HOD model using VPF, WP Keras.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load the autoreload extension\n",
    "%reload_ext autoreload\n",
    "# Set extension to reload modules every time before executing code\n",
    "%autoreload 2\n",
    "\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import os\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/home/ajana/.local/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow import keras\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = '/mnt/data4/Abhishek/WP/'\n",
    "path_vpf = '/mnt/data4/Abhishek/VPF/random/'\n",
    "f = np.loadtxt('/home/ajana/github/HODProject/parameters.txt')\n",
    "\n",
    "def extractIndex(filename):\n",
    "    return int(filename.split('_')[-1][:4])\n",
    "\n",
    "random_data = []\n",
    "\n",
    "for filename in os.listdir(path):\n",
    "    if filename.startswith('wp_galaxies'):\n",
    "        pos = extractIndex(filename)\n",
    "        wp,_ = np.load(os.path.join(path,filename))\n",
    "        _,vpf = np.load(os.path.join(path_vpf,'vpf_'+filename[3:]))\n",
    "        _temp = np.concatenate((vpf,wp,f[pos]))\n",
    "        random_data.append(_temp)\n",
    "        \n",
    "full_data = np.vstack((random_data))\n",
    "del random_data\n",
    "np.random.shuffle(full_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_data = full_data[:,:-5]\n",
    "y_data = full_data[:,-5:]\n",
    "from sklearn.preprocessing import  MinMaxScaler\n",
    "sc= MinMaxScaler()\n",
    "X_data= sc.fit_transform(X_data)\n",
    "y_data= y_data.reshape(-1,5)\n",
    "y_data=sc.fit_transform(y_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_corr(pred,test):\n",
    "    sns.jointplot(pred.iloc[:,0],test.iloc[:,0],kind = \"kde\")\n",
    "    sns.jointplot(pred.iloc[:,1],test.iloc[:,1],kind = 'kde',color='purple')\n",
    "    sns.jointplot(pred.iloc[:,2],test.iloc[:,2],kind = 'kde',color='blue')\n",
    "    sns.jointplot(pred.iloc[:,3],test.iloc[:,3],kind = 'kde',color='green')\n",
    "    sns.jointplot(pred.iloc[:,4],test.iloc[:,4],kind = 'kde',color='red')\n",
    "    plt.tight_layout\n",
    "    \n",
    "def plot_error(pred,test,dims=(12,8)):\n",
    "    fig,ax = plt.subplots(figsize=dims)\n",
    "    sns.distplot(pred.iloc[:,0]-test.iloc[:,0],color='blue',ax=ax,kde_kws={\"label\": \"M_cut\"})\n",
    "    sns.distplot(pred.iloc[:,1]-test.iloc[:,1],color='cyan',ax=ax,kde_kws={\"label\": \"M_1\"})\n",
    "    sns.distplot(pred.iloc[:,2]-test.iloc[:,2],color='purple',ax=ax,kde_kws={\"label\": \"Sigma\"})\n",
    "    sns.distplot(pred.iloc[:,3]-test.iloc[:,3],color='yellow',ax=ax,kde_kws={\"label\": \"Kappa\"})\n",
    "    sns.distplot(pred.iloc[:,4]-test.iloc[:,4],color='red',ax=ax,kde_kws={\"label\": \"Alpha\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_data, y_data, test_size=0.1,random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "in_dim = X_train.shape[1]\n",
    "out_dim = y_train.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(128, input_dim=in_dim, activation=\"relu\"))\n",
    "model.add(Dense(64, activation=\"relu\"))\n",
    "model.add(Dense(32, activation=\"relu\"))\n",
    "model.add(Dense(out_dim))\n",
    "model.compile(loss=\"mse\", optimizer=\"adam\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model(in_dim,out_dim):\n",
    "    model = keras.Sequential([\n",
    "        layers.Dense(128, activation='relu', input_dim=in_dim),\n",
    "        layers.Dense(128, activation='relu'),\n",
    "        layers.Dense(64, activation='relu'),\n",
    "        layers.Dense(out_dim)\n",
    "      ])\n",
    "\n",
    "    optimizer = tf.keras.optimizers.RMSprop(0.0001)\n",
    "\n",
    "    model.compile(loss='mse',\n",
    "                    optimizer=optimizer,\n",
    "                    metrics=['mae', 'mse'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_8 (Dense)              (None, 128)               7680      \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 128)               16512     \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 64)                8256      \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 5)                 325       \n",
      "=================================================================\n",
      "Total params: 32,773\n",
      "Trainable params: 32,773\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "mod = build_model(in_dim,out_dim)\n",
    "mod.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#y_rf = pd.DataFrame(y_rf)\n",
    "\n",
    "#y_test.reset_index(drop=True, inplace=True)\n",
    "#y_rf.reset_index(drop=True, inplace=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 3600 samples, validate on 900 samples\n",
      "Epoch 1/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0206 - mean_absolute_error: 0.0935 - mean_squared_error: 0.0206 - val_loss: 0.0293 - val_mean_absolute_error: 0.1279 - val_mean_squared_error: 0.0293\n",
      "Epoch 2/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0205 - mean_absolute_error: 0.0935 - mean_squared_error: 0.0205 - val_loss: 0.0193 - val_mean_absolute_error: 0.0861 - val_mean_squared_error: 0.0193\n",
      "Epoch 3/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0207 - mean_absolute_error: 0.0943 - mean_squared_error: 0.0207 - val_loss: 0.0196 - val_mean_absolute_error: 0.0881 - val_mean_squared_error: 0.0196\n",
      "Epoch 4/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0203 - mean_absolute_error: 0.0930 - mean_squared_error: 0.0203 - val_loss: 0.0204 - val_mean_absolute_error: 0.0930 - val_mean_squared_error: 0.0204\n",
      "Epoch 5/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0205 - mean_absolute_error: 0.0930 - mean_squared_error: 0.0205 - val_loss: 0.0202 - val_mean_absolute_error: 0.0919 - val_mean_squared_error: 0.0202\n",
      "Epoch 6/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0203 - mean_absolute_error: 0.0922 - mean_squared_error: 0.0203 - val_loss: 0.0214 - val_mean_absolute_error: 0.0980 - val_mean_squared_error: 0.0214\n",
      "Epoch 7/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0204 - mean_absolute_error: 0.0925 - mean_squared_error: 0.0204 - val_loss: 0.0198 - val_mean_absolute_error: 0.0905 - val_mean_squared_error: 0.0198\n",
      "Epoch 8/600\n",
      "3600/3600 [==============================] - 0s 84us/sample - loss: 0.0203 - mean_absolute_error: 0.0927 - mean_squared_error: 0.0203 - val_loss: 0.0237 - val_mean_absolute_error: 0.1039 - val_mean_squared_error: 0.0237\n",
      "Epoch 9/600\n",
      "3600/3600 [==============================] - 0s 62us/sample - loss: 0.0204 - mean_absolute_error: 0.0928 - mean_squared_error: 0.0204 - val_loss: 0.0190 - val_mean_absolute_error: 0.0848 - val_mean_squared_error: 0.0190\n",
      "Epoch 10/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0203 - mean_absolute_error: 0.0920 - mean_squared_error: 0.0203 - val_loss: 0.0189 - val_mean_absolute_error: 0.0845 - val_mean_squared_error: 0.0189\n",
      "Epoch 11/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0202 - mean_absolute_error: 0.0913 - mean_squared_error: 0.0202 - val_loss: 0.0189 - val_mean_absolute_error: 0.0846 - val_mean_squared_error: 0.0189\n",
      "Epoch 12/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0202 - mean_absolute_error: 0.0919 - mean_squared_error: 0.0202 - val_loss: 0.0203 - val_mean_absolute_error: 0.0924 - val_mean_squared_error: 0.0203\n",
      "Epoch 13/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0203 - mean_absolute_error: 0.0919 - mean_squared_error: 0.0203 - val_loss: 0.0189 - val_mean_absolute_error: 0.0843 - val_mean_squared_error: 0.0189\n",
      "Epoch 14/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0202 - mean_absolute_error: 0.0918 - mean_squared_error: 0.0202 - val_loss: 0.0224 - val_mean_absolute_error: 0.0993 - val_mean_squared_error: 0.0224\n",
      "Epoch 15/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0202 - mean_absolute_error: 0.0919 - mean_squared_error: 0.0202 - val_loss: 0.0196 - val_mean_absolute_error: 0.0888 - val_mean_squared_error: 0.0196\n",
      "Epoch 16/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0201 - mean_absolute_error: 0.0913 - mean_squared_error: 0.0201 - val_loss: 0.0191 - val_mean_absolute_error: 0.0845 - val_mean_squared_error: 0.0191\n",
      "Epoch 17/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0200 - mean_absolute_error: 0.0912 - mean_squared_error: 0.0200 - val_loss: 0.0191 - val_mean_absolute_error: 0.0860 - val_mean_squared_error: 0.0191\n",
      "Epoch 18/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0200 - mean_absolute_error: 0.0911 - mean_squared_error: 0.0200 - val_loss: 0.0194 - val_mean_absolute_error: 0.0862 - val_mean_squared_error: 0.0194\n",
      "Epoch 19/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0199 - mean_absolute_error: 0.0909 - mean_squared_error: 0.0199 - val_loss: 0.0195 - val_mean_absolute_error: 0.0868 - val_mean_squared_error: 0.0195\n",
      "Epoch 20/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0200 - mean_absolute_error: 0.0912 - mean_squared_error: 0.0200 - val_loss: 0.0197 - val_mean_absolute_error: 0.0889 - val_mean_squared_error: 0.0197\n",
      "Epoch 21/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0199 - mean_absolute_error: 0.0908 - mean_squared_error: 0.0199 - val_loss: 0.0188 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0188\n",
      "Epoch 22/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0199 - mean_absolute_error: 0.0906 - mean_squared_error: 0.0199 - val_loss: 0.0190 - val_mean_absolute_error: 0.0847 - val_mean_squared_error: 0.0190\n",
      "Epoch 23/600\n",
      "3600/3600 [==============================] - 0s 76us/sample - loss: 0.0199 - mean_absolute_error: 0.0909 - mean_squared_error: 0.0199 - val_loss: 0.0217 - val_mean_absolute_error: 0.0977 - val_mean_squared_error: 0.0217\n",
      "Epoch 24/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0199 - mean_absolute_error: 0.0908 - mean_squared_error: 0.0199 - val_loss: 0.0188 - val_mean_absolute_error: 0.0845 - val_mean_squared_error: 0.0188\n",
      "Epoch 25/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0199 - mean_absolute_error: 0.0904 - mean_squared_error: 0.0199 - val_loss: 0.0220 - val_mean_absolute_error: 0.0991 - val_mean_squared_error: 0.0220\n",
      "Epoch 26/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0197 - mean_absolute_error: 0.0901 - mean_squared_error: 0.0197 - val_loss: 0.0208 - val_mean_absolute_error: 0.0949 - val_mean_squared_error: 0.0208\n",
      "Epoch 27/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0197 - mean_absolute_error: 0.0898 - mean_squared_error: 0.0197 - val_loss: 0.0194 - val_mean_absolute_error: 0.0882 - val_mean_squared_error: 0.0194\n",
      "Epoch 28/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0198 - mean_absolute_error: 0.0904 - mean_squared_error: 0.0198 - val_loss: 0.0194 - val_mean_absolute_error: 0.0850 - val_mean_squared_error: 0.0194\n",
      "Epoch 29/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0197 - mean_absolute_error: 0.0895 - mean_squared_error: 0.0197 - val_loss: 0.0196 - val_mean_absolute_error: 0.0894 - val_mean_squared_error: 0.0196\n",
      "Epoch 30/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0196 - mean_absolute_error: 0.0888 - mean_squared_error: 0.0196 - val_loss: 0.0206 - val_mean_absolute_error: 0.0942 - val_mean_squared_error: 0.0206\n",
      "Epoch 31/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0197 - mean_absolute_error: 0.0901 - mean_squared_error: 0.0197 - val_loss: 0.0195 - val_mean_absolute_error: 0.0842 - val_mean_squared_error: 0.0195\n",
      "Epoch 32/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0199 - mean_absolute_error: 0.0905 - mean_squared_error: 0.0199 - val_loss: 0.0187 - val_mean_absolute_error: 0.0849 - val_mean_squared_error: 0.0187\n",
      "Epoch 33/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0195 - mean_absolute_error: 0.0885 - mean_squared_error: 0.0195 - val_loss: 0.0195 - val_mean_absolute_error: 0.0876 - val_mean_squared_error: 0.0195\n",
      "Epoch 34/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0195 - mean_absolute_error: 0.0885 - mean_squared_error: 0.0195 - val_loss: 0.0203 - val_mean_absolute_error: 0.0911 - val_mean_squared_error: 0.0203\n",
      "Epoch 35/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0195 - mean_absolute_error: 0.0887 - mean_squared_error: 0.0195 - val_loss: 0.0186 - val_mean_absolute_error: 0.0832 - val_mean_squared_error: 0.0186\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 36/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0197 - mean_absolute_error: 0.0893 - mean_squared_error: 0.0197 - val_loss: 0.0191 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0191\n",
      "Epoch 37/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0196 - mean_absolute_error: 0.0893 - mean_squared_error: 0.0196 - val_loss: 0.0212 - val_mean_absolute_error: 0.0965 - val_mean_squared_error: 0.0212\n",
      "Epoch 38/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0195 - mean_absolute_error: 0.0889 - mean_squared_error: 0.0195 - val_loss: 0.0204 - val_mean_absolute_error: 0.0900 - val_mean_squared_error: 0.0204\n",
      "Epoch 39/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0196 - mean_absolute_error: 0.0894 - mean_squared_error: 0.0196 - val_loss: 0.0195 - val_mean_absolute_error: 0.0869 - val_mean_squared_error: 0.0195\n",
      "Epoch 40/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0195 - mean_absolute_error: 0.0890 - mean_squared_error: 0.0195 - val_loss: 0.0214 - val_mean_absolute_error: 0.0998 - val_mean_squared_error: 0.0214\n",
      "Epoch 41/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0196 - mean_absolute_error: 0.0889 - mean_squared_error: 0.0196 - val_loss: 0.0193 - val_mean_absolute_error: 0.0869 - val_mean_squared_error: 0.0193\n",
      "Epoch 42/600\n",
      "3600/3600 [==============================] - 0s 76us/sample - loss: 0.0195 - mean_absolute_error: 0.0886 - mean_squared_error: 0.0195 - val_loss: 0.0220 - val_mean_absolute_error: 0.0993 - val_mean_squared_error: 0.0220\n",
      "Epoch 43/600\n",
      "3600/3600 [==============================] - 0s 72us/sample - loss: 0.0195 - mean_absolute_error: 0.0890 - mean_squared_error: 0.0195 - val_loss: 0.0210 - val_mean_absolute_error: 0.0936 - val_mean_squared_error: 0.0210\n",
      "Epoch 44/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0195 - mean_absolute_error: 0.0887 - mean_squared_error: 0.0195 - val_loss: 0.0198 - val_mean_absolute_error: 0.0897 - val_mean_squared_error: 0.0198\n",
      "Epoch 45/600\n",
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0194 - mean_absolute_error: 0.0878 - mean_squared_error: 0.0194 - val_loss: 0.0185 - val_mean_absolute_error: 0.0831 - val_mean_squared_error: 0.0185\n",
      "Epoch 46/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0194 - mean_absolute_error: 0.0885 - mean_squared_error: 0.0194 - val_loss: 0.0188 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0188\n",
      "Epoch 47/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0194 - mean_absolute_error: 0.0880 - mean_squared_error: 0.0194 - val_loss: 0.0198 - val_mean_absolute_error: 0.0899 - val_mean_squared_error: 0.0198\n",
      "Epoch 48/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0194 - mean_absolute_error: 0.0886 - mean_squared_error: 0.0194 - val_loss: 0.0186 - val_mean_absolute_error: 0.0834 - val_mean_squared_error: 0.0186\n",
      "Epoch 49/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0193 - mean_absolute_error: 0.0878 - mean_squared_error: 0.0193 - val_loss: 0.0190 - val_mean_absolute_error: 0.0851 - val_mean_squared_error: 0.0190\n",
      "Epoch 50/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0193 - mean_absolute_error: 0.0878 - mean_squared_error: 0.0193 - val_loss: 0.0197 - val_mean_absolute_error: 0.0893 - val_mean_squared_error: 0.0197\n",
      "Epoch 51/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0193 - mean_absolute_error: 0.0876 - mean_squared_error: 0.0193 - val_loss: 0.0193 - val_mean_absolute_error: 0.0864 - val_mean_squared_error: 0.0193\n",
      "Epoch 52/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0194 - mean_absolute_error: 0.0886 - mean_squared_error: 0.0194 - val_loss: 0.0195 - val_mean_absolute_error: 0.0882 - val_mean_squared_error: 0.0195\n",
      "Epoch 53/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0193 - mean_absolute_error: 0.0872 - mean_squared_error: 0.0193 - val_loss: 0.0199 - val_mean_absolute_error: 0.0909 - val_mean_squared_error: 0.0199\n",
      "Epoch 54/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0193 - mean_absolute_error: 0.0880 - mean_squared_error: 0.0193 - val_loss: 0.0194 - val_mean_absolute_error: 0.0862 - val_mean_squared_error: 0.0194\n",
      "Epoch 55/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0192 - mean_absolute_error: 0.0868 - mean_squared_error: 0.0192 - val_loss: 0.0195 - val_mean_absolute_error: 0.0843 - val_mean_squared_error: 0.0195\n",
      "Epoch 56/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0193 - mean_absolute_error: 0.0880 - mean_squared_error: 0.0193 - val_loss: 0.0192 - val_mean_absolute_error: 0.0862 - val_mean_squared_error: 0.0192\n",
      "Epoch 57/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0193 - mean_absolute_error: 0.0878 - mean_squared_error: 0.0193 - val_loss: 0.0194 - val_mean_absolute_error: 0.0868 - val_mean_squared_error: 0.0194\n",
      "Epoch 58/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0192 - mean_absolute_error: 0.0876 - mean_squared_error: 0.0192 - val_loss: 0.0185 - val_mean_absolute_error: 0.0810 - val_mean_squared_error: 0.0185\n",
      "Epoch 59/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0191 - mean_absolute_error: 0.0871 - mean_squared_error: 0.0191 - val_loss: 0.0214 - val_mean_absolute_error: 0.0976 - val_mean_squared_error: 0.0214\n",
      "Epoch 60/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0192 - mean_absolute_error: 0.0875 - mean_squared_error: 0.0192 - val_loss: 0.0187 - val_mean_absolute_error: 0.0833 - val_mean_squared_error: 0.0187\n",
      "Epoch 61/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0192 - mean_absolute_error: 0.0871 - mean_squared_error: 0.0192 - val_loss: 0.0194 - val_mean_absolute_error: 0.0885 - val_mean_squared_error: 0.0194\n",
      "Epoch 62/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0191 - mean_absolute_error: 0.0871 - mean_squared_error: 0.0191 - val_loss: 0.0186 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0186\n",
      "Epoch 63/600\n",
      "3600/3600 [==============================] - 0s 72us/sample - loss: 0.0192 - mean_absolute_error: 0.0873 - mean_squared_error: 0.0192 - val_loss: 0.0183 - val_mean_absolute_error: 0.0817 - val_mean_squared_error: 0.0183\n",
      "Epoch 64/600\n",
      "3600/3600 [==============================] - 0s 85us/sample - loss: 0.0192 - mean_absolute_error: 0.0871 - mean_squared_error: 0.0192 - val_loss: 0.0231 - val_mean_absolute_error: 0.1014 - val_mean_squared_error: 0.0231\n",
      "Epoch 65/600\n",
      "3600/3600 [==============================] - 0s 83us/sample - loss: 0.0190 - mean_absolute_error: 0.0865 - mean_squared_error: 0.0190 - val_loss: 0.0213 - val_mean_absolute_error: 0.0968 - val_mean_squared_error: 0.0213\n",
      "Epoch 66/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0191 - mean_absolute_error: 0.0869 - mean_squared_error: 0.0191 - val_loss: 0.0209 - val_mean_absolute_error: 0.0920 - val_mean_squared_error: 0.0209\n",
      "Epoch 67/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0192 - mean_absolute_error: 0.0871 - mean_squared_error: 0.0192 - val_loss: 0.0184 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0184\n",
      "Epoch 68/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0190 - mean_absolute_error: 0.0862 - mean_squared_error: 0.0190 - val_loss: 0.0197 - val_mean_absolute_error: 0.0884 - val_mean_squared_error: 0.0197\n",
      "Epoch 69/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0190 - mean_absolute_error: 0.0866 - mean_squared_error: 0.0190 - val_loss: 0.0183 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0183\n",
      "Epoch 70/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0190 - mean_absolute_error: 0.0867 - mean_squared_error: 0.0190 - val_loss: 0.0225 - val_mean_absolute_error: 0.1030 - val_mean_squared_error: 0.0225\n",
      "Epoch 71/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0189 - mean_absolute_error: 0.0860 - mean_squared_error: 0.0189 - val_loss: 0.0185 - val_mean_absolute_error: 0.0825 - val_mean_squared_error: 0.0185\n",
      "Epoch 72/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0191 - mean_absolute_error: 0.0872 - mean_squared_error: 0.0191 - val_loss: 0.0192 - val_mean_absolute_error: 0.0844 - val_mean_squared_error: 0.0192\n",
      "Epoch 73/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0190 - mean_absolute_error: 0.0863 - mean_squared_error: 0.0190 - val_loss: 0.0183 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0183\n",
      "Epoch 74/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0191 - mean_absolute_error: 0.0867 - mean_squared_error: 0.0191 - val_loss: 0.0201 - val_mean_absolute_error: 0.0890 - val_mean_squared_error: 0.0201\n",
      "Epoch 75/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0190 - mean_absolute_error: 0.0863 - mean_squared_error: 0.0190 - val_loss: 0.0193 - val_mean_absolute_error: 0.0871 - val_mean_squared_error: 0.0193\n",
      "Epoch 76/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0190 - mean_absolute_error: 0.0864 - mean_squared_error: 0.0190 - val_loss: 0.0207 - val_mean_absolute_error: 0.0939 - val_mean_squared_error: 0.0207\n",
      "Epoch 77/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0189 - mean_absolute_error: 0.0859 - mean_squared_error: 0.0189 - val_loss: 0.0195 - val_mean_absolute_error: 0.0868 - val_mean_squared_error: 0.0195\n",
      "Epoch 78/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0190 - mean_absolute_error: 0.0867 - mean_squared_error: 0.0190 - val_loss: 0.0181 - val_mean_absolute_error: 0.0795 - val_mean_squared_error: 0.0181\n",
      "Epoch 79/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0188 - mean_absolute_error: 0.0854 - mean_squared_error: 0.0188 - val_loss: 0.0184 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0184\n",
      "Epoch 80/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0190 - mean_absolute_error: 0.0859 - mean_squared_error: 0.0190 - val_loss: 0.0183 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0183\n",
      "Epoch 81/600\n",
      "3600/3600 [==============================] - 0s 62us/sample - loss: 0.0188 - mean_absolute_error: 0.0857 - mean_squared_error: 0.0188 - val_loss: 0.0197 - val_mean_absolute_error: 0.0897 - val_mean_squared_error: 0.0197\n",
      "Epoch 82/600\n",
      "3600/3600 [==============================] - 0s 65us/sample - loss: 0.0189 - mean_absolute_error: 0.0860 - mean_squared_error: 0.0189 - val_loss: 0.0195 - val_mean_absolute_error: 0.0878 - val_mean_squared_error: 0.0195\n",
      "Epoch 83/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0188 - mean_absolute_error: 0.0855 - mean_squared_error: 0.0188 - val_loss: 0.0182 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0182\n",
      "Epoch 84/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0189 - mean_absolute_error: 0.0863 - mean_squared_error: 0.0189 - val_loss: 0.0182 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0182\n",
      "Epoch 85/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0187 - mean_absolute_error: 0.0853 - mean_squared_error: 0.0187 - val_loss: 0.0183 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0183\n",
      "Epoch 86/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0188 - mean_absolute_error: 0.0854 - mean_squared_error: 0.0188 - val_loss: 0.0182 - val_mean_absolute_error: 0.0807 - val_mean_squared_error: 0.0182\n",
      "Epoch 87/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0188 - mean_absolute_error: 0.0857 - mean_squared_error: 0.0188 - val_loss: 0.0203 - val_mean_absolute_error: 0.0869 - val_mean_squared_error: 0.0203\n",
      "Epoch 88/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0188 - mean_absolute_error: 0.0854 - mean_squared_error: 0.0188 - val_loss: 0.0198 - val_mean_absolute_error: 0.0848 - val_mean_squared_error: 0.0198\n",
      "Epoch 89/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0188 - mean_absolute_error: 0.0857 - mean_squared_error: 0.0188 - val_loss: 0.0201 - val_mean_absolute_error: 0.0909 - val_mean_squared_error: 0.0201\n",
      "Epoch 90/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0189 - mean_absolute_error: 0.0861 - mean_squared_error: 0.0189 - val_loss: 0.0206 - val_mean_absolute_error: 0.0949 - val_mean_squared_error: 0.0206\n",
      "Epoch 91/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0188 - mean_absolute_error: 0.0857 - mean_squared_error: 0.0188 - val_loss: 0.0203 - val_mean_absolute_error: 0.0901 - val_mean_squared_error: 0.0203\n",
      "Epoch 92/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0187 - mean_absolute_error: 0.0848 - mean_squared_error: 0.0187 - val_loss: 0.0186 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0186\n",
      "Epoch 93/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0187 - mean_absolute_error: 0.0850 - mean_squared_error: 0.0187 - val_loss: 0.0181 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0181\n",
      "Epoch 94/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0188 - mean_absolute_error: 0.0861 - mean_squared_error: 0.0188 - val_loss: 0.0187 - val_mean_absolute_error: 0.0813 - val_mean_squared_error: 0.0187\n",
      "Epoch 95/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0188 - mean_absolute_error: 0.0856 - mean_squared_error: 0.0188 - val_loss: 0.0201 - val_mean_absolute_error: 0.0890 - val_mean_squared_error: 0.0201\n",
      "Epoch 96/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0186 - mean_absolute_error: 0.0846 - mean_squared_error: 0.0186 - val_loss: 0.0240 - val_mean_absolute_error: 0.1073 - val_mean_squared_error: 0.0240\n",
      "Epoch 97/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0187 - mean_absolute_error: 0.0849 - mean_squared_error: 0.0187 - val_loss: 0.0229 - val_mean_absolute_error: 0.1047 - val_mean_squared_error: 0.0229\n",
      "Epoch 98/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0187 - mean_absolute_error: 0.0849 - mean_squared_error: 0.0187 - val_loss: 0.0194 - val_mean_absolute_error: 0.0873 - val_mean_squared_error: 0.0194\n",
      "Epoch 99/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0186 - mean_absolute_error: 0.0843 - mean_squared_error: 0.0186 - val_loss: 0.0206 - val_mean_absolute_error: 0.0929 - val_mean_squared_error: 0.0206\n",
      "Epoch 100/600\n",
      "3600/3600 [==============================] - 0s 74us/sample - loss: 0.0187 - mean_absolute_error: 0.0854 - mean_squared_error: 0.0187 - val_loss: 0.0186 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0186\n",
      "Epoch 101/600\n",
      "3600/3600 [==============================] - 0s 76us/sample - loss: 0.0186 - mean_absolute_error: 0.0845 - mean_squared_error: 0.0186 - val_loss: 0.0187 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0187\n",
      "Epoch 102/600\n",
      "3600/3600 [==============================] - 0s 83us/sample - loss: 0.0187 - mean_absolute_error: 0.0850 - mean_squared_error: 0.0187 - val_loss: 0.0184 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0184\n",
      "Epoch 103/600\n",
      "3600/3600 [==============================] - 0s 86us/sample - loss: 0.0185 - mean_absolute_error: 0.0843 - mean_squared_error: 0.0185 - val_loss: 0.0201 - val_mean_absolute_error: 0.0909 - val_mean_squared_error: 0.0201\n",
      "Epoch 104/600\n",
      "3600/3600 [==============================] - 0s 83us/sample - loss: 0.0187 - mean_absolute_error: 0.0851 - mean_squared_error: 0.0187 - val_loss: 0.0194 - val_mean_absolute_error: 0.0880 - val_mean_squared_error: 0.0194\n",
      "Epoch 105/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0186 - mean_absolute_error: 0.0848 - mean_squared_error: 0.0186 - val_loss: 0.0189 - val_mean_absolute_error: 0.0834 - val_mean_squared_error: 0.0189\n",
      "Epoch 106/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0186 - mean_absolute_error: 0.0845 - mean_squared_error: 0.0186 - val_loss: 0.0192 - val_mean_absolute_error: 0.0861 - val_mean_squared_error: 0.0192\n",
      "Epoch 107/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0186 - mean_absolute_error: 0.0849 - mean_squared_error: 0.0186 - val_loss: 0.0190 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0190\n",
      "Epoch 108/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0186 - mean_absolute_error: 0.0847 - mean_squared_error: 0.0186 - val_loss: 0.0193 - val_mean_absolute_error: 0.0851 - val_mean_squared_error: 0.0193\n",
      "Epoch 109/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0186 - mean_absolute_error: 0.0849 - mean_squared_error: 0.0186 - val_loss: 0.0188 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0188\n",
      "Epoch 110/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0186 - mean_absolute_error: 0.0846 - mean_squared_error: 0.0186 - val_loss: 0.0185 - val_mean_absolute_error: 0.0812 - val_mean_squared_error: 0.0185\n",
      "Epoch 111/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0185 - mean_absolute_error: 0.0842 - mean_squared_error: 0.0185 - val_loss: 0.0201 - val_mean_absolute_error: 0.0918 - val_mean_squared_error: 0.0201\n",
      "Epoch 112/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0186 - mean_absolute_error: 0.0850 - mean_squared_error: 0.0186 - val_loss: 0.0196 - val_mean_absolute_error: 0.0817 - val_mean_squared_error: 0.0196\n",
      "Epoch 113/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0185 - mean_absolute_error: 0.0841 - mean_squared_error: 0.0185 - val_loss: 0.0181 - val_mean_absolute_error: 0.0789 - val_mean_squared_error: 0.0181\n",
      "Epoch 114/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0185 - mean_absolute_error: 0.0843 - mean_squared_error: 0.0185 - val_loss: 0.0197 - val_mean_absolute_error: 0.0898 - val_mean_squared_error: 0.0197\n",
      "Epoch 115/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0185 - mean_absolute_error: 0.0839 - mean_squared_error: 0.0185 - val_loss: 0.0183 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0183\n",
      "Epoch 116/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0185 - mean_absolute_error: 0.0842 - mean_squared_error: 0.0185 - val_loss: 0.0189 - val_mean_absolute_error: 0.0845 - val_mean_squared_error: 0.0189\n",
      "Epoch 117/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0184 - mean_absolute_error: 0.0841 - mean_squared_error: 0.0184 - val_loss: 0.0197 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0197\n",
      "Epoch 118/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0184 - mean_absolute_error: 0.0837 - mean_squared_error: 0.0184 - val_loss: 0.0180 - val_mean_absolute_error: 0.0784 - val_mean_squared_error: 0.0180\n",
      "Epoch 119/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0185 - mean_absolute_error: 0.0847 - mean_squared_error: 0.0185 - val_loss: 0.0184 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0184\n",
      "Epoch 120/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0185 - mean_absolute_error: 0.0840 - mean_squared_error: 0.0185 - val_loss: 0.0200 - val_mean_absolute_error: 0.0900 - val_mean_squared_error: 0.0200\n",
      "Epoch 121/600\n",
      "3600/3600 [==============================] - 0s 78us/sample - loss: 0.0184 - mean_absolute_error: 0.0839 - mean_squared_error: 0.0184 - val_loss: 0.0200 - val_mean_absolute_error: 0.0891 - val_mean_squared_error: 0.0200\n",
      "Epoch 122/600\n",
      "3600/3600 [==============================] - 0s 90us/sample - loss: 0.0184 - mean_absolute_error: 0.0834 - mean_squared_error: 0.0184 - val_loss: 0.0196 - val_mean_absolute_error: 0.0880 - val_mean_squared_error: 0.0196\n",
      "Epoch 123/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0183 - mean_absolute_error: 0.0830 - mean_squared_error: 0.0183 - val_loss: 0.0207 - val_mean_absolute_error: 0.0939 - val_mean_squared_error: 0.0207\n",
      "Epoch 124/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0184 - mean_absolute_error: 0.0844 - mean_squared_error: 0.0184 - val_loss: 0.0202 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0202\n",
      "Epoch 125/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0184 - mean_absolute_error: 0.0838 - mean_squared_error: 0.0184 - val_loss: 0.0211 - val_mean_absolute_error: 0.0981 - val_mean_squared_error: 0.0211\n",
      "Epoch 126/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0183 - mean_absolute_error: 0.0837 - mean_squared_error: 0.0183 - val_loss: 0.0224 - val_mean_absolute_error: 0.1063 - val_mean_squared_error: 0.0224\n",
      "Epoch 127/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0185 - mean_absolute_error: 0.0845 - mean_squared_error: 0.0185 - val_loss: 0.0207 - val_mean_absolute_error: 0.0931 - val_mean_squared_error: 0.0207\n",
      "Epoch 128/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0837 - mean_squared_error: 0.0183 - val_loss: 0.0184 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0184\n",
      "Epoch 129/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0837 - mean_squared_error: 0.0183 - val_loss: 0.0221 - val_mean_absolute_error: 0.0985 - val_mean_squared_error: 0.0221\n",
      "Epoch 130/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0836 - mean_squared_error: 0.0183 - val_loss: 0.0184 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0184\n",
      "Epoch 131/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0831 - mean_squared_error: 0.0183 - val_loss: 0.0182 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0182\n",
      "Epoch 132/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0184 - mean_absolute_error: 0.0839 - mean_squared_error: 0.0184 - val_loss: 0.0182 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0182\n",
      "Epoch 133/600\n",
      "3600/3600 [==============================] - 0s 57us/sample - loss: 0.0182 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0182 - val_loss: 0.0184 - val_mean_absolute_error: 0.0799 - val_mean_squared_error: 0.0184\n",
      "Epoch 134/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0832 - mean_squared_error: 0.0183 - val_loss: 0.0185 - val_mean_absolute_error: 0.0806 - val_mean_squared_error: 0.0185\n",
      "Epoch 135/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0182 - mean_absolute_error: 0.0832 - mean_squared_error: 0.0182 - val_loss: 0.0191 - val_mean_absolute_error: 0.0857 - val_mean_squared_error: 0.0191\n",
      "Epoch 136/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0183 - mean_absolute_error: 0.0832 - mean_squared_error: 0.0183 - val_loss: 0.0187 - val_mean_absolute_error: 0.0807 - val_mean_squared_error: 0.0187\n",
      "Epoch 137/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0182 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0182 - val_loss: 0.0181 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0181\n",
      "Epoch 138/600\n",
      "3600/3600 [==============================] - 0s 66us/sample - loss: 0.0181 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0181 - val_loss: 0.0212 - val_mean_absolute_error: 0.0965 - val_mean_squared_error: 0.0212\n",
      "Epoch 139/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0182 - mean_absolute_error: 0.0834 - mean_squared_error: 0.0182 - val_loss: 0.0203 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0203\n",
      "Epoch 140/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0182 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0182 - val_loss: 0.0212 - val_mean_absolute_error: 0.0960 - val_mean_squared_error: 0.0212\n",
      "Epoch 141/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0181 - mean_absolute_error: 0.0831 - mean_squared_error: 0.0181 - val_loss: 0.0184 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0184\n",
      "Epoch 142/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0182 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0182 - val_loss: 0.0190 - val_mean_absolute_error: 0.0843 - val_mean_squared_error: 0.0190\n",
      "Epoch 143/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0181 - mean_absolute_error: 0.0824 - mean_squared_error: 0.0181 - val_loss: 0.0203 - val_mean_absolute_error: 0.0914 - val_mean_squared_error: 0.0203\n",
      "Epoch 144/600\n",
      "3600/3600 [==============================] - 0s 95us/sample - loss: 0.0183 - mean_absolute_error: 0.0836 - mean_squared_error: 0.0183 - val_loss: 0.0190 - val_mean_absolute_error: 0.0834 - val_mean_squared_error: 0.0190\n",
      "Epoch 145/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0181 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0181 - val_loss: 0.0188 - val_mean_absolute_error: 0.0843 - val_mean_squared_error: 0.0188\n",
      "Epoch 146/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0182 - mean_absolute_error: 0.0828 - mean_squared_error: 0.0182 - val_loss: 0.0220 - val_mean_absolute_error: 0.1000 - val_mean_squared_error: 0.0220\n",
      "Epoch 147/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0182 - mean_absolute_error: 0.0831 - mean_squared_error: 0.0182 - val_loss: 0.0189 - val_mean_absolute_error: 0.0818 - val_mean_squared_error: 0.0189\n",
      "Epoch 148/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0181 - mean_absolute_error: 0.0823 - mean_squared_error: 0.0181 - val_loss: 0.0212 - val_mean_absolute_error: 0.0958 - val_mean_squared_error: 0.0212\n",
      "Epoch 149/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0182 - mean_absolute_error: 0.0832 - mean_squared_error: 0.0182 - val_loss: 0.0194 - val_mean_absolute_error: 0.0848 - val_mean_squared_error: 0.0194\n",
      "Epoch 150/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0181 - mean_absolute_error: 0.0830 - mean_squared_error: 0.0181 - val_loss: 0.0199 - val_mean_absolute_error: 0.0890 - val_mean_squared_error: 0.0199\n",
      "Epoch 151/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0181 - mean_absolute_error: 0.0823 - mean_squared_error: 0.0181 - val_loss: 0.0203 - val_mean_absolute_error: 0.0941 - val_mean_squared_error: 0.0203\n",
      "Epoch 152/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0181 - mean_absolute_error: 0.0827 - mean_squared_error: 0.0181 - val_loss: 0.0193 - val_mean_absolute_error: 0.0862 - val_mean_squared_error: 0.0193\n",
      "Epoch 153/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0181 - mean_absolute_error: 0.0826 - mean_squared_error: 0.0181 - val_loss: 0.0184 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0184\n",
      "Epoch 154/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0181 - mean_absolute_error: 0.0824 - mean_squared_error: 0.0181 - val_loss: 0.0186 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0186\n",
      "Epoch 155/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0180 - mean_absolute_error: 0.0825 - mean_squared_error: 0.0180 - val_loss: 0.0188 - val_mean_absolute_error: 0.0846 - val_mean_squared_error: 0.0188\n",
      "Epoch 156/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0181 - mean_absolute_error: 0.0822 - mean_squared_error: 0.0181 - val_loss: 0.0182 - val_mean_absolute_error: 0.0784 - val_mean_squared_error: 0.0182\n",
      "Epoch 157/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0181 - mean_absolute_error: 0.0827 - mean_squared_error: 0.0181 - val_loss: 0.0182 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0182\n",
      "Epoch 158/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0180 - mean_absolute_error: 0.0818 - mean_squared_error: 0.0180 - val_loss: 0.0191 - val_mean_absolute_error: 0.0859 - val_mean_squared_error: 0.0191\n",
      "Epoch 159/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0180 - mean_absolute_error: 0.0824 - mean_squared_error: 0.0180 - val_loss: 0.0203 - val_mean_absolute_error: 0.0876 - val_mean_squared_error: 0.0203\n",
      "Epoch 160/600\n",
      "3600/3600 [==============================] - 0s 65us/sample - loss: 0.0181 - mean_absolute_error: 0.0829 - mean_squared_error: 0.0181 - val_loss: 0.0195 - val_mean_absolute_error: 0.0864 - val_mean_squared_error: 0.0195\n",
      "Epoch 161/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0180 - mean_absolute_error: 0.0821 - mean_squared_error: 0.0180 - val_loss: 0.0188 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0188\n",
      "Epoch 162/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0181 - mean_absolute_error: 0.0826 - mean_squared_error: 0.0181 - val_loss: 0.0223 - val_mean_absolute_error: 0.1010 - val_mean_squared_error: 0.0223\n",
      "Epoch 163/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0180 - mean_absolute_error: 0.0823 - mean_squared_error: 0.0180 - val_loss: 0.0192 - val_mean_absolute_error: 0.0873 - val_mean_squared_error: 0.0192\n",
      "Epoch 164/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0179 - mean_absolute_error: 0.0817 - mean_squared_error: 0.0179 - val_loss: 0.0185 - val_mean_absolute_error: 0.0780 - val_mean_squared_error: 0.0185\n",
      "Epoch 165/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0179 - mean_absolute_error: 0.0817 - mean_squared_error: 0.0179 - val_loss: 0.0204 - val_mean_absolute_error: 0.0916 - val_mean_squared_error: 0.0204\n",
      "Epoch 166/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0180 - mean_absolute_error: 0.0823 - mean_squared_error: 0.0180 - val_loss: 0.0180 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0180\n",
      "Epoch 167/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0179 - mean_absolute_error: 0.0817 - mean_squared_error: 0.0179 - val_loss: 0.0185 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0185\n",
      "Epoch 168/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0180 - mean_absolute_error: 0.0821 - mean_squared_error: 0.0180 - val_loss: 0.0180 - val_mean_absolute_error: 0.0773 - val_mean_squared_error: 0.0180\n",
      "Epoch 169/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0179 - mean_absolute_error: 0.0817 - mean_squared_error: 0.0179 - val_loss: 0.0183 - val_mean_absolute_error: 0.0799 - val_mean_squared_error: 0.0183\n",
      "Epoch 170/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0179 - mean_absolute_error: 0.0816 - mean_squared_error: 0.0179 - val_loss: 0.0181 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0181\n",
      "Epoch 171/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0179 - mean_absolute_error: 0.0815 - mean_squared_error: 0.0179 - val_loss: 0.0222 - val_mean_absolute_error: 0.0969 - val_mean_squared_error: 0.0222\n",
      "Epoch 172/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0180 - mean_absolute_error: 0.0818 - mean_squared_error: 0.0180 - val_loss: 0.0201 - val_mean_absolute_error: 0.0886 - val_mean_squared_error: 0.0201\n",
      "Epoch 173/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0179 - mean_absolute_error: 0.0818 - mean_squared_error: 0.0179 - val_loss: 0.0185 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0185\n",
      "Epoch 174/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0180 - mean_absolute_error: 0.0826 - mean_squared_error: 0.0180 - val_loss: 0.0180 - val_mean_absolute_error: 0.0768 - val_mean_squared_error: 0.0180\n",
      "Epoch 175/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0178 - mean_absolute_error: 0.0815 - mean_squared_error: 0.0178 - val_loss: 0.0215 - val_mean_absolute_error: 0.0970 - val_mean_squared_error: 0.0215\n",
      "Epoch 176/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0179 - mean_absolute_error: 0.0814 - mean_squared_error: 0.0179 - val_loss: 0.0208 - val_mean_absolute_error: 0.0951 - val_mean_squared_error: 0.0208\n",
      "Epoch 177/600\n",
      "3600/3600 [==============================] - 0s 73us/sample - loss: 0.0179 - mean_absolute_error: 0.0814 - mean_squared_error: 0.0179 - val_loss: 0.0211 - val_mean_absolute_error: 0.0944 - val_mean_squared_error: 0.0211\n",
      "Epoch 178/600\n",
      "3600/3600 [==============================] - 0s 78us/sample - loss: 0.0178 - mean_absolute_error: 0.0813 - mean_squared_error: 0.0178 - val_loss: 0.0182 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0182\n",
      "Epoch 179/600\n",
      "3600/3600 [==============================] - 0s 84us/sample - loss: 0.0179 - mean_absolute_error: 0.0820 - mean_squared_error: 0.0179 - val_loss: 0.0193 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0193\n",
      "Epoch 180/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0179 - mean_absolute_error: 0.0817 - mean_squared_error: 0.0179 - val_loss: 0.0191 - val_mean_absolute_error: 0.0820 - val_mean_squared_error: 0.0191\n",
      "Epoch 181/600\n",
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0177 - mean_absolute_error: 0.0809 - mean_squared_error: 0.0177 - val_loss: 0.0189 - val_mean_absolute_error: 0.0823 - val_mean_squared_error: 0.0189\n",
      "Epoch 182/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0178 - mean_absolute_error: 0.0810 - mean_squared_error: 0.0178 - val_loss: 0.0187 - val_mean_absolute_error: 0.0818 - val_mean_squared_error: 0.0187\n",
      "Epoch 183/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0177 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0177 - val_loss: 0.0193 - val_mean_absolute_error: 0.0824 - val_mean_squared_error: 0.0193\n",
      "Epoch 184/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0179 - mean_absolute_error: 0.0815 - mean_squared_error: 0.0179 - val_loss: 0.0187 - val_mean_absolute_error: 0.0816 - val_mean_squared_error: 0.0187\n",
      "Epoch 185/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0177 - mean_absolute_error: 0.0809 - mean_squared_error: 0.0177 - val_loss: 0.0195 - val_mean_absolute_error: 0.0869 - val_mean_squared_error: 0.0195\n",
      "Epoch 186/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0178 - mean_absolute_error: 0.0815 - mean_squared_error: 0.0178 - val_loss: 0.0189 - val_mean_absolute_error: 0.0826 - val_mean_squared_error: 0.0189\n",
      "Epoch 187/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0178 - mean_absolute_error: 0.0815 - mean_squared_error: 0.0178 - val_loss: 0.0185 - val_mean_absolute_error: 0.0808 - val_mean_squared_error: 0.0185\n",
      "Epoch 188/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0178 - mean_absolute_error: 0.0810 - mean_squared_error: 0.0178 - val_loss: 0.0192 - val_mean_absolute_error: 0.0854 - val_mean_squared_error: 0.0192\n",
      "Epoch 189/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0177 - mean_absolute_error: 0.0812 - mean_squared_error: 0.0177 - val_loss: 0.0192 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0192\n",
      "Epoch 190/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0178 - mean_absolute_error: 0.0814 - mean_squared_error: 0.0178 - val_loss: 0.0189 - val_mean_absolute_error: 0.0820 - val_mean_squared_error: 0.0189\n",
      "Epoch 191/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0177 - mean_absolute_error: 0.0805 - mean_squared_error: 0.0177 - val_loss: 0.0193 - val_mean_absolute_error: 0.0832 - val_mean_squared_error: 0.0193\n",
      "Epoch 192/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0177 - mean_absolute_error: 0.0807 - mean_squared_error: 0.0177 - val_loss: 0.0202 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0202\n",
      "Epoch 193/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0177 - mean_absolute_error: 0.0809 - mean_squared_error: 0.0177 - val_loss: 0.0199 - val_mean_absolute_error: 0.0894 - val_mean_squared_error: 0.0199\n",
      "Epoch 194/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0178 - mean_absolute_error: 0.0811 - mean_squared_error: 0.0178 - val_loss: 0.0189 - val_mean_absolute_error: 0.0819 - val_mean_squared_error: 0.0189\n",
      "Epoch 195/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0177 - mean_absolute_error: 0.0812 - mean_squared_error: 0.0177 - val_loss: 0.0188 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0188\n",
      "Epoch 196/600\n",
      "3600/3600 [==============================] - 0s 57us/sample - loss: 0.0177 - mean_absolute_error: 0.0811 - mean_squared_error: 0.0177 - val_loss: 0.0181 - val_mean_absolute_error: 0.0775 - val_mean_squared_error: 0.0181\n",
      "Epoch 197/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0178 - mean_absolute_error: 0.0809 - mean_squared_error: 0.0178 - val_loss: 0.0210 - val_mean_absolute_error: 0.0934 - val_mean_squared_error: 0.0210\n",
      "Epoch 198/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0177 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0177 - val_loss: 0.0187 - val_mean_absolute_error: 0.0807 - val_mean_squared_error: 0.0187\n",
      "Epoch 199/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0177 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0177 - val_loss: 0.0191 - val_mean_absolute_error: 0.0845 - val_mean_squared_error: 0.0191\n",
      "Epoch 200/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0176 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0176 - val_loss: 0.0214 - val_mean_absolute_error: 0.0950 - val_mean_squared_error: 0.0214\n",
      "Epoch 201/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0177 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0177 - val_loss: 0.0191 - val_mean_absolute_error: 0.0831 - val_mean_squared_error: 0.0191\n",
      "Epoch 202/600\n",
      "3600/3600 [==============================] - 0s 92us/sample - loss: 0.0176 - mean_absolute_error: 0.0802 - mean_squared_error: 0.0176 - val_loss: 0.0195 - val_mean_absolute_error: 0.0885 - val_mean_squared_error: 0.0195\n",
      "Epoch 203/600\n",
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0176 - mean_absolute_error: 0.0801 - mean_squared_error: 0.0176 - val_loss: 0.0191 - val_mean_absolute_error: 0.0844 - val_mean_squared_error: 0.0191\n",
      "Epoch 204/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0176 - mean_absolute_error: 0.0807 - mean_squared_error: 0.0176 - val_loss: 0.0189 - val_mean_absolute_error: 0.0825 - val_mean_squared_error: 0.0189\n",
      "Epoch 205/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0176 - mean_absolute_error: 0.0807 - mean_squared_error: 0.0176 - val_loss: 0.0183 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0183\n",
      "Epoch 206/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0176 - mean_absolute_error: 0.0803 - mean_squared_error: 0.0176 - val_loss: 0.0190 - val_mean_absolute_error: 0.0832 - val_mean_squared_error: 0.0190\n",
      "Epoch 207/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0177 - mean_absolute_error: 0.0807 - mean_squared_error: 0.0177 - val_loss: 0.0187 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0187\n",
      "Epoch 208/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0176 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0176 - val_loss: 0.0185 - val_mean_absolute_error: 0.0820 - val_mean_squared_error: 0.0185\n",
      "Epoch 209/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0175 - mean_absolute_error: 0.0805 - mean_squared_error: 0.0175 - val_loss: 0.0187 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0187\n",
      "Epoch 210/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0175 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0175 - val_loss: 0.0191 - val_mean_absolute_error: 0.0844 - val_mean_squared_error: 0.0191\n",
      "Epoch 211/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0175 - mean_absolute_error: 0.0802 - mean_squared_error: 0.0175 - val_loss: 0.0199 - val_mean_absolute_error: 0.0873 - val_mean_squared_error: 0.0199\n",
      "Epoch 212/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0176 - mean_absolute_error: 0.0802 - mean_squared_error: 0.0176 - val_loss: 0.0191 - val_mean_absolute_error: 0.0828 - val_mean_squared_error: 0.0191\n",
      "Epoch 213/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0175 - mean_absolute_error: 0.0804 - mean_squared_error: 0.0175 - val_loss: 0.0185 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0185\n",
      "Epoch 214/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0174 - mean_absolute_error: 0.0795 - mean_squared_error: 0.0174 - val_loss: 0.0218 - val_mean_absolute_error: 0.0979 - val_mean_squared_error: 0.0218\n",
      "Epoch 215/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0176 - mean_absolute_error: 0.0806 - mean_squared_error: 0.0176 - val_loss: 0.0189 - val_mean_absolute_error: 0.0823 - val_mean_squared_error: 0.0189\n",
      "Epoch 216/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0175 - mean_absolute_error: 0.0801 - mean_squared_error: 0.0175 - val_loss: 0.0182 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0182\n",
      "Epoch 217/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0174 - mean_absolute_error: 0.0795 - mean_squared_error: 0.0174 - val_loss: 0.0212 - val_mean_absolute_error: 0.0954 - val_mean_squared_error: 0.0212\n",
      "Epoch 218/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0176 - mean_absolute_error: 0.0809 - mean_squared_error: 0.0176 - val_loss: 0.0183 - val_mean_absolute_error: 0.0789 - val_mean_squared_error: 0.0183\n",
      "Epoch 219/600\n",
      "3600/3600 [==============================] - 0s 66us/sample - loss: 0.0174 - mean_absolute_error: 0.0797 - mean_squared_error: 0.0174 - val_loss: 0.0194 - val_mean_absolute_error: 0.0854 - val_mean_squared_error: 0.0194\n",
      "Epoch 220/600\n",
      "3600/3600 [==============================] - 0s 74us/sample - loss: 0.0175 - mean_absolute_error: 0.0799 - mean_squared_error: 0.0175 - val_loss: 0.0181 - val_mean_absolute_error: 0.0775 - val_mean_squared_error: 0.0181\n",
      "Epoch 221/600\n",
      "3600/3600 [==============================] - 0s 86us/sample - loss: 0.0175 - mean_absolute_error: 0.0799 - mean_squared_error: 0.0175 - val_loss: 0.0185 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0185\n",
      "Epoch 222/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0175 - mean_absolute_error: 0.0800 - mean_squared_error: 0.0175 - val_loss: 0.0217 - val_mean_absolute_error: 0.1002 - val_mean_squared_error: 0.0217\n",
      "Epoch 223/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0175 - mean_absolute_error: 0.0797 - mean_squared_error: 0.0175 - val_loss: 0.0185 - val_mean_absolute_error: 0.0786 - val_mean_squared_error: 0.0185\n",
      "Epoch 224/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0175 - mean_absolute_error: 0.0798 - mean_squared_error: 0.0175 - val_loss: 0.0198 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0198\n",
      "Epoch 225/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0173 - mean_absolute_error: 0.0792 - mean_squared_error: 0.0173 - val_loss: 0.0186 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0186\n",
      "Epoch 226/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0176 - mean_absolute_error: 0.0804 - mean_squared_error: 0.0176 - val_loss: 0.0182 - val_mean_absolute_error: 0.0772 - val_mean_squared_error: 0.0182\n",
      "Epoch 227/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0174 - mean_absolute_error: 0.0801 - mean_squared_error: 0.0174 - val_loss: 0.0185 - val_mean_absolute_error: 0.0785 - val_mean_squared_error: 0.0185\n",
      "Epoch 228/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0175 - mean_absolute_error: 0.0803 - mean_squared_error: 0.0175 - val_loss: 0.0199 - val_mean_absolute_error: 0.0880 - val_mean_squared_error: 0.0199\n",
      "Epoch 229/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0174 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0174 - val_loss: 0.0222 - val_mean_absolute_error: 0.0994 - val_mean_squared_error: 0.0222\n",
      "Epoch 230/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0174 - mean_absolute_error: 0.0796 - mean_squared_error: 0.0174 - val_loss: 0.0197 - val_mean_absolute_error: 0.0863 - val_mean_squared_error: 0.0197\n",
      "Epoch 231/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0174 - mean_absolute_error: 0.0798 - mean_squared_error: 0.0174 - val_loss: 0.0188 - val_mean_absolute_error: 0.0823 - val_mean_squared_error: 0.0188\n",
      "Epoch 232/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0173 - mean_absolute_error: 0.0796 - mean_squared_error: 0.0173 - val_loss: 0.0193 - val_mean_absolute_error: 0.0841 - val_mean_squared_error: 0.0193\n",
      "Epoch 233/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0174 - mean_absolute_error: 0.0797 - mean_squared_error: 0.0174 - val_loss: 0.0181 - val_mean_absolute_error: 0.0770 - val_mean_squared_error: 0.0181\n",
      "Epoch 234/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0172 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0172 - val_loss: 0.0191 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0191\n",
      "Epoch 235/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0174 - mean_absolute_error: 0.0798 - mean_squared_error: 0.0174 - val_loss: 0.0181 - val_mean_absolute_error: 0.0764 - val_mean_squared_error: 0.0181\n",
      "Epoch 236/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0174 - mean_absolute_error: 0.0797 - mean_squared_error: 0.0174 - val_loss: 0.0193 - val_mean_absolute_error: 0.0859 - val_mean_squared_error: 0.0193\n",
      "Epoch 237/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0173 - mean_absolute_error: 0.0798 - mean_squared_error: 0.0173 - val_loss: 0.0183 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0183\n",
      "Epoch 238/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0174 - mean_absolute_error: 0.0799 - mean_squared_error: 0.0174 - val_loss: 0.0182 - val_mean_absolute_error: 0.0773 - val_mean_squared_error: 0.0182\n",
      "Epoch 239/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0173 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0173 - val_loss: 0.0185 - val_mean_absolute_error: 0.0784 - val_mean_squared_error: 0.0185\n",
      "Epoch 240/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0173 - mean_absolute_error: 0.0792 - mean_squared_error: 0.0173 - val_loss: 0.0196 - val_mean_absolute_error: 0.0819 - val_mean_squared_error: 0.0196\n",
      "Epoch 241/600\n",
      "3600/3600 [==============================] - 0s 96us/sample - loss: 0.0174 - mean_absolute_error: 0.0796 - mean_squared_error: 0.0174 - val_loss: 0.0185 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0185\n",
      "Epoch 242/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0173 - mean_absolute_error: 0.0790 - mean_squared_error: 0.0173 - val_loss: 0.0232 - val_mean_absolute_error: 0.1026 - val_mean_squared_error: 0.0232\n",
      "Epoch 243/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0173 - mean_absolute_error: 0.0794 - mean_squared_error: 0.0173 - val_loss: 0.0184 - val_mean_absolute_error: 0.0786 - val_mean_squared_error: 0.0184\n",
      "Epoch 244/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0172 - mean_absolute_error: 0.0791 - mean_squared_error: 0.0172 - val_loss: 0.0189 - val_mean_absolute_error: 0.0824 - val_mean_squared_error: 0.0189\n",
      "Epoch 245/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0172 - mean_absolute_error: 0.0789 - mean_squared_error: 0.0172 - val_loss: 0.0188 - val_mean_absolute_error: 0.0819 - val_mean_squared_error: 0.0188\n",
      "Epoch 246/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0174 - mean_absolute_error: 0.0798 - mean_squared_error: 0.0174 - val_loss: 0.0212 - val_mean_absolute_error: 0.0964 - val_mean_squared_error: 0.0212\n",
      "Epoch 247/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0173 - mean_absolute_error: 0.0791 - mean_squared_error: 0.0173 - val_loss: 0.0188 - val_mean_absolute_error: 0.0814 - val_mean_squared_error: 0.0188\n",
      "Epoch 248/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0172 - mean_absolute_error: 0.0788 - mean_squared_error: 0.0172 - val_loss: 0.0182 - val_mean_absolute_error: 0.0772 - val_mean_squared_error: 0.0182\n",
      "Epoch 249/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0172 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0172 - val_loss: 0.0186 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0186\n",
      "Epoch 250/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0172 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0172 - val_loss: 0.0207 - val_mean_absolute_error: 0.0908 - val_mean_squared_error: 0.0207\n",
      "Epoch 251/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0173 - mean_absolute_error: 0.0795 - mean_squared_error: 0.0173 - val_loss: 0.0193 - val_mean_absolute_error: 0.0810 - val_mean_squared_error: 0.0193\n",
      "Epoch 252/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0172 - mean_absolute_error: 0.0792 - mean_squared_error: 0.0172 - val_loss: 0.0186 - val_mean_absolute_error: 0.0793 - val_mean_squared_error: 0.0186\n",
      "Epoch 253/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0172 - mean_absolute_error: 0.0787 - mean_squared_error: 0.0172 - val_loss: 0.0199 - val_mean_absolute_error: 0.0838 - val_mean_squared_error: 0.0199\n",
      "Epoch 254/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0172 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0172 - val_loss: 0.0183 - val_mean_absolute_error: 0.0783 - val_mean_squared_error: 0.0183\n",
      "Epoch 255/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0171 - mean_absolute_error: 0.0784 - mean_squared_error: 0.0171 - val_loss: 0.0194 - val_mean_absolute_error: 0.0853 - val_mean_squared_error: 0.0194\n",
      "Epoch 256/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0172 - mean_absolute_error: 0.0789 - mean_squared_error: 0.0172 - val_loss: 0.0190 - val_mean_absolute_error: 0.0814 - val_mean_squared_error: 0.0190\n",
      "Epoch 257/600\n",
      "3600/3600 [==============================] - 0s 82us/sample - loss: 0.0171 - mean_absolute_error: 0.0788 - mean_squared_error: 0.0171 - val_loss: 0.0193 - val_mean_absolute_error: 0.0833 - val_mean_squared_error: 0.0193\n",
      "Epoch 258/600\n",
      "3600/3600 [==============================] - 0s 82us/sample - loss: 0.0172 - mean_absolute_error: 0.0793 - mean_squared_error: 0.0172 - val_loss: 0.0183 - val_mean_absolute_error: 0.0775 - val_mean_squared_error: 0.0183\n",
      "Epoch 259/600\n",
      "3600/3600 [==============================] - 0s 88us/sample - loss: 0.0171 - mean_absolute_error: 0.0788 - mean_squared_error: 0.0171 - val_loss: 0.0187 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0187\n",
      "Epoch 260/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0171 - mean_absolute_error: 0.0786 - mean_squared_error: 0.0171 - val_loss: 0.0196 - val_mean_absolute_error: 0.0862 - val_mean_squared_error: 0.0196\n",
      "Epoch 261/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0172 - mean_absolute_error: 0.0790 - mean_squared_error: 0.0172 - val_loss: 0.0193 - val_mean_absolute_error: 0.0842 - val_mean_squared_error: 0.0193\n",
      "Epoch 262/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0172 - mean_absolute_error: 0.0791 - mean_squared_error: 0.0172 - val_loss: 0.0187 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0187\n",
      "Epoch 263/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0171 - mean_absolute_error: 0.0787 - mean_squared_error: 0.0171 - val_loss: 0.0196 - val_mean_absolute_error: 0.0833 - val_mean_squared_error: 0.0196\n",
      "Epoch 264/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0170 - mean_absolute_error: 0.0784 - mean_squared_error: 0.0170 - val_loss: 0.0189 - val_mean_absolute_error: 0.0806 - val_mean_squared_error: 0.0189\n",
      "Epoch 265/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0172 - mean_absolute_error: 0.0784 - mean_squared_error: 0.0172 - val_loss: 0.0188 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0188\n",
      "Epoch 266/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0172 - mean_absolute_error: 0.0789 - mean_squared_error: 0.0172 - val_loss: 0.0193 - val_mean_absolute_error: 0.0834 - val_mean_squared_error: 0.0193\n",
      "Epoch 267/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0171 - mean_absolute_error: 0.0787 - mean_squared_error: 0.0171 - val_loss: 0.0201 - val_mean_absolute_error: 0.0906 - val_mean_squared_error: 0.0201\n",
      "Epoch 268/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0171 - mean_absolute_error: 0.0789 - mean_squared_error: 0.0171 - val_loss: 0.0186 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0186\n",
      "Epoch 269/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0171 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0171 - val_loss: 0.0213 - val_mean_absolute_error: 0.0939 - val_mean_squared_error: 0.0213\n",
      "Epoch 270/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0171 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0171 - val_loss: 0.0192 - val_mean_absolute_error: 0.0813 - val_mean_squared_error: 0.0192\n",
      "Epoch 271/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0171 - mean_absolute_error: 0.0786 - mean_squared_error: 0.0171 - val_loss: 0.0194 - val_mean_absolute_error: 0.0849 - val_mean_squared_error: 0.0194\n",
      "Epoch 272/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0170 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0170 - val_loss: 0.0187 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0187\n",
      "Epoch 273/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0170 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0170 - val_loss: 0.0215 - val_mean_absolute_error: 0.0933 - val_mean_squared_error: 0.0215\n",
      "Epoch 274/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0171 - mean_absolute_error: 0.0789 - mean_squared_error: 0.0171 - val_loss: 0.0197 - val_mean_absolute_error: 0.0856 - val_mean_squared_error: 0.0197\n",
      "Epoch 275/600\n",
      "3600/3600 [==============================] - 0s 66us/sample - loss: 0.0171 - mean_absolute_error: 0.0791 - mean_squared_error: 0.0171 - val_loss: 0.0194 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0194\n",
      "Epoch 276/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0170 - mean_absolute_error: 0.0781 - mean_squared_error: 0.0170 - val_loss: 0.0183 - val_mean_absolute_error: 0.0764 - val_mean_squared_error: 0.0183\n",
      "Epoch 277/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0170 - mean_absolute_error: 0.0781 - mean_squared_error: 0.0170 - val_loss: 0.0185 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0185\n",
      "Epoch 278/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0170 - mean_absolute_error: 0.0779 - mean_squared_error: 0.0170 - val_loss: 0.0186 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0186\n",
      "Epoch 279/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0169 - mean_absolute_error: 0.0780 - mean_squared_error: 0.0169 - val_loss: 0.0182 - val_mean_absolute_error: 0.0766 - val_mean_squared_error: 0.0182\n",
      "Epoch 280/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0170 - mean_absolute_error: 0.0778 - mean_squared_error: 0.0170 - val_loss: 0.0204 - val_mean_absolute_error: 0.0889 - val_mean_squared_error: 0.0204\n",
      "Epoch 281/600\n",
      "3600/3600 [==============================] - 0s 62us/sample - loss: 0.0170 - mean_absolute_error: 0.0779 - mean_squared_error: 0.0170 - val_loss: 0.0185 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0185\n",
      "Epoch 282/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0170 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0170 - val_loss: 0.0185 - val_mean_absolute_error: 0.0796 - val_mean_squared_error: 0.0185\n",
      "Epoch 283/600\n",
      "3600/3600 [==============================] - 0s 58us/sample - loss: 0.0170 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0170 - val_loss: 0.0187 - val_mean_absolute_error: 0.0816 - val_mean_squared_error: 0.0187\n",
      "Epoch 284/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0170 - mean_absolute_error: 0.0784 - mean_squared_error: 0.0170 - val_loss: 0.0187 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0187\n",
      "Epoch 285/600\n",
      "3600/3600 [==============================] - 0s 57us/sample - loss: 0.0169 - mean_absolute_error: 0.0780 - mean_squared_error: 0.0169 - val_loss: 0.0183 - val_mean_absolute_error: 0.0767 - val_mean_squared_error: 0.0183\n",
      "Epoch 286/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0170 - mean_absolute_error: 0.0782 - mean_squared_error: 0.0170 - val_loss: 0.0207 - val_mean_absolute_error: 0.0929 - val_mean_squared_error: 0.0207\n",
      "Epoch 287/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0170 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0170 - val_loss: 0.0184 - val_mean_absolute_error: 0.0766 - val_mean_squared_error: 0.0184\n",
      "Epoch 288/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0170 - mean_absolute_error: 0.0782 - mean_squared_error: 0.0170 - val_loss: 0.0185 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0185\n",
      "Epoch 289/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0169 - mean_absolute_error: 0.0780 - mean_squared_error: 0.0169 - val_loss: 0.0185 - val_mean_absolute_error: 0.0784 - val_mean_squared_error: 0.0185\n",
      "Epoch 290/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0169 - mean_absolute_error: 0.0779 - mean_squared_error: 0.0169 - val_loss: 0.0202 - val_mean_absolute_error: 0.0886 - val_mean_squared_error: 0.0202\n",
      "Epoch 291/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0169 - mean_absolute_error: 0.0777 - mean_squared_error: 0.0169 - val_loss: 0.0194 - val_mean_absolute_error: 0.0825 - val_mean_squared_error: 0.0194\n",
      "Epoch 292/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0168 - mean_absolute_error: 0.0776 - mean_squared_error: 0.0168 - val_loss: 0.0194 - val_mean_absolute_error: 0.0841 - val_mean_squared_error: 0.0194\n",
      "Epoch 293/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0170 - mean_absolute_error: 0.0782 - mean_squared_error: 0.0170 - val_loss: 0.0187 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0187\n",
      "Epoch 294/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0169 - mean_absolute_error: 0.0783 - mean_squared_error: 0.0169 - val_loss: 0.0191 - val_mean_absolute_error: 0.0843 - val_mean_squared_error: 0.0191\n",
      "Epoch 295/600\n",
      "3600/3600 [==============================] - 0s 88us/sample - loss: 0.0168 - mean_absolute_error: 0.0781 - mean_squared_error: 0.0168 - val_loss: 0.0196 - val_mean_absolute_error: 0.0855 - val_mean_squared_error: 0.0196\n",
      "Epoch 296/600\n",
      "3600/3600 [==============================] - 0s 85us/sample - loss: 0.0168 - mean_absolute_error: 0.0775 - mean_squared_error: 0.0168 - val_loss: 0.0228 - val_mean_absolute_error: 0.0976 - val_mean_squared_error: 0.0228\n",
      "Epoch 297/600\n",
      "3600/3600 [==============================] - 0s 84us/sample - loss: 0.0169 - mean_absolute_error: 0.0779 - mean_squared_error: 0.0169 - val_loss: 0.0206 - val_mean_absolute_error: 0.0902 - val_mean_squared_error: 0.0206\n",
      "Epoch 298/600\n",
      "3600/3600 [==============================] - 0s 93us/sample - loss: 0.0169 - mean_absolute_error: 0.0781 - mean_squared_error: 0.0169 - val_loss: 0.0187 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0187\n",
      "Epoch 299/600\n",
      "3600/3600 [==============================] - 0s 86us/sample - loss: 0.0168 - mean_absolute_error: 0.0775 - mean_squared_error: 0.0168 - val_loss: 0.0185 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0185\n",
      "Epoch 300/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0168 - mean_absolute_error: 0.0771 - mean_squared_error: 0.0168 - val_loss: 0.0186 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0186\n",
      "Epoch 301/600\n",
      "3600/3600 [==============================] - 0s 57us/sample - loss: 0.0168 - mean_absolute_error: 0.0776 - mean_squared_error: 0.0168 - val_loss: 0.0190 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0190\n",
      "Epoch 302/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0168 - mean_absolute_error: 0.0776 - mean_squared_error: 0.0168 - val_loss: 0.0197 - val_mean_absolute_error: 0.0849 - val_mean_squared_error: 0.0197\n",
      "Epoch 303/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0168 - mean_absolute_error: 0.0777 - mean_squared_error: 0.0168 - val_loss: 0.0203 - val_mean_absolute_error: 0.0831 - val_mean_squared_error: 0.0203\n",
      "Epoch 304/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0168 - mean_absolute_error: 0.0776 - mean_squared_error: 0.0168 - val_loss: 0.0191 - val_mean_absolute_error: 0.0828 - val_mean_squared_error: 0.0191\n",
      "Epoch 305/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0168 - mean_absolute_error: 0.0777 - mean_squared_error: 0.0168 - val_loss: 0.0194 - val_mean_absolute_error: 0.0846 - val_mean_squared_error: 0.0194\n",
      "Epoch 306/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0168 - mean_absolute_error: 0.0773 - mean_squared_error: 0.0168 - val_loss: 0.0219 - val_mean_absolute_error: 0.0947 - val_mean_squared_error: 0.0219\n",
      "Epoch 307/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0167 - mean_absolute_error: 0.0774 - mean_squared_error: 0.0167 - val_loss: 0.0189 - val_mean_absolute_error: 0.0813 - val_mean_squared_error: 0.0189\n",
      "Epoch 308/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0167 - mean_absolute_error: 0.0769 - mean_squared_error: 0.0167 - val_loss: 0.0200 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0200\n",
      "Epoch 309/600\n",
      "3600/3600 [==============================] - 0s 62us/sample - loss: 0.0167 - mean_absolute_error: 0.0769 - mean_squared_error: 0.0167 - val_loss: 0.0191 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0191\n",
      "Epoch 310/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0168 - mean_absolute_error: 0.0773 - mean_squared_error: 0.0168 - val_loss: 0.0189 - val_mean_absolute_error: 0.0771 - val_mean_squared_error: 0.0189\n",
      "Epoch 311/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0168 - mean_absolute_error: 0.0777 - mean_squared_error: 0.0168 - val_loss: 0.0186 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0186\n",
      "Epoch 312/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0168 - mean_absolute_error: 0.0772 - mean_squared_error: 0.0168 - val_loss: 0.0195 - val_mean_absolute_error: 0.0835 - val_mean_squared_error: 0.0195\n",
      "Epoch 313/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0167 - mean_absolute_error: 0.0772 - mean_squared_error: 0.0167 - val_loss: 0.0190 - val_mean_absolute_error: 0.0799 - val_mean_squared_error: 0.0190\n",
      "Epoch 314/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0167 - mean_absolute_error: 0.0773 - mean_squared_error: 0.0167 - val_loss: 0.0186 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0186\n",
      "Epoch 315/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0167 - mean_absolute_error: 0.0771 - mean_squared_error: 0.0167 - val_loss: 0.0206 - val_mean_absolute_error: 0.0910 - val_mean_squared_error: 0.0206\n",
      "Epoch 316/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0166 - mean_absolute_error: 0.0769 - mean_squared_error: 0.0166 - val_loss: 0.0185 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0185\n",
      "Epoch 317/600\n",
      "3600/3600 [==============================] - 0s 73us/sample - loss: 0.0166 - mean_absolute_error: 0.0768 - mean_squared_error: 0.0166 - val_loss: 0.0185 - val_mean_absolute_error: 0.0773 - val_mean_squared_error: 0.0185\n",
      "Epoch 318/600\n",
      "3600/3600 [==============================] - 0s 89us/sample - loss: 0.0167 - mean_absolute_error: 0.0774 - mean_squared_error: 0.0167 - val_loss: 0.0209 - val_mean_absolute_error: 0.0902 - val_mean_squared_error: 0.0209\n",
      "Epoch 319/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0167 - mean_absolute_error: 0.0773 - mean_squared_error: 0.0167 - val_loss: 0.0188 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0188\n",
      "Epoch 320/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0166 - mean_absolute_error: 0.0773 - mean_squared_error: 0.0166 - val_loss: 0.0188 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0188\n",
      "Epoch 321/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0167 - mean_absolute_error: 0.0770 - mean_squared_error: 0.0167 - val_loss: 0.0189 - val_mean_absolute_error: 0.0808 - val_mean_squared_error: 0.0189\n",
      "Epoch 322/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0167 - mean_absolute_error: 0.0774 - mean_squared_error: 0.0167 - val_loss: 0.0184 - val_mean_absolute_error: 0.0780 - val_mean_squared_error: 0.0184\n",
      "Epoch 323/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0166 - mean_absolute_error: 0.0772 - mean_squared_error: 0.0166 - val_loss: 0.0214 - val_mean_absolute_error: 0.0948 - val_mean_squared_error: 0.0214\n",
      "Epoch 324/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0166 - mean_absolute_error: 0.0765 - mean_squared_error: 0.0166 - val_loss: 0.0198 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0198\n",
      "Epoch 325/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0165 - mean_absolute_error: 0.0764 - mean_squared_error: 0.0165 - val_loss: 0.0198 - val_mean_absolute_error: 0.0806 - val_mean_squared_error: 0.0198\n",
      "Epoch 326/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0167 - mean_absolute_error: 0.0772 - mean_squared_error: 0.0167 - val_loss: 0.0196 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0196\n",
      "Epoch 327/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0167 - mean_absolute_error: 0.0772 - mean_squared_error: 0.0167 - val_loss: 0.0189 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0189\n",
      "Epoch 328/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0166 - mean_absolute_error: 0.0765 - mean_squared_error: 0.0166 - val_loss: 0.0195 - val_mean_absolute_error: 0.0824 - val_mean_squared_error: 0.0195\n",
      "Epoch 329/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0166 - mean_absolute_error: 0.0770 - mean_squared_error: 0.0166 - val_loss: 0.0205 - val_mean_absolute_error: 0.0891 - val_mean_squared_error: 0.0205\n",
      "Epoch 330/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0166 - mean_absolute_error: 0.0769 - mean_squared_error: 0.0166 - val_loss: 0.0186 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0186\n",
      "Epoch 331/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0166 - mean_absolute_error: 0.0770 - mean_squared_error: 0.0166 - val_loss: 0.0186 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0186\n",
      "Epoch 332/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0165 - mean_absolute_error: 0.0763 - mean_squared_error: 0.0165 - val_loss: 0.0193 - val_mean_absolute_error: 0.0841 - val_mean_squared_error: 0.0193\n",
      "Epoch 333/600\n",
      "3600/3600 [==============================] - 0s 88us/sample - loss: 0.0166 - mean_absolute_error: 0.0769 - mean_squared_error: 0.0166 - val_loss: 0.0193 - val_mean_absolute_error: 0.0816 - val_mean_squared_error: 0.0193\n",
      "Epoch 334/600\n",
      "3600/3600 [==============================] - 0s 89us/sample - loss: 0.0165 - mean_absolute_error: 0.0768 - mean_squared_error: 0.0165 - val_loss: 0.0202 - val_mean_absolute_error: 0.0896 - val_mean_squared_error: 0.0202\n",
      "Epoch 335/600\n",
      "3600/3600 [==============================] - 0s 89us/sample - loss: 0.0166 - mean_absolute_error: 0.0767 - mean_squared_error: 0.0166 - val_loss: 0.0206 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0206\n",
      "Epoch 336/600\n",
      "3600/3600 [==============================] - 0s 91us/sample - loss: 0.0165 - mean_absolute_error: 0.0762 - mean_squared_error: 0.0165 - val_loss: 0.0187 - val_mean_absolute_error: 0.0780 - val_mean_squared_error: 0.0187\n",
      "Epoch 337/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0166 - mean_absolute_error: 0.0766 - mean_squared_error: 0.0166 - val_loss: 0.0185 - val_mean_absolute_error: 0.0770 - val_mean_squared_error: 0.0185\n",
      "Epoch 338/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0165 - mean_absolute_error: 0.0766 - mean_squared_error: 0.0165 - val_loss: 0.0186 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0186\n",
      "Epoch 339/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0165 - mean_absolute_error: 0.0765 - mean_squared_error: 0.0165 - val_loss: 0.0193 - val_mean_absolute_error: 0.0823 - val_mean_squared_error: 0.0193\n",
      "Epoch 340/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0165 - mean_absolute_error: 0.0766 - mean_squared_error: 0.0165 - val_loss: 0.0192 - val_mean_absolute_error: 0.0802 - val_mean_squared_error: 0.0192\n",
      "Epoch 341/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0166 - mean_absolute_error: 0.0771 - mean_squared_error: 0.0166 - val_loss: 0.0187 - val_mean_absolute_error: 0.0780 - val_mean_squared_error: 0.0187\n",
      "Epoch 342/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0165 - mean_absolute_error: 0.0763 - mean_squared_error: 0.0165 - val_loss: 0.0184 - val_mean_absolute_error: 0.0762 - val_mean_squared_error: 0.0184\n",
      "Epoch 343/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0164 - mean_absolute_error: 0.0763 - mean_squared_error: 0.0164 - val_loss: 0.0185 - val_mean_absolute_error: 0.0763 - val_mean_squared_error: 0.0185\n",
      "Epoch 344/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0164 - mean_absolute_error: 0.0760 - mean_squared_error: 0.0164 - val_loss: 0.0206 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0206\n",
      "Epoch 345/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0165 - mean_absolute_error: 0.0768 - mean_squared_error: 0.0165 - val_loss: 0.0193 - val_mean_absolute_error: 0.0793 - val_mean_squared_error: 0.0193\n",
      "Epoch 346/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0164 - mean_absolute_error: 0.0764 - mean_squared_error: 0.0164 - val_loss: 0.0187 - val_mean_absolute_error: 0.0783 - val_mean_squared_error: 0.0187\n",
      "Epoch 347/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0165 - mean_absolute_error: 0.0766 - mean_squared_error: 0.0165 - val_loss: 0.0189 - val_mean_absolute_error: 0.0798 - val_mean_squared_error: 0.0189\n",
      "Epoch 348/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0164 - mean_absolute_error: 0.0766 - mean_squared_error: 0.0164 - val_loss: 0.0202 - val_mean_absolute_error: 0.0825 - val_mean_squared_error: 0.0202\n",
      "Epoch 349/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0165 - mean_absolute_error: 0.0768 - mean_squared_error: 0.0165 - val_loss: 0.0188 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0188\n",
      "Epoch 350/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0164 - mean_absolute_error: 0.0765 - mean_squared_error: 0.0164 - val_loss: 0.0184 - val_mean_absolute_error: 0.0764 - val_mean_squared_error: 0.0184\n",
      "Epoch 351/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0163 - mean_absolute_error: 0.0760 - mean_squared_error: 0.0163 - val_loss: 0.0188 - val_mean_absolute_error: 0.0785 - val_mean_squared_error: 0.0188\n",
      "Epoch 352/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0164 - mean_absolute_error: 0.0761 - mean_squared_error: 0.0164 - val_loss: 0.0193 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0193\n",
      "Epoch 353/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0163 - mean_absolute_error: 0.0760 - mean_squared_error: 0.0163 - val_loss: 0.0192 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0192\n",
      "Epoch 354/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0164 - mean_absolute_error: 0.0760 - mean_squared_error: 0.0164 - val_loss: 0.0191 - val_mean_absolute_error: 0.0799 - val_mean_squared_error: 0.0191\n",
      "Epoch 355/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0164 - mean_absolute_error: 0.0764 - mean_squared_error: 0.0164 - val_loss: 0.0193 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0193\n",
      "Epoch 356/600\n",
      "3600/3600 [==============================] - 0s 89us/sample - loss: 0.0163 - mean_absolute_error: 0.0759 - mean_squared_error: 0.0163 - val_loss: 0.0190 - val_mean_absolute_error: 0.0808 - val_mean_squared_error: 0.0190\n",
      "Epoch 357/600\n",
      "3600/3600 [==============================] - 0s 88us/sample - loss: 0.0164 - mean_absolute_error: 0.0762 - mean_squared_error: 0.0164 - val_loss: 0.0196 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0196\n",
      "Epoch 358/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0164 - mean_absolute_error: 0.0761 - mean_squared_error: 0.0164 - val_loss: 0.0191 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0191\n",
      "Epoch 359/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0163 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0163 - val_loss: 0.0219 - val_mean_absolute_error: 0.0968 - val_mean_squared_error: 0.0219\n",
      "Epoch 360/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0163 - mean_absolute_error: 0.0759 - mean_squared_error: 0.0163 - val_loss: 0.0191 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0191\n",
      "Epoch 361/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0163 - mean_absolute_error: 0.0761 - mean_squared_error: 0.0163 - val_loss: 0.0194 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0194\n",
      "Epoch 362/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0164 - mean_absolute_error: 0.0761 - mean_squared_error: 0.0164 - val_loss: 0.0198 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0198\n",
      "Epoch 363/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0164 - mean_absolute_error: 0.0761 - mean_squared_error: 0.0164 - val_loss: 0.0205 - val_mean_absolute_error: 0.0855 - val_mean_squared_error: 0.0205\n",
      "Epoch 364/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0163 - mean_absolute_error: 0.0759 - mean_squared_error: 0.0163 - val_loss: 0.0185 - val_mean_absolute_error: 0.0765 - val_mean_squared_error: 0.0185\n",
      "Epoch 365/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0163 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0163 - val_loss: 0.0211 - val_mean_absolute_error: 0.0932 - val_mean_squared_error: 0.0211\n",
      "Epoch 366/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0163 - mean_absolute_error: 0.0760 - mean_squared_error: 0.0163 - val_loss: 0.0200 - val_mean_absolute_error: 0.0870 - val_mean_squared_error: 0.0200\n",
      "Epoch 367/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0163 - mean_absolute_error: 0.0763 - mean_squared_error: 0.0163 - val_loss: 0.0202 - val_mean_absolute_error: 0.0876 - val_mean_squared_error: 0.0202\n",
      "Epoch 368/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0163 - mean_absolute_error: 0.0759 - mean_squared_error: 0.0163 - val_loss: 0.0203 - val_mean_absolute_error: 0.0850 - val_mean_squared_error: 0.0203\n",
      "Epoch 369/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0162 - mean_absolute_error: 0.0755 - mean_squared_error: 0.0162 - val_loss: 0.0212 - val_mean_absolute_error: 0.0807 - val_mean_squared_error: 0.0212\n",
      "Epoch 370/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0163 - mean_absolute_error: 0.0755 - mean_squared_error: 0.0163 - val_loss: 0.0194 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0194\n",
      "Epoch 371/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0162 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0162 - val_loss: 0.0198 - val_mean_absolute_error: 0.0793 - val_mean_squared_error: 0.0198\n",
      "Epoch 372/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0162 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0162 - val_loss: 0.0187 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0187\n",
      "Epoch 373/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0163 - mean_absolute_error: 0.0754 - mean_squared_error: 0.0163 - val_loss: 0.0189 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0189\n",
      "Epoch 374/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0162 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0162 - val_loss: 0.0193 - val_mean_absolute_error: 0.0806 - val_mean_squared_error: 0.0193\n",
      "Epoch 375/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0162 - mean_absolute_error: 0.0754 - mean_squared_error: 0.0162 - val_loss: 0.0199 - val_mean_absolute_error: 0.0796 - val_mean_squared_error: 0.0199\n",
      "Epoch 376/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0162 - mean_absolute_error: 0.0755 - mean_squared_error: 0.0162 - val_loss: 0.0192 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0192\n",
      "Epoch 377/600\n",
      "3600/3600 [==============================] - 0s 91us/sample - loss: 0.0162 - mean_absolute_error: 0.0759 - mean_squared_error: 0.0162 - val_loss: 0.0213 - val_mean_absolute_error: 0.0897 - val_mean_squared_error: 0.0213\n",
      "Epoch 378/600\n",
      "3600/3600 [==============================] - 0s 96us/sample - loss: 0.0161 - mean_absolute_error: 0.0754 - mean_squared_error: 0.0161 - val_loss: 0.0192 - val_mean_absolute_error: 0.0785 - val_mean_squared_error: 0.0192\n",
      "Epoch 379/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0162 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0162 - val_loss: 0.0196 - val_mean_absolute_error: 0.0830 - val_mean_squared_error: 0.0196\n",
      "Epoch 380/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0162 - mean_absolute_error: 0.0756 - mean_squared_error: 0.0162 - val_loss: 0.0189 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0189\n",
      "Epoch 381/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0161 - mean_absolute_error: 0.0755 - mean_squared_error: 0.0161 - val_loss: 0.0193 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0193\n",
      "Epoch 382/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0161 - mean_absolute_error: 0.0754 - mean_squared_error: 0.0161 - val_loss: 0.0194 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0194\n",
      "Epoch 383/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0162 - mean_absolute_error: 0.0755 - mean_squared_error: 0.0162 - val_loss: 0.0189 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0189\n",
      "Epoch 384/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0161 - mean_absolute_error: 0.0758 - mean_squared_error: 0.0161 - val_loss: 0.0210 - val_mean_absolute_error: 0.0880 - val_mean_squared_error: 0.0210\n",
      "Epoch 385/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0161 - mean_absolute_error: 0.0753 - mean_squared_error: 0.0161 - val_loss: 0.0195 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0195\n",
      "Epoch 386/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0160 - mean_absolute_error: 0.0749 - mean_squared_error: 0.0160 - val_loss: 0.0198 - val_mean_absolute_error: 0.0847 - val_mean_squared_error: 0.0198\n",
      "Epoch 387/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0162 - mean_absolute_error: 0.0757 - mean_squared_error: 0.0162 - val_loss: 0.0212 - val_mean_absolute_error: 0.0896 - val_mean_squared_error: 0.0212\n",
      "Epoch 388/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0160 - mean_absolute_error: 0.0748 - mean_squared_error: 0.0160 - val_loss: 0.0208 - val_mean_absolute_error: 0.0893 - val_mean_squared_error: 0.0208\n",
      "Epoch 389/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0161 - mean_absolute_error: 0.0753 - mean_squared_error: 0.0161 - val_loss: 0.0186 - val_mean_absolute_error: 0.0766 - val_mean_squared_error: 0.0186\n",
      "Epoch 390/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0161 - mean_absolute_error: 0.0750 - mean_squared_error: 0.0161 - val_loss: 0.0198 - val_mean_absolute_error: 0.0840 - val_mean_squared_error: 0.0198\n",
      "Epoch 391/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0160 - mean_absolute_error: 0.0749 - mean_squared_error: 0.0160 - val_loss: 0.0189 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0189\n",
      "Epoch 392/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0161 - mean_absolute_error: 0.0753 - mean_squared_error: 0.0161 - val_loss: 0.0194 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0194\n",
      "Epoch 393/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0160 - mean_absolute_error: 0.0748 - mean_squared_error: 0.0160 - val_loss: 0.0213 - val_mean_absolute_error: 0.0912 - val_mean_squared_error: 0.0213\n",
      "Epoch 394/600\n",
      "3600/3600 [==============================] - 0s 66us/sample - loss: 0.0160 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0160 - val_loss: 0.0216 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0216\n",
      "Epoch 395/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0160 - mean_absolute_error: 0.0752 - mean_squared_error: 0.0160 - val_loss: 0.0187 - val_mean_absolute_error: 0.0768 - val_mean_squared_error: 0.0187\n",
      "Epoch 396/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0160 - mean_absolute_error: 0.0753 - mean_squared_error: 0.0160 - val_loss: 0.0191 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0191\n",
      "Epoch 397/600\n",
      "3600/3600 [==============================] - 0s 76us/sample - loss: 0.0161 - mean_absolute_error: 0.0756 - mean_squared_error: 0.0161 - val_loss: 0.0195 - val_mean_absolute_error: 0.0830 - val_mean_squared_error: 0.0195\n",
      "Epoch 398/600\n",
      "3600/3600 [==============================] - 0s 95us/sample - loss: 0.0159 - mean_absolute_error: 0.0744 - mean_squared_error: 0.0159 - val_loss: 0.0216 - val_mean_absolute_error: 0.0904 - val_mean_squared_error: 0.0216\n",
      "Epoch 399/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0161 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0161 - val_loss: 0.0186 - val_mean_absolute_error: 0.0763 - val_mean_squared_error: 0.0186\n",
      "Epoch 400/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0160 - mean_absolute_error: 0.0746 - mean_squared_error: 0.0160 - val_loss: 0.0188 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0188\n",
      "Epoch 401/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0161 - mean_absolute_error: 0.0750 - mean_squared_error: 0.0161 - val_loss: 0.0188 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0188\n",
      "Epoch 402/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0159 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0159 - val_loss: 0.0199 - val_mean_absolute_error: 0.0827 - val_mean_squared_error: 0.0199\n",
      "Epoch 403/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0160 - mean_absolute_error: 0.0752 - mean_squared_error: 0.0160 - val_loss: 0.0194 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0194\n",
      "Epoch 404/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0160 - mean_absolute_error: 0.0749 - mean_squared_error: 0.0160 - val_loss: 0.0190 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0190\n",
      "Epoch 405/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0159 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0159 - val_loss: 0.0200 - val_mean_absolute_error: 0.0849 - val_mean_squared_error: 0.0200\n",
      "Epoch 406/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0160 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0160 - val_loss: 0.0201 - val_mean_absolute_error: 0.0850 - val_mean_squared_error: 0.0201\n",
      "Epoch 407/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0160 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0160 - val_loss: 0.0194 - val_mean_absolute_error: 0.0795 - val_mean_squared_error: 0.0194\n",
      "Epoch 408/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0160 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0160 - val_loss: 0.0188 - val_mean_absolute_error: 0.0765 - val_mean_squared_error: 0.0188\n",
      "Epoch 409/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0159 - mean_absolute_error: 0.0744 - mean_squared_error: 0.0159 - val_loss: 0.0198 - val_mean_absolute_error: 0.0824 - val_mean_squared_error: 0.0198\n",
      "Epoch 410/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0159 - mean_absolute_error: 0.0750 - mean_squared_error: 0.0159 - val_loss: 0.0188 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0188\n",
      "Epoch 411/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0159 - mean_absolute_error: 0.0744 - mean_squared_error: 0.0159 - val_loss: 0.0199 - val_mean_absolute_error: 0.0846 - val_mean_squared_error: 0.0199\n",
      "Epoch 412/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0159 - mean_absolute_error: 0.0745 - mean_squared_error: 0.0159 - val_loss: 0.0187 - val_mean_absolute_error: 0.0762 - val_mean_squared_error: 0.0187\n",
      "Epoch 413/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0160 - mean_absolute_error: 0.0752 - mean_squared_error: 0.0160 - val_loss: 0.0188 - val_mean_absolute_error: 0.0765 - val_mean_squared_error: 0.0188\n",
      "Epoch 414/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0159 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0159 - val_loss: 0.0198 - val_mean_absolute_error: 0.0842 - val_mean_squared_error: 0.0198\n",
      "Epoch 415/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0159 - mean_absolute_error: 0.0748 - mean_squared_error: 0.0159 - val_loss: 0.0201 - val_mean_absolute_error: 0.0868 - val_mean_squared_error: 0.0201\n",
      "Epoch 416/600\n",
      "3600/3600 [==============================] - 0s 65us/sample - loss: 0.0159 - mean_absolute_error: 0.0742 - mean_squared_error: 0.0159 - val_loss: 0.0187 - val_mean_absolute_error: 0.0766 - val_mean_squared_error: 0.0187\n",
      "Epoch 417/600\n",
      "3600/3600 [==============================] - 0s 66us/sample - loss: 0.0159 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0159 - val_loss: 0.0198 - val_mean_absolute_error: 0.0839 - val_mean_squared_error: 0.0198\n",
      "Epoch 418/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0158 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0158 - val_loss: 0.0191 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0191\n",
      "Epoch 419/600\n",
      "3600/3600 [==============================] - 0s 76us/sample - loss: 0.0157 - mean_absolute_error: 0.0742 - mean_squared_error: 0.0157 - val_loss: 0.0214 - val_mean_absolute_error: 0.0882 - val_mean_squared_error: 0.0214\n",
      "Epoch 420/600\n",
      "3600/3600 [==============================] - 0s 82us/sample - loss: 0.0159 - mean_absolute_error: 0.0751 - mean_squared_error: 0.0159 - val_loss: 0.0193 - val_mean_absolute_error: 0.0812 - val_mean_squared_error: 0.0193\n",
      "Epoch 421/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0159 - mean_absolute_error: 0.0747 - mean_squared_error: 0.0159 - val_loss: 0.0193 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0193\n",
      "Epoch 422/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0157 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0157 - val_loss: 0.0196 - val_mean_absolute_error: 0.0824 - val_mean_squared_error: 0.0196\n",
      "Epoch 423/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0158 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0158 - val_loss: 0.0223 - val_mean_absolute_error: 0.1001 - val_mean_squared_error: 0.0223\n",
      "Epoch 424/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0158 - mean_absolute_error: 0.0748 - mean_squared_error: 0.0158 - val_loss: 0.0196 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0196\n",
      "Epoch 425/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0158 - mean_absolute_error: 0.0744 - mean_squared_error: 0.0158 - val_loss: 0.0206 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0206\n",
      "Epoch 426/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0157 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0157 - val_loss: 0.0195 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0195\n",
      "Epoch 427/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0158 - mean_absolute_error: 0.0748 - mean_squared_error: 0.0158 - val_loss: 0.0201 - val_mean_absolute_error: 0.0798 - val_mean_squared_error: 0.0201\n",
      "Epoch 428/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0158 - mean_absolute_error: 0.0745 - mean_squared_error: 0.0158 - val_loss: 0.0190 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0190\n",
      "Epoch 429/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0158 - mean_absolute_error: 0.0745 - mean_squared_error: 0.0158 - val_loss: 0.0194 - val_mean_absolute_error: 0.0812 - val_mean_squared_error: 0.0194\n",
      "Epoch 430/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0157 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0157 - val_loss: 0.0192 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0192\n",
      "Epoch 431/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0158 - mean_absolute_error: 0.0741 - mean_squared_error: 0.0158 - val_loss: 0.0191 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0191\n",
      "Epoch 432/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0157 - mean_absolute_error: 0.0741 - mean_squared_error: 0.0157 - val_loss: 0.0205 - val_mean_absolute_error: 0.0875 - val_mean_squared_error: 0.0205\n",
      "Epoch 433/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0157 - mean_absolute_error: 0.0741 - mean_squared_error: 0.0157 - val_loss: 0.0213 - val_mean_absolute_error: 0.0868 - val_mean_squared_error: 0.0213\n",
      "Epoch 434/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0157 - mean_absolute_error: 0.0739 - mean_squared_error: 0.0157 - val_loss: 0.0192 - val_mean_absolute_error: 0.0793 - val_mean_squared_error: 0.0192\n",
      "Epoch 435/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0157 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0157 - val_loss: 0.0190 - val_mean_absolute_error: 0.0789 - val_mean_squared_error: 0.0190\n",
      "Epoch 436/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0158 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0158 - val_loss: 0.0192 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0192\n",
      "Epoch 437/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0157 - mean_absolute_error: 0.0742 - mean_squared_error: 0.0157 - val_loss: 0.0196 - val_mean_absolute_error: 0.0842 - val_mean_squared_error: 0.0196\n",
      "Epoch 438/600\n",
      "3600/3600 [==============================] - 0s 71us/sample - loss: 0.0156 - mean_absolute_error: 0.0735 - mean_squared_error: 0.0156 - val_loss: 0.0199 - val_mean_absolute_error: 0.0832 - val_mean_squared_error: 0.0199\n",
      "Epoch 439/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0157 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0157 - val_loss: 0.0207 - val_mean_absolute_error: 0.0802 - val_mean_squared_error: 0.0207\n",
      "Epoch 440/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0156 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0156 - val_loss: 0.0191 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0191\n",
      "Epoch 441/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0157 - mean_absolute_error: 0.0739 - mean_squared_error: 0.0157 - val_loss: 0.0207 - val_mean_absolute_error: 0.0876 - val_mean_squared_error: 0.0207\n",
      "Epoch 442/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0157 - mean_absolute_error: 0.0736 - mean_squared_error: 0.0157 - val_loss: 0.0224 - val_mean_absolute_error: 0.0966 - val_mean_squared_error: 0.0224\n",
      "Epoch 443/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0157 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0157 - val_loss: 0.0188 - val_mean_absolute_error: 0.0762 - val_mean_squared_error: 0.0188\n",
      "Epoch 444/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0156 - mean_absolute_error: 0.0739 - mean_squared_error: 0.0156 - val_loss: 0.0193 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0193\n",
      "Epoch 445/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0157 - mean_absolute_error: 0.0738 - mean_squared_error: 0.0157 - val_loss: 0.0193 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0193\n",
      "Epoch 446/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0156 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0156 - val_loss: 0.0201 - val_mean_absolute_error: 0.0860 - val_mean_squared_error: 0.0201\n",
      "Epoch 447/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0156 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0156 - val_loss: 0.0208 - val_mean_absolute_error: 0.0888 - val_mean_squared_error: 0.0208\n",
      "Epoch 448/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0156 - mean_absolute_error: 0.0737 - mean_squared_error: 0.0156 - val_loss: 0.0193 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0193\n",
      "Epoch 449/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0156 - mean_absolute_error: 0.0737 - mean_squared_error: 0.0156 - val_loss: 0.0190 - val_mean_absolute_error: 0.0767 - val_mean_squared_error: 0.0190\n",
      "Epoch 450/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0157 - mean_absolute_error: 0.0738 - mean_squared_error: 0.0157 - val_loss: 0.0206 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0206\n",
      "Epoch 451/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0156 - mean_absolute_error: 0.0738 - mean_squared_error: 0.0156 - val_loss: 0.0210 - val_mean_absolute_error: 0.0825 - val_mean_squared_error: 0.0210\n",
      "Epoch 452/600\n",
      "3600/3600 [==============================] - 0s 85us/sample - loss: 0.0156 - mean_absolute_error: 0.0735 - mean_squared_error: 0.0156 - val_loss: 0.0200 - val_mean_absolute_error: 0.0835 - val_mean_squared_error: 0.0200\n",
      "Epoch 453/600\n",
      "3600/3600 [==============================] - 0s 77us/sample - loss: 0.0156 - mean_absolute_error: 0.0743 - mean_squared_error: 0.0156 - val_loss: 0.0193 - val_mean_absolute_error: 0.0783 - val_mean_squared_error: 0.0193\n",
      "Epoch 454/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0155 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0155 - val_loss: 0.0210 - val_mean_absolute_error: 0.0902 - val_mean_squared_error: 0.0210\n",
      "Epoch 455/600\n",
      "3600/3600 [==============================] - 0s 74us/sample - loss: 0.0155 - mean_absolute_error: 0.0738 - mean_squared_error: 0.0155 - val_loss: 0.0194 - val_mean_absolute_error: 0.0786 - val_mean_squared_error: 0.0194\n",
      "Epoch 456/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0155 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0155 - val_loss: 0.0190 - val_mean_absolute_error: 0.0771 - val_mean_squared_error: 0.0190\n",
      "Epoch 457/600\n",
      "3600/3600 [==============================] - 0s 84us/sample - loss: 0.0157 - mean_absolute_error: 0.0740 - mean_squared_error: 0.0157 - val_loss: 0.0215 - val_mean_absolute_error: 0.0914 - val_mean_squared_error: 0.0215\n",
      "Epoch 458/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0155 - mean_absolute_error: 0.0733 - mean_squared_error: 0.0155 - val_loss: 0.0205 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0205\n",
      "Epoch 459/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0154 - mean_absolute_error: 0.0731 - mean_squared_error: 0.0154 - val_loss: 0.0193 - val_mean_absolute_error: 0.0770 - val_mean_squared_error: 0.0193\n",
      "Epoch 460/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0155 - mean_absolute_error: 0.0734 - mean_squared_error: 0.0155 - val_loss: 0.0195 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0195\n",
      "Epoch 461/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0155 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0155 - val_loss: 0.0196 - val_mean_absolute_error: 0.0816 - val_mean_squared_error: 0.0196\n",
      "Epoch 462/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0154 - mean_absolute_error: 0.0734 - mean_squared_error: 0.0154 - val_loss: 0.0214 - val_mean_absolute_error: 0.0917 - val_mean_squared_error: 0.0214\n",
      "Epoch 463/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0155 - mean_absolute_error: 0.0739 - mean_squared_error: 0.0155 - val_loss: 0.0194 - val_mean_absolute_error: 0.0808 - val_mean_squared_error: 0.0194\n",
      "Epoch 464/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0155 - mean_absolute_error: 0.0738 - mean_squared_error: 0.0155 - val_loss: 0.0193 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0193\n",
      "Epoch 465/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0155 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0155 - val_loss: 0.0196 - val_mean_absolute_error: 0.0812 - val_mean_squared_error: 0.0196\n",
      "Epoch 466/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0155 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0155 - val_loss: 0.0189 - val_mean_absolute_error: 0.0768 - val_mean_squared_error: 0.0189\n",
      "Epoch 467/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0155 - mean_absolute_error: 0.0736 - mean_squared_error: 0.0155 - val_loss: 0.0207 - val_mean_absolute_error: 0.0886 - val_mean_squared_error: 0.0207\n",
      "Epoch 468/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0155 - mean_absolute_error: 0.0735 - mean_squared_error: 0.0155 - val_loss: 0.0194 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0194\n",
      "Epoch 469/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0154 - mean_absolute_error: 0.0730 - mean_squared_error: 0.0154 - val_loss: 0.0195 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0195\n",
      "Epoch 470/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0155 - mean_absolute_error: 0.0734 - mean_squared_error: 0.0155 - val_loss: 0.0215 - val_mean_absolute_error: 0.0914 - val_mean_squared_error: 0.0215\n",
      "Epoch 471/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0154 - mean_absolute_error: 0.0731 - mean_squared_error: 0.0154 - val_loss: 0.0195 - val_mean_absolute_error: 0.0802 - val_mean_squared_error: 0.0195\n",
      "Epoch 472/600\n",
      "3600/3600 [==============================] - 0s 79us/sample - loss: 0.0155 - mean_absolute_error: 0.0736 - mean_squared_error: 0.0155 - val_loss: 0.0192 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0192\n",
      "Epoch 473/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0154 - mean_absolute_error: 0.0734 - mean_squared_error: 0.0154 - val_loss: 0.0195 - val_mean_absolute_error: 0.0803 - val_mean_squared_error: 0.0195\n",
      "Epoch 474/600\n",
      "3600/3600 [==============================] - 0s 70us/sample - loss: 0.0154 - mean_absolute_error: 0.0729 - mean_squared_error: 0.0154 - val_loss: 0.0191 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0191\n",
      "Epoch 475/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0154 - mean_absolute_error: 0.0735 - mean_squared_error: 0.0154 - val_loss: 0.0206 - val_mean_absolute_error: 0.0872 - val_mean_squared_error: 0.0206\n",
      "Epoch 476/600\n",
      "3600/3600 [==============================] - 0s 72us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0198 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0198\n",
      "Epoch 477/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0155 - mean_absolute_error: 0.0735 - mean_squared_error: 0.0155 - val_loss: 0.0199 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0199\n",
      "Epoch 478/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0153 - mean_absolute_error: 0.0727 - mean_squared_error: 0.0153 - val_loss: 0.0191 - val_mean_absolute_error: 0.0773 - val_mean_squared_error: 0.0191\n",
      "Epoch 479/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0153 - mean_absolute_error: 0.0730 - mean_squared_error: 0.0153 - val_loss: 0.0193 - val_mean_absolute_error: 0.0789 - val_mean_squared_error: 0.0193\n",
      "Epoch 480/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0154 - mean_absolute_error: 0.0731 - mean_squared_error: 0.0154 - val_loss: 0.0193 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0193\n",
      "Epoch 481/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0153 - mean_absolute_error: 0.0730 - mean_squared_error: 0.0153 - val_loss: 0.0200 - val_mean_absolute_error: 0.0831 - val_mean_squared_error: 0.0200\n",
      "Epoch 482/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0221 - val_mean_absolute_error: 0.0957 - val_mean_squared_error: 0.0221\n",
      "Epoch 483/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0208 - val_mean_absolute_error: 0.0880 - val_mean_squared_error: 0.0208\n",
      "Epoch 484/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0154 - mean_absolute_error: 0.0732 - mean_squared_error: 0.0154 - val_loss: 0.0212 - val_mean_absolute_error: 0.0859 - val_mean_squared_error: 0.0212\n",
      "Epoch 485/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0153 - mean_absolute_error: 0.0731 - mean_squared_error: 0.0153 - val_loss: 0.0193 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0193\n",
      "Epoch 486/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0153 - mean_absolute_error: 0.0726 - mean_squared_error: 0.0153 - val_loss: 0.0196 - val_mean_absolute_error: 0.0795 - val_mean_squared_error: 0.0196\n",
      "Epoch 487/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0193 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0193\n",
      "Epoch 488/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0152 - mean_absolute_error: 0.0726 - mean_squared_error: 0.0152 - val_loss: 0.0195 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0195\n",
      "Epoch 489/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0153 - mean_absolute_error: 0.0729 - mean_squared_error: 0.0153 - val_loss: 0.0199 - val_mean_absolute_error: 0.0785 - val_mean_squared_error: 0.0199\n",
      "Epoch 490/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0153 - mean_absolute_error: 0.0726 - mean_squared_error: 0.0153 - val_loss: 0.0195 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0195\n",
      "Epoch 491/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0198 - val_mean_absolute_error: 0.0795 - val_mean_squared_error: 0.0198\n",
      "Epoch 492/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0152 - mean_absolute_error: 0.0726 - mean_squared_error: 0.0152 - val_loss: 0.0200 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0200\n",
      "Epoch 493/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0152 - mean_absolute_error: 0.0723 - mean_squared_error: 0.0152 - val_loss: 0.0209 - val_mean_absolute_error: 0.0844 - val_mean_squared_error: 0.0209\n",
      "Epoch 494/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0152 - mean_absolute_error: 0.0725 - mean_squared_error: 0.0152 - val_loss: 0.0194 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0194\n",
      "Epoch 495/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0153 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0153 - val_loss: 0.0201 - val_mean_absolute_error: 0.0786 - val_mean_squared_error: 0.0201\n",
      "Epoch 496/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0152 - mean_absolute_error: 0.0729 - mean_squared_error: 0.0152 - val_loss: 0.0200 - val_mean_absolute_error: 0.0830 - val_mean_squared_error: 0.0200\n",
      "Epoch 497/600\n",
      "3600/3600 [==============================] - 0s 78us/sample - loss: 0.0152 - mean_absolute_error: 0.0727 - mean_squared_error: 0.0152 - val_loss: 0.0192 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0192\n",
      "Epoch 498/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0152 - mean_absolute_error: 0.0730 - mean_squared_error: 0.0152 - val_loss: 0.0196 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0196\n",
      "Epoch 499/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0152 - mean_absolute_error: 0.0725 - mean_squared_error: 0.0152 - val_loss: 0.0197 - val_mean_absolute_error: 0.0780 - val_mean_squared_error: 0.0197\n",
      "Epoch 500/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0152 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0152 - val_loss: 0.0199 - val_mean_absolute_error: 0.0826 - val_mean_squared_error: 0.0199\n",
      "Epoch 501/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0151 - mean_absolute_error: 0.0722 - mean_squared_error: 0.0151 - val_loss: 0.0199 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0199\n",
      "Epoch 502/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0152 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0152 - val_loss: 0.0193 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0193\n",
      "Epoch 503/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0153 - mean_absolute_error: 0.0727 - mean_squared_error: 0.0153 - val_loss: 0.0193 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0193\n",
      "Epoch 504/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0151 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0151 - val_loss: 0.0193 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0193\n",
      "Epoch 505/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0151 - mean_absolute_error: 0.0722 - mean_squared_error: 0.0151 - val_loss: 0.0208 - val_mean_absolute_error: 0.0855 - val_mean_squared_error: 0.0208\n",
      "Epoch 506/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0152 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0152 - val_loss: 0.0193 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0193\n",
      "Epoch 507/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0150 - mean_absolute_error: 0.0720 - mean_squared_error: 0.0150 - val_loss: 0.0200 - val_mean_absolute_error: 0.0798 - val_mean_squared_error: 0.0200\n",
      "Epoch 508/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0151 - mean_absolute_error: 0.0723 - mean_squared_error: 0.0151 - val_loss: 0.0215 - val_mean_absolute_error: 0.0886 - val_mean_squared_error: 0.0215\n",
      "Epoch 509/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0151 - mean_absolute_error: 0.0726 - mean_squared_error: 0.0151 - val_loss: 0.0207 - val_mean_absolute_error: 0.0855 - val_mean_squared_error: 0.0207\n",
      "Epoch 510/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0151 - mean_absolute_error: 0.0723 - mean_squared_error: 0.0151 - val_loss: 0.0193 - val_mean_absolute_error: 0.0768 - val_mean_squared_error: 0.0193\n",
      "Epoch 511/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0152 - mean_absolute_error: 0.0725 - mean_squared_error: 0.0152 - val_loss: 0.0200 - val_mean_absolute_error: 0.0819 - val_mean_squared_error: 0.0200\n",
      "Epoch 512/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0151 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0151 - val_loss: 0.0201 - val_mean_absolute_error: 0.0791 - val_mean_squared_error: 0.0201\n",
      "Epoch 513/600\n",
      "3600/3600 [==============================] - 0s 62us/sample - loss: 0.0151 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0151 - val_loss: 0.0200 - val_mean_absolute_error: 0.0823 - val_mean_squared_error: 0.0200\n",
      "Epoch 514/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0151 - mean_absolute_error: 0.0725 - mean_squared_error: 0.0151 - val_loss: 0.0217 - val_mean_absolute_error: 0.0960 - val_mean_squared_error: 0.0217\n",
      "Epoch 515/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0151 - mean_absolute_error: 0.0720 - mean_squared_error: 0.0151 - val_loss: 0.0200 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0200\n",
      "Epoch 516/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0151 - mean_absolute_error: 0.0728 - mean_squared_error: 0.0151 - val_loss: 0.0205 - val_mean_absolute_error: 0.0836 - val_mean_squared_error: 0.0205\n",
      "Epoch 517/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0152 - mean_absolute_error: 0.0723 - mean_squared_error: 0.0152 - val_loss: 0.0206 - val_mean_absolute_error: 0.0859 - val_mean_squared_error: 0.0206\n",
      "Epoch 518/600\n",
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0198 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0198\n",
      "Epoch 519/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0201 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0201\n",
      "Epoch 520/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0150 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0150 - val_loss: 0.0200 - val_mean_absolute_error: 0.0826 - val_mean_squared_error: 0.0200\n",
      "Epoch 521/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0151 - mean_absolute_error: 0.0720 - mean_squared_error: 0.0151 - val_loss: 0.0196 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0196\n",
      "Epoch 522/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0150 - mean_absolute_error: 0.0720 - mean_squared_error: 0.0150 - val_loss: 0.0216 - val_mean_absolute_error: 0.0926 - val_mean_squared_error: 0.0216\n",
      "Epoch 523/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0150 - mean_absolute_error: 0.0719 - mean_squared_error: 0.0150 - val_loss: 0.0202 - val_mean_absolute_error: 0.0781 - val_mean_squared_error: 0.0202\n",
      "Epoch 524/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0150 - mean_absolute_error: 0.0719 - mean_squared_error: 0.0150 - val_loss: 0.0198 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0198\n",
      "Epoch 525/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0196 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0196\n",
      "Epoch 526/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0150 - mean_absolute_error: 0.0725 - mean_squared_error: 0.0150 - val_loss: 0.0199 - val_mean_absolute_error: 0.0789 - val_mean_squared_error: 0.0199\n",
      "Epoch 527/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0199 - val_mean_absolute_error: 0.0826 - val_mean_squared_error: 0.0199\n",
      "Epoch 528/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0150 - mean_absolute_error: 0.0724 - mean_squared_error: 0.0150 - val_loss: 0.0208 - val_mean_absolute_error: 0.0818 - val_mean_squared_error: 0.0208\n",
      "Epoch 529/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0150 - mean_absolute_error: 0.0718 - mean_squared_error: 0.0150 - val_loss: 0.0233 - val_mean_absolute_error: 0.1010 - val_mean_squared_error: 0.0233\n",
      "Epoch 530/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0149 - mean_absolute_error: 0.0715 - mean_squared_error: 0.0149 - val_loss: 0.0207 - val_mean_absolute_error: 0.0815 - val_mean_squared_error: 0.0207\n",
      "Epoch 531/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0197 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0197\n",
      "Epoch 532/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0149 - mean_absolute_error: 0.0717 - mean_squared_error: 0.0149 - val_loss: 0.0209 - val_mean_absolute_error: 0.0841 - val_mean_squared_error: 0.0209\n",
      "Epoch 533/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0150 - mean_absolute_error: 0.0721 - mean_squared_error: 0.0150 - val_loss: 0.0203 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0203\n",
      "Epoch 534/600\n",
      "3600/3600 [==============================] - 0s 73us/sample - loss: 0.0150 - mean_absolute_error: 0.0718 - mean_squared_error: 0.0150 - val_loss: 0.0200 - val_mean_absolute_error: 0.0831 - val_mean_squared_error: 0.0200\n",
      "Epoch 535/600\n",
      "3600/3600 [==============================] - 0s 72us/sample - loss: 0.0149 - mean_absolute_error: 0.0714 - mean_squared_error: 0.0149 - val_loss: 0.0200 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0200\n",
      "Epoch 536/600\n",
      "3600/3600 [==============================] - 0s 79us/sample - loss: 0.0149 - mean_absolute_error: 0.0718 - mean_squared_error: 0.0149 - val_loss: 0.0200 - val_mean_absolute_error: 0.0797 - val_mean_squared_error: 0.0200\n",
      "Epoch 537/600\n",
      "3600/3600 [==============================] - 0s 89us/sample - loss: 0.0148 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0148 - val_loss: 0.0225 - val_mean_absolute_error: 0.0925 - val_mean_squared_error: 0.0225\n",
      "Epoch 538/600\n",
      "3600/3600 [==============================] - 0s 87us/sample - loss: 0.0149 - mean_absolute_error: 0.0716 - mean_squared_error: 0.0149 - val_loss: 0.0195 - val_mean_absolute_error: 0.0774 - val_mean_squared_error: 0.0195\n",
      "Epoch 539/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0149 - mean_absolute_error: 0.0715 - mean_squared_error: 0.0149 - val_loss: 0.0205 - val_mean_absolute_error: 0.0805 - val_mean_squared_error: 0.0205\n",
      "Epoch 540/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0148 - mean_absolute_error: 0.0715 - mean_squared_error: 0.0148 - val_loss: 0.0200 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0200\n",
      "Epoch 541/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0149 - mean_absolute_error: 0.0718 - mean_squared_error: 0.0149 - val_loss: 0.0207 - val_mean_absolute_error: 0.0884 - val_mean_squared_error: 0.0207\n",
      "Epoch 542/600\n",
      "3600/3600 [==============================] - 0s 57us/sample - loss: 0.0149 - mean_absolute_error: 0.0719 - mean_squared_error: 0.0149 - val_loss: 0.0201 - val_mean_absolute_error: 0.0808 - val_mean_squared_error: 0.0201\n",
      "Epoch 543/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0148 - mean_absolute_error: 0.0715 - mean_squared_error: 0.0148 - val_loss: 0.0208 - val_mean_absolute_error: 0.0866 - val_mean_squared_error: 0.0208\n",
      "Epoch 544/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0149 - mean_absolute_error: 0.0716 - mean_squared_error: 0.0149 - val_loss: 0.0202 - val_mean_absolute_error: 0.0788 - val_mean_squared_error: 0.0202\n",
      "Epoch 545/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0149 - mean_absolute_error: 0.0714 - mean_squared_error: 0.0149 - val_loss: 0.0219 - val_mean_absolute_error: 0.0932 - val_mean_squared_error: 0.0219\n",
      "Epoch 546/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0148 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0148 - val_loss: 0.0206 - val_mean_absolute_error: 0.0853 - val_mean_squared_error: 0.0206\n",
      "Epoch 547/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0147 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0147 - val_loss: 0.0208 - val_mean_absolute_error: 0.0802 - val_mean_squared_error: 0.0208\n",
      "Epoch 548/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0149 - mean_absolute_error: 0.0714 - mean_squared_error: 0.0149 - val_loss: 0.0199 - val_mean_absolute_error: 0.0785 - val_mean_squared_error: 0.0199\n",
      "Epoch 549/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0148 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0148 - val_loss: 0.0200 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0200\n",
      "Epoch 550/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0149 - mean_absolute_error: 0.0716 - mean_squared_error: 0.0149 - val_loss: 0.0202 - val_mean_absolute_error: 0.0820 - val_mean_squared_error: 0.0202\n",
      "Epoch 551/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0148 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0148 - val_loss: 0.0200 - val_mean_absolute_error: 0.0800 - val_mean_squared_error: 0.0200\n",
      "Epoch 552/600\n",
      "3600/3600 [==============================] - 0s 55us/sample - loss: 0.0148 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0148 - val_loss: 0.0200 - val_mean_absolute_error: 0.0819 - val_mean_squared_error: 0.0200\n",
      "Epoch 553/600\n",
      "3600/3600 [==============================] - 0s 75us/sample - loss: 0.0148 - mean_absolute_error: 0.0716 - mean_squared_error: 0.0148 - val_loss: 0.0206 - val_mean_absolute_error: 0.0846 - val_mean_squared_error: 0.0206\n",
      "Epoch 554/600\n",
      "3600/3600 [==============================] - 0s 72us/sample - loss: 0.0147 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0147 - val_loss: 0.0206 - val_mean_absolute_error: 0.0835 - val_mean_squared_error: 0.0206\n",
      "Epoch 555/600\n",
      "3600/3600 [==============================] - 0s 77us/sample - loss: 0.0147 - mean_absolute_error: 0.0714 - mean_squared_error: 0.0147 - val_loss: 0.0205 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0205\n",
      "Epoch 556/600\n",
      "3600/3600 [==============================] - 0s 83us/sample - loss: 0.0148 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0148 - val_loss: 0.0206 - val_mean_absolute_error: 0.0837 - val_mean_squared_error: 0.0206\n",
      "Epoch 557/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0147 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0147 - val_loss: 0.0197 - val_mean_absolute_error: 0.0776 - val_mean_squared_error: 0.0197\n",
      "Epoch 558/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0148 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0148 - val_loss: 0.0203 - val_mean_absolute_error: 0.0817 - val_mean_squared_error: 0.0203\n",
      "Epoch 559/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0148 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0148 - val_loss: 0.0199 - val_mean_absolute_error: 0.0811 - val_mean_squared_error: 0.0199\n",
      "Epoch 560/600\n",
      "3600/3600 [==============================] - 0s 48us/sample - loss: 0.0146 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0146 - val_loss: 0.0198 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0198\n",
      "Epoch 561/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0147 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0147 - val_loss: 0.0203 - val_mean_absolute_error: 0.0783 - val_mean_squared_error: 0.0203\n",
      "Epoch 562/600\n",
      "3600/3600 [==============================] - 0s 49us/sample - loss: 0.0148 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0148 - val_loss: 0.0196 - val_mean_absolute_error: 0.0794 - val_mean_squared_error: 0.0196\n",
      "Epoch 563/600\n",
      "3600/3600 [==============================] - 0s 52us/sample - loss: 0.0147 - mean_absolute_error: 0.0713 - mean_squared_error: 0.0147 - val_loss: 0.0204 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0204\n",
      "Epoch 564/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0147 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0147 - val_loss: 0.0196 - val_mean_absolute_error: 0.0772 - val_mean_squared_error: 0.0196\n",
      "Epoch 565/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0147 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0147 - val_loss: 0.0204 - val_mean_absolute_error: 0.0801 - val_mean_squared_error: 0.0204\n",
      "Epoch 566/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0146 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0146 - val_loss: 0.0219 - val_mean_absolute_error: 0.0850 - val_mean_squared_error: 0.0219\n",
      "Epoch 567/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0146 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0146 - val_loss: 0.0215 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0215\n",
      "Epoch 568/600\n",
      "3600/3600 [==============================] - 0s 51us/sample - loss: 0.0147 - mean_absolute_error: 0.0712 - mean_squared_error: 0.0147 - val_loss: 0.0209 - val_mean_absolute_error: 0.0863 - val_mean_squared_error: 0.0209\n",
      "Epoch 569/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0146 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0146 - val_loss: 0.0198 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0198\n",
      "Epoch 570/600\n",
      "3600/3600 [==============================] - 0s 81us/sample - loss: 0.0146 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0146 - val_loss: 0.0214 - val_mean_absolute_error: 0.0906 - val_mean_squared_error: 0.0214\n",
      "Epoch 571/600\n",
      "3600/3600 [==============================] - 0s 60us/sample - loss: 0.0147 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0147 - val_loss: 0.0230 - val_mean_absolute_error: 0.0996 - val_mean_squared_error: 0.0230\n",
      "Epoch 572/600\n",
      "3600/3600 [==============================] - 0s 59us/sample - loss: 0.0146 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0146 - val_loss: 0.0212 - val_mean_absolute_error: 0.0804 - val_mean_squared_error: 0.0212\n",
      "Epoch 573/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0146 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0146 - val_loss: 0.0206 - val_mean_absolute_error: 0.0860 - val_mean_squared_error: 0.0206\n",
      "Epoch 574/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0145 - mean_absolute_error: 0.0706 - mean_squared_error: 0.0145 - val_loss: 0.0222 - val_mean_absolute_error: 0.0922 - val_mean_squared_error: 0.0222\n",
      "Epoch 575/600\n",
      "3600/3600 [==============================] - 0s 93us/sample - loss: 0.0146 - mean_absolute_error: 0.0711 - mean_squared_error: 0.0146 - val_loss: 0.0205 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0205\n",
      "Epoch 576/600\n",
      "3600/3600 [==============================] - 0s 94us/sample - loss: 0.0147 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0147 - val_loss: 0.0198 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0198\n",
      "Epoch 577/600\n",
      "3600/3600 [==============================] - 0s 56us/sample - loss: 0.0146 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0146 - val_loss: 0.0200 - val_mean_absolute_error: 0.0787 - val_mean_squared_error: 0.0200\n",
      "Epoch 578/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0146 - mean_absolute_error: 0.0705 - mean_squared_error: 0.0146 - val_loss: 0.0202 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0202\n",
      "Epoch 579/600\n",
      "3600/3600 [==============================] - 0s 69us/sample - loss: 0.0146 - mean_absolute_error: 0.0709 - mean_squared_error: 0.0146 - val_loss: 0.0200 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0200\n",
      "Epoch 580/600\n",
      "3600/3600 [==============================] - 0s 61us/sample - loss: 0.0145 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0145 - val_loss: 0.0210 - val_mean_absolute_error: 0.0809 - val_mean_squared_error: 0.0210\n",
      "Epoch 581/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0146 - mean_absolute_error: 0.0704 - mean_squared_error: 0.0146 - val_loss: 0.0207 - val_mean_absolute_error: 0.0798 - val_mean_squared_error: 0.0207\n",
      "Epoch 582/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0146 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0146 - val_loss: 0.0197 - val_mean_absolute_error: 0.0778 - val_mean_squared_error: 0.0197\n",
      "Epoch 583/600\n",
      "3600/3600 [==============================] - 0s 68us/sample - loss: 0.0145 - mean_absolute_error: 0.0703 - mean_squared_error: 0.0145 - val_loss: 0.0213 - val_mean_absolute_error: 0.0889 - val_mean_squared_error: 0.0213\n",
      "Epoch 584/600\n",
      "3600/3600 [==============================] - 0s 67us/sample - loss: 0.0146 - mean_absolute_error: 0.0705 - mean_squared_error: 0.0146 - val_loss: 0.0241 - val_mean_absolute_error: 0.0997 - val_mean_squared_error: 0.0241\n",
      "Epoch 585/600\n",
      "3600/3600 [==============================] - 0s 63us/sample - loss: 0.0145 - mean_absolute_error: 0.0708 - mean_squared_error: 0.0145 - val_loss: 0.0205 - val_mean_absolute_error: 0.0802 - val_mean_squared_error: 0.0205\n",
      "Epoch 586/600\n",
      "3600/3600 [==============================] - 0s 65us/sample - loss: 0.0145 - mean_absolute_error: 0.0706 - mean_squared_error: 0.0145 - val_loss: 0.0210 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0210\n",
      "Epoch 587/600\n",
      "3600/3600 [==============================] - 0s 53us/sample - loss: 0.0145 - mean_absolute_error: 0.0710 - mean_squared_error: 0.0145 - val_loss: 0.0208 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0208\n",
      "Epoch 588/600\n",
      "3600/3600 [==============================] - 0s 65us/sample - loss: 0.0145 - mean_absolute_error: 0.0706 - mean_squared_error: 0.0145 - val_loss: 0.0202 - val_mean_absolute_error: 0.0813 - val_mean_squared_error: 0.0202\n",
      "Epoch 589/600\n",
      "3600/3600 [==============================] - 0s 64us/sample - loss: 0.0145 - mean_absolute_error: 0.0702 - mean_squared_error: 0.0145 - val_loss: 0.0215 - val_mean_absolute_error: 0.0901 - val_mean_squared_error: 0.0215\n",
      "Epoch 590/600\n",
      "3600/3600 [==============================] - 0s 50us/sample - loss: 0.0145 - mean_absolute_error: 0.0707 - mean_squared_error: 0.0145 - val_loss: 0.0222 - val_mean_absolute_error: 0.0829 - val_mean_squared_error: 0.0222\n",
      "Epoch 591/600\n",
      "3600/3600 [==============================] - 0s 54us/sample - loss: 0.0145 - mean_absolute_error: 0.0706 - mean_squared_error: 0.0145 - val_loss: 0.0198 - val_mean_absolute_error: 0.0771 - val_mean_squared_error: 0.0198\n",
      "Epoch 592/600\n",
      "3600/3600 [==============================] - 0s 78us/sample - loss: 0.0144 - mean_absolute_error: 0.0704 - mean_squared_error: 0.0144 - val_loss: 0.0200 - val_mean_absolute_error: 0.0777 - val_mean_squared_error: 0.0200\n",
      "Epoch 593/600\n",
      "3600/3600 [==============================] - 0s 88us/sample - loss: 0.0144 - mean_absolute_error: 0.0702 - mean_squared_error: 0.0144 - val_loss: 0.0202 - val_mean_absolute_error: 0.0782 - val_mean_squared_error: 0.0202\n",
      "Epoch 594/600\n",
      "3600/3600 [==============================] - 0s 80us/sample - loss: 0.0145 - mean_absolute_error: 0.0703 - mean_squared_error: 0.0145 - val_loss: 0.0203 - val_mean_absolute_error: 0.0793 - val_mean_squared_error: 0.0203\n",
      "Epoch 595/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0145 - mean_absolute_error: 0.0701 - mean_squared_error: 0.0145 - val_loss: 0.0204 - val_mean_absolute_error: 0.0822 - val_mean_squared_error: 0.0204\n",
      "Epoch 596/600\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0145 - mean_absolute_error: 0.0709 - mean_squared_error: 0.0145 - val_loss: 0.0201 - val_mean_absolute_error: 0.0779 - val_mean_squared_error: 0.0201\n",
      "Epoch 597/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0144 - mean_absolute_error: 0.0701 - mean_squared_error: 0.0144 - val_loss: 0.0203 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0203\n",
      "Epoch 598/600\n",
      "3600/3600 [==============================] - 0s 45us/sample - loss: 0.0144 - mean_absolute_error: 0.0706 - mean_squared_error: 0.0144 - val_loss: 0.0202 - val_mean_absolute_error: 0.0799 - val_mean_squared_error: 0.0202\n",
      "Epoch 599/600\n",
      "3600/3600 [==============================] - 0s 46us/sample - loss: 0.0144 - mean_absolute_error: 0.0699 - mean_squared_error: 0.0144 - val_loss: 0.0202 - val_mean_absolute_error: 0.0790 - val_mean_squared_error: 0.0202\n",
      "Epoch 600/600\n",
      "3600/3600 [==============================] - 0s 47us/sample - loss: 0.0145 - mean_absolute_error: 0.0703 - mean_squared_error: 0.0145 - val_loss: 0.0201 - val_mean_absolute_error: 0.0792 - val_mean_squared_error: 0.0201\n"
     ]
    }
   ],
   "source": [
    "history = mod.fit(X_train, y_train, epochs=600,validation_split=0.2, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#pd.concat([y_test,y_rf] ,axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = mod.predict(X_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "ypred = pd.DataFrame(y_pred)\n",
    "ytest = pd.DataFrame(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y1 MSE:0.0018\n",
      "y2 MSE:0.0041\n",
      "y3 MSE:0.0003\n",
      "y4 MSE:0.0870\n",
      "y5 MSE:0.0037\n"
     ]
    }
   ],
   "source": [
    "print(\"y1 MSE:%.4f\" % mean_squared_error(ytest.iloc[:,0], y_pred[:,0])) \n",
    "print(\"y2 MSE:%.4f\" % mean_squared_error(ytest.iloc[:,1], y_pred[:,1]))\n",
    "print(\"y3 MSE:%.4f\" % mean_squared_error(ytest.iloc[:,2], y_pred[:,2])) \n",
    "print(\"y4 MSE:%.4f\" % mean_squared_error(ytest.iloc[:,3], y_pred[:,3]))\n",
    "print(\"y5 MSE:%.4f\" % mean_squared_error(ytest.iloc[:,4], y_pred[:,4])) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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": [
    "x_ax = range(len(X_test))\n",
    "plt.scatter(x_ax[:20], ytest.iloc[:20,0],  s=6, label=\"y1-test\")\n",
    "plt.plot(x_ax[:20], ypred.iloc[:20,0], label=\"y1-pred\")\n",
    "plt.scatter(x_ax[:20], ytest.iloc[:20,1],  s=6, label=\"y3-test\")\n",
    "plt.plot(x_ax[:20], ypred.iloc[:20,1], label=\"y3-pred\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['loss', 'mean_absolute_error', 'mean_squared_error', 'val_loss', 'val_mean_absolute_error', 'val_mean_squared_error'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(history.history.keys())\n",
    "# \"Loss\"\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'validation'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbEAAAGoCAYAAADICdviAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3debgcVZ3/8fc3NysJZCEBYsKWISMEwnpFhEFZNYAGHBEBceIAZkBwYBhAGPwBogOICziKSER2DTAoghJQdhcIQxQJsiUhLAkBckNCQsx+7/n9cfqSTqf7dnV3VZ2u6s/refrpvt1V1V+am/7cc+rUOeacQ0REJIt6hS5ARESkXgoxERHJLIWYiIhklkJMREQySyEmIiKZ1Tt0AQnQcEsRyRsLXUCzUktMREQyK48tMWlm69bBzJnw7LPwwgvw5puwcCGsXAmdndCnDwwduv42bBh84AMwahSMHu3vN9003pq6umDRIl/LW2/525tvQkcHrF4Na9f625o10Ls3DBiw/jZwIGy2Wc+3AQPAiv6Qds4f9+9/h+XL/X33bcWK9beVK/3NzH8uffv6+0039Z/F1lvDFltAW1u8n4dIhlgOL3bO3X9Q5i1fDnfeCb/6FTzyCCxb5p/v0wc23xwGD4Z+/fyX8dq1/sv8vff8fqtWbXy87i/x4mAbNQpGjoT+/X3QdN9WrYKlS+Hdd/394sXrg2rBAn+/cKEP0FL9+/vg6N3b19rW5gNv9er1t3L7pal3bx/yu+8O7e3woQ/BXnvBiBFh65K4qTuxAoWYJGf2bLjiCpg61QfTVlv5L9g994QddvDBU60VsXq1byV1dGx4v2gRvPPO+p+7uqLV1KuXb911t/K6b5tvvuHPw4b5FlQ1a9b41lJ3K6rc/erVvvUF6+/79fMhOWDAhvfdt759198751uw3bfly9d/Fh0dvtX48svw+uvrjz92LBx+OEyYAB/7WLT/FmlmCrEKFGISv/nz4bzzfHj16QOHHOK/THfeecNutbh0dsKSJT7UOjvX39at8yEwcCAMGuTvN9nEB1ke/f3v/g+HF1+Ep5+GZ57xAdq/Pxx4IBxzDHz6077lK1mjEKtAISbxWbcOrrwSvv51//ioo/wX57BhoStrTatX+yB78kl44gnfYuvXz7fQjjsOPvlJtdCyQyFWgUJM4vHKK/D5z/svy/32g9NO8+eopDk45wfSPPwwPPqob7UOHQqTJsG//RvsuGPoCqVnCrEKFGLSuLvugi9+0Xfh/cd/wMEHh65IetLZCX/9K9x7L/zhD77V/NGPwqmnwtFH+8Ei0mwUYhUoxKR+zsFll8EFF8BOO8GFF/rBG5IdS5bA/ffDb37jR2tuuy2cey7867+qq7G5KMQqUIhJfdatg8mT4YYbfMvr3HP9IArJpq4u3xX885/D88/768/OPBNOPz3+6/KkHgqxChRiUrs1a+D44+EXv/DnVCZNSmbUoaTPOT8YZOpU+L//82F2ySVw0knqZgxL/8AqUIhJbVavhn/+Z5g2zQ/eOPro0BVJUl58Ea65xs+wstNO8N3v+ksl9AdLCPrQK1CISXTr1vkh83fdBWedBZ/6VOiKJGnOwR//CFOm+Ov/JkyAa6+FbbYJXVmrUYhVoBCTaLq6fLfhrbf68ySf+UzoiiRNa9fC3XfD9df7bsXvfAe+9CW1ytKjD7oChZhEc/75cPnl/tzICSeErkZCeest+Pa34S9/8QN6rrsOttsudFWtQCFWgUJMqrvhBjjxRN99+B//ob++W51z8Otf+27FXr3gxz/2F7pLkvSPrgKFmPTs0Ufh0EP9LOmXXaYRarLeW2/534mZM/2F0lde6ae1kiQoxCpQiElls2bBPvv4NbF++EM/ia5IsXXrfJfi7bf7pWDuvNNfMC1xU4hVoBCT8t55Bz78Yb/+1tVXax5E6dkf/gDf+pZviU2dCh//eOiK8kYhVkFO16SQhnR2wrHHwrx5/kJXBZhUs//+/tzYkCF+lvwf/zh0RdIiFGKysYsvhgcfhDPOgF12CV2NZMXo0fCDH/jVpU89Fc4+O/pipSJ1UneibOg3v/GjEA8/HM45J3Q1kkWdnf4c6q9+5Wd3ueUWvxipNELdiRUoxGS9uXNhr71gxAj/F7VGmkm9nPNza/7oR37Ax733+t8rqZdCrAKFmHgrV8K++8LLL/vzGR/4QOiKJA/++Ef45jf9iMUHHtB0VfVTiFWgc2LifeUrfqHE889XgEl8/umf/AwfCxb4P5JefDF0RZIzCjGBn/7U3044AT7ykdDVSN6MH+8vhF65EvbbD2bMCF2R5Ii6E1vd00/74Np5Z7jiCmhrC12R5NUbb/jBQu+95ycTPuig0BVliboTK1CItbIlS2DPPWHFCj8P3pAhoSuSvFu0yK8C/sYbcNtt8OlPh64oKxRiFag7sVV1dfnuw/nz4aKLFGCSjuHD4fvfh7Fj/YKq118fuiLJOIVYq/rGN/zqzKefDuPGha5GWsmmm/rBHu3tfmmf73wndEWSYepObEXTpsEnP+lnpz/vPC2tImGsXetnwX/kEf97eOml+l2sTB9MBQqxVqMLmqWZdHbC//wP3HMPTJ7sL47W4KJyFGIVaHGoVrJihZ8GqLPTz4+oAJPQ2trgzDP9cj9TpvjBRrfeCn37hq5MMkIh1iqcg5NP9gsYXnqpLmiW5mHmz41tuilccw0sXeqnrNL6dRKBBna0iosv9us8nXiiX+hSpNkcc4wffv/gg/587eLFoSuSDNA5sVZw880waRJMmOC/JHTyXJrZH//o17EbMwbuuw+23z50Rc1A/2grUIjl3aOP+lV2x4+Hyy+HPn1CVyRS3TPPwIUXQv/+fnmgD30odEWhKcQqUIjl2Ysv+imlhgzxIxF1jkGy5PXX/YTUS5b42T0mTgxdUUgKsQp0Tiyv3nwTjjjCdx1eeqkCTLJnm238H1/bbuunp/rBD/wAJZEiCrE86uiAQw7xQfbNb8LIkaErEqnPsGF+BvyPfAT+/d/9wKRVq0JXJU1E3Yl58847cPDBvivx8sth991DVyTSuK4uP0Dpppv8xfq//GWrLbCp7sQKFGJ5smCBH8Qxe7afG3HvvUNXJBKvxx/33eMDBsDtt7fSci4KsQrUnZgXL7/sV9F95RXfAlOASR7tu6+/IHrQIH8t2UUX+TkYpWWpJZYHjz0Gn/mM/8d8+eWw006hKxJJ1ooVcNVV8MADvnvx5pvzvhqDWmIVqCWWZc75xSwPPRQGDoSrr1aASWvYZBP4r/+Cr3/d90LsuSd873v+3Jm0FLXEsmrJEj/r9513+gtBL7xQw+ilNS1e7APsT3/y3Y0/+hHstlvoquKmllgFCrEsuvtuOO00ePttP+T4c5+DXmpUSwtzDn77W98zsWwZnHKK/8Nuyy1DVxYXhVgFCrEseekl+OpXfYiNGQNnn63uQ5Fi770H118Pv/61n7LqzDP9bfjw0JU1SiFWgUIsC2bNgiuugBtv9OssfeEL8NnPQm+tpCNS1rx5cMMNftXoAQP8Ui9f+Qr84z+GrqxeCrEKFGLNas0auP9+3z0ybZqfuPfII+H442Ho0NDViWTDq6/668kefBDWrYMDDvArOhx1lJ9TNDsUYhUoxJrJO+/AQw/55SfuvtsP3hg2zE98+qlP+cciUrvFi/0fhffe6ycF6NPHz2wzYYKfom3HHf0q081LIVaBQiyE1ath/nyYMwf+8hd/+/Of/YXK4Fe4/fCH/T+uvfZSt6FIXJzzU7I98ghMn+67HcGP7N1zT2hv9//mPvhB2G47/4djc6y/1xRFNCOFGMDf/+4DxDl/nUnxfbnnil9bs8aH0qpVG96vXOmXWX/3XX9bssRPyDtvnh9VWGz0aPiHf/D99bvtloW/CkXy4a234K9/9YOmZs3yf1iuWbP+9UGD/Cz6o0f7QBs6dP39gAF+8EjprW9fP1q4+9bWtv7xsGH+WLVTiFWgEAO/kuz++ydQCrDZZv42aJAfITVyJGy1lb8fNcoH1uDByby3iNRm7VqYO9evZfbGG77r8Y03/MoQy5b5P0iXLav/oupJk/wArdopxCrIXYiZ2f1AM4ynHQ4sCl1EBFmpE1RrUlRrMuKsdZFzbkJMx8qV3IVYszCzGc659tB1VJOVOkG1JkW1JiNLtWaZpnkQEZHMUoiJiEhmKcSSMyV0ARFlpU5QrUlRrcnIUq2ZpXNiIiKSWWqJiYhIZinEREQksxRiIiKSWQoxERHJrNyF2IQJExx+6inddNNNt7zcIsvpd2BFuQuxRYuyMiONiEj8Wu07MHchJiIirUMhJiIimaUQExGRzFKIiYhIZinEREQksxRiIiKSWQoxERHJLIWYiIhklkJMREQySyEmIiKZpRATEZHMUoiJiOSJq2m+4MzrHbqApjF/fugKRCTvRo9O/j3efjv592giaomJiORJi7XEFGIiInnS1RW6glQpxERE8kQtMRERySyFmIiIZJZCTEREMkshJiIimaUQS4+ZXW9mC83sbxVe/7yZzSzcHjez3dKuUUQkUzo7Q1eQqtAtsRuBCT28/grwMefcrsA3gClpFCUiklktFmJBZ+xwzv3ezLbr4fXHi36cDqRwubuISIa1WIiFbonV4iTgvnIvmNlkM5thZjM6OjpSLktEJKzi78DONWtCl5OqTISYmR2ID7GvlnvdOTfFOdfunGsfMWJEusWJiARW/B3Y1tXVUoM7mj7EzGxX4DrgSOfcO6HrERFpal1dsHRp6CpS09QhZmbbAL8EvuCcmxW6HhGRTGihVTmCDuwws6nAAcBwM5sPXAT0AXDO/Ri4ENgc+JGZAaxzzrWHqVZEJCPmzYNddgldRSpCj048rsrrJwMnp1SOiEg+vPJK6ApS09TdiSIiUqNevWDmzNBVpEYhJiKSJ/36wdNPh64iNQoxEZE86d/ft8TWrQtdSSoUYiIiedKvH6xaBbNaY0C3QkxEJE/69/f3v/992DpSohATEcmTvn1hyy3hwQdDV5IKhZiISN60t8PvfgerV4euJHEKMRGRvPnoR+G99+CBB0JXkjiFmIhI3uyxB2y2Gfz856ErSZxCTEQkb/r0gYMPhl/8At7J97zpCjERkTz65CdhzRq4/vrQlSRKISYikkdjxsCee8KVV+Z6gEfQCYBFpEVMn97z6/vsk04dreb44+Hss+Gmm2Dy5NDVJEIhJiLxqRZWteynYGvcnnvCuHHw9a/D5z8PAweGrih26k4UkcZMn77+loXjthIzOOUUWLDAdyvmkFpiIlK7tIOl+/3UOqvd+PGw//5w2WVwwgmw3XahK4qVWmIiEl3ollHo98+q008H5+DLX/b3OaIQE5Hqmi08mq2eZrfFFnDiiXDffXDzzaGriZVCTEQqa/awaObams2nPw277gpf+Qq88kroamKjEBORjTV7eBXLSp2htbXB+edDV5cfqbh2beiKYqEQE5ENJREKjz++4S1uCrJottoK/vM/4Ykn4JxzQlcTC41OFBEvjiCIGlCVttt33/rfe/p0jV6M4sAD4bnn4Pvfh7339hdEZ5hCTEQaC7A4W1bFx6on0BRk0fzbv8Hs2X6wx7bbwn77ha6obupOFGl19QRYkl2Dpe9RK3UtVtenD1xyiR+1eNRR8PLLoSuqm0JMpJXV+oWfdHBVek+J3+DBcOmlfqb7I46AxYtDV1QXhZhIq6olwEKEVyPvr9ZYNKNHwze+AXPnwj//cyZnu1eIibSiWgOsWTRTLXmx665+pOJjj8G//At0doauqCYa2CHSaqIGWFrno2odiPH449EGfWiQR3SHHupXgL72Wt/NeO21fvLgDFCIibSSJAKs0a47Te7bHI49Ft57D37yExg6FL71rdAVRRK0O9HMrjezhWb2twqvm5n9j5nNMbOZZrZn2jWK5EbcARb3rB61HK+WGiW6k0+GiRPhiisyE2KhW2I3Aj8EKs1IeRgwtnD7MHBN4V5EahHly7xZgiFqyyxqt6JEZwZnnAHLl8N558GQIf6asiYWNMScc783s+162ORI4GbnnAOmm9kQMxvpnHszlQJF8iCuAAuxhpi6GNPXq5efY3HFCjj1VH+O7NhjQ1dVUbOPThwFzCv6eX7huQ2Y2WQzm2FmMzo6OlIrTqTpxRFgIScDrva+zRi+AWzwHbh0aeMH7N0bLr7Yj1z8whdg2rTGj5mQ0N2J1ZQbHrPRim7OuSnAFID29vZ8rfgmUq+4Aiyi5TNnRt520K67Rt5WqtvgO/CDH4znO7BfP/jv/4azzoKjj4YHH2zK7ttmb4nNB7Yu+nk0sCBQLSLZkVKALZ858/1bLWrapwVaUk1r4EC4/HLYfHM/q8ffyo7BC6rZQ+we4F8KoxT3AZbqfJhIFSkEWD3BldhxdAF0soYO9aMV29rg4x+H114LXdEGQg+xnwo8AXzQzOab2UlmdoqZnVLYZBowF5gD/AT4cqBSRfKjpy/9Kue/4gqvcsftkVpjYY0c6YNs+XJ/YXQTjT0IPTrxuCqvO+C0lMoRyb5GBkJEaH0lafnMmcmdK9NIx8aNGePPkZ1zju9afOwxGDAgdFVN350oIlE10loJ0PqSDBo/Hi64AJ56yq9F5sKPo1OIieRBI+fBqgRYVM/+/e9lb7Xo8f3Updgc9t8fvvQluO02+OY3Q1fT9EPsRaSagAEWJaS6txk/cGDVbSUjjjsOXn0VLrwQdtwRPvvZYKUoxESyLFCA1drCKt5HYZYDZnD22fDmmzBpkj9fttdeQUpRd6JIViUQYNXOf9XTRSg51bcvXHIJbLYZHHOMnwE/AIWYSBYlFGCVxBleCsEcGTrUD/R49VU4LcxAcoWYSNbEOSN9QbUAyzwNr0/O+PF+RehbboGf/Sz1t1eIiWRJowFWZv8QAZaLYJT1TjjBh9mpp8Lcuam+tUJMJCtyEmB1U2uqebW1wX/9l79u7IQToKsrtbdWiIlkQYoBpsEbUpettvLnxZ54Am64IbW31RB7kWaXcoBFKqnC80HaSk24PEjL+sQn4L774Nxz4aij/Oz3CVNLTKSZJRBglUQJsOlUDjCqvBaMuiHTYwZnnAFLl/rVoVOgEBNpVgkFWLlWWLUAqxZepdvGTkGUHWPGwGc+Az/5SSpThSnERJpRkwVY3CrN2lHzLPbqSmxOkybB8OHw1a8mPkmwQkyk2eQ8wFKlFlwYm2wCxx8Pv/89PPRQom+lEBNpJk0SYLV0HyZOQZRNRxwBW2wBX/taoq0xhZhIs2iiAEuSuhJbRN++8PnPw5NPwv33J/Y2CjGRZtAiAVazRlphasGFd9hhMHIkXHRRYq0xhZhIaAldB9aMAaZWWIvp0wc+9zm/EvSf/pTIWyjEREJqkguZ4wyw2No/lVpSUQJMrbDm8YlP+OVavvvdRA6vEBMJpQkCrBkGcNTcCpNs6d8fJk6Eu++GOXNiP7xCTCSERtcDiynA0lSuK7FigKkVli9HHQW9e8NVV8V+aIWYSNpSXNAy7QCrFB+VzoWVP4gGc+TO5pvDQQfBTTfFvgK0QkwkTSkFWE8z0YfuPuymwRwtZuJEWL489oUzFWIiaUkxwCqWUL2CuiXaClM3YvbttBPssANcc02sw+0VYiJpqBZgjz+eaIA1wwCOYmVbYepGzDcz3xqbOdNfAB0ThZhI0qIEWA371noNWBrhVUsrTN2ILezgg/28itdcE9shFWIiSUogwMqJI8B6qKQu6kaUjWyyCRxyCNx+OyxeHMshFWIiSUkhwCoN4Ijaffh40a3cz1HUEiE1dSMqwPLpU5+C1avhxhtjOVzQEDOzCWb2kpnNMbPzyry+jZk9YmZPm9lMMzs8RJ0iNUspwMru3vM7+7cn/pZXscjdiDoP1np22AHGjYPrr49lgEewEDOzNuBq4DBgHHCcmY0r2exrwB3OuT2AY4EfpVulSB3qXc22houY6w2wuMOrXIzoPJhU9YlPwHPPwYwZDR8qZEtsb2COc26uc24NcBtwZMk2Dtis8HgwsCDF+kSSUa4V1uAIxGrdh0m0vBpuB6kbsXUddBD06wc33NDwoUKG2ChgXtHP8wvPFbsYOMHM5gPTgK+UO5CZTTazGWY2o6OjI4laRaJppBuxRC0BVvHtSLbbsFTD3YgKsLps8B24dGnocqobNAj23x+mToVVqxo6VMgQszLPlXaQHgfc6JwbDRwO3GJmG9XsnJvinGt3zrWPGDEigVJFIqg3wCJ0IfY0gKPi2/VcTUMa6kbUebDYbfAdOHhw6HKimTAB3n0XfvWrhg4TMsTmA1sX/TyajbsLTwLuAHDOPQH0B4anUp1ILRIOsI12o3KApd36ik21VpgCLF/22AO22soP8GhAyBB7ChhrZtubWV/8wI17SrZ5HTgYwMx2woeY+guluQQIsLJvQ3zh1VOcJNIK00CO1tOrF3z84/DggzB/fv2HibGkmjjn1gGnA78FXsCPQnzOzC4xs4mFzf4T+JKZPQNMBb7oXEJrXIskIcUAS0OwAFMrLJ8OOcQPs7/zzroP0TvGcmrmnJuGH7BR/NyFRY+fB/ZLuy6RyOoZTp/RAIuqpuH0CrDWtvXW/rqx22+HM8+s6xCasUOkXvV0I8YYYEmd+6oUK1FbYeV3rjOIFGD5d8AB/t/F66/XtbtCTKQeTRBgoSXejagAaw0HHODv6+xSVIiJ1Cqma8GSDrA4l1+JEicKMKnLqFHwD/8Ad99d1+4KMZE4RRzIUWk2+g12KXf4CPuUhlctQVYuWhrqRiz7JhqJKCU+9CF44gl4772ad1WIidQipoEcpUpbYbUGWLVWV5xriiV+UbNaYa2nvR3WroXHHqt5V4WYSFQpnQerJ8CiqLZd1FZYKXUjSsPGj/dzKf7udzXvGnSIvUhmpHQerJYAS2PF5lKRuhEVYFKrvn1ht93qCjG1xESqiWlGjtAB1lNERGmFNbTEigJMqhk3DmbNqvm8mEJMpCcxTilVq5AtsNi7EUWqGTvWz94RYdBTMYWYSCWNLG5ZotZWWJoBFmWsYGkrTOfBJHZjx/r7p5+uaTeFmEg5UQIsoYEcoc+BRelG3HgnBZg0aPhwGDJEISbSsHoDrIx6zoOVLSnidpVUiop6rtjSeTBJhBlsvz288EJNuynERIo1EmAJnQdLqgVWz2CO2M6DKcCknBEjal6WRSEm0i3mAIurG7FRUeMi1W5EkXKGD4c334Sursi7KMREIHiAVSwr4na1iqUbUefBJG4jRsC6dbBwYeRdFGIiKQdY2cOXO3T1qqqKqxWm82CSiuHD/f0bb0TeRSEmrS3hACu7W+nhq1dQl6iDOWI7D6YAk0YNGuTvly2LvItCTFpXCgHW7N2IdcWKwkiS0q+fv1+5MvIuCjFpTSmMQqy3GzEO5WImlouadR5MktQdYitWRN5FISatJ8brwLrVs8BlUhc16zyYZJZCTKSKRgMspvNgSdF5MMm0Pn38/Zo1kXfRUizSOhLoQqz3PFgSrbB6AyzawXVBs6TAOX9vFnkXtcSkNWQgwBrRyLRSiZ0HE6lVd4j1ih5NCjHJv0ABlpZaAkznwaSpKcRESgQMsFpbYXEvdFltW50Hk6bTPd2UuhNFImqSAKtHLSs1K8AkU9QSEyH2VZnzEmDR3kBhJAHU0RLT6ETJpyYKsDhVi5YoAaYLmiVPgrbEzGyCmb1kZnPM7LwK2xxjZs+b2XNm9vO0a5QMarIAi6MVtg8pBVjFgyvAJAXdLbDuAR4RBGuJmVkbcDVwKDAfeMrM7nHOPV+0zVjgfGA/59wSM9siTLWSGTkLsKjREMtQetB5MAkrSyEG7A3Mcc7NBTCz24AjgeeLtvkScLVzbgmAcy76IjPSemIMsEqzcCQZYPVEQaV40UAOybSMhNgoYF7Rz/OBD5ds848AZvYnoA242Dl3fzrlSaYkHGCVrgFLsguxmsQDrBoFmMQtYy2xcsNPSivvDYwFDgBGA38ws12cc+9ucCCzycBkgG222Sb+SiWfGgiwniKzmQOs/E6akSPrNvgO3HLLwNU0oHtofQ0hFnJgx3xg66KfRwMLymxzt3NurXPuFeAlfKhtwDk3xTnX7pxrHzFiRGIFS5OqpxWWQIA9TvMHmEYi5tMG34GDB4cup3GdnZE3DRliTwFjzWx7M+sLHAvcU7LNr4ADAcxsOL57cW6qVUpzq7cbsUToARxRKcAk17qXYlm1KvIuwULMObcOOB34LfACcIdz7jkzu8TMJhY2+y3wjpk9DzwCnOOceydMxdJ0Yh6JWCz3AVbxTRRgEtCAAf5++fLIuwS92Nk5Nw2YVvLchUWPHXBW4SayXoIDOVoiwDQSUZpRv35+cEcNIaZppyR/FGDvU4BJppjBJptADStCaNopyZ4oM9NH2KfeAEsrvCDacirQYIBVowCTNA0cCO++W327AoWYZEtCIxGTCLDi/eO6kDmRANNQemkmQ4bA229H3lwhJtmRgQCrVOF0aguypgkwtcIkbUOHKsQkhxIaSh9XgMU5e33UdpECTHJp6FD4298ib64Qk+YX00jESvMhbrBLucPXuH0jUhnEAQowaV5Dh8LChX7Wjgjriml0omRbAwFW2grLSoCVowCT3Bg6FNasiTy4QyEmzS2mkYil0g6wWtcC62m/hib0VYBJsxsyxN9HPC+mEJPmldJAjloCbHqF7ZOgAJOWNGyYv1eISaY1aYAloZGRiNHfRMPoJSOGDvX3CjHJrJRGIqYVYD21b5riYuZG9hOJm0JMMi3FkYgbHbpSSTUfKRpdCyZSxmab+XXFFi6MtLlCTJpHirPSl+6RVIA1Gg8KMGk5bW0weHByIWZmN9dclEg1Kc5Kn1aA9SRKK0wBJi1ryJDIIdbjxc5mVrpIpQEHmtkQAOfcxI33EqlRPcPoK+xXz3mwsoeur6INVIqIqN2IpRRg0jJqmHqq2owdo4HngesAhw+xduC7jdQn8r4oAZbySMS0htB301B6kRJDhsCrr0batFp3YjvwZ+ACYKlz7lFgpXPuMefcY43UKFJ3gJUR10COuAKsllZYKQWYtLzuqaci6LEl5pzrAq40s/8t3L9dbR+RSBoJsBimlEpzTbBu9Z4H23gnLakiObfppituKYkAABh1SURBVH51585OP9CjB5ECyTk3H/ismR0BLIuhRGllMQZYOSHPg0H0EYl1D+QoJ0qAqRUmWdG/v79fscIHWg9qalU55+4F7q23LpG4A6zZzoNF7UaMfSBHNQowyZIaQkzXiUl6Ug6wsocvd+jqVTVE58FEatSvn7+P8G9aISbpCBBgaZ8HKxcV9ZwHU4BJy+sOsRUrqm6qEJPkJRxgZXer/o41bVdNXOfBFGAi+AUxwU8/VYVCTJKVQoA1ujZYo+o9D5boSEQFmGRZd4hVGZkICjFJUpMHWJqzctQ9kKPswRVgknNdXf5eLTEJJkCAlX2LCs+nPa1Uqbq7EXUtmLSCzk5/r5aYBJHwdWAQbSBHkhqZF7Gh82DVqBUmebB6tb8fMKDqpgoxSV+NAVbPSERIf3kVjUQUicl77/n77gUye6AQk3jFvCZY3AHWKAWYSAqWL/cXPHdf9NyDoCFmZhPM7CUzm2Nm5/Ww3dFm5sysPc36pEb1BlgFSQRYI62wOAOs/BsowEQA3xIbMiTSpsFCzMzagKuBw4BxwHFmNq7MdpsC/w48mW6FUpNGAizha8HiEHeAbdQKU4CJrLdsGQwbFmnTkC2xvYE5zrm5zrk1wG3AkWW2+wZwBbAqzeKkBikEWNRJfZNohSUeYJUowKRVdXTANttE2jRkiI0C5hX9PL/w3PvMbA9ga+fcb3o6kJlNNrMZZjajo6Mj/kqlsgABVvGtIm0V3T4EXJ1ZQ+mlBht8By5dGrqcxi1cmIkQszLPufdfNOsFXAn8Z7UDOeemOOfanXPtI0aMiLFE6VHE4fBR940aYPUM5Ki10p4CKWqAJXotmFphUmSD78DBg0OX05jVq2HJkkyE2Hxg66KfRwMLin7eFNgFeNTMXsV/T9yjwR1NItDFzGmMRAwaYFEowCTPuld0jhhiIVdpfgoYa2bbA28AxwLHd7/onFsKDO/+2cweBc52zs1IuU4pldMAqxYNqQSYzoNJq5s/39+PGRNp82AtMefcOuB04LfAC8AdzrnnzOwSM5sYqi6JQUoBVouevvp7OvfVTQEmkpLXX/f3O+0UafOQLTGcc9OAaSXPXVhh2wPSqEmqiPFi5kYDrNFWWNRIUICJpOi112DEiMhD7IOGmGRMxgOsnhhQgImkbN48GLfRJcMVadopiSbjAVYPBZhIypzzLbGIXYmglphE0WIBVilWFGAiCVuwwE85tccekXdRS0ySoQAr7KgAE4ls1ix/v9dekXdRiEnPYp7Ut5gCrNIbKsCkRc2eDX36wC67RN5FISaVJdiNqACr9IYKMGlhs2b5AOvXL/IuCjEpL+HzYFEpwERahHM+xNprm5RJAzskPgEvZq5HI+EFMQeYSKt76y0/qKOG82GglpiUU08rLGMDOZouwNQKk1ZXx6AOUIhJqZgGcijAut9MASYSyaxZflDH+PE17aYQk+hqOA9WSgFW6Y0VYCKAD7Gdd65pUAcoxKRYPeuDZWgkogJMpEnVOagDFGLSLcHzYBvtVuktqu5ZPwWYSBN7+21Ytqzm82GgEJMoYjwPpgDrYV+RVjV7tr/fc8+ad1WISWzdiKXKdSOWowATaXGzZ0NbW82DOkAhJgl1I4ZYmbmUAkwkI2bP9jPXDxhQ864KMamsgW7EUmlfzKwAE8mQOXPq6koEhVhrS6kbMe3zYAowkQxZvBgWLapp+ZViCrFWlWI3YtnDR9qqdgowkYzpHtShEJPYJNyNqAATkfe9/LK/3333unZXiLWigN2ICjAR2cDrr8PIkTB4cF27K8RaTeBuxDQpwEQyYN48PzKxTgoxWS9H3YgKMJEMcM6H2Ac/WPchFGKtpKdWWIyT+4YOsHIUYCJN6N13/RpiCjGpKsXJfdMS9TyYAkykSb3+ur/fcce6D6EQawUJrRHWDLNylIoSJQowkSYxb56/V0tM6haxGzEv58FqCrAoFGAi9VuwwC+Euc02dR9CIZZ3gSf3TUIjAzlqe6MqrTAFmEhjOjpg1CjoVX8UKcTyLMeT+1aT+HkwBZhI4xYtgq23bugQQUPMzCaY2UtmNsfMzivz+llm9ryZzTSzh8xs2xB15lLOh9OXUoCJNKHullgDgoWYmbUBVwOHAeOA48xsXMlmTwPtzrldgTuBK9KtMsPUjfg+BZhIE3LOh9jo0Q0dJmRLbG9gjnNurnNuDXAbcGTxBs65R5xzKwo/Tgca+69tFTnsRqz3ejAFmEiTWrYM1qzJdIiNAuYV/Ty/8FwlJwH3lXvBzCab2Qwzm9HR0RFjiTmU0W7EcqJcD7bxTg0MpRdpUht8By5dGrqcaLq/qzMcYlbmOVd2Q7MTgHbg2+Ved85Ncc61O+faR4wYEWOJGdTC3YilyrbCyr6BrgWTbNvgO7DOiXRTF1OI9Y6hlHrNB4qHpYwGFpRuZGaHABcAH3POrU6ptmxqkW7E1K8HU4CJxC8HLbGngLFmtr2Z9QWOBe4p3sDM9gCuBSY65xYGqDE/ctSNWCrRbkQFmEgyOjqgrQ223LKhwwQLMefcOuB04LfAC8AdzrnnzOwSM5tY2OzbwCDgf83sr2Z2T4XDScA1wpISpRUW6XowBZhI83nrLT+8vndjHYIhuxNxzk0DppU8d2HR40NSLyqLAq8RFrIbsVRs58EUYCLJevtt2G67hg+jGTvyLuFuxCQkPpxeASYS3sKFCjEh+GjEuFthjSyvsvFOCiORprR2rT8ntm3jkzApxPIsJ0usqBtRJGdeew26umBc6SRNtVOIZVlCrbCNdqn9XeqS+HB6BZhIc5gzx9/vsUfDh1KI5VXGBnOkMq1UTxRgIumZPRs22QR22KHhQynEsqqeVliJZhnMkcp5ME0rJdI85syBXXf114k1SCGWRxFbYaVCtMKiBlg5Og8mkkFr1sCsWfChD8VyOIVYFuWoFVaOzoOJ5Nizz8KqVXDoobEcTiGWNxFHJJZqllZYlAArS+fBRLJh+nTo0wcOPDCWwynEWkETjkisdyAHxDitlIikq7MTHn0UDj8cBg2K5ZAKsazpKZBivC4sSToPJtKi/vIXWLQIjj8+tkMqxPIuhvNn3eLoSqwlwDScXiRnfvELGDECJk6svm1ECjHZSFJdiY0EWPkd6+hGVICJhPHaa/Dkk3DaadC/f2yHVYhlSUqjEpPQaIDFukqziKTvhhv8ebAvfznWwyrE8iLhUYmNSCTANJxeJDuefx4eewzOPtt3J8ZIIZZnMZ4Pg+gjChvdp5jOg4lkXGcnfP/7fgXns86K/fBBF8WU/KoWXrEP5FA3okhzuvNOP0PHHXfAppvGfni1xCR2cQRY+R3rDDC1wkTCePllfy7syCPh6KMTeQu1xLKiSQZ1dMdF6Rm4qO2g1AdyKMBEwli5Ei65BIYOhSlTwCyRt1GISV3q6bzTQA6RFtHVBVdcAfPnw0MPwRZbJPZWCjFJXKUoaSjARKR53Xijn17qW9+CAw5I9K10Tkw2EmdkJBZgaoWJNKf774dbboETT4Rzzkn87RRiLSTyeaaYKMBEWswf/gDf+Q4cdBBcc01i58GKKcRaXKVRgY3EwD497K8AE8mpGTPgG9+AvfaCu++Gvn1TeVudE5OKuuMg6rjIavHR0ChE0LVgIs3qqafg//0/2HFHuO++2JZZiUIhJowfOLDH6ad6CrOo7Z6aAqze1pRaYSLpe/JJuPBCH2APPQTDhqX69gqxrNhnn1iuFRu06651Xy9WT0RU6q6sOcDUjSjSfB5+GC67DHbdFR54IPUAA50Ty7cavtgjzZhRIwWYSI7ddRd885vwkY8EaYF1U0tM3letW7GW41SiABPJOOf8VFK33OKnk5o6FQYMCFZO0JaYmU0ws5fMbI6ZnVfm9X5mdnvh9SfNbLv0q8yIGgY99DSYYvzAgXW3yqrtqwATybjOTvje99ZfB3bnnUEDDAKGmJm1AVcDhwHjgOPMbFzJZicBS5xzOwBXAt9Kt8ocqPBFX21UYNQw696uWnjFHmAikq41a+DrX4ff/AbOPx+uuw56h+/MC1nB3sAc59xcADO7DTgSeL5omyOBiwuP7wR+aGbmnHNpFto0Yhrc0S3KII9Gz5XVNQIxSoCpFSaSnvfeg699DZ59Fq66Cs44I3RF7wvZnTgKmFf08/zCc2W3cc6tA5YCm5ceyMwmm9kMM5vR0dGRULkZUOnLv4cv/KRm8air9QUKMJE6bPAduHRpvAdfuNCH1osvwm23NVWAQdgQKzcfSWkLK8o2OOemOOfanXPtI2Je+jo3UgqyHsOrSh0KMJH6bPAdOHhwfAd+5RU4/XR45x0/J+Ixx8R37JiEDLH5wNZFP48GFlTaxsx6A4OBxalU16yqfYn3FARVgqyRMIu0vwJMJDueeca3unr39nMiHnhg6IrKCnlO7ClgrJltD7wBHAscX7LNPcAk4AngaODhlj0fFpcq59VKg6jSObOaAq+R4I16DBGJz2OPwaWXwpgxvgW27bahK6ooWIg559aZ2enAb4E24Hrn3HNmdgkwwzl3D/BT4BYzm4NvgR0bqt6mUm2Ax777wuOlay/XsH+RhrsaFWAi2XLXXfCDH/iLmO+5BzbfaBhCUwk6PtI5Nw2YVvLchUWPVwGfTbuuXIgxyOoSR3hFOY6IxOemm/yClhMn+kEcga8Bi0LTTmVVlC/3KBcLxxkS3cdTgIlkS/csHDfeCJMmwS9+kYkAA4VY/kXtrmskMGrZXwEm0ly6A+zmm/0sHNdf3xQXMUeVnUplY1G7BKt1LRYfLym1zMChABNJh3M+tG69FU4+Ga69Fnplq22TrWplY3G3gJKgABNpTj/9qQ+wL30pkwEGaom1lu4widIqi/P9olKAiaTnjjvgZz+DyZPhmmsyGWCgEMuHWkcaRu1erEc9LT6Fl0i6Hn7YB9fRR8OPfpTZAAOFWH7UE2Td4gi0ersrFWAi6Xr+ebj8cvjoR/2SKm1toStqiEIsT+q99qvWQIvr/JoCTCRdixfDRRfBqFH+oub+/UNX1DCFWN40ehFzGgNAFF4i6evshEsugRUr4KGHYNiw0BXFQiGWR90hkeSMHPVQeImEM3Wqn9T3ppsgoSWYQsju2TyprplCo5lqEWk1s2b58Prc5+ALXwhdTazUEsu70K0yhZdIWJ2dcOWVfiLfa64BK7dMY3YpxFpF0hP+lr6XiDSHe+/1qzL/7GcwdGjoamKnEGslxeGSRKApvESay8qVflLfj30MjjsudDWJUIi1qtLAqSfUFFoize2Xv4QlS/x1YTnrRuymEBNPgSSSL6tX+6mljjgi1/++NTpRRCSPHnwQli2Dc88NXUmiFGIiInl0990wfjzsv3/oShKlEBMRyZtXX4XZs/0SKzk9F9ZNISYikjcPP+xnpj/mmNCVJE4hJiKSN0884bsRt9wydCWJU4iJiOTJunUwZw5MmBC6klQoxERE8mTlSn9/6KFh60iJQkxEJE9WroQBA2C33UJXkgqFmIhInqxaBXvtBb1bYy4LhZiISJ6sWZOr9cKqUYiJiORJVxfsvHPoKlKjEBMRyZsxY0JXkBqFmIhI3my9degKUhMkxMxsmJk9YGazC/cbrdRmZrub2RNm9pyZzTSzz4WoVUQkc0aPDl1BakK1xM4DHnLOjQUeKvxcagXwL865nYEJwFVmNiTFGkVEsqdXL9hss9BVpCZUiB0J3FR4fBNwVOkGzrlZzrnZhccLgIXAiNQqFBHJoj59cj/pb7FQIbalc+5NgML9Fj1tbGZ7A32Blyu8PtnMZpjZjI6OjtiLFRFpZsXfgeucC11OqhK7Gs7MHgS2KvPSBTUeZyRwCzDJOddVbhvn3BRgCkB7e3tr/R8UkZa3wXfgZpu11HdgYiHmnDuk0mtm9raZjXTOvVkIqYUVttsMuBf4mnNuekKliojkR1tb6ApSFao78R5gUuHxJODu0g3MrC9wF3Czc+5/U6xNRCS7FGKpuBw41MxmA4cWfsbM2s3susI2xwAfBb5oZn8t3HYPU66ISEb0aq3Lf4PMEOmcewc4uMzzM4CTC49vBW5NuTQRkWxroZGJoBk7RETyRSEmIiKZpRATEZHMUoiJiEhmKcRERCSzWizEWmP96ihaaNZnEcmxYcNCV5AqtcRERPKkb9/QFaRKISYiIpmlEBMRkcxSiImISGYpxEREJLMUYiIiklkKMRERySyFmIiIZJZCTEREMkshJiIimaUQExGRzFKIiYhIZinEREQks8w5F7qGWJlZB/Ba6DqA4cCi0EVEkJU6QbUmRbUmI85aFznnJkTZ0Mzuj7ptHuQuxJqFmc1wzrWHrqOarNQJqjUpqjUZWao1y9SdKCIimaUQExGRzFKIJWdK6AIiykqdoFqTolqTkaVaM0vnxEREJLPUEhMRkcxSiImISGYpxOpkZsPM7AEzm124H1pmm93N7Akze87MZprZ54peu9HMXjGzvxZuuydQ4wQze8nM5pjZeWVe72dmtxdef9LMtit67fzC8y+Z2Sfirq2OWs8ys+cLn+NDZrZt0WudRZ/jPU1Q6xfNrKOoppOLXptU+J2ZbWaTAtd5ZVGNs8zs3aLX0v5MrzezhWb2twqvm5n9T+G/ZaaZ7Vn0WpqfabU6P1+ob6aZPW5muxW99qqZPVv4TGckWWdLcc7pVscNuAI4r/D4POBbZbb5R2Bs4fEHgDeBIYWfbwSOTrC+NuBlYAzQF3gGGFeyzZeBHxceHwvcXng8rrB9P2D7wnHaAtd6ILBJ4fGp3bUWfl6e4v/3KLV+EfhhmX2HAXML90MLj4eGqrNk+68A14f4TAvv91FgT+BvFV4/HLgPMGAf4Mm0P9OIde7b/f7AYd11Fn5+FRie5ufaCje1xOp3JHBT4fFNwFGlGzjnZjnnZhceLwAWAiNSqm9vYI5zbq5zbg1wG77mYsX/DXcCB5uZFZ6/zTm32jn3CjCncLxgtTrnHnHOrSj8OB0YnWA9PYnyuVbyCeAB59xi59wS4AEgqZkVaq3zOGBqQrVU5Zz7PbC4h02OBG523nRgiJmNJN3PtGqdzrnHC3VA2N/TlqEQq9+Wzrk3AQr3W/S0sZntjf+L+OWip/+70O1wpZn1i7m+UcC8op/nF54ru41zbh2wFNg84r5xqvX9TsL/Vd6tv5nNMLPpZrbRHxMxi1rrZwr/b+80s61r3DcOkd+r0DW7PfBw0dNpfqZRVPrvSft3tRalv6cO+J2Z/dnMJgeqKXd6hy6gmZnZg8BWZV66oMbjjARuASY557oKT58PvIUPtinAV4FL6q9247ct81zp9RSVtomyb5wiv5+ZnQC0Ax8renob59wCMxsDPGxmzzrnXi63fwyi1PprYKpzbrWZnYJv7R4Ucd+41PJexwJ3Ouc6i55L8zONoll+VyMxswPxIfZPRU/vV/hMtwAeMLMXCy07aYBaYj1wzh3inNulzO1u4O1COHWH1MJyxzCzzYB7ga8VukG6j/1moWtkNXAD8XfXzQe2Lvp5NLCg0jZm1hsYjO8qibJvnCK9n5kdgv8DYmLhcwPe76rFOTcXeBTYI2Stzrl3iur7CbBX1H3TrLPIsZR0Jab8mUZR6b8n7d/VqsxsV+A64Ejn3Dvdzxd9pguBu0i2i751hD4pl9Ub8G02HNhxRZlt+gIPAWeWeW1k4d6Aq4DLY66vN/4k9/asP7G/c8k2p7HhwI47Co93ZsOBHXNJdmBHlFr3wHfFji15fijQr/B4ODCbHgYwpFTryKLHnwamFx4PA14p1Dy08HhYqDoL230QP+DAQn2mRe+7HZUHTBzBhgM7/i/tzzRindvgzyHvW/L8QGDTosePAxOS/kxb4Ra8gKze8OeOHir8A3+o+x8OvqvrusLjE4C1wF+LbrsXXnsYeBb4G3ArMCiBGg8HZhW+/C8oPHcJviUD0B/438I/uv8DxhTte0Fhv5eAw1L4PKvV+iDwdtHneE/h+X0Ln+MzhfuTmqDWy4DnCjU9AuxYtO+Jhc97DvCvIess/HwxJX9ABfpMp+JH767Ft65OAk4BTim8bsDVhf+WZ4H2QJ9ptTqvA5YU/Z7OKDw/pvB5PlP43bgg6c+0VW6adkpERDJL58RERCSzFGIiIpJZCjEREckshZiIiGSWQkxERDJLISZSJzNrM7Onzew3oWsRaVUKMZH6nQG8ELoIkVamEBOpg5mNxs8icV3oWkRamUJMpD5XAecCXdU2FJHkKMREamRmnwQWOuf+HLoWkVanEBOp3X7ARDN7Fb/Y5EFmdmvYkkRak+ZOFGmAmR0AnO2c+2ToWkRakVpiIiKSWWqJiYhIZqklJiIimaUQExGRzFKIiYhIZinEREQksxRiIiKSWQoxERHJLIWYiIhk1v8HHSuvxgBJhOAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_corr(ypred,ytest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "scrolled": true
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.252721</td>\n",
       "      <td>0.599728</td>\n",
       "      <td>0.512525</td>\n",
       "      <td>0.981790</td>\n",
       "      <td>0.657591</td>\n",
       "      <td>0.231055</td>\n",
       "      <td>0.576148</td>\n",
       "      <td>0.487459</td>\n",
       "      <td>0.518398</td>\n",
       "      <td>0.572734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.584286</td>\n",
       "      <td>0.358530</td>\n",
       "      <td>0.932715</td>\n",
       "      <td>0.567207</td>\n",
       "      <td>0.325643</td>\n",
       "      <td>0.601309</td>\n",
       "      <td>0.409537</td>\n",
       "      <td>0.925475</td>\n",
       "      <td>0.435044</td>\n",
       "      <td>0.311838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.248941</td>\n",
       "      <td>0.529385</td>\n",
       "      <td>0.143769</td>\n",
       "      <td>0.492893</td>\n",
       "      <td>0.188981</td>\n",
       "      <td>0.189050</td>\n",
       "      <td>0.555611</td>\n",
       "      <td>0.135954</td>\n",
       "      <td>0.490172</td>\n",
       "      <td>0.159177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.785817</td>\n",
       "      <td>0.841516</td>\n",
       "      <td>0.847347</td>\n",
       "      <td>0.866447</td>\n",
       "      <td>0.471485</td>\n",
       "      <td>0.857280</td>\n",
       "      <td>0.836506</td>\n",
       "      <td>0.868012</td>\n",
       "      <td>0.548071</td>\n",
       "      <td>0.342384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.627867</td>\n",
       "      <td>0.674821</td>\n",
       "      <td>0.528824</td>\n",
       "      <td>0.535871</td>\n",
       "      <td>0.752388</td>\n",
       "      <td>0.573493</td>\n",
       "      <td>0.613921</td>\n",
       "      <td>0.489683</td>\n",
       "      <td>0.488304</td>\n",
       "      <td>0.640762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>495</th>\n",
       "      <td>0.527802</td>\n",
       "      <td>0.397016</td>\n",
       "      <td>0.204618</td>\n",
       "      <td>0.512096</td>\n",
       "      <td>0.753966</td>\n",
       "      <td>0.514856</td>\n",
       "      <td>0.374162</td>\n",
       "      <td>0.164782</td>\n",
       "      <td>0.498816</td>\n",
       "      <td>0.688411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>496</th>\n",
       "      <td>0.055696</td>\n",
       "      <td>0.581310</td>\n",
       "      <td>0.477440</td>\n",
       "      <td>0.098886</td>\n",
       "      <td>0.553291</td>\n",
       "      <td>0.052919</td>\n",
       "      <td>0.495894</td>\n",
       "      <td>0.478110</td>\n",
       "      <td>0.508168</td>\n",
       "      <td>0.426957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>497</th>\n",
       "      <td>0.217943</td>\n",
       "      <td>0.324533</td>\n",
       "      <td>0.455735</td>\n",
       "      <td>0.183283</td>\n",
       "      <td>0.974808</td>\n",
       "      <td>0.208432</td>\n",
       "      <td>0.250947</td>\n",
       "      <td>0.460487</td>\n",
       "      <td>0.556596</td>\n",
       "      <td>0.874762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>498</th>\n",
       "      <td>0.279715</td>\n",
       "      <td>0.812766</td>\n",
       "      <td>0.733105</td>\n",
       "      <td>0.355676</td>\n",
       "      <td>0.268785</td>\n",
       "      <td>0.233802</td>\n",
       "      <td>0.742213</td>\n",
       "      <td>0.722258</td>\n",
       "      <td>0.505625</td>\n",
       "      <td>0.164122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>499</th>\n",
       "      <td>0.381543</td>\n",
       "      <td>0.193880</td>\n",
       "      <td>0.691562</td>\n",
       "      <td>0.186422</td>\n",
       "      <td>0.460305</td>\n",
       "      <td>0.336501</td>\n",
       "      <td>0.177748</td>\n",
       "      <td>0.650133</td>\n",
       "      <td>0.448598</td>\n",
       "      <td>0.479279</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>500 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            0         1         2         3         4         0         1  \\\n",
       "0    0.252721  0.599728  0.512525  0.981790  0.657591  0.231055  0.576148   \n",
       "1    0.584286  0.358530  0.932715  0.567207  0.325643  0.601309  0.409537   \n",
       "2    0.248941  0.529385  0.143769  0.492893  0.188981  0.189050  0.555611   \n",
       "3    0.785817  0.841516  0.847347  0.866447  0.471485  0.857280  0.836506   \n",
       "4    0.627867  0.674821  0.528824  0.535871  0.752388  0.573493  0.613921   \n",
       "..        ...       ...       ...       ...       ...       ...       ...   \n",
       "495  0.527802  0.397016  0.204618  0.512096  0.753966  0.514856  0.374162   \n",
       "496  0.055696  0.581310  0.477440  0.098886  0.553291  0.052919  0.495894   \n",
       "497  0.217943  0.324533  0.455735  0.183283  0.974808  0.208432  0.250947   \n",
       "498  0.279715  0.812766  0.733105  0.355676  0.268785  0.233802  0.742213   \n",
       "499  0.381543  0.193880  0.691562  0.186422  0.460305  0.336501  0.177748   \n",
       "\n",
       "            2         3         4  \n",
       "0    0.487459  0.518398  0.572734  \n",
       "1    0.925475  0.435044  0.311838  \n",
       "2    0.135954  0.490172  0.159177  \n",
       "3    0.868012  0.548071  0.342384  \n",
       "4    0.489683  0.488304  0.640762  \n",
       "..        ...       ...       ...  \n",
       "495  0.164782  0.498816  0.688411  \n",
       "496  0.478110  0.508168  0.426957  \n",
       "497  0.460487  0.556596  0.874762  \n",
       "498  0.722258  0.505625  0.164122  \n",
       "499  0.650133  0.448598  0.479279  \n",
       "\n",
       "[500 rows x 10 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ytest.reset_index(drop=True, inplace=True)\n",
    "ypred.reset_index(drop=True, inplace=True)\n",
    "pd.concat([ytest,ypred] ,axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'y_rf' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-54-dd7165d366df>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0msns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjointplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_rf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0my_rf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0my_test\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mkind\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'kde'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mcolor\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'green'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m: name 'y_rf' is not defined"
     ]
    }
   ],
   "source": [
    "sns.jointplot(y_rf.iloc[:,0],y_rf.iloc[:,0]-y_test.iloc[:,0],kind = 'kde',color='green')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.jointplot(y_rf.iloc[:,1],y_rf.iloc[:,1]-y_test.iloc[:,1],kind = 'kde',color='green')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.jointplot(y_rf.iloc[:,2],y_rf.iloc[:,2]-y_test.iloc[:,2],kind = 'kde',color='green')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.jointplot(y_rf.iloc[:,3],y_rf.iloc[:,3]-y_test.iloc[:,3],kind = 'kde',color='green')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.jointplot(y_rf.iloc[:,4],y_rf.iloc[:,4]-y_test.iloc[:,4],kind = 'kde',color='green')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}