Predict Beauty.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import keras\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.applications import ResNet50\n",
    "from keras.layers import Dense, Dropout, Activation, Flatten\n",
    "from keras.layers import Convolution2D, MaxPooling2D\n",
    "from keras.utils import np_utils\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.optimizers import SGD, Adam\n",
    "import matplotlib.pyplot as plt\n",
    "import glob, os\n",
    "import PIL.Image\n",
    "\n",
    "import csv\n",
    "import pandas as pd\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "330001it [00:00, 481330.14it/s]\n"
     ]
    }
   ],
   "source": [
    "all_labels = {}\n",
    "label_dis = np.zeros(5)\n",
    "with open(\"/Users/leon/Downloads/SCUT-FBP5500_v2/All_Ratings.csv\") as csv_file:\n",
    "    csv_reader = csv.reader(csv_file, delimiter=';')\n",
    "    line_count = 0\n",
    "    for row in tqdm(csv_reader):\n",
    "        if line_count == 0:\n",
    "            line_count += 1\n",
    "        else:\n",
    "            line_count += 1\n",
    "            all_labels[row[1].split(\".\")[0]] = row[2]\n",
    "            label_dis[int(row[2]) -1] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 19794.  67634. 159328.  62272.  20972.]\n"
     ]
    }
   ],
   "source": [
    "print(label_dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def to_array(i):\n",
    "    out = np.zeros(5)\n",
    "    out[i] = 1\n",
    "    return (i - 2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5491/5491 [00:11<00:00, 484.50it/s]\n"
     ]
    }
   ],
   "source": [
    "path = \"/Users/leon/Downloads/SCUT-FBP5500_v2/Scaled/\"\n",
    "valid_images = \".png\"\n",
    "images = []\n",
    "labels = []\n",
    "for f in tqdm(os.listdir(path)):\n",
    "    if f.endswith(valid_images):\n",
    "        img = PIL.Image.open(path + f).convert('RGB')\n",
    "        images.append(np.array(img).reshape(128,128,3))\n",
    "        labels.append(to_array(int(all_labels[f[:-7]]) - 1))\n",
    "        \n",
    "images = np.array(images)#.reshape(-1,128,128,3)\n",
    "labels = np.array(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]\n",
      "\n",
      " [[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]\n",
      "\n",
      " [[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]\n",
      "\n",
      " [[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]\n",
      "\n",
      " [[255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  ...\n",
      "  [255 255 255]\n",
      "  [255 255 255]\n",
      "  [255 255 255]]]\n"
     ]
    }
   ],
   "source": [
    "print(images[3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "(trainX, testX, trainY, testY) = train_test_split(images, labels, test_size=0.25, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(keras.layers.Conv2D(4, [3,3], activation='relu', padding='same'))\n",
    "model.add(keras.layers.Conv2D(8, [3,3], activation='relu', padding='same'))\n",
    "model.add(keras.layers.Conv2D(16, [3,3], activation='relu', padding='same'))\n",
    "\n",
    "model.add(keras.layers.BatchNormalization())\n",
    "model.add(keras.layers.Dropout(0.15))\n",
    "model.add(keras.layers.Activation('relu'))\n",
    "model.add(keras.layers.Flatten())\n",
    "model.add(keras.layers.Dense(50))\n",
    "model.add(keras.layers.Dense(1, activation=tf.nn.sigmoid))\n",
    "#model.add(keras.layers.Activation('sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt = keras.optimizers.RMSprop(0.001)\n",
    "model.compile(loss=\"mse\", optimizer=opt, metrics=['mae', 'mse'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4117 samples, validate on 1373 samples\n",
      "Epoch 1/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.6005 - mae: 0.5159 - mse: 0.6005 - val_loss: 0.5871 - val_mae: 0.5295 - val_mse: 0.5871\n",
      "Epoch 2/5000\n",
      "4117/4117 [==============================] - 3s 749us/step - loss: 0.6076 - mae: 0.5160 - mse: 0.6076 - val_loss: 0.6365 - val_mae: 0.5607 - val_mse: 0.6365\n",
      "Epoch 3/5000\n",
      "4117/4117 [==============================] - 3s 748us/step - loss: 0.6087 - mae: 0.5184 - mse: 0.6087 - val_loss: 0.6377 - val_mae: 0.5403 - val_mse: 0.6377\n",
      "Epoch 4/5000\n",
      "4117/4117 [==============================] - 3s 780us/step - loss: 0.5936 - mae: 0.5099 - mse: 0.5936 - val_loss: 0.7021 - val_mae: 0.5677 - val_mse: 0.7021\n",
      "Epoch 5/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.6007 - mae: 0.5100 - mse: 0.6007 - val_loss: 0.6312 - val_mae: 0.5537 - val_mse: 0.6312\n",
      "Epoch 6/5000\n",
      "4117/4117 [==============================] - 3s 789us/step - loss: 0.5877 - mae: 0.5034 - mse: 0.5877 - val_loss: 0.7824 - val_mae: 0.6494 - val_mse: 0.7824\n",
      "Epoch 7/5000\n",
      "4117/4117 [==============================] - 3s 759us/step - loss: 0.5873 - mae: 0.5069 - mse: 0.5873 - val_loss: 0.5910 - val_mae: 0.5217 - val_mse: 0.5910\n",
      "Epoch 8/5000\n",
      "4117/4117 [==============================] - 3s 781us/step - loss: 0.5861 - mae: 0.4995 - mse: 0.5861 - val_loss: 0.6538 - val_mae: 0.5667 - val_mse: 0.6538\n",
      "Epoch 9/5000\n",
      "4117/4117 [==============================] - 3s 762us/step - loss: 0.5803 - mae: 0.4972 - mse: 0.5803 - val_loss: 0.7427 - val_mae: 0.5792 - val_mse: 0.7427\n",
      "Epoch 10/5000\n",
      "4117/4117 [==============================] - 3s 758us/step - loss: 0.5841 - mae: 0.5011 - mse: 0.5841 - val_loss: 0.6036 - val_mae: 0.5268 - val_mse: 0.6036\n",
      "Epoch 11/5000\n",
      "4117/4117 [==============================] - 3s 759us/step - loss: 0.5811 - mae: 0.5027 - mse: 0.5811 - val_loss: 0.6586 - val_mae: 0.5762 - val_mse: 0.6586\n",
      "Epoch 12/5000\n",
      "4117/4117 [==============================] - 3s 755us/step - loss: 0.5786 - mae: 0.4955 - mse: 0.5786 - val_loss: 0.6318 - val_mae: 0.5541 - val_mse: 0.6318\n",
      "Epoch 13/5000\n",
      "4117/4117 [==============================] - 3s 757us/step - loss: 0.5784 - mae: 0.4986 - mse: 0.5784 - val_loss: 0.7354 - val_mae: 0.5762 - val_mse: 0.7354\n",
      "Epoch 14/5000\n",
      "4117/4117 [==============================] - 3s 759us/step - loss: 0.5860 - mae: 0.5028 - mse: 0.5860 - val_loss: 0.6546 - val_mae: 0.5686 - val_mse: 0.6546\n",
      "Epoch 15/5000\n",
      "4117/4117 [==============================] - 3s 757us/step - loss: 0.5696 - mae: 0.4893 - mse: 0.5696 - val_loss: 0.5869 - val_mae: 0.5105 - val_mse: 0.5869\n",
      "Epoch 16/5000\n",
      "4117/4117 [==============================] - 3s 790us/step - loss: 0.5689 - mae: 0.4918 - mse: 0.5689 - val_loss: 0.6128 - val_mae: 0.5515 - val_mse: 0.6128\n",
      "Epoch 17/5000\n",
      "4117/4117 [==============================] - 3s 761us/step - loss: 0.5663 - mae: 0.4904 - mse: 0.5663 - val_loss: 0.5907 - val_mae: 0.5234 - val_mse: 0.5907\n",
      "Epoch 18/5000\n",
      "4117/4117 [==============================] - 3s 789us/step - loss: 0.5714 - mae: 0.4960 - mse: 0.5714 - val_loss: 0.8930 - val_mae: 0.7180 - val_mse: 0.8930\n",
      "Epoch 19/5000\n",
      "4117/4117 [==============================] - 3s 764us/step - loss: 0.5667 - mae: 0.4892 - mse: 0.5667 - val_loss: 0.6460 - val_mae: 0.5387 - val_mse: 0.6460\n",
      "Epoch 20/5000\n",
      "4117/4117 [==============================] - 3s 775us/step - loss: 0.5626 - mae: 0.4876 - mse: 0.5626 - val_loss: 0.6200 - val_mae: 0.5526 - val_mse: 0.6200\n",
      "Epoch 21/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.5666 - mae: 0.4894 - mse: 0.5666 - val_loss: 0.5960 - val_mae: 0.5236 - val_mse: 0.5960\n",
      "Epoch 22/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.5645 - mae: 0.4870 - mse: 0.5645 - val_loss: 0.5987 - val_mae: 0.5132 - val_mse: 0.5987\n",
      "Epoch 23/5000\n",
      "4117/4117 [==============================] - 3s 788us/step - loss: 0.5580 - mae: 0.4863 - mse: 0.5580 - val_loss: 0.5962 - val_mae: 0.5161 - val_mse: 0.5962\n",
      "Epoch 24/5000\n",
      "4117/4117 [==============================] - 3s 789us/step - loss: 0.5611 - mae: 0.4840 - mse: 0.5611 - val_loss: 0.6177 - val_mae: 0.5419 - val_mse: 0.6177\n",
      "Epoch 25/5000\n",
      "4117/4117 [==============================] - 3s 795us/step - loss: 0.5616 - mae: 0.4885 - mse: 0.5616 - val_loss: 0.6625 - val_mae: 0.5787 - val_mse: 0.6625\n",
      "Epoch 26/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.5599 - mae: 0.4820 - mse: 0.5599 - val_loss: 0.6241 - val_mae: 0.5236 - val_mse: 0.6241\n",
      "Epoch 27/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.5609 - mae: 0.4865 - mse: 0.5609 - val_loss: 0.5878 - val_mae: 0.5162 - val_mse: 0.5878\n",
      "Epoch 28/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.5524 - mae: 0.4809 - mse: 0.5524 - val_loss: 0.5897 - val_mae: 0.5121 - val_mse: 0.5897\n",
      "Epoch 29/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.5623 - mae: 0.4851 - mse: 0.5623 - val_loss: 0.8391 - val_mae: 0.6947 - val_mse: 0.8391\n",
      "Epoch 30/5000\n",
      "4117/4117 [==============================] - 3s 839us/step - loss: 0.5577 - mae: 0.4844 - mse: 0.5577 - val_loss: 0.5997 - val_mae: 0.5213 - val_mse: 0.5997\n",
      "Epoch 31/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.5510 - mae: 0.4802 - mse: 0.5510 - val_loss: 0.6571 - val_mae: 0.5439 - val_mse: 0.6571\n",
      "Epoch 32/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.5551 - mae: 0.4830 - mse: 0.5551 - val_loss: 0.6217 - val_mae: 0.5329 - val_mse: 0.6217\n",
      "Epoch 33/5000\n",
      "4117/4117 [==============================] - 4s 864us/step - loss: 0.5533 - mae: 0.4805 - mse: 0.5533 - val_loss: 0.5864 - val_mae: 0.5160 - val_mse: 0.5864\n",
      "Epoch 34/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.5506 - mae: 0.4810 - mse: 0.5506 - val_loss: 0.5863 - val_mae: 0.5157 - val_mse: 0.5863\n",
      "Epoch 35/5000\n",
      "4117/4117 [==============================] - 4s 859us/step - loss: 0.5486 - mae: 0.4783 - mse: 0.5486 - val_loss: 0.5892 - val_mae: 0.5147 - val_mse: 0.5892\n",
      "Epoch 36/5000\n",
      "4117/4117 [==============================] - 4s 867us/step - loss: 0.5483 - mae: 0.4766 - mse: 0.5483 - val_loss: 0.6108 - val_mae: 0.5201 - val_mse: 0.6108\n",
      "Epoch 37/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.5550 - mae: 0.4846 - mse: 0.5550 - val_loss: 0.6093 - val_mae: 0.5393 - val_mse: 0.6093\n",
      "Epoch 38/5000\n",
      "4117/4117 [==============================] - 4s 875us/step - loss: 0.5460 - mae: 0.4802 - mse: 0.5460 - val_loss: 0.5969 - val_mae: 0.5219 - val_mse: 0.5969\n",
      "Epoch 39/5000\n",
      "4117/4117 [==============================] - 4s 881us/step - loss: 0.5372 - mae: 0.4721 - mse: 0.5372 - val_loss: 0.6642 - val_mae: 0.5403 - val_mse: 0.6642\n",
      "Epoch 40/5000\n",
      "4117/4117 [==============================] - 4s 869us/step - loss: 0.5452 - mae: 0.4746 - mse: 0.5452 - val_loss: 0.6708 - val_mae: 0.5752 - val_mse: 0.6708\n",
      "Epoch 41/5000\n",
      "4117/4117 [==============================] - 4s 868us/step - loss: 0.5400 - mae: 0.4718 - mse: 0.5400 - val_loss: 0.6307 - val_mae: 0.5280 - val_mse: 0.6307\n",
      "Epoch 42/5000\n",
      "4117/4117 [==============================] - 4s 895us/step - loss: 0.5413 - mae: 0.4746 - mse: 0.5413 - val_loss: 0.6095 - val_mae: 0.5379 - val_mse: 0.6095\n",
      "Epoch 43/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5410 - mae: 0.4739 - mse: 0.5410 - val_loss: 0.6346 - val_mae: 0.5291 - val_mse: 0.6346\n",
      "Epoch 44/5000\n",
      "4117/4117 [==============================] - 4s 885us/step - loss: 0.5356 - mae: 0.4682 - mse: 0.5356 - val_loss: 0.6230 - val_mae: 0.5502 - val_mse: 0.6230\n",
      "Epoch 45/5000\n",
      "4117/4117 [==============================] - 4s 893us/step - loss: 0.5455 - mae: 0.4720 - mse: 0.5455 - val_loss: 0.6199 - val_mae: 0.5250 - val_mse: 0.6199\n",
      "Epoch 46/5000\n",
      "4117/4117 [==============================] - 4s 907us/step - loss: 0.5406 - mae: 0.4735 - mse: 0.5406 - val_loss: 0.5779 - val_mae: 0.5136 - val_mse: 0.5779\n",
      "Epoch 47/5000\n",
      "4117/4117 [==============================] - 4s 909us/step - loss: 0.5334 - mae: 0.4684 - mse: 0.5334 - val_loss: 0.6115 - val_mae: 0.5317 - val_mse: 0.6115\n",
      "Epoch 48/5000\n",
      "4117/4117 [==============================] - 4s 887us/step - loss: 0.5369 - mae: 0.4707 - mse: 0.5369 - val_loss: 0.6040 - val_mae: 0.5143 - val_mse: 0.6040\n",
      "Epoch 49/5000\n",
      "4117/4117 [==============================] - 4s 925us/step - loss: 0.5402 - mae: 0.4721 - mse: 0.5402 - val_loss: 0.6034 - val_mae: 0.5158 - val_mse: 0.6034\n",
      "Epoch 50/5000\n",
      "4117/4117 [==============================] - 4s 948us/step - loss: 0.5351 - mae: 0.4679 - mse: 0.5351 - val_loss: 0.5958 - val_mae: 0.5196 - val_mse: 0.5958\n",
      "Epoch 51/5000\n",
      "4117/4117 [==============================] - 4s 927us/step - loss: 0.5310 - mae: 0.4655 - mse: 0.5310 - val_loss: 0.6043 - val_mae: 0.5138 - val_mse: 0.6043\n",
      "Epoch 52/5000\n",
      "4117/4117 [==============================] - 4s 920us/step - loss: 0.5370 - mae: 0.4741 - mse: 0.5370 - val_loss: 0.6060 - val_mae: 0.5146 - val_mse: 0.6060\n",
      "Epoch 53/5000\n",
      "4117/4117 [==============================] - 4s 892us/step - loss: 0.5315 - mae: 0.4639 - mse: 0.5315 - val_loss: 0.6047 - val_mae: 0.5130 - val_mse: 0.6047\n",
      "Epoch 54/5000\n",
      "4117/4117 [==============================] - 4s 907us/step - loss: 0.5346 - mae: 0.4674 - mse: 0.5346 - val_loss: 0.5868 - val_mae: 0.5097 - val_mse: 0.5868\n",
      "Epoch 55/5000\n",
      "4117/4117 [==============================] - 4s 962us/step - loss: 0.5317 - mae: 0.4672 - mse: 0.5317 - val_loss: 0.5844 - val_mae: 0.5173 - val_mse: 0.5844\n",
      "Epoch 56/5000\n",
      "4117/4117 [==============================] - 4s 908us/step - loss: 0.5301 - mae: 0.4660 - mse: 0.5301 - val_loss: 0.5923 - val_mae: 0.5187 - val_mse: 0.5923\n",
      "Epoch 57/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5344 - mae: 0.4703 - mse: 0.5344 - val_loss: 0.5970 - val_mae: 0.5157 - val_mse: 0.5970\n",
      "Epoch 58/5000\n",
      "4117/4117 [==============================] - 4s 889us/step - loss: 0.5297 - mae: 0.4617 - mse: 0.5297 - val_loss: 0.6291 - val_mae: 0.5333 - val_mse: 0.6291\n",
      "Epoch 59/5000\n",
      "4117/4117 [==============================] - 4s 919us/step - loss: 0.5269 - mae: 0.4624 - mse: 0.5269 - val_loss: 0.7057 - val_mae: 0.5917 - val_mse: 0.7057\n",
      "Epoch 60/5000\n",
      "4117/4117 [==============================] - 4s 886us/step - loss: 0.5253 - mae: 0.4604 - mse: 0.5253 - val_loss: 0.5802 - val_mae: 0.5065 - val_mse: 0.5802\n",
      "Epoch 61/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5316 - mae: 0.4662 - mse: 0.5316 - val_loss: 0.5878 - val_mae: 0.5205 - val_mse: 0.5878\n",
      "Epoch 62/5000\n",
      "4117/4117 [==============================] - 4s 876us/step - loss: 0.5303 - mae: 0.4644 - mse: 0.5303 - val_loss: 0.5999 - val_mae: 0.5202 - val_mse: 0.5999\n",
      "Epoch 63/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5280 - mae: 0.4635 - mse: 0.5280 - val_loss: 0.5951 - val_mae: 0.5172 - val_mse: 0.5951\n",
      "Epoch 64/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.5241 - mae: 0.4578 - mse: 0.5241 - val_loss: 0.6090 - val_mae: 0.5181 - val_mse: 0.6090\n",
      "Epoch 65/5000\n",
      "4117/4117 [==============================] - 4s 886us/step - loss: 0.5243 - mae: 0.4604 - mse: 0.5243 - val_loss: 0.6018 - val_mae: 0.5218 - val_mse: 0.6018\n",
      "Epoch 66/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.5195 - mae: 0.4550 - mse: 0.5195 - val_loss: 0.5958 - val_mae: 0.5187 - val_mse: 0.5958\n",
      "Epoch 67/5000\n",
      "4117/4117 [==============================] - 4s 874us/step - loss: 0.5254 - mae: 0.4595 - mse: 0.5254 - val_loss: 0.5987 - val_mae: 0.5162 - val_mse: 0.5987\n",
      "Epoch 68/5000\n",
      "4117/4117 [==============================] - 4s 887us/step - loss: 0.5292 - mae: 0.4614 - mse: 0.5292 - val_loss: 0.6055 - val_mae: 0.5275 - val_mse: 0.6055\n",
      "Epoch 69/5000\n",
      "4117/4117 [==============================] - 4s 887us/step - loss: 0.5325 - mae: 0.4656 - mse: 0.5325 - val_loss: 0.6015 - val_mae: 0.5257 - val_mse: 0.6015\n",
      "Epoch 70/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.5365 - mae: 0.4649 - mse: 0.5365 - val_loss: 0.5951 - val_mae: 0.5239 - val_mse: 0.5951\n",
      "Epoch 71/5000\n",
      "4117/4117 [==============================] - 4s 881us/step - loss: 0.5183 - mae: 0.4562 - mse: 0.5183 - val_loss: 0.6163 - val_mae: 0.5218 - val_mse: 0.6163\n",
      "Epoch 72/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.5201 - mae: 0.4531 - mse: 0.5201 - val_loss: 0.6608 - val_mae: 0.5607 - val_mse: 0.6608\n",
      "Epoch 73/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.5225 - mae: 0.4545 - mse: 0.5225 - val_loss: 0.6064 - val_mae: 0.5172 - val_mse: 0.6064\n",
      "Epoch 74/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.5234 - mae: 0.4565 - mse: 0.5234 - val_loss: 0.6212 - val_mae: 0.5331 - val_mse: 0.6212\n",
      "Epoch 75/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.5192 - mae: 0.4577 - mse: 0.5192 - val_loss: 0.5967 - val_mae: 0.5141 - val_mse: 0.5967\n",
      "Epoch 76/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5177 - mae: 0.4510 - mse: 0.5177 - val_loss: 0.6039 - val_mae: 0.5217 - val_mse: 0.6039\n",
      "Epoch 77/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.5151 - mae: 0.4512 - mse: 0.5151 - val_loss: 0.5967 - val_mae: 0.5133 - val_mse: 0.5967\n",
      "Epoch 78/5000\n",
      "4117/4117 [==============================] - 4s 888us/step - loss: 0.5164 - mae: 0.4539 - mse: 0.5164 - val_loss: 0.6016 - val_mae: 0.5234 - val_mse: 0.6016\n",
      "Epoch 79/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.5201 - mae: 0.4547 - mse: 0.5201 - val_loss: 0.6520 - val_mae: 0.5400 - val_mse: 0.6520\n",
      "Epoch 80/5000\n",
      "4117/4117 [==============================] - 4s 859us/step - loss: 0.5197 - mae: 0.4528 - mse: 0.5197 - val_loss: 0.5865 - val_mae: 0.5117 - val_mse: 0.5865\n",
      "Epoch 81/5000\n",
      "4117/4117 [==============================] - 4s 891us/step - loss: 0.5197 - mae: 0.4530 - mse: 0.5197 - val_loss: 0.6120 - val_mae: 0.5239 - val_mse: 0.6120\n",
      "Epoch 82/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.5147 - mae: 0.4515 - mse: 0.5147 - val_loss: 0.6206 - val_mae: 0.5239 - val_mse: 0.6206\n",
      "Epoch 83/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5181 - mae: 0.4509 - mse: 0.5181 - val_loss: 0.6024 - val_mae: 0.5205 - val_mse: 0.6024\n",
      "Epoch 84/5000\n",
      "4117/4117 [==============================] - 4s 882us/step - loss: 0.5142 - mae: 0.4511 - mse: 0.5142 - val_loss: 0.6047 - val_mae: 0.5181 - val_mse: 0.6047\n",
      "Epoch 85/5000\n",
      "4117/4117 [==============================] - 4s 867us/step - loss: 0.5169 - mae: 0.4521 - mse: 0.5169 - val_loss: 0.6238 - val_mae: 0.5399 - val_mse: 0.6238\n",
      "Epoch 86/5000\n",
      "4117/4117 [==============================] - 4s 858us/step - loss: 0.5149 - mae: 0.4513 - mse: 0.5149 - val_loss: 0.5991 - val_mae: 0.5219 - val_mse: 0.5991\n",
      "Epoch 87/5000\n",
      "4117/4117 [==============================] - 4s 865us/step - loss: 0.5181 - mae: 0.4534 - mse: 0.5181 - val_loss: 0.5921 - val_mae: 0.5129 - val_mse: 0.5921\n",
      "Epoch 88/5000\n",
      "4117/4117 [==============================] - 4s 881us/step - loss: 0.5116 - mae: 0.4478 - mse: 0.5116 - val_loss: 0.5896 - val_mae: 0.5148 - val_mse: 0.5896\n",
      "Epoch 89/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.5111 - mae: 0.4473 - mse: 0.5111 - val_loss: 0.6095 - val_mae: 0.5207 - val_mse: 0.6095\n",
      "Epoch 90/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.5136 - mae: 0.4482 - mse: 0.5136 - val_loss: 0.6426 - val_mae: 0.5323 - val_mse: 0.6426\n",
      "Epoch 91/5000\n",
      "4117/4117 [==============================] - 4s 862us/step - loss: 0.5125 - mae: 0.4474 - mse: 0.5125 - val_loss: 0.6071 - val_mae: 0.5145 - val_mse: 0.6071\n",
      "Epoch 92/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.5137 - mae: 0.4469 - mse: 0.5137 - val_loss: 0.6618 - val_mae: 0.5635 - val_mse: 0.6618\n",
      "Epoch 93/5000\n",
      "4117/4117 [==============================] - 4s 859us/step - loss: 0.5146 - mae: 0.4491 - mse: 0.5146 - val_loss: 0.6158 - val_mae: 0.5193 - val_mse: 0.6158\n",
      "Epoch 94/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5166 - mae: 0.4502 - mse: 0.5166 - val_loss: 0.5984 - val_mae: 0.5203 - val_mse: 0.5984\n",
      "Epoch 95/5000\n",
      "4117/4117 [==============================] - 4s 869us/step - loss: 0.5130 - mae: 0.4493 - mse: 0.5130 - val_loss: 0.6125 - val_mae: 0.5205 - val_mse: 0.6125\n",
      "Epoch 96/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.5202 - mae: 0.4548 - mse: 0.5202 - val_loss: 0.6027 - val_mae: 0.5163 - val_mse: 0.6027\n",
      "Epoch 97/5000\n",
      "4117/4117 [==============================] - 4s 853us/step - loss: 0.5134 - mae: 0.4481 - mse: 0.5134 - val_loss: 0.6001 - val_mae: 0.5118 - val_mse: 0.6001\n",
      "Epoch 98/5000\n",
      "4117/4117 [==============================] - 4s 850us/step - loss: 0.5106 - mae: 0.4451 - mse: 0.5106 - val_loss: 0.6156 - val_mae: 0.5397 - val_mse: 0.6156\n",
      "Epoch 99/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.5168 - mae: 0.4507 - mse: 0.5168 - val_loss: 0.6266 - val_mae: 0.5209 - val_mse: 0.6266\n",
      "Epoch 100/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.5162 - mae: 0.4484 - mse: 0.5162 - val_loss: 0.6124 - val_mae: 0.5263 - val_mse: 0.6124\n",
      "Epoch 101/5000\n",
      "4117/4117 [==============================] - 4s 854us/step - loss: 0.5167 - mae: 0.4487 - mse: 0.5167 - val_loss: 0.6045 - val_mae: 0.5141 - val_mse: 0.6045\n",
      "Epoch 102/5000\n",
      "4117/4117 [==============================] - 4s 859us/step - loss: 0.5146 - mae: 0.4485 - mse: 0.5146 - val_loss: 0.6096 - val_mae: 0.5211 - val_mse: 0.6096\n",
      "Epoch 103/5000\n",
      "4117/4117 [==============================] - 4s 888us/step - loss: 0.5106 - mae: 0.4456 - mse: 0.5106 - val_loss: 0.6431 - val_mae: 0.5372 - val_mse: 0.6431\n",
      "Epoch 104/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5198 - mae: 0.4500 - mse: 0.5198 - val_loss: 0.6025 - val_mae: 0.5200 - val_mse: 0.6025\n",
      "Epoch 105/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.5128 - mae: 0.4470 - mse: 0.5128 - val_loss: 0.6002 - val_mae: 0.5134 - val_mse: 0.6002\n",
      "Epoch 106/5000\n",
      "4117/4117 [==============================] - 4s 863us/step - loss: 0.5032 - mae: 0.4382 - mse: 0.5032 - val_loss: 0.6073 - val_mae: 0.5238 - val_mse: 0.6073\n",
      "Epoch 107/5000\n",
      "4117/4117 [==============================] - 4s 874us/step - loss: 0.5118 - mae: 0.4457 - mse: 0.5118 - val_loss: 0.5965 - val_mae: 0.5143 - val_mse: 0.5965\n",
      "Epoch 108/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.5101 - mae: 0.4429 - mse: 0.5101 - val_loss: 0.5959 - val_mae: 0.5161 - val_mse: 0.5959\n",
      "Epoch 109/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.5081 - mae: 0.4432 - mse: 0.5081 - val_loss: 0.6073 - val_mae: 0.5252 - val_mse: 0.6073\n",
      "Epoch 110/5000\n",
      "4117/4117 [==============================] - 4s 870us/step - loss: 0.5084 - mae: 0.4438 - mse: 0.5084 - val_loss: 0.5971 - val_mae: 0.5189 - val_mse: 0.5971\n",
      "Epoch 111/5000\n",
      "4117/4117 [==============================] - 4s 858us/step - loss: 0.5130 - mae: 0.4458 - mse: 0.5130 - val_loss: 0.6042 - val_mae: 0.5259 - val_mse: 0.6042\n",
      "Epoch 112/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.5090 - mae: 0.4452 - mse: 0.5090 - val_loss: 0.6359 - val_mae: 0.5331 - val_mse: 0.6359\n",
      "Epoch 113/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.5106 - mae: 0.4456 - mse: 0.5106 - val_loss: 0.5933 - val_mae: 0.5159 - val_mse: 0.5933\n",
      "Epoch 114/5000\n",
      "4117/4117 [==============================] - 4s 869us/step - loss: 0.5105 - mae: 0.4459 - mse: 0.5105 - val_loss: 0.6139 - val_mae: 0.5251 - val_mse: 0.6139\n",
      "Epoch 115/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5062 - mae: 0.4397 - mse: 0.5062 - val_loss: 0.6139 - val_mae: 0.5313 - val_mse: 0.6139\n",
      "Epoch 116/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.5060 - mae: 0.4428 - mse: 0.5060 - val_loss: 0.6308 - val_mae: 0.5258 - val_mse: 0.6308\n",
      "Epoch 117/5000\n",
      "4117/4117 [==============================] - 4s 879us/step - loss: 0.5054 - mae: 0.4395 - mse: 0.5054 - val_loss: 0.6005 - val_mae: 0.5185 - val_mse: 0.6005\n",
      "Epoch 118/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.5080 - mae: 0.4441 - mse: 0.5080 - val_loss: 0.6249 - val_mae: 0.5257 - val_mse: 0.6249\n",
      "Epoch 119/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.5032 - mae: 0.4407 - mse: 0.5032 - val_loss: 0.6097 - val_mae: 0.5265 - val_mse: 0.6097\n",
      "Epoch 120/5000\n",
      "4117/4117 [==============================] - 4s 891us/step - loss: 0.5088 - mae: 0.4398 - mse: 0.5088 - val_loss: 0.6304 - val_mae: 0.5394 - val_mse: 0.6305\n",
      "Epoch 121/5000\n",
      "4117/4117 [==============================] - 4s 865us/step - loss: 0.5070 - mae: 0.4418 - mse: 0.5070 - val_loss: 0.6103 - val_mae: 0.5225 - val_mse: 0.6103\n",
      "Epoch 122/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.5047 - mae: 0.4399 - mse: 0.5047 - val_loss: 0.6062 - val_mae: 0.5234 - val_mse: 0.6062\n",
      "Epoch 123/5000\n",
      "4117/4117 [==============================] - 4s 862us/step - loss: 0.5051 - mae: 0.4393 - mse: 0.5051 - val_loss: 0.6173 - val_mae: 0.5352 - val_mse: 0.6173\n",
      "Epoch 124/5000\n",
      "4117/4117 [==============================] - 4s 876us/step - loss: 0.5043 - mae: 0.4387 - mse: 0.5043 - val_loss: 0.6184 - val_mae: 0.5240 - val_mse: 0.6184\n",
      "Epoch 125/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.5050 - mae: 0.4401 - mse: 0.5050 - val_loss: 0.5963 - val_mae: 0.5146 - val_mse: 0.5963\n",
      "Epoch 126/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5024 - mae: 0.4380 - mse: 0.5024 - val_loss: 0.6138 - val_mae: 0.5250 - val_mse: 0.6138\n",
      "Epoch 127/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.5012 - mae: 0.4380 - mse: 0.5012 - val_loss: 0.6095 - val_mae: 0.5170 - val_mse: 0.6095\n",
      "Epoch 128/5000\n",
      "4117/4117 [==============================] - 4s 854us/step - loss: 0.5110 - mae: 0.4399 - mse: 0.5110 - val_loss: 0.6107 - val_mae: 0.5195 - val_mse: 0.6107\n",
      "Epoch 129/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.5022 - mae: 0.4391 - mse: 0.5022 - val_loss: 0.6227 - val_mae: 0.5309 - val_mse: 0.6227\n",
      "Epoch 130/5000\n",
      "4117/4117 [==============================] - 4s 890us/step - loss: 0.5040 - mae: 0.4393 - mse: 0.5040 - val_loss: 0.6234 - val_mae: 0.5356 - val_mse: 0.6234\n",
      "Epoch 131/5000\n",
      "4117/4117 [==============================] - 4s 863us/step - loss: 0.5061 - mae: 0.4402 - mse: 0.5061 - val_loss: 0.6316 - val_mae: 0.5258 - val_mse: 0.6316\n",
      "Epoch 132/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.5052 - mae: 0.4405 - mse: 0.5052 - val_loss: 0.6418 - val_mae: 0.5498 - val_mse: 0.6418\n",
      "Epoch 133/5000\n",
      "4117/4117 [==============================] - 4s 858us/step - loss: 0.5029 - mae: 0.4387 - mse: 0.5029 - val_loss: 0.6342 - val_mae: 0.5232 - val_mse: 0.6342\n",
      "Epoch 134/5000\n",
      "4117/4117 [==============================] - 4s 875us/step - loss: 0.5057 - mae: 0.4412 - mse: 0.5057 - val_loss: 0.6128 - val_mae: 0.5226 - val_mse: 0.6128\n",
      "Epoch 135/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.5026 - mae: 0.4360 - mse: 0.5026 - val_loss: 0.6329 - val_mae: 0.5343 - val_mse: 0.6329\n",
      "Epoch 136/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.5022 - mae: 0.4355 - mse: 0.5022 - val_loss: 0.6475 - val_mae: 0.5518 - val_mse: 0.6475\n",
      "Epoch 137/5000\n",
      "4117/4117 [==============================] - 4s 879us/step - loss: 0.5060 - mae: 0.4383 - mse: 0.5060 - val_loss: 0.5951 - val_mae: 0.5123 - val_mse: 0.5951\n",
      "Epoch 138/5000\n",
      "4117/4117 [==============================] - 4s 888us/step - loss: 0.5073 - mae: 0.4375 - mse: 0.5073 - val_loss: 0.6065 - val_mae: 0.5165 - val_mse: 0.6065\n",
      "Epoch 139/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.5042 - mae: 0.4354 - mse: 0.5042 - val_loss: 0.6011 - val_mae: 0.5212 - val_mse: 0.6011\n",
      "Epoch 140/5000\n",
      "4117/4117 [==============================] - 4s 867us/step - loss: 0.5065 - mae: 0.4392 - mse: 0.5065 - val_loss: 0.6044 - val_mae: 0.5197 - val_mse: 0.6044\n",
      "Epoch 141/5000\n",
      "4117/4117 [==============================] - 4s 876us/step - loss: 0.5040 - mae: 0.4378 - mse: 0.5040 - val_loss: 0.6165 - val_mae: 0.5204 - val_mse: 0.6165\n",
      "Epoch 142/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4996 - mae: 0.4356 - mse: 0.4996 - val_loss: 0.6348 - val_mae: 0.5264 - val_mse: 0.6348\n",
      "Epoch 143/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.5044 - mae: 0.4375 - mse: 0.5044 - val_loss: 0.6456 - val_mae: 0.5400 - val_mse: 0.6456\n",
      "Epoch 144/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.5005 - mae: 0.4352 - mse: 0.5005 - val_loss: 0.6515 - val_mae: 0.5363 - val_mse: 0.6515\n",
      "Epoch 145/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4989 - mae: 0.4340 - mse: 0.4989 - val_loss: 0.6007 - val_mae: 0.5153 - val_mse: 0.6007\n",
      "Epoch 146/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4981 - mae: 0.4329 - mse: 0.4981 - val_loss: 0.6214 - val_mae: 0.5278 - val_mse: 0.6214\n",
      "Epoch 147/5000\n",
      "4117/4117 [==============================] - 4s 874us/step - loss: 0.5048 - mae: 0.4374 - mse: 0.5048 - val_loss: 0.6140 - val_mae: 0.5229 - val_mse: 0.6140\n",
      "Epoch 148/5000\n",
      "4117/4117 [==============================] - 4s 863us/step - loss: 0.5031 - mae: 0.4350 - mse: 0.5031 - val_loss: 0.6239 - val_mae: 0.5364 - val_mse: 0.6239\n",
      "Epoch 149/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4999 - mae: 0.4355 - mse: 0.4999 - val_loss: 0.6008 - val_mae: 0.5132 - val_mse: 0.6008\n",
      "Epoch 150/5000\n",
      "4117/4117 [==============================] - 4s 860us/step - loss: 0.4971 - mae: 0.4321 - mse: 0.4971 - val_loss: 0.6030 - val_mae: 0.5155 - val_mse: 0.6030\n",
      "Epoch 151/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.5012 - mae: 0.4346 - mse: 0.5012 - val_loss: 0.6138 - val_mae: 0.5225 - val_mse: 0.6138\n",
      "Epoch 152/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.5008 - mae: 0.4339 - mse: 0.5008 - val_loss: 0.6106 - val_mae: 0.5251 - val_mse: 0.6106\n",
      "Epoch 153/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4998 - mae: 0.4360 - mse: 0.4998 - val_loss: 0.6123 - val_mae: 0.5186 - val_mse: 0.6123\n",
      "Epoch 154/5000\n",
      "4117/4117 [==============================] - 4s 893us/step - loss: 0.4998 - mae: 0.4330 - mse: 0.4998 - val_loss: 0.6080 - val_mae: 0.5182 - val_mse: 0.6080\n",
      "Epoch 155/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.4966 - mae: 0.4318 - mse: 0.4966 - val_loss: 0.6196 - val_mae: 0.5291 - val_mse: 0.6196\n",
      "Epoch 156/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.5024 - mae: 0.4350 - mse: 0.5024 - val_loss: 0.5990 - val_mae: 0.5184 - val_mse: 0.5990\n",
      "Epoch 157/5000\n",
      "4117/4117 [==============================] - 4s 878us/step - loss: 0.5027 - mae: 0.4345 - mse: 0.5027 - val_loss: 0.6441 - val_mae: 0.5371 - val_mse: 0.6441\n",
      "Epoch 158/5000\n",
      "4117/4117 [==============================] - 4s 864us/step - loss: 0.4978 - mae: 0.4308 - mse: 0.4978 - val_loss: 0.6100 - val_mae: 0.5189 - val_mse: 0.6100\n",
      "Epoch 159/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.5011 - mae: 0.4331 - mse: 0.5011 - val_loss: 0.5993 - val_mae: 0.5165 - val_mse: 0.5993\n",
      "Epoch 160/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.5010 - mae: 0.4347 - mse: 0.5010 - val_loss: 0.6148 - val_mae: 0.5181 - val_mse: 0.6148\n",
      "Epoch 161/5000\n",
      "4117/4117 [==============================] - 4s 874us/step - loss: 0.5002 - mae: 0.4348 - mse: 0.5002 - val_loss: 0.6142 - val_mae: 0.5259 - val_mse: 0.6142\n",
      "Epoch 162/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.5009 - mae: 0.4342 - mse: 0.5009 - val_loss: 0.6153 - val_mae: 0.5249 - val_mse: 0.6153\n",
      "Epoch 163/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.5007 - mae: 0.4353 - mse: 0.5007 - val_loss: 0.6428 - val_mae: 0.5370 - val_mse: 0.6428\n",
      "Epoch 164/5000\n",
      "4117/4117 [==============================] - 4s 882us/step - loss: 0.4991 - mae: 0.4317 - mse: 0.4991 - val_loss: 0.6038 - val_mae: 0.5171 - val_mse: 0.6038\n",
      "Epoch 165/5000\n",
      "4117/4117 [==============================] - 4s 854us/step - loss: 0.5028 - mae: 0.4352 - mse: 0.5028 - val_loss: 0.6175 - val_mae: 0.5235 - val_mse: 0.6175\n",
      "Epoch 166/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.5025 - mae: 0.4355 - mse: 0.5025 - val_loss: 0.6149 - val_mae: 0.5189 - val_mse: 0.6149\n",
      "Epoch 167/5000\n",
      "4117/4117 [==============================] - 4s 867us/step - loss: 0.4965 - mae: 0.4318 - mse: 0.4965 - val_loss: 0.6127 - val_mae: 0.5208 - val_mse: 0.6127\n",
      "Epoch 168/5000\n",
      "4117/4117 [==============================] - 4s 869us/step - loss: 0.4958 - mae: 0.4306 - mse: 0.4958 - val_loss: 0.6461 - val_mae: 0.5511 - val_mse: 0.6461\n",
      "Epoch 169/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.5019 - mae: 0.4350 - mse: 0.5019 - val_loss: 0.6184 - val_mae: 0.5305 - val_mse: 0.6184\n",
      "Epoch 170/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4997 - mae: 0.4335 - mse: 0.4997 - val_loss: 0.6373 - val_mae: 0.5288 - val_mse: 0.6373\n",
      "Epoch 171/5000\n",
      "4117/4117 [==============================] - 4s 883us/step - loss: 0.4983 - mae: 0.4328 - mse: 0.4983 - val_loss: 0.6148 - val_mae: 0.5233 - val_mse: 0.6148\n",
      "Epoch 172/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.4951 - mae: 0.4304 - mse: 0.4951 - val_loss: 0.6112 - val_mae: 0.5209 - val_mse: 0.6112\n",
      "Epoch 173/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4984 - mae: 0.4314 - mse: 0.4984 - val_loss: 0.6034 - val_mae: 0.5163 - val_mse: 0.6034\n",
      "Epoch 174/5000\n",
      "4117/4117 [==============================] - 4s 879us/step - loss: 0.4975 - mae: 0.4301 - mse: 0.4975 - val_loss: 0.6083 - val_mae: 0.5196 - val_mse: 0.6083\n",
      "Epoch 175/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4956 - mae: 0.4293 - mse: 0.4956 - val_loss: 0.6114 - val_mae: 0.5244 - val_mse: 0.6114\n",
      "Epoch 176/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4993 - mae: 0.4336 - mse: 0.4993 - val_loss: 0.6243 - val_mae: 0.5312 - val_mse: 0.6243\n",
      "Epoch 177/5000\n",
      "4117/4117 [==============================] - 4s 870us/step - loss: 0.4977 - mae: 0.4302 - mse: 0.4977 - val_loss: 0.6141 - val_mae: 0.5222 - val_mse: 0.6141\n",
      "Epoch 178/5000\n",
      "4117/4117 [==============================] - 4s 864us/step - loss: 0.5005 - mae: 0.4310 - mse: 0.5005 - val_loss: 0.6066 - val_mae: 0.5149 - val_mse: 0.6066\n",
      "Epoch 179/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.5005 - mae: 0.4334 - mse: 0.5005 - val_loss: 0.6355 - val_mae: 0.5366 - val_mse: 0.6355\n",
      "Epoch 180/5000\n",
      "4117/4117 [==============================] - 4s 858us/step - loss: 0.4934 - mae: 0.4290 - mse: 0.4934 - val_loss: 0.6095 - val_mae: 0.5200 - val_mse: 0.6095\n",
      "Epoch 181/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4995 - mae: 0.4338 - mse: 0.4995 - val_loss: 0.6087 - val_mae: 0.5215 - val_mse: 0.6087\n",
      "Epoch 182/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4967 - mae: 0.4293 - mse: 0.4967 - val_loss: 0.6423 - val_mae: 0.5315 - val_mse: 0.6423\n",
      "Epoch 183/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4969 - mae: 0.4314 - mse: 0.4969 - val_loss: 0.6238 - val_mae: 0.5282 - val_mse: 0.6238\n",
      "Epoch 184/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4965 - mae: 0.4315 - mse: 0.4965 - val_loss: 0.6132 - val_mae: 0.5214 - val_mse: 0.6132\n",
      "Epoch 185/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.5010 - mae: 0.4321 - mse: 0.5010 - val_loss: 0.6136 - val_mae: 0.5249 - val_mse: 0.6136\n",
      "Epoch 186/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4928 - mae: 0.4284 - mse: 0.4928 - val_loss: 0.6279 - val_mae: 0.5386 - val_mse: 0.6279\n",
      "Epoch 187/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4950 - mae: 0.4318 - mse: 0.4950 - val_loss: 0.6605 - val_mae: 0.5421 - val_mse: 0.6605\n",
      "Epoch 188/5000\n",
      "4117/4117 [==============================] - 4s 865us/step - loss: 0.4969 - mae: 0.4313 - mse: 0.4969 - val_loss: 0.6090 - val_mae: 0.5183 - val_mse: 0.6090\n",
      "Epoch 189/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.4937 - mae: 0.4278 - mse: 0.4937 - val_loss: 0.6176 - val_mae: 0.5250 - val_mse: 0.6176\n",
      "Epoch 190/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.4959 - mae: 0.4310 - mse: 0.4959 - val_loss: 0.6231 - val_mae: 0.5262 - val_mse: 0.6231\n",
      "Epoch 191/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.4990 - mae: 0.4313 - mse: 0.4990 - val_loss: 0.6015 - val_mae: 0.5095 - val_mse: 0.6015\n",
      "Epoch 192/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4935 - mae: 0.4282 - mse: 0.4935 - val_loss: 0.6166 - val_mae: 0.5205 - val_mse: 0.6166\n",
      "Epoch 193/5000\n",
      "4117/4117 [==============================] - 4s 853us/step - loss: 0.4949 - mae: 0.4262 - mse: 0.4949 - val_loss: 0.6130 - val_mae: 0.5179 - val_mse: 0.6130\n",
      "Epoch 194/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4960 - mae: 0.4295 - mse: 0.4960 - val_loss: 0.6248 - val_mae: 0.5308 - val_mse: 0.6248\n",
      "Epoch 195/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4983 - mae: 0.4312 - mse: 0.4983 - val_loss: 0.6438 - val_mae: 0.5321 - val_mse: 0.6438\n",
      "Epoch 196/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4963 - mae: 0.4293 - mse: 0.4963 - val_loss: 0.6102 - val_mae: 0.5170 - val_mse: 0.6102\n",
      "Epoch 197/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.4974 - mae: 0.4319 - mse: 0.4974 - val_loss: 0.6148 - val_mae: 0.5266 - val_mse: 0.6148\n",
      "Epoch 198/5000\n",
      "4117/4117 [==============================] - 4s 875us/step - loss: 0.4916 - mae: 0.4261 - mse: 0.4916 - val_loss: 0.6134 - val_mae: 0.5174 - val_mse: 0.6134\n",
      "Epoch 199/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4954 - mae: 0.4294 - mse: 0.4954 - val_loss: 0.6092 - val_mae: 0.5203 - val_mse: 0.6092\n",
      "Epoch 200/5000\n",
      "4117/4117 [==============================] - 4s 873us/step - loss: 0.4922 - mae: 0.4242 - mse: 0.4922 - val_loss: 0.6124 - val_mae: 0.5161 - val_mse: 0.6124\n",
      "Epoch 201/5000\n",
      "4117/4117 [==============================] - 4s 866us/step - loss: 0.4982 - mae: 0.4318 - mse: 0.4982 - val_loss: 0.6644 - val_mae: 0.5617 - val_mse: 0.6644\n",
      "Epoch 202/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4940 - mae: 0.4286 - mse: 0.4940 - val_loss: 0.6543 - val_mae: 0.5349 - val_mse: 0.6543\n",
      "Epoch 203/5000\n",
      "4117/4117 [==============================] - 4s 857us/step - loss: 0.4951 - mae: 0.4276 - mse: 0.4951 - val_loss: 0.6221 - val_mae: 0.5261 - val_mse: 0.6221\n",
      "Epoch 204/5000\n",
      "4117/4117 [==============================] - 4s 883us/step - loss: 0.4964 - mae: 0.4307 - mse: 0.4964 - val_loss: 0.6179 - val_mae: 0.5222 - val_mse: 0.6179\n",
      "Epoch 205/5000\n",
      "4117/4117 [==============================] - 4s 850us/step - loss: 0.4952 - mae: 0.4295 - mse: 0.4952 - val_loss: 0.6257 - val_mae: 0.5282 - val_mse: 0.6257\n",
      "Epoch 206/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.4937 - mae: 0.4273 - mse: 0.4937 - val_loss: 0.6179 - val_mae: 0.5244 - val_mse: 0.6179\n",
      "Epoch 207/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.4898 - mae: 0.4242 - mse: 0.4898 - val_loss: 0.6192 - val_mae: 0.5287 - val_mse: 0.6192\n",
      "Epoch 208/5000\n",
      "4117/4117 [==============================] - 4s 859us/step - loss: 0.4960 - mae: 0.4292 - mse: 0.4960 - val_loss: 0.6090 - val_mae: 0.5246 - val_mse: 0.6090\n",
      "Epoch 209/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4932 - mae: 0.4276 - mse: 0.4932 - val_loss: 0.6105 - val_mae: 0.5200 - val_mse: 0.6105\n",
      "Epoch 210/5000\n",
      "4117/4117 [==============================] - 4s 875us/step - loss: 0.4959 - mae: 0.4293 - mse: 0.4959 - val_loss: 0.6108 - val_mae: 0.5138 - val_mse: 0.6108\n",
      "Epoch 211/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.4951 - mae: 0.4295 - mse: 0.4951 - val_loss: 0.6065 - val_mae: 0.5208 - val_mse: 0.6065\n",
      "Epoch 212/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4942 - mae: 0.4267 - mse: 0.4942 - val_loss: 0.6131 - val_mae: 0.5217 - val_mse: 0.6131\n",
      "Epoch 213/5000\n",
      "4117/4117 [==============================] - 4s 865us/step - loss: 0.4939 - mae: 0.4278 - mse: 0.4939 - val_loss: 0.6420 - val_mae: 0.5456 - val_mse: 0.6420\n",
      "Epoch 214/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.4970 - mae: 0.4304 - mse: 0.4970 - val_loss: 0.6228 - val_mae: 0.5309 - val_mse: 0.6228\n",
      "Epoch 215/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4950 - mae: 0.4277 - mse: 0.4950 - val_loss: 0.6011 - val_mae: 0.5114 - val_mse: 0.6011\n",
      "Epoch 216/5000\n",
      "4117/4117 [==============================] - 4s 860us/step - loss: 0.4939 - mae: 0.4266 - mse: 0.4939 - val_loss: 0.6103 - val_mae: 0.5175 - val_mse: 0.6103\n",
      "Epoch 217/5000\n",
      "4117/4117 [==============================] - 4s 877us/step - loss: 0.4915 - mae: 0.4262 - mse: 0.4915 - val_loss: 0.6207 - val_mae: 0.5278 - val_mse: 0.6207\n",
      "Epoch 218/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4978 - mae: 0.4302 - mse: 0.4978 - val_loss: 0.6308 - val_mae: 0.5368 - val_mse: 0.6308\n",
      "Epoch 219/5000\n",
      "4117/4117 [==============================] - 4s 854us/step - loss: 0.4945 - mae: 0.4285 - mse: 0.4945 - val_loss: 0.6102 - val_mae: 0.5272 - val_mse: 0.6102\n",
      "Epoch 220/5000\n",
      "4117/4117 [==============================] - 4s 1ms/step - loss: 0.4906 - mae: 0.4243 - mse: 0.4906 - val_loss: 0.6074 - val_mae: 0.5175 - val_mse: 0.6074\n",
      "Epoch 221/5000\n",
      "4117/4117 [==============================] - 4s 1ms/step - loss: 0.4942 - mae: 0.4262 - mse: 0.4942 - val_loss: 0.6137 - val_mae: 0.5223 - val_mse: 0.6137\n",
      "Epoch 222/5000\n",
      "4117/4117 [==============================] - 5s 1ms/step - loss: 0.4901 - mae: 0.4249 - mse: 0.4901 - val_loss: 0.6241 - val_mae: 0.5261 - val_mse: 0.6241\n",
      "Epoch 223/5000\n",
      "4117/4117 [==============================] - 5s 1ms/step - loss: 0.4922 - mae: 0.4264 - mse: 0.4922 - val_loss: 0.6223 - val_mae: 0.5236 - val_mse: 0.6223\n",
      "Epoch 224/5000\n",
      "4117/4117 [==============================] - 4s 956us/step - loss: 0.4949 - mae: 0.4262 - mse: 0.4949 - val_loss: 0.6124 - val_mae: 0.5216 - val_mse: 0.6124\n",
      "Epoch 225/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4936 - mae: 0.4276 - mse: 0.4936 - val_loss: 0.6279 - val_mae: 0.5329 - val_mse: 0.6279\n",
      "Epoch 226/5000\n",
      "4117/4117 [==============================] - 3s 841us/step - loss: 0.4936 - mae: 0.4262 - mse: 0.4936 - val_loss: 0.6187 - val_mae: 0.5268 - val_mse: 0.6187\n",
      "Epoch 227/5000\n",
      "4117/4117 [==============================] - 4s 869us/step - loss: 0.4900 - mae: 0.4251 - mse: 0.4900 - val_loss: 0.6253 - val_mae: 0.5271 - val_mse: 0.6253\n",
      "Epoch 228/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.4935 - mae: 0.4265 - mse: 0.4935 - val_loss: 0.6359 - val_mae: 0.5350 - val_mse: 0.6359\n",
      "Epoch 229/5000\n",
      "4117/4117 [==============================] - 3s 841us/step - loss: 0.4992 - mae: 0.4330 - mse: 0.4992 - val_loss: 0.6152 - val_mae: 0.5242 - val_mse: 0.6152\n",
      "Epoch 230/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4885 - mae: 0.4245 - mse: 0.4885 - val_loss: 0.6464 - val_mae: 0.5334 - val_mse: 0.6464\n",
      "Epoch 231/5000\n",
      "4117/4117 [==============================] - 4s 876us/step - loss: 0.4939 - mae: 0.4281 - mse: 0.4939 - val_loss: 0.6241 - val_mae: 0.5256 - val_mse: 0.6241\n",
      "Epoch 232/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4921 - mae: 0.4253 - mse: 0.4921 - val_loss: 0.6090 - val_mae: 0.5164 - val_mse: 0.6090\n",
      "Epoch 233/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4913 - mae: 0.4254 - mse: 0.4913 - val_loss: 0.6156 - val_mae: 0.5257 - val_mse: 0.6156\n",
      "Epoch 234/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4909 - mae: 0.4244 - mse: 0.4909 - val_loss: 0.6172 - val_mae: 0.5263 - val_mse: 0.6172\n",
      "Epoch 235/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4920 - mae: 0.4254 - mse: 0.4920 - val_loss: 0.6269 - val_mae: 0.5336 - val_mse: 0.6269\n",
      "Epoch 236/5000\n",
      "4117/4117 [==============================] - 4s 862us/step - loss: 0.4927 - mae: 0.4256 - mse: 0.4927 - val_loss: 0.6182 - val_mae: 0.5313 - val_mse: 0.6182\n",
      "Epoch 237/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.4936 - mae: 0.4274 - mse: 0.4936 - val_loss: 0.6192 - val_mae: 0.5207 - val_mse: 0.6192\n",
      "Epoch 238/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4979 - mae: 0.4302 - mse: 0.4979 - val_loss: 0.6144 - val_mae: 0.5186 - val_mse: 0.6144\n",
      "Epoch 239/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4917 - mae: 0.4265 - mse: 0.4917 - val_loss: 0.6420 - val_mae: 0.5443 - val_mse: 0.6420\n",
      "Epoch 240/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4906 - mae: 0.4237 - mse: 0.4906 - val_loss: 0.6213 - val_mae: 0.5236 - val_mse: 0.6213\n",
      "Epoch 241/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.4941 - mae: 0.4280 - mse: 0.4941 - val_loss: 0.6169 - val_mae: 0.5252 - val_mse: 0.6169\n",
      "Epoch 242/5000\n",
      "4117/4117 [==============================] - 3s 840us/step - loss: 0.4906 - mae: 0.4254 - mse: 0.4906 - val_loss: 0.5964 - val_mae: 0.5116 - val_mse: 0.5964\n",
      "Epoch 243/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4893 - mae: 0.4247 - mse: 0.4893 - val_loss: 0.6118 - val_mae: 0.5273 - val_mse: 0.6118\n",
      "Epoch 244/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4915 - mae: 0.4250 - mse: 0.4915 - val_loss: 0.6377 - val_mae: 0.5416 - val_mse: 0.6377\n",
      "Epoch 245/5000\n",
      "4117/4117 [==============================] - 3s 840us/step - loss: 0.4938 - mae: 0.4270 - mse: 0.4938 - val_loss: 0.6196 - val_mae: 0.5207 - val_mse: 0.6196\n",
      "Epoch 246/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4884 - mae: 0.4244 - mse: 0.4884 - val_loss: 0.6154 - val_mae: 0.5198 - val_mse: 0.6154\n",
      "Epoch 247/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4913 - mae: 0.4265 - mse: 0.4913 - val_loss: 0.6031 - val_mae: 0.5100 - val_mse: 0.6031\n",
      "Epoch 248/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4904 - mae: 0.4233 - mse: 0.4904 - val_loss: 0.6091 - val_mae: 0.5191 - val_mse: 0.6091\n",
      "Epoch 249/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4911 - mae: 0.4265 - mse: 0.4911 - val_loss: 0.6114 - val_mae: 0.5139 - val_mse: 0.6114\n",
      "Epoch 250/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4957 - mae: 0.4305 - mse: 0.4957 - val_loss: 0.6185 - val_mae: 0.5240 - val_mse: 0.6185\n",
      "Epoch 251/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4914 - mae: 0.4268 - mse: 0.4914 - val_loss: 0.6053 - val_mae: 0.5141 - val_mse: 0.6053\n",
      "Epoch 252/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4930 - mae: 0.4275 - mse: 0.4930 - val_loss: 0.6243 - val_mae: 0.5345 - val_mse: 0.6243\n",
      "Epoch 253/5000\n",
      "4117/4117 [==============================] - 3s 840us/step - loss: 0.4892 - mae: 0.4241 - mse: 0.4892 - val_loss: 0.6121 - val_mae: 0.5172 - val_mse: 0.6121\n",
      "Epoch 254/5000\n",
      "4117/4117 [==============================] - 4s 854us/step - loss: 0.4907 - mae: 0.4246 - mse: 0.4907 - val_loss: 0.6230 - val_mae: 0.5211 - val_mse: 0.6230\n",
      "Epoch 255/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4917 - mae: 0.4252 - mse: 0.4917 - val_loss: 0.6044 - val_mae: 0.5220 - val_mse: 0.6044\n",
      "Epoch 256/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4921 - mae: 0.4257 - mse: 0.4921 - val_loss: 0.6167 - val_mae: 0.5226 - val_mse: 0.6167\n",
      "Epoch 257/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.4912 - mae: 0.4241 - mse: 0.4912 - val_loss: 0.6132 - val_mae: 0.5176 - val_mse: 0.6132\n",
      "Epoch 258/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4928 - mae: 0.4268 - mse: 0.4928 - val_loss: 0.6230 - val_mae: 0.5336 - val_mse: 0.6230\n",
      "Epoch 259/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4884 - mae: 0.4216 - mse: 0.4884 - val_loss: 0.6121 - val_mae: 0.5202 - val_mse: 0.6121\n",
      "Epoch 260/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4901 - mae: 0.4238 - mse: 0.4901 - val_loss: 0.6218 - val_mae: 0.5237 - val_mse: 0.6218\n",
      "Epoch 261/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4894 - mae: 0.4233 - mse: 0.4894 - val_loss: 0.6258 - val_mae: 0.5254 - val_mse: 0.6258\n",
      "Epoch 262/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4898 - mae: 0.4234 - mse: 0.4898 - val_loss: 0.6681 - val_mae: 0.5456 - val_mse: 0.6681\n",
      "Epoch 263/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4910 - mae: 0.4256 - mse: 0.4910 - val_loss: 0.6107 - val_mae: 0.5184 - val_mse: 0.6107\n",
      "Epoch 264/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4875 - mae: 0.4210 - mse: 0.4875 - val_loss: 0.6183 - val_mae: 0.5222 - val_mse: 0.6183\n",
      "Epoch 265/5000\n",
      "4117/4117 [==============================] - 4s 852us/step - loss: 0.4904 - mae: 0.4239 - mse: 0.4904 - val_loss: 0.6135 - val_mae: 0.5155 - val_mse: 0.6135\n",
      "Epoch 266/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.4891 - mae: 0.4246 - mse: 0.4891 - val_loss: 0.6277 - val_mae: 0.5244 - val_mse: 0.6277\n",
      "Epoch 267/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4973 - mae: 0.4264 - mse: 0.4973 - val_loss: 0.6056 - val_mae: 0.5156 - val_mse: 0.6056\n",
      "Epoch 268/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4899 - mae: 0.4253 - mse: 0.4899 - val_loss: 0.6254 - val_mae: 0.5274 - val_mse: 0.6254\n",
      "Epoch 269/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4905 - mae: 0.4237 - mse: 0.4905 - val_loss: 0.6182 - val_mae: 0.5268 - val_mse: 0.6182\n",
      "Epoch 270/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4868 - mae: 0.4226 - mse: 0.4868 - val_loss: 0.6260 - val_mae: 0.5269 - val_mse: 0.6260\n",
      "Epoch 271/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4901 - mae: 0.4226 - mse: 0.4901 - val_loss: 0.6159 - val_mae: 0.5198 - val_mse: 0.6159\n",
      "Epoch 272/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4880 - mae: 0.4216 - mse: 0.4880 - val_loss: 0.6193 - val_mae: 0.5221 - val_mse: 0.6193\n",
      "Epoch 273/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4879 - mae: 0.4205 - mse: 0.4879 - val_loss: 0.6340 - val_mae: 0.5358 - val_mse: 0.6340\n",
      "Epoch 274/5000\n",
      "4117/4117 [==============================] - 3s 841us/step - loss: 0.4947 - mae: 0.4261 - mse: 0.4947 - val_loss: 0.6235 - val_mae: 0.5216 - val_mse: 0.6235\n",
      "Epoch 275/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4914 - mae: 0.4239 - mse: 0.4914 - val_loss: 0.6236 - val_mae: 0.5274 - val_mse: 0.6236\n",
      "Epoch 276/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.4887 - mae: 0.4233 - mse: 0.4887 - val_loss: 0.6124 - val_mae: 0.5215 - val_mse: 0.6124\n",
      "Epoch 277/5000\n",
      "4117/4117 [==============================] - 3s 841us/step - loss: 0.4885 - mae: 0.4237 - mse: 0.4885 - val_loss: 0.6199 - val_mae: 0.5291 - val_mse: 0.6199\n",
      "Epoch 278/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4927 - mae: 0.4267 - mse: 0.4927 - val_loss: 0.6204 - val_mae: 0.5243 - val_mse: 0.6204\n",
      "Epoch 279/5000\n",
      "4117/4117 [==============================] - 4s 862us/step - loss: 0.4842 - mae: 0.4179 - mse: 0.4842 - val_loss: 0.6047 - val_mae: 0.5134 - val_mse: 0.6047\n",
      "Epoch 280/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4865 - mae: 0.4210 - mse: 0.4865 - val_loss: 0.6618 - val_mae: 0.5441 - val_mse: 0.6618\n",
      "Epoch 281/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4857 - mae: 0.4192 - mse: 0.4857 - val_loss: 0.6135 - val_mae: 0.5195 - val_mse: 0.6135\n",
      "Epoch 282/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4852 - mae: 0.4184 - mse: 0.4852 - val_loss: 0.6170 - val_mae: 0.5232 - val_mse: 0.6170\n",
      "Epoch 283/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4912 - mae: 0.4228 - mse: 0.4912 - val_loss: 0.6250 - val_mae: 0.5288 - val_mse: 0.6250\n",
      "Epoch 284/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4915 - mae: 0.4252 - mse: 0.4915 - val_loss: 0.6221 - val_mae: 0.5297 - val_mse: 0.6221\n",
      "Epoch 285/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4920 - mae: 0.4247 - mse: 0.4920 - val_loss: 0.6086 - val_mae: 0.5203 - val_mse: 0.6086\n",
      "Epoch 286/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.4842 - mae: 0.4175 - mse: 0.4842 - val_loss: 0.6271 - val_mae: 0.5301 - val_mse: 0.6271\n",
      "Epoch 287/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4946 - mae: 0.4247 - mse: 0.4946 - val_loss: 0.6260 - val_mae: 0.5254 - val_mse: 0.6260\n",
      "Epoch 288/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4847 - mae: 0.4208 - mse: 0.4847 - val_loss: 0.6145 - val_mae: 0.5184 - val_mse: 0.6145\n",
      "Epoch 289/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4835 - mae: 0.4171 - mse: 0.4835 - val_loss: 0.6105 - val_mae: 0.5174 - val_mse: 0.6105\n",
      "Epoch 290/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4874 - mae: 0.4213 - mse: 0.4874 - val_loss: 0.6088 - val_mae: 0.5186 - val_mse: 0.6088\n",
      "Epoch 291/5000\n",
      "4117/4117 [==============================] - 3s 849us/step - loss: 0.4891 - mae: 0.4225 - mse: 0.4891 - val_loss: 0.6248 - val_mae: 0.5240 - val_mse: 0.6248\n",
      "Epoch 292/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4896 - mae: 0.4233 - mse: 0.4896 - val_loss: 0.6283 - val_mae: 0.5293 - val_mse: 0.6283\n",
      "Epoch 293/5000\n",
      "4117/4117 [==============================] - 4s 871us/step - loss: 0.4921 - mae: 0.4235 - mse: 0.4921 - val_loss: 0.6116 - val_mae: 0.5169 - val_mse: 0.6116\n",
      "Epoch 294/5000\n",
      "4117/4117 [==============================] - 3s 844us/step - loss: 0.4954 - mae: 0.4279 - mse: 0.4954 - val_loss: 0.6090 - val_mae: 0.5195 - val_mse: 0.6090\n",
      "Epoch 295/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4904 - mae: 0.4239 - mse: 0.4904 - val_loss: 0.6162 - val_mae: 0.5198 - val_mse: 0.6162\n",
      "Epoch 296/5000\n",
      "4117/4117 [==============================] - 3s 846us/step - loss: 0.4859 - mae: 0.4192 - mse: 0.4859 - val_loss: 0.6146 - val_mae: 0.5199 - val_mse: 0.6146\n",
      "Epoch 297/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4918 - mae: 0.4220 - mse: 0.4918 - val_loss: 0.6281 - val_mae: 0.5271 - val_mse: 0.6281\n",
      "Epoch 298/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4880 - mae: 0.4212 - mse: 0.4880 - val_loss: 0.6149 - val_mae: 0.5229 - val_mse: 0.6149\n",
      "Epoch 299/5000\n",
      "4117/4117 [==============================] - 4s 856us/step - loss: 0.4872 - mae: 0.4225 - mse: 0.4872 - val_loss: 0.6099 - val_mae: 0.5170 - val_mse: 0.6099\n",
      "Epoch 300/5000\n",
      "4117/4117 [==============================] - 4s 861us/step - loss: 0.4890 - mae: 0.4232 - mse: 0.4890 - val_loss: 0.6130 - val_mae: 0.5180 - val_mse: 0.6130\n",
      "Epoch 301/5000\n",
      "4117/4117 [==============================] - 4s 857us/step - loss: 0.4869 - mae: 0.4206 - mse: 0.4869 - val_loss: 0.6166 - val_mae: 0.5221 - val_mse: 0.6166\n",
      "Epoch 302/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4871 - mae: 0.4211 - mse: 0.4871 - val_loss: 0.6093 - val_mae: 0.5183 - val_mse: 0.6093\n",
      "Epoch 303/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4867 - mae: 0.4225 - mse: 0.4867 - val_loss: 0.6068 - val_mae: 0.5208 - val_mse: 0.6068\n",
      "Epoch 304/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4862 - mae: 0.4203 - mse: 0.4862 - val_loss: 0.6129 - val_mae: 0.5220 - val_mse: 0.6129\n",
      "Epoch 305/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4886 - mae: 0.4216 - mse: 0.4886 - val_loss: 0.6234 - val_mae: 0.5201 - val_mse: 0.6234\n",
      "Epoch 306/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4904 - mae: 0.4240 - mse: 0.4904 - val_loss: 0.6379 - val_mae: 0.5408 - val_mse: 0.6379\n",
      "Epoch 307/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4862 - mae: 0.4198 - mse: 0.4862 - val_loss: 0.6257 - val_mae: 0.5304 - val_mse: 0.6257\n",
      "Epoch 308/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4877 - mae: 0.4224 - mse: 0.4877 - val_loss: 0.6012 - val_mae: 0.5123 - val_mse: 0.6012\n",
      "Epoch 309/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4867 - mae: 0.4210 - mse: 0.4867 - val_loss: 0.6282 - val_mae: 0.5273 - val_mse: 0.6282\n",
      "Epoch 310/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4928 - mae: 0.4269 - mse: 0.4928 - val_loss: 0.6077 - val_mae: 0.5183 - val_mse: 0.6077\n",
      "Epoch 311/5000\n",
      "4117/4117 [==============================] - 3s 789us/step - loss: 0.4850 - mae: 0.4211 - mse: 0.4850 - val_loss: 0.6079 - val_mae: 0.5217 - val_mse: 0.6079\n",
      "Epoch 312/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4856 - mae: 0.4190 - mse: 0.4856 - val_loss: 0.6174 - val_mae: 0.5237 - val_mse: 0.6174\n",
      "Epoch 313/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4814 - mae: 0.4171 - mse: 0.4814 - val_loss: 0.6221 - val_mae: 0.5202 - val_mse: 0.6221\n",
      "Epoch 314/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4844 - mae: 0.4183 - mse: 0.4844 - val_loss: 0.6411 - val_mae: 0.5355 - val_mse: 0.6411\n",
      "Epoch 315/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4851 - mae: 0.4202 - mse: 0.4851 - val_loss: 0.5984 - val_mae: 0.5124 - val_mse: 0.5984\n",
      "Epoch 316/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4861 - mae: 0.4202 - mse: 0.4861 - val_loss: 0.6167 - val_mae: 0.5210 - val_mse: 0.6167\n",
      "Epoch 317/5000\n",
      "4117/4117 [==============================] - 3s 795us/step - loss: 0.4881 - mae: 0.4222 - mse: 0.4881 - val_loss: 0.6096 - val_mae: 0.5172 - val_mse: 0.6096\n",
      "Epoch 318/5000\n",
      "4117/4117 [==============================] - 3s 803us/step - loss: 0.4871 - mae: 0.4221 - mse: 0.4871 - val_loss: 0.6435 - val_mae: 0.5409 - val_mse: 0.6435\n",
      "Epoch 319/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4855 - mae: 0.4184 - mse: 0.4855 - val_loss: 0.6243 - val_mae: 0.5325 - val_mse: 0.6243\n",
      "Epoch 320/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4837 - mae: 0.4197 - mse: 0.4837 - val_loss: 0.6127 - val_mae: 0.5263 - val_mse: 0.6127\n",
      "Epoch 321/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4912 - mae: 0.4241 - mse: 0.4912 - val_loss: 0.6257 - val_mae: 0.5301 - val_mse: 0.6257\n",
      "Epoch 322/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4881 - mae: 0.4218 - mse: 0.4881 - val_loss: 0.6238 - val_mae: 0.5307 - val_mse: 0.6238\n",
      "Epoch 323/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4874 - mae: 0.4186 - mse: 0.4874 - val_loss: 0.6156 - val_mae: 0.5189 - val_mse: 0.6156\n",
      "Epoch 324/5000\n",
      "4117/4117 [==============================] - 3s 795us/step - loss: 0.4844 - mae: 0.4179 - mse: 0.4844 - val_loss: 0.6227 - val_mae: 0.5207 - val_mse: 0.6227\n",
      "Epoch 325/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4881 - mae: 0.4222 - mse: 0.4881 - val_loss: 0.6107 - val_mae: 0.5192 - val_mse: 0.6107\n",
      "Epoch 326/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4844 - mae: 0.4192 - mse: 0.4844 - val_loss: 0.6170 - val_mae: 0.5246 - val_mse: 0.6170\n",
      "Epoch 327/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4866 - mae: 0.4184 - mse: 0.4866 - val_loss: 0.6303 - val_mae: 0.5242 - val_mse: 0.6303\n",
      "Epoch 328/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4892 - mae: 0.4234 - mse: 0.4892 - val_loss: 0.6467 - val_mae: 0.5426 - val_mse: 0.6467\n",
      "Epoch 329/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4889 - mae: 0.4203 - mse: 0.4889 - val_loss: 0.6109 - val_mae: 0.5179 - val_mse: 0.6109\n",
      "Epoch 330/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4915 - mae: 0.4231 - mse: 0.4915 - val_loss: 0.6161 - val_mae: 0.5214 - val_mse: 0.6161\n",
      "Epoch 331/5000\n",
      "4117/4117 [==============================] - 3s 795us/step - loss: 0.4862 - mae: 0.4191 - mse: 0.4862 - val_loss: 0.6192 - val_mae: 0.5250 - val_mse: 0.6192\n",
      "Epoch 332/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4879 - mae: 0.4209 - mse: 0.4879 - val_loss: 0.6335 - val_mae: 0.5309 - val_mse: 0.6335\n",
      "Epoch 333/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4851 - mae: 0.4195 - mse: 0.4851 - val_loss: 0.6098 - val_mae: 0.5166 - val_mse: 0.6098\n",
      "Epoch 334/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4828 - mae: 0.4174 - mse: 0.4828 - val_loss: 0.6177 - val_mae: 0.5289 - val_mse: 0.6177\n",
      "Epoch 335/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4868 - mae: 0.4216 - mse: 0.4868 - val_loss: 0.5980 - val_mae: 0.5119 - val_mse: 0.5980\n",
      "Epoch 336/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4859 - mae: 0.4213 - mse: 0.4859 - val_loss: 0.6155 - val_mae: 0.5228 - val_mse: 0.6155\n",
      "Epoch 337/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4871 - mae: 0.4205 - mse: 0.4871 - val_loss: 0.6068 - val_mae: 0.5159 - val_mse: 0.6068\n",
      "Epoch 338/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4855 - mae: 0.4201 - mse: 0.4855 - val_loss: 0.6246 - val_mae: 0.5281 - val_mse: 0.6246\n",
      "Epoch 339/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4871 - mae: 0.4215 - mse: 0.4871 - val_loss: 0.6165 - val_mae: 0.5272 - val_mse: 0.6165\n",
      "Epoch 340/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4888 - mae: 0.4244 - mse: 0.4888 - val_loss: 0.6303 - val_mae: 0.5267 - val_mse: 0.6303\n",
      "Epoch 341/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4856 - mae: 0.4182 - mse: 0.4856 - val_loss: 0.6148 - val_mae: 0.5259 - val_mse: 0.6148\n",
      "Epoch 342/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4829 - mae: 0.4180 - mse: 0.4829 - val_loss: 0.6071 - val_mae: 0.5164 - val_mse: 0.6071\n",
      "Epoch 343/5000\n",
      "4117/4117 [==============================] - 3s 835us/step - loss: 0.4813 - mae: 0.4169 - mse: 0.4813 - val_loss: 0.6167 - val_mae: 0.5280 - val_mse: 0.6167\n",
      "Epoch 344/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4846 - mae: 0.4195 - mse: 0.4846 - val_loss: 0.6682 - val_mae: 0.5465 - val_mse: 0.6682\n",
      "Epoch 345/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4875 - mae: 0.4194 - mse: 0.4875 - val_loss: 0.6165 - val_mae: 0.5245 - val_mse: 0.6165\n",
      "Epoch 346/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4851 - mae: 0.4180 - mse: 0.4851 - val_loss: 0.6142 - val_mae: 0.5188 - val_mse: 0.6142\n",
      "Epoch 347/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4848 - mae: 0.4205 - mse: 0.4848 - val_loss: 0.6039 - val_mae: 0.5183 - val_mse: 0.6039\n",
      "Epoch 348/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4853 - mae: 0.4188 - mse: 0.4853 - val_loss: 0.6242 - val_mae: 0.5251 - val_mse: 0.6242\n",
      "Epoch 349/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4850 - mae: 0.4210 - mse: 0.4850 - val_loss: 0.6142 - val_mae: 0.5193 - val_mse: 0.6142\n",
      "Epoch 350/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4898 - mae: 0.4245 - mse: 0.4898 - val_loss: 0.6076 - val_mae: 0.5209 - val_mse: 0.6076\n",
      "Epoch 351/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4843 - mae: 0.4195 - mse: 0.4843 - val_loss: 0.6046 - val_mae: 0.5228 - val_mse: 0.6046\n",
      "Epoch 352/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4816 - mae: 0.4169 - mse: 0.4816 - val_loss: 0.6175 - val_mae: 0.5223 - val_mse: 0.6175\n",
      "Epoch 353/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4828 - mae: 0.4191 - mse: 0.4828 - val_loss: 0.6351 - val_mae: 0.5291 - val_mse: 0.6351\n",
      "Epoch 354/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4865 - mae: 0.4205 - mse: 0.4865 - val_loss: 0.6180 - val_mae: 0.5243 - val_mse: 0.6180\n",
      "Epoch 355/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4805 - mae: 0.4153 - mse: 0.4805 - val_loss: 0.6153 - val_mae: 0.5203 - val_mse: 0.6153\n",
      "Epoch 356/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4862 - mae: 0.4180 - mse: 0.4862 - val_loss: 0.6197 - val_mae: 0.5263 - val_mse: 0.6197\n",
      "Epoch 357/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4855 - mae: 0.4205 - mse: 0.4855 - val_loss: 0.6233 - val_mae: 0.5280 - val_mse: 0.6233\n",
      "Epoch 358/5000\n",
      "4117/4117 [==============================] - 4s 850us/step - loss: 0.4765 - mae: 0.4090 - mse: 0.4765 - val_loss: 0.6178 - val_mae: 0.5186 - val_mse: 0.6178\n",
      "Epoch 639/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4735 - mae: 0.4084 - mse: 0.4735 - val_loss: 0.6267 - val_mae: 0.5255 - val_mse: 0.6267\n",
      "Epoch 640/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4710 - mae: 0.4069 - mse: 0.4710 - val_loss: 0.6267 - val_mae: 0.5263 - val_mse: 0.6267\n",
      "Epoch 641/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4735 - mae: 0.4071 - mse: 0.4735 - val_loss: 0.6344 - val_mae: 0.5334 - val_mse: 0.6344\n",
      "Epoch 642/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4779 - mae: 0.4114 - mse: 0.4779 - val_loss: 0.6149 - val_mae: 0.5185 - val_mse: 0.6149\n",
      "Epoch 643/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4750 - mae: 0.4084 - mse: 0.4750 - val_loss: 0.6189 - val_mae: 0.5237 - val_mse: 0.6189\n",
      "Epoch 644/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4734 - mae: 0.4066 - mse: 0.4734 - val_loss: 0.6212 - val_mae: 0.5217 - val_mse: 0.6212\n",
      "Epoch 645/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4745 - mae: 0.4103 - mse: 0.4745 - val_loss: 0.6130 - val_mae: 0.5176 - val_mse: 0.6130\n",
      "Epoch 646/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4762 - mae: 0.4092 - mse: 0.4762 - val_loss: 0.6142 - val_mae: 0.5160 - val_mse: 0.6142\n",
      "Epoch 647/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4748 - mae: 0.4091 - mse: 0.4748 - val_loss: 0.6193 - val_mae: 0.5195 - val_mse: 0.6193\n",
      "Epoch 648/5000\n",
      "4117/4117 [==============================] - 3s 802us/step - loss: 0.4730 - mae: 0.4077 - mse: 0.4730 - val_loss: 0.6193 - val_mae: 0.5223 - val_mse: 0.6193\n",
      "Epoch 649/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4710 - mae: 0.4057 - mse: 0.4710 - val_loss: 0.6260 - val_mae: 0.5265 - val_mse: 0.6260\n",
      "Epoch 650/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4743 - mae: 0.4084 - mse: 0.4743 - val_loss: 0.6203 - val_mae: 0.5213 - val_mse: 0.6203\n",
      "Epoch 651/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4747 - mae: 0.4078 - mse: 0.4747 - val_loss: 0.6270 - val_mae: 0.5270 - val_mse: 0.6270\n",
      "Epoch 652/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4740 - mae: 0.4083 - mse: 0.4740 - val_loss: 0.6241 - val_mae: 0.5275 - val_mse: 0.6241\n",
      "Epoch 653/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4759 - mae: 0.4085 - mse: 0.4759 - val_loss: 0.6290 - val_mae: 0.5300 - val_mse: 0.6290\n",
      "Epoch 654/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4748 - mae: 0.4089 - mse: 0.4748 - val_loss: 0.6214 - val_mae: 0.5232 - val_mse: 0.6214\n",
      "Epoch 655/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4742 - mae: 0.4078 - mse: 0.4742 - val_loss: 0.6186 - val_mae: 0.5199 - val_mse: 0.6186\n",
      "Epoch 656/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4758 - mae: 0.4096 - mse: 0.4758 - val_loss: 0.6229 - val_mae: 0.5213 - val_mse: 0.6229\n",
      "Epoch 657/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4763 - mae: 0.4104 - mse: 0.4763 - val_loss: 0.6603 - val_mae: 0.5532 - val_mse: 0.6603\n",
      "Epoch 658/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4729 - mae: 0.4086 - mse: 0.4729 - val_loss: 0.6148 - val_mae: 0.5178 - val_mse: 0.6148\n",
      "Epoch 659/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4757 - mae: 0.4107 - mse: 0.4757 - val_loss: 0.6119 - val_mae: 0.5193 - val_mse: 0.6119\n",
      "Epoch 660/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4756 - mae: 0.4076 - mse: 0.4756 - val_loss: 0.6243 - val_mae: 0.5224 - val_mse: 0.6243\n",
      "Epoch 661/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4737 - mae: 0.4060 - mse: 0.4737 - val_loss: 0.6213 - val_mae: 0.5167 - val_mse: 0.6213\n",
      "Epoch 708/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4764 - mae: 0.4088 - mse: 0.4764 - val_loss: 0.6262 - val_mae: 0.5235 - val_mse: 0.6262\n",
      "Epoch 709/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4755 - mae: 0.4094 - mse: 0.4755 - val_loss: 0.6242 - val_mae: 0.5223 - val_mse: 0.6242\n",
      "Epoch 710/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4731 - mae: 0.4048 - mse: 0.4731 - val_loss: 0.6201 - val_mae: 0.5178 - val_mse: 0.6201\n",
      "Epoch 711/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4745 - mae: 0.4091 - mse: 0.4745 - val_loss: 0.6271 - val_mae: 0.5208 - val_mse: 0.6271\n",
      "Epoch 712/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4727 - mae: 0.4073 - mse: 0.4727 - val_loss: 0.6219 - val_mae: 0.5240 - val_mse: 0.6219\n",
      "Epoch 713/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4731 - mae: 0.4060 - mse: 0.4731 - val_loss: 0.6177 - val_mae: 0.5210 - val_mse: 0.6177\n",
      "Epoch 714/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4707 - mae: 0.4058 - mse: 0.4707 - val_loss: 0.6080 - val_mae: 0.5137 - val_mse: 0.6080\n",
      "Epoch 715/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4700 - mae: 0.4041 - mse: 0.4700 - val_loss: 0.6115 - val_mae: 0.5146 - val_mse: 0.6115\n",
      "Epoch 716/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4709 - mae: 0.4058 - mse: 0.4709 - val_loss: 0.6206 - val_mae: 0.5217 - val_mse: 0.6206\n",
      "Epoch 717/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4723 - mae: 0.4057 - mse: 0.4723 - val_loss: 0.6247 - val_mae: 0.5215 - val_mse: 0.6247\n",
      "Epoch 718/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4722 - mae: 0.4057 - mse: 0.4722 - val_loss: 0.6202 - val_mae: 0.5217 - val_mse: 0.6202\n",
      "Epoch 719/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4753 - mae: 0.4101 - mse: 0.4753 - val_loss: 0.6342 - val_mae: 0.5366 - val_mse: 0.6342\n",
      "Epoch 720/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4733 - mae: 0.4070 - mse: 0.4733 - val_loss: 0.6210 - val_mae: 0.5244 - val_mse: 0.6210\n",
      "Epoch 721/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4753 - mae: 0.4090 - mse: 0.4753 - val_loss: 0.6225 - val_mae: 0.5224 - val_mse: 0.6225\n",
      "Epoch 722/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4696 - mae: 0.4041 - mse: 0.4696 - val_loss: 0.6526 - val_mae: 0.5353 - val_mse: 0.6526\n",
      "Epoch 723/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4710 - mae: 0.4049 - mse: 0.4710 - val_loss: 0.6213 - val_mae: 0.5204 - val_mse: 0.6213\n",
      "Epoch 724/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4683 - mae: 0.4045 - mse: 0.4683 - val_loss: 0.6186 - val_mae: 0.5162 - val_mse: 0.6186\n",
      "Epoch 725/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4716 - mae: 0.4065 - mse: 0.4716 - val_loss: 0.6293 - val_mae: 0.5233 - val_mse: 0.6293\n",
      "Epoch 726/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4701 - mae: 0.4045 - mse: 0.4701 - val_loss: 0.6221 - val_mae: 0.5233 - val_mse: 0.6221\n",
      "Epoch 727/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4728 - mae: 0.4067 - mse: 0.4728 - val_loss: 0.6211 - val_mae: 0.5233 - val_mse: 0.6211\n",
      "Epoch 728/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4715 - mae: 0.4058 - mse: 0.4715 - val_loss: 0.6375 - val_mae: 0.5287 - val_mse: 0.6375\n",
      "Epoch 729/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4750 - mae: 0.4088 - mse: 0.4750 - val_loss: 0.6568 - val_mae: 0.5434 - val_mse: 0.6568\n",
      "Epoch 730/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4707 - mae: 0.4049 - mse: 0.4707 - val_loss: 0.6216 - val_mae: 0.5222 - val_mse: 0.6216\n",
      "Epoch 731/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4741 - mae: 0.4085 - mse: 0.4741 - val_loss: 0.6202 - val_mae: 0.5240 - val_mse: 0.6202\n",
      "Epoch 791/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4731 - mae: 0.4056 - mse: 0.4731 - val_loss: 0.6265 - val_mae: 0.5275 - val_mse: 0.6265\n",
      "Epoch 792/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4722 - mae: 0.4088 - mse: 0.4722 - val_loss: 0.6367 - val_mae: 0.5344 - val_mse: 0.6367\n",
      "Epoch 793/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4757 - mae: 0.4092 - mse: 0.4757 - val_loss: 0.6292 - val_mae: 0.5308 - val_mse: 0.6292\n",
      "Epoch 794/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4722 - mae: 0.4072 - mse: 0.4722 - val_loss: 0.6195 - val_mae: 0.5207 - val_mse: 0.6195\n",
      "Epoch 795/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4696 - mae: 0.4055 - mse: 0.4696 - val_loss: 0.6263 - val_mae: 0.5249 - val_mse: 0.6263\n",
      "Epoch 796/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4716 - mae: 0.4055 - mse: 0.4716 - val_loss: 0.6192 - val_mae: 0.5177 - val_mse: 0.6192\n",
      "Epoch 797/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4671 - mae: 0.4035 - mse: 0.4671 - val_loss: 0.6283 - val_mae: 0.5229 - val_mse: 0.6283\n",
      "Epoch 798/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4726 - mae: 0.4081 - mse: 0.4726 - val_loss: 0.6212 - val_mae: 0.5230 - val_mse: 0.6212\n",
      "Epoch 799/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4690 - mae: 0.4039 - mse: 0.4690 - val_loss: 0.6252 - val_mae: 0.5252 - val_mse: 0.6252\n",
      "Epoch 800/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4706 - mae: 0.4050 - mse: 0.4706 - val_loss: 0.6217 - val_mae: 0.5214 - val_mse: 0.6217\n",
      "Epoch 801/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4728 - mae: 0.4056 - mse: 0.4728 - val_loss: 0.6254 - val_mae: 0.5225 - val_mse: 0.6254\n",
      "Epoch 802/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4735 - mae: 0.4066 - mse: 0.4735 - val_loss: 0.6318 - val_mae: 0.5265 - val_mse: 0.6318\n",
      "Epoch 803/5000\n",
      "4117/4117 [==============================] - 4s 851us/step - loss: 0.4725 - mae: 0.4074 - mse: 0.4725 - val_loss: 0.6250 - val_mae: 0.5260 - val_mse: 0.6250\n",
      "Epoch 804/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4720 - mae: 0.4066 - mse: 0.4720 - val_loss: 0.6211 - val_mae: 0.5198 - val_mse: 0.6211\n",
      "Epoch 805/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4707 - mae: 0.4046 - mse: 0.4707 - val_loss: 0.6300 - val_mae: 0.5315 - val_mse: 0.6300\n",
      "Epoch 806/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4713 - mae: 0.4055 - mse: 0.4713 - val_loss: 0.6696 - val_mae: 0.5526 - val_mse: 0.6696\n",
      "Epoch 807/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4745 - mae: 0.4067 - mse: 0.4745 - val_loss: 0.6226 - val_mae: 0.5250 - val_mse: 0.6226\n",
      "Epoch 808/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4736 - mae: 0.4064 - mse: 0.4736 - val_loss: 0.6074 - val_mae: 0.5135 - val_mse: 0.6074\n",
      "Epoch 809/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4672 - mae: 0.4021 - mse: 0.4672 - val_loss: 0.6211 - val_mae: 0.5231 - val_mse: 0.6211\n",
      "Epoch 810/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4714 - mae: 0.4068 - mse: 0.4714 - val_loss: 0.6364 - val_mae: 0.5318 - val_mse: 0.6364\n",
      "Epoch 811/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4731 - mae: 0.4066 - mse: 0.4731 - val_loss: 0.6140 - val_mae: 0.5187 - val_mse: 0.6140\n",
      "Epoch 812/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4695 - mae: 0.4039 - mse: 0.4695 - val_loss: 0.6392 - val_mae: 0.5360 - val_mse: 0.6392\n",
      "Epoch 813/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4697 - mae: 0.4032 - mse: 0.4697 - val_loss: 0.6202 - val_mae: 0.5162 - val_mse: 0.6202\n",
      "Epoch 814/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4697 - mae: 0.4040 - mse: 0.4697 - val_loss: 0.6190 - val_mae: 0.5173 - val_mse: 0.6190\n",
      "Epoch 880/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4709 - mae: 0.4053 - mse: 0.4709 - val_loss: 0.6225 - val_mae: 0.5233 - val_mse: 0.6225\n",
      "Epoch 881/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4720 - mae: 0.4042 - mse: 0.4720 - val_loss: 0.6226 - val_mae: 0.5215 - val_mse: 0.6226\n",
      "Epoch 882/5000\n",
      "4117/4117 [==============================] - 4s 866us/step - loss: 0.4712 - mae: 0.4063 - mse: 0.4712 - val_loss: 0.6125 - val_mae: 0.5159 - val_mse: 0.6125\n",
      "Epoch 883/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4686 - mae: 0.4029 - mse: 0.4686 - val_loss: 0.6356 - val_mae: 0.5318 - val_mse: 0.6356\n",
      "Epoch 884/5000\n",
      "4117/4117 [==============================] - 3s 793us/step - loss: 0.4664 - mae: 0.4018 - mse: 0.4664 - val_loss: 0.6279 - val_mae: 0.5218 - val_mse: 0.6279\n",
      "Epoch 885/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4675 - mae: 0.4021 - mse: 0.4675 - val_loss: 0.6193 - val_mae: 0.5220 - val_mse: 0.6193\n",
      "Epoch 886/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4676 - mae: 0.4012 - mse: 0.4676 - val_loss: 0.6211 - val_mae: 0.5183 - val_mse: 0.6211\n",
      "Epoch 887/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4685 - mae: 0.4017 - mse: 0.4685 - val_loss: 0.6342 - val_mae: 0.5310 - val_mse: 0.6342\n",
      "Epoch 888/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4710 - mae: 0.4036 - mse: 0.4710 - val_loss: 0.6246 - val_mae: 0.5219 - val_mse: 0.6246\n",
      "Epoch 889/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4664 - mae: 0.4008 - mse: 0.4664 - val_loss: 0.6203 - val_mae: 0.5196 - val_mse: 0.6203\n",
      "Epoch 890/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4741 - mae: 0.4067 - mse: 0.4741 - val_loss: 0.6251 - val_mae: 0.5207 - val_mse: 0.6251\n",
      "Epoch 891/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4714 - mae: 0.4060 - mse: 0.4714 - val_loss: 0.6839 - val_mae: 0.5629 - val_mse: 0.6839\n",
      "Epoch 892/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4691 - mae: 0.4025 - mse: 0.4691 - val_loss: 0.6251 - val_mae: 0.5187 - val_mse: 0.6251\n",
      "Epoch 893/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4709 - mae: 0.4046 - mse: 0.4709 - val_loss: 0.6296 - val_mae: 0.5265 - val_mse: 0.6296\n",
      "Epoch 894/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4731 - mae: 0.4061 - mse: 0.4731 - val_loss: 0.6243 - val_mae: 0.5226 - val_mse: 0.6243\n",
      "Epoch 895/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4677 - mae: 0.4013 - mse: 0.4677 - val_loss: 0.6101 - val_mae: 0.5108 - val_mse: 0.6101\n",
      "Epoch 896/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4657 - mae: 0.4004 - mse: 0.4657 - val_loss: 0.6116 - val_mae: 0.5169 - val_mse: 0.6116\n",
      "Epoch 897/5000\n",
      "4117/4117 [==============================] - 4s 876us/step - loss: 0.4721 - mae: 0.4040 - mse: 0.4721 - val_loss: 0.6368 - val_mae: 0.5307 - val_mse: 0.6368\n",
      "Epoch 898/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4711 - mae: 0.4033 - mse: 0.4711 - val_loss: 0.6161 - val_mae: 0.5160 - val_mse: 0.6161\n",
      "Epoch 899/5000\n",
      "4117/4117 [==============================] - 3s 801us/step - loss: 0.4690 - mae: 0.4021 - mse: 0.4690 - val_loss: 0.6096 - val_mae: 0.5121 - val_mse: 0.6096\n",
      "Epoch 900/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4687 - mae: 0.4019 - mse: 0.4687 - val_loss: 0.6095 - val_mae: 0.5162 - val_mse: 0.6095\n",
      "Epoch 901/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4744 - mae: 0.4071 - mse: 0.4744 - val_loss: 0.6196 - val_mae: 0.5249 - val_mse: 0.6196\n",
      "Epoch 902/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4716 - mae: 0.4049 - mse: 0.4716 - val_loss: 0.6166 - val_mae: 0.5219 - val_mse: 0.6166\n",
      "Epoch 903/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4694 - mae: 0.4024 - mse: 0.4694 - val_loss: 0.6191 - val_mae: 0.5171 - val_mse: 0.6191\n",
      "Epoch 974/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4693 - mae: 0.4032 - mse: 0.4693 - val_loss: 0.6232 - val_mae: 0.5230 - val_mse: 0.6232\n",
      "Epoch 975/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4666 - mae: 0.4020 - mse: 0.4666 - val_loss: 0.6183 - val_mae: 0.5200 - val_mse: 0.6183\n",
      "Epoch 976/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4695 - mae: 0.4027 - mse: 0.4695 - val_loss: 0.6313 - val_mae: 0.5276 - val_mse: 0.6313\n",
      "Epoch 977/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4695 - mae: 0.4034 - mse: 0.4695 - val_loss: 0.6299 - val_mae: 0.5246 - val_mse: 0.6299\n",
      "Epoch 978/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4651 - mae: 0.3985 - mse: 0.4651 - val_loss: 0.6240 - val_mae: 0.5229 - val_mse: 0.6240\n",
      "Epoch 979/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4686 - mae: 0.4014 - mse: 0.4686 - val_loss: 0.6169 - val_mae: 0.5189 - val_mse: 0.6169\n",
      "Epoch 980/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4787 - mae: 0.4074 - mse: 0.4787 - val_loss: 0.6235 - val_mae: 0.5225 - val_mse: 0.6235\n",
      "Epoch 981/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4694 - mae: 0.4019 - mse: 0.4694 - val_loss: 0.6453 - val_mae: 0.5366 - val_mse: 0.6453\n",
      "Epoch 982/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4705 - mae: 0.4033 - mse: 0.4705 - val_loss: 0.6214 - val_mae: 0.5225 - val_mse: 0.6214\n",
      "Epoch 983/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4724 - mae: 0.4059 - mse: 0.4724 - val_loss: 0.6144 - val_mae: 0.5188 - val_mse: 0.6144\n",
      "Epoch 984/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4716 - mae: 0.4052 - mse: 0.4716 - val_loss: 0.6189 - val_mae: 0.5237 - val_mse: 0.6189\n",
      "Epoch 985/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4694 - mae: 0.4029 - mse: 0.4694 - val_loss: 0.6196 - val_mae: 0.5190 - val_mse: 0.6196\n",
      "Epoch 986/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4686 - mae: 0.4013 - mse: 0.4686 - val_loss: 0.6186 - val_mae: 0.5216 - val_mse: 0.6186\n",
      "Epoch 987/5000\n",
      "4117/4117 [==============================] - 3s 792us/step - loss: 0.4711 - mae: 0.4038 - mse: 0.4711 - val_loss: 0.6262 - val_mae: 0.5258 - val_mse: 0.6262\n",
      "Epoch 988/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4666 - mae: 0.3999 - mse: 0.4666 - val_loss: 0.6085 - val_mae: 0.5123 - val_mse: 0.6085\n",
      "Epoch 989/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4717 - mae: 0.4034 - mse: 0.4717 - val_loss: 0.6166 - val_mae: 0.5181 - val_mse: 0.6166\n",
      "Epoch 990/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4698 - mae: 0.4040 - mse: 0.4698 - val_loss: 0.6234 - val_mae: 0.5198 - val_mse: 0.6234\n",
      "Epoch 991/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4674 - mae: 0.4022 - mse: 0.4674 - val_loss: 0.6167 - val_mae: 0.5212 - val_mse: 0.6167\n",
      "Epoch 992/5000\n",
      "4117/4117 [==============================] - 3s 792us/step - loss: 0.4694 - mae: 0.4031 - mse: 0.4694 - val_loss: 0.6226 - val_mae: 0.5238 - val_mse: 0.6226\n",
      "Epoch 993/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4682 - mae: 0.4025 - mse: 0.4682 - val_loss: 0.6392 - val_mae: 0.5323 - val_mse: 0.6392\n",
      "Epoch 994/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4645 - mae: 0.3980 - mse: 0.4645 - val_loss: 0.6138 - val_mae: 0.5188 - val_mse: 0.6138\n",
      "Epoch 995/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4655 - mae: 0.3999 - mse: 0.4655 - val_loss: 0.6349 - val_mae: 0.5320 - val_mse: 0.6349\n",
      "Epoch 996/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4693 - mae: 0.4004 - mse: 0.4693 - val_loss: 0.6244 - val_mae: 0.5240 - val_mse: 0.6244\n",
      "Epoch 997/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4738 - mae: 0.4047 - mse: 0.4738 - val_loss: 0.6239 - val_mae: 0.5203 - val_mse: 0.6239\n",
      "Epoch 1074/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4671 - mae: 0.4010 - mse: 0.4671 - val_loss: 0.6263 - val_mae: 0.5248 - val_mse: 0.6263\n",
      "Epoch 1075/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4690 - mae: 0.4030 - mse: 0.4690 - val_loss: 0.6195 - val_mae: 0.5195 - val_mse: 0.6195\n",
      "Epoch 1076/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4671 - mae: 0.4002 - mse: 0.4671 - val_loss: 0.6275 - val_mae: 0.5278 - val_mse: 0.6275\n",
      "Epoch 1077/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4686 - mae: 0.4021 - mse: 0.4686 - val_loss: 0.6229 - val_mae: 0.5210 - val_mse: 0.6229\n",
      "Epoch 1078/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4649 - mae: 0.3982 - mse: 0.4649 - val_loss: 0.6275 - val_mae: 0.5230 - val_mse: 0.6275\n",
      "Epoch 1079/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4648 - mae: 0.3978 - mse: 0.4648 - val_loss: 0.6217 - val_mae: 0.5211 - val_mse: 0.6217\n",
      "Epoch 1080/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4660 - mae: 0.3996 - mse: 0.4660 - val_loss: 0.6285 - val_mae: 0.5247 - val_mse: 0.6285\n",
      "Epoch 1081/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4684 - mae: 0.4018 - mse: 0.4684 - val_loss: 0.6280 - val_mae: 0.5219 - val_mse: 0.6280\n",
      "Epoch 1082/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4688 - mae: 0.4016 - mse: 0.4688 - val_loss: 0.6241 - val_mae: 0.5272 - val_mse: 0.6241\n",
      "Epoch 1083/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4640 - mae: 0.3978 - mse: 0.4640 - val_loss: 0.6272 - val_mae: 0.5224 - val_mse: 0.6272\n",
      "Epoch 1084/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4661 - mae: 0.3993 - mse: 0.4661 - val_loss: 0.6207 - val_mae: 0.5190 - val_mse: 0.6207\n",
      "Epoch 1085/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4636 - mae: 0.3990 - mse: 0.4636 - val_loss: 0.6103 - val_mae: 0.5114 - val_mse: 0.6103\n",
      "Epoch 1086/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4677 - mae: 0.4001 - mse: 0.4677 - val_loss: 0.6217 - val_mae: 0.5222 - val_mse: 0.6217\n",
      "Epoch 1087/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4715 - mae: 0.4041 - mse: 0.4715 - val_loss: 0.6175 - val_mae: 0.5178 - val_mse: 0.6175\n",
      "Epoch 1088/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4709 - mae: 0.4035 - mse: 0.4709 - val_loss: 0.6294 - val_mae: 0.5265 - val_mse: 0.6294\n",
      "Epoch 1089/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4687 - mae: 0.4010 - mse: 0.4687 - val_loss: 0.6098 - val_mae: 0.5156 - val_mse: 0.6098\n",
      "Epoch 1090/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4679 - mae: 0.4004 - mse: 0.4679 - val_loss: 0.6271 - val_mae: 0.5212 - val_mse: 0.6271\n",
      "Epoch 1091/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4705 - mae: 0.4025 - mse: 0.4705 - val_loss: 0.6234 - val_mae: 0.5212 - val_mse: 0.6234\n",
      "Epoch 1092/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4632 - mae: 0.3972 - mse: 0.4632 - val_loss: 0.6332 - val_mae: 0.5273 - val_mse: 0.6332\n",
      "Epoch 1093/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4690 - mae: 0.4013 - mse: 0.4690 - val_loss: 0.6123 - val_mae: 0.5177 - val_mse: 0.6123\n",
      "Epoch 1094/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4647 - mae: 0.4000 - mse: 0.4647 - val_loss: 0.6257 - val_mae: 0.5232 - val_mse: 0.6257\n",
      "Epoch 1095/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4672 - mae: 0.4005 - mse: 0.4672 - val_loss: 0.6283 - val_mae: 0.5255 - val_mse: 0.6283\n",
      "Epoch 1096/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4631 - mae: 0.3985 - mse: 0.4631 - val_loss: 0.6235 - val_mae: 0.5252 - val_mse: 0.6235\n",
      "Epoch 1178/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4686 - mae: 0.4011 - mse: 0.4686 - val_loss: 0.6177 - val_mae: 0.5173 - val_mse: 0.6177\n",
      "Epoch 1179/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4631 - mae: 0.3978 - mse: 0.4631 - val_loss: 0.6180 - val_mae: 0.5170 - val_mse: 0.6180\n",
      "Epoch 1180/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4676 - mae: 0.4029 - mse: 0.4676 - val_loss: 0.6136 - val_mae: 0.5183 - val_mse: 0.6136\n",
      "Epoch 1181/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4653 - mae: 0.3992 - mse: 0.4653 - val_loss: 0.6209 - val_mae: 0.5192 - val_mse: 0.6209\n",
      "Epoch 1182/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4707 - mae: 0.4019 - mse: 0.4707 - val_loss: 0.6176 - val_mae: 0.5223 - val_mse: 0.6176\n",
      "Epoch 1183/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4668 - mae: 0.4005 - mse: 0.4668 - val_loss: 0.6229 - val_mae: 0.5198 - val_mse: 0.6229\n",
      "Epoch 1184/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4695 - mae: 0.4024 - mse: 0.4695 - val_loss: 0.6113 - val_mae: 0.5128 - val_mse: 0.6113\n",
      "Epoch 1185/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4647 - mae: 0.4000 - mse: 0.4647 - val_loss: 0.6127 - val_mae: 0.5186 - val_mse: 0.6127\n",
      "Epoch 1186/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4661 - mae: 0.3996 - mse: 0.4661 - val_loss: 0.6148 - val_mae: 0.5171 - val_mse: 0.6148\n",
      "Epoch 1187/5000\n",
      "4117/4117 [==============================] - 3s 802us/step - loss: 0.4616 - mae: 0.3971 - mse: 0.4616 - val_loss: 0.6275 - val_mae: 0.5258 - val_mse: 0.6275\n",
      "Epoch 1188/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4652 - mae: 0.3981 - mse: 0.4652 - val_loss: 0.6309 - val_mae: 0.5239 - val_mse: 0.6309\n",
      "Epoch 1189/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4658 - mae: 0.3990 - mse: 0.4658 - val_loss: 0.6250 - val_mae: 0.5219 - val_mse: 0.6250\n",
      "Epoch 1190/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4693 - mae: 0.4011 - mse: 0.4693 - val_loss: 0.6233 - val_mae: 0.5219 - val_mse: 0.6233\n",
      "Epoch 1191/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4659 - mae: 0.4000 - mse: 0.4659 - val_loss: 0.6297 - val_mae: 0.5251 - val_mse: 0.6297\n",
      "Epoch 1192/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4665 - mae: 0.3994 - mse: 0.4665 - val_loss: 0.6222 - val_mae: 0.5262 - val_mse: 0.6222\n",
      "Epoch 1193/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4668 - mae: 0.4011 - mse: 0.4668 - val_loss: 0.6321 - val_mae: 0.5261 - val_mse: 0.6321\n",
      "Epoch 1194/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4648 - mae: 0.3996 - mse: 0.4648 - val_loss: 0.6106 - val_mae: 0.5133 - val_mse: 0.6106\n",
      "Epoch 1195/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4695 - mae: 0.4040 - mse: 0.4695 - val_loss: 0.6315 - val_mae: 0.5258 - val_mse: 0.6315\n",
      "Epoch 1196/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4661 - mae: 0.4007 - mse: 0.4661 - val_loss: 0.6284 - val_mae: 0.5267 - val_mse: 0.6284\n",
      "Epoch 1197/5000\n",
      "4117/4117 [==============================] - 3s 845us/step - loss: 0.4635 - mae: 0.3979 - mse: 0.4635 - val_loss: 0.6104 - val_mae: 0.5154 - val_mse: 0.6104\n",
      "Epoch 1198/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4675 - mae: 0.4008 - mse: 0.4675 - val_loss: 0.6170 - val_mae: 0.5203 - val_mse: 0.6170\n",
      "Epoch 1199/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4680 - mae: 0.4006 - mse: 0.4680 - val_loss: 0.6179 - val_mae: 0.5187 - val_mse: 0.6179\n",
      "Epoch 1200/5000\n",
      "4117/4117 [==============================] - 4s 860us/step - loss: 0.4653 - mae: 0.3995 - mse: 0.4653 - val_loss: 0.6119 - val_mae: 0.5179 - val_mse: 0.6119\n",
      "Epoch 1287/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4620 - mae: 0.3958 - mse: 0.4620 - val_loss: 0.6328 - val_mae: 0.5272 - val_mse: 0.6328\n",
      "Epoch 1288/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4673 - mae: 0.4000 - mse: 0.4673 - val_loss: 0.6155 - val_mae: 0.5183 - val_mse: 0.6155\n",
      "Epoch 1289/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4664 - mae: 0.3996 - mse: 0.4664 - val_loss: 0.6184 - val_mae: 0.5217 - val_mse: 0.6184\n",
      "Epoch 1290/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4638 - mae: 0.3961 - mse: 0.4638 - val_loss: 0.6341 - val_mae: 0.5265 - val_mse: 0.6341\n",
      "Epoch 1291/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4644 - mae: 0.3973 - mse: 0.4644 - val_loss: 0.6343 - val_mae: 0.5277 - val_mse: 0.6343\n",
      "Epoch 1292/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4672 - mae: 0.3997 - mse: 0.4672 - val_loss: 0.6193 - val_mae: 0.5184 - val_mse: 0.6193\n",
      "Epoch 1293/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4666 - mae: 0.3987 - mse: 0.4666 - val_loss: 0.6209 - val_mae: 0.5211 - val_mse: 0.6209\n",
      "Epoch 1294/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4643 - mae: 0.3988 - mse: 0.4643 - val_loss: 0.6204 - val_mae: 0.5233 - val_mse: 0.6204\n",
      "Epoch 1295/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4655 - mae: 0.3993 - mse: 0.4655 - val_loss: 0.6273 - val_mae: 0.5261 - val_mse: 0.6273\n",
      "Epoch 1296/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4625 - mae: 0.3969 - mse: 0.4625 - val_loss: 0.6249 - val_mae: 0.5226 - val_mse: 0.6249\n",
      "Epoch 1297/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4673 - mae: 0.4007 - mse: 0.4673 - val_loss: 0.6279 - val_mae: 0.5235 - val_mse: 0.6279\n",
      "Epoch 1298/5000\n",
      "4117/4117 [==============================] - 3s 834us/step - loss: 0.4645 - mae: 0.3991 - mse: 0.4645 - val_loss: 0.6147 - val_mae: 0.5204 - val_mse: 0.6147\n",
      "Epoch 1299/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4648 - mae: 0.3976 - mse: 0.4648 - val_loss: 0.6286 - val_mae: 0.5228 - val_mse: 0.6286\n",
      "Epoch 1300/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4647 - mae: 0.3971 - mse: 0.4647 - val_loss: 0.6322 - val_mae: 0.5250 - val_mse: 0.6322\n",
      "Epoch 1301/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4637 - mae: 0.3976 - mse: 0.4637 - val_loss: 0.6217 - val_mae: 0.5205 - val_mse: 0.6217\n",
      "Epoch 1302/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4691 - mae: 0.4009 - mse: 0.4691 - val_loss: 0.6211 - val_mae: 0.5200 - val_mse: 0.6211\n",
      "Epoch 1303/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4623 - mae: 0.3958 - mse: 0.4623 - val_loss: 0.6139 - val_mae: 0.5145 - val_mse: 0.6139\n",
      "Epoch 1304/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4659 - mae: 0.3990 - mse: 0.4659 - val_loss: 0.6177 - val_mae: 0.5163 - val_mse: 0.6177\n",
      "Epoch 1305/5000\n",
      "4117/4117 [==============================] - 3s 803us/step - loss: 0.4689 - mae: 0.4010 - mse: 0.4689 - val_loss: 0.6422 - val_mae: 0.5336 - val_mse: 0.6422\n",
      "Epoch 1306/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4697 - mae: 0.3989 - mse: 0.4697 - val_loss: 0.6263 - val_mae: 0.5231 - val_mse: 0.6263\n",
      "Epoch 1307/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4667 - mae: 0.3985 - mse: 0.4667 - val_loss: 0.6270 - val_mae: 0.5230 - val_mse: 0.6270\n",
      "Epoch 1308/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4686 - mae: 0.4010 - mse: 0.4686 - val_loss: 0.6296 - val_mae: 0.5241 - val_mse: 0.6296\n",
      "Epoch 1309/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4649 - mae: 0.3975 - mse: 0.4649 - val_loss: 0.6244 - val_mae: 0.5198 - val_mse: 0.6244\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4682 - mae: 0.3985 - mse: 0.4682 - val_loss: 0.6198 - val_mae: 0.5172 - val_mse: 0.6198\n",
      "Epoch 1402/5000\n",
      "4117/4117 [==============================] - 3s 794us/step - loss: 0.4704 - mae: 0.4044 - mse: 0.4704 - val_loss: 0.6360 - val_mae: 0.5276 - val_mse: 0.6360\n",
      "Epoch 1403/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4625 - mae: 0.3964 - mse: 0.4625 - val_loss: 0.6250 - val_mae: 0.5217 - val_mse: 0.6250\n",
      "Epoch 1404/5000\n",
      "4117/4117 [==============================] - 3s 834us/step - loss: 0.4629 - mae: 0.3959 - mse: 0.4629 - val_loss: 0.6253 - val_mae: 0.5225 - val_mse: 0.6253\n",
      "Epoch 1405/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4676 - mae: 0.3991 - mse: 0.4676 - val_loss: 0.6346 - val_mae: 0.5299 - val_mse: 0.6346\n",
      "Epoch 1406/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4703 - mae: 0.4012 - mse: 0.4703 - val_loss: 0.6235 - val_mae: 0.5227 - val_mse: 0.6235\n",
      "Epoch 1407/5000\n",
      "4117/4117 [==============================] - 3s 801us/step - loss: 0.4660 - mae: 0.3982 - mse: 0.4660 - val_loss: 0.6259 - val_mae: 0.5210 - val_mse: 0.6259\n",
      "Epoch 1408/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4644 - mae: 0.3983 - mse: 0.4644 - val_loss: 0.6268 - val_mae: 0.5228 - val_mse: 0.6268\n",
      "Epoch 1409/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4639 - mae: 0.3976 - mse: 0.4639 - val_loss: 0.6246 - val_mae: 0.5177 - val_mse: 0.6246\n",
      "Epoch 1410/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4660 - mae: 0.3995 - mse: 0.4660 - val_loss: 0.6391 - val_mae: 0.5306 - val_mse: 0.6391\n",
      "Epoch 1411/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4593 - mae: 0.3944 - mse: 0.4593 - val_loss: 0.6305 - val_mae: 0.5244 - val_mse: 0.6305\n",
      "Epoch 1412/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4700 - mae: 0.4010 - mse: 0.4700 - val_loss: 0.6446 - val_mae: 0.5379 - val_mse: 0.6446\n",
      "Epoch 1413/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4670 - mae: 0.3997 - mse: 0.4670 - val_loss: 0.6372 - val_mae: 0.5346 - val_mse: 0.6372\n",
      "Epoch 1414/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4715 - mae: 0.4035 - mse: 0.4715 - val_loss: 0.6177 - val_mae: 0.5185 - val_mse: 0.6177\n",
      "Epoch 1415/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4603 - mae: 0.3954 - mse: 0.4603 - val_loss: 0.6451 - val_mae: 0.5346 - val_mse: 0.6451\n",
      "Epoch 1416/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4653 - mae: 0.3970 - mse: 0.4653 - val_loss: 0.6176 - val_mae: 0.5180 - val_mse: 0.6176\n",
      "Epoch 1417/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4638 - mae: 0.3970 - mse: 0.4638 - val_loss: 0.6096 - val_mae: 0.5130 - val_mse: 0.6096\n",
      "Epoch 1418/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4665 - mae: 0.3993 - mse: 0.4665 - val_loss: 0.6224 - val_mae: 0.5170 - val_mse: 0.6224\n",
      "Epoch 1419/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4658 - mae: 0.3993 - mse: 0.4658 - val_loss: 0.6236 - val_mae: 0.5196 - val_mse: 0.6236\n",
      "Epoch 1420/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4624 - mae: 0.3964 - mse: 0.4624 - val_loss: 0.6143 - val_mae: 0.5153 - val_mse: 0.6143\n",
      "Epoch 1421/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4647 - mae: 0.3981 - mse: 0.4647 - val_loss: 0.6202 - val_mae: 0.5186 - val_mse: 0.6202\n",
      "Epoch 1422/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4639 - mae: 0.3969 - mse: 0.4639 - val_loss: 0.6273 - val_mae: 0.5211 - val_mse: 0.6273\n",
      "Epoch 1423/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4647 - mae: 0.3973 - mse: 0.4647 - val_loss: 0.6214 - val_mae: 0.5190 - val_mse: 0.6214\n",
      "Epoch 1424/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4615 - mae: 0.3946 - mse: 0.4615 - val_loss: 0.6304 - val_mae: 0.5238 - val_mse: 0.6304\n",
      "Epoch 1521/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4649 - mae: 0.3985 - mse: 0.4649 - val_loss: 0.6387 - val_mae: 0.5293 - val_mse: 0.6387\n",
      "Epoch 1522/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4653 - mae: 0.3981 - mse: 0.4653 - val_loss: 0.6326 - val_mae: 0.5245 - val_mse: 0.6326\n",
      "Epoch 1523/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4647 - mae: 0.3985 - mse: 0.4647 - val_loss: 0.6337 - val_mae: 0.5272 - val_mse: 0.6337\n",
      "Epoch 1524/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4674 - mae: 0.3994 - mse: 0.4674 - val_loss: 0.6422 - val_mae: 0.5316 - val_mse: 0.6422\n",
      "Epoch 1525/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4633 - mae: 0.3956 - mse: 0.4633 - val_loss: 0.6273 - val_mae: 0.5197 - val_mse: 0.6273\n",
      "Epoch 1526/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4648 - mae: 0.3970 - mse: 0.4648 - val_loss: 0.6256 - val_mae: 0.5233 - val_mse: 0.6256\n",
      "Epoch 1527/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4697 - mae: 0.4021 - mse: 0.4697 - val_loss: 0.6436 - val_mae: 0.5326 - val_mse: 0.6436\n",
      "Epoch 1528/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4647 - mae: 0.3979 - mse: 0.4647 - val_loss: 0.6462 - val_mae: 0.5350 - val_mse: 0.6462\n",
      "Epoch 1529/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4618 - mae: 0.3954 - mse: 0.4618 - val_loss: 0.6228 - val_mae: 0.5220 - val_mse: 0.6228\n",
      "Epoch 1530/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4628 - mae: 0.3957 - mse: 0.4628 - val_loss: 0.6370 - val_mae: 0.5290 - val_mse: 0.6370\n",
      "Epoch 1531/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4632 - mae: 0.3967 - mse: 0.4632 - val_loss: 0.6262 - val_mae: 0.5229 - val_mse: 0.6262\n",
      "Epoch 1532/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4602 - mae: 0.3932 - mse: 0.4602 - val_loss: 0.6269 - val_mae: 0.5221 - val_mse: 0.6269\n",
      "Epoch 1533/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4687 - mae: 0.3994 - mse: 0.4687 - val_loss: 0.6289 - val_mae: 0.5224 - val_mse: 0.6289\n",
      "Epoch 1534/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4669 - mae: 0.3993 - mse: 0.4669 - val_loss: 0.6464 - val_mae: 0.5333 - val_mse: 0.6464\n",
      "Epoch 1535/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4656 - mae: 0.3985 - mse: 0.4656 - val_loss: 0.6200 - val_mae: 0.5217 - val_mse: 0.6200\n",
      "Epoch 1536/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4635 - mae: 0.3971 - mse: 0.4635 - val_loss: 0.6317 - val_mae: 0.5244 - val_mse: 0.6317\n",
      "Epoch 1537/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4639 - mae: 0.3949 - mse: 0.4639 - val_loss: 0.6244 - val_mae: 0.5193 - val_mse: 0.6244\n",
      "Epoch 1538/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4669 - mae: 0.3975 - mse: 0.4669 - val_loss: 0.6318 - val_mae: 0.5253 - val_mse: 0.6318\n",
      "Epoch 1539/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4614 - mae: 0.3945 - mse: 0.4614 - val_loss: 0.6281 - val_mae: 0.5238 - val_mse: 0.6281\n",
      "Epoch 1540/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4628 - mae: 0.3946 - mse: 0.4628 - val_loss: 0.6208 - val_mae: 0.5213 - val_mse: 0.6208\n",
      "Epoch 1541/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4683 - mae: 0.3996 - mse: 0.4683 - val_loss: 0.6166 - val_mae: 0.5186 - val_mse: 0.6166\n",
      "Epoch 1542/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4608 - mae: 0.3950 - mse: 0.4608 - val_loss: 0.6375 - val_mae: 0.5280 - val_mse: 0.6375\n",
      "Epoch 1543/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4623 - mae: 0.3960 - mse: 0.4623 - val_loss: 0.6667 - val_mae: 0.5380 - val_mse: 0.6667\n",
      "Epoch 1544/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4633 - mae: 0.3962 - mse: 0.4633 - val_loss: 0.6464 - val_mae: 0.5351 - val_mse: 0.6464\n",
      "Epoch 1646/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4668 - mae: 0.4002 - mse: 0.4668 - val_loss: 0.6371 - val_mae: 0.5287 - val_mse: 0.6371\n",
      "Epoch 1647/5000\n",
      "4117/4117 [==============================] - 3s 802us/step - loss: 0.4700 - mae: 0.4015 - mse: 0.4700 - val_loss: 0.6329 - val_mae: 0.5237 - val_mse: 0.6329\n",
      "Epoch 1648/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4676 - mae: 0.3999 - mse: 0.4676 - val_loss: 0.6298 - val_mae: 0.5249 - val_mse: 0.6298\n",
      "Epoch 1649/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4628 - mae: 0.3961 - mse: 0.4628 - val_loss: 0.6227 - val_mae: 0.5176 - val_mse: 0.6227\n",
      "Epoch 1650/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4649 - mae: 0.3963 - mse: 0.4649 - val_loss: 0.6258 - val_mae: 0.5197 - val_mse: 0.6258\n",
      "Epoch 1651/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4661 - mae: 0.3984 - mse: 0.4661 - val_loss: 0.6195 - val_mae: 0.5181 - val_mse: 0.6195\n",
      "Epoch 1652/5000\n",
      "4117/4117 [==============================] - 3s 792us/step - loss: 0.4665 - mae: 0.4002 - mse: 0.4665 - val_loss: 0.6256 - val_mae: 0.5203 - val_mse: 0.6256\n",
      "Epoch 1653/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4671 - mae: 0.3980 - mse: 0.4671 - val_loss: 0.6186 - val_mae: 0.5119 - val_mse: 0.6186\n",
      "Epoch 1654/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4643 - mae: 0.3966 - mse: 0.4643 - val_loss: 0.6362 - val_mae: 0.5309 - val_mse: 0.6362\n",
      "Epoch 1655/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4622 - mae: 0.3954 - mse: 0.4622 - val_loss: 0.6192 - val_mae: 0.5151 - val_mse: 0.6192\n",
      "Epoch 1656/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4630 - mae: 0.3955 - mse: 0.4630 - val_loss: 0.6175 - val_mae: 0.5166 - val_mse: 0.6175\n",
      "Epoch 1657/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4663 - mae: 0.3974 - mse: 0.4663 - val_loss: 0.6376 - val_mae: 0.5306 - val_mse: 0.6376\n",
      "Epoch 1658/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4647 - mae: 0.3970 - mse: 0.4647 - val_loss: 0.6392 - val_mae: 0.5273 - val_mse: 0.6392\n",
      "Epoch 1659/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4647 - mae: 0.3975 - mse: 0.4647 - val_loss: 0.6444 - val_mae: 0.5292 - val_mse: 0.6444\n",
      "Epoch 1660/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4658 - mae: 0.3982 - mse: 0.4658 - val_loss: 0.6270 - val_mae: 0.5220 - val_mse: 0.6270\n",
      "Epoch 1661/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4668 - mae: 0.3983 - mse: 0.4668 - val_loss: 0.6424 - val_mae: 0.5290 - val_mse: 0.6424\n",
      "Epoch 1662/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4642 - mae: 0.3959 - mse: 0.4642 - val_loss: 0.6434 - val_mae: 0.5275 - val_mse: 0.6434\n",
      "Epoch 1663/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4610 - mae: 0.3942 - mse: 0.4610 - val_loss: 0.6476 - val_mae: 0.5341 - val_mse: 0.6476\n",
      "Epoch 1664/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4644 - mae: 0.3958 - mse: 0.4644 - val_loss: 0.6300 - val_mae: 0.5217 - val_mse: 0.6300\n",
      "Epoch 1665/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4610 - mae: 0.3938 - mse: 0.4610 - val_loss: 0.6460 - val_mae: 0.5329 - val_mse: 0.6460\n",
      "Epoch 1666/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4653 - mae: 0.3982 - mse: 0.4653 - val_loss: 0.6339 - val_mae: 0.5251 - val_mse: 0.6339\n",
      "Epoch 1667/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4660 - mae: 0.3987 - mse: 0.4660 - val_loss: 0.6321 - val_mae: 0.5262 - val_mse: 0.6321\n",
      "Epoch 1668/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4665 - mae: 0.3970 - mse: 0.4665 - val_loss: 0.6790 - val_mae: 0.5512 - val_mse: 0.6790\n",
      "Epoch 1669/5000\n",
      "4117/4117 [==============================] - 3s 834us/step - loss: 0.4618 - mae: 0.3946 - mse: 0.4618 - val_loss: 0.6296 - val_mae: 0.5215 - val_mse: 0.6296\n",
      "Epoch 1777/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4656 - mae: 0.3958 - mse: 0.4656 - val_loss: 0.6333 - val_mae: 0.5257 - val_mse: 0.6333\n",
      "Epoch 1778/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4622 - mae: 0.3945 - mse: 0.4622 - val_loss: 0.6320 - val_mae: 0.5222 - val_mse: 0.6320\n",
      "Epoch 1779/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4666 - mae: 0.3972 - mse: 0.4666 - val_loss: 0.6288 - val_mae: 0.5230 - val_mse: 0.6288\n",
      "Epoch 1780/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4625 - mae: 0.3957 - mse: 0.4625 - val_loss: 0.6355 - val_mae: 0.5268 - val_mse: 0.6355\n",
      "Epoch 1781/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4643 - mae: 0.3976 - mse: 0.4643 - val_loss: 0.6380 - val_mae: 0.5321 - val_mse: 0.6380\n",
      "Epoch 1782/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4625 - mae: 0.3953 - mse: 0.4625 - val_loss: 0.6450 - val_mae: 0.5331 - val_mse: 0.6450\n",
      "Epoch 1783/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4655 - mae: 0.3975 - mse: 0.4655 - val_loss: 0.6340 - val_mae: 0.5221 - val_mse: 0.6340\n",
      "Epoch 1784/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4621 - mae: 0.3953 - mse: 0.4621 - val_loss: 0.6346 - val_mae: 0.5264 - val_mse: 0.6346\n",
      "Epoch 1785/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4655 - mae: 0.3963 - mse: 0.4655 - val_loss: 0.6397 - val_mae: 0.5286 - val_mse: 0.6397\n",
      "Epoch 1786/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4677 - mae: 0.4004 - mse: 0.4677 - val_loss: 0.6489 - val_mae: 0.5334 - val_mse: 0.6489\n",
      "Epoch 1787/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4631 - mae: 0.3953 - mse: 0.4631 - val_loss: 0.6222 - val_mae: 0.5184 - val_mse: 0.6222\n",
      "Epoch 1788/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4640 - mae: 0.3958 - mse: 0.4640 - val_loss: 0.6208 - val_mae: 0.5143 - val_mse: 0.6208\n",
      "Epoch 1789/5000\n",
      "4117/4117 [==============================] - 3s 798us/step - loss: 0.4652 - mae: 0.3979 - mse: 0.4652 - val_loss: 0.6367 - val_mae: 0.5266 - val_mse: 0.6367\n",
      "Epoch 1790/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4671 - mae: 0.3989 - mse: 0.4671 - val_loss: 0.6275 - val_mae: 0.5232 - val_mse: 0.6275\n",
      "Epoch 1791/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4620 - mae: 0.3948 - mse: 0.4620 - val_loss: 0.6324 - val_mae: 0.5253 - val_mse: 0.6324\n",
      "Epoch 1792/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4658 - mae: 0.3983 - mse: 0.4658 - val_loss: 0.6252 - val_mae: 0.5249 - val_mse: 0.6252\n",
      "Epoch 1793/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4639 - mae: 0.3958 - mse: 0.4639 - val_loss: 0.6234 - val_mae: 0.5206 - val_mse: 0.6234\n",
      "Epoch 1794/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4607 - mae: 0.3949 - mse: 0.4607 - val_loss: 0.6180 - val_mae: 0.5182 - val_mse: 0.6180\n",
      "Epoch 1795/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4660 - mae: 0.3994 - mse: 0.4660 - val_loss: 0.6299 - val_mae: 0.5270 - val_mse: 0.6299\n",
      "Epoch 1796/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4647 - mae: 0.3968 - mse: 0.4647 - val_loss: 0.6282 - val_mae: 0.5220 - val_mse: 0.6282\n",
      "Epoch 1797/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4619 - mae: 0.3948 - mse: 0.4619 - val_loss: 0.6311 - val_mae: 0.5211 - val_mse: 0.6311\n",
      "Epoch 1798/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4658 - mae: 0.3968 - mse: 0.4658 - val_loss: 0.6333 - val_mae: 0.5287 - val_mse: 0.6333\n",
      "Epoch 1799/5000\n",
      "4117/4117 [==============================] - 3s 838us/step - loss: 0.4606 - mae: 0.3933 - mse: 0.4606 - val_loss: 0.6267 - val_mae: 0.5251 - val_mse: 0.6267\n",
      "Epoch 1913/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4634 - mae: 0.3972 - mse: 0.4634 - val_loss: 0.6270 - val_mae: 0.5241 - val_mse: 0.6270\n",
      "Epoch 1914/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4641 - mae: 0.3953 - mse: 0.4641 - val_loss: 0.6273 - val_mae: 0.5259 - val_mse: 0.6273\n",
      "Epoch 1915/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4600 - mae: 0.3926 - mse: 0.4600 - val_loss: 0.6172 - val_mae: 0.5179 - val_mse: 0.6172\n",
      "Epoch 1916/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4642 - mae: 0.3952 - mse: 0.4642 - val_loss: 0.6365 - val_mae: 0.5277 - val_mse: 0.6365\n",
      "Epoch 1917/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4625 - mae: 0.3938 - mse: 0.4625 - val_loss: 0.6305 - val_mae: 0.5259 - val_mse: 0.6305\n",
      "Epoch 1918/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4631 - mae: 0.3954 - mse: 0.4631 - val_loss: 0.6393 - val_mae: 0.5289 - val_mse: 0.6393\n",
      "Epoch 1919/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4641 - mae: 0.3956 - mse: 0.4641 - val_loss: 0.6353 - val_mae: 0.5262 - val_mse: 0.6353\n",
      "Epoch 1920/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4625 - mae: 0.3957 - mse: 0.4625 - val_loss: 0.6303 - val_mae: 0.5220 - val_mse: 0.6303\n",
      "Epoch 1921/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4651 - mae: 0.3971 - mse: 0.4651 - val_loss: 0.6727 - val_mae: 0.5526 - val_mse: 0.6727\n",
      "Epoch 1922/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4627 - mae: 0.3944 - mse: 0.4627 - val_loss: 0.6399 - val_mae: 0.5307 - val_mse: 0.6399\n",
      "Epoch 1923/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4616 - mae: 0.3947 - mse: 0.4616 - val_loss: 0.6283 - val_mae: 0.5238 - val_mse: 0.6283\n",
      "Epoch 1924/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4646 - mae: 0.3972 - mse: 0.4646 - val_loss: 0.6358 - val_mae: 0.5253 - val_mse: 0.6358\n",
      "Epoch 1925/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4638 - mae: 0.3964 - mse: 0.4638 - val_loss: 0.6281 - val_mae: 0.5262 - val_mse: 0.6281\n",
      "Epoch 1926/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4641 - mae: 0.3954 - mse: 0.4641 - val_loss: 0.6417 - val_mae: 0.5315 - val_mse: 0.6417\n",
      "Epoch 1927/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4589 - mae: 0.3917 - mse: 0.4589 - val_loss: 0.6462 - val_mae: 0.5315 - val_mse: 0.6462\n",
      "Epoch 1928/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4641 - mae: 0.3961 - mse: 0.4641 - val_loss: 0.6334 - val_mae: 0.5253 - val_mse: 0.6334\n",
      "Epoch 1929/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4673 - mae: 0.4003 - mse: 0.4673 - val_loss: 0.6306 - val_mae: 0.5251 - val_mse: 0.6306\n",
      "Epoch 1930/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4681 - mae: 0.3989 - mse: 0.4681 - val_loss: 0.6335 - val_mae: 0.5231 - val_mse: 0.6335\n",
      "Epoch 1931/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4666 - mae: 0.3970 - mse: 0.4666 - val_loss: 0.6528 - val_mae: 0.5371 - val_mse: 0.6528\n",
      "Epoch 1932/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4644 - mae: 0.3972 - mse: 0.4644 - val_loss: 0.6309 - val_mae: 0.5201 - val_mse: 0.6309\n",
      "Epoch 1933/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4631 - mae: 0.3953 - mse: 0.4631 - val_loss: 0.6659 - val_mae: 0.5454 - val_mse: 0.6659\n",
      "Epoch 1934/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4632 - mae: 0.3954 - mse: 0.4632 - val_loss: 0.6319 - val_mae: 0.5268 - val_mse: 0.6319\n",
      "Epoch 1935/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4655 - mae: 0.3969 - mse: 0.4655 - val_loss: 0.6415 - val_mae: 0.5309 - val_mse: 0.6415\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4653 - mae: 0.3967 - mse: 0.4653 - val_loss: 0.6362 - val_mae: 0.5296 - val_mse: 0.6362\n",
      "Epoch 2054/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4646 - mae: 0.3972 - mse: 0.4646 - val_loss: 0.6383 - val_mae: 0.5258 - val_mse: 0.6383\n",
      "Epoch 2055/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4708 - mae: 0.3971 - mse: 0.4708 - val_loss: 0.6221 - val_mae: 0.5197 - val_mse: 0.6221\n",
      "Epoch 2056/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4606 - mae: 0.3943 - mse: 0.4606 - val_loss: 0.6391 - val_mae: 0.5272 - val_mse: 0.6391\n",
      "Epoch 2057/5000\n",
      "4117/4117 [==============================] - 3s 798us/step - loss: 0.4636 - mae: 0.3949 - mse: 0.4636 - val_loss: 0.6244 - val_mae: 0.5177 - val_mse: 0.6244\n",
      "Epoch 2058/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4644 - mae: 0.3985 - mse: 0.4644 - val_loss: 0.6335 - val_mae: 0.5263 - val_mse: 0.6335\n",
      "Epoch 2059/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4643 - mae: 0.3956 - mse: 0.4643 - val_loss: 0.6588 - val_mae: 0.5452 - val_mse: 0.6588\n",
      "Epoch 2060/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4644 - mae: 0.3958 - mse: 0.4644 - val_loss: 0.6315 - val_mae: 0.5203 - val_mse: 0.6315\n",
      "Epoch 2061/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4674 - mae: 0.3980 - mse: 0.4674 - val_loss: 0.6262 - val_mae: 0.5207 - val_mse: 0.6262\n",
      "Epoch 2062/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4634 - mae: 0.3954 - mse: 0.4634 - val_loss: 0.6347 - val_mae: 0.5229 - val_mse: 0.6347\n",
      "Epoch 2063/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4644 - mae: 0.3960 - mse: 0.4644 - val_loss: 0.6329 - val_mae: 0.5242 - val_mse: 0.6329\n",
      "Epoch 2064/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4624 - mae: 0.3941 - mse: 0.4624 - val_loss: 0.6275 - val_mae: 0.5204 - val_mse: 0.6275\n",
      "Epoch 2065/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4649 - mae: 0.3972 - mse: 0.4649 - val_loss: 0.6381 - val_mae: 0.5241 - val_mse: 0.6381\n",
      "Epoch 2066/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4621 - mae: 0.3945 - mse: 0.4621 - val_loss: 0.6396 - val_mae: 0.5285 - val_mse: 0.6396\n",
      "Epoch 2067/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4658 - mae: 0.3973 - mse: 0.4658 - val_loss: 0.6416 - val_mae: 0.5272 - val_mse: 0.6416\n",
      "Epoch 2068/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4608 - mae: 0.3935 - mse: 0.4608 - val_loss: 0.6445 - val_mae: 0.5331 - val_mse: 0.6445\n",
      "Epoch 2069/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4704 - mae: 0.4008 - mse: 0.4704 - val_loss: 0.6453 - val_mae: 0.5272 - val_mse: 0.6453\n",
      "Epoch 2070/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4628 - mae: 0.3963 - mse: 0.4628 - val_loss: 0.6378 - val_mae: 0.5260 - val_mse: 0.6378\n",
      "Epoch 2071/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4627 - mae: 0.3958 - mse: 0.4627 - val_loss: 0.6216 - val_mae: 0.5182 - val_mse: 0.6216\n",
      "Epoch 2072/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4635 - mae: 0.3956 - mse: 0.4635 - val_loss: 0.6344 - val_mae: 0.5190 - val_mse: 0.6344\n",
      "Epoch 2073/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4626 - mae: 0.3949 - mse: 0.4626 - val_loss: 0.6317 - val_mae: 0.5215 - val_mse: 0.6317\n",
      "Epoch 2074/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4601 - mae: 0.3921 - mse: 0.4601 - val_loss: 0.6391 - val_mae: 0.5276 - val_mse: 0.6391\n",
      "Epoch 2075/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4627 - mae: 0.3957 - mse: 0.4627 - val_loss: 0.6304 - val_mae: 0.5215 - val_mse: 0.6304\n",
      "Epoch 2076/5000\n",
      "4117/4117 [==============================] - 3s 837us/step - loss: 0.4631 - mae: 0.3966 - mse: 0.4631 - val_loss: 0.6273 - val_mae: 0.5213 - val_mse: 0.6273\n",
      "Epoch 2200/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4659 - mae: 0.3966 - mse: 0.4659 - val_loss: 0.6264 - val_mae: 0.5235 - val_mse: 0.6264\n",
      "Epoch 2201/5000\n",
      "4117/4117 [==============================] - 3s 794us/step - loss: 0.4600 - mae: 0.3931 - mse: 0.4600 - val_loss: 0.6485 - val_mae: 0.5323 - val_mse: 0.6485\n",
      "Epoch 2202/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4678 - mae: 0.3988 - mse: 0.4678 - val_loss: 0.6466 - val_mae: 0.5304 - val_mse: 0.6466\n",
      "Epoch 2203/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4659 - mae: 0.3979 - mse: 0.4659 - val_loss: 0.6338 - val_mae: 0.5269 - val_mse: 0.6338\n",
      "Epoch 2204/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4625 - mae: 0.3961 - mse: 0.4625 - val_loss: 0.6500 - val_mae: 0.5314 - val_mse: 0.6500\n",
      "Epoch 2205/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4610 - mae: 0.3934 - mse: 0.4610 - val_loss: 0.6403 - val_mae: 0.5279 - val_mse: 0.6403\n",
      "Epoch 2206/5000\n",
      "4117/4117 [==============================] - 3s 802us/step - loss: 0.4621 - mae: 0.3949 - mse: 0.4621 - val_loss: 0.6459 - val_mae: 0.5320 - val_mse: 0.6459\n",
      "Epoch 2207/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4620 - mae: 0.3943 - mse: 0.4620 - val_loss: 0.6403 - val_mae: 0.5282 - val_mse: 0.6403\n",
      "Epoch 2208/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4639 - mae: 0.3952 - mse: 0.4639 - val_loss: 0.6263 - val_mae: 0.5198 - val_mse: 0.6263\n",
      "Epoch 2209/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4606 - mae: 0.3942 - mse: 0.4606 - val_loss: 0.6291 - val_mae: 0.5261 - val_mse: 0.6291\n",
      "Epoch 2210/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4627 - mae: 0.3949 - mse: 0.4627 - val_loss: 0.6156 - val_mae: 0.5138 - val_mse: 0.6156\n",
      "Epoch 2211/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4613 - mae: 0.3939 - mse: 0.4613 - val_loss: 0.6316 - val_mae: 0.5222 - val_mse: 0.6316\n",
      "Epoch 2212/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4608 - mae: 0.3945 - mse: 0.4608 - val_loss: 0.6328 - val_mae: 0.5275 - val_mse: 0.6328\n",
      "Epoch 2213/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4601 - mae: 0.3937 - mse: 0.4601 - val_loss: 0.6407 - val_mae: 0.5289 - val_mse: 0.6407\n",
      "Epoch 2214/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4640 - mae: 0.3969 - mse: 0.4640 - val_loss: 0.6401 - val_mae: 0.5255 - val_mse: 0.6401\n",
      "Epoch 2215/5000\n",
      "4117/4117 [==============================] - 3s 803us/step - loss: 0.4671 - mae: 0.3971 - mse: 0.4671 - val_loss: 0.6378 - val_mae: 0.5264 - val_mse: 0.6378\n",
      "Epoch 2216/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4629 - mae: 0.3953 - mse: 0.4629 - val_loss: 0.6451 - val_mae: 0.5319 - val_mse: 0.6451\n",
      "Epoch 2217/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4604 - mae: 0.3943 - mse: 0.4604 - val_loss: 0.6390 - val_mae: 0.5251 - val_mse: 0.6390\n",
      "Epoch 2218/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4621 - mae: 0.3949 - mse: 0.4621 - val_loss: 0.6366 - val_mae: 0.5235 - val_mse: 0.6366\n",
      "Epoch 2219/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4673 - mae: 0.3975 - mse: 0.4673 - val_loss: 0.6255 - val_mae: 0.5201 - val_mse: 0.6255\n",
      "Epoch 2220/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4655 - mae: 0.3975 - mse: 0.4655 - val_loss: 0.6355 - val_mae: 0.5271 - val_mse: 0.6355\n",
      "Epoch 2221/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4634 - mae: 0.3970 - mse: 0.4634 - val_loss: 0.6209 - val_mae: 0.5202 - val_mse: 0.6209\n",
      "Epoch 2222/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4626 - mae: 0.3960 - mse: 0.4626 - val_loss: 0.6484 - val_mae: 0.5326 - val_mse: 0.6484\n",
      "Epoch 2351/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4658 - mae: 0.3972 - mse: 0.4658 - val_loss: 0.6434 - val_mae: 0.5310 - val_mse: 0.6434\n",
      "Epoch 2352/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4641 - mae: 0.3951 - mse: 0.4641 - val_loss: 0.6448 - val_mae: 0.5312 - val_mse: 0.6448\n",
      "Epoch 2353/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4645 - mae: 0.3952 - mse: 0.4645 - val_loss: 0.6265 - val_mae: 0.5214 - val_mse: 0.6265\n",
      "Epoch 2354/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4639 - mae: 0.3955 - mse: 0.4639 - val_loss: 0.6277 - val_mae: 0.5210 - val_mse: 0.6277\n",
      "Epoch 2355/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4610 - mae: 0.3933 - mse: 0.4610 - val_loss: 0.7025 - val_mae: 0.5685 - val_mse: 0.7025\n",
      "Epoch 2356/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4673 - mae: 0.3969 - mse: 0.4673 - val_loss: 0.6561 - val_mae: 0.5361 - val_mse: 0.6561\n",
      "Epoch 2357/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4641 - mae: 0.3951 - mse: 0.4641 - val_loss: 0.6363 - val_mae: 0.5285 - val_mse: 0.6363\n",
      "Epoch 2358/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4641 - mae: 0.3969 - mse: 0.4641 - val_loss: 0.6567 - val_mae: 0.5369 - val_mse: 0.6567\n",
      "Epoch 2359/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4645 - mae: 0.3969 - mse: 0.4645 - val_loss: 0.6545 - val_mae: 0.5383 - val_mse: 0.6545\n",
      "Epoch 2360/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4591 - mae: 0.3926 - mse: 0.4591 - val_loss: 0.6409 - val_mae: 0.5285 - val_mse: 0.6409\n",
      "Epoch 2361/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4592 - mae: 0.3928 - mse: 0.4592 - val_loss: 0.6567 - val_mae: 0.5386 - val_mse: 0.6567\n",
      "Epoch 2362/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4675 - mae: 0.3982 - mse: 0.4675 - val_loss: 0.6582 - val_mae: 0.5395 - val_mse: 0.6582\n",
      "Epoch 2363/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4581 - mae: 0.3921 - mse: 0.4581 - val_loss: 0.6322 - val_mae: 0.5228 - val_mse: 0.6322\n",
      "Epoch 2364/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4621 - mae: 0.3941 - mse: 0.4621 - val_loss: 0.6522 - val_mae: 0.5343 - val_mse: 0.6522\n",
      "Epoch 2365/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4606 - mae: 0.3932 - mse: 0.4606 - val_loss: 0.6417 - val_mae: 0.5276 - val_mse: 0.6417\n",
      "Epoch 2366/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4588 - mae: 0.3917 - mse: 0.4588 - val_loss: 0.6458 - val_mae: 0.5343 - val_mse: 0.6458\n",
      "Epoch 2367/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4633 - mae: 0.3937 - mse: 0.4633 - val_loss: 0.6326 - val_mae: 0.5224 - val_mse: 0.6326\n",
      "Epoch 2368/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4685 - mae: 0.3979 - mse: 0.4685 - val_loss: 0.6445 - val_mae: 0.5332 - val_mse: 0.6445\n",
      "Epoch 2369/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4651 - mae: 0.3957 - mse: 0.4651 - val_loss: 0.6411 - val_mae: 0.5304 - val_mse: 0.6411\n",
      "Epoch 2370/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4629 - mae: 0.3962 - mse: 0.4629 - val_loss: 0.6504 - val_mae: 0.5357 - val_mse: 0.6504\n",
      "Epoch 2371/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4661 - mae: 0.3965 - mse: 0.4661 - val_loss: 0.6512 - val_mae: 0.5338 - val_mse: 0.6512\n",
      "Epoch 2372/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4675 - mae: 0.3978 - mse: 0.4675 - val_loss: 0.6495 - val_mae: 0.5347 - val_mse: 0.6495\n",
      "Epoch 2373/5000\n",
      "4117/4117 [==============================] - 3s 836us/step - loss: 0.4649 - mae: 0.3964 - mse: 0.4649 - val_loss: 0.6566 - val_mae: 0.5384 - val_mse: 0.6566\n",
      "Epoch 2507/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4668 - mae: 0.3985 - mse: 0.4668 - val_loss: 0.6460 - val_mae: 0.5303 - val_mse: 0.6460\n",
      "Epoch 2508/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4674 - mae: 0.3983 - mse: 0.4674 - val_loss: 0.6422 - val_mae: 0.5363 - val_mse: 0.6422\n",
      "Epoch 2509/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4627 - mae: 0.3938 - mse: 0.4627 - val_loss: 0.6337 - val_mae: 0.5233 - val_mse: 0.6337\n",
      "Epoch 2510/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4615 - mae: 0.3943 - mse: 0.4615 - val_loss: 0.6414 - val_mae: 0.5290 - val_mse: 0.6414\n",
      "Epoch 2511/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4708 - mae: 0.3993 - mse: 0.4708 - val_loss: 0.6394 - val_mae: 0.5276 - val_mse: 0.6394\n",
      "Epoch 2512/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4613 - mae: 0.3936 - mse: 0.4613 - val_loss: 0.6271 - val_mae: 0.5260 - val_mse: 0.6271\n",
      "Epoch 2513/5000\n",
      "4117/4117 [==============================] - 3s 798us/step - loss: 0.4631 - mae: 0.3952 - mse: 0.4631 - val_loss: 0.6450 - val_mae: 0.5312 - val_mse: 0.6450\n",
      "Epoch 2514/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4624 - mae: 0.3954 - mse: 0.4624 - val_loss: 0.6415 - val_mae: 0.5286 - val_mse: 0.6415\n",
      "Epoch 2515/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4622 - mae: 0.3957 - mse: 0.4622 - val_loss: 0.6543 - val_mae: 0.5366 - val_mse: 0.6543\n",
      "Epoch 2516/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4610 - mae: 0.3939 - mse: 0.4610 - val_loss: 0.6294 - val_mae: 0.5258 - val_mse: 0.6294\n",
      "Epoch 2517/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4603 - mae: 0.3925 - mse: 0.4603 - val_loss: 0.6451 - val_mae: 0.5308 - val_mse: 0.6451\n",
      "Epoch 2518/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4661 - mae: 0.3976 - mse: 0.4661 - val_loss: 0.6470 - val_mae: 0.5274 - val_mse: 0.6470\n",
      "Epoch 2519/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4596 - mae: 0.3926 - mse: 0.4596 - val_loss: 0.6491 - val_mae: 0.5348 - val_mse: 0.6491\n",
      "Epoch 2520/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4677 - mae: 0.3981 - mse: 0.4677 - val_loss: 0.6351 - val_mae: 0.5296 - val_mse: 0.6351\n",
      "Epoch 2521/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4667 - mae: 0.3972 - mse: 0.4667 - val_loss: 0.6570 - val_mae: 0.5368 - val_mse: 0.6570\n",
      "Epoch 2522/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4674 - mae: 0.3981 - mse: 0.4674 - val_loss: 0.6473 - val_mae: 0.5289 - val_mse: 0.6473\n",
      "Epoch 2523/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4642 - mae: 0.3943 - mse: 0.4642 - val_loss: 0.6415 - val_mae: 0.5283 - val_mse: 0.6415\n",
      "Epoch 2524/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4627 - mae: 0.3940 - mse: 0.4627 - val_loss: 0.6417 - val_mae: 0.5276 - val_mse: 0.6417\n",
      "Epoch 2525/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4632 - mae: 0.3942 - mse: 0.4632 - val_loss: 0.6320 - val_mae: 0.5230 - val_mse: 0.6320\n",
      "Epoch 2526/5000\n",
      "4117/4117 [==============================] - 3s 811us/step - loss: 0.4619 - mae: 0.3935 - mse: 0.4619 - val_loss: 0.6478 - val_mae: 0.5323 - val_mse: 0.6478\n",
      "Epoch 2527/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4684 - mae: 0.3981 - mse: 0.4684 - val_loss: 0.6365 - val_mae: 0.5286 - val_mse: 0.6365\n",
      "Epoch 2528/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4623 - mae: 0.3947 - mse: 0.4623 - val_loss: 0.6425 - val_mae: 0.5271 - val_mse: 0.6425\n",
      "Epoch 2529/5000\n",
      "4117/4117 [==============================] - 4s 860us/step - loss: 0.4621 - mae: 0.3952 - mse: 0.4621 - val_loss: 0.6454 - val_mae: 0.5282 - val_mse: 0.6454\n",
      "Epoch 2668/5000\n",
      "4117/4117 [==============================] - 3s 838us/step - loss: 0.4639 - mae: 0.3957 - mse: 0.4639 - val_loss: 0.6428 - val_mae: 0.5275 - val_mse: 0.6428\n",
      "Epoch 2669/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4629 - mae: 0.3938 - mse: 0.4629 - val_loss: 0.6467 - val_mae: 0.5297 - val_mse: 0.6467\n",
      "Epoch 2670/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4659 - mae: 0.3970 - mse: 0.4659 - val_loss: 0.6391 - val_mae: 0.5246 - val_mse: 0.6391\n",
      "Epoch 2671/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4637 - mae: 0.3950 - mse: 0.4637 - val_loss: 0.6429 - val_mae: 0.5283 - val_mse: 0.6429\n",
      "Epoch 2672/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4624 - mae: 0.3945 - mse: 0.4624 - val_loss: 0.6452 - val_mae: 0.5312 - val_mse: 0.6452\n",
      "Epoch 2673/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4649 - mae: 0.3941 - mse: 0.4649 - val_loss: 0.6379 - val_mae: 0.5266 - val_mse: 0.6379\n",
      "Epoch 2674/5000\n",
      "4117/4117 [==============================] - 3s 841us/step - loss: 0.4620 - mae: 0.3932 - mse: 0.4620 - val_loss: 0.6317 - val_mae: 0.5274 - val_mse: 0.6317\n",
      "Epoch 2675/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4628 - mae: 0.3937 - mse: 0.4628 - val_loss: 0.6450 - val_mae: 0.5272 - val_mse: 0.6450\n",
      "Epoch 2676/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4621 - mae: 0.3927 - mse: 0.4621 - val_loss: 0.6258 - val_mae: 0.5210 - val_mse: 0.6258\n",
      "Epoch 2677/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4655 - mae: 0.3971 - mse: 0.4655 - val_loss: 0.6504 - val_mae: 0.5341 - val_mse: 0.6504\n",
      "Epoch 2678/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4649 - mae: 0.3951 - mse: 0.4649 - val_loss: 0.6288 - val_mae: 0.5243 - val_mse: 0.6288\n",
      "Epoch 2679/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4646 - mae: 0.3966 - mse: 0.4646 - val_loss: 0.6417 - val_mae: 0.5264 - val_mse: 0.6417\n",
      "Epoch 2680/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4647 - mae: 0.3962 - mse: 0.4647 - val_loss: 0.6341 - val_mae: 0.5238 - val_mse: 0.6341\n",
      "Epoch 2681/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4620 - mae: 0.3941 - mse: 0.4620 - val_loss: 0.6336 - val_mae: 0.5196 - val_mse: 0.6336\n",
      "Epoch 2682/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4597 - mae: 0.3930 - mse: 0.4597 - val_loss: 0.6326 - val_mae: 0.5200 - val_mse: 0.6326\n",
      "Epoch 2683/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4627 - mae: 0.3939 - mse: 0.4627 - val_loss: 0.6287 - val_mae: 0.5189 - val_mse: 0.6287\n",
      "Epoch 2684/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4595 - mae: 0.3932 - mse: 0.4595 - val_loss: 0.6205 - val_mae: 0.5189 - val_mse: 0.6205\n",
      "Epoch 2685/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4607 - mae: 0.3923 - mse: 0.4607 - val_loss: 0.6313 - val_mae: 0.5205 - val_mse: 0.6313\n",
      "Epoch 2686/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.4654 - mae: 0.3975 - mse: 0.4654 - val_loss: 0.6368 - val_mae: 0.5233 - val_mse: 0.6368\n",
      "Epoch 2687/5000\n",
      "4117/4117 [==============================] - 3s 834us/step - loss: 0.4618 - mae: 0.3948 - mse: 0.4618 - val_loss: 0.6386 - val_mae: 0.5277 - val_mse: 0.6386\n",
      "Epoch 2688/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4592 - mae: 0.3925 - mse: 0.4592 - val_loss: 0.6281 - val_mae: 0.5237 - val_mse: 0.6281\n",
      "Epoch 2689/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4644 - mae: 0.3957 - mse: 0.4644 - val_loss: 0.6340 - val_mae: 0.5259 - val_mse: 0.6340\n",
      "Epoch 2690/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4685 - mae: 0.3968 - mse: 0.4685 - val_loss: 0.6454 - val_mae: 0.5294 - val_mse: 0.6454\n",
      "Epoch 2834/5000\n",
      "4117/4117 [==============================] - 3s 837us/step - loss: 0.4641 - mae: 0.3944 - mse: 0.4641 - val_loss: 0.6375 - val_mae: 0.5278 - val_mse: 0.6375\n",
      "Epoch 2835/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4639 - mae: 0.3949 - mse: 0.4639 - val_loss: 0.6386 - val_mae: 0.5255 - val_mse: 0.6386\n",
      "Epoch 2836/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4673 - mae: 0.3978 - mse: 0.4673 - val_loss: 0.6265 - val_mae: 0.5197 - val_mse: 0.6265\n",
      "Epoch 2837/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4671 - mae: 0.3960 - mse: 0.4671 - val_loss: 0.6321 - val_mae: 0.5219 - val_mse: 0.6321\n",
      "Epoch 2838/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4621 - mae: 0.3936 - mse: 0.4621 - val_loss: 0.6184 - val_mae: 0.5135 - val_mse: 0.6184\n",
      "Epoch 2839/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4592 - mae: 0.3920 - mse: 0.4592 - val_loss: 0.6248 - val_mae: 0.5185 - val_mse: 0.6248\n",
      "Epoch 2840/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4616 - mae: 0.3944 - mse: 0.4616 - val_loss: 0.6304 - val_mae: 0.5202 - val_mse: 0.6304\n",
      "Epoch 2841/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4666 - mae: 0.3982 - mse: 0.4666 - val_loss: 0.6316 - val_mae: 0.5285 - val_mse: 0.6316\n",
      "Epoch 2842/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4642 - mae: 0.3942 - mse: 0.4642 - val_loss: 0.6370 - val_mae: 0.5234 - val_mse: 0.6370\n",
      "Epoch 2843/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4684 - mae: 0.3979 - mse: 0.4684 - val_loss: 0.6170 - val_mae: 0.5138 - val_mse: 0.6170\n",
      "Epoch 2844/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4656 - mae: 0.3966 - mse: 0.4656 - val_loss: 0.6400 - val_mae: 0.5278 - val_mse: 0.6400\n",
      "Epoch 2845/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4644 - mae: 0.3947 - mse: 0.4644 - val_loss: 0.6341 - val_mae: 0.5247 - val_mse: 0.6341\n",
      "Epoch 2846/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4630 - mae: 0.3952 - mse: 0.4630 - val_loss: 0.6223 - val_mae: 0.5175 - val_mse: 0.6223\n",
      "Epoch 2847/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4698 - mae: 0.3992 - mse: 0.4698 - val_loss: 0.6233 - val_mae: 0.5191 - val_mse: 0.6233\n",
      "Epoch 2848/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4639 - mae: 0.3955 - mse: 0.4639 - val_loss: 0.6319 - val_mae: 0.5214 - val_mse: 0.6319\n",
      "Epoch 2849/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4670 - mae: 0.3961 - mse: 0.4670 - val_loss: 0.6447 - val_mae: 0.5334 - val_mse: 0.6447\n",
      "Epoch 2850/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4697 - mae: 0.4003 - mse: 0.4697 - val_loss: 0.6269 - val_mae: 0.5242 - val_mse: 0.6269\n",
      "Epoch 2851/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4635 - mae: 0.3936 - mse: 0.4635 - val_loss: 0.6353 - val_mae: 0.5249 - val_mse: 0.6353\n",
      "Epoch 2852/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4621 - mae: 0.3947 - mse: 0.4621 - val_loss: 0.6510 - val_mae: 0.5320 - val_mse: 0.6510\n",
      "Epoch 2853/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4657 - mae: 0.3963 - mse: 0.4657 - val_loss: 0.6288 - val_mae: 0.5218 - val_mse: 0.6288\n",
      "Epoch 2854/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4618 - mae: 0.3935 - mse: 0.4618 - val_loss: 0.6332 - val_mae: 0.5238 - val_mse: 0.6332\n",
      "Epoch 2855/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4633 - mae: 0.3935 - mse: 0.4633 - val_loss: 0.6234 - val_mae: 0.5212 - val_mse: 0.6234\n",
      "Epoch 2856/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4602 - mae: 0.3933 - mse: 0.4602 - val_loss: 0.6256 - val_mae: 0.5174 - val_mse: 0.6256\n",
      "Epoch 2857/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4653 - mae: 0.3961 - mse: 0.4653 - val_loss: 0.6304 - val_mae: 0.5259 - val_mse: 0.6304\n",
      "Epoch 3008/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4688 - mae: 0.3988 - mse: 0.4688 - val_loss: 0.6435 - val_mae: 0.5322 - val_mse: 0.6435\n",
      "Epoch 3009/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4633 - mae: 0.3934 - mse: 0.4633 - val_loss: 0.6184 - val_mae: 0.5224 - val_mse: 0.6184\n",
      "Epoch 3010/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4669 - mae: 0.3979 - mse: 0.4669 - val_loss: 0.6390 - val_mae: 0.5227 - val_mse: 0.6390\n",
      "Epoch 3011/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.4732 - mae: 0.3995 - mse: 0.4732 - val_loss: 0.6493 - val_mae: 0.5318 - val_mse: 0.6493\n",
      "Epoch 3012/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4725 - mae: 0.3986 - mse: 0.4725 - val_loss: 0.6472 - val_mae: 0.5253 - val_mse: 0.6472\n",
      "Epoch 3013/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4617 - mae: 0.3930 - mse: 0.4617 - val_loss: 0.6550 - val_mae: 0.5373 - val_mse: 0.6550\n",
      "Epoch 3014/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4658 - mae: 0.3959 - mse: 0.4658 - val_loss: 0.6571 - val_mae: 0.5340 - val_mse: 0.6571\n",
      "Epoch 3015/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4671 - mae: 0.3964 - mse: 0.4671 - val_loss: 0.6457 - val_mae: 0.5264 - val_mse: 0.6457\n",
      "Epoch 3016/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4628 - mae: 0.3954 - mse: 0.4628 - val_loss: 0.6412 - val_mae: 0.5279 - val_mse: 0.6412\n",
      "Epoch 3017/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4688 - mae: 0.3984 - mse: 0.4688 - val_loss: 0.6381 - val_mae: 0.5287 - val_mse: 0.6381\n",
      "Epoch 3018/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4659 - mae: 0.3971 - mse: 0.4659 - val_loss: 0.6409 - val_mae: 0.5283 - val_mse: 0.6409\n",
      "Epoch 3019/5000\n",
      "4117/4117 [==============================] - 3s 801us/step - loss: 0.4685 - mae: 0.3991 - mse: 0.4685 - val_loss: 0.6676 - val_mae: 0.5424 - val_mse: 0.6676\n",
      "Epoch 3020/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4666 - mae: 0.3976 - mse: 0.4666 - val_loss: 0.6386 - val_mae: 0.5241 - val_mse: 0.6386\n",
      "Epoch 3021/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4654 - mae: 0.3954 - mse: 0.4654 - val_loss: 0.6393 - val_mae: 0.5243 - val_mse: 0.6393\n",
      "Epoch 3022/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4713 - mae: 0.3989 - mse: 0.4713 - val_loss: 0.6447 - val_mae: 0.5293 - val_mse: 0.6447\n",
      "Epoch 3023/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4699 - mae: 0.3962 - mse: 0.4699 - val_loss: 0.6478 - val_mae: 0.5271 - val_mse: 0.6478\n",
      "Epoch 3024/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4661 - mae: 0.3959 - mse: 0.4661 - val_loss: 0.6387 - val_mae: 0.5289 - val_mse: 0.6387\n",
      "Epoch 3025/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4607 - mae: 0.3932 - mse: 0.4607 - val_loss: 0.6454 - val_mae: 0.5346 - val_mse: 0.6454\n",
      "Epoch 3026/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4645 - mae: 0.3948 - mse: 0.4645 - val_loss: 0.6484 - val_mae: 0.5335 - val_mse: 0.6484\n",
      "Epoch 3027/5000\n",
      "4117/4117 [==============================] - 4s 867us/step - loss: 0.4711 - mae: 0.3988 - mse: 0.4711 - val_loss: 0.6300 - val_mae: 0.5232 - val_mse: 0.6300\n",
      "Epoch 3028/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4594 - mae: 0.3909 - mse: 0.4594 - val_loss: 0.6519 - val_mae: 0.5340 - val_mse: 0.6519\n",
      "Epoch 3029/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4658 - mae: 0.3971 - mse: 0.4658 - val_loss: 0.6447 - val_mae: 0.5286 - val_mse: 0.6447\n",
      "Epoch 3030/5000\n",
      "4117/4117 [==============================] - 3s 838us/step - loss: 0.4662 - mae: 0.3958 - mse: 0.4662 - val_loss: 0.6474 - val_mae: 0.5335 - val_mse: 0.6474\n",
      "Epoch 3185/5000\n",
      "4117/4117 [==============================] - 3s 832us/step - loss: 0.4648 - mae: 0.3968 - mse: 0.4648 - val_loss: 0.6274 - val_mae: 0.5187 - val_mse: 0.6274\n",
      "Epoch 3186/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4626 - mae: 0.3946 - mse: 0.4626 - val_loss: 0.6393 - val_mae: 0.5260 - val_mse: 0.6393\n",
      "Epoch 3187/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4725 - mae: 0.3981 - mse: 0.4725 - val_loss: 0.6292 - val_mae: 0.5216 - val_mse: 0.6292\n",
      "Epoch 3188/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4642 - mae: 0.3952 - mse: 0.4642 - val_loss: 0.6246 - val_mae: 0.5185 - val_mse: 0.6246\n",
      "Epoch 3189/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4636 - mae: 0.3955 - mse: 0.4636 - val_loss: 0.6302 - val_mae: 0.5238 - val_mse: 0.6302\n",
      "Epoch 3190/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4620 - mae: 0.3937 - mse: 0.4620 - val_loss: 0.6274 - val_mae: 0.5209 - val_mse: 0.6274\n",
      "Epoch 3191/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4703 - mae: 0.4001 - mse: 0.4703 - val_loss: 0.6398 - val_mae: 0.5275 - val_mse: 0.6398\n",
      "Epoch 3192/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4682 - mae: 0.3988 - mse: 0.4682 - val_loss: 0.6482 - val_mae: 0.5315 - val_mse: 0.6482\n",
      "Epoch 3193/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4695 - mae: 0.3996 - mse: 0.4695 - val_loss: 0.6341 - val_mae: 0.5247 - val_mse: 0.6341\n",
      "Epoch 3194/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4646 - mae: 0.3959 - mse: 0.4646 - val_loss: 0.6327 - val_mae: 0.5220 - val_mse: 0.6327\n",
      "Epoch 3195/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4645 - mae: 0.3958 - mse: 0.4645 - val_loss: 0.6333 - val_mae: 0.5248 - val_mse: 0.6333\n",
      "Epoch 3196/5000\n",
      "4117/4117 [==============================] - 3s 835us/step - loss: 0.4620 - mae: 0.3926 - mse: 0.4620 - val_loss: 0.6367 - val_mae: 0.5241 - val_mse: 0.6367\n",
      "Epoch 3197/5000\n",
      "4117/4117 [==============================] - 3s 834us/step - loss: 0.4676 - mae: 0.3967 - mse: 0.4676 - val_loss: 0.6385 - val_mae: 0.5304 - val_mse: 0.6385\n",
      "Epoch 3198/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4643 - mae: 0.3962 - mse: 0.4643 - val_loss: 0.6385 - val_mae: 0.5274 - val_mse: 0.6385\n",
      "Epoch 3199/5000\n",
      "4117/4117 [==============================] - 3s 842us/step - loss: 0.4605 - mae: 0.3923 - mse: 0.4605 - val_loss: 0.6513 - val_mae: 0.5334 - val_mse: 0.6513\n",
      "Epoch 3200/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4670 - mae: 0.3960 - mse: 0.4670 - val_loss: 0.6368 - val_mae: 0.5256 - val_mse: 0.6368\n",
      "Epoch 3201/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4627 - mae: 0.3931 - mse: 0.4627 - val_loss: 0.6438 - val_mae: 0.5285 - val_mse: 0.6438\n",
      "Epoch 3202/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4617 - mae: 0.3936 - mse: 0.4617 - val_loss: 0.6375 - val_mae: 0.5271 - val_mse: 0.6375\n",
      "Epoch 3203/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4648 - mae: 0.3959 - mse: 0.4648 - val_loss: 0.6273 - val_mae: 0.5195 - val_mse: 0.6273\n",
      "Epoch 3204/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4650 - mae: 0.3959 - mse: 0.4650 - val_loss: 0.6411 - val_mae: 0.5257 - val_mse: 0.6411\n",
      "Epoch 3205/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4690 - mae: 0.3990 - mse: 0.4690 - val_loss: 0.6452 - val_mae: 0.5287 - val_mse: 0.6452\n",
      "Epoch 3206/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4639 - mae: 0.3949 - mse: 0.4639 - val_loss: 0.6340 - val_mae: 0.5252 - val_mse: 0.6340\n",
      "Epoch 3207/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4699 - mae: 0.4006 - mse: 0.4699 - val_loss: 0.6372 - val_mae: 0.5283 - val_mse: 0.6372\n",
      "Epoch 3208/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4684 - mae: 0.3974 - mse: 0.4684 - val_loss: 0.6369 - val_mae: 0.5240 - val_mse: 0.6369\n",
      "Epoch 3368/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4658 - mae: 0.3958 - mse: 0.4658 - val_loss: 0.6318 - val_mae: 0.5242 - val_mse: 0.6318\n",
      "Epoch 3369/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4676 - mae: 0.3966 - mse: 0.4676 - val_loss: 0.6302 - val_mae: 0.5238 - val_mse: 0.6302\n",
      "Epoch 3370/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4742 - mae: 0.4013 - mse: 0.4742 - val_loss: 0.6408 - val_mae: 0.5256 - val_mse: 0.6408\n",
      "Epoch 3371/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4734 - mae: 0.4012 - mse: 0.4734 - val_loss: 0.6264 - val_mae: 0.5238 - val_mse: 0.6264\n",
      "Epoch 3372/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4745 - mae: 0.4001 - mse: 0.4745 - val_loss: 0.6209 - val_mae: 0.5223 - val_mse: 0.6209\n",
      "Epoch 3373/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4623 - mae: 0.3940 - mse: 0.4623 - val_loss: 0.6233 - val_mae: 0.5207 - val_mse: 0.6233\n",
      "Epoch 3374/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4716 - mae: 0.4002 - mse: 0.4716 - val_loss: 0.6378 - val_mae: 0.5247 - val_mse: 0.6378\n",
      "Epoch 3375/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4722 - mae: 0.4003 - mse: 0.4722 - val_loss: 0.6403 - val_mae: 0.5253 - val_mse: 0.6403\n",
      "Epoch 3376/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4669 - mae: 0.3985 - mse: 0.4669 - val_loss: 0.6205 - val_mae: 0.5180 - val_mse: 0.6205\n",
      "Epoch 3377/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4687 - mae: 0.3984 - mse: 0.4687 - val_loss: 0.6274 - val_mae: 0.5189 - val_mse: 0.6274\n",
      "Epoch 3378/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4671 - mae: 0.3965 - mse: 0.4671 - val_loss: 0.6210 - val_mae: 0.5193 - val_mse: 0.6210\n",
      "Epoch 3379/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4615 - mae: 0.3935 - mse: 0.4615 - val_loss: 0.6231 - val_mae: 0.5192 - val_mse: 0.6231\n",
      "Epoch 3380/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4637 - mae: 0.3945 - mse: 0.4637 - val_loss: 0.6224 - val_mae: 0.5176 - val_mse: 0.6224\n",
      "Epoch 3381/5000\n",
      "4117/4117 [==============================] - 3s 843us/step - loss: 0.4654 - mae: 0.3969 - mse: 0.4654 - val_loss: 0.6261 - val_mae: 0.5193 - val_mse: 0.6261\n",
      "Epoch 3382/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4616 - mae: 0.3931 - mse: 0.4616 - val_loss: 0.6208 - val_mae: 0.5196 - val_mse: 0.6208\n",
      "Epoch 3383/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4651 - mae: 0.3938 - mse: 0.4651 - val_loss: 0.6273 - val_mae: 0.5210 - val_mse: 0.6273\n",
      "Epoch 3384/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4639 - mae: 0.3952 - mse: 0.4639 - val_loss: 0.6397 - val_mae: 0.5284 - val_mse: 0.6397\n",
      "Epoch 3385/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4640 - mae: 0.3959 - mse: 0.4640 - val_loss: 0.6269 - val_mae: 0.5220 - val_mse: 0.6269\n",
      "Epoch 3386/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4670 - mae: 0.3961 - mse: 0.4670 - val_loss: 0.6272 - val_mae: 0.5252 - val_mse: 0.6272\n",
      "Epoch 3387/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4611 - mae: 0.3928 - mse: 0.4611 - val_loss: 0.6304 - val_mae: 0.5224 - val_mse: 0.6304\n",
      "Epoch 3388/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4655 - mae: 0.3960 - mse: 0.4655 - val_loss: 0.6416 - val_mae: 0.5280 - val_mse: 0.6416\n",
      "Epoch 3389/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4727 - mae: 0.3987 - mse: 0.4727 - val_loss: 0.6356 - val_mae: 0.5223 - val_mse: 0.6356\n",
      "Epoch 3390/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4688 - mae: 0.3960 - mse: 0.4688 - val_loss: 0.6480 - val_mae: 0.5330 - val_mse: 0.6480\n",
      "Epoch 3391/5000\n",
      "4117/4117 [==============================] - 3s 847us/step - loss: 0.4700 - mae: 0.4014 - mse: 0.4700 - val_loss: 0.6203 - val_mae: 0.5135 - val_mse: 0.6203\n",
      "Epoch 3556/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4655 - mae: 0.3968 - mse: 0.4655 - val_loss: 0.6297 - val_mae: 0.5193 - val_mse: 0.6297\n",
      "Epoch 3557/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4675 - mae: 0.3966 - mse: 0.4675 - val_loss: 0.6301 - val_mae: 0.5208 - val_mse: 0.6301\n",
      "Epoch 3558/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4706 - mae: 0.4003 - mse: 0.4706 - val_loss: 0.6220 - val_mae: 0.5177 - val_mse: 0.6220\n",
      "Epoch 3559/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4686 - mae: 0.3982 - mse: 0.4686 - val_loss: 0.6266 - val_mae: 0.5158 - val_mse: 0.6266\n",
      "Epoch 3560/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4737 - mae: 0.4008 - mse: 0.4737 - val_loss: 0.6316 - val_mae: 0.5171 - val_mse: 0.6316\n",
      "Epoch 3561/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4746 - mae: 0.4019 - mse: 0.4746 - val_loss: 0.6310 - val_mae: 0.5161 - val_mse: 0.6310\n",
      "Epoch 3562/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4674 - mae: 0.3990 - mse: 0.4674 - val_loss: 0.6176 - val_mae: 0.5107 - val_mse: 0.6176\n",
      "Epoch 3563/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4663 - mae: 0.3962 - mse: 0.4663 - val_loss: 0.6355 - val_mae: 0.5195 - val_mse: 0.6355\n",
      "Epoch 3564/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4682 - mae: 0.3991 - mse: 0.4682 - val_loss: 0.6213 - val_mae: 0.5174 - val_mse: 0.6213\n",
      "Epoch 3565/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4634 - mae: 0.3950 - mse: 0.4634 - val_loss: 0.6373 - val_mae: 0.5209 - val_mse: 0.6373\n",
      "Epoch 3566/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4699 - mae: 0.4001 - mse: 0.4699 - val_loss: 0.6385 - val_mae: 0.5268 - val_mse: 0.6385\n",
      "Epoch 3567/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4681 - mae: 0.3980 - mse: 0.4681 - val_loss: 0.6389 - val_mae: 0.5244 - val_mse: 0.6389\n",
      "Epoch 3568/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4631 - mae: 0.3946 - mse: 0.4631 - val_loss: 0.6364 - val_mae: 0.5206 - val_mse: 0.6364\n",
      "Epoch 3569/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4672 - mae: 0.3985 - mse: 0.4672 - val_loss: 0.6379 - val_mae: 0.5227 - val_mse: 0.6379\n",
      "Epoch 3570/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4690 - mae: 0.3986 - mse: 0.4690 - val_loss: 0.6257 - val_mae: 0.5157 - val_mse: 0.6257\n",
      "Epoch 3571/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4653 - mae: 0.3955 - mse: 0.4653 - val_loss: 0.6235 - val_mae: 0.5158 - val_mse: 0.6235\n",
      "Epoch 3572/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4652 - mae: 0.3968 - mse: 0.4652 - val_loss: 0.6147 - val_mae: 0.5120 - val_mse: 0.6147\n",
      "Epoch 3573/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4679 - mae: 0.3975 - mse: 0.4679 - val_loss: 0.6290 - val_mae: 0.5177 - val_mse: 0.6290\n",
      "Epoch 3574/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4724 - mae: 0.4007 - mse: 0.4724 - val_loss: 0.6273 - val_mae: 0.5218 - val_mse: 0.6273\n",
      "Epoch 3575/5000\n",
      "4117/4117 [==============================] - 3s 795us/step - loss: 0.4725 - mae: 0.3994 - mse: 0.4725 - val_loss: 0.6342 - val_mae: 0.5223 - val_mse: 0.6342\n",
      "Epoch 3576/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4744 - mae: 0.4014 - mse: 0.4744 - val_loss: 0.6317 - val_mae: 0.5229 - val_mse: 0.6317\n",
      "Epoch 3577/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4682 - mae: 0.3969 - mse: 0.4682 - val_loss: 0.6378 - val_mae: 0.5223 - val_mse: 0.6378\n",
      "Epoch 3578/5000\n",
      "4117/4117 [==============================] - 3s 850us/step - loss: 0.4656 - mae: 0.3953 - mse: 0.4656 - val_loss: 0.6324 - val_mae: 0.5228 - val_mse: 0.6324\n",
      "Epoch 3749/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4699 - mae: 0.3980 - mse: 0.4699 - val_loss: 0.6396 - val_mae: 0.5247 - val_mse: 0.6396\n",
      "Epoch 3750/5000\n",
      "4117/4117 [==============================] - 3s 798us/step - loss: 0.4677 - mae: 0.3957 - mse: 0.4677 - val_loss: 0.6304 - val_mae: 0.5200 - val_mse: 0.6304\n",
      "Epoch 3751/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4676 - mae: 0.3984 - mse: 0.4676 - val_loss: 0.6379 - val_mae: 0.5218 - val_mse: 0.6379\n",
      "Epoch 3752/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4630 - mae: 0.3946 - mse: 0.4630 - val_loss: 0.6346 - val_mae: 0.5256 - val_mse: 0.6346\n",
      "Epoch 3753/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4688 - mae: 0.3972 - mse: 0.4688 - val_loss: 0.6448 - val_mae: 0.5298 - val_mse: 0.6448\n",
      "Epoch 3754/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4693 - mae: 0.3993 - mse: 0.4693 - val_loss: 0.6357 - val_mae: 0.5242 - val_mse: 0.6357\n",
      "Epoch 3755/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4651 - mae: 0.3964 - mse: 0.4651 - val_loss: 0.6423 - val_mae: 0.5297 - val_mse: 0.6423\n",
      "Epoch 3756/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4686 - mae: 0.3978 - mse: 0.4686 - val_loss: 0.6466 - val_mae: 0.5282 - val_mse: 0.6466\n",
      "Epoch 3757/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4688 - mae: 0.3965 - mse: 0.4688 - val_loss: 0.6396 - val_mae: 0.5237 - val_mse: 0.6396\n",
      "Epoch 3758/5000\n",
      "4117/4117 [==============================] - 3s 807us/step - loss: 0.4626 - mae: 0.3928 - mse: 0.4626 - val_loss: 0.6265 - val_mae: 0.5192 - val_mse: 0.6265\n",
      "Epoch 3759/5000\n",
      "4117/4117 [==============================] - 3s 810us/step - loss: 0.4609 - mae: 0.3932 - mse: 0.4609 - val_loss: 0.6297 - val_mae: 0.5214 - val_mse: 0.6297\n",
      "Epoch 3760/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4719 - mae: 0.3993 - mse: 0.4719 - val_loss: 0.6345 - val_mae: 0.5239 - val_mse: 0.6345\n",
      "Epoch 3761/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4654 - mae: 0.3961 - mse: 0.4654 - val_loss: 0.6378 - val_mae: 0.5277 - val_mse: 0.6378\n",
      "Epoch 3762/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4728 - mae: 0.3998 - mse: 0.4728 - val_loss: 0.6330 - val_mae: 0.5217 - val_mse: 0.6330\n",
      "Epoch 3763/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4617 - mae: 0.3933 - mse: 0.4617 - val_loss: 0.6290 - val_mae: 0.5197 - val_mse: 0.6290\n",
      "Epoch 3764/5000\n",
      "4117/4117 [==============================] - 4s 872us/step - loss: 0.4687 - mae: 0.3968 - mse: 0.4687 - val_loss: 0.6368 - val_mae: 0.5248 - val_mse: 0.6368\n",
      "Epoch 3765/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4698 - mae: 0.3998 - mse: 0.4698 - val_loss: 0.6361 - val_mae: 0.5231 - val_mse: 0.6361\n",
      "Epoch 3766/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4675 - mae: 0.3957 - mse: 0.4675 - val_loss: 0.6279 - val_mae: 0.5160 - val_mse: 0.6279\n",
      "Epoch 3767/5000\n",
      "4117/4117 [==============================] - 4s 855us/step - loss: 0.4695 - mae: 0.3980 - mse: 0.4695 - val_loss: 0.6268 - val_mae: 0.5190 - val_mse: 0.6268\n",
      "Epoch 3768/5000\n",
      "4117/4117 [==============================] - 3s 829us/step - loss: 0.4657 - mae: 0.3975 - mse: 0.4657 - val_loss: 0.6262 - val_mae: 0.5192 - val_mse: 0.6262\n",
      "Epoch 3769/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4656 - mae: 0.3964 - mse: 0.4656 - val_loss: 0.6364 - val_mae: 0.5204 - val_mse: 0.6364\n",
      "Epoch 3770/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4690 - mae: 0.3995 - mse: 0.4690 - val_loss: 0.6362 - val_mae: 0.5213 - val_mse: 0.6362\n",
      "Epoch 3771/5000\n",
      "4117/4117 [==============================] - 4s 860us/step - loss: 0.4673 - mae: 0.3961 - mse: 0.4673 - val_loss: 0.6379 - val_mae: 0.5215 - val_mse: 0.6379\n",
      "Epoch 3947/5000\n",
      "4117/4117 [==============================] - 3s 833us/step - loss: 0.4685 - mae: 0.3988 - mse: 0.4685 - val_loss: 0.6360 - val_mae: 0.5185 - val_mse: 0.6360\n",
      "Epoch 3948/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4717 - mae: 0.3988 - mse: 0.4717 - val_loss: 0.6194 - val_mae: 0.5121 - val_mse: 0.6194\n",
      "Epoch 3949/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4702 - mae: 0.3995 - mse: 0.4702 - val_loss: 0.6361 - val_mae: 0.5243 - val_mse: 0.6361\n",
      "Epoch 3950/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4738 - mae: 0.4017 - mse: 0.4738 - val_loss: 0.6316 - val_mae: 0.5184 - val_mse: 0.6316\n",
      "Epoch 3951/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4677 - mae: 0.3988 - mse: 0.4677 - val_loss: 0.6279 - val_mae: 0.5197 - val_mse: 0.6279\n",
      "Epoch 3952/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4708 - mae: 0.3992 - mse: 0.4708 - val_loss: 0.6336 - val_mae: 0.5218 - val_mse: 0.6336\n",
      "Epoch 3953/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4675 - mae: 0.3965 - mse: 0.4675 - val_loss: 0.6355 - val_mae: 0.5211 - val_mse: 0.6355\n",
      "Epoch 3954/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4651 - mae: 0.3962 - mse: 0.4651 - val_loss: 0.6453 - val_mae: 0.5262 - val_mse: 0.6453\n",
      "Epoch 3955/5000\n",
      "4117/4117 [==============================] - 3s 802us/step - loss: 0.4728 - mae: 0.4012 - mse: 0.4728 - val_loss: 0.6372 - val_mae: 0.5237 - val_mse: 0.6372\n",
      "Epoch 3956/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4714 - mae: 0.3991 - mse: 0.4714 - val_loss: 0.6445 - val_mae: 0.5280 - val_mse: 0.6445\n",
      "Epoch 3957/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4696 - mae: 0.3996 - mse: 0.4696 - val_loss: 0.6395 - val_mae: 0.5258 - val_mse: 0.6395\n",
      "Epoch 3958/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4637 - mae: 0.3954 - mse: 0.4637 - val_loss: 0.6324 - val_mae: 0.5210 - val_mse: 0.6324\n",
      "Epoch 3959/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4697 - mae: 0.4001 - mse: 0.4697 - val_loss: 0.6214 - val_mae: 0.5139 - val_mse: 0.6214\n",
      "Epoch 3960/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4641 - mae: 0.3951 - mse: 0.4641 - val_loss: 0.6280 - val_mae: 0.5191 - val_mse: 0.6280\n",
      "Epoch 3961/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4676 - mae: 0.3975 - mse: 0.4676 - val_loss: 0.6337 - val_mae: 0.5206 - val_mse: 0.6337\n",
      "Epoch 3962/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4715 - mae: 0.3989 - mse: 0.4715 - val_loss: 0.6272 - val_mae: 0.5200 - val_mse: 0.6272\n",
      "Epoch 3963/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4702 - mae: 0.3978 - mse: 0.4702 - val_loss: 0.6256 - val_mae: 0.5161 - val_mse: 0.6256\n",
      "Epoch 3964/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4689 - mae: 0.3972 - mse: 0.4689 - val_loss: 0.6405 - val_mae: 0.5245 - val_mse: 0.6405\n",
      "Epoch 3965/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4668 - mae: 0.3962 - mse: 0.4668 - val_loss: 0.6306 - val_mae: 0.5205 - val_mse: 0.6306\n",
      "Epoch 3966/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4693 - mae: 0.3968 - mse: 0.4693 - val_loss: 0.6140 - val_mae: 0.5094 - val_mse: 0.6140\n",
      "Epoch 3967/5000\n",
      "4117/4117 [==============================] - 3s 803us/step - loss: 0.4656 - mae: 0.3953 - mse: 0.4656 - val_loss: 0.6278 - val_mae: 0.5219 - val_mse: 0.6278\n",
      "Epoch 3968/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4645 - mae: 0.3938 - mse: 0.4645 - val_loss: 0.6305 - val_mae: 0.5190 - val_mse: 0.6305\n",
      "Epoch 3969/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4692 - mae: 0.3971 - mse: 0.4692 - val_loss: 0.6234 - val_mae: 0.5178 - val_mse: 0.6234\n",
      "Epoch 4151/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4663 - mae: 0.3965 - mse: 0.4663 - val_loss: 0.6338 - val_mae: 0.5208 - val_mse: 0.6338\n",
      "Epoch 4152/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4655 - mae: 0.3953 - mse: 0.4655 - val_loss: 0.6422 - val_mae: 0.5250 - val_mse: 0.6422\n",
      "Epoch 4153/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4701 - mae: 0.3955 - mse: 0.4701 - val_loss: 0.6266 - val_mae: 0.5166 - val_mse: 0.6266\n",
      "Epoch 4154/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4709 - mae: 0.4007 - mse: 0.4709 - val_loss: 0.6295 - val_mae: 0.5235 - val_mse: 0.6295\n",
      "Epoch 4155/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4637 - mae: 0.3944 - mse: 0.4637 - val_loss: 0.6429 - val_mae: 0.5295 - val_mse: 0.6429\n",
      "Epoch 4156/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4629 - mae: 0.3941 - mse: 0.4629 - val_loss: 0.6380 - val_mae: 0.5218 - val_mse: 0.6380\n",
      "Epoch 4157/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4658 - mae: 0.3970 - mse: 0.4658 - val_loss: 0.6341 - val_mae: 0.5277 - val_mse: 0.6341\n",
      "Epoch 4158/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4679 - mae: 0.3997 - mse: 0.4679 - val_loss: 0.6363 - val_mae: 0.5186 - val_mse: 0.6363\n",
      "Epoch 4159/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4709 - mae: 0.3999 - mse: 0.4709 - val_loss: 0.6218 - val_mae: 0.5176 - val_mse: 0.6218\n",
      "Epoch 4160/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4632 - mae: 0.3941 - mse: 0.4632 - val_loss: 0.6221 - val_mae: 0.5146 - val_mse: 0.6221\n",
      "Epoch 4161/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4689 - mae: 0.3983 - mse: 0.4689 - val_loss: 0.6365 - val_mae: 0.5261 - val_mse: 0.6365\n",
      "Epoch 4162/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4706 - mae: 0.3973 - mse: 0.4706 - val_loss: 0.6530 - val_mae: 0.5303 - val_mse: 0.6530\n",
      "Epoch 4163/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4718 - mae: 0.3986 - mse: 0.4718 - val_loss: 0.6296 - val_mae: 0.5228 - val_mse: 0.6296\n",
      "Epoch 4164/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4643 - mae: 0.3959 - mse: 0.4643 - val_loss: 0.6412 - val_mae: 0.5263 - val_mse: 0.6412\n",
      "Epoch 4165/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4709 - mae: 0.3990 - mse: 0.4709 - val_loss: 0.6377 - val_mae: 0.5274 - val_mse: 0.6377\n",
      "Epoch 4166/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4695 - mae: 0.3992 - mse: 0.4695 - val_loss: 0.6404 - val_mae: 0.5256 - val_mse: 0.6404\n",
      "Epoch 4167/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4710 - mae: 0.3989 - mse: 0.4710 - val_loss: 0.6335 - val_mae: 0.5200 - val_mse: 0.6335\n",
      "Epoch 4168/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4699 - mae: 0.3990 - mse: 0.4699 - val_loss: 0.6383 - val_mae: 0.5274 - val_mse: 0.6383\n",
      "Epoch 4169/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4641 - mae: 0.3948 - mse: 0.4641 - val_loss: 0.6311 - val_mae: 0.5223 - val_mse: 0.6311\n",
      "Epoch 4170/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4715 - mae: 0.4003 - mse: 0.4715 - val_loss: 0.6297 - val_mae: 0.5209 - val_mse: 0.6297\n",
      "Epoch 4171/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4666 - mae: 0.3954 - mse: 0.4666 - val_loss: 0.6224 - val_mae: 0.5184 - val_mse: 0.6224\n",
      "Epoch 4172/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4669 - mae: 0.3976 - mse: 0.4669 - val_loss: 0.6366 - val_mae: 0.5224 - val_mse: 0.6366\n",
      "Epoch 4173/5000\n",
      "4117/4117 [==============================] - 4s 885us/step - loss: 0.4709 - mae: 0.3990 - mse: 0.4709 - val_loss: 0.6317 - val_mae: 0.5233 - val_mse: 0.6317\n",
      "Epoch 4359/5000\n",
      "4117/4117 [==============================] - 4s 880us/step - loss: 0.4678 - mae: 0.3955 - mse: 0.4678 - val_loss: 0.6395 - val_mae: 0.5276 - val_mse: 0.6395\n",
      "Epoch 4360/5000\n",
      "4117/4117 [==============================] - 3s 848us/step - loss: 0.4761 - mae: 0.4048 - mse: 0.4761 - val_loss: 0.6336 - val_mae: 0.5262 - val_mse: 0.6336\n",
      "Epoch 4361/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4761 - mae: 0.4027 - mse: 0.4761 - val_loss: 0.6395 - val_mae: 0.5269 - val_mse: 0.6395\n",
      "Epoch 4362/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4660 - mae: 0.3982 - mse: 0.4660 - val_loss: 0.6290 - val_mae: 0.5185 - val_mse: 0.6290\n",
      "Epoch 4363/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4688 - mae: 0.3991 - mse: 0.4688 - val_loss: 0.6343 - val_mae: 0.5259 - val_mse: 0.6343\n",
      "Epoch 4364/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4722 - mae: 0.4010 - mse: 0.4722 - val_loss: 0.6318 - val_mae: 0.5177 - val_mse: 0.6318\n",
      "Epoch 4365/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4672 - mae: 0.3969 - mse: 0.4672 - val_loss: 0.6368 - val_mae: 0.5279 - val_mse: 0.6368\n",
      "Epoch 4366/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4673 - mae: 0.3975 - mse: 0.4673 - val_loss: 0.6462 - val_mae: 0.5329 - val_mse: 0.6462\n",
      "Epoch 4367/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4677 - mae: 0.3985 - mse: 0.4677 - val_loss: 0.6340 - val_mae: 0.5268 - val_mse: 0.6340\n",
      "Epoch 4368/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4660 - mae: 0.3974 - mse: 0.4660 - val_loss: 0.6311 - val_mae: 0.5237 - val_mse: 0.6311\n",
      "Epoch 4369/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4655 - mae: 0.3946 - mse: 0.4655 - val_loss: 0.6259 - val_mae: 0.5170 - val_mse: 0.6259\n",
      "Epoch 4370/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4651 - mae: 0.3963 - mse: 0.4651 - val_loss: 0.6274 - val_mae: 0.5195 - val_mse: 0.6274\n",
      "Epoch 4371/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4746 - mae: 0.4024 - mse: 0.4746 - val_loss: 0.6244 - val_mae: 0.5183 - val_mse: 0.6244\n",
      "Epoch 4372/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4649 - mae: 0.3951 - mse: 0.4649 - val_loss: 0.6249 - val_mae: 0.5156 - val_mse: 0.6249\n",
      "Epoch 4373/5000\n",
      "4117/4117 [==============================] - 3s 820us/step - loss: 0.4748 - mae: 0.4041 - mse: 0.4748 - val_loss: 0.6321 - val_mae: 0.5185 - val_mse: 0.6321\n",
      "Epoch 4374/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4679 - mae: 0.3972 - mse: 0.4679 - val_loss: 0.6329 - val_mae: 0.5223 - val_mse: 0.6329\n",
      "Epoch 4375/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4663 - mae: 0.3970 - mse: 0.4663 - val_loss: 0.6411 - val_mae: 0.5293 - val_mse: 0.6411\n",
      "Epoch 4376/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4661 - mae: 0.3957 - mse: 0.4661 - val_loss: 0.6311 - val_mae: 0.5192 - val_mse: 0.6311\n",
      "Epoch 4377/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4678 - mae: 0.3984 - mse: 0.4678 - val_loss: 0.6337 - val_mae: 0.5255 - val_mse: 0.6337\n",
      "Epoch 4378/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4772 - mae: 0.4029 - mse: 0.4772 - val_loss: 0.6368 - val_mae: 0.5278 - val_mse: 0.6368\n",
      "Epoch 4379/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4673 - mae: 0.3994 - mse: 0.4673 - val_loss: 0.6413 - val_mae: 0.5280 - val_mse: 0.6413\n",
      "Epoch 4380/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4639 - mae: 0.3956 - mse: 0.4639 - val_loss: 0.6158 - val_mae: 0.5128 - val_mse: 0.6158\n",
      "Epoch 4381/5000\n",
      "4117/4117 [==============================] - 3s 839us/step - loss: 0.4768 - mae: 0.4051 - mse: 0.4768 - val_loss: 0.6309 - val_mae: 0.5217 - val_mse: 0.6309\n",
      "Epoch 4573/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4698 - mae: 0.3988 - mse: 0.4698 - val_loss: 0.6371 - val_mae: 0.5224 - val_mse: 0.6371\n",
      "Epoch 4574/5000\n",
      "4117/4117 [==============================] - 3s 799us/step - loss: 0.4664 - mae: 0.3962 - mse: 0.4664 - val_loss: 0.6490 - val_mae: 0.5359 - val_mse: 0.6490\n",
      "Epoch 4575/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4612 - mae: 0.3923 - mse: 0.4612 - val_loss: 0.6416 - val_mae: 0.5289 - val_mse: 0.6416\n",
      "Epoch 4576/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4627 - mae: 0.3953 - mse: 0.4627 - val_loss: 0.6361 - val_mae: 0.5267 - val_mse: 0.6361\n",
      "Epoch 4577/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4645 - mae: 0.3961 - mse: 0.4645 - val_loss: 0.6459 - val_mae: 0.5264 - val_mse: 0.6459\n",
      "Epoch 4578/5000\n",
      "4117/4117 [==============================] - 3s 796us/step - loss: 0.4683 - mae: 0.3981 - mse: 0.4683 - val_loss: 0.6448 - val_mae: 0.5294 - val_mse: 0.6448\n",
      "Epoch 4579/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4799 - mae: 0.4043 - mse: 0.4799 - val_loss: 0.6351 - val_mae: 0.5218 - val_mse: 0.6351\n",
      "Epoch 4580/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4685 - mae: 0.3979 - mse: 0.4685 - val_loss: 0.6323 - val_mae: 0.5176 - val_mse: 0.6323\n",
      "Epoch 4581/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4751 - mae: 0.4035 - mse: 0.4751 - val_loss: 0.6405 - val_mae: 0.5243 - val_mse: 0.6405\n",
      "Epoch 4582/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4762 - mae: 0.4029 - mse: 0.4762 - val_loss: 0.6368 - val_mae: 0.5268 - val_mse: 0.6368\n",
      "Epoch 4583/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4712 - mae: 0.4015 - mse: 0.4712 - val_loss: 0.6308 - val_mae: 0.5225 - val_mse: 0.6308\n",
      "Epoch 4584/5000\n",
      "4117/4117 [==============================] - 3s 812us/step - loss: 0.4716 - mae: 0.4008 - mse: 0.4716 - val_loss: 0.6368 - val_mae: 0.5245 - val_mse: 0.6368\n",
      "Epoch 4585/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4679 - mae: 0.3966 - mse: 0.4679 - val_loss: 0.6349 - val_mae: 0.5233 - val_mse: 0.6349\n",
      "Epoch 4586/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4654 - mae: 0.3958 - mse: 0.4654 - val_loss: 0.6540 - val_mae: 0.5350 - val_mse: 0.6540\n",
      "Epoch 4587/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4695 - mae: 0.3996 - mse: 0.4695 - val_loss: 0.6294 - val_mae: 0.5165 - val_mse: 0.6294\n",
      "Epoch 4588/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4706 - mae: 0.3986 - mse: 0.4706 - val_loss: 0.6277 - val_mae: 0.5157 - val_mse: 0.6277\n",
      "Epoch 4589/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4748 - mae: 0.4018 - mse: 0.4748 - val_loss: 0.6335 - val_mae: 0.5241 - val_mse: 0.6335\n",
      "Epoch 4793/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4739 - mae: 0.4030 - mse: 0.4739 - val_loss: 0.6332 - val_mae: 0.5234 - val_mse: 0.6332\n",
      "Epoch 4794/5000\n",
      "4117/4117 [==============================] - 3s 831us/step - loss: 0.4726 - mae: 0.4009 - mse: 0.4726 - val_loss: 0.6338 - val_mae: 0.5252 - val_mse: 0.6338\n",
      "Epoch 4795/5000\n",
      "4117/4117 [==============================] - 3s 826us/step - loss: 0.4727 - mae: 0.4001 - mse: 0.4727 - val_loss: 0.6480 - val_mae: 0.5331 - val_mse: 0.6480\n",
      "Epoch 4796/5000\n",
      "4117/4117 [==============================] - 3s 813us/step - loss: 0.4712 - mae: 0.3999 - mse: 0.4712 - val_loss: 0.6265 - val_mae: 0.5191 - val_mse: 0.6265\n",
      "Epoch 4797/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4708 - mae: 0.4001 - mse: 0.4708 - val_loss: 0.6272 - val_mae: 0.5181 - val_mse: 0.6272\n",
      "Epoch 4798/5000\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4725 - mae: 0.4017 - mse: 0.4725 - val_loss: 0.6406 - val_mae: 0.5289 - val_mse: 0.6406\n",
      "Epoch 4799/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4672 - mae: 0.3978 - mse: 0.4672 - val_loss: 0.6352 - val_mae: 0.5211 - val_mse: 0.6352\n",
      "Epoch 4800/5000\n",
      "4117/4117 [==============================] - 3s 816us/step - loss: 0.4680 - mae: 0.3975 - mse: 0.4680 - val_loss: 0.6330 - val_mae: 0.5226 - val_mse: 0.6330\n",
      "Epoch 4801/5000\n",
      "4117/4117 [==============================] - 3s 797us/step - loss: 0.4733 - mae: 0.4015 - mse: 0.4733 - val_loss: 0.6499 - val_mae: 0.5324 - val_mse: 0.6499\n",
      "Epoch 4802/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4713 - mae: 0.4010 - mse: 0.4713 - val_loss: 0.6505 - val_mae: 0.5274 - val_mse: 0.6505\n",
      "Epoch 4803/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4687 - mae: 0.3978 - mse: 0.4687 - val_loss: 0.6423 - val_mae: 0.5264 - val_mse: 0.6423\n",
      "Epoch 4804/5000\n",
      "4117/4117 [==============================] - 3s 809us/step - loss: 0.4790 - mae: 0.4033 - mse: 0.4790 - val_loss: 0.6410 - val_mae: 0.5254 - val_mse: 0.6410\n",
      "Epoch 4805/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4756 - mae: 0.4023 - mse: 0.4756 - val_loss: 0.6460 - val_mae: 0.5290 - val_mse: 0.6460\n",
      "Epoch 4806/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4690 - mae: 0.3990 - mse: 0.4690 - val_loss: 0.6343 - val_mae: 0.5259 - val_mse: 0.6343\n",
      "Epoch 4807/5000\n",
      "4117/4117 [==============================] - 3s 818us/step - loss: 0.4800 - mae: 0.4055 - mse: 0.4800 - val_loss: 0.6310 - val_mae: 0.5171 - val_mse: 0.6310\n",
      "Epoch 4808/5000\n",
      "4117/4117 [==============================] - 3s 805us/step - loss: 0.4881 - mae: 0.4100 - mse: 0.4881 - val_loss: 0.6382 - val_mae: 0.5234 - val_mse: 0.6382\n",
      "Epoch 4809/5000\n",
      "4117/4117 [==============================] - 3s 808us/step - loss: 0.4719 - mae: 0.4001 - mse: 0.4719 - val_loss: 0.6351 - val_mae: 0.5231 - val_mse: 0.6351\n",
      "Epoch 4810/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4814 - mae: 0.4046 - mse: 0.4814 - val_loss: 0.6405 - val_mae: 0.5272 - val_mse: 0.6405\n",
      "Epoch 4811/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4703 - mae: 0.3989 - mse: 0.4703 - val_loss: 0.6359 - val_mae: 0.5284 - val_mse: 0.6359\n",
      "Epoch 4812/5000\n",
      "4117/4117 [==============================] - 3s 804us/step - loss: 0.4741 - mae: 0.4014 - mse: 0.4741 - val_loss: 0.6358 - val_mae: 0.5245 - val_mse: 0.6358\n",
      "Epoch 4813/5000\n",
      "4117/4117 [==============================] - 3s 817us/step - loss: 0.4752 - mae: 0.4017 - mse: 0.4752 - val_loss: 0.6373 - val_mae: 0.5229 - val_mse: 0.6373\n",
      "Epoch 4814/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4768 - mae: 0.4036 - mse: 0.4768 - val_loss: 0.6319 - val_mae: 0.5184 - val_mse: 0.6319\n",
      "Epoch 4815/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4695 - mae: 0.3993 - mse: 0.4695 - val_loss: 0.6400 - val_mae: 0.5253 - val_mse: 0.6400\n",
      "4117/4117 [==============================] - 3s 830us/step - loss: 0.4673 - mae: 0.3987 - mse: 0.4673 - val_loss: 0.6216 - val_mae: 0.5169 - val_mse: 0.6216\n",
      "Epoch 4958/5000\n",
      "4117/4117 [==============================] - 3s 823us/step - loss: 0.4776 - mae: 0.4034 - mse: 0.4776 - val_loss: 0.6367 - val_mae: 0.5225 - val_mse: 0.6367\n",
      "Epoch 4959/5000\n",
      "4117/4117 [==============================] - 3s 821us/step - loss: 0.4732 - mae: 0.4016 - mse: 0.4732 - val_loss: 0.6305 - val_mae: 0.5215 - val_mse: 0.6305\n",
      "Epoch 4960/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4791 - mae: 0.4049 - mse: 0.4791 - val_loss: 0.6342 - val_mae: 0.5215 - val_mse: 0.6342\n",
      "Epoch 4961/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4825 - mae: 0.4070 - mse: 0.4825 - val_loss: 0.6329 - val_mae: 0.5251 - val_mse: 0.6329\n",
      "Epoch 4962/5000\n",
      "4117/4117 [==============================] - 3s 806us/step - loss: 0.4763 - mae: 0.4030 - mse: 0.4763 - val_loss: 0.6265 - val_mae: 0.5240 - val_mse: 0.6265\n",
      "Epoch 4963/5000\n",
      "4117/4117 [==============================] - 3s 814us/step - loss: 0.4726 - mae: 0.3993 - mse: 0.4726 - val_loss: 0.6380 - val_mae: 0.5245 - val_mse: 0.6380\n",
      "Epoch 4964/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4750 - mae: 0.4016 - mse: 0.4750 - val_loss: 0.6606 - val_mae: 0.5430 - val_mse: 0.6606\n",
      "Epoch 4965/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4779 - mae: 0.4063 - mse: 0.4779 - val_loss: 0.6286 - val_mae: 0.5217 - val_mse: 0.6286\n",
      "Epoch 4966/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4782 - mae: 0.4052 - mse: 0.4782 - val_loss: 0.6193 - val_mae: 0.5194 - val_mse: 0.6193\n",
      "Epoch 4967/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4697 - mae: 0.3993 - mse: 0.4697 - val_loss: 0.6319 - val_mae: 0.5206 - val_mse: 0.6319\n",
      "Epoch 4968/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4752 - mae: 0.4039 - mse: 0.4752 - val_loss: 0.6244 - val_mae: 0.5177 - val_mse: 0.6244\n",
      "Epoch 4969/5000\n",
      "4117/4117 [==============================] - 3s 815us/step - loss: 0.4708 - mae: 0.3991 - mse: 0.4708 - val_loss: 0.6174 - val_mae: 0.5187 - val_mse: 0.6174\n",
      "Epoch 4970/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4691 - mae: 0.3991 - mse: 0.4691 - val_loss: 0.6193 - val_mae: 0.5161 - val_mse: 0.6193\n",
      "Epoch 4971/5000\n",
      "4117/4117 [==============================] - 3s 822us/step - loss: 0.4718 - mae: 0.4000 - mse: 0.4718 - val_loss: 0.6256 - val_mae: 0.5193 - val_mse: 0.6256\n",
      "Epoch 4972/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4707 - mae: 0.4001 - mse: 0.4707 - val_loss: 0.6300 - val_mae: 0.5199 - val_mse: 0.6300\n",
      "Epoch 4973/5000\n",
      "4117/4117 [==============================] - 3s 824us/step - loss: 0.4690 - mae: 0.3992 - mse: 0.4690 - val_loss: 0.6319 - val_mae: 0.5213 - val_mse: 0.6319\n",
      "Epoch 4974/5000\n",
      "4117/4117 [==============================] - 3s 828us/step - loss: 0.4647 - mae: 0.3964 - mse: 0.4647 - val_loss: 0.6228 - val_mae: 0.5183 - val_mse: 0.6228\n",
      "Epoch 4975/5000\n",
      "4117/4117 [==============================] - 3s 827us/step - loss: 0.4730 - mae: 0.4040 - mse: 0.4730 - val_loss: 0.6328 - val_mae: 0.5222 - val_mse: 0.6328\n",
      "Epoch 4976/5000\n",
      "4117/4117 [==============================] - 3s 835us/step - loss: 0.4764 - mae: 0.4037 - mse: 0.4764 - val_loss: 0.6321 - val_mae: 0.5215 - val_mse: 0.6321\n",
      "Epoch 4977/5000\n",
      "4117/4117 [==============================] - 3s 819us/step - loss: 0.4760 - mae: 0.4026 - mse: 0.4760 - val_loss: 0.6263 - val_mae: 0.5185 - val_mse: 0.6263\n",
      "Epoch 4978/5000\n",
      "4117/4117 [==============================] - 3s 800us/step - loss: 0.4748 - mae: 0.4038 - mse: 0.4748 - val_loss: 0.6711 - val_mae: 0.5406 - val_mse: 0.6711\n",
      "Epoch 4979/5000\n",
      "4117/4117 [==============================] - 3s 825us/step - loss: 0.4830 - mae: 0.4083 - mse: 0.4830 - val_loss: 0.6657 - val_mae: 0.5379 - val_mse: 0.6657\n",
      "Epoch 4980/5000\n",
      "1780/4117 [===========>..................] - ETA: 1s - loss: 0.4614 - mae: 0.3955 - mse: 0.4614"
     ]
    }
   ],
   "source": [
    "# train the neural network\n",
    "H = model.fit(x=trainX, y=trainY, validation_data=(testX, testY), epochs=5000, batch_size=20)\n",
    "train_mse = model.evaluate(trainX, trainY, verbose=0)\n",
    "test_mse = model.evaluate(testX, testY, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_8\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "resnet50 (Model)             (None, 2048)              23587712  \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 1)                 2049      \n",
      "=================================================================\n",
      "Total params: 23,589,761\n",
      "Trainable params: 2,049\n",
      "Non-trainable params: 23,587,712\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "resnet = ResNet50(include_top=False, pooling=\"avg\")\n",
    "model = Sequential()\n",
    "model.add(resnet)\n",
    "model.add(Dense(1))\n",
    "model.layers[0].trainable = False\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 4117 samples, validate on 1373 samples\n",
      "Epoch 1/30\n",
      "4117/4117 [==============================] - 261s 63ms/step - loss: 0.8323 - val_loss: 1.1889\n",
      "Epoch 2/30\n",
      "4117/4117 [==============================] - 288s 70ms/step - loss: 0.6682 - val_loss: 1.0790\n",
      "Epoch 3/30\n",
      "4117/4117 [==============================] - 322s 78ms/step - loss: 0.6107 - val_loss: 1.3984\n",
      "Epoch 4/30\n",
      "4117/4117 [==============================] - 277s 67ms/step - loss: 0.5288 - val_loss: 1.8008\n",
      "Epoch 5/30\n",
      " 832/4117 [=====>........................] - ETA: 2:36 - loss: 0.5123"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-76-962782e162fb>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'mean_squared_error'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mAdam\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrainX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrainY\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidation_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtestX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtestY\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m30\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, max_queue_size, workers, use_multiprocessing, **kwargs)\u001b[0m\n\u001b[1;32m   1237\u001b[0m                                         \u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1238\u001b[0m                                         \u001b[0mvalidation_steps\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidation_steps\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1239\u001b[0;31m                                         validation_freq=validation_freq)\n\u001b[0m\u001b[1;32m   1240\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1241\u001b[0m     def evaluate(self,\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/keras/engine/training_arrays.py\u001b[0m in \u001b[0;36mfit_loop\u001b[0;34m(model, fit_function, fit_inputs, out_labels, batch_size, epochs, verbose, callbacks, val_function, val_inputs, shuffle, initial_epoch, steps_per_epoch, validation_steps, validation_freq)\u001b[0m\n\u001b[1;32m    194\u001b[0m                     \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtoarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    195\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 196\u001b[0;31m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfit_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    197\u001b[0m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mto_list\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    198\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0ml\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mout_labels\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/keras/backend.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   3474\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3475\u001b[0m     fetched = self._callable_fn(*array_vals,\n\u001b[0;32m-> 3476\u001b[0;31m                                 run_metadata=self.run_metadata)\n\u001b[0m\u001b[1;32m   3477\u001b[0m     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call_fetch_callbacks\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfetched\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_fetches\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3478\u001b[0m     output_structure = nest.pack_sequence_as(\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/client/session.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1470\u001b[0m         ret = tf_session.TF_SessionRunCallable(self._session._session,\n\u001b[1;32m   1471\u001b[0m                                                \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_handle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1472\u001b[0;31m                                                run_metadata_ptr)\n\u001b[0m\u001b[1;32m   1473\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1474\u001b[0m           \u001b[0mproto_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "model.compile(loss='mean_squared_error', optimizer=Adam())\n",
    "model.fit(batch_size=32, x=trainX, y=trainY, validation_data=(testX, testY), epochs=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]]\n"
     ]
    }
   ],
   "source": [
    "example_batch = testX[1000:1010]\n",
    "example_result = model.predict(example_batch)\n",
    "print(example_result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('models/beauty_pred.ker') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=L size=128x128 at 0x7F878327EA10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred = all_labels[\"AF14\"]\n",
    "img = PIL.Image.open(path + \"AF14_01.png\").convert('L')\n",
    "print(pred)\n",
    "display(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]]\n"
     ]
    }
   ],
   "source": [
    "print(model.predict(example_batch))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}