keras-nonlinear-regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# imports\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dropout\n",
    "from sklearn.metrics import r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy\n",
    "from keras.optimizers import Adam\n",
    "import keras\n",
    "from matplotlib import pyplot\n",
    "from keras.callbacks import EarlyStopping\n",
    "import pandas as pd \n",
    "from sklearn.preprocessing import LabelEncoder "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read data from csv file for training and validation data\n",
    "TrainingSet = numpy.genfromtxt(\"./training.csv\", delimiter=\",\", skip_header=True)\n",
    "ValidationSet = numpy.genfromtxt(\"./validation.csv\", delimiter=\",\", skip_header=True)\n",
    "\n",
    "# Split into input (X) and output (Y) variables\n",
    "X1 = TrainingSet[:,0:6]\n",
    "Y1 = TrainingSet[:,6]\n",
    "\n",
    "X2 = ValidationSet[:,0:6]\n",
    "Y2 = ValidationSet[:,6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 232 samples, validate on 76 samples\n",
      "Epoch 1/10000000\n",
      " - 1s - loss: 404.3374 - val_loss: 309.6100\n",
      "Epoch 2/10000000\n",
      " - 0s - loss: 375.2555 - val_loss: 294.8144\n",
      "Epoch 3/10000000\n",
      " - 0s - loss: 359.7111 - val_loss: 285.7164\n",
      "Epoch 4/10000000\n",
      " - 0s - loss: 351.3803 - val_loss: 280.9324\n",
      "Epoch 5/10000000\n",
      " - 0s - loss: 347.6015 - val_loss: 278.0652\n",
      "Epoch 6/10000000\n",
      " - 0s - loss: 344.8125 - val_loss: 275.5017\n",
      "Epoch 7/10000000\n",
      " - 0s - loss: 341.8539 - val_loss: 272.8559\n",
      "Epoch 8/10000000\n",
      " - 0s - loss: 338.6591 - val_loss: 270.0061\n",
      "Epoch 9/10000000\n",
      " - 0s - loss: 335.5329 - val_loss: 266.7927\n",
      "Epoch 10/10000000\n",
      " - 0s - loss: 331.8977 - val_loss: 263.3298\n",
      "Epoch 11/10000000\n",
      " - 0s - loss: 327.9559 - val_loss: 259.5031\n",
      "Epoch 12/10000000\n",
      " - 0s - loss: 323.5365 - val_loss: 255.3663\n",
      "Epoch 13/10000000\n",
      " - 0s - loss: 318.8910 - val_loss: 250.9605\n",
      "Epoch 14/10000000\n",
      " - 0s - loss: 313.6974 - val_loss: 246.3042\n",
      "Epoch 15/10000000\n",
      " - 0s - loss: 308.3256 - val_loss: 241.3783\n",
      "Epoch 16/10000000\n",
      " - 0s - loss: 302.5909 - val_loss: 236.3427\n",
      "Epoch 17/10000000\n",
      " - 0s - loss: 296.4475 - val_loss: 231.2525\n",
      "Epoch 18/10000000\n",
      " - 0s - loss: 290.4416 - val_loss: 226.0038\n",
      "Epoch 19/10000000\n",
      " - 0s - loss: 284.2799 - val_loss: 220.7450\n",
      "Epoch 20/10000000\n",
      " - 0s - loss: 277.7946 - val_loss: 215.7167\n",
      "Epoch 21/10000000\n",
      " - 0s - loss: 271.8400 - val_loss: 210.9422\n",
      "Epoch 22/10000000\n",
      " - 0s - loss: 265.8304 - val_loss: 206.7351\n",
      "Epoch 23/10000000\n",
      " - 0s - loss: 260.4738 - val_loss: 203.1385\n",
      "Epoch 24/10000000\n",
      " - 0s - loss: 255.6395 - val_loss: 200.2638\n",
      "Epoch 25/10000000\n",
      " - 0s - loss: 251.9976 - val_loss: 198.2166\n",
      "Epoch 26/10000000\n",
      " - 0s - loss: 248.2025 - val_loss: 197.0029\n",
      "Epoch 27/10000000\n",
      " - 0s - loss: 245.9034 - val_loss: 196.4062\n",
      "Epoch 28/10000000\n",
      " - 0s - loss: 243.8562 - val_loss: 196.4463\n",
      "Epoch 29/10000000\n",
      " - 0s - loss: 242.0419 - val_loss: 196.9243\n",
      "Epoch 30/10000000\n",
      " - 0s - loss: 241.4939 - val_loss: 197.5282\n",
      "Epoch 31/10000000\n",
      " - 0s - loss: 241.0920 - val_loss: 198.0245\n",
      "Epoch 32/10000000\n",
      " - 0s - loss: 240.8747 - val_loss: 198.4490\n",
      "Epoch 33/10000000\n",
      " - 0s - loss: 240.7563 - val_loss: 198.6839\n",
      "Epoch 34/10000000\n",
      " - 0s - loss: 240.5559 - val_loss: 198.7330\n",
      "Epoch 35/10000000\n",
      " - 0s - loss: 240.5655 - val_loss: 198.9215\n",
      "Epoch 36/10000000\n",
      " - 0s - loss: 240.3128 - val_loss: 198.6680\n",
      "Epoch 37/10000000\n",
      " - 0s - loss: 240.1912 - val_loss: 198.4865\n",
      "Epoch 38/10000000\n",
      " - 0s - loss: 240.1707 - val_loss: 198.5354\n",
      "Epoch 39/10000000\n",
      " - 0s - loss: 239.8521 - val_loss: 198.2461\n",
      "Epoch 40/10000000\n",
      " - 0s - loss: 239.7904 - val_loss: 197.8781\n",
      "Epoch 41/10000000\n",
      " - 0s - loss: 239.9121 - val_loss: 197.5384\n",
      "Epoch 42/10000000\n",
      " - 0s - loss: 239.8051 - val_loss: 197.5764\n",
      "Epoch 43/10000000\n",
      " - 0s - loss: 239.6411 - val_loss: 197.8012\n",
      "Epoch 44/10000000\n",
      " - 0s - loss: 239.4158 - val_loss: 198.0564\n",
      "Epoch 45/10000000\n",
      " - 0s - loss: 239.2494 - val_loss: 198.6649\n",
      "Epoch 46/10000000\n",
      " - 0s - loss: 239.0791 - val_loss: 199.2236\n",
      "Epoch 47/10000000\n",
      " - 0s - loss: 239.0726 - val_loss: 199.7865\n",
      "Epoch 48/10000000\n",
      " - 0s - loss: 239.0262 - val_loss: 199.9812\n",
      "Epoch 49/10000000\n",
      " - 0s - loss: 238.9372 - val_loss: 199.9056\n",
      "Epoch 50/10000000\n",
      " - 0s - loss: 238.7978 - val_loss: 199.7509\n",
      "Epoch 51/10000000\n",
      " - 0s - loss: 238.6228 - val_loss: 199.5372\n",
      "Epoch 52/10000000\n",
      " - 0s - loss: 238.4473 - val_loss: 199.2227\n",
      "Epoch 53/10000000\n",
      " - 0s - loss: 238.3821 - val_loss: 198.8013\n",
      "Epoch 54/10000000\n",
      " - 0s - loss: 238.1814 - val_loss: 198.5798\n",
      "Epoch 55/10000000\n",
      " - 0s - loss: 238.0966 - val_loss: 198.1750\n",
      "Epoch 56/10000000\n",
      " - 0s - loss: 238.0117 - val_loss: 198.0500\n",
      "Epoch 57/10000000\n",
      " - 0s - loss: 237.8618 - val_loss: 198.2554\n",
      "Epoch 58/10000000\n",
      " - 0s - loss: 237.7092 - val_loss: 198.5746\n",
      "Epoch 59/10000000\n",
      " - 0s - loss: 237.7295 - val_loss: 198.9153\n",
      "Epoch 60/10000000\n",
      " - 0s - loss: 237.5467 - val_loss: 198.8940\n",
      "Epoch 61/10000000\n",
      " - 0s - loss: 237.4320 - val_loss: 198.6800\n",
      "Epoch 62/10000000\n",
      " - 0s - loss: 237.3223 - val_loss: 198.5539\n",
      "Epoch 63/10000000\n",
      " - 0s - loss: 237.3038 - val_loss: 198.5236\n",
      "Epoch 64/10000000\n",
      " - 0s - loss: 237.1237 - val_loss: 198.9835\n",
      "Epoch 65/10000000\n",
      " - 0s - loss: 237.0398 - val_loss: 199.3556\n",
      "Epoch 66/10000000\n",
      " - 0s - loss: 236.9926 - val_loss: 199.5638\n",
      "Epoch 67/10000000\n",
      " - 0s - loss: 236.9869 - val_loss: 199.8619\n",
      "Epoch 68/10000000\n",
      " - 0s - loss: 236.9322 - val_loss: 199.9018\n",
      "Epoch 69/10000000\n",
      " - 0s - loss: 236.8326 - val_loss: 199.5444\n",
      "Epoch 70/10000000\n",
      " - 0s - loss: 236.6006 - val_loss: 198.9766\n",
      "Epoch 71/10000000\n",
      " - 0s - loss: 236.3486 - val_loss: 198.1425\n",
      "Epoch 72/10000000\n",
      " - 0s - loss: 236.4781 - val_loss: 197.4667\n",
      "Epoch 73/10000000\n",
      " - 0s - loss: 236.3946 - val_loss: 197.2709\n",
      "Epoch 74/10000000\n",
      " - 0s - loss: 236.3165 - val_loss: 197.2895\n",
      "Epoch 75/10000000\n",
      " - 0s - loss: 236.1749 - val_loss: 197.4305\n",
      "Epoch 76/10000000\n",
      " - 0s - loss: 235.9903 - val_loss: 197.5942\n",
      "Epoch 77/10000000\n",
      " - 0s - loss: 236.1782 - val_loss: 198.0923\n",
      "Epoch 78/10000000\n",
      " - 0s - loss: 235.7132 - val_loss: 198.0497\n",
      "Epoch 79/10000000\n",
      " - 0s - loss: 235.6577 - val_loss: 198.0000\n",
      "Epoch 80/10000000\n",
      " - 0s - loss: 235.5127 - val_loss: 197.8643\n",
      "Epoch 81/10000000\n",
      " - 0s - loss: 235.4015 - val_loss: 197.8828\n",
      "Epoch 82/10000000\n",
      " - 0s - loss: 235.3294 - val_loss: 197.9406\n",
      "Epoch 83/10000000\n",
      " - 0s - loss: 235.2051 - val_loss: 198.4744\n",
      "Epoch 84/10000000\n",
      " - 0s - loss: 235.1108 - val_loss: 198.5955\n",
      "Epoch 85/10000000\n",
      " - 0s - loss: 234.9862 - val_loss: 198.3631\n",
      "Epoch 86/10000000\n",
      " - 0s - loss: 234.8161 - val_loss: 198.2144\n",
      "Epoch 87/10000000\n",
      " - 0s - loss: 234.7039 - val_loss: 198.0336\n",
      "Epoch 88/10000000\n",
      " - 0s - loss: 234.5378 - val_loss: 198.0311\n",
      "Epoch 89/10000000\n",
      " - 0s - loss: 234.3574 - val_loss: 198.3667\n",
      "Epoch 90/10000000\n",
      " - 0s - loss: 234.4761 - val_loss: 198.7534\n",
      "Epoch 91/10000000\n",
      " - 0s - loss: 234.3007 - val_loss: 198.6729\n",
      "Epoch 92/10000000\n",
      " - 0s - loss: 234.1115 - val_loss: 198.5859\n",
      "Epoch 93/10000000\n",
      " - 0s - loss: 233.9980 - val_loss: 198.0532\n",
      "Epoch 94/10000000\n",
      " - 0s - loss: 233.7466 - val_loss: 197.6463\n",
      "Epoch 95/10000000\n",
      " - 0s - loss: 233.7787 - val_loss: 197.3472\n",
      "Epoch 96/10000000\n",
      " - 0s - loss: 233.4425 - val_loss: 197.5734\n",
      "Epoch 97/10000000\n",
      " - 0s - loss: 233.3487 - val_loss: 197.8008\n",
      "Epoch 98/10000000\n",
      " - 0s - loss: 233.2328 - val_loss: 197.7281\n",
      "Epoch 99/10000000\n",
      " - 0s - loss: 233.1117 - val_loss: 197.6817\n",
      "Epoch 100/10000000\n",
      " - 0s - loss: 232.9602 - val_loss: 197.9137\n",
      "Epoch 101/10000000\n",
      " - 0s - loss: 232.8739 - val_loss: 197.8880\n",
      "Epoch 102/10000000\n",
      " - 0s - loss: 232.7895 - val_loss: 197.4438\n",
      "Epoch 103/10000000\n",
      " - 0s - loss: 232.7038 - val_loss: 197.4013\n",
      "Epoch 104/10000000\n",
      " - 0s - loss: 232.4241 - val_loss: 198.1633\n",
      "Epoch 105/10000000\n",
      " - 0s - loss: 232.5389 - val_loss: 198.1818\n",
      "Epoch 106/10000000\n",
      " - 0s - loss: 232.2177 - val_loss: 197.4664\n",
      "Epoch 107/10000000\n",
      " - 0s - loss: 231.9705 - val_loss: 196.8382\n",
      "Epoch 108/10000000\n",
      " - 0s - loss: 231.6963 - val_loss: 196.3968\n",
      "Epoch 109/10000000\n",
      " - 0s - loss: 231.5513 - val_loss: 195.9977\n",
      "Epoch 110/10000000\n",
      " - 0s - loss: 231.5665 - val_loss: 195.7375\n",
      "Epoch 111/10000000\n",
      " - 0s - loss: 231.4251 - val_loss: 195.9104\n",
      "Epoch 112/10000000\n",
      " - 0s - loss: 230.9907 - val_loss: 196.2386\n",
      "Epoch 113/10000000\n",
      " - 0s - loss: 230.6764 - val_loss: 196.7135\n",
      "Epoch 114/10000000\n",
      " - 0s - loss: 230.3076 - val_loss: 196.1538\n",
      "Epoch 115/10000000\n",
      " - 0s - loss: 229.7232 - val_loss: 195.9030\n",
      "Epoch 116/10000000\n",
      " - 0s - loss: 229.4335 - val_loss: 195.2312\n",
      "Epoch 117/10000000\n",
      " - 0s - loss: 229.0130 - val_loss: 194.5650\n",
      "Epoch 118/10000000\n",
      " - 0s - loss: 229.5841 - val_loss: 194.3074\n",
      "Epoch 119/10000000\n",
      " - 0s - loss: 230.0447 - val_loss: 194.1773\n",
      "Epoch 120/10000000\n",
      " - 0s - loss: 229.6446 - val_loss: 194.1019\n",
      "Epoch 121/10000000\n",
      " - 0s - loss: 228.8661 - val_loss: 194.3277\n",
      "Epoch 122/10000000\n",
      " - 0s - loss: 228.0737 - val_loss: 195.0831\n",
      "Epoch 123/10000000\n",
      " - 0s - loss: 227.9146 - val_loss: 195.6099\n",
      "Epoch 124/10000000\n",
      " - 0s - loss: 228.3039 - val_loss: 195.4970\n",
      "Epoch 125/10000000\n",
      " - 0s - loss: 227.4324 - val_loss: 194.2471\n",
      "Epoch 126/10000000\n",
      " - 0s - loss: 226.9927 - val_loss: 193.5786\n",
      "Epoch 127/10000000\n",
      " - 0s - loss: 227.0249 - val_loss: 192.9961\n",
      "Epoch 128/10000000\n",
      " - 0s - loss: 226.8555 - val_loss: 192.9233\n",
      "Epoch 129/10000000\n",
      " - 0s - loss: 226.3521 - val_loss: 193.2070\n",
      "Epoch 130/10000000\n",
      " - 0s - loss: 226.2942 - val_loss: 194.4982\n",
      "Epoch 131/10000000\n",
      " - 0s - loss: 226.1386 - val_loss: 195.4029\n",
      "Epoch 132/10000000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 0s - loss: 225.8677 - val_loss: 194.0932\n",
      "Epoch 133/10000000\n",
      " - 0s - loss: 225.0746 - val_loss: 192.8357\n",
      "Epoch 134/10000000\n",
      " - 0s - loss: 224.8184 - val_loss: 191.6137\n",
      "Epoch 135/10000000\n",
      " - 0s - loss: 224.4828 - val_loss: 191.6492\n",
      "Epoch 136/10000000\n",
      " - 0s - loss: 223.8026 - val_loss: 192.1965\n",
      "Epoch 137/10000000\n",
      " - 0s - loss: 223.7140 - val_loss: 193.6813\n",
      "Epoch 138/10000000\n",
      " - 0s - loss: 224.0164 - val_loss: 193.0983\n",
      "Epoch 139/10000000\n",
      " - 0s - loss: 223.2894 - val_loss: 192.5623\n",
      "Epoch 140/10000000\n",
      " - 0s - loss: 222.5569 - val_loss: 190.9665\n",
      "Epoch 141/10000000\n",
      " - 0s - loss: 221.9068 - val_loss: 189.9485\n",
      "Epoch 142/10000000\n",
      " - 0s - loss: 221.5390 - val_loss: 189.2122\n",
      "Epoch 143/10000000\n",
      " - 0s - loss: 221.4248 - val_loss: 188.5644\n",
      "Epoch 144/10000000\n",
      " - 0s - loss: 221.0609 - val_loss: 188.0770\n",
      "Epoch 145/10000000\n",
      " - 0s - loss: 220.7368 - val_loss: 187.8320\n",
      "Epoch 146/10000000\n",
      " - 0s - loss: 220.1484 - val_loss: 187.6898\n",
      "Epoch 147/10000000\n",
      " - 0s - loss: 219.4844 - val_loss: 187.3777\n",
      "Epoch 148/10000000\n",
      " - 0s - loss: 219.0358 - val_loss: 187.2445\n",
      "Epoch 149/10000000\n",
      " - 0s - loss: 218.5353 - val_loss: 186.8537\n",
      "Epoch 150/10000000\n",
      " - 0s - loss: 217.9757 - val_loss: 186.5437\n",
      "Epoch 151/10000000\n",
      " - 0s - loss: 217.4511 - val_loss: 186.0903\n",
      "Epoch 152/10000000\n",
      " - 0s - loss: 216.9645 - val_loss: 184.9469\n",
      "Epoch 153/10000000\n",
      " - 0s - loss: 216.3938 - val_loss: 184.5468\n",
      "Epoch 154/10000000\n",
      " - 0s - loss: 215.6852 - val_loss: 184.9159\n",
      "Epoch 155/10000000\n",
      " - 0s - loss: 215.2845 - val_loss: 186.8736\n",
      "Epoch 156/10000000\n",
      " - 0s - loss: 216.6804 - val_loss: 187.6324\n",
      "Epoch 157/10000000\n",
      " - 0s - loss: 214.9999 - val_loss: 183.1812\n",
      "Epoch 158/10000000\n",
      " - 0s - loss: 212.9916 - val_loss: 180.8576\n",
      "Epoch 159/10000000\n",
      " - 0s - loss: 213.1575 - val_loss: 180.1301\n",
      "Epoch 160/10000000\n",
      " - 0s - loss: 213.7664 - val_loss: 179.4552\n",
      "Epoch 161/10000000\n",
      " - 0s - loss: 212.8575 - val_loss: 178.8515\n",
      "Epoch 162/10000000\n",
      " - 0s - loss: 211.3917 - val_loss: 178.7752\n",
      "Epoch 163/10000000\n",
      " - 0s - loss: 209.6184 - val_loss: 180.3216\n",
      "Epoch 164/10000000\n",
      " - 0s - loss: 209.6680 - val_loss: 182.8222\n",
      "Epoch 165/10000000\n",
      " - 0s - loss: 209.8520 - val_loss: 180.5202\n",
      "Epoch 166/10000000\n",
      " - 0s - loss: 208.1876 - val_loss: 177.9664\n",
      "Epoch 167/10000000\n",
      " - 0s - loss: 207.2842 - val_loss: 175.1391\n",
      "Epoch 168/10000000\n",
      " - 0s - loss: 206.4750 - val_loss: 174.3483\n",
      "Epoch 169/10000000\n",
      " - 0s - loss: 205.5183 - val_loss: 174.2364\n",
      "Epoch 170/10000000\n",
      " - 0s - loss: 204.2340 - val_loss: 174.5261\n",
      "Epoch 171/10000000\n",
      " - 0s - loss: 203.1987 - val_loss: 174.0214\n",
      "Epoch 172/10000000\n",
      " - 0s - loss: 202.5007 - val_loss: 173.6355\n",
      "Epoch 173/10000000\n",
      " - 0s - loss: 201.4743 - val_loss: 171.6443\n",
      "Epoch 174/10000000\n",
      " - 0s - loss: 200.3985 - val_loss: 170.1512\n",
      "Epoch 175/10000000\n",
      " - 0s - loss: 199.5050 - val_loss: 168.3026\n",
      "Epoch 176/10000000\n",
      " - 0s - loss: 199.2813 - val_loss: 167.4313\n",
      "Epoch 177/10000000\n",
      " - 0s - loss: 197.2954 - val_loss: 167.9148\n",
      "Epoch 178/10000000\n",
      " - 0s - loss: 195.9012 - val_loss: 168.3747\n",
      "Epoch 179/10000000\n",
      " - 0s - loss: 195.0270 - val_loss: 166.2235\n",
      "Epoch 180/10000000\n",
      " - 0s - loss: 193.7709 - val_loss: 164.5722\n",
      "Epoch 181/10000000\n",
      " - 0s - loss: 192.1652 - val_loss: 162.0016\n",
      "Epoch 182/10000000\n",
      " - 0s - loss: 191.8638 - val_loss: 160.7229\n",
      "Epoch 183/10000000\n",
      " - 0s - loss: 191.4880 - val_loss: 159.6008\n",
      "Epoch 184/10000000\n",
      " - 0s - loss: 189.7686 - val_loss: 159.0300\n",
      "Epoch 185/10000000\n",
      " - 0s - loss: 186.8606 - val_loss: 162.4144\n",
      "Epoch 186/10000000\n",
      " - 0s - loss: 187.5038 - val_loss: 159.8384\n",
      "Epoch 187/10000000\n",
      " - 0s - loss: 185.2482 - val_loss: 157.3847\n",
      "Epoch 188/10000000\n",
      " - 0s - loss: 183.0341 - val_loss: 156.9012\n",
      "Epoch 189/10000000\n",
      " - 0s - loss: 181.9437 - val_loss: 155.4670\n",
      "Epoch 190/10000000\n",
      " - 0s - loss: 180.2125 - val_loss: 152.1032\n",
      "Epoch 191/10000000\n",
      " - 0s - loss: 178.1442 - val_loss: 149.4443\n",
      "Epoch 192/10000000\n",
      " - 0s - loss: 178.5215 - val_loss: 148.2245\n",
      "Epoch 193/10000000\n",
      " - 0s - loss: 175.9819 - val_loss: 151.6729\n",
      "Epoch 194/10000000\n",
      " - 0s - loss: 174.3932 - val_loss: 148.3694\n",
      "Epoch 195/10000000\n",
      " - 0s - loss: 171.7957 - val_loss: 144.8136\n",
      "Epoch 196/10000000\n",
      " - 0s - loss: 170.3044 - val_loss: 143.2381\n",
      "Epoch 197/10000000\n",
      " - 0s - loss: 168.4197 - val_loss: 143.8532\n",
      "Epoch 198/10000000\n",
      " - 0s - loss: 166.6741 - val_loss: 143.6439\n",
      "Epoch 199/10000000\n",
      " - 0s - loss: 165.5343 - val_loss: 139.8479\n",
      "Epoch 200/10000000\n",
      " - 0s - loss: 163.2186 - val_loss: 136.9763\n",
      "Epoch 201/10000000\n",
      " - 0s - loss: 160.8097 - val_loss: 138.7165\n",
      "Epoch 202/10000000\n",
      " - 0s - loss: 160.2502 - val_loss: 136.5220\n",
      "Epoch 203/10000000\n",
      " - 0s - loss: 157.0403 - val_loss: 130.9358\n",
      "Epoch 204/10000000\n",
      " - 0s - loss: 155.6575 - val_loss: 129.3727\n",
      "Epoch 205/10000000\n",
      " - 0s - loss: 153.8686 - val_loss: 128.9259\n",
      "Epoch 206/10000000\n",
      " - 0s - loss: 152.7503 - val_loss: 131.1255\n",
      "Epoch 207/10000000\n",
      " - 0s - loss: 149.8259 - val_loss: 123.8214\n",
      "Epoch 208/10000000\n",
      " - 0s - loss: 147.6574 - val_loss: 122.7466\n",
      "Epoch 209/10000000\n",
      " - 0s - loss: 145.5030 - val_loss: 125.7544\n",
      "Epoch 210/10000000\n",
      " - 0s - loss: 144.0790 - val_loss: 118.8526\n",
      "Epoch 211/10000000\n",
      " - 0s - loss: 141.1175 - val_loss: 116.0498\n",
      "Epoch 212/10000000\n",
      " - 0s - loss: 140.3155 - val_loss: 115.9177\n",
      "Epoch 213/10000000\n",
      " - 0s - loss: 137.0851 - val_loss: 120.5550\n",
      "Epoch 214/10000000\n",
      " - 0s - loss: 136.7862 - val_loss: 113.4455\n",
      "Epoch 215/10000000\n",
      " - 0s - loss: 132.6787 - val_loss: 109.7631\n",
      "Epoch 216/10000000\n",
      " - 0s - loss: 131.0077 - val_loss: 109.1655\n",
      "Epoch 217/10000000\n",
      " - 0s - loss: 129.0218 - val_loss: 107.0975\n",
      "Epoch 218/10000000\n",
      " - 0s - loss: 126.6121 - val_loss: 104.1593\n",
      "Epoch 219/10000000\n",
      " - 0s - loss: 125.5135 - val_loss: 105.3840\n",
      "Epoch 220/10000000\n",
      " - 0s - loss: 123.9512 - val_loss: 104.5540\n",
      "Epoch 221/10000000\n",
      " - 0s - loss: 120.7989 - val_loss: 99.6564\n",
      "Epoch 222/10000000\n",
      " - 0s - loss: 125.8678 - val_loss: 97.8695\n",
      "Epoch 223/10000000\n",
      " - 0s - loss: 119.2765 - val_loss: 114.9480\n",
      "Epoch 224/10000000\n",
      " - 0s - loss: 125.0163 - val_loss: 95.8539\n",
      "Epoch 225/10000000\n",
      " - 0s - loss: 120.7481 - val_loss: 95.9972\n",
      "Epoch 226/10000000\n",
      " - 0s - loss: 118.3717 - val_loss: 97.2311\n",
      "Epoch 227/10000000\n",
      " - 0s - loss: 112.2732 - val_loss: 99.4042\n",
      "Epoch 228/10000000\n",
      " - 0s - loss: 111.3849 - val_loss: 91.0918\n",
      "Epoch 229/10000000\n",
      " - 0s - loss: 109.5693 - val_loss: 88.5779\n",
      "Epoch 230/10000000\n",
      " - 0s - loss: 107.1510 - val_loss: 93.7407\n",
      "Epoch 231/10000000\n",
      " - 0s - loss: 108.4626 - val_loss: 91.1903\n",
      "Epoch 232/10000000\n",
      " - 0s - loss: 105.6543 - val_loss: 85.2376\n",
      "Epoch 233/10000000\n",
      " - 0s - loss: 104.0284 - val_loss: 88.3317\n",
      "Epoch 234/10000000\n",
      " - 0s - loss: 103.3770 - val_loss: 89.4946\n",
      "Epoch 235/10000000\n",
      " - 0s - loss: 101.0594 - val_loss: 82.4076\n",
      "Epoch 236/10000000\n",
      " - 0s - loss: 99.0198 - val_loss: 83.2899\n",
      "Epoch 237/10000000\n",
      " - 0s - loss: 97.0308 - val_loss: 84.6584\n",
      "Epoch 238/10000000\n",
      " - 0s - loss: 95.7771 - val_loss: 79.4402\n",
      "Epoch 239/10000000\n",
      " - 0s - loss: 95.5814 - val_loss: 78.9086\n",
      "Epoch 240/10000000\n",
      " - 0s - loss: 93.5693 - val_loss: 82.4913\n",
      "Epoch 241/10000000\n",
      " - 0s - loss: 93.4106 - val_loss: 79.0517\n",
      "Epoch 242/10000000\n",
      " - 0s - loss: 91.9907 - val_loss: 77.1644\n",
      "Epoch 243/10000000\n",
      " - 0s - loss: 91.4818 - val_loss: 77.0786\n",
      "Epoch 244/10000000\n",
      " - 0s - loss: 91.1469 - val_loss: 79.9624\n",
      "Epoch 245/10000000\n",
      " - 0s - loss: 89.2218 - val_loss: 75.0374\n",
      "Epoch 246/10000000\n",
      " - 0s - loss: 92.9519 - val_loss: 74.9174\n",
      "Epoch 247/10000000\n",
      " - 0s - loss: 88.9368 - val_loss: 82.6417\n",
      "Epoch 248/10000000\n",
      " - 0s - loss: 90.1105 - val_loss: 74.1661\n",
      "Epoch 249/10000000\n",
      " - 0s - loss: 87.1979 - val_loss: 74.8223\n",
      "Epoch 250/10000000\n",
      " - 0s - loss: 86.1758 - val_loss: 75.3091\n",
      "Epoch 251/10000000\n",
      " - 0s - loss: 85.9226 - val_loss: 74.3616\n",
      "Epoch 252/10000000\n",
      " - 0s - loss: 85.0691 - val_loss: 73.8962\n",
      "Epoch 253/10000000\n",
      " - 0s - loss: 85.8900 - val_loss: 73.9822\n",
      "Epoch 254/10000000\n",
      " - 0s - loss: 85.1015 - val_loss: 77.6231\n",
      "Epoch 255/10000000\n",
      " - 0s - loss: 84.5553 - val_loss: 71.5928\n",
      "Epoch 256/10000000\n",
      " - 0s - loss: 84.7012 - val_loss: 72.5828\n",
      "Epoch 257/10000000\n",
      " - 0s - loss: 82.2304 - val_loss: 80.7227\n",
      "Epoch 258/10000000\n",
      " - 0s - loss: 85.6695 - val_loss: 71.5649\n",
      "Epoch 259/10000000\n",
      " - 0s - loss: 84.1242 - val_loss: 71.8647\n",
      "Epoch 260/10000000\n",
      " - 0s - loss: 85.1095 - val_loss: 76.8812\n",
      "Epoch 261/10000000\n",
      " - 0s - loss: 88.0713 - val_loss: 81.2362\n",
      "Epoch 262/10000000\n",
      " - 0s - loss: 85.7494 - val_loss: 70.9554\n",
      "Epoch 263/10000000\n",
      " - 0s - loss: 84.0437 - val_loss: 73.5596\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 264/10000000\n",
      " - 0s - loss: 81.8762 - val_loss: 72.9376\n",
      "Epoch 265/10000000\n",
      " - 0s - loss: 83.0389 - val_loss: 69.7595\n",
      "Epoch 266/10000000\n",
      " - 0s - loss: 81.1133 - val_loss: 72.8278\n",
      "Epoch 267/10000000\n",
      " - 0s - loss: 80.8147 - val_loss: 71.7250\n",
      "Epoch 268/10000000\n",
      " - 0s - loss: 80.2131 - val_loss: 70.1438\n",
      "Epoch 269/10000000\n",
      " - 0s - loss: 80.2669 - val_loss: 69.7171\n",
      "Epoch 270/10000000\n",
      " - 0s - loss: 79.9815 - val_loss: 71.5652\n",
      "Epoch 271/10000000\n",
      " - 0s - loss: 79.8216 - val_loss: 72.3033\n",
      "Epoch 272/10000000\n",
      " - 0s - loss: 80.0324 - val_loss: 70.5617\n",
      "Epoch 273/10000000\n",
      " - 0s - loss: 79.5343 - val_loss: 69.3942\n",
      "Epoch 274/10000000\n",
      " - 0s - loss: 80.4780 - val_loss: 69.5802\n",
      "Epoch 275/10000000\n",
      " - 0s - loss: 79.0237 - val_loss: 74.1403\n",
      "Epoch 276/10000000\n",
      " - 0s - loss: 80.2075 - val_loss: 70.0187\n",
      "Epoch 277/10000000\n",
      " - 0s - loss: 78.5222 - val_loss: 69.2006\n",
      "Epoch 278/10000000\n",
      " - 0s - loss: 81.8816 - val_loss: 70.3821\n",
      "Epoch 279/10000000\n",
      " - 0s - loss: 83.0620 - val_loss: 79.5702\n",
      "Epoch 280/10000000\n",
      " - 0s - loss: 80.0081 - val_loss: 67.9033\n",
      "Epoch 281/10000000\n",
      " - 0s - loss: 84.3505 - val_loss: 67.2808\n",
      "Epoch 282/10000000\n",
      " - 0s - loss: 77.9080 - val_loss: 74.6755\n",
      "Epoch 283/10000000\n",
      " - 0s - loss: 80.1767 - val_loss: 69.9488\n",
      "Epoch 284/10000000\n",
      " - 0s - loss: 75.0663 - val_loss: 63.6831\n",
      "Epoch 285/10000000\n",
      " - 0s - loss: 77.9508 - val_loss: 62.2906\n",
      "Epoch 286/10000000\n",
      " - 0s - loss: 73.2548 - val_loss: 69.7801\n",
      "Epoch 287/10000000\n",
      " - 0s - loss: 75.0809 - val_loss: 62.1624\n",
      "Epoch 288/10000000\n",
      " - 0s - loss: 71.4248 - val_loss: 59.5791\n",
      "Epoch 289/10000000\n",
      " - 0s - loss: 71.0929 - val_loss: 60.3116\n",
      "Epoch 290/10000000\n",
      " - 0s - loss: 70.3271 - val_loss: 60.8598\n",
      "Epoch 291/10000000\n",
      " - 0s - loss: 71.1784 - val_loss: 63.3779\n",
      "Epoch 292/10000000\n",
      " - 0s - loss: 70.3899 - val_loss: 59.5239\n",
      "Epoch 293/10000000\n",
      " - 0s - loss: 69.4709 - val_loss: 58.4136\n",
      "Epoch 294/10000000\n",
      " - 0s - loss: 68.4567 - val_loss: 61.7049\n",
      "Epoch 295/10000000\n",
      " - 0s - loss: 71.2738 - val_loss: 59.5847\n",
      "Epoch 296/10000000\n",
      " - 0s - loss: 68.3289 - val_loss: 58.8522\n",
      "Epoch 297/10000000\n",
      " - 0s - loss: 71.5201 - val_loss: 57.5557\n",
      "Epoch 298/10000000\n",
      " - 0s - loss: 67.4160 - val_loss: 59.0558\n",
      "Epoch 299/10000000\n",
      " - 0s - loss: 67.2132 - val_loss: 57.8518\n",
      "Epoch 300/10000000\n",
      " - 0s - loss: 66.9274 - val_loss: 57.0729\n",
      "Epoch 301/10000000\n",
      " - 0s - loss: 67.6819 - val_loss: 57.4972\n",
      "Epoch 302/10000000\n",
      " - 0s - loss: 67.3145 - val_loss: 57.1518\n",
      "Epoch 303/10000000\n",
      " - 0s - loss: 66.1540 - val_loss: 61.5868\n",
      "Epoch 304/10000000\n",
      " - 0s - loss: 67.1594 - val_loss: 56.3556\n",
      "Epoch 305/10000000\n",
      " - 0s - loss: 66.9374 - val_loss: 56.1487\n",
      "Epoch 306/10000000\n",
      " - 0s - loss: 66.1014 - val_loss: 60.3481\n",
      "Epoch 307/10000000\n",
      " - 0s - loss: 66.6644 - val_loss: 57.1693\n",
      "Epoch 308/10000000\n",
      " - 0s - loss: 65.2367 - val_loss: 55.8538\n",
      "Epoch 309/10000000\n",
      " - 0s - loss: 66.8406 - val_loss: 55.8118\n",
      "Epoch 310/10000000\n",
      " - 0s - loss: 65.6111 - val_loss: 57.5310\n",
      "Epoch 311/10000000\n",
      " - 0s - loss: 64.9846 - val_loss: 56.1618\n",
      "Epoch 312/10000000\n",
      " - 0s - loss: 64.7348 - val_loss: 55.6899\n",
      "Epoch 313/10000000\n",
      " - 0s - loss: 64.6827 - val_loss: 56.8709\n",
      "Epoch 314/10000000\n",
      " - 0s - loss: 65.2264 - val_loss: 56.1530\n",
      "Epoch 315/10000000\n",
      " - 0s - loss: 64.2943 - val_loss: 56.7960\n",
      "Epoch 316/10000000\n",
      " - 0s - loss: 64.4068 - val_loss: 56.7646\n",
      "Epoch 317/10000000\n",
      " - 0s - loss: 64.4326 - val_loss: 55.4357\n",
      "Epoch 318/10000000\n",
      " - 0s - loss: 64.1518 - val_loss: 55.3105\n",
      "Epoch 319/10000000\n",
      " - 0s - loss: 65.1449 - val_loss: 54.8110\n",
      "Epoch 320/10000000\n",
      " - 0s - loss: 63.7353 - val_loss: 57.5032\n",
      "Epoch 321/10000000\n",
      " - 0s - loss: 63.9537 - val_loss: 55.2648\n",
      "Epoch 322/10000000\n",
      " - 0s - loss: 63.3634 - val_loss: 54.8250\n",
      "Epoch 323/10000000\n",
      " - 0s - loss: 63.3533 - val_loss: 55.5467\n",
      "Epoch 324/10000000\n",
      " - 0s - loss: 63.2443 - val_loss: 55.3938\n",
      "Epoch 325/10000000\n",
      " - 0s - loss: 63.0467 - val_loss: 55.7194\n",
      "Epoch 326/10000000\n",
      " - 0s - loss: 62.7823 - val_loss: 54.2783\n",
      "Epoch 327/10000000\n",
      " - 0s - loss: 64.7053 - val_loss: 54.2452\n",
      "Epoch 328/10000000\n",
      " - 0s - loss: 64.5199 - val_loss: 57.3929\n",
      "Epoch 329/10000000\n",
      " - 0s - loss: 62.7227 - val_loss: 54.1579\n",
      "Epoch 330/10000000\n",
      " - 0s - loss: 63.8097 - val_loss: 54.0885\n",
      "Epoch 331/10000000\n",
      " - 0s - loss: 62.5286 - val_loss: 56.2687\n",
      "Epoch 332/10000000\n",
      " - 0s - loss: 63.8416 - val_loss: 57.1309\n",
      "Epoch 333/10000000\n",
      " - 0s - loss: 63.5279 - val_loss: 54.2627\n",
      "Epoch 334/10000000\n",
      " - 0s - loss: 62.2556 - val_loss: 54.1753\n",
      "Epoch 335/10000000\n",
      " - 0s - loss: 62.1480 - val_loss: 54.0872\n",
      "Epoch 336/10000000\n",
      " - 0s - loss: 61.9420 - val_loss: 54.8801\n",
      "Epoch 337/10000000\n",
      " - 0s - loss: 62.9101 - val_loss: 55.1268\n",
      "Epoch 338/10000000\n",
      " - 0s - loss: 61.8930 - val_loss: 53.7495\n",
      "Epoch 339/10000000\n",
      " - 0s - loss: 61.9445 - val_loss: 54.0086\n",
      "Epoch 340/10000000\n",
      " - 0s - loss: 61.6666 - val_loss: 53.4591\n",
      "Epoch 341/10000000\n",
      " - 0s - loss: 61.5856 - val_loss: 54.8164\n",
      "Epoch 342/10000000\n",
      " - 0s - loss: 63.5259 - val_loss: 55.0151\n",
      "Epoch 343/10000000\n",
      " - 0s - loss: 62.4853 - val_loss: 53.7095\n",
      "Epoch 344/10000000\n",
      " - 0s - loss: 61.8771 - val_loss: 57.2880\n",
      "Epoch 345/10000000\n",
      " - 0s - loss: 65.7725 - val_loss: 57.1885\n",
      "Epoch 346/10000000\n",
      " - 0s - loss: 65.3458 - val_loss: 54.3291\n",
      "Epoch 347/10000000\n",
      " - 0s - loss: 63.4227 - val_loss: 55.3697\n",
      "Epoch 348/10000000\n",
      " - 0s - loss: 62.5200 - val_loss: 55.5982\n",
      "Epoch 349/10000000\n",
      " - 0s - loss: 61.5239 - val_loss: 52.9108\n",
      "Epoch 350/10000000\n",
      " - 0s - loss: 62.2195 - val_loss: 53.0569\n",
      "Epoch 351/10000000\n",
      " - 0s - loss: 60.6915 - val_loss: 53.1101\n",
      "Epoch 352/10000000\n",
      " - 0s - loss: 60.6414 - val_loss: 53.6939\n",
      "Epoch 353/10000000\n",
      " - 0s - loss: 61.1932 - val_loss: 55.2446\n",
      "Epoch 354/10000000\n",
      " - 0s - loss: 60.4140 - val_loss: 52.5642\n",
      "Epoch 355/10000000\n",
      " - 0s - loss: 61.0043 - val_loss: 52.5251\n",
      "Epoch 356/10000000\n",
      " - 0s - loss: 60.9754 - val_loss: 54.3206\n",
      "Epoch 357/10000000\n",
      " - 0s - loss: 60.0788 - val_loss: 52.7561\n",
      "Epoch 358/10000000\n",
      " - 0s - loss: 60.3500 - val_loss: 53.0253\n",
      "Epoch 359/10000000\n",
      " - 0s - loss: 59.6992 - val_loss: 56.0185\n",
      "Epoch 360/10000000\n",
      " - 0s - loss: 60.4860 - val_loss: 52.1981\n",
      "Epoch 361/10000000\n",
      " - 0s - loss: 62.6488 - val_loss: 52.3978\n",
      "Epoch 362/10000000\n",
      " - 0s - loss: 58.6446 - val_loss: 62.9310\n",
      "Epoch 363/10000000\n",
      " - 0s - loss: 65.8543 - val_loss: 54.6803\n",
      "Epoch 364/10000000\n",
      " - 0s - loss: 59.8142 - val_loss: 52.9893\n",
      "Epoch 365/10000000\n",
      " - 0s - loss: 61.6210 - val_loss: 53.4816\n",
      "Epoch 366/10000000\n",
      " - 0s - loss: 61.1580 - val_loss: 56.8128\n",
      "Epoch 367/10000000\n",
      " - 0s - loss: 61.4421 - val_loss: 51.8346\n",
      "Epoch 368/10000000\n",
      " - 0s - loss: 59.4228 - val_loss: 52.7083\n",
      "Epoch 369/10000000\n",
      " - 0s - loss: 58.7494 - val_loss: 55.9930\n",
      "Epoch 370/10000000\n",
      " - 0s - loss: 60.1795 - val_loss: 53.2655\n",
      "Epoch 371/10000000\n",
      " - 0s - loss: 59.2597 - val_loss: 51.9789\n",
      "Epoch 372/10000000\n",
      " - 0s - loss: 59.3399 - val_loss: 52.4442\n",
      "Epoch 373/10000000\n",
      " - 0s - loss: 59.7245 - val_loss: 51.2823\n",
      "Epoch 374/10000000\n",
      " - 0s - loss: 58.3431 - val_loss: 55.2744\n",
      "Epoch 375/10000000\n",
      " - 0s - loss: 61.4009 - val_loss: 53.4603\n",
      "Epoch 376/10000000\n",
      " - 0s - loss: 58.8338 - val_loss: 52.1415\n",
      "Epoch 377/10000000\n",
      " - 0s - loss: 61.0111 - val_loss: 51.5777\n",
      "Epoch 378/10000000\n",
      " - 0s - loss: 59.3303 - val_loss: 57.5802\n",
      "Epoch 379/10000000\n",
      " - 0s - loss: 59.6570 - val_loss: 51.0886\n",
      "Epoch 380/10000000\n",
      " - 0s - loss: 59.9617 - val_loss: 51.1781\n",
      "Epoch 381/10000000\n",
      " - 0s - loss: 58.5649 - val_loss: 54.8613\n",
      "Epoch 382/10000000\n",
      " - 0s - loss: 61.1434 - val_loss: 51.9995\n",
      "Epoch 383/10000000\n",
      " - 0s - loss: 58.5067 - val_loss: 51.7075\n",
      "Epoch 384/10000000\n",
      " - 0s - loss: 60.1086 - val_loss: 52.0020\n",
      "Epoch 385/10000000\n",
      " - 0s - loss: 58.0633 - val_loss: 51.8389\n",
      "Epoch 386/10000000\n",
      " - 0s - loss: 57.1480 - val_loss: 50.4453\n",
      "Epoch 387/10000000\n",
      " - 0s - loss: 59.5953 - val_loss: 50.2750\n",
      "Epoch 388/10000000\n",
      " - 0s - loss: 57.7760 - val_loss: 56.0920\n",
      "Epoch 389/10000000\n",
      " - 0s - loss: 60.0605 - val_loss: 51.2149\n",
      "Epoch 390/10000000\n",
      " - 0s - loss: 57.2120 - val_loss: 50.1055\n",
      "Epoch 391/10000000\n",
      " - 0s - loss: 57.6820 - val_loss: 50.1896\n",
      "Epoch 392/10000000\n",
      " - 0s - loss: 56.6753 - val_loss: 54.6551\n",
      "Epoch 393/10000000\n",
      " - 0s - loss: 60.1308 - val_loss: 51.5381\n",
      "Epoch 394/10000000\n",
      " - 0s - loss: 56.6198 - val_loss: 49.8391\n",
      "Epoch 395/10000000\n",
      " - 0s - loss: 58.6413 - val_loss: 49.6475\n",
      "Epoch 396/10000000\n",
      " - 0s - loss: 56.5766 - val_loss: 52.0953\n",
      "Epoch 397/10000000\n",
      " - 0s - loss: 57.8526 - val_loss: 51.4278\n",
      "Epoch 398/10000000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 0s - loss: 57.3627 - val_loss: 49.5722\n",
      "Epoch 399/10000000\n",
      " - 0s - loss: 57.6261 - val_loss: 49.9393\n",
      "Epoch 400/10000000\n",
      " - 0s - loss: 56.4614 - val_loss: 50.4169\n",
      "Epoch 401/10000000\n",
      " - 0s - loss: 56.6917 - val_loss: 51.8208\n",
      "Epoch 402/10000000\n",
      " - 0s - loss: 57.1966 - val_loss: 49.4367\n",
      "Epoch 403/10000000\n",
      " - 0s - loss: 56.2179 - val_loss: 49.9755\n",
      "Epoch 404/10000000\n",
      " - 0s - loss: 56.1844 - val_loss: 49.5251\n",
      "Epoch 405/10000000\n",
      " - 0s - loss: 56.3043 - val_loss: 50.0773\n",
      "Epoch 406/10000000\n",
      " - 0s - loss: 56.6836 - val_loss: 50.3680\n",
      "Epoch 407/10000000\n",
      " - 0s - loss: 55.9443 - val_loss: 48.9125\n",
      "Epoch 408/10000000\n",
      " - 0s - loss: 56.8805 - val_loss: 48.9223\n",
      "Epoch 409/10000000\n",
      " - 0s - loss: 55.5438 - val_loss: 51.5834\n",
      "Epoch 410/10000000\n",
      " - 0s - loss: 56.3839 - val_loss: 49.1182\n",
      "Epoch 411/10000000\n",
      " - 0s - loss: 56.3195 - val_loss: 48.8493\n",
      "Epoch 412/10000000\n",
      " - 0s - loss: 55.4827 - val_loss: 53.4909\n",
      "Epoch 413/10000000\n",
      " - 0s - loss: 56.9031 - val_loss: 48.9527\n",
      "Epoch 414/10000000\n",
      " - 0s - loss: 55.5117 - val_loss: 48.4251\n",
      "Epoch 415/10000000\n",
      " - 0s - loss: 55.8730 - val_loss: 49.2347\n",
      "Epoch 416/10000000\n",
      " - 0s - loss: 55.2258 - val_loss: 51.3050\n",
      "Epoch 417/10000000\n",
      " - 0s - loss: 55.6793 - val_loss: 48.3772\n",
      "Epoch 418/10000000\n",
      " - 0s - loss: 54.9728 - val_loss: 48.4591\n",
      "Epoch 419/10000000\n",
      " - 0s - loss: 55.1496 - val_loss: 48.6517\n",
      "Epoch 420/10000000\n",
      " - 0s - loss: 54.6653 - val_loss: 50.7062\n",
      "Epoch 421/10000000\n",
      " - 0s - loss: 55.4263 - val_loss: 48.7379\n",
      "Epoch 422/10000000\n",
      " - 0s - loss: 54.7362 - val_loss: 48.0617\n",
      "Epoch 423/10000000\n",
      " - 0s - loss: 54.4959 - val_loss: 48.6898\n",
      "Epoch 424/10000000\n",
      " - 0s - loss: 54.1683 - val_loss: 47.8851\n",
      "Epoch 425/10000000\n",
      " - 0s - loss: 54.1330 - val_loss: 47.5516\n",
      "Epoch 426/10000000\n",
      " - 0s - loss: 54.4413 - val_loss: 47.8463\n",
      "Epoch 427/10000000\n",
      " - 0s - loss: 54.0072 - val_loss: 48.8165\n",
      "Epoch 428/10000000\n",
      " - 0s - loss: 54.0346 - val_loss: 47.7969\n",
      "Epoch 429/10000000\n",
      " - 0s - loss: 54.1703 - val_loss: 47.1862\n",
      "Epoch 430/10000000\n",
      " - 0s - loss: 54.4834 - val_loss: 47.3757\n",
      "Epoch 431/10000000\n",
      " - 0s - loss: 54.5875 - val_loss: 48.8997\n",
      "Epoch 432/10000000\n",
      " - 0s - loss: 53.9261 - val_loss: 47.2871\n",
      "Epoch 433/10000000\n",
      " - 0s - loss: 53.5757 - val_loss: 47.4052\n",
      "Epoch 434/10000000\n",
      " - 0s - loss: 53.2059 - val_loss: 49.2545\n",
      "Epoch 435/10000000\n",
      " - 0s - loss: 54.0051 - val_loss: 47.0872\n",
      "Epoch 436/10000000\n",
      " - 0s - loss: 55.9076 - val_loss: 46.8741\n",
      "Epoch 437/10000000\n",
      " - 0s - loss: 54.6919 - val_loss: 50.3118\n",
      "Epoch 438/10000000\n",
      " - 0s - loss: 53.8955 - val_loss: 46.6968\n",
      "Epoch 439/10000000\n",
      " - 0s - loss: 54.6705 - val_loss: 47.1576\n",
      "Epoch 440/10000000\n",
      " - 0s - loss: 54.0059 - val_loss: 50.9124\n",
      "Epoch 441/10000000\n",
      " - 0s - loss: 54.5727 - val_loss: 47.8060\n",
      "Epoch 442/10000000\n",
      " - 0s - loss: 53.4631 - val_loss: 46.6083\n",
      "Epoch 443/10000000\n",
      " - 0s - loss: 52.5124 - val_loss: 48.6509\n",
      "Epoch 444/10000000\n",
      " - 0s - loss: 54.0888 - val_loss: 47.3617\n",
      "Epoch 445/10000000\n",
      " - 0s - loss: 52.8079 - val_loss: 47.9495\n",
      "Epoch 446/10000000\n",
      " - 0s - loss: 55.8784 - val_loss: 46.5915\n",
      "Epoch 447/10000000\n",
      " - 0s - loss: 52.3287 - val_loss: 50.0995\n",
      "Epoch 448/10000000\n",
      " - 0s - loss: 53.9758 - val_loss: 46.8885\n",
      "Epoch 449/10000000\n",
      " - 0s - loss: 52.1042 - val_loss: 45.7973\n",
      "Epoch 450/10000000\n",
      " - 0s - loss: 52.1410 - val_loss: 46.0600\n",
      "Epoch 451/10000000\n",
      " - 0s - loss: 51.8740 - val_loss: 46.3637\n",
      "Epoch 452/10000000\n",
      " - 0s - loss: 51.6648 - val_loss: 45.5591\n",
      "Epoch 453/10000000\n",
      " - 0s - loss: 51.7578 - val_loss: 45.2883\n",
      "Epoch 454/10000000\n",
      " - 0s - loss: 51.7081 - val_loss: 45.1605\n",
      "Epoch 455/10000000\n",
      " - 0s - loss: 51.8862 - val_loss: 45.8474\n",
      "Epoch 456/10000000\n",
      " - 0s - loss: 51.6564 - val_loss: 46.3957\n",
      "Epoch 457/10000000\n",
      " - 0s - loss: 51.7121 - val_loss: 45.4366\n",
      "Epoch 458/10000000\n",
      " - 0s - loss: 51.5253 - val_loss: 45.1847\n",
      "Epoch 459/10000000\n",
      " - 0s - loss: 50.9862 - val_loss: 47.6703\n",
      "Epoch 460/10000000\n",
      " - 0s - loss: 51.9415 - val_loss: 45.5844\n",
      "Epoch 461/10000000\n",
      " - 0s - loss: 51.5780 - val_loss: 44.6610\n",
      "Epoch 462/10000000\n",
      " - 0s - loss: 51.2951 - val_loss: 47.4278\n",
      "Epoch 463/10000000\n",
      " - 0s - loss: 51.7215 - val_loss: 46.4014\n",
      "Epoch 464/10000000\n",
      " - 0s - loss: 51.0430 - val_loss: 44.9216\n",
      "Epoch 465/10000000\n",
      " - 0s - loss: 50.4058 - val_loss: 44.8725\n",
      "Epoch 466/10000000\n",
      " - 0s - loss: 50.3920 - val_loss: 44.6545\n",
      "Epoch 467/10000000\n",
      " - 0s - loss: 50.5572 - val_loss: 44.4574\n",
      "Epoch 468/10000000\n",
      " - 0s - loss: 50.6514 - val_loss: 44.7838\n",
      "Epoch 469/10000000\n",
      " - 0s - loss: 50.2285 - val_loss: 44.3914\n",
      "Epoch 470/10000000\n",
      " - 0s - loss: 50.1980 - val_loss: 44.3025\n",
      "Epoch 471/10000000\n",
      " - 0s - loss: 49.8706 - val_loss: 45.4408\n",
      "Epoch 472/10000000\n",
      " - 0s - loss: 49.9412 - val_loss: 43.8580\n",
      "Epoch 473/10000000\n",
      " - 0s - loss: 50.5610 - val_loss: 43.8383\n",
      "Epoch 474/10000000\n",
      " - 0s - loss: 50.8941 - val_loss: 43.6760\n",
      "Epoch 475/10000000\n",
      " - 0s - loss: 49.4571 - val_loss: 45.3870\n",
      "Epoch 476/10000000\n",
      " - 0s - loss: 50.1080 - val_loss: 44.7527\n",
      "Epoch 477/10000000\n",
      " - 0s - loss: 49.3307 - val_loss: 43.3585\n",
      "Epoch 478/10000000\n",
      " - 0s - loss: 49.5381 - val_loss: 43.2390\n",
      "Epoch 479/10000000\n",
      " - 0s - loss: 49.0989 - val_loss: 45.4245\n",
      "Epoch 480/10000000\n",
      " - 0s - loss: 50.3841 - val_loss: 45.7162\n",
      "Epoch 481/10000000\n",
      " - 0s - loss: 50.3953 - val_loss: 43.1250\n",
      "Epoch 482/10000000\n",
      " - 0s - loss: 48.7288 - val_loss: 45.2715\n",
      "Epoch 483/10000000\n",
      " - 0s - loss: 49.4793 - val_loss: 43.9653\n",
      "Epoch 484/10000000\n",
      " - 0s - loss: 48.7170 - val_loss: 42.8438\n",
      "Epoch 485/10000000\n",
      " - 0s - loss: 48.9205 - val_loss: 42.9120\n",
      "Epoch 486/10000000\n",
      " - 0s - loss: 48.6162 - val_loss: 44.1381\n",
      "Epoch 487/10000000\n",
      " - 0s - loss: 48.6685 - val_loss: 44.0259\n",
      "Epoch 488/10000000\n",
      " - 0s - loss: 48.4127 - val_loss: 42.7583\n",
      "Epoch 489/10000000\n",
      " - 0s - loss: 48.0126 - val_loss: 42.5899\n",
      "Epoch 490/10000000\n",
      " - 0s - loss: 50.5934 - val_loss: 42.2597\n",
      "Epoch 491/10000000\n",
      " - 0s - loss: 48.5715 - val_loss: 47.6801\n",
      "Epoch 492/10000000\n",
      " - 0s - loss: 50.5373 - val_loss: 42.5741\n",
      "Epoch 493/10000000\n",
      " - 0s - loss: 48.0124 - val_loss: 42.0312\n",
      "Epoch 494/10000000\n",
      " - 0s - loss: 48.1448 - val_loss: 42.9234\n",
      "Epoch 495/10000000\n",
      " - 0s - loss: 47.6891 - val_loss: 44.8762\n",
      "Epoch 496/10000000\n",
      " - 0s - loss: 48.1864 - val_loss: 41.6997\n",
      "Epoch 497/10000000\n",
      " - 0s - loss: 50.0593 - val_loss: 41.6572\n",
      "Epoch 498/10000000\n",
      " - 0s - loss: 47.0988 - val_loss: 48.9984\n",
      "Epoch 499/10000000\n",
      " - 0s - loss: 52.4313 - val_loss: 41.9214\n",
      "Epoch 500/10000000\n",
      " - 0s - loss: 52.4982 - val_loss: 46.8336\n",
      "Epoch 501/10000000\n",
      " - 0s - loss: 52.8972 - val_loss: 43.8065\n",
      "Epoch 502/10000000\n",
      " - 0s - loss: 50.6345 - val_loss: 49.8608\n",
      "Epoch 503/10000000\n",
      " - 0s - loss: 49.4933 - val_loss: 42.4264\n",
      "Epoch 504/10000000\n",
      " - 0s - loss: 51.5394 - val_loss: 41.7772\n",
      "Epoch 505/10000000\n",
      " - 0s - loss: 47.0302 - val_loss: 47.7760\n",
      "Epoch 506/10000000\n",
      " - 0s - loss: 50.7563 - val_loss: 42.0184\n",
      "Epoch 507/10000000\n",
      " - 0s - loss: 47.8048 - val_loss: 41.6850\n",
      "Epoch 508/10000000\n",
      " - 0s - loss: 47.3292 - val_loss: 43.8529\n",
      "Epoch 509/10000000\n",
      " - 0s - loss: 48.0199 - val_loss: 42.8508\n",
      "Epoch 510/10000000\n",
      " - 0s - loss: 46.4946 - val_loss: 40.5939\n",
      "Epoch 511/10000000\n",
      " - 0s - loss: 46.7184 - val_loss: 40.6292\n",
      "Epoch 512/10000000\n",
      " - 0s - loss: 46.9841 - val_loss: 42.1164\n",
      "Epoch 513/10000000\n",
      " - 0s - loss: 46.1113 - val_loss: 40.3570\n",
      "Epoch 514/10000000\n",
      " - 0s - loss: 45.9990 - val_loss: 40.4610\n",
      "Epoch 515/10000000\n",
      " - 0s - loss: 46.1314 - val_loss: 41.3798\n",
      "Epoch 516/10000000\n",
      " - 0s - loss: 45.7764 - val_loss: 40.0617\n",
      "Epoch 517/10000000\n",
      " - 0s - loss: 45.6163 - val_loss: 40.7348\n",
      "Epoch 518/10000000\n",
      " - 0s - loss: 45.4371 - val_loss: 41.8863\n",
      "Epoch 519/10000000\n",
      " - 0s - loss: 47.2269 - val_loss: 40.4595\n",
      "Epoch 520/10000000\n",
      " - 0s - loss: 48.1412 - val_loss: 40.1435\n",
      "Epoch 521/10000000\n",
      " - 0s - loss: 46.6859 - val_loss: 42.9431\n",
      "Epoch 522/10000000\n",
      " - 0s - loss: 46.1292 - val_loss: 40.1548\n",
      "Epoch 523/10000000\n",
      " - 0s - loss: 44.6216 - val_loss: 39.3922\n",
      "Epoch 524/10000000\n",
      " - 0s - loss: 45.6007 - val_loss: 39.2289\n",
      "Epoch 525/10000000\n",
      " - 0s - loss: 44.9125 - val_loss: 39.8547\n",
      "Epoch 526/10000000\n",
      " - 0s - loss: 44.7261 - val_loss: 41.3140\n",
      "Epoch 527/10000000\n",
      " - 0s - loss: 44.8419 - val_loss: 39.0210\n",
      "Epoch 528/10000000\n",
      " - 0s - loss: 44.4183 - val_loss: 38.9085\n",
      "Epoch 529/10000000\n",
      " - 0s - loss: 44.4812 - val_loss: 39.4799\n",
      "Epoch 530/10000000\n",
      " - 0s - loss: 44.1855 - val_loss: 38.9784\n",
      "Epoch 531/10000000\n",
      " - 0s - loss: 44.1115 - val_loss: 39.5426\n",
      "Epoch 532/10000000\n",
      " - 0s - loss: 44.1222 - val_loss: 40.0976\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 533/10000000\n",
      " - 0s - loss: 44.4090 - val_loss: 39.3877\n",
      "Epoch 534/10000000\n",
      " - 0s - loss: 43.9356 - val_loss: 38.3515\n",
      "Epoch 535/10000000\n",
      " - 0s - loss: 44.1925 - val_loss: 38.3825\n",
      "Epoch 536/10000000\n",
      " - 0s - loss: 43.5761 - val_loss: 39.7465\n",
      "Epoch 537/10000000\n",
      " - 0s - loss: 43.9470 - val_loss: 39.4896\n",
      "Epoch 538/10000000\n",
      " - 0s - loss: 43.9855 - val_loss: 38.6534\n",
      "Epoch 539/10000000\n",
      " - 0s - loss: 43.5445 - val_loss: 37.8950\n",
      "Epoch 540/10000000\n",
      " - 0s - loss: 43.7675 - val_loss: 37.6924\n",
      "Epoch 541/10000000\n",
      " - 0s - loss: 43.8592 - val_loss: 39.0426\n",
      "Epoch 542/10000000\n",
      " - 0s - loss: 42.7757 - val_loss: 37.5506\n",
      "Epoch 543/10000000\n",
      " - 0s - loss: 45.9564 - val_loss: 37.4781\n",
      "Epoch 544/10000000\n",
      " - 0s - loss: 43.7439 - val_loss: 39.6255\n",
      "Epoch 545/10000000\n",
      " - 0s - loss: 42.9138 - val_loss: 37.4169\n",
      "Epoch 546/10000000\n",
      " - 0s - loss: 43.3157 - val_loss: 37.4114\n",
      "Epoch 547/10000000\n",
      " - 0s - loss: 42.5154 - val_loss: 39.7301\n",
      "Epoch 548/10000000\n",
      " - 0s - loss: 44.1312 - val_loss: 38.0989\n",
      "Epoch 549/10000000\n",
      " - 0s - loss: 44.9011 - val_loss: 37.9881\n",
      "Epoch 550/10000000\n",
      " - 0s - loss: 43.5098 - val_loss: 38.5084\n",
      "Epoch 551/10000000\n",
      " - 0s - loss: 42.6401 - val_loss: 40.1905\n",
      "Epoch 552/10000000\n",
      " - 0s - loss: 43.3522 - val_loss: 36.8808\n",
      "Epoch 553/10000000\n",
      " - 0s - loss: 42.6025 - val_loss: 37.2333\n",
      "Epoch 554/10000000\n",
      " - 0s - loss: 42.4873 - val_loss: 42.5783\n",
      "Epoch 555/10000000\n",
      " - 0s - loss: 46.0663 - val_loss: 36.8986\n",
      "Epoch 556/10000000\n",
      " - 0s - loss: 44.3091 - val_loss: 39.6837\n",
      "Epoch 557/10000000\n",
      " - 0s - loss: 46.9807 - val_loss: 37.7624\n",
      "Epoch 558/10000000\n",
      " - 0s - loss: 41.8210 - val_loss: 36.4532\n",
      "Epoch 559/10000000\n",
      " - 0s - loss: 41.5070 - val_loss: 36.5608\n",
      "Epoch 560/10000000\n",
      " - 0s - loss: 41.8712 - val_loss: 37.2800\n",
      "Epoch 561/10000000\n",
      " - 0s - loss: 41.2799 - val_loss: 38.2398\n",
      "Epoch 562/10000000\n",
      " - 0s - loss: 41.1530 - val_loss: 35.9534\n",
      "Epoch 563/10000000\n",
      " - 0s - loss: 41.5569 - val_loss: 36.0592\n",
      "Epoch 564/10000000\n",
      " - 0s - loss: 41.3282 - val_loss: 37.1219\n",
      "Epoch 565/10000000\n",
      " - 0s - loss: 40.7337 - val_loss: 40.2904\n",
      "Epoch 566/10000000\n",
      " - 0s - loss: 42.7244 - val_loss: 36.3654\n",
      "Epoch 567/10000000\n",
      " - 0s - loss: 42.1171 - val_loss: 35.8687\n",
      "Epoch 568/10000000\n",
      " - 0s - loss: 43.1826 - val_loss: 37.2063\n",
      "Epoch 569/10000000\n",
      " - 0s - loss: 40.2915 - val_loss: 35.2211\n",
      "Epoch 570/10000000\n",
      " - 0s - loss: 41.4292 - val_loss: 35.1444\n",
      "Epoch 571/10000000\n",
      " - 0s - loss: 39.9594 - val_loss: 37.6754\n",
      "Epoch 572/10000000\n",
      " - 0s - loss: 41.0702 - val_loss: 37.3637\n",
      "Epoch 573/10000000\n",
      " - 0s - loss: 40.1456 - val_loss: 34.8915\n",
      "Epoch 574/10000000\n",
      " - 0s - loss: 40.3457 - val_loss: 34.8293\n",
      "Epoch 575/10000000\n",
      " - 0s - loss: 40.3507 - val_loss: 35.3009\n",
      "Epoch 576/10000000\n",
      " - 0s - loss: 39.5893 - val_loss: 34.7667\n",
      "Epoch 577/10000000\n",
      " - 0s - loss: 39.4662 - val_loss: 34.7310\n",
      "Epoch 578/10000000\n",
      " - 0s - loss: 39.5851 - val_loss: 34.6704\n",
      "Epoch 579/10000000\n",
      " - 0s - loss: 40.0071 - val_loss: 34.5031\n",
      "Epoch 580/10000000\n",
      " - 0s - loss: 40.4894 - val_loss: 37.2882\n",
      "Epoch 581/10000000\n",
      " - 0s - loss: 40.2013 - val_loss: 34.2801\n",
      "Epoch 582/10000000\n",
      " - 0s - loss: 39.0111 - val_loss: 34.2425\n",
      "Epoch 583/10000000\n",
      " - 0s - loss: 39.0200 - val_loss: 33.9647\n",
      "Epoch 584/10000000\n",
      " - 0s - loss: 39.1800 - val_loss: 33.8753\n",
      "Epoch 585/10000000\n",
      " - 0s - loss: 39.3881 - val_loss: 34.5478\n",
      "Epoch 586/10000000\n",
      " - 0s - loss: 39.0717 - val_loss: 33.7637\n",
      "Epoch 587/10000000\n",
      " - 0s - loss: 38.3797 - val_loss: 35.7211\n",
      "Epoch 588/10000000\n",
      " - 0s - loss: 39.0330 - val_loss: 34.1193\n",
      "Epoch 589/10000000\n",
      " - 0s - loss: 38.1347 - val_loss: 33.4503\n",
      "Epoch 590/10000000\n",
      " - 0s - loss: 38.7356 - val_loss: 33.3745\n",
      "Epoch 591/10000000\n",
      " - 0s - loss: 38.0715 - val_loss: 34.6323\n",
      "Epoch 592/10000000\n",
      " - 0s - loss: 38.5164 - val_loss: 34.6722\n",
      "Epoch 593/10000000\n",
      " - 0s - loss: 38.3797 - val_loss: 34.0291\n",
      "Epoch 594/10000000\n",
      " - 0s - loss: 38.2082 - val_loss: 33.0835\n",
      "Epoch 595/10000000\n",
      " - 0s - loss: 37.6259 - val_loss: 34.2865\n",
      "Epoch 596/10000000\n",
      " - 0s - loss: 38.1263 - val_loss: 33.7832\n",
      "Epoch 597/10000000\n",
      " - 0s - loss: 37.7360 - val_loss: 32.7868\n",
      "Epoch 598/10000000\n",
      " - 0s - loss: 37.8655 - val_loss: 33.1834\n",
      "Epoch 599/10000000\n",
      " - 0s - loss: 37.2498 - val_loss: 34.9120\n",
      "Epoch 600/10000000\n",
      " - 0s - loss: 37.8771 - val_loss: 32.7121\n",
      "Epoch 601/10000000\n",
      " - 0s - loss: 36.8914 - val_loss: 32.8895\n",
      "Epoch 602/10000000\n",
      " - 0s - loss: 38.7483 - val_loss: 33.0640\n",
      "Epoch 603/10000000\n",
      " - 0s - loss: 38.3094 - val_loss: 36.9875\n",
      "Epoch 604/10000000\n",
      " - 0s - loss: 37.7924 - val_loss: 32.7102\n",
      "Epoch 605/10000000\n",
      " - 0s - loss: 39.8996 - val_loss: 32.1398\n",
      "Epoch 606/10000000\n",
      " - 0s - loss: 37.1522 - val_loss: 35.9430\n",
      "Epoch 607/10000000\n",
      " - 0s - loss: 38.1023 - val_loss: 32.2205\n",
      "Epoch 608/10000000\n",
      " - 0s - loss: 36.4021 - val_loss: 31.9865\n",
      "Epoch 609/10000000\n",
      " - 0s - loss: 38.5816 - val_loss: 31.9915\n",
      "Epoch 610/10000000\n",
      " - 0s - loss: 36.4124 - val_loss: 31.6173\n",
      "Epoch 611/10000000\n",
      " - 0s - loss: 36.7262 - val_loss: 31.9326\n",
      "Epoch 612/10000000\n",
      " - 0s - loss: 36.3028 - val_loss: 33.0279\n",
      "Epoch 613/10000000\n",
      " - 0s - loss: 36.4814 - val_loss: 31.7927\n",
      "Epoch 614/10000000\n",
      " - 0s - loss: 35.7956 - val_loss: 32.0581\n",
      "Epoch 615/10000000\n",
      " - 0s - loss: 35.8612 - val_loss: 31.2732\n",
      "Epoch 616/10000000\n",
      " - 0s - loss: 36.4619 - val_loss: 31.1381\n",
      "Epoch 617/10000000\n",
      " - 0s - loss: 35.6618 - val_loss: 31.5098\n",
      "Epoch 618/10000000\n",
      " - 0s - loss: 35.5659 - val_loss: 31.6977\n",
      "Epoch 619/10000000\n",
      " - 0s - loss: 35.5246 - val_loss: 32.6620\n",
      "Epoch 620/10000000\n",
      " - 0s - loss: 35.7560 - val_loss: 31.8747\n",
      "Epoch 621/10000000\n",
      " - 0s - loss: 35.4552 - val_loss: 31.0884\n",
      "Epoch 622/10000000\n",
      " - 0s - loss: 35.3598 - val_loss: 31.1065\n",
      "Epoch 623/10000000\n",
      " - 0s - loss: 35.5319 - val_loss: 30.7639\n",
      "Epoch 624/10000000\n",
      " - 0s - loss: 35.4550 - val_loss: 30.9851\n",
      "Epoch 625/10000000\n",
      " - 0s - loss: 35.1073 - val_loss: 30.9854\n",
      "Epoch 626/10000000\n",
      " - 0s - loss: 34.4445 - val_loss: 30.5919\n",
      "Epoch 627/10000000\n",
      " - 0s - loss: 35.6751 - val_loss: 30.2337\n",
      "Epoch 628/10000000\n",
      " - 0s - loss: 34.9684 - val_loss: 32.8660\n",
      "Epoch 629/10000000\n",
      " - 0s - loss: 34.8096 - val_loss: 30.0168\n",
      "Epoch 630/10000000\n",
      " - 0s - loss: 35.4496 - val_loss: 29.9435\n",
      "Epoch 631/10000000\n",
      " - 0s - loss: 34.2266 - val_loss: 33.8997\n",
      "Epoch 632/10000000\n",
      " - 0s - loss: 35.1721 - val_loss: 29.9375\n",
      "Epoch 633/10000000\n",
      " - 0s - loss: 35.4320 - val_loss: 29.9497\n",
      "Epoch 634/10000000\n",
      " - 0s - loss: 33.5192 - val_loss: 33.1474\n",
      "Epoch 635/10000000\n",
      " - 0s - loss: 35.7433 - val_loss: 30.5436\n",
      "Epoch 636/10000000\n",
      " - 0s - loss: 34.1494 - val_loss: 30.3174\n",
      "Epoch 637/10000000\n",
      " - 0s - loss: 35.4151 - val_loss: 29.8321\n",
      "Epoch 638/10000000\n",
      " - 0s - loss: 34.8562 - val_loss: 33.4501\n",
      "Epoch 639/10000000\n",
      " - 0s - loss: 35.3817 - val_loss: 29.3899\n",
      "Epoch 640/10000000\n",
      " - 0s - loss: 34.4456 - val_loss: 28.9501\n",
      "Epoch 641/10000000\n",
      " - 0s - loss: 32.9580 - val_loss: 29.8233\n",
      "Epoch 642/10000000\n",
      " - 0s - loss: 33.5611 - val_loss: 29.9037\n",
      "Epoch 643/10000000\n",
      " - 0s - loss: 32.9348 - val_loss: 28.6483\n",
      "Epoch 644/10000000\n",
      " - 0s - loss: 33.3564 - val_loss: 28.6011\n",
      "Epoch 645/10000000\n",
      " - 0s - loss: 33.1229 - val_loss: 30.1682\n",
      "Epoch 646/10000000\n",
      " - 0s - loss: 33.0439 - val_loss: 29.3825\n",
      "Epoch 647/10000000\n",
      " - 0s - loss: 32.7995 - val_loss: 28.4940\n",
      "Epoch 648/10000000\n",
      " - 0s - loss: 33.3330 - val_loss: 28.3466\n",
      "Epoch 649/10000000\n",
      " - 0s - loss: 32.2112 - val_loss: 30.8094\n",
      "Epoch 650/10000000\n",
      " - 0s - loss: 33.3878 - val_loss: 28.8084\n",
      "Epoch 651/10000000\n",
      " - 0s - loss: 32.4881 - val_loss: 28.5370\n",
      "Epoch 652/10000000\n",
      " - 0s - loss: 33.5150 - val_loss: 28.4791\n",
      "Epoch 653/10000000\n",
      " - 0s - loss: 33.1716 - val_loss: 30.0679\n",
      "Epoch 654/10000000\n",
      " - 0s - loss: 32.0833 - val_loss: 27.7898\n",
      "Epoch 655/10000000\n",
      " - 0s - loss: 32.4708 - val_loss: 27.6365\n",
      "Epoch 656/10000000\n",
      " - 0s - loss: 32.1201 - val_loss: 28.5038\n",
      "Epoch 657/10000000\n",
      " - 0s - loss: 31.5110 - val_loss: 27.6014\n",
      "Epoch 658/10000000\n",
      " - 0s - loss: 31.4339 - val_loss: 27.5281\n",
      "Epoch 659/10000000\n",
      " - 0s - loss: 32.0997 - val_loss: 27.6060\n",
      "Epoch 660/10000000\n",
      " - 0s - loss: 31.9980 - val_loss: 31.0972\n",
      "Epoch 661/10000000\n",
      " - 0s - loss: 32.4230 - val_loss: 27.2167\n",
      "Epoch 662/10000000\n",
      " - 0s - loss: 31.8997 - val_loss: 27.8153\n",
      "Epoch 663/10000000\n",
      " - 0s - loss: 32.2404 - val_loss: 30.0519\n",
      "Epoch 664/10000000\n",
      " - 0s - loss: 32.2937 - val_loss: 28.3736\n",
      "Epoch 665/10000000\n",
      " - 0s - loss: 30.8618 - val_loss: 26.9613\n",
      "Epoch 666/10000000\n",
      " - 0s - loss: 32.0037 - val_loss: 27.1102\n",
      "Epoch 667/10000000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 0s - loss: 30.6087 - val_loss: 29.8113\n",
      "Epoch 668/10000000\n",
      " - 0s - loss: 32.9195 - val_loss: 29.7136\n",
      "Epoch 669/10000000\n",
      " - 0s - loss: 31.3009 - val_loss: 26.6765\n",
      "Epoch 670/10000000\n",
      " - 0s - loss: 30.9495 - val_loss: 26.6389\n",
      "Epoch 671/10000000\n",
      " - 0s - loss: 29.9115 - val_loss: 30.0596\n",
      "Epoch 672/10000000\n",
      " - 0s - loss: 32.6541 - val_loss: 27.2927\n",
      "Epoch 673/10000000\n",
      " - 0s - loss: 29.8676 - val_loss: 26.5845\n",
      "Epoch 674/10000000\n",
      " - 0s - loss: 31.2479 - val_loss: 26.3071\n",
      "Epoch 675/10000000\n",
      " - 0s - loss: 29.6489 - val_loss: 31.3787\n",
      "Epoch 676/10000000\n",
      " - 0s - loss: 32.5563 - val_loss: 26.6966\n",
      "Epoch 677/10000000\n",
      " - 0s - loss: 29.6853 - val_loss: 25.8467\n",
      "Epoch 678/10000000\n",
      " - 0s - loss: 30.2753 - val_loss: 26.1335\n",
      "Epoch 679/10000000\n",
      " - 0s - loss: 29.7688 - val_loss: 26.0660\n",
      "Epoch 680/10000000\n",
      " - 0s - loss: 29.4099 - val_loss: 25.7254\n",
      "Epoch 681/10000000\n",
      " - 0s - loss: 30.5197 - val_loss: 25.9001\n",
      "Epoch 682/10000000\n",
      " - 0s - loss: 29.6967 - val_loss: 28.6877\n",
      "Epoch 683/10000000\n",
      " - 0s - loss: 29.7637 - val_loss: 25.5571\n",
      "Epoch 684/10000000\n",
      " - 0s - loss: 30.5601 - val_loss: 25.3073\n",
      "Epoch 685/10000000\n",
      " - 0s - loss: 29.1961 - val_loss: 28.5965\n",
      "Epoch 686/10000000\n",
      " - 0s - loss: 30.2524 - val_loss: 25.8250\n",
      "Epoch 687/10000000\n",
      " - 0s - loss: 29.3624 - val_loss: 25.1044\n",
      "Epoch 688/10000000\n",
      " - 0s - loss: 29.3730 - val_loss: 25.7144\n",
      "Epoch 689/10000000\n",
      " - 0s - loss: 28.5161 - val_loss: 24.8712\n",
      "Epoch 690/10000000\n",
      " - 0s - loss: 29.0762 - val_loss: 24.7828\n",
      "Epoch 691/10000000\n",
      " - 0s - loss: 28.6009 - val_loss: 26.2627\n",
      "Epoch 692/10000000\n",
      " - 0s - loss: 29.9335 - val_loss: 26.1854\n",
      "Epoch 693/10000000\n",
      " - 0s - loss: 28.9901 - val_loss: 24.9880\n",
      "Epoch 694/10000000\n",
      " - 0s - loss: 29.0507 - val_loss: 25.9113\n",
      "Epoch 695/10000000\n",
      " - 0s - loss: 28.8689 - val_loss: 26.0019\n",
      "Epoch 696/10000000\n",
      " - 0s - loss: 28.0079 - val_loss: 24.5968\n",
      "Epoch 697/10000000\n",
      " - 0s - loss: 28.7768 - val_loss: 24.4804\n",
      "Epoch 698/10000000\n",
      " - 0s - loss: 27.9009 - val_loss: 27.3487\n",
      "Epoch 699/10000000\n",
      " - 0s - loss: 29.2096 - val_loss: 24.5197\n",
      "Epoch 700/10000000\n",
      " - 0s - loss: 27.7884 - val_loss: 25.7604\n",
      "Epoch 701/10000000\n",
      " - 0s - loss: 31.4068 - val_loss: 23.9151\n",
      "Epoch 702/10000000\n",
      " - 0s - loss: 27.9165 - val_loss: 26.9490\n",
      "Epoch 703/10000000\n",
      " - 0s - loss: 28.8251 - val_loss: 24.1663\n",
      "Epoch 704/10000000\n",
      " - 0s - loss: 27.5114 - val_loss: 23.7295\n",
      "Epoch 705/10000000\n",
      " - 0s - loss: 28.2602 - val_loss: 24.2703\n",
      "Epoch 706/10000000\n",
      " - 0s - loss: 27.2702 - val_loss: 23.6318\n",
      "Epoch 707/10000000\n",
      " - 0s - loss: 27.1974 - val_loss: 23.7035\n",
      "Epoch 708/10000000\n",
      " - 0s - loss: 27.4909 - val_loss: 23.6323\n",
      "Epoch 709/10000000\n",
      " - 0s - loss: 27.6809 - val_loss: 23.4785\n",
      "Epoch 710/10000000\n",
      " - 0s - loss: 27.1358 - val_loss: 24.4871\n",
      "Epoch 711/10000000\n",
      " - 0s - loss: 26.9634 - val_loss: 23.6837\n",
      "Epoch 712/10000000\n",
      " - 0s - loss: 26.6737 - val_loss: 23.1477\n",
      "Epoch 713/10000000\n",
      " - 0s - loss: 26.9595 - val_loss: 23.5053\n",
      "Epoch 714/10000000\n",
      " - 0s - loss: 27.2649 - val_loss: 25.6302\n",
      "Epoch 715/10000000\n",
      " - 0s - loss: 26.9907 - val_loss: 22.8533\n",
      "Epoch 716/10000000\n",
      " - 0s - loss: 27.6653 - val_loss: 22.7810\n",
      "Epoch 717/10000000\n",
      " - 0s - loss: 27.8643 - val_loss: 25.3534\n",
      "Epoch 718/10000000\n",
      " - 0s - loss: 26.6980 - val_loss: 22.9178\n",
      "Epoch 719/10000000\n",
      " - 0s - loss: 27.1137 - val_loss: 22.6060\n",
      "Epoch 720/10000000\n",
      " - 0s - loss: 26.0906 - val_loss: 25.2510\n",
      "Epoch 721/10000000\n",
      " - 0s - loss: 26.7817 - val_loss: 22.4993\n",
      "Epoch 722/10000000\n",
      " - 0s - loss: 26.8110 - val_loss: 22.6004\n",
      "Epoch 723/10000000\n",
      " - 0s - loss: 26.0588 - val_loss: 24.7273\n",
      "Epoch 724/10000000\n",
      " - 0s - loss: 27.0661 - val_loss: 22.3043\n",
      "Epoch 725/10000000\n",
      " - 0s - loss: 25.8184 - val_loss: 22.2867\n",
      "Epoch 726/10000000\n",
      " - 0s - loss: 25.8242 - val_loss: 23.5167\n",
      "Epoch 727/10000000\n",
      " - 0s - loss: 25.7145 - val_loss: 22.3214\n",
      "Epoch 728/10000000\n",
      " - 0s - loss: 25.3271 - val_loss: 21.8892\n",
      "Epoch 729/10000000\n",
      " - 0s - loss: 25.4795 - val_loss: 22.6980\n",
      "Epoch 730/10000000\n",
      " - 0s - loss: 25.0892 - val_loss: 21.7827\n",
      "Epoch 731/10000000\n",
      " - 0s - loss: 25.0160 - val_loss: 21.8049\n",
      "Epoch 732/10000000\n",
      " - 0s - loss: 24.8869 - val_loss: 22.0986\n",
      "Epoch 733/10000000\n",
      " - 0s - loss: 25.0237 - val_loss: 21.6556\n",
      "Epoch 734/10000000\n",
      " - 0s - loss: 24.6006 - val_loss: 21.4219\n",
      "Epoch 735/10000000\n",
      " - 0s - loss: 25.3699 - val_loss: 21.2926\n",
      "Epoch 736/10000000\n",
      " - 0s - loss: 24.7002 - val_loss: 21.8598\n",
      "Epoch 737/10000000\n",
      " - 0s - loss: 25.5274 - val_loss: 21.3989\n",
      "Epoch 738/10000000\n",
      " - 0s - loss: 24.4262 - val_loss: 21.9065\n",
      "Epoch 739/10000000\n",
      " - 0s - loss: 25.6839 - val_loss: 21.6284\n",
      "Epoch 740/10000000\n",
      " - 0s - loss: 24.6776 - val_loss: 22.4319\n",
      "Epoch 741/10000000\n",
      " - 0s - loss: 24.2802 - val_loss: 20.8636\n",
      "Epoch 742/10000000\n",
      " - 0s - loss: 25.0776 - val_loss: 21.6364\n",
      "Epoch 743/10000000\n",
      " - 0s - loss: 25.6642 - val_loss: 25.3038\n",
      "Epoch 744/10000000\n",
      " - 0s - loss: 24.8586 - val_loss: 21.3748\n",
      "Epoch 745/10000000\n",
      " - 0s - loss: 27.5410 - val_loss: 20.9045\n",
      "Epoch 746/10000000\n",
      " - 0s - loss: 24.5573 - val_loss: 24.6313\n",
      "Epoch 747/10000000\n",
      " - 0s - loss: 25.5365 - val_loss: 20.6252\n",
      "Epoch 748/10000000\n",
      " - 0s - loss: 23.9305 - val_loss: 20.6522\n",
      "Epoch 749/10000000\n",
      " - 0s - loss: 24.0405 - val_loss: 21.1689\n",
      "Epoch 750/10000000\n",
      " - 0s - loss: 23.5161 - val_loss: 20.5290\n",
      "Epoch 751/10000000\n",
      " - 0s - loss: 23.4155 - val_loss: 20.4936\n",
      "Epoch 752/10000000\n",
      " - 0s - loss: 23.3809 - val_loss: 20.6717\n",
      "Epoch 753/10000000\n",
      " - 0s - loss: 23.1126 - val_loss: 20.1382\n",
      "Epoch 754/10000000\n",
      " - 0s - loss: 23.2107 - val_loss: 20.1820\n",
      "Epoch 755/10000000\n",
      " - 0s - loss: 22.9634 - val_loss: 20.1897\n",
      "Epoch 756/10000000\n",
      " - 0s - loss: 22.9937 - val_loss: 20.4083\n",
      "Epoch 757/10000000\n",
      " - 0s - loss: 23.2671 - val_loss: 20.5406\n",
      "Epoch 758/10000000\n",
      " - 0s - loss: 22.8233 - val_loss: 19.8933\n",
      "Epoch 759/10000000\n",
      " - 0s - loss: 23.2438 - val_loss: 19.7453\n",
      "Epoch 760/10000000\n",
      " - 0s - loss: 22.7872 - val_loss: 21.6490\n",
      "Epoch 761/10000000\n",
      " - 0s - loss: 23.2631 - val_loss: 19.5362\n",
      "Epoch 762/10000000\n",
      " - 0s - loss: 22.5655 - val_loss: 19.6762\n",
      "Epoch 763/10000000\n",
      " - 0s - loss: 22.4861 - val_loss: 19.6426\n",
      "Epoch 764/10000000\n",
      " - 0s - loss: 22.3962 - val_loss: 19.6404\n",
      "Epoch 765/10000000\n",
      " - 0s - loss: 22.2646 - val_loss: 19.2288\n",
      "Epoch 766/10000000\n",
      " - 0s - loss: 22.2097 - val_loss: 20.0277\n",
      "Epoch 767/10000000\n",
      " - 0s - loss: 22.2323 - val_loss: 19.7768\n",
      "Epoch 768/10000000\n",
      " - 0s - loss: 22.4490 - val_loss: 19.2523\n",
      "Epoch 769/10000000\n",
      " - 0s - loss: 22.5760 - val_loss: 19.1430\n",
      "Epoch 770/10000000\n",
      " - 0s - loss: 21.9095 - val_loss: 20.4901\n",
      "Epoch 771/10000000\n",
      " - 0s - loss: 22.6566 - val_loss: 19.5234\n",
      "Epoch 772/10000000\n",
      " - 0s - loss: 21.8575 - val_loss: 18.9588\n",
      "Epoch 773/10000000\n",
      " - 0s - loss: 21.7401 - val_loss: 19.8039\n",
      "Epoch 774/10000000\n",
      " - 0s - loss: 22.0680 - val_loss: 20.9123\n",
      "Epoch 775/10000000\n",
      " - 0s - loss: 22.3034 - val_loss: 18.7284\n",
      "Epoch 776/10000000\n",
      " - 0s - loss: 21.5383 - val_loss: 19.2898\n",
      "Epoch 777/10000000\n",
      " - 0s - loss: 21.3252 - val_loss: 18.7669\n",
      "Epoch 778/10000000\n",
      " - 0s - loss: 21.7963 - val_loss: 18.5170\n",
      "Epoch 779/10000000\n",
      " - 0s - loss: 21.2055 - val_loss: 19.6394\n",
      "Epoch 780/10000000\n",
      " - 0s - loss: 21.7032 - val_loss: 18.5021\n",
      "Epoch 781/10000000\n",
      " - 0s - loss: 21.1241 - val_loss: 18.4440\n",
      "Epoch 782/10000000\n",
      " - 0s - loss: 21.2770 - val_loss: 18.8223\n",
      "Epoch 783/10000000\n",
      " - 0s - loss: 21.5436 - val_loss: 19.9323\n",
      "Epoch 784/10000000\n",
      " - 0s - loss: 20.7349 - val_loss: 18.5084\n",
      "Epoch 785/10000000\n",
      " - 0s - loss: 21.8825 - val_loss: 18.1441\n",
      "Epoch 786/10000000\n",
      " - 0s - loss: 21.1420 - val_loss: 19.5939\n",
      "Epoch 787/10000000\n",
      " - 0s - loss: 20.9727 - val_loss: 18.2904\n",
      "Epoch 788/10000000\n",
      " - 0s - loss: 20.5647 - val_loss: 17.8689\n",
      "Epoch 789/10000000\n",
      " - 0s - loss: 21.1670 - val_loss: 17.7831\n",
      "Epoch 790/10000000\n",
      " - 0s - loss: 20.6098 - val_loss: 17.7652\n",
      "Epoch 791/10000000\n",
      " - 0s - loss: 20.5475 - val_loss: 18.3958\n",
      "Epoch 792/10000000\n",
      " - 0s - loss: 20.3766 - val_loss: 18.2227\n",
      "Epoch 793/10000000\n",
      " - 0s - loss: 20.6159 - val_loss: 17.6712\n",
      "Epoch 794/10000000\n",
      " - 0s - loss: 20.1360 - val_loss: 18.3858\n",
      "Epoch 795/10000000\n",
      " - 0s - loss: 20.5734 - val_loss: 18.0824\n",
      "Epoch 796/10000000\n",
      " - 0s - loss: 19.5640 - val_loss: 17.9445\n",
      "Epoch 797/10000000\n",
      " - 0s - loss: 21.1035 - val_loss: 17.4411\n",
      "Epoch 798/10000000\n",
      " - 0s - loss: 19.7528 - val_loss: 18.9797\n",
      "Epoch 799/10000000\n",
      " - 0s - loss: 20.1566 - val_loss: 17.3238\n",
      "Epoch 800/10000000\n",
      " - 0s - loss: 19.4962 - val_loss: 17.5925\n",
      "Epoch 801/10000000\n",
      " - 0s - loss: 20.7258 - val_loss: 17.1615\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 802/10000000\n",
      " - 0s - loss: 19.8665 - val_loss: 17.3853\n",
      "Epoch 803/10000000\n",
      " - 0s - loss: 19.4683 - val_loss: 17.1663\n",
      "Epoch 804/10000000\n",
      " - 0s - loss: 19.2933 - val_loss: 16.9227\n",
      "Epoch 805/10000000\n",
      " - 0s - loss: 19.5463 - val_loss: 17.2156\n",
      "Epoch 806/10000000\n",
      " - 0s - loss: 19.8774 - val_loss: 17.0710\n",
      "Epoch 807/10000000\n",
      " - 0s - loss: 19.6649 - val_loss: 16.8830\n",
      "Epoch 808/10000000\n",
      " - 0s - loss: 19.0174 - val_loss: 18.4748\n",
      "Epoch 809/10000000\n",
      " - 0s - loss: 19.9685 - val_loss: 17.3752\n",
      "Epoch 810/10000000\n",
      " - 0s - loss: 19.2507 - val_loss: 17.1456\n",
      "Epoch 811/10000000\n",
      " - 0s - loss: 19.5613 - val_loss: 17.3937\n",
      "Epoch 812/10000000\n",
      " - 0s - loss: 19.3270 - val_loss: 17.2942\n",
      "Epoch 813/10000000\n",
      " - 0s - loss: 18.6998 - val_loss: 16.6173\n",
      "Epoch 814/10000000\n",
      " - 0s - loss: 19.3615 - val_loss: 16.2984\n",
      "Epoch 815/10000000\n",
      " - 0s - loss: 18.4755 - val_loss: 17.3780\n",
      "Epoch 816/10000000\n",
      " - 0s - loss: 18.8318 - val_loss: 16.2087\n",
      "Epoch 817/10000000\n",
      " - 0s - loss: 18.4228 - val_loss: 16.1822\n",
      "Epoch 818/10000000\n",
      " - 0s - loss: 18.7755 - val_loss: 16.8457\n",
      "Epoch 819/10000000\n",
      " - 0s - loss: 18.8626 - val_loss: 16.2233\n",
      "Epoch 820/10000000\n",
      " - 0s - loss: 20.3428 - val_loss: 16.1050\n",
      "Epoch 821/10000000\n",
      " - 0s - loss: 17.8823 - val_loss: 20.1634\n",
      "Epoch 822/10000000\n",
      " - 0s - loss: 20.6731 - val_loss: 15.9884\n",
      "Epoch 823/10000000\n",
      " - 0s - loss: 19.3431 - val_loss: 16.7358\n",
      "Epoch 824/10000000\n",
      " - 0s - loss: 19.6219 - val_loss: 17.9950\n",
      "Epoch 825/10000000\n",
      " - 0s - loss: 18.8223 - val_loss: 15.7841\n",
      "Epoch 826/10000000\n",
      " - 0s - loss: 17.7959 - val_loss: 15.6737\n",
      "Epoch 827/10000000\n",
      " - 0s - loss: 17.7960 - val_loss: 15.9420\n",
      "Epoch 828/10000000\n",
      " - 0s - loss: 18.0404 - val_loss: 15.6406\n",
      "Epoch 829/10000000\n",
      " - 0s - loss: 17.8720 - val_loss: 17.1163\n",
      "Epoch 830/10000000\n",
      " - 0s - loss: 17.8159 - val_loss: 15.3587\n",
      "Epoch 831/10000000\n",
      " - 0s - loss: 17.7934 - val_loss: 15.4325\n",
      "Epoch 832/10000000\n",
      " - 0s - loss: 17.7844 - val_loss: 16.1082\n",
      "Epoch 833/10000000\n",
      " - 0s - loss: 17.5250 - val_loss: 15.9207\n",
      "Epoch 834/10000000\n",
      " - 0s - loss: 17.3562 - val_loss: 15.6726\n",
      "Epoch 835/10000000\n",
      " - 0s - loss: 17.2645 - val_loss: 15.3365\n",
      "Epoch 836/10000000\n",
      " - 0s - loss: 17.1287 - val_loss: 15.6022\n",
      "Epoch 837/10000000\n",
      " - 0s - loss: 17.0034 - val_loss: 14.9097\n",
      "Epoch 838/10000000\n",
      " - 0s - loss: 17.6134 - val_loss: 14.8784\n",
      "Epoch 839/10000000\n",
      " - 0s - loss: 17.1098 - val_loss: 15.3866\n",
      "Epoch 840/10000000\n",
      " - 0s - loss: 16.9627 - val_loss: 15.0454\n",
      "Epoch 841/10000000\n",
      " - 0s - loss: 17.1682 - val_loss: 14.6306\n",
      "Epoch 842/10000000\n",
      " - 0s - loss: 16.9637 - val_loss: 15.4504\n",
      "Epoch 843/10000000\n",
      " - 0s - loss: 16.6066 - val_loss: 14.6445\n",
      "Epoch 844/10000000\n",
      " - 0s - loss: 16.9983 - val_loss: 14.5442\n",
      "Epoch 845/10000000\n",
      " - 0s - loss: 16.4076 - val_loss: 15.2978\n",
      "Epoch 846/10000000\n",
      " - 0s - loss: 16.5278 - val_loss: 14.4898\n",
      "Epoch 847/10000000\n",
      " - 0s - loss: 16.8390 - val_loss: 14.4343\n",
      "Epoch 848/10000000\n",
      " - 0s - loss: 16.2355 - val_loss: 15.0744\n",
      "Epoch 849/10000000\n",
      " - 0s - loss: 17.1639 - val_loss: 14.3930\n",
      "Epoch 850/10000000\n",
      " - 0s - loss: 17.8655 - val_loss: 15.9101\n",
      "Epoch 851/10000000\n",
      " - 0s - loss: 17.7997 - val_loss: 16.2441\n",
      "Epoch 852/10000000\n",
      " - 0s - loss: 17.2457 - val_loss: 14.7374\n",
      "Epoch 853/10000000\n",
      " - 0s - loss: 16.1634 - val_loss: 14.1880\n",
      "Epoch 854/10000000\n",
      " - 0s - loss: 15.9709 - val_loss: 15.0170\n",
      "Epoch 855/10000000\n",
      " - 0s - loss: 16.1146 - val_loss: 14.3175\n",
      "Epoch 856/10000000\n",
      " - 0s - loss: 16.2978 - val_loss: 14.0858\n",
      "Epoch 857/10000000\n",
      " - 0s - loss: 16.3550 - val_loss: 14.4390\n",
      "Epoch 858/10000000\n",
      " - 0s - loss: 15.6024 - val_loss: 14.2908\n",
      "Epoch 859/10000000\n",
      " - 0s - loss: 16.5638 - val_loss: 14.1781\n",
      "Epoch 860/10000000\n",
      " - 0s - loss: 16.0041 - val_loss: 15.2730\n",
      "Epoch 861/10000000\n",
      " - 0s - loss: 16.3836 - val_loss: 13.8468\n",
      "Epoch 862/10000000\n",
      " - 0s - loss: 15.2553 - val_loss: 14.5861\n",
      "Epoch 863/10000000\n",
      " - 0s - loss: 16.3192 - val_loss: 14.9569\n",
      "Epoch 864/10000000\n",
      " - 0s - loss: 16.9312 - val_loss: 14.5168\n",
      "Epoch 865/10000000\n",
      " - 0s - loss: 15.4503 - val_loss: 15.4184\n",
      "Epoch 866/10000000\n",
      " - 0s - loss: 17.7642 - val_loss: 13.7175\n",
      "Epoch 867/10000000\n",
      " - 0s - loss: 16.2931 - val_loss: 15.2142\n",
      "Epoch 868/10000000\n",
      " - 0s - loss: 14.8689 - val_loss: 14.4078\n",
      "Epoch 869/10000000\n",
      " - 0s - loss: 17.6013 - val_loss: 13.2405\n",
      "Epoch 870/10000000\n",
      " - 0s - loss: 14.9985 - val_loss: 16.1048\n",
      "Epoch 871/10000000\n",
      " - 0s - loss: 16.5506 - val_loss: 13.1571\n",
      "Epoch 872/10000000\n",
      " - 0s - loss: 15.7211 - val_loss: 13.2822\n",
      "Epoch 873/10000000\n",
      " - 0s - loss: 14.8710 - val_loss: 14.9243\n",
      "Epoch 874/10000000\n",
      " - 0s - loss: 15.4453 - val_loss: 13.1292\n",
      "Epoch 875/10000000\n",
      " - 0s - loss: 14.5916 - val_loss: 13.0490\n",
      "Epoch 876/10000000\n",
      " - 0s - loss: 14.8411 - val_loss: 13.6083\n",
      "Epoch 877/10000000\n",
      " - 0s - loss: 15.6989 - val_loss: 13.1393\n",
      "Epoch 878/10000000\n",
      " - 0s - loss: 15.1620 - val_loss: 13.6072\n",
      "Epoch 879/10000000\n",
      " - 0s - loss: 15.3042 - val_loss: 14.4799\n",
      "Epoch 880/10000000\n",
      " - 0s - loss: 16.7545 - val_loss: 14.8439\n",
      "Epoch 881/10000000\n",
      " - 0s - loss: 15.3256 - val_loss: 13.5358\n",
      "Epoch 882/10000000\n",
      " - 0s - loss: 15.8476 - val_loss: 13.0286\n",
      "Epoch 883/10000000\n",
      " - 0s - loss: 14.7343 - val_loss: 14.4178\n",
      "Epoch 884/10000000\n",
      " - 0s - loss: 14.1677 - val_loss: 13.4501\n",
      "Epoch 885/10000000\n",
      " - 0s - loss: 16.9275 - val_loss: 12.5643\n",
      "Epoch 886/10000000\n",
      " - 0s - loss: 14.5967 - val_loss: 15.8108\n",
      "Epoch 887/10000000\n",
      " - 0s - loss: 15.6423 - val_loss: 12.4602\n",
      "Epoch 888/10000000\n",
      " - 0s - loss: 16.4599 - val_loss: 12.7329\n",
      "Epoch 889/10000000\n",
      " - 0s - loss: 16.4793 - val_loss: 16.3716\n",
      "Epoch 890/10000000\n",
      " - 0s - loss: 14.9090 - val_loss: 12.9652\n",
      "Epoch 891/10000000\n",
      " - 0s - loss: 16.8460 - val_loss: 12.6571\n",
      "Epoch 892/10000000\n",
      " - 0s - loss: 14.6157 - val_loss: 15.8268\n",
      "Epoch 893/10000000\n",
      " - 0s - loss: 15.3261 - val_loss: 12.5078\n",
      "Epoch 894/10000000\n",
      " - 0s - loss: 16.9412 - val_loss: 12.8733\n",
      "Epoch 895/10000000\n",
      " - 0s - loss: 16.2139 - val_loss: 16.2994\n",
      "Epoch 896/10000000\n",
      " - 0s - loss: 15.0407 - val_loss: 12.5389\n",
      "Epoch 897/10000000\n",
      " - 0s - loss: 15.3779 - val_loss: 11.9943\n",
      "Epoch 898/10000000\n",
      " - 0s - loss: 13.1257 - val_loss: 16.3192\n",
      "Epoch 899/10000000\n",
      " - 0s - loss: 16.2022 - val_loss: 12.2172\n",
      "Epoch 900/10000000\n",
      " - 0s - loss: 13.3467 - val_loss: 11.9488\n",
      "Epoch 901/10000000\n",
      " - 0s - loss: 13.6189 - val_loss: 12.1087\n",
      "Epoch 902/10000000\n",
      " - 0s - loss: 13.1560 - val_loss: 11.7889\n",
      "Epoch 903/10000000\n",
      " - 0s - loss: 14.0906 - val_loss: 12.0947\n",
      "Epoch 904/10000000\n",
      " - 0s - loss: 13.8051 - val_loss: 13.7690\n",
      "Epoch 905/10000000\n",
      " - 0s - loss: 13.4748 - val_loss: 12.0838\n",
      "Epoch 906/10000000\n",
      " - 0s - loss: 14.3687 - val_loss: 11.5488\n",
      "Epoch 907/10000000\n",
      " - 0s - loss: 13.1275 - val_loss: 12.0518\n",
      "Epoch 908/10000000\n",
      " - 0s - loss: 12.9488 - val_loss: 11.4482\n",
      "Epoch 909/10000000\n",
      " - 0s - loss: 12.9128 - val_loss: 11.5289\n",
      "Epoch 910/10000000\n",
      " - 0s - loss: 12.8663 - val_loss: 11.7822\n",
      "Epoch 911/10000000\n",
      " - 0s - loss: 12.6232 - val_loss: 11.4088\n",
      "Epoch 912/10000000\n",
      " - 0s - loss: 13.5936 - val_loss: 11.6099\n",
      "Epoch 913/10000000\n",
      " - 0s - loss: 13.2133 - val_loss: 13.6750\n",
      "Epoch 914/10000000\n",
      " - 0s - loss: 13.3527 - val_loss: 11.5604\n",
      "Epoch 915/10000000\n",
      " - 0s - loss: 13.5683 - val_loss: 11.3901\n",
      "Epoch 916/10000000\n",
      " - 0s - loss: 12.4995 - val_loss: 13.6252\n",
      "Epoch 917/10000000\n",
      " - 0s - loss: 13.1108 - val_loss: 11.3174\n",
      "Epoch 918/10000000\n",
      " - 0s - loss: 13.9445 - val_loss: 11.2035\n",
      "Epoch 919/10000000\n",
      " - 0s - loss: 12.6859 - val_loss: 12.8389\n",
      "Epoch 920/10000000\n",
      " - 0s - loss: 13.1915 - val_loss: 11.0151\n",
      "Epoch 921/10000000\n",
      " - 0s - loss: 12.3204 - val_loss: 11.1363\n",
      "Epoch 922/10000000\n",
      " - 0s - loss: 12.1770 - val_loss: 10.8731\n",
      "Epoch 923/10000000\n",
      " - 0s - loss: 12.3167 - val_loss: 10.8511\n",
      "Epoch 924/10000000\n",
      " - 0s - loss: 12.2399 - val_loss: 10.9760\n",
      "Epoch 925/10000000\n",
      " - 0s - loss: 12.2710 - val_loss: 10.8033\n",
      "Epoch 926/10000000\n",
      " - 0s - loss: 12.1804 - val_loss: 11.4541\n",
      "Epoch 927/10000000\n",
      " - 0s - loss: 12.2108 - val_loss: 11.9952\n",
      "Epoch 928/10000000\n",
      " - 0s - loss: 12.4828 - val_loss: 10.7368\n",
      "Epoch 929/10000000\n",
      " - 0s - loss: 11.9100 - val_loss: 11.4245\n",
      "Epoch 930/10000000\n",
      " - 0s - loss: 12.7646 - val_loss: 10.7077\n",
      "Epoch 931/10000000\n",
      " - 0s - loss: 12.0095 - val_loss: 11.5102\n",
      "Epoch 932/10000000\n",
      " - 0s - loss: 12.8177 - val_loss: 12.0719\n",
      "Epoch 933/10000000\n",
      " - 0s - loss: 13.1136 - val_loss: 11.4551\n",
      "Epoch 934/10000000\n",
      " - 0s - loss: 11.9000 - val_loss: 11.1887\n",
      "Epoch 935/10000000\n",
      " - 0s - loss: 12.7995 - val_loss: 11.0165\n",
      "Epoch 936/10000000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " - 0s - loss: 11.8279 - val_loss: 12.3256\n",
      "Epoch 937/10000000\n",
      " - 0s - loss: 12.3382 - val_loss: 10.3997\n",
      "Epoch 938/10000000\n",
      " - 0s - loss: 11.9398 - val_loss: 10.4382\n",
      "Epoch 939/10000000\n",
      " - 0s - loss: 12.6542 - val_loss: 11.5473\n",
      "Epoch 940/10000000\n",
      " - 0s - loss: 12.2535 - val_loss: 10.5585\n",
      "Epoch 941/10000000\n",
      " - 0s - loss: 11.8319 - val_loss: 10.7189\n",
      "Epoch 942/10000000\n",
      " - 0s - loss: 11.8024 - val_loss: 11.0484\n",
      "Epoch 943/10000000\n",
      " - 0s - loss: 11.6419 - val_loss: 10.8282\n",
      "Epoch 944/10000000\n",
      " - 0s - loss: 12.2797 - val_loss: 10.4333\n",
      "Epoch 945/10000000\n",
      " - 0s - loss: 11.8923 - val_loss: 10.3076\n",
      "Epoch 946/10000000\n",
      " - 0s - loss: 10.9064 - val_loss: 10.6424\n",
      "Epoch 947/10000000\n",
      " - 0s - loss: 12.4064 - val_loss: 10.5742\n",
      "Epoch 948/10000000\n",
      " - 0s - loss: 11.9683 - val_loss: 10.9068\n",
      "Epoch 949/10000000\n",
      " - 0s - loss: 12.9133 - val_loss: 10.2823\n",
      "Epoch 950/10000000\n",
      " - 0s - loss: 10.8446 - val_loss: 13.2464\n",
      "Epoch 951/10000000\n",
      " - 0s - loss: 12.9546 - val_loss: 10.0711\n",
      "Epoch 952/10000000\n",
      " - 0s - loss: 12.0083 - val_loss: 11.0015\n",
      "Epoch 953/10000000\n",
      " - 0s - loss: 11.4462 - val_loss: 12.0013\n",
      "Epoch 954/10000000\n",
      " - 0s - loss: 13.0050 - val_loss: 10.4218\n",
      "Epoch 955/10000000\n",
      " - 0s - loss: 11.9834 - val_loss: 9.8836\n",
      "Epoch 956/10000000\n",
      " - 0s - loss: 12.1193 - val_loss: 11.2102\n",
      "Epoch 957/10000000\n",
      " - 0s - loss: 11.5940 - val_loss: 9.7312\n",
      "Epoch 958/10000000\n",
      " - 0s - loss: 10.7179 - val_loss: 11.0946\n",
      "Epoch 959/10000000\n",
      " - 0s - loss: 11.6576 - val_loss: 10.1923\n",
      "Epoch 960/10000000\n",
      " - 0s - loss: 11.0688 - val_loss: 9.9179\n",
      "Epoch 961/10000000\n",
      " - 0s - loss: 11.2334 - val_loss: 10.5510\n",
      "Epoch 962/10000000\n",
      " - 0s - loss: 11.2882 - val_loss: 10.5093\n",
      "Epoch 963/10000000\n",
      " - 0s - loss: 10.5315 - val_loss: 10.3780\n",
      "Epoch 964/10000000\n",
      " - 0s - loss: 12.1245 - val_loss: 9.9440\n",
      "Epoch 965/10000000\n",
      " - 0s - loss: 10.8647 - val_loss: 10.9863\n",
      "Epoch 966/10000000\n",
      " - 0s - loss: 10.5881 - val_loss: 10.0633\n",
      "Epoch 967/10000000\n",
      " - 0s - loss: 11.3106 - val_loss: 9.7997\n",
      "Epoch 968/10000000\n",
      " - 0s - loss: 10.4115 - val_loss: 11.1594\n",
      "Epoch 969/10000000\n",
      " - 0s - loss: 10.9755 - val_loss: 9.4401\n",
      "Epoch 970/10000000\n",
      " - 0s - loss: 10.2182 - val_loss: 9.3404\n",
      "Epoch 971/10000000\n",
      " - 0s - loss: 10.1738 - val_loss: 9.5824\n",
      "Epoch 972/10000000\n",
      " - 0s - loss: 10.3079 - val_loss: 9.6156\n",
      "Epoch 973/10000000\n",
      " - 0s - loss: 10.1746 - val_loss: 10.2735\n",
      "Epoch 974/10000000\n",
      " - 0s - loss: 10.2170 - val_loss: 9.2193\n",
      "Epoch 975/10000000\n",
      " - 0s - loss: 10.1683 - val_loss: 9.1656\n",
      "Epoch 976/10000000\n",
      " - 0s - loss: 9.9860 - val_loss: 9.4909\n",
      "Epoch 977/10000000\n",
      " - 0s - loss: 9.9467 - val_loss: 9.1527\n",
      "Epoch 978/10000000\n",
      " - 0s - loss: 10.1045 - val_loss: 9.1047\n",
      "Epoch 979/10000000\n",
      " - 0s - loss: 9.8621 - val_loss: 9.9681\n",
      "Epoch 980/10000000\n",
      " - 0s - loss: 10.0044 - val_loss: 8.9903\n",
      "Epoch 981/10000000\n",
      " - 0s - loss: 9.8525 - val_loss: 8.9927\n",
      "Epoch 982/10000000\n",
      " - 0s - loss: 9.8861 - val_loss: 8.9885\n",
      "Epoch 983/10000000\n",
      " - 0s - loss: 9.6275 - val_loss: 9.5100\n",
      "Epoch 984/10000000\n",
      " - 0s - loss: 9.9188 - val_loss: 8.8987\n",
      "Epoch 985/10000000\n",
      " - 0s - loss: 10.4281 - val_loss: 8.7567\n",
      "Epoch 986/10000000\n",
      " - 0s - loss: 9.6798 - val_loss: 11.1063\n",
      "Epoch 987/10000000\n",
      " - 0s - loss: 10.8624 - val_loss: 8.8009\n",
      "Epoch 988/10000000\n",
      " - 0s - loss: 11.0662 - val_loss: 8.8754\n",
      "Epoch 989/10000000\n",
      " - 0s - loss: 9.7833 - val_loss: 10.7007\n",
      "Epoch 990/10000000\n",
      " - 0s - loss: 10.4028 - val_loss: 8.7811\n",
      "Epoch 991/10000000\n",
      " - 0s - loss: 9.3605 - val_loss: 8.6166\n",
      "Epoch 992/10000000\n",
      " - 0s - loss: 9.6170 - val_loss: 8.9168\n",
      "Epoch 993/10000000\n",
      " - 0s - loss: 9.2305 - val_loss: 10.7192\n",
      "Epoch 994/10000000\n",
      " - 0s - loss: 10.4474 - val_loss: 8.7769\n",
      "Epoch 995/10000000\n",
      " - 0s - loss: 9.3367 - val_loss: 8.5162\n",
      "Epoch 996/10000000\n",
      " - 0s - loss: 9.2215 - val_loss: 8.8815\n",
      "Epoch 997/10000000\n",
      " - 0s - loss: 9.2328 - val_loss: 8.3532\n",
      "Epoch 998/10000000\n",
      " - 0s - loss: 9.3088 - val_loss: 8.4506\n",
      "Epoch 999/10000000\n",
      " - 0s - loss: 9.0984 - val_loss: 9.3297\n",
      "Epoch 1000/10000000\n",
      " - 0s - loss: 9.4723 - val_loss: 8.5369\n",
      "Epoch 1001/10000000\n",
      " - 0s - loss: 9.2420 - val_loss: 8.2450\n",
      "Epoch 1002/10000000\n",
      " - 0s - loss: 9.1801 - val_loss: 8.5782\n",
      "Epoch 1003/10000000\n",
      " - 0s - loss: 9.4883 - val_loss: 8.3310\n",
      "Epoch 1004/10000000\n",
      " - 0s - loss: 8.8871 - val_loss: 8.0896\n",
      "Epoch 1005/10000000\n",
      " - 0s - loss: 9.2623 - val_loss: 8.2275\n",
      "Epoch 1006/10000000\n",
      " - 0s - loss: 8.9819 - val_loss: 9.0393\n",
      "Epoch 1007/10000000\n",
      " - 0s - loss: 9.2960 - val_loss: 8.2933\n",
      "Epoch 1008/10000000\n",
      " - 0s - loss: 8.7474 - val_loss: 8.8567\n",
      "Epoch 1009/10000000\n",
      " - 0s - loss: 8.9852 - val_loss: 8.2739\n",
      "Epoch 1010/10000000\n",
      " - 0s - loss: 8.7543 - val_loss: 8.0649\n",
      "Epoch 1011/10000000\n",
      " - 0s - loss: 8.8270 - val_loss: 7.9920\n",
      "Epoch 1012/10000000\n",
      " - 0s - loss: 8.7180 - val_loss: 8.5151\n",
      "Epoch 1013/10000000\n",
      " - 0s - loss: 9.3165 - val_loss: 8.1772\n",
      "Epoch 1014/10000000\n",
      " - 0s - loss: 8.7761 - val_loss: 7.9184\n",
      "Epoch 1015/10000000\n",
      " - 0s - loss: 9.1944 - val_loss: 8.6794\n",
      "Epoch 1016/10000000\n",
      " - 0s - loss: 9.0224 - val_loss: 7.9433\n",
      "Epoch 1017/10000000\n",
      " - 0s - loss: 8.6937 - val_loss: 8.9281\n",
      "Epoch 1018/10000000\n",
      " - 0s - loss: 9.0177 - val_loss: 7.8216\n",
      "Epoch 1019/10000000\n",
      " - 0s - loss: 8.9139 - val_loss: 7.7433\n",
      "Epoch 1020/10000000\n",
      " - 0s - loss: 8.6345 - val_loss: 8.3875\n",
      "Epoch 1021/10000000\n",
      " - 0s - loss: 8.6482 - val_loss: 7.9406\n",
      "Epoch 1022/10000000\n",
      " - 0s - loss: 8.6977 - val_loss: 7.9867\n",
      "Epoch 1023/10000000\n",
      " - 0s - loss: 8.4544 - val_loss: 7.7386\n",
      "Epoch 1024/10000000\n",
      " - 0s - loss: 8.4229 - val_loss: 8.3588\n",
      "Epoch 1025/10000000\n",
      " - 0s - loss: 8.4891 - val_loss: 7.7082\n",
      "Epoch 1026/10000000\n",
      " - 0s - loss: 8.6581 - val_loss: 7.7701\n",
      "Epoch 1027/10000000\n",
      " - 0s - loss: 8.9976 - val_loss: 8.6369\n",
      "Epoch 1028/10000000\n",
      " - 0s - loss: 9.0615 - val_loss: 7.5585\n",
      "Epoch 1029/10000000\n",
      " - 0s - loss: 7.9730 - val_loss: 9.8018\n",
      "Epoch 1030/10000000\n",
      " - 0s - loss: 9.9748 - val_loss: 7.8033\n",
      "Epoch 1031/10000000\n",
      " - 0s - loss: 8.1631 - val_loss: 7.7855\n",
      "Epoch 1032/10000000\n",
      " - 0s - loss: 8.6653 - val_loss: 8.5888\n",
      "Epoch 1033/10000000\n",
      " - 0s - loss: 8.4321 - val_loss: 7.4880\n",
      "Epoch 1034/10000000\n",
      " - 0s - loss: 8.8364 - val_loss: 7.3635\n",
      "Epoch 1035/10000000\n",
      " - 0s - loss: 8.4003 - val_loss: 9.1880\n",
      "Epoch 1036/10000000\n",
      " - 0s - loss: 8.5324 - val_loss: 7.3723\n",
      "Epoch 1037/10000000\n",
      " - 0s - loss: 9.2408 - val_loss: 7.5403\n",
      "Epoch 1038/10000000\n",
      " - 0s - loss: 8.1333 - val_loss: 11.4461\n",
      "Epoch 1039/10000000\n",
      " - 0s - loss: 9.4180 - val_loss: 8.2309\n",
      "Epoch 1040/10000000\n",
      " - 0s - loss: 12.5216 - val_loss: 7.2283\n",
      "Epoch 1041/10000000\n",
      " - 0s - loss: 10.0606 - val_loss: 11.9495\n",
      "Epoch 1042/10000000\n",
      " - 0s - loss: 10.1541 - val_loss: 7.8286\n",
      "Epoch 1043/10000000\n",
      " - 0s - loss: 9.1229 - val_loss: 7.4472\n",
      "Epoch 1044/10000000\n",
      " - 0s - loss: 8.2957 - val_loss: 7.7529\n",
      "Epoch 1045/10000000\n",
      " - 0s - loss: 7.6484 - val_loss: 7.4215\n",
      "Epoch 1046/10000000\n",
      " - 0s - loss: 8.7674 - val_loss: 7.2988\n",
      "Epoch 1047/10000000\n",
      " - 0s - loss: 7.7276 - val_loss: 7.9520\n",
      "Epoch 1048/10000000\n",
      " - 0s - loss: 7.9071 - val_loss: 7.0812\n",
      "Epoch 1049/10000000\n",
      " - 0s - loss: 7.9481 - val_loss: 7.0457\n",
      "Epoch 1050/10000000\n",
      " - 0s - loss: 7.7285 - val_loss: 7.1758\n",
      "Epoch 1051/10000000\n",
      " - 0s - loss: 7.9599 - val_loss: 6.9355\n",
      "Epoch 1052/10000000\n",
      " - 0s - loss: 9.2700 - val_loss: 6.8916\n",
      "Epoch 1053/10000000\n",
      " - 0s - loss: 8.2818 - val_loss: 10.6753\n",
      "Epoch 1054/10000000\n",
      " - 0s - loss: 8.9554 - val_loss: 7.3693\n",
      "Epoch 1055/10000000\n",
      " - 0s - loss: 8.8684 - val_loss: 6.9468\n",
      "Epoch 1056/10000000\n",
      " - 0s - loss: 7.5448 - val_loss: 8.7415\n",
      "Epoch 1057/10000000\n",
      " - 0s - loss: 7.8449 - val_loss: 7.3562\n",
      "Epoch 1058/10000000\n",
      " - 0s - loss: 8.7362 - val_loss: 6.9142\n",
      "Epoch 1059/10000000\n",
      " - 0s - loss: 7.6128 - val_loss: 8.0398\n",
      "Epoch 1060/10000000\n",
      " - 0s - loss: 7.9429 - val_loss: 6.8457\n",
      "Epoch 1061/10000000\n",
      " - 0s - loss: 7.6644 - val_loss: 6.7984\n",
      "Epoch 1062/10000000\n",
      " - 0s - loss: 7.6317 - val_loss: 7.6616\n",
      "Epoch 1063/10000000\n",
      " - 0s - loss: 7.4303 - val_loss: 6.7210\n",
      "Epoch 1064/10000000\n",
      " - 0s - loss: 7.5502 - val_loss: 6.6756\n",
      "Epoch 1065/10000000\n",
      " - 0s - loss: 7.2384 - val_loss: 7.2434\n",
      "Epoch 1066/10000000\n",
      " - 0s - loss: 7.3179 - val_loss: 6.7082\n",
      "Epoch 1067/10000000\n",
      " - 0s - loss: 7.3105 - val_loss: 6.6348\n",
      "Epoch 1068/10000000\n",
      " - 0s - loss: 7.4021 - val_loss: 6.6887\n",
      "Epoch 1069/10000000\n",
      " - 0s - loss: 7.1766 - val_loss: 6.5985\n",
      "Epoch 1070/10000000\n",
      " - 0s - loss: 7.4049 - val_loss: 6.8902\n",
      "Epoch 1071/10000000\n",
      " - 0s - loss: 7.1148 - val_loss: 6.5541\n",
      "Epoch 1072/10000000\n",
      " - 0s - loss: 7.1981 - val_loss: 6.8520\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1073/10000000\n",
      " - 0s - loss: 7.1126 - val_loss: 6.7266\n",
      "Epoch 1074/10000000\n",
      " - 0s - loss: 7.0353 - val_loss: 6.9727\n",
      "Epoch 1075/10000000\n",
      " - 0s - loss: 7.0506 - val_loss: 6.5876\n",
      "Epoch 1076/10000000\n",
      " - 0s - loss: 7.0929 - val_loss: 6.6103\n",
      "Epoch 1077/10000000\n",
      " - 0s - loss: 7.0441 - val_loss: 7.0209\n",
      "Epoch 1078/10000000\n",
      " - 0s - loss: 7.3995 - val_loss: 6.6645\n",
      "Epoch 1079/10000000\n",
      " - 0s - loss: 6.9025 - val_loss: 6.5147\n",
      "Epoch 1080/10000000\n",
      " - 0s - loss: 7.1923 - val_loss: 6.9448\n",
      "Epoch 1081/10000000\n",
      " - 0s - loss: 7.0143 - val_loss: 6.7769\n",
      "Epoch 1082/10000000\n",
      " - 0s - loss: 6.9422 - val_loss: 6.5516\n",
      "Epoch 1083/10000000\n",
      " - 0s - loss: 6.9258 - val_loss: 6.5684\n",
      "Epoch 1084/10000000\n",
      " - 0s - loss: 6.8762 - val_loss: 6.4114\n",
      "Epoch 1085/10000000\n",
      " - 0s - loss: 6.8413 - val_loss: 6.4650\n",
      "Epoch 1086/10000000\n",
      " - 0s - loss: 7.0338 - val_loss: 6.3168\n",
      "Epoch 1087/10000000\n",
      " - 0s - loss: 6.9174 - val_loss: 6.2455\n",
      "Epoch 1088/10000000\n",
      " - 0s - loss: 6.8089 - val_loss: 7.5160\n",
      "Epoch 1089/10000000\n",
      " - 0s - loss: 7.2199 - val_loss: 6.3315\n",
      "Epoch 1090/10000000\n",
      " - 0s - loss: 6.7431 - val_loss: 6.2323\n",
      "Epoch 1091/10000000\n",
      " - 0s - loss: 6.7318 - val_loss: 6.4316\n",
      "Epoch 1092/10000000\n",
      " - 0s - loss: 6.7580 - val_loss: 6.4160\n",
      "Epoch 1093/10000000\n",
      " - 0s - loss: 7.0724 - val_loss: 6.3893\n",
      "Epoch 1094/10000000\n",
      " - 0s - loss: 6.6537 - val_loss: 6.1094\n",
      "Epoch 1095/10000000\n",
      " - 0s - loss: 6.7571 - val_loss: 7.1213\n",
      "Epoch 1096/10000000\n",
      " - 0s - loss: 6.9945 - val_loss: 6.1571\n",
      "Epoch 1097/10000000\n",
      " - 0s - loss: 6.5405 - val_loss: 6.0576\n",
      "Epoch 1098/10000000\n",
      " - 0s - loss: 6.5386 - val_loss: 6.5969\n",
      "Epoch 1099/10000000\n",
      " - 0s - loss: 7.0885 - val_loss: 6.2547\n",
      "Epoch 1100/10000000\n",
      " - 0s - loss: 6.6852 - val_loss: 5.9354\n",
      "Epoch 1101/10000000\n",
      " - 0s - loss: 6.5944 - val_loss: 6.7160\n",
      "Epoch 1102/10000000\n",
      " - 0s - loss: 6.5254 - val_loss: 5.8852\n",
      "Epoch 1103/10000000\n",
      " - 0s - loss: 7.5517 - val_loss: 5.8155\n",
      "Epoch 1104/10000000\n",
      " - 0s - loss: 6.5595 - val_loss: 6.9059\n",
      "Epoch 1105/10000000\n",
      " - 0s - loss: 6.4927 - val_loss: 5.9181\n",
      "Epoch 1106/10000000\n",
      " - 0s - loss: 7.4271 - val_loss: 5.9430\n",
      "Epoch 1107/10000000\n",
      " - 0s - loss: 6.3968 - val_loss: 7.5077\n",
      "Epoch 1108/10000000\n",
      " - 0s - loss: 6.9146 - val_loss: 5.9621\n",
      "Epoch 1109/10000000\n",
      " - 0s - loss: 7.1136 - val_loss: 6.0998\n",
      "Epoch 1110/10000000\n",
      " - 0s - loss: 6.5124 - val_loss: 6.0768\n",
      "Epoch 1111/10000000\n",
      " - 0s - loss: 6.4584 - val_loss: 5.7700\n",
      "Epoch 1112/10000000\n",
      " - 0s - loss: 6.3258 - val_loss: 6.6934\n",
      "Epoch 1113/10000000\n",
      " - 0s - loss: 6.6165 - val_loss: 5.7399\n",
      "Epoch 1114/10000000\n",
      " - 0s - loss: 7.0081 - val_loss: 5.7036\n",
      "Epoch 1115/10000000\n",
      " - 0s - loss: 6.3125 - val_loss: 7.2417\n",
      "Epoch 1116/10000000\n",
      " - 0s - loss: 6.8519 - val_loss: 6.1670\n",
      "Epoch 1117/10000000\n",
      " - 0s - loss: 7.0528 - val_loss: 6.0171\n",
      "Epoch 1118/10000000\n",
      " - 0s - loss: 7.2181 - val_loss: 5.8175\n",
      "Epoch 1119/10000000\n",
      " - 0s - loss: 6.1625 - val_loss: 7.0628\n",
      "Epoch 1120/10000000\n",
      " - 0s - loss: 7.4607 - val_loss: 7.5151\n",
      "Epoch 1121/10000000\n",
      " - 0s - loss: 7.4190 - val_loss: 5.6891\n",
      "Epoch 1122/10000000\n",
      " - 0s - loss: 7.7969 - val_loss: 5.6175\n",
      "Epoch 1123/10000000\n",
      " - 0s - loss: 6.8284 - val_loss: 8.6095\n",
      "Epoch 1124/10000000\n",
      " - 0s - loss: 7.7345 - val_loss: 6.2320\n",
      "Epoch 1125/10000000\n",
      " - 0s - loss: 6.9093 - val_loss: 6.3444\n",
      "Epoch 1126/10000000\n",
      " - 0s - loss: 6.5787 - val_loss: 5.8967\n",
      "Epoch 1127/10000000\n",
      " - 0s - loss: 6.4969 - val_loss: 5.4541\n",
      "Epoch 1128/10000000\n",
      " - 0s - loss: 6.0308 - val_loss: 5.9296\n",
      "Epoch 1129/10000000\n",
      " - 0s - loss: 6.0245 - val_loss: 5.4427\n",
      "Epoch 1130/10000000\n",
      " - 0s - loss: 6.0023 - val_loss: 5.3838\n",
      "Epoch 1131/10000000\n",
      " - 0s - loss: 6.0245 - val_loss: 5.4500\n",
      "Epoch 1132/10000000\n",
      " - 0s - loss: 5.9120 - val_loss: 5.3454\n",
      "Epoch 1133/10000000\n",
      " - 0s - loss: 5.9650 - val_loss: 5.4925\n",
      "Epoch 1134/10000000\n",
      " - 0s - loss: 6.0444 - val_loss: 5.6349\n",
      "Epoch 1135/10000000\n",
      " - 0s - loss: 5.9618 - val_loss: 5.4620\n",
      "Epoch 1136/10000000\n",
      " - 0s - loss: 5.8627 - val_loss: 5.7956\n",
      "Epoch 1137/10000000\n",
      " - 0s - loss: 5.8093 - val_loss: 5.2866\n",
      "Epoch 1138/10000000\n",
      " - 0s - loss: 6.1066 - val_loss: 5.5359\n",
      "Epoch 1139/10000000\n",
      " - 0s - loss: 5.8303 - val_loss: 5.9411\n",
      "Epoch 1140/10000000\n",
      " - 0s - loss: 6.0966 - val_loss: 5.2651\n",
      "Epoch 1141/10000000\n",
      " - 0s - loss: 5.7223 - val_loss: 6.3941\n",
      "Epoch 1142/10000000\n",
      " - 0s - loss: 6.4091 - val_loss: 5.2813\n",
      "Epoch 1143/10000000\n",
      " - 0s - loss: 5.7530 - val_loss: 5.2522\n",
      "Epoch 1144/10000000\n",
      " - 0s - loss: 6.3529 - val_loss: 5.5816\n",
      "Epoch 1145/10000000\n",
      " - 0s - loss: 5.8162 - val_loss: 5.2431\n",
      "Epoch 1146/10000000\n",
      " - 0s - loss: 5.6252 - val_loss: 5.7886\n",
      "Epoch 1147/10000000\n",
      " - 0s - loss: 5.8701 - val_loss: 5.4409\n",
      "Epoch 1148/10000000\n",
      " - 0s - loss: 5.6557 - val_loss: 5.2852\n",
      "Epoch 1149/10000000\n",
      " - 0s - loss: 5.7757 - val_loss: 5.1745\n",
      "Epoch 1150/10000000\n",
      " - 0s - loss: 5.5280 - val_loss: 5.9814\n",
      "Epoch 1151/10000000\n",
      " - 0s - loss: 5.8995 - val_loss: 5.0181\n",
      "Epoch 1152/10000000\n",
      " - 0s - loss: 5.7304 - val_loss: 5.0959\n",
      "Epoch 1153/10000000\n",
      " - 0s - loss: 6.0896 - val_loss: 5.0208\n",
      "Epoch 1154/10000000\n",
      " - 0s - loss: 5.5984 - val_loss: 5.2206\n",
      "Epoch 1155/10000000\n",
      " - 0s - loss: 5.6843 - val_loss: 5.0671\n",
      "Epoch 1156/10000000\n",
      " - 0s - loss: 6.0492 - val_loss: 4.9855\n",
      "Epoch 1157/10000000\n",
      " - 0s - loss: 5.5248 - val_loss: 5.8764\n",
      "Epoch 1158/10000000\n",
      " - 0s - loss: 5.6560 - val_loss: 5.0147\n",
      "Epoch 1159/10000000\n",
      " - 0s - loss: 6.0889 - val_loss: 5.2420\n",
      "Epoch 1160/10000000\n",
      " - 0s - loss: 6.0505 - val_loss: 6.0390\n",
      "Epoch 1161/10000000\n",
      " - 0s - loss: 5.8662 - val_loss: 4.8911\n",
      "Epoch 1162/10000000\n",
      " - 0s - loss: 5.8013 - val_loss: 5.1345\n",
      "Epoch 1163/10000000\n",
      " - 0s - loss: 5.5355 - val_loss: 4.9277\n",
      "Epoch 1164/10000000\n",
      " - 0s - loss: 5.4783 - val_loss: 5.2166\n",
      "Epoch 1165/10000000\n",
      " - 0s - loss: 5.2691 - val_loss: 4.9819\n",
      "Epoch 1166/10000000\n",
      " - 0s - loss: 5.7735 - val_loss: 4.9380\n",
      "Epoch 1167/10000000\n",
      " - 0s - loss: 5.4265 - val_loss: 4.8854\n",
      "Epoch 1168/10000000\n",
      " - 0s - loss: 5.3469 - val_loss: 4.8245\n",
      "Epoch 1169/10000000\n",
      " - 0s - loss: 5.6829 - val_loss: 4.8191\n",
      "Epoch 1170/10000000\n",
      " - 0s - loss: 5.3759 - val_loss: 4.8387\n",
      "Epoch 1171/10000000\n",
      " - 0s - loss: 5.5684 - val_loss: 4.8799\n",
      "Epoch 1172/10000000\n",
      " - 0s - loss: 5.3505 - val_loss: 5.1684\n",
      "Epoch 1173/10000000\n",
      " - 0s - loss: 5.2363 - val_loss: 4.7120\n",
      "Epoch 1174/10000000\n",
      " - 0s - loss: 5.7396 - val_loss: 5.7778\n",
      "Epoch 1175/10000000\n",
      " - 0s - loss: 6.6173 - val_loss: 5.2082\n",
      "Epoch 1176/10000000\n",
      " - 0s - loss: 5.9836 - val_loss: 6.1546\n",
      "Epoch 1177/10000000\n",
      " - 0s - loss: 7.1282 - val_loss: 5.5862\n",
      "Epoch 1178/10000000\n",
      " - 0s - loss: 5.8479 - val_loss: 4.7036\n",
      "Epoch 1179/10000000\n",
      " - 0s - loss: 5.3720 - val_loss: 4.6842\n",
      "Epoch 1180/10000000\n",
      " - 0s - loss: 5.7158 - val_loss: 5.0440\n",
      "Epoch 1181/10000000\n",
      " - 0s - loss: 5.1478 - val_loss: 4.5538\n",
      "Epoch 1182/10000000\n",
      " - 0s - loss: 5.7363 - val_loss: 4.6964\n",
      "Epoch 1183/10000000\n",
      " - 0s - loss: 5.1779 - val_loss: 5.2061\n",
      "Epoch 1184/10000000\n",
      " - 0s - loss: 5.4871 - val_loss: 4.6166\n",
      "Epoch 1185/10000000\n",
      " - 0s - loss: 5.3668 - val_loss: 5.2080\n",
      "Epoch 1186/10000000\n",
      " - 0s - loss: 5.3449 - val_loss: 4.6042\n",
      "Epoch 1187/10000000\n",
      " - 0s - loss: 5.0841 - val_loss: 4.5770\n",
      "Epoch 1188/10000000\n",
      " - 0s - loss: 5.0977 - val_loss: 5.0767\n",
      "Epoch 1189/10000000\n",
      " - 0s - loss: 5.3421 - val_loss: 4.4427\n",
      "Epoch 1190/10000000\n",
      " - 0s - loss: 5.1363 - val_loss: 4.3998\n",
      "Epoch 1191/10000000\n",
      " - 0s - loss: 5.1461 - val_loss: 4.9015\n",
      "Epoch 1192/10000000\n",
      " - 0s - loss: 4.9277 - val_loss: 4.4699\n",
      "Epoch 1193/10000000\n",
      " - 0s - loss: 5.9204 - val_loss: 4.4314\n",
      "Epoch 1194/10000000\n",
      " - 0s - loss: 5.1603 - val_loss: 5.1618\n",
      "Epoch 1195/10000000\n",
      " - 0s - loss: 5.5965 - val_loss: 4.3587\n",
      "Epoch 1196/10000000\n",
      " - 0s - loss: 5.2638 - val_loss: 4.9320\n",
      "Epoch 1197/10000000\n",
      " - 0s - loss: 5.0845 - val_loss: 4.3340\n",
      "Epoch 1198/10000000\n",
      " - 0s - loss: 5.0289 - val_loss: 4.7478\n",
      "Epoch 1199/10000000\n",
      " - 0s - loss: 5.0251 - val_loss: 4.5188\n",
      "Epoch 1200/10000000\n",
      " - 0s - loss: 4.9090 - val_loss: 4.2936\n",
      "Epoch 1201/10000000\n",
      " - 0s - loss: 4.9193 - val_loss: 4.4232\n",
      "Epoch 1202/10000000\n",
      " - 0s - loss: 4.8766 - val_loss: 4.2722\n",
      "Epoch 1203/10000000\n",
      " - 0s - loss: 4.8755 - val_loss: 4.5277\n",
      "Epoch 1204/10000000\n",
      " - 0s - loss: 4.8555 - val_loss: 4.2818\n",
      "Epoch 1205/10000000\n",
      " - 0s - loss: 4.8622 - val_loss: 4.2269\n",
      "Epoch 1206/10000000\n",
      " - 0s - loss: 4.8210 - val_loss: 4.6739\n",
      "Epoch 1207/10000000\n",
      " - 0s - loss: 4.8735 - val_loss: 4.6915\n",
      "Epoch 1208/10000000\n",
      " - 0s - loss: 4.7944 - val_loss: 4.2322\n",
      "Epoch 1209/10000000\n",
      " - 0s - loss: 4.9545 - val_loss: 4.1996\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1210/10000000\n",
      " - 0s - loss: 4.8871 - val_loss: 4.6599\n",
      "Epoch 1211/10000000\n",
      " - 0s - loss: 4.8459 - val_loss: 4.2380\n",
      "Epoch 1212/10000000\n",
      " - 0s - loss: 4.7134 - val_loss: 4.5379\n",
      "Epoch 1213/10000000\n",
      " - 0s - loss: 4.9625 - val_loss: 4.2001\n",
      "Epoch 1214/10000000\n",
      " - 0s - loss: 5.6640 - val_loss: 4.1797\n",
      "Epoch 1215/10000000\n",
      " - 0s - loss: 4.7665 - val_loss: 5.2348\n",
      "Epoch 1216/10000000\n",
      " - 0s - loss: 5.1092 - val_loss: 4.1993\n",
      "Epoch 1217/10000000\n",
      " - 0s - loss: 4.7074 - val_loss: 4.5000\n",
      "Epoch 1218/10000000\n",
      " - 0s - loss: 4.7616 - val_loss: 4.2145\n",
      "Epoch 1219/10000000\n",
      " - 0s - loss: 4.6962 - val_loss: 4.2080\n",
      "Epoch 1220/10000000\n",
      " - 0s - loss: 4.5925 - val_loss: 4.9511\n",
      "Epoch 1221/10000000\n",
      " - 0s - loss: 5.4358 - val_loss: 4.1291\n",
      "Epoch 1222/10000000\n",
      " - 0s - loss: 5.4968 - val_loss: 4.1000\n",
      "Epoch 1223/10000000\n",
      " - 0s - loss: 4.9367 - val_loss: 5.2729\n",
      "Epoch 1224/10000000\n",
      " - 0s - loss: 4.8999 - val_loss: 4.9103\n",
      "Epoch 1225/10000000\n",
      " - 0s - loss: 5.4134 - val_loss: 4.8376\n",
      "Epoch 1226/10000000\n",
      " - 0s - loss: 5.0979 - val_loss: 4.3516\n",
      "Epoch 1227/10000000\n",
      " - 0s - loss: 4.9992 - val_loss: 4.1934\n",
      "Epoch 1228/10000000\n",
      " - 0s - loss: 5.0998 - val_loss: 4.8844\n",
      "Epoch 1229/10000000\n",
      " - 0s - loss: 4.6707 - val_loss: 4.2604\n",
      "Epoch 1230/10000000\n",
      " - 0s - loss: 5.2529 - val_loss: 4.0561\n",
      "Epoch 1231/10000000\n",
      " - 0s - loss: 4.9002 - val_loss: 4.1151\n",
      "Epoch 1232/10000000\n",
      " - 0s - loss: 4.6715 - val_loss: 4.0586\n",
      "Epoch 1233/10000000\n",
      " - 0s - loss: 4.6326 - val_loss: 5.4901\n",
      "Epoch 1234/10000000\n",
      " - 0s - loss: 5.0412 - val_loss: 3.9576\n",
      "Epoch 1235/10000000\n",
      " - 0s - loss: 5.1113 - val_loss: 3.8908\n",
      "Epoch 1236/10000000\n",
      " - 0s - loss: 4.6563 - val_loss: 4.6413\n",
      "Epoch 1237/10000000\n",
      " - 0s - loss: 4.8586 - val_loss: 3.8752\n",
      "Epoch 1238/10000000\n",
      " - 0s - loss: 4.4810 - val_loss: 4.4714\n",
      "Epoch 1239/10000000\n",
      " - 0s - loss: 4.6365 - val_loss: 3.9127\n",
      "Epoch 1240/10000000\n",
      " - 0s - loss: 4.4572 - val_loss: 4.0040\n",
      "Epoch 1241/10000000\n",
      " - 0s - loss: 4.4302 - val_loss: 4.1273\n",
      "Epoch 1242/10000000\n",
      " - 0s - loss: 4.7119 - val_loss: 3.8887\n",
      "Epoch 1243/10000000\n",
      " - 0s - loss: 4.4080 - val_loss: 3.8320\n",
      "Epoch 1244/10000000\n",
      " - 0s - loss: 4.5348 - val_loss: 3.8042\n",
      "Epoch 1245/10000000\n",
      " - 0s - loss: 4.4624 - val_loss: 3.7746\n",
      "Epoch 1246/10000000\n",
      " - 0s - loss: 4.6808 - val_loss: 4.4631\n",
      "Epoch 1247/10000000\n",
      " - 0s - loss: 4.5397 - val_loss: 3.7246\n",
      "Epoch 1248/10000000\n",
      " - 0s - loss: 4.4826 - val_loss: 3.9691\n",
      "Epoch 1249/10000000\n",
      " - 0s - loss: 4.4107 - val_loss: 4.0228\n",
      "Epoch 1250/10000000\n",
      " - 0s - loss: 4.5015 - val_loss: 3.7352\n",
      "Epoch 1251/10000000\n",
      " - 0s - loss: 4.5577 - val_loss: 3.7612\n",
      "Epoch 1252/10000000\n",
      " - 0s - loss: 4.3795 - val_loss: 4.1317\n",
      "Epoch 1253/10000000\n",
      " - 0s - loss: 4.4622 - val_loss: 3.7691\n",
      "Epoch 1254/10000000\n",
      " - 0s - loss: 4.6233 - val_loss: 3.7671\n",
      "Epoch 1255/10000000\n",
      " - 0s - loss: 4.4122 - val_loss: 3.8403\n",
      "Epoch 1256/10000000\n",
      " - 0s - loss: 4.4754 - val_loss: 4.0901\n",
      "Epoch 1257/10000000\n",
      " - 0s - loss: 4.2747 - val_loss: 3.7881\n",
      "Epoch 1258/10000000\n",
      " - 0s - loss: 4.4624 - val_loss: 3.9444\n",
      "Epoch 1259/10000000\n",
      " - 0s - loss: 4.5996 - val_loss: 3.6797\n",
      "Epoch 1260/10000000\n",
      " - 0s - loss: 4.9126 - val_loss: 3.7220\n",
      "Epoch 1261/10000000\n",
      " - 0s - loss: 4.6216 - val_loss: 4.9806\n",
      "Epoch 1262/10000000\n",
      " - 0s - loss: 4.9096 - val_loss: 3.6752\n",
      "Epoch 1263/10000000\n",
      " - 0s - loss: 4.2747 - val_loss: 4.4240\n",
      "Epoch 1264/10000000\n",
      " - 0s - loss: 4.4669 - val_loss: 3.6362\n",
      "Epoch 1265/10000000\n",
      " - 0s - loss: 4.2349 - val_loss: 3.8691\n",
      "Epoch 1266/10000000\n",
      " - 0s - loss: 4.2255 - val_loss: 3.8124\n",
      "Epoch 1267/10000000\n",
      " - 0s - loss: 4.1930 - val_loss: 3.6571\n",
      "Epoch 1268/10000000\n",
      " - 0s - loss: 4.1372 - val_loss: 3.5928\n",
      "Epoch 1269/10000000\n",
      " - 0s - loss: 4.1603 - val_loss: 3.6786\n",
      "Epoch 1270/10000000\n",
      " - 0s - loss: 4.1128 - val_loss: 4.0671\n",
      "Epoch 1271/10000000\n",
      " - 0s - loss: 4.2501 - val_loss: 3.5249\n",
      "Epoch 1272/10000000\n",
      " - 0s - loss: 4.1378 - val_loss: 3.4759\n",
      "Epoch 1273/10000000\n",
      " - 0s - loss: 4.1205 - val_loss: 3.8014\n",
      "Epoch 1274/10000000\n",
      " - 0s - loss: 4.1543 - val_loss: 3.5306\n",
      "Epoch 1275/10000000\n",
      " - 0s - loss: 4.2070 - val_loss: 3.8341\n",
      "Epoch 1276/10000000\n",
      " - 0s - loss: 4.5871 - val_loss: 3.7059\n",
      "Epoch 1277/10000000\n",
      " - 0s - loss: 4.0684 - val_loss: 3.6692\n",
      "Epoch 1278/10000000\n",
      " - 0s - loss: 4.2705 - val_loss: 3.9492\n",
      "Epoch 1279/10000000\n",
      " - 0s - loss: 4.3097 - val_loss: 3.4824\n",
      "Epoch 1280/10000000\n",
      " - 0s - loss: 4.2819 - val_loss: 3.4861\n",
      "Epoch 1281/10000000\n",
      " - 0s - loss: 4.0150 - val_loss: 4.3188\n",
      "Epoch 1282/10000000\n",
      " - 0s - loss: 4.3668 - val_loss: 3.6599\n",
      "Epoch 1283/10000000\n",
      " - 0s - loss: 4.8884 - val_loss: 3.4313\n",
      "Epoch 1284/10000000\n",
      " - 0s - loss: 4.3288 - val_loss: 4.3370\n",
      "Epoch 1285/10000000\n",
      " - 0s - loss: 4.1264 - val_loss: 3.6232\n",
      "Epoch 1286/10000000\n",
      " - 0s - loss: 4.5405 - val_loss: 3.4346\n",
      "Epoch 1287/10000000\n",
      " - 0s - loss: 3.9981 - val_loss: 4.0332\n",
      "Epoch 1288/10000000\n",
      " - 0s - loss: 4.3550 - val_loss: 3.5932\n",
      "Epoch 1289/10000000\n",
      " - 0s - loss: 4.6569 - val_loss: 3.3661\n",
      "Epoch 1290/10000000\n",
      " - 0s - loss: 3.8593 - val_loss: 4.3924\n",
      "Epoch 1291/10000000\n",
      " - 0s - loss: 4.2684 - val_loss: 3.5134\n",
      "Epoch 1292/10000000\n",
      " - 0s - loss: 5.2572 - val_loss: 3.5742\n",
      "Epoch 1293/10000000\n",
      " - 0s - loss: 5.3020 - val_loss: 3.8947\n",
      "Epoch 1294/10000000\n",
      " - 0s - loss: 4.5458 - val_loss: 5.5284\n",
      "Epoch 1295/10000000\n",
      " - 0s - loss: 5.9447 - val_loss: 4.6889\n",
      "Epoch 1296/10000000\n",
      " - 0s - loss: 4.7314 - val_loss: 3.4215\n",
      "Epoch 1297/10000000\n",
      " - 0s - loss: 3.8204 - val_loss: 3.4764\n",
      "Epoch 1298/10000000\n",
      " - 0s - loss: 4.0077 - val_loss: 4.0517\n",
      "Epoch 1299/10000000\n",
      " - 0s - loss: 4.8298 - val_loss: 3.3926\n",
      "Epoch 1300/10000000\n",
      " - 0s - loss: 4.5462 - val_loss: 3.5346\n",
      "Epoch 1301/10000000\n",
      " - 0s - loss: 3.9115 - val_loss: 6.2905\n",
      "Epoch 1302/10000000\n",
      " - 0s - loss: 5.5078 - val_loss: 4.3187\n",
      "Epoch 1303/10000000\n",
      " - 0s - loss: 5.1521 - val_loss: 3.3196\n",
      "Epoch 1304/10000000\n",
      " - 0s - loss: 3.9342 - val_loss: 4.2894\n",
      "Epoch 1305/10000000\n",
      " - 0s - loss: 3.9665 - val_loss: 3.6192\n",
      "Epoch 1306/10000000\n",
      " - 0s - loss: 4.2596 - val_loss: 3.4284\n",
      "Epoch 1307/10000000\n",
      " - 0s - loss: 4.0675 - val_loss: 3.4309\n",
      "Epoch 1308/10000000\n",
      " - 0s - loss: 3.8256 - val_loss: 3.3673\n",
      "Epoch 1309/10000000\n",
      " - 0s - loss: 3.8653 - val_loss: 4.2781\n",
      "Epoch 1310/10000000\n",
      " - 0s - loss: 4.0857 - val_loss: 3.2591\n",
      "Epoch 1311/10000000\n",
      " - 0s - loss: 4.5594 - val_loss: 3.9059\n",
      "Epoch 1312/10000000\n",
      " - 0s - loss: 5.6960 - val_loss: 3.2564\n",
      "Epoch 1313/10000000\n",
      " - 0s - loss: 4.9727 - val_loss: 6.6143\n",
      "Epoch 1314/10000000\n",
      " - 0s - loss: 6.0454 - val_loss: 6.6844\n",
      "Epoch 1315/10000000\n",
      " - 0s - loss: 5.9583 - val_loss: 3.7816\n",
      "Epoch 1316/10000000\n",
      " - 0s - loss: 5.8726 - val_loss: 3.2531\n",
      "Epoch 1317/10000000\n",
      " - 0s - loss: 4.4245 - val_loss: 4.5733\n",
      "Epoch 1318/10000000\n",
      " - 0s - loss: 4.1741 - val_loss: 3.9802\n",
      "Epoch 1319/10000000\n",
      " - 0s - loss: 4.4235 - val_loss: 3.4951\n",
      "Epoch 1320/10000000\n",
      " - 0s - loss: 3.7164 - val_loss: 3.1864\n",
      "Epoch 1321/10000000\n",
      " - 0s - loss: 3.6876 - val_loss: 3.2849\n",
      "Epoch 1322/10000000\n",
      " - 0s - loss: 3.5764 - val_loss: 3.7751\n",
      "Epoch 1323/10000000\n",
      " - 0s - loss: 3.7293 - val_loss: 3.2196\n",
      "Epoch 1324/10000000\n",
      " - 0s - loss: 3.6656 - val_loss: 3.1651\n",
      "Epoch 1325/10000000\n",
      " - 0s - loss: 3.6452 - val_loss: 3.7073\n",
      "Epoch 1326/10000000\n",
      " - 0s - loss: 3.9408 - val_loss: 3.1338\n",
      "Epoch 1327/10000000\n",
      " - 0s - loss: 3.7121 - val_loss: 3.1617\n",
      "Epoch 1328/10000000\n",
      " - 0s - loss: 3.4249 - val_loss: 3.8800\n",
      "Epoch 1329/10000000\n",
      " - 0s - loss: 3.8924 - val_loss: 3.1415\n",
      "Epoch 1330/10000000\n",
      " - 0s - loss: 3.5917 - val_loss: 3.1056\n",
      "Epoch 1331/10000000\n",
      " - 0s - loss: 3.4169 - val_loss: 3.4484\n",
      "Epoch 1332/10000000\n",
      " - 0s - loss: 3.6190 - val_loss: 3.1884\n",
      "Epoch 1333/10000000\n",
      " - 0s - loss: 3.4938 - val_loss: 3.3360\n",
      "Epoch 1334/10000000\n",
      " - 0s - loss: 3.5593 - val_loss: 3.1004\n",
      "Epoch 1335/10000000\n",
      " - 0s - loss: 3.3670 - val_loss: 3.1055\n",
      "Epoch 1336/10000000\n",
      " - 0s - loss: 3.4028 - val_loss: 3.2072\n",
      "Epoch 1337/10000000\n",
      " - 0s - loss: 3.4594 - val_loss: 3.0638\n",
      "Epoch 1338/10000000\n",
      " - 0s - loss: 3.3138 - val_loss: 3.0330\n",
      "Epoch 1339/10000000\n",
      " - 0s - loss: 3.3097 - val_loss: 3.0437\n",
      "Epoch 1340/10000000\n",
      " - 0s - loss: 3.3760 - val_loss: 3.1637\n",
      "Epoch 1341/10000000\n",
      " - 0s - loss: 3.3140 - val_loss: 3.0297\n",
      "Epoch 1342/10000000\n",
      " - 0s - loss: 3.3018 - val_loss: 3.0238\n",
      "Epoch 1343/10000000\n",
      " - 0s - loss: 3.3177 - val_loss: 3.2961\n",
      "Epoch 1344/10000000\n",
      " - 0s - loss: 3.5127 - val_loss: 3.0134\n",
      "Epoch 1345/10000000\n",
      " - 0s - loss: 3.8881 - val_loss: 3.0934\n",
      "Epoch 1346/10000000\n",
      " - 0s - loss: 3.3394 - val_loss: 4.2742\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1347/10000000\n",
      " - 0s - loss: 3.6903 - val_loss: 3.7226\n",
      "Epoch 1348/10000000\n",
      " - 0s - loss: 4.8460 - val_loss: 3.2658\n",
      "Epoch 1349/10000000\n",
      " - 0s - loss: 4.1240 - val_loss: 3.1851\n",
      "Epoch 1350/10000000\n",
      " - 0s - loss: 3.6349 - val_loss: 3.0715\n",
      "Epoch 1351/10000000\n",
      " - 0s - loss: 3.0529 - val_loss: 5.3122\n",
      "Epoch 1352/10000000\n",
      " - 0s - loss: 4.4652 - val_loss: 3.8978\n",
      "Epoch 1353/10000000\n",
      " - 0s - loss: 6.0339 - val_loss: 2.9468\n",
      "Epoch 1354/10000000\n",
      " - 0s - loss: 4.6072 - val_loss: 4.1561\n",
      "Epoch 1355/10000000\n",
      " - 0s - loss: 4.1199 - val_loss: 3.5735\n",
      "Epoch 1356/10000000\n",
      " - 0s - loss: 3.9340 - val_loss: 4.3880\n",
      "Epoch 1357/10000000\n",
      " - 0s - loss: 3.7959 - val_loss: 3.2653\n",
      "Epoch 1358/10000000\n",
      " - 0s - loss: 3.8037 - val_loss: 3.1700\n",
      "Epoch 1359/10000000\n",
      " - 0s - loss: 3.8360 - val_loss: 2.9121\n",
      "Epoch 1360/10000000\n",
      " - 0s - loss: 4.1982 - val_loss: 3.0833\n",
      "Epoch 1361/10000000\n",
      " - 0s - loss: 3.6434 - val_loss: 4.8926\n",
      "Epoch 1362/10000000\n",
      " - 0s - loss: 4.0725 - val_loss: 3.3390\n",
      "Epoch 1363/10000000\n",
      " - 0s - loss: 4.0195 - val_loss: 2.9908\n",
      "Epoch 1364/10000000\n",
      " - 0s - loss: 3.7485 - val_loss: 2.9524\n",
      "Epoch 1365/10000000\n",
      " - 0s - loss: 3.4766 - val_loss: 3.3875\n",
      "Epoch 1366/10000000\n",
      " - 0s - loss: 3.6842 - val_loss: 4.9816\n",
      "Epoch 1367/10000000\n",
      " - 0s - loss: 5.3987 - val_loss: 2.8852\n",
      "Epoch 1368/10000000\n",
      " - 0s - loss: 4.4071 - val_loss: 3.9784\n",
      "Epoch 1369/10000000\n",
      " - 0s - loss: 4.0054 - val_loss: 5.9012\n",
      "Epoch 1370/10000000\n",
      " - 0s - loss: 4.7846 - val_loss: 3.4001\n",
      "Epoch 1371/10000000\n",
      " - 0s - loss: 4.1505 - val_loss: 3.5445\n",
      "Epoch 1372/10000000\n",
      " - 0s - loss: 4.8317 - val_loss: 3.2932\n",
      "Epoch 1373/10000000\n",
      " - 0s - loss: 3.6263 - val_loss: 4.0860\n",
      "Epoch 1374/10000000\n",
      " - 0s - loss: 4.1201 - val_loss: 4.7514\n",
      "Epoch 1375/10000000\n",
      " - 0s - loss: 4.3354 - val_loss: 3.0442\n",
      "Epoch 1376/10000000\n",
      " - 0s - loss: 3.6481 - val_loss: 2.7730\n",
      "Epoch 1377/10000000\n",
      " - 0s - loss: 3.2624 - val_loss: 3.4476\n",
      "Epoch 1378/10000000\n",
      " - 0s - loss: 3.2851 - val_loss: 3.0308\n",
      "Epoch 1379/10000000\n",
      " - 0s - loss: 3.4688 - val_loss: 2.7752\n",
      "Epoch 1380/10000000\n",
      " - 0s - loss: 3.0998 - val_loss: 2.8293\n",
      "Epoch 1381/10000000\n",
      " - 0s - loss: 3.1532 - val_loss: 2.8587\n",
      "Epoch 1382/10000000\n",
      " - 0s - loss: 3.0994 - val_loss: 3.1347\n",
      "Epoch 1383/10000000\n",
      " - 0s - loss: 3.1234 - val_loss: 2.8203\n",
      "Epoch 1384/10000000\n",
      " - 0s - loss: 3.1956 - val_loss: 2.8114\n",
      "Epoch 1385/10000000\n",
      " - 0s - loss: 3.0485 - val_loss: 3.0180\n",
      "Epoch 1386/10000000\n",
      " - 0s - loss: 3.1382 - val_loss: 2.7802\n",
      "Epoch 1387/10000000\n",
      " - 0s - loss: 3.2617 - val_loss: 2.7889\n",
      "Epoch 1388/10000000\n",
      " - 0s - loss: 3.1638 - val_loss: 3.0514\n",
      "Epoch 1389/10000000\n",
      " - 0s - loss: 3.3157 - val_loss: 2.8029\n",
      "Epoch 1390/10000000\n",
      " - 0s - loss: 3.0228 - val_loss: 3.2313\n",
      "Epoch 1391/10000000\n",
      " - 0s - loss: 3.0941 - val_loss: 2.8108\n",
      "Epoch 1392/10000000\n",
      " - 0s - loss: 3.1618 - val_loss: 2.7629\n",
      "Epoch 1393/10000000\n",
      " - 0s - loss: 3.2283 - val_loss: 2.7647\n",
      "Epoch 1394/10000000\n",
      " - 0s - loss: 3.0523 - val_loss: 2.8075\n",
      "Epoch 1395/10000000\n",
      " - 0s - loss: 3.2038 - val_loss: 2.7975\n",
      "Epoch 1396/10000000\n",
      " - 0s - loss: 2.9214 - val_loss: 3.1794\n",
      "Epoch 1397/10000000\n",
      " - 0s - loss: 3.1971 - val_loss: 2.7492\n",
      "Epoch 1398/10000000\n",
      " - 0s - loss: 2.9107 - val_loss: 2.7956\n",
      "Epoch 1399/10000000\n",
      " - 0s - loss: 2.9544 - val_loss: 3.0197\n",
      "Epoch 1400/10000000\n",
      " - 0s - loss: 3.0392 - val_loss: 2.9406\n",
      "Epoch 1401/10000000\n",
      " - 0s - loss: 2.8920 - val_loss: 2.7942\n",
      "Epoch 1402/10000000\n",
      " - 0s - loss: 3.0880 - val_loss: 2.8775\n",
      "Epoch 1403/10000000\n",
      " - 0s - loss: 2.9401 - val_loss: 2.8416\n",
      "Epoch 1404/10000000\n",
      " - 0s - loss: 2.8088 - val_loss: 2.9551\n",
      "Epoch 1405/10000000\n",
      " - 0s - loss: 3.2797 - val_loss: 2.7578\n",
      "Epoch 1406/10000000\n",
      " - 0s - loss: 2.8459 - val_loss: 2.9026\n",
      "Epoch 1407/10000000\n",
      " - 0s - loss: 2.9211 - val_loss: 2.8002\n",
      "Epoch 1408/10000000\n",
      " - 0s - loss: 2.8621 - val_loss: 2.7338\n",
      "Epoch 1409/10000000\n",
      " - 0s - loss: 2.9574 - val_loss: 2.7876\n",
      "Epoch 1410/10000000\n",
      " - 0s - loss: 2.8603 - val_loss: 2.8595\n",
      "Epoch 1411/10000000\n",
      " - 0s - loss: 2.8466 - val_loss: 2.7524\n",
      "Epoch 1412/10000000\n",
      " - 0s - loss: 2.9725 - val_loss: 2.7307\n",
      "Epoch 1413/10000000\n",
      " - 0s - loss: 3.0671 - val_loss: 2.7333\n",
      "Epoch 1414/10000000\n",
      " - 0s - loss: 2.8564 - val_loss: 2.7822\n",
      "Epoch 1415/10000000\n",
      " - 0s - loss: 2.8171 - val_loss: 2.7468\n",
      "Epoch 1416/10000000\n",
      " - 0s - loss: 2.9885 - val_loss: 2.8477\n",
      "Epoch 1417/10000000\n",
      " - 0s - loss: 3.2158 - val_loss: 2.8022\n",
      "Epoch 1418/10000000\n",
      " - 0s - loss: 3.0318 - val_loss: 2.9414\n",
      "Epoch 1419/10000000\n",
      " - 0s - loss: 3.0531 - val_loss: 2.8022\n",
      "Epoch 1420/10000000\n",
      " - 0s - loss: 2.8209 - val_loss: 3.4281\n",
      "Epoch 1421/10000000\n",
      " - 0s - loss: 3.1874 - val_loss: 2.7397\n",
      "Epoch 1422/10000000\n",
      " - 0s - loss: 2.8433 - val_loss: 2.7316\n",
      "Epoch 1423/10000000\n",
      " - 0s - loss: 2.8363 - val_loss: 2.7057\n",
      "Epoch 1424/10000000\n",
      " - 0s - loss: 3.0849 - val_loss: 2.7003\n",
      "Epoch 1425/10000000\n",
      " - 0s - loss: 2.8024 - val_loss: 3.0443\n",
      "Epoch 1426/10000000\n",
      " - 0s - loss: 2.9571 - val_loss: 2.8341\n",
      "Epoch 1427/10000000\n",
      " - 0s - loss: 3.3075 - val_loss: 2.6576\n",
      "Epoch 1428/10000000\n",
      " - 0s - loss: 2.7519 - val_loss: 2.7772\n",
      "Epoch 1429/10000000\n",
      " - 0s - loss: 2.8096 - val_loss: 2.6550\n",
      "Epoch 1430/10000000\n",
      " - 0s - loss: 2.8535 - val_loss: 2.7320\n",
      "Epoch 1431/10000000\n",
      " - 0s - loss: 2.7387 - val_loss: 2.8357\n",
      "Epoch 1432/10000000\n",
      " - 0s - loss: 2.8070 - val_loss: 2.7181\n",
      "Epoch 1433/10000000\n",
      " - 0s - loss: 3.1399 - val_loss: 2.7763\n",
      "Epoch 1434/10000000\n",
      " - 0s - loss: 2.9187 - val_loss: 3.1697\n",
      "Epoch 1435/10000000\n",
      " - 0s - loss: 3.0392 - val_loss: 2.7646\n",
      "Epoch 1436/10000000\n",
      " - 0s - loss: 2.8562 - val_loss: 3.5602\n",
      "Epoch 1437/10000000\n",
      " - 0s - loss: 3.3948 - val_loss: 2.7918\n",
      "Epoch 1438/10000000\n",
      " - 0s - loss: 3.7751 - val_loss: 2.5906\n",
      "Epoch 1439/10000000\n",
      " - 0s - loss: 3.3281 - val_loss: 3.0531\n",
      "Epoch 1440/10000000\n",
      " - 0s - loss: 3.5258 - val_loss: 3.6178\n",
      "Epoch 1441/10000000\n",
      " - 0s - loss: 3.6539 - val_loss: 3.6653\n",
      "Epoch 1442/10000000\n",
      " - 0s - loss: 3.3693 - val_loss: 2.6196\n",
      "Epoch 1443/10000000\n",
      " - 0s - loss: 2.7301 - val_loss: 2.8900\n",
      "Epoch 1444/10000000\n",
      " - 0s - loss: 2.9258 - val_loss: 2.8064\n",
      "Epoch 1445/10000000\n",
      " - 0s - loss: 3.2214 - val_loss: 2.6580\n",
      "Epoch 1446/10000000\n",
      " - 0s - loss: 2.8144 - val_loss: 3.0560\n",
      "Epoch 1447/10000000\n",
      " - 0s - loss: 2.7637 - val_loss: 2.6782\n",
      "Epoch 1448/10000000\n",
      " - 0s - loss: 2.8340 - val_loss: 2.7750\n",
      "Epoch 1449/10000000\n",
      " - 0s - loss: 2.7400 - val_loss: 2.7367\n",
      "Epoch 1450/10000000\n",
      " - 0s - loss: 2.7389 - val_loss: 2.6554\n",
      "Epoch 1451/10000000\n",
      " - 0s - loss: 2.7390 - val_loss: 2.8728\n",
      "Epoch 1452/10000000\n",
      " - 0s - loss: 2.8946 - val_loss: 2.7631\n",
      "Epoch 1453/10000000\n",
      " - 0s - loss: 2.7198 - val_loss: 2.5584\n",
      "Epoch 1454/10000000\n",
      " - 0s - loss: 2.6723 - val_loss: 2.5991\n",
      "Epoch 1455/10000000\n",
      " - 0s - loss: 2.6564 - val_loss: 2.5945\n",
      "Epoch 1456/10000000\n",
      " - 0s - loss: 2.9270 - val_loss: 2.6354\n",
      "Epoch 1457/10000000\n",
      " - 0s - loss: 3.2035 - val_loss: 2.5448\n",
      "Epoch 1458/10000000\n",
      " - 0s - loss: 3.0863 - val_loss: 2.6051\n",
      "Epoch 1459/10000000\n",
      " - 0s - loss: 2.7814 - val_loss: 3.0573\n",
      "Epoch 1460/10000000\n",
      " - 0s - loss: 2.7582 - val_loss: 2.9266\n",
      "Epoch 1461/10000000\n",
      " - 0s - loss: 3.5904 - val_loss: 2.5373\n",
      "Epoch 1462/10000000\n",
      " - 0s - loss: 3.0894 - val_loss: 3.1940\n",
      "Epoch 1463/10000000\n",
      " - 0s - loss: 2.7223 - val_loss: 2.8612\n",
      "Epoch 1464/10000000\n",
      " - 0s - loss: 3.1293 - val_loss: 2.6828\n",
      "Epoch 1465/10000000\n",
      " - 0s - loss: 2.7333 - val_loss: 2.4869\n",
      "Epoch 1466/10000000\n",
      " - 0s - loss: 2.7963 - val_loss: 2.4871\n",
      "Epoch 1467/10000000\n",
      " - 0s - loss: 2.6440 - val_loss: 2.6821\n",
      "Epoch 1468/10000000\n",
      " - 0s - loss: 2.6124 - val_loss: 2.4870\n",
      "Epoch 1469/10000000\n",
      " - 0s - loss: 2.7300 - val_loss: 2.9423\n",
      "Epoch 1470/10000000\n",
      " - 0s - loss: 3.3573 - val_loss: 2.6821\n",
      "Epoch 1471/10000000\n",
      " - 0s - loss: 2.6266 - val_loss: 2.6197\n",
      "Epoch 1472/10000000\n",
      " - 0s - loss: 2.6498 - val_loss: 2.9380\n",
      "Epoch 1473/10000000\n",
      " - 0s - loss: 2.8003 - val_loss: 2.5993\n",
      "Epoch 1474/10000000\n",
      " - 0s - loss: 2.9199 - val_loss: 2.5210\n",
      "Epoch 1475/10000000\n",
      " - 0s - loss: 2.9970 - val_loss: 2.4777\n",
      "Epoch 1476/10000000\n",
      " - 0s - loss: 3.1053 - val_loss: 2.5526\n",
      "Epoch 1477/10000000\n",
      " - 0s - loss: 2.8125 - val_loss: 3.3429\n",
      "Epoch 1478/10000000\n",
      " - 0s - loss: 2.8322 - val_loss: 2.8383\n",
      "Epoch 1479/10000000\n",
      " - 0s - loss: 3.1318 - val_loss: 2.8334\n",
      "Epoch 1480/10000000\n",
      " - 0s - loss: 2.8646 - val_loss: 2.4451\n",
      "Epoch 1481/10000000\n",
      " - 0s - loss: 2.5108 - val_loss: 2.4288\n",
      "Epoch 1482/10000000\n",
      " - 0s - loss: 2.6373 - val_loss: 2.4535\n",
      "Epoch 1483/10000000\n",
      " - 0s - loss: 2.5409 - val_loss: 2.4595\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1484/10000000\n",
      " - 0s - loss: 2.8952 - val_loss: 2.4910\n",
      "Epoch 1485/10000000\n",
      " - 0s - loss: 2.7863 - val_loss: 2.6719\n",
      "Epoch 1486/10000000\n",
      " - 0s - loss: 2.6013 - val_loss: 2.5096\n",
      "Epoch 1487/10000000\n",
      " - 0s - loss: 2.6962 - val_loss: 2.7205\n",
      "Epoch 1488/10000000\n",
      " - 0s - loss: 2.7127 - val_loss: 2.4071\n",
      "Epoch 1489/10000000\n",
      " - 0s - loss: 2.4658 - val_loss: 2.4665\n",
      "Epoch 1490/10000000\n",
      " - 0s - loss: 2.7103 - val_loss: 2.5893\n",
      "Epoch 1491/10000000\n",
      " - 0s - loss: 2.6949 - val_loss: 2.4126\n",
      "Epoch 1492/10000000\n",
      " - 0s - loss: 2.7541 - val_loss: 2.5028\n",
      "Epoch 1493/10000000\n",
      " - 0s - loss: 2.7892 - val_loss: 2.9719\n",
      "Epoch 1494/10000000\n",
      " - 0s - loss: 2.6146 - val_loss: 2.8132\n",
      "Epoch 1495/10000000\n",
      " - 0s - loss: 3.0843 - val_loss: 3.2242\n",
      "Epoch 1496/10000000\n",
      " - 0s - loss: 2.7241 - val_loss: 2.6253\n",
      "Epoch 1497/10000000\n",
      " - 0s - loss: 3.6451 - val_loss: 3.0720\n",
      "Epoch 1498/10000000\n",
      " - 0s - loss: 4.3070 - val_loss: 2.3793\n",
      "Epoch 1499/10000000\n",
      " - 0s - loss: 3.4287 - val_loss: 3.2731\n",
      "Epoch 1500/10000000\n",
      " - 0s - loss: 3.1896 - val_loss: 6.0110\n",
      "Epoch 1501/10000000\n",
      " - 0s - loss: 4.8340 - val_loss: 2.7792\n",
      "Epoch 1502/10000000\n",
      " - 0s - loss: 3.2309 - val_loss: 2.3201\n",
      "Epoch 1503/10000000\n",
      " - 0s - loss: 2.5862 - val_loss: 2.2643\n",
      "Epoch 1504/10000000\n",
      " - 0s - loss: 2.7623 - val_loss: 2.2845\n",
      "Epoch 1505/10000000\n",
      " - 0s - loss: 2.5147 - val_loss: 3.1318\n",
      "Epoch 1506/10000000\n",
      " - 0s - loss: 2.6944 - val_loss: 2.4878\n",
      "Epoch 1507/10000000\n",
      " - 0s - loss: 2.9180 - val_loss: 2.5241\n",
      "Epoch 1508/10000000\n",
      " - 0s - loss: 2.6349 - val_loss: 2.2402\n",
      "Epoch 1509/10000000\n",
      " - 0s - loss: 2.9336 - val_loss: 2.2997\n",
      "Epoch 1510/10000000\n",
      " - 0s - loss: 3.1144 - val_loss: 2.6909\n",
      "Epoch 1511/10000000\n",
      " - 0s - loss: 2.8391 - val_loss: 2.7587\n",
      "Epoch 1512/10000000\n",
      " - 0s - loss: 2.6325 - val_loss: 4.6382\n",
      "Epoch 1513/10000000\n",
      " - 0s - loss: 3.8817 - val_loss: 2.6552\n",
      "Epoch 1514/10000000\n",
      " - 0s - loss: 3.7319 - val_loss: 2.2602\n",
      "Epoch 1515/10000000\n",
      " - 0s - loss: 2.3447 - val_loss: 3.5821\n",
      "Epoch 1516/10000000\n",
      " - 0s - loss: 2.9965 - val_loss: 2.7116\n",
      "Epoch 1517/10000000\n",
      " - 0s - loss: 3.1120 - val_loss: 2.2975\n",
      "Epoch 1518/10000000\n",
      " - 0s - loss: 2.3510 - val_loss: 2.8211\n",
      "Epoch 1519/10000000\n",
      " - 0s - loss: 2.6597 - val_loss: 2.2912\n",
      "Epoch 1520/10000000\n",
      " - 0s - loss: 2.4844 - val_loss: 2.2803\n",
      "Epoch 1521/10000000\n",
      " - 0s - loss: 2.4062 - val_loss: 2.3657\n",
      "Epoch 1522/10000000\n",
      " - 0s - loss: 2.3727 - val_loss: 2.2389\n",
      "Epoch 1523/10000000\n",
      " - 0s - loss: 2.5376 - val_loss: 2.2154\n",
      "Epoch 1524/10000000\n",
      " - 0s - loss: 2.3917 - val_loss: 2.2115\n",
      "Epoch 1525/10000000\n",
      " - 0s - loss: 2.3889 - val_loss: 2.5110\n",
      "Epoch 1526/10000000\n",
      " - 0s - loss: 2.5221 - val_loss: 2.2400\n",
      "Epoch 1527/10000000\n",
      " - 0s - loss: 2.3666 - val_loss: 2.2534\n",
      "Epoch 1528/10000000\n",
      " - 0s - loss: 2.3735 - val_loss: 2.6583\n",
      "Epoch 1529/10000000\n",
      " - 0s - loss: 2.5385 - val_loss: 2.2772\n",
      "Epoch 1530/10000000\n",
      " - 0s - loss: 2.3407 - val_loss: 2.2388\n",
      "Epoch 1531/10000000\n",
      " - 0s - loss: 2.4569 - val_loss: 2.3469\n",
      "Epoch 1532/10000000\n",
      " - 0s - loss: 2.5455 - val_loss: 2.2067\n",
      "Epoch 1533/10000000\n",
      " - 0s - loss: 2.4824 - val_loss: 2.3026\n",
      "Epoch 1534/10000000\n",
      " - 0s - loss: 2.3927 - val_loss: 3.1191\n",
      "Epoch 1535/10000000\n",
      " - 0s - loss: 2.8622 - val_loss: 2.2393\n",
      "Epoch 1536/10000000\n",
      " - 0s - loss: 2.4921 - val_loss: 2.4912\n",
      "Epoch 1537/10000000\n",
      " - 0s - loss: 2.4510 - val_loss: 2.3163\n",
      "Epoch 1538/10000000\n",
      " - 0s - loss: 2.4183 - val_loss: 2.2807\n",
      "Epoch 1539/10000000\n",
      " - 0s - loss: 2.7930 - val_loss: 2.5331\n",
      "Epoch 1540/10000000\n",
      " - 0s - loss: 3.5057 - val_loss: 2.3229\n",
      "Epoch 1541/10000000\n",
      " - 0s - loss: 2.5347 - val_loss: 2.8127\n",
      "Epoch 1542/10000000\n",
      " - 0s - loss: 2.8504 - val_loss: 2.9424\n",
      "Epoch 1543/10000000\n",
      " - 0s - loss: 2.8565 - val_loss: 2.1556\n",
      "Epoch 1544/10000000\n",
      " - 0s - loss: 2.3186 - val_loss: 2.1954\n",
      "Epoch 1545/10000000\n",
      " - 0s - loss: 2.3884 - val_loss: 2.2025\n",
      "Epoch 1546/10000000\n",
      " - 0s - loss: 2.4444 - val_loss: 2.2440\n",
      "Epoch 1547/10000000\n",
      " - 0s - loss: 2.5685 - val_loss: 2.2228\n",
      "Epoch 1548/10000000\n",
      " - 0s - loss: 2.3618 - val_loss: 2.2126\n",
      "Epoch 1549/10000000\n",
      " - 0s - loss: 2.2682 - val_loss: 2.4303\n",
      "Epoch 1550/10000000\n",
      " - 0s - loss: 2.4758 - val_loss: 2.2164\n",
      "Epoch 1551/10000000\n",
      " - 0s - loss: 2.3277 - val_loss: 2.3806\n",
      "Epoch 1552/10000000\n",
      " - 0s - loss: 2.2966 - val_loss: 2.1989\n",
      "Epoch 1553/10000000\n",
      " - 0s - loss: 2.3825 - val_loss: 2.6335\n",
      "Epoch 1554/10000000\n",
      " - 0s - loss: 2.5815 - val_loss: 2.2187\n",
      "Epoch 1555/10000000\n",
      " - 0s - loss: 2.5764 - val_loss: 2.1729\n",
      "Epoch 1556/10000000\n",
      " - 0s - loss: 2.2514 - val_loss: 2.9308\n",
      "Epoch 1557/10000000\n",
      " - 0s - loss: 2.6998 - val_loss: 2.2738\n",
      "Epoch 1558/10000000\n",
      " - 0s - loss: 2.4783 - val_loss: 2.3693\n",
      "Epoch 1559/10000000\n",
      " - 0s - loss: 2.3248 - val_loss: 2.2461\n",
      "Epoch 1560/10000000\n",
      " - 0s - loss: 2.2679 - val_loss: 2.2293\n",
      "Epoch 1561/10000000\n",
      " - 0s - loss: 2.3010 - val_loss: 2.2008\n",
      "Epoch 1562/10000000\n",
      " - 0s - loss: 2.2694 - val_loss: 2.2040\n",
      "Epoch 1563/10000000\n",
      " - 0s - loss: 2.3186 - val_loss: 2.2406\n",
      "Epoch 1564/10000000\n",
      " - 0s - loss: 2.3665 - val_loss: 2.1851\n",
      "Epoch 1565/10000000\n",
      " - 0s - loss: 2.2472 - val_loss: 2.2153\n",
      "Epoch 1566/10000000\n",
      " - 0s - loss: 2.2579 - val_loss: 2.2139\n",
      "Epoch 1567/10000000\n",
      " - 0s - loss: 2.2610 - val_loss: 2.1894\n",
      "Epoch 1568/10000000\n",
      " - 0s - loss: 2.5357 - val_loss: 2.1634\n",
      "Epoch 1569/10000000\n",
      " - 0s - loss: 2.4973 - val_loss: 2.2673\n",
      "Epoch 1570/10000000\n",
      " - 0s - loss: 2.3800 - val_loss: 2.1463\n",
      "Epoch 1571/10000000\n",
      " - 0s - loss: 2.3862 - val_loss: 2.2357\n",
      "Epoch 1572/10000000\n",
      " - 0s - loss: 2.3419 - val_loss: 2.2267\n",
      "Epoch 1573/10000000\n",
      " - 0s - loss: 2.2771 - val_loss: 2.6598\n",
      "Epoch 1574/10000000\n",
      " - 0s - loss: 2.3612 - val_loss: 2.4153\n",
      "Epoch 1575/10000000\n",
      " - 0s - loss: 2.5579 - val_loss: 2.3517\n",
      "Epoch 1576/10000000\n",
      " - 0s - loss: 2.3356 - val_loss: 2.1651\n",
      "Epoch 1577/10000000\n",
      " - 0s - loss: 2.3142 - val_loss: 2.2258\n",
      "Epoch 1578/10000000\n",
      " - 0s - loss: 2.4478 - val_loss: 2.4285\n",
      "Epoch 1579/10000000\n",
      " - 0s - loss: 2.3318 - val_loss: 2.0926\n",
      "Epoch 1580/10000000\n",
      " - 0s - loss: 2.2819 - val_loss: 2.1462\n",
      "Epoch 1581/10000000\n",
      " - 0s - loss: 2.2570 - val_loss: 2.1254\n",
      "Epoch 1582/10000000\n",
      " - 0s - loss: 2.2215 - val_loss: 2.4577\n",
      "Epoch 1583/10000000\n",
      " - 0s - loss: 2.3074 - val_loss: 2.3520\n",
      "Epoch 1584/10000000\n",
      " - 0s - loss: 3.1031 - val_loss: 2.3992\n",
      "Epoch 1585/10000000\n",
      " - 0s - loss: 2.8401 - val_loss: 2.2056\n",
      "Epoch 1586/10000000\n",
      " - 0s - loss: 2.4092 - val_loss: 3.1707\n",
      "Epoch 1587/10000000\n",
      " - 0s - loss: 3.0579 - val_loss: 4.8376\n",
      "Epoch 1588/10000000\n",
      " - 0s - loss: 3.6776 - val_loss: 3.3756\n",
      "Epoch 1589/10000000\n",
      " - 0s - loss: 4.1451 - val_loss: 2.1394\n",
      "Epoch 1590/10000000\n",
      " - 0s - loss: 2.7807 - val_loss: 2.4069\n",
      "Epoch 1591/10000000\n",
      " - 0s - loss: 2.4442 - val_loss: 2.6208\n",
      "Epoch 1592/10000000\n",
      " - 0s - loss: 2.6533 - val_loss: 2.8163\n",
      "Epoch 1593/10000000\n",
      " - 0s - loss: 2.5038 - val_loss: 2.5063\n",
      "Epoch 1594/10000000\n",
      " - 0s - loss: 2.8341 - val_loss: 2.2563\n",
      "Epoch 1595/10000000\n",
      " - 0s - loss: 2.4627 - val_loss: 2.1116\n",
      "Epoch 1596/10000000\n",
      " - 0s - loss: 2.3147 - val_loss: 2.0648\n",
      "Epoch 1597/10000000\n",
      " - 0s - loss: 2.2167 - val_loss: 2.2830\n",
      "Epoch 1598/10000000\n",
      " - 0s - loss: 2.2105 - val_loss: 2.0760\n",
      "Epoch 1599/10000000\n",
      " - 0s - loss: 2.2424 - val_loss: 2.3876\n",
      "Epoch 1600/10000000\n",
      " - 0s - loss: 2.3893 - val_loss: 2.0980\n",
      "Epoch 1601/10000000\n",
      " - 0s - loss: 2.2342 - val_loss: 2.2683\n",
      "Epoch 1602/10000000\n",
      " - 0s - loss: 2.3527 - val_loss: 2.6784\n",
      "Epoch 1603/10000000\n",
      " - 0s - loss: 2.4671 - val_loss: 2.1582\n",
      "Epoch 1604/10000000\n",
      " - 0s - loss: 2.3946 - val_loss: 2.1395\n",
      "Epoch 1605/10000000\n",
      " - 0s - loss: 2.2739 - val_loss: 2.1418\n",
      "Epoch 1606/10000000\n",
      " - 0s - loss: 2.3218 - val_loss: 2.1113\n",
      "Epoch 1607/10000000\n",
      " - 0s - loss: 2.2237 - val_loss: 2.4955\n",
      "Epoch 1608/10000000\n",
      " - 0s - loss: 2.2711 - val_loss: 2.1471\n",
      "Epoch 1609/10000000\n",
      " - 0s - loss: 2.3487 - val_loss: 2.1778\n",
      "Epoch 1610/10000000\n",
      " - 0s - loss: 2.2982 - val_loss: 2.1013\n",
      "Epoch 1611/10000000\n",
      " - 0s - loss: 2.4091 - val_loss: 2.1431\n",
      "Epoch 1612/10000000\n",
      " - 0s - loss: 2.4863 - val_loss: 2.2412\n",
      "Epoch 1613/10000000\n",
      " - 0s - loss: 2.3057 - val_loss: 2.1355\n",
      "Epoch 1614/10000000\n",
      " - 0s - loss: 2.1663 - val_loss: 2.9007\n",
      "Epoch 1615/10000000\n",
      " - 0s - loss: 2.4628 - val_loss: 2.3157\n",
      "Epoch 1616/10000000\n",
      " - 0s - loss: 2.5230 - val_loss: 2.8615\n",
      "Epoch 1617/10000000\n",
      " - 0s - loss: 2.7848 - val_loss: 2.1642\n",
      "Epoch 1618/10000000\n",
      " - 0s - loss: 2.6845 - val_loss: 2.1381\n",
      "Epoch 1619/10000000\n",
      " - 0s - loss: 2.7668 - val_loss: 2.1359\n",
      "Epoch 1620/10000000\n",
      " - 0s - loss: 2.3614 - val_loss: 3.0387\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1621/10000000\n",
      " - 0s - loss: 2.6920 - val_loss: 4.1844\n",
      "Epoch 1622/10000000\n",
      " - 0s - loss: 3.2646 - val_loss: 2.6684\n",
      "Epoch 1623/10000000\n",
      " - 0s - loss: 3.0284 - val_loss: 2.5911\n",
      "Epoch 1624/10000000\n",
      " - 0s - loss: 2.8739 - val_loss: 2.0734\n",
      "Epoch 1625/10000000\n",
      " - 0s - loss: 2.8522 - val_loss: 2.1815\n",
      "Epoch 1626/10000000\n",
      " - 0s - loss: 2.5266 - val_loss: 4.1539\n",
      "Epoch 1627/10000000\n",
      " - 0s - loss: 2.9130 - val_loss: 2.8047\n",
      "Epoch 1628/10000000\n",
      " - 0s - loss: 3.2727 - val_loss: 2.5079\n",
      "Epoch 1629/10000000\n",
      " - 0s - loss: 2.3573 - val_loss: 2.4039\n",
      "Epoch 1630/10000000\n",
      " - 0s - loss: 2.2043 - val_loss: 2.2784\n",
      "Epoch 1631/10000000\n",
      " - 0s - loss: 2.4643 - val_loss: 2.1570\n",
      "Epoch 1632/10000000\n",
      " - 0s - loss: 2.1295 - val_loss: 2.0621\n",
      "Epoch 1633/10000000\n",
      " - 0s - loss: 2.1308 - val_loss: 2.0175\n",
      "Epoch 1634/10000000\n",
      " - 0s - loss: 2.0867 - val_loss: 2.2017\n",
      "Epoch 1635/10000000\n",
      " - 0s - loss: 2.1881 - val_loss: 2.0336\n",
      "Epoch 1636/10000000\n",
      " - 0s - loss: 2.1560 - val_loss: 2.0814\n",
      "Epoch 1637/10000000\n",
      " - 0s - loss: 2.2111 - val_loss: 2.0598\n",
      "Epoch 1638/10000000\n",
      " - 0s - loss: 2.1541 - val_loss: 2.0850\n",
      "Epoch 1639/10000000\n",
      " - 0s - loss: 2.2525 - val_loss: 2.2718\n",
      "Epoch 1640/10000000\n",
      " - 0s - loss: 2.0679 - val_loss: 2.2951\n",
      "Epoch 1641/10000000\n",
      " - 0s - loss: 2.2905 - val_loss: 2.8071\n",
      "Epoch 1642/10000000\n",
      " - 0s - loss: 2.7128 - val_loss: 2.0371\n",
      "Epoch 1643/10000000\n",
      " - 0s - loss: 2.8094 - val_loss: 2.0297\n",
      "Epoch 1644/10000000\n",
      " - 0s - loss: 2.2089 - val_loss: 2.3915\n",
      "Epoch 1645/10000000\n",
      " - 0s - loss: 2.1639 - val_loss: 2.1387\n",
      "Epoch 1646/10000000\n",
      " - 0s - loss: 2.1050 - val_loss: 2.6877\n",
      "Epoch 1647/10000000\n",
      " - 0s - loss: 2.4190 - val_loss: 2.0947\n",
      "Epoch 1648/10000000\n",
      " - 0s - loss: 2.1799 - val_loss: 2.0752\n",
      "Epoch 1649/10000000\n",
      " - 0s - loss: 2.1036 - val_loss: 2.1225\n",
      "Epoch 1650/10000000\n",
      " - 0s - loss: 2.0850 - val_loss: 2.1784\n",
      "Epoch 1651/10000000\n",
      " - 0s - loss: 2.1780 - val_loss: 2.0552\n",
      "Epoch 1652/10000000\n",
      " - 0s - loss: 2.1802 - val_loss: 2.0138\n",
      "Epoch 1653/10000000\n",
      " - 0s - loss: 2.5203 - val_loss: 2.1026\n",
      "Epoch 1654/10000000\n",
      " - 0s - loss: 2.2423 - val_loss: 2.0140\n",
      "Epoch 1655/10000000\n",
      " - 0s - loss: 2.1538 - val_loss: 2.2098\n",
      "Epoch 1656/10000000\n",
      " - 0s - loss: 2.1987 - val_loss: 2.4731\n",
      "Epoch 1657/10000000\n",
      " - 0s - loss: 2.4071 - val_loss: 3.3148\n",
      "Epoch 1658/10000000\n",
      " - 0s - loss: 3.2208 - val_loss: 2.4809\n",
      "Epoch 1659/10000000\n",
      " - 0s - loss: 3.9197 - val_loss: 2.4342\n",
      "Epoch 1660/10000000\n",
      " - 0s - loss: 4.3531 - val_loss: 2.3882\n",
      "Epoch 1661/10000000\n",
      " - 0s - loss: 2.8403 - val_loss: 2.6861\n",
      "Epoch 1662/10000000\n",
      " - 0s - loss: 2.5899 - val_loss: 3.8790\n",
      "Epoch 1663/10000000\n",
      " - 0s - loss: 2.6935 - val_loss: 3.3083\n",
      "Epoch 1664/10000000\n",
      " - 0s - loss: 3.1696 - val_loss: 2.7456\n",
      "Epoch 1665/10000000\n",
      " - 0s - loss: 2.6912 - val_loss: 2.0654\n",
      "Epoch 1666/10000000\n",
      " - 0s - loss: 2.0324 - val_loss: 2.2260\n",
      "Epoch 1667/10000000\n",
      " - 0s - loss: 2.2192 - val_loss: 2.3398\n",
      "Epoch 1668/10000000\n",
      " - 0s - loss: 2.1766 - val_loss: 2.1367\n",
      "Epoch 1669/10000000\n",
      " - 0s - loss: 2.1808 - val_loss: 2.0144\n",
      "Epoch 1670/10000000\n",
      " - 0s - loss: 2.0007 - val_loss: 1.9888\n",
      "Epoch 1671/10000000\n",
      " - 0s - loss: 2.0944 - val_loss: 2.0707\n",
      "Epoch 1672/10000000\n",
      " - 0s - loss: 2.1074 - val_loss: 2.0458\n",
      "Epoch 1673/10000000\n",
      " - 0s - loss: 2.1268 - val_loss: 2.0216\n",
      "Epoch 1674/10000000\n",
      " - 0s - loss: 2.0702 - val_loss: 2.0456\n",
      "Epoch 1675/10000000\n",
      " - 0s - loss: 2.1634 - val_loss: 2.1170\n",
      "Epoch 1676/10000000\n",
      " - 0s - loss: 2.0350 - val_loss: 2.4896\n",
      "Epoch 1677/10000000\n",
      " - 0s - loss: 2.1824 - val_loss: 2.1089\n",
      "Epoch 1678/10000000\n",
      " - 0s - loss: 2.3645 - val_loss: 2.1995\n",
      "Epoch 1679/10000000\n",
      " - 0s - loss: 2.1895 - val_loss: 2.1410\n",
      "Epoch 1680/10000000\n",
      " - 0s - loss: 2.9141 - val_loss: 2.2361\n",
      "Epoch 1681/10000000\n",
      " - 0s - loss: 2.8955 - val_loss: 1.9980\n",
      "Epoch 1682/10000000\n",
      " - 0s - loss: 2.7023 - val_loss: 2.8413\n",
      "Epoch 1683/10000000\n",
      " - 0s - loss: 2.7918 - val_loss: 3.4430\n",
      "Epoch 1684/10000000\n",
      " - 0s - loss: 2.7078 - val_loss: 2.7156\n",
      "Epoch 1685/10000000\n",
      " - 0s - loss: 2.7199 - val_loss: 2.1890\n",
      "Epoch 1686/10000000\n",
      " - 0s - loss: 2.2158 - val_loss: 1.9876\n",
      "Epoch 1687/10000000\n",
      " - 0s - loss: 2.2332 - val_loss: 2.0550\n",
      "Epoch 1688/10000000\n",
      " - 0s - loss: 2.0779 - val_loss: 2.3246\n",
      "Epoch 1689/10000000\n",
      " - 0s - loss: 2.0733 - val_loss: 2.1353\n",
      "Epoch 1690/10000000\n",
      " - 0s - loss: 2.0871 - val_loss: 2.2473\n",
      "Epoch 1691/10000000\n",
      " - 0s - loss: 2.0113 - val_loss: 2.2750\n",
      "Epoch 1692/10000000\n",
      " - 0s - loss: 2.2921 - val_loss: 2.1291\n",
      "Epoch 1693/10000000\n",
      " - 0s - loss: 2.1715 - val_loss: 2.0302\n",
      "Epoch 1694/10000000\n",
      " - 0s - loss: 2.2541 - val_loss: 2.0353\n",
      "Epoch 1695/10000000\n",
      " - 0s - loss: 2.6459 - val_loss: 2.0495\n",
      "Epoch 1696/10000000\n",
      " - 0s - loss: 2.7047 - val_loss: 2.1349\n",
      "Epoch 1697/10000000\n",
      " - 0s - loss: 2.3825 - val_loss: 3.0541\n",
      "Epoch 1698/10000000\n",
      " - 0s - loss: 2.5212 - val_loss: 2.3736\n",
      "Epoch 1699/10000000\n",
      " - 0s - loss: 2.3234 - val_loss: 2.2666\n",
      "Epoch 1700/10000000\n",
      " - 0s - loss: 1.9573 - val_loss: 2.2790\n",
      "Epoch 1701/10000000\n",
      " - 0s - loss: 2.2389 - val_loss: 2.3943\n",
      "Epoch 1702/10000000\n",
      " - 0s - loss: 2.5397 - val_loss: 1.9478\n",
      "Epoch 1703/10000000\n",
      " - 0s - loss: 2.0419 - val_loss: 1.9452\n",
      "Epoch 1704/10000000\n",
      " - 0s - loss: 2.1966 - val_loss: 2.1835\n",
      "Epoch 1705/10000000\n",
      " - 0s - loss: 2.0039 - val_loss: 2.0419\n",
      "Epoch 1706/10000000\n",
      " - 0s - loss: 2.0391 - val_loss: 2.0598\n",
      "Epoch 1707/10000000\n",
      " - 0s - loss: 2.0315 - val_loss: 2.0114\n",
      "Epoch 1708/10000000\n",
      " - 0s - loss: 1.9478 - val_loss: 1.9962\n",
      "Epoch 1709/10000000\n",
      " - 0s - loss: 2.0090 - val_loss: 2.1593\n",
      "Epoch 1710/10000000\n",
      " - 0s - loss: 1.9835 - val_loss: 1.9816\n",
      "Epoch 1711/10000000\n",
      " - 0s - loss: 2.0068 - val_loss: 2.3449\n",
      "Epoch 1712/10000000\n",
      " - 0s - loss: 2.4529 - val_loss: 1.9991\n",
      "Epoch 1713/10000000\n",
      " - 0s - loss: 1.9377 - val_loss: 1.9436\n",
      "Epoch 1714/10000000\n",
      " - 0s - loss: 2.0174 - val_loss: 2.0402\n",
      "Epoch 1715/10000000\n",
      " - 0s - loss: 1.9493 - val_loss: 1.9716\n",
      "Epoch 1716/10000000\n",
      " - 0s - loss: 1.9812 - val_loss: 2.0496\n",
      "Epoch 1717/10000000\n",
      " - 0s - loss: 2.0015 - val_loss: 1.9234\n",
      "Epoch 1718/10000000\n",
      " - 0s - loss: 1.9340 - val_loss: 1.9806\n",
      "Epoch 1719/10000000\n",
      " - 0s - loss: 1.9433 - val_loss: 1.9335\n",
      "Epoch 1720/10000000\n",
      " - 0s - loss: 1.9209 - val_loss: 1.9709\n",
      "Epoch 1721/10000000\n",
      " - 0s - loss: 1.9558 - val_loss: 2.2099\n",
      "Epoch 1722/10000000\n",
      " - 0s - loss: 2.0702 - val_loss: 1.9696\n",
      "Epoch 1723/10000000\n",
      " - 0s - loss: 1.9338 - val_loss: 2.0138\n",
      "Epoch 1724/10000000\n",
      " - 0s - loss: 1.9310 - val_loss: 2.1143\n",
      "Epoch 1725/10000000\n",
      " - 0s - loss: 1.9446 - val_loss: 2.0180\n",
      "Epoch 1726/10000000\n",
      " - 0s - loss: 1.9214 - val_loss: 2.0120\n",
      "Epoch 1727/10000000\n",
      " - 0s - loss: 1.8991 - val_loss: 2.1985\n",
      "Epoch 1728/10000000\n",
      " - 0s - loss: 2.0268 - val_loss: 2.0081\n",
      "Epoch 1729/10000000\n",
      " - 0s - loss: 2.0100 - val_loss: 1.9500\n",
      "Epoch 1730/10000000\n",
      " - 0s - loss: 1.9147 - val_loss: 1.9754\n",
      "Epoch 1731/10000000\n",
      " - 0s - loss: 1.8954 - val_loss: 1.9695\n",
      "Epoch 1732/10000000\n",
      " - 0s - loss: 2.0211 - val_loss: 2.1700\n",
      "Epoch 1733/10000000\n",
      " - 0s - loss: 2.1196 - val_loss: 2.1791\n",
      "Epoch 1734/10000000\n",
      " - 0s - loss: 1.9789 - val_loss: 2.0418\n",
      "Epoch 1735/10000000\n",
      " - 0s - loss: 2.0459 - val_loss: 2.1170\n",
      "Epoch 1736/10000000\n",
      " - 0s - loss: 2.2343 - val_loss: 2.0065\n",
      "Epoch 1737/10000000\n",
      " - 0s - loss: 2.0861 - val_loss: 2.0251\n",
      "Epoch 1738/10000000\n",
      " - 0s - loss: 2.0700 - val_loss: 2.0505\n",
      "Epoch 1739/10000000\n",
      " - 0s - loss: 2.2042 - val_loss: 2.1731\n",
      "Epoch 1740/10000000\n",
      " - 0s - loss: 2.4603 - val_loss: 1.9000\n",
      "Epoch 1741/10000000\n",
      " - 0s - loss: 2.6774 - val_loss: 1.9479\n",
      "Epoch 1742/10000000\n",
      " - 0s - loss: 2.1185 - val_loss: 2.1174\n",
      "Epoch 1743/10000000\n",
      " - 0s - loss: 1.9364 - val_loss: 2.2521\n",
      "Epoch 1744/10000000\n",
      " - 0s - loss: 1.9779 - val_loss: 2.8968\n",
      "Epoch 1745/10000000\n",
      " - 0s - loss: 2.5543 - val_loss: 2.4227\n",
      "Epoch 1746/10000000\n",
      " - 0s - loss: 2.6536 - val_loss: 2.7782\n",
      "Epoch 1747/10000000\n",
      " - 0s - loss: 3.7326 - val_loss: 2.0577\n",
      "Epoch 1748/10000000\n",
      " - 0s - loss: 2.5434 - val_loss: 2.2006\n",
      "Epoch 1749/10000000\n",
      " - 0s - loss: 2.2456 - val_loss: 2.2972\n",
      "Epoch 1750/10000000\n",
      " - 0s - loss: 1.9060 - val_loss: 2.8057\n",
      "Epoch 1751/10000000\n",
      " - 0s - loss: 2.4632 - val_loss: 2.5852\n",
      "Epoch 1752/10000000\n",
      " - 0s - loss: 2.2705 - val_loss: 2.0942\n",
      "Epoch 1753/10000000\n",
      " - 0s - loss: 2.0740 - val_loss: 1.9816\n",
      "Epoch 1754/10000000\n",
      " - 0s - loss: 1.8684 - val_loss: 2.0276\n",
      "Epoch 1755/10000000\n",
      " - 0s - loss: 1.9342 - val_loss: 2.1945\n",
      "Epoch 1756/10000000\n",
      " - 0s - loss: 2.5653 - val_loss: 2.0700\n",
      "Epoch 1757/10000000\n",
      " - 0s - loss: 2.1906 - val_loss: 2.1093\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1758/10000000\n",
      " - 0s - loss: 2.0416 - val_loss: 2.2114\n",
      "Epoch 1759/10000000\n",
      " - 0s - loss: 1.9916 - val_loss: 2.3107\n",
      "Epoch 1760/10000000\n",
      " - 0s - loss: 2.0262 - val_loss: 2.0372\n",
      "Epoch 1761/10000000\n",
      " - 0s - loss: 1.9480 - val_loss: 2.0025\n",
      "Epoch 1762/10000000\n",
      " - 0s - loss: 1.9620 - val_loss: 2.0090\n",
      "Epoch 1763/10000000\n",
      " - 0s - loss: 1.9590 - val_loss: 1.9465\n",
      "Epoch 1764/10000000\n",
      " - 0s - loss: 1.8736 - val_loss: 1.9813\n",
      "Epoch 1765/10000000\n",
      " - 0s - loss: 1.9319 - val_loss: 1.9941\n",
      "Epoch 1766/10000000\n",
      " - 0s - loss: 1.8749 - val_loss: 2.0441\n",
      "Epoch 1767/10000000\n",
      " - 0s - loss: 1.9635 - val_loss: 2.1145\n",
      "Epoch 1768/10000000\n",
      " - 0s - loss: 1.8290 - val_loss: 2.1069\n",
      "Epoch 1769/10000000\n",
      " - 0s - loss: 2.0763 - val_loss: 2.0625\n",
      "Epoch 1770/10000000\n",
      " - 0s - loss: 1.8460 - val_loss: 2.0544\n",
      "Epoch 1771/10000000\n",
      " - 0s - loss: 1.8522 - val_loss: 2.1477\n",
      "Epoch 1772/10000000\n",
      " - 0s - loss: 1.9642 - val_loss: 1.9985\n",
      "Epoch 1773/10000000\n",
      " - 0s - loss: 1.9773 - val_loss: 1.9110\n",
      "Epoch 1774/10000000\n",
      " - 0s - loss: 2.1061 - val_loss: 1.9409\n",
      "Epoch 1775/10000000\n",
      " - 0s - loss: 1.8449 - val_loss: 2.1521\n",
      "Epoch 1776/10000000\n",
      " - 0s - loss: 2.0037 - val_loss: 2.1246\n",
      "Epoch 1777/10000000\n",
      " - 0s - loss: 1.9502 - val_loss: 1.9523\n",
      "Epoch 1778/10000000\n",
      " - 0s - loss: 1.8480 - val_loss: 1.9743\n",
      "Epoch 1779/10000000\n",
      " - 0s - loss: 1.8603 - val_loss: 2.0309\n",
      "Epoch 1780/10000000\n",
      " - 0s - loss: 1.8426 - val_loss: 2.0200\n",
      "Epoch 1781/10000000\n",
      " - 0s - loss: 1.8043 - val_loss: 2.2358\n",
      "Epoch 1782/10000000\n",
      " - 0s - loss: 1.9006 - val_loss: 2.0058\n",
      "Epoch 1783/10000000\n",
      " - 0s - loss: 1.8855 - val_loss: 1.9163\n",
      "Epoch 1784/10000000\n",
      " - 0s - loss: 1.9048 - val_loss: 1.9366\n",
      "Epoch 1785/10000000\n",
      " - 0s - loss: 1.8770 - val_loss: 1.9531\n",
      "Epoch 1786/10000000\n",
      " - 0s - loss: 1.8432 - val_loss: 2.0499\n",
      "Epoch 1787/10000000\n",
      " - 0s - loss: 1.9822 - val_loss: 1.9938\n",
      "Epoch 1788/10000000\n",
      " - 0s - loss: 1.9237 - val_loss: 2.1111\n",
      "Epoch 1789/10000000\n",
      " - 0s - loss: 1.8356 - val_loss: 2.2475\n",
      "Epoch 1790/10000000\n",
      " - 0s - loss: 2.1428 - val_loss: 2.3104\n",
      "Epoch 1791/10000000\n",
      " - 0s - loss: 2.1729 - val_loss: 2.0394\n",
      "Epoch 1792/10000000\n",
      " - 0s - loss: 2.2945 - val_loss: 2.2947\n",
      "Epoch 1793/10000000\n",
      " - 0s - loss: 2.6792 - val_loss: 2.1390\n",
      "Epoch 1794/10000000\n",
      " - 0s - loss: 1.8759 - val_loss: 2.1611\n",
      "Epoch 1795/10000000\n",
      " - 0s - loss: 1.9112 - val_loss: 2.8127\n",
      "Epoch 1796/10000000\n",
      " - 0s - loss: 2.1993 - val_loss: 2.4949\n",
      "Epoch 1797/10000000\n",
      " - 0s - loss: 2.3671 - val_loss: 2.1793\n",
      "Epoch 1798/10000000\n",
      " - 0s - loss: 1.9411 - val_loss: 1.9742\n",
      "Epoch 1799/10000000\n",
      " - 0s - loss: 1.8185 - val_loss: 1.9380\n",
      "Epoch 1800/10000000\n",
      " - 0s - loss: 1.7736 - val_loss: 2.1302\n",
      "Epoch 1801/10000000\n",
      " - 0s - loss: 1.9549 - val_loss: 1.9329\n",
      "Epoch 1802/10000000\n",
      " - 0s - loss: 1.9691 - val_loss: 1.8627\n",
      "Epoch 1803/10000000\n",
      " - 0s - loss: 1.8411 - val_loss: 1.9370\n",
      "Epoch 1804/10000000\n",
      " - 0s - loss: 1.8019 - val_loss: 2.0753\n",
      "Epoch 1805/10000000\n",
      " - 0s - loss: 1.8862 - val_loss: 2.2371\n",
      "Epoch 1806/10000000\n",
      " - 0s - loss: 2.0026 - val_loss: 2.0623\n",
      "Epoch 1807/10000000\n",
      " - 0s - loss: 1.8280 - val_loss: 2.0517\n",
      "Epoch 1808/10000000\n",
      " - 0s - loss: 1.9711 - val_loss: 2.0417\n",
      "Epoch 1809/10000000\n",
      " - 0s - loss: 1.8324 - val_loss: 2.0799\n",
      "Epoch 1810/10000000\n",
      " - 0s - loss: 2.0840 - val_loss: 2.1557\n",
      "Epoch 1811/10000000\n",
      " - 0s - loss: 2.0075 - val_loss: 2.0557\n",
      "Epoch 1812/10000000\n",
      " - 0s - loss: 2.7574 - val_loss: 2.1399\n",
      "Epoch 1813/10000000\n",
      " - 0s - loss: 2.6589 - val_loss: 1.9361\n",
      "Epoch 1814/10000000\n",
      " - 0s - loss: 2.1732 - val_loss: 2.3341\n",
      "Epoch 1815/10000000\n",
      " - 0s - loss: 2.1489 - val_loss: 2.6448\n",
      "Epoch 1816/10000000\n",
      " - 0s - loss: 2.0715 - val_loss: 2.3697\n",
      "Epoch 1817/10000000\n",
      " - 0s - loss: 2.0354 - val_loss: 2.4326\n",
      "Epoch 1818/10000000\n",
      " - 0s - loss: 2.0489 - val_loss: 1.9561\n",
      "Epoch 1819/10000000\n",
      " - 0s - loss: 1.8059 - val_loss: 2.2342\n",
      "Epoch 1820/10000000\n",
      " - 0s - loss: 1.9669 - val_loss: 1.9519\n",
      "Epoch 1821/10000000\n",
      " - 0s - loss: 1.8184 - val_loss: 1.9336\n",
      "Epoch 1822/10000000\n",
      " - 0s - loss: 1.7795 - val_loss: 1.9355\n",
      "Epoch 1823/10000000\n",
      " - 0s - loss: 1.9436 - val_loss: 2.1734\n",
      "Epoch 1824/10000000\n",
      " - 0s - loss: 2.1896 - val_loss: 1.8428\n",
      "Epoch 1825/10000000\n",
      " - 0s - loss: 1.7844 - val_loss: 2.3881\n",
      "Epoch 1826/10000000\n",
      " - 0s - loss: 1.9942 - val_loss: 3.0806\n",
      "Epoch 1827/10000000\n",
      " - 0s - loss: 2.8285 - val_loss: 2.1734\n",
      "Epoch 1828/10000000\n",
      " - 0s - loss: 2.2964 - val_loss: 2.0023\n",
      "Epoch 1829/10000000\n",
      " - 0s - loss: 1.7685 - val_loss: 2.1358\n",
      "Epoch 1830/10000000\n",
      " - 0s - loss: 1.7791 - val_loss: 2.1328\n",
      "Epoch 1831/10000000\n",
      " - 0s - loss: 1.8883 - val_loss: 2.0947\n",
      "Epoch 1832/10000000\n",
      " - 0s - loss: 1.7775 - val_loss: 2.0145\n",
      "Epoch 1833/10000000\n",
      " - 0s - loss: 1.7920 - val_loss: 1.9858\n",
      "Epoch 1834/10000000\n",
      " - 0s - loss: 1.7651 - val_loss: 1.9608\n",
      "Epoch 1835/10000000\n",
      " - 0s - loss: 1.8016 - val_loss: 2.0392\n",
      "Epoch 1836/10000000\n",
      " - 0s - loss: 1.8341 - val_loss: 2.1018\n",
      "Epoch 1837/10000000\n",
      " - 0s - loss: 1.9108 - val_loss: 2.0960\n",
      "Epoch 1838/10000000\n",
      " - 0s - loss: 2.1416 - val_loss: 2.2812\n",
      "Epoch 1839/10000000\n",
      " - 0s - loss: 2.5124 - val_loss: 1.9931\n",
      "Epoch 1840/10000000\n",
      " - 0s - loss: 2.1275 - val_loss: 2.0903\n",
      "Epoch 1841/10000000\n",
      " - 0s - loss: 1.9065 - val_loss: 2.3519\n",
      "Epoch 1842/10000000\n",
      " - 0s - loss: 1.8586 - val_loss: 2.1037\n",
      "Epoch 1843/10000000\n",
      " - 0s - loss: 1.8700 - val_loss: 2.1923\n",
      "Epoch 1844/10000000\n",
      " - 0s - loss: 1.9556 - val_loss: 1.9127\n",
      "Epoch 1845/10000000\n",
      " - 0s - loss: 1.7518 - val_loss: 1.9891\n",
      "Epoch 1846/10000000\n",
      " - 0s - loss: 1.8064 - val_loss: 1.8792\n",
      "Epoch 1847/10000000\n",
      " - 0s - loss: 1.7955 - val_loss: 1.9123\n",
      "Epoch 1848/10000000\n",
      " - 0s - loss: 1.7632 - val_loss: 1.9632\n",
      "Epoch 1849/10000000\n",
      " - 0s - loss: 1.7507 - val_loss: 2.1146\n",
      "Epoch 1850/10000000\n",
      " - 0s - loss: 1.7969 - val_loss: 2.0394\n",
      "Epoch 1851/10000000\n",
      " - 0s - loss: 1.8104 - val_loss: 1.9954\n",
      "Epoch 1852/10000000\n",
      " - 0s - loss: 1.7670 - val_loss: 1.9620\n",
      "Epoch 1853/10000000\n",
      " - 0s - loss: 1.7846 - val_loss: 1.9857\n",
      "Epoch 1854/10000000\n",
      " - 0s - loss: 1.8900 - val_loss: 1.9520\n",
      "Epoch 1855/10000000\n",
      " - 0s - loss: 1.8150 - val_loss: 1.9597\n",
      "Epoch 1856/10000000\n",
      " - 0s - loss: 2.0028 - val_loss: 2.0112\n",
      "Epoch 1857/10000000\n",
      " - 0s - loss: 2.1240 - val_loss: 1.9795\n",
      "Epoch 1858/10000000\n",
      " - 0s - loss: 1.6746 - val_loss: 2.6674\n",
      "Epoch 1859/10000000\n",
      " - 0s - loss: 2.2281 - val_loss: 2.3153\n",
      "Epoch 1860/10000000\n",
      " - 0s - loss: 2.1489 - val_loss: 2.0260\n",
      "Epoch 1861/10000000\n",
      " - 0s - loss: 1.7699 - val_loss: 1.9225\n",
      "Epoch 1862/10000000\n",
      " - 0s - loss: 1.7966 - val_loss: 1.9332\n",
      "Epoch 1863/10000000\n",
      " - 0s - loss: 1.8650 - val_loss: 2.3439\n",
      "Epoch 1864/10000000\n",
      " - 0s - loss: 2.0821 - val_loss: 1.9003\n",
      "Epoch 1865/10000000\n",
      " - 0s - loss: 1.8011 - val_loss: 1.8800\n",
      "Epoch 1866/10000000\n",
      " - 0s - loss: 1.7221 - val_loss: 2.0531\n",
      "Epoch 1867/10000000\n",
      " - 0s - loss: 1.7893 - val_loss: 2.1796\n",
      "Epoch 1868/10000000\n",
      " - 0s - loss: 2.0074 - val_loss: 1.9827\n",
      "Epoch 1869/10000000\n",
      " - 0s - loss: 1.7743 - val_loss: 1.9103\n",
      "Epoch 1870/10000000\n",
      " - 0s - loss: 1.7927 - val_loss: 1.9643\n",
      "Epoch 1871/10000000\n",
      " - 0s - loss: 1.7441 - val_loss: 2.1353\n",
      "Epoch 1872/10000000\n",
      " - 0s - loss: 1.7594 - val_loss: 2.2324\n",
      "Epoch 1873/10000000\n",
      " - 0s - loss: 1.9706 - val_loss: 3.0631\n",
      "Epoch 1874/10000000\n",
      " - 0s - loss: 2.7748 - val_loss: 2.3867\n",
      "Epoch 1875/10000000\n",
      " - 0s - loss: 3.0503 - val_loss: 1.9034\n",
      "Epoch 1876/10000000\n",
      " - 0s - loss: 3.0940 - val_loss: 1.8320\n",
      "Epoch 1877/10000000\n",
      " - 0s - loss: 2.5306 - val_loss: 1.8440\n",
      "Epoch 1878/10000000\n",
      " - 0s - loss: 2.0988 - val_loss: 2.1503\n",
      "Epoch 1879/10000000\n",
      " - 0s - loss: 1.8402 - val_loss: 2.0433\n",
      "Epoch 1880/10000000\n",
      " - 0s - loss: 1.8191 - val_loss: 2.0265\n",
      "Epoch 1881/10000000\n",
      " - 0s - loss: 1.7926 - val_loss: 1.9539\n",
      "Epoch 1882/10000000\n",
      " - 0s - loss: 1.7233 - val_loss: 2.0365\n",
      "Epoch 1883/10000000\n",
      " - 0s - loss: 1.7631 - val_loss: 1.9038\n",
      "Epoch 1884/10000000\n",
      " - 0s - loss: 1.7334 - val_loss: 1.9557\n",
      "Epoch 1885/10000000\n",
      " - 0s - loss: 1.7755 - val_loss: 1.9105\n",
      "Epoch 1886/10000000\n",
      " - 0s - loss: 1.7084 - val_loss: 2.0024\n",
      "Epoch 1887/10000000\n",
      " - 0s - loss: 1.7612 - val_loss: 2.2258\n",
      "Epoch 1888/10000000\n",
      " - 0s - loss: 1.8676 - val_loss: 2.0108\n",
      "Epoch 1889/10000000\n",
      " - 0s - loss: 1.9732 - val_loss: 2.0128\n",
      "Epoch 1890/10000000\n",
      " - 0s - loss: 1.7573 - val_loss: 2.1375\n",
      "Epoch 1891/10000000\n",
      " - 0s - loss: 1.7243 - val_loss: 2.1738\n",
      "Epoch 1892/10000000\n",
      " - 0s - loss: 2.1326 - val_loss: 2.0649\n",
      "Epoch 1893/10000000\n",
      " - 0s - loss: 1.9741 - val_loss: 1.8376\n",
      "Epoch 1894/10000000\n",
      " - 0s - loss: 1.7098 - val_loss: 1.9094\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1895/10000000\n",
      " - 0s - loss: 1.8020 - val_loss: 2.0749\n",
      "Epoch 1896/10000000\n",
      " - 0s - loss: 1.8394 - val_loss: 1.9253\n",
      "Epoch 1897/10000000\n",
      " - 0s - loss: 1.7668 - val_loss: 2.0523\n",
      "Epoch 1898/10000000\n",
      " - 0s - loss: 1.8070 - val_loss: 1.9704\n",
      "Epoch 1899/10000000\n",
      " - 0s - loss: 1.9003 - val_loss: 2.0871\n",
      "Epoch 1900/10000000\n",
      " - 0s - loss: 1.8252 - val_loss: 2.0496\n",
      "Epoch 1901/10000000\n",
      " - 0s - loss: 1.7463 - val_loss: 2.0086\n",
      "Epoch 1902/10000000\n",
      " - 0s - loss: 1.7404 - val_loss: 2.2608\n",
      "Epoch 1903/10000000\n",
      " - 0s - loss: 1.9383 - val_loss: 1.9250\n",
      "Epoch 1904/10000000\n",
      " - 0s - loss: 1.7441 - val_loss: 2.0440\n",
      "Epoch 1905/10000000\n",
      " - 0s - loss: 1.7421 - val_loss: 2.3438\n",
      "Epoch 1906/10000000\n",
      " - 0s - loss: 1.9804 - val_loss: 2.4038\n",
      "Epoch 1907/10000000\n",
      " - 0s - loss: 2.0626 - val_loss: 2.0299\n",
      "Epoch 1908/10000000\n",
      " - 0s - loss: 1.7913 - val_loss: 2.0017\n",
      "Epoch 1909/10000000\n",
      " - 0s - loss: 1.9646 - val_loss: 1.9691\n",
      "Epoch 1910/10000000\n",
      " - 0s - loss: 1.8787 - val_loss: 2.0070\n",
      "Epoch 1911/10000000\n",
      " - 0s - loss: 1.7349 - val_loss: 2.0224\n",
      "Epoch 1912/10000000\n",
      " - 0s - loss: 1.8949 - val_loss: 1.9417\n",
      "Epoch 1913/10000000\n",
      " - 0s - loss: 1.7391 - val_loss: 1.9188\n",
      "Epoch 1914/10000000\n",
      " - 0s - loss: 1.7936 - val_loss: 1.8979\n",
      "Epoch 1915/10000000\n",
      " - 0s - loss: 1.7235 - val_loss: 1.9816\n",
      "Epoch 1916/10000000\n",
      " - 0s - loss: 1.7415 - val_loss: 2.0182\n",
      "Epoch 1917/10000000\n",
      " - 0s - loss: 1.7493 - val_loss: 1.9593\n",
      "Epoch 1918/10000000\n",
      " - 0s - loss: 1.7717 - val_loss: 1.9447\n",
      "Epoch 1919/10000000\n",
      " - 0s - loss: 1.8871 - val_loss: 1.9547\n",
      "Epoch 1920/10000000\n",
      " - 0s - loss: 1.7118 - val_loss: 1.9681\n",
      "Epoch 1921/10000000\n",
      " - 0s - loss: 1.6915 - val_loss: 2.0066\n",
      "Epoch 1922/10000000\n",
      " - 0s - loss: 1.6897 - val_loss: 2.0974\n",
      "Epoch 1923/10000000\n",
      " - 0s - loss: 1.7822 - val_loss: 1.9696\n",
      "Epoch 1924/10000000\n",
      " - 0s - loss: 1.7895 - val_loss: 2.0910\n",
      "Epoch 1925/10000000\n",
      " - 0s - loss: 1.9814 - val_loss: 1.9912\n",
      "Epoch 1926/10000000\n",
      " - 0s - loss: 1.8720 - val_loss: 1.9411\n",
      "Epoch 1927/10000000\n",
      " - 0s - loss: 1.6991 - val_loss: 1.9681\n",
      "Epoch 1928/10000000\n",
      " - 0s - loss: 1.7220 - val_loss: 1.9578\n",
      "Epoch 1929/10000000\n",
      " - 0s - loss: 1.9557 - val_loss: 2.2541\n",
      "Epoch 1930/10000000\n",
      " - 0s - loss: 2.3578 - val_loss: 2.0155\n",
      "Epoch 1931/10000000\n",
      " - 0s - loss: 2.6821 - val_loss: 1.9758\n",
      "Epoch 1932/10000000\n",
      " - 0s - loss: 2.6313 - val_loss: 2.1782\n",
      "Epoch 1933/10000000\n",
      " - 0s - loss: 2.5374 - val_loss: 1.9315\n",
      "Epoch 1934/10000000\n",
      " - 0s - loss: 1.8772 - val_loss: 2.0500\n",
      "Epoch 1935/10000000\n",
      " - 0s - loss: 1.8410 - val_loss: 2.0273\n",
      "Epoch 1936/10000000\n",
      " - 0s - loss: 1.8355 - val_loss: 2.1276\n",
      "Epoch 1937/10000000\n",
      " - 0s - loss: 1.9257 - val_loss: 2.0291\n",
      "Epoch 1938/10000000\n",
      " - 0s - loss: 2.1127 - val_loss: 2.1274\n",
      "Epoch 1939/10000000\n",
      " - 0s - loss: 1.8755 - val_loss: 2.0893\n",
      "Epoch 1940/10000000\n",
      " - 0s - loss: 1.9294 - val_loss: 2.1170\n",
      "Epoch 1941/10000000\n",
      " - 0s - loss: 1.7050 - val_loss: 2.2339\n",
      "Epoch 1942/10000000\n",
      " - 0s - loss: 1.8413 - val_loss: 2.3971\n",
      "Epoch 1943/10000000\n",
      " - 0s - loss: 1.9279 - val_loss: 2.6675\n",
      "Epoch 1944/10000000\n",
      " - 0s - loss: 2.7721 - val_loss: 2.3698\n",
      "Epoch 1945/10000000\n",
      " - 0s - loss: 2.2045 - val_loss: 1.9541\n",
      "Epoch 1946/10000000\n",
      " - 0s - loss: 1.6628 - val_loss: 1.9909\n",
      "Epoch 1947/10000000\n",
      " - 0s - loss: 1.6859 - val_loss: 2.0204\n",
      "Epoch 1948/10000000\n",
      " - 0s - loss: 1.6704 - val_loss: 2.0973\n",
      "Epoch 1949/10000000\n",
      " - 0s - loss: 1.6739 - val_loss: 2.0293\n",
      "Epoch 1950/10000000\n",
      " - 0s - loss: 1.6733 - val_loss: 1.9472\n",
      "Epoch 1951/10000000\n",
      " - 0s - loss: 1.7044 - val_loss: 2.0138\n",
      "Epoch 1952/10000000\n",
      " - 0s - loss: 1.7901 - val_loss: 1.9644\n",
      "Epoch 1953/10000000\n",
      " - 0s - loss: 1.7128 - val_loss: 1.9966\n",
      "Epoch 1954/10000000\n",
      " - 0s - loss: 1.8142 - val_loss: 1.9723\n",
      "Epoch 1955/10000000\n",
      " - 0s - loss: 1.9786 - val_loss: 2.0301\n",
      "Epoch 1956/10000000\n",
      " - 0s - loss: 2.2040 - val_loss: 2.1477\n",
      "Epoch 1957/10000000\n",
      " - 0s - loss: 2.0979 - val_loss: 2.0723\n",
      "Epoch 1958/10000000\n",
      " - 0s - loss: 1.9967 - val_loss: 2.1900\n",
      "Epoch 1959/10000000\n",
      " - 0s - loss: 1.8238 - val_loss: 2.5007\n",
      "Epoch 1960/10000000\n",
      " - 0s - loss: 2.2387 - val_loss: 2.1618\n",
      "Epoch 1961/10000000\n",
      " - 0s - loss: 1.9964 - val_loss: 2.0473\n",
      "Epoch 1962/10000000\n",
      " - 0s - loss: 1.7125 - val_loss: 2.4567\n",
      "Epoch 1963/10000000\n",
      " - 0s - loss: 1.9411 - val_loss: 2.3587\n",
      "Epoch 1964/10000000\n",
      " - 0s - loss: 1.8556 - val_loss: 2.0150\n",
      "Epoch 1965/10000000\n",
      " - 0s - loss: 1.6691 - val_loss: 2.2246\n",
      "Epoch 1966/10000000\n",
      " - 0s - loss: 1.8528 - val_loss: 1.9848\n",
      "Epoch 1967/10000000\n",
      " - 0s - loss: 1.6684 - val_loss: 2.1494\n",
      "Epoch 1968/10000000\n",
      " - 0s - loss: 1.7698 - val_loss: 2.3671\n",
      "Epoch 1969/10000000\n",
      " - 0s - loss: 2.0300 - val_loss: 1.9264\n",
      "Epoch 1970/10000000\n",
      " - 0s - loss: 1.7990 - val_loss: 1.9793\n",
      "Epoch 1971/10000000\n",
      " - 0s - loss: 1.9100 - val_loss: 1.9963\n",
      "Epoch 1972/10000000\n",
      " - 0s - loss: 1.9327 - val_loss: 1.9367\n",
      "Epoch 1973/10000000\n",
      " - 0s - loss: 1.6855 - val_loss: 1.9263\n",
      "Epoch 1974/10000000\n",
      " - 0s - loss: 1.5936 - val_loss: 2.1352\n",
      "Epoch 1975/10000000\n",
      " - 0s - loss: 1.7503 - val_loss: 2.1407\n",
      "Epoch 1976/10000000\n",
      " - 0s - loss: 1.8348 - val_loss: 2.1137\n",
      "Epoch 1977/10000000\n",
      " - 0s - loss: 1.7938 - val_loss: 2.0103\n",
      "Epoch 1978/10000000\n",
      " - 0s - loss: 1.6536 - val_loss: 1.9538\n",
      "Epoch 1979/10000000\n",
      " - 0s - loss: 1.6505 - val_loss: 2.0679\n",
      "Epoch 1980/10000000\n",
      " - 0s - loss: 1.9114 - val_loss: 2.0321\n",
      "Epoch 1981/10000000\n",
      " - 0s - loss: 1.6644 - val_loss: 2.0329\n",
      "Epoch 1982/10000000\n",
      " - 0s - loss: 1.7631 - val_loss: 1.9697\n",
      "Epoch 1983/10000000\n",
      " - 0s - loss: 1.6787 - val_loss: 1.8716\n",
      "Epoch 1984/10000000\n",
      " - 0s - loss: 1.6664 - val_loss: 1.8617\n",
      "Epoch 1985/10000000\n",
      " - 0s - loss: 1.6567 - val_loss: 1.9094\n",
      "Epoch 1986/10000000\n",
      " - 0s - loss: 1.7581 - val_loss: 1.9387\n",
      "Epoch 1987/10000000\n",
      " - 0s - loss: 1.6555 - val_loss: 1.9127\n",
      "Epoch 1988/10000000\n",
      " - 0s - loss: 1.6422 - val_loss: 1.9603\n",
      "Epoch 1989/10000000\n",
      " - 0s - loss: 1.6750 - val_loss: 2.0003\n",
      "Epoch 1990/10000000\n",
      " - 0s - loss: 1.6222 - val_loss: 2.0244\n",
      "Epoch 1991/10000000\n",
      " - 0s - loss: 1.6731 - val_loss: 2.0245\n",
      "Epoch 1992/10000000\n",
      " - 0s - loss: 1.7124 - val_loss: 1.9585\n",
      "Epoch 1993/10000000\n",
      " - 0s - loss: 1.7286 - val_loss: 1.9881\n",
      "Epoch 1994/10000000\n",
      " - 0s - loss: 1.8578 - val_loss: 2.2422\n",
      "Epoch 1995/10000000\n",
      " - 0s - loss: 1.9232 - val_loss: 2.1110\n",
      "Epoch 1996/10000000\n",
      " - 0s - loss: 2.3113 - val_loss: 2.4873\n",
      "Epoch 1997/10000000\n",
      " - 0s - loss: 2.6567 - val_loss: 2.2546\n",
      "Epoch 1998/10000000\n",
      " - 0s - loss: 2.6008 - val_loss: 1.9511\n",
      "Epoch 1999/10000000\n",
      " - 0s - loss: 1.9074 - val_loss: 1.8518\n",
      "Epoch 2000/10000000\n",
      " - 0s - loss: 1.7503 - val_loss: 1.8327\n",
      "Epoch 2001/10000000\n",
      " - 0s - loss: 1.9630 - val_loss: 1.9384\n",
      "Epoch 2002/10000000\n",
      " - 0s - loss: 1.8867 - val_loss: 2.1321\n",
      "Epoch 2003/10000000\n",
      " - 0s - loss: 1.8579 - val_loss: 2.2050\n",
      "Epoch 2004/10000000\n",
      " - 0s - loss: 1.9148 - val_loss: 2.1727\n",
      "Epoch 2005/10000000\n",
      " - 0s - loss: 2.2985 - val_loss: 2.0347\n",
      "Epoch 2006/10000000\n",
      " - 0s - loss: 1.8917 - val_loss: 2.0225\n",
      "Epoch 2007/10000000\n",
      " - 0s - loss: 1.8024 - val_loss: 2.1524\n",
      "Epoch 2008/10000000\n",
      " - 0s - loss: 1.9954 - val_loss: 1.8473\n",
      "Epoch 2009/10000000\n",
      " - 0s - loss: 1.6728 - val_loss: 2.0109\n",
      "Epoch 2010/10000000\n",
      " - 0s - loss: 1.7875 - val_loss: 1.9417\n",
      "Epoch 2011/10000000\n",
      " - 0s - loss: 1.8007 - val_loss: 2.0511\n",
      "Epoch 2012/10000000\n",
      " - 0s - loss: 1.8708 - val_loss: 2.3201\n",
      "Epoch 2013/10000000\n",
      " - 0s - loss: 1.8852 - val_loss: 2.3283\n",
      "Epoch 2014/10000000\n",
      " - 0s - loss: 1.8135 - val_loss: 2.0598\n",
      "Epoch 2015/10000000\n",
      " - 0s - loss: 1.6927 - val_loss: 2.0287\n",
      "Epoch 2016/10000000\n",
      " - 0s - loss: 1.8673 - val_loss: 1.9588\n",
      "Epoch 2017/10000000\n",
      " - 0s - loss: 1.6819 - val_loss: 1.9647\n",
      "Epoch 2018/10000000\n",
      " - 0s - loss: 1.7000 - val_loss: 1.8573\n",
      "Epoch 2019/10000000\n",
      " - 0s - loss: 1.6991 - val_loss: 1.8309\n",
      "Epoch 2020/10000000\n",
      " - 0s - loss: 1.7044 - val_loss: 1.8348\n",
      "Epoch 2021/10000000\n",
      " - 0s - loss: 1.6439 - val_loss: 1.9172\n",
      "Epoch 2022/10000000\n",
      " - 0s - loss: 1.8091 - val_loss: 2.0178\n",
      "Epoch 2023/10000000\n",
      " - 0s - loss: 1.6266 - val_loss: 2.2286\n",
      "Epoch 2024/10000000\n",
      " - 0s - loss: 1.8539 - val_loss: 2.3235\n",
      "Epoch 2025/10000000\n",
      " - 0s - loss: 2.0838 - val_loss: 2.0952\n",
      "Epoch 2026/10000000\n",
      " - 0s - loss: 1.9015 - val_loss: 2.2463\n",
      "Epoch 2027/10000000\n",
      " - 0s - loss: 2.0124 - val_loss: 2.0311\n",
      "Epoch 2028/10000000\n",
      " - 0s - loss: 1.8834 - val_loss: 1.9426\n",
      "Epoch 2029/10000000\n",
      " - 0s - loss: 1.6109 - val_loss: 1.8930\n",
      "Epoch 2030/10000000\n",
      " - 0s - loss: 1.7912 - val_loss: 1.9481\n",
      "Epoch 2031/10000000\n",
      " - 0s - loss: 1.7985 - val_loss: 2.0433\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2032/10000000\n",
      " - 0s - loss: 1.8413 - val_loss: 2.0180\n",
      "Epoch 2033/10000000\n",
      " - 0s - loss: 1.9487 - val_loss: 1.9795\n",
      "Epoch 2034/10000000\n",
      " - 0s - loss: 2.1496 - val_loss: 2.4359\n",
      "Epoch 2035/10000000\n",
      " - 0s - loss: 2.7646 - val_loss: 1.8527\n",
      "Epoch 2036/10000000\n",
      " - 0s - loss: 2.0840 - val_loss: 1.8994\n",
      "Epoch 2037/10000000\n",
      " - 0s - loss: 1.8379 - val_loss: 2.0031\n",
      "Epoch 2038/10000000\n",
      " - 0s - loss: 1.9602 - val_loss: 2.0976\n",
      "Epoch 2039/10000000\n",
      " - 0s - loss: 1.8996 - val_loss: 2.0604\n",
      "Epoch 2040/10000000\n",
      " - 0s - loss: 1.5723 - val_loss: 2.3546\n",
      "Epoch 2041/10000000\n",
      " - 0s - loss: 1.9528 - val_loss: 1.9736\n",
      "Epoch 2042/10000000\n",
      " - 0s - loss: 1.7779 - val_loss: 1.9938\n",
      "Epoch 2043/10000000\n",
      " - 0s - loss: 1.9458 - val_loss: 2.0935\n",
      "Epoch 2044/10000000\n",
      " - 0s - loss: 2.1444 - val_loss: 2.0993\n",
      "Epoch 2045/10000000\n",
      " - 0s - loss: 2.2990 - val_loss: 2.0528\n",
      "Epoch 2046/10000000\n",
      " - 0s - loss: 2.4036 - val_loss: 2.0376\n",
      "Epoch 2047/10000000\n",
      " - 0s - loss: 1.6687 - val_loss: 2.0632\n",
      "Epoch 2048/10000000\n",
      " - 0s - loss: 1.5771 - val_loss: 2.1138\n",
      "Epoch 2049/10000000\n",
      " - 0s - loss: 1.6386 - val_loss: 2.0923\n",
      "Epoch 2050/10000000\n",
      " - 0s - loss: 1.7044 - val_loss: 2.0137\n",
      "Epoch 2051/10000000\n",
      " - 0s - loss: 1.6029 - val_loss: 2.2085\n",
      "Epoch 2052/10000000\n",
      " - 0s - loss: 1.8079 - val_loss: 2.2761\n",
      "Epoch 2053/10000000\n",
      " - 0s - loss: 1.7525 - val_loss: 2.3227\n",
      "Epoch 2054/10000000\n",
      " - 0s - loss: 2.0954 - val_loss: 2.4881\n",
      "Epoch 2055/10000000\n",
      " - 0s - loss: 2.4012 - val_loss: 2.0592\n",
      "Epoch 2056/10000000\n",
      " - 0s - loss: 1.9841 - val_loss: 1.8852\n",
      "Epoch 2057/10000000\n",
      " - 0s - loss: 1.9529 - val_loss: 2.2708\n",
      "Epoch 2058/10000000\n",
      " - 0s - loss: 1.8997 - val_loss: 1.9215\n",
      "Epoch 2059/10000000\n",
      " - 0s - loss: 1.6992 - val_loss: 1.9045\n",
      "Epoch 2060/10000000\n",
      " - 0s - loss: 1.7679 - val_loss: 1.9718\n",
      "Epoch 2061/10000000\n",
      " - 0s - loss: 1.6374 - val_loss: 1.9653\n",
      "Epoch 2062/10000000\n",
      " - 0s - loss: 1.5752 - val_loss: 1.9460\n",
      "Epoch 2063/10000000\n",
      " - 0s - loss: 1.7224 - val_loss: 2.0857\n",
      "Epoch 2064/10000000\n",
      " - 0s - loss: 2.1723 - val_loss: 3.1073\n",
      "Epoch 2065/10000000\n",
      " - 0s - loss: 3.4545 - val_loss: 3.0479\n",
      "Epoch 2066/10000000\n",
      " - 0s - loss: 3.6521 - val_loss: 2.1030\n",
      "Epoch 2067/10000000\n",
      " - 0s - loss: 3.0198 - val_loss: 1.9368\n",
      "Epoch 2068/10000000\n",
      " - 0s - loss: 2.9358 - val_loss: 1.9756\n",
      "Epoch 2069/10000000\n",
      " - 0s - loss: 3.5526 - val_loss: 1.9143\n",
      "Epoch 2070/10000000\n",
      " - 0s - loss: 3.6920 - val_loss: 2.5361\n",
      "Epoch 2071/10000000\n",
      " - 0s - loss: 4.7037 - val_loss: 2.4114\n",
      "Epoch 2072/10000000\n",
      " - 0s - loss: 3.7331 - val_loss: 2.0824\n",
      "Epoch 2073/10000000\n",
      " - 0s - loss: 2.5883 - val_loss: 2.1929\n",
      "Epoch 2074/10000000\n",
      " - 0s - loss: 2.4469 - val_loss: 1.9448\n",
      "Epoch 2075/10000000\n",
      " - 0s - loss: 2.0105 - val_loss: 1.9991\n",
      "Epoch 2076/10000000\n",
      " - 0s - loss: 2.0675 - val_loss: 2.2103\n",
      "Epoch 2077/10000000\n",
      " - 0s - loss: 1.9056 - val_loss: 2.0406\n",
      "Epoch 2078/10000000\n",
      " - 0s - loss: 1.6903 - val_loss: 2.3961\n",
      "Epoch 2079/10000000\n",
      " - 0s - loss: 1.7943 - val_loss: 2.3079\n",
      "Epoch 2080/10000000\n",
      " - 0s - loss: 1.7937 - val_loss: 2.3146\n",
      "Epoch 2081/10000000\n",
      " - 0s - loss: 1.9928 - val_loss: 2.3425\n",
      "Epoch 2082/10000000\n",
      " - 0s - loss: 2.3050 - val_loss: 2.5564\n",
      "Epoch 2083/10000000\n",
      " - 0s - loss: 2.9430 - val_loss: 2.2250\n",
      "Epoch 2084/10000000\n",
      " - 0s - loss: 2.3286 - val_loss: 2.0106\n",
      "Epoch 2085/10000000\n",
      " - 0s - loss: 1.9624 - val_loss: 1.9674\n",
      "Epoch 2086/10000000\n",
      " - 0s - loss: 1.8893 - val_loss: 2.0824\n",
      "Epoch 2087/10000000\n",
      " - 0s - loss: 1.6899 - val_loss: 2.6104\n",
      "Epoch 2088/10000000\n",
      " - 0s - loss: 2.1132 - val_loss: 2.1619\n",
      "Epoch 2089/10000000\n",
      " - 0s - loss: 1.7201 - val_loss: 2.1547\n",
      "Epoch 2090/10000000\n",
      " - 0s - loss: 1.8504 - val_loss: 1.8499\n",
      "Epoch 2091/10000000\n",
      " - 0s - loss: 1.7152 - val_loss: 1.8412\n",
      "Epoch 2092/10000000\n",
      " - 0s - loss: 1.5450 - val_loss: 2.0240\n",
      "Epoch 2093/10000000\n",
      " - 0s - loss: 1.5990 - val_loss: 2.1146\n",
      "Epoch 2094/10000000\n",
      " - 0s - loss: 1.6380 - val_loss: 2.0257\n",
      "Epoch 2095/10000000\n",
      " - 0s - loss: 1.5956 - val_loss: 1.9518\n",
      "Epoch 2096/10000000\n",
      " - 0s - loss: 1.6184 - val_loss: 1.8726\n",
      "Epoch 2097/10000000\n",
      " - 0s - loss: 1.5523 - val_loss: 1.9081\n",
      "Epoch 2098/10000000\n",
      " - 0s - loss: 1.6239 - val_loss: 2.0518\n",
      "Epoch 2099/10000000\n",
      " - 0s - loss: 1.7524 - val_loss: 2.0478\n",
      "Epoch 2100/10000000\n",
      " - 0s - loss: 2.0981 - val_loss: 1.9283\n",
      "Epoch 2101/10000000\n",
      " - 0s - loss: 1.5857 - val_loss: 1.9444\n",
      "Epoch 2102/10000000\n",
      " - 0s - loss: 1.7910 - val_loss: 1.9612\n",
      "Epoch 2103/10000000\n",
      " - 0s - loss: 1.8348 - val_loss: 1.9964\n",
      "Epoch 2104/10000000\n",
      " - 0s - loss: 1.7966 - val_loss: 1.9594\n",
      "Epoch 2105/10000000\n",
      " - 0s - loss: 1.5655 - val_loss: 1.8634\n",
      "Epoch 2106/10000000\n",
      " - 0s - loss: 1.6129 - val_loss: 1.8435\n",
      "Epoch 2107/10000000\n",
      " - 0s - loss: 1.6498 - val_loss: 1.8885\n",
      "Epoch 2108/10000000\n",
      " - 0s - loss: 1.6318 - val_loss: 1.9514\n",
      "Epoch 2109/10000000\n",
      " - 0s - loss: 1.6855 - val_loss: 1.9752\n",
      "Epoch 2110/10000000\n",
      " - 0s - loss: 1.5497 - val_loss: 2.0493\n",
      "Epoch 2111/10000000\n",
      " - 0s - loss: 1.6955 - val_loss: 1.9653\n",
      "Epoch 2112/10000000\n",
      " - 0s - loss: 1.6197 - val_loss: 1.9541\n",
      "Epoch 2113/10000000\n",
      " - 0s - loss: 1.5583 - val_loss: 2.1046\n",
      "Epoch 2114/10000000\n",
      " - 0s - loss: 1.7534 - val_loss: 1.9596\n",
      "Epoch 2115/10000000\n",
      " - 0s - loss: 1.5421 - val_loss: 1.9469\n",
      "Epoch 2116/10000000\n",
      " - 0s - loss: 1.6111 - val_loss: 1.8181\n",
      "Epoch 2117/10000000\n",
      " - 0s - loss: 1.6464 - val_loss: 1.8864\n",
      "Epoch 2118/10000000\n",
      " - 0s - loss: 2.2700 - val_loss: 1.9237\n",
      "Epoch 2119/10000000\n",
      " - 0s - loss: 1.8495 - val_loss: 1.8072\n",
      "Epoch 2120/10000000\n",
      " - 0s - loss: 1.8209 - val_loss: 1.8243\n",
      "Epoch 2121/10000000\n",
      " - 0s - loss: 1.5045 - val_loss: 2.1676\n",
      "Epoch 2122/10000000\n",
      " - 0s - loss: 1.6522 - val_loss: 2.5743\n",
      "Epoch 2123/10000000\n",
      " - 0s - loss: 2.1428 - val_loss: 2.1433\n",
      "Epoch 2124/10000000\n",
      " - 0s - loss: 1.8458 - val_loss: 2.0397\n",
      "Epoch 2125/10000000\n",
      " - 0s - loss: 1.6696 - val_loss: 2.1882\n",
      "Epoch 2126/10000000\n",
      " - 0s - loss: 1.7802 - val_loss: 1.9953\n",
      "Epoch 2127/10000000\n",
      " - 0s - loss: 1.6385 - val_loss: 2.0156\n",
      "Epoch 2128/10000000\n",
      " - 0s - loss: 1.5678 - val_loss: 2.0991\n",
      "Epoch 2129/10000000\n",
      " - 0s - loss: 1.6103 - val_loss: 1.9442\n",
      "Epoch 2130/10000000\n",
      " - 0s - loss: 1.5317 - val_loss: 1.9115\n",
      "Epoch 2131/10000000\n",
      " - 0s - loss: 1.6997 - val_loss: 1.8276\n",
      "Epoch 2132/10000000\n",
      " - 0s - loss: 1.6156 - val_loss: 1.8127\n",
      "Epoch 2133/10000000\n",
      " - 0s - loss: 1.5582 - val_loss: 1.9329\n",
      "Epoch 2134/10000000\n",
      " - 0s - loss: 1.6435 - val_loss: 1.8267\n",
      "Epoch 2135/10000000\n",
      " - 0s - loss: 1.5638 - val_loss: 2.0098\n",
      "Epoch 2136/10000000\n",
      " - 0s - loss: 1.6012 - val_loss: 2.0434\n",
      "Epoch 2137/10000000\n",
      " - 0s - loss: 1.5836 - val_loss: 1.9944\n",
      "Epoch 2138/10000000\n",
      " - 0s - loss: 1.5764 - val_loss: 2.0724\n",
      "Epoch 2139/10000000\n",
      " - 0s - loss: 1.6990 - val_loss: 2.1490\n",
      "Epoch 2140/10000000\n",
      " - 0s - loss: 2.0202 - val_loss: 1.9197\n",
      "Epoch 2141/10000000\n",
      " - 0s - loss: 1.6483 - val_loss: 1.8704\n",
      "Epoch 2142/10000000\n",
      " - 0s - loss: 1.6252 - val_loss: 2.0117\n",
      "Epoch 2143/10000000\n",
      " - 0s - loss: 1.6723 - val_loss: 1.9350\n",
      "Epoch 2144/10000000\n",
      " - 0s - loss: 1.6269 - val_loss: 1.9746\n",
      "Epoch 2145/10000000\n",
      " - 0s - loss: 1.5092 - val_loss: 2.0817\n",
      "Epoch 2146/10000000\n",
      " - 0s - loss: 1.5845 - val_loss: 2.3986\n",
      "Epoch 2147/10000000\n",
      " - 0s - loss: 1.9515 - val_loss: 2.3925\n",
      "Epoch 2148/10000000\n",
      " - 0s - loss: 1.9621 - val_loss: 1.9169\n",
      "Epoch 2149/10000000\n",
      " - 0s - loss: 1.5261 - val_loss: 1.9007\n",
      "Epoch 2150/10000000\n",
      " - 0s - loss: 1.5774 - val_loss: 1.9090\n",
      "Epoch 2151/10000000\n",
      " - 0s - loss: 1.5960 - val_loss: 1.9595\n",
      "Epoch 2152/10000000\n",
      " - 0s - loss: 1.6199 - val_loss: 1.9879\n",
      "Epoch 2153/10000000\n",
      " - 0s - loss: 1.7028 - val_loss: 1.8680\n",
      "Epoch 2154/10000000\n",
      " - 0s - loss: 1.6225 - val_loss: 1.8304\n",
      "Epoch 2155/10000000\n",
      " - 0s - loss: 1.5626 - val_loss: 1.8289\n",
      "Epoch 2156/10000000\n",
      " - 0s - loss: 1.6400 - val_loss: 1.9244\n",
      "Epoch 2157/10000000\n",
      " - 0s - loss: 1.5138 - val_loss: 2.0287\n",
      "Epoch 2158/10000000\n",
      " - 0s - loss: 1.6629 - val_loss: 1.9842\n",
      "Epoch 2159/10000000\n",
      " - 0s - loss: 1.5439 - val_loss: 1.9725\n",
      "Epoch 2160/10000000\n",
      " - 0s - loss: 1.6236 - val_loss: 1.9729\n",
      "Epoch 2161/10000000\n",
      " - 0s - loss: 1.6967 - val_loss: 2.1258\n",
      "Epoch 2162/10000000\n",
      " - 0s - loss: 1.9111 - val_loss: 1.8956\n",
      "Epoch 2163/10000000\n",
      " - 0s - loss: 2.3231 - val_loss: 1.8118\n",
      "Epoch 2164/10000000\n",
      " - 0s - loss: 2.1830 - val_loss: 2.0437\n",
      "Epoch 2165/10000000\n",
      " - 0s - loss: 2.0106 - val_loss: 1.9194\n",
      "Epoch 2166/10000000\n",
      " - 0s - loss: 1.6911 - val_loss: 2.1782\n",
      "Epoch 2167/10000000\n",
      " - 0s - loss: 2.2168 - val_loss: 1.8939\n",
      "Epoch 2168/10000000\n",
      " - 0s - loss: 1.6701 - val_loss: 2.1100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2169/10000000\n",
      " - 0s - loss: 1.5959 - val_loss: 1.9557\n",
      "Epoch 2170/10000000\n",
      " - 0s - loss: 1.5803 - val_loss: 1.8228\n",
      "Epoch 2171/10000000\n",
      " - 0s - loss: 1.5584 - val_loss: 1.8490\n",
      "Epoch 2172/10000000\n",
      " - 0s - loss: 1.4882 - val_loss: 1.8465\n",
      "Epoch 2173/10000000\n",
      " - 0s - loss: 1.4999 - val_loss: 1.9331\n",
      "Epoch 2174/10000000\n",
      " - 0s - loss: 1.5526 - val_loss: 1.9489\n",
      "Epoch 2175/10000000\n",
      " - 0s - loss: 1.5897 - val_loss: 2.1656\n",
      "Epoch 2176/10000000\n",
      " - 0s - loss: 1.8106 - val_loss: 2.3629\n",
      "Epoch 2177/10000000\n",
      " - 0s - loss: 2.5310 - val_loss: 2.4947\n",
      "Epoch 2178/10000000\n",
      " - 0s - loss: 3.0093 - val_loss: 2.2232\n",
      "Epoch 2179/10000000\n",
      " - 0s - loss: 2.9030 - val_loss: 1.8902\n",
      "Epoch 2180/10000000\n",
      " - 0s - loss: 2.6398 - val_loss: 2.0279\n",
      "Epoch 2181/10000000\n",
      " - 0s - loss: 2.3641 - val_loss: 2.0027\n",
      "Epoch 2182/10000000\n",
      " - 0s - loss: 1.9548 - val_loss: 2.7580\n",
      "Epoch 2183/10000000\n",
      " - 0s - loss: 2.5836 - val_loss: 2.1155\n",
      "Epoch 2184/10000000\n",
      " - 0s - loss: 2.0125 - val_loss: 2.2717\n",
      "Epoch 2185/10000000\n",
      " - 0s - loss: 1.9922 - val_loss: 2.1781\n",
      "Epoch 2186/10000000\n",
      " - 0s - loss: 1.9151 - val_loss: 1.9686\n",
      "Epoch 2187/10000000\n",
      " - 0s - loss: 1.8667 - val_loss: 1.8671\n",
      "Epoch 2188/10000000\n",
      " - 0s - loss: 1.7885 - val_loss: 2.0367\n",
      "Epoch 2189/10000000\n",
      " - 0s - loss: 1.5891 - val_loss: 2.3529\n",
      "Epoch 2190/10000000\n",
      " - 0s - loss: 1.6014 - val_loss: 2.4722\n",
      "Epoch 2191/10000000\n",
      " - 0s - loss: 2.0210 - val_loss: 2.2587\n",
      "Epoch 2192/10000000\n",
      " - 0s - loss: 1.8627 - val_loss: 2.2802\n",
      "Epoch 2193/10000000\n",
      " - 0s - loss: 2.0455 - val_loss: 1.8586\n",
      "Epoch 2194/10000000\n",
      " - 0s - loss: 1.6984 - val_loss: 1.8000\n",
      "Epoch 2195/10000000\n",
      " - 0s - loss: 1.7735 - val_loss: 1.7700\n",
      "Epoch 2196/10000000\n",
      " - 0s - loss: 1.5413 - val_loss: 1.9269\n",
      "Epoch 2197/10000000\n",
      " - 0s - loss: 1.6001 - val_loss: 2.0739\n",
      "Epoch 2198/10000000\n",
      " - 0s - loss: 1.5927 - val_loss: 1.9923\n",
      "Epoch 2199/10000000\n",
      " - 0s - loss: 1.5681 - val_loss: 1.8409\n",
      "Epoch 2200/10000000\n",
      " - 0s - loss: 1.6356 - val_loss: 1.8943\n",
      "Epoch 2201/10000000\n",
      " - 0s - loss: 1.7819 - val_loss: 2.0394\n",
      "Epoch 2202/10000000\n",
      " - 0s - loss: 2.1849 - val_loss: 2.3701\n",
      "Epoch 2203/10000000\n",
      " - 0s - loss: 2.7662 - val_loss: 2.0421\n",
      "Epoch 2204/10000000\n",
      " - 0s - loss: 2.5499 - val_loss: 1.8560\n",
      "Epoch 2205/10000000\n",
      " - 0s - loss: 1.9565 - val_loss: 2.1115\n",
      "Epoch 2206/10000000\n",
      " - 0s - loss: 1.5523 - val_loss: 2.8357\n",
      "Epoch 2207/10000000\n",
      " - 0s - loss: 1.9360 - val_loss: 3.1077\n",
      "Epoch 2208/10000000\n",
      " - 0s - loss: 2.6541 - val_loss: 3.2881\n",
      "Epoch 2209/10000000\n",
      " - 0s - loss: 2.5477 - val_loss: 2.5244\n",
      "Epoch 2210/10000000\n",
      " - 0s - loss: 2.2266 - val_loss: 2.0460\n",
      "Epoch 2211/10000000\n",
      " - 0s - loss: 1.7698 - val_loss: 1.9973\n",
      "Epoch 2212/10000000\n",
      " - 0s - loss: 1.8874 - val_loss: 1.8979\n",
      "Epoch 2213/10000000\n",
      " - 0s - loss: 1.7298 - val_loss: 1.8168\n",
      "Epoch 2214/10000000\n",
      " - 0s - loss: 1.5706 - val_loss: 1.8767\n",
      "Epoch 2215/10000000\n",
      " - 0s - loss: 1.5384 - val_loss: 1.9850\n",
      "Epoch 2216/10000000\n",
      " - 0s - loss: 1.5398 - val_loss: 1.9789\n",
      "Epoch 2217/10000000\n",
      " - 0s - loss: 1.5882 - val_loss: 1.8624\n",
      "Epoch 2218/10000000\n",
      " - 0s - loss: 1.4406 - val_loss: 2.0919\n",
      "Epoch 2219/10000000\n",
      " - 0s - loss: 1.6259 - val_loss: 2.0702\n",
      "Epoch 2220/10000000\n",
      " - 0s - loss: 1.6937 - val_loss: 1.9448\n",
      "Epoch 2221/10000000\n",
      " - 0s - loss: 1.6601 - val_loss: 2.0953\n",
      "Epoch 2222/10000000\n",
      " - 0s - loss: 1.8949 - val_loss: 1.8264\n",
      "Epoch 2223/10000000\n",
      " - 0s - loss: 1.4666 - val_loss: 2.0651\n",
      "Epoch 2224/10000000\n",
      " - 0s - loss: 1.7400 - val_loss: 2.1130\n",
      "Epoch 2225/10000000\n",
      " - 0s - loss: 1.8741 - val_loss: 1.9976\n",
      "Epoch 2226/10000000\n",
      " - 0s - loss: 1.5914 - val_loss: 1.8860\n",
      "Epoch 2227/10000000\n",
      " - 0s - loss: 1.5292 - val_loss: 1.9797\n",
      "Epoch 2228/10000000\n",
      " - 0s - loss: 1.8279 - val_loss: 2.0671\n",
      "Epoch 2229/10000000\n",
      " - 0s - loss: 1.7685 - val_loss: 2.0916\n",
      "Epoch 2230/10000000\n",
      " - 0s - loss: 1.7830 - val_loss: 1.9841\n",
      "Epoch 2231/10000000\n",
      " - 0s - loss: 1.7057 - val_loss: 1.9841\n",
      "Epoch 2232/10000000\n",
      " - 0s - loss: 1.6662 - val_loss: 1.8762\n",
      "Epoch 2233/10000000\n",
      " - 0s - loss: 1.4552 - val_loss: 2.0337\n",
      "Epoch 2234/10000000\n",
      " - 0s - loss: 1.5738 - val_loss: 1.9910\n",
      "Epoch 2235/10000000\n",
      " - 0s - loss: 1.5774 - val_loss: 1.7970\n",
      "Epoch 2236/10000000\n",
      " - 0s - loss: 1.4887 - val_loss: 1.8010\n",
      "Epoch 2237/10000000\n",
      " - 0s - loss: 1.5102 - val_loss: 1.8023\n",
      "Epoch 2238/10000000\n",
      " - 0s - loss: 1.6388 - val_loss: 1.8218\n",
      "Epoch 2239/10000000\n",
      " - 0s - loss: 1.4943 - val_loss: 1.8698\n",
      "Epoch 2240/10000000\n",
      " - 0s - loss: 1.4671 - val_loss: 1.9241\n",
      "Epoch 2241/10000000\n",
      " - 0s - loss: 1.5128 - val_loss: 1.9330\n",
      "Epoch 2242/10000000\n",
      " - 0s - loss: 1.4690 - val_loss: 1.9515\n",
      "Epoch 2243/10000000\n",
      " - 0s - loss: 1.4776 - val_loss: 1.8765\n",
      "Epoch 2244/10000000\n",
      " - 0s - loss: 1.4609 - val_loss: 1.8496\n",
      "Epoch 2245/10000000\n",
      " - 0s - loss: 1.5470 - val_loss: 1.7488\n",
      "Epoch 2246/10000000\n",
      " - 0s - loss: 1.4803 - val_loss: 1.7868\n",
      "Epoch 2247/10000000\n",
      " - 0s - loss: 1.4792 - val_loss: 1.9247\n",
      "Epoch 2248/10000000\n",
      " - 0s - loss: 1.5284 - val_loss: 2.0630\n",
      "Epoch 2249/10000000\n",
      " - 0s - loss: 1.6618 - val_loss: 2.3764\n",
      "Epoch 2250/10000000\n",
      " - 0s - loss: 2.1382 - val_loss: 2.1211\n",
      "Epoch 2251/10000000\n",
      " - 0s - loss: 2.0913 - val_loss: 1.8981\n",
      "Epoch 2252/10000000\n",
      " - 0s - loss: 1.7255 - val_loss: 1.8000\n",
      "Epoch 2253/10000000\n",
      " - 0s - loss: 1.4875 - val_loss: 1.7880\n",
      "Epoch 2254/10000000\n",
      " - 0s - loss: 1.4722 - val_loss: 2.1386\n",
      "Epoch 2255/10000000\n",
      " - 0s - loss: 1.8635 - val_loss: 1.9657\n",
      "Epoch 2256/10000000\n",
      " - 0s - loss: 1.7039 - val_loss: 2.1992\n",
      "Epoch 2257/10000000\n",
      " - 0s - loss: 1.9566 - val_loss: 2.6463\n",
      "Epoch 2258/10000000\n",
      " - 0s - loss: 2.4188 - val_loss: 2.0308\n",
      "Epoch 2259/10000000\n",
      " - 0s - loss: 1.6861 - val_loss: 1.8924\n",
      "Epoch 2260/10000000\n",
      " - 0s - loss: 1.6764 - val_loss: 2.0053\n",
      "Epoch 2261/10000000\n",
      " - 0s - loss: 1.6251 - val_loss: 1.8682\n",
      "Epoch 2262/10000000\n",
      " - 0s - loss: 1.4887 - val_loss: 1.8500\n",
      "Epoch 2263/10000000\n",
      " - 0s - loss: 1.7595 - val_loss: 1.9842\n",
      "Epoch 2264/10000000\n",
      " - 0s - loss: 1.6211 - val_loss: 1.8496\n",
      "Epoch 2265/10000000\n",
      " - 0s - loss: 1.4889 - val_loss: 2.0464\n",
      "Epoch 2266/10000000\n",
      " - 0s - loss: 1.6501 - val_loss: 1.8115\n",
      "Epoch 2267/10000000\n",
      " - 0s - loss: 1.5253 - val_loss: 1.8824\n",
      "Epoch 2268/10000000\n",
      " - 0s - loss: 1.4508 - val_loss: 1.8504\n",
      "Epoch 2269/10000000\n",
      " - 0s - loss: 1.4674 - val_loss: 1.8109\n",
      "Epoch 2270/10000000\n",
      " - 0s - loss: 1.4478 - val_loss: 1.8136\n",
      "Epoch 2271/10000000\n",
      " - 0s - loss: 1.4705 - val_loss: 1.8575\n",
      "Epoch 2272/10000000\n",
      " - 0s - loss: 1.4283 - val_loss: 1.9671\n",
      "Epoch 2273/10000000\n",
      " - 0s - loss: 1.6360 - val_loss: 2.1548\n",
      "Epoch 2274/10000000\n",
      " - 0s - loss: 1.7937 - val_loss: 2.1942\n",
      "Epoch 2275/10000000\n",
      " - 0s - loss: 1.8655 - val_loss: 2.3416\n",
      "Epoch 2276/10000000\n",
      " - 0s - loss: 2.0046 - val_loss: 1.8835\n",
      "Epoch 2277/10000000\n",
      " - 0s - loss: 1.6092 - val_loss: 1.8813\n",
      "Epoch 2278/10000000\n",
      " - 0s - loss: 1.7237 - val_loss: 1.8213\n",
      "Epoch 2279/10000000\n",
      " - 0s - loss: 1.4956 - val_loss: 1.8519\n",
      "Epoch 2280/10000000\n",
      " - 0s - loss: 1.5566 - val_loss: 1.8205\n",
      "Epoch 2281/10000000\n",
      " - 0s - loss: 1.6816 - val_loss: 1.7978\n",
      "Epoch 2282/10000000\n",
      " - 0s - loss: 1.6299 - val_loss: 1.8356\n",
      "Epoch 2283/10000000\n",
      " - 0s - loss: 1.7894 - val_loss: 1.8037\n",
      "Epoch 2284/10000000\n",
      " - 0s - loss: 1.9351 - val_loss: 2.0255\n",
      "Epoch 2285/10000000\n",
      " - 0s - loss: 2.5130 - val_loss: 2.0154\n",
      "Epoch 2286/10000000\n",
      " - 0s - loss: 2.5269 - val_loss: 2.0226\n",
      "Epoch 2287/10000000\n",
      " - 0s - loss: 2.6041 - val_loss: 1.9562\n",
      "Epoch 2288/10000000\n",
      " - 0s - loss: 1.7216 - val_loss: 2.0351\n",
      "Epoch 2289/10000000\n",
      " - 0s - loss: 1.6088 - val_loss: 2.1914\n",
      "Epoch 2290/10000000\n",
      " - 0s - loss: 1.7362 - val_loss: 2.3122\n",
      "Epoch 2291/10000000\n",
      " - 0s - loss: 1.6942 - val_loss: 2.2109\n",
      "Epoch 2292/10000000\n",
      " - 0s - loss: 1.5948 - val_loss: 2.0265\n",
      "Epoch 2293/10000000\n",
      " - 0s - loss: 1.5761 - val_loss: 2.0706\n",
      "Epoch 2294/10000000\n",
      " - 0s - loss: 1.6340 - val_loss: 1.9466\n",
      "Epoch 2295/10000000\n",
      " - 0s - loss: 1.6149 - val_loss: 1.9046\n",
      "Epoch 2296/10000000\n",
      " - 0s - loss: 1.7512 - val_loss: 2.0209\n",
      "Epoch 2297/10000000\n",
      " - 0s - loss: 1.9690 - val_loss: 1.9467\n",
      "Epoch 2298/10000000\n",
      " - 0s - loss: 1.5973 - val_loss: 1.8592\n",
      "Epoch 2299/10000000\n",
      " - 0s - loss: 1.5325 - val_loss: 1.9248\n",
      "Epoch 2300/10000000\n",
      " - 0s - loss: 1.5469 - val_loss: 1.9214\n",
      "Epoch 2301/10000000\n",
      " - 0s - loss: 1.4397 - val_loss: 1.8844\n",
      "Epoch 2302/10000000\n",
      " - 0s - loss: 1.4218 - val_loss: 1.8811\n",
      "Epoch 2303/10000000\n",
      " - 0s - loss: 1.4307 - val_loss: 2.0087\n",
      "Epoch 2304/10000000\n",
      " - 0s - loss: 1.5516 - val_loss: 1.9122\n",
      "Epoch 2305/10000000\n",
      " - 0s - loss: 1.4520 - val_loss: 1.8362\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2306/10000000\n",
      " - 0s - loss: 1.4519 - val_loss: 1.8495\n",
      "Epoch 2307/10000000\n",
      " - 0s - loss: 1.5024 - val_loss: 1.8539\n",
      "Epoch 2308/10000000\n",
      " - 0s - loss: 1.5074 - val_loss: 2.0750\n",
      "Epoch 2309/10000000\n",
      " - 0s - loss: 1.6212 - val_loss: 1.9061\n",
      "Epoch 2310/10000000\n",
      " - 0s - loss: 1.4630 - val_loss: 1.8400\n",
      "Epoch 2311/10000000\n",
      " - 0s - loss: 1.6500 - val_loss: 1.7375\n",
      "Epoch 2312/10000000\n",
      " - 0s - loss: 1.4591 - val_loss: 1.7601\n",
      "Epoch 2313/10000000\n",
      " - 0s - loss: 1.4243 - val_loss: 1.8598\n",
      "Epoch 2314/10000000\n",
      " - 0s - loss: 1.4717 - val_loss: 1.8333\n",
      "Epoch 2315/10000000\n",
      " - 0s - loss: 1.4970 - val_loss: 1.9610\n",
      "Epoch 2316/10000000\n",
      " - 0s - loss: 1.6540 - val_loss: 1.8641\n",
      "Epoch 2317/10000000\n",
      " - 0s - loss: 1.5338 - val_loss: 1.9105\n",
      "Epoch 2318/10000000\n",
      " - 0s - loss: 1.6340 - val_loss: 1.9012\n",
      "Epoch 2319/10000000\n",
      " - 0s - loss: 1.5693 - val_loss: 1.8313\n",
      "Epoch 2320/10000000\n",
      " - 0s - loss: 1.4394 - val_loss: 1.7744\n",
      "Epoch 2321/10000000\n",
      " - 0s - loss: 1.4168 - val_loss: 2.0637\n",
      "Epoch 2322/10000000\n",
      " - 0s - loss: 1.7231 - val_loss: 1.7611\n",
      "Epoch 2323/10000000\n",
      " - 0s - loss: 1.4126 - val_loss: 1.7754\n",
      "Epoch 2324/10000000\n",
      " - 0s - loss: 1.4430 - val_loss: 1.8033\n",
      "Epoch 2325/10000000\n",
      " - 0s - loss: 1.4607 - val_loss: 1.7838\n",
      "Epoch 2326/10000000\n",
      " - 0s - loss: 1.4144 - val_loss: 1.8213\n",
      "Epoch 2327/10000000\n",
      " - 0s - loss: 1.4298 - val_loss: 1.9706\n",
      "Epoch 2328/10000000\n",
      " - 0s - loss: 1.6056 - val_loss: 1.8413\n",
      "Epoch 2329/10000000\n",
      " - 0s - loss: 1.4619 - val_loss: 1.7924\n",
      "Epoch 2330/10000000\n",
      " - 0s - loss: 1.5440 - val_loss: 1.9043\n",
      "Epoch 2331/10000000\n",
      " - 0s - loss: 1.5638 - val_loss: 1.8641\n",
      "Epoch 2332/10000000\n",
      " - 0s - loss: 1.4881 - val_loss: 1.9827\n",
      "Epoch 2333/10000000\n",
      " - 0s - loss: 1.5776 - val_loss: 1.8680\n",
      "Epoch 2334/10000000\n",
      " - 0s - loss: 1.4434 - val_loss: 1.9435\n",
      "Epoch 2335/10000000\n",
      " - 0s - loss: 1.4472 - val_loss: 1.9132\n",
      "Epoch 2336/10000000\n",
      " - 0s - loss: 1.5847 - val_loss: 2.3259\n",
      "Epoch 2337/10000000\n",
      " - 0s - loss: 1.8998 - val_loss: 2.3416\n",
      "Epoch 2338/10000000\n",
      " - 0s - loss: 2.0290 - val_loss: 1.9354\n",
      "Epoch 2339/10000000\n",
      " - 0s - loss: 1.6327 - val_loss: 1.8282\n",
      "Epoch 2340/10000000\n",
      " - 0s - loss: 1.5075 - val_loss: 1.8798\n",
      "Epoch 2341/10000000\n",
      " - 0s - loss: 1.4673 - val_loss: 1.8692\n",
      "Epoch 2342/10000000\n",
      " - 0s - loss: 1.4195 - val_loss: 1.8691\n",
      "Epoch 2343/10000000\n",
      " - 0s - loss: 1.4418 - val_loss: 1.8553\n",
      "Epoch 2344/10000000\n",
      " - 0s - loss: 1.5117 - val_loss: 1.8780\n",
      "Epoch 2345/10000000\n",
      " - 0s - loss: 1.4803 - val_loss: 1.8551\n",
      "Epoch 2346/10000000\n",
      " - 0s - loss: 1.4912 - val_loss: 1.9400\n",
      "Epoch 2347/10000000\n",
      " - 0s - loss: 1.6065 - val_loss: 1.7687\n",
      "Epoch 2348/10000000\n",
      " - 0s - loss: 1.4891 - val_loss: 1.7935\n",
      "Epoch 2349/10000000\n",
      " - 0s - loss: 1.4191 - val_loss: 1.8752\n",
      "Epoch 2350/10000000\n",
      " - 0s - loss: 1.4183 - val_loss: 1.9441\n",
      "Epoch 2351/10000000\n",
      " - 0s - loss: 1.5131 - val_loss: 1.9488\n",
      "Epoch 2352/10000000\n",
      " - 0s - loss: 1.5909 - val_loss: 2.1270\n",
      "Epoch 2353/10000000\n",
      " - 0s - loss: 1.9173 - val_loss: 2.6034\n",
      "Epoch 2354/10000000\n",
      " - 0s - loss: 2.7293 - val_loss: 3.1345\n",
      "Epoch 2355/10000000\n",
      " - 0s - loss: 3.1165 - val_loss: 2.6937\n",
      "Epoch 2356/10000000\n",
      " - 0s - loss: 2.9370 - val_loss: 3.6321\n",
      "Epoch 2357/10000000\n",
      " - 0s - loss: 3.9215 - val_loss: 2.9088\n",
      "Epoch 2358/10000000\n",
      " - 0s - loss: 4.2239 - val_loss: 2.4535\n",
      "Epoch 2359/10000000\n",
      " - 0s - loss: 2.3861 - val_loss: 2.0131\n",
      "Epoch 2360/10000000\n",
      " - 0s - loss: 2.6017 - val_loss: 1.8863\n",
      "Epoch 2361/10000000\n",
      " - 0s - loss: 1.8183 - val_loss: 1.8338\n",
      "Epoch 2362/10000000\n",
      " - 0s - loss: 1.5733 - val_loss: 1.7318\n",
      "Epoch 2363/10000000\n",
      " - 0s - loss: 1.7996 - val_loss: 2.1814\n",
      "Epoch 2364/10000000\n",
      " - 0s - loss: 2.2541 - val_loss: 2.2459\n",
      "Epoch 2365/10000000\n",
      " - 0s - loss: 3.2636 - val_loss: 1.9399\n",
      "Epoch 2366/10000000\n",
      " - 0s - loss: 2.0297 - val_loss: 1.7940\n",
      "Epoch 2367/10000000\n",
      " - 0s - loss: 2.0619 - val_loss: 1.7898\n",
      "Epoch 2368/10000000\n",
      " - 0s - loss: 2.0737 - val_loss: 1.8483\n",
      "Epoch 2369/10000000\n",
      " - 0s - loss: 2.1493 - val_loss: 1.9453\n",
      "Epoch 2370/10000000\n",
      " - 0s - loss: 1.7397 - val_loss: 2.2754\n",
      "Epoch 2371/10000000\n",
      " - 0s - loss: 2.1714 - val_loss: 1.9489\n",
      "Epoch 2372/10000000\n",
      " - 0s - loss: 2.3268 - val_loss: 1.9214\n",
      "Epoch 2373/10000000\n",
      " - 0s - loss: 1.6656 - val_loss: 2.1026\n",
      "Epoch 2374/10000000\n",
      " - 0s - loss: 1.6155 - val_loss: 1.9373\n",
      "Epoch 2375/10000000\n",
      " - 0s - loss: 1.7735 - val_loss: 1.9843\n",
      "Epoch 2376/10000000\n",
      " - 0s - loss: 1.4365 - val_loss: 2.2442\n",
      "Epoch 2377/10000000\n",
      " - 0s - loss: 1.7194 - val_loss: 2.1103\n",
      "Epoch 2378/10000000\n",
      " - 0s - loss: 1.5241 - val_loss: 1.9611\n",
      "Epoch 2379/10000000\n",
      " - 0s - loss: 1.5526 - val_loss: 1.7708\n",
      "Epoch 2380/10000000\n",
      " - 0s - loss: 1.4453 - val_loss: 1.8061\n",
      "Epoch 2381/10000000\n",
      " - 0s - loss: 1.4865 - val_loss: 1.8599\n",
      "Epoch 2382/10000000\n",
      " - 0s - loss: 1.7081 - val_loss: 1.7359\n",
      "Epoch 2383/10000000\n",
      " - 0s - loss: 1.6672 - val_loss: 1.8695\n",
      "Epoch 2384/10000000\n",
      " - 0s - loss: 1.5675 - val_loss: 1.7863\n",
      "Epoch 2385/10000000\n",
      " - 0s - loss: 1.5652 - val_loss: 1.8767\n",
      "Epoch 2386/10000000\n",
      " - 0s - loss: 1.5320 - val_loss: 1.9765\n",
      "Epoch 2387/10000000\n",
      " - 0s - loss: 2.0927 - val_loss: 2.2817\n",
      "Epoch 2388/10000000\n",
      " - 0s - loss: 2.2986 - val_loss: 1.9039\n",
      "Epoch 2389/10000000\n",
      " - 0s - loss: 1.8892 - val_loss: 1.7400\n",
      "Epoch 2390/10000000\n",
      " - 0s - loss: 1.5425 - val_loss: 1.9914\n",
      "Epoch 2391/10000000\n",
      " - 0s - loss: 1.4280 - val_loss: 2.4569\n",
      "Epoch 2392/10000000\n",
      " - 0s - loss: 1.7286 - val_loss: 2.5570\n",
      "Epoch 2393/10000000\n",
      " - 0s - loss: 1.9596 - val_loss: 2.3498\n",
      "Epoch 2394/10000000\n",
      " - 0s - loss: 1.8993 - val_loss: 2.1255\n",
      "Epoch 2395/10000000\n",
      " - 0s - loss: 1.7996 - val_loss: 1.7244\n",
      "Epoch 2396/10000000\n",
      " - 0s - loss: 1.4148 - val_loss: 1.8416\n",
      "Epoch 2397/10000000\n",
      " - 0s - loss: 1.4246 - val_loss: 1.8485\n",
      "Epoch 2398/10000000\n",
      " - 0s - loss: 1.4726 - val_loss: 1.7822\n",
      "Epoch 2399/10000000\n",
      " - 0s - loss: 1.3892 - val_loss: 1.7748\n",
      "Epoch 2400/10000000\n",
      " - 0s - loss: 1.4213 - val_loss: 1.8217\n",
      "Epoch 2401/10000000\n",
      " - 0s - loss: 1.4004 - val_loss: 1.7587\n",
      "Epoch 2402/10000000\n",
      " - 0s - loss: 1.3795 - val_loss: 1.7344\n",
      "Epoch 2403/10000000\n",
      " - 0s - loss: 1.5304 - val_loss: 1.9026\n",
      "Epoch 2404/10000000\n",
      " - 0s - loss: 1.6726 - val_loss: 1.9218\n",
      "Epoch 2405/10000000\n",
      " - 0s - loss: 1.7284 - val_loss: 1.8071\n",
      "Epoch 2406/10000000\n",
      " - 0s - loss: 1.4765 - val_loss: 1.8655\n",
      "Epoch 2407/10000000\n",
      " - 0s - loss: 1.3756 - val_loss: 1.8816\n",
      "Epoch 2408/10000000\n",
      " - 0s - loss: 1.4494 - val_loss: 2.1726\n",
      "Epoch 2409/10000000\n",
      " - 0s - loss: 1.9215 - val_loss: 2.1639\n",
      "Epoch 2410/10000000\n",
      " - 0s - loss: 1.7533 - val_loss: 1.8546\n",
      "Epoch 2411/10000000\n",
      " - 0s - loss: 1.3937 - val_loss: 1.7707\n",
      "Epoch 2412/10000000\n",
      " - 0s - loss: 1.4330 - val_loss: 1.7271\n",
      "Epoch 2413/10000000\n",
      " - 0s - loss: 1.3831 - val_loss: 1.7476\n",
      "Epoch 2414/10000000\n",
      " - 0s - loss: 1.3973 - val_loss: 1.7676\n",
      "Epoch 2415/10000000\n",
      " - 0s - loss: 1.4176 - val_loss: 1.8233\n",
      "Epoch 2416/10000000\n",
      " - 0s - loss: 1.4498 - val_loss: 1.8232\n",
      "Epoch 2417/10000000\n",
      " - 0s - loss: 1.4838 - val_loss: 1.7307\n",
      "Epoch 2418/10000000\n",
      " - 0s - loss: 1.3752 - val_loss: 1.7304\n",
      "Epoch 2419/10000000\n",
      " - 0s - loss: 1.4431 - val_loss: 1.7931\n",
      "Epoch 2420/10000000\n",
      " - 0s - loss: 1.3678 - val_loss: 1.9314\n",
      "Epoch 2421/10000000\n",
      " - 0s - loss: 1.4646 - val_loss: 1.8649\n",
      "Epoch 2422/10000000\n",
      " - 0s - loss: 1.3951 - val_loss: 1.8257\n",
      "Epoch 2423/10000000\n",
      " - 0s - loss: 1.3731 - val_loss: 1.8057\n",
      "Epoch 2424/10000000\n",
      " - 0s - loss: 1.3736 - val_loss: 1.7671\n",
      "Epoch 2425/10000000\n",
      " - 0s - loss: 1.3555 - val_loss: 1.7955\n",
      "Epoch 2426/10000000\n",
      " - 0s - loss: 1.3787 - val_loss: 1.7753\n",
      "Epoch 2427/10000000\n",
      " - 0s - loss: 1.4772 - val_loss: 1.8263\n",
      "Epoch 2428/10000000\n",
      " - 0s - loss: 1.5085 - val_loss: 1.8094\n",
      "Epoch 2429/10000000\n",
      " - 0s - loss: 1.5784 - val_loss: 1.7254\n",
      "Epoch 2430/10000000\n",
      " - 0s - loss: 1.4223 - val_loss: 1.7807\n",
      "Epoch 2431/10000000\n",
      " - 0s - loss: 1.4331 - val_loss: 1.7838\n",
      "Epoch 2432/10000000\n",
      " - 0s - loss: 1.3680 - val_loss: 1.8745\n",
      "Epoch 2433/10000000\n",
      " - 0s - loss: 1.4125 - val_loss: 2.0535\n",
      "Epoch 2434/10000000\n",
      " - 0s - loss: 1.5946 - val_loss: 1.9600\n",
      "Epoch 2435/10000000\n",
      " - 0s - loss: 1.6187 - val_loss: 1.7933\n",
      "Epoch 2436/10000000\n",
      " - 0s - loss: 1.4279 - val_loss: 1.8509\n",
      "Epoch 2437/10000000\n",
      " - 0s - loss: 1.7801 - val_loss: 2.0439\n",
      "Epoch 2438/10000000\n",
      " - 0s - loss: 2.2337 - val_loss: 1.8318\n",
      "Epoch 2439/10000000\n",
      " - 0s - loss: 1.7305 - val_loss: 2.0842\n",
      "Epoch 2440/10000000\n",
      " - 0s - loss: 2.3566 - val_loss: 1.7934\n",
      "Epoch 2441/10000000\n",
      " - 0s - loss: 1.8210 - val_loss: 1.7915\n",
      "Epoch 2442/10000000\n",
      " - 0s - loss: 1.5802 - val_loss: 1.7791\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2443/10000000\n",
      " - 0s - loss: 1.4278 - val_loss: 1.7996\n",
      "Epoch 2444/10000000\n",
      " - 0s - loss: 1.5404 - val_loss: 1.9461\n",
      "Epoch 2445/10000000\n",
      " - 0s - loss: 2.0245 - val_loss: 1.7649\n",
      "Epoch 2446/10000000\n",
      " - 0s - loss: 1.3574 - val_loss: 2.1296\n",
      "Epoch 2447/10000000\n",
      " - 0s - loss: 1.6287 - val_loss: 2.1295\n",
      "Epoch 2448/10000000\n",
      " - 0s - loss: 1.6135 - val_loss: 1.9965\n",
      "Epoch 2449/10000000\n",
      " - 0s - loss: 1.5548 - val_loss: 2.1875\n",
      "Epoch 2450/10000000\n",
      " - 0s - loss: 1.8868 - val_loss: 2.0116\n",
      "Epoch 2451/10000000\n",
      " - 0s - loss: 1.7383 - val_loss: 1.7794\n",
      "Epoch 2452/10000000\n",
      " - 0s - loss: 1.4809 - val_loss: 1.7529\n",
      "Epoch 2453/10000000\n",
      " - 0s - loss: 1.3916 - val_loss: 1.7586\n",
      "Epoch 2454/10000000\n",
      " - 0s - loss: 1.3331 - val_loss: 1.9621\n",
      "Epoch 2455/10000000\n",
      " - 0s - loss: 1.4941 - val_loss: 2.1190\n",
      "Epoch 2456/10000000\n",
      " - 0s - loss: 1.8805 - val_loss: 1.8786\n",
      "Epoch 2457/10000000\n",
      " - 0s - loss: 1.7914 - val_loss: 1.7830\n",
      "Epoch 2458/10000000\n",
      " - 0s - loss: 1.5962 - val_loss: 1.8920\n",
      "Epoch 2459/10000000\n",
      " - 0s - loss: 1.6384 - val_loss: 1.8215\n",
      "Epoch 2460/10000000\n",
      " - 0s - loss: 1.9561 - val_loss: 2.1426\n",
      "Epoch 2461/10000000\n",
      " - 0s - loss: 2.1368 - val_loss: 1.9853\n",
      "Epoch 2462/10000000\n",
      " - 0s - loss: 1.9826 - val_loss: 2.0176\n",
      "Epoch 2463/10000000\n",
      " - 0s - loss: 1.9304 - val_loss: 1.7477\n",
      "Epoch 2464/10000000\n",
      " - 0s - loss: 1.5385 - val_loss: 1.7640\n",
      "Epoch 2465/10000000\n",
      " - 0s - loss: 1.4542 - val_loss: 1.8220\n",
      "Epoch 2466/10000000\n",
      " - 0s - loss: 1.6904 - val_loss: 1.7493\n",
      "Epoch 2467/10000000\n",
      " - 0s - loss: 1.4969 - val_loss: 1.9184\n",
      "Epoch 2468/10000000\n",
      " - 0s - loss: 1.5487 - val_loss: 1.8223\n",
      "Epoch 2469/10000000\n",
      " - 0s - loss: 1.3932 - val_loss: 1.9882\n",
      "Epoch 2470/10000000\n",
      " - 0s - loss: 1.5667 - val_loss: 1.8436\n",
      "Epoch 2471/10000000\n",
      " - 0s - loss: 1.5957 - val_loss: 1.7303\n",
      "Epoch 2472/10000000\n",
      " - 0s - loss: 1.3919 - val_loss: 1.7942\n",
      "Epoch 2473/10000000\n",
      " - 0s - loss: 1.3736 - val_loss: 1.8910\n",
      "Epoch 2474/10000000\n",
      " - 0s - loss: 1.4149 - val_loss: 2.0162\n",
      "Epoch 2475/10000000\n",
      " - 0s - loss: 1.4632 - val_loss: 1.8352\n",
      "Epoch 2476/10000000\n",
      " - 0s - loss: 1.3951 - val_loss: 1.8290\n",
      "Epoch 2477/10000000\n",
      " - 0s - loss: 1.4107 - val_loss: 1.7612\n",
      "Epoch 2478/10000000\n",
      " - 0s - loss: 1.4005 - val_loss: 1.9145\n",
      "Epoch 2479/10000000\n",
      " - 0s - loss: 1.4991 - val_loss: 1.6610\n",
      "Epoch 2480/10000000\n",
      " - 0s - loss: 1.3818 - val_loss: 1.7033\n",
      "Epoch 2481/10000000\n",
      " - 0s - loss: 1.5171 - val_loss: 1.6934\n",
      "Epoch 2482/10000000\n",
      " - 0s - loss: 1.4181 - val_loss: 1.8872\n",
      "Epoch 2483/10000000\n",
      " - 0s - loss: 1.4841 - val_loss: 1.8840\n",
      "Epoch 2484/10000000\n",
      " - 0s - loss: 1.4326 - val_loss: 1.9768\n",
      "Epoch 2485/10000000\n",
      " - 0s - loss: 1.8456 - val_loss: 2.2204\n",
      "Epoch 2486/10000000\n",
      " - 0s - loss: 2.5119 - val_loss: 2.4549\n",
      "Epoch 2487/10000000\n",
      " - 0s - loss: 2.1388 - val_loss: 1.7301\n",
      "Epoch 2488/10000000\n",
      " - 0s - loss: 1.6270 - val_loss: 2.0326\n",
      "Epoch 2489/10000000\n",
      " - 0s - loss: 2.1610 - val_loss: 2.1380\n",
      "Epoch 2490/10000000\n",
      " - 0s - loss: 1.9593 - val_loss: 1.7625\n",
      "Epoch 2491/10000000\n",
      " - 0s - loss: 1.5133 - val_loss: 1.8139\n",
      "Epoch 2492/10000000\n",
      " - 0s - loss: 1.4578 - val_loss: 1.8073\n",
      "Epoch 2493/10000000\n",
      " - 0s - loss: 1.5891 - val_loss: 1.8514\n",
      "Epoch 2494/10000000\n",
      " - 0s - loss: 1.4260 - val_loss: 1.7176\n",
      "Epoch 2495/10000000\n",
      " - 0s - loss: 1.3927 - val_loss: 1.7059\n",
      "Epoch 2496/10000000\n",
      " - 0s - loss: 1.5527 - val_loss: 1.7609\n",
      "Epoch 2497/10000000\n",
      " - 0s - loss: 1.5063 - val_loss: 1.9171\n",
      "Epoch 2498/10000000\n",
      " - 0s - loss: 1.7610 - val_loss: 1.7867\n",
      "Epoch 2499/10000000\n",
      " - 0s - loss: 1.5560 - val_loss: 1.8387\n",
      "Epoch 2500/10000000\n",
      " - 0s - loss: 1.6862 - val_loss: 1.7915\n",
      "Epoch 2501/10000000\n",
      " - 0s - loss: 1.3676 - val_loss: 1.9290\n",
      "Epoch 2502/10000000\n",
      " - 0s - loss: 1.4062 - val_loss: 2.0502\n",
      "Epoch 2503/10000000\n",
      " - 0s - loss: 1.5411 - val_loss: 2.0422\n",
      "Epoch 2504/10000000\n",
      " - 0s - loss: 1.6270 - val_loss: 2.7928\n",
      "Epoch 2505/10000000\n",
      " - 0s - loss: 2.4719 - val_loss: 2.4782\n",
      "Epoch 2506/10000000\n",
      " - 0s - loss: 2.7811 - val_loss: 3.2249\n",
      "Epoch 2507/10000000\n",
      " - 0s - loss: 3.3043 - val_loss: 2.1703\n",
      "Epoch 2508/10000000\n",
      " - 0s - loss: 3.4010 - val_loss: 1.9629\n",
      "Epoch 2509/10000000\n",
      " - 0s - loss: 2.2870 - val_loss: 1.9178\n",
      "Epoch 2510/10000000\n",
      " - 0s - loss: 1.6805 - val_loss: 1.8020\n",
      "Epoch 2511/10000000\n",
      " - 0s - loss: 1.8247 - val_loss: 1.9689\n",
      "Epoch 2512/10000000\n",
      " - 0s - loss: 1.7691 - val_loss: 1.9312\n",
      "Epoch 2513/10000000\n",
      " - 0s - loss: 1.5103 - val_loss: 1.9480\n",
      "Epoch 2514/10000000\n",
      " - 0s - loss: 1.6092 - val_loss: 1.8246\n",
      "Epoch 2515/10000000\n",
      " - 0s - loss: 1.3541 - val_loss: 1.8912\n",
      "Epoch 2516/10000000\n",
      " - 0s - loss: 1.5483 - val_loss: 2.0212\n",
      "Epoch 2517/10000000\n",
      " - 0s - loss: 1.6029 - val_loss: 2.1892\n",
      "Epoch 2518/10000000\n",
      " - 0s - loss: 1.7744 - val_loss: 1.8108\n",
      "Epoch 2519/10000000\n",
      " - 0s - loss: 1.3670 - val_loss: 1.8803\n",
      "Epoch 2520/10000000\n",
      " - 0s - loss: 1.4326 - val_loss: 1.7554\n",
      "Epoch 2521/10000000\n",
      " - 0s - loss: 1.3835 - val_loss: 1.7219\n",
      "Epoch 2522/10000000\n",
      " - 0s - loss: 1.3460 - val_loss: 1.7360\n",
      "Epoch 2523/10000000\n",
      " - 0s - loss: 1.4072 - val_loss: 1.7026\n",
      "Epoch 2524/10000000\n",
      " - 0s - loss: 1.3764 - val_loss: 1.6934\n",
      "Epoch 2525/10000000\n",
      " - 0s - loss: 1.3943 - val_loss: 1.6796\n",
      "Epoch 2526/10000000\n",
      " - 0s - loss: 1.3687 - val_loss: 1.6326\n",
      "Epoch 2527/10000000\n",
      " - 0s - loss: 1.3584 - val_loss: 1.6630\n",
      "Epoch 2528/10000000\n",
      " - 0s - loss: 1.3679 - val_loss: 1.7693\n",
      "Epoch 2529/10000000\n",
      " - 0s - loss: 1.3453 - val_loss: 1.9926\n",
      "Epoch 2530/10000000\n",
      " - 0s - loss: 1.6037 - val_loss: 1.9407\n",
      "Epoch 2531/10000000\n",
      " - 0s - loss: 1.4637 - val_loss: 2.0834\n",
      "Epoch 2532/10000000\n",
      " - 0s - loss: 1.7956 - val_loss: 2.0085\n",
      "Epoch 2533/10000000\n",
      " - 0s - loss: 1.8283 - val_loss: 1.7008\n",
      "Epoch 2534/10000000\n",
      " - 0s - loss: 1.3175 - val_loss: 1.7788\n",
      "Epoch 2535/10000000\n",
      " - 0s - loss: 1.3663 - val_loss: 1.7381\n",
      "Epoch 2536/10000000\n",
      " - 0s - loss: 1.3400 - val_loss: 1.7436\n",
      "Epoch 2537/10000000\n",
      " - 0s - loss: 1.3849 - val_loss: 1.7632\n",
      "Epoch 2538/10000000\n",
      " - 0s - loss: 1.3825 - val_loss: 1.8045\n",
      "Epoch 2539/10000000\n",
      " - 0s - loss: 1.3660 - val_loss: 1.8778\n",
      "Epoch 2540/10000000\n",
      " - 0s - loss: 1.3805 - val_loss: 1.7428\n",
      "Epoch 2541/10000000\n",
      " - 0s - loss: 1.3443 - val_loss: 1.7005\n",
      "Epoch 2542/10000000\n",
      " - 0s - loss: 1.3718 - val_loss: 1.7077\n",
      "Epoch 2543/10000000\n",
      " - 0s - loss: 1.3268 - val_loss: 1.8381\n",
      "Epoch 2544/10000000\n",
      " - 0s - loss: 1.4694 - val_loss: 1.7394\n",
      "Epoch 2545/10000000\n",
      " - 0s - loss: 1.3292 - val_loss: 1.6760\n",
      "Epoch 2546/10000000\n",
      " - 0s - loss: 1.3517 - val_loss: 1.6526\n",
      "Epoch 2547/10000000\n",
      " - 0s - loss: 1.4592 - val_loss: 1.9479\n",
      "Epoch 2548/10000000\n",
      " - 0s - loss: 1.9658 - val_loss: 2.3549\n",
      "Epoch 2549/10000000\n",
      " - 0s - loss: 2.1445 - val_loss: 1.9510\n",
      "Epoch 2550/10000000\n",
      " - 0s - loss: 1.7381 - val_loss: 1.8604\n",
      "Epoch 2551/10000000\n",
      " - 0s - loss: 1.6144 - val_loss: 1.8051\n",
      "Epoch 2552/10000000\n",
      " - 0s - loss: 1.4375 - val_loss: 1.8125\n",
      "Epoch 2553/10000000\n",
      " - 0s - loss: 1.3336 - val_loss: 1.8297\n",
      "Epoch 2554/10000000\n",
      " - 0s - loss: 1.4050 - val_loss: 1.7730\n",
      "Epoch 2555/10000000\n",
      " - 0s - loss: 1.3347 - val_loss: 1.8488\n",
      "Epoch 2556/10000000\n",
      " - 0s - loss: 1.4382 - val_loss: 1.7848\n",
      "Epoch 2557/10000000\n",
      " - 0s - loss: 1.7301 - val_loss: 2.1148\n",
      "Epoch 2558/10000000\n",
      " - 0s - loss: 2.3282 - val_loss: 2.0605\n",
      "Epoch 2559/10000000\n",
      " - 0s - loss: 1.7162 - val_loss: 2.3702\n",
      "Epoch 2560/10000000\n",
      " - 0s - loss: 2.2475 - val_loss: 2.7202\n",
      "Epoch 2561/10000000\n",
      " - 0s - loss: 2.6892 - val_loss: 2.3693\n",
      "Epoch 2562/10000000\n",
      " - 0s - loss: 1.8468 - val_loss: 1.8671\n",
      "Epoch 2563/10000000\n",
      " - 0s - loss: 1.6075 - val_loss: 1.9036\n",
      "Epoch 2564/10000000\n",
      " - 0s - loss: 1.7954 - val_loss: 1.8127\n",
      "Epoch 2565/10000000\n",
      " - 0s - loss: 1.5740 - val_loss: 1.6914\n",
      "Epoch 2566/10000000\n",
      " - 0s - loss: 1.3764 - val_loss: 1.6512\n",
      "Epoch 2567/10000000\n",
      " - 0s - loss: 1.3110 - val_loss: 1.7547\n",
      "Epoch 2568/10000000\n",
      " - 0s - loss: 1.4774 - val_loss: 1.7262\n",
      "Epoch 2569/10000000\n",
      " - 0s - loss: 1.4222 - val_loss: 1.8916\n",
      "Epoch 2570/10000000\n",
      " - 0s - loss: 1.5819 - val_loss: 1.7147\n",
      "Epoch 2571/10000000\n",
      " - 0s - loss: 1.3891 - val_loss: 1.7405\n",
      "Epoch 2572/10000000\n",
      " - 0s - loss: 1.3327 - val_loss: 1.7903\n",
      "Epoch 2573/10000000\n",
      " - 0s - loss: 1.3290 - val_loss: 1.7861\n",
      "Epoch 2574/10000000\n",
      " - 0s - loss: 1.3452 - val_loss: 1.8954\n",
      "Epoch 2575/10000000\n",
      " - 0s - loss: 1.6255 - val_loss: 1.6396\n",
      "Epoch 2576/10000000\n",
      " - 0s - loss: 1.3511 - val_loss: 1.7424\n",
      "Epoch 2577/10000000\n",
      " - 0s - loss: 1.4582 - val_loss: 1.8587\n",
      "Epoch 2578/10000000\n",
      " - 0s - loss: 1.4265 - val_loss: 2.0653\n",
      "Epoch 2579/10000000\n",
      " - 0s - loss: 2.0478 - val_loss: 1.9843\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2580/10000000\n",
      " - 0s - loss: 1.9038 - val_loss: 1.7195\n",
      "Epoch 2581/10000000\n",
      " - 0s - loss: 1.4520 - val_loss: 1.6614\n",
      "Epoch 2582/10000000\n",
      " - 0s - loss: 1.4714 - val_loss: 1.7760\n",
      "Epoch 2583/10000000\n",
      " - 0s - loss: 1.5573 - val_loss: 1.7176\n",
      "Epoch 2584/10000000\n",
      " - 0s - loss: 1.4456 - val_loss: 1.7054\n",
      "Epoch 2585/10000000\n",
      " - 0s - loss: 1.3424 - val_loss: 1.7746\n",
      "Epoch 2586/10000000\n",
      " - 0s - loss: 1.4076 - val_loss: 2.0301\n",
      "Epoch 2587/10000000\n",
      " - 0s - loss: 1.4674 - val_loss: 1.9713\n",
      "Epoch 2588/10000000\n",
      " - 0s - loss: 1.5974 - val_loss: 1.7151\n",
      "Epoch 2589/10000000\n",
      " - 0s - loss: 1.4525 - val_loss: 1.6827\n",
      "Epoch 2590/10000000\n",
      " - 0s - loss: 1.3590 - val_loss: 1.6954\n",
      "Epoch 2591/10000000\n",
      " - 0s - loss: 1.3671 - val_loss: 1.6896\n",
      "Epoch 2592/10000000\n",
      " - 0s - loss: 1.3403 - val_loss: 1.7094\n",
      "Epoch 2593/10000000\n",
      " - 0s - loss: 1.2935 - val_loss: 1.7258\n",
      "Epoch 2594/10000000\n",
      " - 0s - loss: 1.3568 - val_loss: 1.7205\n",
      "Epoch 2595/10000000\n",
      " - 0s - loss: 1.3755 - val_loss: 1.7782\n",
      "Epoch 2596/10000000\n",
      " - 0s - loss: 1.3133 - val_loss: 1.7882\n",
      "Epoch 2597/10000000\n",
      " - 0s - loss: 1.3858 - val_loss: 2.0510\n",
      "Epoch 2598/10000000\n",
      " - 0s - loss: 1.7314 - val_loss: 2.0366\n",
      "Epoch 2599/10000000\n",
      " - 0s - loss: 1.8343 - val_loss: 2.1730\n",
      "Epoch 2600/10000000\n",
      " - 0s - loss: 1.9908 - val_loss: 1.7440\n",
      "Epoch 2601/10000000\n",
      " - 0s - loss: 1.4913 - val_loss: 1.6604\n",
      "Epoch 2602/10000000\n",
      " - 0s - loss: 1.3815 - val_loss: 1.6387\n",
      "Epoch 2603/10000000\n",
      " - 0s - loss: 1.3252 - val_loss: 1.6901\n",
      "Epoch 2604/10000000\n",
      " - 0s - loss: 1.3236 - val_loss: 1.7163\n",
      "Epoch 2605/10000000\n",
      " - 0s - loss: 1.2904 - val_loss: 1.7234\n",
      "Epoch 2606/10000000\n",
      " - 0s - loss: 1.3115 - val_loss: 1.8633\n",
      "Epoch 2607/10000000\n",
      " - 0s - loss: 1.4661 - val_loss: 1.7071\n",
      "Epoch 2608/10000000\n",
      " - 0s - loss: 1.3656 - val_loss: 1.8612\n",
      "Epoch 2609/10000000\n",
      " - 0s - loss: 1.4484 - val_loss: 1.8549\n",
      "Epoch 2610/10000000\n",
      " - 0s - loss: 1.5578 - val_loss: 1.9748\n",
      "Epoch 2611/10000000\n",
      " - 0s - loss: 1.4999 - val_loss: 1.7021\n",
      "Epoch 2612/10000000\n",
      " - 0s - loss: 1.3990 - val_loss: 1.7475\n",
      "Epoch 2613/10000000\n",
      " - 0s - loss: 1.5110 - val_loss: 1.9226\n",
      "Epoch 2614/10000000\n",
      " - 0s - loss: 1.7579 - val_loss: 2.2847\n",
      "Epoch 2615/10000000\n",
      " - 0s - loss: 2.1236 - val_loss: 1.8474\n",
      "Epoch 2616/10000000\n",
      " - 0s - loss: 1.8753 - val_loss: 1.8974\n",
      "Epoch 2617/10000000\n",
      " - 0s - loss: 1.9509 - val_loss: 2.4225\n",
      "Epoch 2618/10000000\n",
      " - 0s - loss: 2.4968 - val_loss: 2.2511\n",
      "Epoch 2619/10000000\n",
      " - 0s - loss: 2.0199 - val_loss: 1.8133\n",
      "Epoch 2620/10000000\n",
      " - 0s - loss: 1.7018 - val_loss: 1.7072\n",
      "Epoch 2621/10000000\n",
      " - 0s - loss: 1.5933 - val_loss: 1.8674\n",
      "Epoch 2622/10000000\n",
      " - 0s - loss: 1.9925 - val_loss: 1.8292\n",
      "Epoch 2623/10000000\n",
      " - 0s - loss: 1.9283 - val_loss: 1.6376\n",
      "Epoch 2624/10000000\n",
      " - 0s - loss: 1.6115 - val_loss: 1.6630\n",
      "Epoch 2625/10000000\n",
      " - 0s - loss: 1.5035 - val_loss: 1.7444\n",
      "Epoch 2626/10000000\n",
      " - 0s - loss: 1.3946 - val_loss: 1.7642\n",
      "Epoch 2627/10000000\n",
      " - 0s - loss: 1.5142 - val_loss: 1.8023\n",
      "Epoch 2628/10000000\n",
      " - 0s - loss: 1.4749 - val_loss: 1.7048\n",
      "Epoch 2629/10000000\n",
      " - 0s - loss: 1.3417 - val_loss: 1.7269\n",
      "Epoch 2630/10000000\n",
      " - 0s - loss: 1.3157 - val_loss: 1.7613\n",
      "Epoch 2631/10000000\n",
      " - 0s - loss: 1.3352 - val_loss: 1.6474\n",
      "Epoch 2632/10000000\n",
      " - 0s - loss: 1.3182 - val_loss: 1.7612\n",
      "Epoch 2633/10000000\n",
      " - 0s - loss: 1.4595 - val_loss: 1.9944\n",
      "Epoch 2634/10000000\n",
      " - 0s - loss: 1.9530 - val_loss: 2.6760\n",
      "Epoch 2635/10000000\n",
      " - 0s - loss: 2.1688 - val_loss: 2.3726\n",
      "Epoch 2636/10000000\n",
      " - 0s - loss: 1.8665 - val_loss: 1.7761\n",
      "Epoch 2637/10000000\n",
      " - 0s - loss: 1.3545 - val_loss: 1.7323\n",
      "Epoch 2638/10000000\n",
      " - 0s - loss: 1.2964 - val_loss: 1.8770\n",
      "Epoch 2639/10000000\n",
      " - 0s - loss: 1.4383 - val_loss: 1.8806\n",
      "Epoch 2640/10000000\n",
      " - 0s - loss: 1.3350 - val_loss: 1.7390\n",
      "Epoch 2641/10000000\n",
      " - 0s - loss: 1.3680 - val_loss: 1.8659\n",
      "Epoch 2642/10000000\n",
      " - 0s - loss: 1.6375 - val_loss: 2.0070\n",
      "Epoch 2643/10000000\n",
      " - 0s - loss: 1.8820 - val_loss: 2.1540\n",
      "Epoch 2644/10000000\n",
      " - 0s - loss: 1.8142 - val_loss: 2.1498\n",
      "Epoch 2645/10000000\n",
      " - 0s - loss: 2.1107 - val_loss: 1.8377\n",
      "Epoch 2646/10000000\n",
      " - 0s - loss: 2.0673 - val_loss: 1.8828\n",
      "Epoch 2647/10000000\n",
      " - 0s - loss: 1.8820 - val_loss: 2.1981\n",
      "Epoch 2648/10000000\n",
      " - 0s - loss: 1.8395 - val_loss: 1.8116\n",
      "Epoch 2649/10000000\n",
      " - 0s - loss: 1.6659 - val_loss: 1.7052\n",
      "Epoch 2650/10000000\n",
      " - 0s - loss: 1.4295 - val_loss: 1.6872\n",
      "Epoch 2651/10000000\n",
      " - 0s - loss: 1.3133 - val_loss: 1.6969\n",
      "Epoch 2652/10000000\n",
      " - 0s - loss: 1.2881 - val_loss: 1.6993\n",
      "Epoch 2653/10000000\n",
      " - 0s - loss: 1.2794 - val_loss: 1.8014\n",
      "Epoch 2654/10000000\n",
      " - 0s - loss: 1.4171 - val_loss: 1.6737\n",
      "Epoch 2655/10000000\n",
      " - 0s - loss: 1.2889 - val_loss: 1.7458\n",
      "Epoch 2656/10000000\n",
      " - 0s - loss: 1.3455 - val_loss: 2.0307\n",
      "Epoch 2657/10000000\n",
      " - 0s - loss: 1.6738 - val_loss: 1.7636\n",
      "Epoch 2658/10000000\n",
      " - 0s - loss: 1.4283 - val_loss: 1.9444\n",
      "Epoch 2659/10000000\n",
      " - 0s - loss: 1.5220 - val_loss: 1.7359\n",
      "Epoch 2660/10000000\n",
      " - 0s - loss: 1.5382 - val_loss: 2.0097\n",
      "Epoch 2661/10000000\n",
      " - 0s - loss: 1.6512 - val_loss: 1.9112\n",
      "Epoch 2662/10000000\n",
      " - 0s - loss: 1.7590 - val_loss: 1.9051\n",
      "Epoch 2663/10000000\n",
      " - 0s - loss: 1.5622 - val_loss: 2.0280\n",
      "Epoch 2664/10000000\n",
      " - 0s - loss: 2.1341 - val_loss: 3.0081\n",
      "Epoch 2665/10000000\n",
      " - 0s - loss: 3.6139 - val_loss: 2.4902\n",
      "Epoch 2666/10000000\n",
      " - 0s - loss: 2.6819 - val_loss: 2.1257\n",
      "Epoch 2667/10000000\n",
      " - 0s - loss: 2.2597 - val_loss: 1.7845\n",
      "Epoch 2668/10000000\n",
      " - 0s - loss: 1.4266 - val_loss: 1.7123\n",
      "Epoch 2669/10000000\n",
      " - 0s - loss: 1.3580 - val_loss: 1.6769\n",
      "Epoch 2670/10000000\n",
      " - 0s - loss: 1.3233 - val_loss: 1.7959\n",
      "Epoch 2671/10000000\n",
      " - 0s - loss: 1.3385 - val_loss: 1.8059\n",
      "Epoch 2672/10000000\n",
      " - 0s - loss: 1.3702 - val_loss: 1.7578\n",
      "Epoch 2673/10000000\n",
      " - 0s - loss: 1.4960 - val_loss: 1.7638\n",
      "Epoch 2674/10000000\n",
      " - 0s - loss: 1.4693 - val_loss: 1.6516\n",
      "Epoch 2675/10000000\n",
      " - 0s - loss: 1.3668 - val_loss: 2.0237\n",
      "Epoch 2676/10000000\n",
      " - 0s - loss: 1.6042 - val_loss: 1.8863\n",
      "Epoch 2677/10000000\n",
      " - 0s - loss: 1.5139 - val_loss: 1.7115\n",
      "Epoch 2678/10000000\n",
      " - 0s - loss: 1.2814 - val_loss: 1.7299\n",
      "Epoch 2679/10000000\n",
      " - 0s - loss: 1.3832 - val_loss: 1.6815\n",
      "Epoch 2680/10000000\n",
      " - 0s - loss: 1.3439 - val_loss: 1.5507\n",
      "Epoch 2681/10000000\n",
      " - 0s - loss: 1.3101 - val_loss: 1.5709\n",
      "Epoch 2682/10000000\n",
      " - 0s - loss: 1.3361 - val_loss: 1.6539\n",
      "Epoch 2683/10000000\n",
      " - 0s - loss: 1.3498 - val_loss: 1.7758\n",
      "Epoch 2684/10000000\n",
      " - 0s - loss: 1.3546 - val_loss: 1.6947\n",
      "Epoch 2685/10000000\n",
      " - 0s - loss: 1.3992 - val_loss: 2.1328\n",
      "Epoch 2686/10000000\n",
      " - 0s - loss: 1.9859 - val_loss: 2.5587\n",
      "Epoch 2687/10000000\n",
      " - 0s - loss: 2.2975 - val_loss: 1.8994\n",
      "Epoch 2688/10000000\n",
      " - 0s - loss: 1.4075 - val_loss: 1.9519\n",
      "Epoch 2689/10000000\n",
      " - 0s - loss: 1.6037 - val_loss: 1.9050\n",
      "Epoch 2690/10000000\n",
      " - 0s - loss: 1.7992 - val_loss: 2.5948\n",
      "Epoch 2691/10000000\n",
      " - 0s - loss: 2.1612 - val_loss: 1.9776\n",
      "Epoch 2692/10000000\n",
      " - 0s - loss: 2.5116 - val_loss: 2.0189\n",
      "Epoch 2693/10000000\n",
      " - 0s - loss: 2.0127 - val_loss: 1.7876\n",
      "Epoch 2694/10000000\n",
      " - 0s - loss: 1.9245 - val_loss: 1.7243\n",
      "Epoch 2695/10000000\n",
      " - 0s - loss: 1.4332 - val_loss: 1.7470\n",
      "Epoch 2696/10000000\n",
      " - 0s - loss: 1.2769 - val_loss: 1.6787\n",
      "Epoch 2697/10000000\n",
      " - 0s - loss: 1.3161 - val_loss: 1.7597\n",
      "Epoch 2698/10000000\n",
      " - 0s - loss: 1.5012 - val_loss: 2.0963\n",
      "Epoch 2699/10000000\n",
      " - 0s - loss: 2.2042 - val_loss: 2.0780\n",
      "Epoch 2700/10000000\n",
      " - 0s - loss: 2.3353 - val_loss: 2.0333\n",
      "Epoch 2701/10000000\n",
      " - 0s - loss: 1.7567 - val_loss: 1.7995\n",
      "Epoch 2702/10000000\n",
      " - 0s - loss: 1.5870 - val_loss: 1.7552\n",
      "Epoch 2703/10000000\n",
      " - 0s - loss: 1.7350 - val_loss: 2.1224\n",
      "Epoch 2704/10000000\n",
      " - 0s - loss: 2.2736 - val_loss: 1.6717\n",
      "Epoch 2705/10000000\n",
      " - 0s - loss: 1.4388 - val_loss: 1.6904\n",
      "Epoch 2706/10000000\n",
      " - 0s - loss: 1.3632 - val_loss: 1.6345\n",
      "Epoch 2707/10000000\n",
      " - 0s - loss: 1.3858 - val_loss: 1.6243\n",
      "Epoch 2708/10000000\n",
      " - 0s - loss: 1.2532 - val_loss: 1.8136\n",
      "Epoch 2709/10000000\n",
      " - 0s - loss: 1.3782 - val_loss: 1.7818\n",
      "Epoch 2710/10000000\n",
      " - 0s - loss: 1.2912 - val_loss: 1.8519\n",
      "Epoch 2711/10000000\n",
      " - 0s - loss: 1.5315 - val_loss: 1.8824\n",
      "Epoch 2712/10000000\n",
      " - 0s - loss: 1.5243 - val_loss: 1.9035\n",
      "Epoch 2713/10000000\n",
      " - 0s - loss: 1.9555 - val_loss: 1.8471\n",
      "Epoch 2714/10000000\n",
      " - 0s - loss: 1.4179 - val_loss: 1.6933\n",
      "Epoch 2715/10000000\n",
      " - 0s - loss: 1.3175 - val_loss: 1.6602\n",
      "Epoch 2716/10000000\n",
      " - 0s - loss: 1.3257 - val_loss: 1.8071\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2717/10000000\n",
      " - 0s - loss: 1.4621 - val_loss: 1.7248\n",
      "Epoch 2718/10000000\n",
      " - 0s - loss: 1.3649 - val_loss: 1.6728\n",
      "Epoch 2719/10000000\n",
      " - 0s - loss: 1.3278 - val_loss: 1.5935\n",
      "Epoch 2720/10000000\n",
      " - 0s - loss: 1.2696 - val_loss: 1.6241\n",
      "Epoch 2721/10000000\n",
      " - 0s - loss: 1.3824 - val_loss: 1.6690\n",
      "Epoch 2722/10000000\n",
      " - 0s - loss: 1.4808 - val_loss: 1.9612\n",
      "Epoch 2723/10000000\n",
      " - 0s - loss: 2.0078 - val_loss: 2.0428\n",
      "Epoch 2724/10000000\n",
      " - 0s - loss: 1.7225 - val_loss: 1.9720\n",
      "Epoch 2725/10000000\n",
      " - 0s - loss: 1.6575 - val_loss: 1.7224\n",
      "Epoch 2726/10000000\n",
      " - 0s - loss: 1.4810 - val_loss: 2.1401\n",
      "Epoch 2727/10000000\n",
      " - 0s - loss: 2.5481 - val_loss: 2.2685\n",
      "Epoch 2728/10000000\n",
      " - 0s - loss: 2.0189 - val_loss: 2.3888\n",
      "Epoch 2729/10000000\n",
      " - 0s - loss: 1.9512 - val_loss: 1.9907\n",
      "Epoch 2730/10000000\n",
      " - 0s - loss: 1.9495 - val_loss: 1.9014\n",
      "Epoch 2731/10000000\n",
      " - 0s - loss: 1.9201 - val_loss: 1.7898\n",
      "Epoch 2732/10000000\n",
      " - 0s - loss: 1.5941 - val_loss: 1.5639\n",
      "Epoch 2733/10000000\n",
      " - 0s - loss: 1.7719 - val_loss: 1.5085\n",
      "Epoch 2734/10000000\n",
      " - 0s - loss: 1.3881 - val_loss: 1.5673\n",
      "Epoch 2735/10000000\n",
      " - 0s - loss: 1.4961 - val_loss: 1.6601\n",
      "Epoch 2736/10000000\n",
      " - 0s - loss: 1.4734 - val_loss: 1.7338\n",
      "Epoch 2737/10000000\n",
      " - 0s - loss: 1.6202 - val_loss: 1.7692\n",
      "Epoch 2738/10000000\n",
      " - 0s - loss: 1.3751 - val_loss: 1.7544\n",
      "Epoch 2739/10000000\n",
      " - 0s - loss: 1.2832 - val_loss: 1.9397\n",
      "Epoch 2740/10000000\n",
      " - 0s - loss: 1.3573 - val_loss: 1.7156\n",
      "Epoch 2741/10000000\n",
      " - 0s - loss: 1.3403 - val_loss: 1.6237\n",
      "Epoch 2742/10000000\n",
      " - 0s - loss: 1.3898 - val_loss: 1.6656\n",
      "Epoch 2743/10000000\n",
      " - 0s - loss: 1.4530 - val_loss: 1.5715\n",
      "Epoch 2744/10000000\n",
      " - 0s - loss: 1.3561 - val_loss: 1.6900\n",
      "Epoch 2745/10000000\n",
      " - 0s - loss: 1.5142 - val_loss: 1.7292\n",
      "Epoch 2746/10000000\n",
      " - 0s - loss: 1.5262 - val_loss: 1.9920\n",
      "Epoch 2747/10000000\n",
      " - 0s - loss: 1.8200 - val_loss: 1.8662\n",
      "Epoch 2748/10000000\n",
      " - 0s - loss: 1.5235 - val_loss: 1.7875\n",
      "Epoch 2749/10000000\n",
      " - 0s - loss: 1.3494 - val_loss: 1.6770\n",
      "Epoch 2750/10000000\n",
      " - 0s - loss: 1.3644 - val_loss: 1.7284\n",
      "Epoch 2751/10000000\n",
      " - 0s - loss: 1.3788 - val_loss: 1.7016\n",
      "Epoch 2752/10000000\n",
      " - 0s - loss: 1.3726 - val_loss: 1.6601\n",
      "Epoch 2753/10000000\n",
      " - 0s - loss: 1.3929 - val_loss: 1.7341\n",
      "Epoch 2754/10000000\n",
      " - 0s - loss: 1.5856 - val_loss: 1.7640\n",
      "Epoch 2755/10000000\n",
      " - 0s - loss: 1.2881 - val_loss: 1.7716\n",
      "Epoch 2756/10000000\n",
      " - 0s - loss: 1.2891 - val_loss: 1.7684\n",
      "Epoch 2757/10000000\n",
      " - 0s - loss: 1.3976 - val_loss: 1.7436\n",
      "Epoch 2758/10000000\n",
      " - 0s - loss: 1.5414 - val_loss: 1.6824\n",
      "Epoch 2759/10000000\n",
      " - 0s - loss: 1.2989 - val_loss: 1.7084\n",
      "Epoch 2760/10000000\n",
      " - 0s - loss: 1.2762 - val_loss: 1.6645\n",
      "Epoch 2761/10000000\n",
      " - 0s - loss: 1.3223 - val_loss: 1.8003\n",
      "Epoch 2762/10000000\n",
      " - 0s - loss: 1.4863 - val_loss: 1.7093\n",
      "Epoch 2763/10000000\n",
      " - 0s - loss: 1.2519 - val_loss: 1.7691\n",
      "Epoch 2764/10000000\n",
      " - 0s - loss: 1.3343 - val_loss: 1.7121\n",
      "Epoch 2765/10000000\n",
      " - 0s - loss: 1.5155 - val_loss: 1.7490\n",
      "Epoch 2766/10000000\n",
      " - 0s - loss: 1.3947 - val_loss: 1.7062\n",
      "Epoch 2767/10000000\n",
      " - 0s - loss: 1.3074 - val_loss: 1.6083\n",
      "Epoch 2768/10000000\n",
      " - 0s - loss: 1.2570 - val_loss: 1.5890\n",
      "Epoch 2769/10000000\n",
      " - 0s - loss: 1.2864 - val_loss: 1.6405\n",
      "Epoch 2770/10000000\n",
      " - 0s - loss: 1.3394 - val_loss: 1.6422\n",
      "Epoch 2771/10000000\n",
      " - 0s - loss: 1.2242 - val_loss: 1.8831\n",
      "Epoch 2772/10000000\n",
      " - 0s - loss: 1.3772 - val_loss: 1.9201\n",
      "Epoch 2773/10000000\n",
      " - 0s - loss: 1.5178 - val_loss: 1.7058\n",
      "Epoch 2774/10000000\n",
      " - 0s - loss: 1.2343 - val_loss: 1.7822\n",
      "Epoch 2775/10000000\n",
      " - 0s - loss: 1.3075 - val_loss: 1.7008\n",
      "Epoch 2776/10000000\n",
      " - 0s - loss: 1.3582 - val_loss: 1.6766\n",
      "Epoch 2777/10000000\n",
      " - 0s - loss: 1.4868 - val_loss: 1.6059\n",
      "Epoch 2778/10000000\n",
      " - 0s - loss: 1.2460 - val_loss: 1.7025\n",
      "Epoch 2779/10000000\n",
      " - 0s - loss: 1.3232 - val_loss: 1.7128\n",
      "Epoch 2780/10000000\n",
      " - 0s - loss: 1.2829 - val_loss: 1.9018\n",
      "Epoch 2781/10000000\n",
      " - 0s - loss: 1.7518 - val_loss: 2.1795\n",
      "Epoch 2782/10000000\n",
      " - 0s - loss: 1.7503 - val_loss: 1.8218\n",
      "Epoch 2783/10000000\n",
      " - 0s - loss: 1.3769 - val_loss: 1.7318\n",
      "Epoch 2784/10000000\n",
      " - 0s - loss: 1.2918 - val_loss: 1.7843\n",
      "Epoch 2785/10000000\n",
      " - 0s - loss: 1.5589 - val_loss: 2.2807\n",
      "Epoch 2786/10000000\n",
      " - 0s - loss: 2.4714 - val_loss: 2.4806\n",
      "Epoch 2787/10000000\n",
      " - 0s - loss: 2.7383 - val_loss: 1.6318\n",
      "Epoch 2788/10000000\n",
      " - 0s - loss: 1.2920 - val_loss: 1.5860\n",
      "Epoch 2789/10000000\n",
      " - 0s - loss: 1.2901 - val_loss: 1.6352\n",
      "Epoch 2790/10000000\n",
      " - 0s - loss: 1.2668 - val_loss: 1.9053\n",
      "Epoch 2791/10000000\n",
      " - 0s - loss: 1.5307 - val_loss: 2.0806\n",
      "Epoch 2792/10000000\n",
      " - 0s - loss: 1.6399 - val_loss: 2.1848\n",
      "Epoch 2793/10000000\n",
      " - 0s - loss: 2.0518 - val_loss: 2.5081\n",
      "Epoch 2794/10000000\n",
      " - 0s - loss: 2.4650 - val_loss: 2.5516\n",
      "Epoch 2795/10000000\n",
      " - 0s - loss: 3.0896 - val_loss: 2.1727\n",
      "Epoch 2796/10000000\n",
      " - 0s - loss: 1.9797 - val_loss: 1.9002\n",
      "Epoch 2797/10000000\n",
      " - 0s - loss: 1.8894 - val_loss: 2.2238\n",
      "Epoch 2798/10000000\n",
      " - 0s - loss: 2.8851 - val_loss: 2.2565\n",
      "Epoch 2799/10000000\n",
      " - 0s - loss: 2.7828 - val_loss: 2.2883\n",
      "Epoch 2800/10000000\n",
      " - 0s - loss: 2.3647 - val_loss: 2.5730\n",
      "Epoch 2801/10000000\n",
      " - 0s - loss: 2.7514 - val_loss: 1.8881\n",
      "Epoch 2802/10000000\n",
      " - 0s - loss: 2.0915 - val_loss: 1.8161\n",
      "Epoch 2803/10000000\n",
      " - 0s - loss: 2.3071 - val_loss: 1.6709\n",
      "Epoch 2804/10000000\n",
      " - 0s - loss: 1.7951 - val_loss: 1.7065\n",
      "Epoch 2805/10000000\n",
      " - 0s - loss: 2.0056 - val_loss: 1.8020\n",
      "Epoch 2806/10000000\n",
      " - 0s - loss: 2.1479 - val_loss: 1.6633\n",
      "Epoch 2807/10000000\n",
      " - 0s - loss: 1.6840 - val_loss: 1.6726\n",
      "Epoch 2808/10000000\n",
      " - 0s - loss: 1.5702 - val_loss: 2.0728\n",
      "Epoch 2809/10000000\n",
      " - 0s - loss: 1.4627 - val_loss: 2.2191\n",
      "Epoch 2810/10000000\n",
      " - 0s - loss: 1.7289 - val_loss: 1.8059\n",
      "Epoch 2811/10000000\n",
      " - 0s - loss: 1.4504 - val_loss: 1.6365\n",
      "Epoch 2812/10000000\n",
      " - 0s - loss: 1.2797 - val_loss: 1.6881\n",
      "Epoch 2813/10000000\n",
      " - 0s - loss: 1.3002 - val_loss: 1.9270\n",
      "Epoch 2814/10000000\n",
      " - 0s - loss: 1.4221 - val_loss: 1.8359\n",
      "Epoch 2815/10000000\n",
      " - 0s - loss: 1.2741 - val_loss: 2.0244\n",
      "Epoch 2816/10000000\n",
      " - 0s - loss: 1.6824 - val_loss: 2.0849\n",
      "Epoch 2817/10000000\n",
      " - 0s - loss: 1.5600 - val_loss: 1.8524\n",
      "Epoch 2818/10000000\n",
      " - 0s - loss: 1.4361 - val_loss: 1.6896\n",
      "Epoch 2819/10000000\n",
      " - 0s - loss: 1.2753 - val_loss: 1.6727\n",
      "Epoch 2820/10000000\n",
      " - 0s - loss: 1.3120 - val_loss: 1.5970\n",
      "Epoch 2821/10000000\n",
      " - 0s - loss: 1.2564 - val_loss: 1.7691\n",
      "Epoch 2822/10000000\n",
      " - 0s - loss: 1.4261 - val_loss: 1.7250\n",
      "Epoch 2823/10000000\n",
      " - 0s - loss: 1.5542 - val_loss: 1.9200\n",
      "Epoch 2824/10000000\n",
      " - 0s - loss: 1.8042 - val_loss: 1.9407\n",
      "Epoch 2825/10000000\n",
      " - 0s - loss: 2.0226 - val_loss: 1.7337\n",
      "Epoch 2826/10000000\n",
      " - 0s - loss: 1.3587 - val_loss: 1.7781\n",
      "Epoch 2827/10000000\n",
      " - 0s - loss: 1.4486 - val_loss: 1.7229\n",
      "Epoch 2828/10000000\n",
      " - 0s - loss: 1.2646 - val_loss: 1.7716\n",
      "Epoch 2829/10000000\n",
      " - 0s - loss: 1.2240 - val_loss: 1.8915\n",
      "Epoch 2830/10000000\n",
      " - 0s - loss: 1.3944 - val_loss: 1.9768\n",
      "Epoch 2831/10000000\n",
      " - 0s - loss: 1.5086 - val_loss: 1.8417\n",
      "Epoch 2832/10000000\n",
      " - 0s - loss: 1.4303 - val_loss: 1.6513\n",
      "Epoch 2833/10000000\n",
      " - 0s - loss: 1.2740 - val_loss: 1.7168\n",
      "Epoch 2834/10000000\n",
      " - 0s - loss: 1.2845 - val_loss: 1.7378\n",
      "Epoch 2835/10000000\n",
      " - 0s - loss: 1.2430 - val_loss: 1.6671\n",
      "Epoch 2836/10000000\n",
      " - 0s - loss: 1.3087 - val_loss: 1.6003\n",
      "Epoch 2837/10000000\n",
      " - 0s - loss: 1.2634 - val_loss: 1.5856\n",
      "Epoch 2838/10000000\n",
      " - 0s - loss: 1.2234 - val_loss: 1.5814\n",
      "Epoch 2839/10000000\n",
      " - 0s - loss: 1.2464 - val_loss: 1.5688\n",
      "Epoch 2840/10000000\n",
      " - 0s - loss: 1.2262 - val_loss: 1.6295\n",
      "Epoch 2841/10000000\n",
      " - 0s - loss: 1.2454 - val_loss: 1.7173\n",
      "Epoch 2842/10000000\n",
      " - 0s - loss: 1.3148 - val_loss: 1.7689\n",
      "Epoch 2843/10000000\n",
      " - 0s - loss: 1.3170 - val_loss: 1.6865\n",
      "Epoch 2844/10000000\n",
      " - 0s - loss: 1.2286 - val_loss: 1.6372\n",
      "Epoch 2845/10000000\n",
      " - 0s - loss: 1.2354 - val_loss: 1.6543\n",
      "Epoch 2846/10000000\n",
      " - 0s - loss: 1.2809 - val_loss: 1.7583\n",
      "Epoch 2847/10000000\n",
      " - 0s - loss: 1.4116 - val_loss: 1.6750\n",
      "Epoch 2848/10000000\n",
      " - 0s - loss: 1.3596 - val_loss: 1.7114\n",
      "Epoch 2849/10000000\n",
      " - 0s - loss: 1.3105 - val_loss: 1.6621\n",
      "Epoch 2850/10000000\n",
      " - 0s - loss: 1.2576 - val_loss: 1.6816\n",
      "Epoch 2851/10000000\n",
      " - 0s - loss: 1.2970 - val_loss: 1.6773\n",
      "Epoch 2852/10000000\n",
      " - 0s - loss: 1.3091 - val_loss: 1.7283\n",
      "Epoch 2853/10000000\n",
      " - 0s - loss: 1.4163 - val_loss: 1.8677\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2854/10000000\n",
      " - 0s - loss: 1.5100 - val_loss: 1.6936\n",
      "Epoch 2855/10000000\n",
      " - 0s - loss: 1.2755 - val_loss: 1.5986\n",
      "Epoch 2856/10000000\n",
      " - 0s - loss: 1.2047 - val_loss: 1.7891\n",
      "Epoch 2857/10000000\n",
      " - 0s - loss: 1.3790 - val_loss: 2.1319\n",
      "Epoch 2858/10000000\n",
      " - 0s - loss: 1.7727 - val_loss: 2.3161\n",
      "Epoch 2859/10000000\n",
      " - 0s - loss: 2.1040 - val_loss: 2.7135\n",
      "Epoch 2860/10000000\n",
      " - 0s - loss: 2.5121 - val_loss: 2.0968\n",
      "Epoch 2861/10000000\n",
      " - 0s - loss: 1.9862 - val_loss: 1.7401\n",
      "Epoch 2862/10000000\n",
      " - 0s - loss: 1.4731 - val_loss: 1.7472\n",
      "Epoch 2863/10000000\n",
      " - 0s - loss: 1.4607 - val_loss: 1.8978\n",
      "Epoch 2864/10000000\n",
      " - 0s - loss: 2.1380 - val_loss: 1.9579\n",
      "Epoch 2865/10000000\n",
      " - 0s - loss: 2.2277 - val_loss: 1.9058\n",
      "Epoch 2866/10000000\n",
      " - 0s - loss: 1.9395 - val_loss: 2.0367\n",
      "Epoch 2867/10000000\n",
      " - 0s - loss: 2.1432 - val_loss: 1.7846\n",
      "Epoch 2868/10000000\n",
      " - 0s - loss: 1.5815 - val_loss: 1.5813\n",
      "Epoch 2869/10000000\n",
      " - 0s - loss: 1.3837 - val_loss: 1.6219\n",
      "Epoch 2870/10000000\n",
      " - 0s - loss: 1.5442 - val_loss: 1.6348\n",
      "Epoch 2871/10000000\n",
      " - 0s - loss: 1.3181 - val_loss: 1.7539\n",
      "Epoch 2872/10000000\n",
      " - 0s - loss: 1.2961 - val_loss: 1.7880\n",
      "Epoch 2873/10000000\n",
      " - 0s - loss: 1.2647 - val_loss: 1.6432\n",
      "Epoch 2874/10000000\n",
      " - 0s - loss: 1.2804 - val_loss: 2.1699\n",
      "Epoch 2875/10000000\n",
      " - 0s - loss: 2.1029 - val_loss: 2.1830\n",
      "Epoch 2876/10000000\n",
      " - 0s - loss: 1.7057 - val_loss: 1.8764\n",
      "Epoch 2877/10000000\n",
      " - 0s - loss: 1.5514 - val_loss: 1.7045\n",
      "Epoch 2878/10000000\n",
      " - 0s - loss: 1.3338 - val_loss: 1.6684\n",
      "Epoch 2879/10000000\n",
      " - 0s - loss: 1.3123 - val_loss: 1.6467\n",
      "Epoch 2880/10000000\n",
      " - 0s - loss: 1.2818 - val_loss: 1.6616\n",
      "Epoch 2881/10000000\n",
      " - 0s - loss: 1.2472 - val_loss: 1.6100\n",
      "Epoch 2882/10000000\n",
      " - 0s - loss: 1.2557 - val_loss: 1.6055\n",
      "Epoch 2883/10000000\n",
      " - 0s - loss: 1.4362 - val_loss: 2.0891\n",
      "Epoch 2884/10000000\n",
      " - 0s - loss: 2.0429 - val_loss: 1.9522\n",
      "Epoch 2885/10000000\n",
      " - 0s - loss: 1.7187 - val_loss: 1.6589\n",
      "Epoch 2886/10000000\n",
      " - 0s - loss: 1.4107 - val_loss: 1.6713\n",
      "Epoch 2887/10000000\n",
      " - 0s - loss: 1.2293 - val_loss: 1.7911\n",
      "Epoch 2888/10000000\n",
      " - 0s - loss: 1.3573 - val_loss: 1.6730\n",
      "Epoch 2889/10000000\n",
      " - 0s - loss: 1.2317 - val_loss: 1.7985\n",
      "Epoch 2890/10000000\n",
      " - 0s - loss: 1.2874 - val_loss: 1.7737\n",
      "Epoch 2891/10000000\n",
      " - 0s - loss: 1.2967 - val_loss: 1.7868\n",
      "Epoch 2892/10000000\n",
      " - 0s - loss: 1.3952 - val_loss: 1.6812\n",
      "Epoch 2893/10000000\n",
      " - 0s - loss: 1.2081 - val_loss: 1.6854\n",
      "Epoch 2894/10000000\n",
      " - 0s - loss: 1.2446 - val_loss: 1.6388\n",
      "Epoch 2895/10000000\n",
      " - 0s - loss: 1.2554 - val_loss: 1.6857\n",
      "Epoch 2896/10000000\n",
      " - 0s - loss: 1.4464 - val_loss: 1.6381\n",
      "Epoch 2897/10000000\n",
      " - 0s - loss: 1.3243 - val_loss: 1.5752\n",
      "Epoch 2898/10000000\n",
      " - 0s - loss: 1.2213 - val_loss: 1.6341\n",
      "Epoch 2899/10000000\n",
      " - 0s - loss: 1.2571 - val_loss: 1.5832\n",
      "Epoch 2900/10000000\n",
      " - 0s - loss: 1.3243 - val_loss: 1.5670\n",
      "Epoch 2901/10000000\n",
      " - 0s - loss: 1.2035 - val_loss: 1.6549\n",
      "Epoch 2902/10000000\n",
      " - 0s - loss: 1.2669 - val_loss: 1.6388\n",
      "Epoch 2903/10000000\n",
      " - 0s - loss: 1.2295 - val_loss: 1.6497\n",
      "Epoch 2904/10000000\n",
      " - 0s - loss: 1.2436 - val_loss: 1.6380\n",
      "Epoch 2905/10000000\n",
      " - 0s - loss: 1.2798 - val_loss: 1.5797\n",
      "Epoch 2906/10000000\n",
      " - 0s - loss: 1.2648 - val_loss: 1.6013\n",
      "Epoch 2907/10000000\n",
      " - 0s - loss: 1.2643 - val_loss: 1.5884\n",
      "Epoch 2908/10000000\n",
      " - 0s - loss: 1.2261 - val_loss: 1.5533\n",
      "Epoch 2909/10000000\n",
      " - 0s - loss: 1.2886 - val_loss: 1.5651\n",
      "Epoch 2910/10000000\n",
      " - 0s - loss: 1.2317 - val_loss: 1.6152\n",
      "Epoch 2911/10000000\n",
      " - 0s - loss: 1.3004 - val_loss: 1.7180\n",
      "Epoch 2912/10000000\n",
      " - 0s - loss: 1.4056 - val_loss: 1.7790\n",
      "Epoch 2913/10000000\n",
      " - 0s - loss: 1.4889 - val_loss: 2.1128\n",
      "Epoch 2914/10000000\n",
      " - 0s - loss: 2.3980 - val_loss: 2.2963\n",
      "Epoch 2915/10000000\n",
      " - 0s - loss: 2.3543 - val_loss: 2.3428\n",
      "Epoch 2916/10000000\n",
      " - 0s - loss: 2.3619 - val_loss: 1.8337\n",
      "Epoch 2917/10000000\n",
      " - 0s - loss: 1.8441 - val_loss: 1.6354\n",
      "Epoch 2918/10000000\n",
      " - 0s - loss: 1.4318 - val_loss: 1.7877\n",
      "Epoch 2919/10000000\n",
      " - 0s - loss: 1.3174 - val_loss: 1.7808\n",
      "Epoch 2920/10000000\n",
      " - 0s - loss: 1.2819 - val_loss: 2.2757\n",
      "Epoch 2921/10000000\n",
      " - 0s - loss: 2.0185 - val_loss: 2.4325\n",
      "Epoch 2922/10000000\n",
      " - 0s - loss: 2.3058 - val_loss: 1.8756\n",
      "Epoch 2923/10000000\n",
      " - 0s - loss: 1.4975 - val_loss: 1.7289\n",
      "Epoch 2924/10000000\n",
      " - 0s - loss: 1.2469 - val_loss: 1.6524\n",
      "Epoch 2925/10000000\n",
      " - 0s - loss: 1.2286 - val_loss: 1.6308\n",
      "Epoch 2926/10000000\n",
      " - 0s - loss: 1.1900 - val_loss: 1.6834\n",
      "Epoch 2927/10000000\n",
      " - 0s - loss: 1.2365 - val_loss: 1.7037\n",
      "Epoch 2928/10000000\n",
      " - 0s - loss: 1.2640 - val_loss: 1.8274\n",
      "Epoch 2929/10000000\n",
      " - 0s - loss: 1.4737 - val_loss: 2.2565\n",
      "Epoch 2930/10000000\n",
      " - 0s - loss: 2.1426 - val_loss: 2.6056\n",
      "Epoch 2931/10000000\n",
      " - 0s - loss: 2.5200 - val_loss: 1.9239\n",
      "Epoch 2932/10000000\n",
      " - 0s - loss: 1.5277 - val_loss: 1.7396\n",
      "Epoch 2933/10000000\n",
      " - 0s - loss: 1.2773 - val_loss: 1.6526\n",
      "Epoch 02933: early stopping\n"
     ]
    }
   ],
   "source": [
    "# Create model\n",
    "model = Sequential()\n",
    "model.add(Dense(128, activation=\"relu\", input_dim=6))\n",
    "model.add(Dense(32, activation=\"relu\"))\n",
    "model.add(Dense(8, activation=\"relu\"))\n",
    "# Since the regression is performed, a Dense layer containing a single neuron with a linear activation function.\n",
    "# Typically ReLu-based activation are used but since it is performed regression, it is needed a linear activation.\n",
    "model.add(Dense(1, activation=\"linear\"))\n",
    "\n",
    "# Compile model: The model is initialized with the Adam optimizer and then it is compiled.\n",
    "model.compile(loss='mean_squared_error', optimizer=Adam(lr=1e-3, decay=1e-3 / 200))\n",
    "\n",
    "# Patient early stopping\n",
    "es = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=200)\n",
    "\n",
    "# Fit the model\n",
    "history = model.fit(X1, Y1, validation_data=(X2, Y2), epochs=10000000, batch_size=100, verbose=2, callbacks=[es])\n",
    "\n",
    "# Calculate predictions\n",
    "PredTestSet = model.predict(X1)\n",
    "PredValSet = model.predict(X2)\n",
    "\n",
    "# Save predictions\n",
    "numpy.savetxt(\"trainresults.csv\", PredTestSet, delimiter=\",\")\n",
    "numpy.savetxt(\"valresults.csv\", PredValSet, delimiter=\",\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fad781ca110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot training history\n",
    "pyplot.plot(history.history['loss'], label='train')\n",
    "pyplot.plot(history.history['val_loss'], label='test')\n",
    "pyplot.legend()\n",
    "plt.title('Training History'),\n",
    "plt.xlabel('Epoch'),\n",
    "plt.ylabel('Validation Loss')\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Training Set R-Square=', 0.9949287381599607)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fad387f8090>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot actual vs prediction for training set\n",
    "TestResults = numpy.genfromtxt(\"trainresults.csv\", delimiter=\",\")\n",
    "plt.plot(Y1,TestResults,'ro')\n",
    "plt.title('Training Set')\n",
    "plt.xlabel('Actual')\n",
    "plt.ylabel('Predicted')\n",
    "\n",
    "# Compute R-Square value for training set\n",
    "TestR2Value = r2_score(Y1,TestResults)\n",
    "print(\"Training Set R-Square=\", TestR2Value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('Validation Set R-Square=', 0.9916464732839869)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fad35ef78d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot actual vs prediction for validation set\n",
    "ValResults = numpy.genfromtxt(\"valresults.csv\", delimiter=\",\")\n",
    "plt.plot(Y2,ValResults,'ro')\n",
    "plt.title('Validation Set')\n",
    "plt.xlabel('Actual')\n",
    "plt.ylabel('Predicted')\n",
    "\n",
    "# Compute R-Square value for validation set\n",
    "ValR2Value = r2_score(Y2,ValResults)\n",
    "print(\"Validation Set R-Square=\",ValR2Value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}