ANN_LevyFlights_HGWO_airfoil_dataset.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4cd1c3db",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import modules\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler, QuantileTransformer, MaxAbsScaler\n",
    "from sklearn.metrics import r2_score, mean_absolute_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "71ceeff6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Input, Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3fd2fe26",
   "metadata": {},
   "outputs": [],
   "source": [
    "import Grey_Wolves as GWO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "114db60b",
   "metadata": {},
   "outputs": [],
   "source": [
    "url_file = 'https://archive.ics.uci.edu/ml/machine-learning-databases/00291/airfoil_self_noise.dat'\n",
    "columns = ['frequency',\n",
    "           'angle_of_attack', \n",
    "           'chord_length', \n",
    "           'free_stream_velocity',\n",
    "           'suction_side_displacement_thickness',\n",
    "           'scaled_sound_pressure_level']\n",
    "\n",
    "features = ['frequency','angle_of_attack', \n",
    "            'chord_length', \n",
    "            'free_stream_velocity',\n",
    "            'suction_side_displacement_thickness']\n",
    "\n",
    "df = pd.read_csv(url_file, sep='\\t', header=None, names=columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0c6abd23",
   "metadata": {},
   "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>frequency</th>\n",
       "      <th>angle_of_attack</th>\n",
       "      <th>chord_length</th>\n",
       "      <th>free_stream_velocity</th>\n",
       "      <th>suction_side_displacement_thickness</th>\n",
       "      <th>scaled_sound_pressure_level</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>800</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3048</td>\n",
       "      <td>71.3</td>\n",
       "      <td>0.002663</td>\n",
       "      <td>126.201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3048</td>\n",
       "      <td>71.3</td>\n",
       "      <td>0.002663</td>\n",
       "      <td>125.201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3048</td>\n",
       "      <td>71.3</td>\n",
       "      <td>0.002663</td>\n",
       "      <td>125.951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3048</td>\n",
       "      <td>71.3</td>\n",
       "      <td>0.002663</td>\n",
       "      <td>127.591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3048</td>\n",
       "      <td>71.3</td>\n",
       "      <td>0.002663</td>\n",
       "      <td>127.461</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   frequency  angle_of_attack  chord_length  free_stream_velocity   \n",
       "0        800              0.0        0.3048                  71.3  \\\n",
       "1       1000              0.0        0.3048                  71.3   \n",
       "2       1250              0.0        0.3048                  71.3   \n",
       "3       1600              0.0        0.3048                  71.3   \n",
       "4       2000              0.0        0.3048                  71.3   \n",
       "\n",
       "   suction_side_displacement_thickness  scaled_sound_pressure_level  \n",
       "0                             0.002663                      126.201  \n",
       "1                             0.002663                      125.201  \n",
       "2                             0.002663                      125.951  \n",
       "3                             0.002663                      127.591  \n",
       "4                             0.002663                      127.461  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b927a7dd",
   "metadata": {},
   "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>frequency</th>\n",
       "      <th>angle_of_attack</th>\n",
       "      <th>chord_length</th>\n",
       "      <th>free_stream_velocity</th>\n",
       "      <th>suction_side_displacement_thickness</th>\n",
       "      <th>scaled_sound_pressure_level</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1498</th>\n",
       "      <td>2500</td>\n",
       "      <td>15.6</td>\n",
       "      <td>0.1016</td>\n",
       "      <td>39.6</td>\n",
       "      <td>0.052849</td>\n",
       "      <td>110.264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1499</th>\n",
       "      <td>3150</td>\n",
       "      <td>15.6</td>\n",
       "      <td>0.1016</td>\n",
       "      <td>39.6</td>\n",
       "      <td>0.052849</td>\n",
       "      <td>109.254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1500</th>\n",
       "      <td>4000</td>\n",
       "      <td>15.6</td>\n",
       "      <td>0.1016</td>\n",
       "      <td>39.6</td>\n",
       "      <td>0.052849</td>\n",
       "      <td>106.604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1501</th>\n",
       "      <td>5000</td>\n",
       "      <td>15.6</td>\n",
       "      <td>0.1016</td>\n",
       "      <td>39.6</td>\n",
       "      <td>0.052849</td>\n",
       "      <td>106.224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1502</th>\n",
       "      <td>6300</td>\n",
       "      <td>15.6</td>\n",
       "      <td>0.1016</td>\n",
       "      <td>39.6</td>\n",
       "      <td>0.052849</td>\n",
       "      <td>104.204</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      frequency  angle_of_attack  chord_length  free_stream_velocity   \n",
       "1498       2500             15.6        0.1016                  39.6  \\\n",
       "1499       3150             15.6        0.1016                  39.6   \n",
       "1500       4000             15.6        0.1016                  39.6   \n",
       "1501       5000             15.6        0.1016                  39.6   \n",
       "1502       6300             15.6        0.1016                  39.6   \n",
       "\n",
       "      suction_side_displacement_thickness  scaled_sound_pressure_level  \n",
       "1498                             0.052849                      110.264  \n",
       "1499                             0.052849                      109.254  \n",
       "1500                             0.052849                      106.604  \n",
       "1501                             0.052849                      106.224  \n",
       "1502                             0.052849                      104.204  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "ba40ec39",
   "metadata": {},
   "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>frequency</th>\n",
       "      <th>angle_of_attack</th>\n",
       "      <th>chord_length</th>\n",
       "      <th>free_stream_velocity</th>\n",
       "      <th>suction_side_displacement_thickness</th>\n",
       "      <th>scaled_sound_pressure_level</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1503.000000</td>\n",
       "      <td>1503.000000</td>\n",
       "      <td>1503.000000</td>\n",
       "      <td>1503.000000</td>\n",
       "      <td>1503.000000</td>\n",
       "      <td>1503.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2886.380572</td>\n",
       "      <td>6.782302</td>\n",
       "      <td>0.136548</td>\n",
       "      <td>50.860745</td>\n",
       "      <td>0.011140</td>\n",
       "      <td>124.835943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>3152.573137</td>\n",
       "      <td>5.918128</td>\n",
       "      <td>0.093541</td>\n",
       "      <td>15.572784</td>\n",
       "      <td>0.013150</td>\n",
       "      <td>6.898657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>200.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.025400</td>\n",
       "      <td>31.700000</td>\n",
       "      <td>0.000401</td>\n",
       "      <td>103.380000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>800.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.050800</td>\n",
       "      <td>39.600000</td>\n",
       "      <td>0.002535</td>\n",
       "      <td>120.191000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1600.000000</td>\n",
       "      <td>5.400000</td>\n",
       "      <td>0.101600</td>\n",
       "      <td>39.600000</td>\n",
       "      <td>0.004957</td>\n",
       "      <td>125.721000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>4000.000000</td>\n",
       "      <td>9.900000</td>\n",
       "      <td>0.228600</td>\n",
       "      <td>71.300000</td>\n",
       "      <td>0.015576</td>\n",
       "      <td>129.995500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>20000.000000</td>\n",
       "      <td>22.200000</td>\n",
       "      <td>0.304800</td>\n",
       "      <td>71.300000</td>\n",
       "      <td>0.058411</td>\n",
       "      <td>140.987000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          frequency  angle_of_attack  chord_length  free_stream_velocity   \n",
       "count   1503.000000      1503.000000   1503.000000           1503.000000  \\\n",
       "mean    2886.380572         6.782302      0.136548             50.860745   \n",
       "std     3152.573137         5.918128      0.093541             15.572784   \n",
       "min      200.000000         0.000000      0.025400             31.700000   \n",
       "25%      800.000000         2.000000      0.050800             39.600000   \n",
       "50%     1600.000000         5.400000      0.101600             39.600000   \n",
       "75%     4000.000000         9.900000      0.228600             71.300000   \n",
       "max    20000.000000        22.200000      0.304800             71.300000   \n",
       "\n",
       "       suction_side_displacement_thickness  scaled_sound_pressure_level  \n",
       "count                          1503.000000                  1503.000000  \n",
       "mean                              0.011140                   124.835943  \n",
       "std                               0.013150                     6.898657  \n",
       "min                               0.000401                   103.380000  \n",
       "25%                               0.002535                   120.191000  \n",
       "50%                               0.004957                   125.721000  \n",
       "75%                               0.015576                   129.995500  \n",
       "max                               0.058411                   140.987000  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "cfc7d504",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1503 entries, 0 to 1502\n",
      "Data columns (total 6 columns):\n",
      " #   Column                               Non-Null Count  Dtype  \n",
      "---  ------                               --------------  -----  \n",
      " 0   frequency                            1503 non-null   int64  \n",
      " 1   angle_of_attack                      1503 non-null   float64\n",
      " 2   chord_length                         1503 non-null   float64\n",
      " 3   free_stream_velocity                 1503 non-null   float64\n",
      " 4   suction_side_displacement_thickness  1503 non-null   float64\n",
      " 5   scaled_sound_pressure_level          1503 non-null   float64\n",
      "dtypes: float64(5), int64(1)\n",
      "memory usage: 70.6 KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d5f5320f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess_input(df):\n",
    "    \n",
    "    train_data = df.copy()\n",
    "    train_set = train_data.drop(\"scaled_sound_pressure_level\", axis = 1)\n",
    "    train_labels = train_data[\"scaled_sound_pressure_level\"].copy()\n",
    "    \n",
    "    train_set = QuantileTransformer(random_state=0).fit_transform(train_set)\n",
    "    train_set = MaxAbsScaler().fit_transform(train_set)\n",
    "    train_set = StandardScaler().fit_transform(train_set)\n",
    "    \n",
    "    # Train-test split\n",
    "    X_train, X_test, y_train, y_test = train_test_split(train_set, train_labels, test_size=0.2)\n",
    "    \n",
    "    return X_train, X_test, y_train, y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b577ebf7",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = preprocess_input(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d486047c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1202, 5)\n",
      "(301, 5)\n",
      "(1202,)\n",
      "(301,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "print(y_train.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5b3219e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense (Dense)               (None, 48)                288       \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 1)                 49        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 337 (1.32 KB)\n",
      "Trainable params: 337 (1.32 KB)\n",
      "Non-trainable params: 0 (0.00 Byte)\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Input(X_train.shape[1],))\n",
    "\n",
    "model.add(Dense(48, activation = 'relu'))\n",
    "\n",
    "model.add(Dense(1))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5f0f3766",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss = 'mean_squared_error', metrics = tf.keras.metrics.RootMeanSquaredError())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "38595e5b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_shape(model):\n",
    "    weights_layer = model.get_weights()\n",
    "    shapes = []\n",
    "    for weights in weights_layer:\n",
    "        shapes.append(weights.shape)\n",
    "    return shapes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7c976809",
   "metadata": {},
   "outputs": [],
   "source": [
    "def set_shape(weights,shapes):\n",
    "    new_weights = []\n",
    "    index=0\n",
    "    for shape in shapes:\n",
    "        if(len(shape)>1):\n",
    "            n_nodes = np.prod(shape) + index\n",
    "        else:\n",
    "            n_nodes=shape[0]+index\n",
    "        tmp = np.array(weights[index:n_nodes]).reshape(shape)\n",
    "        new_weights.append(tmp)\n",
    "        index=n_nodes\n",
    "    return new_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "de82e63a",
   "metadata": {},
   "outputs": [],
   "source": [
    "shape = get_shape(model)\n",
    "num_solutions = 20\n",
    "num_dimensions = 337"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "0d660063",
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_nn(W):\n",
    "    model.set_weights(set_shape(W,shape))\n",
    "    score = model.evaluate(X_train, y_train, verbose = 0)\n",
    "    return score[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b6bde5d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from math import gamma\n",
    "from typing import List, Union, Tuple, Dict\n",
    "\n",
    "def get_levy_flight_step(beta: float = 1.0, multiplier: float = 0.001, \n",
    "                             size: Union[List, Tuple, np.ndarray] = None, case: int = 0) -> Union[float, List, np.ndarray]:\n",
    "        \"\"\"\n",
    "        Get the Levy-flight step size\n",
    "\n",
    "        Args:\n",
    "            beta (float): Should be in range [0, 2].\n",
    "\n",
    "                * 0-1: small range --> exploit\n",
    "                * 1-2: large range --> explore\n",
    "\n",
    "            multiplier (float): default = 0.001\n",
    "            size (tuple, list): size of levy-flight steps, for example: (3, 2), 5, (4, )\n",
    "            case (int): Should be one of these value [0, 1, -1].\n",
    "\n",
    "                * 0: return multiplier * s * self.generator.uniform()\n",
    "                * 1: return multiplier * s * self.generator.normal(0, 1)\n",
    "                * -1: return multiplier * s\n",
    "\n",
    "        Returns:\n",
    "            float, list, np.ndarray: The step size of Levy-flight trajectory\n",
    "        \"\"\"\n",
    "        # u and v are two random variables which follow self.generator.normal distribution\n",
    "        # sigma_u : standard deviation of u\n",
    "        sigma_u = np.power(gamma(1. + beta) * np.sin(np.pi * beta / 2) / (gamma((1 + beta) / 2.) * beta * np.power(2., (beta - 1) / 2)), 1. / beta)\n",
    "        # sigma_v : standard deviation of v\n",
    "        sigma_v = 1\n",
    "        size = 1 if size is None else size\n",
    "        u = np.random.normal(0, sigma_u ** 2, size)\n",
    "        v = np.random.normal(0, sigma_v ** 2, size)\n",
    "        s = u / np.power(np.abs(v), 1 / beta)\n",
    "        if case == 0:\n",
    "            step = multiplier * s * np.random.uniform()\n",
    "        elif case == 1:\n",
    "            step = multiplier * s * np.random.normal(0, 1)\n",
    "        else:\n",
    "            step = multiplier * s\n",
    "        return step[0] if size == 1 else step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "02fb40af",
   "metadata": {},
   "outputs": [],
   "source": [
    "def sech(x):\n",
    "    return 1 / np.cosh(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "61e842d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def MHGWO(agent_vector, agent_error, X_alpha, X_beta, X_delta, a_alpha, a_beta, a_delta, E):\n",
    "    \n",
    "#     R1, R2, R3 = np.random.uniform(), np.random.uniform(), np.random.uniform()\n",
    "#     S = R1 + R2 + R3\n",
    "#     R1, R2, R3 = R1/S, R2/S, R3/S\n",
    "    \n",
    "    A1 = a_alpha * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    A2 = a_beta * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    A3 = a_delta * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    C1 = 2 * np.random.rand(num_dimensions)\n",
    "    C2 = 2 * np.random.rand(num_dimensions)\n",
    "    C3 = 2 * np.random.rand(num_dimensions)\n",
    "    X1 = X_alpha - (A1 * np.abs((C1 * X_alpha) - agent_vector))\n",
    "    X2 = X_beta - (A2 * np.abs((C2 * X_beta) - agent_vector))\n",
    "    X3 = X_delta - (A3 * np.abs((C3 * X_delta) - agent_vector))\n",
    "\n",
    "#     Y = (R1 * X1) + (R2 * X2) + (R3 * X3)\n",
    "    Y = (X1 + X2 + X3) / 3\n",
    "    y_error = evaluate_nn(Y)\n",
    "    if y_error < agent_error:\n",
    "        agent_vector = Y\n",
    "        agent_error = y_error\n",
    "    \n",
    "    # mutation driven approach\n",
    "    levy = get_levy_flight_step(multiplier = 0.1, size = num_dimensions, case = 1)\n",
    "    Z = agent_vector + (E * levy)\n",
    "    z_error = evaluate_nn(Z)\n",
    "    if z_error < agent_error:\n",
    "        agent_vector = Z\n",
    "        agent_error = z_error\n",
    "\n",
    "    return agent_vector, agent_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "05f13b65",
   "metadata": {},
   "outputs": [],
   "source": [
    "def custom_gwo(agent_vector, agent_error, epoch, best_solution, X_alpha, X_beta, X_delta, a_alpha, a_beta, a_delta):\n",
    "    \n",
    "    A1 = a_alpha * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    A2 = a_beta * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    A3 = a_delta * (2 * np.random.rand(num_dimensions) - 1)\n",
    "    C1 = 2 * np.random.rand(num_dimensions)\n",
    "    C2 = 2 * np.random.rand(num_dimensions)\n",
    "    C3 = 2 * np.random.rand(num_dimensions)\n",
    "    X1 = X_alpha - (A1 * np.abs((C1 * X_alpha) - agent_vector))\n",
    "    X2 = X_beta - (A2 * np.abs((C2 * X_beta) - agent_vector))\n",
    "    X3 = X_delta - (A3 * np.abs((C3 * X_delta) - agent_vector))\n",
    "\n",
    "    Y = (X1 + X2 + X3) / 3\n",
    "    y_error = evaluate_nn(Y)\n",
    "    if y_error < agent_error:\n",
    "        agent_vector = Y\n",
    "        agent_error = y_error\n",
    "        \n",
    "    # mutation\n",
    "    \n",
    "    # approach 2\n",
    "    \"\"\"\n",
    "    levy = get_levy_flight_step(beta = 1, multiplier = 0.1, case = -1)\n",
    "    pos_new = agent_vector + ((1/np.sqrt(epoch + 0.01)) * levy * (best_solution - agent_vector))\n",
    "    \n",
    "    pos_new_error = evaluate_nn(pos_new)\n",
    "    if pos_new_error < agent_error:\n",
    "        agent_vector = pos_new\n",
    "        agent_error = pos_new_error\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    # approach 3\n",
    "    if np.random.random() < 0.5:  \n",
    "        pos_new = best_solution + (np.random.normal(0, 1, num_dimensions) * (best_solution - agent_vector))\n",
    "    else:\n",
    "        beta = np.random.uniform(0.01, 1)\n",
    "        levy = get_levy_flight_step(beta = beta, multiplier = 0.05, size = num_dimensions, case = -1) \n",
    "        pos_new = best_solution + (levy * (best_solution - agent_vector))\n",
    "        \n",
    "    pos_new_error = evaluate_nn(pos_new)\n",
    "    if pos_new_error < agent_error:\n",
    "        agent_vector = pos_new\n",
    "        agent_error = pos_new_error\n",
    "    \n",
    "    \n",
    "    # approach 4\n",
    "    \"\"\"\n",
    "     if np.random.random() < 0.5:\n",
    "        levy_step = get_levy_flight_step(multiplier = 0.01, size = num_dimensions, case = -1)\n",
    "        Z = agent_vector + levy_step\n",
    "        z_error = evaluate_nn(Z)\n",
    "        if z_error < agent_error:\n",
    "            agent_vector = Z\n",
    "            agent_error = z_error\n",
    "    \"\"\"\n",
    "    \n",
    "    return agent_vector, agent_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "87c926ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(20, 337)\n"
     ]
    }
   ],
   "source": [
    "# generate the solutions\n",
    "current_population_vectors = np.random.uniform(size = (num_solutions, num_dimensions))\n",
    "print(current_population_vectors.shape)\n",
    "\n",
    "# set parameters\n",
    "epochs, iteration_errors = 1000, []\n",
    "\n",
    "# calculate error of each solution\n",
    "curr_errors = np.empty(shape=(num_solutions))\n",
    "for i in range(num_solutions):\n",
    "    curr_errors[i] = evaluate_nn(current_population_vectors[i])\n",
    "    \n",
    "# get alpha, beta and delta wolves\n",
    "X_alpha, X_beta, X_delta, x_alpha_score, x_beta_score, x_delta_score = GWO.find_alpha_beta_delta(current_population_vectors, curr_errors)\n",
    "\n",
    "# set the best solution and best error to alpha wolf and alpha wolf's error respectively\n",
    "best_solution, best_error = X_alpha, x_alpha_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "c0457e85",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch  1\n",
      "Top three least errors :  102.77053833007812 103.96216583251953 105.27889251708984\n",
      "Best error :  95.11406707763672 \n",
      "\n",
      "Epoch  2\n",
      "Top three least errors :  95.11406707763672 95.76801300048828 97.6406021118164\n",
      "Best error :  63.5240592956543 \n",
      "\n",
      "Epoch  3\n",
      "Top three least errors :  63.5240592956543 84.4583969116211 90.982666015625\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\admin\\anaconda3\\lib\\site-packages\\keras\\src\\backend.py:4312: RuntimeWarning: overflow encountered in cast\n",
      "  value = np.asarray(value, dtype=dtype_numpy(x))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  58.549354553222656 \n",
      "\n",
      "Epoch  4\n",
      "Top three least errors :  58.549354553222656 63.5240592956543 63.574161529541016\n",
      "Best error :  48.73828887939453 \n",
      "\n",
      "Epoch  5\n",
      "Top three least errors :  48.73828887939453 52.8521614074707 53.642127990722656\n",
      "Best error :  41.37184524536133 \n",
      "\n",
      "Epoch  6\n",
      "Top three least errors :  41.37184524536133 44.2247428894043 44.531707763671875\n",
      "Best error :  35.11841583251953 \n",
      "\n",
      "Epoch  7\n",
      "Top three least errors :  35.11841583251953 35.41050338745117 36.63443374633789\n",
      "Best error :  33.023189544677734 \n",
      "\n",
      "Epoch  8\n",
      "Top three least errors :  33.023189544677734 33.479339599609375 33.5479850769043\n",
      "Best error :  30.676668167114258 \n",
      "\n",
      "Epoch  9\n",
      "Top three least errors :  30.676668167114258 32.04003143310547 32.49953079223633\n",
      "Best error :  24.13945770263672 \n",
      "\n",
      "Epoch  10\n",
      "Top three least errors :  24.13945770263672 26.053823471069336 27.955970764160156\n",
      "Best error :  20.15976333618164 \n",
      "\n",
      "Epoch  11\n",
      "Top three least errors :  20.15976333618164 22.84790802001953 23.125232696533203\n",
      "Best error :  18.53851890563965 \n",
      "\n",
      "Epoch  12\n",
      "Top three least errors :  18.53851890563965 19.5285701751709 19.696903228759766\n",
      "Best error :  15.84785270690918 \n",
      "\n",
      "Epoch  13\n",
      "Top three least errors :  15.84785270690918 16.002349853515625 18.20430564880371\n",
      "Best error :  15.547158241271973 \n",
      "\n",
      "Epoch  14\n",
      "Top three least errors :  15.547158241271973 15.84785270690918 16.002349853515625\n",
      "Best error :  12.30342960357666 \n",
      "\n",
      "Epoch  15\n",
      "Top three least errors :  12.30342960357666 13.4376802444458 14.88050651550293\n",
      "Best error :  11.796198844909668 \n",
      "\n",
      "Epoch  16\n",
      "Top three least errors :  11.796198844909668 12.167247772216797 12.30342960357666\n",
      "Best error :  11.796198844909668 \n",
      "\n",
      "Epoch  17\n",
      "Top three least errors :  11.796198844909668 12.167247772216797 12.30342960357666\n",
      "Best error :  11.796198844909668 \n",
      "\n",
      "Epoch  18\n",
      "Top three least errors :  11.796198844909668 11.831992149353027 12.103232383728027\n",
      "Best error :  9.964892387390137 \n",
      "\n",
      "Epoch  19\n",
      "Top three least errors :  9.964892387390137 10.89315128326416 11.796198844909668\n",
      "Best error :  9.964892387390137 \n",
      "\n",
      "Epoch  20\n",
      "Top three least errors :  9.964892387390137 10.25664234161377 10.825033187866211\n",
      "Best error :  9.964892387390137 \n",
      "\n",
      "Epoch  21\n",
      "Top three least errors :  9.964892387390137 10.25664234161377 10.31927490234375\n",
      "Best error :  9.448408126831055 \n",
      "\n",
      "Epoch  22\n",
      "Top three least errors :  9.448408126831055 9.65689468383789 9.945308685302734\n",
      "Best error :  9.448408126831055 \n",
      "\n",
      "Epoch  23\n",
      "Top three least errors :  9.448408126831055 9.65689468383789 9.945308685302734\n",
      "Best error :  9.417584419250488 \n",
      "\n",
      "Epoch  24\n",
      "Top three least errors :  9.417584419250488 9.448408126831055 9.535862922668457\n",
      "Best error :  9.417584419250488 \n",
      "\n",
      "Epoch  25\n",
      "Top three least errors :  9.417584419250488 9.448408126831055 9.471901893615723\n",
      "Best error :  8.96142578125 \n",
      "\n",
      "Epoch  26\n",
      "Top three least errors :  8.96142578125 9.417584419250488 9.448408126831055\n",
      "Best error :  8.96142578125 \n",
      "\n",
      "Epoch  27\n",
      "Top three least errors :  8.96142578125 9.417584419250488 9.448408126831055\n",
      "Best error :  8.96142578125 \n",
      "\n",
      "Epoch  28\n",
      "Top three least errors :  8.96142578125 9.0327730178833 9.417584419250488\n",
      "Best error :  8.911701202392578 \n",
      "\n",
      "Epoch  29\n",
      "Top three least errors :  8.911701202392578 8.96142578125 9.417584419250488\n",
      "Best error :  8.911701202392578 \n",
      "\n",
      "Epoch  30\n",
      "Top three least errors :  8.911701202392578 8.96142578125 9.417584419250488\n",
      "Best error :  8.634029388427734 \n",
      "\n",
      "Epoch  31\n",
      "Top three least errors :  8.634029388427734 8.911701202392578 8.96142578125\n",
      "Best error :  8.512866973876953 \n",
      "\n",
      "Epoch  32\n",
      "Top three least errors :  8.512866973876953 8.634029388427734 8.885384559631348\n",
      "Best error :  8.512866973876953 \n",
      "\n",
      "Epoch  33\n",
      "Top three least errors :  8.512866973876953 8.592300415039062 8.885384559631348\n",
      "Best error :  8.512866973876953 \n",
      "\n",
      "Epoch  34\n",
      "Top three least errors :  8.512866973876953 8.563512802124023 8.592300415039062\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  35\n",
      "Top three least errors :  8.394697189331055 8.512866973876953 8.563512802124023\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  36\n",
      "Top three least errors :  8.394697189331055 8.512866973876953 8.563512802124023\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  37\n",
      "Top three least errors :  8.394697189331055 8.512866973876953 8.563512802124023\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  38\n",
      "Top three least errors :  8.394697189331055 8.512866973876953 8.563512802124023\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  39\n",
      "Top three least errors :  8.394697189331055 8.48093032836914 8.512866973876953\n",
      "Best error :  8.394697189331055 \n",
      "\n",
      "Epoch  40\n",
      "Top three least errors :  8.394697189331055 8.48093032836914 8.512866973876953\n",
      "Best error :  8.3550443649292 \n",
      "\n",
      "Epoch  41\n",
      "Top three least errors :  8.3550443649292 8.394697189331055 8.443262100219727\n",
      "Best error :  8.255215644836426 \n",
      "\n",
      "Epoch  42\n",
      "Top three least errors :  8.255215644836426 8.3550443649292 8.443262100219727\n",
      "Best error :  7.649556636810303 \n",
      "\n",
      "Epoch  43\n",
      "Top three least errors :  7.649556636810303 7.826550483703613 8.143139839172363\n",
      "Best error :  7.611993789672852 \n",
      "\n",
      "Epoch  44\n",
      "Top three least errors :  7.611993789672852 7.649556636810303 7.825693130493164\n",
      "Best error :  7.611993789672852 \n",
      "\n",
      "Epoch  45\n",
      "Top three least errors :  7.611993789672852 7.649556636810303 7.825693130493164\n",
      "Best error :  7.454042434692383 \n",
      "\n",
      "Epoch  46\n",
      "Top three least errors :  7.454042434692383 7.611993789672852 7.825693130493164\n",
      "Best error :  6.918899059295654 \n",
      "\n",
      "Epoch  47\n",
      "Top three least errors :  6.918899059295654 7.241711616516113 7.454042434692383\n",
      "Best error :  6.899960041046143 \n",
      "\n",
      "Epoch  48\n",
      "Top three least errors :  6.899960041046143 6.918899059295654 7.241711616516113\n",
      "Best error :  6.899960041046143 \n",
      "\n",
      "Epoch  49\n",
      "Top three least errors :  6.899960041046143 6.918899059295654 7.241711616516113\n",
      "Best error :  6.899960041046143 \n",
      "\n",
      "Epoch  50\n",
      "Top three least errors :  6.899960041046143 6.918899059295654 7.241711616516113\n",
      "Best error :  6.899960041046143 \n",
      "\n",
      "Epoch  51\n",
      "Top three least errors :  6.899960041046143 6.918899059295654 7.241711616516113\n",
      "Best error :  6.899960041046143 \n",
      "\n",
      "Epoch  52\n",
      "Top three least errors :  6.899960041046143 6.918899059295654 6.960538864135742\n",
      "Best error :  6.566112995147705 \n",
      "\n",
      "Epoch  53\n",
      "Top three least errors :  6.566112995147705 6.899960041046143 6.918899059295654\n",
      "Best error :  6.354241371154785 \n",
      "\n",
      "Epoch  54\n",
      "Top three least errors :  6.354241371154785 6.5474677085876465 6.566112995147705\n",
      "Best error :  6.354241371154785 \n",
      "\n",
      "Epoch  55\n",
      "Top three least errors :  6.354241371154785 6.5474677085876465 6.566112995147705\n",
      "Best error :  6.354241371154785 \n",
      "\n",
      "Epoch  56\n",
      "Top three least errors :  6.354241371154785 6.5474677085876465 6.566112995147705\n",
      "Best error :  6.354241371154785 \n",
      "\n",
      "Epoch  57\n",
      "Top three least errors :  6.354241371154785 6.4970526695251465 6.5474677085876465\n",
      "Best error :  6.354241371154785 \n",
      "\n",
      "Epoch  58\n",
      "Top three least errors :  6.354241371154785 6.4970526695251465 6.5474677085876465\n",
      "Best error :  6.291489124298096 \n",
      "\n",
      "Epoch  59\n",
      "Top three least errors :  6.291489124298096 6.354241371154785 6.4970526695251465\n",
      "Best error :  6.10119104385376 \n",
      "\n",
      "Epoch  60\n",
      "Top three least errors :  6.10119104385376 6.291489124298096 6.354241371154785\n",
      "Best error :  6.10119104385376 \n",
      "\n",
      "Epoch  61\n",
      "Top three least errors :  6.10119104385376 6.256809234619141 6.354241371154785\n",
      "Best error :  5.6783223152160645 \n",
      "\n",
      "Epoch  62\n",
      "Top three least errors :  5.6783223152160645 6.055045127868652 6.10119104385376\n",
      "Best error :  5.6783223152160645 \n",
      "\n",
      "Epoch  63\n",
      "Top three least errors :  5.6783223152160645 6.055045127868652 6.056789398193359\n",
      "Best error :  5.6783223152160645 \n",
      "\n",
      "Epoch  64\n",
      "Top three least errors :  5.6783223152160645 6.055045127868652 6.056789398193359\n",
      "Best error :  5.6783223152160645 \n",
      "\n",
      "Epoch  65\n",
      "Top three least errors :  5.6783223152160645 5.685658931732178 5.783554553985596\n",
      "Best error :  5.5441203117370605 \n",
      "\n",
      "Epoch  66\n",
      "Top three least errors :  5.5441203117370605 5.627438068389893 5.6783223152160645\n",
      "Best error :  5.475144863128662 \n",
      "\n",
      "Epoch  67\n",
      "Top three least errors :  5.475144863128662 5.498756408691406 5.5441203117370605\n",
      "Best error :  5.192662239074707 \n",
      "\n",
      "Epoch  68\n",
      "Top three least errors :  5.192662239074707 5.475144863128662 5.488275527954102\n",
      "Best error :  5.192662239074707 \n",
      "\n",
      "Epoch  69\n",
      "Top three least errors :  5.192662239074707 5.475144863128662 5.488275527954102\n",
      "Best error :  5.192662239074707 \n",
      "\n",
      "Epoch  70\n",
      "Top three least errors :  5.192662239074707 5.4123334884643555 5.475144863128662\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  5.099549293518066 \n",
      "\n",
      "Epoch  71\n",
      "Top three least errors :  5.099549293518066 5.192662239074707 5.244421005249023\n",
      "Best error :  5.099549293518066 \n",
      "\n",
      "Epoch  72\n",
      "Top three least errors :  5.099549293518066 5.192662239074707 5.244421005249023\n",
      "Best error :  5.099549293518066 \n",
      "\n",
      "Epoch  73\n",
      "Top three least errors :  5.099549293518066 5.192662239074707 5.244421005249023\n",
      "Best error :  5.099549293518066 \n",
      "\n",
      "Epoch  74\n",
      "Top three least errors :  5.099549293518066 5.192662239074707 5.244421005249023\n",
      "Best error :  5.099549293518066 \n",
      "\n",
      "Epoch  75\n",
      "Top three least errors :  5.099549293518066 5.192662239074707 5.193145275115967\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  76\n",
      "Top three least errors :  4.756775379180908 5.099549293518066 5.192662239074707\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  77\n",
      "Top three least errors :  4.756775379180908 5.099549293518066 5.192662239074707\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  78\n",
      "Top three least errors :  4.756775379180908 5.099549293518066 5.192662239074707\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  79\n",
      "Top three least errors :  4.756775379180908 5.099549293518066 5.192662239074707\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  80\n",
      "Top three least errors :  4.756775379180908 4.947261810302734 5.099549293518066\n",
      "Best error :  4.756775379180908 \n",
      "\n",
      "Epoch  81\n",
      "Top three least errors :  4.756775379180908 4.947261810302734 5.099549293518066\n",
      "Best error :  4.723586559295654 \n",
      "\n",
      "Epoch  82\n",
      "Top three least errors :  4.723586559295654 4.756775379180908 4.838592529296875\n",
      "Best error :  4.5102128982543945 \n",
      "\n",
      "Epoch  83\n",
      "Top three least errors :  4.5102128982543945 4.675743579864502 4.697213649749756\n",
      "Best error :  4.4491424560546875 \n",
      "\n",
      "Epoch  84\n",
      "Top three least errors :  4.4491424560546875 4.4539475440979 4.461118698120117\n",
      "Best error :  4.4491424560546875 \n",
      "\n",
      "Epoch  85\n",
      "Top three least errors :  4.4491424560546875 4.4539475440979 4.461118698120117\n",
      "Best error :  4.398185729980469 \n",
      "\n",
      "Epoch  86\n",
      "Top three least errors :  4.398185729980469 4.4409284591674805 4.4491424560546875\n",
      "Best error :  4.398185729980469 \n",
      "\n",
      "Epoch  87\n",
      "Top three least errors :  4.398185729980469 4.4005560874938965 4.421098709106445\n",
      "Best error :  4.398185729980469 \n",
      "\n",
      "Epoch  88\n",
      "Top three least errors :  4.398185729980469 4.4005560874938965 4.421098709106445\n",
      "Best error :  4.376874923706055 \n",
      "\n",
      "Epoch  89\n",
      "Top three least errors :  4.376874923706055 4.398185729980469 4.41108512878418\n",
      "Best error :  4.376874923706055 \n",
      "\n",
      "Epoch  90\n",
      "Top three least errors :  4.376874923706055 4.398185729980469 4.41108512878418\n",
      "Best error :  4.265321731567383 \n",
      "\n",
      "Epoch  91\n",
      "Top three least errors :  4.265321731567383 4.322284698486328 4.373569488525391\n",
      "Best error :  4.251017093658447 \n",
      "\n",
      "Epoch  92\n",
      "Top three least errors :  4.251017093658447 4.263426303863525 4.265321731567383\n",
      "Best error :  4.241235733032227 \n",
      "\n",
      "Epoch  93\n",
      "Top three least errors :  4.241235733032227 4.251017093658447 4.263426303863525\n",
      "Best error :  4.218403339385986 \n",
      "\n",
      "Epoch  94\n",
      "Top three least errors :  4.218403339385986 4.232015609741211 4.241235733032227\n",
      "Best error :  4.213274955749512 \n",
      "\n",
      "Epoch  95\n",
      "Top three least errors :  4.213274955749512 4.214990139007568 4.218403339385986\n",
      "Best error :  4.213274955749512 \n",
      "\n",
      "Epoch  96\n",
      "Top three least errors :  4.213274955749512 4.214990139007568 4.218033313751221\n",
      "Best error :  4.213274955749512 \n",
      "\n",
      "Epoch  97\n",
      "Top three least errors :  4.213274955749512 4.214990139007568 4.215278625488281\n",
      "Best error :  4.207553386688232 \n",
      "\n",
      "Epoch  98\n",
      "Top three least errors :  4.207553386688232 4.208273410797119 4.213274955749512\n",
      "Best error :  4.204929828643799 \n",
      "\n",
      "Epoch  99\n",
      "Top three least errors :  4.204929828643799 4.207360744476318 4.207553386688232\n",
      "Best error :  4.202852249145508 \n",
      "\n",
      "Epoch  100\n",
      "Top three least errors :  4.202852249145508 4.204656600952148 4.204929828643799\n",
      "Best error :  4.202852249145508 \n",
      "\n",
      "Epoch  101\n",
      "Top three least errors :  4.202852249145508 4.203858852386475 4.204656600952148\n",
      "Best error :  4.1977033615112305 \n",
      "\n",
      "Epoch  102\n",
      "Top three least errors :  4.1977033615112305 4.199740409851074 4.200136661529541\n",
      "Best error :  4.1777567863464355 \n",
      "\n",
      "Epoch  103\n",
      "Top three least errors :  4.1777567863464355 4.1917948722839355 4.197193622589111\n",
      "Best error :  4.1777567863464355 \n",
      "\n",
      "Epoch  104\n",
      "Top three least errors :  4.1777567863464355 4.190601348876953 4.1917948722839355\n",
      "Best error :  4.1777567863464355 \n",
      "\n",
      "Epoch  105\n",
      "Top three least errors :  4.1777567863464355 4.189011573791504 4.1917948722839355\n",
      "Best error :  4.1777567863464355 \n",
      "\n",
      "Epoch  106\n",
      "Top three least errors :  4.1777567863464355 4.184793472290039 4.1917948722839355\n",
      "Best error :  4.1777567863464355 \n",
      "\n",
      "Epoch  107\n",
      "Top three least errors :  4.1777567863464355 4.184793472290039 4.186474323272705\n",
      "Best error :  4.168973445892334 \n",
      "\n",
      "Epoch  108\n",
      "Top three least errors :  4.168973445892334 4.1721343994140625 4.1721930503845215\n",
      "Best error :  4.1665754318237305 \n",
      "\n",
      "Epoch  109\n",
      "Top three least errors :  4.1665754318237305 4.168973445892334 4.1721343994140625\n",
      "Best error :  4.164788722991943 \n",
      "\n",
      "Epoch  110\n",
      "Top three least errors :  4.164788722991943 4.1665754318237305 4.16767692565918\n",
      "Best error :  4.154513835906982 \n",
      "\n",
      "Epoch  111\n",
      "Top three least errors :  4.154513835906982 4.162586688995361 4.164788722991943\n",
      "Best error :  4.152602672576904 \n",
      "\n",
      "Epoch  112\n",
      "Top three least errors :  4.152602672576904 4.153843879699707 4.154513835906982\n",
      "Best error :  4.152602672576904 \n",
      "\n",
      "Epoch  113\n",
      "Top three least errors :  4.152602672576904 4.153843879699707 4.154513835906982\n",
      "Best error :  4.1525444984436035 \n",
      "\n",
      "Epoch  114\n",
      "Top three least errors :  4.1525444984436035 4.152602672576904 4.153843879699707\n",
      "Best error :  4.151485443115234 \n",
      "\n",
      "Epoch  115\n",
      "Top three least errors :  4.151485443115234 4.1525444984436035 4.152602672576904\n",
      "Best error :  4.14292573928833 \n",
      "\n",
      "Epoch  116\n",
      "Top three least errors :  4.14292573928833 4.148271560668945 4.151485443115234\n",
      "Best error :  4.12432336807251 \n",
      "\n",
      "Epoch  117\n",
      "Top three least errors :  4.12432336807251 4.139204502105713 4.14292573928833\n",
      "Best error :  4.099991798400879 \n",
      "\n",
      "Epoch  118\n",
      "Top three least errors :  4.099991798400879 4.12432336807251 4.128673076629639\n",
      "Best error :  4.0965895652771 \n",
      "\n",
      "Epoch  119\n",
      "Top three least errors :  4.0965895652771 4.099991798400879 4.118297100067139\n",
      "Best error :  4.065582275390625 \n",
      "\n",
      "Epoch  120\n",
      "Top three least errors :  4.065582275390625 4.096230506896973 4.0965895652771\n",
      "Best error :  4.065582275390625 \n",
      "\n",
      "Epoch  121\n",
      "Top three least errors :  4.065582275390625 4.087090969085693 4.088082790374756\n",
      "Best error :  4.065582275390625 \n",
      "\n",
      "Epoch  122\n",
      "Top three least errors :  4.065582275390625 4.076854705810547 4.08577299118042\n",
      "Best error :  4.062844276428223 \n",
      "\n",
      "Epoch  123\n",
      "Top three least errors :  4.062844276428223 4.065582275390625 4.0712103843688965\n",
      "Best error :  4.062844276428223 \n",
      "\n",
      "Epoch  124\n",
      "Top three least errors :  4.062844276428223 4.065582275390625 4.068197727203369\n",
      "Best error :  4.057925701141357 \n",
      "\n",
      "Epoch  125\n",
      "Top three least errors :  4.057925701141357 4.061948776245117 4.062844276428223\n",
      "Best error :  4.055353164672852 \n",
      "\n",
      "Epoch  126\n",
      "Top three least errors :  4.055353164672852 4.057925701141357 4.058225154876709\n",
      "Best error :  4.055353164672852 \n",
      "\n",
      "Epoch  127\n",
      "Top three least errors :  4.055353164672852 4.055463790893555 4.057925701141357\n",
      "Best error :  4.055353164672852 \n",
      "\n",
      "Epoch  128\n",
      "Top three least errors :  4.055353164672852 4.055463790893555 4.057925701141357\n",
      "Best error :  4.055353164672852 \n",
      "\n",
      "Epoch  129\n",
      "Top three least errors :  4.055353164672852 4.055463790893555 4.056037425994873\n",
      "Best error :  4.055353164672852 \n",
      "\n",
      "Epoch  130\n",
      "Top three least errors :  4.055353164672852 4.055463790893555 4.056037425994873\n",
      "Best error :  4.051827907562256 \n",
      "\n",
      "Epoch  131\n",
      "Top three least errors :  4.051827907562256 4.0536909103393555 4.055353164672852\n",
      "Best error :  4.050251483917236 \n",
      "\n",
      "Epoch  132\n",
      "Top three least errors :  4.050251483917236 4.051827907562256 4.055353164672852\n",
      "Best error :  4.049957275390625 \n",
      "\n",
      "Epoch  133\n",
      "Top three least errors :  4.049957275390625 4.050251483917236 4.050468444824219\n",
      "Best error :  4.049957275390625 \n",
      "\n",
      "Epoch  134\n",
      "Top three least errors :  4.049957275390625 4.050251483917236 4.050468444824219\n",
      "Best error :  4.049957275390625 \n",
      "\n",
      "Epoch  135\n",
      "Top three least errors :  4.049957275390625 4.050251483917236 4.050468444824219\n",
      "Best error :  4.049957275390625 \n",
      "\n",
      "Epoch  136\n",
      "Top three least errors :  4.049957275390625 4.050144672393799 4.050468444824219\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  4.046730041503906 \n",
      "\n",
      "Epoch  137\n",
      "Top three least errors :  4.046730041503906 4.04793643951416 4.049957275390625\n",
      "Best error :  4.046730041503906 \n",
      "\n",
      "Epoch  138\n",
      "Top three least errors :  4.046730041503906 4.04793643951416 4.049957275390625\n",
      "Best error :  4.045923233032227 \n",
      "\n",
      "Epoch  139\n",
      "Top three least errors :  4.045923233032227 4.046730041503906 4.04793643951416\n",
      "Best error :  4.043523788452148 \n",
      "\n",
      "Epoch  140\n",
      "Top three least errors :  4.043523788452148 4.045796871185303 4.045923233032227\n",
      "Best error :  4.041772365570068 \n",
      "\n",
      "Epoch  141\n",
      "Top three least errors :  4.041772365570068 4.043523788452148 4.0453081130981445\n",
      "Best error :  4.040634632110596 \n",
      "\n",
      "Epoch  142\n",
      "Top three least errors :  4.040634632110596 4.040805816650391 4.041772365570068\n",
      "Best error :  4.032131195068359 \n",
      "\n",
      "Epoch  143\n",
      "Top three least errors :  4.032131195068359 4.040634632110596 4.040805816650391\n",
      "Best error :  4.024343490600586 \n",
      "\n",
      "Epoch  144\n",
      "Top three least errors :  4.024343490600586 4.0291829109191895 4.0304341316223145\n",
      "Best error :  4.022097110748291 \n",
      "\n",
      "Epoch  145\n",
      "Top three least errors :  4.022097110748291 4.0230255126953125 4.024343490600586\n",
      "Best error :  4.014875411987305 \n",
      "\n",
      "Epoch  146\n",
      "Top three least errors :  4.014875411987305 4.017874717712402 4.021586894989014\n",
      "Best error :  4.006505966186523 \n",
      "\n",
      "Epoch  147\n",
      "Top three least errors :  4.006505966186523 4.007729530334473 4.011122703552246\n",
      "Best error :  4.006505966186523 \n",
      "\n",
      "Epoch  148\n",
      "Top three least errors :  4.006505966186523 4.007729530334473 4.010470867156982\n",
      "Best error :  4.005089282989502 \n",
      "\n",
      "Epoch  149\n",
      "Top three least errors :  4.005089282989502 4.006505966186523 4.007469177246094\n",
      "Best error :  4.001806735992432 \n",
      "\n",
      "Epoch  150\n",
      "Top three least errors :  4.001806735992432 4.003345966339111 4.005089282989502\n",
      "Best error :  4.001806735992432 \n",
      "\n",
      "Epoch  151\n",
      "Top three least errors :  4.001806735992432 4.002587795257568 4.002618789672852\n",
      "Best error :  3.995518684387207 \n",
      "\n",
      "Epoch  152\n",
      "Top three least errors :  3.995518684387207 3.9998974800109863 4.001754283905029\n",
      "Best error :  3.992584705352783 \n",
      "\n",
      "Epoch  153\n",
      "Top three least errors :  3.992584705352783 3.995518684387207 3.9964382648468018\n",
      "Best error :  3.988269090652466 \n",
      "\n",
      "Epoch  154\n",
      "Top three least errors :  3.988269090652466 3.9903366565704346 3.992584705352783\n",
      "Best error :  3.986246109008789 \n",
      "\n",
      "Epoch  155\n",
      "Top three least errors :  3.986246109008789 3.9870247840881348 3.988269090652466\n",
      "Best error :  3.9852638244628906 \n",
      "\n",
      "Epoch  156\n",
      "Top three least errors :  3.9852638244628906 3.986246109008789 3.9864702224731445\n",
      "Best error :  3.9828784465789795 \n",
      "\n",
      "Epoch  157\n",
      "Top three least errors :  3.9828784465789795 3.9837639331817627 3.984490156173706\n",
      "Best error :  3.981478452682495 \n",
      "\n",
      "Epoch  158\n",
      "Top three least errors :  3.981478452682495 3.9828784465789795 3.9843101501464844\n",
      "Best error :  3.9789977073669434 \n",
      "\n",
      "Epoch  159\n",
      "Top three least errors :  3.9789977073669434 3.9791057109832764 3.980959177017212\n",
      "Best error :  3.9726157188415527 \n",
      "\n",
      "Epoch  160\n",
      "Top three least errors :  3.9726157188415527 3.9760069847106934 3.9780333042144775\n",
      "Best error :  3.9576008319854736 \n",
      "\n",
      "Epoch  161\n",
      "Top three least errors :  3.9576008319854736 3.9610443115234375 3.9660725593566895\n",
      "Best error :  3.9358971118927 \n",
      "\n",
      "Epoch  162\n",
      "Top three least errors :  3.9358971118927 3.941533327102661 3.9455363750457764\n",
      "Best error :  3.9239532947540283 \n",
      "\n",
      "Epoch  163\n",
      "Top three least errors :  3.9239532947540283 3.9259262084960938 3.9358971118927\n",
      "Best error :  3.872906446456909 \n",
      "\n",
      "Epoch  164\n",
      "Top three least errors :  3.872906446456909 3.8820595741271973 3.9099087715148926\n",
      "Best error :  3.863610029220581 \n",
      "\n",
      "Epoch  165\n",
      "Top three least errors :  3.863610029220581 3.872906446456909 3.8820595741271973\n",
      "Best error :  3.86249041557312 \n",
      "\n",
      "Epoch  166\n",
      "Top three least errors :  3.86249041557312 3.863610029220581 3.864527463912964\n",
      "Best error :  3.849872589111328 \n",
      "\n",
      "Epoch  167\n",
      "Top three least errors :  3.849872589111328 3.850564956665039 3.862025737762451\n",
      "Best error :  3.849872589111328 \n",
      "\n",
      "Epoch  168\n",
      "Top three least errors :  3.849872589111328 3.850564956665039 3.8555617332458496\n",
      "Best error :  3.8487772941589355 \n",
      "\n",
      "Epoch  169\n",
      "Top three least errors :  3.8487772941589355 3.849872589111328 3.850564956665039\n",
      "Best error :  3.8460512161254883 \n",
      "\n",
      "Epoch  170\n",
      "Top three least errors :  3.8460512161254883 3.846799373626709 3.848372220993042\n",
      "Best error :  3.8345839977264404 \n",
      "\n",
      "Epoch  171\n",
      "Top three least errors :  3.8345839977264404 3.836172342300415 3.8406941890716553\n",
      "Best error :  3.8345839977264404 \n",
      "\n",
      "Epoch  172\n",
      "Top three least errors :  3.8345839977264404 3.8346478939056396 3.835197687149048\n",
      "Best error :  3.807809352874756 \n",
      "\n",
      "Epoch  173\n",
      "Top three least errors :  3.807809352874756 3.8295254707336426 3.8345839977264404\n",
      "Best error :  3.8007700443267822 \n",
      "\n",
      "Epoch  174\n",
      "Top three least errors :  3.8007700443267822 3.807809352874756 3.81146502494812\n",
      "Best error :  3.7890076637268066 \n",
      "\n",
      "Epoch  175\n",
      "Top three least errors :  3.7890076637268066 3.8002734184265137 3.8007700443267822\n",
      "Best error :  3.7890076637268066 \n",
      "\n",
      "Epoch  176\n",
      "Top three least errors :  3.7890076637268066 3.7890989780426025 3.791085720062256\n",
      "Best error :  3.7817063331604004 \n",
      "\n",
      "Epoch  177\n",
      "Top three least errors :  3.7817063331604004 3.7819855213165283 3.7890076637268066\n",
      "Best error :  3.7683703899383545 \n",
      "\n",
      "Epoch  178\n",
      "Top three least errors :  3.7683703899383545 3.780141592025757 3.781601667404175\n",
      "Best error :  3.7267558574676514 \n",
      "\n",
      "Epoch  179\n",
      "Top three least errors :  3.7267558574676514 3.746062994003296 3.7524008750915527\n",
      "Best error :  3.7267558574676514 \n",
      "\n",
      "Epoch  180\n",
      "Top three least errors :  3.7267558574676514 3.742929458618164 3.746062994003296\n",
      "Best error :  3.7127537727355957 \n",
      "\n",
      "Epoch  181\n",
      "Top three least errors :  3.7127537727355957 3.723658561706543 3.7267558574676514\n",
      "Best error :  3.7099802494049072 \n",
      "\n",
      "Epoch  182\n",
      "Top three least errors :  3.7099802494049072 3.7127537727355957 3.714611291885376\n",
      "Best error :  3.6829426288604736 \n",
      "\n",
      "Epoch  183\n",
      "Top three least errors :  3.6829426288604736 3.7077786922454834 3.7099802494049072\n",
      "Best error :  3.6762735843658447 \n",
      "\n",
      "Epoch  184\n",
      "Top three least errors :  3.6762735843658447 3.6829426288604736 3.691504716873169\n",
      "Best error :  3.670279026031494 \n",
      "\n",
      "Epoch  185\n",
      "Top three least errors :  3.670279026031494 3.671581268310547 3.6740498542785645\n",
      "Best error :  3.663794994354248 \n",
      "\n",
      "Epoch  186\n",
      "Top three least errors :  3.663794994354248 3.670279026031494 3.671509265899658\n",
      "Best error :  3.6614177227020264 \n",
      "\n",
      "Epoch  187\n",
      "Top three least errors :  3.6614177227020264 3.663794994354248 3.6641290187835693\n",
      "Best error :  3.6583611965179443 \n",
      "\n",
      "Epoch  188\n",
      "Top three least errors :  3.6583611965179443 3.6614177227020264 3.663794994354248\n",
      "Best error :  3.657107353210449 \n",
      "\n",
      "Epoch  189\n",
      "Top three least errors :  3.657107353210449 3.6583611965179443 3.6614177227020264\n",
      "Best error :  3.6501071453094482 \n",
      "\n",
      "Epoch  190\n",
      "Top three least errors :  3.6501071453094482 3.6518361568450928 3.652224063873291\n",
      "Best error :  3.6501071453094482 \n",
      "\n",
      "Epoch  191\n",
      "Top three least errors :  3.6501071453094482 3.6515958309173584 3.6518361568450928\n",
      "Best error :  3.6470961570739746 \n",
      "\n",
      "Epoch  192\n",
      "Top three least errors :  3.6470961570739746 3.6501071453094482 3.6509921550750732\n",
      "Best error :  3.6409363746643066 \n",
      "\n",
      "Epoch  193\n",
      "Top three least errors :  3.6409363746643066 3.6457138061523438 3.6470961570739746\n",
      "Best error :  3.6252381801605225 \n",
      "\n",
      "Epoch  194\n",
      "Top three least errors :  3.6252381801605225 3.6409363746643066 3.642282485961914\n",
      "Best error :  3.613158702850342 \n",
      "\n",
      "Epoch  195\n",
      "Top three least errors :  3.613158702850342 3.6178693771362305 3.6232423782348633\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  196\n",
      "Top three least errors :  3.5718491077423096 3.6127259731292725 3.613158702850342\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  197\n",
      "Top three least errors :  3.5718491077423096 3.6127259731292725 3.613158702850342\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  198\n",
      "Top three least errors :  3.5718491077423096 3.6127259731292725 3.613158702850342\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  199\n",
      "Top three least errors :  3.5718491077423096 3.590285062789917 3.5988333225250244\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  200\n",
      "Top three least errors :  3.5718491077423096 3.574700355529785 3.590285062789917\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  201\n",
      "Top three least errors :  3.5718491077423096 3.574700355529785 3.590285062789917\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  202\n",
      "Top three least errors :  3.5718491077423096 3.574700355529785 3.580674171447754\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  203\n",
      "Top three least errors :  3.5718491077423096 3.574700355529785 3.580674171447754\n",
      "Best error :  3.5718491077423096 \n",
      "\n",
      "Epoch  204\n",
      "Top three least errors :  3.5718491077423096 3.574700355529785 3.580674171447754\n",
      "Best error :  3.5608487129211426 \n",
      "\n",
      "Epoch  205\n",
      "Top three least errors :  3.5608487129211426 3.5718491077423096 3.573890209197998\n",
      "Best error :  3.5608487129211426 \n",
      "\n",
      "Epoch  206\n",
      "Top three least errors :  3.5608487129211426 3.5683159828186035 3.573890209197998\n",
      "Best error :  3.5608487129211426 \n",
      "\n",
      "Epoch  207\n",
      "Top three least errors :  3.5608487129211426 3.56683611869812 3.5683159828186035\n",
      "Best error :  3.5608487129211426 \n",
      "\n",
      "Epoch  208\n",
      "Top three least errors :  3.5608487129211426 3.5616443157196045 3.5638673305511475\n",
      "Best error :  3.5608487129211426 \n",
      "\n",
      "Epoch  209\n",
      "Top three least errors :  3.5608487129211426 3.5616443157196045 3.5635666847229004\n",
      "Best error :  3.5596187114715576 \n",
      "\n",
      "Epoch  210\n",
      "Top three least errors :  3.5596187114715576 3.5608487129211426 3.5616443157196045\n",
      "Best error :  3.5580694675445557 \n",
      "\n",
      "Epoch  211\n",
      "Top three least errors :  3.5580694675445557 3.55947208404541 3.5596187114715576\n",
      "Best error :  3.557469129562378 \n",
      "\n",
      "Epoch  212\n",
      "Top three least errors :  3.557469129562378 3.5580694675445557 3.5583372116088867\n",
      "Best error :  3.557469129562378 \n",
      "\n",
      "Epoch  213\n",
      "Top three least errors :  3.557469129562378 3.557654619216919 3.5580694675445557\n",
      "Best error :  3.5544707775115967 \n",
      "\n",
      "Epoch  214\n",
      "Top three least errors :  3.5544707775115967 3.557469129562378 3.557654619216919\n",
      "Best error :  3.5544707775115967 \n",
      "\n",
      "Epoch  215\n",
      "Top three least errors :  3.5544707775115967 3.557469129562378 3.557654619216919\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  216\n",
      "Top three least errors :  3.538555383682251 3.5544707775115967 3.557469129562378\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  217\n",
      "Top three least errors :  3.538555383682251 3.5516374111175537 3.5544707775115967\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  218\n",
      "Top three least errors :  3.538555383682251 3.5512046813964844 3.5516374111175537\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  219\n",
      "Top three least errors :  3.538555383682251 3.548755407333374 3.5512046813964844\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  220\n",
      "Top three least errors :  3.538555383682251 3.54193377494812 3.545076608657837\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  221\n",
      "Top three least errors :  3.538555383682251 3.54193377494812 3.543835163116455\n",
      "Best error :  3.538555383682251 \n",
      "\n",
      "Epoch  222\n",
      "Top three least errors :  3.538555383682251 3.539313554763794 3.539613723754883\n",
      "Best error :  3.529594898223877 \n",
      "\n",
      "Epoch  223\n",
      "Top three least errors :  3.529594898223877 3.5353996753692627 3.538555383682251\n",
      "Best error :  3.529594898223877 \n",
      "\n",
      "Epoch  224\n",
      "Top three least errors :  3.529594898223877 3.5317413806915283 3.5353996753692627\n",
      "Best error :  3.5208663940429688 \n",
      "\n",
      "Epoch  225\n",
      "Top three least errors :  3.5208663940429688 3.5274198055267334 3.529594898223877\n",
      "Best error :  3.5197200775146484 \n",
      "\n",
      "Epoch  226\n",
      "Top three least errors :  3.5197200775146484 3.520608901977539 3.5208663940429688\n",
      "Best error :  3.5193123817443848 \n",
      "\n",
      "Epoch  227\n",
      "Top three least errors :  3.5193123817443848 3.5197200775146484 3.520608901977539\n",
      "Best error :  3.5193123817443848 \n",
      "\n",
      "Epoch  228\n",
      "Top three least errors :  3.5193123817443848 3.5196568965911865 3.5197200775146484\n",
      "Best error :  3.5193123817443848 \n",
      "\n",
      "Epoch  229\n",
      "Top three least errors :  3.5193123817443848 3.5196568965911865 3.5197200775146484\n",
      "Best error :  3.511364221572876 \n",
      "\n",
      "Epoch  230\n",
      "Top three least errors :  3.511364221572876 3.5162787437438965 3.5193123817443848\n",
      "Best error :  3.511364221572876 \n",
      "\n",
      "Epoch  231\n",
      "Top three least errors :  3.511364221572876 3.5162787437438965 3.5193123817443848\n",
      "Best error :  3.511364221572876 \n",
      "\n",
      "Epoch  232\n",
      "Top three least errors :  3.511364221572876 3.5126261711120605 3.5162787437438965\n",
      "Best error :  3.509819269180298 \n",
      "\n",
      "Epoch  233\n",
      "Top three least errors :  3.509819269180298 3.511098861694336 3.511364221572876\n",
      "Best error :  3.50840163230896 \n",
      "\n",
      "Epoch  234\n",
      "Top three least errors :  3.50840163230896 3.5096445083618164 3.509819269180298\n",
      "Best error :  3.50840163230896 \n",
      "\n",
      "Epoch  235\n",
      "Top three least errors :  3.50840163230896 3.5090885162353516 3.5096445083618164\n",
      "Best error :  3.503288745880127 \n",
      "\n",
      "Epoch  236\n",
      "Top three least errors :  3.503288745880127 3.50840163230896 3.5090885162353516\n",
      "Best error :  3.4987599849700928 \n",
      "\n",
      "Epoch  237\n",
      "Top three least errors :  3.4987599849700928 3.5021955966949463 3.5025880336761475\n",
      "Best error :  3.4986982345581055 \n",
      "\n",
      "Epoch  238\n",
      "Top three least errors :  3.4986982345581055 3.4987599849700928 3.5005674362182617\n",
      "Best error :  3.4967143535614014 \n",
      "\n",
      "Epoch  239\n",
      "Top three least errors :  3.4967143535614014 3.49703049659729 3.4986982345581055\n",
      "Best error :  3.4967143535614014 \n",
      "\n",
      "Epoch  240\n",
      "Top three least errors :  3.4967143535614014 3.49703049659729 3.4985432624816895\n",
      "Best error :  3.494994640350342 \n",
      "\n",
      "Epoch  241\n",
      "Top three least errors :  3.494994640350342 3.496354579925537 3.4967143535614014\n",
      "Best error :  3.482875108718872 \n",
      "\n",
      "Epoch  242\n",
      "Top three least errors :  3.482875108718872 3.494994640350342 3.4951281547546387\n",
      "Best error :  3.482875108718872 \n",
      "\n",
      "Epoch  243\n",
      "Top three least errors :  3.482875108718872 3.4831063747406006 3.4871973991394043\n",
      "Best error :  3.482875108718872 \n",
      "\n",
      "Epoch  244\n",
      "Top three least errors :  3.482875108718872 3.4831063747406006 3.4833648204803467\n",
      "Best error :  3.482875108718872 \n",
      "\n",
      "Epoch  245\n",
      "Top three least errors :  3.482875108718872 3.483062744140625 3.4831063747406006\n",
      "Best error :  3.4820356369018555 \n",
      "\n",
      "Epoch  246\n",
      "Top three least errors :  3.4820356369018555 3.482224941253662 3.4823660850524902\n",
      "Best error :  3.4820356369018555 \n",
      "\n",
      "Epoch  247\n",
      "Top three least errors :  3.4820356369018555 3.482224941253662 3.4823660850524902\n",
      "Best error :  3.4807777404785156 \n",
      "\n",
      "Epoch  248\n",
      "Top three least errors :  3.4807777404785156 3.4820356369018555 3.482224941253662\n",
      "Best error :  3.4772467613220215 \n",
      "\n",
      "Epoch  249\n",
      "Top three least errors :  3.4772467613220215 3.4789652824401855 3.4807777404785156\n",
      "Best error :  3.4719491004943848 \n",
      "\n",
      "Epoch  250\n",
      "Top three least errors :  3.4719491004943848 3.474337339401245 3.4751057624816895\n",
      "Best error :  3.4719491004943848 \n",
      "\n",
      "Epoch  251\n",
      "Top three least errors :  3.4719491004943848 3.474337339401245 3.4751057624816895\n",
      "Best error :  3.4673986434936523 \n",
      "\n",
      "Epoch  252\n",
      "Top three least errors :  3.4673986434936523 3.471346616744995 3.4719491004943848\n",
      "Best error :  3.4630775451660156 \n",
      "\n",
      "Epoch  253\n",
      "Top three least errors :  3.4630775451660156 3.4673986434936523 3.471346616744995\n",
      "Best error :  3.4630775451660156 \n",
      "\n",
      "Epoch  254\n",
      "Top three least errors :  3.4630775451660156 3.4673986434936523 3.468762159347534\n",
      "Best error :  3.4562714099884033 \n",
      "\n",
      "Epoch  255\n",
      "Top three least errors :  3.4562714099884033 3.45910382270813 3.4614744186401367\n",
      "Best error :  3.449679374694824 \n",
      "\n",
      "Epoch  256\n",
      "Top three least errors :  3.449679374694824 3.4562714099884033 3.45910382270813\n",
      "Best error :  3.4487993717193604 \n",
      "\n",
      "Epoch  257\n",
      "Top three least errors :  3.4487993717193604 3.449679374694824 3.4533445835113525\n",
      "Best error :  3.4439632892608643 \n",
      "\n",
      "Epoch  258\n",
      "Top three least errors :  3.4439632892608643 3.4475104808807373 3.4487993717193604\n",
      "Best error :  3.442563772201538 \n",
      "\n",
      "Epoch  259\n",
      "Top three least errors :  3.442563772201538 3.4439632892608643 3.448460578918457\n",
      "Best error :  3.4375100135803223 \n",
      "\n",
      "Epoch  260\n",
      "Top three least errors :  3.4375100135803223 3.4400713443756104 3.4424891471862793\n",
      "Best error :  3.4329094886779785 \n",
      "\n",
      "Epoch  261\n",
      "Top three least errors :  3.4329094886779785 3.436635732650757 3.4375100135803223\n",
      "Best error :  3.425168037414551 \n",
      "\n",
      "Epoch  262\n",
      "Top three least errors :  3.425168037414551 3.432518720626831 3.4329094886779785\n",
      "Best error :  3.423433780670166 \n",
      "\n",
      "Epoch  263\n",
      "Top three least errors :  3.423433780670166 3.425168037414551 3.426192045211792\n",
      "Best error :  3.422199249267578 \n",
      "\n",
      "Epoch  264\n",
      "Top three least errors :  3.422199249267578 3.423433780670166 3.424152374267578\n",
      "Best error :  3.4183974266052246 \n",
      "\n",
      "Epoch  265\n",
      "Top three least errors :  3.4183974266052246 3.419523239135742 3.422199249267578\n",
      "Best error :  3.4183974266052246 \n",
      "\n",
      "Epoch  266\n",
      "Top three least errors :  3.4183974266052246 3.419523239135742 3.420978546142578\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.4183974266052246 \n",
      "\n",
      "Epoch  267\n",
      "Top three least errors :  3.4183974266052246 3.419523239135742 3.420978546142578\n",
      "Best error :  3.4157228469848633 \n",
      "\n",
      "Epoch  268\n",
      "Top three least errors :  3.4157228469848633 3.4182567596435547 3.4183974266052246\n",
      "Best error :  3.4157228469848633 \n",
      "\n",
      "Epoch  269\n",
      "Top three least errors :  3.4157228469848633 3.415984869003296 3.416800022125244\n",
      "Best error :  3.413149118423462 \n",
      "\n",
      "Epoch  270\n",
      "Top three least errors :  3.413149118423462 3.4150967597961426 3.4152345657348633\n",
      "Best error :  3.4118170738220215 \n",
      "\n",
      "Epoch  271\n",
      "Top three least errors :  3.4118170738220215 3.413149118423462 3.41438627243042\n",
      "Best error :  3.4118170738220215 \n",
      "\n",
      "Epoch  272\n",
      "Top three least errors :  3.4118170738220215 3.413149118423462 3.41438627243042\n",
      "Best error :  3.4118170738220215 \n",
      "\n",
      "Epoch  273\n",
      "Top three least errors :  3.4118170738220215 3.4118785858154297 3.413149118423462\n",
      "Best error :  3.409968137741089 \n",
      "\n",
      "Epoch  274\n",
      "Top three least errors :  3.409968137741089 3.4118170738220215 3.4118785858154297\n",
      "Best error :  3.4088516235351562 \n",
      "\n",
      "Epoch  275\n",
      "Top three least errors :  3.4088516235351562 3.409205675125122 3.409968137741089\n",
      "Best error :  3.407595157623291 \n",
      "\n",
      "Epoch  276\n",
      "Top three least errors :  3.407595157623291 3.407916784286499 3.4088516235351562\n",
      "Best error :  3.407595157623291 \n",
      "\n",
      "Epoch  277\n",
      "Top three least errors :  3.407595157623291 3.407916784286499 3.4088516235351562\n",
      "Best error :  3.406000852584839 \n",
      "\n",
      "Epoch  278\n",
      "Top three least errors :  3.406000852584839 3.4075374603271484 3.4075658321380615\n",
      "Best error :  3.397369146347046 \n",
      "\n",
      "Epoch  279\n",
      "Top three least errors :  3.397369146347046 3.4033455848693848 3.406000852584839\n",
      "Best error :  3.395489454269409 \n",
      "\n",
      "Epoch  280\n",
      "Top three least errors :  3.395489454269409 3.3957204818725586 3.3958473205566406\n",
      "Best error :  3.39452862739563 \n",
      "\n",
      "Epoch  281\n",
      "Top three least errors :  3.39452862739563 3.395489454269409 3.3957204818725586\n",
      "Best error :  3.39452862739563 \n",
      "\n",
      "Epoch  282\n",
      "Top three least errors :  3.39452862739563 3.395489454269409 3.3957204818725586\n",
      "Best error :  3.39314341545105 \n",
      "\n",
      "Epoch  283\n",
      "Top three least errors :  3.39314341545105 3.3935298919677734 3.393690347671509\n",
      "Best error :  3.391036033630371 \n",
      "\n",
      "Epoch  284\n",
      "Top three least errors :  3.391036033630371 3.393082618713379 3.39314341545105\n",
      "Best error :  3.3901896476745605 \n",
      "\n",
      "Epoch  285\n",
      "Top three least errors :  3.3901896476745605 3.391036033630371 3.391526699066162\n",
      "Best error :  3.3893237113952637 \n",
      "\n",
      "Epoch  286\n",
      "Top three least errors :  3.3893237113952637 3.389660596847534 3.3898701667785645\n",
      "Best error :  3.3858425617218018 \n",
      "\n",
      "Epoch  287\n",
      "Top three least errors :  3.3858425617218018 3.387115716934204 3.3885436058044434\n",
      "Best error :  3.3810794353485107 \n",
      "\n",
      "Epoch  288\n",
      "Top three least errors :  3.3810794353485107 3.3858425617218018 3.386449098587036\n",
      "Best error :  3.379709243774414 \n",
      "\n",
      "Epoch  289\n",
      "Top three least errors :  3.379709243774414 3.3810794353485107 3.3832521438598633\n",
      "Best error :  3.379709243774414 \n",
      "\n",
      "Epoch  290\n",
      "Top three least errors :  3.379709243774414 3.3810794353485107 3.383195638656616\n",
      "Best error :  3.379709243774414 \n",
      "\n",
      "Epoch  291\n",
      "Top three least errors :  3.379709243774414 3.3800578117370605 3.3810794353485107\n",
      "Best error :  3.3740298748016357 \n",
      "\n",
      "Epoch  292\n",
      "Top three least errors :  3.3740298748016357 3.3788230419158936 3.379709243774414\n",
      "Best error :  3.3681423664093018 \n",
      "\n",
      "Epoch  293\n",
      "Top three least errors :  3.3681423664093018 3.3712925910949707 3.371892213821411\n",
      "Best error :  3.367433786392212 \n",
      "\n",
      "Epoch  294\n",
      "Top three least errors :  3.367433786392212 3.3681423664093018 3.3690474033355713\n",
      "Best error :  3.367356061935425 \n",
      "\n",
      "Epoch  295\n",
      "Top three least errors :  3.367356061935425 3.367433786392212 3.367661714553833\n",
      "Best error :  3.3664872646331787 \n",
      "\n",
      "Epoch  296\n",
      "Top three least errors :  3.3664872646331787 3.3667967319488525 3.3672900199890137\n",
      "Best error :  3.3664872646331787 \n",
      "\n",
      "Epoch  297\n",
      "Top three least errors :  3.3664872646331787 3.3666903972625732 3.366727590560913\n",
      "Best error :  3.3664236068725586 \n",
      "\n",
      "Epoch  298\n",
      "Top three least errors :  3.3664236068725586 3.3664872646331787 3.3666903972625732\n",
      "Best error :  3.3623385429382324 \n",
      "\n",
      "Epoch  299\n",
      "Top three least errors :  3.3623385429382324 3.3654873371124268 3.3660082817077637\n",
      "Best error :  3.358001708984375 \n",
      "\n",
      "Epoch  300\n",
      "Top three least errors :  3.358001708984375 3.3587777614593506 3.360521078109741\n",
      "Best error :  3.3541462421417236 \n",
      "\n",
      "Epoch  301\n",
      "Top three least errors :  3.3541462421417236 3.3563530445098877 3.3565115928649902\n",
      "Best error :  3.3539540767669678 \n",
      "\n",
      "Epoch  302\n",
      "Top three least errors :  3.3539540767669678 3.3541462421417236 3.3548455238342285\n",
      "Best error :  3.3522396087646484 \n",
      "\n",
      "Epoch  303\n",
      "Top three least errors :  3.3522396087646484 3.352877378463745 3.353130578994751\n",
      "Best error :  3.3509409427642822 \n",
      "\n",
      "Epoch  304\n",
      "Top three least errors :  3.3509409427642822 3.35101580619812 3.351757287979126\n",
      "Best error :  3.350480318069458 \n",
      "\n",
      "Epoch  305\n",
      "Top three least errors :  3.350480318069458 3.3509409427642822 3.35101580619812\n",
      "Best error :  3.349180221557617 \n",
      "\n",
      "Epoch  306\n",
      "Top three least errors :  3.349180221557617 3.349701166152954 3.349876642227173\n",
      "Best error :  3.347106695175171 \n",
      "\n",
      "Epoch  307\n",
      "Top three least errors :  3.347106695175171 3.3483006954193115 3.349180221557617\n",
      "Best error :  3.346848726272583 \n",
      "\n",
      "Epoch  308\n",
      "Top three least errors :  3.346848726272583 3.347106695175171 3.3471767902374268\n",
      "Best error :  3.3447492122650146 \n",
      "\n",
      "Epoch  309\n",
      "Top three least errors :  3.3447492122650146 3.346144914627075 3.346848726272583\n",
      "Best error :  3.3447492122650146 \n",
      "\n",
      "Epoch  310\n",
      "Top three least errors :  3.3447492122650146 3.3455939292907715 3.345632791519165\n",
      "Best error :  3.3437626361846924 \n",
      "\n",
      "Epoch  311\n",
      "Top three least errors :  3.3437626361846924 3.3442037105560303 3.3447492122650146\n",
      "Best error :  3.3432722091674805 \n",
      "\n",
      "Epoch  312\n",
      "Top three least errors :  3.3432722091674805 3.343647003173828 3.3437068462371826\n",
      "Best error :  3.3406076431274414 \n",
      "\n",
      "Epoch  313\n",
      "Top three least errors :  3.3406076431274414 3.3412668704986572 3.3427743911743164\n",
      "Best error :  3.3394477367401123 \n",
      "\n",
      "Epoch  314\n",
      "Top three least errors :  3.3394477367401123 3.3406076431274414 3.3412668704986572\n",
      "Best error :  3.3375139236450195 \n",
      "\n",
      "Epoch  315\n",
      "Top three least errors :  3.3375139236450195 3.3394477367401123 3.3406076431274414\n",
      "Best error :  3.335529327392578 \n",
      "\n",
      "Epoch  316\n",
      "Top three least errors :  3.335529327392578 3.3375139236450195 3.3381664752960205\n",
      "Best error :  3.335529327392578 \n",
      "\n",
      "Epoch  317\n",
      "Top three least errors :  3.335529327392578 3.3373515605926514 3.3375139236450195\n",
      "Best error :  3.3289639949798584 \n",
      "\n",
      "Epoch  318\n",
      "Top three least errors :  3.3289639949798584 3.3347277641296387 3.3354246616363525\n",
      "Best error :  3.3289639949798584 \n",
      "\n",
      "Epoch  319\n",
      "Top three least errors :  3.3289639949798584 3.3293354511260986 3.3316142559051514\n",
      "Best error :  3.328547239303589 \n",
      "\n",
      "Epoch  320\n",
      "Top three least errors :  3.328547239303589 3.3286314010620117 3.3289639949798584\n",
      "Best error :  3.325941801071167 \n",
      "\n",
      "Epoch  321\n",
      "Top three least errors :  3.325941801071167 3.3283779621124268 3.328547239303589\n",
      "Best error :  3.323312520980835 \n",
      "\n",
      "Epoch  322\n",
      "Top three least errors :  3.323312520980835 3.324680805206299 3.325112819671631\n",
      "Best error :  3.3222880363464355 \n",
      "\n",
      "Epoch  323\n",
      "Top three least errors :  3.3222880363464355 3.3222944736480713 3.322781562805176\n",
      "Best error :  3.31986665725708 \n",
      "\n",
      "Epoch  324\n",
      "Top three least errors :  3.31986665725708 3.3201611042022705 3.3222880363464355\n",
      "Best error :  3.3195204734802246 \n",
      "\n",
      "Epoch  325\n",
      "Top three least errors :  3.3195204734802246 3.319732666015625 3.31986665725708\n",
      "Best error :  3.317521095275879 \n",
      "\n",
      "Epoch  326\n",
      "Top three least errors :  3.317521095275879 3.3183224201202393 3.3195204734802246\n",
      "Best error :  3.3165721893310547 \n",
      "\n",
      "Epoch  327\n",
      "Top three least errors :  3.3165721893310547 3.317521095275879 3.3180930614471436\n",
      "Best error :  3.31573224067688 \n",
      "\n",
      "Epoch  328\n",
      "Top three least errors :  3.31573224067688 3.3165109157562256 3.3165721893310547\n",
      "Best error :  3.3142549991607666 \n",
      "\n",
      "Epoch  329\n",
      "Top three least errors :  3.3142549991607666 3.31573224067688 3.315857172012329\n",
      "Best error :  3.3141181468963623 \n",
      "\n",
      "Epoch  330\n",
      "Top three least errors :  3.3141181468963623 3.314239025115967 3.3142549991607666\n",
      "Best error :  3.313403844833374 \n",
      "\n",
      "Epoch  331\n",
      "Top three least errors :  3.313403844833374 3.3140361309051514 3.3141181468963623\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.312720775604248 \n",
      "\n",
      "Epoch  332\n",
      "Top three least errors :  3.312720775604248 3.3129372596740723 3.313403844833374\n",
      "Best error :  3.312103748321533 \n",
      "\n",
      "Epoch  333\n",
      "Top three least errors :  3.312103748321533 3.312720775604248 3.3129372596740723\n",
      "Best error :  3.3120803833007812 \n",
      "\n",
      "Epoch  334\n",
      "Top three least errors :  3.3120803833007812 3.312103748321533 3.3129372596740723\n",
      "Best error :  3.3115711212158203 \n",
      "\n",
      "Epoch  335\n",
      "Top three least errors :  3.3115711212158203 3.3119192123413086 3.3120803833007812\n",
      "Best error :  3.3103559017181396 \n",
      "\n",
      "Epoch  336\n",
      "Top three least errors :  3.3103559017181396 3.3111469745635986 3.3113489151000977\n",
      "Best error :  3.3100156784057617 \n",
      "\n",
      "Epoch  337\n",
      "Top three least errors :  3.3100156784057617 3.3103559017181396 3.310577392578125\n",
      "Best error :  3.3100156784057617 \n",
      "\n",
      "Epoch  338\n",
      "Top three least errors :  3.3100156784057617 3.3103559017181396 3.310577392578125\n",
      "Best error :  3.303936719894409 \n",
      "\n",
      "Epoch  339\n",
      "Top three least errors :  3.303936719894409 3.3086962699890137 3.309246301651001\n",
      "Best error :  3.301373243331909 \n",
      "\n",
      "Epoch  340\n",
      "Top three least errors :  3.301373243331909 3.303936719894409 3.3085739612579346\n",
      "Best error :  3.2986059188842773 \n",
      "\n",
      "Epoch  341\n",
      "Top three least errors :  3.2986059188842773 3.2995548248291016 3.301373243331909\n",
      "Best error :  3.2983686923980713 \n",
      "\n",
      "Epoch  342\n",
      "Top three least errors :  3.2983686923980713 3.2986059188842773 3.2995548248291016\n",
      "Best error :  3.297971248626709 \n",
      "\n",
      "Epoch  343\n",
      "Top three least errors :  3.297971248626709 3.2983686923980713 3.2986059188842773\n",
      "Best error :  3.296394109725952 \n",
      "\n",
      "Epoch  344\n",
      "Top three least errors :  3.296394109725952 3.2976794242858887 3.297971248626709\n",
      "Best error :  3.2912344932556152 \n",
      "\n",
      "Epoch  345\n",
      "Top three least errors :  3.2912344932556152 3.2950143814086914 3.296394109725952\n",
      "Best error :  3.290295124053955 \n",
      "\n",
      "Epoch  346\n",
      "Top three least errors :  3.290295124053955 3.2912344932556152 3.2950143814086914\n",
      "Best error :  3.2876081466674805 \n",
      "\n",
      "Epoch  347\n",
      "Top three least errors :  3.2876081466674805 3.2879247665405273 3.290295124053955\n",
      "Best error :  3.286942958831787 \n",
      "\n",
      "Epoch  348\n",
      "Top three least errors :  3.286942958831787 3.2875735759735107 3.2876081466674805\n",
      "Best error :  3.2860190868377686 \n",
      "\n",
      "Epoch  349\n",
      "Top three least errors :  3.2860190868377686 3.286942958831787 3.2870545387268066\n",
      "Best error :  3.2860190868377686 \n",
      "\n",
      "Epoch  350\n",
      "Top three least errors :  3.2860190868377686 3.286942958831787 3.2870545387268066\n",
      "Best error :  3.2854933738708496 \n",
      "\n",
      "Epoch  351\n",
      "Top three least errors :  3.2854933738708496 3.2860190868377686 3.2868919372558594\n",
      "Best error :  3.2839813232421875 \n",
      "\n",
      "Epoch  352\n",
      "Top three least errors :  3.2839813232421875 3.2854933738708496 3.2860190868377686\n",
      "Best error :  3.282346487045288 \n",
      "\n",
      "Epoch  353\n",
      "Top three least errors :  3.282346487045288 3.283198595046997 3.2836008071899414\n",
      "Best error :  3.282346487045288 \n",
      "\n",
      "Epoch  354\n",
      "Top three least errors :  3.282346487045288 3.2827203273773193 3.2836008071899414\n",
      "Best error :  3.2821974754333496 \n",
      "\n",
      "Epoch  355\n",
      "Top three least errors :  3.2821974754333496 3.282346487045288 3.2827203273773193\n",
      "Best error :  3.2796008586883545 \n",
      "\n",
      "Epoch  356\n",
      "Top three least errors :  3.2796008586883545 3.281571626663208 3.2815816402435303\n",
      "Best error :  3.2779271602630615 \n",
      "\n",
      "Epoch  357\n",
      "Top three least errors :  3.2779271602630615 3.2790956497192383 3.279209613800049\n",
      "Best error :  3.2778961658477783 \n",
      "\n",
      "Epoch  358\n",
      "Top three least errors :  3.2778961658477783 3.2779271602630615 3.278614044189453\n",
      "Best error :  3.2778961658477783 \n",
      "\n",
      "Epoch  359\n",
      "Top three least errors :  3.2778961658477783 3.2779271602630615 3.278614044189453\n",
      "Best error :  3.2778961658477783 \n",
      "\n",
      "Epoch  360\n",
      "Top three least errors :  3.2778961658477783 3.2779271602630615 3.278614044189453\n",
      "Best error :  3.2778961658477783 \n",
      "\n",
      "Epoch  361\n",
      "Top three least errors :  3.2778961658477783 3.2779271602630615 3.278614044189453\n",
      "Best error :  3.2778563499450684 \n",
      "\n",
      "Epoch  362\n",
      "Top three least errors :  3.2778563499450684 3.2778961658477783 3.2779271602630615\n",
      "Best error :  3.277743101119995 \n",
      "\n",
      "Epoch  363\n",
      "Top three least errors :  3.277743101119995 3.27779483795166 3.2778563499450684\n",
      "Best error :  3.2777366638183594 \n",
      "\n",
      "Epoch  364\n",
      "Top three least errors :  3.2777366638183594 3.277743101119995 3.27779483795166\n",
      "Best error :  3.2777366638183594 \n",
      "\n",
      "Epoch  365\n",
      "Top three least errors :  3.2777366638183594 3.277743101119995 3.27779483795166\n",
      "Best error :  3.277115821838379 \n",
      "\n",
      "Epoch  366\n",
      "Top three least errors :  3.277115821838379 3.2777278423309326 3.2777366638183594\n",
      "Best error :  3.27404522895813 \n",
      "\n",
      "Epoch  367\n",
      "Top three least errors :  3.27404522895813 3.276930332183838 3.2770485877990723\n",
      "Best error :  3.2727181911468506 \n",
      "\n",
      "Epoch  368\n",
      "Top three least errors :  3.2727181911468506 3.27404522895813 3.274109363555908\n",
      "Best error :  3.268681526184082 \n",
      "\n",
      "Epoch  369\n",
      "Top three least errors :  3.268681526184082 3.2712650299072266 3.2721776962280273\n",
      "Best error :  3.2675654888153076 \n",
      "\n",
      "Epoch  370\n",
      "Top three least errors :  3.2675654888153076 3.2677602767944336 3.2684903144836426\n",
      "Best error :  3.266550064086914 \n",
      "\n",
      "Epoch  371\n",
      "Top three least errors :  3.266550064086914 3.2669999599456787 3.2675654888153076\n",
      "Best error :  3.2661473751068115 \n",
      "\n",
      "Epoch  372\n",
      "Top three least errors :  3.2661473751068115 3.266550064086914 3.2669999599456787\n",
      "Best error :  3.2657437324523926 \n",
      "\n",
      "Epoch  373\n",
      "Top three least errors :  3.2657437324523926 3.2657902240753174 3.265821933746338\n",
      "Best error :  3.2655727863311768 \n",
      "\n",
      "Epoch  374\n",
      "Top three least errors :  3.2655727863311768 3.2657437324523926 3.2657902240753174\n",
      "Best error :  3.2655727863311768 \n",
      "\n",
      "Epoch  375\n",
      "Top three least errors :  3.2655727863311768 3.2657437324523926 3.2657902240753174\n",
      "Best error :  3.2632956504821777 \n",
      "\n",
      "Epoch  376\n",
      "Top three least errors :  3.2632956504821777 3.264991044998169 3.2655727863311768\n",
      "Best error :  3.2610580921173096 \n",
      "\n",
      "Epoch  377\n",
      "Top three least errors :  3.2610580921173096 3.2628283500671387 3.2632956504821777\n",
      "Best error :  3.2592272758483887 \n",
      "\n",
      "Epoch  378\n",
      "Top three least errors :  3.2592272758483887 3.260422706604004 3.2610580921173096\n",
      "Best error :  3.2591428756713867 \n",
      "\n",
      "Epoch  379\n",
      "Top three least errors :  3.2591428756713867 3.2592272758483887 3.2592456340789795\n",
      "Best error :  3.2584476470947266 \n",
      "\n",
      "Epoch  380\n",
      "Top three least errors :  3.2584476470947266 3.258605480194092 3.258793592453003\n",
      "Best error :  3.257714033126831 \n",
      "\n",
      "Epoch  381\n",
      "Top three least errors :  3.257714033126831 3.2581231594085693 3.258363723754883\n",
      "Best error :  3.2553727626800537 \n",
      "\n",
      "Epoch  382\n",
      "Top three least errors :  3.2553727626800537 3.2570083141326904 3.2572693824768066\n",
      "Best error :  3.25477933883667 \n",
      "\n",
      "Epoch  383\n",
      "Top three least errors :  3.25477933883667 3.2553727626800537 3.255682945251465\n",
      "Best error :  3.254011392593384 \n",
      "\n",
      "Epoch  384\n",
      "Top three least errors :  3.254011392593384 3.2545266151428223 3.25477933883667\n",
      "Best error :  3.25345778465271 \n",
      "\n",
      "Epoch  385\n",
      "Top three least errors :  3.25345778465271 3.253699541091919 3.2538585662841797\n",
      "Best error :  3.250560998916626 \n",
      "\n",
      "Epoch  386\n",
      "Top three least errors :  3.250560998916626 3.252884864807129 3.253061294555664\n",
      "Best error :  3.248819351196289 \n",
      "\n",
      "Epoch  387\n",
      "Top three least errors :  3.248819351196289 3.2502801418304443 3.250560998916626\n",
      "Best error :  3.2468326091766357 \n",
      "\n",
      "Epoch  388\n",
      "Top three least errors :  3.2468326091766357 3.2474050521850586 3.248819351196289\n",
      "Best error :  3.2448644638061523 \n",
      "\n",
      "Epoch  389\n",
      "Top three least errors :  3.2448644638061523 3.2465856075286865 3.2467758655548096\n",
      "Best error :  3.2448644638061523 \n",
      "\n",
      "Epoch  390\n",
      "Top three least errors :  3.2448644638061523 3.24570894241333 3.245734214782715\n",
      "Best error :  3.242178201675415 \n",
      "\n",
      "Epoch  391\n",
      "Top three least errors :  3.242178201675415 3.244318962097168 3.244419813156128\n",
      "Best error :  3.234957695007324 \n",
      "\n",
      "Epoch  392\n",
      "Top three least errors :  3.234957695007324 3.242178201675415 3.2427215576171875\n",
      "Best error :  3.232121706008911 \n",
      "\n",
      "Epoch  393\n",
      "Top three least errors :  3.232121706008911 3.2346653938293457 3.234957695007324\n",
      "Best error :  3.232121706008911 \n",
      "\n",
      "Epoch  394\n",
      "Top three least errors :  3.232121706008911 3.2346653938293457 3.2347118854522705\n",
      "Best error :  3.232121706008911 \n",
      "\n",
      "Epoch  395\n",
      "Top three least errors :  3.232121706008911 3.2326650619506836 3.2346653938293457\n",
      "Best error :  3.232121706008911 \n",
      "\n",
      "Epoch  396\n",
      "Top three least errors :  3.232121706008911 3.232316732406616 3.2326650619506836\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.232121706008911 \n",
      "\n",
      "Epoch  397\n",
      "Top three least errors :  3.232121706008911 3.232184648513794 3.232274293899536\n",
      "Best error :  3.232116937637329 \n",
      "\n",
      "Epoch  398\n",
      "Top three least errors :  3.232116937637329 3.232121706008911 3.232164144515991\n",
      "Best error :  3.2312967777252197 \n",
      "\n",
      "Epoch  399\n",
      "Top three least errors :  3.2312967777252197 3.2315125465393066 3.2320361137390137\n",
      "Best error :  3.228309392929077 \n",
      "\n",
      "Epoch  400\n",
      "Top three least errors :  3.228309392929077 3.2308719158172607 3.2312967777252197\n",
      "Best error :  3.2229812145233154 \n",
      "\n",
      "Epoch  401\n",
      "Top three least errors :  3.2229812145233154 3.228309392929077 3.228646755218506\n",
      "Best error :  3.2229812145233154 \n",
      "\n",
      "Epoch  402\n",
      "Top three least errors :  3.2229812145233154 3.2243669033050537 3.228309392929077\n",
      "Best error :  3.2229812145233154 \n",
      "\n",
      "Epoch  403\n",
      "Top three least errors :  3.2229812145233154 3.223540782928467 3.2243669033050537\n",
      "Best error :  3.222561836242676 \n",
      "\n",
      "Epoch  404\n",
      "Top three least errors :  3.222561836242676 3.2229812145233154 3.223116159439087\n",
      "Best error :  3.222561836242676 \n",
      "\n",
      "Epoch  405\n",
      "Top three least errors :  3.222561836242676 3.2229297161102295 3.222956895828247\n",
      "Best error :  3.222381830215454 \n",
      "\n",
      "Epoch  406\n",
      "Top three least errors :  3.222381830215454 3.2225239276885986 3.222561836242676\n",
      "Best error :  3.222381830215454 \n",
      "\n",
      "Epoch  407\n",
      "Top three least errors :  3.222381830215454 3.2225239276885986 3.222561836242676\n",
      "Best error :  3.222280263900757 \n",
      "\n",
      "Epoch  408\n",
      "Top three least errors :  3.222280263900757 3.222381830215454 3.2225239276885986\n",
      "Best error :  3.2220396995544434 \n",
      "\n",
      "Epoch  409\n",
      "Top three least errors :  3.2220396995544434 3.222280263900757 3.222381830215454\n",
      "Best error :  3.2207119464874268 \n",
      "\n",
      "Epoch  410\n",
      "Top three least errors :  3.2207119464874268 3.221829652786255 3.2220396995544434\n",
      "Best error :  3.2207119464874268 \n",
      "\n",
      "Epoch  411\n",
      "Top three least errors :  3.2207119464874268 3.220905303955078 3.221384286880493\n",
      "Best error :  3.219007730484009 \n",
      "\n",
      "Epoch  412\n",
      "Top three least errors :  3.219007730484009 3.220088481903076 3.2205095291137695\n",
      "Best error :  3.2165472507476807 \n",
      "\n",
      "Epoch  413\n",
      "Top three least errors :  3.2165472507476807 3.219007730484009 3.2190425395965576\n",
      "Best error :  3.2159945964813232 \n",
      "\n",
      "Epoch  414\n",
      "Top three least errors :  3.2159945964813232 3.2165472507476807 3.2166779041290283\n",
      "Best error :  3.2146646976470947 \n",
      "\n",
      "Epoch  415\n",
      "Top three least errors :  3.2146646976470947 3.215301990509033 3.21592116355896\n",
      "Best error :  3.21195387840271 \n",
      "\n",
      "Epoch  416\n",
      "Top three least errors :  3.21195387840271 3.2135462760925293 3.2146646976470947\n",
      "Best error :  3.21195387840271 \n",
      "\n",
      "Epoch  417\n",
      "Top three least errors :  3.21195387840271 3.2135462760925293 3.2146646976470947\n",
      "Best error :  3.210641860961914 \n",
      "\n",
      "Epoch  418\n",
      "Top three least errors :  3.210641860961914 3.21195387840271 3.2122087478637695\n",
      "Best error :  3.210641860961914 \n",
      "\n",
      "Epoch  419\n",
      "Top three least errors :  3.210641860961914 3.2107834815979004 3.211487293243408\n",
      "Best error :  3.210641860961914 \n",
      "\n",
      "Epoch  420\n",
      "Top three least errors :  3.210641860961914 3.2107834815979004 3.210905075073242\n",
      "Best error :  3.210536241531372 \n",
      "\n",
      "Epoch  421\n",
      "Top three least errors :  3.210536241531372 3.2106378078460693 3.210641860961914\n",
      "Best error :  3.2101991176605225 \n",
      "\n",
      "Epoch  422\n",
      "Top three least errors :  3.2101991176605225 3.2103538513183594 3.2103593349456787\n",
      "Best error :  3.2075729370117188 \n",
      "\n",
      "Epoch  423\n",
      "Top three least errors :  3.2075729370117188 3.2101991176605225 3.2103538513183594\n",
      "Best error :  3.2063963413238525 \n",
      "\n",
      "Epoch  424\n",
      "Top three least errors :  3.2063963413238525 3.207263708114624 3.2075631618499756\n",
      "Best error :  3.204869270324707 \n",
      "\n",
      "Epoch  425\n",
      "Top three least errors :  3.204869270324707 3.205223798751831 3.205871105194092\n",
      "Best error :  3.200237512588501 \n",
      "\n",
      "Epoch  426\n",
      "Top three least errors :  3.200237512588501 3.2003540992736816 3.204869270324707\n",
      "Best error :  3.200237512588501 \n",
      "\n",
      "Epoch  427\n",
      "Top three least errors :  3.200237512588501 3.200326442718506 3.2003540992736816\n",
      "Best error :  3.1997547149658203 \n",
      "\n",
      "Epoch  428\n",
      "Top three least errors :  3.1997547149658203 3.2000277042388916 3.2000701427459717\n",
      "Best error :  3.1988720893859863 \n",
      "\n",
      "Epoch  429\n",
      "Top three least errors :  3.1988720893859863 3.1997547149658203 3.199954032897949\n",
      "Best error :  3.1988720893859863 \n",
      "\n",
      "Epoch  430\n",
      "Top three least errors :  3.1988720893859863 3.1997547149658203 3.199954032897949\n",
      "Best error :  3.1976470947265625 \n",
      "\n",
      "Epoch  431\n",
      "Top three least errors :  3.1976470947265625 3.1981043815612793 3.1985578536987305\n",
      "Best error :  3.1955533027648926 \n",
      "\n",
      "Epoch  432\n",
      "Top three least errors :  3.1955533027648926 3.1969552040100098 3.19759202003479\n",
      "Best error :  3.1955533027648926 \n",
      "\n",
      "Epoch  433\n",
      "Top three least errors :  3.1955533027648926 3.196037530899048 3.1964476108551025\n",
      "Best error :  3.195143222808838 \n",
      "\n",
      "Epoch  434\n",
      "Top three least errors :  3.195143222808838 3.1954894065856934 3.1955533027648926\n",
      "Best error :  3.194720506668091 \n",
      "\n",
      "Epoch  435\n",
      "Top three least errors :  3.194720506668091 3.194817543029785 3.195021867752075\n",
      "Best error :  3.1945102214813232 \n",
      "\n",
      "Epoch  436\n",
      "Top three least errors :  3.1945102214813232 3.194720506668091 3.194742441177368\n",
      "Best error :  3.194148302078247 \n",
      "\n",
      "Epoch  437\n",
      "Top three least errors :  3.194148302078247 3.1943047046661377 3.1944944858551025\n",
      "Best error :  3.1929972171783447 \n",
      "\n",
      "Epoch  438\n",
      "Top three least errors :  3.1929972171783447 3.193849802017212 3.194148302078247\n",
      "Best error :  3.1927716732025146 \n",
      "\n",
      "Epoch  439\n",
      "Top three least errors :  3.1927716732025146 3.1929972171783447 3.193272352218628\n",
      "Best error :  3.192549705505371 \n",
      "\n",
      "Epoch  440\n",
      "Top three least errors :  3.192549705505371 3.1926050186157227 3.1927669048309326\n",
      "Best error :  3.1923060417175293 \n",
      "\n",
      "Epoch  441\n",
      "Top three least errors :  3.1923060417175293 3.192410945892334 3.1925151348114014\n",
      "Best error :  3.1919877529144287 \n",
      "\n",
      "Epoch  442\n",
      "Top three least errors :  3.1919877529144287 3.1920363903045654 3.1922953128814697\n",
      "Best error :  3.191483736038208 \n",
      "\n",
      "Epoch  443\n",
      "Top three least errors :  3.191483736038208 3.1917128562927246 3.1919116973876953\n",
      "Best error :  3.1908488273620605 \n",
      "\n",
      "Epoch  444\n",
      "Top three least errors :  3.1908488273620605 3.1910934448242188 3.191483736038208\n",
      "Best error :  3.1908488273620605 \n",
      "\n",
      "Epoch  445\n",
      "Top three least errors :  3.1908488273620605 3.1909689903259277 3.1909804344177246\n",
      "Best error :  3.1903109550476074 \n",
      "\n",
      "Epoch  446\n",
      "Top three least errors :  3.1903109550476074 3.190620183944702 3.1906614303588867\n",
      "Best error :  3.189854621887207 \n",
      "\n",
      "Epoch  447\n",
      "Top three least errors :  3.189854621887207 3.1900768280029297 3.190196990966797\n",
      "Best error :  3.1891489028930664 \n",
      "\n",
      "Epoch  448\n",
      "Top three least errors :  3.1891489028930664 3.1891610622406006 3.189464569091797\n",
      "Best error :  3.1888909339904785 \n",
      "\n",
      "Epoch  449\n",
      "Top three least errors :  3.1888909339904785 3.188910722732544 3.18900203704834\n",
      "Best error :  3.188080310821533 \n",
      "\n",
      "Epoch  450\n",
      "Top three least errors :  3.188080310821533 3.1881062984466553 3.1884925365448\n",
      "Best error :  3.18682861328125 \n",
      "\n",
      "Epoch  451\n",
      "Top three least errors :  3.18682861328125 3.1879825592041016 3.188074827194214\n",
      "Best error :  3.1861960887908936 \n",
      "\n",
      "Epoch  452\n",
      "Top three least errors :  3.1861960887908936 3.1864986419677734 3.18682861328125\n",
      "Best error :  3.185473918914795 \n",
      "\n",
      "Epoch  453\n",
      "Top three least errors :  3.185473918914795 3.185758113861084 3.1861164569854736\n",
      "Best error :  3.184624671936035 \n",
      "\n",
      "Epoch  454\n",
      "Top three least errors :  3.184624671936035 3.184802770614624 3.1850192546844482\n",
      "Best error :  3.1843881607055664 \n",
      "\n",
      "Epoch  455\n",
      "Top three least errors :  3.1843881607055664 3.184453248977661 3.184467077255249\n",
      "Best error :  3.184213161468506 \n",
      "\n",
      "Epoch  456\n",
      "Top three least errors :  3.184213161468506 3.1843388080596924 3.1843881607055664\n",
      "Best error :  3.1839234828948975 \n",
      "\n",
      "Epoch  457\n",
      "Top three least errors :  3.1839234828948975 3.184213161468506 3.1843388080596924\n",
      "Best error :  3.183596134185791 \n",
      "\n",
      "Epoch  458\n",
      "Top three least errors :  3.183596134185791 3.1837592124938965 3.1839234828948975\n",
      "Best error :  3.183493137359619 \n",
      "\n",
      "Epoch  459\n",
      "Top three least errors :  3.183493137359619 3.183596134185791 3.1837377548217773\n",
      "Best error :  3.183387279510498 \n",
      "\n",
      "Epoch  460\n",
      "Top three least errors :  3.183387279510498 3.1834492683410645 3.1834540367126465\n",
      "Best error :  3.1828861236572266 \n",
      "\n",
      "Epoch  461\n",
      "Top three least errors :  3.1828861236572266 3.1833231449127197 3.1833767890930176\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.1826140880584717 \n",
      "\n",
      "Epoch  462\n",
      "Top three least errors :  3.1826140880584717 3.1828861236572266 3.1829030513763428\n",
      "Best error :  3.1821651458740234 \n",
      "\n",
      "Epoch  463\n",
      "Top three least errors :  3.1821651458740234 3.182349920272827 3.1823959350585938\n",
      "Best error :  3.1820547580718994 \n",
      "\n",
      "Epoch  464\n",
      "Top three least errors :  3.1820547580718994 3.1821651458740234 3.1821842193603516\n",
      "Best error :  3.1820547580718994 \n",
      "\n",
      "Epoch  465\n",
      "Top three least errors :  3.1820547580718994 3.182065010070801 3.1820764541625977\n",
      "Best error :  3.1817619800567627 \n",
      "\n",
      "Epoch  466\n",
      "Top three least errors :  3.1817619800567627 3.1820433139801025 3.1820547580718994\n",
      "Best error :  3.1816599369049072 \n",
      "\n",
      "Epoch  467\n",
      "Top three least errors :  3.1816599369049072 3.1817619800567627 3.181851625442505\n",
      "Best error :  3.181452751159668 \n",
      "\n",
      "Epoch  468\n",
      "Top three least errors :  3.181452751159668 3.1814680099487305 3.181528329849243\n",
      "Best error :  3.180508613586426 \n",
      "\n",
      "Epoch  469\n",
      "Top three least errors :  3.180508613586426 3.1808695793151855 3.1811020374298096\n",
      "Best error :  3.1787309646606445 \n",
      "\n",
      "Epoch  470\n",
      "Top three least errors :  3.1787309646606445 3.1803531646728516 3.180379629135132\n",
      "Best error :  3.177321434020996 \n",
      "\n",
      "Epoch  471\n",
      "Top three least errors :  3.177321434020996 3.1780848503112793 3.178351879119873\n",
      "Best error :  3.1757402420043945 \n",
      "\n",
      "Epoch  472\n",
      "Top three least errors :  3.1757402420043945 3.177321434020996 3.177623987197876\n",
      "Best error :  3.1742539405822754 \n",
      "\n",
      "Epoch  473\n",
      "Top three least errors :  3.1742539405822754 3.175724983215332 3.1757402420043945\n",
      "Best error :  3.1742122173309326 \n",
      "\n",
      "Epoch  474\n",
      "Top three least errors :  3.1742122173309326 3.1742539405822754 3.175724983215332\n",
      "Best error :  3.1702237129211426 \n",
      "\n",
      "Epoch  475\n",
      "Top three least errors :  3.1702237129211426 3.1733553409576416 3.1742122173309326\n",
      "Best error :  3.1640491485595703 \n",
      "\n",
      "Epoch  476\n",
      "Top three least errors :  3.1640491485595703 3.1698060035705566 3.1702237129211426\n",
      "Best error :  3.1640491485595703 \n",
      "\n",
      "Epoch  477\n",
      "Top three least errors :  3.1640491485595703 3.164890766143799 3.1698060035705566\n",
      "Best error :  3.1640491485595703 \n",
      "\n",
      "Epoch  478\n",
      "Top three least errors :  3.1640491485595703 3.1640801429748535 3.164890766143799\n",
      "Best error :  3.1635305881500244 \n",
      "\n",
      "Epoch  479\n",
      "Top three least errors :  3.1635305881500244 3.1639328002929688 3.1640491485595703\n",
      "Best error :  3.162539482116699 \n",
      "\n",
      "Epoch  480\n",
      "Top three least errors :  3.162539482116699 3.1635305881500244 3.1639328002929688\n",
      "Best error :  3.162017583847046 \n",
      "\n",
      "Epoch  481\n",
      "Top three least errors :  3.162017583847046 3.162539482116699 3.1626617908477783\n",
      "Best error :  3.161038875579834 \n",
      "\n",
      "Epoch  482\n",
      "Top three least errors :  3.161038875579834 3.162017583847046 3.162539482116699\n",
      "Best error :  3.161038875579834 \n",
      "\n",
      "Epoch  483\n",
      "Top three least errors :  3.161038875579834 3.162017583847046 3.162040948867798\n",
      "Best error :  3.161038875579834 \n",
      "\n",
      "Epoch  484\n",
      "Top three least errors :  3.161038875579834 3.162017583847046 3.162040948867798\n",
      "Best error :  3.1603548526763916 \n",
      "\n",
      "Epoch  485\n",
      "Top three least errors :  3.1603548526763916 3.161031723022461 3.161038875579834\n",
      "Best error :  3.159599781036377 \n",
      "\n",
      "Epoch  486\n",
      "Top three least errors :  3.159599781036377 3.160202980041504 3.1603548526763916\n",
      "Best error :  3.1570842266082764 \n",
      "\n",
      "Epoch  487\n",
      "Top three least errors :  3.1570842266082764 3.158846139907837 3.159332752227783\n",
      "Best error :  3.1570842266082764 \n",
      "\n",
      "Epoch  488\n",
      "Top three least errors :  3.1570842266082764 3.157294988632202 3.1582095623016357\n",
      "Best error :  3.1561570167541504 \n",
      "\n",
      "Epoch  489\n",
      "Top three least errors :  3.1561570167541504 3.15625262260437 3.1565282344818115\n",
      "Best error :  3.1558661460876465 \n",
      "\n",
      "Epoch  490\n",
      "Top three least errors :  3.1558661460876465 3.1561570167541504 3.15625262260437\n",
      "Best error :  3.154771327972412 \n",
      "\n",
      "Epoch  491\n",
      "Top three least errors :  3.154771327972412 3.1548256874084473 3.1556215286254883\n",
      "Best error :  3.154771327972412 \n",
      "\n",
      "Epoch  492\n",
      "Top three least errors :  3.154771327972412 3.1548256874084473 3.1552250385284424\n",
      "Best error :  3.15468430519104 \n",
      "\n",
      "Epoch  493\n",
      "Top three least errors :  3.15468430519104 3.154771327972412 3.1548256874084473\n",
      "Best error :  3.1543891429901123 \n",
      "\n",
      "Epoch  494\n",
      "Top three least errors :  3.1543891429901123 3.15468430519104 3.154717445373535\n",
      "Best error :  3.154268980026245 \n",
      "\n",
      "Epoch  495\n",
      "Top three least errors :  3.154268980026245 3.1543891429901123 3.1546199321746826\n",
      "Best error :  3.154167890548706 \n",
      "\n",
      "Epoch  496\n",
      "Top three least errors :  3.154167890548706 3.154268980026245 3.1543891429901123\n",
      "Best error :  3.1540184020996094 \n",
      "\n",
      "Epoch  497\n",
      "Top three least errors :  3.1540184020996094 3.1541128158569336 3.154167890548706\n",
      "Best error :  3.1540184020996094 \n",
      "\n",
      "Epoch  498\n",
      "Top three least errors :  3.1540184020996094 3.1540379524230957 3.1541128158569336\n",
      "Best error :  3.152222156524658 \n",
      "\n",
      "Epoch  499\n",
      "Top three least errors :  3.152222156524658 3.1540184020996094 3.1540379524230957\n",
      "Best error :  3.1519486904144287 \n",
      "\n",
      "Epoch  500\n",
      "Top three least errors :  3.1519486904144287 3.152193546295166 3.152222156524658\n",
      "Best error :  3.150123357772827 \n",
      "\n",
      "Epoch  501\n",
      "Top three least errors :  3.150123357772827 3.1503310203552246 3.1518847942352295\n",
      "Best error :  3.150123357772827 \n",
      "\n",
      "Epoch  502\n",
      "Top three least errors :  3.150123357772827 3.1503310203552246 3.1518847942352295\n",
      "Best error :  3.150123357772827 \n",
      "\n",
      "Epoch  503\n",
      "Top three least errors :  3.150123357772827 3.1503310203552246 3.1506733894348145\n",
      "Best error :  3.1483333110809326 \n",
      "\n",
      "Epoch  504\n",
      "Top three least errors :  3.1483333110809326 3.1496758460998535 3.1497573852539062\n",
      "Best error :  3.1476688385009766 \n",
      "\n",
      "Epoch  505\n",
      "Top three least errors :  3.1476688385009766 3.14827823638916 3.1483333110809326\n",
      "Best error :  3.145881414413452 \n",
      "\n",
      "Epoch  506\n",
      "Top three least errors :  3.145881414413452 3.1476688385009766 3.14827823638916\n",
      "Best error :  3.1450984477996826 \n",
      "\n",
      "Epoch  507\n",
      "Top three least errors :  3.1450984477996826 3.1455962657928467 3.145881414413452\n",
      "Best error :  3.142951726913452 \n",
      "\n",
      "Epoch  508\n",
      "Top three least errors :  3.142951726913452 3.1448280811309814 3.1450984477996826\n",
      "Best error :  3.142951726913452 \n",
      "\n",
      "Epoch  509\n",
      "Top three least errors :  3.142951726913452 3.1448280811309814 3.1450984477996826\n",
      "Best error :  3.142909288406372 \n",
      "\n",
      "Epoch  510\n",
      "Top three least errors :  3.142909288406372 3.142951726913452 3.1434335708618164\n",
      "Best error :  3.1416635513305664 \n",
      "\n",
      "Epoch  511\n",
      "Top three least errors :  3.1416635513305664 3.1424951553344727 3.142909288406372\n",
      "Best error :  3.1416635513305664 \n",
      "\n",
      "Epoch  512\n",
      "Top three least errors :  3.1416635513305664 3.141705274581909 3.141819715499878\n",
      "Best error :  3.1416287422180176 \n",
      "\n",
      "Epoch  513\n",
      "Top three least errors :  3.1416287422180176 3.1416635513305664 3.141705274581909\n",
      "Best error :  3.1406607627868652 \n",
      "\n",
      "Epoch  514\n",
      "Top three least errors :  3.1406607627868652 3.1416287422180176 3.141662359237671\n",
      "Best error :  3.1399545669555664 \n",
      "\n",
      "Epoch  515\n",
      "Top three least errors :  3.1399545669555664 3.14056396484375 3.1406607627868652\n",
      "Best error :  3.138566255569458 \n",
      "\n",
      "Epoch  516\n",
      "Top three least errors :  3.138566255569458 3.1390485763549805 3.1399006843566895\n",
      "Best error :  3.1383183002471924 \n",
      "\n",
      "Epoch  517\n",
      "Top three least errors :  3.1383183002471924 3.138566255569458 3.1387298107147217\n",
      "Best error :  3.137784004211426 \n",
      "\n",
      "Epoch  518\n",
      "Top three least errors :  3.137784004211426 3.1381242275238037 3.1382970809936523\n",
      "Best error :  3.1374571323394775 \n",
      "\n",
      "Epoch  519\n",
      "Top three least errors :  3.1374571323394775 3.137784004211426 3.1379759311676025\n",
      "Best error :  3.137432813644409 \n",
      "\n",
      "Epoch  520\n",
      "Top three least errors :  3.137432813644409 3.1374571323394775 3.1379475593566895\n",
      "Best error :  3.137432813644409 \n",
      "\n",
      "Epoch  521\n",
      "Top three least errors :  3.137432813644409 3.13744854927063 3.1375839710235596\n",
      "Best error :  3.1361916065216064 \n",
      "\n",
      "Epoch  522\n",
      "Top three least errors :  3.1361916065216064 3.1374318599700928 3.137432813644409\n",
      "Best error :  3.134664535522461 \n",
      "\n",
      "Epoch  523\n",
      "Top three least errors :  3.134664535522461 3.1361916065216064 3.1364920139312744\n",
      "Best error :  3.134664535522461 \n",
      "\n",
      "Epoch  524\n",
      "Top three least errors :  3.134664535522461 3.1346874237060547 3.1348071098327637\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\admin\\AppData\\Local\\Temp\\ipykernel_17232\\2435813276.py:34: RuntimeWarning: divide by zero encountered in divide\n",
      "  s = u / np.power(np.abs(v), 1 / beta)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.1345410346984863 \n",
      "\n",
      "Epoch  525\n",
      "Top three least errors :  3.1345410346984863 3.134664535522461 3.1346874237060547\n",
      "Best error :  3.1344077587127686 \n",
      "\n",
      "Epoch  526\n",
      "Top three least errors :  3.1344077587127686 3.1345410346984863 3.134617567062378\n",
      "Best error :  3.13419771194458 \n",
      "\n",
      "Epoch  527\n",
      "Top three least errors :  3.13419771194458 3.1344025135040283 3.1344077587127686\n",
      "Best error :  3.133823871612549 \n",
      "\n",
      "Epoch  528\n",
      "Top three least errors :  3.133823871612549 3.1339821815490723 3.1340911388397217\n",
      "Best error :  3.1337497234344482 \n",
      "\n",
      "Epoch  529\n",
      "Top three least errors :  3.1337497234344482 3.133823871612549 3.1338915824890137\n",
      "Best error :  3.1330673694610596 \n",
      "\n",
      "Epoch  530\n",
      "Top three least errors :  3.1330673694610596 3.133357286453247 3.133662700653076\n",
      "Best error :  3.132798910140991 \n",
      "\n",
      "Epoch  531\n",
      "Top three least errors :  3.132798910140991 3.133051872253418 3.1330673694610596\n",
      "Best error :  3.132798910140991 \n",
      "\n",
      "Epoch  532\n",
      "Top three least errors :  3.132798910140991 3.132901430130005 3.133051872253418\n",
      "Best error :  3.13275408744812 \n",
      "\n",
      "Epoch  533\n",
      "Top three least errors :  3.13275408744812 3.132798910140991 3.132823944091797\n",
      "Best error :  3.1326565742492676 \n",
      "\n",
      "Epoch  534\n",
      "Top three least errors :  3.1326565742492676 3.13275408744812 3.132823944091797\n",
      "Best error :  3.132462978363037 \n",
      "\n",
      "Epoch  535\n",
      "Top three least errors :  3.132462978363037 3.132481336593628 3.1326565742492676\n",
      "Best error :  3.132462978363037 \n",
      "\n",
      "Epoch  536\n",
      "Top three least errors :  3.132462978363037 3.132481336593628 3.1326565742492676\n",
      "Best error :  3.132161855697632 \n",
      "\n",
      "Epoch  537\n",
      "Top three least errors :  3.132161855697632 3.1323156356811523 3.132462978363037\n",
      "Best error :  3.1318182945251465 \n",
      "\n",
      "Epoch  538\n",
      "Top three least errors :  3.1318182945251465 3.132005453109741 3.132161855697632\n",
      "Best error :  3.1313915252685547 \n",
      "\n",
      "Epoch  539\n",
      "Top three least errors :  3.1313915252685547 3.131679058074951 3.1318182945251465\n",
      "Best error :  3.1313915252685547 \n",
      "\n",
      "Epoch  540\n",
      "Top three least errors :  3.1313915252685547 3.131434679031372 3.1316123008728027\n",
      "Best error :  3.1313283443450928 \n",
      "\n",
      "Epoch  541\n",
      "Top three least errors :  3.1313283443450928 3.1313915252685547 3.131392002105713\n",
      "Best error :  3.130657911300659 \n",
      "\n",
      "Epoch  542\n",
      "Top three least errors :  3.130657911300659 3.1312241554260254 3.1313283443450928\n",
      "Best error :  3.130358934402466 \n",
      "\n",
      "Epoch  543\n",
      "Top three least errors :  3.130358934402466 3.1305880546569824 3.130613088607788\n",
      "Best error :  3.1297855377197266 \n",
      "\n",
      "Epoch  544\n",
      "Top three least errors :  3.1297855377197266 3.1298201084136963 3.1302478313446045\n",
      "Best error :  3.1294760704040527 \n",
      "\n",
      "Epoch  545\n",
      "Top three least errors :  3.1294760704040527 3.1297855377197266 3.129796028137207\n",
      "Best error :  3.129422187805176 \n",
      "\n",
      "Epoch  546\n",
      "Top three least errors :  3.129422187805176 3.1294760704040527 3.1297855377197266\n",
      "Best error :  3.129422187805176 \n",
      "\n",
      "Epoch  547\n",
      "Top three least errors :  3.129422187805176 3.1294760704040527 3.1296494007110596\n",
      "Best error :  3.1291654109954834 \n",
      "\n",
      "Epoch  548\n",
      "Top three least errors :  3.1291654109954834 3.1293838024139404 3.129422187805176\n",
      "Best error :  3.129126787185669 \n",
      "\n",
      "Epoch  549\n",
      "Top three least errors :  3.129126787185669 3.1291654109954834 3.1291909217834473\n",
      "Best error :  3.128800392150879 \n",
      "\n",
      "Epoch  550\n",
      "Top three least errors :  3.128800392150879 3.1288647651672363 3.129009485244751\n",
      "Best error :  3.1282219886779785 \n",
      "\n",
      "Epoch  551\n",
      "Top three least errors :  3.1282219886779785 3.128617525100708 3.128800392150879\n",
      "Best error :  3.1271960735321045 \n",
      "\n",
      "Epoch  552\n",
      "Top three least errors :  3.1271960735321045 3.127790689468384 3.1282219886779785\n",
      "Best error :  3.126082181930542 \n",
      "\n",
      "Epoch  553\n",
      "Top three least errors :  3.126082181930542 3.1271960735321045 3.1272032260894775\n",
      "Best error :  3.1252758502960205 \n",
      "\n",
      "Epoch  554\n",
      "Top three least errors :  3.1252758502960205 3.126082181930542 3.1268317699432373\n",
      "Best error :  3.1252758502960205 \n",
      "\n",
      "Epoch  555\n",
      "Top three least errors :  3.1252758502960205 3.1254966259002686 3.126008987426758\n",
      "Best error :  3.125065565109253 \n",
      "\n",
      "Epoch  556\n",
      "Top three least errors :  3.125065565109253 3.1252667903900146 3.1252713203430176\n",
      "Best error :  3.124857187271118 \n",
      "\n",
      "Epoch  557\n",
      "Top three least errors :  3.124857187271118 3.125042676925659 3.125065565109253\n",
      "Best error :  3.124173164367676 \n",
      "\n",
      "Epoch  558\n",
      "Top three least errors :  3.124173164367676 3.124857187271118 3.125042676925659\n",
      "Best error :  3.1238996982574463 \n",
      "\n",
      "Epoch  559\n",
      "Top three least errors :  3.1238996982574463 3.124173164367676 3.1242246627807617\n",
      "Best error :  3.123832941055298 \n",
      "\n",
      "Epoch  560\n",
      "Top three least errors :  3.123832941055298 3.1238996982574463 3.1242246627807617\n",
      "Best error :  3.123832941055298 \n",
      "\n",
      "Epoch  561\n",
      "Top three least errors :  3.123832941055298 3.1238996982574463 3.1240832805633545\n",
      "Best error :  3.1235294342041016 \n",
      "\n",
      "Epoch  562\n",
      "Top three least errors :  3.1235294342041016 3.123826742172241 3.123832941055298\n",
      "Best error :  3.1235060691833496 \n",
      "\n",
      "Epoch  563\n",
      "Top three least errors :  3.1235060691833496 3.1235294342041016 3.123826742172241\n",
      "Best error :  3.1235013008117676 \n",
      "\n",
      "Epoch  564\n",
      "Top three least errors :  3.1235013008117676 3.1235060691833496 3.123826742172241\n",
      "Best error :  3.1235013008117676 \n",
      "\n",
      "Epoch  565\n",
      "Top three least errors :  3.1235013008117676 3.1235060691833496 3.123631715774536\n",
      "Best error :  3.122629404067993 \n",
      "\n",
      "Epoch  566\n",
      "Top three least errors :  3.122629404067993 3.1232004165649414 3.1232500076293945\n",
      "Best error :  3.1225569248199463 \n",
      "\n",
      "Epoch  567\n",
      "Top three least errors :  3.1225569248199463 3.122629404067993 3.1226747035980225\n",
      "Best error :  3.1222479343414307 \n",
      "\n",
      "Epoch  568\n",
      "Top three least errors :  3.1222479343414307 3.1224756240844727 3.1225430965423584\n",
      "Best error :  3.1218605041503906 \n",
      "\n",
      "Epoch  569\n",
      "Top three least errors :  3.1218605041503906 3.1222479343414307 3.1222634315490723\n",
      "Best error :  3.1214542388916016 \n",
      "\n",
      "Epoch  570\n",
      "Top three least errors :  3.1214542388916016 3.121485948562622 3.121657133102417\n",
      "Best error :  3.1213343143463135 \n",
      "\n",
      "Epoch  571\n",
      "Top three least errors :  3.1213343143463135 3.1214542388916016 3.121657133102417\n",
      "Best error :  3.1208205223083496 \n",
      "\n",
      "Epoch  572\n",
      "Top three least errors :  3.1208205223083496 3.1209537982940674 3.12127947807312\n",
      "Best error :  3.1208205223083496 \n",
      "\n",
      "Epoch  573\n",
      "Top three least errors :  3.1208205223083496 3.1209537982940674 3.12127947807312\n",
      "Best error :  3.1204373836517334 \n",
      "\n",
      "Epoch  574\n",
      "Top three least errors :  3.1204373836517334 3.1208205223083496 3.1209537982940674\n",
      "Best error :  3.118844509124756 \n",
      "\n",
      "Epoch  575\n",
      "Top three least errors :  3.118844509124756 3.1199026107788086 3.1200897693634033\n",
      "Best error :  3.118844509124756 \n",
      "\n",
      "Epoch  576\n",
      "Top three least errors :  3.118844509124756 3.1188576221466064 3.1189115047454834\n",
      "Best error :  3.118781328201294 \n",
      "\n",
      "Epoch  577\n",
      "Top three least errors :  3.118781328201294 3.118834972381592 3.118844509124756\n",
      "Best error :  3.118715763092041 \n",
      "\n",
      "Epoch  578\n",
      "Top three least errors :  3.118715763092041 3.118781328201294 3.118834972381592\n",
      "Best error :  3.116913318634033 \n",
      "\n",
      "Epoch  579\n",
      "Top three least errors :  3.116913318634033 3.1177308559417725 3.1187100410461426\n",
      "Best error :  3.116913318634033 \n",
      "\n",
      "Epoch  580\n",
      "Top three least errors :  3.116913318634033 3.117022752761841 3.1171112060546875\n",
      "Best error :  3.115248203277588 \n",
      "\n",
      "Epoch  581\n",
      "Top three least errors :  3.115248203277588 3.116802215576172 3.116913318634033\n",
      "Best error :  3.1149725914001465 \n",
      "\n",
      "Epoch  582\n",
      "Top three least errors :  3.1149725914001465 3.115248203277588 3.11541748046875\n",
      "Best error :  3.1141724586486816 \n",
      "\n",
      "Epoch  583\n",
      "Top three least errors :  3.1141724586486816 3.1149725914001465 3.115248203277588\n",
      "Best error :  3.1136388778686523 \n",
      "\n",
      "Epoch  584\n",
      "Top three least errors :  3.1136388778686523 3.1141724586486816 3.114365339279175\n",
      "Best error :  3.1118311882019043 \n",
      "\n",
      "Epoch  585\n",
      "Top three least errors :  3.1118311882019043 3.1123745441436768 3.113433361053467\n",
      "Best error :  3.1118311882019043 \n",
      "\n",
      "Epoch  586\n",
      "Top three least errors :  3.1118311882019043 3.112178325653076 3.1123745441436768\n",
      "Best error :  3.1107840538024902 \n",
      "\n",
      "Epoch  587\n",
      "Top three least errors :  3.1107840538024902 3.111656427383423 3.1118311882019043\n",
      "Best error :  3.1103594303131104 \n",
      "\n",
      "Epoch  588\n",
      "Top three least errors :  3.1103594303131104 3.1107840538024902 3.1111698150634766\n",
      "Best error :  3.1095457077026367 \n",
      "\n",
      "Epoch  589\n",
      "Top three least errors :  3.1095457077026367 3.1100189685821533 3.1105270385742188\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.1095457077026367 \n",
      "\n",
      "Epoch  590\n",
      "Top three least errors :  3.1095457077026367 3.1100189685821533 3.1105270385742188\n",
      "Best error :  3.1095457077026367 \n",
      "\n",
      "Epoch  591\n",
      "Top three least errors :  3.1095457077026367 3.1100189685821533 3.1105270385742188\n",
      "Best error :  3.1059350967407227 \n",
      "\n",
      "Epoch  592\n",
      "Top three least errors :  3.1059350967407227 3.1093242168426514 3.1095457077026367\n",
      "Best error :  3.1031367778778076 \n",
      "\n",
      "Epoch  593\n",
      "Top three least errors :  3.1031367778778076 3.1059350967407227 3.1093242168426514\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  594\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  595\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  596\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  597\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  598\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  599\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  600\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  601\n",
      "Top three least errors :  3.1005702018737793 3.1031367778778076 3.103224992752075\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  602\n",
      "Top three least errors :  3.1005702018737793 3.1021440029144287 3.103046178817749\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  603\n",
      "Top three least errors :  3.1005702018737793 3.100621223449707 3.1021440029144287\n",
      "Best error :  3.1005702018737793 \n",
      "\n",
      "Epoch  604\n",
      "Top three least errors :  3.1005702018737793 3.100621223449707 3.1021440029144287\n",
      "Best error :  3.099207878112793 \n",
      "\n",
      "Epoch  605\n",
      "Top three least errors :  3.099207878112793 3.099940061569214 3.1005702018737793\n",
      "Best error :  3.099207878112793 \n",
      "\n",
      "Epoch  606\n",
      "Top three least errors :  3.099207878112793 3.099940061569214 3.1005702018737793\n",
      "Best error :  3.0991129875183105 \n",
      "\n",
      "Epoch  607\n",
      "Top three least errors :  3.0991129875183105 3.099207878112793 3.099940061569214\n",
      "Best error :  3.099015474319458 \n",
      "\n",
      "Epoch  608\n",
      "Top three least errors :  3.099015474319458 3.0991129875183105 3.0997722148895264\n",
      "Best error :  3.0989623069763184 \n",
      "\n",
      "Epoch  609\n",
      "Top three least errors :  3.0989623069763184 3.099015474319458 3.0991055965423584\n",
      "Best error :  3.0989623069763184 \n",
      "\n",
      "Epoch  610\n",
      "Top three least errors :  3.0989623069763184 3.099012851715088 3.0991055965423584\n",
      "Best error :  3.098933696746826 \n",
      "\n",
      "Epoch  611\n",
      "Top three least errors :  3.098933696746826 3.0989623069763184 3.0989863872528076\n",
      "Best error :  3.0989162921905518 \n",
      "\n",
      "Epoch  612\n",
      "Top three least errors :  3.0989162921905518 3.0989325046539307 3.098933696746826\n",
      "Best error :  3.0989162921905518 \n",
      "\n",
      "Epoch  613\n",
      "Top three least errors :  3.0989162921905518 3.0989325046539307 3.098933696746826\n",
      "Best error :  3.0988664627075195 \n",
      "\n",
      "Epoch  614\n",
      "Top three least errors :  3.0988664627075195 3.0989162921905518 3.0989325046539307\n",
      "Best error :  3.0988152027130127 \n",
      "\n",
      "Epoch  615\n",
      "Top three least errors :  3.0988152027130127 3.0988664627075195 3.0989325046539307\n",
      "Best error :  3.0988152027130127 \n",
      "\n",
      "Epoch  616\n",
      "Top three least errors :  3.0988152027130127 3.0988664627075195 3.0989325046539307\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  617\n",
      "Top three least errors :  3.098161458969116 3.0987157821655273 3.0987837314605713\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  618\n",
      "Top three least errors :  3.098161458969116 3.0983517169952393 3.0986180305480957\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  619\n",
      "Top three least errors :  3.098161458969116 3.098306655883789 3.0983517169952393\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  620\n",
      "Top three least errors :  3.098161458969116 3.0982277393341064 3.098306655883789\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  621\n",
      "Top three least errors :  3.098161458969116 3.0982277393341064 3.098306655883789\n",
      "Best error :  3.098161458969116 \n",
      "\n",
      "Epoch  622\n",
      "Top three least errors :  3.098161458969116 3.0981733798980713 3.0982277393341064\n",
      "Best error :  3.097853899002075 \n",
      "\n",
      "Epoch  623\n",
      "Top three least errors :  3.097853899002075 3.0979504585266113 3.098161458969116\n",
      "Best error :  3.097797393798828 \n",
      "\n",
      "Epoch  624\n",
      "Top three least errors :  3.097797393798828 3.097853899002075 3.097930431365967\n",
      "Best error :  3.0977816581726074 \n",
      "\n",
      "Epoch  625\n",
      "Top three least errors :  3.0977816581726074 3.097797393798828 3.0978450775146484\n",
      "Best error :  3.097315788269043 \n",
      "\n",
      "Epoch  626\n",
      "Top three least errors :  3.097315788269043 3.097778081893921 3.0977816581726074\n",
      "Best error :  3.0953562259674072 \n",
      "\n",
      "Epoch  627\n",
      "Top three least errors :  3.0953562259674072 3.0972495079040527 3.097315788269043\n",
      "Best error :  3.0947365760803223 \n",
      "\n",
      "Epoch  628\n",
      "Top three least errors :  3.0947365760803223 3.0947587490081787 3.0953562259674072\n",
      "Best error :  3.0944392681121826 \n",
      "\n",
      "Epoch  629\n",
      "Top three least errors :  3.0944392681121826 3.0947365760803223 3.0947587490081787\n",
      "Best error :  3.0937211513519287 \n",
      "\n",
      "Epoch  630\n",
      "Top three least errors :  3.0937211513519287 3.0941824913024902 3.0944392681121826\n",
      "Best error :  3.0937211513519287 \n",
      "\n",
      "Epoch  631\n",
      "Top three least errors :  3.0937211513519287 3.0940096378326416 3.0944392681121826\n",
      "Best error :  3.093393325805664 \n",
      "\n",
      "Epoch  632\n",
      "Top three least errors :  3.093393325805664 3.0937211513519287 3.0940096378326416\n",
      "Best error :  3.0921919345855713 \n",
      "\n",
      "Epoch  633\n",
      "Top three least errors :  3.0921919345855713 3.093393325805664 3.093559741973877\n",
      "Best error :  3.09143328666687 \n",
      "\n",
      "Epoch  634\n",
      "Top three least errors :  3.09143328666687 3.0920169353485107 3.0921919345855713\n",
      "Best error :  3.09143328666687 \n",
      "\n",
      "Epoch  635\n",
      "Top three least errors :  3.09143328666687 3.0914433002471924 3.091499090194702\n",
      "Best error :  3.09143328666687 \n",
      "\n",
      "Epoch  636\n",
      "Top three least errors :  3.09143328666687 3.0914433002471924 3.091499090194702\n",
      "Best error :  3.0913753509521484 \n",
      "\n",
      "Epoch  637\n",
      "Top three least errors :  3.0913753509521484 3.09143328666687 3.091499090194702\n",
      "Best error :  3.0911335945129395 \n",
      "\n",
      "Epoch  638\n",
      "Top three least errors :  3.0911335945129395 3.0913753509521484 3.0914058685302734\n",
      "Best error :  3.090754747390747 \n",
      "\n",
      "Epoch  639\n",
      "Top three least errors :  3.090754747390747 3.090958833694458 3.091127395629883\n",
      "Best error :  3.090402364730835 \n",
      "\n",
      "Epoch  640\n",
      "Top three least errors :  3.090402364730835 3.090754747390747 3.0907704830169678\n",
      "Best error :  3.090402364730835 \n",
      "\n",
      "Epoch  641\n",
      "Top three least errors :  3.090402364730835 3.090693950653076 3.09073805809021\n",
      "Best error :  3.0896453857421875 \n",
      "\n",
      "Epoch  642\n",
      "Top three least errors :  3.0896453857421875 3.090317487716675 3.090346574783325\n",
      "Best error :  3.0896453857421875 \n",
      "\n",
      "Epoch  643\n",
      "Top three least errors :  3.0896453857421875 3.090317487716675 3.090346574783325\n",
      "Best error :  3.0887527465820312 \n",
      "\n",
      "Epoch  644\n",
      "Top three least errors :  3.0887527465820312 3.0896453857421875 3.0901622772216797\n",
      "Best error :  3.0867104530334473 \n",
      "\n",
      "Epoch  645\n",
      "Top three least errors :  3.0867104530334473 3.0880279541015625 3.0887527465820312\n",
      "Best error :  3.0867104530334473 \n",
      "\n",
      "Epoch  646\n",
      "Top three least errors :  3.0867104530334473 3.0880279541015625 3.0887527465820312\n",
      "Best error :  3.0867104530334473 \n",
      "\n",
      "Epoch  647\n",
      "Top three least errors :  3.0867104530334473 3.0880279541015625 3.0887527465820312\n",
      "Best error :  3.0858523845672607 \n",
      "\n",
      "Epoch  648\n",
      "Top three least errors :  3.0858523845672607 3.0867104530334473 3.0880279541015625\n",
      "Best error :  3.0858523845672607 \n",
      "\n",
      "Epoch  649\n",
      "Top three least errors :  3.0858523845672607 3.0867104530334473 3.0880279541015625\n",
      "Best error :  3.0858523845672607 \n",
      "\n",
      "Epoch  650\n",
      "Top three least errors :  3.0858523845672607 3.0867104530334473 3.0880279541015625\n",
      "Best error :  3.0858523845672607 \n",
      "\n",
      "Epoch  651\n",
      "Top three least errors :  3.0858523845672607 3.0861151218414307 3.0867104530334473\n",
      "Best error :  3.0858356952667236 \n",
      "\n",
      "Epoch  652\n",
      "Top three least errors :  3.0858356952667236 3.0858523845672607 3.0861151218414307\n",
      "Best error :  3.0858356952667236 \n",
      "\n",
      "Epoch  653\n",
      "Top three least errors :  3.0858356952667236 3.0858523845672607 3.0861151218414307\n",
      "Best error :  3.0858356952667236 \n",
      "\n",
      "Epoch  654\n",
      "Top three least errors :  3.0858356952667236 3.0858523845672607 3.0861151218414307\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.0822083950042725 \n",
      "\n",
      "Epoch  655\n",
      "Top three least errors :  3.0822083950042725 3.085782289505005 3.0858356952667236\n",
      "Best error :  3.0822083950042725 \n",
      "\n",
      "Epoch  656\n",
      "Top three least errors :  3.0822083950042725 3.083909749984741 3.084761619567871\n",
      "Best error :  3.0805060863494873 \n",
      "\n",
      "Epoch  657\n",
      "Top three least errors :  3.0805060863494873 3.0822083950042725 3.083909749984741\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  658\n",
      "Top three least errors :  3.0758793354034424 3.079681396484375 3.0805060863494873\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  659\n",
      "Top three least errors :  3.0758793354034424 3.079561471939087 3.079681396484375\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  660\n",
      "Top three least errors :  3.0758793354034424 3.0783114433288574 3.079561471939087\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  661\n",
      "Top three least errors :  3.0758793354034424 3.0781009197235107 3.0783114433288574\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  662\n",
      "Top three least errors :  3.0758793354034424 3.0781009197235107 3.0783114433288574\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  663\n",
      "Top three least errors :  3.0758793354034424 3.0765249729156494 3.0781009197235107\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  664\n",
      "Top three least errors :  3.0758793354034424 3.0765249729156494 3.0779170989990234\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  665\n",
      "Top three least errors :  3.0758793354034424 3.0765249729156494 3.0779170989990234\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  666\n",
      "Top three least errors :  3.0758793354034424 3.0765249729156494 3.0779170989990234\n",
      "Best error :  3.0758793354034424 \n",
      "\n",
      "Epoch  667\n",
      "Top three least errors :  3.0758793354034424 3.0760905742645264 3.076378583908081\n",
      "Best error :  3.075774669647217 \n",
      "\n",
      "Epoch  668\n",
      "Top three least errors :  3.075774669647217 3.0758793354034424 3.0760555267333984\n",
      "Best error :  3.075774669647217 \n",
      "\n",
      "Epoch  669\n",
      "Top three least errors :  3.075774669647217 3.0758793354034424 3.0760555267333984\n",
      "Best error :  3.075744867324829 \n",
      "\n",
      "Epoch  670\n",
      "Top three least errors :  3.075744867324829 3.075774669647217 3.0760555267333984\n",
      "Best error :  3.075744867324829 \n",
      "\n",
      "Epoch  671\n",
      "Top three least errors :  3.075744867324829 3.075774669647217 3.0760555267333984\n",
      "Best error :  3.0743601322174072 \n",
      "\n",
      "Epoch  672\n",
      "Top three least errors :  3.0743601322174072 3.0756373405456543 3.075744867324829\n",
      "Best error :  3.0721590518951416 \n",
      "\n",
      "Epoch  673\n",
      "Top three least errors :  3.0721590518951416 3.0740573406219482 3.0743601322174072\n",
      "Best error :  3.0721590518951416 \n",
      "\n",
      "Epoch  674\n",
      "Top three least errors :  3.0721590518951416 3.0740573406219482 3.0743601322174072\n",
      "Best error :  3.0719332695007324 \n",
      "\n",
      "Epoch  675\n",
      "Top three least errors :  3.0719332695007324 3.0721590518951416 3.0727295875549316\n",
      "Best error :  3.071120023727417 \n",
      "\n",
      "Epoch  676\n",
      "Top three least errors :  3.071120023727417 3.0719332695007324 3.0721590518951416\n",
      "Best error :  3.069565773010254 \n",
      "\n",
      "Epoch  677\n",
      "Top three least errors :  3.069565773010254 3.071120023727417 3.07181453704834\n",
      "Best error :  3.0665760040283203 \n",
      "\n",
      "Epoch  678\n",
      "Top three least errors :  3.0665760040283203 3.068946361541748 3.069565773010254\n",
      "Best error :  3.0625579357147217 \n",
      "\n",
      "Epoch  679\n",
      "Top three least errors :  3.0625579357147217 3.063181161880493 3.065784454345703\n",
      "Best error :  3.061007261276245 \n",
      "\n",
      "Epoch  680\n",
      "Top three least errors :  3.061007261276245 3.0625579357147217 3.063181161880493\n",
      "Best error :  3.0600502490997314 \n",
      "\n",
      "Epoch  681\n",
      "Top three least errors :  3.0600502490997314 3.061007261276245 3.0625579357147217\n",
      "Best error :  3.0571579933166504 \n",
      "\n",
      "Epoch  682\n",
      "Top three least errors :  3.0571579933166504 3.0594193935394287 3.0600502490997314\n",
      "Best error :  3.0538480281829834 \n",
      "\n",
      "Epoch  683\n",
      "Top three least errors :  3.0538480281829834 3.055624008178711 3.0571579933166504\n",
      "Best error :  3.0530364513397217 \n",
      "\n",
      "Epoch  684\n",
      "Top three least errors :  3.0530364513397217 3.0538480281829834 3.05391788482666\n",
      "Best error :  3.052497625350952 \n",
      "\n",
      "Epoch  685\n",
      "Top three least errors :  3.052497625350952 3.0530364513397217 3.0538480281829834\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\admin\\AppData\\Local\\Temp\\ipykernel_17232\\2435813276.py:34: RuntimeWarning: overflow encountered in divide\n",
      "  s = u / np.power(np.abs(v), 1 / beta)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.047666311264038 \n",
      "\n",
      "Epoch  686\n",
      "Top three least errors :  3.047666311264038 3.052497625350952 3.0530364513397217\n",
      "Best error :  3.047666311264038 \n",
      "\n",
      "Epoch  687\n",
      "Top three least errors :  3.047666311264038 3.0505237579345703 3.052497625350952\n",
      "Best error :  3.047666311264038 \n",
      "\n",
      "Epoch  688\n",
      "Top three least errors :  3.047666311264038 3.0505237579345703 3.050808906555176\n",
      "Best error :  3.047666311264038 \n",
      "\n",
      "Epoch  689\n",
      "Top three least errors :  3.047666311264038 3.0505237579345703 3.050808906555176\n",
      "Best error :  3.046079158782959 \n",
      "\n",
      "Epoch  690\n",
      "Top three least errors :  3.046079158782959 3.0463650226593018 3.047666311264038\n",
      "Best error :  3.046079158782959 \n",
      "\n",
      "Epoch  691\n",
      "Top three least errors :  3.046079158782959 3.0460827350616455 3.0463650226593018\n",
      "Best error :  3.045689105987549 \n",
      "\n",
      "Epoch  692\n",
      "Top three least errors :  3.045689105987549 3.046079158782959 3.0460827350616455\n",
      "Best error :  3.045689105987549 \n",
      "\n",
      "Epoch  693\n",
      "Top three least errors :  3.045689105987549 3.046079158782959 3.0460827350616455\n",
      "Best error :  3.045689105987549 \n",
      "\n",
      "Epoch  694\n",
      "Top three least errors :  3.045689105987549 3.046079158782959 3.0460827350616455\n",
      "Best error :  3.0437331199645996 \n",
      "\n",
      "Epoch  695\n",
      "Top three least errors :  3.0437331199645996 3.045689105987549 3.046079158782959\n",
      "Best error :  3.0437331199645996 \n",
      "\n",
      "Epoch  696\n",
      "Top three least errors :  3.0437331199645996 3.0438244342803955 3.0443923473358154\n",
      "Best error :  3.043396472930908 \n",
      "\n",
      "Epoch  697\n",
      "Top three least errors :  3.043396472930908 3.0437331199645996 3.0438244342803955\n",
      "Best error :  3.042536735534668 \n",
      "\n",
      "Epoch  698\n",
      "Top three least errors :  3.042536735534668 3.043396472930908 3.0436055660247803\n",
      "Best error :  3.042536735534668 \n",
      "\n",
      "Epoch  699\n",
      "Top three least errors :  3.042536735534668 3.0428171157836914 3.0429575443267822\n",
      "Best error :  3.0422627925872803 \n",
      "\n",
      "Epoch  700\n",
      "Top three least errors :  3.0422627925872803 3.042536735534668 3.0425801277160645\n",
      "Best error :  3.041890859603882 \n",
      "\n",
      "Epoch  701\n",
      "Top three least errors :  3.041890859603882 3.0422627925872803 3.042536735534668\n",
      "Best error :  3.041890859603882 \n",
      "\n",
      "Epoch  702\n",
      "Top three least errors :  3.041890859603882 3.0421571731567383 3.0422275066375732\n",
      "Best error :  3.041890859603882 \n",
      "\n",
      "Epoch  703\n",
      "Top three least errors :  3.041890859603882 3.0420596599578857 3.0421571731567383\n",
      "Best error :  3.041890859603882 \n",
      "\n",
      "Epoch  704\n",
      "Top three least errors :  3.041890859603882 3.04195237159729 3.0420596599578857\n",
      "Best error :  3.0417351722717285 \n",
      "\n",
      "Epoch  705\n",
      "Top three least errors :  3.0417351722717285 3.041890859603882 3.0420596599578857\n",
      "Best error :  3.041330575942993 \n",
      "\n",
      "Epoch  706\n",
      "Top three least errors :  3.041330575942993 3.0417351722717285 3.041877508163452\n",
      "Best error :  3.041330575942993 \n",
      "\n",
      "Epoch  707\n",
      "Top three least errors :  3.041330575942993 3.041409730911255 3.0417351722717285\n",
      "Best error :  3.0400195121765137 \n",
      "\n",
      "Epoch  708\n",
      "Top three least errors :  3.0400195121765137 3.0401387214660645 3.040571451187134\n",
      "Best error :  3.0362908840179443 \n",
      "\n",
      "Epoch  709\n",
      "Top three least errors :  3.0362908840179443 3.0400195121765137 3.0401387214660645\n",
      "Best error :  3.0333261489868164 \n",
      "\n",
      "Epoch  710\n",
      "Top three least errors :  3.0333261489868164 3.0358800888061523 3.0362908840179443\n",
      "Best error :  3.0333261489868164 \n",
      "\n",
      "Epoch  711\n",
      "Top three least errors :  3.0333261489868164 3.0352087020874023 3.0358800888061523\n",
      "Best error :  3.0333261489868164 \n",
      "\n",
      "Epoch  712\n",
      "Top three least errors :  3.0333261489868164 3.034790277481079 3.0348315238952637\n",
      "Best error :  3.0333261489868164 \n",
      "\n",
      "Epoch  713\n",
      "Top three least errors :  3.0333261489868164 3.033418893814087 3.0336530208587646\n",
      "Best error :  3.033064365386963 \n",
      "\n",
      "Epoch  714\n",
      "Top three least errors :  3.033064365386963 3.0333261489868164 3.033418893814087\n",
      "Best error :  3.032893180847168 \n",
      "\n",
      "Epoch  715\n",
      "Top three least errors :  3.032893180847168 3.033064365386963 3.0332343578338623\n",
      "Best error :  3.031353235244751 \n",
      "\n",
      "Epoch  716\n",
      "Top three least errors :  3.031353235244751 3.0322318077087402 3.0328423976898193\n",
      "Best error :  3.031137466430664 \n",
      "\n",
      "Epoch  717\n",
      "Top three least errors :  3.031137466430664 3.031353235244751 3.0319671630859375\n",
      "Best error :  3.030783176422119 \n",
      "\n",
      "Epoch  718\n",
      "Top three least errors :  3.030783176422119 3.030810594558716 3.031137466430664\n",
      "Best error :  3.0307793617248535 \n",
      "\n",
      "Epoch  719\n",
      "Top three least errors :  3.0307793617248535 3.030783176422119 3.030810594558716\n",
      "Best error :  3.03037166595459 \n",
      "\n",
      "Epoch  720\n",
      "Top three least errors :  3.03037166595459 3.0307793617248535 3.030783176422119\n",
      "Best error :  3.030026912689209 \n",
      "\n",
      "Epoch  721\n",
      "Top three least errors :  3.030026912689209 3.03037166595459 3.0305399894714355\n",
      "Best error :  3.0297365188598633 \n",
      "\n",
      "Epoch  722\n",
      "Top three least errors :  3.0297365188598633 3.0298538208007812 3.0299570560455322\n",
      "Best error :  3.0293283462524414 \n",
      "\n",
      "Epoch  723\n",
      "Top three least errors :  3.0293283462524414 3.0297365188598633 3.0298538208007812\n",
      "Best error :  3.02927827835083 \n",
      "\n",
      "Epoch  724\n",
      "Top three least errors :  3.02927827835083 3.0293283462524414 3.029573440551758\n",
      "Best error :  3.0260729789733887 \n",
      "\n",
      "Epoch  725\n",
      "Top three least errors :  3.0260729789733887 3.0288541316986084 3.028961420059204\n",
      "Best error :  3.0257391929626465 \n",
      "\n",
      "Epoch  726\n",
      "Top three least errors :  3.0257391929626465 3.0260729789733887 3.028053045272827\n",
      "Best error :  3.0240561962127686 \n",
      "\n",
      "Epoch  727\n",
      "Top three least errors :  3.0240561962127686 3.0252842903137207 3.0253829956054688\n",
      "Best error :  3.0219924449920654 \n",
      "\n",
      "Epoch  728\n",
      "Top three least errors :  3.0219924449920654 3.0226542949676514 3.0238544940948486\n",
      "Best error :  3.0219924449920654 \n",
      "\n",
      "Epoch  729\n",
      "Top three least errors :  3.0219924449920654 3.0219955444335938 3.0226542949676514\n",
      "Best error :  3.021052837371826 \n",
      "\n",
      "Epoch  730\n",
      "Top three least errors :  3.021052837371826 3.0219924449920654 3.0219955444335938\n",
      "Best error :  3.0197219848632812 \n",
      "\n",
      "Epoch  731\n",
      "Top three least errors :  3.0197219848632812 3.0205421447753906 3.021052837371826\n",
      "Best error :  3.0197219848632812 \n",
      "\n",
      "Epoch  732\n",
      "Top three least errors :  3.0197219848632812 3.0205421447753906 3.021052837371826\n",
      "Best error :  3.019474744796753 \n",
      "\n",
      "Epoch  733\n",
      "Top three least errors :  3.019474744796753 3.0197219848632812 3.0199966430664062\n",
      "Best error :  3.017971992492676 \n",
      "\n",
      "Epoch  734\n",
      "Top three least errors :  3.017971992492676 3.019474744796753 3.0197219848632812\n",
      "Best error :  3.017815113067627 \n",
      "\n",
      "Epoch  735\n",
      "Top three least errors :  3.017815113067627 3.017971992492676 3.019092321395874\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  736\n",
      "Top three least errors :  3.0165538787841797 3.017704725265503 3.017815113067627\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  737\n",
      "Top three least errors :  3.0165538787841797 3.017704725265503 3.017815113067627\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  738\n",
      "Top three least errors :  3.0165538787841797 3.017704725265503 3.017815113067627\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  739\n",
      "Top three least errors :  3.0165538787841797 3.017704725265503 3.017815113067627\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  740\n",
      "Top three least errors :  3.0165538787841797 3.0166070461273193 3.017704725265503\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  741\n",
      "Top three least errors :  3.0165538787841797 3.0166070461273193 3.017704725265503\n",
      "Best error :  3.0165538787841797 \n",
      "\n",
      "Epoch  742\n",
      "Top three least errors :  3.0165538787841797 3.0166070461273193 3.017704725265503\n",
      "Best error :  3.016486406326294 \n",
      "\n",
      "Epoch  743\n",
      "Top three least errors :  3.016486406326294 3.0165538787841797 3.0166070461273193\n",
      "Best error :  3.016486406326294 \n",
      "\n",
      "Epoch  744\n",
      "Top three least errors :  3.016486406326294 3.0165538787841797 3.0166070461273193\n",
      "Best error :  3.013787269592285 \n",
      "\n",
      "Epoch  745\n",
      "Top three least errors :  3.013787269592285 3.014986991882324 3.0153093338012695\n",
      "Best error :  3.01374888420105 \n",
      "\n",
      "Epoch  746\n",
      "Top three least errors :  3.01374888420105 3.013787269592285 3.0142505168914795\n",
      "Best error :  3.01374888420105 \n",
      "\n",
      "Epoch  747\n",
      "Top three least errors :  3.01374888420105 3.013787269592285 3.0141117572784424\n",
      "Best error :  3.01374888420105 \n",
      "\n",
      "Epoch  748\n",
      "Top three least errors :  3.01374888420105 3.013787269592285 3.0141117572784424\n",
      "Best error :  3.012639045715332 \n",
      "\n",
      "Epoch  749\n",
      "Top three least errors :  3.012639045715332 3.01374888420105 3.013787269592285\n",
      "Best error :  3.0103726387023926 \n",
      "\n",
      "Epoch  750\n",
      "Top three least errors :  3.0103726387023926 3.012639045715332 3.013075828552246\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  3.0103726387023926 \n",
      "\n",
      "Epoch  751\n",
      "Top three least errors :  3.0103726387023926 3.0110201835632324 3.0117878913879395\n",
      "Best error :  3.0101211071014404 \n",
      "\n",
      "Epoch  752\n",
      "Top three least errors :  3.0101211071014404 3.0102782249450684 3.010343551635742\n",
      "Best error :  3.008171796798706 \n",
      "\n",
      "Epoch  753\n",
      "Top three least errors :  3.008171796798706 3.009690999984741 3.0098485946655273\n",
      "Best error :  3.0062694549560547 \n",
      "\n",
      "Epoch  754\n",
      "Top three least errors :  3.0062694549560547 3.008171796798706 3.009690999984741\n",
      "Best error :  3.006251096725464 \n",
      "\n",
      "Epoch  755\n",
      "Top three least errors :  3.006251096725464 3.0062694549560547 3.007556438446045\n",
      "Best error :  3.0048604011535645 \n",
      "\n",
      "Epoch  756\n",
      "Top three least errors :  3.0048604011535645 3.0059006214141846 3.006251096725464\n",
      "Best error :  3.0031912326812744 \n",
      "\n",
      "Epoch  757\n",
      "Top three least errors :  3.0031912326812744 3.0033373832702637 3.0033481121063232\n",
      "Best error :  3.0031912326812744 \n",
      "\n",
      "Epoch  758\n",
      "Top three least errors :  3.0031912326812744 3.0033373832702637 3.0033481121063232\n",
      "Best error :  3.0031912326812744 \n",
      "\n",
      "Epoch  759\n",
      "Top three least errors :  3.0031912326812744 3.003249406814575 3.0033373832702637\n",
      "Best error :  3.0031912326812744 \n",
      "\n",
      "Epoch  760\n",
      "Top three least errors :  3.0031912326812744 3.003249406814575 3.0033373832702637\n",
      "Best error :  3.002971649169922 \n",
      "\n",
      "Epoch  761\n",
      "Top three least errors :  3.002971649169922 3.0031912326812744 3.003249406814575\n",
      "Best error :  3.002971649169922 \n",
      "\n",
      "Epoch  762\n",
      "Top three least errors :  3.002971649169922 3.002974033355713 3.0031912326812744\n",
      "Best error :  3.002971649169922 \n",
      "\n",
      "Epoch  763\n",
      "Top three least errors :  3.002971649169922 3.002974033355713 3.002998113632202\n",
      "Best error :  3.002535820007324 \n",
      "\n",
      "Epoch  764\n",
      "Top three least errors :  3.002535820007324 3.0026817321777344 3.002971649169922\n",
      "Best error :  3.001659393310547 \n",
      "\n",
      "Epoch  765\n",
      "Top three least errors :  3.001659393310547 3.0023839473724365 3.002535820007324\n",
      "Best error :  3.0009377002716064 \n",
      "\n",
      "Epoch  766\n",
      "Top three least errors :  3.0009377002716064 3.001659393310547 3.0018796920776367\n",
      "Best error :  3.0009377002716064 \n",
      "\n",
      "Epoch  767\n",
      "Top three least errors :  3.0009377002716064 3.0010201930999756 3.0015621185302734\n",
      "Best error :  2.998924493789673 \n",
      "\n",
      "Epoch  768\n",
      "Top three least errors :  2.998924493789673 3.00059175491333 3.0009377002716064\n",
      "Best error :  2.998924493789673 \n",
      "\n",
      "Epoch  769\n",
      "Top three least errors :  2.998924493789673 2.99945068359375 3.0000317096710205\n",
      "Best error :  2.997577428817749 \n",
      "\n",
      "Epoch  770\n",
      "Top three least errors :  2.997577428817749 2.998361825942993 2.998924493789673\n",
      "Best error :  2.997577428817749 \n",
      "\n",
      "Epoch  771\n",
      "Top three least errors :  2.997577428817749 2.998361825942993 2.998924493789673\n",
      "Best error :  2.9962728023529053 \n",
      "\n",
      "Epoch  772\n",
      "Top three least errors :  2.9962728023529053 2.997480630874634 2.997577428817749\n",
      "Best error :  2.9925577640533447 \n",
      "\n",
      "Epoch  773\n",
      "Top three least errors :  2.9925577640533447 2.9962728023529053 2.997480630874634\n",
      "Best error :  2.9925577640533447 \n",
      "\n",
      "Epoch  774\n",
      "Top three least errors :  2.9925577640533447 2.9930999279022217 2.9931576251983643\n",
      "Best error :  2.9925577640533447 \n",
      "\n",
      "Epoch  775\n",
      "Top three least errors :  2.9925577640533447 2.9925642013549805 2.9930567741394043\n",
      "Best error :  2.9925153255462646 \n",
      "\n",
      "Epoch  776\n",
      "Top three least errors :  2.9925153255462646 2.9925577640533447 2.9925642013549805\n",
      "Best error :  2.992156505584717 \n",
      "\n",
      "Epoch  777\n",
      "Top three least errors :  2.992156505584717 2.9925153255462646 2.9925577640533447\n",
      "Best error :  2.992156505584717 \n",
      "\n",
      "Epoch  778\n",
      "Top three least errors :  2.992156505584717 2.9922525882720947 2.9922702312469482\n",
      "Best error :  2.991124153137207 \n",
      "\n",
      "Epoch  779\n",
      "Top three least errors :  2.991124153137207 2.9916276931762695 2.991717576980591\n",
      "Best error :  2.991124153137207 \n",
      "\n",
      "Epoch  780\n",
      "Top three least errors :  2.991124153137207 2.9916276931762695 2.991717576980591\n",
      "Best error :  2.9904541969299316 \n",
      "\n",
      "Epoch  781\n",
      "Top three least errors :  2.9904541969299316 2.991124153137207 2.9916276931762695\n",
      "Best error :  2.9904541969299316 \n",
      "\n",
      "Epoch  782\n",
      "Top three least errors :  2.9904541969299316 2.991124153137207 2.9916276931762695\n",
      "Best error :  2.9904541969299316 \n",
      "\n",
      "Epoch  783\n",
      "Top three least errors :  2.9904541969299316 2.9905567169189453 2.991124153137207\n",
      "Best error :  2.990436553955078 \n",
      "\n",
      "Epoch  784\n",
      "Top three least errors :  2.990436553955078 2.9904541969299316 2.9905595779418945\n",
      "Best error :  2.9902303218841553 \n",
      "\n",
      "Epoch  785\n",
      "Top three least errors :  2.9902303218841553 2.990330696105957 2.9904069900512695\n",
      "Best error :  2.9893200397491455 \n",
      "\n",
      "Epoch  786\n",
      "Top three least errors :  2.9893200397491455 2.9902303218841553 2.990330696105957\n",
      "Best error :  2.9852890968322754 \n",
      "\n",
      "Epoch  787\n",
      "Top three least errors :  2.9852890968322754 2.987548828125 2.9893200397491455\n",
      "Best error :  2.9852890968322754 \n",
      "\n",
      "Epoch  788\n",
      "Top three least errors :  2.9852890968322754 2.9861302375793457 2.986337661743164\n",
      "Best error :  2.9840216636657715 \n",
      "\n",
      "Epoch  789\n",
      "Top three least errors :  2.9840216636657715 2.9847564697265625 2.9852890968322754\n",
      "Best error :  2.9840216636657715 \n",
      "\n",
      "Epoch  790\n",
      "Top three least errors :  2.9840216636657715 2.9843244552612305 2.9846537113189697\n",
      "Best error :  2.9840216636657715 \n",
      "\n",
      "Epoch  791\n",
      "Top three least errors :  2.9840216636657715 2.9843244552612305 2.9846537113189697\n",
      "Best error :  2.9833524227142334 \n",
      "\n",
      "Epoch  792\n",
      "Top three least errors :  2.9833524227142334 2.9840216636657715 2.9842638969421387\n",
      "Best error :  2.982901096343994 \n",
      "\n",
      "Epoch  793\n",
      "Top three least errors :  2.982901096343994 2.9833524227142334 2.9840216636657715\n",
      "Best error :  2.9826245307922363 \n",
      "\n",
      "Epoch  794\n",
      "Top three least errors :  2.9826245307922363 2.982783794403076 2.982901096343994\n",
      "Best error :  2.9826245307922363 \n",
      "\n",
      "Epoch  795\n",
      "Top three least errors :  2.9826245307922363 2.982783794403076 2.982901096343994\n",
      "Best error :  2.9823858737945557 \n",
      "\n",
      "Epoch  796\n",
      "Top three least errors :  2.9823858737945557 2.9825167655944824 2.9826245307922363\n",
      "Best error :  2.9821114540100098 \n",
      "\n",
      "Epoch  797\n",
      "Top three least errors :  2.9821114540100098 2.982351541519165 2.9823858737945557\n",
      "Best error :  2.979964256286621 \n",
      "\n",
      "Epoch  798\n",
      "Top three least errors :  2.979964256286621 2.981600761413574 2.9821114540100098\n",
      "Best error :  2.979964256286621 \n",
      "\n",
      "Epoch  799\n",
      "Top three least errors :  2.979964256286621 2.981600761413574 2.9817845821380615\n",
      "Best error :  2.9796760082244873 \n",
      "\n",
      "Epoch  800\n",
      "Top three least errors :  2.9796760082244873 2.979964256286621 2.981600761413574\n",
      "Best error :  2.9789509773254395 \n",
      "\n",
      "Epoch  801\n",
      "Top three least errors :  2.9789509773254395 2.9796760082244873 2.979964256286621\n",
      "Best error :  2.9772276878356934 \n",
      "\n",
      "Epoch  802\n",
      "Top three least errors :  2.9772276878356934 2.9774363040924072 2.9779696464538574\n",
      "Best error :  2.9771690368652344 \n",
      "\n",
      "Epoch  803\n",
      "Top three least errors :  2.9771690368652344 2.9772276878356934 2.9774363040924072\n",
      "Best error :  2.975599527359009 \n",
      "\n",
      "Epoch  804\n",
      "Top three least errors :  2.975599527359009 2.9760310649871826 2.9771690368652344\n",
      "Best error :  2.9710357189178467 \n",
      "\n",
      "Epoch  805\n",
      "Top three least errors :  2.9710357189178467 2.975599527359009 2.9760310649871826\n",
      "Best error :  2.9710357189178467 \n",
      "\n",
      "Epoch  806\n",
      "Top three least errors :  2.9710357189178467 2.97338604927063 2.9734203815460205\n",
      "Best error :  2.9689416885375977 \n",
      "\n",
      "Epoch  807\n",
      "Top three least errors :  2.9689416885375977 2.970629930496216 2.9710357189178467\n",
      "Best error :  2.9676578044891357 \n",
      "\n",
      "Epoch  808\n",
      "Top three least errors :  2.9676578044891357 2.968574285507202 2.9689416885375977\n",
      "Best error :  2.9667019844055176 \n",
      "\n",
      "Epoch  809\n",
      "Top three least errors :  2.9667019844055176 2.9673078060150146 2.9676578044891357\n",
      "Best error :  2.9665327072143555 \n",
      "\n",
      "Epoch  810\n",
      "Top three least errors :  2.9665327072143555 2.9667019844055176 2.9673004150390625\n",
      "Best error :  2.963796854019165 \n",
      "\n",
      "Epoch  811\n",
      "Top three least errors :  2.963796854019165 2.9665327072143555 2.9667019844055176\n",
      "Best error :  2.9635982513427734 \n",
      "\n",
      "Epoch  812\n",
      "Top three least errors :  2.9635982513427734 2.963796854019165 2.9665327072143555\n",
      "Best error :  2.9635982513427734 \n",
      "\n",
      "Epoch  813\n",
      "Top three least errors :  2.9635982513427734 2.963796854019165 2.965193748474121\n",
      "Best error :  2.962242841720581 \n",
      "\n",
      "Epoch  814\n",
      "Top three least errors :  2.962242841720581 2.9624509811401367 2.9631595611572266\n",
      "Best error :  2.9620437622070312 \n",
      "\n",
      "Epoch  815\n",
      "Top three least errors :  2.9620437622070312 2.962242841720581 2.9624509811401367\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  2.9620437622070312 \n",
      "\n",
      "Epoch  816\n",
      "Top three least errors :  2.9620437622070312 2.962242841720581 2.9624509811401367\n",
      "Best error :  2.957608222961426 \n",
      "\n",
      "Epoch  817\n",
      "Top three least errors :  2.957608222961426 2.9610683917999268 2.9620394706726074\n",
      "Best error :  2.9573686122894287 \n",
      "\n",
      "Epoch  818\n",
      "Top three least errors :  2.9573686122894287 2.957608222961426 2.9580726623535156\n",
      "Best error :  2.953841209411621 \n",
      "\n",
      "Epoch  819\n",
      "Top three least errors :  2.953841209411621 2.956244468688965 2.956939697265625\n",
      "Best error :  2.953458547592163 \n",
      "\n",
      "Epoch  820\n",
      "Top three least errors :  2.953458547592163 2.953841209411621 2.954582691192627\n",
      "Best error :  2.953000783920288 \n",
      "\n",
      "Epoch  821\n",
      "Top three least errors :  2.953000783920288 2.953402280807495 2.953458547592163\n",
      "Best error :  2.9473483562469482 \n",
      "\n",
      "Epoch  822\n",
      "Top three least errors :  2.9473483562469482 2.952239751815796 2.952472686767578\n",
      "Best error :  2.9455411434173584 \n",
      "\n",
      "Epoch  823\n",
      "Top three least errors :  2.9455411434173584 2.9473483562469482 2.949242353439331\n",
      "Best error :  2.9436306953430176 \n",
      "\n",
      "Epoch  824\n",
      "Top three least errors :  2.9436306953430176 2.944406747817993 2.944479465484619\n",
      "Best error :  2.9434897899627686 \n",
      "\n",
      "Epoch  825\n",
      "Top three least errors :  2.9434897899627686 2.944406747817993 2.944479465484619\n",
      "Best error :  2.9434897899627686 \n",
      "\n",
      "Epoch  826\n",
      "Top three least errors :  2.9434897899627686 2.94364595413208 2.944406747817993\n",
      "Best error :  2.941812038421631 \n",
      "\n",
      "Epoch  827\n",
      "Top three least errors :  2.941812038421631 2.9418582916259766 2.9424750804901123\n",
      "Best error :  2.940023899078369 \n",
      "\n",
      "Epoch  828\n",
      "Top three least errors :  2.940023899078369 2.941812038421631 2.9418582916259766\n",
      "Best error :  2.940023899078369 \n",
      "\n",
      "Epoch  829\n",
      "Top three least errors :  2.940023899078369 2.9412012100219727 2.941795825958252\n",
      "Best error :  2.9394233226776123 \n",
      "\n",
      "Epoch  830\n",
      "Top three least errors :  2.9394233226776123 2.940023899078369 2.9400672912597656\n",
      "Best error :  2.9389636516571045 \n",
      "\n",
      "Epoch  831\n",
      "Top three least errors :  2.9389636516571045 2.9394233226776123 2.9400672912597656\n",
      "Best error :  2.938282012939453 \n",
      "\n",
      "Epoch  832\n",
      "Top three least errors :  2.938282012939453 2.9389636516571045 2.9390034675598145\n",
      "Best error :  2.9372427463531494 \n",
      "\n",
      "Epoch  833\n",
      "Top three least errors :  2.9372427463531494 2.938091516494751 2.938282012939453\n",
      "Best error :  2.9347751140594482 \n",
      "\n",
      "Epoch  834\n",
      "Top three least errors :  2.9347751140594482 2.9348902702331543 2.935899257659912\n",
      "Best error :  2.932741165161133 \n",
      "\n",
      "Epoch  835\n",
      "Top three least errors :  2.932741165161133 2.932936906814575 2.934614419937134\n",
      "Best error :  2.92928409576416 \n",
      "\n",
      "Epoch  836\n",
      "Top three least errors :  2.92928409576416 2.931342601776123 2.932741165161133\n",
      "Best error :  2.9287877082824707 \n",
      "\n",
      "Epoch  837\n",
      "Top three least errors :  2.9287877082824707 2.92928409576416 2.931342601776123\n",
      "Best error :  2.9250376224517822 \n",
      "\n",
      "Epoch  838\n",
      "Top three least errors :  2.9250376224517822 2.9263916015625 2.927976131439209\n",
      "Best error :  2.9181442260742188 \n",
      "\n",
      "Epoch  839\n",
      "Top three least errors :  2.9181442260742188 2.9234304428100586 2.9250376224517822\n",
      "Best error :  2.9181442260742188 \n",
      "\n",
      "Epoch  840\n",
      "Top three least errors :  2.9181442260742188 2.918165922164917 2.918826103210449\n",
      "Best error :  2.9165380001068115 \n",
      "\n",
      "Epoch  841\n",
      "Top three least errors :  2.9165380001068115 2.9176084995269775 2.9181442260742188\n",
      "Best error :  2.9153823852539062 \n",
      "\n",
      "Epoch  842\n",
      "Top three least errors :  2.9153823852539062 2.9161059856414795 2.9165380001068115\n",
      "Best error :  2.914240837097168 \n",
      "\n",
      "Epoch  843\n",
      "Top three least errors :  2.914240837097168 2.9153823852539062 2.9161059856414795\n",
      "Best error :  2.9131646156311035 \n",
      "\n",
      "Epoch  844\n",
      "Top three least errors :  2.9131646156311035 2.9135305881500244 2.914240837097168\n",
      "Best error :  2.912647008895874 \n",
      "\n",
      "Epoch  845\n",
      "Top three least errors :  2.912647008895874 2.9131646156311035 2.9135305881500244\n",
      "Best error :  2.912217855453491 \n",
      "\n",
      "Epoch  846\n",
      "Top three least errors :  2.912217855453491 2.9124581813812256 2.912647008895874\n",
      "Best error :  2.9102375507354736 \n",
      "\n",
      "Epoch  847\n",
      "Top three least errors :  2.9102375507354736 2.9107658863067627 2.911726236343384\n",
      "Best error :  2.9102375507354736 \n",
      "\n",
      "Epoch  848\n",
      "Top three least errors :  2.9102375507354736 2.9106032848358154 2.9107658863067627\n",
      "Best error :  2.9098727703094482 \n",
      "\n",
      "Epoch  849\n",
      "Top three least errors :  2.9098727703094482 2.910060167312622 2.9102375507354736\n",
      "Best error :  2.909306049346924 \n",
      "\n",
      "Epoch  850\n",
      "Top three least errors :  2.909306049346924 2.9098727703094482 2.910060167312622\n",
      "Best error :  2.908686876296997 \n",
      "\n",
      "Epoch  851\n",
      "Top three least errors :  2.908686876296997 2.909306049346924 2.9095377922058105\n",
      "Best error :  2.908066749572754 \n",
      "\n",
      "Epoch  852\n",
      "Top three least errors :  2.908066749572754 2.908503770828247 2.908686876296997\n",
      "Best error :  2.908066749572754 \n",
      "\n",
      "Epoch  853\n",
      "Top three least errors :  2.908066749572754 2.908503770828247 2.908686876296997\n",
      "Best error :  2.907341480255127 \n",
      "\n",
      "Epoch  854\n",
      "Top three least errors :  2.907341480255127 2.908066749572754 2.9081296920776367\n",
      "Best error :  2.9072723388671875 \n",
      "\n",
      "Epoch  855\n",
      "Top three least errors :  2.9072723388671875 2.907341480255127 2.907423496246338\n",
      "Best error :  2.9066920280456543 \n",
      "\n",
      "Epoch  856\n",
      "Top three least errors :  2.9066920280456543 2.9072723388671875 2.907341480255127\n",
      "Best error :  2.906188726425171 \n",
      "\n",
      "Epoch  857\n",
      "Top three least errors :  2.906188726425171 2.9066920280456543 2.9072723388671875\n",
      "Best error :  2.903628349304199 \n",
      "\n",
      "Epoch  858\n",
      "Top three least errors :  2.903628349304199 2.9060282707214355 2.906073570251465\n",
      "Best error :  2.903628349304199 \n",
      "\n",
      "Epoch  859\n",
      "Top three least errors :  2.903628349304199 2.9045228958129883 2.905740976333618\n",
      "Best error :  2.9032325744628906 \n",
      "\n",
      "Epoch  860\n",
      "Top three least errors :  2.9032325744628906 2.903628349304199 2.9036617279052734\n",
      "Best error :  2.902251958847046 \n",
      "\n",
      "Epoch  861\n",
      "Top three least errors :  2.902251958847046 2.902620792388916 2.9026663303375244\n",
      "Best error :  2.9014809131622314 \n",
      "\n",
      "Epoch  862\n",
      "Top three least errors :  2.9014809131622314 2.902251958847046 2.902620792388916\n",
      "Best error :  2.9014809131622314 \n",
      "\n",
      "Epoch  863\n",
      "Top three least errors :  2.9014809131622314 2.902231454849243 2.902251958847046\n",
      "Best error :  2.9011447429656982 \n",
      "\n",
      "Epoch  864\n",
      "Top three least errors :  2.9011447429656982 2.9014809131622314 2.901895761489868\n",
      "Best error :  2.9011447429656982 \n",
      "\n",
      "Epoch  865\n",
      "Top three least errors :  2.9011447429656982 2.901318311691284 2.9014809131622314\n",
      "Best error :  2.9010393619537354 \n",
      "\n",
      "Epoch  866\n",
      "Top three least errors :  2.9010393619537354 2.9011447429656982 2.901318311691284\n",
      "Best error :  2.899902582168579 \n",
      "\n",
      "Epoch  867\n",
      "Top three least errors :  2.899902582168579 2.9000957012176514 2.9007294178009033\n",
      "Best error :  2.8993029594421387 \n",
      "\n",
      "Epoch  868\n",
      "Top three least errors :  2.8993029594421387 2.899902582168579 2.9000957012176514\n",
      "Best error :  2.8993029594421387 \n",
      "\n",
      "Epoch  869\n",
      "Top three least errors :  2.8993029594421387 2.899902582168579 2.9000957012176514\n",
      "Best error :  2.8988442420959473 \n",
      "\n",
      "Epoch  870\n",
      "Top three least errors :  2.8988442420959473 2.8993029594421387 2.8996195793151855\n",
      "Best error :  2.8988373279571533 \n",
      "\n",
      "Epoch  871\n",
      "Top three least errors :  2.8988373279571533 2.8988442420959473 2.8991456031799316\n",
      "Best error :  2.8983843326568604 \n",
      "\n",
      "Epoch  872\n",
      "Top three least errors :  2.8983843326568604 2.8988373279571533 2.8988442420959473\n",
      "Best error :  2.8971450328826904 \n",
      "\n",
      "Epoch  873\n",
      "Top three least errors :  2.8971450328826904 2.8983843326568604 2.8988373279571533\n",
      "Best error :  2.8971450328826904 \n",
      "\n",
      "Epoch  874\n",
      "Top three least errors :  2.8971450328826904 2.8975906372070312 2.897613048553467\n",
      "Best error :  2.8971450328826904 \n",
      "\n",
      "Epoch  875\n",
      "Top three least errors :  2.8971450328826904 2.8974006175994873 2.897587299346924\n",
      "Best error :  2.896806478500366 \n",
      "\n",
      "Epoch  876\n",
      "Top three least errors :  2.896806478500366 2.897059202194214 2.8971450328826904\n",
      "Best error :  2.896806478500366 \n",
      "\n",
      "Epoch  877\n",
      "Top three least errors :  2.896806478500366 2.897059202194214 2.8971242904663086\n",
      "Best error :  2.8964004516601562 \n",
      "\n",
      "Epoch  878\n",
      "Top three least errors :  2.8964004516601562 2.896806478500366 2.8969004154205322\n",
      "Best error :  2.8959710597991943 \n",
      "\n",
      "Epoch  879\n",
      "Top three least errors :  2.8959710597991943 2.8963112831115723 2.8964004516601562\n",
      "Best error :  2.8955330848693848 \n",
      "\n",
      "Epoch  880\n",
      "Top three least errors :  2.8955330848693848 2.89575457572937 2.895890235900879\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  2.8955330848693848 \n",
      "\n",
      "Epoch  881\n",
      "Top three least errors :  2.8955330848693848 2.89575457572937 2.8957972526550293\n",
      "Best error :  2.89542555809021 \n",
      "\n",
      "Epoch  882\n",
      "Top three least errors :  2.89542555809021 2.8955330848693848 2.895761489868164\n",
      "Best error :  2.8953895568847656 \n",
      "\n",
      "Epoch  883\n",
      "Top three least errors :  2.8953895568847656 2.89542555809021 2.895685911178589\n",
      "Best error :  2.8953895568847656 \n",
      "\n",
      "Epoch  884\n",
      "Top three least errors :  2.8953895568847656 2.89542555809021 2.895685911178589\n",
      "Best error :  2.8953895568847656 \n",
      "\n",
      "Epoch  885\n",
      "Top three least errors :  2.8953895568847656 2.89542555809021 2.8956291675567627\n",
      "Best error :  2.8953895568847656 \n",
      "\n",
      "Epoch  886\n",
      "Top three least errors :  2.8953895568847656 2.89542555809021 2.8955602645874023\n",
      "Best error :  2.895354986190796 \n",
      "\n",
      "Epoch  887\n",
      "Top three least errors :  2.895354986190796 2.8953895568847656 2.895416021347046\n",
      "Best error :  2.894406318664551 \n",
      "\n",
      "Epoch  888\n",
      "Top three least errors :  2.894406318664551 2.894421100616455 2.8947155475616455\n",
      "Best error :  2.894406318664551 \n",
      "\n",
      "Epoch  889\n",
      "Top three least errors :  2.894406318664551 2.894421100616455 2.8947155475616455\n",
      "Best error :  2.894406318664551 \n",
      "\n",
      "Epoch  890\n",
      "Top three least errors :  2.894406318664551 2.894421100616455 2.8947155475616455\n",
      "Best error :  2.894406318664551 \n",
      "\n",
      "Epoch  891\n",
      "Top three least errors :  2.894406318664551 2.894411563873291 2.894421100616455\n",
      "Best error :  2.8944051265716553 \n",
      "\n",
      "Epoch  892\n",
      "Top three least errors :  2.8944051265716553 2.894406318664551 2.894411563873291\n",
      "Best error :  2.8932275772094727 \n",
      "\n",
      "Epoch  893\n",
      "Top three least errors :  2.8932275772094727 2.8944051265716553 2.894406318664551\n",
      "Best error :  2.8932275772094727 \n",
      "\n",
      "Epoch  894\n",
      "Top three least errors :  2.8932275772094727 2.893360137939453 2.893719434738159\n",
      "Best error :  2.8932275772094727 \n",
      "\n",
      "Epoch  895\n",
      "Top three least errors :  2.8932275772094727 2.893360137939453 2.8934450149536133\n",
      "Best error :  2.8931884765625 \n",
      "\n",
      "Epoch  896\n",
      "Top three least errors :  2.8931884765625 2.8932275772094727 2.893360137939453\n",
      "Best error :  2.892313241958618 \n",
      "\n",
      "Epoch  897\n",
      "Top three least errors :  2.892313241958618 2.8931405544281006 2.8931884765625\n",
      "Best error :  2.891932964324951 \n",
      "\n",
      "Epoch  898\n",
      "Top three least errors :  2.891932964324951 2.892313241958618 2.8931405544281006\n",
      "Best error :  2.8906097412109375 \n",
      "\n",
      "Epoch  899\n",
      "Top three least errors :  2.8906097412109375 2.891615390777588 2.8916707038879395\n",
      "Best error :  2.889134168624878 \n",
      "\n",
      "Epoch  900\n",
      "Top three least errors :  2.889134168624878 2.8898255825042725 2.8901119232177734\n",
      "Best error :  2.889134168624878 \n",
      "\n",
      "Epoch  901\n",
      "Top three least errors :  2.889134168624878 2.8894190788269043 2.8896725177764893\n",
      "Best error :  2.888317584991455 \n",
      "\n",
      "Epoch  902\n",
      "Top three least errors :  2.888317584991455 2.8885385990142822 2.888779878616333\n",
      "Best error :  2.888171434402466 \n",
      "\n",
      "Epoch  903\n",
      "Top three least errors :  2.888171434402466 2.888317584991455 2.8885385990142822\n",
      "Best error :  2.887507438659668 \n",
      "\n",
      "Epoch  904\n",
      "Top three least errors :  2.887507438659668 2.8880350589752197 2.888171434402466\n",
      "Best error :  2.8868184089660645 \n",
      "\n",
      "Epoch  905\n",
      "Top three least errors :  2.8868184089660645 2.886855125427246 2.887282133102417\n",
      "Best error :  2.8861727714538574 \n",
      "\n",
      "Epoch  906\n",
      "Top three least errors :  2.8861727714538574 2.8862192630767822 2.8863837718963623\n",
      "Best error :  2.8860085010528564 \n",
      "\n",
      "Epoch  907\n",
      "Top three least errors :  2.8860085010528564 2.886073350906372 2.8861727714538574\n",
      "Best error :  2.885392189025879 \n",
      "\n",
      "Epoch  908\n",
      "Top three least errors :  2.885392189025879 2.8859341144561768 2.8860085010528564\n",
      "Best error :  2.884312629699707 \n",
      "\n",
      "Epoch  909\n",
      "Top three least errors :  2.884312629699707 2.884840965270996 2.8849244117736816\n",
      "Best error :  2.884312629699707 \n",
      "\n",
      "Epoch  910\n",
      "Top three least errors :  2.884312629699707 2.884509801864624 2.884840965270996\n",
      "Best error :  2.884312629699707 \n",
      "\n",
      "Epoch  911\n",
      "Top three least errors :  2.884312629699707 2.884413242340088 2.8844358921051025\n",
      "Best error :  2.88356614112854 \n",
      "\n",
      "Epoch  912\n",
      "Top three least errors :  2.88356614112854 2.883835554122925 2.884066581726074\n",
      "Best error :  2.8823719024658203 \n",
      "\n",
      "Epoch  913\n",
      "Top three least errors :  2.8823719024658203 2.882723569869995 2.883429765701294\n",
      "Best error :  2.8823719024658203 \n",
      "\n",
      "Epoch  914\n",
      "Top three least errors :  2.8823719024658203 2.8824522495269775 2.8825042247772217\n",
      "Best error :  2.8823719024658203 \n",
      "\n",
      "Epoch  915\n",
      "Top three least errors :  2.8823719024658203 2.882402181625366 2.8824310302734375\n",
      "Best error :  2.8823065757751465 \n",
      "\n",
      "Epoch  916\n",
      "Top three least errors :  2.8823065757751465 2.8823719024658203 2.882402181625366\n",
      "Best error :  2.8816733360290527 \n",
      "\n",
      "Epoch  917\n",
      "Top three least errors :  2.8816733360290527 2.881742000579834 2.8821117877960205\n",
      "Best error :  2.88165020942688 \n",
      "\n",
      "Epoch  918\n",
      "Top three least errors :  2.88165020942688 2.8816733360290527 2.881742000579834\n",
      "Best error :  2.881268262863159 \n",
      "\n",
      "Epoch  919\n",
      "Top three least errors :  2.881268262863159 2.88165020942688 2.8816733360290527\n",
      "Best error :  2.8812029361724854 \n",
      "\n",
      "Epoch  920\n",
      "Top three least errors :  2.8812029361724854 2.881268262863159 2.881516695022583\n",
      "Best error :  2.8809733390808105 \n",
      "\n",
      "Epoch  921\n",
      "Top three least errors :  2.8809733390808105 2.8812029361724854 2.881268262863159\n",
      "Best error :  2.8808541297912598 \n",
      "\n",
      "Epoch  922\n",
      "Top three least errors :  2.8808541297912598 2.880901336669922 2.8809614181518555\n",
      "Best error :  2.8808541297912598 \n",
      "\n",
      "Epoch  923\n",
      "Top three least errors :  2.8808541297912598 2.880901336669922 2.8809614181518555\n",
      "Best error :  2.8808541297912598 \n",
      "\n",
      "Epoch  924\n",
      "Top three least errors :  2.8808541297912598 2.880901336669922 2.8809218406677246\n",
      "Best error :  2.8808541297912598 \n",
      "\n",
      "Epoch  925\n",
      "Top three least errors :  2.8808541297912598 2.880901336669922 2.8809218406677246\n",
      "Best error :  2.880525827407837 \n",
      "\n",
      "Epoch  926\n",
      "Top three least errors :  2.880525827407837 2.88081431388855 2.8808326721191406\n",
      "Best error :  2.8804943561553955 \n",
      "\n",
      "Epoch  927\n",
      "Top three least errors :  2.8804943561553955 2.880525827407837 2.88081431388855\n",
      "Best error :  2.8799102306365967 \n",
      "\n",
      "Epoch  928\n",
      "Top three least errors :  2.8799102306365967 2.8803536891937256 2.8804285526275635\n",
      "Best error :  2.879049062728882 \n",
      "\n",
      "Epoch  929\n",
      "Top three least errors :  2.879049062728882 2.8797531127929688 2.8798515796661377\n",
      "Best error :  2.8786001205444336 \n",
      "\n",
      "Epoch  930\n",
      "Top three least errors :  2.8786001205444336 2.879049062728882 2.8791232109069824\n",
      "Best error :  2.8774354457855225 \n",
      "\n",
      "Epoch  931\n",
      "Top three least errors :  2.8774354457855225 2.8776631355285645 2.8781778812408447\n",
      "Best error :  2.876410961151123 \n",
      "\n",
      "Epoch  932\n",
      "Top three least errors :  2.876410961151123 2.877427577972412 2.8774354457855225\n",
      "Best error :  2.875715732574463 \n",
      "\n",
      "Epoch  933\n",
      "Top three least errors :  2.875715732574463 2.8758177757263184 2.8759288787841797\n",
      "Best error :  2.874152898788452 \n",
      "\n",
      "Epoch  934\n",
      "Top three least errors :  2.874152898788452 2.8743457794189453 2.8744354248046875\n",
      "Best error :  2.872195243835449 \n",
      "\n",
      "Epoch  935\n",
      "Top three least errors :  2.872195243835449 2.8730056285858154 2.873318672180176\n",
      "Best error :  2.871657133102417 \n",
      "\n",
      "Epoch  936\n",
      "Top three least errors :  2.871657133102417 2.871748924255371 2.87212872505188\n",
      "Best error :  2.871188163757324 \n",
      "\n",
      "Epoch  937\n",
      "Top three least errors :  2.871188163757324 2.8713836669921875 2.8715665340423584\n",
      "Best error :  2.8706071376800537 \n",
      "\n",
      "Epoch  938\n",
      "Top three least errors :  2.8706071376800537 2.870934247970581 2.8710596561431885\n",
      "Best error :  2.8693032264709473 \n",
      "\n",
      "Epoch  939\n",
      "Top three least errors :  2.8693032264709473 2.8703389167785645 2.870537042617798\n",
      "Best error :  2.8691539764404297 \n",
      "\n",
      "Epoch  940\n",
      "Top three least errors :  2.8691539764404297 2.8693032264709473 2.8694286346435547\n",
      "Best error :  2.867570400238037 \n",
      "\n",
      "Epoch  941\n",
      "Top three least errors :  2.867570400238037 2.8678407669067383 2.8686673641204834\n",
      "Best error :  2.8672196865081787 \n",
      "\n",
      "Epoch  942\n",
      "Top three least errors :  2.8672196865081787 2.867570400238037 2.8676743507385254\n",
      "Best error :  2.8672196865081787 \n",
      "\n",
      "Epoch  943\n",
      "Top three least errors :  2.8672196865081787 2.867400884628296 2.867570400238037\n",
      "Best error :  2.8671388626098633 \n",
      "\n",
      "Epoch  944\n",
      "Top three least errors :  2.8671388626098633 2.867218494415283 2.8672196865081787\n",
      "Best error :  2.865902900695801 \n",
      "\n",
      "Epoch  945\n",
      "Top three least errors :  2.865902900695801 2.867053508758545 2.8671388626098633\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best error :  2.864943504333496 \n",
      "\n",
      "Epoch  946\n",
      "Top three least errors :  2.864943504333496 2.865902900695801 2.8661258220672607\n",
      "Best error :  2.864943504333496 \n",
      "\n",
      "Epoch  947\n",
      "Top three least errors :  2.864943504333496 2.8650519847869873 2.8650755882263184\n",
      "Best error :  2.8646178245544434 \n",
      "\n",
      "Epoch  948\n",
      "Top three least errors :  2.8646178245544434 2.864805221557617 2.864816188812256\n",
      "Best error :  2.863661050796509 \n",
      "\n",
      "Epoch  949\n",
      "Top three least errors :  2.863661050796509 2.8637967109680176 2.8639578819274902\n",
      "Best error :  2.8630499839782715 \n",
      "\n",
      "Epoch  950\n",
      "Top three least errors :  2.8630499839782715 2.863618850708008 2.8636910915374756\n",
      "Best error :  2.862419605255127 \n",
      "\n",
      "Epoch  951\n",
      "Top three least errors :  2.862419605255127 2.8628509044647217 2.8630499839782715\n",
      "Best error :  2.862334728240967 \n",
      "\n",
      "Epoch  952\n",
      "Top three least errors :  2.862334728240967 2.862419605255127 2.862478494644165\n",
      "Best error :  2.861821174621582 \n",
      "\n",
      "Epoch  953\n",
      "Top three least errors :  2.861821174621582 2.861896514892578 2.8621456623077393\n",
      "Best error :  2.8611176013946533 \n",
      "\n",
      "Epoch  954\n",
      "Top three least errors :  2.8611176013946533 2.8613524436950684 2.861787796020508\n",
      "Best error :  2.8594250679016113 \n",
      "\n",
      "Epoch  955\n",
      "Top three least errors :  2.8594250679016113 2.8606314659118652 2.8607256412506104\n",
      "Best error :  2.859187126159668 \n",
      "\n",
      "Epoch  956\n",
      "Top three least errors :  2.859187126159668 2.8594250679016113 2.859675645828247\n",
      "Best error :  2.858262538909912 \n",
      "\n",
      "Epoch  957\n",
      "Top three least errors :  2.858262538909912 2.8583555221557617 2.858381748199463\n",
      "Best error :  2.857501745223999 \n",
      "\n",
      "Epoch  958\n",
      "Top three least errors :  2.857501745223999 2.857503652572632 2.857806921005249\n",
      "Best error :  2.8570878505706787 \n",
      "\n",
      "Epoch  959\n",
      "Top three least errors :  2.8570878505706787 2.8573219776153564 2.8574378490448\n",
      "Best error :  2.8566484451293945 \n",
      "\n",
      "Epoch  960\n",
      "Top three least errors :  2.8566484451293945 2.8568105697631836 2.8570048809051514\n",
      "Best error :  2.8561818599700928 \n",
      "\n",
      "Epoch  961\n",
      "Top three least errors :  2.8561818599700928 2.856452703475952 2.8564956188201904\n",
      "Best error :  2.8556125164031982 \n",
      "\n",
      "Epoch  962\n",
      "Top three least errors :  2.8556125164031982 2.856055498123169 2.8561182022094727\n",
      "Best error :  2.8542282581329346 \n",
      "\n",
      "Epoch  963\n",
      "Top three least errors :  2.8542282581329346 2.8546133041381836 2.85475754737854\n",
      "Best error :  2.854217767715454 \n",
      "\n",
      "Epoch  964\n",
      "Top three least errors :  2.854217767715454 2.8542282581329346 2.8543801307678223\n",
      "Best error :  2.8539271354675293 \n",
      "\n",
      "Epoch  965\n",
      "Top three least errors :  2.8539271354675293 2.85398006439209 2.854217767715454\n",
      "Best error :  2.8530805110931396 \n",
      "\n",
      "Epoch  966\n",
      "Top three least errors :  2.8530805110931396 2.8533458709716797 2.853541851043701\n",
      "Best error :  2.852341651916504 \n",
      "\n",
      "Epoch  967\n",
      "Top three least errors :  2.852341651916504 2.8528409004211426 2.852926254272461\n",
      "Best error :  2.851844072341919 \n",
      "\n",
      "Epoch  968\n",
      "Top three least errors :  2.851844072341919 2.852064371109009 2.852268934249878\n",
      "Best error :  2.8511338233947754 \n",
      "\n",
      "Epoch  969\n",
      "Top three least errors :  2.8511338233947754 2.851539373397827 2.851682662963867\n",
      "Best error :  2.8508894443511963 \n",
      "\n",
      "Epoch  970\n",
      "Top three least errors :  2.8508894443511963 2.8511338233947754 2.851357936859131\n",
      "Best error :  2.850170612335205 \n",
      "\n",
      "Epoch  971\n",
      "Top three least errors :  2.850170612335205 2.850508213043213 2.850782632827759\n",
      "Best error :  2.8488025665283203 \n",
      "\n",
      "Epoch  972\n",
      "Top three least errors :  2.8488025665283203 2.8495097160339355 2.8500702381134033\n",
      "Best error :  2.8484299182891846 \n",
      "\n",
      "Epoch  973\n",
      "Top three least errors :  2.8484299182891846 2.848586082458496 2.848615884780884\n",
      "Best error :  2.8482513427734375 \n",
      "\n",
      "Epoch  974\n",
      "Top three least errors :  2.8482513427734375 2.8483452796936035 2.8483564853668213\n",
      "Best error :  2.8471016883850098 \n",
      "\n",
      "Epoch  975\n",
      "Top three least errors :  2.8471016883850098 2.8474795818328857 2.84753680229187\n",
      "Best error :  2.8468382358551025 \n",
      "\n",
      "Epoch  976\n",
      "Top three least errors :  2.8468382358551025 2.8471016883850098 2.847170829772949\n",
      "Best error :  2.8463363647460938 \n",
      "\n",
      "Epoch  977\n",
      "Top three least errors :  2.8463363647460938 2.846463680267334 2.846778631210327\n",
      "Best error :  2.845372438430786 \n",
      "\n",
      "Epoch  978\n",
      "Top three least errors :  2.845372438430786 2.845400810241699 2.8456385135650635\n",
      "Best error :  2.845200777053833 \n",
      "\n",
      "Epoch  979\n",
      "Top three least errors :  2.845200777053833 2.8452553749084473 2.845264196395874\n",
      "Best error :  2.8443801403045654 \n",
      "\n",
      "Epoch  980\n",
      "Top three least errors :  2.8443801403045654 2.8447206020355225 2.8449056148529053\n",
      "Best error :  2.844193458557129 \n",
      "\n",
      "Epoch  981\n",
      "Top three least errors :  2.844193458557129 2.844193935394287 2.8443801403045654\n",
      "Best error :  2.8440825939178467 \n",
      "\n",
      "Epoch  982\n",
      "Top three least errors :  2.8440825939178467 2.84415864944458 2.8441836833953857\n",
      "Best error :  2.8437952995300293 \n",
      "\n",
      "Epoch  983\n",
      "Top three least errors :  2.8437952995300293 2.8440825939178467 2.84415864944458\n",
      "Best error :  2.8437418937683105 \n",
      "\n",
      "Epoch  984\n",
      "Top three least errors :  2.8437418937683105 2.8437952995300293 2.8438289165496826\n",
      "Best error :  2.842755079269409 \n",
      "\n",
      "Epoch  985\n",
      "Top three least errors :  2.842755079269409 2.8436474800109863 2.8437418937683105\n",
      "Best error :  2.842755079269409 \n",
      "\n",
      "Epoch  986\n",
      "Top three least errors :  2.842755079269409 2.8428590297698975 2.8431737422943115\n",
      "Best error :  2.841581344604492 \n",
      "\n",
      "Epoch  987\n",
      "Top three least errors :  2.841581344604492 2.84177565574646 2.8424534797668457\n",
      "Best error :  2.8400981426239014 \n",
      "\n",
      "Epoch  988\n",
      "Top three least errors :  2.8400981426239014 2.8409159183502197 2.841548442840576\n",
      "Best error :  2.838815450668335 \n",
      "\n",
      "Epoch  989\n",
      "Top three least errors :  2.838815450668335 2.839916706085205 2.8400981426239014\n",
      "Best error :  2.838247537612915 \n",
      "\n",
      "Epoch  990\n",
      "Top three least errors :  2.838247537612915 2.838815450668335 2.839050769805908\n",
      "Best error :  2.8381271362304688 \n",
      "\n",
      "Epoch  991\n",
      "Top three least errors :  2.8381271362304688 2.8382019996643066 2.838247537612915\n",
      "Best error :  2.837582588195801 \n",
      "\n",
      "Epoch  992\n",
      "Top three least errors :  2.837582588195801 2.837857961654663 2.8379480838775635\n",
      "Best error :  2.8372511863708496 \n",
      "\n",
      "Epoch  993\n",
      "Top three least errors :  2.8372511863708496 2.837282180786133 2.8375141620635986\n",
      "Best error :  2.8368773460388184 \n",
      "\n",
      "Epoch  994\n",
      "Top three least errors :  2.8368773460388184 2.8370842933654785 2.8372511863708496\n",
      "Best error :  2.836343288421631 \n",
      "\n",
      "Epoch  995\n",
      "Top three least errors :  2.836343288421631 2.836716413497925 2.8367223739624023\n",
      "Best error :  2.8356878757476807 \n",
      "\n",
      "Epoch  996\n",
      "Top three least errors :  2.8356878757476807 2.835963010787964 2.8360259532928467\n",
      "Best error :  2.834376811981201 \n",
      "\n",
      "Epoch  997\n",
      "Top three least errors :  2.834376811981201 2.835625648498535 2.8356878757476807\n",
      "Best error :  2.8336899280548096 \n",
      "\n",
      "Epoch  998\n",
      "Top three least errors :  2.8336899280548096 2.8339297771453857 2.834376811981201\n",
      "Best error :  2.832646131515503 \n",
      "\n",
      "Epoch  999\n",
      "Top three least errors :  2.832646131515503 2.8332064151763916 2.833235502243042\n",
      "Best error :  2.8323347568511963 \n",
      "\n",
      "Epoch  1000\n",
      "Top three least errors :  2.8323347568511963 2.832646131515503 2.8326902389526367\n",
      "Best error :  2.831439733505249 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "for epoch in range(epochs):\n",
    "    \n",
    "    print(\"Epoch \", epoch + 1)\n",
    "    print(\"Top three least errors : \", x_alpha_score, x_beta_score, x_delta_score)\n",
    "    \n",
    "    # Generate a_alpha, a_beta, a_delta and E\n",
    "    x = epoch/epochs\n",
    "    a_alpha = np.power(sech(4 * x), 2)\n",
    "    a_delta = np.power(sech(3 * x), 2)\n",
    "    a_beta = (a_alpha + a_delta) * 0.5\n",
    "    \n",
    "    # MGWO (levy case = 1,multiplier = 0.1)  RMSE = 3.17, 3.58\n",
    "    \n",
    "    \"\"\"\n",
    "    custom_gwo (levy case = -1, multiplier = 0.01, Z = agent_vector + levy_step)  \n",
    "    RMSE = 3.52, 3.48   R2 = 0.737, 0.747  MAE = 2.679, 2.638\n",
    "    \"\"\"\n",
    "    \n",
    "    \"\"\"\n",
    "    custom_gwo (levy case = -1, multiplier = 0.01, approach 3 eqn)\n",
    "    RMSE = 3.217, 3.36  R2 = 0.788, 0.725  MAE = 2.54, 2.61\n",
    "    \"\"\"\n",
    "    \n",
    "    \"\"\"\n",
    "    custom_gwo (levy case = -1, multiplier = 0.05, approach 3 eqn)\n",
    "    RMSE = 2.83. 2.71, 3.36  R2 = 0.826, 0.86  MAE = 2.22, 2.1658\n",
    "    \"\"\"\n",
    "    \n",
    "    \n",
    "    # update each wolf\n",
    "    for i in range(num_solutions):\n",
    "        \n",
    "        current_population_vectors[i], curr_errors[i] = custom_gwo(current_population_vectors[i], curr_errors[i], epoch,\n",
    "                                                                   best_solution, X_alpha, X_beta, X_delta, a_alpha, a_beta, a_delta) \n",
    "    \n",
    "    # update the alpha, beta and delta wolves\n",
    "    X_alpha, X_beta, X_delta, x_alpha_score, x_beta_score, x_delta_score = GWO.find_alpha_beta_delta(current_population_vectors, curr_errors)\n",
    "    \n",
    "    # set the best solution and best error to alpha wolf and alpha wolf's error respectively\n",
    "    best_solution, best_error = X_alpha, x_alpha_score\n",
    "    print('Best error : ', best_error, '\\n')\n",
    "    iteration_errors.append(best_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "74579590",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.set_weights(set_shape(best_solution, shape))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c80edfc4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "38/38 [==============================] - 0s 747us/step - loss: 8.0171 - root_mean_squared_error: 2.8314\n",
      "10/10 [==============================] - 0s 887us/step - loss: 7.3722 - root_mean_squared_error: 2.7152\n",
      "Train RMSE =  2.831439733505249\n",
      "Test RMSE =  2.71518611907959\n",
      "---------------------------\n",
      "Train MSE =  8.017050743103027\n",
      "Test MSE =  7.3722357749938965\n",
      "---------------------------\n",
      "38/38 [==============================] - 0s 652us/step\n",
      "10/10 [==============================] - 0s 777us/step\n",
      "Train R2 =  0.8264881730079064\n",
      "Test R2 =  0.8603976645277887\n",
      "-----------------------------\n",
      "Train MAE =  2.2217829678705248\n",
      "Test MAE =  2.1658707056394046\n"
     ]
    }
   ],
   "source": [
    "# RMSE\n",
    "score = model.evaluate(X_train, y_train)\n",
    "train_rmse, train_mse = score[1], score[0]\n",
    "\n",
    "score = model.evaluate(X_test, y_test)\n",
    "test_rmse, test_mse = score[1], score[0]\n",
    "\n",
    "print(\"Train RMSE = \", train_rmse)\n",
    "print(\"Test RMSE = \", test_rmse)\n",
    "print(\"---------------------------\")\n",
    "\n",
    "#MSE\n",
    "print(\"Train MSE = \", train_mse)\n",
    "print(\"Test MSE = \", test_mse)\n",
    "print(\"---------------------------\")\n",
    "\n",
    "#R2 Score\n",
    "y_train_pred = model.predict(X_train)\n",
    "y_test_pred = model.predict(X_test)\n",
    "train_r2_score = r2_score(y_train, y_train_pred)\n",
    "test_r2_score = r2_score(y_test, y_test_pred)\n",
    "\n",
    "print(\"Train R2 = \", train_r2_score)\n",
    "print(\"Test R2 = \", test_r2_score)\n",
    "print(\"-----------------------------\")\n",
    "\n",
    "# MAE\n",
    "train_mae = mean_absolute_error(y_train, y_train_pred)\n",
    "test_mae = mean_absolute_error(y_test, y_test_pred)\n",
    "print(\"Train MAE = \", train_mae)\n",
    "print(\"Test MAE = \", test_mae)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4f9bc36f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# graph of iteration errors\n",
    "\n",
    "x_values = range(len(iteration_errors))\n",
    "plt.plot(x_values, iteration_errors)\n",
    "plt.xlabel('Iteration')\n",
    "plt.ylabel('Best wolf RMSE error')\n",
    "plt.title('Error plot')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "81cc6527",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scatter and line plot for train set\n",
    "\n",
    "x_values = np.array([i for i in range(X_train.shape[0])])\n",
    "plt.scatter(x_values, y_train, color = 'black', label = 'Actual values')\n",
    "plt.plot(x_values, y_train_pred, color = 'red', label = 'Predicted values')\n",
    "plt.title('Actual vs Predicted values for train set')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "ba6e5025",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scatter and line plot for test set\n",
    "\n",
    "x_values = np.array([i for i in range(X_test.shape[0])])\n",
    "plt.scatter(x_values, y_test, color = 'black', label = 'Actual values')\n",
    "plt.plot(x_values, y_test_pred, color = 'red', label = 'Predicted values')\n",
    "plt.title('Actual vs Predicted values for test set')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad3ba915",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b211be4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}