklık değişimi uygulayan fonksiyonu uygulayıp\n",
    "# resim matrisini dönem bir fonksiyon\n",
    "\n",
    "def get_matrix(fname):\n",
    "    img = cv2.imread(fname)\n",
    "    h, w = img.shape[:2]\n",
    "    #img = img[160:340,:,:] # crop only\n",
    "    if w != 640 or h != 360:\n",
    "        img = cv2.resize(img, (640,360))\n",
    "    img = cv2.resize(img[160:340,:,:], (args['image_width'],args['image_height'])) #crop then resize\n",
    "    if random.randint(0,2) == 1:\n",
    "        img = augment_brightness(img)   \n",
    "    if random.randint(0,2) == 1:\n",
    "        img = random_translate(img,100,0)  \n",
    "    #if random.randint(0,2) == 1:\n",
    "    #    img = random_shadow(img,320,90)\n",
    "    return img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Bütün veriyi hafızaya almamız mümkün değil\n",
    "# Ek olarak bazen çeşitli değişimler - Augmentation - uygulamakda istiyebiliriz\n",
    "# python generator ile gerektiğinde veri okunur düzenlenir ve eğitim veya test için \n",
    "# sisteme verilir\n",
    "# alttaki fonksiyonlar bu işi yapar\n",
    "\n",
    "# Generate data for training\n",
    "def generate_data():\n",
    "    i = 0\n",
    "    while True:\n",
    "        x = []\n",
    "        y = []\n",
    "        for j in range(i,i+bsize):  \n",
    "            ix = reindex[j]\n",
    "            img = get_matrix(images[ix])\n",
    "            lbl = np.array([labels[ix]])\n",
    "            flip = random.randint(0,1)\n",
    "            if flip == 1:\n",
    "                img = cv2.flip(img,1)\n",
    "                lbl = lbl*-1.0\n",
    "            x.append(img)\n",
    "            y.append(lbl)\n",
    "        x = np.array(x)\n",
    "        y = np.array(y)\n",
    "        #print(\"#------ Sending TRAINING batch ------#\")\n",
    "        yield (x,y)    \n",
    "        i +=bsize\n",
    "        if i+bsize > splitpoint:\n",
    "            i = 0\n",
    "            \n",
    "# Generate data for validation                  \n",
    "def generate_data_val():\n",
    "    i = splitpoint\n",
    "    while True:\n",
    "        x = []\n",
    "        y = []\n",
    "        for j in range(i,i+bsize): \n",
    "            ix = reindex[j]\n",
    "            x.append(get_matrix(images[ix]))\n",
    "            y.append(np.array([labels[ix]]))\n",
    "        x = np.array(x)\n",
    "        y = np.array(y)\n",
    "        #print(\"#------ Sending VALIDATION batch ------#\")\n",
    "        yield (x,y)    \n",
    "        i +=bsize\n",
    "        if i+bsize > dlen:\n",
    "            i = splitpoint\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# Keras için gerekenler\n",
    "from keras.models import Sequential, load_model\n",
    "from keras.layers import Dense, Dropout, Flatten, Lambda\n",
    "from keras.layers import Conv2D, MaxPooling2D, Cropping2D, Reshape\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.optimizers import SGD, RMSprop, Adadelta, Adagrad, Adam, Adamax, Nadam\n",
    "from keras.regularizers import l2\n",
    "from keras import backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'float16'"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Destroy the current TF graph and create a new one\n",
    "K.clear_session()\n",
    "#import keras\n",
    "#print(keras.__version__)\n",
    "# make sure soft-placement is off\n",
    "\n",
    "# allow_soft_placement: an op will be placed on CPU if not possible on GPU\n",
    "# allow_growth: attempts to allocate only as much GPU memory based on runtime allocations \n",
    "# per_process_gpu_memory_fraction: set the fraction of the overall memory that each GPU should be allocated\n",
    "\n",
    "tf_config = tf.ConfigProto(allow_soft_placement=False)\n",
    "tf_config.gpu_options.allow_growth = True\n",
    "#tf_config.gpu_options.per_process_gpu_memory_fraction = 0.5\n",
    "s = tf.Session(config=tf_config)\n",
    "K.set_session(s)\n",
    "\n",
    "# enable 16-bit training\n",
    "K.set_floatx(args['FP'])\n",
    "if args['FP'] == \"float16\":\n",
    "    K.set_epsilon(1e-4)\n",
    "K.floatx()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Model based on NVIDIA's End to End Learning for Self-Driving Cars model\n",
    "#input shape, original paper uses 60*200, openzeka uses 80*320\n",
    "shape=(args['image_height'],args['image_width'],3)\n",
    "# Sıralı bir keras modeli tanılıyoruz\n",
    "model = Sequential()\n",
    "# Cropping\n",
    "# gelen resmin istediğimiz ksımını kırpmak için bu katmanı yazıyoruz\n",
    "# Cropping2D(cropping((top_crop, bottom_crop), (left_crop, right_crop)))\n",
    "# aşağıdaki satırda \n",
    "# üstten 150 alttan 20 piksel \n",
    "# soldan 0 sağdan 640 piksel kırpılıyor\n",
    "#model.add(Cropping2D(cropping=((150,20),(0,640)), input_shape=shape))\n",
    "# Normalization\n",
    "# 0 - 255 arası değerler -1 ila 1 arasına çekiliyor\n",
    "model.add(Lambda(lambda x: (2*x / 255.0) - 1.0, input_shape=shape))\n",
    "# lambda + cast\n",
    "#model.add(Lambda(lambda x: tf.cast((2*x / 255.0) - 1.0 ,dtype=tf.float16)))\n",
    "# cast to float16\n",
    "#model.add(Lambda(lambda x: tf.cast(x,dtype=tf.float16)))\n",
    "# Doesnt work, requires numpy array as input\n",
    "#model.add(Lambda(lambda x: K.cast_to_floatx(x)))\n",
    "\n",
    "# Evrişim katmanı (5, 5) lik 24 tane 2 şer piksel kayarak\n",
    "model.add(Conv2D(24, (5, 5), activation=\"relu\", strides=(2, 2), kernel_regularizer=l2(0.001)))\n",
    "model.add(Conv2D(36, (5, 5), activation=\"relu\", strides=(2, 2), kernel_regularizer=l2(0.001)))\n",
    "model.add(Conv2D(48, (5, 5), activation=\"relu\", strides=(2, 2), kernel_regularizer=l2(0.001)))\n",
    "model.add(Conv2D(64, (3, 3), activation=\"relu\", kernel_regularizer=l2(0.001)))\n",
    "model.add(Conv2D(64, (3, 3), activation=\"relu\", kernel_regularizer=l2(0.001)))\n",
    "#Flattendan once kullanilmayan baglantilari drop out yaptik(sonradan ekledik)\n",
    "model.add(Dropout(0.5))\n",
    "# Ağın çıkışı burda vectöre çevriliyor\n",
    "model.add(Flatten())\n",
    "# Yapay Sinir ağı kısmı\n",
    "model.add(Dense(100))\n",
    "model.add(Dense(50))\n",
    "model.add(Dense(10))\n",
    "# Ağın çıkışı Açı \n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mse', optimizer=args['optimizer'], metrics=['mse', 'msle', 'mae', 'mape', 'cosine'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lambda_1 (Lambda)            (None, 90, 320, 3)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_1 (Conv2D)            (None, 43, 158, 24)       1824      \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 20, 77, 36)        21636     \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 8, 37, 48)         43248     \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 6, 35, 64)         27712     \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 4, 33, 64)         36928     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 4, 33, 64)         0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 8448)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 100)               844900    \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 50)                5050      \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 10)                510       \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 1)                 11        \n",
      "=================================================================\n",
      "Total params: 981,819\n",
      "Trainable params: 981,819\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Tanımladığımız ağın yapsı\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"lambda_1/sub:0\", shape=(?, 90, 320, 3), dtype=float16)\n",
      "Tensor(\"conv2d_1/Relu:0\", shape=(?, 43, 158, 24), dtype=float16)\n",
      "Tensor(\"conv2d_2/Relu:0\", shape=(?, 20, 77, 36), dtype=float16)\n",
      "Tensor(\"conv2d_3/Relu:0\", shape=(?, 8, 37, 48), dtype=float16)\n",
      "Tensor(\"conv2d_4/Relu:0\", shape=(?, 6, 35, 64), dtype=float16)\n",
      "Tensor(\"conv2d_5/Relu:0\", shape=(?, 4, 33, 64), dtype=float16)\n",
      "Tensor(\"dropout_1/cond/Merge:0\", shape=(?, 4, 33, 64), dtype=float16)\n",
      "Tensor(\"flatten_1/Reshape:0\", shape=(?, ?), dtype=float16)\n",
      "Tensor(\"dense_1/BiasAdd:0\", shape=(?, 100), dtype=float16)\n",
      "Tensor(\"dense_2/BiasAdd:0\", shape=(?, 50), dtype=float16)\n",
      "Tensor(\"dense_3/BiasAdd:0\", shape=(?, 10), dtype=float16)\n",
      "Tensor(\"dense_4/BiasAdd:0\", shape=(?, 1), dtype=float16)\n"
     ]
    }
   ],
   "source": [
    "for layer in model.layers:\n",
    "    print(layer.output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<type 'float'>\n",
      "<type 'numpy.float16'>\n"
     ]
    }
   ],
   "source": [
    "# Açı değerlerinide -0.3 ila 0.3 aralığından -1 ila 1 aralığına çekebilmek için 3 ile çarpıyoruz\n",
    "print(type(labels[1]))\n",
    "labels = 3*np.array(labels).astype(args['FP'])\n",
    "print(type(labels[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_name_full = args['model_name'] + '_' + args['FP'] + '_' + args['optimizer']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Eğitim esnasında test hata değeri en düşük değeri kaydeden bir fonksiyon\n",
    "model_checkpoint = ModelCheckpoint('models/' + model_name_full + '_weights_{epoch:03d}_{val_loss:.2f}.h5', monitor='val_loss', save_best_only=True, period=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "if args['resume']:\n",
    "    model = load_model(\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/250\n",
      "131/131 [==============================] - 337s 3s/step - loss: 0.3016 - mean_squared_error: 0.1410 - mean_squared_logarithmic_error: 0.0319 - mean_absolute_error: 0.2855 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.2069 - val_mean_squared_error: 0.0928 - val_mean_squared_logarithmic_error: 0.0192 - val_mean_absolute_error: 0.2266 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 2/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.1753 - mean_squared_error: 0.0905 - mean_squared_logarithmic_error: 0.0204 - mean_absolute_error: 0.2269 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.1418 - val_mean_squared_error: 0.0776 - val_mean_squared_logarithmic_error: 0.0159 - val_mean_absolute_error: 0.2068 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 3/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.1346 - mean_squared_error: 0.0814 - mean_squared_logarithmic_error: 0.0184 - mean_absolute_error: 0.2128 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.1232 - val_mean_squared_error: 0.0785 - val_mean_squared_logarithmic_error: 0.0170 - val_mean_absolute_error: 0.2038 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 4/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.1197 - mean_squared_error: 0.0794 - mean_squared_logarithmic_error: 0.0181 - mean_absolute_error: 0.2098 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.1110 - val_mean_squared_error: 0.0747 - val_mean_squared_logarithmic_error: 0.0146 - val_mean_absolute_error: 0.2038 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 5/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.1102 - mean_squared_error: 0.0767 - mean_squared_logarithmic_error: 0.0173 - mean_absolute_error: 0.2049 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.1018 - val_mean_squared_error: 0.0709 - val_mean_squared_logarithmic_error: 0.0142 - val_mean_absolute_error: 0.1951 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 6/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.1045 - mean_squared_error: 0.0750 - mean_squared_logarithmic_error: 0.0170 - mean_absolute_error: 0.2024 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.1011 - val_mean_squared_error: 0.0730 - val_mean_squared_logarithmic_error: 0.0151 - val_mean_absolute_error: 0.1957 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 7/250\n",
      "131/131 [==============================] - 310s 2s/step - loss: 0.1011 - mean_squared_error: 0.0742 - mean_squared_logarithmic_error: 0.0168 - mean_absolute_error: 0.2009 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0927 - val_mean_squared_error: 0.0669 - val_mean_squared_logarithmic_error: 0.0137 - val_mean_absolute_error: 0.1861 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 8/250\n",
      "131/131 [==============================] - 310s 2s/step - loss: 0.0973 - mean_squared_error: 0.0722 - mean_squared_logarithmic_error: 0.0164 - mean_absolute_error: 0.1973 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0932 - val_mean_squared_error: 0.0689 - val_mean_squared_logarithmic_error: 0.0144 - val_mean_absolute_error: 0.1881 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 9/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0950 - mean_squared_error: 0.0714 - mean_squared_logarithmic_error: 0.0163 - mean_absolute_error: 0.1962 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0910 - val_mean_squared_error: 0.0677 - val_mean_squared_logarithmic_error: 0.0140 - val_mean_absolute_error: 0.1878 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 10/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0931 - mean_squared_error: 0.0703 - mean_squared_logarithmic_error: 0.0161 - mean_absolute_error: 0.1944 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0897 - val_mean_squared_error: 0.0671 - val_mean_squared_logarithmic_error: 0.0135 - val_mean_absolute_error: 0.1867 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 11/250\n",
      "131/131 [==============================] - 315s 2s/step - loss: 0.0920 - mean_squared_error: 0.0698 - mean_squared_logarithmic_error: 0.0159 - mean_absolute_error: 0.1932 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0855 - val_mean_squared_error: 0.0637 - val_mean_squared_logarithmic_error: 0.0131 - val_mean_absolute_error: 0.1813 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 12/250\n",
      "131/131 [==============================] - 317s 2s/step - loss: 0.0908 - mean_squared_error: 0.0692 - mean_squared_logarithmic_error: 0.0158 - mean_absolute_error: 0.1921 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0872 - val_mean_squared_error: 0.0659 - val_mean_squared_logarithmic_error: 0.0133 - val_mean_absolute_error: 0.1851 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 13/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0900 - mean_squared_error: 0.0688 - mean_squared_logarithmic_error: 0.0158 - mean_absolute_error: 0.1918 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0859 - val_mean_squared_error: 0.0648 - val_mean_squared_logarithmic_error: 0.0134 - val_mean_absolute_error: 0.1804 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 14/250\n",
      "131/131 [==============================] - 319s 2s/step - loss: 0.0890 - mean_squared_error: 0.0684 - mean_squared_logarithmic_error: 0.0156 - mean_absolute_error: 0.1907 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0839 - val_mean_squared_error: 0.0635 - val_mean_squared_logarithmic_error: 0.0131 - val_mean_absolute_error: 0.1794 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 15/250\n",
      "131/131 [==============================] - 319s 2s/step - loss: 0.0869 - mean_squared_error: 0.0666 - mean_squared_logarithmic_error: 0.0152 - mean_absolute_error: 0.1874 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0840 - val_mean_squared_error: 0.0639 - val_mean_squared_logarithmic_error: 0.0132 - val_mean_absolute_error: 0.1855 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 16/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0868 - mean_squared_error: 0.0667 - mean_squared_logarithmic_error: 0.0154 - mean_absolute_error: 0.1879 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0834 - val_mean_squared_error: 0.0636 - val_mean_squared_logarithmic_error: 0.0127 - val_mean_absolute_error: 0.1800 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 17/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0868 - mean_squared_error: 0.0668 - mean_squared_logarithmic_error: 0.0152 - mean_absolute_error: 0.1881 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0829 - val_mean_squared_error: 0.0630 - val_mean_squared_logarithmic_error: 0.0130 - val_mean_absolute_error: 0.1803 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 18/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0861 - mean_squared_error: 0.0663 - mean_squared_logarithmic_error: 0.0152 - mean_absolute_error: 0.1874 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0827 - val_mean_squared_error: 0.0629 - val_mean_squared_logarithmic_error: 0.0128 - val_mean_absolute_error: 0.1832 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 19/250\n",
      "131/131 [==============================] - 316s 2s/step - loss: 0.0856 - mean_squared_error: 0.0660 - mean_squared_logarithmic_error: 0.0151 - mean_absolute_error: 0.1867 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0842 - val_mean_squared_error: 0.0647 - val_mean_squared_logarithmic_error: 0.0138 - val_mean_absolute_error: 0.1816 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 20/250\n",
      "131/131 [==============================] - 321s 2s/step - loss: 0.0855 - mean_squared_error: 0.0658 - mean_squared_logarithmic_error: 0.0151 - mean_absolute_error: 0.1869 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0808 - val_mean_squared_error: 0.0612 - val_mean_squared_logarithmic_error: 0.0126 - val_mean_absolute_error: 0.1768 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 21/250\n",
      "131/131 [==============================] - 316s 2s/step - loss: 0.0843 - mean_squared_error: 0.0649 - mean_squared_logarithmic_error: 0.0148 - mean_absolute_error: 0.1848 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0807 - val_mean_squared_error: 0.0612 - val_mean_squared_logarithmic_error: 0.0126 - val_mean_absolute_error: 0.1757 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 22/250\n",
      "131/131 [==============================] - 317s 2s/step - loss: 0.0834 - mean_squared_error: 0.0641 - mean_squared_logarithmic_error: 0.0146 - mean_absolute_error: 0.1836 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0788 - val_mean_squared_error: 0.0597 - val_mean_squared_logarithmic_error: 0.0122 - val_mean_absolute_error: 0.1750 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 23/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0830 - mean_squared_error: 0.0638 - mean_squared_logarithmic_error: 0.0147 - mean_absolute_error: 0.1828 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0805 - val_mean_squared_error: 0.0610 - val_mean_squared_logarithmic_error: 0.0125 - val_mean_absolute_error: 0.1778 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 24/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0837 - mean_squared_error: 0.0645 - mean_squared_logarithmic_error: 0.0148 - mean_absolute_error: 0.1840 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0807 - val_mean_squared_error: 0.0614 - val_mean_squared_logarithmic_error: 0.0128 - val_mean_absolute_error: 0.1788 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 25/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0828 - mean_squared_error: 0.0637 - mean_squared_logarithmic_error: 0.0146 - mean_absolute_error: 0.1826 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0786 - val_mean_squared_error: 0.0593 - val_mean_squared_logarithmic_error: 0.0120 - val_mean_absolute_error: 0.1736 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 26/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0825 - mean_squared_error: 0.0633 - mean_squared_logarithmic_error: 0.0145 - mean_absolute_error: 0.1822 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0796 - val_mean_squared_error: 0.0605 - val_mean_squared_logarithmic_error: 0.0122 - val_mean_absolute_error: 0.1770 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 27/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0820 - mean_squared_error: 0.0629 - mean_squared_logarithmic_error: 0.0144 - mean_absolute_error: 0.1810 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0773 - val_mean_squared_error: 0.0584 - val_mean_squared_logarithmic_error: 0.0121 - val_mean_absolute_error: 0.1718 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 28/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0816 - mean_squared_error: 0.0625 - mean_squared_logarithmic_error: 0.0143 - mean_absolute_error: 0.1805 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0771 - val_mean_squared_error: 0.0580 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1707 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 29/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0815 - mean_squared_error: 0.0625 - mean_squared_logarithmic_error: 0.0142 - mean_absolute_error: 0.1807 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0790 - val_mean_squared_error: 0.0596 - val_mean_squared_logarithmic_error: 0.0123 - val_mean_absolute_error: 0.1748 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 30/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0810 - mean_squared_error: 0.0619 - mean_squared_logarithmic_error: 0.0143 - mean_absolute_error: 0.1796 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0776 - val_mean_squared_error: 0.0588 - val_mean_squared_logarithmic_error: 0.0123 - val_mean_absolute_error: 0.1717 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 31/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0814 - mean_squared_error: 0.0624 - mean_squared_logarithmic_error: 0.0143 - mean_absolute_error: 0.1803 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0779 - val_mean_squared_error: 0.0589 - val_mean_squared_logarithmic_error: 0.0124 - val_mean_absolute_error: 0.1732 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 32/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0808 - mean_squared_error: 0.0618 - mean_squared_logarithmic_error: 0.0143 - mean_absolute_error: 0.1792 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0773 - val_mean_squared_error: 0.0583 - val_mean_squared_logarithmic_error: 0.0122 - val_mean_absolute_error: 0.1725 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 33/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0808 - mean_squared_error: 0.0620 - mean_squared_logarithmic_error: 0.0142 - mean_absolute_error: 0.1797 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0826 - val_mean_squared_error: 0.0638 - val_mean_squared_logarithmic_error: 0.0123 - val_mean_absolute_error: 0.1847 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 34/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0796 - mean_squared_error: 0.0609 - mean_squared_logarithmic_error: 0.0140 - mean_absolute_error: 0.1777 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0767 - val_mean_squared_error: 0.0579 - val_mean_squared_logarithmic_error: 0.0119 - val_mean_absolute_error: 0.1719 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 35/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0804 - mean_squared_error: 0.0618 - mean_squared_logarithmic_error: 0.0143 - mean_absolute_error: 0.1798 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0759 - val_mean_squared_error: 0.0572 - val_mean_squared_logarithmic_error: 0.0121 - val_mean_absolute_error: 0.1713 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 36/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0801 - mean_squared_error: 0.0615 - mean_squared_logarithmic_error: 0.0140 - mean_absolute_error: 0.1790 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0763 - val_mean_squared_error: 0.0575 - val_mean_squared_logarithmic_error: 0.0118 - val_mean_absolute_error: 0.1708 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 37/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 311s 2s/step - loss: 0.0792 - mean_squared_error: 0.0606 - mean_squared_logarithmic_error: 0.0139 - mean_absolute_error: 0.1771 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0755 - val_mean_squared_error: 0.0571 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1695 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 38/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0789 - mean_squared_error: 0.0605 - mean_squared_logarithmic_error: 0.0139 - mean_absolute_error: 0.1767 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0767 - val_mean_squared_error: 0.0581 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1727 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 39/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0786 - mean_squared_error: 0.0602 - mean_squared_logarithmic_error: 0.0140 - mean_absolute_error: 0.1765 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0779 - val_mean_squared_error: 0.0594 - val_mean_squared_logarithmic_error: 0.0119 - val_mean_absolute_error: 0.1715 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 40/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0785 - mean_squared_error: 0.0601 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1764 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0759 - val_mean_squared_error: 0.0575 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1709 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 41/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0782 - mean_squared_error: 0.0598 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1760 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0734 - val_mean_squared_error: 0.0551 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1627 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 42/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0779 - mean_squared_error: 0.0595 - mean_squared_logarithmic_error: 0.0138 - mean_absolute_error: 0.1758 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0733 - val_mean_squared_error: 0.0551 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1646 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 43/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0777 - mean_squared_error: 0.0594 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1752 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0757 - val_mean_squared_error: 0.0575 - val_mean_squared_logarithmic_error: 0.0119 - val_mean_absolute_error: 0.1680 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 44/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0784 - mean_squared_error: 0.0600 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1763 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0742 - val_mean_squared_error: 0.0558 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1666 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 45/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0776 - mean_squared_error: 0.0592 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1748 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0737 - val_mean_squared_error: 0.0554 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1639 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 46/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0781 - mean_squared_error: 0.0597 - mean_squared_logarithmic_error: 0.0138 - mean_absolute_error: 0.1761 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0726 - val_mean_squared_error: 0.0545 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1634 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 47/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0773 - mean_squared_error: 0.0590 - mean_squared_logarithmic_error: 0.0136 - mean_absolute_error: 0.1747 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0736 - val_mean_squared_error: 0.0556 - val_mean_squared_logarithmic_error: 0.0116 - val_mean_absolute_error: 0.1636 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 48/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0766 - mean_squared_error: 0.0585 - mean_squared_logarithmic_error: 0.0135 - mean_absolute_error: 0.1732 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0724 - val_mean_squared_error: 0.0545 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1635 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 49/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0773 - mean_squared_error: 0.0591 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1748 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0732 - val_mean_squared_error: 0.0550 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1645 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 50/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0771 - mean_squared_error: 0.0588 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1740 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0729 - val_mean_squared_error: 0.0547 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1645 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 51/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0770 - mean_squared_error: 0.0588 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1742 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0748 - val_mean_squared_error: 0.0565 - val_mean_squared_logarithmic_error: 0.0116 - val_mean_absolute_error: 0.1692 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 52/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0765 - mean_squared_error: 0.0583 - mean_squared_logarithmic_error: 0.0135 - mean_absolute_error: 0.1733 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0732 - val_mean_squared_error: 0.0550 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1677 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 53/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0765 - mean_squared_error: 0.0583 - mean_squared_logarithmic_error: 0.0136 - mean_absolute_error: 0.1735 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0741 - val_mean_squared_error: 0.0560 - val_mean_squared_logarithmic_error: 0.0118 - val_mean_absolute_error: 0.1663 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 54/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0766 - mean_squared_error: 0.0584 - mean_squared_logarithmic_error: 0.0137 - mean_absolute_error: 0.1735 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0728 - val_mean_squared_error: 0.0547 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1632 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 55/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0761 - mean_squared_error: 0.0580 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1726 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0730 - val_mean_squared_error: 0.0551 - val_mean_squared_logarithmic_error: 0.0116 - val_mean_absolute_error: 0.1675 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 56/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0762 - mean_squared_error: 0.0581 - mean_squared_logarithmic_error: 0.0134 - mean_absolute_error: 0.1730 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0736 - val_mean_squared_error: 0.0555 - val_mean_squared_logarithmic_error: 0.0116 - val_mean_absolute_error: 0.1651 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 57/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0763 - mean_squared_error: 0.0582 - mean_squared_logarithmic_error: 0.0135 - mean_absolute_error: 0.1731 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0736 - val_mean_squared_error: 0.0554 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1666 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 58/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0758 - mean_squared_error: 0.0576 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1719 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0722 - val_mean_squared_error: 0.0541 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1624 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 59/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0758 - mean_squared_error: 0.0576 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1721 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0716 - val_mean_squared_error: 0.0535 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1622 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 60/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0755 - mean_squared_error: 0.0573 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1717 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0731 - val_mean_squared_error: 0.0549 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1668 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 61/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0756 - mean_squared_error: 0.0575 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1720 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0714 - val_mean_squared_error: 0.0532 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1605 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 62/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0756 - mean_squared_error: 0.0574 - mean_squared_logarithmic_error: 0.0134 - mean_absolute_error: 0.1718 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0722 - val_mean_squared_error: 0.0540 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1631 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 63/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0755 - mean_squared_error: 0.0573 - mean_squared_logarithmic_error: 0.0134 - mean_absolute_error: 0.1717 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0758 - val_mean_squared_error: 0.0577 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1769 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 64/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0750 - mean_squared_error: 0.0569 - mean_squared_logarithmic_error: 0.0134 - mean_absolute_error: 0.1711 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0714 - val_mean_squared_error: 0.0533 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1619 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 65/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0756 - mean_squared_error: 0.0574 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1720 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0712 - val_mean_squared_error: 0.0531 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1612 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 66/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0750 - mean_squared_error: 0.0568 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1710 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0713 - val_mean_squared_error: 0.0532 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1621 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 67/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0763 - mean_squared_error: 0.0580 - mean_squared_logarithmic_error: 0.0135 - mean_absolute_error: 0.1730 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0741 - val_mean_squared_error: 0.0558 - val_mean_squared_logarithmic_error: 0.0117 - val_mean_absolute_error: 0.1687 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 68/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0757 - mean_squared_error: 0.0574 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1718 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0713 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1614 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 69/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0755 - mean_squared_error: 0.0572 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1720 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0727 - val_mean_squared_error: 0.0544 - val_mean_squared_logarithmic_error: 0.0112 - val_mean_absolute_error: 0.1629 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 70/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0754 - mean_squared_error: 0.0570 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1714 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0743 - val_mean_squared_error: 0.0560 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1718 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 71/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0753 - mean_squared_error: 0.0570 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1717 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0719 - val_mean_squared_error: 0.0537 - val_mean_squared_logarithmic_error: 0.0112 - val_mean_absolute_error: 0.1649 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 72/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0749 - mean_squared_error: 0.0566 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1702 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0709 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1604 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 73/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 313s 2s/step - loss: 0.0752 - mean_squared_error: 0.0569 - mean_squared_logarithmic_error: 0.0133 - mean_absolute_error: 0.1710 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0732 - val_mean_squared_error: 0.0549 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1690 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 74/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0757 - mean_squared_error: 0.0573 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1717 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0714 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1648 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 75/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0751 - mean_squared_error: 0.0568 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1709 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0717 - val_mean_squared_error: 0.0533 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1626 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 76/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0746 - mean_squared_error: 0.0562 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1699 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0708 - val_mean_squared_error: 0.0525 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1611 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 77/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0740 - mean_squared_error: 0.0558 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1689 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0723 - val_mean_squared_error: 0.0540 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1643 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 78/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0747 - mean_squared_error: 0.0564 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1703 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0707 - val_mean_squared_error: 0.0524 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1585 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 79/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0748 - mean_squared_error: 0.0565 - mean_squared_logarithmic_error: 0.0132 - mean_absolute_error: 0.1708 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0724 - val_mean_squared_error: 0.0540 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1672 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 80/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0743 - mean_squared_error: 0.0560 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1698 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0729 - val_mean_squared_error: 0.0546 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1663 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 81/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0743 - mean_squared_error: 0.0559 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1694 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0727 - val_mean_squared_error: 0.0544 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1681 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 82/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0741 - mean_squared_error: 0.0558 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1692 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0712 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1627 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 83/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0741 - mean_squared_error: 0.0558 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1692 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0721 - val_mean_squared_error: 0.0540 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1643 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 84/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0741 - mean_squared_error: 0.0557 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1692 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0712 - val_mean_squared_error: 0.0529 - val_mean_squared_logarithmic_error: 0.0114 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 85/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0742 - mean_squared_error: 0.0558 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1691 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0724 - val_mean_squared_error: 0.0540 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1649 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 86/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0739 - mean_squared_error: 0.0556 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1689 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0701 - val_mean_squared_error: 0.0518 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1589 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 87/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0741 - mean_squared_error: 0.0557 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1689 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0752 - val_mean_squared_error: 0.0569 - val_mean_squared_logarithmic_error: 0.0112 - val_mean_absolute_error: 0.1719 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 88/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0745 - mean_squared_error: 0.0561 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1702 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0735 - val_mean_squared_error: 0.0552 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1687 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 89/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0746 - mean_squared_error: 0.0561 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1698 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0698 - val_mean_squared_error: 0.0514 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1589 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 90/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0741 - mean_squared_error: 0.0557 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1691 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0715 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1622 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 91/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0742 - mean_squared_error: 0.0557 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1694 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0719 - val_mean_squared_error: 0.0536 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1642 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 92/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0744 - mean_squared_error: 0.0560 - mean_squared_logarithmic_error: 0.0131 - mean_absolute_error: 0.1700 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0714 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1634 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 93/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0736 - mean_squared_error: 0.0552 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1681 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0711 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1610 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 94/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0734 - mean_squared_error: 0.0550 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1674 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0695 - val_mean_squared_error: 0.0512 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1571 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 95/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0736 - mean_squared_error: 0.0553 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1682 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0701 - val_mean_squared_error: 0.0518 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1585 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 96/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0733 - mean_squared_error: 0.0550 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1675 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0708 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1621 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 97/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0744 - mean_squared_error: 0.0561 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1700 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0715 - val_mean_squared_error: 0.0531 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1636 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 98/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0741 - mean_squared_error: 0.0556 - mean_squared_logarithmic_error: 0.0130 - mean_absolute_error: 0.1693 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0710 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1622 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 99/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0741 - mean_squared_error: 0.0556 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1695 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0722 - val_mean_squared_error: 0.0537 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1677 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 100/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0740 - mean_squared_error: 0.0554 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1688 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0735 - val_mean_squared_error: 0.0548 - val_mean_squared_logarithmic_error: 0.0112 - val_mean_absolute_error: 0.1702 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 101/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0735 - mean_squared_error: 0.0549 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1679 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0714 - val_mean_squared_error: 0.0529 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1623 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 102/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0731 - mean_squared_error: 0.0546 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1670 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0736 - val_mean_squared_error: 0.0552 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1689 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 103/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0737 - mean_squared_error: 0.0552 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1684 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0731 - val_mean_squared_error: 0.0546 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1691 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 104/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0738 - mean_squared_error: 0.0553 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1686 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0732 - val_mean_squared_error: 0.0547 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1680 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 105/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0730 - mean_squared_error: 0.0545 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1667 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0510 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 106/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0740 - mean_squared_error: 0.0556 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1689 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0715 - val_mean_squared_error: 0.0531 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1656 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 107/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0735 - mean_squared_error: 0.0551 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1679 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1565 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 108/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0730 - mean_squared_error: 0.0546 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1672 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0716 - val_mean_squared_error: 0.0533 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1665 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 109/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 312s 2s/step - loss: 0.0731 - mean_squared_error: 0.0547 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1674 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0707 - val_mean_squared_error: 0.0523 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1619 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 110/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0727 - mean_squared_error: 0.0543 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1665 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0700 - val_mean_squared_error: 0.0515 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1607 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 111/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0732 - mean_squared_error: 0.0547 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1677 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0717 - val_mean_squared_error: 0.0533 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1624 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 112/250\n",
      "131/131 [==============================] - 315s 2s/step - loss: 0.0730 - mean_squared_error: 0.0545 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1671 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1567 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 113/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0725 - mean_squared_error: 0.0541 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1664 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0700 - val_mean_squared_error: 0.0516 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 114/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0726 - mean_squared_error: 0.0542 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1663 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0703 - val_mean_squared_error: 0.0520 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1623 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 115/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0737 - mean_squared_error: 0.0553 - mean_squared_logarithmic_error: 0.0129 - mean_absolute_error: 0.1688 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0705 - val_mean_squared_error: 0.0521 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1628 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 116/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0729 - mean_squared_error: 0.0544 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1666 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0738 - val_mean_squared_error: 0.0553 - val_mean_squared_logarithmic_error: 0.0112 - val_mean_absolute_error: 0.1720 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 117/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0728 - mean_squared_error: 0.0543 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1669 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0715 - val_mean_squared_error: 0.0530 - val_mean_squared_logarithmic_error: 0.0115 - val_mean_absolute_error: 0.1593 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 118/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0728 - mean_squared_error: 0.0543 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1664 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0709 - val_mean_squared_error: 0.0524 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1639 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 119/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0727 - mean_squared_error: 0.0542 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1665 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0711 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1625 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 120/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0729 - mean_squared_error: 0.0544 - mean_squared_logarithmic_error: 0.0128 - mean_absolute_error: 0.1672 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0719 - val_mean_squared_error: 0.0534 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1642 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 121/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0723 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1657 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0692 - val_mean_squared_error: 0.0509 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1566 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 122/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0728 - mean_squared_error: 0.0544 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1670 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0692 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1577 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 123/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0728 - mean_squared_error: 0.0542 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1668 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0697 - val_mean_squared_error: 0.0511 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1593 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 124/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0722 - mean_squared_error: 0.0537 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1656 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0712 - val_mean_squared_error: 0.0527 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1634 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 125/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0731 - mean_squared_error: 0.0545 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1675 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0698 - val_mean_squared_error: 0.0513 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1597 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 126/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0725 - mean_squared_error: 0.0540 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1657 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0701 - val_mean_squared_error: 0.0517 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1600 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 127/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0726 - mean_squared_error: 0.0541 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1662 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0687 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1566 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 128/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0720 - mean_squared_error: 0.0536 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1652 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0697 - val_mean_squared_error: 0.0513 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 129/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0719 - mean_squared_error: 0.0533 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1648 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1561 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 130/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0722 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1656 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0691 - val_mean_squared_error: 0.0506 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1571 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 131/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0726 - mean_squared_error: 0.0541 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1667 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0509 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1589 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 132/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0728 - mean_squared_error: 0.0544 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1671 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0698 - val_mean_squared_error: 0.0514 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1598 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 133/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0727 - mean_squared_error: 0.0543 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1666 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0697 - val_mean_squared_error: 0.0513 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1586 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 134/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0719 - mean_squared_error: 0.0535 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1652 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0696 - val_mean_squared_error: 0.0512 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1593 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 135/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0723 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1662 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1573 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 136/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0724 - mean_squared_error: 0.0539 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1661 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0691 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1552 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 137/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0723 - mean_squared_error: 0.0539 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1660 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0506 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1562 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 138/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0715 - mean_squared_error: 0.0531 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1643 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0506 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1590 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 139/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0726 - mean_squared_error: 0.0542 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1668 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0704 - val_mean_squared_error: 0.0519 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1600 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 140/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0725 - mean_squared_error: 0.0540 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1660 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0505 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1572 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 141/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0719 - mean_squared_error: 0.0535 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1654 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0702 - val_mean_squared_error: 0.0517 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1604 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 142/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0725 - mean_squared_error: 0.0540 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1665 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1589 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 143/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0716 - mean_squared_error: 0.0531 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1646 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0688 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1567 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 144/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0720 - mean_squared_error: 0.0535 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1656 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0505 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1558 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 145/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 312s 2s/step - loss: 0.0718 - mean_squared_error: 0.0534 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1652 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0720 - val_mean_squared_error: 0.0535 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1659 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 146/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0720 - mean_squared_error: 0.0534 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1655 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0505 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1574 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 147/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0723 - mean_squared_error: 0.0536 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1659 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0713 - val_mean_squared_error: 0.0526 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1638 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 148/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0718 - mean_squared_error: 0.0532 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1648 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 149/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0726 - mean_squared_error: 0.0540 - mean_squared_logarithmic_error: 0.0127 - mean_absolute_error: 0.1666 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0695 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1581 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 150/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0725 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1659 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0719 - val_mean_squared_error: 0.0531 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1675 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 151/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0722 - mean_squared_error: 0.0535 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1654 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0695 - val_mean_squared_error: 0.0510 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1591 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 152/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0719 - mean_squared_error: 0.0532 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1651 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0697 - val_mean_squared_error: 0.0510 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1602 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 153/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0719 - mean_squared_error: 0.0532 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1650 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1552 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 154/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0724 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1659 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0696 - val_mean_squared_error: 0.0510 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1571 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 155/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0726 - mean_squared_error: 0.0538 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1660 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0506 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1572 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 156/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0709 - mean_squared_error: 0.0523 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1629 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0684 - val_mean_squared_error: 0.0498 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1546 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 157/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0716 - mean_squared_error: 0.0530 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1647 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0683 - val_mean_squared_error: 0.0498 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1555 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 158/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0711 - mean_squared_error: 0.0526 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1636 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0685 - val_mean_squared_error: 0.0499 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1567 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 159/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0723 - mean_squared_error: 0.0537 - mean_squared_logarithmic_error: 0.0126 - mean_absolute_error: 0.1661 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1595 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 160/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0716 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1640 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0688 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1569 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 161/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0528 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1637 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1561 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 162/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0714 - mean_squared_error: 0.0528 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1641 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0683 - val_mean_squared_error: 0.0497 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1558 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 163/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1637 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1554 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 164/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0715 - mean_squared_error: 0.0530 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1641 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0705 - val_mean_squared_error: 0.0519 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1615 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 165/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1640 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0505 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1557 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 166/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1640 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0679 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1553 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 167/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0526 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1639 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0682 - val_mean_squared_error: 0.0497 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1538 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 168/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0712 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1638 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1572 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 169/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0717 - mean_squared_error: 0.0531 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1648 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0685 - val_mean_squared_error: 0.0498 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1556 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 170/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0715 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1643 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0108 - val_mean_absolute_error: 0.1564 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 171/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0717 - mean_squared_error: 0.0531 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1650 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0683 - val_mean_squared_error: 0.0496 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1540 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 172/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0716 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1643 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0686 - val_mean_squared_error: 0.0499 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1548 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 173/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0719 - mean_squared_error: 0.0532 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1647 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0684 - val_mean_squared_error: 0.0497 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1557 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 174/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0526 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1642 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0704 - val_mean_squared_error: 0.0517 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1613 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 175/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0525 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1636 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0700 - val_mean_squared_error: 0.0512 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1608 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 176/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0710 - mean_squared_error: 0.0523 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1636 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0697 - val_mean_squared_error: 0.0510 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1602 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 177/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0710 - mean_squared_error: 0.0523 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1630 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0681 - val_mean_squared_error: 0.0495 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1549 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 178/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0715 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1645 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0681 - val_mean_squared_error: 0.0494 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1557 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 179/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0707 - mean_squared_error: 0.0521 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1627 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1567 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 180/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0714 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0124 - mean_absolute_error: 0.1643 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0686 - val_mean_squared_error: 0.0498 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1568 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 181/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0525 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1635 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0686 - val_mean_squared_error: 0.0500 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1579 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 182/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0717 - mean_squared_error: 0.0530 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1649 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0674 - val_mean_squared_error: 0.0487 - val_mean_squared_logarithmic_error: 0.0100 - val_mean_absolute_error: 0.1543 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 183/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0716 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1645 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0681 - val_mean_squared_error: 0.0494 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1536 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 184/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0718 - mean_squared_error: 0.0530 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1646 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1586 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 185/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0708 - mean_squared_error: 0.0521 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1629 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0692 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1571 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 186/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0717 - mean_squared_error: 0.0529 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1646 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0684 - val_mean_squared_error: 0.0496 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1570 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 187/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0713 - mean_squared_error: 0.0525 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1638 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0693 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1577 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 188/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0526 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1636 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0505 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1584 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 189/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0705 - mean_squared_error: 0.0518 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0711 - val_mean_squared_error: 0.0525 - val_mean_squared_logarithmic_error: 0.0113 - val_mean_absolute_error: 0.1668 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 190/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0717 - mean_squared_error: 0.0530 - mean_squared_logarithmic_error: 0.0125 - mean_absolute_error: 0.1651 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0700 - val_mean_squared_error: 0.0511 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1600 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 191/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0714 - mean_squared_error: 0.0526 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1641 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0674 - val_mean_squared_error: 0.0487 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1545 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 192/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0708 - mean_squared_error: 0.0521 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1629 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0670 - val_mean_squared_error: 0.0484 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1525 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 193/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0706 - mean_squared_error: 0.0520 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1625 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0677 - val_mean_squared_error: 0.0490 - val_mean_squared_logarithmic_error: 0.0100 - val_mean_absolute_error: 0.1534 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 194/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0701 - mean_squared_error: 0.0514 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1615 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0111 - val_mean_absolute_error: 0.1566 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 195/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0710 - mean_squared_error: 0.0523 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1635 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0676 - val_mean_squared_error: 0.0489 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1514 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 196/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0703 - mean_squared_error: 0.0517 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1622 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1566 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 197/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0706 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0680 - val_mean_squared_error: 0.0494 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1562 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 198/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0708 - mean_squared_error: 0.0521 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1630 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0689 - val_mean_squared_error: 0.0502 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1552 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 199/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0704 - mean_squared_error: 0.0517 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1624 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0680 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1564 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 200/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0705 - mean_squared_error: 0.0518 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1622 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0678 - val_mean_squared_error: 0.0492 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1532 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 201/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0703 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1621 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0683 - val_mean_squared_error: 0.0496 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1562 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 202/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0525 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1638 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0677 - val_mean_squared_error: 0.0489 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1546 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 203/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0711 - mean_squared_error: 0.0524 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1635 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0673 - val_mean_squared_error: 0.0486 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1530 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 204/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0712 - mean_squared_error: 0.0524 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1635 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0681 - val_mean_squared_error: 0.0494 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1563 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 205/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0706 - mean_squared_error: 0.0518 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1624 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0691 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1577 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 206/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0708 - mean_squared_error: 0.0520 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1624 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0678 - val_mean_squared_error: 0.0490 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1538 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 207/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0706 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1628 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0684 - val_mean_squared_error: 0.0497 - val_mean_squared_logarithmic_error: 0.0107 - val_mean_absolute_error: 0.1548 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 208/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0701 - mean_squared_error: 0.0514 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1614 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1612 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 209/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0709 - mean_squared_error: 0.0522 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1635 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1584 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 210/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0707 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1629 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0666 - val_mean_squared_error: 0.0479 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1516 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 211/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0702 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1622 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0696 - val_mean_squared_error: 0.0508 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1586 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 212/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0711 - mean_squared_error: 0.0523 - mean_squared_logarithmic_error: 0.0123 - mean_absolute_error: 0.1639 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0666 - val_mean_squared_error: 0.0478 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1523 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 213/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0703 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0667 - val_mean_squared_error: 0.0481 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1514 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 214/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0707 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1627 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0670 - val_mean_squared_error: 0.0482 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1521 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 215/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0715 - mean_squared_error: 0.0527 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1643 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0682 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1548 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 216/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0706 - mean_squared_error: 0.0518 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1627 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0671 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1523 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 217/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "131/131 [==============================] - 312s 2s/step - loss: 0.0705 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0676 - val_mean_squared_error: 0.0487 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1535 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 218/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0705 - mean_squared_error: 0.0517 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0676 - val_mean_squared_error: 0.0489 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1541 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 219/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0707 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1626 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0692 - val_mean_squared_error: 0.0504 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1603 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 220/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0671 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1525 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 221/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0675 - val_mean_squared_error: 0.0488 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1521 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 222/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0705 - mean_squared_error: 0.0517 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1626 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0687 - val_mean_squared_error: 0.0499 - val_mean_squared_logarithmic_error: 0.0105 - val_mean_absolute_error: 0.1541 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 223/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0700 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1611 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0694 - val_mean_squared_error: 0.0506 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1590 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 224/250\n",
      "131/131 [==============================] - 311s 2s/step - loss: 0.0700 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1614 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0678 - val_mean_squared_error: 0.0490 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1543 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 225/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0704 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1625 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0663 - val_mean_squared_error: 0.0476 - val_mean_squared_logarithmic_error: 0.0100 - val_mean_absolute_error: 0.1510 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 226/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1619 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0677 - val_mean_squared_error: 0.0489 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1524 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 227/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0708 - mean_squared_error: 0.0520 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1631 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0670 - val_mean_squared_error: 0.0482 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1518 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 228/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0703 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0661 - val_mean_squared_error: 0.0473 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1504 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 229/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0707 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1627 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0672 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1517 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 230/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0699 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0118 - mean_absolute_error: 0.1614 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0690 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0110 - val_mean_absolute_error: 0.1538 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 231/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0700 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1616 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0687 - val_mean_squared_error: 0.0499 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1560 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 232/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0697 - mean_squared_error: 0.0509 - mean_squared_logarithmic_error: 0.0118 - mean_absolute_error: 0.1610 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0669 - val_mean_squared_error: 0.0481 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1512 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 233/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0679 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1551 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 234/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0705 - mean_squared_error: 0.0517 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1625 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0680 - val_mean_squared_error: 0.0492 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1540 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 235/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0695 - mean_squared_error: 0.0509 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1608 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0678 - val_mean_squared_error: 0.0492 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1548 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 236/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0702 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1618 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0676 - val_mean_squared_error: 0.0489 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1537 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 237/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0702 - mean_squared_error: 0.0514 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0682 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1565 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 238/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0704 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1621 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0680 - val_mean_squared_error: 0.0493 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1544 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 239/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0703 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1623 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0668 - val_mean_squared_error: 0.0481 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1520 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 240/250\n",
      "131/131 [==============================] - 314s 2s/step - loss: 0.0701 - mean_squared_error: 0.0513 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1618 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0670 - val_mean_squared_error: 0.0482 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1529 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 241/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0121 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0672 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1520 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 242/250\n",
      "131/131 [==============================] - 313s 2s/step - loss: 0.0698 - mean_squared_error: 0.0510 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1609 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0668 - val_mean_squared_error: 0.0480 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1498 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 243/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0699 - mean_squared_error: 0.0511 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1612 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0691 - val_mean_squared_error: 0.0503 - val_mean_squared_logarithmic_error: 0.0109 - val_mean_absolute_error: 0.1562 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 244/250\n",
      "131/131 [==============================] - 315s 2s/step - loss: 0.0707 - mean_squared_error: 0.0519 - mean_squared_logarithmic_error: 0.0122 - mean_absolute_error: 0.1632 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0696 - val_mean_squared_error: 0.0507 - val_mean_squared_logarithmic_error: 0.0106 - val_mean_absolute_error: 0.1609 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 245/250\n",
      "131/131 [==============================] - 315s 2s/step - loss: 0.0703 - mean_squared_error: 0.0515 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1620 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0668 - val_mean_squared_error: 0.0480 - val_mean_squared_logarithmic_error: 0.0100 - val_mean_absolute_error: 0.1508 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 246/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0704 - mean_squared_error: 0.0516 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1626 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0671 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0102 - val_mean_absolute_error: 0.1517 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 247/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0698 - mean_squared_error: 0.0509 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1611 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0671 - val_mean_squared_error: 0.0483 - val_mean_squared_logarithmic_error: 0.0101 - val_mean_absolute_error: 0.1509 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 248/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0700 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1615 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0673 - val_mean_squared_error: 0.0485 - val_mean_squared_logarithmic_error: 0.0099 - val_mean_absolute_error: 0.1540 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 249/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0697 - mean_squared_error: 0.0510 - mean_squared_logarithmic_error: 0.0119 - mean_absolute_error: 0.1610 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0674 - val_mean_squared_error: 0.0487 - val_mean_squared_logarithmic_error: 0.0103 - val_mean_absolute_error: 0.1536 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n",
      "Epoch 250/250\n",
      "131/131 [==============================] - 312s 2s/step - loss: 0.0700 - mean_squared_error: 0.0512 - mean_squared_logarithmic_error: 0.0120 - mean_absolute_error: 0.1612 - mean_absolute_percentage_error: inf - cosine_proximity: nan - val_loss: 0.0665 - val_mean_squared_error: 0.0478 - val_mean_squared_logarithmic_error: 0.0104 - val_mean_absolute_error: 0.1501 - val_mean_absolute_percentage_error: inf - val_cosine_proximity: nan\n"
     ]
    }
   ],
   "source": [
    "# Eğitim fonksiyonu \n",
    "hs = model.fit_generator(generate_data(),steps_per_epoch=int(splitpoint/ bsize),\n",
    "                    validation_data=generate_data_val(), \n",
    "                    validation_steps=(dlen-splitpoint)/bsize, epochs=args['epochs_number'], callbacks=[model_checkpoint])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['loss', 'mean_squared_logarithmic_error', 'val_mean_absolute_percentage_error', 'val_cosine_proximity', 'val_mean_squared_logarithmic_error', 'mean_absolute_percentage_error', 'val_mean_squared_error', 'val_mean_absolute_error', 'cosine_proximity', 'mean_squared_error', 'val_loss', 'mean_absolute_error']\n"
     ]
    }
   ],
   "source": [
    "# Eğitim işleminin gidişatını grafik olarak görüyoruz\n",
    "# Train and validation loss charts\n",
    "print(hs.history.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEWCAYAAABi5jCmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzt3Xu8VXWd//HXO0AIRDhcvHATLHsIhztHcH4MhoGIlZe80mSBkzGZ1VgjE1YTRvrLilGn8pJKjTqWIoUymYOXpHQ04yIaKMpFjIsiIKKIouhn/thfTnsd97lw9j5sjryfj8d+sNb6ftd3f79rw37vtb6btRURmJmZ7faBcnfAzMz2LQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDAeDlYWk3pJCUssSt7tG0thStmm2v3Ew2H5rfwoRSaMlrSt3P6x5cDCY7SOU43+TiaQWDdlWTxslPSPdX/gvYTOXPvVOkfSkpNclzZR0iKR7JL0m6X5JFanuMZIekfSKpCckjc5r51xJT6d9Vkv6p7yy0ZLWSfoXSS9JekHSuQ3o2yckPS7pVUlrJV1SoNo/StqQ2rwob9/hkhamfTdKuiKv7GRJy9I45kvqW8vz/6ekS2uOIy3fAvQC/lvSdkn/Wt8xqmOc8yV9X9KfU3/vktQpr7yu4z5f0mWS/hfYARwhqZOkX6TjslXSnXn1PylpSWrrEUkD88rWSLoo/V3YJul2SW0ktQPuAbqlsW6X1C0d40dTWy9I+qmkA/LaGyfpmdTWNZL+IOm8vPJ/TH9ntkqaJ+nwBhyroyTdJ+nl1PZZNV6vayX9TtLrwHG1bOsg6WZJmyQ9L+nbSoEqaZKk/5V0paQtwCX19ckKiAg/mvEDWAP8CTgE6A68BCwGhgBtgN8D01LZFuDj5D4QHJ/Wu6Z2PgF8CBDwUXJvUkNT2WhgFzAdaJXa2AFU1NO30cCA9HwDgY3AqamsNxDAr4B2qd4mYGwqfxT4bFo+EDgmLX8EeD31vxXwr8BK4IC847G7jf8ELq3Rn3U1jt3YvPU6j1Ed45wPrAf6p7H8GvivhrSZ9v0rUAm0TGO6G7gdqEjrH011h6TXdwTQApiYxtA6bzx/BroBnYCngS8WGnvaNgw4Jj1v71T/wlTWBXgVOC2V/zPwNnBeKj8lHfe+qfzbwCP1HKd2wFrg3LTPEGAz0C/v9doGjEzHqk0t224G7gLap34/C3w+tTGJ3N/Vr6Tn+GC5/402x0fZO+BHkS9g7s3gM3nrvwauzVv/CnAn8A3glhr7zgMm1tLuncA/p+XRwBtAy7zyl0hv1nvQ16uAK9Nyb3LBcFRe+Q+BmWn5j8B3gS412vg3YFbe+gfIvSmPzjsejQ2GPTpGeXXmA5fnrfcD3iL35l1nm2nf6XllhwHvUiB0gWuB79XY9gx/C441wDk1jud1hcZeyzguBOak5c8Bj+aVidyb+u5guGf3m3He67ADOLyO9s8GHqqx7WfAtLzX6+Ya5Zlt6Zi+RQqTtO2fgPlpeRLw11L+G9sfH76U9P6wMW/5jQLrBwKHA2emywavSHoF+Htyb0RIOlHSn9Ip/ivkPuF2yWtnS0TsylvfkdqtlaQRkh5Mp/zbgC/WaBNybza7PU/u0y7A58mdHSyXtEDSJ9P2bqkeABHxbmqje119aaA6j1E9ao6jFbmxNqTN/H17Ai9HxNZa+vcvNdrqyd+OGcCLect1vkaSPiLpt5JelPQq8P/52+vTLb9fkXvXzZ+8Phz4j7x+vEwuPOp6HQ4HRtTo/2eAQ/PqrC2wX/62LuSO7fN5256v8byF2rA94ImZ/cdacp9cv1CzQFJrcmcanwPuioi303VtFfmcvwR+CpwYEW9Kuor3BkNPYHla7gVsAIiIFcCn07Xj04DZkjqn8gF5fVdqY32B538daJu3fmiN8pq3Fq71GDVAz7zlXuQuu2xuYJv5/VgLdJLUMSJeKdC/yyLiskb0r9BtlK8FHgc+HRGvSboQOCOVvQD02F0xHeceefvu7sute9CHtcAfIuL4Pexn/rbN5I7t4cBTaVsvsq+/bxldJJ8x7D/+CzhJ0gmSWqRJydGSegAHAK3JXePfJelEYFwJnrM9uU+/b0oaDvxDgTr/JqmtpEpy155vB5B0jqSu6Yxg9xvku8As4BOSxkhqBfwLsBN4pEDbS4CPp8ncQ8ldKsm3ETgib72uY1SfcyT1k9SW3FzM7Ih4Z0/bjIgXyF2muUZShaRWko5NxTcAX0xnYpLUTrkJ/vYN6N9GoLOkDnnb2pObR9gu6Sjg/Lyyu4EBkk5V7ps9F5AN1uuAi9PrRpoQPrOePvwW+Iikz6ZxtZJ0tGr58kAh6ZjOAi6T1D5NeH+d3HG2EnEw7CciYi25CcNvkguAtcAU4AMR8RrwVXL/4LaSewOfW4Kn/RIwXdJrwHdS+zX9gdwk5gPAjIi4N20fDyyTtB34D2BCRLwREc8A5wA/Iffp8STgpIh4q0DbtwBPkLv2fi8pdPJ8H/h2uqxxUV3HqAFjvYXc9fAXyU2QfhXqPu51tPVZcp+Kl5Oby7kwtbUQ+AK5s7Ct5I7bpAb0jYhYTm6if3UabzfgInKv9WvkQuf2vPqbgTPJzVNsITdvspBcCBMRc4AfALely1BLgRPr6cNr5D5wTCB35vdiaqN1Q8aQ5yvkzgZXAw+TOzP9+R62YXVQmrAxs0aSNJ/ct5BuLHdfmkq6pLeO3BcdHix3f6xp+YzBzApKl786pjmob5Kbc/pTmbtle4Enn60okpaRmwis6Z/2cGJyn5YuaRVS5+WTZu7vyF2mOYDcRO+pEfFGXTtIGkVujuQ9IqLOb7HZvsOXkszMLMOXkszMLKNZXkrq0qVL9O7du9zdMDNrVhYtWrQ5IrrWV69ZBkPv3r1ZuHBhubthZtasSHq+/lq+lGRmZjU4GMzMLMPBYGZmGc1yjsHMyu/tt99m3bp1vPnmm+XuitXQpk0bevToQatWrRq1v4PBzBpl3bp1tG/fnt69e5O7+artCyKCLVu2sG7dOvr06dOoNnwpycwa5c0336Rz584OhX2MJDp37lzUmZyDwcwazaGwbyr2dXEwmJlZhoPBzJqlV155hWuuuaZR+3784x/nlVdq/kBe1ne+8x3uv//+RrVfjDvvvJOnnnqq/opNyMFgZs1SXcGwa9eugtt3+93vfkfHjh3rrDN9+nTGjh3b6P41loPBzKyRpk6dyqpVqxg8eDBTpkxh/vz5jBo1ipNPPpl+/foBcOqppzJs2DAqKyu5/vrrq/ft3bs3mzdvZs2aNfTt25cvfOELVFZWMm7cON54I3dn8UmTJjF79uzq+tOmTWPo0KEMGDCA5ctzP1O+adMmjj/+eCorKznvvPM4/PDD2bx5c6af77zzDpMmTaJ///4MGDCAK6+8EoBVq1Yxfvx4hg0bxqhRo1i+fDmPPPIIc+fOZcqUKQwePJhVq1Y1+XEsxF9XNbOiffe/l/HUhldL2ma/bgcx7aTKWssvv/xyli5dypIlSwCYP38+ixcvZunSpdVf0/z5z39Op06deOONNzj66KM5/fTT6dy5c6adFStW8Ktf/YobbriBs846i1//+tecc84573m+Ll26sHjxYq655hpmzJjBjTfeyHe/+10+9rGPcfHFF/M///M/zJw58z37LVmyhPXr17N06VKA6ktYkydP5rrrruPII4/kscce40tf+hK///3vOfnkk/nkJz/JGWec0bgDVwIOBjN73xg+fHjmu/s//vGPmTNnDgBr165lxYoV7wmGPn36MHjwYACGDRvGmjVrCrZ92mmnVdf5zW9+A8DDDz9c3f748eOpqKh4z35HHHEEq1ev5itf+Qqf+MQnGDduHNu3b+eRRx7hzDPPrK63c+fORo669BwMZla0uj7Z703t2rWrXp4/fz73338/jz76KG3btmX06NEFv9vfunXr6uUWLVpUX0qqrV6LFi3qncPIV1FRwRNPPMG8efO47rrrmDVrFldddRUdO3asPtvZ13iOwcyapfbt2/Paa6/VWr5t2zYqKipo27Yty5cv509/Kv3PVY8cOZJZs2YBcO+997J169b31Nm8eTPvvvsup59+OpdeeimLFy/moIMOok+fPtxxxx1A7n8rP/HEEw0a197gYDCzZqlz586MHDmS/v37M2XKlPeUjx8/nl27dtG3b1+mTp3KMcccU/I+TJs2jXvvvZf+/ftzxx13cOihh9K+fftMnfXr1zN69GgGDx7MOeecw/e//30Abr31VmbOnMmgQYOorKzkrrvuAmDChAn86Ec/YsiQIWWbfG6Wv/lcVVUV/qEes/J6+umn6du3b7m7UVY7d+6kRYsWtGzZkkcffZTzzz9/n7k8VOj1kbQoIqrq29dzDGZmjfTXv/6Vs846i3fffZcDDjiAG264odxdKgkHg5lZIx155JE8/vjj5e5GyXmOwczMMhwMZmaW4WAwM7MMB4OZmWU4GMxsv3HggQcCsGHDhlrvRTR69Gjq+zr8VVddxY4dO6rXG3Ib71Jbs2YNv/zlL5ukbQeDme13unXrVn3n1MaoGQwNuY13qe3zwSBpvKRnJK2UNLVAeWtJt6fyxyT1rlHeS9J2SReVoj9m9v43depUrr766ur1Sy65hBkzZrB9+3bGjBlTfYvs3f+jON+aNWvo378/AG+88QYTJkygb9++fOpTn8rcK+n888+nqqqKyspKpk2bBuRuzLdhwwaOO+44jjvuOOBvt/EGuOKKK+jfvz/9+/fnqquuqn6+2m7vne+OO+6gf//+DBo0iGOPPRbI3bZ7ypQpHH300QwcOJCf/exn1eN/6KGHGDx4cPWtvEul6P/HIKkFcDVwPLAOWCBpbkTk/9LE54GtEfFhSROAHwBn55VfAdxTbF/MrEzumQov/qW0bR46AE68vNbis88+mwsvvJALLrgAgFmzZjFv3jzatGnDnDlzOOigg9i8eTPHHHMMJ598cq2/g3zttdfStm1bnn76aZ588kmGDh1aXXbZZZfRqVMn3nnnHcaMGcOTTz7JV7/6Va644goefPBBunTpkmlr0aJF/OIXv+Cxxx4jIhgxYgQf/ehHqaioaNDtvadPn868efPo3r179aWpmTNn0qFDBxYsWMDOnTsZOXIk48aN4/LLL2fGjBn89re/bdThrUspzhiGAysjYnVEvAXcBpxSo84pwE1peTYwRulVknQq8BywrAR9MbP9xJAhQ3jppZfYsGEDTzzxBBUVFfTs2ZOI4Jvf/CYDBw5k7NixrF+/no0bN9bazh//+MfqN+iBAwcycODA6rJZs2YxdOhQhgwZwrJly+r9ZbWHH36YT33qU7Rr144DDzyQ0047jYceegho2O29R44cyaRJk7jhhht45513gNzN+W6++WYGDx7MiBEj2LJlCytWrNijY7WnSvE/n7sDa/PW1wEjaqsTEbskbQM6S3oT+Aa5s406LyNJmgxMBujVq1cJum1mJVPHJ/umdOaZZzJ79mxefPFFzj47dxHi1ltvZdOmTSxatIhWrVrRu3fvgrfbrs9zzz3HjBkzWLBgARUVFUyaNKlR7ezWkNt7X3fddTz22GPcfffdDBs2jEWLFhER/OQnP+GEE07I1J0/f36j+1Kfck8+XwJcGRHb66sYEddHRFVEVHXt2rXpe2Zm+7yzzz6b2267jdmzZ1f/6M22bds4+OCDadWqFQ8++CDPP/98nW0ce+yx1ZO4S5cu5cknnwTg1VdfpV27dnTo0IGNGzdyzz1/u9pd262xR40axZ133smOHTt4/fXXmTNnDqNGjWrweFatWsWIESOYPn06Xbt2Ze3atZxwwglce+21vP322wA8++yzvP766016e+5SnDGsB3rmrfdI2wrVWSepJdAB2ELuzOIMST8EOgLvSnozIn5agn6Z2ftcZWUlr732Gt27d+ewww4D4DOf+QwnnXQSAwYMoKqqiqOOOqrONs4//3zOPfdc+vbtS9++fRk2bBgAgwYNYsiQIRx11FH07NmTkSNHVu8zefJkxo8fT7du3XjwwQertw8dOpRJkyYxfPhwAM477zyGDBlS66/C1TRlyhRWrFhBRDBmzBgGDRrEwIEDWbNmDUOHDiUi6Nq1K3feeScDBw6kRYsWDBo0iEmTJvG1r31tTw5dnYq+7XZ6o38WGEMuABYA/xARy/LqXAAMiIgvpsnn0yLirBrtXAJsj4gZ9T2nb7ttVn6+7fa+ray33U5zBl8G5gEtgJ9HxDJJ04GFETEXmAncImkl8DIwodjnNTOzplGS225HxO+A39XY9p285TeBM2vuV6P+JaXoi5mZFafck89m1ow1x1+A3B8U+7o4GMysUdq0acOWLVscDvuYiGDLli20adOm0W34F9zMrFF69OjBunXr2LRpU7m7YjW0adOGHj16NHp/B4OZNUqrVq3o06dPubthTcCXkszMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDAeDmZllOBjMzCzDwWBmZhkOBjMzy3AwmJlZhoPBzMwyHAxmZpbhYDAzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJiZWYaDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLMPBYGZmGSUJBknjJT0jaaWkqQXKW0u6PZU/Jql32n68pEWS/pL+/Fgp+mNmZo1XdDBIagFcDZwI9AM+LalfjWqfB7ZGxIeBK4EfpO2bgZMiYgAwEbil2P6YmVlxSnHGMBxYGRGrI+It4DbglBp1TgFuSsuzgTGSFBGPR8SGtH0Z8EFJrUvQJzMza6RSBEN3YG3e+rq0rWCdiNgFbAM616hzOrA4InaWoE9mZtZILcvdAQBJleQuL42ro85kYDJAr1699lLPzMz2P6U4Y1gP9Mxb75G2FawjqSXQAdiS1nsAc4DPRcSq2p4kIq6PiKqIqOratWsJum1mZoWUIhgWAEdK6iPpAGACMLdGnbnkJpcBzgB+HxEhqSNwNzA1Iv63BH0xM7MiFR0Mac7gy8A84GlgVkQskzRd0smp2kygs6SVwNeB3V9p/TLwYeA7kpakx8HF9snMzBpPEVHuPuyxqqqqWLhwYbm7YWbWrEhaFBFV9dXz/3w2M7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDAeDmZllOBjMzCzDwWBmZhkOBjMzy3AwmJlZhoPBzMwyHAxmZpbhYDAzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJiZWYaDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDJKEgySxkt6RtJKSVMLlLeWdHsqf0xS77yyi9P2ZySdUIr+mJlZ4xUdDJJaAFcDJwL9gE9L6lej2ueBrRHxYeBK4Adp337ABKASGA9ck9ozM7MyKcUZw3BgZUSsjoi3gNuAU2rUOQW4KS3PBsZIUtp+W0TsjIjngJWpPTMzK5NSBEN3YG3e+rq0rWCdiNgFbAM6N3BfACRNlrRQ0sJNmzaVoNtmZlZIs5l8jojrI6IqIqq6du1a7u6Ymb1vlSIY1gM989Z7pG0F60hqCXQAtjRwXzMz24tKEQwLgCMl9ZF0ALnJ5Lk16swFJqblM4DfR0Sk7RPSt5b6AEcCfy5Bn8zMrJFaFttAROyS9GVgHtAC+HlELJM0HVgYEXOBmcAtklYCL5MLD1K9WcBTwC7ggoh4p9g+mZlZ4yn3wb15qaqqioULF5a7G2ZmzYqkRRFRVV+9ZjP5bGZme4eDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDAeDmZllOBjMzCzDwWBmZhkOBjMzy3AwmJlZhoPBzMwyHAxmZpbhYDAzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJiZWYaDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLKOoYJDUSdJ9klakPytqqTcx1VkhaWLa1lbS3ZKWS1om6fJi+mJmZqVR7BnDVOCBiDgSeCCtZ0jqBEwDRgDDgWl5ATIjIo4ChgAjJZ1YZH/MzKxIxQbDKcBNafkm4NQCdU4A7ouIlyNiK3AfMD4idkTEgwAR8RawGOhRZH/MzKxIxQbDIRHxQlp+ETikQJ3uwNq89XVpWzVJHYGTyJ11mJlZGbWsr4Kk+4FDCxR9K38lIkJS7GkHJLUEfgX8OCJW11FvMjAZoFevXnv6NGZm1kD1BkNEjK2tTNJGSYdFxAuSDgNeKlBtPTA6b70HMD9v/XpgRURcVU8/rk91qaqq2uMAMjOzhin2UtJcYGJangjcVaDOPGCcpIo06TwubUPSpUAH4MIi+2FmZiVSbDBcDhwvaQUwNq0jqUrSjQAR8TLwPWBBekyPiJcl9SB3OaofsFjSEknnFdkfMzMrkiKa31WZqqqqWLhwYbm7YWbWrEhaFBFV9dXz/3w2M7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDAeDmZllOBjMzCzDwWBmZhkOBjMzy3AwmJlZhoPBzMwyHAxmZpbhYDAzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJiZWYaDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDKKCgZJnSTdJ2lF+rOilnoTU50VkiYWKJ8raWkxfTEzs9Io9oxhKvBARBwJPJDWMyR1AqYBI4DhwLT8AJF0GrC9yH6YmVmJFBsMpwA3peWbgFML1DkBuC8iXo6IrcB9wHgASQcCXwcuLbIfZmZWIsUGwyER8UJafhE4pECd7sDavPV1aRvA94B/B3bU90SSJktaKGnhpk2biuiymZnVpWV9FSTdDxxaoOhb+SsREZKioU8saTDwoYj4mqTe9dWPiOuB6wGqqqoa/DxmZrZn6g2GiBhbW5mkjZIOi4gXJB0GvFSg2npgdN56D2A+8HdAlaQ1qR8HS5ofEaMxM7OyKfZS0lxg97eMJgJ3FagzDxgnqSJNOo8D5kXEtRHRLSJ6A38PPOtQMDMrv2KD4XLgeEkrgLFpHUlVkm4EiIiXyc0lLEiP6WmbmZntgxTR/C7XV1VVxcKFC8vdDTOzZkXSooioqq+e/+ezmZllOBjMzCzDwWBmZhkOBjMzy3AwmJlZhoPBzMwyHAxmZpbhYDAzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJiZWYaDwczMMhwMZmaW4WAwM7MMB4OZmWU4GMzMLMPBYGZmGQ4GMzPLcDCYmVmGg8HMzDIcDGZmluFgMDOzDEVEufuwxyRtAp4vdz/2UBdgc7k7sZd5zPsHj7n5ODwiutZXqVkGQ3MkaWFEVJW7H3uTx7x/8Jjff3wpyczMMhwMZmaW4WDYe64vdwfKwGPeP3jM7zOeYzAzswyfMZiZWYaDwczMMhwMJSSpk6T7JK1If1bUUm9iqrNC0sQC5XMlLW36HhevmDFLaivpbknLJS2TdPne7f2ekTRe0jOSVkqaWqC8taTbU/ljknrnlV2ctj8j6YS92e9iNHbMko6XtEjSX9KfH9vbfW+MYl7jVN5L0nZJF+2tPjeJiPCjRA/gh8DUtDwV+EGBOp2A1enPirRckVd+GvBLYGm5x9PUYwbaAselOgcADwEnlntMtYyzBbAKOCL19QmgX406XwKuS8sTgNvTcr9UvzXQJ7XTotxjauIxDwG6peX+wPpyj6cpx5tXPhu4A7io3OMp5uEzhtI6BbgpLd8EnFqgzgnAfRHxckRsBe4DxgNIOhD4OnDpXuhrqTR6zBGxIyIeBIiIt4DFQI+90OfGGA6sjIjVqa+3kRt7vvxjMRsYI0lp+20RsTMingNWpvb2dY0ec0Q8HhEb0vZlwAcltd4rvW68Yl5jJJ0KPEduvM2ag6G0DomIF9Lyi8AhBep0B9bmra9L2wC+B/w7sKPJelh6xY4ZAEkdgZOAB5qikyVQ7xjy60TELmAb0LmB++6LihlzvtOBxRGxs4n6WSqNHm/6UPcN4Lt7oZ9NrmW5O9DcSLofOLRA0bfyVyIiJDX4u8CSBgMfioiv1bxuWW5NNea89lsCvwJ+HBGrG9dL2xdJqgR+AIwrd1+a2CXAlRGxPZ1ANGsOhj0UEWNrK5O0UdJhEfGCpMOAlwpUWw+MzlvvAcwH/g6okrSG3OtysKT5ETGaMmvCMe92PbAiIq4qQXebynqgZ956j7StUJ11Kew6AFsauO++qJgxI6kHMAf4XESsavruFq2Y8Y4AzpD0Q6Aj8K6kNyPip03f7SZQ7kmO99MD+BHZidgfFqjTidx1yIr0eA7oVKNOb5rP5HNRYyY3n/Jr4APlHks942xJbtK8D3+bmKysUecCshOTs9JyJdnJ59U0j8nnYsbcMdU/rdzj2BvjrVHnEpr55HPZO/B+epC7tvoAsAK4P+/Nrwq4Ma/eP5KbgFwJnFugneYUDI0eM7lPZAE8DSxJj/PKPaY6xvpx4Fly31z5Vto2HTg5Lbch942UlcCfgSPy9v1W2u8Z9tFvXpVyzMC3gdfzXtclwMHlHk9TvsZ5bTT7YPAtMczMLMPfSjIzswwHg5mZZTgYzMwsw8FgZmYZDgYzM8twMJjtRZJGS/ptufthVhcHg5mZZTgYzAqQdI6kP0taIulnklqk++xfmX474gFJXVPdwZL+JOlJSXN2/yaFpA9Lul/SE5IWS/pQav5ASbPT71DcuvvunGb7CgeDWQ2S+gJnAyMjYjDwDvAZoB2wMCIqgT8A09IuNwPfiIiBwF/ytt8KXB0Rg4D/B+y+C+0Q4EJyv9NwBDCyyQdltgd8Ez2z9xoDDAMWpA/zHyR3c8B3gdtTnf8CfiOpA9AxIv6Qtt8E3CGpPdA9IuYARMSbAKm9P0fEurS+hNwtUB5u+mGZNYyDwey9BNwUERdnNkr/VqNeY+8nk/+7BO/gf4e2j/GlJLP3eoDcLZQPhurftT6c3L+XM1KdfwAejohtwFZJo9L2zwJ/iIjXyN2a+dTURmtJbffa0WExAAAAeUlEQVTqKMwayZ9UzGqIiKckfRu4V9IHgLfJ3W75dWB4KnuJ3DwEwETguvTGvxo4N23/LPAzSdNTG2fuxWGYNZrvrmrWQJK2R8SB5e6HWVPzpSQzM8vwGYOZmWX4jMHMzDIcDGZmluFgMDOzDAeDmZllOBjMzCzj/wB4JDTgHc6f7QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEWCAYAAABi5jCmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAHaBJREFUeJzt3XucVXW9//HXO0AJQRguKlfB5PcQhjsjeA4HQ1HAyvuNjhfwZDyOWv20R5yw+oWSnchIOZWKKHawTEQSo9RQTErTiEtooOgAYgwoAiqKCAp+fn/sxbTXOMMMs/ewZ+D9fDz2Y9blu9b+fPeG/d5rffdeWxGBmZnZXp8qdAFmZla/OBjMzCzFwWBmZikOBjMzS3EwmJlZioPBzMxSHAx2UJL0mKQxha4jHyRdIunxWm47VNLL+a7JDm7y9xjMDh2S1gFXRsSCQtdi9ZePGMzqmDL8f80aDP9jtXpDUmdJD0naLGmrpJ9J+pSk70h6TdKbku6V1DJp31TSL5O270haLOnoZN1CSVcm02MlPSNpiqS3Jb0q6Yys+20paYak1yVtkHSTpEbV1DpW0p+TGrdJWiVpeNb6hZK+L+nPwA7gOEkdJM2T9Jak1ZK+nNX+UUk/zpqfJeme7Pqz1oWkqyWVSnpP0vckfUbSs5LelTRb0mFJ22GSypLpXwBdgN9K2i7pvyQ9IumrFfr2gqRz9/f5s4OHg8HqheSF+HfAa0BXoCMwCxib3E4BjgOaAz9LNhsDtAQ6A22A/wQ+qOIuBgMvA22Bm4EZkpSs+19gN3A80B8YAVxZg7IHA2uSfU4EHpLUOmv9ZcA4oEXSr1lAGdABuAD4b0mnJm3/A7hM0qmSLgEGAf93H/c9EhgInAT8FzAduJTMY9EL+GLFDSLiMuAfwJkR0TwibgZmJtsBIKkvmcf+kRr03w5SDgarLwaRecEcHxHvR8TOiHgGuAS4JSLWRsR24HpgtKTGwEdkAuH4iNgTEUsj4t0q9v9aRNwVEXvIvBi2B45OjjA+B1yb3O+bwK3A6BrU/CYwNSI+iogHyATP57PW/29ErIyI3cAxwBDgm0nflgN3A5cDRMQbwFVJbf8DXB4R7+3jvm+OiHcjYiWwAng8eYy2AY+RCbiamAf8H0ndk/nLgAci4sMabm8HIQeD1Redybx4766wvAOZd9t7vQY0Bo4GfgHMB2ZJ2ijpZklNqtj/G3snImJHMtkcOBZoAryenI56B7gTOKoGNW+I9Kc3Xkvq3Wt9hX68VeHF/jUy7873+i3QCHg5CcV92ZQ1/UEl882r2R6AiNgJPABcmoyDfJHM42qHMAeD1RfrgS7JkUC2jWRevPfqQua0z6bknfqNEdET+FfgCyTvwPfzfncBbSOiVXI7MiKKa7Btx6zTUXtr25g1nx0aG4HWklpUaL8ha/77wEtAe0mfOBWUJ5V9DHEmmSOz4cCOiHiuju7bGggHg9UXfwVeByZLOiIZWB4C3A9cJ6mbpObAf5M51bFb0imSeifjE++SObX08f7caUS8DjwO/FjSkclg92ckfbYGmx8FfE1SE0kXAj2AR6u4n/XAs8APkr71Ab4E/BJA0snAFWSCbQzwU0kdK9tXjjaRGavJru05Mo/bj/HRguFgsHoiOfd/JpkB4H+QGaS9GLiHzIvVn4BXgZ3A3k/RHAPMIRMKLwF/pHYvbJcDhwEvAm8n+2xfg+0WAd2BLWTe7V8QEVv30f6LZAbWNwJzgYkRsUDSkcC9wFciYkNEPA3MAH5e4YgkH34AfCc5bfaNrOX3Ar1JgsoObf6Cm1ktSBpL5oti/1boWvJB0uXAuIOlP5YbHzGYHeIkNQOuJvORVzMHg1lVJE1LvghW8Tat0LXli6SRwGYyYw+/KnA5Vk/4VJKZmaX4iMHMzFIqfma8QWjbtm107dq10GWYmTUoS5cu3RIR7apr1yCDoWvXrixZsqTQZZiZNSiSXqu+lU8lmZlZBQ4GMzNLcTCYmVlKgxxjMLPC++ijjygrK2Pnzp2FLsUqaNq0KZ06daJJk6ouNrxvDgYzq5WysjJatGhB165dyf8lnay2IoKtW7dSVlZGt27darUPn0oys1rZuXMnbdq0cSjUM5Jo06ZNTkdyDgYzqzWHQv2U6/PiYDAzsxQHg5k1SO+88w633357rbb93Oc+xzvvvLPPNt/97ndZsGBBrfafi4cffpgXX3zxgN9vNgeDmTVI+wqG3bsr/nR42qOPPkqrVq322WbSpEmcdtppta6vthwMZma1NGHCBNasWUO/fv0YP348CxcuZOjQoZx11ln07NkTgHPOOYeBAwdSXFzM9On//LmJrl27smXLFtatW0ePHj348pe/THFxMSNGjOCDDz4AYOzYscyZM6e8/cSJExkwYAC9e/dm1apVAGzevJnTTz+d4uJirrzySo499li2bNmSqnPPnj2MHTuWXr160bt3b2699VYA1qxZw6hRoxg4cCBDhw5l1apVPPvss8ybN4/x48fTr18/1qxZU+ePY2X8cVUzy9mNv13Jixvfzes+e3Y4kolnFle5fvLkyaxYsYLly5cDsHDhQpYtW8aKFSvKP6Z5zz330Lp1az744ANOPPFEzj//fNq0aZPaT2lpKffffz933XUXF110Eb/+9a+59NJLP3F/bdu2ZdmyZdx+++1MmTKFu+++mxtvvJFTTz2V66+/nt///vfMmDHjE9stX76cDRs2sGLFCoDyU1jjxo1j2rRpdO/enUWLFnH11Vfzhz/8gbPOOosvfOELXHDBBbV74PLAwWBmB41BgwalPrv/k5/8hLlz5wKwfv16SktLPxEM3bp1o1+/fgAMHDiQdevWVbrv8847r7zNQw89BMAzzzxTvv9Ro0ZRVFT0ie2OO+441q5dy1e/+lU+//nPM2LECLZv386zzz7LhRdeWN5u165dtex1/jkYzCxn+3pnfyAdccQR5dMLFy5kwYIFPPfcczRr1oxhw4ZV+tn+ww8/vHy6UaNG5aeSqmrXqFGjascwshUVFfH8888zf/58pk2bxuzZs5k6dSqtWrUqP9qpbzzGYGYNUosWLXjvvfeqXL9t2zaKiopo1qwZq1at4i9/+UveaxgyZAizZ88G4PHHH+ftt9/+RJstW7bw8ccfc/7553PTTTexbNkyjjzySLp168aDDz4IZL6t/Pzzz9eoXweCg8HMGqQ2bdowZMgQevXqxfjx4z+xftSoUezevZsePXowYcIETjrppLzXMHHiRB5//HF69erFgw8+yDHHHEOLFi1SbTZs2MCwYcPo168fl156KT/4wQ8AuO+++5gxYwZ9+/aluLiY3/zmNwCMHj2aH/3oR/Tv379gg88N8jefS0pKwj/UY1ZYL730Ej169Ch0GQW1a9cuGjVqROPGjXnuuee46qqr6s3pocqeH0lLI6Kkum09xmBmVkv/+Mc/uOiii/j444857LDDuOuuuwpdUl44GMzMaql79+787W9/K3QZeecxBjMzS3EwmJlZioPBzMxSHAxmZpbiYDCzQ0bz5s0B2LhxY5XXIho2bBjVfRx+6tSp7Nixo3y+Jpfxzrd169bxq1/9qk727WAws0NOhw4dyq+cWhsVg6Eml/HOt3ofDJJGSXpZ0mpJEypZf7ikB5L1iyR1rbC+i6Ttkr6Rj3rM7OA3YcIEbrvttvL5G264gSlTprB9+3aGDx9efonsvd8ozrZu3Tp69eoFwAcffMDo0aPp0aMH5557bupaSVdddRUlJSUUFxczceJEIHNhvo0bN3LKKadwyimnAP+8jDfALbfcQq9evejVqxdTp04tv7+qLu+d7cEHH6RXr1707duXk08+Gchctnv8+PGceOKJ9OnThzvvvLO8/08//TT9+vUrv5R3vuT8PQZJjYDbgNOBMmCxpHkRkf1LE18C3o6I4yWNBn4IXJy1/hbgsVxrMbMCeWwCvPH3/O7zmN5wxuQqV1988cVce+21XHPNNQDMnj2b+fPn07RpU+bOncuRRx7Jli1bOOmkkzjrrLOq/B3kO+64g2bNmvHSSy/xwgsvMGDAgPJ13//+92ndujV79uxh+PDhvPDCC3zta1/jlltu4amnnqJt27apfS1dupSf//znLFq0iIhg8ODBfPazn6WoqKhGl/eeNGkS8+fPp2PHjuWnpmbMmEHLli1ZvHgxu3btYsiQIYwYMYLJkyczZcoUfve739Xq4d2XfBwxDAJWR8TaiPgQmAWcXaHN2cDMZHoOMFzJsyTpHOBVYGUeajGzQ0T//v1588032bhxI88//zxFRUV07tyZiOBb3/oWffr04bTTTmPDhg1s2rSpyv386U9/Kn+B7tOnD3369ClfN3v2bAYMGED//v1ZuXJltb+s9swzz3DuuedyxBFH0Lx5c8477zyefvppoGaX9x4yZAhjx47lrrvuYs+ePUDm4nz33nsv/fr1Y/DgwWzdupXS0tL9eqz2Vz6++dwRWJ81XwYMrqpNROyWtA1oI2kn8E0yRxv7PI0kaRwwDqBLly55KNvM8mYf7+zr0oUXXsicOXN44403uPjizEmI++67j82bN7N06VKaNGlC165dK73cdnVeffVVpkyZwuLFiykqKmLs2LG12s9eNbm897Rp01i0aBGPPPIIAwcOZOnSpUQEP/3pTxk5cmSq7cKFC2tdS3UKPfh8A3BrRGyvrmFETI+IkogoadeuXd1XZmb13sUXX8ysWbOYM2dO+Y/ebNu2jaOOOoomTZrw1FNP8dprr+1zHyeffHL5IO6KFSt44YUXAHj33Xc54ogjaNmyJZs2beKxx/55truqS2MPHTqUhx9+mB07dvD+++8zd+5chg4dWuP+rFmzhsGDBzNp0iTatWvH+vXrGTlyJHfccQcfffQRAK+88grvv/9+nV6eOx9HDBuAzlnznZJllbUpk9QYaAlsJXNkcYGkm4FWwMeSdkbEz/JQl5kd5IqLi3nvvffo2LEj7du3B+CSSy7hzDPPpHfv3pSUlHDCCSfscx9XXXUVV1xxBT169KBHjx4MHDgQgL59+9K/f39OOOEEOnfuzJAhQ8q3GTduHKNGjaJDhw489dRT5csHDBjA2LFjGTRoEABXXnkl/fv3r/JX4SoaP348paWlRATDhw+nb9++9OnTh3Xr1jFgwAAignbt2vHwww/Tp08fGjVqRN++fRk7dizXXXfd/jx0+5TzZbeTF/pXgOFkAmAx8O8RsTKrzTVA74j4z2Tw+byIuKjCfm4AtkfElOru05fdNis8X3a7fivoZbeTMYOvAPOBRsA9EbFS0iRgSUTMA2YAv5C0GngLGJ3r/ZqZWd3Iy2W3I+JR4NEKy76bNb0TuLDidhXa35CPWszMLDeFHnw2swasIf4C5KEg1+fFwWBmtdK0aVO2bt3qcKhnIoKtW7fStGnTWu/Dv+BmZrXSqVMnysrK2Lx5c6FLsQqaNm1Kp06dar29g8HMaqVJkyZ069at0GVYHfCpJDMzS3EwmJlZioPBzMxSHAxmZpbiYDAzsxQHg5mZpTgYzMwsxcFgZmYpDgYzM0txMJiZWYqDwczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzFwWBmZikOBjMzS3EwmJlZSl6CQdIoSS9LWi1pQiXrD5f0QLJ+kaSuyfLTJS2V9Pfk76n5qMfMzGov52CQ1Ai4DTgD6Al8UVLPCs2+BLwdEccDtwI/TJZvAc6MiN7AGOAXudZjZma5yccRwyBgdUSsjYgPgVnA2RXanA3MTKbnAMMlKSL+FhEbk+UrgU9LOjwPNZmZWS3lIxg6Auuz5suSZZW2iYjdwDagTYU25wPLImJXHmoyM7NaalzoAgAkFZM5vTRiH23GAeMAunTpcoAqMzM79OTjiGED0DlrvlOyrNI2khoDLYGtyXwnYC5weUSsqepOImJ6RJREREm7du3yULaZmVUmH8GwGOguqZukw4DRwLwKbeaRGVwGuAD4Q0SEpFbAI8CEiPhzHmoxM7Mc5RwMyZjBV4D5wEvA7IhYKWmSpLOSZjOANpJWA18H9n6k9SvA8cB3JS1PbkflWpOZmdWeIqLQNey3kpKSWLJkSaHLMDNrUCQtjYiS6tr5m89mZpbiYDAzsxQHg5mZpTgYzMwsxcFgZmYpDgYzM0txMJiZWYqDwczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzFwWBmZikOBjMzS3EwmJlZioPBzMxSHAxmZpbiYDAzsxQHg5mZpTgYzMwsxcFgZmYpDgYzM0txMJiZWUpegkHSKEkvS1otaUIl6w+X9ECyfpGkrlnrrk+WvyxpZD7qMTOz2ss5GCQ1Am4DzgB6Al+U1LNCsy8Bb0fE8cCtwA+TbXsCo4FiYBRwe7I/MzMrkHwcMQwCVkfE2oj4EJgFnF2hzdnAzGR6DjBckpLlsyJiV0S8CqxO9mdmZgWSj2DoCKzPmi9LllXaJiJ2A9uANjXcFgBJ4yQtkbRk8+bNeSjbzMwq02AGnyNiekSURERJu3btCl2OmdlBKx/BsAHonDXfKVlWaRtJjYGWwNYabmtmZgdQPoJhMdBdUjdJh5EZTJ5Xoc08YEwyfQHwh4iIZPno5FNL3YDuwF/zUJOZmdVS41x3EBG7JX0FmA80Au6JiJWSJgFLImIeMAP4haTVwFtkwoOk3WzgRWA3cE1E7Mm1JjMzqz1l3rg3LCUlJbFkyZJCl2Fm1qBIWhoRJdW1azCDz2ZmdmA4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzFwWBmZikOBjMzS3EwmJlZioPBzMxSHAxmZpbiYDAzsxQHg5mZpTgYzMwsxcFgZmYpDgYzM0txMJiZWYqDwczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFJyCgZJrSU9Iak0+VtURbsxSZtSSWOSZc0kPSJplaSVkibnUouZmeVHrkcME4AnI6I78GQynyKpNTARGAwMAiZmBciUiDgB6A8MkXRGjvWYmVmOcg2Gs4GZyfRM4JxK2owEnoiItyLibeAJYFRE7IiIpwAi4kNgGdApx3rMzCxHuQbD0RHxejL9BnB0JW06Auuz5suSZeUktQLOJHPUYWZmBdS4ugaSFgDHVLLq29kzERGSYn8LkNQYuB/4SUSs3Ue7ccA4gC5duuzv3ZiZWQ1VGwwRcVpV6yRtktQ+Il6X1B54s5JmG4BhWfOdgIVZ89OB0oiYWk0d05O2lJSU7HcAmZlZzeR6KmkeMCaZHgP8ppI284ERkoqSQecRyTIk3QS0BK7NsQ4zM8uTXINhMnC6pFLgtGQeSSWS7gaIiLeA7wGLk9ukiHhLUicyp6N6AsskLZd0ZY71mJlZjhTR8M7KlJSUxJIlSwpdhplZgyJpaUSUVNfO33w2M7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzFwWBmZikOBjMzS3EwmJlZioPBzMxSHAxmZpbiYDAzsxQHg5mZpTgYzMwsxcFgZmYpDgYzM0txMJiZWYqDwczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFJyCgZJrSU9Iak0+VtURbsxSZtSSWMqWT9P0opcajEzs/zI9YhhAvBkRHQHnkzmUyS1BiYCg4FBwMTsAJF0HrA9xzrMzCxPcg2Gs4GZyfRM4JxK2owEnoiItyLibeAJYBSApObA14GbcqzDzMzyJNdgODoiXk+m3wCOrqRNR2B91nxZsgzge8CPgR3V3ZGkcZKWSFqyefPmHEo2M7N9aVxdA0kLgGMqWfXt7JmICElR0zuW1A/4TERcJ6lrde0jYjowHaCkpKTG92NmZvun2mCIiNOqWidpk6T2EfG6pPbAm5U02wAMy5rvBCwE/gUokbQuqeMoSQsjYhhmZlYwuZ5Kmgfs/ZTRGOA3lbSZD4yQVJQMOo8A5kfEHRHRISK6Av8GvOJQMDMrvFyDYTJwuqRS4LRkHkklku4GiIi3yIwlLE5uk5JlZmZWDymi4Z2uLykpiSVLlhS6DDOzBkXS0ogoqa6dv/lsZmYpDgYzM0txMJiZWYqDwczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYGZmKQ4GMzNLcTCYmVmKg8HMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzFwWBmZikOBjMzS3EwmJlZioPBzMxSHAxmZpbiYDAzsxQHg5mZpTgYzMwsRRFR6Br2m6TNwGuFrmM/tQW2FLqIA8x9PjS4zw3HsRHRrrpGDTIYGiJJSyKipNB1HEju86HBfT74+FSSmZmlOBjMzCzFwXDgTC90AQXgPh8a3OeDjMcYzMwsxUcMZmaW4mAwM7MUB0MeSWot6QlJpcnfoirajUnalEoaU8n6eZJW1H3Fuculz5KaSXpE0ipJKyVNPrDV7x9JoyS9LGm1pAmVrD9c0gPJ+kWSumatuz5Z/rKkkQey7lzUts+STpe0VNLfk7+nHujaayOX5zhZ30XSdknfOFA114mI8C1PN+BmYEIyPQH4YSVtWgNrk79FyXRR1vrzgF8BKwrdn7ruM9AMOCVpcxjwNHBGoftURT8bAWuA45Janwd6VmhzNTAtmR4NPJBM90zaHw50S/bTqNB9quM+9wc6JNO9gA2F7k9d9jdr/RzgQeAbhe5PLjcfMeTX2cDMZHomcE4lbUYCT0TEWxHxNvAEMApAUnPg68BNB6DWfKl1nyNiR0Q8BRARHwLLgE4HoObaGASsjoi1Sa2zyPQ9W/ZjMQcYLknJ8lkRsSsiXgVWJ/ur72rd54j4W0RsTJavBD4t6fADUnXt5fIcI+kc4FUy/W3QHAz5dXREvJ5MvwEcXUmbjsD6rPmyZBnA94AfAzvqrML8y7XPAEhqBZwJPFkXReZBtX3IbhMRu4FtQJsablsf5dLnbOcDyyJiVx3VmS+17m/ypu6bwI0HoM4617jQBTQ0khYAx1Sy6tvZMxERkmr8WWBJ/YDPRMR1Fc9bFlpd9Tlr/42B+4GfRMTa2lVp9ZGkYuCHwIhC11LHbgBujYjtyQFEg+Zg2E8RcVpV6yRtktQ+Il6X1B54s5JmG4BhWfOdgIXAvwAlktaReV6OkrQwIoZRYHXY572mA6URMTUP5daVDUDnrPlOybLK2pQlYdcS2FrDbeujXPqMpE7AXODyiFhT9+XmLJf+DgYukHQz0Ar4WNLOiPhZ3ZddBwo9yHEw3YAfkR6IvbmSNq3JnIcsSm6vAq0rtOlKwxl8zqnPZMZTfg18qtB9qaafjckMmnfjnwOTxRXaXEN6YHJ2Ml1MevB5LQ1j8DmXPrdK2p9X6H4ciP5WaHMDDXzwueAFHEw3MudWnwRKgQVZL34lwN1Z7f6DzADkauCKSvbTkIKh1n0m844sgJeA5cntykL3aR99/RzwCplPrnw7WTYJOCuZbkrmEymrgb8Cx2Vt++1ku5epp5+8ymefge8A72c9r8uBowrdn7p8jrP20eCDwZfEMDOzFH8qyczMUhwMZmaW4mAwM7MUB4OZmaU4GMzMLMXBYHYASRom6XeFrsNsXxwMZmaW4mAwq4SkSyX9VdJySXdKapRcZ//W5LcjnpTULmnbT9JfJL0gae7e36SQdLykBZKel7RM0meS3TeXNCf5HYr79l6d06y+cDCYVSCpB3AxMCQi+gF7gEuAI4AlEVEM/BGYmGxyL/DNiOgD/D1r+X3AbRHRF/hXYO9VaPsD15L5nYbjgCF13imz/eCL6Jl90nBgILA4eTP/aTIXB/wYeCBp80vgIUktgVYR8cdk+UzgQUktgI4RMRcgInYCJPv7a0SUJfPLyVwC5Zm675ZZzTgYzD5JwMyIuD61UPp/FdrV9noy2b9LsAf/P7R6xqeSzD7pSTKXUD4Kyn/X+lgy/18uSNr8O/BMRGwD3pY0NFl+GfDHiHiPzKWZz0n2cbikZge0F2a15HcqZhVExIuSvgM8LulTwEdkLrf8PjAoWfcmmXEIgDHAtOSFfy1wRbL8MuBOSZOSfVx4ALthVmu+uqpZDUnaHhHNC12HWV3zqSQzM0vxEYOZmaX4iMHMzFIcDGZmluJgMDOzFAeDmZmlOBjMzCzl/wOXrGnlZQ/6GQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#for val in df.columns[1:]:\n",
    "#    plt.title(val)\n",
    "#    plt.plot(df.epoch,df[val])\n",
    "#    plt.show()\n",
    "\n",
    "for val in hs.history.keys():\n",
    "    if \"val_\" not in val:\n",
    "        plt.title(val)\n",
    "        plt.plot(hs.history[val])\n",
    "        plt.plot(hs.history['val_' + val])\n",
    "        plt.xlabel('epoch')\n",
    "        plt.legend(['training set', 'validation set'], loc='upper right')\n",
    "        plt.savefig('models/' + model_name_full + '_' + val + '.png')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "try:\n",
    "    plt.plot(hs.history['lr'])\n",
    "    plt.title('Learning rate')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.show()\n",
    "except:\n",
    "    pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Eğittiğimiz modeli kaydediyoruz\n",
    "# Ağ yapsını json olarak\n",
    "# Ağ parametre değerlerini h5 uzantılı olarak\n",
    "import json \n",
    "# Save model weights and json.\n",
    "model.save_weights('models/' + model_name_full + '_model.h5')\n",
    "model_json  = model.to_json()\n",
    "with open('models/' + model_name_full + '_model.json', 'w') as outfile:\n",
    "    json.dump(model_json, outfile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0.0, ' - > ', 0.3427734375)\n",
      "(-0.258572280407, ' - > ', -0.19303385416666666)\n",
      "(-0.120483756065, ' - > ', -0.1484375)\n",
      "(0.0, ' - > ', -0.032206217447916664)\n",
      "(-0.0921533256769, ' - > ', -0.12451171875)\n",
      "(0.15944202244300001, ' - > ', 0.16748046875)\n",
      "(-0.180676743388, ' - > ', -0.17333984375)\n",
      "(-0.155888929963, ' - > ', -0.16959635416666666)\n",
      "(0.0, ' - > ', 0.0023066202799479165)\n",
      "(-0.0992385521531, ' - > ', -0.07478841145833333)\n",
      "(-0.173591509461, ' - > ', -0.19401041666666666)\n",
      "(-0.0779933556914, ' - > ', -0.06363932291666667)\n",
      "(-0.15943153202500002, ' - > ', -0.15999348958333334)\n",
      "(-0.11694113910199999, ' - > ', -0.17626953125)\n",
      "(0.340000003576, ' - > ', 0.23828125)\n",
      "(0.0, ' - > ', -0.035990397135416664)\n",
      "(-0.340000003576, ' - > ', -0.24772135416666666)\n",
      "(-0.10278116911599999, ' - > ', -0.155029296875)\n",
      "(-0.0602907724679, ' - > ', -0.0811767578125)\n",
      "(0.0, ' - > ', -0.08528645833333333)\n",
      "(0.0567586384714, ' - > ', 0.085693359375)\n",
      "(0.021353468298900002, ' - > ', 0.14705403645833334)\n",
      "(-0.0284282118082, ' - > ', 0.021952311197916668)\n",
      "(-0.053216021508, ' - > ', -0.010045369466145834)\n",
      "(-0.322297394276, ' - > ', -0.2693684895833333)\n",
      "(0.0, ' - > ', 0.15909830729166666)\n",
      "(-0.127568975091, ' - > ', -0.19287109375)\n",
      "(0.0, ' - > ', -0.021809895833333332)\n",
      "(0.102791652083, ' - > ', 0.12923177083333334)\n",
      "(-0.166516765952, ' - > ', -0.13167317708333334)\n",
      "(-0.187751471996, ' - > ', -0.13264973958333334)\n",
      "(-0.152346313, ' - > ', -0.16471354166666666)\n",
      "(-0.15943153202500002, ' - > ', -0.16927083333333334)\n",
      "(-0.170048907399, ' - > ', -0.16072591145833334)\n",
      "(0.0, ' - > ', 0.0499267578125)\n",
      "(-0.120483756065, ' - > ', -0.12353515625)\n",
      "(0.0, ' - > ', -0.024088541666666668)\n",
      "(-0.212539300323, ' - > ', -0.19954427083333334)\n",
      "(0.0, ' - > ', -0.06742350260416667)\n",
      "(0.0, ' - > ', -0.035664876302083336)\n",
      "(-0.152346313, ' - > ', -0.11873372395833333)\n",
      "(0.0, ' - > ', -0.08699544270833333)\n",
      "(0.0, ' - > ', -0.162841796875)\n",
      "(0.0, ' - > ', 0.038248697916666664)\n",
      "(-0.340000003576, ' - > ', -0.19986979166666666)\n",
      "(0.0, ' - > ', 0.2841796875)\n",
      "(0.0, ' - > ', -0.07755533854166667)\n",
      "(0.0, ' - > ', -0.00348663330078125)\n",
      "(-0.308137446642, ' - > ', -0.24674479166666666)\n",
      "(-0.155888929963, ' - > ', -0.15681966145833334)\n",
      "(-0.0992385521531, ' - > ', -0.11344401041666667)\n",
      "(-0.340000003576, ' - > ', -0.3356119791666667)\n",
      "(0.0, ' - > ', 0.035481770833333336)\n",
      "(0.0, ' - > ', -0.0594482421875)\n",
      "(0.131111592054, ' - > ', 0.1103515625)\n",
      "(0.340000003576, ' - > ', 0.23990885416666666)\n",
      "(0.0, ' - > ', 0.058878580729166664)\n",
      "(-0.106323778629, ' - > ', -0.16715494791666666)\n",
      "(0.0, ' - > ', 0.028035481770833332)\n",
      "(0.0, ' - > ', -0.0445556640625)\n",
      "(-0.14881418645400002, ' - > ', -0.15462239583333334)\n",
      "(0.0, ' - > ', -0.014139811197916666)\n",
      "(-0.141728952527, ' - > ', -0.132568359375)\n",
      "(-0.297520071268, ' - > ', -0.2916666666666667)\n",
      "(0.340000003576, ' - > ', 0.314453125)\n",
      "(0.0, ' - > ', -0.28955078125)\n",
      "(-0.04967341199519999, ' - > ', -0.06522623697916667)\n",
      "(-0.269189655781, ' - > ', -0.2578125)\n",
      "(-0.198379307985, ' - > ', -0.16845703125)\n",
      "(0.0, ' - > ', 0.052042643229166664)\n",
      "(0.15944202244300001, ' - > ', 0.123046875)\n",
      "(0.0, ' - > ', -0.08199055989583333)\n",
      "(0.187761962414, ' - > ', 0.15706380208333334)\n",
      "(0.0, ' - > ', 0.0355224609375)\n",
      "(0.0, ' - > ', 0.2643229166666667)\n",
      "(0.131111592054, ' - > ', 0.11515299479166667)\n",
      "(0.0, ' - > ', -0.01080322265625)\n",
      "(0.0, ' - > ', 0.006988525390625)\n",
      "(0.0, ' - > ', 0.08748372395833333)\n",
      "(0.0, ' - > ', 0.044474283854166664)\n",
      "(-0.20192193984999998, ' - > ', -0.16552734375)\n",
      "(-0.0107256285846, ' - > ', -0.005498250325520833)\n",
      "(-0.21962451934799998, ' - > ', -0.22526041666666666)\n",
      "(0.0, ' - > ', 0.0506591796875)\n",
      "(0.0, ' - > ', 0.028238932291666668)\n",
      "(0.340000003576, ' - > ', 0.20556640625)\n",
      "(0.0, ' - > ', 0.24007161458333334)\n",
      "(0.0, ' - > ', -0.041524251302083336)\n",
      "(0.0, ' - > ', -0.18489583333333334)\n",
      "(0.102791652083, ' - > ', 0.16861979166666666)\n",
      "(0.0, ' - > ', 0.24723307291666666)\n",
      "(0.0, ' - > ', 0.10595703125)\n",
      "(-0.194836705923, ' - > ', -0.21630859375)\n",
      "(0.0, ' - > ', -0.020853678385416668)\n",
      "(-0.07445074617860001, ' - > ', -0.14794921875)\n",
      "(-0.127568975091, ' - > ', -0.134033203125)\n",
      "(-0.109866395593, ' - > ', -0.14436848958333334)\n",
      "(0.0, ' - > ', 0.004987080891927083)\n",
      "(0.0, ' - > ', 0.09993489583333333)\n",
      "(-0.053216021508, ' - > ', -0.06803385416666667)\n"
     ]
    }
   ],
   "source": [
    "# rastgele 10 resim seçip modelimiz hesapladığı sonuçla gerçeğine bakıyoruz \n",
    "# Eğer sonuçlar iyi ise kullanabiliriz\n",
    "# Sonuççlar kötüyse eğitim aşamasına dönmemiz lazım\n",
    "# Compare actual and predicted steering\n",
    "for i in range(100):\n",
    "    ix = random.randint(0,len(df)-1)\n",
    "    out = model.predict(get_matrix(df.FileName[ix]).reshape(1,args['image_height'],args['image_width'],3))\n",
    "    print(df.Angle[ix], ' - > ', out[0][0]/3)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}