Supervised_Learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use first 500,000 rows for training and testing\n",
    "data = pd.read_csv('train.csv',nrows=500000)\n",
    "train_data = data.iloc[:450000] # use first 450,000 obs. for training\n",
    "test_data = data.iloc[450000:] # use last 50,000 obs for testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get train and test target from their respective datasets\n",
    "train_target = train_data.pop('target')\n",
    "test_target = test_data.pop('target')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove id columns from datasets\n",
    "train_row_id = train_data.pop('row_id')\n",
    "train_time_id = train_data.pop('time_id')\n",
    "train_investment_id = train_data.pop('investment_id')\n",
    "\n",
    "test_row_id = test_data.pop('row_id')\n",
    "test_time_id = test_data.pop('time_id')\n",
    "test_investment_id = test_data.pop('investment_id')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Deep Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import DNN libraries\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.layers import BatchNormalization, Dense, Dropout\n",
    "from tensorflow.keras import Sequential \n",
    "from scipy.stats import pearsonr # pearson correlation coef"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create DNN \n",
    "def create_dnn(nodes,layers,dropout,activation):\n",
    "    model = Sequential()\n",
    "    \n",
    "    for i in list(range(1,layers+1)):\n",
    "        if i == 1:\n",
    "            model.add(Dense(nodes, input_dim = (train_data.shape[1]), activation = 'swish'))\n",
    "            model.add(BatchNormalization())\n",
    "            model.add(Dropout(0.3))\n",
    "        else:\n",
    "            model.add(Dense(i/2, activation='swish'))\n",
    "            model.add(BatchNormalization())\n",
    "            model.add(Dropout(0.3))\n",
    "\n",
    "    model.add(BatchNormalization())\n",
    "    model.add(Dense(1))\n",
    "    model.compile(loss = 'mse', optimizer = 'adam', metrics = ['mae'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/40\n",
      "110/110 [==============================] - 6s 58ms/step - loss: 2.2331 - mae: 0.8520\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 6s 58ms/step - loss: 1.6926 - mae: 0.8052\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 6s 58ms/step - loss: 1.4158 - mae: 0.7747\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 6s 58ms/step - loss: 1.2270 - mae: 0.7482\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 7s 60ms/step - loss: 1.1020 - mae: 0.7256\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 1.0264 - mae: 0.7074\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 7s 60ms/step - loss: 0.9124 - mae: 0.6630\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 0.8777 - mae: 0.6448\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 0.8713 - mae: 0.6409\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 0.8700 - mae: 0.6400\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 0.8687 - mae: 0.6393\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 7s 61ms/step - loss: 0.8682 - mae: 0.6389\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8679 - mae: 0.6390\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8679 - mae: 0.6390\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8677 - mae: 0.6388\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8676 - mae: 0.6387\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8673 - mae: 0.6385\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8678 - mae: 0.6388\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8672 - mae: 0.6386\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8670 - mae: 0.6385\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8672 - mae: 0.6386\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8665 - mae: 0.6385\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8670 - mae: 0.6384\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8665 - mae: 0.6386\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8661 - mae: 0.6383\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8657 - mae: 0.6381\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8658 - mae: 0.6382\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8658 - mae: 0.6380\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8656 - mae: 0.6380\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 0.8650 - mae: 0.6376\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8647 - mae: 0.6379\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8644 - mae: 0.6376\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8644 - mae: 0.6376\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8642 - mae: 0.6376\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8638 - mae: 0.6376\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8630 - mae: 0.6373\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8630 - mae: 0.6373\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8628 - mae: 0.6372\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8622 - mae: 0.6371\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8623 - mae: 0.6369\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 1.7590 - mae: 0.9459\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 1.4259 - mae: 0.8538\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 1.2618 - mae: 0.8029\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 1.1474 - mae: 0.7617\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 1.0731 - mae: 0.7331\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 1.0146 - mae: 0.7093\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 0.8842 - mae: 0.6474\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 0.8685 - mae: 0.6389\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8685 - mae: 0.6390\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8677 - mae: 0.6386\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8677 - mae: 0.6386\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8671 - mae: 0.6384\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8665 - mae: 0.6382\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8664 - mae: 0.6381\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8662 - mae: 0.6382\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8659 - mae: 0.6380\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8657 - mae: 0.6378\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8648 - mae: 0.6376\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8648 - mae: 0.6372\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8640 - mae: 0.6373\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8639 - mae: 0.6372\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8631 - mae: 0.6369\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8627 - mae: 0.6366\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8627 - mae: 0.6367\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8622 - mae: 0.6367\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8625 - mae: 0.6367\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8614 - mae: 0.6364\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8605 - mae: 0.6361\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8606 - mae: 0.6359\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8599 - mae: 0.6359\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8592 - mae: 0.6356\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8589 - mae: 0.6353\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8590 - mae: 0.6356\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8579 - mae: 0.6354\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8574 - mae: 0.6349\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8577 - mae: 0.6349\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8569 - mae: 0.6349\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8564 - mae: 0.6348\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8552 - mae: 0.6346\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8551 - mae: 0.6345\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 1.1683 - mae: 0.7760\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 1.0205 - mae: 0.7163\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.9451 - mae: 0.6808\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.9066 - mae: 0.6620\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 0.8858 - mae: 0.6509\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8730 - mae: 0.6445\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8642 - mae: 0.6390\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8573 - mae: 0.6350\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8567 - mae: 0.6347\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8561 - mae: 0.6345\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8559 - mae: 0.6345\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8548 - mae: 0.6341\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8540 - mae: 0.6337\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8533 - mae: 0.6337\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8533 - mae: 0.6336\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8526 - mae: 0.6334\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8525 - mae: 0.6336\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8511 - mae: 0.6332\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8506 - mae: 0.6331\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8497 - mae: 0.6326\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8489 - mae: 0.6326\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8482 - mae: 0.6322\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8474 - mae: 0.6323\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8471 - mae: 0.6320\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8465 - mae: 0.6321\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8446 - mae: 0.6312\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8437 - mae: 0.6316\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8435 - mae: 0.6315\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8407 - mae: 0.6305\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8410 - mae: 0.6308\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8400 - mae: 0.6304\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8398 - mae: 0.6302\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8379 - mae: 0.6301\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8366 - mae: 0.6297\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8371 - mae: 0.6299\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 7s 64ms/step - loss: 0.8360 - mae: 0.6292\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8343 - mae: 0.6289\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8334 - mae: 0.6287\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 7s 63ms/step - loss: 0.8328 - mae: 0.6290\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 7s 62ms/step - loss: 0.8320 - mae: 0.6284\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 2.1672 - mae: 0.8053\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.5760 - mae: 0.7407\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.2984 - mae: 0.7215\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.1572 - mae: 0.7076\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.0829 - mae: 0.7000\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.0265 - mae: 0.6897\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.9913 - mae: 0.6811\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.9466 - mae: 0.6680\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.9021 - mae: 0.6525\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8902 - mae: 0.6480\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8767 - mae: 0.6424\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8710 - mae: 0.6400\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8710 - mae: 0.6401\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8707 - mae: 0.6398\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8706 - mae: 0.6400\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8706 - mae: 0.6397\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8703 - mae: 0.6398\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8701 - mae: 0.6396\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8699 - mae: 0.6397\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8699 - mae: 0.6397\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8700 - mae: 0.6395\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8696 - mae: 0.6395\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8698 - mae: 0.6396\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8694 - mae: 0.6394\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8697 - mae: 0.6397\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8694 - mae: 0.6395\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8690 - mae: 0.6392\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8690 - mae: 0.6394\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8690 - mae: 0.6392\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8688 - mae: 0.6393\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8686 - mae: 0.6393\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8687 - mae: 0.6392\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8683 - mae: 0.6392\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8682 - mae: 0.6392\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8684 - mae: 0.6389\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8682 - mae: 0.6389\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8680 - mae: 0.6391\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8677 - mae: 0.6389\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8680 - mae: 0.6388\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8676 - mae: 0.6389\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 2.0785 - mae: 0.9366\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.5460 - mae: 0.8823\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.3287 - mae: 0.8260\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.2146 - mae: 0.7902\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.1342 - mae: 0.7590\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 1.0734 - mae: 0.7336\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.9896 - mae: 0.6998\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8983 - mae: 0.6567\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8800 - mae: 0.6460\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8736 - mae: 0.6418\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8709 - mae: 0.6403\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8694 - mae: 0.6394\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8685 - mae: 0.6391\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8681 - mae: 0.6387\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8676 - mae: 0.6387\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8673 - mae: 0.6386\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8673 - mae: 0.6387\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8667 - mae: 0.6383\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8670 - mae: 0.6385\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8665 - mae: 0.6383\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8660 - mae: 0.6381\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8664 - mae: 0.6382\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8657 - mae: 0.6377\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8654 - mae: 0.6379\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8646 - mae: 0.6376\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8647 - mae: 0.6374\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8646 - mae: 0.6374\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8638 - mae: 0.6375\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8636 - mae: 0.6371\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8633 - mae: 0.6372\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8626 - mae: 0.6369\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8624 - mae: 0.6369\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8622 - mae: 0.6367\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8613 - mae: 0.6364\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8607 - mae: 0.6362\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8608 - mae: 0.6362\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8603 - mae: 0.6361\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8598 - mae: 0.6360\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8594 - mae: 0.6357\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8590 - mae: 0.6357\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8890 - mae: 0.6485\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8673 - mae: 0.6395\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8626 - mae: 0.6371\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8619 - mae: 0.6365\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8609 - mae: 0.6363\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8603 - mae: 0.6359\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8597 - mae: 0.6356\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8587 - mae: 0.6353\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8580 - mae: 0.6352\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8578 - mae: 0.6351\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8567 - mae: 0.6347\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8555 - mae: 0.6346\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8551 - mae: 0.6344\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8547 - mae: 0.6339\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8536 - mae: 0.6336\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8533 - mae: 0.6338\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8528 - mae: 0.6335\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8520 - mae: 0.6334\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8516 - mae: 0.6333\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8504 - mae: 0.6331\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8498 - mae: 0.6329\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8488 - mae: 0.6323\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8482 - mae: 0.6325\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8468 - mae: 0.6320\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8469 - mae: 0.6319\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8459 - mae: 0.6319\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8448 - mae: 0.6315\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8451 - mae: 0.6319\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8436 - mae: 0.6312\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8436 - mae: 0.6314\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8423 - mae: 0.6310\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8410 - mae: 0.6306\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 3s 28ms/step - loss: 0.8419 - mae: 0.6307\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 3s 29ms/step - loss: 0.8411 - mae: 0.6308\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 3s 30ms/step - loss: 0.8392 - mae: 0.6302\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 3s 24ms/step - loss: 0.8396 - mae: 0.6300\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 3s 24ms/step - loss: 0.8379 - mae: 0.6300\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 3s 24ms/step - loss: 0.8374 - mae: 0.6298\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 3s 24ms/step - loss: 0.8373 - mae: 0.6298\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 3s 23ms/step - loss: 0.8357 - mae: 0.6293\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 2.0738 - mae: 0.9035\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.4467 - mae: 0.7250\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.2150 - mae: 0.7105\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.0935 - mae: 0.6993\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.0147 - mae: 0.6867\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9762 - mae: 0.6804\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9513 - mae: 0.6715\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9335 - mae: 0.6672\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9022 - mae: 0.6567\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8911 - mae: 0.6512\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8857 - mae: 0.6474\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8811 - mae: 0.6451\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8785 - mae: 0.6436\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8757 - mae: 0.6424\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8752 - mae: 0.6420\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8740 - mae: 0.6414\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8723 - mae: 0.6405\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8720 - mae: 0.6405\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8718 - mae: 0.6404\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6403\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8720 - mae: 0.6404\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6404\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8718 - mae: 0.6403\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8718 - mae: 0.6403\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8719 - mae: 0.6402\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8718 - mae: 0.6404\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8717 - mae: 0.6404\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8716 - mae: 0.6402\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8715 - mae: 0.6402\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8714 - mae: 0.6402\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 2s 20ms/step - loss: 1.2027 - mae: 0.8066\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.0192 - mae: 0.7245\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9509 - mae: 0.6881\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.9265 - mae: 0.6745\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.9105 - mae: 0.6650\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8988 - mae: 0.6583\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8912 - mae: 0.6529\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8842 - mae: 0.6493\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8796 - mae: 0.6462\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8706 - mae: 0.6402\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8680 - mae: 0.6389\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8673 - mae: 0.6386\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8668 - mae: 0.6385\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8664 - mae: 0.6382\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8664 - mae: 0.6380\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8657 - mae: 0.6380\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8654 - mae: 0.6378\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8640 - mae: 0.6373\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8633 - mae: 0.6370\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8635 - mae: 0.6369\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8628 - mae: 0.6368\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8624 - mae: 0.6365\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8623 - mae: 0.6366\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8621 - mae: 0.6364\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8618 - mae: 0.6363\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8613 - mae: 0.6363\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8609 - mae: 0.6362\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8606 - mae: 0.6360\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8604 - mae: 0.6358\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8600 - mae: 0.6358\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8595 - mae: 0.6357\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8593 - mae: 0.6356\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8591 - mae: 0.6355\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8583 - mae: 0.6354\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8586 - mae: 0.6353\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8567 - mae: 0.6349\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8578 - mae: 0.6352\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8565 - mae: 0.6351\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8563 - mae: 0.6346\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8562 - mae: 0.6350\n",
      "Epoch 1/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8714 - mae: 0.6406\n",
      "Epoch 2/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8677 - mae: 0.6385\n",
      "Epoch 3/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8653 - mae: 0.6377\n",
      "Epoch 4/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8640 - mae: 0.6373\n",
      "Epoch 5/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8625 - mae: 0.6365\n",
      "Epoch 6/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8618 - mae: 0.6364\n",
      "Epoch 7/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8607 - mae: 0.6359\n",
      "Epoch 8/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8598 - mae: 0.6355\n",
      "Epoch 9/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8599 - mae: 0.6358\n",
      "Epoch 10/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8584 - mae: 0.6348\n",
      "Epoch 11/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8583 - mae: 0.6353\n",
      "Epoch 12/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8574 - mae: 0.6348\n",
      "Epoch 13/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8572 - mae: 0.6349\n",
      "Epoch 14/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8562 - mae: 0.6344\n",
      "Epoch 15/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8555 - mae: 0.6342\n",
      "Epoch 16/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8548 - mae: 0.6343\n",
      "Epoch 17/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8554 - mae: 0.6342\n",
      "Epoch 18/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8542 - mae: 0.6340\n",
      "Epoch 19/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8535 - mae: 0.6337\n",
      "Epoch 20/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8533 - mae: 0.6337\n",
      "Epoch 21/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8524 - mae: 0.6334\n",
      "Epoch 22/40\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8516 - mae: 0.6333\n",
      "Epoch 23/40\n",
      "110/110 [==============================] - 1s 14ms/step - loss: 0.8510 - mae: 0.6332\n",
      "Epoch 24/40\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8513 - mae: 0.6332\n",
      "Epoch 25/40\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8506 - mae: 0.6330\n",
      "Epoch 26/40\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8494 - mae: 0.6327\n",
      "Epoch 27/40\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8494 - mae: 0.6325\n",
      "Epoch 28/40\n",
      "110/110 [==============================] - 2s 18ms/step - loss: 0.8490 - mae: 0.6326\n",
      "Epoch 29/40\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8485 - mae: 0.6325\n",
      "Epoch 30/40\n",
      "110/110 [==============================] - 2s 19ms/step - loss: 0.8476 - mae: 0.6321\n",
      "Epoch 31/40\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8473 - mae: 0.6325\n",
      "Epoch 32/40\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8465 - mae: 0.6320\n",
      "Epoch 33/40\n",
      "110/110 [==============================] - 2s 18ms/step - loss: 0.8464 - mae: 0.6319\n",
      "Epoch 34/40\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8465 - mae: 0.6322\n",
      "Epoch 35/40\n",
      "110/110 [==============================] - 2s 18ms/step - loss: 0.8451 - mae: 0.6318\n",
      "Epoch 36/40\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8446 - mae: 0.6315\n",
      "Epoch 37/40\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8437 - mae: 0.6316\n",
      "Epoch 38/40\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8442 - mae: 0.6317\n",
      "Epoch 39/40\n",
      "110/110 [==============================] - 2s 18ms/step - loss: 0.8431 - mae: 0.6313\n",
      "Epoch 40/40\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8430 - mae: 0.6315\n"
     ]
    }
   ],
   "source": [
    "models = {}\n",
    "for i in [512,256,128]:\n",
    "    for j in [5,4,3]:\n",
    "                model = create_dnn(i,j,0.3,'swish') # create a DNN with the given number of initial nodes and layers\n",
    "            \n",
    "                model.fit(train_data, train_target, epochs = 40, batch_size = 4096, verbose = 1) # fit the DNN\n",
    "                \n",
    "                comb = str(i)+'-'+str(j)\n",
    "                models[comb] = model # store the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get PCC and p-value on test set for each model\n",
    "pearson_coef = []\n",
    "pearson_p = []\n",
    "for i in models:\n",
    "    pred = models[i].predict(test_data).reshape(50000,)\n",
    "    targ = np.array(test_target).reshape(50000,).astype('float32')\n",
    "    results = pearsonr(pred, targ)\n",
    "    pearson_coef.append(results[0])\n",
    "    pearson_p.append(results[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model:  512-5\n",
      "Pearson Coef:  0.10435927508124318\n",
      "Pearson P-Value:  4.398409931642727e-121\n",
      "Model:  512-4\n",
      "Pearson Coef:  0.12227179884800957\n",
      "Pearson P-Value:  8.388047922672366e-166\n",
      "Model:  512-3\n",
      "Pearson Coef:  0.11576269055212128\n",
      "Pearson P-Value:  1.0253749826401292e-148\n",
      "Model:  256-5\n",
      "Pearson Coef:  0.1192094740967064\n",
      "Pearson P-Value:  1.208023395294987e-157\n",
      "Model:  256-4\n",
      "Pearson Coef:  0.1095912628794523\n",
      "Pearson P-Value:  2.1293748299609102e-133\n",
      "Model:  256-3\n",
      "Pearson Coef:  0.12035871192711757\n",
      "Pearson P-Value:  1.1098585533898039e-160\n",
      "Model:  128-5\n",
      "Pearson Coef:  0.049417309945315235\n",
      "Pearson P-Value:  2.038422867454673e-28\n",
      "Model:  128-4\n",
      "Pearson Coef:  0.12771977811513616\n",
      "Pearson P-Value:  7.645201554990986e-181\n",
      "Model:  128-3\n",
      "Pearson Coef:  0.1217218129144381\n",
      "Pearson P-Value:  2.5383371265869945e-164\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>512-5</th>\n",
       "      <th>512-4</th>\n",
       "      <th>512-3</th>\n",
       "      <th>256-5</th>\n",
       "      <th>256-4</th>\n",
       "      <th>256-3</th>\n",
       "      <th>128-5</th>\n",
       "      <th>128-4</th>\n",
       "      <th>128-3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PCC</th>\n",
       "      <td>1.043593e-01</td>\n",
       "      <td>1.222718e-01</td>\n",
       "      <td>1.157627e-01</td>\n",
       "      <td>1.192095e-01</td>\n",
       "      <td>1.095913e-01</td>\n",
       "      <td>1.203587e-01</td>\n",
       "      <td>4.941731e-02</td>\n",
       "      <td>1.277198e-01</td>\n",
       "      <td>1.217218e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P-Value</th>\n",
       "      <td>4.398410e-121</td>\n",
       "      <td>8.388048e-166</td>\n",
       "      <td>1.025375e-148</td>\n",
       "      <td>1.208023e-157</td>\n",
       "      <td>2.129375e-133</td>\n",
       "      <td>1.109859e-160</td>\n",
       "      <td>2.038423e-28</td>\n",
       "      <td>7.645202e-181</td>\n",
       "      <td>2.538337e-164</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 512-5          512-4          512-3          256-5  \\\n",
       "PCC       1.043593e-01   1.222718e-01   1.157627e-01   1.192095e-01   \n",
       "P-Value  4.398410e-121  8.388048e-166  1.025375e-148  1.208023e-157   \n",
       "\n",
       "                 256-4          256-3         128-5          128-4  \\\n",
       "PCC       1.095913e-01   1.203587e-01  4.941731e-02   1.277198e-01   \n",
       "P-Value  2.129375e-133  1.109859e-160  2.038423e-28  7.645202e-181   \n",
       "\n",
       "                 128-3  \n",
       "PCC       1.217218e-01  \n",
       "P-Value  2.538337e-164  "
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show results\n",
    "n = 0\n",
    "for i in models:\n",
    "    print(\"Model: \",i)\n",
    "    print(\"Pearson Coef: \",pearson_coef[n])\n",
    "    print(\"Pearson P-Value: \",pearson_p[n])\n",
    "    n+=1\n",
    "    \n",
    "DNN_results = pd.DataFrame([pearson_coef,pearson_p],columns= models.keys(),index=[\"PCC\",\"P-Value\"])\n",
    "DNN_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "models_rf = {}\n",
    "\n",
    "\n",
    "for i in [2,5,10]:\n",
    "    model= RandomForestRegressor(max_depth=i,n_estimators=17)\n",
    "    model.fit(train_data,train_target)\n",
    "    comb = str(i)\n",
    "    models_rf[comb] = model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [],
   "source": [
    "pearson_coef_rf = []\n",
    "pearson_p_rf = []\n",
    "for i in models_rf:\n",
    "    pred = models_rf[i].predict(test_data).reshape(50000,)\n",
    "    targ = np.array(test_target).reshape(50000,).astype('float32')\n",
    "    results = pearsonr(pred, targ)\n",
    "    pearson_coef_rf.append(results[0])\n",
    "    pearson_p_rf.append(results[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model:  2\n",
      "Pearson Coef:  0.07259279055511453\n",
      "Pearson P-Value:  2.1172311983616614e-59\n",
      "Model:  5\n",
      "Pearson Coef:  0.08145395648163131\n",
      "Pearson P-Value:  2.3295223367972207e-74\n",
      "Model:  10\n",
      "Pearson Coef:  0.08946673155727483\n",
      "Pearson P-Value:  2.2284561640876307e-89\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>2</th>\n",
       "      <th>5</th>\n",
       "      <th>10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PCC</th>\n",
       "      <td>7.259279e-02</td>\n",
       "      <td>8.145396e-02</td>\n",
       "      <td>8.946673e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P-Value</th>\n",
       "      <td>2.117231e-59</td>\n",
       "      <td>2.329522e-74</td>\n",
       "      <td>2.228456e-89</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    2             5            10\n",
       "PCC      7.259279e-02  8.145396e-02  8.946673e-02\n",
       "P-Value  2.117231e-59  2.329522e-74  2.228456e-89"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show results\n",
    "n = 0\n",
    "for i in models_rf:\n",
    "    print(\"Model: \",i)\n",
    "    print(\"Pearson Coef: \",pearson_coef_rf[n])\n",
    "    print(\"Pearson P-Value: \",pearson_p_rf[n])\n",
    "    n+=1\n",
    "    \n",
    "rf_results = pd.DataFrame([pearson_coef_rf,pearson_p_rf],columns= models_rf.keys(),index=[\"PCC\",\"P-Value\"])\n",
    "rf_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ensemble"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn.ensemble import VotingRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "mod1 = GradientBoostingRegressor(n_estimators=10)\n",
    "mod2 = RandomForestRegressor(n_estimators=10)\n",
    "mod3 = LinearRegression()\n",
    "ensemble = VotingRegressor(estimators=[('gb', mod1), ('rf', mod2), ('lr', mod3)])\n",
    "ensemble_model = ensemble.fit(train_data, train_target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred = ensemble_model.predict(test_data).reshape(50000,)\n",
    "targ = np.array(test_target).reshape(50000,).astype('float32')\n",
    "results = pearsonr(pred, targ)\n",
    "pearson_coef_en = results[0]\n",
    "pearson_p_en = results[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DNN - Improved"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 1.3485 - mae: 0.7957\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 1.0815 - mae: 0.7369\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 1.0219 - mae: 0.7085\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.9803 - mae: 0.6914\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.9522 - mae: 0.6784\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.9315 - mae: 0.6685\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.9152 - mae: 0.6618\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8967 - mae: 0.6539\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8776 - mae: 0.6441\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8728 - mae: 0.6411\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8712 - mae: 0.6400\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8704 - mae: 0.6398\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8700 - mae: 0.6396\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8699 - mae: 0.6394\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8697 - mae: 0.6394\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8694 - mae: 0.6395\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8693 - mae: 0.6394\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8693 - mae: 0.6393\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8692 - mae: 0.6393\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8691 - mae: 0.6392\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8689 - mae: 0.6390\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8686 - mae: 0.6391\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8685 - mae: 0.6390\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8683 - mae: 0.6390\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8681 - mae: 0.6388\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8681 - mae: 0.6389\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8679 - mae: 0.6387\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8678 - mae: 0.6387\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8676 - mae: 0.6387\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8676 - mae: 0.6388\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8673 - mae: 0.6384\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8671 - mae: 0.6385\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8672 - mae: 0.6386\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8667 - mae: 0.6384\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8665 - mae: 0.6386\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8662 - mae: 0.6380\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8666 - mae: 0.6384\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8660 - mae: 0.6382\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8658 - mae: 0.6381\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8651 - mae: 0.6379\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 2s 18ms/step - loss: 0.8648 - mae: 0.6377\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8649 - mae: 0.6377\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8646 - mae: 0.6378\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8640 - mae: 0.6372\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8637 - mae: 0.6373\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8632 - mae: 0.6370\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8634 - mae: 0.6371\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8626 - mae: 0.6371\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8623 - mae: 0.6369\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8617 - mae: 0.6366\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8617 - mae: 0.6367\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8616 - mae: 0.6365\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8610 - mae: 0.6365\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8607 - mae: 0.6362\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8602 - mae: 0.6361\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8601 - mae: 0.6361\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8597 - mae: 0.6360\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8594 - mae: 0.6359\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8596 - mae: 0.6360\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8590 - mae: 0.6360\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8593 - mae: 0.6358\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8582 - mae: 0.6358\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8574 - mae: 0.6354\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8579 - mae: 0.6354\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8583 - mae: 0.6358\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8577 - mae: 0.6354\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8571 - mae: 0.6353\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8564 - mae: 0.6352\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8563 - mae: 0.6354\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8571 - mae: 0.6351\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8560 - mae: 0.6348\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8561 - mae: 0.6351\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8556 - mae: 0.6349\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8555 - mae: 0.6348\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8557 - mae: 0.6350\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8548 - mae: 0.6347\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8548 - mae: 0.6347\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8546 - mae: 0.6346\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8538 - mae: 0.6344\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8543 - mae: 0.6347\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8534 - mae: 0.6342\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8541 - mae: 0.6348\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8534 - mae: 0.6345\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8535 - mae: 0.6344\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8529 - mae: 0.6343\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8526 - mae: 0.6345\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8519 - mae: 0.6339\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8526 - mae: 0.6344\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8518 - mae: 0.6342\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8513 - mae: 0.6340\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8514 - mae: 0.6340\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8507 - mae: 0.6339\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8507 - mae: 0.6338\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8509 - mae: 0.6338\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8506 - mae: 0.6337\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8499 - mae: 0.6338\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8502 - mae: 0.6337\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8496 - mae: 0.6336\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8495 - mae: 0.6336\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8490 - mae: 0.6331\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.9489 - mae: 0.6849\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.9026 - mae: 0.6602\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8877 - mae: 0.6517\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8780 - mae: 0.6458\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8731 - mae: 0.6430\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8689 - mae: 0.6404\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8673 - mae: 0.6396\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8660 - mae: 0.6385\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8651 - mae: 0.6378\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8641 - mae: 0.6375\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8637 - mae: 0.6372\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8633 - mae: 0.6370\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8625 - mae: 0.6367\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8625 - mae: 0.6367\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8620 - mae: 0.6364\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8615 - mae: 0.6362\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8612 - mae: 0.6363\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8609 - mae: 0.6362\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8606 - mae: 0.6362\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8604 - mae: 0.6361\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8596 - mae: 0.6358\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8593 - mae: 0.6355\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8588 - mae: 0.6356\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8579 - mae: 0.6353\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8587 - mae: 0.6355\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8576 - mae: 0.6353\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8574 - mae: 0.6353\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8566 - mae: 0.6349\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8565 - mae: 0.6348\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8567 - mae: 0.6349\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8559 - mae: 0.6346\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8557 - mae: 0.6348\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8550 - mae: 0.6345\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8555 - mae: 0.6346\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8550 - mae: 0.6346\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8546 - mae: 0.6344\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8541 - mae: 0.6342\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8532 - mae: 0.6341\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8533 - mae: 0.6339\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8531 - mae: 0.6340\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8527 - mae: 0.6338\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8526 - mae: 0.6339\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8521 - mae: 0.6339\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8518 - mae: 0.6337\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8507 - mae: 0.6333\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8506 - mae: 0.6335\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8510 - mae: 0.6336\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8499 - mae: 0.6333\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8492 - mae: 0.6332\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8494 - mae: 0.6332\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8492 - mae: 0.6332\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8491 - mae: 0.6329\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8489 - mae: 0.6330\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8482 - mae: 0.6330\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8481 - mae: 0.6325\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8476 - mae: 0.6328\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8471 - mae: 0.6326\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8473 - mae: 0.6324\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8469 - mae: 0.6326\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8464 - mae: 0.6325\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8466 - mae: 0.6323\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8451 - mae: 0.6323\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8448 - mae: 0.6323\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8453 - mae: 0.6318\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8442 - mae: 0.6318\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8442 - mae: 0.6320\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8451 - mae: 0.6320\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8440 - mae: 0.6319\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8443 - mae: 0.6318\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8433 - mae: 0.6317\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8435 - mae: 0.6316\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8439 - mae: 0.6315\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8428 - mae: 0.6316\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8418 - mae: 0.6315\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8433 - mae: 0.6320\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8417 - mae: 0.6311\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8414 - mae: 0.6317\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8418 - mae: 0.6313\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8409 - mae: 0.6313\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8408 - mae: 0.6312\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8410 - mae: 0.6314\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8404 - mae: 0.6310\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8403 - mae: 0.6309\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8383 - mae: 0.6304\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8404 - mae: 0.6311\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8400 - mae: 0.6308\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8404 - mae: 0.6311\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8399 - mae: 0.6308\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8390 - mae: 0.6308\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8387 - mae: 0.6306\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8387 - mae: 0.6308\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8374 - mae: 0.6301\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8388 - mae: 0.6308\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8376 - mae: 0.6303\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8370 - mae: 0.6304\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8379 - mae: 0.6303\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8379 - mae: 0.6305\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8358 - mae: 0.6299\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8368 - mae: 0.6299\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8362 - mae: 0.6301\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 1.1439 - mae: 0.7812\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 1.0020 - mae: 0.7154\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.9344 - mae: 0.6803\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.9011 - mae: 0.6608\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8836 - mae: 0.6506\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8738 - mae: 0.6448\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8679 - mae: 0.6411\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8648 - mae: 0.6391\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8621 - mae: 0.6376\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8606 - mae: 0.6367\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8597 - mae: 0.6355\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8592 - mae: 0.6355\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8590 - mae: 0.6356\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8584 - mae: 0.6354\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8585 - mae: 0.6350\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8573 - mae: 0.6349\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8574 - mae: 0.6349\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8573 - mae: 0.6349\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8565 - mae: 0.6348\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8563 - mae: 0.6347\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8563 - mae: 0.6344\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8547 - mae: 0.6341\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8548 - mae: 0.6343\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8540 - mae: 0.6340\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8531 - mae: 0.6336\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8534 - mae: 0.6338\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8527 - mae: 0.6338\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8524 - mae: 0.6336\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8519 - mae: 0.6334\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8520 - mae: 0.6334\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8514 - mae: 0.6333\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8505 - mae: 0.6329\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8496 - mae: 0.6327\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8500 - mae: 0.6327\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8486 - mae: 0.6323\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8487 - mae: 0.6328\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8489 - mae: 0.6327\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8476 - mae: 0.6322\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8473 - mae: 0.6324\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8468 - mae: 0.6323\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8467 - mae: 0.6323\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8448 - mae: 0.6319\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8457 - mae: 0.6322\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8453 - mae: 0.6316\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8443 - mae: 0.6317\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8437 - mae: 0.6317\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8438 - mae: 0.6315\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8433 - mae: 0.6315\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8425 - mae: 0.6312\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8423 - mae: 0.6312\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8424 - mae: 0.6311\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8418 - mae: 0.6311\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8405 - mae: 0.6308\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8404 - mae: 0.6309\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8398 - mae: 0.6305\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8388 - mae: 0.6302\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8395 - mae: 0.6303\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8398 - mae: 0.6308\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8386 - mae: 0.6301\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8382 - mae: 0.6303\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8379 - mae: 0.6301\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8380 - mae: 0.6299\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8375 - mae: 0.6298\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8370 - mae: 0.6299\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8361 - mae: 0.6302\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8366 - mae: 0.6296\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8350 - mae: 0.6297\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8353 - mae: 0.6298\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8353 - mae: 0.6294\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8343 - mae: 0.6293\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8351 - mae: 0.6295\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8341 - mae: 0.6293\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8341 - mae: 0.6292\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8336 - mae: 0.6293\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8337 - mae: 0.6288\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8333 - mae: 0.6291\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8328 - mae: 0.6291\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8326 - mae: 0.6289\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8325 - mae: 0.6288\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8323 - mae: 0.6286\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8315 - mae: 0.6285\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8311 - mae: 0.6288\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8305 - mae: 0.6285\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8306 - mae: 0.6285\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8298 - mae: 0.6286\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8304 - mae: 0.6280\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8301 - mae: 0.6282\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8308 - mae: 0.6286\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8295 - mae: 0.6282\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8284 - mae: 0.6280\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8287 - mae: 0.6282\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8290 - mae: 0.6278\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8284 - mae: 0.6282\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8278 - mae: 0.6277\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8277 - mae: 0.6278\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 2s 16ms/step - loss: 0.8268 - mae: 0.6277\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8279 - mae: 0.6279\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8265 - mae: 0.6275\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 2s 15ms/step - loss: 0.8273 - mae: 0.6278\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8264 - mae: 0.6275\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 1.1539 - mae: 0.7596\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 1.0062 - mae: 0.7087\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.9336 - mae: 0.6763\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8946 - mae: 0.6570\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8778 - mae: 0.6476\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8657 - mae: 0.6414\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8605 - mae: 0.6381\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8563 - mae: 0.6360\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8537 - mae: 0.6344\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8527 - mae: 0.6338\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8511 - mae: 0.6330\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8502 - mae: 0.6326\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8502 - mae: 0.6326\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8488 - mae: 0.6320\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8486 - mae: 0.6318\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8476 - mae: 0.6314\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8460 - mae: 0.6311\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8460 - mae: 0.6312\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8447 - mae: 0.6307\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8441 - mae: 0.6307\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8437 - mae: 0.6304\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8428 - mae: 0.6304\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8431 - mae: 0.6304\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8414 - mae: 0.6297\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8407 - mae: 0.6295\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8392 - mae: 0.6295\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8394 - mae: 0.6295\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8379 - mae: 0.6290\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8376 - mae: 0.6288\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8357 - mae: 0.6285\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8353 - mae: 0.6284\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8353 - mae: 0.6284\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8335 - mae: 0.6278\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8336 - mae: 0.6281\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8319 - mae: 0.6276\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8315 - mae: 0.6276\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8299 - mae: 0.6271\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8288 - mae: 0.6265\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8299 - mae: 0.6274\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8287 - mae: 0.6270\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8282 - mae: 0.6269\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8264 - mae: 0.6264\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8262 - mae: 0.6262\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8265 - mae: 0.6265\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8251 - mae: 0.6260\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8233 - mae: 0.6256\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8236 - mae: 0.6259\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8230 - mae: 0.6253\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8214 - mae: 0.6253\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8224 - mae: 0.6254\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8204 - mae: 0.6250\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8197 - mae: 0.6242\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8204 - mae: 0.6249\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8193 - mae: 0.6246\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8188 - mae: 0.6246\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8178 - mae: 0.6240\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8174 - mae: 0.6241\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8171 - mae: 0.6240\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8158 - mae: 0.6240\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8153 - mae: 0.6238\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8161 - mae: 0.6239\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8147 - mae: 0.6235\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8148 - mae: 0.6234\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8136 - mae: 0.6234\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8130 - mae: 0.6231\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 2s 14ms/step - loss: 0.8134 - mae: 0.6233\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8131 - mae: 0.6231\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8124 - mae: 0.6228\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8108 - mae: 0.6226\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8111 - mae: 0.6227\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8116 - mae: 0.6227\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8096 - mae: 0.6225\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8099 - mae: 0.6227\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8091 - mae: 0.6223\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8088 - mae: 0.6220\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8087 - mae: 0.6219\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8078 - mae: 0.6218\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8085 - mae: 0.6220\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8066 - mae: 0.6213\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8077 - mae: 0.6222\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8079 - mae: 0.6221\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8063 - mae: 0.6212\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8050 - mae: 0.6211\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8056 - mae: 0.6214\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8051 - mae: 0.6216\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8049 - mae: 0.6210\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8041 - mae: 0.6210\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8036 - mae: 0.6208\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8045 - mae: 0.6212\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8037 - mae: 0.6209\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8024 - mae: 0.6207\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8027 - mae: 0.6208\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8017 - mae: 0.6204\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8014 - mae: 0.6204\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8002 - mae: 0.6199\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8008 - mae: 0.6203\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8028 - mae: 0.6206\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.7994 - mae: 0.6201\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.7996 - mae: 0.6203\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.7988 - mae: 0.6198\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 3.5049 - mae: 1.2976\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 2.0807 - mae: 0.9390\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.2983 - mae: 0.8208\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 1.2001 - mae: 0.7890\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.1330 - mae: 0.7623\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.0889 - mae: 0.7446\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.0538 - mae: 0.7277\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.0016 - mae: 0.7052\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.9532 - mae: 0.6819\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.9355 - mae: 0.6732\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.9207 - mae: 0.6661\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.9098 - mae: 0.6608\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.9032 - mae: 0.6576\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8970 - mae: 0.6545\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8913 - mae: 0.6517\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8882 - mae: 0.6500\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8843 - mae: 0.6480\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8822 - mae: 0.6467\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8803 - mae: 0.6455\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8782 - mae: 0.6445\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8771 - mae: 0.6437\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8758 - mae: 0.6432\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8747 - mae: 0.6426\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8740 - mae: 0.6420\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8739 - mae: 0.6418\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8729 - mae: 0.6414\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8720 - mae: 0.6408\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8720 - mae: 0.6409\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8715 - mae: 0.6403\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8712 - mae: 0.6403\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8708 - mae: 0.6400\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8705 - mae: 0.6398\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8702 - mae: 0.6397\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8703 - mae: 0.6398\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8700 - mae: 0.6395\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8696 - mae: 0.6394\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8697 - mae: 0.6395\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8692 - mae: 0.6392\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8691 - mae: 0.6393\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8691 - mae: 0.6391\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8688 - mae: 0.6392\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8688 - mae: 0.6389\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8685 - mae: 0.6389\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8682 - mae: 0.6388\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8679 - mae: 0.6386\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8679 - mae: 0.6387\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8680 - mae: 0.6387\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8679 - mae: 0.6386\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8679 - mae: 0.6389\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8672 - mae: 0.6383\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8675 - mae: 0.6386\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8669 - mae: 0.6383\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8670 - mae: 0.6384\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8665 - mae: 0.6383\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8668 - mae: 0.6383\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8665 - mae: 0.6383\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8660 - mae: 0.6380\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8659 - mae: 0.6381\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8654 - mae: 0.6379\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8650 - mae: 0.6379\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8653 - mae: 0.6378\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8647 - mae: 0.6378\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8645 - mae: 0.6375\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8638 - mae: 0.6374\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8639 - mae: 0.6374\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8638 - mae: 0.6372\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8631 - mae: 0.6372\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8631 - mae: 0.6372\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8627 - mae: 0.6370\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8634 - mae: 0.6371\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8627 - mae: 0.6369\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8627 - mae: 0.6366\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8624 - mae: 0.6368\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8617 - mae: 0.6367\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8618 - mae: 0.6367\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8618 - mae: 0.6366\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8614 - mae: 0.6367\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8613 - mae: 0.6366\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8612 - mae: 0.6366\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8605 - mae: 0.6364\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8601 - mae: 0.6362\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8611 - mae: 0.6363\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8607 - mae: 0.6364\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8601 - mae: 0.6363\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8610 - mae: 0.6365\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8603 - mae: 0.6362\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8588 - mae: 0.6358\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8600 - mae: 0.6362\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8595 - mae: 0.6360\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8594 - mae: 0.6359\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8589 - mae: 0.6359\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8592 - mae: 0.6361\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8593 - mae: 0.6359\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8587 - mae: 0.6359\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8587 - mae: 0.6359\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8590 - mae: 0.6358\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8579 - mae: 0.6356\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8578 - mae: 0.6357\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8580 - mae: 0.6357\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8574 - mae: 0.6354\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8771 - mae: 0.6448\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8677 - mae: 0.6386\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8670 - mae: 0.6382\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8662 - mae: 0.6380\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8650 - mae: 0.6376\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8646 - mae: 0.6374\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8642 - mae: 0.6372\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8638 - mae: 0.6372\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8633 - mae: 0.6368\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8627 - mae: 0.6368\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8621 - mae: 0.6366\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8620 - mae: 0.6367\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8618 - mae: 0.6365\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8617 - mae: 0.6364\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8616 - mae: 0.6364\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8617 - mae: 0.6364\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8605 - mae: 0.6359\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8604 - mae: 0.6359\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8607 - mae: 0.6362\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8599 - mae: 0.6357\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8595 - mae: 0.6358\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8592 - mae: 0.6358\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8592 - mae: 0.6355\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8588 - mae: 0.6358\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8588 - mae: 0.6353\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8580 - mae: 0.6356\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8579 - mae: 0.6352\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8584 - mae: 0.6356\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8575 - mae: 0.6355\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8571 - mae: 0.6351\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8574 - mae: 0.6352\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8576 - mae: 0.6351\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8568 - mae: 0.6350\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8562 - mae: 0.6349\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8567 - mae: 0.6349\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8568 - mae: 0.6349\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8561 - mae: 0.6346\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8560 - mae: 0.6349\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8556 - mae: 0.6349\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8556 - mae: 0.6346\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8549 - mae: 0.6344\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8548 - mae: 0.6346\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8551 - mae: 0.6346\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8543 - mae: 0.6344\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8546 - mae: 0.6344\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8544 - mae: 0.6345\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8543 - mae: 0.6345\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8536 - mae: 0.6345\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8536 - mae: 0.6340\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8533 - mae: 0.6342\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8531 - mae: 0.6343\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8539 - mae: 0.6344\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8532 - mae: 0.6342\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8524 - mae: 0.6340\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8526 - mae: 0.6342\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8523 - mae: 0.6338\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8528 - mae: 0.6340\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8519 - mae: 0.6338\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8523 - mae: 0.6338\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8517 - mae: 0.6338\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8517 - mae: 0.6338\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8510 - mae: 0.6335\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8516 - mae: 0.6339\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8514 - mae: 0.6340\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8503 - mae: 0.6336\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8505 - mae: 0.6337\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8505 - mae: 0.6335\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8511 - mae: 0.6336\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8493 - mae: 0.6334\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8498 - mae: 0.6334\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8497 - mae: 0.6331\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8492 - mae: 0.6334\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8492 - mae: 0.6333\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8491 - mae: 0.6334\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8491 - mae: 0.6335\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8504 - mae: 0.6333\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8491 - mae: 0.6330\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8481 - mae: 0.6331\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8487 - mae: 0.6331\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8487 - mae: 0.6331\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8483 - mae: 0.6332\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8488 - mae: 0.6331\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8477 - mae: 0.6331\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8475 - mae: 0.6330\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8475 - mae: 0.6329\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8480 - mae: 0.6332\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8484 - mae: 0.6332\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8475 - mae: 0.6327\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8472 - mae: 0.6328\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8472 - mae: 0.6327\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8480 - mae: 0.6332\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8464 - mae: 0.6328\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8477 - mae: 0.6329\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8467 - mae: 0.6329\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8470 - mae: 0.6328\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8464 - mae: 0.6327\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8476 - mae: 0.6330\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8460 - mae: 0.6326\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8470 - mae: 0.6329\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8462 - mae: 0.6324\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.9018 - mae: 0.6554\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8735 - mae: 0.6422\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8671 - mae: 0.6390\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8653 - mae: 0.6381\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8647 - mae: 0.6377\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8635 - mae: 0.6374\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8633 - mae: 0.6371\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8622 - mae: 0.6367\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8616 - mae: 0.6365\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8606 - mae: 0.6362\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8602 - mae: 0.6362\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8606 - mae: 0.6360\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8592 - mae: 0.6356\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8588 - mae: 0.6355\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8590 - mae: 0.6355\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8585 - mae: 0.6355\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8580 - mae: 0.6353\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8572 - mae: 0.6350\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8576 - mae: 0.6351\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8573 - mae: 0.6350\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8567 - mae: 0.6348\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8565 - mae: 0.6348\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8555 - mae: 0.6345\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8560 - mae: 0.6347\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8551 - mae: 0.6346\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8544 - mae: 0.6341\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8546 - mae: 0.6344\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8541 - mae: 0.6341\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8539 - mae: 0.6340\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8536 - mae: 0.6342\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8541 - mae: 0.6340\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8527 - mae: 0.6337\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8530 - mae: 0.6341\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8524 - mae: 0.6336\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8519 - mae: 0.6336\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8513 - mae: 0.6335\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8515 - mae: 0.6337\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8510 - mae: 0.6333\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8510 - mae: 0.6334\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8508 - mae: 0.6328\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8506 - mae: 0.6333\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8502 - mae: 0.6331\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8494 - mae: 0.6330\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8491 - mae: 0.6329\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8490 - mae: 0.6332\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8489 - mae: 0.6328\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8485 - mae: 0.6326\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8489 - mae: 0.6328\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8485 - mae: 0.6326\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8477 - mae: 0.6325\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8471 - mae: 0.6324\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8481 - mae: 0.6327\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8469 - mae: 0.6321\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8468 - mae: 0.6321\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8464 - mae: 0.6321\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8471 - mae: 0.6323\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8458 - mae: 0.6320\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8471 - mae: 0.6322\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8461 - mae: 0.6324\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8464 - mae: 0.6320\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8460 - mae: 0.6322\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8450 - mae: 0.6319\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8460 - mae: 0.6322\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8445 - mae: 0.6317\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8461 - mae: 0.6321\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8445 - mae: 0.6318\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8458 - mae: 0.6323\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8448 - mae: 0.6318\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8444 - mae: 0.6317\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8442 - mae: 0.6316\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8444 - mae: 0.6318\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8443 - mae: 0.6317\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8437 - mae: 0.6312\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8442 - mae: 0.6317\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8434 - mae: 0.6315\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8440 - mae: 0.6316\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8437 - mae: 0.6316\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8427 - mae: 0.6312\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8427 - mae: 0.6314\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8430 - mae: 0.6313\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8429 - mae: 0.6310\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8426 - mae: 0.6314\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8428 - mae: 0.6315\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8431 - mae: 0.6313\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8428 - mae: 0.6311\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8422 - mae: 0.6312\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8415 - mae: 0.6309\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8421 - mae: 0.6311\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8411 - mae: 0.6308\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8421 - mae: 0.6311\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8418 - mae: 0.6312\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8416 - mae: 0.6311\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8412 - mae: 0.6308\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8404 - mae: 0.6306\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8407 - mae: 0.6310\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8407 - mae: 0.6313\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8405 - mae: 0.6308\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8404 - mae: 0.6307\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8408 - mae: 0.6307\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8400 - mae: 0.6307\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.0374 - mae: 0.7120\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.9356 - mae: 0.6747\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8933 - mae: 0.6552\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8735 - mae: 0.6448\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8649 - mae: 0.6401\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8597 - mae: 0.6369\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8573 - mae: 0.6359\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8556 - mae: 0.6347\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8553 - mae: 0.6344\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8542 - mae: 0.6340\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8536 - mae: 0.6338\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8540 - mae: 0.6335\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8527 - mae: 0.6333\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8527 - mae: 0.6334\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8516 - mae: 0.6327\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8518 - mae: 0.6330\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8500 - mae: 0.6325\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8501 - mae: 0.6325\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8496 - mae: 0.6322\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8491 - mae: 0.6321\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8483 - mae: 0.6321\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8481 - mae: 0.6319\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8475 - mae: 0.6317\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8468 - mae: 0.6314\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8462 - mae: 0.6311\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8461 - mae: 0.6315\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8457 - mae: 0.6312\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8448 - mae: 0.6308\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8442 - mae: 0.6308\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8439 - mae: 0.6307\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8435 - mae: 0.6307\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8440 - mae: 0.6309\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8418 - mae: 0.6298\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8420 - mae: 0.6303\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8412 - mae: 0.6301\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8412 - mae: 0.6304\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8406 - mae: 0.6301\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8400 - mae: 0.6295\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8402 - mae: 0.6297\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8386 - mae: 0.6298\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8385 - mae: 0.6293\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8383 - mae: 0.6294\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8381 - mae: 0.6294\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8376 - mae: 0.6291\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8368 - mae: 0.6290\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8365 - mae: 0.6288\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8360 - mae: 0.6289\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8353 - mae: 0.6289\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8355 - mae: 0.6286\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8348 - mae: 0.6288\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8345 - mae: 0.6283\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8349 - mae: 0.6288\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8336 - mae: 0.6284\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8336 - mae: 0.6283\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8328 - mae: 0.6283\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8318 - mae: 0.6280\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8325 - mae: 0.6283\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8322 - mae: 0.6281\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8317 - mae: 0.6278\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8320 - mae: 0.6280\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8302 - mae: 0.6278\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8304 - mae: 0.6278\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8299 - mae: 0.6277\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8299 - mae: 0.6277\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8292 - mae: 0.6273\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8297 - mae: 0.6274\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8291 - mae: 0.6274\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8289 - mae: 0.6273\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8284 - mae: 0.6272\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8279 - mae: 0.6272\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8281 - mae: 0.6269\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8268 - mae: 0.6267\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8274 - mae: 0.6273\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8274 - mae: 0.6272\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8273 - mae: 0.6272\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8270 - mae: 0.6268\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8253 - mae: 0.6266\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8263 - mae: 0.6268\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8267 - mae: 0.6267\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8265 - mae: 0.6269\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8256 - mae: 0.6264\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8253 - mae: 0.6264\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8251 - mae: 0.6269\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8241 - mae: 0.6262\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8238 - mae: 0.6262\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8241 - mae: 0.6261\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8252 - mae: 0.6269\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8235 - mae: 0.6260\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8232 - mae: 0.6262\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8228 - mae: 0.6261\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8243 - mae: 0.6263\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8234 - mae: 0.6261\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8233 - mae: 0.6262\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8230 - mae: 0.6261\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8226 - mae: 0.6258\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8226 - mae: 0.6263\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8224 - mae: 0.6260\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8206 - mae: 0.6255\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8222 - mae: 0.6259\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8220 - mae: 0.6257\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 2.2616 - mae: 1.0213\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 1.3755 - mae: 0.8557\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.1991 - mae: 0.7847\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.1266 - mae: 0.7522\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.0784 - mae: 0.7381\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.0373 - mae: 0.7170\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.0036 - mae: 0.7032\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.9513 - mae: 0.6825\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8787 - mae: 0.6448\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8714 - mae: 0.6402\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8713 - mae: 0.6402\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8709 - mae: 0.6401\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8709 - mae: 0.6400\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8709 - mae: 0.6401\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8707 - mae: 0.6400\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8705 - mae: 0.6399\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8707 - mae: 0.6400\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8703 - mae: 0.6399\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8703 - mae: 0.6398\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8703 - mae: 0.6398\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8702 - mae: 0.6398\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8697 - mae: 0.6395\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8695 - mae: 0.6397\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8695 - mae: 0.6395\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8692 - mae: 0.6396\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8689 - mae: 0.6394\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8688 - mae: 0.6395\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8691 - mae: 0.6394\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8688 - mae: 0.6394\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8688 - mae: 0.6392\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8682 - mae: 0.6393\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8678 - mae: 0.6389\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8680 - mae: 0.6391\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8680 - mae: 0.6391\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8679 - mae: 0.6388\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8677 - mae: 0.6389\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8676 - mae: 0.6389\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8670 - mae: 0.6387\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8669 - mae: 0.6388\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8670 - mae: 0.6386\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8667 - mae: 0.6386\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 2s 17ms/step - loss: 0.8662 - mae: 0.6384\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8665 - mae: 0.6384\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8657 - mae: 0.6383\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8658 - mae: 0.6381\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8659 - mae: 0.6383\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8660 - mae: 0.6382\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8655 - mae: 0.6382\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8652 - mae: 0.6379\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8656 - mae: 0.6382\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8649 - mae: 0.6380\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8654 - mae: 0.6378\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8653 - mae: 0.6379\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8649 - mae: 0.6378\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8645 - mae: 0.6376\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8651 - mae: 0.6379\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8644 - mae: 0.6376\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8645 - mae: 0.6377\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8645 - mae: 0.6377\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8640 - mae: 0.6375\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8638 - mae: 0.6374\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8644 - mae: 0.6376\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8639 - mae: 0.6375\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8639 - mae: 0.6373\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8636 - mae: 0.6373\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8636 - mae: 0.6373\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8635 - mae: 0.6374\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8632 - mae: 0.6371\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8636 - mae: 0.6372\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8633 - mae: 0.6374\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8630 - mae: 0.6372\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8629 - mae: 0.6373\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8627 - mae: 0.6370\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8628 - mae: 0.6373\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8627 - mae: 0.6369\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8625 - mae: 0.6371\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8623 - mae: 0.6369\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8617 - mae: 0.6368\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8623 - mae: 0.6369\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8622 - mae: 0.6369\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8618 - mae: 0.6367\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8621 - mae: 0.6368\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8620 - mae: 0.6370\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8618 - mae: 0.6368\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8616 - mae: 0.6369\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8608 - mae: 0.6364\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8614 - mae: 0.6366\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8611 - mae: 0.6368\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8613 - mae: 0.6365\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8609 - mae: 0.6365\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8603 - mae: 0.6363\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8608 - mae: 0.6364\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8603 - mae: 0.6363\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8605 - mae: 0.6363\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8611 - mae: 0.6364\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8606 - mae: 0.6365\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8608 - mae: 0.6364\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8603 - mae: 0.6363\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8596 - mae: 0.6362\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8605 - mae: 0.6364\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8943 - mae: 0.6550\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8760 - mae: 0.6432\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8717 - mae: 0.6405\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8705 - mae: 0.6398\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8698 - mae: 0.6395\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8695 - mae: 0.6394\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8688 - mae: 0.6393\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8682 - mae: 0.6388\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8676 - mae: 0.6390\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8671 - mae: 0.6386\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8661 - mae: 0.6383\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8656 - mae: 0.6379\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8651 - mae: 0.6378\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8648 - mae: 0.6378\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8643 - mae: 0.6375\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8635 - mae: 0.6373\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8633 - mae: 0.6372\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8636 - mae: 0.6371\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8636 - mae: 0.6371\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8628 - mae: 0.6371\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8631 - mae: 0.6368\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8624 - mae: 0.6368\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8624 - mae: 0.6368\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8624 - mae: 0.6367\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 13ms/step - loss: 0.8619 - mae: 0.6365\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8623 - mae: 0.6366\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8613 - mae: 0.6364\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8614 - mae: 0.6366\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 12ms/step - loss: 0.8615 - mae: 0.6365\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8611 - mae: 0.6363\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8610 - mae: 0.6363\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8610 - mae: 0.6364\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8607 - mae: 0.6361\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8605 - mae: 0.6361\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8603 - mae: 0.6361\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8608 - mae: 0.6363\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8603 - mae: 0.6360\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8594 - mae: 0.6359\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8598 - mae: 0.6360\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8598 - mae: 0.6358\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8594 - mae: 0.6362\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8599 - mae: 0.6358\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8599 - mae: 0.6359\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8594 - mae: 0.6356\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8594 - mae: 0.6359\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8588 - mae: 0.6356\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8588 - mae: 0.6356\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8591 - mae: 0.6357\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8592 - mae: 0.6358\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8582 - mae: 0.6355\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8587 - mae: 0.6356\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 0.8584 - mae: 0.6356\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8580 - mae: 0.6354\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8581 - mae: 0.6353\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8572 - mae: 0.6352\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8583 - mae: 0.6354\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8578 - mae: 0.6355\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8573 - mae: 0.6354\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8573 - mae: 0.6351\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8572 - mae: 0.6352\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8572 - mae: 0.6350\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8571 - mae: 0.6352\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8569 - mae: 0.6351\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8570 - mae: 0.6354\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8570 - mae: 0.6352\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8565 - mae: 0.6350\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8562 - mae: 0.6350\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8566 - mae: 0.6351\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8560 - mae: 0.6349\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8561 - mae: 0.6351\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8557 - mae: 0.6348\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8557 - mae: 0.6349\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8564 - mae: 0.6352\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8560 - mae: 0.6349\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8560 - mae: 0.6349\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8558 - mae: 0.6349\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8558 - mae: 0.6349\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8553 - mae: 0.6348\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8556 - mae: 0.6347\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8552 - mae: 0.6348\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8551 - mae: 0.6348\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8552 - mae: 0.6347\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8556 - mae: 0.6347\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8555 - mae: 0.6348\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8557 - mae: 0.6346\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8552 - mae: 0.6347\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8555 - mae: 0.6347\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8548 - mae: 0.6347\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8553 - mae: 0.6348\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8549 - mae: 0.6344\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8547 - mae: 0.6347\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8541 - mae: 0.6345\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8552 - mae: 0.6350\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8546 - mae: 0.6343\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8547 - mae: 0.6346\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8543 - mae: 0.6346\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8540 - mae: 0.6345\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8547 - mae: 0.6346\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8547 - mae: 0.6346\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8543 - mae: 0.6343\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 2.0834 - mae: 0.9116\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.6482 - mae: 0.8598\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.3854 - mae: 0.8135\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 11ms/step - loss: 1.2180 - mae: 0.7706\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 1.1078 - mae: 0.7410\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 1.0267 - mae: 0.7128\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8661 - mae: 0.6386\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8631 - mae: 0.6373\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8627 - mae: 0.6372\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8628 - mae: 0.6370\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8624 - mae: 0.6369\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8627 - mae: 0.6370\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8621 - mae: 0.6370\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8622 - mae: 0.6370\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8620 - mae: 0.6369\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8617 - mae: 0.6368\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8622 - mae: 0.6367\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8621 - mae: 0.6369\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8617 - mae: 0.6365\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8616 - mae: 0.6366\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8611 - mae: 0.6364\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8614 - mae: 0.6364\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8603 - mae: 0.6362\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8604 - mae: 0.6361\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8604 - mae: 0.6362\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8608 - mae: 0.6362\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8607 - mae: 0.6363\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8597 - mae: 0.6361\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8594 - mae: 0.6357\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8601 - mae: 0.6361\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8593 - mae: 0.6359\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8593 - mae: 0.6358\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8593 - mae: 0.6357\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8589 - mae: 0.6356\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8593 - mae: 0.6355\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8586 - mae: 0.6354\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8584 - mae: 0.6354\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8580 - mae: 0.6353\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8581 - mae: 0.6354\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8583 - mae: 0.6352\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8575 - mae: 0.6352\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8576 - mae: 0.6353\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8574 - mae: 0.6351\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8571 - mae: 0.6351\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8565 - mae: 0.6349\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8571 - mae: 0.6350\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8564 - mae: 0.6349\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8568 - mae: 0.6349\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8558 - mae: 0.6347\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8562 - mae: 0.6348\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8561 - mae: 0.6348\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8555 - mae: 0.6346\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8559 - mae: 0.6348\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8554 - mae: 0.6346\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8558 - mae: 0.6348\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8555 - mae: 0.6345\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8547 - mae: 0.6346\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8551 - mae: 0.6344\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8546 - mae: 0.6344\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8543 - mae: 0.6342\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8536 - mae: 0.6341\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8534 - mae: 0.6342\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8535 - mae: 0.6339\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8537 - mae: 0.6343\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8536 - mae: 0.6338\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8540 - mae: 0.6339\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8536 - mae: 0.6343\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8535 - mae: 0.6341\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8535 - mae: 0.6339\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8532 - mae: 0.6341\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8527 - mae: 0.6338\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8525 - mae: 0.6339\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8526 - mae: 0.6337\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8525 - mae: 0.6337\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8520 - mae: 0.6339\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8525 - mae: 0.6336\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8526 - mae: 0.6338\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8520 - mae: 0.6335\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8523 - mae: 0.6335\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8518 - mae: 0.6334\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8517 - mae: 0.6336\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8522 - mae: 0.6340\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8513 - mae: 0.6334\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8519 - mae: 0.6338\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8521 - mae: 0.6336\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8513 - mae: 0.6335\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8517 - mae: 0.6336\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8516 - mae: 0.6335\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8515 - mae: 0.6333\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8517 - mae: 0.6335\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8513 - mae: 0.6335\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8512 - mae: 0.6334\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8507 - mae: 0.6334\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8504 - mae: 0.6330\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8512 - mae: 0.6337\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8510 - mae: 0.6335\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8508 - mae: 0.6334\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8506 - mae: 0.6331\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8502 - mae: 0.6331\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8495 - mae: 0.6333\n",
      "Epoch 1/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8672 - mae: 0.6390\n",
      "Epoch 2/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8629 - mae: 0.6373\n",
      "Epoch 3/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8606 - mae: 0.6364\n",
      "Epoch 4/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8589 - mae: 0.6356\n",
      "Epoch 5/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8579 - mae: 0.6350\n",
      "Epoch 6/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8575 - mae: 0.6350\n",
      "Epoch 7/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8568 - mae: 0.6347\n",
      "Epoch 8/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8558 - mae: 0.6344\n",
      "Epoch 9/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8550 - mae: 0.6341\n",
      "Epoch 10/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8544 - mae: 0.6337\n",
      "Epoch 11/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8540 - mae: 0.6340\n",
      "Epoch 12/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8536 - mae: 0.6335\n",
      "Epoch 13/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8537 - mae: 0.6336\n",
      "Epoch 14/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8533 - mae: 0.6334\n",
      "Epoch 15/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8523 - mae: 0.6332\n",
      "Epoch 16/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8521 - mae: 0.6330\n",
      "Epoch 17/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8509 - mae: 0.6327\n",
      "Epoch 18/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8505 - mae: 0.6328\n",
      "Epoch 19/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8511 - mae: 0.6330\n",
      "Epoch 20/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8506 - mae: 0.6325\n",
      "Epoch 21/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8502 - mae: 0.6328\n",
      "Epoch 22/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8502 - mae: 0.6326\n",
      "Epoch 23/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8495 - mae: 0.6325\n",
      "Epoch 24/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8493 - mae: 0.6323\n",
      "Epoch 25/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8490 - mae: 0.6322\n",
      "Epoch 26/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8487 - mae: 0.6320\n",
      "Epoch 27/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8482 - mae: 0.6322\n",
      "Epoch 28/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8478 - mae: 0.6319\n",
      "Epoch 29/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8479 - mae: 0.6319\n",
      "Epoch 30/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8472 - mae: 0.6319\n",
      "Epoch 31/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8470 - mae: 0.6317\n",
      "Epoch 32/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8467 - mae: 0.6318\n",
      "Epoch 33/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8464 - mae: 0.6317\n",
      "Epoch 34/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8463 - mae: 0.6315\n",
      "Epoch 35/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8461 - mae: 0.6316\n",
      "Epoch 36/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8452 - mae: 0.6313\n",
      "Epoch 37/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8458 - mae: 0.6315\n",
      "Epoch 38/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8453 - mae: 0.6311\n",
      "Epoch 39/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8452 - mae: 0.6315\n",
      "Epoch 40/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8453 - mae: 0.6313\n",
      "Epoch 41/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8450 - mae: 0.6315\n",
      "Epoch 42/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8445 - mae: 0.6309\n",
      "Epoch 43/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8441 - mae: 0.6312\n",
      "Epoch 44/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8437 - mae: 0.6311\n",
      "Epoch 45/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8440 - mae: 0.6308\n",
      "Epoch 46/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8442 - mae: 0.6309\n",
      "Epoch 47/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8433 - mae: 0.6309\n",
      "Epoch 48/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8432 - mae: 0.6308\n",
      "Epoch 49/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8432 - mae: 0.6309\n",
      "Epoch 50/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8432 - mae: 0.6310\n",
      "Epoch 51/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8422 - mae: 0.6306\n",
      "Epoch 52/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8429 - mae: 0.6308\n",
      "Epoch 53/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8430 - mae: 0.6305\n",
      "Epoch 54/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8425 - mae: 0.6305\n",
      "Epoch 55/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8423 - mae: 0.6307\n",
      "Epoch 56/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8421 - mae: 0.6304\n",
      "Epoch 57/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8423 - mae: 0.6307\n",
      "Epoch 58/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8417 - mae: 0.6305\n",
      "Epoch 59/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8423 - mae: 0.6307\n",
      "Epoch 60/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8410 - mae: 0.6303\n",
      "Epoch 61/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8411 - mae: 0.6304\n",
      "Epoch 62/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8414 - mae: 0.6304\n",
      "Epoch 63/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8410 - mae: 0.6304\n",
      "Epoch 64/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8410 - mae: 0.6301\n",
      "Epoch 65/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8413 - mae: 0.6304\n",
      "Epoch 66/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8403 - mae: 0.6302\n",
      "Epoch 67/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8402 - mae: 0.6302\n",
      "Epoch 68/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8407 - mae: 0.6301\n",
      "Epoch 69/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8397 - mae: 0.6299\n",
      "Epoch 70/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8406 - mae: 0.6303\n",
      "Epoch 71/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8403 - mae: 0.6305\n",
      "Epoch 72/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8396 - mae: 0.6300\n",
      "Epoch 73/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8401 - mae: 0.6302\n",
      "Epoch 74/100\n",
      "110/110 [==============================] - 1s 10ms/step - loss: 0.8395 - mae: 0.6301\n",
      "Epoch 75/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8396 - mae: 0.6301\n",
      "Epoch 76/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8387 - mae: 0.6298\n",
      "Epoch 77/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8387 - mae: 0.6298\n",
      "Epoch 78/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8391 - mae: 0.6298\n",
      "Epoch 79/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8396 - mae: 0.6303\n",
      "Epoch 80/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8386 - mae: 0.6300\n",
      "Epoch 81/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8387 - mae: 0.6300\n",
      "Epoch 82/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8392 - mae: 0.6300\n",
      "Epoch 83/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8384 - mae: 0.6298\n",
      "Epoch 84/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8385 - mae: 0.6296\n",
      "Epoch 85/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8381 - mae: 0.6299\n",
      "Epoch 86/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8388 - mae: 0.6298\n",
      "Epoch 87/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8384 - mae: 0.6298\n",
      "Epoch 88/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8382 - mae: 0.6297\n",
      "Epoch 89/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8377 - mae: 0.6295\n",
      "Epoch 90/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8381 - mae: 0.6294\n",
      "Epoch 91/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8384 - mae: 0.6298\n",
      "Epoch 92/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8377 - mae: 0.6295\n",
      "Epoch 93/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8381 - mae: 0.6298\n",
      "Epoch 94/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8391 - mae: 0.6301\n",
      "Epoch 95/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8382 - mae: 0.6299\n",
      "Epoch 96/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8375 - mae: 0.6297\n",
      "Epoch 97/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8368 - mae: 0.6295\n",
      "Epoch 98/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8377 - mae: 0.6296\n",
      "Epoch 99/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8378 - mae: 0.6297\n",
      "Epoch 100/100\n",
      "110/110 [==============================] - 1s 9ms/step - loss: 0.8377 - mae: 0.6298\n"
     ]
    }
   ],
   "source": [
    "models_improved = {}\n",
    "for i in [128,64,32]:\n",
    "    for j in [5,4,3,2]:\n",
    "                model = create_dnn(i,j,0.3,'swish') # create a DNN with the given number of initial nodes and layers\n",
    "\n",
    "                model.fit(train_data, train_target, epochs = 100, batch_size = 4096, verbose = 1) # fit the DNN\n",
    "                \n",
    "                comb = str(i)+'-'+str(j)\n",
    "                models_improved[comb] = model # store the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get PCC and p-value on test set for each model\n",
    "pearson_coef_imp = []\n",
    "pearson_p_imp = []\n",
    "for i in models_improved:\n",
    "    pred = models_improved[i].predict(test_data).reshape(50000,)\n",
    "    targ = np.array(test_target).reshape(50000,).astype('float32')\n",
    "    results = pearsonr(pred, targ)\n",
    "    pearson_coef_imp.append(results[0])\n",
    "    pearson_p_imp.append(results[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.10743922196923479,\n",
       " 0.11347369519451647,\n",
       " 0.11492025134806846,\n",
       " 0.11225180171056587,\n",
       " 0.12389366078207467,\n",
       " 0.11734464508021099,\n",
       " 0.11614679820352401,\n",
       " 0.1228070722234781,\n",
       " 0.13433897056388372,\n",
       " 0.1318593599667382,\n",
       " 0.12806720395610785,\n",
       " 0.12040990904751069]"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pearson_coef_imp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Save Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DNN\n",
    "with open('dnn_keys.data', 'wb') as fi:\n",
    "    pickle.dump(list(models.keys()), fi)\n",
    "with open('dnn_coefs.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_coef),fi)\n",
    "with open('dnn_p.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_p),fi)\n",
    "    \n",
    "# RF\n",
    "with open('rf_keys.data', 'wb') as fi:\n",
    "    pickle.dump(list(models_rf.keys()), fi)\n",
    "with open('rf_coefs.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_coef_rf),fi)\n",
    "with open('rf_p.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_p_rf),fi)\n",
    "    \n",
    "    \n",
    "# DNN - Improved\n",
    "with open('dnn_imp_keys.data', 'wb') as fi:\n",
    "    pickle.dump(list(models_improved.keys()), fi)\n",
    "with open('dnn_imp_coefs.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_coef_imp),fi)\n",
    "with open('dnn_imp_p.data','wb') as fi:\n",
    "    pickle.dump(list(pearson_p_imp),fi)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['128-5',\n",
       " '128-4',\n",
       " '128-3',\n",
       " '128-2',\n",
       " '64-5',\n",
       " '64-4',\n",
       " '64-3',\n",
       " '64-2',\n",
       " '32-5',\n",
       " '32-4',\n",
       " '32-3',\n",
       " '32-2']"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open('dnn_imp_keys.data', 'rb') as fi:\n",
    "   x = pickle.load(fi)\n",
    "x"
   ]
  },
  {
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}