Model Generation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fgZ9gjmPfSnK"
   },
   "source": [
    "## Importing Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "baYFZMW_bJHh"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Make numpy values easier to read.\n",
    "np.set_printoptions(precision=3, suppress=True)\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.layers.experimental import preprocessing\n",
    "from tensorflow.keras import regularizers\n",
    "#from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B87Rd1SOUv02"
   },
   "source": [
    "## Loading Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 226
    },
    "id": "GS-dBMpuYMnz",
    "outputId": "9d60321b-6130-4f5d-8216-a6a4fee19ca0"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>SoldPr</th>\n",
       "      <th>Type</th>\n",
       "      <th>Zip</th>\n",
       "      <th>Area</th>\n",
       "      <th>Rooms</th>\n",
       "      <th>FullBaths</th>\n",
       "      <th>HalfBaths</th>\n",
       "      <th>BsmtBth</th>\n",
       "      <th>Beds</th>\n",
       "      <th>BsmtBeds</th>\n",
       "      <th>GarageSpaces</th>\n",
       "      <th>houseEra</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3500</td>\n",
       "      <td>294718.041797</td>\n",
       "      <td>SFH</td>\n",
       "      <td>60077</td>\n",
       "      <td>1100</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>19A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6859</td>\n",
       "      <td>330946.875000</td>\n",
       "      <td>SFH</td>\n",
       "      <td>60025</td>\n",
       "      <td>1600</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>No</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>19B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6623</td>\n",
       "      <td>415125.000000</td>\n",
       "      <td>SFH</td>\n",
       "      <td>60025</td>\n",
       "      <td>1500</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>19A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8672</td>\n",
       "      <td>336662.931641</td>\n",
       "      <td>SFH</td>\n",
       "      <td>60076</td>\n",
       "      <td>1422</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.5</td>\n",
       "      <td>19A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4572</td>\n",
       "      <td>571775.077344</td>\n",
       "      <td>SFH</td>\n",
       "      <td>60062</td>\n",
       "      <td>2668</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>No</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>19A</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0         SoldPr Type    Zip  Area  Rooms  FullBaths  HalfBaths  \\\n",
       "0        3500  294718.041797  SFH  60077  1100      7          1          0   \n",
       "1        6859  330946.875000  SFH  60025  1600      6          2          0   \n",
       "2        6623  415125.000000  SFH  60025  1500      7          2          0   \n",
       "3        8672  336662.931641  SFH  60076  1422      9          1          1   \n",
       "4        4572  571775.077344  SFH  60062  2668      8          2          1   \n",
       "\n",
       "  BsmtBth  Beds  BsmtBeds  GarageSpaces houseEra  \n",
       "0      No   3.0         0           2.0      19A  \n",
       "1      No   3.0         0           2.0      19B  \n",
       "2     Yes   3.0         0           2.0      19A  \n",
       "3      No   3.0         0           2.5      19A  \n",
       "4      No   4.0         0           2.0      19A  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('cleaned_dataset.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Deleting First Column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(\"Unnamed: 0\" , inplace = True , axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SoldPr          float64\n",
       "Type             object\n",
       "Zip               int64\n",
       "Area              int64\n",
       "Rooms             int64\n",
       "FullBaths         int64\n",
       "HalfBaths         int64\n",
       "BsmtBth          object\n",
       "Beds            float64\n",
       "BsmtBeds          int64\n",
       "GarageSpaces    float64\n",
       "houseEra         object\n",
       "dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "nngEWwnpx6eJ",
    "outputId": "a0506047-f179-4b92-c184-db0ab2086dd3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SoldPr          float64\n",
       "Type             object\n",
       "Zip              object\n",
       "Area              int64\n",
       "Rooms             int64\n",
       "FullBaths         int64\n",
       "HalfBaths         int64\n",
       "BsmtBth          object\n",
       "Beds            float64\n",
       "BsmtBeds          int64\n",
       "GarageSpaces    float64\n",
       "houseEra         object\n",
       "dtype: object"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Zip'] = df['Zip'].astype(str)\n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Checking The Dataset :"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 9358 entries, 0 to 9357\n",
      "Data columns (total 12 columns):\n",
      " #   Column        Non-Null Count  Dtype  \n",
      "---  ------        --------------  -----  \n",
      " 0   SoldPr        9358 non-null   float64\n",
      " 1   Type          9358 non-null   object \n",
      " 2   Zip           9358 non-null   object \n",
      " 3   Area          9358 non-null   int64  \n",
      " 4   Rooms         9358 non-null   int64  \n",
      " 5   FullBaths     9358 non-null   int64  \n",
      " 6   HalfBaths     9358 non-null   int64  \n",
      " 7   BsmtBth       9358 non-null   object \n",
      " 8   Beds          9358 non-null   float64\n",
      " 9   BsmtBeds      9358 non-null   int64  \n",
      " 10  GarageSpaces  9358 non-null   float64\n",
      " 11  houseEra      9358 non-null   object \n",
      "dtypes: float64(3), int64(5), object(4)\n",
      "memory usage: 877.4+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yPhMIKlWmWo_",
    "outputId": "eaa24a1e-64e7-4b04-ff91-87e6e34bd1b8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5988 train examples\n",
      "1498 validation examples\n",
      "1872 test examples\n"
     ]
    }
   ],
   "source": [
    "train, test = train_test_split(df, test_size=0.2)\n",
    "train, val = train_test_split(train, test_size=0.2)\n",
    "print(len(train), 'train examples')\n",
    "print(len(val), 'validation examples')\n",
    "print(len(test), 'test examples')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "44RzokpwQbAJ"
   },
   "source": [
    "Next, we will separate the training and validation dataframes into features and targets and do a rough normalisation of price."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "D8rCGIK1ZzKx"
   },
   "outputs": [],
   "source": [
    "housePrices_features = train.copy()\n",
    "housePrices_labels = housePrices_features.pop('SoldPr')\n",
    "housePrices_labels = housePrices_labels/100000\n",
    "\n",
    "val_features = val.copy()\n",
    "val_labels = val.pop('SoldPr')\n",
    "val_labels = val_labels/100000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3VshL1mDKSf6",
    "outputId": "7c65fe1a-a165-40b9-c17d-658cf22407b2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Type             object\n",
      "Zip              object\n",
      "Area              int64\n",
      "Rooms             int64\n",
      "FullBaths         int64\n",
      "HalfBaths         int64\n",
      "BsmtBth          object\n",
      "Beds            float64\n",
      "BsmtBeds          int64\n",
      "GarageSpaces    float64\n",
      "houseEra         object\n",
      "dtype: object\n",
      "float64\n"
     ]
    }
   ],
   "source": [
    "print(housePrices_features.dtypes)\n",
    "print(housePrices_labels.dtypes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5WODe_1da3yw",
    "outputId": "438c856c-3d61-4564-c8d5-c4837ce328f2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Type': <KerasTensor: shape=(None, 1) dtype=string (created by layer 'Type')>,\n",
       " 'Zip': <KerasTensor: shape=(None, 1) dtype=string (created by layer 'Zip')>,\n",
       " 'Area': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'Area')>,\n",
       " 'Rooms': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'Rooms')>,\n",
       " 'FullBaths': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'FullBaths')>,\n",
       " 'HalfBaths': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'HalfBaths')>,\n",
       " 'BsmtBth': <KerasTensor: shape=(None, 1) dtype=string (created by layer 'BsmtBth')>,\n",
       " 'Beds': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'Beds')>,\n",
       " 'BsmtBeds': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'BsmtBeds')>,\n",
       " 'GarageSpaces': <KerasTensor: shape=(None, 1) dtype=float32 (created by layer 'GarageSpaces')>,\n",
       " 'houseEra': <KerasTensor: shape=(None, 1) dtype=string (created by layer 'houseEra')>}"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inputs = {}\n",
    "\n",
    "for name, column in housePrices_features.items():\n",
    "  dtype = column.dtype\n",
    "  if dtype == object:\n",
    "    dtype = tf.string\n",
    "  else:\n",
    "    dtype = tf.float32\n",
    "\n",
    "  inputs[name] = tf.keras.Input(shape=(1,), name=name, dtype=dtype)\n",
    "\n",
    "inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "wPRC_E6rkp8D",
    "outputId": "d07d5c2f-f181-49d7-fe67-7dd3055a0d90"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<KerasTensor: shape=(None, 7) dtype=float32 (created by layer 'normalization_1')>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_inputs = {name:input for name,input in inputs.items()\n",
    "                  if input.dtype==tf.float32}\n",
    "\n",
    "x = layers.Concatenate()(list(numeric_inputs.values()))\n",
    "norm = preprocessing.Normalization()\n",
    "norm.adapt(np.array(df[numeric_inputs.keys()]))\n",
    "all_numeric_inputs = norm(x)\n",
    "\n",
    "all_numeric_inputs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-JoR45Uj712l"
   },
   "source": [
    "Collect all the symbolic preprocessing results, to concatenate them later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "M7jIJw5XntdN"
   },
   "outputs": [],
   "source": [
    "preprocessed_inputs = [all_numeric_inputs]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "r0Hryylyosfm"
   },
   "source": [
    "String fields will first use a StringLookup function to map the strings to their corresponding index in a vocabulary based on the different data strings. Second, using the index and default settings, the Tensorflow's pre-processing category encoding will generate a one-hot vector.\n",
    "\n",
    "Together with the numeric field handling, we then transform the input data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "79fi1Cgan2YV"
   },
   "outputs": [],
   "source": [
    "for name, input in inputs.items():\n",
    "  if input.dtype == tf.float32:\n",
    "    continue\n",
    "\n",
    "  lookup = preprocessing.StringLookup(vocabulary=np.unique(housePrices_features[name]))\n",
    "  one_hot = preprocessing.CategoryEncoding(num_tokens =lookup.vocabulary_size())\n",
    "\n",
    "  x = lookup(input)\n",
    "  x = one_hot(x)\n",
    "  preprocessed_inputs.append(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Wnhv0T7itnc7"
   },
   "source": [
    "Given the preprocessed inputs (for normalisation and categorisation) and the corresponding input fields we can build a model the preprocesses the input data. We can visualise the model, as shown below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 525
    },
    "id": "XJRzUTe8ukXc",
    "outputId": "7c43d279-442c-40cf-c4d4-324f70befb8c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n"
     ]
    }
   ],
   "source": [
    "preprocessed_inputs_cat = layers.Concatenate()(preprocessed_inputs)\n",
    "\n",
    "housePrices_preprocessing = tf.keras.Model(inputs, preprocessed_inputs_cat)\n",
    "\n",
    "tf.keras.utils.plot_model(model = housePrices_preprocessing , rankdir=\"LR\", dpi=72, show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "id": "5YjdYyMEacwQ"
   },
   "outputs": [],
   "source": [
    "housePrices_features_dict = {name: np.array(value)\n",
    "                         for name, value in housePrices_features.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "SjnmU8PSv8T3",
    "outputId": "4bf7b766-0371-48fc-872c-84bd2ae5a597"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: shape=(1, 69), dtype=float32, numpy=\n",
       "array([[-0.443, -0.399, -1.248,  0.767, -0.226, -0.374,  0.243,  0.   ,\n",
       "         0.   ,  0.   ,  1.   ,  0.   ,  0.   ,  1.   ,  0.   ,  0.   ,\n",
       "         0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,\n",
       "         0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  1.   ,  0.   ,  0.   ,\n",
       "         1.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,\n",
       "         0.   ,  1.   ,  0.   ,  0.   ,  1.   ,  0.   ,  0.   ,  0.   ,\n",
       "         0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,\n",
       "         0.   ,  0.   ,  0.   ,  0.   ,  1.   ,  0.   ,  0.   ,  1.   ,\n",
       "         0.   ,  0.   ,  0.   ,  0.   ,  0.   ]], dtype=float32)>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_dict = {name:values[:1] for name, values in housePrices_features_dict.items()}\n",
    "housePrices_preprocessing(features_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "coIPtGaCzUV7"
   },
   "outputs": [],
   "source": [
    "def housePrices_model(preprocessing_head, inputs):\n",
    "  body = tf.keras.Sequential([\n",
    "    layers.Dense(128,activation='relu', kernel_regularizer=regularizers.l2(0.001)),\n",
    "    layers.Dense(64,activation='relu'),\n",
    "    #layers.Dense(64,activation='relu'),\n",
    "    layers.Dense(1)\n",
    "  ])\n",
    "\n",
    "  preprocessed_inputs = preprocessing_head(inputs)\n",
    "  result = body(preprocessed_inputs)\n",
    "  model = tf.keras.Model(inputs, result)\n",
    "\n",
    "  model.compile(loss='mse', optimizer='adam', metrics=['mae'])\n",
    "  return model\n",
    "\n",
    "housePrices_model = housePrices_model(housePrices_preprocessing, inputs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LK5uBQQF2KbZ"
   },
   "source": [
    "When you train the model, pass the dictionary of features as 'x', and the label as 'y'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "VpMHKYTqomAL"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/250\n",
      "188/188 [==============================] - 3s 13ms/step - loss: 5.0737 - mae: 1.3223 - val_loss: 2.0653 - val_mae: 0.8644\n",
      "Epoch 2/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 2.1224 - mae: 0.8547 - val_loss: 1.7354 - val_mae: 0.7682\n",
      "Epoch 3/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.9978 - mae: 0.8361 - val_loss: 1.7121 - val_mae: 0.7688\n",
      "Epoch 4/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.8955 - mae: 0.8102 - val_loss: 1.6625 - val_mae: 0.7603\n",
      "Epoch 5/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.8635 - mae: 0.8073 - val_loss: 1.6242 - val_mae: 0.7424\n",
      "Epoch 6/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.7945 - mae: 0.7850 - val_loss: 1.7151 - val_mae: 0.7665\n",
      "Epoch 7/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.7828 - mae: 0.7932 - val_loss: 1.7042 - val_mae: 0.7638\n",
      "Epoch 8/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.7065 - mae: 0.7695 - val_loss: 1.6550 - val_mae: 0.7704\n",
      "Epoch 9/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.6512 - mae: 0.7608 - val_loss: 1.5885 - val_mae: 0.7391\n",
      "Epoch 10/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.6272 - mae: 0.7525 - val_loss: 1.7512 - val_mae: 0.7751\n",
      "Epoch 11/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.6470 - mae: 0.7717 - val_loss: 1.6055 - val_mae: 0.7570\n",
      "Epoch 12/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5637 - mae: 0.7497 - val_loss: 1.6772 - val_mae: 0.7706\n",
      "Epoch 13/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5979 - mae: 0.7533 - val_loss: 2.0433 - val_mae: 0.8755\n",
      "Epoch 14/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5699 - mae: 0.7555 - val_loss: 1.7139 - val_mae: 0.7671\n",
      "Epoch 15/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5618 - mae: 0.7581 - val_loss: 1.5756 - val_mae: 0.7377\n",
      "Epoch 16/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5189 - mae: 0.7379 - val_loss: 1.6568 - val_mae: 0.7488\n",
      "Epoch 17/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.4888 - mae: 0.7449 - val_loss: 1.8047 - val_mae: 0.8441\n",
      "Epoch 18/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5338 - mae: 0.7568 - val_loss: 1.5655 - val_mae: 0.7354\n",
      "Epoch 19/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.4482 - mae: 0.7266 - val_loss: 1.7041 - val_mae: 0.8053\n",
      "Epoch 20/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.4138 - mae: 0.7168 - val_loss: 1.7818 - val_mae: 0.8247\n",
      "Epoch 21/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.4138 - mae: 0.7251 - val_loss: 1.8161 - val_mae: 0.8189\n",
      "Epoch 22/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.5309 - mae: 0.7642 - val_loss: 1.7390 - val_mae: 0.7755\n",
      "Epoch 23/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.3961 - mae: 0.7220 - val_loss: 1.6779 - val_mae: 0.7654\n",
      "Epoch 24/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.3845 - mae: 0.7190 - val_loss: 1.6486 - val_mae: 0.7569\n",
      "Epoch 25/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.3200 - mae: 0.7021 - val_loss: 1.6562 - val_mae: 0.7816\n",
      "Epoch 26/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.3404 - mae: 0.7119 - val_loss: 1.6772 - val_mae: 0.7587\n",
      "Epoch 27/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2943 - mae: 0.6987 - val_loss: 1.9114 - val_mae: 0.8320\n",
      "Epoch 28/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.3192 - mae: 0.7132 - val_loss: 1.7617 - val_mae: 0.7839\n",
      "Epoch 29/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2770 - mae: 0.6986 - val_loss: 1.6085 - val_mae: 0.7493\n",
      "Epoch 30/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2339 - mae: 0.6853 - val_loss: 1.6966 - val_mae: 0.7857\n",
      "Epoch 31/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2592 - mae: 0.6942 - val_loss: 2.0076 - val_mae: 0.8592\n",
      "Epoch 32/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2460 - mae: 0.6995 - val_loss: 2.0704 - val_mae: 0.8221\n",
      "Epoch 33/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2212 - mae: 0.6918 - val_loss: 1.7252 - val_mae: 0.7652\n",
      "Epoch 34/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.1722 - mae: 0.6768 - val_loss: 2.1097 - val_mae: 0.8427\n",
      "Epoch 35/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 1.2580 - mae: 0.7010 - val_loss: 1.6929 - val_mae: 0.7719\n",
      "Epoch 36/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 1.1918 - mae: 0.6849 - val_loss: 1.9526 - val_mae: 0.8298\n",
      "Epoch 37/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.2583 - mae: 0.6969 - val_loss: 1.9191 - val_mae: 0.8013\n",
      "Epoch 38/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.1241 - mae: 0.6656 - val_loss: 1.7922 - val_mae: 0.7642\n",
      "Epoch 39/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 1.1202 - mae: 0.6719 - val_loss: 1.8439 - val_mae: 0.8026\n",
      "Epoch 40/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.1449 - mae: 0.6733 - val_loss: 1.7553 - val_mae: 0.7564\n",
      "Epoch 41/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0781 - mae: 0.6569 - val_loss: 1.7954 - val_mae: 0.7633\n",
      "Epoch 42/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0976 - mae: 0.6689 - val_loss: 1.7444 - val_mae: 0.7768\n",
      "Epoch 43/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0533 - mae: 0.6504 - val_loss: 1.9779 - val_mae: 0.8021\n",
      "Epoch 44/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0797 - mae: 0.6610 - val_loss: 1.7290 - val_mae: 0.7703\n",
      "Epoch 45/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.1068 - mae: 0.6736 - val_loss: 1.8560 - val_mae: 0.7751\n",
      "Epoch 46/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0550 - mae: 0.6568 - val_loss: 1.7221 - val_mae: 0.7595\n",
      "Epoch 47/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0179 - mae: 0.6417 - val_loss: 1.7100 - val_mae: 0.7736\n",
      "Epoch 48/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0357 - mae: 0.6547 - val_loss: 1.7696 - val_mae: 0.7760\n",
      "Epoch 49/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0062 - mae: 0.6387 - val_loss: 1.8498 - val_mae: 0.7847\n",
      "Epoch 50/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 1.0066 - mae: 0.6424 - val_loss: 1.8728 - val_mae: 0.7778\n",
      "Epoch 51/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9684 - mae: 0.6308 - val_loss: 1.8010 - val_mae: 0.7971\n",
      "Epoch 52/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9576 - mae: 0.6230 - val_loss: 1.7926 - val_mae: 0.7722\n",
      "Epoch 53/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9862 - mae: 0.6356 - val_loss: 2.0020 - val_mae: 0.7994\n",
      "Epoch 54/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9896 - mae: 0.6418 - val_loss: 1.8593 - val_mae: 0.8213\n",
      "Epoch 55/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9228 - mae: 0.6196 - val_loss: 1.7410 - val_mae: 0.7638\n",
      "Epoch 56/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9358 - mae: 0.6268 - val_loss: 1.9921 - val_mae: 0.8025\n",
      "Epoch 57/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9351 - mae: 0.6276 - val_loss: 1.8078 - val_mae: 0.7935\n",
      "Epoch 58/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9150 - mae: 0.6175 - val_loss: 1.7669 - val_mae: 0.7645\n",
      "Epoch 59/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9347 - mae: 0.6208 - val_loss: 1.8608 - val_mae: 0.8066\n",
      "Epoch 60/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8928 - mae: 0.6117 - val_loss: 1.9553 - val_mae: 0.8013\n",
      "Epoch 61/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8762 - mae: 0.6036 - val_loss: 1.7877 - val_mae: 0.7743\n",
      "Epoch 62/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8758 - mae: 0.6059 - val_loss: 1.7684 - val_mae: 0.7674\n",
      "Epoch 63/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8864 - mae: 0.6034 - val_loss: 2.0399 - val_mae: 0.8145\n",
      "Epoch 64/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8784 - mae: 0.6018 - val_loss: 1.7748 - val_mae: 0.7731\n",
      "Epoch 65/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.9183 - mae: 0.6164 - val_loss: 1.7129 - val_mae: 0.7721\n",
      "Epoch 66/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8738 - mae: 0.6091 - val_loss: 1.7923 - val_mae: 0.7857\n",
      "Epoch 67/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8543 - mae: 0.6046 - val_loss: 1.8522 - val_mae: 0.7719\n",
      "Epoch 68/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8394 - mae: 0.5978 - val_loss: 1.8295 - val_mae: 0.7842\n",
      "Epoch 69/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8617 - mae: 0.6091 - val_loss: 1.8862 - val_mae: 0.8274\n",
      "Epoch 70/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8348 - mae: 0.5927 - val_loss: 1.8183 - val_mae: 0.7860\n",
      "Epoch 71/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8249 - mae: 0.5884 - val_loss: 1.8507 - val_mae: 0.7901\n",
      "Epoch 72/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8093 - mae: 0.5846 - val_loss: 1.9261 - val_mae: 0.7911\n",
      "Epoch 73/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8237 - mae: 0.5909 - val_loss: 1.9702 - val_mae: 0.8183\n",
      "Epoch 74/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8340 - mae: 0.5883 - val_loss: 1.7894 - val_mae: 0.7839\n",
      "Epoch 75/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8165 - mae: 0.5866 - val_loss: 1.8654 - val_mae: 0.7817\n",
      "Epoch 76/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7959 - mae: 0.5778 - val_loss: 1.9495 - val_mae: 0.8043\n",
      "Epoch 77/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7680 - mae: 0.5717 - val_loss: 1.7927 - val_mae: 0.7922\n",
      "Epoch 78/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8614 - mae: 0.5971 - val_loss: 1.9644 - val_mae: 0.8006\n",
      "Epoch 79/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7665 - mae: 0.5661 - val_loss: 1.7694 - val_mae: 0.7760\n",
      "Epoch 80/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7425 - mae: 0.5605 - val_loss: 1.7972 - val_mae: 0.7881\n",
      "Epoch 81/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7958 - mae: 0.5852 - val_loss: 2.0626 - val_mae: 0.8141\n",
      "Epoch 82/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7660 - mae: 0.5666 - val_loss: 2.0056 - val_mae: 0.8288\n",
      "Epoch 83/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8166 - mae: 0.5868 - val_loss: 2.2891 - val_mae: 0.8659\n",
      "Epoch 84/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8063 - mae: 0.5834 - val_loss: 1.9085 - val_mae: 0.8026\n",
      "Epoch 85/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7559 - mae: 0.5648 - val_loss: 1.9040 - val_mae: 0.7879\n",
      "Epoch 86/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7247 - mae: 0.5561 - val_loss: 2.0426 - val_mae: 0.7930\n",
      "Epoch 87/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7401 - mae: 0.5577 - val_loss: 2.0742 - val_mae: 0.8015\n",
      "Epoch 88/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7422 - mae: 0.5640 - val_loss: 1.9285 - val_mae: 0.7884\n",
      "Epoch 89/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7354 - mae: 0.5571 - val_loss: 1.9150 - val_mae: 0.7929\n",
      "Epoch 90/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7146 - mae: 0.5503 - val_loss: 1.9433 - val_mae: 0.8046\n",
      "Epoch 91/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7024 - mae: 0.5457 - val_loss: 1.9622 - val_mae: 0.8058\n",
      "Epoch 92/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7342 - mae: 0.5602 - val_loss: 2.0378 - val_mae: 0.8243\n",
      "Epoch 93/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7696 - mae: 0.5689 - val_loss: 1.9317 - val_mae: 0.7999\n",
      "Epoch 94/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7065 - mae: 0.5464 - val_loss: 1.9495 - val_mae: 0.8139\n",
      "Epoch 95/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7151 - mae: 0.5489 - val_loss: 1.8976 - val_mae: 0.7983\n",
      "Epoch 96/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6840 - mae: 0.5372 - val_loss: 1.9951 - val_mae: 0.8136\n",
      "Epoch 97/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6768 - mae: 0.5369 - val_loss: 2.1081 - val_mae: 0.8063\n",
      "Epoch 98/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7387 - mae: 0.5617 - val_loss: 2.0861 - val_mae: 0.8490\n",
      "Epoch 99/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7949 - mae: 0.5735 - val_loss: 2.2176 - val_mae: 0.8388\n",
      "Epoch 100/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7110 - mae: 0.5477 - val_loss: 2.0615 - val_mae: 0.8013\n",
      "Epoch 101/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6903 - mae: 0.5380 - val_loss: 1.9909 - val_mae: 0.8134\n",
      "Epoch 102/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6978 - mae: 0.5464 - val_loss: 1.9951 - val_mae: 0.8299\n",
      "Epoch 103/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6730 - mae: 0.5403 - val_loss: 1.9780 - val_mae: 0.8163\n",
      "Epoch 104/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6793 - mae: 0.5363 - val_loss: 1.9369 - val_mae: 0.7995\n",
      "Epoch 105/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6914 - mae: 0.5370 - val_loss: 2.0212 - val_mae: 0.8251\n",
      "Epoch 106/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6869 - mae: 0.5355 - val_loss: 2.0190 - val_mae: 0.8089\n",
      "Epoch 107/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7330 - mae: 0.5526 - val_loss: 2.2331 - val_mae: 0.8588\n",
      "Epoch 108/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6456 - mae: 0.5224 - val_loss: 2.0331 - val_mae: 0.8519\n",
      "Epoch 109/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6507 - mae: 0.5265 - val_loss: 2.0113 - val_mae: 0.8145\n",
      "Epoch 110/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6644 - mae: 0.5327 - val_loss: 2.0420 - val_mae: 0.8429\n",
      "Epoch 111/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7047 - mae: 0.5458 - val_loss: 2.0506 - val_mae: 0.8167\n",
      "Epoch 112/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6549 - mae: 0.5249 - val_loss: 2.0318 - val_mae: 0.8125\n",
      "Epoch 113/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6490 - mae: 0.5279 - val_loss: 1.9866 - val_mae: 0.8036\n",
      "Epoch 114/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6553 - mae: 0.5272 - val_loss: 2.0362 - val_mae: 0.8169\n",
      "Epoch 115/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6428 - mae: 0.5241 - val_loss: 2.1309 - val_mae: 0.8250\n",
      "Epoch 116/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6655 - mae: 0.5338 - val_loss: 1.9577 - val_mae: 0.8153\n",
      "Epoch 117/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6551 - mae: 0.5304 - val_loss: 2.0584 - val_mae: 0.8267\n",
      "Epoch 118/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6285 - mae: 0.5173 - val_loss: 2.0636 - val_mae: 0.8268\n",
      "Epoch 119/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7633 - mae: 0.5625 - val_loss: 2.0809 - val_mae: 0.8186\n",
      "Epoch 120/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6440 - mae: 0.5255 - val_loss: 2.1179 - val_mae: 0.8154\n",
      "Epoch 121/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6243 - mae: 0.5136 - val_loss: 1.9740 - val_mae: 0.8114\n",
      "Epoch 122/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6426 - mae: 0.5201 - val_loss: 2.0587 - val_mae: 0.8218\n",
      "Epoch 123/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6150 - mae: 0.5129 - val_loss: 2.0082 - val_mae: 0.8169\n",
      "Epoch 124/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6602 - mae: 0.5303 - val_loss: 2.1084 - val_mae: 0.8155\n",
      "Epoch 125/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6747 - mae: 0.5357 - val_loss: 2.0620 - val_mae: 0.8215\n",
      "Epoch 126/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6300 - mae: 0.5128 - val_loss: 2.0202 - val_mae: 0.8208\n",
      "Epoch 127/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6132 - mae: 0.5083 - val_loss: 2.0852 - val_mae: 0.8177\n",
      "Epoch 128/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5996 - mae: 0.5005 - val_loss: 2.0888 - val_mae: 0.8459\n",
      "Epoch 129/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6107 - mae: 0.5126 - val_loss: 2.0390 - val_mae: 0.8136\n",
      "Epoch 130/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5903 - mae: 0.4965 - val_loss: 2.1297 - val_mae: 0.8375\n",
      "Epoch 131/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.7075 - mae: 0.5423 - val_loss: 2.0740 - val_mae: 0.8359\n",
      "Epoch 132/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6377 - mae: 0.5203 - val_loss: 2.0271 - val_mae: 0.8200\n",
      "Epoch 133/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6297 - mae: 0.5210 - val_loss: 1.9106 - val_mae: 0.8103\n",
      "Epoch 134/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5972 - mae: 0.4989 - val_loss: 2.1830 - val_mae: 0.8443\n",
      "Epoch 135/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5975 - mae: 0.5031 - val_loss: 2.0959 - val_mae: 0.8436\n",
      "Epoch 136/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6113 - mae: 0.5059 - val_loss: 2.0535 - val_mae: 0.8391\n",
      "Epoch 137/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.6106 - mae: 0.5061 - val_loss: 2.1361 - val_mae: 0.8353\n",
      "Epoch 138/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.6582 - mae: 0.5247 - val_loss: 2.0961 - val_mae: 0.8270\n",
      "Epoch 139/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6007 - mae: 0.5025 - val_loss: 2.1092 - val_mae: 0.8374\n",
      "Epoch 140/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5903 - mae: 0.4982 - val_loss: 2.1952 - val_mae: 0.8464\n",
      "Epoch 141/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6038 - mae: 0.5016 - val_loss: 2.0714 - val_mae: 0.8284\n",
      "Epoch 142/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6314 - mae: 0.5145 - val_loss: 2.0972 - val_mae: 0.8267\n",
      "Epoch 143/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5942 - mae: 0.5002 - val_loss: 2.0841 - val_mae: 0.8283\n",
      "Epoch 144/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6035 - mae: 0.5079 - val_loss: 2.1218 - val_mae: 0.8345\n",
      "Epoch 145/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5706 - mae: 0.4916 - val_loss: 2.1207 - val_mae: 0.8292\n",
      "Epoch 146/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.5660 - mae: 0.4849 - val_loss: 2.0923 - val_mae: 0.8271\n",
      "Epoch 147/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5984 - mae: 0.4993 - val_loss: 2.0677 - val_mae: 0.8298\n",
      "Epoch 148/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5892 - mae: 0.4965 - val_loss: 2.2777 - val_mae: 0.8584\n",
      "Epoch 149/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6483 - mae: 0.5188 - val_loss: 2.0365 - val_mae: 0.8161\n",
      "Epoch 150/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6073 - mae: 0.5053 - val_loss: 2.1672 - val_mae: 0.8393\n",
      "Epoch 151/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5782 - mae: 0.4943 - val_loss: 2.0135 - val_mae: 0.8252\n",
      "Epoch 152/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5710 - mae: 0.4921 - val_loss: 2.2769 - val_mae: 0.8448\n",
      "Epoch 153/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5596 - mae: 0.4860 - val_loss: 2.1687 - val_mae: 0.8423\n",
      "Epoch 154/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.5680 - mae: 0.4918 - val_loss: 2.0944 - val_mae: 0.8445\n",
      "Epoch 155/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6149 - mae: 0.5048 - val_loss: 2.1927 - val_mae: 0.8414\n",
      "Epoch 156/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5789 - mae: 0.4928 - val_loss: 2.0896 - val_mae: 0.8245\n",
      "Epoch 157/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5777 - mae: 0.4896 - val_loss: 2.2863 - val_mae: 0.8827\n",
      "Epoch 158/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5965 - mae: 0.5028 - val_loss: 2.0661 - val_mae: 0.8293\n",
      "Epoch 159/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5744 - mae: 0.4877 - val_loss: 2.3304 - val_mae: 0.8652\n",
      "Epoch 160/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5678 - mae: 0.4837 - val_loss: 2.0928 - val_mae: 0.8381\n",
      "Epoch 161/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5721 - mae: 0.4869 - val_loss: 2.2627 - val_mae: 0.8654\n",
      "Epoch 162/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5574 - mae: 0.4798 - val_loss: 2.1745 - val_mae: 0.8395\n",
      "Epoch 163/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.5548 - mae: 0.4789 - val_loss: 2.1148 - val_mae: 0.8389\n",
      "Epoch 164/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5419 - mae: 0.4777 - val_loss: 2.3348 - val_mae: 0.8615\n",
      "Epoch 165/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5906 - mae: 0.4967 - val_loss: 2.1472 - val_mae: 0.8414\n",
      "Epoch 166/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5542 - mae: 0.4815 - val_loss: 2.1796 - val_mae: 0.8312\n",
      "Epoch 167/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5588 - mae: 0.4845 - val_loss: 2.2294 - val_mae: 0.8554\n",
      "Epoch 168/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5452 - mae: 0.4781 - val_loss: 2.2550 - val_mae: 0.8450\n",
      "Epoch 169/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.5397 - mae: 0.4761 - val_loss: 2.1898 - val_mae: 0.8392\n",
      "Epoch 170/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5450 - mae: 0.4762 - val_loss: 2.2698 - val_mae: 0.8550\n",
      "Epoch 171/250\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5494 - mae: 0.4805 - val_loss: 2.1584 - val_mae: 0.8406\n",
      "Epoch 172/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6120 - mae: 0.5010 - val_loss: 2.1874 - val_mae: 0.8411\n",
      "Epoch 173/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5453 - mae: 0.4812 - val_loss: 2.2911 - val_mae: 0.8563\n",
      "Epoch 174/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5497 - mae: 0.4774 - val_loss: 2.3337 - val_mae: 0.8514\n",
      "Epoch 175/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5527 - mae: 0.4827 - val_loss: 2.2481 - val_mae: 0.8613\n",
      "Epoch 176/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5422 - mae: 0.4738 - val_loss: 2.3456 - val_mae: 0.8506\n",
      "Epoch 177/250\n",
      "188/188 [==============================] - 2s 12ms/step - loss: 0.7353 - mae: 0.5493 - val_loss: 2.2179 - val_mae: 0.8579\n",
      "Epoch 178/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5836 - mae: 0.4921 - val_loss: 2.1770 - val_mae: 0.8533\n",
      "Epoch 179/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5347 - mae: 0.4694 - val_loss: 2.3026 - val_mae: 0.8456\n",
      "Epoch 180/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5314 - mae: 0.4684 - val_loss: 2.1112 - val_mae: 0.8552\n",
      "Epoch 181/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5667 - mae: 0.4831 - val_loss: 2.1851 - val_mae: 0.8408\n",
      "Epoch 182/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5437 - mae: 0.4718 - val_loss: 2.1183 - val_mae: 0.8411\n",
      "Epoch 183/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5435 - mae: 0.4748 - val_loss: 2.2569 - val_mae: 0.8492\n",
      "Epoch 184/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5207 - mae: 0.4664 - val_loss: 2.1521 - val_mae: 0.8381\n",
      "Epoch 185/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5161 - mae: 0.4625 - val_loss: 2.2095 - val_mae: 0.8605\n",
      "Epoch 186/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5385 - mae: 0.4709 - val_loss: 2.1779 - val_mae: 0.8587\n",
      "Epoch 187/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5231 - mae: 0.4671 - val_loss: 2.0693 - val_mae: 0.8379\n",
      "Epoch 188/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5203 - mae: 0.4672 - val_loss: 2.2714 - val_mae: 0.8513\n",
      "Epoch 189/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5349 - mae: 0.4699 - val_loss: 2.3898 - val_mae: 0.9036\n",
      "Epoch 190/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5390 - mae: 0.4703 - val_loss: 2.0721 - val_mae: 0.8445\n",
      "Epoch 191/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5197 - mae: 0.4669 - val_loss: 2.3286 - val_mae: 0.8530\n",
      "Epoch 192/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5394 - mae: 0.4698 - val_loss: 2.2039 - val_mae: 0.8516\n",
      "Epoch 193/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5499 - mae: 0.4753 - val_loss: 2.3169 - val_mae: 0.8827\n",
      "Epoch 194/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5445 - mae: 0.4777 - val_loss: 2.3298 - val_mae: 0.8645\n",
      "Epoch 195/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5067 - mae: 0.4576 - val_loss: 2.1908 - val_mae: 0.8503\n",
      "Epoch 196/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5331 - mae: 0.4716 - val_loss: 2.2578 - val_mae: 0.8597\n",
      "Epoch 197/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5081 - mae: 0.4553 - val_loss: 2.2421 - val_mae: 0.8554\n",
      "Epoch 198/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5570 - mae: 0.4795 - val_loss: 2.2245 - val_mae: 0.8516\n",
      "Epoch 199/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5241 - mae: 0.4657 - val_loss: 2.1865 - val_mae: 0.8417\n",
      "Epoch 200/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5183 - mae: 0.4666 - val_loss: 2.3794 - val_mae: 0.8696\n",
      "Epoch 201/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5011 - mae: 0.4537 - val_loss: 2.3442 - val_mae: 0.8656\n",
      "Epoch 202/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.6109 - mae: 0.5021 - val_loss: 2.3175 - val_mae: 0.8467\n",
      "Epoch 203/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5254 - mae: 0.4678 - val_loss: 2.2227 - val_mae: 0.8729\n",
      "Epoch 204/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5135 - mae: 0.4608 - val_loss: 2.1743 - val_mae: 0.8442\n",
      "Epoch 205/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4921 - mae: 0.4512 - val_loss: 2.2785 - val_mae: 0.8546\n",
      "Epoch 206/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5219 - mae: 0.4648 - val_loss: 2.1163 - val_mae: 0.8376\n",
      "Epoch 207/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5204 - mae: 0.4632 - val_loss: 2.3447 - val_mae: 0.8849\n",
      "Epoch 208/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5595 - mae: 0.4822 - val_loss: 2.1575 - val_mae: 0.8513\n",
      "Epoch 209/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5983 - mae: 0.4965 - val_loss: 2.2211 - val_mae: 0.8443\n",
      "Epoch 210/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5375 - mae: 0.4709 - val_loss: 2.0543 - val_mae: 0.8319\n",
      "Epoch 211/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5095 - mae: 0.4564 - val_loss: 2.2413 - val_mae: 0.8569\n",
      "Epoch 212/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4902 - mae: 0.4460 - val_loss: 2.3378 - val_mae: 0.8854\n",
      "Epoch 213/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5314 - mae: 0.4697 - val_loss: 2.2224 - val_mae: 0.8709\n",
      "Epoch 214/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5031 - mae: 0.4538 - val_loss: 2.1384 - val_mae: 0.8471\n",
      "Epoch 215/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4911 - mae: 0.4513 - val_loss: 2.2418 - val_mae: 0.8542\n",
      "Epoch 216/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5052 - mae: 0.4586 - val_loss: 2.3236 - val_mae: 0.8525\n",
      "Epoch 217/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5079 - mae: 0.4571 - val_loss: 2.1985 - val_mae: 0.8581\n",
      "Epoch 218/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5635 - mae: 0.4828 - val_loss: 2.2034 - val_mae: 0.8464\n",
      "Epoch 219/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5262 - mae: 0.4681 - val_loss: 2.2528 - val_mae: 0.8463\n",
      "Epoch 220/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4984 - mae: 0.4544 - val_loss: 2.1106 - val_mae: 0.8453\n",
      "Epoch 221/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4853 - mae: 0.4465 - val_loss: 2.1550 - val_mae: 0.8415\n",
      "Epoch 222/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4933 - mae: 0.4504 - val_loss: 2.3051 - val_mae: 0.8663\n",
      "Epoch 223/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5053 - mae: 0.4559 - val_loss: 2.2170 - val_mae: 0.8638\n",
      "Epoch 224/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5162 - mae: 0.4635 - val_loss: 2.3057 - val_mae: 0.8704\n",
      "Epoch 225/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5054 - mae: 0.4563 - val_loss: 2.4462 - val_mae: 0.8674\n",
      "Epoch 226/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.8715 - mae: 0.5609 - val_loss: 2.2549 - val_mae: 0.8663\n",
      "Epoch 227/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5482 - mae: 0.4703 - val_loss: 2.1448 - val_mae: 0.8468\n",
      "Epoch 228/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4788 - mae: 0.4435 - val_loss: 2.2360 - val_mae: 0.8531\n",
      "Epoch 229/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4623 - mae: 0.4329 - val_loss: 2.1716 - val_mae: 0.8389\n",
      "Epoch 230/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4885 - mae: 0.4487 - val_loss: 2.1457 - val_mae: 0.8395\n",
      "Epoch 231/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4647 - mae: 0.4346 - val_loss: 2.2211 - val_mae: 0.8503\n",
      "Epoch 232/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4801 - mae: 0.4426 - val_loss: 2.1483 - val_mae: 0.8472\n",
      "Epoch 233/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4958 - mae: 0.4496 - val_loss: 2.2308 - val_mae: 0.8576\n",
      "Epoch 234/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4981 - mae: 0.4538 - val_loss: 2.2384 - val_mae: 0.8552\n",
      "Epoch 235/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4958 - mae: 0.4507 - val_loss: 2.2697 - val_mae: 0.8584\n",
      "Epoch 236/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4949 - mae: 0.4527 - val_loss: 2.3275 - val_mae: 0.8673\n",
      "Epoch 237/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4995 - mae: 0.4514 - val_loss: 2.3491 - val_mae: 0.8719\n",
      "Epoch 238/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5223 - mae: 0.4636 - val_loss: 2.1862 - val_mae: 0.8418\n",
      "Epoch 239/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4618 - mae: 0.4357 - val_loss: 2.2231 - val_mae: 0.8611\n",
      "Epoch 240/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4572 - mae: 0.4315 - val_loss: 2.2421 - val_mae: 0.8509\n",
      "Epoch 241/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5109 - mae: 0.4588 - val_loss: 2.2463 - val_mae: 0.8512\n",
      "Epoch 242/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4993 - mae: 0.4525 - val_loss: 2.3066 - val_mae: 0.8687\n",
      "Epoch 243/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4749 - mae: 0.4431 - val_loss: 2.2166 - val_mae: 0.8554\n",
      "Epoch 244/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5005 - mae: 0.4503 - val_loss: 2.1777 - val_mae: 0.8368\n",
      "Epoch 245/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4761 - mae: 0.4433 - val_loss: 2.2146 - val_mae: 0.8492\n",
      "Epoch 246/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4801 - mae: 0.4421 - val_loss: 2.3475 - val_mae: 0.8616\n",
      "Epoch 247/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.4911 - mae: 0.4505 - val_loss: 2.2908 - val_mae: 0.8683\n",
      "Epoch 248/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5135 - mae: 0.4614 - val_loss: 2.3626 - val_mae: 0.8676\n",
      "Epoch 249/250\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5290 - mae: 0.4630 - val_loss: 2.2152 - val_mae: 0.8547\n",
      "Epoch 250/250\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.4697 - mae: 0.4385 - val_loss: 2.4007 - val_mae: 0.8655\n"
     ]
    }
   ],
   "source": [
    "val_features_dict = {name: np.array(value)\n",
    "                         for name, value in val.items()}\n",
    "history_1 = housePrices_model.fit(x=housePrices_features_dict, y=housePrices_labels,epochs=250,\n",
    "                        validation_data=(val_features_dict, val_labels))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PxDJzn1LtK0c"
   },
   "source": [
    "It is crucial to visualise the metrics produced during training; this will quickly indicate how quickly the model converges, potential issues of under and overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 573
    },
    "id": "v52vc1fw49Um",
    "outputId": "4afc78c8-a765-403a-9faa-4b4546ad49c6"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Draw a graph of the loss, which is the distance between\n",
    "# the predicted and actual values during training and validation.\n",
    "import matplotlib.pyplot as plt\n",
    "train_loss = history_1.history['mae']\n",
    "val_loss = history_1.history['mae']\n",
    "epochs = range(1, len(train_loss) + 1)\n",
    "\n",
    "plt.plot(epochs, train_loss, 'g.', label='Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# Exclude the first few epochs so the graph is easier to read\n",
    "SKIP = 20\n",
    "\n",
    "plt.plot(epochs[SKIP:], train_loss[SKIP:], 'g.', label='Training loss')\n",
    "plt.plot(epochs[SKIP:], val_loss[SKIP:], 'b.', label='Validation loss')\n",
    "plt.title('Training and validation loss')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LxgJarZk3bfH"
   },
   "source": [
    "By implementing pre-processing as part of the model, we can now save the model and deploy it elsewhere as a simple package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Saving Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "housePrices_model.save('test.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "id": "Ay-8ymNA2ZCh"
   },
   "outputs": [],
   "source": [
    "reloaded = tf.keras.models.load_model('test.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ojBbrVDFYnnR"
   },
   "outputs": [],
   "source": [
    "! zip -r test.zip test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "id": "Qm6jMTpD20lK"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tf.Tensor([[2.092]], shape=(1, 1), dtype=float32)\n",
      "tf.Tensor([[2.092]], shape=(1, 1), dtype=float32)\n"
     ]
    }
   ],
   "source": [
    "features_dict = {name:values[:1] for name, values in housePrices_features_dict.items()}\n",
    "\n",
    "before = housePrices_model(features_dict)\n",
    "after = reloaded(features_dict)\n",
    "assert (before-after)<1e-3\n",
    "print(before)\n",
    "print(after)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WA-S29Bfu7BF"
   },
   "source": [
    "Finally, we can run predictions on our own data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "pLOZuDVtNIow",
    "outputId": "8540bac1-c30f-4e7f-852c-7880b25dffd7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 268ms/step\n",
      "[[14.554]\n",
      " [ 3.946]\n",
      " [ 3.435]\n",
      " [ 3.391]\n",
      " [ 4.035]]\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "houseData_own2 = {'Type': np.array(['SFH', 'SFH', 'SFH', 'Condo', 'Condo']),\n",
    "  'houseEra': np.array(['recent', '19A', '20A', '20A', '19B']),\n",
    "\t'Area': np.array([8410, 1400, 1500, 1500, 1600]),\n",
    "  'Zip': np.array(['60062', '60062', '60076', '60076', '60202']),\n",
    "\t'Rooms': np.array([16, 6, 7, 7, 7]),\n",
    "  'FullBaths': np.array([6.0, 2.0, 2.0, 2.5, 2.0]),\n",
    "  'HalfBaths': np.array([0.0, 1.0, 1.0, 0.0, 0.0]),\n",
    "  'BsmtBth': np.array(['Yes', 'No', 'No', 'No', 'No']),\n",
    "  'Beds': np.array([5, 3, 3, 3, 3]),\n",
    "  'BsmtBeds': np.array([1.0, 0.0, 0.0, 0.0, 0.0]),\n",
    "  'GarageSpaces': np.array([3, 2, 0, 0, 0])  }\n",
    "\n",
    "ans = reloaded.predict(houseData_own2)\n",
    "\n",
    "print(ans)\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "history_visible": true,
   "name": "housePricePrediction2.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}