Proof of Concept/Public data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "4277037c",
   "metadata": {},
   "source": [
    "<h1 align=\"center\">\n",
    "        <picture>\n",
    "         <source media=\"(prefers-color-scheme: dark)\" srcset=\"https://user-images.githubusercontent.com/16140783/218357099-29d4848f-89ee-463e-9ead-40f27c976f61.png\">\n",
    "         <img width=\"300\" src=\"https://user-images.githubusercontent.com/16140783/218437368-1aa8506c-1ed8-460a-99de-d2c081557170.png\" align=\"center\"></img>\n",
    "    </picture><br/><br/>Subvocalization EMG - PoC<sub><sup><br/>(Public data)</sup></sub>\n",
    "</h1>\n",
    "\n",
    "This PoC is using the same technique in the Complete processing PoC, but using public EMG data instead of Synthetic data. The dataset (and its research) used in this ipynb can be found at https://zenodo.org/record/7668251.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d29bb079",
   "metadata": {},
   "source": [
    "\n",
    "### Packages\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e00ad932",
   "metadata": {},
   "outputs": [],
   "source": [
    "from brainflow.data_filter import DataFilter, FilterTypes, DetrendOperations\n",
    "from sklearn.model_selection import train_test_split\n",
    "from keras.layers import Conv1D, MaxPooling1D, GlobalAveragePooling1D, Dense, Dropout\n",
    "from keras.models import Sequential\n",
    "from keras.utils import to_categorical\n",
    "from more_itertools import sliced\n",
    "from tensorflow import keras\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os, os.path\n",
    "import np_utils\n",
    "import random"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a21f8dbd",
   "metadata": {},
   "source": [
    "### Reading raw EMG data\n",
    "\n",
    "The sEMG data of the subjects downloaded from Zenodo should be extracted to a folder named `dataset`. We will then parse it and change its structure to match the CSV generated by the subvocalization software."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5cc9f43c",
   "metadata": {},
   "outputs": [],
   "source": [
    "path = \"./dataset\"\n",
    "data = [name for name in os.walk(path)]\n",
    "gestures = [\"Open hand\", \"Closed hand\", \"Lateral pinch\", \"Signaling sign\", \"Rock sign\"]\n",
    "dataframes = [[],[],[],[],[]]\n",
    "\n",
    "emg_channels = [1, 2, 3, 4]\n",
    "df_channels = list(map(lambda x: \"Channel_%s\" % (x), emg_channels))\n",
    "\n",
    "for i in range(1,len(data)):\n",
    "    for j in range(0,len(data[i][2])):\n",
    "        file_name = data[i][2][j]\n",
    "        gesture = int(os.path.splitext(file_name)[0])\n",
    "        dataframe = pd.read_csv(data[i][0] + '/' + data[i][2][j], sep='\\t', header=None, names=df_channels)\n",
    "        dataframes[gesture].append(dataframe)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7106c05d",
   "metadata": {},
   "source": [
    "This completes the importing stage, now all subjects' datasets are stored in their respectively gesture slot. We can render the channels from a single subject just like in the Complete processing PoC using data from the Cyton Board:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "286e0382",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 3, 4]\n",
      "['Channel_1', 'Channel_2', 'Channel_3', 'Channel_4']\n",
      "\n",
      "Plot channels:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "random_subject_open_hand = random.choice(dataframes[0])\n",
    "\n",
    "print(emg_channels)\n",
    "print(df_channels)\n",
    "print(\"\\nPlot channels:\")\n",
    "\n",
    "axs = random_subject_open_hand[df_channels].plot(subplots=True)\n",
    "for i,ax in enumerate(axs):\n",
    "    axs[i].legend(loc=\"upper right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "529fc1b7-8be1-4eb3-a5ad-eb00a426e808",
   "metadata": {},
   "source": [
    "However, this dataset has the whole recording in a single stream, therefore we need to break it into intervals. The authors state this dataset has a window length of 200:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "12396d15-f50a-4d78-bd05-fc4e6e072006",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Plot channels:\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "window_length = 200\n",
    "\n",
    "chunk = next(sliced(random_subject_open_hand, window_length))\n",
    "\n",
    "print(\"\\nPlot channels:\")\n",
    "axs = chunk[df_channels].plot(subplots=True)\n",
    "for i,ax in enumerate(axs):\n",
    "    axs[i].legend(loc=\"upper right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "daee6969-ba48-4048-b35e-72af7ed4926b",
   "metadata": {},
   "source": [
    "Lastly, we need to parse this data to make it as similar as our dataset, adding the word (gesture) in the first column. In this step we perform the segmentation and merge it all into a single dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8c1bd646-bbac-44da-b1d2-e66df7f9d574",
   "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>WORD</th>\n",
       "      <th>Channel_1</th>\n",
       "      <th>Channel_2</th>\n",
       "      <th>Channel_3</th>\n",
       "      <th>Channel_4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Open hand</td>\n",
       "      <td>0.004898</td>\n",
       "      <td>0.004003</td>\n",
       "      <td>-0.004031</td>\n",
       "      <td>-0.001103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Open hand</td>\n",
       "      <td>0.003446</td>\n",
       "      <td>-0.004977</td>\n",
       "      <td>0.001012</td>\n",
       "      <td>-0.002247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Open hand</td>\n",
       "      <td>-0.003934</td>\n",
       "      <td>-0.009833</td>\n",
       "      <td>0.007323</td>\n",
       "      <td>-0.003097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Open hand</td>\n",
       "      <td>-0.002250</td>\n",
       "      <td>-0.007743</td>\n",
       "      <td>-0.004610</td>\n",
       "      <td>-0.000540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Open hand</td>\n",
       "      <td>-0.003268</td>\n",
       "      <td>-0.006243</td>\n",
       "      <td>0.002576</td>\n",
       "      <td>-0.003321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759555</th>\n",
       "      <td>Rock sign</td>\n",
       "      <td>0.034819</td>\n",
       "      <td>-0.008418</td>\n",
       "      <td>-0.006649</td>\n",
       "      <td>0.027904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759556</th>\n",
       "      <td>Rock sign</td>\n",
       "      <td>0.043744</td>\n",
       "      <td>-0.007028</td>\n",
       "      <td>-0.001524</td>\n",
       "      <td>0.050832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759557</th>\n",
       "      <td>Rock sign</td>\n",
       "      <td>0.055965</td>\n",
       "      <td>-0.002811</td>\n",
       "      <td>0.007052</td>\n",
       "      <td>0.015889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759558</th>\n",
       "      <td>Rock sign</td>\n",
       "      <td>0.054340</td>\n",
       "      <td>-0.014413</td>\n",
       "      <td>-0.030651</td>\n",
       "      <td>-0.001178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759559</th>\n",
       "      <td>Rock sign</td>\n",
       "      <td>0.049726</td>\n",
       "      <td>-0.020201</td>\n",
       "      <td>-0.052454</td>\n",
       "      <td>-0.040208</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>759560 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             WORD  Channel_1  Channel_2  Channel_3  Channel_4\n",
       "0       Open hand   0.004898   0.004003  -0.004031  -0.001103\n",
       "1       Open hand   0.003446  -0.004977   0.001012  -0.002247\n",
       "2       Open hand  -0.003934  -0.009833   0.007323  -0.003097\n",
       "3       Open hand  -0.002250  -0.007743  -0.004610  -0.000540\n",
       "4       Open hand  -0.003268  -0.006243   0.002576  -0.003321\n",
       "...           ...        ...        ...        ...        ...\n",
       "759555  Rock sign   0.034819  -0.008418  -0.006649   0.027904\n",
       "759556  Rock sign   0.043744  -0.007028  -0.001524   0.050832\n",
       "759557  Rock sign   0.055965  -0.002811   0.007052   0.015889\n",
       "759558  Rock sign   0.054340  -0.014413  -0.030651  -0.001178\n",
       "759559  Rock sign   0.049726  -0.020201  -0.052454  -0.040208\n",
       "\n",
       "[759560 rows x 5 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subjects_count = len(data) - 1\n",
    "restored_data = pd.DataFrame(columns=['WORD'] + df_channels)\n",
    "min_attr = float('inf')\n",
    "for gesture_frames in dataframes:\n",
    "    for df in gesture_frames:\n",
    "        min_attr = min(min_attr, len(df))\n",
    "rows = range(min_attr)\n",
    "\n",
    "# Iterate subjects\n",
    "for k in range(subjects_count):\n",
    "    # Iterate window (segmentation)\n",
    "    for index_slice in sliced(rows, window_length):\n",
    "        # Iterate gestures\n",
    "        for g in range(len(gestures)):\n",
    "            df = dataframes[g][k]\n",
    "            df['WORD'] = gestures[g]\n",
    "            chunk = df.iloc[index_slice]\n",
    "            restored_data = pd.concat([restored_data, chunk], ignore_index = True)\n",
    "\n",
    "restored_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7efe82d8",
   "metadata": {},
   "source": [
    "### Preprocessing\n",
    "\n",
    "With our 4 channels data in hand, we can perform a Butterworth filter to select the frequency range we want to keep, here we want a lowcut of 3.0 and a highcut of 45.0:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "e6688254",
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000914</td>\n",
       "      <td>0.000747</td>\n",
       "      <td>-0.000752</td>\n",
       "      <td>-0.000206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.002628</td>\n",
       "      <td>0.000694</td>\n",
       "      <td>-0.001444</td>\n",
       "      <td>-0.000868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.001910</td>\n",
       "      <td>-0.002834</td>\n",
       "      <td>0.000752</td>\n",
       "      <td>-0.001772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.001237</td>\n",
       "      <td>-0.006782</td>\n",
       "      <td>0.002450</td>\n",
       "      <td>-0.001909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.003539</td>\n",
       "      <td>-0.007481</td>\n",
       "      <td>0.001235</td>\n",
       "      <td>-0.001372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759555</th>\n",
       "      <td>0.040822</td>\n",
       "      <td>-0.001697</td>\n",
       "      <td>-0.022724</td>\n",
       "      <td>0.047444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759556</th>\n",
       "      <td>0.040447</td>\n",
       "      <td>-0.015736</td>\n",
       "      <td>-0.018161</td>\n",
       "      <td>0.021748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759557</th>\n",
       "      <td>0.043213</td>\n",
       "      <td>-0.017485</td>\n",
       "      <td>-0.013564</td>\n",
       "      <td>0.015942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759558</th>\n",
       "      <td>0.047369</td>\n",
       "      <td>-0.009081</td>\n",
       "      <td>-0.020551</td>\n",
       "      <td>0.009154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759559</th>\n",
       "      <td>0.043137</td>\n",
       "      <td>-0.006726</td>\n",
       "      <td>-0.034349</td>\n",
       "      <td>-0.008207</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>759560 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "               0         1         2         3\n",
       "0       0.000914  0.000747 -0.000752 -0.000206\n",
       "1       0.002628  0.000694 -0.001444 -0.000868\n",
       "2       0.001910 -0.002834  0.000752 -0.001772\n",
       "3      -0.001237 -0.006782  0.002450 -0.001909\n",
       "4      -0.003539 -0.007481  0.001235 -0.001372\n",
       "...          ...       ...       ...       ...\n",
       "759555  0.040822 -0.001697 -0.022724  0.047444\n",
       "759556  0.040447 -0.015736 -0.018161  0.021748\n",
       "759557  0.043213 -0.017485 -0.013564  0.015942\n",
       "759558  0.047369 -0.009081 -0.020551  0.009154\n",
       "759559  0.043137 -0.006726 -0.034349 -0.008207\n",
       "\n",
       "[759560 rows x 4 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sampling_rate = window_length\n",
    "exg_channels = [1, 2, 3, 4]\n",
    "\n",
    "curves = [[], [], [], []]\n",
    "for count, channel in enumerate(df_channels):\n",
    "    if count >= 8:\n",
    "        continue\n",
    "    channel_data = restored_data[channel].to_numpy()\n",
    "    DataFilter.detrend(channel_data, DetrendOperations.CONSTANT.value)\n",
    "    DataFilter.perform_bandpass(channel_data, sampling_rate, 3.0, 45.0, 2,\n",
    "                                FilterTypes.BUTTERWORTH.value, 0)\n",
    "    DataFilter.perform_bandstop(channel_data, sampling_rate, 48.0, 52.0, 2,\n",
    "                                FilterTypes.BUTTERWORTH.value, 0)\n",
    "    DataFilter.perform_bandstop(channel_data, sampling_rate, 58.0, 62.0, 2,\n",
    "                                FilterTypes.BUTTERWORTH.value, 0)\n",
    "    curves[count] = channel_data\n",
    "\n",
    "preprocessed_data = pd.DataFrame(np.transpose(curves))\n",
    "\n",
    "preprocessed_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "caaeaeff",
   "metadata": {},
   "source": [
    "We can also plot these curves to see the difference between the unprocessed data, note the scales also were reduced since we removed unused information:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "b54317f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGdCAYAAAD60sxaAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOydd3xT9frHPydJ03TvTaFlljLKLgVlaKHggutCHHjRHy7wqlzRi3pxXnGLG/dGEQcqIluGUFahzLYU6N4zndnn98c335OkzWyTJi3f9+uVV9KTk5PvSZPveb7P83meh+N5ngeDwWAwGAzGJYTI3QNgMBgMBoPB6GmYAcRgMBgMBuOSgxlADAaDwWAwLjmYAcRgMBgMBuOSgxlADAaDwWAwLjmYAcRgMBgMBuOSgxlADAaDwWAwLjmYAcRgMBgMBuOSQ+LuAXgqOp0O5eXlCAgIAMdx7h4Og8FgMBgMO+B5Hs3NzYiNjYVIZNnPwwwgC5SXlyM+Pt7dw2AwGAwGg9EFSkpK0K9fP4vPMwPIAgEBAQDIBxgYGOjm0TAYDAaDwbCHpqYmxMfHC9dxSzADyAI07BUYGMgMIMYlAc/zeHvnefh5i3H3ZYks9MtgMHo1tuYwZgAxGAwAQF5VM97ccQ4AIPMS4/bJA9w8IgaDwXAdLAuMwWAAAA5eqBMeP/v7GWQVNbhxNAwGg+FamAHEYDAAAIcK6gEAgTIJ1Foe93+TBXm72s2jYjAYDNfAQmAMBgM8zwsG0Du3jsNTG0+hpL4d+/JrcM3oWDePjsFgOALP89BoNNBqte4eiksQi8WQSCTd1ikyA4jBYCC/ugX1rSrIvERIGxiGGUMj8fXBIpwoaWQGEIPRi1CpVKioqEBbW5u7h+JSfH19ERMTA6lU2uVjMAOIwWDg0EWi/xk/IARSiQgp8cF6A0ju5pExGAx70el0KCgogFgsRmxsLKRSaZ/L5uR5HiqVCjU1NSgoKMCQIUOsFju0BjOAGAwGDl4k4a/UxDAAwJj4IADAqTI5NFodJGImF2QwPB2VSgWdTof4+Hj4+vq6ezguw8fHB15eXigqKoJKpYJMJuvScdisxmBc4hD9D/EATR5IDKCB4f7w95agXa1FfnWLO4fHYDAcpKsekd6EM86x739KDAbDKhdqWlHbooK3RIQUvedHJOIwuh95nF3S6MbRMRgMhmtgBhCDcYlzrJjU+xkTHwxviVjYnhIfDAA4wQwgBoPRB2EGEINxiXNBH+IaHmPa8iWlXzAA5gFiMBh9E2YAMRiXOOf1BtCgSH+T7WP0HqBzVc1oU2l6elgMBuMS47333kNCQgJkMhlSU1Nx+PBhl74fM4AYjEucCzV6AyjCz2R7dJAMUYHe0PHA6bImdwyNwWD0UbQ6HXieF/5ev349li9fjqeffhrHjh1DSkoKMjIyUF1d7bIxMAOIwbiEUWq0KK4nBdMGR/h3ep6GwZgOiMHonfA8jzaVpsdvxsaNOcobFcivbkGrkniX33jjDSxZsgSLFy9GcnIy1q5dC19fX3z22Wcu+2xYHSAG4xKmqK4NOh4I8JYgIsC70/MjYoOw7WwVciub3TA6BoPRXdrVWiSv2trj73v2uQz4Ss2bGO0qLRraVAAAjiP1i7KysrBy5UphH5FIhPT0dGRmZrpsjD3iAXI0rrdhwwYkJSVBJpNh1KhR2Lx5s8nzPM9j1apViImJgY+PD9LT05Gfn2+yT0JCAjiOM7m99NJLTj83BqM3Y6z/MVcxdrBeF3S+htUCYjAYzqGqSQEACPLxgq9UgtraWmi1WkRFRZnsFxUVhcrKSpeNw+UeIBrXW7t2LVJTU7FmzRpkZGQgLy8PkZGRnfY/cOAAFi5ciNWrV+Oaa67BunXrMH/+fBw7dgwjR44EALzyyit4++238eWXXyIxMRH//e9/kZGRgbNnz5pUhHzuueewZMkS4e+AgABXny6D0augGWCDzIS/AIMBdKG6BTzP97my+gxGX8fHS4yzz2W45X3N0arUoEmhBgcO0YFdq+DsLFzuAXI0rvfWW29hzpw5WLFiBYYPH47nn38e48aNw7vvvguAeH/WrFmDp556CvPmzcPo0aPx1Vdfoby8HBs3bjQ5VkBAAKKjo4Wbn5+fmXdkMC5dBAF0pPnfRkK4L0Qc0KLUoKpJ2ZND8ygq5Qoc19dLYjB6ExzHwVcq6fGbucUSz/OolBPvT4ifF7z1RlJ4eDjEYjGqqqpM9q+qqkJ0dLTLPhuXGkA0rpeenm54QxtxvczMTJP9ASAjI0PYv6CgAJWVlSb7BAUFITU1tdMxX3rpJYSFhWHs2LF49dVXodFYTuVVKpVoamoyuTEYfR0a2jIngAYAb4kYA8KIcXT+Em2JodPxuPWTg7j+gwM4XcaawzIYXUWp0aFVpQHHcYgKMHh/pFIpxo8fj507dwrbdDoddu7cibS0NJeNx6UGUFfiepWVlVb3p/e2jvmvf/0L33//Pf766y/ce++9ePHFF/HYY49ZHOvq1asRFBQk3OLj4+0/UQajF6LT8bhQ3Qqgcw0gY2h47Hz1pSmE3pNfg4s1reB5YEdOle0XMBgMs9CMLz+pGF4SU/Nj+fLl+Pjjj/Hll18iJycH999/P1pbW7F48WKXjafPZoEtX75ceDx69GhIpVLce++9WL16Nby9O2e7rFy50uQ1TU1NzAhi9GkqmxRoV2shEXHoH2q5c/SQKH/syKm6ZIXQ32QWCY/3nKvBw+lD3TgaBqP30qbSAoDZ7LAFCxagpqYGq1atQmVlJcaMGYMtW7Z0cnY4E5d6gLoS14uOjra6P713NFaYmpoKjUaDwsJCs897e3sjMDDQ5MZg9GWo/mdAmC+8xJangsGCB6j3G0C1LUq8uDkH+VX2ebNK6tuwK89QiO1ESSMa9em7DAbDMVr1FeX9vM0LpJctW4aioiIolUocOnQIqampLh2PSw2grsT10tLSTPYHgO3btwv7JyYmIjo62mSfpqYmHDp0yGqsMDs7GyKRyGzmGYNxKUINmsFWwl/Gz5/Xh8t6KzzP45H12fho70Xc900WVBqdzdd8d7gYPA9cNjgcQ6P8oeOBv8/X9sBoGYy+hVqrE35zPlLzBlBP4/IQ2PLly3HnnXdiwoQJmDRpEtasWWMS11u0aBHi4uKwevVqAMBDDz2E6dOn4/XXX8fVV1+N77//HkePHsVHH30EgCjaH374YbzwwgsYMmSIkAYfGxuL+fPnAyBC6kOHDmHmzJkICAhAZmYmHnnkEdx+++0ICQlx9SkzGL2C8zZS4ClUH1TbooS8TY0gXy+Xj80VrDtcjH35xHi5UNOKT/6+iAdmDLa4v0qjww9HSwAAt0/uj6OFDThX1YK952pwzejYHhkzwz4OF9Tjje15aFdpwQO4+7JEzBsT5+5hMYyg4S+ZlxgSkWc0oXC5AWQrrldcXAyR0YcxZcoUrFu3Dk899RSeeOIJDBkyBBs3bhRqAAHAY489htbWVtxzzz1obGzEZZddhi1btgg1gLy9vfH999/jmWeegVKpRGJiIh555BETjU9vhtVjYTiDsxUk03FYtPX6WP7eEsQEyVAhV+B8TTPGDwjtieE5lZL6NvzvjxwAwOVDwrEvvxZv78zHdSmx6BdiXv+UXdKI2hYVwvykSB8eBT9vCT75uwB7ztWw36CH8fKWXGQVGcoU/OenUxjXPwTxVrRtjJ6lTS+A9vUQ7w/QQyLoZcuWYdmyZWaf2717d6dtN910E2666SaLx+M4Ds899xyee+45s8+PGzcOBw8e7NJYPZ3l67ORVdyAj+6YYPPCxWBYQqPV4Ww5MYBGxQXZ3H9wpD8xgKpbeqUB9Mb2c2hTaTEpIRRfLJ6EhR8fxOGCerywKQdr7xhv9jWHC+oAAJMHhkEiFmFiQihkXiJUNSlxrqqF/f48hJL6NmQVNYDjgHcWjsWXBwpxpLABq349jc/+OdEuQ7WqSYFmhcZmOJjRdVr1HiA/C+0x3IFn+KEYdqHW6vDbiXIU1bXhlo8ycaac1STxFDafqsBXmYXuHobd5Fe3QKnRwd9bgoQw2wVCB/VyITT9rdw/cxDEIg7PzRsBANh2thL1reZFzYcK6gEAqQOJwSfzEmPywDAAwL78Gofef82Oc1j58yko1NoujZ9hmd9OlAMApgwKwzWjY7H6+tGQikX4K68Gf5yqsPraLw8UYs6avUh9cSfS39iDrKL6nhiyy7HViLSn0el4tOu/+74WBNCO4oxzZAZQL6K0oR0aHfmnN7SpcevHh1BU1z1hqlbH42RpI9Ra24JQhnnyq5qxbN0xrPr1TK8plHdKP84RsYEQiWyvkA1C6N5nAPE8L3S8T9Qbe0nRgRgeEwgdD+zKre70GrVWJ4RUJiUaPF5j4oMBAOfszCIDgPpWFdbsyMd3h4tx3zdZUGqYEeQIJ0sb8b8/zqLBgqH6WzYxgOalEM3P4Eh/3D9jEADgud/PWhS71zQr8fRvZ0wa/f55ynV9p3oCLy+iz2tra3PzSExpV2vB8zwkIhGkVjJOHYGeIz3nrsAMoF5EYS0xdhLD/TAyLhDydjU2Hi/v1jE/2XcR1727H9e/fwAXL9E6L91l9Z+50Nul+MvMxdQToYba6H62w1+AwQN0sbb3ZYLVNCuhUOsg4oDYYB9h+6xkokPcdqbzRe9MeRPaVFoE+3phaKQh1JUYTgyoizX2fw7ZJQZtyu68GixbdxwatuCwC41Wh399dxwf7yvA8h+yO636cyqakFfVDKlYhIyRhjIoD8wchIgAb1Q3Ky0WrzxaSLw9gyL88MJ8ojGlInlP5KesUnyy76JVz4dYLEZwcDCqq6tRV1eH9vZ2KBQKt98am1vBa1Tw5jRQKpXdOlZ7ezvq6upQXV2N4OBgiMVd9yh5TjDuEqO4rg1vbM/DkmkDMSLWvosQvfgkRQdgTHwwTpc1Ib+b1Xl/PlYGgHgErn77b7xy42hcm8IyXOzlwIVaEw/CX3nVePDKIW4ckX2cLCUG0Eg79D+A4cJf2tAOtVZntW6Qp0G9PzFBPpAaVZ+dnRyFt3fmY29+DdpVWpPUXKr/mZgQauIh64oheLy4EQDxtuVXt2D72SpsP1uFuaNiunxOlwq/nShHYR35//2VV4PP9xfirssShed/1Xt/ZiZFIMjH4Anwlohx0/h+eH/3BXx/pARXmfmsjxQSw3TKoHBcPSoG//31NPKqmlHdpECkm5t0dqS2RYkVP56AjideyAkJlnV4tB5edbXnLMZqmpVQanRQ+XpB1egcsyM4OLjbfcKYAeQmPthzHhuzy5FV3IAtD02Dn7ftf4WxB2hIVPdDEuerW5BX1QyJiMP4ASE4VFCP5T9kIyZIZvUHxiDodDxe3Ewyi2YnR2Hb2SpklzSioVWFED+pm0dnGY1Wh5wK+wXQABAV6A0fLzHa1VqU1LdhoI3UeU+iSH8BHRBmmhE0IjYQccE+KGtsx9/nawWPEAAcuqjX/ySa/g6oIVjfqkJjmwrBvrb/z9kljQCAhZP640x5E747XIzT5XJmANlAq+Px7q7zAICU+GCcKGnES3/mYlJiKEbGBUGt1WHjcbKAM5fyvmBiPN7ffQH78mtQUt/WKSPsiN4DNCEhBCF+UoyKC8LJUjn25dfihvH9XHx2jrHjbJXgZf7ucInV+ZnjOMTExCAyMhJqtbqHRmiZVqUa975/AFodj6/vSkVsiI/tF9nAy8urW54fSu9ZxvUheJ7Hnjwioiypb8fLW3Ltel2B3gBKCPfDEL1b/mJNa5fd6X/qBYJTB4dj3ZLJuHp0DNRaHvd9cwwl9W04eLEOm09V9IigTqvjkVPRBJ3Os8R7xvA8L/SyAYBfT5ThdFkT/L0lWH39KAyLCoCOB/Y6KJDtaRwVQANkUk3QX/wLu6k762moB6hjuw+O48yGwbQ6Hof1F8dJHQwgP28JovXegQt2hMF0Oh7Zeg/QmPhgDNUvXPKrWLjZFr+fKMfF2lYE+3rhm7snIX14FFRaHZ7aeBo8z2PzqQpUNikQ7i/FFUmdC9wOCPPDlEFh4HlgQ1apyXMtSo0gjKf/48sGhwPwzEKXW42+n3+cKkeTwrZhIxaLIZPJ3H47VtaGYrkGEqk3BsaEOOWYzjB+AGYAuYULNS0olysg1rvWv8osQuaFOpuvowbQwHA/xAX7QOYlgkqrEyZ4R6EZElePioFYxOHVG0cjKToAtS1KXP7KX7jlo4N44Ntj+PZQcZeO7wif7y/A3Lf24V47K/T2NCqNDku+Oopxz2/Hn6cqoFBr8drWcwCA+2cMQpi/N2YMiwAAwbj1VE6VOiaApiSGEwPCEf2LJ0B/H+ZqwszWG0A7c6uh1RvfeZXNaFZo4O8tQXJM55Y4AyOoDsi2EXOxtgXNSg1kXiIkRQcIC5f8Xigmd5TGNpVQasFReJ7He38R78+SywciQOaFF68fCR8vMbJLGrH1TBU+2VcAAFiUlgCZl/kL4i2T+gMANhwtEf6/AJBd3AgdD8QF+yAmiHgkLhtCDKB9+bUelUXVrFBj/3lyfQj3l0Kh1gmhv94AzZi8XP/5ehLMAHIDu/UXyCmDwnBrKvmBvmTDC6RQa1EubwdAPEAiESdk5nRlMr1Y04LcShL+mj2CXAR8pRJ8eMd4BOsr/froJ5VP9l00mTxcwbazRKi4/WwVlq075tKsNJ7ncfBiHepalHbv//hPJ7EjpxpKjQ6P/JCNJ385jbLGdsQEyXC3XpMwYxhZhe45V2PVk/VrdhmmvrQLx4obLO7jSk45KICmUG+RKzxATQo1Mi/U4bvDxaiUK5x6bGoAdQyBAcDExFAE+3qhvlWF7frv4KaT5OIyISEEEjNaJ8EAskMHdEzv/RkdFwyJWCR4gIrqWvt0SrxCrcWNazNx1dv7BE+zI+RVNSO/ugXeEhEWpQ0AAEQGGH5rK38+iVNlcnhLRLh98gCLx5mdHIVgXy9UyBXYc86giTHn4Rs/IAQ+XmLUtihNMsPcze68Gqi0OgwM98P9+srl3x92/aLUWVBh+bQhEW4eSWeYAeQG9pwjBtD0oRG4d9pAAECujfBPcX0beB4IkEkQpteX0NVkV3RAf54mLtW0QWEmOoYBYX7Y9vA0bH14Go48lY5AmQSFdW0WMymcQbtKi+N6Y0Ai4rDtbBX+u/G0y95vzY583PLRQWSs2YtDF2173l7blodfjpdBLOIwKi4ICrUOPx0jLvVHZw8TVp8TEkLg7y1BXatKMDI60q7S4rnfz6KssR2f/l3gvJNygJNljgmgKUIIrNa5KbZfZRYi5dltWPjxQaz8+RQe/+mkU49vKQQGAF5iEW7TL0I+2H0e8nY1vtZ3f79V7z3oyMBwYsQU2OEJo/qfMf2DAQARAd4I8vGCjvdMT5paq8Mdnx5Cxpt70abS2H6BBV7dmifMS09uPI2aZvsWG5RtZ8h8c/mQcATIDOLme6YPRIivFxraSAjohvH9EGpFbyfzEuP6sUTP8/3hEmH7USP9D8VbIhZqPjla56krnKtqxv/+OCtokSyxRR/+yhgZjevHxkEqFuFMeVO3S25sOV2Bma/txuEC19U+KqlvQ0FtK8QiDmmDwlz2Pl2FGUA9jEKtFb5w04dGoF+IL6RiEZQaHcoa2y2+rsBIAE0rmwoeIAdqkgCAUqPF+iNkMrjajBAzMlCGYdEB8PeWCKurT/ZddOg9HCGrqAFqLY/oQBk+XjQBAInZV8gtfx5dZcfZKry1Mx8AUNuiwm2fHMJ3VlZTpQ1teH/3BQDAS9ePwrdLUoXPPTkmEP8YaxBfeolFgo5gt4Uw2LeHilCnr2fyV2412lU96wU4V9WMEyWN4Diy4nUEKgAucGIqPM/z+HDPRfA8EBNEtDV/n691Wsf1NpVGuPgOCDWvd1o8NREyLxFOlMqxbN0xNCs1GBYVgPThUWb3N3iAbC88aAbYWH39II7jMETw3HqOl4Hy1o587MuvRV5VM/ae65oW5nBBPT7bT4z76EAZ6ltVeOKXUw6FlbadJRf92cmmWT6BMi8snWno33a3UUaYJW6ZFA+AhDmrmxRQa3XC/2VSBzHx9KHES7H1jOsWfPJ2NZavz0bGmr34eF8BHt1wwuJno1BrsVufZZoxIhohflJcOZx4mruzKG1sU2Hlz6dQUNuK/23OcVnIj+ohx/UPNjFkPQVmAPUwBy/WQanRISZIhsGR/hCLOME1b82lLgigjUSrQ7oYAvv07wIU17chMsDbZsr7nVMS4CXmcKSwQfDSOJvMi2SinTIoDDOTIjF5YCi0Oh7rnKw9KqhtxSPrswGQjJxrU2Kh0fF48pdTOG/hYvRTVhl4noztpgnxCJR54YvFE3Fn2gC8dcuYThoaqgP6K69zCqpCrcWHe4khKeJIc0DqDXQW56qa8Xd+rUVv4vt6XcWcEdEWe2BZghpA5fJ2p4VvjhU3oKyxHX5SMXb9ewaSogOg1fFCOKq7lNQTIzpQJrHYxDXc3xu3TCTeHuquf2DmIIv6KJoKX1jXZjU03KbSIK+SaGCoBwgAhkTpdUAeJoQ+WliP93efF/7ebeY7bAu1VocVP54AzwMLJsTj88UT4SXmsP1sVSchsiXKGttxuqwJIg7Cxd6Y2ycPwA3j+mFFxjCbjXwBYGhUAMb1D4ZWx+PHY6XYlVuNdjWp8dTx9XNHxoDjyKLM2oK0O7z0Zy5+Pk7mFY4jWYoXLOjJsooa0KrSIirQG6P1HtvL9aGkA3boRi3xxvZzghftREkjDl50vheI53n8pP+fX+6B4S+AGUA9Dl1VTR8aIXhy7BFVGqfAU+hEeqGmxe7sqaomhZBa+p+5STbT76MCZUKK6dcHi+x6D0ehP+TJehfporQEAMB3h4udVjWX6nialRpMGBCCZ68bgbdvGYP04VHQ8cCb2/M7vUan47Ehi3jKbp4QL2zvF+KLZ+eNFD5/Y6brDaATpY2dWiysP1KCmmYl4oJ9cIfes/bnacf1EZZoUWpw4wcHcPunh/CP9/d3Cu8V1bUKbQOMV9H2EuYnRYC3BDyPLgvvO0LFnBkjouEjFWOOvpjdVjPFCbuCEP4yo/8xZsm0gZDoDZ6EMF+r3d5jg0k9IZVGh7IGyxdJKrSNCvQWhLYABB2QI9WkXY1CrcUjP2RDxwPD9N/r3Xk1DnsGMi/UoaiuDaF+Ujx1zXAMjwnEI7OGAgCe+e2MXcLx7fr//YQBoQjz9+70vMxLjNdvTnHoO0zF0F/sL8TD32cDAK4ZHdPJyI0OkgleoT9OOl9orNbqhN/8B7eNE3Qx28+aNzZPlDYCACYY1aOaop8njxc3dMmDnFPRhG/0c/lYvWG+ds8Fh49ji1+zy3GsuBE+XmLcNMGzygpQmAHUw1Ah3rShBouY1lSxpgm4aMYAig8hE7FCbT18RqmQt+O/G0+jTaXF2P7BmG+mdoY56MV/Z06108XJLUqNUJSP/rBnJUchOlCG2haV00rT/3GqAocL6iHzEuGthWMhlYjAcRxWZAwDx5HnO8bUD16sQ2lDOwJkEuHCbIuYIB8kRQeA5011BGqtDh/qJ5n7ZwzCdfrPfmdOtdO8Kb9ml6FJQXQbJ0rlWPDRQZPQ5do9F6DjiZfKUf0PYJoK74wwmFqrwx8nycXgujHE4Jg7koRk9+bXosWo5ADP86hqclwcTVvFWAp/UeKCfQQt0COzhgoZmuYQizihpcYFK2GwHTnkt37ZYNPVrydmgmVerENJfTsiArzx7ZJUyLxEqGxSOCwGpoZrxogoIeRx77RBSBsYhjaVFg9+d9zmooYmRNDkDGdwzegY+HtLUN2sRLtai+lDI/Dfa5LN76v3iv9+wnmLE8qhi/VobFMjzE+KWclRSNd7uHZaCGfRjM3RRr/XAWG+iAv2gVrL42gXepfRyvVXjYrGWwvGQsQRXaoze0u2KDVCjbRlVww2WQB4Ej1iAL333ntISEiATCZDamoqDh8+bHX/DRs2ICkpCTKZDKNGjcLmzZtNnud5HqtWrUJMTAx8fHyQnp6O/HzTFXx9fT1uu+02BAYGIjg4GHfffTdaWtw74fA8j/9ek4zFUxMwdZAhJXBguG1NgTkPkEQsEl5rTU9Q16LErR8fRNrqXcLk8sy1I+xOgR4/IAShflLI29U44mTB3JHCemh1POJDfYSQjLEw1RkNRttVWrz4B/kx3j99MOKM2iEMiw7AdfoJ7/VteSav++Eo8f5cmxJrMc3WHDQbzLgtxtYzlSiXk5olN47vh7HxwYgOlKFFqcHfTii/z/OGkOH9MwbhZv2K64U/cvDR3gt4fVseNhwl7uhlXfD+UAxC6O4bQPvP16KuVYUwPymm6rVTQ6P8kRjuB5VGZ/L5rdmRj9QXd+LjvY5p0UqspMB3ZNW1I7DvsZlmi+p1xOC1Nf858DyP7TnEGDAusAjAIzPBTpaQi99lg8MR7u+NKYOsa9nMoTMKXWaMMCwYxCIOby4YgxBfL5wpb8JTv5y2WLussU0lNKHt+Ll1B1+pBNePI//XSYmhWHv7eHhLzP+m546MhljE4VSZ3Cnfc2M2670/s0dEQyIW4Uq9ziyruMFsVipNphhllLHJcQZBsaNhMHmbGn/rF2aPZSShf5gvrtZ7Oz/d17WkDJ2Ox/7ztZC3G2oTvbXjHKqblegf6muXTstduNwAWr9+PZYvX46nn34ax44dQ0pKCjIyMiyW6T5w4AAWLlyIu+++G8ePH8f8+fMxf/58nD5tyAp65ZVX8Pbbb2Pt2rU4dOgQ/Pz8kJGRAYXCsEK87bbbcObMGWzfvh2bNm3C3r17cc8997j6dK3CcRxmDIvE09eOMNEj2PIAtSg1qNYLORPCTVey9ugJ3t6ZjwMX6sBxwIQBIXjv1nFI0Ysy7UEs4oSVyjY7tBllje0WGxd2hNY/ShtomiFwy6T+kIg4HCtuRHGdY+GWhlYVciqacPBiHTYcLcHyH7JRLlcgLtgH904f2Gn/R9LJiv+vvBp8e4i4hivk7UKmnHH4yx5m6sNge420OJ/pM75uSx0AmZcYIhGHuaPIRYK+T3c4WSrHmfImSMUiLLl8IF6+YbRg6Ly4ORfv7DoPjY7HvDGx3arynagPJTnDA0SbWF49OkZorcFxnHDxpNkvhbWt+EAvRH9lay5yK+2vLWMtBb4jYhFnl6EE2A5b51Y2o6S+Hd4SEaYNNa1/4omZYKfKGgEYKoPPtKJls8TxkkZUNysR4C0RDChKdJAMr92UAoAkONz5+WGzQvdNJyug1fFIig7AADuLdNrL43OS8P5t4/Dl4kkmbU86QgzAMP14nBcG0+p4bNX/1ufqPcqxwT5IjgkEz5NWH8bUt6pQqg+xdvTYTumiAbT/Qi10POl/Rq8lNBy/55zjIU8AeH/3edz2ySFc+foe/H6iHMt/yMbHemPqv9ckO7R47GlcbgC98cYbWLJkCRYvXozk5GSsXbsWvr6++Oyzz8zu/9Zbb2HOnDlYsWIFhg8fjueffx7jxo3Du+++C4CsrNasWYOnnnoK8+bNw+jRo/HVV1+hvLwcGzduBADk5ORgy5Yt+OSTT5CamorLLrsM77zzDr7//nuUl3teAalB+sm0Qq4wm3pKVyGhflKTfjeAQQh9zoIBVClX4Dt9+udXd03Cj/dPwdWjHS/BTy9K285UWv2R1LYoMeuNPZj+6l92pZLSFNCOKZIRAd5C5+2DdqSqUw6cr8X4F7Zj7lv7cMtHB7Hix5OCgfHEVcPN/hgTwv1wn94wevKX03jil1O46q19UGp0SIoOQIqD9XLGDQhBgLcE9a0qnCyTI7ukEceKGyEVi3DbZENqNQ2DntTH+bsD9f7MHRWNUD8pOI7Dv2cPFQy+AWG+WHv7eKxZMKZb75MY4ZwQGM/z2KW/uHYU4tOLw9bTlfj9RDle+CMHKq0OYhEHtZbHoxtO2B2KLbKSAt8dqHg2q6jB7O+BekIuHxIOX6mpzs7TMsF4nscJfaglJZ5816kXM6uowa6qw4Ah/DUzKdKk5xrlyuFRWHv7ePhKxdh/vg43fHCgkweMel1vdEErCj9vCa4aFWPV+KFcq/eKbDrpvDDY4YJ61LWqEOTjZTLfpes9XTs6LC6p92dguB8CO2RQ0defKm20+/8DGIq0Th9qEJenxAdBKhGhrlXlcLPjJoUaH+m9srUtSjz43XH8fKwMIg54bM4wYeHsqbjUAFKpVMjKykJ6errhDUUipKenIzMz0+xrMjMzTfYHgIyMDGH/goICVFZWmuwTFBSE1NRUYZ/MzEwEBwdjwoQJwj7p6ekQiUQ4dOiQ2fdVKpVoamoyufUUwb5SoZaFuRWhtVVsUjTxAFmK376/+zxUWh0mJYYKKdpdYergcPhKxSiXK3C6zPJnszuvBm0qLZoUGvzz8yM2Q1i0T9OwqM4VdyfrvUKZDhhA6w4XQ8cDAd4SDAwnpfDvmDwAH90x3qrh9+jsYXhgxiByjEPFaGhTIzkmEB/cPl4Qq9uLl1gkVJV9bWseXtEXubwmJQaRAYYmi1RsWlDb2q3q180KtSBuNq5dw3EcVs4djj0rZmD7I9MxZ2S0w+fSEWcVQ6xsUqCxTQ2xiOtUkHF0vyBcPy4OGh2Pf31/HDtyqiARcfj67kkI8vHC6bImrN1tW7Sp0/Eo1WeBOdsAmj40An5SMXIrm7HZjE6NGkAd07gp1HNrjxCaZtPsdXLGIKWqSYmaZiXEIg7JMeR/ER/qi4ERftDqeOy3I0TL87xgAFnTy80ZGY2fH5iCiABvXKhpxRcHCoXnciqacLJUDomIMykv4Q5mj4iCWMQht7LZYQ+0Jaj4eXZylEkzYWok7M2vMTEIT+kXRub0ejFBPhgY7gcdDxy2M4OL53khLZ0mawCk/hFdbDoqcfhyfyGaFBoMifTH0pmDIBZxiAr0xndLJuOBGYO7Pd+4GpcaQLW1tdBqtYiKMo3lRkVFobLSvNu/srLS6v703tY+kZGmlqdEIkFoaKjF9129ejWCgoKEW3y8Y2GP7mLQAXW+sAjNHM1M4jS9Nq+q2UQ0CpAwDi3+9XD6kG59GWVeYqFGBq3RYQ46SUcHyqDV8Vj16xmh8WZHWpQaIVMqPrSzSI6ucg5erLPLNduu0gqd2b/5v1TsenQG1i2ZjOfnj8TsEdZFzBzH4bE5SVg5NwlBPl64d9pA/LJ0ionmyhHoReDv87WCm/quqaax8JggGfy9JdDoeEGs2xX+zq9Fu1qLxHC/Tr2rAFLc0tyKvCvQz6OqSdnp++YIeXpxbWK4XyctBsdxePXGFCycFA/6b1+UloApg8Lx7HUjAAAf7Llgs7heZZMCKq0OEhEn1BhyFmH+3vi/y4l37dWtuVBrdWhTaXBa7/E7VSYHxwFXWFgBJ8cQA+hEiW3h6Z5zNfj3hhP4vy+PoroLQnBb0EyjIZH+Jt4RGpbOtsNDmVfVjKK6NkglImGesERSdCD+MycJACnLQENhVKOWPjzKbPZXTxLsKxWywazNd/ZS1aTAz8dI49aOnelHxgYhLtgHbSqtSW0fmhxiqWI7nR932RmmzK9uQYVcAW+JqFOT34n6gpBHCu0vddKsUOMTfWj/wSuHYEVGEjJXXoHdj85EagdJg6fCssD0rFy5EnK5XLiVlJTYfpETsaYpKK4nF8f+ZmLikQEy9AvxAc8DJ/VVZymf/V0AlVaH1MTQTjH5rkCzMizVaNHqeCHs9fbCsYLHyVKfMypQDfb1Mlska1z/EHiJOVTIFYIRaI3dedVoU2nRL8TH4TYPlHunD0L2qllYedVwiyJJe7guJRbf3J2Kf05JwNAof9yW2r/TSo7jDO1MLIUw7YEaWMalFVxFsK8UUYHk4pTngBanI9QAGmamlABA9Dgv/mMUHp+ThPljYvHwrCEAgHljYpESH4w2lVboFWUJ6jmNC/Ex29KiuyyZNhBhflIU1rXh/m+OIW31Llzzzt+Y/95+AERvF27hQk4vEEeL6q1mRel0PF7dSsT5Kq3OJdXDT1m40I6IJX/b089rn768x9RBYTZLawDA/LFxSIoOQJNCg/d3X4BSo8Uvx4kBtGBizy4+LUHnO3t0j7Z49vczaFFqMCY+2CQDGABEIg7zx5KQ2y96IwmAkJU6ykLGJjWkfjxaitIG2/MjXZymDgzrJAWYqDf2bFWlNubrg0WQt6sxKMJPKKgbGSCzK8ToKbjUAAoPD4dYLEZVlekXqKqqCtHR5lfk0dHRVven97b26Siy1mg0qK+vt/i+3t7eCAwMNLn1JNaE0NY8QAAwtj+x3o17S+l0vBAWcZYKn8aNcyubzQoYT5fJ0dCmRoC3BGP7BwsrFEs9r4QMHQsF+XykYoyNJ+dmTxiMNne9alRMtwwBZxgRHMfhsiHheOa6Edj2yHT87x+jzO7njJowBy6Qi09PlZofrm8Q2tVGl4CRARRt3gACyGd4/4xBWHPLWEEDwXEcHssYBoBU1S6xUo/IWgsMZ+DvLcGyK4jQfEdOFeTtagTIJJCKRRCLOKs9qoZE+guNLa15gf48XYkz5U1CWv43B4sgb7Nf82EP1AM0ul+wyfbkWMP/2ZYHll44J9u58heLODyu9wJ99ncBLn/5LzS0qREV6O0xTTNpFtrRwvpONb0cYWdOFTafqhSMenMlFmjIb8+5GtS1kJBkuVwBjgNGWDCApgwKw5RBYVBpdXhj+zmb4zBuwdSR8QNCIOLIb8bechO0WvY90wZaLRvhybjUAJJKpRg/fjx27twpbNPpdNi5cyfS0tLMviYtLc1kfwDYvn27sH9iYiKio6NN9mlqasKhQ4eEfdLS0tDY2IisrCxhn127dkGn0yE1NdVp5+dMrKXCCwaQhUwWWmaflncHyIRU1aREgExiEu/tDqF+UsFTlVXU2aihP7Cpg8PhJRYJRbaOmdkXAEr0GQ7mwl+UyUZhMGsYh7/MtffwVIbSLL4uimGrmhS4UNMKjgMmJ/awAVTRdaMt1w4DyBJTB4dj6uAwqLU83txheeKn2g1XGUAAcGtqf1w+JBwp8cF4Z+FYZK+ajbwX5iDnuTlW0+k5jhO8QJY8pBqtDq9vJ96fpTMHIyk6AK0qrVNKQ1B4nrfYHHdYVABEHFDXqhKyUC0d46j+N+5IhuGMYRG4bHA4NDpeOP6itASXeOu6Qr8QXyTHBELHW67TYw2tjsev2WX4z8+nAAD/d1miYFR2ZHBkAEbFBUGj47HppKEm2aAIf/hb8KhxnMGI/OV4mUWpAUDCVYeEFkydDcwAmReSosnY7OkNptPxQgum8QO6nlXqblz+TVu+fDk+/vhjfPnll8jJycH999+P1tZWLF68GACwaNEirFy5Utj/oYcewpYtW/D6668jNzcXzzzzDI4ePYply5YBIP/0hx9+GC+88AJ+++03nDp1CosWLUJsbCzmz58PABg+fDjmzJmDJUuW4PDhw9i/fz+WLVuGW265BbGx1ls/uAvqASqoaTVZbak0OqEnlqVqttTQOF7SKLz295OGCrvdCeV0ZIK+f9RRM0YNdbFSF29Kv2CIRRzK5Qqzfb1seYAAYLK+OWHmBes6IGeEv9xBd0Ng9OI5MjbIYqsHZ5OsN4CsTbjW0Gh1OK8P9SZ1wQACgBUZZOL/+VgZ1uw4Z/a74WoPEEAEpF/fnYpfl07FtSmxEIs4cBxnl+YqTRD5mxcZbzpZgYs1rQjx9cKSyxNxv16k//mBQrsqALcoNfgxqxStVrRaJfXtaGxTQyoWdTJGfaRiIdvNmrfvQk0r6ltV8JaILIZrzMFxHD5aNB4/3JuGX5dOxe5HZwiJCJ6CrbC/JY4W1mPWm3vw0PfZqGlWYmC4Hx5KH2L1NdQL9OWBQjy36SwAy+EvSkp8MK4eFQOeBx7/6aTFvpDv/nUeKo0OgyP9LbYPofrBo3aEwUob2tGm0kIqESHBjhITnorLDaAFCxbgtddew6pVqzBmzBhkZ2djy5Ytgoi5uLgYFRWGVMMpU6Zg3bp1+Oijj5CSkoIff/wRGzduxMiRI4V9HnvsMTz44IO45557MHHiRLS0tGDLli2QyQxCx2+//RZJSUm48sorcdVVV+Gyyy7DRx995OrT7TL9Q30hEXFoVWlRLje4IEsb2qDjAV+pGBEW9ATJsYGQikWob1WhuL4NGq1OqKBsq9eXo9AVXscfibxdjeN6DRKte+LnLREucMeKGjsdi8at+1m5QI3rHwKpRITqZqXV1OtNTgp/9TTUA1TYxUwwGv6a0oOdlqkHKLeyyWovLEsU1rVBpdHBx0ts1fi1xpj4YPxb32JhzY58rPr1TCcjyJEaQO7AECJuNFsQkTbpXTw1EQEyL1w9KgYxQaS5qK0KwGqtDnd/cQSPbjiB5T9kW9zvpL7+T1JMgNmFEvVYWKsSTMNfY+KDHRbb+0olmJQYipT4YCQYNXr2FGgYbG9+jV1Gp1bH4+2d+bj5w0xcrGlFsK8XHp09FL8um9qpHEJHqAF9sbYVBbWtiAjwNlu3rCOPZgyDn1SMk6VyzHlrHx7/8SSOFtYLNcgKaluFOmRPXJVk8TOmOqDDdgihaS2uwRH+HuOx6wo9MvJly5ahqKgISqUShw4dMglD7d69G1988YXJ/jfddBPy8vKgVCpx+vRpXHXVVSbPcxyH5557DpWVlVAoFNixYweGDh1qsk9oaCjWrVuH5uZmyOVyfPbZZ/D3t904z11IJSIhNfaUUdaFcR0TS19cb4kYI+LIRHW8uBGZF+tQ16pCiK+X0y+M1AN0olRuIt7MvFALrY7HoAg/kyabtOO4uZAZbVQZH2I5BCbzEmO8XuNkqSpti1IjuKivtdLDyROJCZIhQJ8J1pXUciqA7in9D0Ayt2RepAVLV8ZM9T9Do/ztrkZujgevHILn5o0AxxFBZsfvR7EDVaDdwcBwP0QGeEOl0XXSyRXWtuJQQT04DkIfJYlYJKQr59loUfHs72eEkMfWM1XYYqHn3FH9xS6lg/6HMoLqgKx4+6gBZC4DsbeTHBOImCAZFGqd2TmsI2/tOIc3tp+Djicenb2PzcSyK4bY1Qk9IsAb88fEgeOARWkDsPPf04WwlDUSw/3wx78ux5wR0dDqeKw/WoIb12bi8lf+wuo/c/Dfjaeh1vKYMSwCVyRZrq49MZHMs7mVTTazDalmsSshbE+i95pufZAx+iJktCgZYL+OgYqF9+XX4ssDpJrx3FExJvUmnEFiuB/C/KRQaXQm9YBoN+GpHWoNjTMj0AaIbqCkwb4LFHVDbz5lfhLferoSCrUOAyP8MDKuZ8Xr3YXjOAzuohC6pL4NpQ3tkIg4YfXWE4hFHIZFd10InefEyXNRWoLQyX2PUZ2cZoVaEK66MgTWHYxbGhzsoAP6Ud9Fe9qQCJM+SvQzs2YAfX+4GN8cLAbHGcLR//31jFnxNNXWWRIv07pAZ6z8n6kB1J0K456K8f/IUqiSwvO88H/77zXJeHPBmE4FDG3x8g2jcObZDDw3b6RDr00I98PaO8Zjw31puH5sHPykYpQ1tuPDPRfx9/laSEQcnrrafO8zSmSADOP6B4PnISTQWCJPH7IfaiGLs7fADCAPgmZhGFcGtiWAplAd0E/HSoVaEq4oJsZxHMZRHZBRGIwK5zquAqkH6Ey53MTNX9+qQpvepWzcm8sctEHm0aIGs1qiX/U/VrJ68iwXuj0MjaRF8RzTAdHw15j4YLtSj50JrWPTFR0QTZ931uRJRZ37zxsuUNT7E+ontWv17S6oDmhXXrUQstDqDBfSjm1YaNmAPCvG8nu7SXmAf88aio/uGI+B4X6oaVbi5a25JvvVtSgFMTrV2nWEhsCK6trQbKbicKVcgZL6dog4YJx+Dupr0P+RrbYTZ8qbUC5XwFcqFnoZOopELLIZKrPGxIRQvLFgDLL+Owvv3TqONKX1luDh9CGC3tAa9Jrxa7YNA0j/G+6qhs9TYAaQB0HFuydL5MJkaK0GkDETEkjNHICk2L6zcKzLvAK0aBYVQjcp1MjR/yAmdXjPfiE+CPf3hlrLm3RbpxlgUYHeNnvFRAfJhPfs2B2+plkpNPe7zsl6p55iiN4DZEnAaAma9dbR69YTdEcITb0X9rj37WHywDBwHCn0Rl33jjRBdSczhpG2EafLmgQDZe+5GlQ2KRDi64X0ZNNCitQDdK6q2az+qqrJYJDcOSUBMi+xUILhx6OlJg03qdc2KTrAYuHBUD+pUETSXGd46v0ZHhPo0YZmd6AeoJOlcqvFP7fpK2FPHxrh9v5XMi8xrh4dgw/vmIBTz2Zg2RXWBdiUq0fHQqJvBHveQmaqSqMTyrUMZQYQw1kMjQqAzEuEZqUGBXptBfUA2VLaxwT54Lslk/Ht/6Vi68PTnC5+NoamPR7T90Ei/ZCIlyoy0LTiLsdxGD8gGICpDsieDDBjaNGvPzqEwTadLIeOJ16Qjo1iewtCQ9tq+z1ACrUWe/XF55zZNdteDKnwjhlAbSqNoGtzln4g2FeKkfqiffv1XjFBAO3hBlB0kAyv3jgaAPDhnou49+ujuP9bUr5j/ti4TsLkAWEG/ZW5GkhU02NskKQNCsPofkFQaXX4QV9tGTCEdGzV7qHG7pmyzkJoagD1ZAi2p+kX4ov+ob7Q6nirrSJowUR3/B6dRaifVKgTtPG4eS9QQW0rNDoeAd4SxDq5wnpPwwwgD8JLLBKqr54sbYROxwsXiwGhti/uExJCMXVweLeEpfYwMi4Q3vrmeSdK5cKkYGkSHBNPhdONwjZ79T+UuSNjwHHEiKJhsKomBb7VNwGdP6Z3en8AQzHEgtpWs9lA5si8UId2tRYxQTJBqNqTJOkvilVNShOvgi3OlDeB54FwfykiApzX7oB6wfafJ2GKoh6oAeQs5o2Jw7+uJCv0rWeqoFDrMLZ/MB6YMbjTvmIRhyH6kKk1j0zH3yItyvjtoSLBc2SvgJ5+v06b0QHRMgwdWyv0NaYIOiDzYbCiulbkVjZDLOJwRZJnNwC1xXx9GGxjdpkQiTCGZoANjQ7olZIDY5gB5GHQbIwTJXJUNSug0pBeRrHBnmNpe0vEgkdm7e4LhiwQCwYQ7TBtXPHWngwwY6KDZEIG2vL1J7BmxzlkrNmL89UtCJRJcE0vDX8BpG9auL83tDrearqxMXS1mT48yi2TkL+3RNClnTLjGbDEvnz7vA6OMnUwOd7+87Xgeb5HagA5k0fSh+D/LkvEzGER+OquSfj5/ikWDUSqnTInhKbp8RP0IWPKtaNjEeTjhdKGduw5V42qJgUu2llAc6z+d3dA/9lSapqVyK9uIcfoJb2fugo1EqnuriO0TlBqYiiCfaU9Ni5XkD48Cv7eEpQ2tJut4k+TNXq7ABpgBpDHIRgLpY3CKtZVvYy6Ay3KtuVMpVCBeqKFVeCouCBwHFDW2C40sLSnBlBHbhhH0oEzL9ZhzY58NLapMTIuED8/MNViz6XeAMdxQgZgth3NMXU6Xkj7T3eju52KQy1l55mD9oqbNsQ51ckpEwaEQioWoUKuQEFtqxAeslQ81NPgOA5PXZOMzxdPwjQbPd2o8DSvytQj06LUCFl5EzpU5/WRinHTePL7+WjvRfyuTxwYERtos4Dm5MQwSCUilMsVOG8UpqUZZEnRgQjx690XfVvQ7/qZ8iaz2XRb9fqf2b04/EXxkYqFDvXm+qDlVXaviKkn4VlXVYaQCXamvAm/ZpPGeJ64ih0aFSD82DU6HuH+3hZ1SgEyLwzWVx+lGW6OaoAA0iRxw31peHT2UGSMiMKjs4fi5/un2pXd4OnQ+i4nOjS0NcepMjmqm5Xw95ZYzN7pCWjGyOZTlXYViZO3qYXzu8zJ/Z58pGIh4/Drg0Uo1YvsPfG3010spcIfL26AjieJB9FmtBm36cNgBy/W44U/cgDAribJPlKxEOIyLjUghND6uPcHACIDZRgU4QeeJ0URjSmqa8WRwgZwHDBrhPlek72NWcnkPHaYM4CqnJvF6U6YAeRhJIT5IlAmgUqjw3eHSUf69OGeuap4YKZBozApMcTqqnW0ENprhFbHo6zRdh+wjnAcqXez7Ioh+PCOCVh2xRCHK896Kil6AyjbDgOIljmYPjTCqW1OHGViQijiQ33QotRg29lKm/sfuFALHQ8MivBDrI3SB13hqlFk0v58fyE0Oh5SsQjRgZ4TOnYWdOVdWNdmohk7ohdAW9LiJYb74a1bxiAlPhgiDuA4YM5I+y7YM4YRj4CxAUQ9QD1Zhdyd0M9q7Z4LJqHA9UfIPH3Z4HCbJT16C9OGhsNLTKpSX6gxeP3qWpSCfIF5gBhOh2RNkZVsRIA33lk4FovSLHeVdidj4oOFzs2XDbYe0hBCPKVy7MuvgVrLI0AmMSnydilDDcTi+jabnaep3qBjinRPIxJx+MdYElb5+ViZyXNnyuV4bWse7vj0EK5952+cLW/CPn2dnsudHP6i3JY6AC/MH4lQfTgmMdzP5QkB7iAiwBvBvl7Q6niTkFSWBf2PMfPGxOHXpVNx4unZyPzPlUKhUlvQzKBDF+vRptKgUh9qFHHAJDd6IXuSuy8bCF+pGGfKm7Ajh5SgUGt12KCv2bRwUtdq/3giATIvpOm9g8Z90IxLJ/SFsGfPVk9j2MWz143EFeeqMW9snMOVRHuadxeOw+5z1bjGRgsK6uE4WdqILw4UAgBuGh8PcR+8QHWFIB8vDIzww8WaVpwoacRMC5kkJfVtQrbJzGHuzza5fmwc3t6Zj335NahuUiAyUIbdedW456ssqLSG3mb//PwwqINwmplu1M5AJOJw++QBuDYlFuuPFPfqLtXW4DgOw6ICcKigHnmVzRgZFwS1VmfQ4tmRkh4g83Kobs+gCD/EBfugrLEdBy/WQd5OdDAj44I8fo5yFqF+Utw5JQEf7L6ANTvOIX14JP7KrUZNsxJhflKP9dR3lVnDI7H3XA12nK3CfdOJ5pOKwHuy9Y4rYR4gD6R/mC/uSEvoFRNLkK8X5o2Js2nIJEWThq2NbWqhZ5OnerbcxRg7wmA0/DVhQIhHZJskhPth/IAQ6Hhg2brj+GTfRdzzNTF+pgwKw/PzR2JYVACqm5WoalLCS8wh1UbWUXcJ8vHCPdMGCZ7UvshIfZdwqsM5cKEObSotwv2lgt7OmXAch+nDDPVh/jhJhO995UJoL0suN3iBXvozF5/sI01Gbxzfr8+E4yk0wSKruAG1+lIXtOyBPdqx3oDL/mP19fW47bbbEBgYiODgYNx9991oabFe6E2hUGDp0qUICwuDv78/brjhBlRVmYqwiouLcfXVV8PX1xeRkZFYsWIFNBpDdc7du3eD47hOt8pK2xoFhuuQSkQYblSvZsawiF5buNBVOGIAeVKxtQdmDIKIAw4XEnGtSqPD7OQofHnXJNwxeQA+XzwRUYEkS2/8gJAeb9vRF6F6lG1nKqFQa7FZb5DMGRntsrDfDH0Y7LcT5UII6FIQQBsT6ifFP6ckAAA+3HsRh/UlQBZMjLfyqt5JTJAPRsYFgueBnTlVqJC34yINe/aRuk8um4luu+02VFRUYPv27VCr1Vi8eDHuuecerFu3zuJrHnnkEfzxxx/YsGEDgoKCsGzZMlx//fXYv38/AECr1eLqq69GdHQ0Dhw4gIqKCixatAheXl548cUXTY6Vl5eHwEDDBTcy0v3hgkudMf2ChCygO/WTCMOAUAOqtBE8z3cSlcvb1Tikj8F7kgF05fAo7Pr3DKw7XIyNx8swMTEUb948RmjEGxvsg6/uSsWrW/Nw92WJbh5t32B8/xDEBMlQIVdgZ041tupF6LQ+lyu4fEgExvUPRklDO2KDZBjdLxiXuaENi7tZPmsoogJlOHixDmcrmjB9aAQGusDr5gnMHRmD02VNWLvnIqjue1RcEIJ8PD86YQ8cbyxndxI5OTlITk7GkSNHMGHCBADAli1bcNVVV6G0tBSxsZ31InK5HBEREVi3bh1uvPFGAEBubi6GDx+OzMxMTJ48GX/++SeuueYalJeXIyqKXADWrl2Lxx9/HDU1NZBKpdi9ezdmzpyJhoYGBAcHd/kcmpqaEBQUBLlcbmJIMbrOr9lleOj7bCSE+WLXv2f0SYFqd1BpdBj59FaotDrsfnRGJw8Z/fyGRPpj+/Lpbholw1N4YdNZfPJ3gaDNCfeX4tAT6UxXx3AazQo1Zr62B7UtSgTKJGhSaHDf9EH4z9wkdw/NKvZev10SAsvMzERwcLBg/ABAeno6RCIRDh06ZPY1WVlZUKvVSE9PF7YlJSWhf//+yMzMFI47atQowfgBgIyMDDQ1NeHMmTMmxxszZgxiYmIwa9YswYNkDaVSiaamJpMbw7lcPSoGT1yVhLV3jGfGjxmkEpHQEHdfh1ojAISwgzuLHzI8B9rvj5aUmDMymhk/DKcSIPMSjJ0mBZGa9KWyBy4xgCorKzuFnCQSCUJDQy1qcSorKyGVSjt5baKiooTXVFZWmhg/9Hn6HADExMRg7dq1+Omnn/DTTz8hPj4eM2bMwLFjx6yOefXq1QgKChJu8fF9L6brbiRiEe6ZNshpXcD7ItS46ViBVanRYnee3gDqY9kmjK4xul+QSaFHV4a/GJcu14+NE/SJXmLOapmF3oZDBtB//vMfswJj41tubq6rxmoXw4YNw7333ovx48djypQp+OyzzzBlyhS8+eabVl+3cuVKyOVy4VZSUtJDI2YwDGToK8lmXqgzKbm/91wtmhUaRAV6Y6x+MmJc2nAch2tGE6Mn3F/q8uw6xqWJSMThuXkj4CsVY3ZyNHylfSeJwaEz+fe//41//vOfVvcZOHAgoqOjUV1dbbJdo9Ggvr4e0dHmK49GR0dDpVKhsbHRxAtUVVUlvCY6OhqHDx82eR3NErN0XACYNGkS/v77b6vj9vb2hrd37+0nxegbJIb7YUikP/KrW7Arr0ooNEjbolwzOpaFDxkCi9IScLigHjeO78fCXwyXMbpfMA49cWWfMn4ABw2giIgIRETYruKalpaGxsZGZGVlYfz48QCAXbt2QafTITU11exrxo8fDy8vL+zcuRM33HADAJLJVVxcjLS0NOG4//vf/1BdXS2E2LZv347AwEAkJydbHE92djZiYph7mNE7yBgRjfzq89h6mhhArUqNkP5+XS/ues9wPtFBMvx4/xR3D4NxCeBI4czegks0QMOHD8ecOXOwZMkSHD58GPv378eyZctwyy23CBlgZWVlSEpKEjw6QUFBuPvuu7F8+XL89ddfyMrKwuLFi5GWlobJkycDAGbPno3k5GTccccdOHHiBLZu3YqnnnoKS5cuFbw3a9aswa+//orz58/j9OnTePjhh7Fr1y4sXbrUFafKYDgdGgbbc64GCrUWO3KqoFDrkBDmK4ikGQwGg9E9XObP+vbbb7Fs2TJceeWVEIlEuOGGG/D2228Lz6vVauTl5aGtrU3Y9uabbwr7KpVKZGRk4P333xeeF4vF2LRpE+6//36kpaXBz88Pd955J5577jlhH5VKhX//+98oKyuDr68vRo8ejR07dmDmzJmuOlUGw6mMjAsUUpu/zizCfn35+etSYq02nGUwGAyG/bikDlBfgNUBYriT1X/m4MM9F0227Vg+DYMje38HZgaDwXAl9l6/+5aiicHoIzySPhReIhG+PVSEhjY1UuKDmfHDYDAYToR5gCzAPEAMT6BdpcX+87UYGReE6CCZu4fDYDAYHg/zADEYfQAfqZhVfmYwGAwXwAwgC1DHGGuJwWAwGAxG74Fet20FuJgBZIHm5mYAYC0xGAwGg8HohTQ3NyMoyHLpEKYBsoBOp0N5eTkCAgKcmnrc1NSE+Ph4lJSU9FltETvH3k9fPz+AnWNfoK+fH9D3z9EV58fzPJqbmxEbGwuRyHK5Q+YBsoBIJEK/fv1cdvzAwMA++WU2hp1j76evnx/AzrEv0NfPD+j75+js87Pm+aG4pBI0g8FgMBgMhifDDCAGg8FgMBiXHMwA6mG8vb3x9NNP9+nO8+wcez99/fwAdo59gb5+fkDfP0d3nh8TQTMYDAaDwbjkYB4gBoPBYDAYlxzMAGIwGAwGg3HJwQwgBoPBYDAYlxzMAGIwGAwGg3HJwQwgBoPBYDAYlxzMAGIwGAwGg3HJwQwgBoPBYDAYlxweYQC99957SEhIgEwmQ2pqKg4fPmx1/w0bNiApKQkymQyjRo3C5s2bhefUajUef/xxjBo1Cn5+foiNjcWiRYtQXl7u6tNgMBgMBoPRS3B7IcT169dj0aJFWLt2LVJTU7FmzRps2LABeXl5iIyM7LT/gQMHMG3aNKxevRrXXHMN1q1bh5dffhnHjh3DyJEjIZfLceONN2LJkiVISUlBQ0MDHnroIWi1Whw9etTucbmqGzyDwWAwGAzXYW83ePBuZtKkSfzSpUuFv7VaLR8bG8uvXr3a7P4333wzf/XVV5tsS01N5e+9916L73H48GEeAF9UVGT3uEpKSngA7MZu7MZu7MZu7NYLbyUlJVav8xK4EZVKhaysLKxcuVLYJhKJkJ6ejszMTLOvyczMxPLly022ZWRkYOPGjRbfRy6Xg+M4BAcH2z22gIAAAEBJSQkCAwPtfh2DwWAwGAz30dTUhPj4eOE6bgm3GkC1tbXQarWIiooy2R4VFYXc3Fyzr6msrDS7f2Vlpdn9FQoFHn/8cSxcuNCqIaNUKqFUKoW/m5ubAQCBgYHMAGIwGF2j6gzQVA6IJEDkcCAg2t0jYjAuGWzJV9xqALkatVqNm2++GTzP44MPPrC67+rVq/Hss8/20MgYDEafpzoH+GCK4W9ZELA8F5D6um9MDAZDwK1ZYOHh4RCLxaiqqjLZXlVVheho8yul6Ohou/anxk9RURG2b99u04uzcuVKyOVy4VZSUtKFM2IwGAw9FSfJvXcg8QAp5EBjsXvHxGAwBNxqAEmlUowfPx47d+4Utul0OuzcuRNpaWlmX5OWlmayPwBs377dZH9q/OTn52PHjh0ICwuzORZvb28h3MXCXgwGo9vI9Yuo4dcCYYPJ4+YK942HwWCY4PYQ2PLly3HnnXdiwoQJmDRpEtasWYPW1lYsXrwYALBo0SLExcVh9erVAICHHnoI06dPx+uvv46rr74a33//PY4ePYqPPvoIADF+brzxRhw7dgybNm2CVqsV9EGhoaGQSqXuOVEGg3FpIS8l90H9iA6oJhdoNq9VZDCcjVarhVqtdvcwXIKXlxfEYnG3j+N2A2jBggWoqanBqlWrUFlZiTFjxmDLli2C0Lm4uNgkj3/KlClYt24dnnrqKTzxxBMYMmQINm7ciJEjRwIAysrK8NtvvwEAxowZY/Jef/31F2bMmNEj58VgMC5xjA2gRr03qJkVZGW4Fp7nUVlZicbGRncPxaUEBwcjOjq6W3X63F4I0VNpampCUFAQ5HI5C4cxGJ5IWz2Qtxk4txXQKIAbPgVkHvRbfW8yUJMD3PELULAP+PsNYNI9wFWvuntkjD5MRUUFGhsbERkZCV9fX+cX8m2tBXgt4BcJuKFIMM/zaGtrQ3V1NYKDgxETE9NpH3uv3273ADEYDIbDaNUkw8pYU3NuCzD6ZveNyRieN2iAguKBAP0kzTRADBei1WoF48ce7avjb6AGlDXkcWA4IPVz/nvYgY+PDwCguroakZGRXQ6HeUQvMAaDwXCI+ovEmBB7AxHD9dsK3DsmYxRyQNVCHgfGGer/MA0Qw4VQzY+vr4tKLSibjR43ueY97ISeY3d0TswAYjAYvY/afHIfmQSMupE8bvAgA4jqf3zDSN0fwQPEDCCG63FZ/0pjA0jhXgPIGefIDCAGg9H7qDtP7sOGAKGJ5LEneYCMBdAAEGhkAOl07hkTg9EdeN7UAFK3kZBYL4YZQAwGo/dRp/cAhQ8BQgeSxx7lATLS/wCAv759j04NtNe7Z0wMRnfQKMn3FxwgkZFtbg6DdRdmADEYjN5HLfUADQZC9B6glipA2eK+MRnT0QMk9gL8IsjjJpYKz+iFqPTeH6kfaesCODUMtnfvXlx77bWIjY0Fx3FWG5w7C2YAMRiM3oexB8gnGPAJIX83FLprRKZ0NIAAJoRm9G5o+Ms7wGAAKZsB3jkh3dbWVqSkpOC9995zyvHsgaXBMxiM3kVbPdBWRx6HDiL3IYlAewMJg0WPdN/YKGYNoBig8hRLhWf0Pnje4F31DgC8fEl/O50GULUB3v7dfou5c+di7ty53T6OIzADiMFg9C6oADog1jDxhg4Eyo95jhBaMIDiDdtYJhijp+F5IlbuLqo2UtaBExsdkwPU7WRBYi4jy8vXLYUSHYEZQAwGo3dBDaDwwYZtQibYxZ4fT0e0GkPLi44eIIB5gBg9h7oNeDHWPe/9RLnbCiXaC9MAMRiM3gWtARQ2xLCNCqE9IROsuYLoIkRepF0AhWmAGAyPgnmAGAxG78JYAE1xRy2gtnqihxB7mW4Xwl9xgFEjZ4MHiGWBMXoIL1/iiekOOg1QdYY8jkgCJN7kcUsN+S7LgoGQAebf28NhBhCDwehd1BoVQaTQWkDyEkCjAiRS146h7Bjw6Sxg9C3A/A5ZK+b0PwDzADF6Ho7rfhiqvQHw8iG1f3xDDdtlSkDRQBYAHh7qsgQLgTEYjJ6huRLY/TLwzgRg82NdO4ZOa9D5hA0ybPePIitOXmcoQuhKTv5AVsanfgDaG02fE4og9jPdTj1ALdVEJ8Rg9AZosUPvANPt1POpc0416JaWFmRnZyM7OxsAUFBQgOzsbBQXFzvl+OZgBlBvh+eB3M3EHc9geCqlWcBbKcDuF0kI68gnJLPEUeQlgFZJmqAG9zds5zggJIE87okwWP42cq9VAXmbTZ+jIm06HopfBMmiAQ+0Vrt6hAxG9+F5QEHr/wSaPifSB5CcZMwfPXoUY8eOxdixYwEAy5cvx9ixY7Fq1SqnHN8czADq7Zz4Hvh+IfDltYBa4e7RMBjmOfopoFEAkSOIZoDXkpo4jkLDX6EDAZHY9LmeEkLXXQDqLxj+PvOL6fNVp8l91AjT7SKRURiMZYIxegEahd7DIwKkHWr9iPQeIF7rlGKIM2bMAM/znW5ffPFFt49tCWYA9XZObSD3VaeB7f9171gYDHPodMD5HeTxnBeB/mnkcfkxx49FjRvj8Belp4TQ1PtDDa4LuwweWK0GqMkjjzsaQIDBAGpiBhCjF0D71nn7mwr6AdMFiK53hnSZAdSbaW8ACvYY/j78EQmHMRieRNVp0qfLy48YP3HjyPayrhhAheQ+2EzWCRUdN5V2aZh2c24ruZ/4f0DUSDL55/5BttVfJKtmLz8gOKHza2lTVBYCY3g6ymaiVwMA37DOz3Oc08NgPQ0zgHozeVvI5BuZDKQtI9s2PUJW3AzH0KqZjspVnN9O7hOnkRRaagB1yQNUSO476msAIFBf8M2VzUaVLUDRfvJ4yGxgxHzy+MzP5F4IfyV3XjEDhoaoLTWuGyOD0V20GqChiDz2CSX99sxBw2DMA8Tocc7+Su6HXwdcuYqsOlsqgZoc946rN/LDncDrwwyre4bzOL+T3A++ktzH6g2guvOdM6hsQSdlcwZQUBy5d6UBVLCHCJ9DEkgdohHXk+0X9xADmtZLMRf+AgB/fWFE5gFieDJNpUT7I/bunM1oDPUAOSkTrKdhBlBvRdFEtAcAkDyPrKzjJ5G/iw64b1y9kaqzQN4f5ML2411A5Wl3j6jvoJADxQfJ48Hp5N431GDAlB+3/1g8DzRSA8hMCCxQbwA1V7jOJX9RH3IePIuEAMIGEWE3rwXObTEygCw0ZKUeoFbmAWK4Dp7nu/5ijYrIKwDyO+uYbGCMmBpAPe8B6tY56mEGUG8lfxtJBw4bDEQOJ9sGTCX31EXPsI8jH5N7Tkwa/q1bADRXuXdMfYWLe4hxEDbYIFIGDF4gR8Jg7Q2GmiTGKfAUv0iyIuV1RHPkCmj2V/Qow7bh15D7nE0GAygy2fzrWQjMuWjV5ILNAAB4eZGQVFtbNxqgttWRe6m/7QKHNATmBg0QPUd6zl2BVYLurZzbQu6HX2fouDtgCrkvOkBWyx7eidcjaG8kpQQA4OYvgR3PkNDMwfeAWc+55j21apICHju27/+PaPYX9f5Q4sYR3YwjQmiaARYQQyrTdkQkIh3i5cVAU5khJOZMGsx4oJKuAfa8TM5VqyTbomwYQCwE1n10WuCr+UBZFnDFk8DkB6x7Ky4BxGIxgoODUV1Nvl++vr7gHJljeB0gryHXD99AQGGjtIqGJzeFAvDumTIsPM+jra0N1dXVCA4Ohljc9f85M4B6KzR0kDDVsC1uPCCWktVv/UXzqcKOomgC9r0GnPqRGASjbuz+MT2J7HWkY3JkMrmQKZqAXx8ASo+65v20GuC7W8jF8qrXgElLXPM+nkJxJrkfOMN0u+ABciAEZi0DjBIUZzCAnI1OBzQWdx5D9CjikaLPBfYDfELMH0PQAF2CHqCmcuDr6wFZEDD6ZmDk9ZY/J3vIXgcU/U0eb3sKOPsbcOt603YNPUXOJqAmF5j6UOfecD1MdDQptUCNIIdQtwGttcQb3iIDOBuJIapW4jGStAD+PeuJCw4OFs61qzADqDeiaDJUm41OMWz3kgFxE4DiA8QL1F0DqGAv0cTQyXr/mr5lAOl0hvDXpCXEGxNLqpCi4gRZYTpzRcnzwOZHDV6RA28DE+7qu6vWtnqg9hx53G+S6XMxKQAnIoZKcxUQEGX7eNYE0BSaCSZ3gQHUUkU8PJzIVBjKcUDStcRrCFgWQAMGD5BCDmiUhsaSlwLHvjIkaJQcBPa+Cty712AUOoKyBdj1PHmcdA0JtZYeBg5+QLxB3UWrIeOLHQsMm2N938Zi4MfFRENYdQa44RO3/qY5jkNMTAwiIyOhVjsoTv7lfqDsCDD+LmDMZbb3L9wPbP83ED4MuOXbrg24C3h5eXXL80NhGqDeCE21DYwD/CNMnzMOg3WXzY8R4yd0EACOhG3kLq6x0pNc2Ek8Zd5BwKibybbwoaSnlKrFYGQ6i4PvA1mfA+BIxl5jcec2Cn2JsixyHzoQ8OtQR8Tb3+BFsfdztpYCTwl0YSYYFWAH9uu8yqc6IMC6ASQLNmTOtNY6dXgeDc8TLzIAjLwRCOpPxOqbV3TtePvfIgZpSCJw42ekwCZAFm3O4MR3wJ6XgB8Wkcrf1tj9MjF+ABLW/XWZR5QiEYvFkMlk9t8UNZDl/QxZSylkY2+y7zWBoZC1lEBWd8ax9+rmzRnGD8AMoN5JxQlyH5PS+TnBAOqmELr2PFmtiSTAkp2GDDOqPeoLHP6I3I+9nVyQAZLVED2aPHYkPGMLVSvwl36SzvgfkHoveXzwA+e9h6dReoTcd/T+UIL1hQvtNaodMoBc4AEyp/+hxKcCvuHksTUDSCS6NHVAladIDziJDLjmTeCWb0iY5exGIOf3zvvrdMQzqG7v/FxjMXDgHfJ41nPEi5Y4jfxddpR4h7oDzwOZem+eVgn89i/LBk1NHnBiHXk8bQU5pxPrgONfd28M7iB3E7kfMMXw27SFn1FI1wlZWT0NM4B6I+XZ5D5mTOfn4ieRH2FjUfe8Nbn6SSlxGonTD5tL/s7rIwZQ3QUgX1+gb+Ldps/RMJgzDaDcP4hXKSSRiDUn/h8xLov2GwzavkbJYXIfP9H880H6TC65nd2eraXAU4RiiC4wgOj7m9MgicTANW8AY24nIRlr+OkNJU/LBKu/aEh/djanfyL3Q2YDskCyeJv6ENm2aTnx6OT8Dmx9Enh3EvBCJPD6UOC1oURfY8yfjwOadpL1Ovxasi0kgeiwdBoSXusOF3aSxZ/Un3iDi/4Gjn9lft+//keEw8OuBq54Cpi5kmw/ub57YzCmtY4soFwNNUTpZ2oP9Lus07juu+NCmAHUG7HmAfIOMGwvyuz6e9BJh/4YhuoNoIK9PfNjdDVHPgXAk3ouHbVSrjCAaKZZyi1EMxIUR+o3AQZPVF9CpzOEwPpZMoD0Ohp7DHWtBmgsIY+teYBcWQzRmgcIIP/P+e8RLZ41/DxQCF2UCbw7Efj2Zucfm+eB0/pK2SNvMGyf/jgJObdWA9tXAetvBzLfBWrzDIX1lE3A+tuAv1YTzVTuHyRsLJIAV79hmkVJvUDdDYNR78+4RcAV+v6K21YR7aUxtfn6YrQcMX4AIGUhuS864JzvYM054M0RxBDc9AhQndv9Y5qjpdogm7BlwBsj8SaidqBXhnSZAdTbULWRCQIwbwAB3Q+DycuIKxkcWdkAQMQwcuHRKoELf3XtuJ6CqhU4/g15TENRxghC6JPOqW/RXAlc1H9mo40uMOPuJPfntnqEZsCp1OaRi5eXHykUaA7qZqeGjTWaSkk9IbE34G8l88OVxRCteYAcwdOqQWvVwB/LySq+9LDzW8KUHiVePqk/MDTDsN1LBiz+E7jyaVLOI3IEMSBu+gJ4+DTwZCWQeh/Zd89LwBvDSTgKAKY8CEQmmb5PghMMoKqzpMAsJyJzQ+q9RAOplBua4FKoUTc43VD2IKgfED8ZAA+c2dj1cVBOfEe8XaoW4OhnwEczXNOyJ28zAJ7MffaGvyi9OKTLDKDeRtUZ4nL1izR0lu6IUBCxi0Jo2tgxPtWQncNxBi9Q3p9dO64r0GpI0z5H2PcGmdBCEoFBV3Z+Pmwwmaw17QZjszuc2kD+Z/GpRBBM6T+ZGAitNUD1me6/jydBw19x4wzVYjviiAdISIHvb77HFsWVxRBteYDshYYNPGXFfGgtUH3W8DfVbjkLqpFJurpz/Sa/cODy5cCCr4EHDgD/WAuM+Ae5CHv5AHNfBuZ/QGo/tdUBbbXkOzDtsc7vk3g5ua844XiLFQoNXSVdTRZ8IjGQfB3ZRudFyplfyP3I6023Uy8XDft1FZ43vMfUh4n4XtMOFO7r3nHN0ZXwF8UTPZp2wgyg3kZFNrmPHWO5iF7/yeS+Nq9rOgOq/xnewRVK00HzPcRjoWwBPp4BrO4HvDIQ+HQ20RDkbLIshCw9Cvz9Bnmc/oz5i6lIZNBXOSMMdkI/qabcYrpd4m2o49TbvWodEQTQFsJfgKF7u7zEtoDSnhR4wFAMEXCuDkirNnSZ764HiF4wWjxgxdxUDux+iTymnrWSQ847fnuDIfw79o6uHWPMrcQjdOsPwMQlwIJvAKlv5/0CY4GwIcT47erijxruQ41S32lIKH87CcMBJBRVk0MqIQ+7yvQYyfOIB6nsqMFw7wqVJ0nxT4mMCKypDrPAyQaQQm5o8TL8Osdf76maNjtgBlBvgxpAlsJfACkERkvxFzuoA2qpIbUdgM6x4P5TAO9AYulTfYe74Hng16UkuwQgq8OSQ0RDsP424LUhwE//ZzoRqtqAX+4lE+SomwydvM0RO4bcd9cAqjoLVJ0iBSpH/KPz8wNnknva162vYI8BRMNV6jbbbn17MsAogg7IiQaQvJR8b8TegL8dNYus4Ukhg+PfkvBKv4kGAW+xEw2gY1+T/2/USCDBjroylhBLSPjs6tesz33UC9SVMJhWbfi9G39vY8cR41DVbDju2Y3kftAVnTulB0QZzpV6cLoCfe2QWSRLlZ6bIx4gjYpkmm57imiITv7QeZ+jnxPNVfgw0uDXUXpxcU+PMIDee+89JCQkQCaTITU1FYcPH7a6/4YNG5CUlASZTIZRo0Zh82bTWio8z2PVqlWIiYmBj48P0tPTkZ+f78pT6DmsCaCN6Wo9oLMbidYidqxp7yYAkEgNHb3PuTkMduAdMlaRF7DoV+C+v4F/fASMX0xCW+o2Enr6fC7w491A9nfAJ+mk5kxADHDVq9aPT3VA5i4Gx78FXh9uXwq7cfaLuaq3g67Qv08moO6ZUvLIfI98Fh9MBT6cbjAinYVCTtKDAesGkJfMYEzIbeiAaA+ujt9JcwiZYE4UQgv6HxshOHugtbs8IQRGa4olzwP6p5HHZVnEGOgKPE88W+p2Ep4+rC80mnpvz7R9oYZHVzLBKk+REJMsiHiSKCIRCYkBhlRxapyYW9QYb8/tYp0v3khDRI81QH9uNbn2ew8Pvgds+Q+ZL49+Bvy8xLTsgLyUtHEBgMse7tpYPcmgdxC3G0Dr16/H8uXL8fTTT+PYsWNISUlBRkaGxTLeBw4cwMKFC3H33Xfj+PHjmD9/PubPn4/Tpw0dvF955RW8/fbbWLt2LQ4dOgQ/Pz9kZGRAYauvSU/QVk+yCboSQmqrNzRbpK0ELNFVITQtVjbqJvPPU3evO9PhK0+Rnl0AMGc1abMQPQpIWQBcuwb413FgyS6SxQEOOP0jsPE+orORBgDXf2S7BP/AmURLUnWKZGJQjn5GWmU0l5OJhYa3zMHzBgPI0kQZMYwYZBqF4966rlB8ENj6BPHQVJ0mHsVNy51bw6M8GwBPjIWOhTo7YhwGs0atfgETPtT2+1PPkjOrQTtL/wN4VgisWl+ZOXI4uejLgokR4KhR3FgMfH4V8L9o4n19dTDxxMqLAZ9Qy/OJs6ELl6qzjhtxtP1Nv4mdjVzBANpM5siaXOLVpWGpjtCFTfmxrmXNVpzQh798gCF64bhfGPGkAfZ5gTRK4OBa8njkDYaEll/uM2STbfkPWSz2TzNksDkK9QCxEJjjvPHGG1iyZAkWL16M5ORkrF27Fr6+vvjss8/M7v/WW29hzpw5WLFiBYYPH47nn38e48aNw7vvvguAeH/WrFmDp556CvPmzcPo0aPx1Vdfoby8HBs3buzBM7PA+tuBl+KB50KAlxMdWyFc2EXc8JHJths99tcbQJWnyIrcHhpL9CsnDhhxvfl9BqeTOkPVZwy9j3qanc8RL9Xwa0k9nY5wHOmLdt07wD1/kUJ8vmHAjJXAwycN6bLW8AszTGKn9G7jY18RNzJgmIh+XQpc3G3+GOXHySTm5Wt5ouS4nguDadWG8SfPB275joyt9LA+ndcIZbPtCriWoB3ebRnpgH1CaJ3WMJawwbaP6YpiiM7KAAMMK+a2WnJu7kKjNHjWIoaTiz4teFpi3Qvfic2PkcWWRr/IVLUYiqZOWGy+ea0rCE4gYXqt0uCFtJdS/TmbK9yZcDmpGN9aDfykrxs26MrO4S9hHAOIca/TdE1TRb00NPxlPA7APh3QqQ1ASyXRxM1fS5o9J1xO/jdfXkO8wDm/k/n86te77qGjXtyWyq693o241QBSqVTIyspCerqhU7RIJEJ6ejoyM82vhjMzM032B4CMjAxh/4KCAlRWVprsExQUhNTUVIvH7FGMVwPt9SQDw14sddY2R2CMPuOItz+mT70VCZeR15vDN9QgsnaHF6hwP0lHFUmA9Gdt/2hjxwL/tx1YcQGY8R/HGiWOXkDuT20gF+g//0P+nvIgcO8+YiTq1MCGfwJNFZ1fTz/PoXMAqZ/l96GGFk2VdxWZ75FsH59QUo036Spgij6teMfTBoFnewPw0UxSF6YrRRqpPixuvO197UmFbywmFzSxN/Eq2cIVxRCd6gHSi0Z5nXuLx9WdJxdo70DDZxafSu4dCSGd30FC4iIJcNc24Klq4I5fiLc4fjKQer/zx24JkchQyb3ypGOvpUZfvwmdn5NIDUkMAbHA+H+SDDVLcJwhHFf4t2PjAEgxRqDzwsleHRDPG6plT76PjF/sRUoMBPcneh2q05t8v/Xq5bYQDCAP8Gg6iFsNoNraWmi1WkRFmYoKo6KiUFlp3pqsrKy0uj+9d+SYAKBUKtHU1GRycwn37Cb1LZboL3ZFB+xL49bpDAbQkFn2vRcNg1nyUHTkNO3Vc4P1/WiGhC0dUNkx4Od7gfdSgZcTDJkGXYXnDaGvcYsca/baldXNsLkkTb2hEFh3C6BuJRP6rOfJRPuPtUSL1d5AQmzGYU2dzihN1sbnOXAGWYVVnnJu8UVj6i8aYv2zXzAYglMeJBNYQyEpRqdqJZqpunziZTvWhZL+ZfpziLPHA2RHCIz2CgsbZF+TSVpqoCbPedmKzvQAib0MIVh3XjSMw1/09yEYQHZ6gLRqYMsT5PGke4D+qSS7cdAVwMLvgLu32g6DOhuqj3TEeG+p1v+POfMGEABkvAg8chZYfha49i3bxnBXDaDWOkO1f+odpgyYQsZYd978ootyfgcJ00kDiLFG8QsH7j8ALPweuPkrcp/+rGPj64gQAqvqde0w3B4C8xRWr16NoKAg4RYf72AxKHvhOOIOjhtHCmzp1PalQFeeIFa71F9faMsOaOw47w/bX8yqs+QCLJIYKhRbwjgds/6i+X1a64Cv5gEnvyc/xPYG0v28O+RvI25qiY/5OiDORupnKAVQdYqktl71quFiIfEGrv+EjOfibtLslFL0N/FAeAfa9tj5RwCjbiSP977m9NOATgtsXEpi/QmXk7Riirc/KUQHEG/ka8PI6pPTTw2nfySZJPbSXEnSxTmR+VYtHbHHAKId5e3NUIlIIqnDyiZDiKe7ONMDBHhG7RRqAEUYFRSMG0eM8aYywzlb48inpNyGbxgwvQd+k/bQFQOIekMikgyVjTsilhDpgb2LKWoAlWU5pgO6+BcAnhSG7OiJ9wkBYvQeLmteIDrXjr+z8/l4B5A5PHkeubdUp8te6HdZqwIUjd07Vg/jVgMoPDwcYrEYVVWmBcuqqqoQHW2+yF90dLTV/em9I8cEgJUrV0Iulwu3khI7qtN2lyGzyX3HCqPmyNd7fwbOIO5Mexh8JbkQNBQaJjtL7F9D7ofNtR0mCh9C9B06NRE+GouEKfteIxegiOHAPH1p+Qt/WRfK2RItHvmU3E+823KIztnQLvEAMOEuw+RDiRhq6ES94xlSK6S9AfjtQbIteZ7t1ggAcPmjADiSZVJ52ubuDnHwA6D4ADGe573XeQIfcytw/cdEO6PSeyNv+JSk/rY3AOe32/9eZXr9T/gwU+2CJagGyFoIjAqgw+w0gJzd0FbVZshwsScN3x48IXW4Ri+EpSUzAGL0Ux2Qrf97ax2wW//dv+Ip24kFPYVgAJ203wNoLfzVVbqqA6IL4kEzzT8v6IAspPqXZ5PnRBIS3nI1XjKDkdXLwmBuNYCkUinGjx+PnTt3Ctt0Oh127tyJtLQ0s69JS0sz2R8Atm/fLuyfmJiI6Ohok32amppw6NAhi8cEAG9vbwQGBprcXM5QagBtt+2hoUaSPfofitSPGExA5yqmxtRdIDoXALj83/Yde+F3xLhpriCp5sY/xvoCQ/rrnBdJt/XYsSSkcuZn88c7/i3wfASpSWGO5kpDCNDYpetqBs4gK7GQBGDmk+b3Gb+YhLl0auD724BvbiBGZ3B/0q3aHiKGGuoS7eumF0irIUbPZ3OI0HGnfgwZL5r3YHAcadGx7Cgw9xXgpi9JdVvqlaKF7OyBCqDt0f8ABg1QW635zt+AYxlgFBp+owZZd6Bif+8g513kheJx7gyB6as/d2wpQRdm52wszP76H0mwiBppaOviCYQPIV5Zdav9HkCaAUaNP2fQFR0QzxuSIag2sCM0icOSB4hqf0Zcb1hguBpaRLO5dwmh3R4CW758OT7++GN8+eWXyMnJwf3334/W1lYsXrwYALBo0SKsXLlS2P+hhx7Cli1b8PrrryM3NxfPPPMMjh49imXLlgEAOI7Dww8/jBdeeAG//fYbTp06hUWLFiE2Nhbz5893xylaZsBUojFpqbQu2Gur1/fmgv36H0rH+hXm2PcGEWQOmW1II7VFQDTwzz/IaqutFvjyWpJOffZXYPOjxBgYdIXhR0w9KeYKcSmbif4EPLDtv0CzmRYGJ9cTA6rfpK4V6+oqYgmpMbTsqGXPGMcB//iQFI7UKonLWyIDbv7aMdH1tBXk/sxGYOfz5P+eswn4YRGw9jLSEPHjK6wbzOXZpDr2lv+QtPrSI2RMQ2brywJYQepL6rVQQ4yKPs9tsV+sKwig7fweyYKJZwqwnAlWRw0gOzLAKDQDrdwZBhANf9khwLYXd4fA1O1koQKYeoAAQ7+ugr2WjdLK00CWfrEy5yX7tFk9hUhMymIA9oXBdDrD/Gvv/GcvjhpANXmkxIZEZtBwdqR/GglTNhR29pw2Fhu0h1Me7NKQu4SgA2IeIIdYsGABXnvtNaxatQpjxoxBdnY2tmzZIoiYi4uLUVFhEHtNmTIF69atw0cffYSUlBT8+OOP2LhxI0aOHCns89hjj+HBBx/EPffcg4kTJ6KlpQVbtmyBTGZHKKInkXgbPDTWVlt5f+rT30c4btEPnQOAI/VezF1gGoqITgdwXFfjF0aMoAl3kb+Pfkou1ud3kPc0FteNvMFQHr5jenXm+8SIAkgIhgqdKTwPZOv7CY29zbExOgORiAhXrSH2Am78nKTmi7yISJJWk7aXqBH6lTRPvECvDCS1VM7+SvRZLVXEwPj2RuCr6zoX+msoAr64huwrCyYXplu+A27dQLw6jgrBo0eR1b1W1TlN3hw8b/C42OsB4jjrOiCF3NDTy94QGGC5oa2ymRj8u14gHcZL7ahoTrUwzhBAU6gw2JmZao5QkweAJxmBfh1EypHJJByqaTefbq1uBzY9TOak5HmGzCRPgoaqaeV8azQWknC9WGqqh3IGjuqAaPbXgCmWSwfIAg1zS0cv0MG1ZKE4cEbncL0rETLBnNx/z8W43QACgGXLlqGoqAhKpRKHDh1Camqq8Nzu3bvxxRdfmOx/0003IS8vD0qlEqdPn8ZVV5n2YuE4Ds899xwqKyuhUCiwY8cODB3qgPu8JxHCYFYMIFp23ZY42Rz+kYbMDnNNTHc+S2LUA2cA8Vaq9lrCO4CkVC/6lbhm+6cRb8PVr5v+AAOigMTp5LFxk8DWOoPLlq5YTqwDSowaMpYdI3oFiY/lgoKegERK+hQ9XtC575e9XPsWOUbYYAA84BtOUtVv/YEUd5zyIJmoC/YCn2YYwkM6HbDxAWJA9ptIPFaT7yep7kNnm++dZA9U8G5PKKn+IhFBiqWWO8Cbw5oOqFafAeYfTSZ+ewkbTATomnaD1gUghs/OZ4G9r5IO49/eYNnLQRE8QAn2v78tqEDc0Xo7zsJY/9PRMOY4I33iVtPnNEpSy6z0CMkwmvW868faFRwRQlfovT+RybYXOo7iqA6I1k6yFP6imKsH1FJt8Mr1pPcHsG0A1V0gnex3/c+xpAoX4xEG0CUN7UZelkUqRHekvdEgirPWu8oaSXoDMXudqSgwfzsxRjgRaQzaHQbOAO78HbhrC3DbBiJU7gitq3P0M0Pbh32vk4t29Ggg/TlgzO1k+5bHDWGe41+R++HXWs7Q8CS8A7r+Wo4j5/nAQWDpEWB5DjD7eRKWiBtPUtgfOEgyCOXFwGcZxIDc9TzJPPPyI5WunZV6THU3NBXdGrR6cNQI+4X6gCFtnYa6jHE0A4wiEhkugjQM1lRu0Jil3Eom7fYG294toRO9Ez1A8ZPI766hwHo6s6uwpP+h0DDYuW2G36GiCdiwmHh4JT7AbT84LyvO2RgbQLaE0DT85QqPiSM6oMrTZGHDGbXesIRxPSD6/9n7Ksn0jJtguK70FLZCYKd+JAkJe18BPrnSdlJOD8EMIHcTHE8uALzWfNuKvM1ETxMxnLRN6AqjbiYXxvJjREsDkG7pm5aTx5MfcH7s2xwjrwcC+xHhdNYXpK3H4Q/Jc+lPk4vWlavIWMuyiHC79jxw/Buyz3gPElq6GrEXEUabMyTCBgF3byP/s7Y60uiQdrjPeMFgUDgDWnnZHgOIXlQdLapGL8LVuZ2fE/Q/XdB9USE0zQTb9wbRQ/WfAsx/H5i0hGw/ar7qvIDgAXLixV4WZNCpFHexc3l3oB49SyGfxGmk8KS8GDj0IdGlfTCFlNQQe5MkCEsaFU8gMployxRyUsLCGvb2V+wq1FtjywCiqevDr7P9G46fTLK85CXEQG8oNBj36U/3TN81Y2x5gIybZ1eeJFpGe8LqLoYZQJ4ADQ2ZS2sUGuLN7/rxA2OA6XqB7fZVRCj32zIyuQX1Jy0iegKJNzDtUfJ43+vA7w8R13DSNYbstoAoQ+rmrufJxV2nIS757nST7mv4hRP91ZyXyGfjHQSMvJFkpDkTWmyypcq8h9IY2qeuo6jWFhHDyb25VaHgAepCCJsa9WXHiP7t2Jfk75kryQVi7B3kIlJyyDB2czTos8Cc6QECDO1qHG1Y3F0qThLPAScyNDfuiHEG6ZbHgQ13kott8ABg0UbLKdqegtjLMF/YKgRLQ2DRrjKAjHRAyhbz+zQWG/owTn3I9jG9/Q06u0NrSTFKnZoUTrSn1Y+zMS6G2BGeNxhAC74l3yt1G9GL7n7JecVKuwAzgDwB+oXtWClZITekRCbP7957TF5KVvOt1cDbY/WZAhxwzRv21WtxFmNu05dir9brCPxJ6rUxUx4kIt6aXFJtmhN7rtbAnUj9iLF42wZgZTFw46fOX/nJggwZS7a8QEJYxUEDiHqAmko7G1m0xpQjAmgKzQSryAbeHEHE3AMuM/zeAqINzX0tlV9obwCU+l569rThcAShYbGLWvQ0VwG//YtoL17qTzRiOh3w95vk+RH/sO5puOoVIG0ZyVYNjAMmLiFVhD3Z82MMNeCsFZptriRzESfqXjsIa4QMIAtNazqggx+QKEDiNPsqqAOG7/GhtcQzBxAPujsI0KfBmzOAGotJkovIiyx0b/vJ0B5l92rDwsQNMAPIE6Bf5OozpoUCczbpw19JlmP19iKRGgwNnYZcUO78zfG0+u4ikZpmm818snNjV59g4LJHDH9PWNz982d0HRp+smYAqdoMadWOXkh8Qgx1RIwbWCrkBg9QV8ITwf1Nq1FLfDpr3WgG48n15rN0aAaYX2TXheSW6K+vS1Z9hpQ8+PNx4MNpRGS863+WvQX2svF+cnEpP04+y+xvgZ/uMqRJX7bc+utDEoCM/wGLN5P2D1e/1rOLpe5C20gUZxo0hx2h3p+wIc7//xpjTQfUVg9k6Y2AqQ/bf8xJ95Aaa4nTSTh12mP2G0/OhobA2uo6F7Sl3p/okaRoolgCzH2JNKseOIMsit1EN2tgM5yCXzgQNYrEqgv2kAJ0PA8c0RcTHHWTc95n8JXAvPeJ+3HcIhKScgcpC0m2g0hCfsTmmHQPcPxrclHqqRAdwzxhg4g+zZoBVJMLIWuNusMdITKJ1MOqyTFkI5YeJccMSeiaqJvjSOZcSzX5rkv9On/nE6cTQ6mxmHgKaOsTiiv0PxT/CBLaqz0H/LjYEKqpOEG6dPM64Mr/du3YBXtJSrVIQmpUKRqBP/5tMH6GziUXpL5MxDAgIIZoDksOGjxCxlS6WP9DSbiMZLeaM4COfEqKNkaPsp39ZYx/pKHKvrvxCSWeel5LalvR5rqAIQkhtoNxNm4RCUP3tF7JCOYB8hSoF4jqgEoOkZWbROZcXcfY24j4013GD0BWALd8C9z8peU+NFJfUoDwwSxD1VyGe6Dhp1ozWVoUQQDtYPiLQsNmxkJo2p+pXzeq84rERAPnG2r+Oy8SAcP0GTfmmvu6ogaQMR0bFk9bQcJOAElptuS5sIZx0+Dxi8mCauL/Eb0Yxd6K770ZjrMdBhME0C6umUM9QOXHTD176nYSwgKI98eNxkC3EIks64Cs1QZz8/kyA8hTGEiF0HvIBEYba46+mRQcvBTx8iGrdoZ7sScTrKqL+h8KzUaihhRgqJHjzPYE5jBO+e4oyHSlBwgwCKEBUgTziqdIAdGgeBJOoC1qHCHndxJ28PIzbVA6+X7iAb72ra7V/OqNUAPIkhBaEEC72AAKGUA8jToNyeylZH9L9DHB/buv83Q31AAyruSv1RiyMO0tjtqDMAPIUxgwhYjEGgqBr+cT/Q8ApN7nzlExGEYG0AXLLTi6KoCmROozwWiBPp3ONf2ZzDFgKima2FrduXmqqz1Ag9OJvihxOnDVq2SbWEI8NgDxDtjqE2iMRkUKPQJA2gOdw5Fjb+vZXnruhhpAFSc696kq3E8MXE7UM1WTx+pb0ex/i/xPdVpDEdi0B7vfld3dmEuFr80jkgupf8+2MLITZgB5Ct4BpHqy2JusVngtmRRdlZnAYNhLSAKJ76tbiZ7CHF2tAUShNa6aK0jxz9o8kn3l5edYVemuIJEatBcdw2Cu9gD5hQH/zgXu2Ggaohu3iIi2q0531o3otEQ0+91CchE35sjHxFPnF9Hz1YA9kYBovb6HB76aZ2gHpFGSdh4A0aH0RCf7Sf9HDIGq06SY5L7XyYLXJ9Q9LX6cjbliiDT8FTvWs/rF6WEGkCcx/k7g3j0kc0XkZWiOyWC4E4nUYACY0wG11hlWfV3tpSQLIqnWAPEC0fBX3LieWRkPnUPuaSsCgKzSaSd4V3mAAHJhEHWYin1DgTELyeOtTxh0I0UHSKbY7/8ioZQvryEFHnU6oLUW2P0y2e/KVb2janpP8I8PgYBY8r36dDZw7GvgrxeJ+NwvApj1rO1jOAOfEIP37feHgL/+Rx7PerZvhPr9zaTC0ybe7spOswEzgDyNyOHAPbtJPylPbDLIuDQJs5IKX60vIhg8oHtp0jQMVn0WKNUbQP16SKsyZDYJhVSeMngJGosBjYJsd7QJsTO4bDngG0Yq5264E9jzKvDF1cSDIAsi4TNeR0Jeb6UA31xPvGbRo92aWuxxRA4nldPDh5Lms78tA/avIc/NealnvD+UtKVkcUub4KYtI96+voC5EBjt6Rg3oefHYwfMAPJEOK57/aQYDGdjTQhd1c3wF4V6j878YsiGdLX+h+IXZsg2O/ubYRwAaTvg7CaZ9hAcT5rgSnxIyOSvF4jBk7IQ+Fc2cNuPpJaKlx+p6k4zmua+7JHhBrcSHE+MoJlPGgTPw64CRt7Qs+MIjCW1ewBSAX/Wcz37/q6kYwisvdEQGu8/2S1DskUvV10xGIweIVxvAJkLgdE6H7S3VVehrSuMW8L0lAcIAEbfROrFHP4ISL0XOPEd2Z5yS8+NoSP9JgA3fQ58fyvRB179uqleZNwichEv3E8ySMMG955KzT2NTwjJipv+GCk+6B3onjTsOauBYXNJoca+ZKhSDxD1bgl1vBK7VhusB2AGEIPBsA3N7qo4QbQxxheO4oPkPj61e++RPA9QtRCdS8VJIGFqz9aASlkI7HyOdGjf+yrRjEhk3evD5wyGzQWWHiY6EeMCcxSpHzB0Nrkx7MM31H3v7eVjKL3Ql4hKJuE9eQmp6F6inxc81PsDMAOIwWDYQ8wYQCwlqeL1Fw1NUpsrDanE3fXWiL2ISNRdadpSP1KL58DbpEcRACRd7RliYg9MIWYwTJAFkbID57eTMLKzFkYuhGmAGAyGbbxkhhAVndgAQ3PHyBGALLDnx+VsJi0hxhwlZaH7xsJg9DaSryP3Z3429ABjBhCDwej1UFd2sVH38mK9AdRTYmVXE9yfiFMBommgDTUZDIZthl1NaoZVnyUFEGVBXS+N0QMwA4jBYNgH7V5OvT7Gjz04zu8wM/4DhA4Epj/e+6vzMhg9iV+Yoe8ZQDIrO9a48iDYr5vBYNgHdWXXniNF96R+htRrD3ZzO0zUCOBfx23vx2AwOpN8HclIBID+nj0veK5pxmAwPAvfUIM7u+QQKXOvUwMBMSR0xGAwGEnXAtBnicZ7tmeYeYAYDIb9xKeS9PDig4YKuvGT3FNPhcFgeB4BUcAVTwJ1Fw1hcw+FGUAMBsN++qcBx74ETv4A6DRkm4ev8hgMRg/TS/pYMgOIwWDYDxU7t1SSe78IIOkq942HwWAwuggzgBgMhv2EJJAGjvUFwKgbST8lL5m7R8VgMBgOwwwgBoNhPxwHZPzP3aNgMBiMbsMMIAvwPA8AaGpqcvNIGAwGg8Fg2Au9btPruCWYAWSB5uZmAEB8fLybR8JgMBgMBsNRmpubERRkuZcfx9sykS5RdDodysvLERAQAM6JKb5NTU2Ij49HSUkJAgP7QO8kM7Bz7P309fMD2Dn2Bfr6+QF9/xxdcX48z6O5uRmxsbEQWalEzTxAFhCJROjXr5/Ljh8YGNgnv8zGsHPs/fT18wPYOfYF+vr5AX3/HJ19ftY8PxRWCZrBYDAYDMYlBzOAGAwGg8FgXHIwA6iH8fb2xtNPPw1vb293D8VlsHPs/fT18wPYOfYF+vr5AX3/HN15fkwEzWAwGAwG45KDeYAYDAaDwWBccjADiMFgMBgMxiUHM4AYDAaDwWBccjADiMFgMBgMxiUHM4AYDAaDwWBccjADiMFgMBgMxiUHM4AYDAaDwWBccjADiMFgMBgMxiUHM4AYDAaDwWBccrBu8BbQ6XQoLy9HQEAAOI5z93AYDAaDwWDYAc/zaG5uRmxsLEQiy34eZgBZoLy8HPHx8e4eBoPBYDAYjC5QUlKCfv36WXyeGUAWCAgIAEA+wMDAQDePhsFgMBgMhj00NTUhPj5euI5bghlAFqBhr8DAQGYAOYlGRSNW7F2B6wZdh2sHXevu4TAYDAajD2NLvuIRIuj33nsPCQkJkMlkSE1NxeHDh63uv2HDBiQlJUEmk2HUqFHYvHmzyfM///wzZs+ejbCwMHAch+zsbBeOnmEvWwq34GDFQaw9sdbdQ2EwGAzGJY7bDaD169dj+fLlePrpp3Hs2DGkpKQgIyMD1dXVZvc/cOAAFi5ciLvvvhvHjx/H/PnzMX/+fJw+fVrYp7W1FZdddhlefvnlnjoNhh3k1ucCAIqbi1GvqHfzaBgMBoNxKcPxPM+7cwCpqamYOHEi3n33XQAk+yo+Ph4PPvgg/vOf/3Taf8GCBWhtbcWmTZuEbZMnT8aYMWOwdq2pZ6GwsBCJiYk4fvw4xowZ49C4mpqaEBQUBLlczkJgTmLhpoU4XUcM1XeueAcz4me4d0AMBoPRR9FqtVCr1e4ehkvw8vKCWCy2+Ly912+3aoBUKhWysrKwcuVKYZtIJEJ6ejoyMzPNviYzMxPLly832ZaRkYGNGze6cqiMbqLRaZDfmC/8faLmBDOAGAwGw8nwPI/Kyko0Nja6eyguJTg4GNHR0d0qU+NWA6i2thZarRZRUVEm26OiopCbm2v2NZWVlWb3r6ys7NZYlEollEql8HdTU1O3jscwpVBeCKXW8PmeqDnhxtEwGAxG34QaP5GRkfD19e1zdex4nkdbW5sgk4mJienysVgWmJ7Vq1fj2Wefdfcw+iy5DcSgDZWFol5Rj9O1p6HRaSARsa8gg8FgOAOtVisYP2FhYe4ejsvw8fEBAFRXVyMyMtJqOMwabhVBh4eHQywWo6qqymR7VVUVoqOjzb4mOjraof3tZeXKlZDL5cKtpKSkW8djmJJXnwcAmDVgFvy9/NGuacf5xvNuHhWDwWD0Hajmx9fX180jcT30HLujc3KrASSVSjF+/Hjs3LlT2KbT6bBz506kpaWZfU1aWprJ/gCwfft2i/vbi7e3t1Dzh9X+cT40A2x46HCMCh8FADhRzcJgDAaD4Wz6WtjLHM44R7enwS9fvhwff/wxvvzyS+Tk5OD+++9Ha2srFi9eDABYtGiRiUj6oYcewpYtW/D6668jNzcXzzzzDI4ePYply5YJ+9TX1yM7Oxtnz54FAOTl5SE7O7vbOiFG1+B5XvAAJYUmYXTEaABMB8RgMBgM9+F2AcaCBQtQU1ODVatWobKyEmPGjMGWLVsEoXNxcbFJM7MpU6Zg3bp1eOqpp/DEE09gyJAh2LhxI0aOHCns89tvvwkGFADccsstAICnn34azzzzTM+cGEOguq0aDcoGiDkxBgUPEmoAMQOIwWAwGO7C7XWAPBVWB8h57C3di6U7l2Jw8GD8Mu8XyJVyXPb9ZQCAPQv2IFQW6uYRMhgMRu9HoVCgoKAAiYmJkMlk7h6O3axevRo///wzcnNz4ePjgylTpuDll1/GsGHDLL7G2rnae/12ewiM0ffJqcsBAAwLJV/mIO8gJAYlAgBO1px027gYDAaD4X727NmDpUuX4uDBg9i+fTvUajVmz56N1tZWl76v20NgjL5PcXMxAGBw8GBhW0pECgrkBawgIoPBYFzibNmyxeTvL774ApGRkcjKysK0adNc9r7MAGK4nNr2WgBAhE+EsC0lIgUbz29kOiAGg8FwITzPo13T3uPv6yPx6XKmllwuBwCEhrpWHsEMIIbLoQZQuE+4sC0lIgUAWEFEBoPBcCHtmnakrkvt8fc9dOsh+Ho5Xo9Ip9Ph4YcfxtSpU02Sm1wB0wAxXI45A2hQ8CChIGJ+Q76llzIYDAbjEmLp0qU4ffo0vv/+e5e/F1t2M1yKRqdBg6IBABDmYyjNLuJEGBU+CpkVmThZcxLDw4a7a4gMBoPRZ/GR+ODQrYfc8r6OsmzZMmzatAl79+5Fv379XDAqU5gBxHApDYoG8OAh4kQI8Q4xeS4lMgWZFZk4UXMCC5IWuGmEDAaD0XfhOK5LoaiehOd5PPjgg/jll1+we/duJCYm9sj7MgOI4VJq2msAAGGyMIhFpg3rRoezitAMBoNxqbN06VKsW7cOv/76KwICAoSuDUFBQULjU1fANEAMl2JO/0OhLTGKm4uF6tAMhiexrXAbMssz3T0MBqNP88EHH0Aul2PGjBmIiYkRbuvXr3fp+zIPEMOl1LXXATDV/1CCvIMwMGggLsovYn/Zflw76NqeHh6DYZFCeSH+veffEHEivD3zbUyPn+7uITEYfRJ3NaRgHiCGS7HmAQKAuYlzAQDr81xr6TMYjnKg/AAAQMfrsGLvCpytO+vmETEYDGfCDCCGS7FlAN049EZIRBKcqDkhtMxgWGdb4TY8uudRlLWUuXsofZpDFSRzxlfii3ZNO5btXIZmVbObR8VgMJwFM4AYLsWWARTuE45Z/WcBAL7Pc33dh95OSXMJnvj7CWwt3Ir7tt/HtFMuQqPT4EjlEQDAmplrEOMXg5r2GmRVZbl5ZAwGw1kwA4jhUqgBZE4DRFk4fCEAYPPFzZAr5T0yrt4Iz/NYfWg1lFolAKCwqRAP7HgAbeo2N4+s75FTl4NmdTMCpAGYFD0J46LGAQDON55388gYDIazYAYQw6XUKYgIOlxm3gMEAGMixmBYyDAotAr8ev7XnhqaWXieR159HnS8zq3jMMeu4l3YV7YPEpEEb898G8HewThTdwZrT6x199A8nh/P/Ygbf7sRFxov2LX/oUoS/poUPQlikVho5Nvbq5aXtZShqKnI3cPoEjpeh72le7G3dC9O1pyEQqNw95AYvRxmADFciq0QGEAKdd0w9AYAwPai7T0yLkt8evpT3Pj7jfjm7DduHUdHtDotXjnyCgBg8YjFmNl/Jp5MfRIA8FfJX+4cWq/gs9OfIa8hD89lPmdXxsnB8oMAgNQY0kNpSPAQAL3bA6TQKLBw00Jc/+v1yK3PdfdwHOaz059h6c6lWLpzKW7bfBvu3HKnu4fksbgrq6onccY5MgOI4TLa1G1oVbcCACJ8I6zuOzN+JgBSFJEaTT2NRqfBdznfAXC/IdaRC/ILKG8th4/EB0tGLwEATImbAjEnRmFTIcpbyt08Qs+lrKUMJc0lAIBj1cewpXCL1f0VGgWOVx8HAEyOmQyA9K4DgAJ5ATQ6jQtH6zqOVB5Bg7IBKp0Kj+19zC0dwrtKbXstPj75MQBgUNAgiDgRztadRUlTiZtH5ll4eXkBANra+n5YnJ4jPeeuwAwghsugNYB8JD7wlVgvxR7tF42RYSPBg8eekj09MbxOHCg/gOr2agDAqdpTaFG1uGUc5siuzgZAqmfTHjuB0kCMCh8FwJCyzegMzeYSc6QS+WtHX7OqmzpefRwqnQqRvpFICEwAAMT6x8JH4gO1To3i5mKXj9kV7C3dKzwukBfgtSOvuXE0jvFe9nto07RhZNhI/DzvZ4yNHAsAyKxgRSqNEYvFCA4ORnV1Nerq6tDe3g6FQtGnbu3t7airq0N1dTWCg4MhFottfzAWYIUQGS6jVqEXQMvCwHGczf1n9p+J03WnsatklxAS60k2nt8oPNbyWmRVZXlM8TvaLiQlMsVk+5S4KciuycaB8gO4ceiN7hiax3OwgoSzFiUvwraibShrKcNXZ7/CfSn3md2fGpuToicJ31sRJ8Lg4ME4VXsK5xvOY2DQwB4Zu7PgeR77yvYBAG4bfhu+zfkWP5z7AdcOuhZjIse4d3A2ONdwDj/n/wwAWDFxBUScCFNipyCrKguZ5Zm4edjNZl9HQyT2zD19iejoaABAdXW1m0fiWoKDg4Vz7SrMAGK4DHv0P8ZcEX8F3jn+Dg6WH0Sbuq1HG/jVK+oFLc3YyLE4Xn0cBysOepwBNCZijMn2KbFT8H72+zhYcRAanQYSEftJG8PzvOABurzf5RgcMhhP/v0kNhdstmgAnajVG5sRpsamYAA1nsdszLbr/XPrcxHlG4UQWYjFfRoVjZBJZJBJZHYdsytclF9EWUsZpCIp/jX2X6htr8XWwq04UH7A4w2gtSfWQsfrMGvALCEbLy0mDe8cfweHKg6Z/d4rNArcuvlWSEVSvHvlu3bPQdZQaVUoairCkJAh3T6WK+E4DjExMYiMjIRarXbZ+6zLWYffL/yOO0fciYyEjB41NL28vLrl+aGwEBjDZThqAA0KHoT+Af2h0qmwv3y/K4fWiT8u/gGNToPksGTcPvx2AAbPgbupV9QLmTu0fxplZNhIBEgD0Kxqxuna0+4YnkeT35iPekU9fCQ+SIlIwcz4mZCIJCiQF+Ci/GKn/XW8DqdqTgHo/FlTHZC9Quicuhzc/PvNWLZrmcV9CuWFmP3TbNy19S5odVp7T8thaPhrYsxE+Hr5YkLUBACe34hYrpRjd8luAMC9o+8VtieHJSNQGohmtfnv/b6yfchvyMeZujNYvGUxqlqrujUOjU6D+3fcj+t/ux5/FvzZrWP1FGKxGDKZzCU3kZcIn+R+gvOt5/Hfw//FC1kvgJfwLnu/jjdnGD8AM4AYLsSeGkDGcBwniKF3Fu902bjMQYWx/xj8DxL6AIfzjedR01bj8veuaq3C9b9djyXblphNvz9ZcxIAMDBoIIK8g0yeE4vEglCX6YA6Q70/4yLHQSqWIkAaIGR27Sre1Wn/oqYiNKma4C327rTSp5lg9qbCH6w4CB48TtacRF59ntl9Pjv9Gdo17ThVewqbCzbbfV6OQg2gaXHTABiMu1M1pzyy5ANle9F2qHVqDA0ZimGhw4Ttxt97czqgLQUGoXthUyHu2npXt+plvX38bRyuPEweH3sbap3rPCu9gQPlB9CsaoavxBciToTfLvyGD7I/cPewHIYZQAyXQUXQjrifZ8TPAABklmf22MTM8zzON5BV/aToSQiWBWN42HAArvcCyZVy3Lv9XuQ35ONgxUEh/doYqknpGJKhTI2dCoAZQOagBhA1egASagWAv4o7lw+gxuaIsBHwEplmlwwOIbWASppLhGKU1jhVe0p4vOnipk7PV7ZW4veLvwt/v5f9HlRalc3jOkqTqknIapvWjxhAQ0OGwkfig2Z1MwrkBU5/T2dBP7drBl7T6bm02DQAZK4wpk3dJhh8b854E+E+4ShuLha+C46ys2gnPj/9OQCS0FHaUorfzv/WpWN5MjzP493j75IQsY2itNQLdsPQG/BM2jMAIGjMehPMAPIgPjv9GV498qpLXeE9iaMhMIBc5H0kPqhX1PdYzZXa9lq0adog4kToF9APgCH92ZUGUIOiAQ/sfAAX5IbifOaawmbXZAOARa3GpOhJAICzdWeh1l7aK1NjtDotjlYdBWBqAFEv48nak6huMxWKUqOlY/gLACJ8IhAoDYSW16JQXmjz/akxBZAQa8ff9ddnv4ZGp0FKRAoifSJR1lKGDec22HdyDnCy5iS0vBYJgQnC91sikmBE2IhO4/QkylvKkVWVBQ6c0DTZGGoAnaw5adKjbU/pHii0CsQHxOPK/lcK4b6uZO8pNAo8k/kMACKiXzaGhDM/PPmhS4xVd5JVlYUPT36I3y78hsf3PY5ZP84yGyZu17QLesm5CXMFneT5xvN2VfJvUbV4TBFLZgB5CBUtFXgz6018dfYrIebd2+mKAeQl9sK4SCJ0PFxx2CXj6khhUyEAIM4/DlKxFIDhgknd3s6E53n8duE3XLfxOpysOYkg7yC8Pv11AMDu0t2obK0U9lXr1DhTewaAZQ9Qv4B+CPIOglqnxrmGc04fb2+lqLkIrepW+Eh8MCzEED6J8I0QDJyOvzVqDNDyAsZwHGeoCN1oPQxW3VaNqrYqiDgRAqWBqGmvMfFAyJVywdi5L+U+3DeGCLI/OvmR1fo83+Z8i5X7Vjpk6NIaUTSln0I/A0/VAdGQ4KToSYj265ztE+cfh4FBA6HltSa6HPp4TsIccByHAYEDABh+546wrWgbGpWNiPWLxcPjH8bNw25GpE8kKlor8Ev+L104K8/l25xvARDvZ5RvFNo17dh8sXNYdk/pHrRr2hHnH4eR4SMRKgsVvlvU02iJqtYqXPXzVbjjzzs8IvTKDCAPYUfxDuHx52c+d+NInAPP8yhtKQUARPlGOfTaSTHEo0HbEbgaOjHSiRIg2VYSToLK1kqnd11//8T7ePLvJ9GobMSQkCH4ZPYnmJ0wGxOiJkDH6/DjuR+FffPq86DQKhAgDUBiUKLZ43Ech5FhIwGYhl0udc7WnQUADAsZBrHIVDRJw2A7igy/uzZ1m2BAmvMAAcDAYJL+frGx88rYGPp/GBQ8SPBeGIe7Np7fiHZNO4aFDMPU2KmYP3g+ov2iUa+otxiqKWkqwStHXsGmi5scqn9DDeqORgQ1qD3VAPrj4h8AgKsHXm1xnxuGkHIZ3+d9D57n0axqxt9lfwMAMhIyABh+18VNjnuA6G/xhqE3wEvkBZlEhrtG3QUA+Cn/J4uva1G14OXDL2PDuQ12hUt7kjez3sS1v1yLT099KtQ6K2spw64Sool7YeoLeGjcQwAgbDOG6qvmJs4VMr/GR40HQAqNWuPjUx+jQdmA3PpcHK086pwT6gbMAPIQjCsPn6g5YdOS9nQqWisgV8oh4SRC9oy9pEYT78vRyqM9UnW3SE4yrIxXyL5evkgOTwYAp3YAL2spw6enPgUAPJDyANZfsx5JoUkAgAVJCwCQiZWKLKmreULUBIg4yz/XkeHEAHIkE0yj0wgVkvsiOXU5ACDouYxJH5AOgAhoaVuIs3VnoeW1iPSNNOtxAIB+/iSEVNFaYfW9hUyy8NG4btB1AIiwnwpxqQD7hqE3gOM4eIm8ML0fCSXQC3hHPj39qbBqPlN3xur7G2PJAKJG3oXGCyYhJE9ArpQLIfAr+l9hcb95g+dBJpYhvyEfx6qP4YszX0CtUyMxKBFDQ4YCQJc9QOcbzuN49XGIOTHmD54vbJ+bOBdiToyc+hyLv58vznyBb3K+wXOZzyHjxwyP8RYpNAp8c/YbFDYVYs2xNcj4KQPfnP0G3+Z8Cx2vw+SYyRgcMhjT+k2DmBMjvyHfpNp2q7oV+0qJ1mdOwhxhOy1MeazKsgFU2lxqYjT+esG9fR8BZgB5BFWtVYLBc3nc5QCIHqg3k1NPLj6DQwYLYSV7SQpNQoBXAFrULT3Ss4hOjB1DBHRV48yVyltZb0GtUyM1JhX3pdxnIrS9Mv5KhMnCUNteix1FO8DzPLYWbgVgWM1awlEDKL8hHwv/WIirfr4K72W/18Wz8WzodzA5LLnTcwMCB2BuAvHMvH3sbQBEEwQQo8USsf6xAGCz9Qj1AI0KH4VR4aMwIHAA2jXt2FG8A/WKekHXRfVIgEHM/nfZ3536HFW2VppcMM7WnrX6/sZQY62jARTuE444/zjw4M16DgvlhfjwxIdu0WvQrLk4/7hOmY/GBHkHCR6iFw6+ILTLuG/0fYJ3ghpA1W3VDrX/+DGfeH9mxM9ApG+ksD1UFooJ0URXZK5ljkanEQyeAK8A1Cnq8PSBpwWj2JXsK92HX/J/sdgnK7smGyqdCkHeQUgITECTqgkvH3kZX5/9GgCEEiBB3kGCdsrYC7S/bD9UOhX6B/QXDEwAQn2mM3VnLH5fPjz5ITQ6DeL84wCQz647mXnOgBlAHgBN+U6JSMGKiSvAgcPukt1C9k9vhK6+qXfDEcQisTDBdDVzwxFojZ0BQQNMttMJwFkeoJM1J/Fn4Z/gwOHRCY92KhzmJfYSqtp+k/MN8hryUNRUBG+xt5AdZwlqAF2UXxT6r1nil/xfsGDTAsG4XHtirRBusAWth7Jk2xKPETKaQ8frDB6g0M4eIABYOnYpxJwY+8r24YvTX+CTU58AsCw2B4wMoFbLBpBWpxUM0ZHhI8FxnJDF9NuF37CvdB90vA7DQ4ebGCWpMamQiCQoaynrJNj9/PTn0Og0gp6OhvfsgXqAYvxiOj1HvUDmhNDPZj6Ld7PfxTc5Pd8YmH437Zk/bkm6BQAR4fLgcfPQm3HVwKuE54O8gwQjyt4wmEKjwG8XSKaXuQrrsweQQpjbCrd1em5P6R5Ut1cjVBaKHTftwNyEueDB4/mDz7sswYXneXx44kM8sPMBrDqwymKDZDqfToubho3zNmJV2irhs+kf0B+X97tc2Hdmf2KcG5eLoMedGT/TZP7q598PET4R0Og0Fo1p+nm+Mu0V9A/oLywI3AkzgDwAuoqYNWAWEoMSBff8/237P5uNG51JvaLeaVlEdAKzdPGxhStFyMaodWqUNhOtUkcP0NjIseDAobi5uFO2UFd4+zjxNFw36DqLE/vNw26Gl8gLJ2tO4vWjRBg9rd80+Hn5WT12uE84YvxiwIO3enFsUbXgxUMvQq1TY3q/6VgwjITdVu1fZVc20LbCbfi77G8crDgonI8nUtpcihZ1C6QiqaDb6ciAwAFCaOP1rNfRrGpGSkSKoCsxR6wfMYCq26ot1oK5KL+INk0bfCQ+gmiaGkCHKw7jh3M/AEAno9bXyxfjI4nX0TgM1qhoFEIHqyavgogTobq92q7vpI7XoaqNFAE0F9azpAOqba8VDH9LF1NXktdAPEDGtX8skRSaJFRITw5LxmOTHuu0D/UC0cWOLf4s+BPNqmbE+sUiLSat0/NX9r8SIk6EM3VnhPmDQnVD8wbPg6+XLx6b9BgCpAHIqc8xm+XZXXiex+rDq/Fu9rvCtg9OfGDWC0QTS1JjUiEWiXHT0Jvw+/zf8eiER/HWzLdMwuxUJ5ddk4269jqodWqhvEDHsCTHcYIXyFwYbEvhFuh4HabGTsXoiNG4dtC1AOD2cgIeYQC99957SEhIgEwmQ2pqKg4ftn7R27BhA5KSkiCTyTBq1Chs3myqVOd5HqtWrUJMTAx8fHyQnp6O/Hz7ipf1NMYTzawBswAAz015DtP6TYNSq8SKPSssfpmdSV59HmZtmIV/bvmnU1b21vQX9kBTu49VHXNpandZcxk0vAY+Eh8TNzcABEgDBEOlu14gHa8TPHp3jbzL4n7hPuG4KpGsXmkK/uwE+9ou2BMG2160HQqtAolBiXjninewctJKzIifAZVOJRhcluB5Hp+c/kT4++uzX/dYpp6jnK0nRuDQkKGd6vkYc1/KfZCKSIj2ivgr8PHsj+Ev9be4f5hPGLxEXtDxOosGCP38R4SNEMTX/QL6YVzkOKEwItDZAAKAqXEkDLa/zFAJfWfxTii1SgwNGYoZ8TOEPmT2eIHqFfVQ69QQcSJE+EZ0ep4aQCdrTprMMbuKd4EH+ftUzSkho7OnEDxAIfZ5kJ+a/BRuGXYL3pr5FrzF3p2eHxCgF0LbkQrP8zy+ziEhoYVJCzsJ6AHyPaAeYuMwWFlLmfC/u3EI8RyF+4TjobFEVPzO8XdQr6i365zsZXPBZnyX+x04cHhw7IPw8/JDbn1uJ8O1WdWM03Xku2lcFiJEFoI7R9wp1LmixPjHYHjocOh4HXYW78SxqmNoUjUhVBZqNiOVZu+a06/S+ZOmzFMD6HDlYVS0WNfTuRK3G0Dr16/H8uXL8fTTT+PYsWNISUlBRkaGxUZuBw4cwMKFC3H33Xfj+PHjmD9/PubPn4/Tpw2T/iuvvIK3334ba9euxaFDh+Dn54eMjAwoFJ7nss8szwQPHsNDhwvudX+pP96e+TYWJS8CALyf/T4e3/u4S0MOn576FCqdCidrT+LZzGe7ZXDVtteiur0aHDiT9GNHGBw8GCHeIVBoFQ4JPh1FCH8FDjArMqY6oO4aQNVt1VBqlZBwEvQP7G9139uTbxce+0h8hOq9tqAGkLVMMOqGvm7QdeA4DmKRGE+mPgmATFzWKl/vLd2L/IZ8+Ep8hcymp/Y/JWSSeBLUADen/zEm2i8aa2etxbNTnsUbM96Aj8TH6v4iTmRTB0S1R/T/QaFiaACI9I006x2lBtCRyiNC9hC9wNK0bnpO9hhANPwV7hNu1hAcFjIM3mJvNKmaTETCxtlxPPgeLc2h0qqELDt7Q+jDQofhyclPWhSv09+cPfWbDlUeQn5DPnwkPvjHkH9Y3I+Gwf4s+FOYLzfkbQAPHqkxqSa/8xuH3ohhIcPQom7B7xd+N3u8rtCqbhUWLg+MeQD3jL4HtybdCqCzFyirKgs6XocBgQMsfk4dmZNIhM5vZL2BL858AQCY3m+6WaOQeoCya7JNQn1qnVrwMFKjMc4/DhOjJ8LPy89mSQlX4nYD6I033sCSJUuwePFiJCcnY+3atfD19cVnn5kXAb/11luYM2cOVqxYgeHDh+P555/HuHHj8O67xP3H8zzWrFmDp556CvPmzcPo0aPx1Vdfoby8HBs3buzBM7OPI5VHABgK71HEIjFWTFyBZ9KegYST4M/CP3HP9nscEvHZS1lLGbYVkVi2mBNj08VNgiiuK9DVW0JQQpcbmnIcJ2QWuDIjzlwKvDH0B9tdITTNFokLiLPZsDQpNAkToycCIJONvZ8hTYWndYM6UtZShqNVR8GBM6msG+0XjdHho8GDt9iChOd5fHyKCEwXDFuAZ9KeQZx/HCpaKzwim6MjjnggJ0ZPxPVDrjc7qZuDamksGUD0ItuxY/yshFmCt6mjhoIyJHgIIn0iodAqcLjiMORKuaDboB5iagDZszCwJICmeIm9hONRz5RcKRfmpXmD5gHo2TDYhcYL0PAaBHkH2X2htgUNb9vjAaJz37xB86wKsNMHpMNb7I2c+hzsLtmN2vZarMtdBwBYOGyhyb403AQYFiFdpbS5FAcrDqJJ1YS1J9aipr0G/QP6C57lRcmLzHqB6PeIetft4Y7hd2Bi9ES0qluFsKyxcN+YIcFD4O/lj1Z1q0k9srN1Z9GuaUeQd5BJRvDzU5/HXzf/JVQndwduNYBUKhWysrKQnp4ubBOJREhPT0dmpvk6F5mZmSb7A0BGRoawf0FBASorK032CQoKQmpqqsVjAoBSqURTU5PJrSegGhd6wevIDUNvwEezP0KANADHq49jxZ4VTk8N/+bsN9DyWqTFpGHFxBUASK2IrrpquyOANkaIKduoLQEA63PXI21dGqZ9Pw3X/nItvjzzpV1eLEsZYB3HcEF+oVthACq+jA+It2v/JyY9gYyEDCwds9Tu90gOSwYHDuWt5WbHSleek2I6F5ajujPjlb8xJ2pO4ETNCUhFUtyRfAd8vXxxR/IdJsf1FHieF0JgXQ3BWoNmsVgSQhc0kdYSCUEJJtsDpYG4fsj1kHASwbDoCMdxgr7i7eNvY3vRdmh4DYaGDBWORys4n607a/M7bk0ATemoA/qr5C/hPReNIF7og+UHeyxjxzj85awO49QbY0sDVCgvFHQuxp5Yc4T5hAm/gTePvYl3j7+Ldk07RoePNpu6PydxDrxEXjjXcK5L2a0bz2/Etb9ci7k/z8WSbUtw+feX46uzXwEAHp/0uJBtGywLFpIpvs/9Xng9ratmHP6yhZfYC2/OeFNYIMrEMkyOnWx2X7FIjJRI8l0ynrOp93x85HgTL3ucfxxkEpndY3EFbjWAamtrodVqERVlWigvKioKlZWVZl9TWVlpdX9678gxAWD16tUICgoSbvHx9l2oukN5SznKWsog5sTChdYcE6Mn4t0r3oW32Bt7Svdg5b6VOFJ5xO4Jied5HK08iuPVx1GvqDeZNOVKuSCw/OeIf+LWpFsxKGgQNLymy54XIf041Hr4wRbUA5RdnW21aqhKq8L7J95Hi7oFDcoGFDYV4rWjr+Fff/0LTSrrhqxxCMwcIbIQIYzXnYw0uvLsH2A9/EUZHDIYr01/rdNF1Br+Un8hjt8xg5DnecFQMXfxpQbQ0aqjaFA0dHqepuPPTpgtaEnmJs6FhJPgTN0Zm4UBnYVaq8aZujNWv/tCDSqRRGhg6kyseYDa1G2C0WHOqP7PpP9g3y37MCqic6Vpyn0p9yFQGojc+ly8dvQ1AIZwC0DCPSJORELNNoTQggfI17InpWMmGDWC0/unY0jwEMT5/3975x3mVJn98e9NT6ZkemMGGNqA9O4gAgrSXJS1orggq6JYEQviKqirC2LFFRuui/tDBSs2RAFFpA3SpCNlhgGmMb2kJ+/vj5v3TTKpk0kmM8P7eR4eIblJ3utN7j33nO/5ng4w2UxuM7fCBQ0OAhFABwr9fVcaKn16HlF37lGZo7yeE5z5e5+/I04Zh/yafHYefXjwwx4DN61Sy3RfX59sWtb0fP15LNy2EAW1BZAKUqRHpcNGbLARG8ZkjXHLotycczMECNhRvANnas+goKaADfH1drPtDa1Si+Vjl6NXQi/8vc/ffZaJqYjfWTJAs+dUTtCaiHgJrLWwYMEC1NTUsD9nz4bPII7W9mn2p3dSb79dPoNSB2HpqKWQCBKsL1iPv//4d4z5dIzXO3Zn3j/4Pmb9OAszfpiB0WtGY9r301BcXwyrzYolu5YwR9rcjFyx9JTqCDyCgWWAEpuXAeqV0AsqqQrVxmqftfsNZzag0lCJFHUKPp/yOZ4Y9gTkEjk2n92MmT/M9Kmdou/rLQMEACMyRgBwH7rYFGgJLNAMULBQIWJjzdLRyqMorCuEWqbG2I5j3V6XFZOFngk9YSVWt3IHIYS1wtJACRD9UEZ2GAnA1eU4HBTUFGDOxjm4bPVlmPbdNNz2w21e5w7RQO+ShEua7EEVCFQD5Em8SQPdOGUc4lXxbs9LJVKfImtAzCw8MuQRAGCWBld1voo9r5apWSnBXxmMZYCi/WeATlSfwMELB1mp46pOV0EQBFby+PXcrz4/K1Q0pQU+UKLkUcxCwFsrvMVmYeM3qIDZHzGKGNzT/x7279GZo5mFhyfozce6/HVNmij/1YmvQEAwKGUQfpv2G3664SdsuGED3rzyTbx4+Ytu23eI7sB+m58d/wwv5L0AQAzsElQJAX8upVNsJ3w65VPMGTDH53bOsgVCCKw2K7uR9vX/JVJENABKSkqCVCpFaWmpy+OlpaVIS/N8x5KWluZze/rfprwnACiVSsTGxrr8CTWEEHx09CNMXTsVFfoKVmcPtCZ7Zccr8eaVb+KqTlchRZ0CvUWPJ357wqfB1venv2ftyimaFAgQcKTiCG5bdxvmbZ6H705/B6kgxUODHmJ3Lc3R3tSaatkIjGBb4ClyqZzdKfsqg9HW0htybkBOQg6m95qO/5v8f0hUJeJk9Uks27vM4+sq9BW4oBdFv409gJyhKd8dRTuCFofTk64/AXRz8daJQTvKhqUN86opGtfRcxnsWOUxFDUUQSVVsWCQQrs5vj31bdhm+1htVjy25TFsPb+VaeBOVJ3AvRvvdcsE1ZvqWVnAXwkjWGgA5GlESiABdSBM7TaVHctucd3c9EQ0u0oNA71R2mBvgfeRAUrRpLCMwsObH4aVWHFl1pUsm0gzBk3xHgoWG7E1qQW+KdDsqzdH6F3Fu1CuL4dWqWXBQyDc1OMmdNV2hUqqwtxBc31uO6LDCCSoElBpqMTWc54dvxtjtVnx1UnRWHFaz2mIUcQAEHVdo7O8awRpGeyjYx9hZ/FOKCQKPDH0iQD3Kjj6JveFXCJHub4cZ+vO4s+qP1Fvrke0PDrohphwEtEASKFQYPDgwdi0ySG8tNls2LRpE3Jz3b0XACA3N9dlewDYsGED2z47OxtpaWku29TW1iIvL8/re7YUBqsBq4+txrn6c3j010dZSaUpKcnLMy/Hq2NexU83/MRa5R/4+QHsL9vPtEGEEJysOol3/3gXT297GgAw85KZ2HTjJvx4/Y/oFtcNZfoy/Hz2Z8gEGZaOWupigDUwWQyAjlQcafIcmz0lYuahU2wnnwLCQKEXAW8W68crj2Nf2T7IBJnLXVvvxN54fuTzAERTQU/ZG5qy7pPYB7EK7wHvoJRBUEqVKNOXeZyO7A9CCMsMhD0DZC+lHqs85hIc0O9aY7G9MzTLsKNoh0t2gwqjL+twmVv6e3TWaMQoYlCqK2UBfbDUm+qx4sAK/HbuN5dAc+3JtThWeQwx8hh8+pdP8eU1X0Kr1OJA+QHcu+lel8611cdXo9ZUi86xnV3KRqGEegGV6Ercgj5v+p+mIhEkeH7k8xiVOQoPD37Y7Xn6/v5GmfgTQVNoGaxUVwqZRMYyUADQPV4sI56uOd2krEUwnK87jwZzA+QSudfZd8FC/5/l1+R7fP67098BELvt5FLv1gmNkUvlWDV5Fb6/7nu3VnK3bSVy1g248vDKgG6othVtQ5muDFql1mP21huXd7gc6VHp7LpwZ987kRUb3vOPUqpkGrW9ZXuxu1Qsfw1MGRhwk0FLEvES2Lx587BixQp8+OGHOHr0KObMmYOGhgbMmjULADBjxgwsWLCAbf/QQw9h/fr1eOWVV3Ds2DE888wz2L17N+6//34Aoohw7ty5eP755/HNN9/g4MGDmDFjBjIyMjB16tRI7CJDLVNj2RXLECWPwu7S3exkQ028moJUIsXSUUuRE5+DCkMF/vbD3zDikxEY++lYDF41GH/95q94c/+bMNvMuKrTVZg3ZB4AMRX+4aQPMSJjBDQyDV4Z84qbz0xmTCYSVYkuk8gDhWYafF1omwILgLxkgGj2Z2ynsW4+JyM7jGRGf09te8rFsMxsM2PNMfG1t/a61ecaVDIVW8f2ou1N3ocKQwX0Fj0kgoQJaMNFWlQaMqIyYCVWNm7BZDWxANKXALKLtguGpw2HhVjwn0P/YY/TAMjTyVcpVbKZQE3VNThjsBhw/8/34419b+DeTfdi2vfTsO70OpTpylgG857+96BXYi90j++Ot8e+DY1Mgz2le/DXb/6Kr058hT2le/Dh4Q8BALP7zQ7bCTdZkwyZIIPFZnGzDaAX1+ZmgAAxWF4+drnHLpnMGHEmma+uJrPVzMTw/gIgZ1+X23rd5pKp7BDdAWqZGmabOaiBok2BygJ6Jfby6d8UDDQjTX8XzujMOuZK7NwhGSjRimg3HzFv/O2Sv0EpVWJv2d6Ahtp+8ad4ozaly5QmlXSlEilzsc6KyWJDXMMNvQlbn7+e/R5bY/kLaAUB0M0334yXX34ZCxcuxIABA7B//36sX7+eiZgLCwtRXOy4Gx0xYgQ+/vhjvPfee+jfvz8+//xzrF27Fn36ODw3Hn/8cTzwwAOYPXs2hg4divr6eqxfvx4qVWQV54A4TXrxyMXs332T+gbdKh4lj8LyscsxruM4xChioLfoUaYXHWoVEgVGZY7CsyOexYujXnRR38cqYvHuVe9i67StHrsVmtOCHuoAqF9yP0gECc7Xn2fpfIrRamQjHGig05h5g+ehc2xnlOnKMH3ddKZr2lS4CWX6MiSqEv3O2QKapwOiF430qPSwaFIaQ09A9Nj9ceEPGKwGJKoSmTOxN+7ufzcAUXNQpitDYW0hTlafhEyQeW1XvbabqGvYcGZDUEM1LTYLHtvyGPaU7oFGpoFapsaRiiOY/9t8jP1sLCoNlegc2xm39HS0FvdN7ouPJn+EXgm9UGOswcLtC3H7+ttRbaxGx5iOzKcoHMgkMqRGieenxp1grATWzAyQP2g5x1cGqFRXCgIChUThV/dBs9AJqgTM7jfb5TmJIGFiciqkDRdUZxSo91VToL+LAxcOuGWyfj77M/QWPTKjMz2a/IWSFE0KK08t37/cZxbogu4C+3/iy6HcGzN7z8TcQXOxfOxyjwaR4YDeLG4r2oZSXSmytdlBrb0l8G1I0kLcf//9LIPTmM2bN7s9duONN+LGG2/0+n6CIOC5557Dc889F6olhpQrOl6BhwY9hGV7l7lM1A2G1KhUvHbFa7ARG/Jr8mGwGpCgTECCOsHvF95XmndAygBsLNzYJCF0aUMpTtechgChyZ0G3ohWiLXjo5VHsatkF9OcAGLNXmfRIUWd4rXDQCPX4P3x7+OBnx/A0cqjuOPHO3BTzk0sOLgx58aAgpLcjFxgj9glZbKamhTItFT5izIwZSC+O/0dy/rQoO3SjEv9thUPSR2CQSmDsLdsL17Z/QoTGg9JG+K1pNkvqR+6arviVM0p/JD/Azu5+6PKUIWvT36NL09+ifyafCgkCrw59k10jeuKj49+jHX569gF/rGhj7l9X7vFd8NHkz/C+wffx8bCjdCZdSAgmD9svl+vpeaSHpWO8/XnUVRfxG4WCCGsqzDU5ZvG0O9SpaESDeYGj00UzlPg/R33ngk98e64d5ERncE0Js50j++OA+UH8GfVn8wczxN6ix4P/fwQFFIFXhj5QpPK4Earkd1AUcfgUNItrhtiFbGoNdXiWMUxl068tSfXAgD+0vUvIWu998Xf+/wdnx3/DAcuHMDW81tdJAjOfH/6e1iJFf2S+/ktr3lCKVXijr53NHe5TWJAygAIEEBAkBWThffHvx8SOUQ4iHgG6GLlzr53Yuu0rS53tc1BIkjQNa4reif2Rnp0erOjfdaCfmF/wMJf6jPRO7F3SL/w9OTQePIy7VYakzXGo4szJTUqFSsnrsQVWVfAZDNh1dFVOFxxGDJBhpt6BHax7h7fHYmqROgtere5Sf5gAugAW+CbCw0GD1w4ALPV7PD/SPPv/yEIAssArMtfh21Foq3/9F7Tfb7muu7XAQC+PPGl38+wERvWHFuDq7+8Gq/seQX5NflQy9R4efTLGJo2FAmqBNw/8H58/9fv8fmUz7Fq8iqv2Se5VI45A+bgi2u+wA/X/4D1169vEWM11gnW4MhOl+nKoLPoIBWkyIoOb7Abo4hBvFLsMvOWBQpU/0MZ0WGE18wV1QH5c+199493saN4B3499ytm/TirSd5Zv5f8Dr1FjxRNSlgEsxJB4rFL8lD5IeQV50EqSL36M4WaJHUSG+L66p5XPWotCSHMZLSl1hUKtEotbuhxA/om9cV/xv8n4NJgJOABUATRKrUtcrcRDL0SekEpVaLaWM2Enf7YWWQvf3kxygoWmiXben4rG7tgIzZmz0+nFvtCI9dg2RXL8Pa4tzE6czQECLgp5yaP85E8IREkrFuvqWMx6AUq3B1glGxtNrRKLQxWAz488iGbTRVoWXJExgi2r+M6jsPnUz73O41+StcpkElETyBfnUk6sw6z1s/C83nPo85ch5z4HCzKXYSfb/zZ7TgKgoCchJywlySCwVMnGP2dZMZkNklEGyw0C+QtAPI1BLWpBFICO1l1kmk+YuQxOFF1Arevvz1gQ9Vfz4qlntGZo8N2XqRlsD1ljt/wewfeAwBMzp7MtFUtwR197kCCKgEnq0/izX1vuj1/vOo4TlafhEKiCKhM35pYmLsQH1/9sU/7hdYAD4A4HpFL5awzZN3pdX62Fu9WQq3/odA2YLPNzLI+h8sP44L+AqLkUQHbCAiCgJEdRuLNsW9iz2178MSwprWEUn+kpuqiWroEJhEkLNuzbO8yNv8n0JORIAhYPnY5frnpF7x2xWsBtSPHq+KZXwztrvPEW/vfwt6yvYiSR2HBsAVY85c1uKHHDX69cVobtBPsfJ0jAApVC3yg0Iu1twCIiv59uUAHCs0Ana8/z7yJnLERG/6585+wEAvGZI3B6r+sRkZUBs7UnsHCbQv9ZpEJIcyBOZwZPGd9nI3Y8GfVn/jl7C8QIODOvneG7XM9EaeKw7MjngUAfHj4Q7cuStpUMCZrTKstIbV1eADE8Qodqve/I//zOSQTEFtkL+gvQClVYkDKgJCuQxAEdgdEHYlpIHRZxmVBCYvlUnmT7zJp+vyPC3+4DPvzBSEEZ2vtGaAWKoEBwD8u/Qfm9J+DrlrRMG9Klyl+XuGKSqZixnGBQm0I1hxfg01n3GeKHa04ilVHVwEAlo5ailt73doqW2MDoUdCDwBi+YR+F6i/TLj1PxSaUfTWmUUDo0Acjf0Rr4pHslrMlnrKAm09vxV7y/ZCLVPjyWFPomNsR7xx5RuQS+T49dyvWH18tdtrKFabFbtKdqGooQhKqbJJoxqayiUJl0AtU6PGWINT1aew4oA4325cp3HoEtfFz6tDz5isMbiu+3UgIHhqq2OwsNlmZqaMzkN0OaGFB0Acr4ztOBb9k/tDb9Fj+f7lPrel2Z/BqYPD0m1AA6BtRdtQY6xhAVAg5a9Q0S2uGxv2F+gE42pjNerMYmdUS6bXE1QJuHfAvVg7dS2237LdrbMnHORm5OL67tfDRmx4bMtjLqNDLDYLnt3xLKzEigmdJ0R0AGIo6BnfE9HyaNSZ65hxXyhb4AOBZhSd7R2coYLsUGUefemAqNB+SpcpLNOYk5DD/IRe/v1lFy8vs82Mnwt/xgObHsDwj4fjzp/E7MuwtGE+Ry00F7lUjn5JYmZ73uZ5WF+wHgBwV9+7wvaZ/nhsyGPoEN0BRQ1FWPr7UgBil2qloRIJqgSM6DDCzztwgqVVdIFxWieCIOCRIY9gxg8z8NXJr9AvuR/SotI8ipxZp1GIy1+UrnFd0S2uG05Wn8TELyai3lwPqSDF5R08d0+EA6lEiv7J/bGtaBv2lu4NyKqfXhwjOfjPU1dPOBAEAU9d+hRqjDXYWLgRd/10FwamDET3+O74pfAXlOnLECOPaXLpsTUilYjz+7ac24LfS35HtjablUabOwImUHxpgAwWA9MAhSIDBIg6oO1F2z1mgKhPV+NuzFt73opt57fht/O/YdaPszC913TIBBm+OfUNKgwVbDuVVIVucd1aJFAflDoIeSV5bK7Ww4MfDsvQ3ECJVkTjhZEvYNb6Wfjq5FfQyDX47Lg4k+yarteE3A+J44BngDg+GZgyEGM7joWN2LBo+yLcveFu/O2Hv7n4aJhtZla/DlcABABXd7kaAFjwc0vPW1q8Nk7Le4HaA1ABcp+kPn62bB/IJDIsGbUEV3W6CgQEe8v2Ys3xNWLwo4jBPy/7Z5NLa62Voami1cPu0t3Yfn479BY90qPSmz0EOFBoAFTcUAyT1eTyHA2KYuQxiFPGheTzWAaoUQDUYG5g87saD3UWBAEvjnoRU7pMgY3Y8H9H/g//PfxfVBgqkKBKwKzes/DFNV9g56078clfPgl5+dwTozPFFvv0qHSsnLgSM3vPDPtn+mNw6mDc3vt2AMBHRz+CyWbCFVlX4N4B90Z2Ye0cngHi+OWJYU9AJVOhTFeGIxVHkF+Tj8///Jy18B8qPwSdRYc4ZVzI5/c4M7P3TGRGZyJZk4xLEi8Ja6rcG6yNtmwPCCF+dUR0wnbfJO/Tv9sbSqkSr455FSUNJfip4CecqT2DERkjcHnm5S1iBNlSUK+rPaV7oJKK2T06QLQlSFQlQi1TQ2/R43z9eRftERXed4ztGLL19IgXdU/Hq47DarMy/dYfF/6AjdiQEZXhseMsRhGDf13+L0zMnohle5chIyoDU7tPxajMURHJbvRO6o11f12HJE1SRM4h3rhv4H3YXrQdx6uO45aet2D+0PltViPXVuABEMcvaVFpWHL5EgDA6mOr8ULeC3jnj3dwTddrECWPYu3vw9OH+/TjaS5yidynCVtL0CepD6SCFGW6MhQ3FLN2aE8QQnCwXBxUezEFQJS0qDTM6D0j0ssIGzkJOYiSR6HOVIefzvwEAG5jZcKJIAjIisnCn1V/4mzdWdcAKAzeU93iu7H9PVZ1zDHzya7taZz9acyozFGtRvsV7plYwaCUKvG/Sf9DYV0hcuJzWq1FSnuCl8A4TeL6HtejU2wnVBoqsfLwSgChH3/RmtHINWymkL92+FJdKcr15ZAK0ohqDDjhQSaRsYygjdiQqklt8UDX20gM5wxQqJBL5KzsR296AMfvgJqncoJHI9egZ0JPHvy0EDwA4jQJuUSOuYPmAhC9K7ae38rKPBdDAAQ4dEDeJtRTaPane3z3VpVq54QO55EvV3W6KqwZUE94E0LTDFCoBNAUanJKO/zMVjP7/XsbR8PhtFZ4AMRpMmM7jkVuei70Fj3mbJwDC7EgMzqzRdu8I8mQVHGysT9H6IMXLt7y18WCcwDUkuUvCi3lUA8iSrjMN+lNzt6yvTBajThSeQQGqwFapbbF/I84nFDBAyBOkxEEAcuuXOZSzw/1+IvWDL3TPVVzChX6Cq/bXcz6n4uFngk9MTx9OC7LuCwiIztox9kfZX/AYrMAEFvg6SDUUGeAumi7IFmdDKPViP1l+9lNwMCUgS2e/eJwmgv/xnKCQi1T4/UrXscNPW6AWqZustNwWyZOFcdagr1lgaw2Kw5XHAbAA6D2jEwiw/vj38c7V70TkQCgZ0JPaJVa1JvrmeUCNUYMZQs8RRAElgX68sSX+ODQBwACG7TL4bQ2eADECRq5RI5FuYuw89adfjtA2htUDNp4fg/lVM0p6C16aGQaXhrghA2pRMpm4dFmhDN1ogN0KFvgnaHZ3nX561BjrEHvxN64MefGkH8OhxNueADEaTYXY+p7SJqoA9pdutvtud0lu/HU1qcAiJ4j3MuDE05yM3IBONzYwz17zjnbE6+Mx2tjXgvL+BsOJ9xcfFcuDicEUB3QyeqTqDJUscf/c/A/mPXjLBytPIpoeTTu7nd3pJbIuUigJakDFw6gwdzgkgEKB6lRqRiQPAByiRwvj36Zzf7icNoa3AiRwwmCBFUCm022p3QPxnUah5KGEry1/y0AwPXdr8eDgx5EgiohwivltHeyYrKQGZ2Jc/XnsLN4Jw6Xi9qzcAVAAPDuVe+i3lyPFE1K2D6Dwwk3PAPE4QQJbYenOqA3970Jk82EIalDsCh3EQ9+OC0GLYM9tfUpHK08CplEFlZjQo1cw4MfTpuHB0AcTpBQD5hvT3+Ld/94F9+c+gYAMG/wPO7kymlRaBms3lwPuUSO18e8HnIPIA6nvcEDIA4nSC7PvByXJF6COlMd3tz/JggIxncaj77JvO2d07IMTx8OtUwtBj9XvI7RWaMjvSQOp9UjEEJIpBfRGqmtrYVWq0VNTQ1iY2MjvRxOK8VsM2PVkVV4+4+3IUDAp1M+Dbn5HIcTCKeqT0EiSLjtAueiJ9DrNw+AvMADIE5TqDHWwGQ1IVmTHOmlcDgczkVNoNdv3gXG4YQArVIb6SVwOBwOpwnwAMgLNDFWW1sb4ZVwOBwOh8MJFHrd9lfg4gGQF+rq6gAAWVm8k4LD4XA4nLZGXV0dtFrv2XmuAfKCzWZDUVERYmJiQtrSXFtbi6ysLJw9e7bdaov4PrZ92vv+AXwf2wPtff+A9r+P4dg/Qgjq6uqQkZEBicR7szvPAHlBIpEgMzMzbO8fGxvbLr/MzvB9bPu09/0D+D62B9r7/gHtfx9DvX++Mj8U7gPE4XA4HA7nooMHQBwOh8PhcC46eADUwiiVSixatAhKpTLSSwkbfB/bPu19/wC+j+2B9r5/QPvfx0juHxdBczgcDofDuejgGSAOh8PhcDgXHTwA4nA4HA6Hc9HBAyAOh8PhcDgXHTwA4nA4HA6Hc9HBAyAOh8PhcDgXHS0SAC1fvhydO3eGSqXC8OHDsWvXLp/bf/bZZ+jZsydUKhX69u2LdevWuTxPCMHChQuRnp4OtVqNcePG4cSJEy7bdO7cGYIguPxZsmRJyPeNw+FwOBxO2yPsAdCaNWswb948LFq0CHv37kX//v0xYcIElJWVedx++/btuOWWW3DHHXdg3759mDp1KqZOnYpDhw6xbZYuXYo33ngD77zzDvLy8hAVFYUJEybAYDC4vNdzzz2H4uJi9ueBBx4I675yOBwOh8NpG4TdB2j48OEYOnQo3nzzTQDikNGsrCw88MADeOKJJ9y2v/nmm9HQ0IDvvvuOPXbppZdiwIABeOedd0AIQUZGBh555BE8+uijAICamhqkpqZi5cqVmDZtGgAxAzR37lzMnTs3nLvH4XA4HA6nDRLWYagmkwl79uzBggUL2GMSiQTjxo3Djh07PL5mx44dmDdvnstjEyZMwNq1awEA+fn5KCkpwbhx49jzWq0Ww4cPx44dO1gABABLlizBP//5T3Ts2BG33norHn74YchknnfZaDTCaDSyf9tsNlRWViIxMTGk0+A5HA6Hw+GEj1YxDb68vBxWqxWpqakuj6empuLYsWMeX1NSUuJx+5KSEvY8fczbNgDw4IMPYtCgQUhISMD27duxYMECFBcX49VXX/X4uYsXL8azzz7btB3kcDgcDofTKjl79iwyMzO9Ph/WACiSOGeR+vXrB4VCgbvvvhuLFy/2OHNkwYIFLq+pqalBx44dcfbsWcTGxrbImjkcDofD4TSP2tpaZGVlISYmxud2YQ2AkpKSIJVKUVpa6vJ4aWkp0tLSPL4mLS3N5/b0v6WlpUhPT3fZZsCAAV7XMnz4cFgsFhQUFCAnJ8fteaVS6TEwio2N5QEQh8MJCeayMuj37kPMhPG8tM7hhBl/v7GwdoEpFAoMHjwYmzZtYo/ZbDZs2rQJubm5Hl+Tm5vrsj0AbNiwgW2fnZ2NtLQ0l21qa2uRl5fn9T0BYP/+/ZBIJEhJSWnOLnE4HE7QFD/xBM7PnYv6zZsjvRQO56In7CWwefPmYebMmRgyZAiGDRuG119/HQ0NDZg1axYAYMaMGejQoQMWL14MAHjooYcwevRovPLKK7j66quxevVq7N69G++99x4AMaKbO3cunn/+eXTv3h3Z2dl4+umnkZGRgalTpwIQhdR5eXm44oorEBMTgx07duDhhx/Gbbfdhvj4+HDvMofD4bhhqapCw848AIDxzxOIueKKCK+Iw7m4CXsAdPPNN+PChQtYuHAhSkpKMGDAAKxfv56JmAsLC11U2iNGjMDHH3+Mp556Ck8++SS6d++OtWvXok+fPmybxx9/HA0NDZg9ezaqq6sxcuRIrF+/HiqVCoBYzlq9ejWeeeYZGI1GZGdn4+GHH3brLuNwOJyWomHLFsBmAwCYz5+P8Go4HE7YfYDaKrW1tdBqtaipqeEaIA6H02zOPfgQ6n76CQAQddll6Pif9yO8Ik57xmq1wmw2R3oZYUEul0MqlXp9PtDrd7vtAuNwOJzWgs1oRP3WrezfPAPECReEEJSUlKC6ujrSSwkrcXFxSEtLa1YzAQ+AOJwWwlJVBfO581D16gnBiyEnp32i27ULRKeDoFKBGAwwFxWB2GwQfJi0cTjBQIOflJQUaDSadtdtSAiBTqdj47Scu8GbCj8LczgtxLl774N+3z5I4+MRM2E8UubOhTQuLtLLuugpmv8EDEcOo/Onn0KiVoflM+p+/hkAEPuXq1Hz1VoQkwmW8nLIeVcqJ4RYzGZUVVUhNTUViYmJkV5O2FDbf6dlZWVISUnxWQ7zBb/94HBaAEIIDEePAgCsVVWoXr0GVZ99FuFVcWwGA2q+/RbGEydhOHw4LJ9BCEH9L5sBALHjx0NubwDhZTBOMNj0ehiOHEFj+a6lqgr598yBpaQEqnaW9fGERqMBgGbpnHgAxOG0AESnAzEYAABxN94AADCdOh3JJXEAGE+eYp1ZpjNnwvIZtvp6WOxjejSDB0OekQEAMBcVheXzOO0Xy4ULyL/xRuRfdz3KXlzKgiBrfQPO3jUbptOnAUJgKS6GzWm2ZXskFKU9XgLjcFoAS2UlAEBQqRA1YgSqP/scpoKCyC6KA+Nxx0zCphwPS0UFzs65F7baWkhiYhA9ZjSS77vP87Z2rYIkJgaSqCjIO3QAdu+G+TwPgDiBYy4tQ+Htt8OUnw8AqFy5EoJcBvWAAah4/z8wHDoESc+esMnkIFYrTAUFUHbtyvWGPuD/ZzgcJ2wGAxp27kT0ZZdBkMtD9r6W8nIAgCwhAYrOnQE07YLbGjGXlkJ/4ABixo1rs0JLw/Hj7O+mgsAzQPW//QbDgQOO9zl4EAnTp3vUdFnso31kqaLeR96hAwBeAuM0jXP33gtTfj5kGemIm/pXlL/1FipWOKwUJFFRSH3uORRJJRAkEhCzGdbqasiSklp8rabCQtgMBii7dGnVARgvgXE4TlSuXIlz98xByXP/DOn7Wu0ZIGliIhQdO4qPVVfDUlUV0s9pKYjZjMLbZ+H8Aw+iYeu2SC8naIzH/2R/b0oJzFIqZnWiLr8c0mTxAmO035k3xmzPAFHBMw+AOE3FWl3NNGqdPvwQyQ8+gNQFTwCCAFlGOuJn/A2d16yGqkd3CBIJJPYhoJEogxGLBdbaWhCTCdbqmoBf9/bbb6Nfv35s/mZubi5++OGHMK6UB0Acjgv6AwcBANWffw7DsWN+tg4cS0UFAECWmAhJVBRkVAgbJt1JuKn69FOWitft2R3h1QQHIQRG5wxQYSGIXQ/kD1rWUvXqBWW3buLr8ws8b2sPlmQp4jHnARCnqZjOngUAyFJSoMjKAgAkzJyJHr/vQrdNm5D25JPsewgAgkIBACAmU4uv1abTsb9bqyrdxNreyMzMxJIlS7Bnzx7s3r0bV155Ja699locDlNzAsADoDaPzWDAhTfegMHpTrYl0O3di9r1P7boZ7YE9KIOQlD64osB/3j94cgAJQAAK4MZ22AZzFpfj/Llb7F/Gw4fieBqgsdSdgHW6mpAKgWkUhCDgQU2/l9Lg5oUKLOzATh9d3xsCwDyTHsAVFQUsu8Xp31jOlMIACx7TJFGR3ssP9PyfaQDIJvRCKLXB/S6KVOmYPLkyejevTt69OiBF154AdHR0di5c2e4lso1QG2d6i++QPlbb6Pm2+/Q9fvvWOQfThrydqHwjjsAiwXKH9axC0Bbh5jN7E4LUil0O3ai/tdfETNmTLPf21JuzwAliN4cis6docvLa5M6oIoV78NaWQlJdDRs9fUwHD4MQkib0wEZ/xSzP4rsziBmM8xnCmEqOAN5Wprf1zqCmmTAHsSYCrwFQI00QKmpgEQCYjTCWl4OWXJys/eF074xFYqZYnmnjn62FJHQAMhshs1qBVqwFGaprITNYIBEowFsNliqqqCwt6wHitVqxWeffYaGhgbk5uaGaaU8AGrz6PfsBQCYz55F1eefI+HWW8P6eabCQpx/8EHAYgEg3v23lwDIdPYcYLFA0GgQP20aKj/4ABUr3g9JAGStFAOgxhmgpghvWwM2oxGV//sfACDt2WdQ9Ph8WCsrYSkuZu3dbQVa4lT1yIG1vk4MgM6cQdSlw/2+1nzBrutJTYVEEwUAMJ72rQGiGSBBLocsLRWWomKYz5/nARDHL2aWAeoU2AukUggSKYjNClttLU7kjgjj6jzTddNGWKuqYKupAUlLgxCAWeHBgweRm5sLg8GA6OhofPXVV7jkkkvCtkZeAmvj6PbvY38vf+ttl/RjqLGZTDg7515YaxzCNuOfLVt6Cyf0Dl7ZuTPirr8OAGA4ejRgXYgvLBViCUyWKApmFZ3FE1m4vGfChamgAESvhyQ2FrGTJ0PZvTsAQB/GOn2gEJMJDTt2gARojEYF0MqePaHo1BlAYMeD2GywXLB39aWkQJltf21hIYj9xsAZqgGiBogAoMgQy2AmrgNqFjXffY+yV18LyW+0NWMqtAdAAWaABEGIqA4IAKRxcRAUChCbDda6uoBek5OTg/379yMvLw9z5szBzJkzceRI+ErsPAPUhjGXlMBSVAxIpZCnpcF8/jwq//c/JN1zT1g+T793L0ynTkEaF4e4m29Gxbvvtq8A6LRoTKjIzoaiY0cIcjmITgdzUTEUdt1GsNAMkMwtA1TQpspHxpMnAUD0FxEEqHpfAuOxYzAcPozYq66K6NqqVq9G6b8WI376dKQ9/ZTf7akAWpXTAyZ7ij6QAMhaVQXYgyxZUpJ4t01nfJ0/D0Unx126GCxdELd1GnvBvYCaj62hAcX/+AeI0YiokZchatiwSC8pbLAAqGNgARAACEoFYNADggQ5e/eEa2kuWCoqYC4thTQmBhKNBtKYGFgqKpgJrD8UCgW62cXcgwcPxu+//45ly5bh3XffDct6eQaoDaPfJ2Z/VDk5SH74YQBAxX9XerwLDQXGP08AANSDByN65GX2x9pPAETbmBXZ2RDkcijspT3jyRPNfm+qAZLa5/MoMjNF4a1eH7DwtjVA3auV3boCANR9+gAADIcinwFiHXyffsrKTt6wmUzseCtzchwB6ZkCv59Dj5c0MRGCXA5BInGI2hsJoa0VFYDVKrYrO/mxyDvY3aB5Biho6n/7DcSubTEcPBjh1YQPa329+D0CIG9KAEQzQGYzJBpNi/wBASQqFaQJCa5ZqCA1SDabDcYw6pd4ANSG0dkDIPXAgYidNFEUpdbUuJi7hRLjCXvJoEd3KHv0ACB2sgSa3mztmOwaDmUXMfChbaXGE80LgIjFInYbQTRCBEQdiCIzU/xcL+3TrRHjqVMAAEUXMQBS9e4NAEwIHUnoXTIxm1G58kPf2xYUABYLJLGxkKWmspKkufCs33JK464uQBRSA47vEIUGYtKkRBdDOPpaapDJaTp1P/3E/k6D32Ax/PknKj/+OODyaUtitn+vpYmJkEZHB/w6RwmsZQTQhBDYdA0AIAZDAASlCkBgfkQLFizAli1bUFBQgIMHD2LBggXYvHkzpk+fHrY18wCoDaPftx+AGAAJUinUgwaKj+8JT7rTYM/2qHr0gFSrZV42/gIE05kzKF28mHnhtFZMThkgQAz0AMBkL/sEi7W6WuwUEgRI4+PZ423REdp0WgyAaAZImZMDyGSwVlXBUlwcyaWxCwUAVK9ezYJOTzBdTkYGBEGAPD1dLHmaTH73o7GxIQCvrfDsc1JSXR6nmUArD4CCwmYwoG7zr+zf+oMHfGztn+J/PIXS5/6JipUrm7my0MPKX3b/n0ARFEoALacBImazWH0QBEjs09olSocOyd+NRVlZGWbMmIGcnByMHTsWv//+O3788UdcFcbSOg+A2ig2vZ5NF9cMHCD+d/AQAIBud+gDIGKzwXTCrv+wC19pFshfGezCm8tR+eH/cOH110O+rkC48MYbODVxEvJvuBFn59zLTijOWKqq2AWTajhoBsjQzAwQFUBL4+JcsgBtLQAiFguM9q41ZZcuAACJUukQQh86FLbPNhcVoezV17w6Z1trahzHLzsbNp0OlR995PX9mC7HXpYSpFLI7RcYf95MnjNAXgIgD9sCDjF8a78paK00bNsGotOJHXSCAEtRcdDZNGtNDQz2727FivddmjxaA8wDKEABNIUFH2Zzi4jEbQ327I9KBUFiDy1kMggSsfvLXyD2n//8BwUFBTAajSgrK8PGjRvDGvwAPABqsxgOHQIsFshSUyGztx9rhgwGAOj27Al5OcJcVASbTieWbmiAYM+Q+AuA9Pv3AwBqf1gPW4CmWKHCePIkyt9+B6aCAhgOHUL9L7+gYsUKt+3ohUuWkc7St8zh99RpEKs16DU0boGn0LKJMb9tTIU3FZ4FzGYIGg1k6enscVVvsU01nIaI5StWoOK993Bh2TLvawMgTUpC0r33AgBqv/3O6/ux2WxOLej0e+1PCG0pcxc1K7LFgLBx8OQ1AEoSM0BtIQCyVlfjwltveW3zjwS1P4omrLGTJ0PRVfx/rw9SB6Tbs4d5Odlqaz2eHyIJ8wBqgv4HgL0VXrzEt0QWiAVATmU6QRBEMTaC1wGFEx4AtVF0zuUveweRqm9fCAoFrBUVIc8qUAG0omtX5jKqsmeADD4CIEtlJcx2c0FbfT3qNm4K6br8Uf7eewAhiBo5EqlPPgkAqP3xJ9ganRBoAKTs7PA0kmdlQVAqQYxGmM+dC3oNzAQx0XUoobJHDgDAeORo0O/dkhhP2TOA2dmOOzwAqp69xOfDKIin3+e6DRs9ivzNZx1dMtTHx1RY6FV70DgDRF8rvpfvY82GmzoHQPZsnrW8HNbaWse6GpkgUqR2Q0yi14fVuqK5WOsbUHjXbJS/8W9ceO3VSC8HgHgxr/9lMwAgZsJ4qPv2AxC8EFqXtwsAoLBnNSv/bxXMJSXNX2iIMNuD+4A9gOyIIuTQl8GIxeKmlSKEwFZfD0AcyuqyDqW4hkjMJfMHD4DaKDRlq+7fnz0mUSig6tcXQOh1QPTiRssdgF3/ATE48pZx0h9wrc3XfPVVSNflC9OZM6j97nsAQPLcuYiffitkKSmw1dai4bffXLY10hZ4+0kQEMsi9O6yOULoxi3wFFWvnoBEAsuFCzCXtv5OsMYdYBT6nWiuWNwXNCixVlRAt9t99pizTkKalARJbCxgs3m9EbCU2wMgpwwQLYGZzrqXSF1e6+wCbUcaHcUCIucymMWDXggAJFEaCCpRINpas0A2kwnnHrifBRatwesJAOo2b4atrg6ylBSoBwyAmp7zghRCN/wuBkBJ994L9eDBIEajXxF9S9JUDyBnBGVovYCI1QrjqVMwnDjhErgTo9Gh/2nk+kwDIJ4B4oQMmqZXdu3i8ni4dEAsAOrhCIAUXboAUilstbXsrrgxBnsApBkirqthxw6YW0gsW75iBWCzIWr0KKj79IYglSJ28mQAQM13ruUR2olFy1IUFb24N0MIzTRA9rt+ikStduiMDrX+Nt7GHWAUZXdxH8znzrE0eCghFovLd6Z2/Xq3bahOQt6pIwRBgLKruEaTfc2NYRmgZOcMkBgA0Ttub1AX6MZlLRoYOmdE2SDUVFcRtCAIkNmF0K21E+zCsmXQ7djJLmiWouKI6GOsNTWo/20ru8mq+eJLAID22mshSCRQ9REDIMPBg00u/VtramA8KjqCa4YNRcKMGQCA+l9+CdXym4VNr2fn1qZ4AFFoJ5jNGJoAyFpVJWZ/bDaYzpxhWR1W/tJoXLLDgKgTBHgAxAkRhBCvxljOOqBQQu/uadkLEDNOTMfipfVe/4cYAMVMngTN0KEAIaj55tuQrs0TlspK1Kz9GgBcjCFj//IXAED9z7/AWi/+aAkhTFBOL5wUBWuFb0YA5CUDBACqvqKPTjgFxKGCBhONM0CyhARI7aUko5eAozmYS0pELx07nspgJlYCE8sELHN30vN6rBfcNUAsA3TunNcLKbFYYLWXNOWNghqV3bLf4ORc600DBIit8QCYx0tro97eZZX27DNszEm4LDa8QQjBublzcfauu1Dx7rswl5ah3p691V73VwCikaUgl8NaU8PK7YFC9T+K7GzIU1IQddkIQC6H6cyZVtGcQGcTSrRaSOPiAn4d/f6GshWe2GwsWBekUhCrFaaCM7CZzY4AKMq9TZ+VwEymkGpTQ/FePABqg1jKLogTdqVS0VHWCfXAgYBEAvPZszB7yco0FeJsGudUAgN864CIzcaEiep+/aGdei0AV/+OcKHLyxOHtebkQDNwoGO9vS+BonNnEKMR9Zs2AhDnqFmKiwG5HOp+/VzeJxReQNZGJojOMCPBg607ACI2m8cyIYVmgcJRBqMXNXnHjpDGxXksgzlmJYlBjLKrfT3+MkAu5oQdAEEA0em8BiWWigpRMCuVQprQqKTZKACymUyiazQ8B0B0MC7NELYmrLW1LOCNuuwyKHv2BAAYj7VsANSwfTt0O8Rp4OXL3xJF8DYb1IMHM+sBQaGA8hJRh9bUMpguLw8AoLG7SEujo6EZLN5E1m/5zevrWgpqlKlodJ73htyuz9TZy1Oh1ABZq6tBLBaxEaZbN3HMhdkEU34BC4Ck0VFurxPkckAQAJstpD5LdB/pPgcDD4DaINStVt6hAxMkU6TR0VD2FLU51Cm6uRjzC0TTuJgYl+4fwLkV3v3CZyo4A1ttLQSlEqqcHoi69FIA4hDKcAs/G3aJdX3NcFd7fEEQEDtFzALV2PVBDTvFE6y6fz+3+rWyu7h/pvz8oH+8lko6B8w9AFIxJ+VDETcS9IW5qAjEYBBPfh78SBw6oOZ5JnmC3gUrOndCzFXjALiWwWw6HQtoaEaUloapb5EztoYG9v2TJjkyQBKFArJ0cRK8yUsZjGV0kpPdUv3KXnYx+PE/QSwW1i0mKBQe794dnWCtrwRGAwl5x46QJSRAZT+nGI4fa7E1EEJw4bXXAUAcNWI2o+ZLsfwVd911LttSIb6n4+2Lhl2/AwCinM4T0aNGAQDqt2wJat2hhAnuG513vSGVShEXF4eysjJUVFTAaLPCaLPBYDRCr9PBYDAE9Uev06GhtBRGmw2W2FiYrFbY0tJgkkphMOhhMJthAmAUBLfXGk0mmKRSGG026Ovqgl4DW4tej4qKCpSVlSEuLg7SAIaseoPPAmuDUP2P88whZzQDBsB45Cj0+/YhduLEZn8evatXdu/uNrPKlxeQ/sAfAMQ7Y0EuhywjA7KUFFjKyqA/dCiss3t0v4snNs3QoW7Paa++GuX/fhMN27fDUlEB3U7xLjBqmPsUcHlGOiSxsbDV1sJw/E+o+/Ru8lpoNqFxxgCwC8nlclirq8U5UnZ36NYG6wLs3NnFy4gSTiG0+Zz9LjgzC1GXj0T1Z5+7ZIBMdoG0c5mAljKNBWfEu1anNbM0vkbjdseqyOooTmk/WwgMGojGeOoAY6/t1AkSjQY2nQ6m/HyH+3dKisdZbw4zxNZXAtP/sR+Ao8lCmdPyGaC6DRtgOHQIgkaDTh+uxJmZt4PodBA0GsROnOCyLS3RmYsC1xfa9HoYj4kBndquUQSA6NGjULZ0KXS7dsGm07ndFLUktBtNnur+ffNGWpoYxJfZg3VzeTlACGSA2w1zoNgMBlgrKwGJBDKZDIL9u00sFjGDabNCUKog89Z0UFkJYjBAajS6tMk3h7i4OLavwcIDoDaImXUFeA6A1AMHourjT6Cz++9QCCGoeG8FBKUCibffHvDn0VlYtBzkDGvlPn0axGx2+YFRATQtKwmCAPWAAaj76Sfo9/8RtgDIUlkJk137oXE6sVEUnTtD1bcvDAcPovaH9Y5s0aXuAZAgkUA9oD8atvwG/d69QQVAvjJAEoUCqh49YDh8GIZDh1ptAKS3D1NU9e/n8XlVWAMgewksM5OVXE1nCllgQ31SnDNTsvR0CBoNiE4HU+FZNt4EcPIASnK1JQAAeVYmkJfHgiq3tXjoAKMIEgmUl/SCfvceGI4cYb45ql49Pb5XJM0QrfUNqPtxPeq3boXxzxNIefQRxFxxBXte/4d480J/uzQDZDxxwi2gDAeEEJT/+00AQMLMGVD37YvUxx9DyTPPIu76691areUZYoakKQ0WltJSgBAIarWrH1SXLpB36ADz+fNo2JmHmCuv8PEu4cVSQm0UAr/QC4KA9PR0pKSkwGw249yrr8F48iRSFy1kWfimUv7ue6hZuxYxEyci5aEHXZ4zxcSg5osvETt5EitLNqbi1y2oXrMGMVdPRsr99we1BmfkcnmzMj8UXgJrg5jsbrzeugLUds2L4chR2Jym8OryduHCa6+hbMmLrGQQ0OfZO6Qad5wB4lBHSVQUYDa7iQapAFrtdNGkawtVec4Tut/F7ICye3fInEZPOKP9y9UAgIoVK2AtL4egVEI9YIDHbTWDBonvu29vk9dia2gQ9VrwHAABTkLogwdBzGYmzm5NNNCMmoeAEnCIxS1lZT5HUAQDDUbkWZliYKNWi983e5nK4ZPi+D0IEgk7GVP/IoqjA8w9iFFkUS8gz63wjrb2VI/PO+uA6uxlupgJnrOwVBQfLhG0bu9e5N9wI2o3bHB7rmj+fBT/4ynU/bAeplOnUPGOY9o2IQQG+tsdIGaA5FlZYkBpMrWIONh85owYTMvlSJw1CwAQP20auv70I1Iff8xte7m9RGQuKgr8M0odFgXOGTpBEBA9ejQAoH7Lrx5f21KYS+0ZoDTP3zdfSKVSqFQqqKKjICkuBvILxH8H8ce0aRMkxcWI69fX7bnYzp2R9cg8aHv18vr66A4dICkuhu333UGvwflPKIIfgAdAbRJWAuvsOQMk79BB7Moxm2Fw8u6oeO899vemeHqYqPjVQ3QvCAIrfzgLoW06HQw0vewUWNATqn7//rBpXnQ0o+Oh/EWJmTRJtNC3lzQ0gwdBYu+YaIx6oBgA6ffsbfKaafZHUCoheEmlUyF07Xff48QVV+LEyJEevW4iha2hgbk8R3n5fyqNjobMfhfeHMsAT1ATSkVWlktgQ/UetCNS3sgnhXarNW6Ft1zwngGiImqvGqAS7yUwwBEA1a77AaYzZyAoFIgeM8bjttIwZ4Cqv/wShkOHcH7uwy5BkKWyEvWbNwMAEmbNAgQB+j/+YOUW85kzsNbUQFAooLJ7fQkSCcvytUQnWP3WbQDEmw9pbCx7XNGxo8cyDiuBlZQEPPaB6blS3YOLqFGXAxBtO0KNubgYF954I6CbUIeNQvClHofDeUFQr7eUlzOJg2a4e5Y8EKhXk+HoUTcT2kjCA6A2hq8WeIogCGw+GB1DoT90GA3bt7NtDAEGQGKrY4H4eR66fwDPQmj9gYOA1QpZWho7OQHi9HBBLoe1qgpmPyMHgoXpf4Z5D4DkKSkuJS+NB/0PRd2vLyCTwVJWBvP5wO8wAYh1c4hjMDzpQACHENpSWgpreTmIwYDzD89rNQZ5uv37xbErGeluXYfOhEMHZK2vZ51Ucnt5UEH1PXZjRnZDkOX6e6B+RXQ7iq8MkNz+HqZzngMgGozJvZQqVb0ucfmMqFGXe+yMAcI/DoNNprdacX7eI6izBz2169cDVitUvXsjdf7jLCtbt0HsiqTlL1Xv3qyNGkDIOsFsOh3KXn8dF956y+sNRcPWrQCAqJGXBfSespQUQCIBzOaAfZUszKXbPQCipT9z4dmQe1tdWL4c5W+9jXNzH/Y5YocQ4tAABZEBolBrCLOHGYiB0GDXSCp79YLMg44xEGgHJzGZmO6qNcADoDaGpawMxGDw2ALvjHqAeFLT2UtNdL4NFfQZDgUWAJnPnxe1PUolSzM3hgVATneGenu5SG0PxCgShQKq3qKOprFGKRRYqqocdys+MkAAoLV7AgFg4xM8IVGr2Z29vollMDq8UxbnuRQHiP//tNdei6iRI5HxystQdOsKy4ULOP/oo82aQRYqaDbKW/aHogpDJxgNOKRxcZDaxZPOHV7EamWu6EonjyrAkQFyK4H50AApssTAxnqh3GOnoom2JWd6/u0pu3ZhvicAEOul/AU4SqK22tqQ3xUTQljmSz1gAGA2o3j+E7BUVaH2+3Xi2q4Wy8Ax48WBk9Segul/nFzmAYSkE8xw5Ajyr78BFe+8i/I3/s0yUS5rN5mYLi965MiA3leQyVggYwlQB0RtQjwJjGUJCZAmJwGEhDyjqbeX6PV79vh0nLbV14PYv4OegrRAoZUCKp1oKg07xSxYsPohwK7/tH+fqDSiNdAiAdDy5cvRuXNnqFQqDB8+HLvsX25vfPbZZ+jZsydUKhX69u2LdevWuTxPCMHChQuRnp4OtVqNcePG4USju87KykpMnz4dsbGxiIuLwx133IF6+6yStgz9EntqgXdGzTJAf0C3dx87uaXMnw/AMUrDH8z7pXNnCF7qrp6Gour2ioGXxl4+cl2bXQfkFACZCgpQ9sqrzXaJpiNAFF27etXcUGKuugrSpCRRXNvbt7iZegnp9jYtALJWVQMApF60SIBYXsh4cQk6vr8C2quvRubrr0NQq6HbsRNVa9Y06fPCga+OOmdYBiiEM8FYxsVJ4EwzkcZTp2E8fhy2+npIoqLYBbrxdqbT+S5lEU9jMChSrRYSrVZ8XSMhNDGZYKF35F5uPgSZjI2IERQKRF/hXUAriY0F7GJimikMFdaqKtG1WRCQ9f4KKHv0gLWmBsVP/kP8jQgCYidPAgDE2idu6/bsgbm0jGnoaLma0txOMP3BQyiYdos4KsS+3xdeX+ZWstLt3Qei00GamMj+XwYC0wEFeA5h5SVveq4eNOALXcnPUl7uMmz3wuuve52lSL9rUq0WErU66M9kM+6Ki5scaBNCoNtuD4BG5Aa9BsDRQEED7NZA2AOgNWvWYN68eVi0aBH27t2L/v37Y8KECaxFrzHbt2/HLbfcgjvuuAP79u3D1KlTMXXqVBxyumAvXboUb7zxBt555x3k5eUhKioKEyZMgMFJ8Dt9+nQcPnwYGzZswHfffYctW7Zg9uzZ4d7dsMM6Xrx0gFFUvXuL7dXl5SicNQsgBDHjx0N7zZQmzZ9iIyK6uOt/2GfZ77zNRUWw1teLBoj24EY9yEMAZNcE6Xb9DmttLRp25iH/pptRsWIFLrzxb79r8gXNeFEzM19IY2PR5dtvkP3lF37bQ+l+6Pc0NQASM0C+AqDGKLt1Q/J94kTz+o0bvW5HzGY07NoV1pq6zWhkglhvAmiKgpoPhlAkS4MQmpkBHC3uptOnWXZKPWiQW2eSIitLNGszGFzS/0wDlOyeAaKvAxzdZxRzcbHYNaRSMedrT6jspny+yl+AGPjSkoLFRys8sVpx7qG5KH72WY+DYD1Bsz/yjAxIo6OR9vRTABwjHjRDhkBubyGWd+ggni9sNhTccAOMJ05AkMvdfrs082YpK4PNLuwPFGKxoHjhQhCTCVEjctH1u28hiY6G8fhx1P7wg8u2Ddvs5a/LRrh5LfmCBUABlqmZpYGX7AqbdXg8dAE9vYFS9uiB6NGjQcxmlC19yeO25hLf6wsUaWKi2KhCSJOdss1nz4rCcrk8oHOqL1gG6MBFlAF69dVXcdddd2HWrFm45JJL8M4770Cj0eCDDz7wuP2yZcswceJEPPbYY+jVqxf++c9/YtCgQXjzTbElkhCC119/HU899RSuvfZa9OvXD//73/9QVFSEtWvXAgCOHj2K9evX4/3338fw4cMxcuRI/Pvf/8bq1atR1IQugdaI2Y8HEEWiVLITMTEaETViBDKWLBbnT9kvIIbD/rNAVADtrb0REMsT9Edq/PMEjCdOwlZXB0GjcbsrBxzZKdPp0zgx8nIU3nknbPYJ2vW//BLwSd4TLH3vpaOrMbL4eBeRpTc0dk8Y44kTLtO+/cECoITAAyAATDir273HpZOPva99SnfhjJk4e8edIZu0rD98GKWLl+DUxEn4c8RlOP/IIyBmM2TJyZD7+c7RYY3W8nJYQ5RtZS7QHRwBkKJjR0Amg02nY2aWnoIzQSaDym5OqHdy2vaVAQLsrfBwF0KbaDaqQwevei4ASJw1CzGTJiLl4Yd97xycx2F4160YT55E3Y8/ovqT1Sj+xz8CEvnSFnw6EkQzdChip0xhz9PyFyVm/HgAonZJotGgw2uvug1wlWq1YtYKjsxcoFSuWgXj0aOQaLXIeOklKDp3RuIdfwcAlL/xb5ffPBVAB1r+ojAhdKAlMKoB8mBpAIgjNoDQZjTpDZRmyGCk2LvZdLt2efyNW+wdYLJm6H8AsfxEGwTozLxAabBnfzT9+zfbD0ndrx8gCDAXFrLmkEgT1gDIZDJhz549GDdunOMDJRKMGzcOO7yo63fs2OGyPQBMmDCBbZ+fn4+SkhKXbbRaLYYPH8622bFjB+Li4jDE6aQ4btw4SCQS5NmtzxtjNBpRW1vr8qc1wgSfAQzGi75cdDSNueoqZL7zNvsCO7sP+8OYTzvAPAugKc6GiEz/06+fR78QeUoK0l94Acru3UWLdosFsZMnQ6rVwlpdHXSLPDGbmbapcfq+uciSkyHv2BEgxKV05w9rtV0D1IQMEGAv4aWkgJhM0Dcqu1kqK1E4cyZ0dgdr3e+/o+jx+c3WC1lranBm2i2o/PBDmAoKYK2sRP3GTQAAzdAhPi/6ACCNiWFmj8EKLhtDxchypwyQIJez7z8btjvUc3ZK1dc+KfyguB2xWmG1j57wpAECvLfCU0NGuRf9D3t9p07IfO01t7lynmBeQD4yQLQUAgA1X3+D0n8t9vu+jhsXx+825bFHIdFqIYmORsyE8S7bx159NQSVCvKOHdFp9SeIaXQOplCfKlMAmYSyZctwYswVOHvPHJTbM7spj8xjpen4v82ANCEBpjNnUPLcP8X2+2PHYLTP5Yu6LDABNKUpXkDEZmOZwMYz3Sg0A2T488+QdazSDJB60GAounRx/MY9nFNoBkjejA4wiqMTrGk6oLpN4u8/UDG6L6QxMaws3VrKYGENgMrLy2G1WpHa6AuWmpqKEqcftTMlJSU+t6f/9bdNSqO7F5lMhoSEBK+fu3jxYmi1WvYny4Pdfyiw6XSo+fa7oH9QzAPISwu8M0l3z0b2V1+iw7LXXVq8VXYzv0Ba4Wknia8SGODQATVs3840BBoPTrqUuOuvQ/Y3XyP7qy+R+c7byHjlZaaXqLNfdJuK4c8/QQwGSGJjoejcOaj38AWbaO8liPaEpbLpJTBAvGuLGjFC/Dyn7j2b0Yizd82G4fBhSOPjkfrkk4Bcjroff8SF15c16TMaQ80spVotOixbhqz33kX0uLGQJich7qabA3oPGpiYQhQA0XJGY4NIZ08qQaFgQX1j1H1dZ61ZKysBmw2QSDw6cwNg3x1DI60La8fvkNn4JUHDSmCV3gMgeiGUJicBgoCqVau86kYonua2yVNS0OXrtcheu9YtIFdkdkD3zb+g63ffugw8bgzVYvkrpVjr6lD5nw9gKSlB/ebNsOl0UA8ahLgbbmDbSKOjkLZoESAIqP70U5x/8EGcuXU6AHE2lz8NX2NkTAPkP8tvraoCzGZAELxmAhVdugBSKWw1Naxc1hxsOh2bE6cZPEjs1rW3lXs6p4QqAwQ4OsGohCIQrNXVzAYgZvwEP1sHRmsrg/EuMDsLFixATU0N+3O2ibXSQCAWC05NvhpFjz0GXZ5vIbgnDMePiy3GTt47vhDkcqh69XKro7MBnIcO+wzErNXVTJyp9BNQRF0qCuTqfvoJtd+LZQm1BwG0y/oEAapevRAzZgwEQUDMuLHie2zaFFSA6Oxe2xTtQKBE5YpdEHQ4YyCwEpiPLjCvn2cXHdY7BUCl/1osBj9xcej00UdImPE3pP/zOQBA9aefNutOlWZtlD17InbCeESPGoWsN99Ej99+89kl5wy70/TScUKsVpQ89xwq/29VQO9Hsx+yNNcORNriDognVW8eTqq+ovDScOQIiNnM2tOliQleRf3UPkH/xx+w1tWxx+lgSm8t8MHASmC+MkD2C2HM2LGstEszPN5gGaBG5qXytDSvHWzSuDiXtndPMJ8kL07ZlLpNm0BMJig6dULqgicQP306Ory01O13GTthPNKff158zYaNsOl00Fx6KTose93n+3tCni6WwCwBaIBoQCNNTPSq/5MoFMxB3BgCIbT+wAHRGiQ9nemV6O+KjuNxhmWAmjnuAXASQjchA1S3aZM4ULpnTxcn9ebAAqAwdAAHQ1gDoKSkJEilUpQ2ip5LS0u9zvBIS0vzuT39r79tGousLRYLKisrvX6uUqlEbGysy59QI8hkiB4jOoxWrVnd5NeXv7kcABA7aaLXlvRAUObkADIZrBUVLun1xtAJ8LK0NDfr+cZEXz4S6f/6l2M7QWhyGSrqsssgqFQwnzsXVN3d4KV9N1TQuzXD0aMBux0HI4KmROWKAZDxyFFYqqpQ8/XXqF6zBhAEZLz0EjspxU6aBEilsNbUMHO3YDCxier+y6veYFoDLxkg/d69qPr4E5QuXuw3S2Stb2AeLI2NB50v7N7KX4CYKZVER4MYjTCePOlogfdy1w+I2SZF586A1epihGc6Tz2AApvMHQiBjMNwvhDS7jMajHnCptczR2Rv3l3BIs8MLANEhc2xU6YgYeZMpD39lNfOubjrr0PaM4sgiY5G/Iy/oeOK95pcMgYcJTBrTY1f7x7WAu/F0JKiZJ1gzdcB6ewdqhoncblmuHhTpT940G3NDpF2CAIge8XAWFAQ8E1S7Q+ik3njuWvNgTXA7NiJ8489HlAjTjgJawCkUCgwePBgbNrkKGnYbDZs2rQJubmeW+pyc3NdtgeADRs2sO2zs7ORlpbmsk1tbS3y8vLYNrm5uaiursYe+xcOAH7++WfYbDYMD9LJMlTET5sGQLzbCdSwCxAnqNdt2AAIApLuvbdZa5CoVGyuF03JeoKWvwKN/uOu+yuyv16LmPHjkXjP3ZDGxDRtXWo1q/vX+eh+8oZ+PxVAhycAkqekQNGtK0AIGgLM4DkCoLgmf54sOZlpq4qeeAJFT/4DAJB0772IvtwhEJUolVBkdwbQvDtVZrDZKfgAiKXavdxpGuz6DthsqPjvf32+Fw3mJFFR7kNLnTJAvrrTBInEZdQIM0H00cUFAFGX252Af9vKHnMMZQ1hCSyAifAsC5YaWABkKigACIFUqw0q8PYF7cYz+RBBW6qq0LBNzFrSVnt/xE+bhh678pD25JNBD+yUxsSwQZtmHzd2gLPDsu/ykqMTLAQZIPv1SD3YEQApMjuIx9RicbPYYEFaCEpgym7dIMjlsBQVo65R150nLFVVjvLXhNAFQKqcHki88w5AEFD77bc4PWkS6n7+JWTv31TCXgKbN28eVqxYgQ8//BBHjx7FnDlz0NDQgFn2+S4zZszAggUL2PYPPfQQ1q9fj1deeQXHjh3DM888g927d+N++wA1QRAwd+5cPP/88/jmm29w8OBBzJgxAxkZGZg6dSoAoFevXpg4cSLuuusu7Nq1C9u2bcP999+PadOmIcPJlTgSqHr2FKNgiwXVX3wZ8OvKl78FQLzb9zSUtKkoAzCtMwUogHZGkZmJzDeWIeWhh4JaV8yVVwIA6rdsadLrLFVV7KJLXVzDAS316fL8l8GI1Sp6sQBBO6gyHdCvWwCrFdprr0HSvXPctguFZwkbKdGMDBArgXnRGhiOHGV/r/nyK5+ZDzaqwMNdurJrF1HQq9X6zfip+9ht+A8eZCaAjV2jG0MDzPptW0EIgU2nYzO7QloCS6BdYN67YpwvhPIO4vnL5CMAYvqfrl39CtebirMGyFs3Wt2GDWLppFcvKJuQgQpF2TrQqfAOF2jfGSBHJ1jzAiBiNkNnv0Fr3E5OHel1Tjogm04HGz13NLMNHhA7+BLvvhsAUPLCv9h5yRv1mzYBVqt4DH10AAdDyqOPovNnn0Hdvz8IIR47hVuKsAdAN998M15++WUsXLgQAwYMwP79+7F+/XomYi4sLESxk2p/xIgR+Pjjj/Hee++hf//++Pzzz7F27Vr0cRI5Pv7443jggQcwe/ZsDB06FPX19Vi/fj1UKhXb5qOPPkLPnj0xduxYTJ48GSNHjsR7TrOwIkncNFFQWv3pp347dwghKH9vhVP2x/3iFww0iDI2mpPkjNE+Ud3TDLBwQQenmk6dbpKehXYDKbp0gdRuZBcOaM2+IQAdkLW2FrDvQ7BrosJwSVQUMl5cgvQlSzxqV0IxpiBQiwVfUI2I9UK5xzIEzQAJKhWI0YjKVd61QL5GFUjUanT+5BN0Xv2J3/Ksyj6HqOb7dWjYvh2CXI6E22f6fI1m6FAICgUsRcUwnT7NMi6SmJiAbBMChbZg+xLZOnRQjgyQxYedR6CNC8EgT08HpFLRFNLLLKvadfbyV4DZn1AiD1AIbaZDbQPNAOUXNMtvy3D4sGjuqNW66TejqBDaSQdEg16JRsOyWs0lcfZdUHTtCmtFBUpf8uw9RGHlrxBmf5xR9+mNTp98jM6rV7uMSmppWkQEff/99+PMmTMwGo3Iy8tzKUNt3rwZK1eudNn+xhtvxPHjx2E0GnHo0CFMnjzZ5XlBEPDcc8+hpKQEBoMBGzduRI9GnQsJCQn4+OOPUVdXh5qaGnzwwQeIDtEXqbnETpwIiVYL8/nzaNi2zet2xGRC8RNP4MKrrwIAEu+8MyTZH8D7mAD22YQwpb6q9yUh+cxAkGdlAYIAm9MMqECgYzXCpf+haIYNAyQSmPLz2UnKG1RALomNDTqtHzV8GDqu/C+6rPse2muv9T5PrJl3qtaaGnZXqGhGB6RUq4U0Lg6Auw7IZjKxgDt5rpghrPr4E48jJwDnDJBnvY6yS3ZAd6dqeys8HSuQcPtMv/soUatZaa3+t98cHkAhzP4AgMIe0FhralwE1xRrfT0LJOWpqWx70/kirzcIRvuQWGUTMreBIshkjiyLBx2QtbaWDSOOndTyARAdyOtvKrzDBdp3BkiWmip6H1ksfoXnvmjY5ZhP2DjTxbSFR46w0TlM/5OWFrIsnkShYA0TNZ9/4bUL2FpTw7rSGtslhBJBImHnrUjBu8AigESlQtzUawGIvh7euPDWW+LzUilSFz6NlEfmhWwNzE331GmPWSjqAyM4ze5qCSRKJWR2oXqgnhXG/HxUrfoIgGgwFk6ksbHs/4e/SdHN0f84E3XppYHfqZ7OD8oUkZr+yZKTm2145vAccQ2AjH+eACwWSLVaJPztb5ClpMBWU+NiUuhMoHfp/pClpjLRszQpCYl33xPQ65x1QA79T+gE0IBd30S9kzzoamj2R6LVQqLRsFZvotN5FeKbTtEW+PBkbpkOyEMnmPHUKcBmgyw9PaRaqUBhnWB+vIBYgOFlDAZFEARmC9AcHZCvcTLy1FQxg2uzof7XXwEgJENQPaEZNIiZYFZ98onHbep//VUsYXbvFvLyV2uDB0ARggrL6rdu9eh8bC4qQuV/VwIAMpa+iIRbbw3p58vpmACj0aOgkprvqfr19dpiHC7oBTQQMz1rfT3O3Xc/bHV1UA8eDK2T2224oN1ZdT9t8LldIINQQ4UsNVWcYWW1sjEITYFqduTNEEBTvHWCGY6KgnvlJb0gSKUskPTW8cfu0pN936X7QxAERNldhVMfe9TneApnokfZA6Dt21H9+ecAXB2pQwXNKnkSFjvM8MQLoUSpZMEcDcqcIRaLOGsLCFm22H293jvBWOOEXZTf0jCNlJ82fUeHlf/vFjNEDLITjFgsTACtGTbM4zYxV4ql7nq7INhsX38oOsAaEz9dvJbUfve9Ry1Q3QaxASXaixlme4IHQBFC3b8/pFqteAfswROh7JVXQYxGaIYNQ2yjEmAoEKRSx1DJk+4XTB21bB8U3oyKJ5iZnh/bdmI24/wjj8B0+jRkaWnIXPa6Xx+TUKC1Z+/qf/lF7LjxQnNa4JuKIAhQNeNETYNNf+LgQHBkgApcHqcOv6peYknV2T3cE6wEFgIRaNrTTyH7m6+hvfbagF+j7NoV8bfdJk4Et9/9h7oEBji6ysweLtqezPBYJ5iHMo+p8CyI2QxBrfbadt5cmBfQOQ8BUBCNE6GEmljSINATNoOBXfgDyS4qc5qXATIcOQKbTgeJVsu+842JvlL0QKvfulU0y/36awCBj/RpCuqBA6Hs0QPEYGCfQ7Hp9aj/7TcA8OoG3p7gAVCEEKRSRI0WR1XU/eLaBqjfv180ExQEpD4xP+SdHBRaBvOkA9Izy3bvbs7hQuHHSwYQO6yK5s9Hw69bICiVyPz3v/22NocKZZcu4qwuQlDx4Ydet2OT4IPsAGvyumgZ7NixJr+WeQCFIANEW+HNjQJY2gFG53NR93DvGSDfs5qagkSj8elw7I3UJxe4zM0KpQeQ4z3tAZDHDJC9FOKUCfDVCk9/y8ouXcJiBgo4ZYAK3QMgYxgF2IFADVutVVUsA9sYKt4WVCo228wX7MYiSH0d1URphgzxekxUvS+BLDUVRKdDyQsvwHzuHKRxceLw6hAjCALibxHtWKpWr3HRkjVs2wZiMECekQHVJS2n/YwUPACKIDH2gZf1m39ljxFCUPryywAA7V//GtYvIRVCm066BkCWigqW2dAMbPkASO5nnAIhBMULF4rdJnI5Mt9YxkYetBQJdhuHmq/Wej3RUhF0czVAgUIFhcGcqJkHUDNa4CksgHXScBGrlbXo0yG9DiuGE24t1YQQmO0XquZqgJqDIJEgY8lixP7lL1BkZ7uY2IUKNnzVQ0bFQieCe8oAeQiA6G+Z/rbDgdyHFxArv4XYgDFQJFFRDg1hfoHHbZzLX4HcXCq7dQMEAdYL5QEP8SSEwFxcDJtOhwYaAPkw7BQEAdH2MliN3R4l7pZpkKjVAX1eU4mdcg0kGg1Mp0+7TCWg5a+Yq8aF7ca7NcEDoAgSNXIkIJXCdOoUuwDp8vKg370HgkKB5IceDOvnK2gGqFEJjA4jVXbvHtaWcq/rYhkEb14yR8SThESCDi+/jOjRo1tyeQDEbg7VJZeAGAxexYTBDkINFmWOoxW+qSMxHB5AwbfAU2gJzHLhAutgMp05A6LXQ1CrWZlCmZ0NyOWw6XRuvi3W6mpxVhP8mxaGG0EuR4eXX0LXH9aFtAWewkpgHjQ95lIqhg0wA0StK8Kk/wEcXYLW8nKXDj5iNrMhqS1pndEYatzqrQxGh6UGOmRUEhXF/I/8OdQbjhxB4ezZ+PPSXJy84kocHzSYmWlGedH/UGLsZTBA/M6FWvfpjDQ6CrHXXgMAqFotTiWwmUyo27xZXMtFUP4CeAAUUaSxscwUq37zryCE4MKbbwIA4m66Kex3vspu9jvw06dd7sCp/kcdhrvdQKBdJtaaGo+dLtS9OurS4YgNY5umLwRBQMLf/w4AqP5ktcdOOksLaoAAQNm9GyCRiOn/Ms8eLZ6wNTTAanclp/qO5iDVaiG1D7I0njgBwKn8lZPDfIwEuZx1mTS+sLjMamphEX5Lw8wFz51zy4SxDJBzCYyZ/blrgIw0A9Q1fAGQNDZWFNzDdSq86exZwGKBoNGERLcVLIrONADy3LZuDsLSQBWgDujCsjfQsOU30cSQ+nURAllGOitRe0MzfBjztIq9ZorPcS2hIH7aLQBE131zWRlqvloLW00NZGlpETv3tzQ8AIow0fYyWOWqVahY8T7L/iTedWfYP1vRMQuQy0GcZgcBgG6vvWNhcGR+BBKNhvlzeCqD0YuqsntkPSRix18FaVwcLBcueDRGZBqgFgqAJCoVm5FlOOy5tdwT9CImjY8PWYZD1Ufs8KIt7voDogtuY0sFb0JoXy7Q7Q15WpqTuaDrSAxP4xCcM0DOmT6XDrDu4QuAnN/fWW/GBrBmZ0e0fMKaO7yUwByeToHruQKZCUYsFuh27wYAdHhjGXru24vuO7aj08cfI3vNGq/DdykShQIJM2dCnpWFpNmzA15bsKhyeoiBjsWC6jWfouL99wEAibNu97vW9gIPgCJM7ORJkGq1MBcWMsPDlsj+AKKpGRUN0jtH4+nTMBwSDbLUEegAo/jqBDP+aQ+AenR3e64lERQK5nZb883Xbs+3ZBcYhU4/pyaWgRCKIaiNYeMnDtkDoH37xccbacq8BUCBDqtsDwgyGStxmZ10QC7jEJxLYHazP1tDA3tefO05EJMJgkoVtg4wCtUmOs8SdAigI6P/odC5eF5LYEHMdAukE8xw9BhsDQ2QxMQgZuxYCAoFZPHx0AwaGHA2J/nBB9Btw0/NcmNvClQMXf7uuzCfPQtpXBzibryxRT67NcADoAgjT0tDl+++RcLMGRAUCki02hbJ/lDonZzB7gpaumQJYLUi+oorQm761hQcXjLuOiCWAQqiqyfUaK8R6+h1Gza6jX6gAVBLaYAAQG0f+2A4cDDg19CLrrwZDtCNYQNIDx2CTa9nIzA0Awe4bMc6wU5cvBkgwLUMRmHjEKKiIHVysZeoVJAmi7oo55lgrPwVxg4wCguADjsCIBpwKCLkAUShAmxTYSGIXUfmTHAlMHuH5cmTHn3bAKdur8GD20wGJWbCBPEGzb5P8X+7rdlGqG0JHgC1AmTJyUhdsADdt/yKrj+sa9GuF/VAscxV/s67KF3yIhq2/AbI5Uid/3iLrcETik6dAbibIVoqKsShlILA2vgjiap/fyg6dQLR610m2NtMJhYQtWwGSAyA9IcOBSyEpiUw2t0TCtT22X2m06dFW32LBbLkZMgazf1R0U6w/AIQp1lLgU7rbi/QcoxzZ5XzOITGKDLchdBUAB3u8hfgFAAdPcp0S6wEFuEMkCw1FYJaLY6vaNSpRiwWh7VAE0wt5VlZENRqEKPRa3cqC4D8iJ1bExKFAnE3XC/+XaNBwvTpEV5Ry8IDoFaENC4u6KnhwRJ/yzSxjGM2o9I+ky3htttYp06k8FYCo9kfeVZWq7hTEQQBsXavDuexJmyOmVQKSUxMi61H1aMHBIUCttpar110jXGUBEKXAZIlJYljGwhB1ccfAxDLX421IbKMDHHYo8XiotnwNwesvaFg7spOGaAi2q3kngWjJS7nVm+aAVKEUQBNUXbpAkGphK2hAebCQhBCYGQZoMgGQIJE4rUMZi4pAaxWsTyVHHh3oSCRMNsGms10hlit0Plxe26tJMyciaiRI5H6jyfZHL+LBR4AXeQIMhkyli5lbtPShISQTZxvDt7MEKlWpPFE5UhCy2ANO3cykzWm/4mLC3s5whlBLmdGg/qDgZXB6EiDULscq+1C6IYtorNsY/0PIAaQzA/ISQdkCdEcsLaCJzNEeqFVdHHPdKrtM+9qvvmGZfrooNlwjcBwRpDJoOxpFwYfOQJrRQVstbWAIEDRuWX0K75QdvbcCs++6x06NPl3SQct6+yDTZ0xHD0GW309JNHRUPXqGcySI4YsKQkd31+BuOuvj/RSWhweAHHsQdCLyHhpKTqtWgVpC2YsvCG3j2SwVla6zKtxdICF/yQfKIqsLLG7yWZDw/btAEI3CDUYVP3sQugAAiBitcJk7wBUhLAEBgAquxCa0lj/w7azd4bR7kPAMQj1YtEAKTyYC9LOOXrhdUZ7zTUQNBqYTp2CLm8XbCaTowQVRhNEZ5yF0Eb7Z8s7dIBEqWyRz/cF6wRrNMHdFIT+hxI1QpwB2LBtm9tzLm7PbUT/w+EBEMeOIJNBO2UKMxGLNNLoKHaScuk0sXeABTPWIJzQYZv1W8WTo0MA3bIlTaBpQmhLWZloOCiXh1xv4+zOLSgUUHpxNacXFt32HQBEQz1rRQWAi0kDJH7XLaWlon7MZILR7p2kHuAeAEmjo9mYhKpPPsGFV14BMRohTU4KewcYxTkAqvnyK/GxVpL9cJTAClwep+XeYEaaRA0bBsjlMJ8755aZZgGQh2nvnNYLD4A4rRZ6IdfbL+TEZnPKALWeEhgARF02AoB9lo7NBuMp8c6TGgK2JGq7ENpw5IjHLhhnmAA6Iz3kd67Onj+qPn0g8WJoqBk6VHREP3MG5vPnxXZqQiAoFBeNJkGakCBq2giB6eRJGO3HThof7zVbEX+L6BRct2EDKj/8HwAg/bnnWiwDQQMg3d59bKhm4l13tchn+4N1gp065dIMQEuMTWmBp0iioqCxZ+NophcQA3bq/9PW9D8XOzwA4rRamKfNQdHTxlwkztaBXB5xkXZjNAMGQKLRwFpZCcPhw6j+4gsAQMzYK1t8LfJOnSCJjQUxmTwKNp1hAugmdMQEilSrZXYGai/lLwCQxsRAbS/bNezYwS6mUZdf3qL6qUgiCAI09kxY7U8/MR8ndf/+Xk0FVTk9RC2QvQsr/rbbEHPFFS2zYNhvQuRyEIMBIASxkyez4xhpFF26QJDLYa2pcekkZS3wQX7fHTc6jgBIt28fbPX1kMbHQ9W7/Q8QbU9cHGcXTpukcSmHCaCzsyHI5RFblycEhQKa4cMBAKWLl8BSUgJpfDxiJkxo+bUIAgs4zs6+G9VffOE2YoESDg8gZ2InTgJkMsT6+f8QNUK8sNRv+Q0134jddHHX/TUsa2qtxE4UTTXrflgP/X6q//EdUCTMmAEAUObkIOWxR8O7xMx97QAAEM1JREFUwEZIFAqmxRPkciTPe7hFP98XEqWSaaca8vLY49Q3KVjBP/2eNuzcyfyAGrZsEZ+7fORFE7C3F/jR4rRaVL16AVIpLGVlMJeWQr9/P4DWYYDoiajLLgMA6PeKs9Tibrjea9kn3KTOnw9l9+6wVlej+B9P4cz029g0dmeCGQvQFJIfehA5v+/ymxmgOqC6DRtgLS+HNCEB0aNGhWVNrZXoMWMgKJUwnTmDul9+AeBZAO1M7Pjx6LTq/9Dp//4XEfGxZog44Tx+xt+CKiuFE1qOol1bNr3eMfMuyO+7qk8fSGJjYaurYy7n9fYux+jLL67va3uAB0CcVotEo2Etvfr9f7DMQMyVLZfmbwrRIy9z/EMQEHfzzRFbi7JLF2R/+QVSHn8cgkYD/b59yL/uelR88F+X7ajvjCJMGSBBIoFErfa7nbpfP3EQpF2voZ0ypdVl+cKNNDqKBX1ErwcEgRlb+kIzZEhYptQHQvKDDyLzrbeQMm9eRD7fF44AaBcIIcw0UhIdzYa5NhVBKkXUpZcCEAMfc0mJmJkWBEQ5//45bQIeAHFaNbQMVrFiBSwlJZBotYgeOzbCq/KMvFMn1oETPWpUxO+IBbkciX+fha7rvhdLcVYryl57Dda6OrZNczURoUKQy10EpNrrrovgaiIHnS0HAIquXVqFJYUvpNHRiLnyilbZ+q0eOACCQgFLWRlMBQUuLfDNGdYabddZVa5ciaqPPhI/q1+/Fh15wwkNPADitGroHTBNN2v/8pdW4TPiCUEQEH/LNAhqNRLvvjvSy2HI09KQuex10RvFbEa9XbNg0+uZcWOoPYCCgZYQVb17Q5XTOsuc4SZ69GgIKhUAQN3Pd/mL4xtnHZBu1+/NaoF3RnvNFGiGDoVNp0PFCnGCetRoXv5qi/AAiNOqaawd0bZyYWzinXciZ+8eaAa5ux5HGtqRVr/pZwAISUkglMTfdCOSH34YGS8tjfRSIoZEo0HsxIkAGpVUOUFBs4r1W7ag9ocfADR/5IsglSJj6Ysuvxmu/2mb8ACI06pRduvG7oiVPXsy75HWTHPS6+Ek+kp7ALRlC4jJ5CgJZGW1ijULCgWS7p4d8WGakSZt0UJ0+mgVYiZN8r8xxycsANq0Cfo9eyDRaEJyEyVPT0f6P58DAMjS03n7extFFukFcDi+EGQyqPv1g27XLsRdd12ruFC3VdT9+0OalARreTkafv/dIYAOUwcYJzgkajU0gwdHehntAvWA/hAUChCTCYJajawV74XMRT52/HjIVv0fpImJvP29jcKPGqfVk7ZoIVKffBLxt94S6aW0aQSJBDFXjAEA1K3/Ebpdoj+KPIRT4Dmc1oREqUTs1VdDEhuLrLffCnlgqRkyBMrs1jE+iNN0eAaI0+pRdu0KZdeWGfDY3om+8kpUf/Y5qj/7jD0WdenwCK6IwwkvGYv/BWI2X3S2Chz/8AwQh3MREZWbC0GjASDOn+rw2quIHj06wqvicMILD344nuAZIA7nIkKiUqHDyy/DcOgQ4v92G/cu4XA4Fy08AOJwLjJirryi1bppczgcTkvBAyAvELslf21tbYRXwuFwOBwOJ1DodZtex73BAyAv1NnHBWSFaUYSh8PhcDic8FFXVwetD5NXgfgLkS5SbDYbioqKEBMTE1LvmdraWmRlZeHs2bOIjdAAw3DD97Ht0973D+D72B5o7/sHtP99DMf+EUJQV1eHjIwMSHx4NPEMkBckEgkywzjMMjY2tl1+mZ3h+9j2ae/7B/B9bA+09/0D2v8+hnr/fGV+KLwNnsPhcDgczkUHD4A4HA6Hw+FcdPAAqIVRKpVYtGgRlEplpJcSNvg+tn3a+/4BfB/bA+19/4D2v4+R3D8uguZwOBwOh3PRwTNAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdALczy5cvRuXNnqFQqDB8+HLt27Yr0koJi8eLFGDp0KGJiYpCSkoKpU6fi+PHjLtuMGTMGgiC4/LnnnnsitOKm88wzz7itv2fPnux5g8GA++67D4mJiYiOjsb111+P0tLSCK646XTu3NltHwVBwH333Qeg7R3DLVu2YMqUKcjIyIAgCFi7dq3L84QQLFy4EOnp6VCr1Rg3bhxOnDjhsk1lZSWmT5+O2NhYxMXF4Y477kB9fX0L7oVvfO2j2WzG/Pnz0bdvX0RFRSEjIwMzZsxAUVGRy3t4Ou5Llixp4T3xjr/jePvtt7utf+LEiS7btObj6G//PP0mBUHASy+9xLZpzccwkOtDIOfPwsJCXH311dBoNEhJScFjjz0Gi8USsnXyAKgFWbNmDebNm4dFixZh79696N+/PyZMmICysrJIL63J/Prrr7jvvvuwc+dObNiwAWazGePHj0dDQ4PLdnfddReKi4vZn6VLl0ZoxcHRu3dvl/Vv3bqVPffwww/j22+/xWeffYZff/0VRUVFuO666yK42qbz+++/u+zfhg0bAAA33ngj26YtHcOGhgb0798fy5cv9/j80qVL8cYbb+Cdd95BXl4eoqKiMGHCBBgMBrbN9OnTcfjwYWzYsAHfffcdtmzZgtmzZ7fULvjF1z7qdDrs3bsXTz/9NPbu3Ysvv/wSx48fxzXXXOO27XPPPedyXB944IGWWH5A+DuOADBx4kSX9X/yyScuz7fm4+hv/5z3q7i4GB988AEEQcD111/vsl1rPYaBXB/8nT+tViuuvvpqmEwmbN++HR9++CFWrlyJhQsXhm6hhNNiDBs2jNx3333s31arlWRkZJDFixdHcFWhoaysjAAgv/76K3ts9OjR5KGHHorcoprJokWLSP/+/T0+V11dTeRyOfnss8/YY0ePHiUAyI4dO1pohaHnoYceIl27diU2m40Q0raPIQDy1VdfsX/bbDaSlpZGXnrpJfZYdXU1USqV5JNPPiGEEHLkyBECgPz+++9smx9++IEIgkDOnz/fYmsPlMb76Ildu3YRAOTMmTPssU6dOpHXXnstvIsLEZ72cebMmeTaa6/1+pq2dBwDOYbXXnstufLKK10ea0vHsPH1IZDz57p164hEIiElJSVsm7fffpvExsYSo9EYknXxDFALYTKZsGfPHowbN449JpFIMG7cOOzYsSOCKwsNNTU1AICEhASXxz/66CMkJSWhT58+WLBgAXQ6XSSWFzQnTpxARkYGunTpgunTp6OwsBAAsGfPHpjNZpfj2bNnT3Ts2LHNHk+TyYRVq1bh73//u8sA4LZ+DCn5+fkoKSlxOWZarRbDhw9nx2zHjh2Ii4vDkCFD2Dbjxo2DRCJBXl5ei685FNTU1EAQBMTFxbk8vmTJEiQmJmLgwIF46aWXQlpaaAk2b96MlJQU5OTkYM6cOaioqGDPtafjWFpaiu+//x533HGH23Nt5Rg2vj4Ecv7csWMH+vbti9TUVLbNhAkTUFtbi8OHD4dkXXwYagtRXl4Oq9XqcjABIDU1FceOHYvQqkKDzWbD3Llzcdlll6FPnz7s8VtvvRWdOnVCRkYGDhw4gPnz5+P48eP48ssvI7jawBk+fDhWrlyJnJwcFBcX49lnn8Xll1+OQ4cOoaSkBAqFwu2ikpqaipKSksgsuJmsXbsW1dXVuP3229ljbf0YOkOPi6ffIH2upKQEKSkpLs/LZDIkJCS0yeNqMBgwf/583HLLLS6DJh988EEMGjQICQkJ2L59OxYsWIDi4mK8+uqrEVxt4EycOBHXXXcdsrOzcerUKTz55JOYNGkSduzYAalU2q6O44cffoiYmBi38npbOYaerg+BnD9LSko8/lbpc6GAB0CcZnPffffh0KFDLvoYAC719r59+yI9PR1jx47FqVOn0LVr15ZeZpOZNGkS+3u/fv0wfPhwdOrUCZ9++inUanUEVxYe/vOf/2DSpEnIyMhgj7X1Y3gxYzabcdNNN4EQgrffftvluXnz5rG/9+vXDwqFAnfffTcWL17cJkYuTJs2jf29b9++6NevH7p27YrNmzdj7NixEVxZ6Pnggw8wffp0qFQql8fbyjH0dn1oDfASWAuRlJQEqVTqpnIvLS1FWlpahFbVfO6//3589913+OWXX5CZmelz2+HDhwMATp482RJLCzlxcXHo0aMHTp48ibS0NJhMJlRXV7ts01aP55kzZ7Bx40bceeedPrdry8eQHhdfv8G0tDS3pgSLxYLKyso2dVxp8HPmzBls2LDBJfvjieHDh8NisaCgoKBlFhhiunTpgqSkJPa9bC/H8bfffsPx48f9/i6B1nkMvV0fAjl/pqWlefyt0udCAQ+AWgiFQoHBgwdj06ZN7DGbzYZNmzYhNzc3gisLDkII7r//fnz11Vf4+eefkZ2d7fc1+/fvBwCkp6eHeXXhob6+HqdOnUJ6ejoGDx4MuVzucjyPHz+OwsLCNnk8//vf/yIlJQVXX321z+3a8jHMzs5GWlqayzGrra1FXl4eO2a5ubmorq7Gnj172DY///wzbDYbC/5aOzT4OXHiBDZu3IjExES/r9m/fz8kEolb2aitcO7cOVRUVLDvZXs4joCYlR08eDD69+/vd9vWdAz9XR8COX/m5ubi4MGDLoEsDeYvueSSkC2U00KsXr2aKJVKsnLlSnLkyBEye/ZsEhcX56JybyvMmTOHaLVasnnzZlJcXMz+6HQ6QgghJ0+eJM899xzZvXs3yc/PJ19//TXp0qULGTVqVIRXHjiPPPII2bx5M8nPzyfbtm0j48aNI0lJSaSsrIwQQsg999xDOnbsSH7++Weye/dukpubS3JzcyO86qZjtVpJx44dyfz5810eb4vHsK6ujuzbt4/s27ePACCvvvoq2bdvH+uAWrJkCYmLiyNff/01OXDgALn22mtJdnY20ev17D0mTpxIBg4cSPLy8sjWrVtJ9+7dyS233BKpXXLD1z6aTCZyzTXXkMzMTLJ//36X3ybtnNm+fTt57bXXyP79+8mpU6fIqlWrSHJyMpkxY0aE98yBr32sq6sjjz76KNmxYwfJz88nGzduJIMGDSLdu3cnBoOBvUdrPo7+vqeEEFJTU0M0Gg15++233V7f2o+hv+sDIf7PnxaLhfTp04eMHz+e7N+/n6xfv54kJyeTBQsWhGydPABqYf7973+Tjh07EoVCQYYNG0Z27twZ6SUFBQCPf/773/8SQggpLCwko0aNIgkJCUSpVJJu3bqRxx57jNTU1ER24U3g5ptvJunp6UShUJAOHTqQm2++mZw8eZI9r9fryb333kvi4+OJRqMhf/3rX0lxcXEEVxwcP/74IwFAjh8/7vJ4WzyGv/zyi8fv5cyZMwkhYiv8008/TVJTU4lSqSRjx4512++Kigpyyy23kOjoaBIbG0tmzZpF6urqIrA3nvG1j/n5+V5/m7/88gshhJA9e/aQ4cOHE61WS1QqFenVqxf517/+5RI8RBpf+6jT6cj48eNJcnIykcvlpFOnTuSuu+5yu5FszcfR3/eUEELeffddolarSXV1tdvrW/sx9Hd9ICSw82dBQQGZNGkSUavVJCkpiTzyyCPEbDaHbJ2CfbEcDofD4XA4Fw1cA8ThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjp4AMThcDgcDueigwdAHA6Hw+FwLjr+Hyzk/CO/bXCrAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "chunk = next(sliced(preprocessed_data, window_length))\n",
    "\n",
    "axs = chunk.plot(subplots=True)\n",
    "for i,ax in enumerate(axs):\n",
    "    axs[i].legend(loc=\"upper right\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96250c3e",
   "metadata": {},
   "source": [
    "In this session we recorded 6 words with the Synthetic board.\n",
    "Now that we have all the session preprocessed, we need to split it in six parts. However, some words might be recorded with an average of ± 2 extra samples, therefore, we also need to map all the splits to have the same size:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3b9652c0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Lowest sample size: '189'\n",
      "\n",
      "EMG data for the first word ('Open hand'):\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000914</td>\n",
       "      <td>0.000747</td>\n",
       "      <td>-0.000752</td>\n",
       "      <td>-0.000206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.002628</td>\n",
       "      <td>0.000694</td>\n",
       "      <td>-0.001444</td>\n",
       "      <td>-0.000868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.001910</td>\n",
       "      <td>-0.002834</td>\n",
       "      <td>0.000752</td>\n",
       "      <td>-0.001772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.001237</td>\n",
       "      <td>-0.006782</td>\n",
       "      <td>0.002450</td>\n",
       "      <td>-0.001909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.003539</td>\n",
       "      <td>-0.007481</td>\n",
       "      <td>0.001235</td>\n",
       "      <td>-0.001372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>0.000800</td>\n",
       "      <td>0.000620</td>\n",
       "      <td>0.002009</td>\n",
       "      <td>-0.000793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>0.001614</td>\n",
       "      <td>0.000603</td>\n",
       "      <td>0.001741</td>\n",
       "      <td>-0.001457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>0.000584</td>\n",
       "      <td>0.002295</td>\n",
       "      <td>0.001405</td>\n",
       "      <td>-0.002100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>187</th>\n",
       "      <td>-0.001439</td>\n",
       "      <td>0.005511</td>\n",
       "      <td>0.002690</td>\n",
       "      <td>-0.001532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>-0.002774</td>\n",
       "      <td>0.008009</td>\n",
       "      <td>0.002525</td>\n",
       "      <td>-0.000686</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>189 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            0         1         2         3\n",
       "0    0.000914  0.000747 -0.000752 -0.000206\n",
       "1    0.002628  0.000694 -0.001444 -0.000868\n",
       "2    0.001910 -0.002834  0.000752 -0.001772\n",
       "3   -0.001237 -0.006782  0.002450 -0.001909\n",
       "4   -0.003539 -0.007481  0.001235 -0.001372\n",
       "..        ...       ...       ...       ...\n",
       "184  0.000800  0.000620  0.002009 -0.000793\n",
       "185  0.001614  0.000603  0.001741 -0.001457\n",
       "186  0.000584  0.002295  0.001405 -0.002100\n",
       "187 -0.001439  0.005511  0.002690 -0.001532\n",
       "188 -0.002774  0.008009  0.002525 -0.000686\n",
       "\n",
       "[189 rows x 4 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emg_words = []\n",
    "emg_data = []\n",
    "curves = pd.DataFrame(curves)\n",
    "\n",
    "word = None\n",
    "data = None\n",
    "for i in range(len(restored_data)):\n",
    "    if i == 0:\n",
    "        word = restored_data.iloc[i, 0]\n",
    "        data = curves[i]\n",
    "    elif restored_data.iloc[i, 0] == word:\n",
    "        data = np.vstack((data, curves[i]))\n",
    "    else:\n",
    "        emg_words.append(word)\n",
    "        emg_data.append(data)\n",
    "        word = restored_data.iloc[i, 0]\n",
    "        data = curves[i]\n",
    "\n",
    "emg_words.append(word)\n",
    "emg_data.append(data)\n",
    "# We need all samples to have the same size, so we standardize it to have the lowest size\n",
    "samples = min([len(x) for x in emg_data])\n",
    "data_reduced = [x[:samples] for x in emg_data]\n",
    "emg_data = np.asarray(data_reduced).astype('float32')\n",
    "\n",
    "print(\"Lowest sample size: '%d'\" % samples)\n",
    "print(\"\\nEMG data for the first word ('%s'):\" % (emg_words[0]))\n",
    "\n",
    "pd.DataFrame(emg_data[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d91f0b2d",
   "metadata": {},
   "source": [
    "### Words visualization\n",
    "\n",
    "These words are now a 2d array (sample x channel) inside a list of words. Which means all the words could be individually seen as a 2-dimensional image heatmap, knowing the columns represents each channel, and the rows are the data recorded over time:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "503e49a1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDMAAAJjCAYAAAARaVp2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACp/ElEQVR4nOzdeXwV5dn/8e9JyE4WAoEQiEAQFwStC1DcQEURXIprtdZ9a0Ws2lr1eapVu2Dr87S2arVWq1arPy0uVR+17lorrrghiuwiCEJCdrKQzO+P+ywzd8KWDJPMOZ/36zWvk1znOsmdcDFzZnJfc0ccx3EEAAAAAAAQEmk9PQAAAAAAAIDtwcUMAAAAAAAQKlzMAAAAAAAAocLFDAAAAAAAECpczAAAAAAAAKHCxQwAAAAAABAqXMwAAAAAAAChwsUMAAAAAAAQKlzMAAAAAAAAoZKSFzMikYguvvjinh6Gx/Dhw3XWWWf19DBCozf/vl599VVFIhG9+uqrW8y77rrrFIlEtH79+mAGtg3uvfdeRSIRLV++vKeHgu2wrTW3PXb0/7HJkydrzJgxO+zrh0Vv2pctX75ckUhE9957bzwW20/1Vr3p94euix173nvvvR3y9Xt7HSNcOI8Iv978++I8Yvt06WLGp59+qu9///saMmSIsrKyVFZWptNOO02ffvqp3+OD5eGHH9b3v/99jRo1SpFIRJMnT+7pIflqyZIluvDCC1VRUaHs7GwVFBTogAMO0B/+8Adt3Lixp4eHqK+//lpXXXWVDjnkEOXn5/t+Ir0j+fmmubGxUdddd11ofvZk8vzzz+vcc8/VmDFjlJ6eruHDh/f0kDw++eQTnXjiiRo2bJiys7M1ZMgQHX744brlllt6emjwSTIej2P7x9jWp08fDRkyRGeddZZWrVrV08MLpYULF+qyyy7T/vvvr+zsbP5gIM4jekplZaVuuukmHXzwwSopKVFRUZG+/e1v6+GHH+7pofmG84hweOmll3TOOedol112UW5urioqKnTeeefp66+/3u6v1Wd7X/DYY4/p1FNPVXFxsc4991yNGDFCy5cv19133605c+bo//2//6fjjjtuuweCbXP77bfr/fff17hx41RZWdnTw/HV//3f/+mkk05SVlaWzjjjDI0ZM0YtLS164403dMUVV+jTTz/VnXfe2dPDhMybs9/85jcaNWqUxo4dq7lz5/b0kHpEY2Ojrr/+eklKihMZt4ULFyotrfdO3nvwwQf18MMPa5999lFZWVlPD8fjzTff1CGHHKKddtpJ559/vkpLS7Vy5Uq99dZb+sMf/qBZs2bFc3v77/lnP/uZrrrqqp4exmb15O8vmY/HN9xwg0aMGKGmpia99dZbuvfee/XGG29o/vz5ys7O7unhbbeerOO5c+fqj3/8o0aPHq3dd99dH374YY+Mo7fgPKLnzJ07V//93/+t6dOn62c/+5n69OmjRx99VKeccooWLFgQfz8TVpxHhMeVV16pqqoqnXTSSRo1apSWLl2qW2+9VU8//bQ+/PBDlZaWbvPX2q6LGUuWLNHpp5+uiooKvf766yopKYk/96Mf/UgHHXSQTj/9dH388ceqqKjYni+NbXT//fdryJAhSktLS6op2suWLdMpp5yiYcOG6eWXX9bgwYPjz82cOVOLFy/W//3f//XgCOG27777qrKyUsXFxZozZ45OOumknh5SUmloaFBeXl6PjiErK6tHv//W/PrXv9Zf/vIXZWRk6Oijj9b8+fN7ekhxv/rVr1RYWKh3331XRUVFnue++eYbz+e9/ffcp08f9emz3X/3CExP/v6S9XgsSdOmTdN+++0nSTrvvPM0YMAA/eY3v9GTTz6pk08+uYdHt/16so6PPfZYVVdXKz8/X//zP/+T0hczOI/oWXvssYcWLVqkYcOGxWMXXXSRpkyZot/85jf66U9/2uPvPbqK84hw+d3vfqcDDzzQ88eII488UpMmTdKtt96qX/7yl9v8tbbrzxk33XSTGhsbdeedd3p2QJI0YMAA/fnPf1ZDQ4N++9vfxuOxfp7PP/9cJ598sgoKCtS/f3/96Ec/UlNTU4fv8cADD2jfffdVTk6OiouLdcopp2jlypWenFiv9YIFC3TIIYcoNzdXQ4YM8XzfbfHEE09ozJgxysrK0h577KHnnnvO8/yKFSt00UUXadddd1VOTo769++vk046qcP0wNi0zP/85z+6/PLLVVJSory8PB133HFat26dJ9dxHP3yl7/U0KFDlZubq0MOOWS7ptWVl5f36r/iddVvf/tb1dfX6+677/bsgGJ23nln/ehHP9ri11i6dKlOOukkFRcXKzc3V9/+9rc73XHdcsst2mOPPZSbm6t+/fppv/3204MPPujJWbVqlc455xwNGjQoXh9//etfO3ytr776SjNmzFBeXp4GDhyoyy67TM3Nzdv1s1dXV+uss85SUVGRCgsLdfbZZ6uxsdGTc8899+jQQw/VwIEDlZWVpdGjR+v222/v8LWGDx+uo48+Wm+88YbGjx+v7OxsVVRU6G9/+1uH3E8//VSHHnqocnJyNHToUP3yl79Ue3v7No05Pz9fxcXF2/VzhklLS4uuvfZa7bvvviosLFReXp4OOuggvfLKK/Gc5cuXx/eD119/fXxa9nXXXRfP+fzzz3XiiSequLhY2dnZ2m+//fTkk096vlds//Haa6/poosu0sCBAzV06FBJ274P2lbbsz+2+0m3Zz8nSc8++6wmTZqk/Px8FRQUaNy4cR3+n0nq8n68rKxMGRkZ2/cLCMiSJUu0xx57dLiQIUkDBw70fN5Z3+7HH3+sSZMmef5v3nPPPR2mp2/r//eqqir95Cc/0dixY9W3b18VFBRo2rRp+uijj7b6s3R2r4FYv/jWjqGS6f3db7/9lJ2drZEjR+rPf/7zNt+/YNGiRTrhhBNUWlqq7OxsDR06VKeccopqamoC/f1tTrIejztz0EEHSTK17fbyyy/roIMOUl5enoqKivSd73xHn332WYfXr1q1Sueee67KysqUlZWlESNG6Ic//KFaWlo2+z03bNig8ePHa+jQoVq4cOFm81pbW3X99ddr1KhRys7OVv/+/XXggQfqhRdeiOd0VnMbN27UJZdcogEDBig/P1/HHnusVq1a1WE/Hnvt4sWLt3qs7kxxcbHy8/O3mpcKOI/o2fOIESNGeC5kSGZ/PmPGDDU3N2vp0qXb9fP3JpxHhOs84uCDD+5w/Dz44INVXFzc6TFkS7brMvVTTz2l4cOHxw9qnQ1s+PDhnf7Dn3zyyRo+fLhmz56tt956S3/84x+1YcMGzy/nV7/6la655hqdfPLJOu+887Ru3TrdcsstOvjgg/XBBx943hhu2LBBRx55pI4//nidfPLJmjNnjq688kqNHTtW06ZN2+rP8sYbb+ixxx7TRRddpPz8fP3xj3/UCSecoC+//FL9+/eXJL377rt68803dcopp2jo0KFavny5br/9dk2ePFkLFixQbm6u52vOmjVL/fr1089//nMtX75cN998sy6++GJPL9q1116rX/7yl5o+fbqmT5+uefPm6YgjjtjiAT0VPPXUU6qoqND+++/fpdevXbtW+++/vxobG3XJJZeof//+uu+++3Tsscdqzpw58SmLf/nLX3TJJZfoxBNPjB8IP/74Y7399tv63ve+F/9a3/72t+Nv2EtKSvTss8/q3HPPVW1trS699FJJ5o3QYYcdpi+//FKXXHKJysrKdP/99+vll1/errGffPLJGjFihGbPnq158+bprrvu0sCBA/Wb3/wmnnP77bdrjz320LHHHqs+ffroqaee0kUXXaT29nbNnDnT8/UWL16sE088Ueeee67OPPNM/fWvf9VZZ52lfffdV3vssYckac2aNTrkkEO0adMmXXXVVcrLy9Odd96pnJycLv3+k01tba3uuusunXrqqTr//PNVV1enu+++W1OnTtU777yjb33rWyopKdHtt9+uH/7whzruuON0/PHHS5L23HNPSWYnf8ABB2jIkCHx3/EjjzyiGTNm6NFHH+0wjfaiiy5SSUmJrr32WjU0NEja/n3QttqW/fHmbMt+7t5779U555yjPfbYQ1dffbWKior0wQcf6Lnnnov/P5O6vx/vrYYNG6a5c+dq/vz52/0X+1WrVumQQw5RJBLR1Vdfrby8PN11112bnYGwLf/fly5dqieeeEInnXSSRowYobVr1+rPf/6zJk2apAULFnSpTWdbjqEffPCBjjzySA0ePFjXX3+92tradMMNN3Q4ielMS0uLpk6dqubmZs2aNUulpaVatWqVnn76aVVXV6uwsDCw3x8UP/nq169fPPbiiy9q2rRpqqio0HXXXaeNGzfqlltu0QEHHKB58+bF72OzevVqjR8/XtXV1brgggu02267adWqVZozZ44aGxuVmZnZ4futX79ehx9+uKqqqvTaa69p5MiRmx3bddddp9mzZ+u8887T+PHjVVtbq/fee0/z5s3T4YcfvtnXnXXWWXrkkUd0+umn69vf/rZee+01HXXUUZvN35ZjNbaM84jeeR6xZs0aSeaCUlhxHhH+84j6+nrV19dvfx0626i6utqR5HznO9/ZYt6xxx7rSHJqa2sdx3Gcn//8544k59hjj/XkXXTRRY4k56OPPnIcx3GWL1/upKenO7/61a88eZ988onTp08fT3zSpEmOJOdvf/tbPNbc3OyUlpY6J5xwwlZ/FklOZmams3jx4njso48+ciQ5t9xySzzW2NjY4bVz587t8L3vueceR5IzZcoUp729PR6/7LLLnPT0dKe6utpxHMf55ptvnMzMTOeoo47y5P3Xf/2XI8k588wztzp2tz322MOZNGnSdr2mN6qpqdmm2nIbNmyY5/d16aWXOpKcf//73/FYXV2dM2LECGf48OFOW1ub4ziO853vfMfZY489tvi1zz33XGfw4MHO+vXrPfFTTjnFKSwsjNfFzTff7EhyHnnkkXhOQ0ODs/POOzuSnFdeeWWL3yf2f+Occ87xxI877jinf//+nlhntTh16lSnoqLCExs2bJgjyXn99dfjsW+++cbJyspyfvzjH8djsd/X22+/7ckrLCx0JDnLli3b4tjd/vGPf2zTz9tbxP6/vvvuu5vN2bRpk9Pc3OyJbdiwwRk0aJDn32vdunWOJOfnP/95h69x2GGHOWPHjnWamprisfb2dmf//fd3Ro0a1WE8Bx54oLNp0ybP19jWfdArr7yyXTW3tf2x43T8P7at+7nq6monPz/fmTBhgrNx40bP93G/rrv7cbejjjrKGTZs2Ha9Zkd6/vnnnfT0dCc9Pd2ZOHGi89Of/tT517/+5bS0tHTItX/Ps2bNciKRiPPBBx/EY5WVlU5xcXGH/5vb+v+9qakpvg+MWbZsmZOVleXccMMNnpgk55577onHYjXjtq3H0GOOOcbJzc11Vq1aFY8tWrTI6dOnT4evafvggw8cSc4//vGPLeYF8fvbFslyPI79P3/xxReddevWOStXrnTmzJnjlJSUOFlZWc7KlSvjud/61recgQMHOpWVlfHYRx995KSlpTlnnHFGPHbGGWc4aWlpne5zY/sE937566+/dvbYYw+noqLCWb58+VbHvNdeezlHHXXUFnPsOn7//fcdSc6ll17qyTvrrLM67NO351i9NTfddNN2H2OTBecRRm86j3Acs38cOHCgc9BBB233a3sLziPCfR4R84tf/MKR5Lz00kvb9bptnh9ZV1cnSVudKhd7vra21hO3r/rEboD2zDPPSDI3BGpvb9fJJ5+s9evXx7fS0lKNGjXKM71bkvr27avvf//78c8zMzM1fvz4bZ4iNWXKFM+V/j333FMFBQWe17uvLrW2tqqyslI777yzioqKNG/evA5f84ILLvBMYzzooIPU1tamFStWSDJ/xWhpadGsWbM8ebErdKkqVivdmYb5zDPPaPz48TrwwAPjsb59++qCCy7Q8uXLtWDBAklSUVGRvvrqK7377rudfh3HcfToo4/qmGOOkeM4nlqcOnWqampq4v/2zzzzjAYPHqwTTzwx/vrc3FxdcMEF2zX2H/zgB57PDzroIFVWVnr+D7lrsaamRuvXr9ekSZO0dOlSz5RrSRo9erTnrx4lJSXaddddPbX9zDPP6Nvf/rbGjx/vyTvttNO2a+zJKj09Pf7Xwvb2dlVVVWnTpk3ab7/9Ov2/b6uqqtLLL7+sk08+WXV1dfEaqqys1NSpU7Vo0aIOKwOcf/75Sk9P98S2dx+0rba2P96Sre3nXnjhBdXV1emqq67qcKNAe5p3d/fjvdXhhx+uuXPn6thjj9VHH32k3/72t5o6daqGDBnSoc3I9txzz2nixIn61re+FY8VFxdv9v/mtvx/z8rKik/nbGtrU2Vlpfr27atdd921y3W0tWNoW1ubXnzxRc2YMcMz82PnnXfepr96xmZe/Otf/9qmqfwxO+L3l4qmTJmikpISlZeX68QTT1ReXp6efPLJeAvc119/rQ8//FBnnXWWp+Vwzz331OGHHx7fl7S3t+uJJ57QMcccE78Hh5u9T/jqq680adIktba26vXXX+8wJb4zRUVF+vTTT7Vo0aJt/vli7QAXXXSRJ+6+Oa9tW47V2DzOI3rfeUR7e7tOO+00VVdXh3qlLc4jwn8e8frrr+v666/XySefrEMPPXS7XrvNFzNiBRLbGW3O5nZWo0aN8nw+cuRIpaWlxacuLlq0SI7jaNSoUSopKfFsn332WYebpg0dOrTDQbBfv37asGHDNv08O+20U4eY/fqNGzfq2muvVXl5ubKysjRgwACVlJSourq6wz98Z18zNh0z9jVjOyP7d1FSUuKZuplqCgoKJG29trZkxYoV2nXXXTvEd9999/jzkrl7bt++fTV+/HiNGjVKM2fO1H/+8594/rp161RdXR3v53RvZ599tqTEDfxWrFihnXfeuUMddjaOLdla3UjSf/7zH02ZMiXel1xSUqL/+q//kqQOtbgttb1ixYoOddiVsSez++67T3vuuWe8B7ukpET/93//1+n/fdvixYvlOI6uueaaDnX085//XFLHG0GOGDGiw9fZ3n3Qttra/nhLtlavsZ76bWmv6O5+vDcbN26cHnvsMW3YsEHvvPOOrr76atXV1enEE0+MvynqTGy/YussJm3b//f29nb9/ve/16hRozx19PHHH3e5jrb2fb/55htt3Lhxu34WtxEjRujyyy/XXXfdpQEDBmjq1Km67bbbtjreHfH7S0W33XabXnjhBc2ZM0fTp0/X+vXrPa06sWPq5o6769evV0NDg9atW6fa2tptbrc6/fTT9c033+i1117TkCFDtuk1N9xwg6qrq7XLLrto7NixuuKKK/Txxx9v8TUrVqxQWlpah/3ulmpzW47V2DzOI3rfecSsWbP03HPP6a677tJee+213a/vLTiPCPd5xOeff67jjjtOY8aM0V133bXdr9/me2YUFhZq8ODBWz1AfPzxxxoyZEi8sDbH/odrb29XJBLRs88+2+Gvk5K5OubWWY5krohti215/axZs3TPPffo0ksv1cSJE1VYWKhIJKJTTjml0xucdHdMqaqgoEBlZWWBrEaw++67a+HChXr66af13HPP6dFHH9Wf/vQnXXvttbr++uvj/67f//73deaZZ3b6NWL3RPDL1upmyZIlOuyww7Tbbrvpd7/7ncrLy5WZmalnnnlGv//97zvUInXYfQ888IDOOusszZgxQ1dccYUGDhyo9PR0zZ49u8MN8DoT+zf5yU9+oqlTp3aaY79p7qzPcHv3QV21LTdjjPGzvlKhVjMzMzVu3DiNGzdOu+yyi84++2z94x//iF/U6q5t+R3++te/1jXXXKNzzjlHv/jFL1RcXKy0tDRdeumlXa6jIP7t/vd//1dnnXWW/vnPf+r555/XJZdcEu+Xj80Q6K5UqMGuGD9+fHwmxYwZM3TggQfqe9/7nhYuXNjh/Zifjj/+eP3tb3/TH/7wB82ePXubXnPwwQdryZIl8Tq566679Pvf/1533HGHzjvvPN/GRq10D+cRves84vrrr9ef/vQn3XjjjTr99NN9//pB4jwivOcRK1eu1BFHHKHCwkI988wzXZpds103AD366KP1l7/8RW+88YZnGk7Mv//9by1fvlwXXnhhh+cWLVrkuQK+ePFitbe3x28QNXLkSDmOoxEjRmiXXXbZzh9jx5gzZ47OPPNM/e///m881tTUpOrq6i59vdh0yUWLFnmWnFq3bl3KX9k/+uijdeedd2ru3LmaOHHidr9+2LBhnd7t/PPPP48/H5OXl6fvfve7+u53v6uWlhYdf/zx+tWvfqWrr75aJSUlys/PV1tbm6ZMmbLV7zl//nw5juM5qG7prutd8dRTT6m5uVlPPvmk52qpPWVyewwbNqzTKbl+jz2s5syZo4qKCj322GOef1v7BHRzFwFi/78zMjK2WkdbG4ef+6CYre2PuyM27Xb+/Pnb9Bf4VBI7Ofz66683mzNs2DAtXry4Q7yz2LaaM2eODjnkEN19992eeHV19Q674dvAgQOVnZ3d7Z9l7NixGjt2rH72s5/pzTff1AEHHKA77rhjs8u27YjfX6qLXcg95JBDdOutt+qqq66KH1M3d9wdMGCA8vLylJOTo4KCgm0+yZg1a5Z23nlnXXvttSosLNRVV121Ta8rLi7W2WefrbPPPlv19fU6+OCDdd111232YsawYcPU3t6uZcuWef66SJ3sWJxH9I7ziNtuu03XXXedLr30Ul155ZVdGktvw3nE5vXW84jKykodccQRam5u1ksvvdTpKjTbYrvWFLviiiuUk5OjCy+8UJWVlZ7nqqqq9IMf/EC5ubm64oorOrz2tttu83we682K9c4ef/zxSk9P1/XXX9/hyo/jOB2+XxDS09M7jOWWW25RW1tbl77elClTlJGRoVtuucXzdW+++ebuDDMpxNa2Pu+887R27doOzy9ZskR/+MMfNvv66dOn65133tHcuXPjsYaGBt15550aPny4Ro8eLUkd6igzM1OjR4+W4zhqbW1Venq6TjjhBD366KOdvvlyL5E1ffp0rV69WnPmzInHYkuO+Sl2hdRdMzU1Nbrnnnu6/DWnT5+ut956S++88048tm7dOv3973/v+kCTSGe/87fffttTX5LidyK335gMHDhQkydP1p///OdOT1w7W8p0c+Pwcx8Us7X9cXccccQRys/P1+zZszssm5cqf8F85ZVXOv1ZY73dW5qGOXXqVM2dO1cffvhhPFZVVdWt/5ud1dE//vGPDvdt8VN6erqmTJmiJ554QqtXr47HFy9erGeffXarr6+trdWmTZs8sbFjxyotLW2Ly9btiN8fzFKW48eP180336ympiYNHjxY3/rWt3Tfffd59n/z58/X888/r+nTp0uS0tLSNGPGDD311FN67733Onzdzv6fXHPNNfrJT36iq6++utOlA232cb1v377aeeedt1onkvSnP/3JEw/zfQPCgPOInj+PePjhh3XJJZfotNNO0+9+97sujaM34jxi83rjeURDQ4OmT5+uVatW6Zlnnum0ZWVbbdfMjFGjRum+++7TaaedprFjx+rcc8/ViBEjtHz5ct19991av369HnrooU6X0Fq2bJmOPfZYHXnkkZo7d64eeOABfe9734v3aI0cOVK//OUvdfXVV2v58uWaMWOG8vPztWzZMj3++OO64IIL9JOf/KTLP2hXHH300br//vtVWFio0aNHa+7cuXrxxRfjSy5tr5KSEv3kJz/R7NmzdfTRR2v69On64IMP9Oyzz27zX8def/11vf7665JM0TQ0NMT/QnXwwQfr4IMP7tLYetrIkSP14IMP6rvf/a523313nXHGGRozZoxaWlr05ptv6h//+IfOOuuszb7+qquu0kMPPaRp06bpkksuUXFxse677z4tW7ZMjz76aPzmd0cccYRKS0t1wAEHaNCgQfrss89066236qijjopPbbrxxhv1yiuvaMKECTr//PM1evRoVVVVad68eXrxxRdVVVUlydyw8dZbb9UZZ5yh999/X4MHD9b999/f5eUyN+eII45QZmamjjnmGF144YWqr6/XX/7yFw0cOHCLf+Hdkp/+9Ke6//77deSRR+pHP/pRfEmlYcOGbXUKaEys7mLrm99///164403JEk/+9nPujSuIP31r3/tsCa8JP3oRz/S0Ucfrccee0zHHXecjjrqKC1btkx33HGHRo8erfr6+nhuTk6ORo8erYcffli77LKLiouLNWbMGI0ZM0a33XabDjzwQI0dO1bnn3++KioqtHbtWs2dO1dfffWVPvroo62O0e99UMzW9sfdUVBQoN///vc677zzNG7cOH3ve99Tv3799NFHH6mxsVH33Xdft7+HZKYix26muXjxYtXU1MRrcq+99tIxxxzjy/fpilmzZqmxsVHHHXecdtttt/h+7OGHH9bw4cPjfbOd+elPf6oHHnhAhx9+uGbNmhVfWnSnnXZSVVXVdrUExRx99NG64YYbdPbZZ2v//ffXJ598or///e+ev+ztCNddd52ef/55HXDAAfrhD3+otrY23XrrrRozZoznYkNnXn75ZV188cU66aSTtMsuu2jTpk26//77428UN2dH/P42J1mPx5tzxRVX6KSTTtK9996rH/zgB7rppps0bdo0TZw4Ueeee258adbCwkJdd9118df9+te/1vPPP69Jkybpggsu0O67766vv/5a//jHP/TGG294lsuMuemmm1RTU6OZM2cqPz/fc6NG2+jRozV58mTtu+++Ki4u1nvvvac5c+bo4osv3uxr9t13X51wwgm6+eabVVlZGV+a9YsvvpC0fa13W1NTUxM/8Y711t96660qKipSUVHRFseZbDiP6NnziHfeeUdnnHGG+vfvr8MOO6zDief++++/w48LOwrnEZvXG88jTjvtNL3zzjs655xz9Nlnn+mzzz6LP9e3b1/NmDFj2wezXWufRH388cfOqaee6gwePNjJyMhwSktLnVNPPdX55JNPOuTGlo1ZsGCBc+KJJzr5+flOv379nIsvvrjDsn2O4ziPPvqoc+CBBzp5eXlOXl6es9tuuzkzZ850Fi5cGM+ZNGlSp8vinHnmmdu0PJ8kZ+bMmR3i9jI9GzZscM4++2xnwIABTt++fZ2pU6c6n3/++WaXLLSXHetsucS2tjbn+uuvdwYPHuzk5OQ4kydPdubPn9/ha25O7PfZ2dbZ8pBh88UXXzjnn3++M3z4cCczM9PJz893DjjgAOeWW27xLHHZ2e9ryZIlzoknnugUFRU52dnZzvjx452nn37ak/PnP//ZOfjgg53+/fs7WVlZzsiRI50rrrjCqamp8eStXbvWmTlzplNeXh6v8cMOO8y58847PXkrVqxwjj32WCc3N9cZMGCA86Mf/ch57rnntmtJpXXr1nnisXpyL2v05JNPOnvuuaeTnZ3tDB8+3PnNb37j/PWvf+10qcHOlqibNGlSh2UDP/74Y2fSpElOdna2M2TIEOcXv/iFc/fdd2/zkkqbq8Mu7lYCE/v9bm5buXKl097e7vz61792hg0b5mRlZTl777238/TTT3e6j3nzzTedfffd18nMzOzw/3DJkiXOGWec4ZSWljoZGRnOkCFDnKOPPtqZM2dOh/F0tmzhtu6Dtndp1m3ZH3dnP+c4pmb3339/JycnxykoKHDGjx/vPPTQQ/Hnu7sf39K/Y1eWp/PTs88+65xzzjnObrvt5vTt29fJzMx0dt55Z2fWrFnO2rVrPbmd7cs++OAD56CDDnKysrKcoUOHOrNnz3b++Mc/OpKcNWvWeF67Lf/fm5qanB//+Mfx484BBxzgzJ07t0Pe9izNui3HUMdxnJdeesnZe++9nczMTGfkyJHOXXfd5fz4xz92srOzN/PbM5YuXeqcc845zsiRI53s7GynuLjYOeSQQ5wXX3xxq9/T79/f5iTj8XhL+6O2tjZn5MiRzsiRI+PLSL/44ovOAQccEP9/fswxxzgLFizo8NoVK1Y4Z5xxRnyJ14qKCmfmzJnxJbA7+75tbW3Oqaee6vTp08d54oknNjvmX/7yl8748eOdoqIiJycnx9ltt92cX/3qV56lkDur44aGBmfmzJlOcXGx07dvX2fGjBnOwoULHUnOjTfe2OG123Ks7kzs/1VnW29aUjpInEf0zHnE1t7/uPf9YcV5RDjOI2JLwPqxX4w4zo6d93vdddfp+uuv17p163ZYby4AYOvYH4fXpZdeqj//+c+qr6/f7M25wmLGjBnbvZRmdyXT7w87zocffqi9995bDzzwAEuV9xIctwBsyXbdMwMAAOxYGzdu9HxeWVmp+++/XwceeGDoTsTtn2XRokV65plnNHny5MC+Z5h/f9hx7DqRzL0H0tLSkq5FCACS1XbdMwMAAOxYEydO1OTJk7X77rtr7dq1uvvuu1VbW6trrrmmp4e23SoqKnTWWWepoqJCK1as0O23367MzEz99Kc/3WHfM5l+f9hxfvvb3+r999/XIYccoj59+ujZZ5/Vs88+qwsuuEDl5eU9PTwAwDbgYgYAAL3I9OnTNWfOHN15552KRCLaZ599dPfdd4fyr8VHHnmkHnroIa1Zs0ZZWVmaOHGifv3rX3frzuVbk0y/P+w4+++/v1544QX94he/UH19vXbaaSddd911+u///u+eHhoAYBvt8HtmAAAAAAAA+Ck098y47bbbNHz4cGVnZ2vChAmedW2BHYGaQ9CoOQSNmkPQqDn0BOoOQaPmghGKixkPP/ywLr/8cv385z/XvHnztNdee2nq1Kn65ptvenpoSFLUHIJGzSFo1ByCRs2hJ1B3CBo1F5xQtJlMmDBB48aN06233ipJam9vV3l5uWbNmqWrrrpqi69tb2/X6tWrlZ+fr0gksk3fz3Ec1dXVqaysTGlpobjeA591p+Zi+dtTd9QcqDkEjZpD0Kg59IQgzyOoOUjUXJB6/Q1AW1pa9P777+vqq6+Ox9LS0jRlyhTNnTu3Q35zc7Oam5vjn69atUqjR4/u0vdeuXKlhg4d2qXXIry2t+Yk/+qOmktN1ByCRs0haNQcekJPnUdQc6mLmgtWr7+YsX79erW1tWnQoEGe+KBBg/T55593yJ89e7auv/76DvGVe0gF6ZLaXcEvraS+5qG2XSpfI+Xn53dv8Ail7a05aQt1N1wqSJNU5woOsJI2RGvuG2ouVflacz+VCrKkF37h+jpWzp5HS7WtUvm/qLlU5WvNjZcK+kjaxRV83ko6MFpzj1NzqcrXmhsSPba6FsX54mVvzi7/kmobpPLjqblU5td5xHGSMiT9zBUbZq/2vFKqbZHKH6TmUplv564Do/u5mkRs40ZvTk4/qdaRyqtTt+Z6/cWM7XX11Vfr8ssvj39eW1ur8vJyFewiFWRIWuhKrrdeHJvJE2282da2FGCzdZcW3RG1uJLTrRfXi5rDdttszdVJBS2S+zZTvznKenG1pE3mQ2oO22qzNVcUPb4e7kqebr34ccX/mEDNYVtttuYGRv9A5fqjVF/rtQVH7C/VbpL0DjWHbba5mrs3IhVEJLUldnSfRF7wvHaNpIbox9QcttUWj63pkly32fjIeu1BzUr5c4hefzFjwIABSk9P19q1az3xtWvXqrS0tEN+VlaWsrKyghoektD21pxE3aF7qDkEjZpD0Kg59ATOIxA0ai5Yvf4uIZmZmdp333310ksvxWPt7e166aWXNHHixB4cGZIVNYegUXMIGjWHoFFz6AnUHYJGzQWr18/MkKTLL79cZ555pvbbbz+NHz9eN998sxoaGnT22Wdv+xepkdRHWuWanzNkL29KW/S5tm6PGGHnS81JppgceVpLNs73prQrMS0Rqcu3mquX1CKd7I5NtXJuFzs6+FdzeTLN5O6a2tvKebx7Y0Vy8K3mCmXewS5KhHbbz05aLu+N0pCqfKm7nRT9E3DihkAr5Z3Wf7ukVj8GjNDzpeY2yezCXOcQDdYurbZRqvVjwCEWiosZ3/3ud7Vu3Tpde+21WrNmjb71rW/pueee63BjFcAv1ByCRs0haNQcgkbNoSdQdwgaNReciOM4Tk8PYkeqra1VYWGhag41d1tf5brD+uZmZtRKKpZUU1OjgoKCoIaKJBKvu2Gd3Im4ypvbLlNzZaLm0HXxmjtdKsiU3r878dy+f7SSb5dq26TCL6g5dF285o6L3gD0aNeT9l/Jf2FWMymcQ82h6+I1Nzm6gs67ridHWcnvlqm2tl2FhWuoOXRZh/dzSxOnTc9EOp+Z8S+xn0PXxWuuIlpzKxLPPWdN/dlf5hyiXKlbc6GYmeGLNyRFpH+6QhdVeFNaohczWgT4pFVmWmJxIvSadTFjrLwrtwLdUiMpQ9r3f12x1T01GKSEJpkWE/dt1u2ay1D8jutAt7VIapcc1x8KIuVWzk2rTW0Cfvha0VUPfxUP2R2cL0hqDm5ESHbrZWquJBHKto6tjZKs1VpTTq+/ASgAAAAAAIAbFzMAAAAAAECopEybyRctUl9J1e7gGm9OToZ5bHVk7iALdFeszcQ1/XXlYm/K/mL2NXy0QWbPfvnBidjhr3tz0gX4Z6PMvm6pK3a7lXNicMNBCvhGUrqZYh2T94KV84I4uMI3q1tMS/CQX/8sHlti5RwuU5P27g/oirZa08GZXrL5nEGScoIaUC/FzAwAAAAAABAqXMwAAAAAAAChkjJtJn1kbqa+jzvYaCVlRR9pM4Ff2mTqKT8Rsm+4XjBAZn1Wa5UToEs2KNpGMj0Rq6TNBDtQWnTLSISesI6vM4rEbf7hn3pJaVJe30Sood6bMk9SQ5BjQkqo/e/Ex3daz5WIBXTgnyZJmZLy1iVi9jlEZIwUaZP0WXDj6m2YmQEAAAAAAEKFixkAAAAAACBUUqbNpKJCKkiTytwrSdhzwTKjj46Ymwh/ODItJO2J0JF9rZxRMm1NtJnAD6skRSTpb4nYaiunXFzKhn9ifZyu+a95ds4UmdbOPwc1KCS1Fpl92PBEKM86hr69min/8E9ZRCqISMpOxNZZ7XQLZBZ2AvyQHd0aahKxkcVWUpnMOQRtJgAAAAAAAOHAxQwAAAAAABAqKdNmosmSMqWcMldssZWTG31sl1kRAOiuPjKXDNtcsf2tnHQxLxG+adhgSirvnQXx2Ma13pycYTItUIAf0qObazWTw4daOdMk1QY3JCS5epl2uhJXrNCbsm41C+jAP+scqcmRBo5KxKZ85M2ZK9MBBfghPU9Kj0iVrpWa8vazkhqV8itwMjMDAAAAAACEChczAAAAAABAqKROm0lsGs7FrtgsK2d09HGTOt79H+iKXJlLhumuWIaVU6eUnyIG/7RFN92diOVY069VJ2/rE9AdGdFtgSv2fStntUzdAT5oaIkeVl138M8b4M05UaYb5Q/BDQtJbIHMKk2HrknExlk5a8UKOvBRvqQ0qd3VZqL+Vs4bSvm2YWZmAAAAAACAUOFiBgAAAAAACJXUaTOZK3Pp5iZXzJ7yOij6yMoS8EuGzFzYSlcs28pZI7OCDuCDgjypICJ9emcitschVtJyUXPwT6bMvm6hK/Y7K+cacZt/+KZGZhe2yBWbbL2nG3eCVNsq6cngxoXkNVxm1r/73MHuSF8udnPwX777k0+8zy1vpIOTmRkAAAAAACBUuJgBAAAAAABCJWXaTBpWmNn+eStdQftu/o3RR9pM4JdWmTpb7opVWDlrlfJ3IoaPRkhKl/7ro0Ton7lWTrqkSIBjQnJrk/nTiHvfNvIOT8oLf/uBGoIcE5JabAEdz8oRw62kCdEE2kzgg2F9TAvnhsZEbJGVUyfaTOCj9ZIi3jaTtvnelDqZVZtSGTMzAAAAAABAqHAxAwAAAAAAhErKtJnEpiRqzhaS1kQfN+3w4SBVNEpKk5Y0J0Ijy62cdTK3Zd8Y3LCQxEolZUgvutpMVGXlVIvVTOCfWJvJd9zBCz0p/9QPmH4N35TkmCn/o11T/jWpx4aDFPDWJilP3gXpiq2csbJan4Bu2LDJHF77ZSRia6xbIZSo4yKJqYaZGQAAAAAAIFS4mAEAAAAAAEIlZdpMMkukzDRJD7uC6VZSrM2E6dfwS7OkiOkkiRm5l5XTKNPa9FZgo0IKGOX+JMN6ktVM4KdMmRrztNA96kkZLTrp4KORktKlnVa5YjtbOZni/Rx8s0hSjqRDXbHOOjgzgxoQkl6LzGmEu5+p1C66Eim3XYlz2BTEzAwAAAAAABAqXMwAAAAAAAChkjJtJuonKV1a8lkiNHKQlRObG9YW0JiQ/NokRaw7Da+wcipk5pLRZgI/NEtqky5xx+zbq4+VaW16NaAxIbnFVjP53BXb60RPyqmSaiX9JLhRIZnlyLyDdbc21Vg5TYrO0Qa6b4ykvvKW3MNWzjpRcvBPjqRcSSpzBS+wkl6TeT9HmwkAAAAAAEA4cDEDAAAAAACESuq0maSbbakrNNLOyY0+0mYCv+RKSpPK6l2x16ycAyW1BjckJLloa9P+7thiK+c4mdYmwA8NMquZrHXFrvSm9NtbSm+T9HFww0ISy5N5B+tele5dK2e1eD8H3xRKypd3Nr/dNdwgDq3wT0GBVBCR1N8VvKHAm3R4bcpPTUjxHx8AAAAAAIQNFzMAAAAAAECopE6bST9JfcwUsbgGK2d19LE9kBEhFUTbTIrcsXQr5z1Rc/BPqyRHqnSFnCpvSqRJtDbBP9E2znirpiQ9beUcLlNztJnAD6UyK9C5e4cXelNqV5gVdAA/rJTpbnIvhLirlbNarGYCHw2UObZmuYPWqXuTzGomKYyZGQAAAAAAIFS4mAEAAAAAAEIlddpMhkrKkIa7Y3nelI3RO7FvDGZESAXpktLMLLCYzFO9KRv/l5qDjzZJcqRsVyiSYeVUiTYT+KdQZsq/q53pw/XelG8NE/Ov4Z92mZVKFrliRd6UT9SxmxjoqixZs/3VoeTUJO/7PaBb0qKb+/3a76y+4Wql/KpNzMwAAAAAAAChwsUMAAAAAAAQKqnTZtIkqU0qHeCKlVkpaxOpgC+a1PGS4fe9n+b8UWp1lPJ3I4ZPotMN993PFfvcyqkT9Qb/ZMu0mbybCL1opXyrTPTTwT9LJKVLq9YmQkOKvSlviPdz8M9AmRUR33PFPrRy8kQ3HXzUrI5tJrOtnHyl/IqIzMwAAAAAAAChwsUMAAAAAAAQKqnTZtIoKUPS/q7YJ96U2OImKX5TWPgpOvWrwD39dZWVUxjNs25QDHRJq0w9neeKXWHlVIodHfzTEn2cmwhNsHOyJTnBDAcpoEZSuumYi7OOoW1iNwf/1EcfF7hi9tu2ViV2h0C3Zcmsiujql1turRQ2vEhSJLgh9UbMzAAAAAAAAKHCxQwAAAAAABAqqdNmkhbdDk2Eap/0psRaATIdSRsCGheSW4HMFLEKV+w2b4qzntnX8FGjzL7O3dqUbeXUKOXvfg0f1UrKkJa47rh+0FVWzgpxm3/4J19SulTijlk9JcfLtAb8IrBBIZk1y5w0DXPF7L8IN4ndHHw0SKbolidC5XZOemCj6bWYmQEAAAAAAEKFixkAAAAAACBUUqfNpFJmKk5+IrTOSimItQK0iTYT+GOAzP+y4YlQ29+8KStl3ZEd6I7Y3a/d86/3snKWijYT+OdLSenWLP9LrJwfy9zqH/DDEEkZUn/3G7kmb8puuVKtI2ljgONC0kqTObSOcsXs9251YtY/fJQhcw6xNhFKH2/lrFDKv59jZgYAAAAAAAgVLmYAAAAAAIBQSZ02ky9lLt38KxFqs3NityhulfRBEINC0suW+V+WkQilW5cQm9q5+zV8lCczz/UgV+xQK6dBZge4MqhBIamtl5Qm7eLet71i5cRWFAP8kCtzXJ3iilktnBtb6TCBfxolRWS6NDcnQ52cWwBd9Y3M+zn3sdNu16xSyi+JyFsLAAAAAAAQKlzMAAAAAAAAoZIybSZV68zMnP4vumJ2Unb0kVsRwy91Mv/Lql2xMm/KqK+k2uBGhGTXKnNna1drkwZZOXmSNgU2IiS7bJk/jRznil1u5RyqlL/jOnxUJ7OPG5cItVjTr3PSpFZHKT8FG/6ol9mFubszS6ycUpl2FMAXq2WOrUWJUJt1G4T0DKU8ZmYAAAAAAIBQ4WIGAAAAAAAIlZRpM1kvqUlSkau3pMNUsBXRR6Zfwy9r1OGS4ZqvvJ+XjpHS2yR9FtSgkNTqZGruIVesQ08d4KOBMu8mLk6EXnjUm3L4EnGbf/hnjUxLcEMilGm1cGqYzPu5dwMbFZJYoUyHZrkrZu/ShspTkkD31MssoZOZCNmL0A0vU8f+pxTDzAwAAAAAABAqXMwAAAAAAAChkjJtJrtEpIKItMF1N/XRdtLy6CN3XIdf1spMEctNhG6wUv6UIS4rwj+x/dfvXbH9e2IgSBl9ZVaWmLxPPFSked6cdnFshX+WyxxbV7hiO1s55TKrO9FmAh8Uyuzqhrtiy62cRrGaCXyULrOfcy15WG7nlMm009FmAgAAAAAAEA49ejFj9uzZGjdunPLz8zVw4EDNmDFDCxcu9ORMnjxZkUjEs/3gBz/ooREj7Kg59ATqDkGj5hA0ag5Bo+YQNGqu9+nRNpPXXntNM2fO1Lhx47Rp0yb913/9l4444ggtWLBAeXl58bzzzz9fN9yQmJyfm5vb2ZfbshGS0qT0xYlQabGV0249IukEWnNu2YkP77Oe+tNaUXNJLtC6K5SULn35XiK0U4WVky7J2f4vjfDomX3dAfGPdrbbTHLFSmFJLsiaa9kgtUjKXOcKpltJeTJJSFpB1lyTzEmTewWTEiunQbSZJLvA38+lSapMhNLtL9OklF8prEcvZjz33HOez++9914NHDhQ77//vg4++OB4PDc3V6Wlpdv0NZubm9Xc3Bz/vLa2dgvZSDU7ouYk6g5bxr4OQaPmEDRqDkGj5hA0aq736VX3zKipqZEkFRd7p0z8/e9/14ABAzRmzBhdffXVamzc/HXP2bNnq7CwML6Vl3e4VQoQ50fNSdQdtg/7OgSNmkPQqDkEjZpD0Ki5nhdxHKdXTDZub2/Xscceq+rqar3xxhvx+J133qlhw4aprKxMH3/8sa688kqNHz9ejz32WKdfp7OrW+Xl5aopkwrSZO5sHbOf9eLV0de0SYUfmwItKCjw5wdEr+NXzUlbqLviaN25Ls7uM9/72nm5Uq0jFW6k5lLBDt/XTZEKMqSnnk3kHnOE9eKq6H7uA2ouFezwmjtUKugj6V+Hx5/7JvKC57UDd4/W3BfUXCrY0TW3RlKBpJwDXcn2n+dGS7UtUuFfqblUsKNrbo5M51KTK9euqGqZNpPTRc2lgh1+bB0WPYdYm8h937omsm9h9ByiNnVrrtcszTpz5kzNnz/fUwySdMEFF8Q/Hjt2rAYPHqzDDjtMS5Ys0ciRIzt8naysLGVlZe3w8SL8/Ko5ibrDtmNfh6BRcwgaNYegUXMIGjXXO/SKNpOLL75YTz/9tF555RUNHTp0i7kTJkyQJC1evHiLecCWUHPoCdQdgkbNIWjUHIJGzSFo1Fzv0aMzMxzH0axZs/T444/r1Vdf1YgRI7b6mg8//FCSNHjw4O36XhtWm5u99nPX29VW0q+ij61Ckgqy5iSZu6unSVqUCP0/OydbZmWJjdv/5REOgdbdYEmZUr47ZrdqrhMr6CS5QGuur6QMSRsSrSV1VsrAfKX8HdeTXZA11xrdcpa6gntbSfPFCjpJLsiaq5c5bLoWp+uwWE6TvG0oSD6BHltbZc4hMhKh1VbK8JqOx9tU06MXM2bOnKkHH3xQ//znP5Wfn681a9ZIkgoLC5WTk6MlS5bowQcf1PTp09W/f399/PHHuuyyy3TwwQdrzz337MmhI6SoOfQE6g5Bo+YQNGoOQaPmEDRqrhdyepDM36M7bPfcc4/jOI7z5ZdfOgcffLBTXFzsZGVlOTvvvLNzxRVXODU1Ndv8PWpqahxJznLJqZIcZ6hre8PappmtZooZx/Z8H4RDEDXnOIm6qymR4wyS42QltoXybk6xnJp+1FwyC3JfV3O6HOdcOa8osTkHWtswOTXl1FwyC7TmjpXjnCDHqUpsi+XdnPFyaval5pJZkDW3UnJqJMcpc21HWduBcmq+Tc0lsyBr7l7JeURynnRtz1rb3yXnL6Lmklmgx9ay6PlqYWJ70qrB9ZKzLMVrrsfbTLakvLxcr732mi/f6zOZuxCXf5WIDT/AumVIXXTeNdMSk1aQNSfJTKt2pDWJmxRrl3usnLtlau4t/74tepdA666vpCxp8gBXzJ6D2CvuloQdKdCay5aZBrulL3exTCvdhf58S/Q+QdZcvaSIpII1rmCmN+eTN0wekleQNdce3fJcsVorJz26IXkFemxtknm/5updGmelrBP7Od7SAgAAAACAUOFiBgAAAAAACJUebTMJ0gCZ2defu2LDP7Ru5x+brshd/uGXfElpUv56V+ysQ7w5b79ibolNmwn8UC9TT991xeZZOcUyrU0rghoUklqTTEvdi4lQqZ1z+p5SbZt04afBjQtJKyO6qdgVtJaR+GfHENBlbeq4INNS6/Ny8Vdi+G+Vq1V9yADvc7XrqblU//kBAAAAAEDIcDEDAAAAAACESsq0mexSIRWkSWmLXcF3raTY/DHaTOCXQknpUl6ZO9jXm1MkqVmAPzbIzL8+2RVbZ+WkS2oNbERIdrE2k38lQnl97aQ9ZIqONhN0X0meVBCR99b+1d6cQTIL6AB+aJb5C7B7BZPFVk6RqDn4KFNSmrTaHVvvTSmRlBXciHolZmYAAAAAAIBQ4WIGAAAAAAAIlZRpM1GFpAxpF/eth1dbOZnRR/t2xUB3ue+4/tRT3uc6u0U20FXrZfbsFa5YvpVTLdpM4LsG15zrDm0mXz8k1QU6HCSzITLtcmNcsSe9KYeKkoN/WmUOrY2u2NBOcji0wjctktJMK0nMHCvlRxlSuiOzQl2KYmYGAAAAAAAIFS5mAAAAAACAUOFiBgAAAAAACJXUuWdGs6Q2aYlr2dWRi6ycyugjS7PCLxtkLhnmumLfs3LOEUuzwj/rZXrJq1yxNVZOnVK6vxI+y5OUYVZojYeGWDl/FPs5+KdY5h1spitmLUE9sliqdWSOw0A3pUW3bFdstJXTKu89NYBu2SgpIg0fkQitW2blpElyAhxTL8TMDAAAAAAAECpczAAAAAAAAKGSOm0mayWlSwtdoZHzvSmV680jS3nBNzWSIvLMS3yi3psyo1Ks5QX/tMos9etehnqBlVMglgOGf/pKyvROv9aJVs6/RM3BP/0kZUhamQitqvKmDLlOpvfpxsBGhSRWKNMx7P4rsP3WrURSQ2AjQtJrkzmHmJ4Iff82Kydbps2kJbBR9TrMzAAAAAAAAKHCxQwAAAAAABAqqdNmskhSRNrZFXI+8qasjj5aXQBA16XLTBFz3ea/v52zUqwsAf9kyNSde/6rNf1aJWLKP/yToY6rmUy0ch4UK4XBP9Gac7fTPW2lXPjzfaTaNulG680e0AV9ZRZuqnHFqq2cMpmyBHzRKnMOUZEI7XaalfO2zLG1NrBR9TrMzAAAAAAAAKHCxQwAAAAAABAqKdNmstqR6hxpl+JEbIk19bow+sgVHvimj0xBuab8H2TPQVwtpl/DP83quBPLsz4vEq1N8E90Kuw6V6j/OiunVKbmVgQ2KiSzDTLH11WJ0GtWyoWaKrNDpM0E3efIvFVzv4VrsnJaxeJ08I/jmC3iDo6zkt4OcEC9FOftAAAAAAAgVLiYAQAAAAAAQiVl2kzaopt2TcTS53pzdope2ql1ZOaTAd3VLDM/zD3N32ozaVvMwhLwUZVMza10xezWpmwxFxb+ibaZeErqAyunSNQc/FOvDqs2Zds5H89meTr4pkXmpGmQK7bSyqHNBH6KFEiRiLa8RE6VUv6clZkZAAAAAAAgVLiYAQAAAAAAQiVl2kyyo5vKErHh9rSd2KUdR2Y+GdBNbbWmhSS9KBH7stGbs1RSQ4BjQop41/Wxfdm6UaxmAv9kmK3EHXvKyukv+ungP9dyEifbz/1A7Ofgm0EyHcPudqZyK6dVlBx81Ffm/Vu6K2atxKkG0WbS0wMAAAAAAADYHlzMAAAAAAAAoZIybSYlfaSCiKR2V3ColRSbukObCXzSJClTUp6rpcnqMlFVJzGgy/IlpUlrXk+ESg+0cprEXFj4J91spa4/jyxf5k0Z3i7v8RfojgZJ6VLD+kToyKu8KWtulOoCHRSSWanM4XWhK7aXlbNQgI82yqxOt9wV+8jKaRdtJj09AAAAAAAAgO3BxQwAAAAAABAqKdNmop1kLt2420fsKa+l0cc2SbVBDArJLi9HyotIKkrE7GmvwyTVBzckJLt+ktKlZ1YnQueUWDnVMlMXAT+0RbdRiVCGPd+6KJqzMqhBIaltkJTmLafdZl/jSVl34y84tsJ3i1wfl1rPZYtFm+AfZ4PpIIk87wp+YiWlK+UxMwMAAAAAAIQKFzMAAAAAAECopE6bSUyz6+Mm67ny6OMmSYuDGQ6SXLnMFLDsRGiBlXKq6GqCjwol9ZFa3TF7daZYWwDgh1xJWZLGJEJDKq2cNKX8HdfhP89+Tid7PivUL/iLHXyzRqZNuNoVW27lFEnKEOCPRpkT9WrXCibPWDnfbzeLnqQy9vMAAAAAACBUuJgBAAAAAABCJfXaTBpcH1dbz+VHH1sF+CNfps2kOhGyZ/xnFkuZjszd2YHuypbUR9rHHWu2cmpEmwn8M1Cm7gpdsaXelA0f0E4HH+Wqkz/Hnev5rFW8nYN/vpQpuwJXzF5IIk38lRj+idWT+3D6Wytnujqukphq+D8HAAAAAABChYsZAAAAAAAgVFKnzaSfOqwq0WZNvbaniwHd1ihTWCsToUl2Tr6kdtFmAn/0kZQhjSt2xfKtnOUyNQf4yX0Qteb3vydvlyfQLa2S0qxuuYvf8aSsFjUH/zTKLMiU64rZK5dkyyyICPghPbq562y1lZMvFgpjZgYAAAAAAAgVLmYAAAAAAIBQSZ02k1yZn3ZYImSvKpFTHf2AOWLwS72kNMmpSYR2mWblVMrMlV0poPtyZeYkTnDFSqycIrGaCfyzXFKmpDWumDW/v1Rmdwj4YqOkiHc39uVtPTUYpJJG18d2G1O2TGkCfuhsRabvWJ8XFMv0maRwqzozMwAAAAAAQKhwMQMAAAAAAIRK6rSZNMn8tPskQjlPWjnV0UemX8Mv0Tuuu6eJZdptJm9E894PbFRIZgNkpvyvc8XWWDkl6jh3EeiqJTLH149cseO8KWObpdp2daxFoCv6SErzLqBTZaWUS6oLbkRIcrnRzd1mss7KSbeeB/zgXsHkJvvJMpnzVtpMAAAAAAAAwoGLGQAAAAAAIFRSp81kg8z8r71csaOtnOXRx02SvghgTEh+aerYZrLQyslVx6V1gK5qj26ZrthHVs5YeednA91RKyldanGt2pT5ByvnHJkdod3eCXRFtjq0mQyyUjIkRYIbEZJcRnQrdMVqrJzOVp8Auip6CuFpZxrykJV0j8x562dBjar3YWYGAAAAAAAIFS5mAAAAAACAUEmdNpNmmUs3a12xcisndmmnRdKbQQwKSa/dPGS4Yy9aOfuIeYnwzxcye/ZRrpjdUtIgMy0R8EOWpHQpM9cVK7jOm1NyHe108E+NpIj3L3JNVkqGAP9USdoo79u1fCuHNhP4KSdHyolIZe4lck55xZv05CEpX3TMzAAAAAAAAKHCxQwAAAAAABAqqdNmkiFz6eZOV8yeH5YdfUzx6TrwUZ2kiJTpvmy4ysrZWdQc/LNSZl/nrqkKK6dSUltgI0Ky6ydzjB3mDj7mzYmtsgP4oKXWdC25u4UXWTk7ZUjpjmipgy/qZUqpyhXbx8ppEG/n4KNCSWnSOHebiSZ7cyg6ZmYAAAAAAIBw4WIGAAAAAAAIldRpMymV+Wndq5QMsnJKoo9Mv4ZP2hpMOaW7b6ueaSWtFTUH/+RISpc+nZsI7XGslcOUf/gpV6bNZKordsfH3pw1SvmpsPBPW3QrGJqINX1lJe0dTXo/sGEhiaVHN/eqOfYuLVN0NcFHLZLSpFLP0ky7enOqlPJFx8wMAAAAAAAQKlzMAAAAAAAAoZI6bSaxabCuu/q3zPemZMamXTP9Gn5zTxEbZj23SJIT4FiQ3HaSlCE98lkidH2VlZMmag7+ic2/vtoVO8vKWS3a6eCbzOim/olYs91mso/MNG3aTOCDPjJv5dJdsQYrp9R6HugWR1K75Lj6mSLHf+HNaVDKH1t7dGbGddddp0gk4tl22223+PNNTU2aOXOm+vfvr759++qEE07Q2rVre3DECDtqDj2BukPQqDkEjZpD0Kg5BI2a6316vM1kjz320Ndffx3f3njjjfhzl112mZ566in94x//0GuvvabVq1fr+OOP78HRIhlQc+gJ1B2CRs0haNQcgkbNIWjUXO/S420mffr0UWlpaYd4TU2N7r77bj344IM69NBDJUn33HOPdt99d7311lv69re/vX3fqFHmpy1JhOZZKd9eF/2A6ddJLbCak5SeJ6VHJBW4gsO9OS0fmJmwSG6B1d0ASZnSKneszsrJFfu5FBBYzTXIzL8eOCsRq77Fm5OllJ8KmwqCqrnYaibproNnh+NohbxLTyApBVVz+TKLheW7Yo1dHzZCLLBja4mkdKnJ1Sr8r8e9KTPGKOWPrT0+M2PRokUqKytTRUWFTjvtNH355ZeSpPfff1+tra2aMmVKPHe33XbTTjvtpLlz527uy6m5uVm1tbWeDXDzu+Yk6g5bx74OQaPmEDRqDkGj5hA0aq536dGLGRMmTNC9996r5557TrfffruWLVumgw46SHV1dVqzZo0yMzNVVFTkec2gQYO0Zs2azX7N2bNnq7CwML6Vl5fv4J8CYbIjak6i7rBl7OsQNGoOQaPmEDRqDkGj5nqfHm0zmTZtWvzjPffcUxMmTNCwYcP0yCOPKCcnp0tf8+qrr9bll18e/7y2ttYURaHMNNhFidxPrNdOiM4XY/Z18toRNSdtpe7SJBW5kq3pYG0dQ0gyge7rojyHQnuqdbGkTV36tgiJQGuuVh3fTbRan4+KxlhZImkFWXOZESkzIs9qJh30l7SxS98WIRFkzfWV6dAc5cq1T0+bRGdTsgv02FomqY+U7lqdzl6cDr2gzcStqKhIu+yyixYvXqzS0lK1tLSourrak7N27dpO+5RisrKyVFBQ4NmAzfGj5iTqDtuHfR2CRs0haNQcgkbNIWjUXM/rVRcz6uvrtWTJEg0ePFj77ruvMjIy9NJLL8WfX7hwob788ktNnDixB0eJZELNoSdQdwgaNYegUXMIGjWHoFFzPa9H20x+8pOf6JhjjtGwYcO0evVq/fznP1d6erpOPfVUFRYW6txzz9Xll1+u4uJiFRQUaNasWZo4cWKXVpVQP0mZklYmQs1WSnX0kduuJK9Aa06SsmUuGRa7YtYcxPbohuQVaN3VSMqQvuOO2auZtIo2kyQXaM01y9TTJtcKJiutnAli2aYkF2jNFUmKSErcZ0/937BylqrjGz0klSBrLldSnryL09kaRGdTsgt0P9cgqY+3FX2qnUOves9ezPjqq6906qmnqrKyUiUlJTrwwAP11ltvqaTErJ/6+9//XmlpaTrhhBPU3NysqVOn6k9/+lNPDhkhR82hJ1B3CBo1h6BRcwgaNYegUXO9T8RxnKS+32Vtba0KCwtV8z2pIFPSnMRzt9Z7c0+LvUbScJn1gulbQlfE665CKkiTuYlPTLY3t+F5U3NloubQdfGamyYVZEgfPpl47ltlVnKFVLtJKnyLmkPXxWtub6kgXZJ75bmhVvKJUm2LVPgXag5dF6+5flJBRNIliefevc6bO+6/pdpmqfB/qDl0XazmHpeZmZHuem6dldsiMzPjQlFz6Lr4fm6cVNBH2ug6tto3AB2yu1TbJhV+kbo116MzMwK1XFIfaaPrAkaJlRIrEHtGNtBlaTJHviJXLMObwgwx+GqDpD7St1zTr/WylZMvig7+KZR5N/FIItSy1puSmS3TFgD4IUvm+Oqac73Xdd6Utl+xm4N/MqKbuy14tJXzrjou5AR02QJJEanRFcq3c2qU8r3qveoGoAAAAAAAAFvDxQwAAAAAABAqqdVmkibluFaVyLAajzpM3QG6K1umzSTPFWvwptBmAl+1yUznvzQR2vCiN6VfpVJ+WiJ8lCkz//pNV2iMlWPPzwa6o05mP+daISdzgDflvPUsoAP/uVc83Mt6Lk9008E/Gxo6nh/Y98zIXs1+jpkZAAAAAAAgVLiYAQAAAAAAQiV12kxi01sPTIQmPOlNGRidL5bdJml+EINC0suQaTNxzxNr9Kaky7vUF9At2TJ7dtdd/vM6y2HKP/xSK1NzH7lit1g5c8RcWPhmY4M5vOY8m4hVrvfmzBUtnPBPe3Rb7YpVWjmxhZ0AP/TrY5agdlxL5NizEN5Vh+71lMPMDAAAAAAAECpczAAAAAAAAKGSOrOhBsrM5d8/ERqy2sqJPdcs2kzgj1aZeYl1rliWN6UgS5IjpmDDH7G+pbWJUOZEK6dRZv61vQ8EumK9TM2tccUmH+LN+eMrZn8I+GCjzBvYHNdKTe9aOUfIvJ1bGNiokAqWuj62D6EdWjqB7thJUpoUcRVagdWqvkIdutdTDjMzAAAAAABAqHAxAwAAAAAAhErqtJkUy/y0g1yxEisndttr7vIPvzTKXDJ03/LarruxMrX3QVCDQlJrkNnX/cEV+66V87CkTYGNCCmisibxcX+97X2yQdQcfNNHZjUT9zz/KivnQkn1km4PalBIai0yNefulsu2chpFNx18lKsOyx2usVKK1aF7PeUwMwMAAAAAAIQKFzMAAAAAAECopE6byVcyU3U+csXs+WGLoo9MhYVfWmUuGbpveZ1p5YyK5tFmAj80y+zDHnHF7rVynhb7OfhnsKQ+UqVr2Yj+f7Lur14nag6+KciTCiLemP2WbrehUm27WLUJvtgoKSKp3BWzVy+pU6JjHei2DJlz19xE6BPr0Dpa3gUTUxEzMwAAAAAAQKhwMQMAAAAAAIRKyrSZVC02M/n7t7iC+1tJsTYT5ojBL7E2E/dqJtadiVUobn8N/2TJ7Nndl6rtadZp4lI2/DNcUqaU+Yor9qKVs06sFAb/jJA5lrqOraPsnNEyrU20mcAHNTIrmrjbTJqsnFxJTmAjQtKLvVdzrYI4f7035cgMqdZRSrdx8nYWAAAAAACEChczAAAAAABAqKRMm8mnMncdnrAsEcsZbiXZ0/+B7mqTmXOY4YqtsXLSxPRr+CdPZs8+0hV728ppFzUH/xRKyvLMhLU+kVQsWjjhn34y+7nlidDYvlbOcJm+AMAHNTKLhbl3bRlWTps4lYCP6mQKKj8RWmfnDJJ5P5fC7XTMzAAAAAAAAKHCxQwAAAAAABAqKdNmslJSjqSxrljO5qbkMBUWfslQhzsRO8u8KZEqcftr+K/C9fE867kMsZ+Df6J3XM9z/3nEbqcbLTPl//3ARoVk1iSpj7ShJhHqd7CVky0pEuCYkNRyZc4j3LKtzxvEoRU+2ihzfM1MhHa1c4pkio42EwAAAAAAgHDgYgYAAAAAAAiVlGkzaZG5IaxnSliFlRS74z9T/uGXbJlLhnmJUKuVkpktVpaAf2pkdnbFrph9R3/7FuyAH4oSHy5/0vvU8Jmi7uCf6H5uqSu0b6mV86GkTYGNCEku1mbiXq3E7qbLEwvowEdNMucQriVMvmPnxNrZU1iK//gAAAAAACBsuJgBAAAAAABCJWXaTIbITP/yXL3Zx0qK3fGfKf/wS5a8cxIlVVkppW2i5uCf5TJ38Hff2dre1y0Vd/mHf9Kj286J0GPveFMuX6mOPXZAV62XlOaZfS1Ve1MqX5fqghsRklyNTAuJa3E6LbJyxoq3c/DRJpn3aq5jZ/9DrJzVSvkldJiZAQAAAAAAQoWLGQAAAAAAIFRSps1kmKS+8k45zCm2kmJLnTBHDH7JlZl+7So8+07XLeu5+zV81CAzLfFzV2yqlbNUgH/aotvoRKjcajNhKiz81FZlysnzds3q4VwqqT64ISHJfSPTObzXFnJWS2oMZjhIBS0y7+eKXLGjrZzblfLnrczMAAAAAAAAocLFDAAAAAAAECop02ZSLqlA0gpXbKB9m2su7cBvaTJtJk2JUJGVUiXuuA4fDZapO/dqJiutnCox5R/+2SApU56Vm066wMr5XObO7IAPVsscN3PdwSJvzrgMqdYRdQdfNMiUkrvm+ls51ZI2BjUgJL2WBtNpkpnnCpZZSdlK+fdznL4DAAAAAIBQ4WIGAAAAAAAIlZRpM8kpk3LSpNavXMGFVlJr9DHF7woLH22UmXrdnAgV2JcQ25kFCx+NkNmzL3DFPrByGsR+Dv6pkZShxIpgkvTAqd6c4x9KHGOBblov071Z5A7a068nyRxcXw1kSEhyhTK7OPeh096lxRZ2AvywXub0oWRtIpb5tpVE0TEzAwAAAAAAhAsXMwAAAAAAQKikTJuJBktKl4a520ze8qY40eecoMaE5Fcnc8nQPQWsyJtSkCEzb3FdUINCUsuRmfI/xhV70copE20m8J9nzvUA73O59vNA19XKHFaHuoP5VtJQmaUAAB98S2Y3VuKK2TP+85XyM/7hoxyZmvPU1N+spNg5RApjZgYAAAAAAAgVLmYAAAAAAIBQSZ02k1ZJ7VKeK1S7zJtSZz0C3ZYuc8nQPb06w8opl5lDRpsJ/LBeZs/+nURow/PelH72LdmB7siW2a81umJv3eLNGS7Pqk5Ad8RKzdO51GQljZNZUezBIEaEZDdSUl9Jw1yxaiunTOZtH+CHfsXRFRBdK4V9/pU3J1+ctzIzAwAAAAAAhAoXMwAAAAAAQKikTptJozrM/bJvrN5mPQLdlitTd0tdsWIrJ00soQP/rJSpKde0xA7TXrnlOvwUO5i6a+oGK+dUmSn/gA9yZdqGPdOr7VbNUZIaghoRUkXEtVBTw3rvc/nir8TwUbHMGzjXEjpLrTaTdeLQyv85AAAAAAAQKlzMAAAAAAAAoZI6bSYZktKl9KGJUH/rbv7Vq82j3X4CdFlfSX2kNa5aKy23curElH/4pm2NKaf0dxOxAnsFnWxJmwIcFJJbbAUdd53Nt3LOEAdX+GawzOHVs4DJUivpBbGCDnyTFt2Un4g1Wm0mdfIu6gR0S5VM0RUmQvaiTfG6TGGp/vMDAAAAAICQ4WIGAAAAAAAIldRpMxkg89O6V5Kw5urE2kzqAxoSUkCbpIhU7QqVllo5CyRZLU9AV62XmVldusIVHGolpYsVdOCf2GphRa5YoZVTJ265Dt8MkJnt7+5sarFam9rmU3LwT2wFHbmOrfZfhJerYxsA0FVVVaY7s39NIraflbPTAKm2XbqwKsiR9S7MzAAAAAAAAKHCxQwAAAAAABAqqdNmMkZSlqR5rli2NyXdegS6bY2kNGuxkkorp0m0mcA3TYpOvc51Be0VdCQuZcM/aTIHTtdd/jusLDFPUktgI0KS6yepQFLE1UL3xVfenPfEyhLw37uu92tF1nOtYtEm+OcrmVWbKl1FtUuWlfQzmTd+VwU2rF6Ht7MAAAAAACBUuJgBAAAAAABCJXXaTA6UmXb9iCs2yJsyOvpYG8yIkAKclZ0sGrHY+jxNtJnANwXRTbu5gm9ZSVaLHdAt2erwbuKtZu/ne93ByhLwT6RUiqTJs58rtdpM0kXbMPyzQaaFpM4VK7Fy0sV+Dv7Lc338jXVsHViilO+nY2YGAAAAAAAIFS5mAAAAAACAUEmdNpNRMreEbXDFrLutZ46IPrZLWhHMsJDc6iRF5L3jdW2VN6dgqGgzgW/6lUkFaZJ2dQWfspKGiJqDf2Lz+fM3n/KkUn4mLPxUIfMOdlgiVLCXN2XSR96WAKA7GmXez7n/Clxq5VRL2hTUgJD0BskcVksLE7GXa7w5hz6mlF9Cp0dnZgwfPlyRSKTDNnPmTEnS5MmTOzz3gx/8oCeHjCRA3SFo1ByCRs0haNQcgkbNIWjUXO/TozMz3n33XbW1tcU/nz9/vg4//HCddNJJ8dj555+vG264If55bm5uoGNE8qHuEDRqDkGj5hA0ag5Bo+YQNGqu9+nRixklJd77AN94440aOXKkJk2aFI/l5uaqtNSeyNUFdTLLSri+1MaF3pSc4u5/G/R+QdZdXnQr6JuIbazvJKlNSGKB7uv2kpQhs3pTTKaVky3mwia5QGsuU6bm9k+Esu/1pmRENySvQGuuTKag3Lf539WbUlor5dI2nNSCrLm+0W2BK2aVnKq7/V3Q2wVZc4MiUkFEnkIre8eb4zzayaqJKabX3AC0paVFDzzwgM455xxFIpF4/O9//7sGDBigMWPG6Oqrr1Zj45a7bpubm1VbW+vZgM2h7hA0ag5Bo+YQNGoOQaPmEDRqrnfoNTcAfeKJJ1RdXa2zzjorHvve976nYcOGqaysTB9//LGuvPJKLVy4UI899thmv87s2bN1/fXXBzBiJAPqDkGj5hA0ag5Bo+YQNGoOQaPmeoeI4zi9YnbK1KlTlZmZqaeesm+7n/Dyyy/rsMMO0+LFizVy5MhOc5qbm9Xc3Bz/vLa2VuXl5ar5g1SQI1VekMgtsl6bHr0rdm27VLhSqqmpUUFBQRd/IoTBDq+7kujKEqNcyfOsF4+RatukwvepuVSww2vuMqkgS97brD9gvbhCqm2VCh+n5lLBDq+5aVJBhqTLErmvHuJ97V6SaiUNFzWXCnZ4zZ0uFWTK26KZbb34leixdTE1lwp2dM3Nl1lZ4gZX7k+s186TWfXkfFFzqWCH7+cKom0mkzp9mSTpmyfNnRR2VurWXK+YmbFixQq9+OKLW7xqJUkTJkyQpC0WRFZWlrKysnwfI5IPdYegUXMIGjWHoFFzCBo1h6BRc71Hr7hnxj333KOBAwfqqKOO2mLehx9+KEkaPHhwAKNCsqPuEDRqDkGj5hA0ag5Bo+YQNGqu9+jxmRnt7e265557dOaZZ6pPn8RwlixZogcffFDTp09X//799fHHH+uyyy7TwQcfrD333HP7v9FcSZnSla7QXSOsnNgNatskrdz+b4HwCKzuCiWlS6pwxT6xcjLFyhIpILCa6ycz3XqpK7bOytlHUsv2f2mES2A1lx7dXDU3wUrJOU1Kb5X0SBd+EIRGYDUXWx6n2hXL86a0LWShsFQQVM1lRjf3obXIyilSx8XDkHwC289lqeO0g99ZY3lSat/+r5xUevxixosvvqgvv/xS55xzjieemZmpF198UTfffLMaGhpUXl6uE044QT/72c96aKRIJtQdgkbNIWjUHIJGzSFo1ByCRs31Lj1+MeOII45QZ/cgLS8v12uvvdYDI0IqoO4QNGoOQaPmEDRqDkGj5hA0aq536fGLGYF5XVKa9JArdFeplTM8+tgq6f0gBoWkF5v7VeSKjbVyGsRcWPinPbq525nsNpMm0WYC/7TJTIV1HWBzTrZypsvc5p82E/ghtp9zz+m3VjOplrnLP+CHJpnOJvepg90B0CbezsFHGTJF5t6RjRztSanUAtUHOaZeqFfcABQAAAAAAGBbcTEDAAAAAACESsq0maxdbWa4emYhVltJsSV+I0GMCCmhVR0vGe5jff5PcSti+KdGZj7s21vIWSRW0IH//u36+IfWc21iPwf/1MtMwW5yxYq9Kf37ShmOTCsn0E3rZcptlCvWZOWslrQxsBEhZbjbTL5Y4HkqtrBTKmNmBgAAAAAACBUuZgAAAAAAgFBJmTaT1ZL6SjrVHVxhJZVHH5l+Db/E7kTsnoto3+p6naSOKzwBXbNWUqb078ZE6CBr+rVWiin/8F+66+N867kFkpoDHAuSW1p0c8+4HmblZMocW2kzgQ+qZHZhWa5YjZXToI6tJ0CXxc4h3Puw270puwyQattlCjRFMTMDAAAAAACEChczAAAAAABAqKRMm8kmmYUlrnbFKhu9Of1jbSd2GwDQVX1kpl67p4gt8qa0NEotAQ4JSW6lpD7eMuvQZiLRZgL/NMscNytcsf5Wzjyxo4N/CmXaSJa7Ym94U5wqOjjhn0qZVRHdC0tUWzklYjUT+KhYHc8hXrFyDpc5wZ0T1KB6H2ZmAAAAAACAUOFiBgAAAAAACJWUaTPJlZQnaci0RGz5s96c4oXmkWmJ8M0mmen8axOhtnd6ajBICa2SHGmcOzbcyqkU7XTwT6PMu4lsV8xeQWJXcZt/+KefzLISrpprec+bkrmrFGmTtDjAcSFpfS1TbktdMXuXNlwsngMf9ZM5tra6YkusnClK+ZXCmJkBAAAAAABChYsZAAAAAAAgVFKmzSQtuun7iVib1WayOvpYJ8AndTKFty4RshYz0W59pUxHzE2EP/pKypDGHuWKZVo5a0SbCfznngpbZD13tMz+8BeBjQbJLENmv1aeCK2r8aYMOVWmD+DG4IaF5PW1TGeT+z2c/bYtX6Y0AV+0RzfXSmAb670pOUvlPfamIGZmAAAAAACAUOFiBgAAAAAACJWUaTNpVPTKjWs1k3QrpzT6mBvIiJASNkmKSKpNhIbZOaUy08iW2k8AXdAu00LiaqnTfVZOm2gzgX9yZd5NVLpiB1o52Ur5qbDwUY1MC0lZIlQ038oplrQxuCEhuW2SOY9od8VWWDn9xW4OPlovc7LamAjNtVIOfU0pvwwnMzMAAAAAAECocDEDAAAAAACESsq0mYyQucuw8hKxYisnPTpdMb1d5m7/gF+aEx/m7NXJ80z5h1/qZPbs7luu2z11+aLm4J/oCjrKdsWyrZw7ZdoCAD+sl1nNZHgilGcvI3G32M/BNyUyu7VdXbF1Vk62TDsK4IvYPRJcUw/mWSkHVnkWO0lJzMwAAAAAAAChwsUMAAAAAAAQKinTZlJcIRWkSborEbNnwaoh+pjid4WFj9pkVjNxT6/Ot3JWyXt7bMAHDdclPs47znqyVNxyHf7pKzPl37WyhO6xct4Q86/hP/ecf+vYuuoj03UH+KFYUo6ksa6YfR5Bmwl8lSvTJuxql7NvkZA5QMpsl1QV3LB6G2ZmAAAAAACAUOFiBgAAAAAACJWUaTPREEl9pFUzE6FSK2VDjXmsDWpMSH6xNpNGV8y+/XWNaG2CfzIk9ZEecIUutO/ynyvaTOCfzOjW3xWz20xaxX4O/nPP8x/kfWpIm1TriDd18EWdzG6swhWzD6PrlOhYB7otQ6bNpDIRGm3nDJE516DNBAAAAAAAIBy4mAEAAAAAAEIlddpM0s32K1foT7nelHXRVoD6oMaE5JcuKSJtdM1FzFlq5RSK1Uzgn+i+zr2whIqsnDWizQT+aYtu7j+PWHOta+uZ7Q8frZJ5B+uuuRIrp0JmP/d8UINCMtso81bN3WZiv51bFc0DfBFdxaTlq0RonJ2ToZSfmpDiPz4AAAAAAAgbLmYAAAAAAIBQSZ02ky8lpUkt7pg1zboo+pgeyICQEqLTYN1dJJ9adbdHpmgzgX+iq5kcU+iKVVs56yRtCmxESHbNMvsw98oS1pT/gqJozuqgBoWk9obMSmHFrpjdZlIkjq3wTY7MLi7PFbNXLkkXHZzwUaGkPqYzOGanC6ycRUr593PMzAAAAAAAAKHCxQwAAAAAABAqqdNm0iQpTTrTHWvzpgwcZh6z2yWtDGZYSHLpktKkvKxE6O1mb8oeTZKcIAeFpNZXptXkO66Yfcv1KnXY/wFd1iRTT+42k7VWzokyfZ7/L6hBIZm95Uh5jjRqfSKWv96bk9Mq2kzgm3yZVhPX27kO5VUqqTGwESHpRd/PNbljP7VyLpJpuUthzMwAAAAAAAChwsUMAAAAAAAQKqnTZlIoKV06aKwr9q6VMzr62CraTOCPaJtJfKkcSZ/Y069pM4GfsmXaTOqsmFudmH4N/8TaTKoToVXWXOsh2eLPJ/DNKkm5Srxtkzp202UsluqDGxKSXKzNZBdXn8lqq224v6g5+KhGUh8p0x3Lt3IaxWomPT0AAAAAAACA7cHFDAAAAAAAECqp02ZSI3Pp5iVXbE8rpzL6yF3+4ZcMmbprSYTy7JwimSn/GwMaE1LDm66P9++xUSAVxKa4ulqb6uyclTItnIAP2qJbkStm19yL4rAK/wyQaW1yrxQ2+hFvDiuZwFcbJKVbncKzrZxqpfx5KzMzAAAAAABAqHAxAwAAAAAAhErqtJlskBSRNHBOItZwojcndits7vIPv6RFt6ZE6JdZVk6FzDTtNYGNCsmsWh1XM2mycoplpiVSc/BDq8yKTOmJUImdk+19HuiOfurYsllufV4spv3DPwWK1lxZImbv0hpEzcFH9ZLSvAuYrLrZmzJkkFL+vJWZGQAAAAAAIFS4mAEAAAAAAEIlddpMmmTaTNz3hG22coZHH9skVQUwJiS/2NQv91zEz62cC8Vd/uGfDTJ7dvel6nVWTpESK1AA3ZUl09pUlAj1H2HljJY55j4d1KCQzEbKTL12d9MV5HpzxjV2sqoO0EVZip5BvJaI2Z2ajaLNBD6qkRSR8goToX/WeFO+VyfT5pnCmJkBAAAAAABChYsZAAAAAAAgVLiYAQAAAAAAQiVl7pmx1JHyHWnkl0dvPmnX6GOrpM+CGBWSXps6LFmoIVd4c9bdZPIAP2ySqTn3PYGWWzljFb2HEOCDIpl7ZrjvWbB/zwwFqaEiRyqISGtcNygosI6j2eJ2VPBPenT78oNEbIWVkyvumQEftcq8VxuTCOW+Y+Vky7zn2xjYqHodZmYAAAAAAIBQ4WIGAAAAAAAIlZRpM9mk6HTDw1zBEiupzpUM+KnF/cnfvc8tVsovqwQfNcnb1iSZ5b3c2kRrE/znnl+9l/Xc22LOP/wzUlK61PJRItTS7E1pim6AH1qj20JXrNrKyRC7OfgoR2bawehEaKTdZpIrqV3ShsBG1eswMwMAAAAAAIQKFzMAAAAAAECopEybyS7lUkGa9OriRGzyCVbS/Ogj06/hl3brUZJuWe1JcaroMoGPGmUuU7tXlsiwcprEfg7+qZapsWxXzK65laLm4J8Bkvp4D612yQ3JkmodWW2eQNc0yiws4d6N2R2d6Z3EgC7Ll3k/NzIRKrNzGpXyJxHMzAAAAAAAAKHCxQwAAAAAABAqKdNmosMlZUq6wxUba+X8K/qY4tN14KMWmUuG7unXN3pTIllShKmw8EurTM2VumL2Zet0sZ+DfzbI1JS75hZaOaygAz+lmc19aI0UWjl7yaxO92Zgo0ISq5dpa3K3M2VZORnq2O4EdFmzOrYNW76sSizGmaqYmQEAAAAAAEKFixkAAAAAACBUUqfNpK+kLGm0O7bOyonNV2yX1BDEoJD00mUuGRYlQpWLvSn9h8rUnHeRE6B73Du7pdZz2TLTrwE/NMrs69z7sDwrp1W0mcA/7WYrdf9JrtHKOVpm5SbaTOCDWKdcqyvWauVEu58Af6yXWUKnKhFaZKUskrQxuBH1Sjv0/9zrr7+uY445RmVlZYpEInriiSc8zzuOo2uvvVaDBw9WTk6OpkyZokWLvP9MVVVVOu2001RQUKCioiKde+65qq+v35HDRohRcwgaNYegUXPoCdQdgkbNIWjUXPjs0IsZDQ0N2muvvXTbbbd1+vxvf/tb/fGPf9Qdd9yht99+W3l5eZo6daqampriOaeddpo+/fRTvfDCC3r66af1+uuv64ILLtiRw0aIUXMIGjWHoFFz6AnUHYJGzSFo1FwIOQGR5Dz++OPxz9vb253S0lLnpptuiseqq6udrKws56GHHnIcx3EWLFjgSHLefffdeM6zzz7rRCIRZ9WqVZ1+n6amJqempia+rVy50pHk1Jwlx7lAztdKbM4R1nas2WqmybympmbH/DIQiKBqznG2UHclcpxBcpy9E9smeTdnLzk1Y6i5ZNAram6YHGeEHOdU1zbe2qbJqZlCzSWDXlFzY8x+zCl2bZdY2yCzP6TmkkOPv6fbX45zsBxnaGJrlndzVsupWUjNJYuerrnHJed5ybl5C9tLkvOkqLlk0dM1t0ZyGiXHOTmx/T95txsl57oUr7kea+1atmyZ1qxZoylTpsRjhYWFmjBhgubOnStJmjt3roqKirTffvvFc6ZMmaK0tDS9/fbbnX7d2bNnq7CwML6Vl5fv2B8EobGjak6i7tA5ag5Bo+bQE3hPh6BRcwgaNdc79djFjDVr1kiSBg0a5IkPGjQo/tyaNWs0cOBAz/N9+vRRcXFxPMd29dVXq6amJr6tXLlyB4weYbSjak6i7tA5ag5Bo+bQE3hPh6BRcwgaNdc7Jd1qJllZWcrKyur4xJeS+kjzXKHp9momB0Ufm3fM2JC8Nlt3GepwybDWSunXIHNndmA7bLbmBsisLLHEFSvp5Atwy3Vsp83WXIHMu4kt1VyRzFIA9nEX2ILN1lx0NZP4KnSSModZOYOLpTxH0oYdN0Aknc3VXG50c59mFls569RxUR1gaza7n4tZuPmnKkTN9djb2dLSUknS2rVrPfG1a9fGnystLdU333zjeX7Tpk2qqqqK5wDbippD0Kg5BI2aQ0+g7hA0ag5Bo+Z6px67mDFixAiVlpbqpZdeisdqa2v19ttva+LEiZKkiRMnqrq6Wu+//3485+WXX1Z7e7smTJgQ+JgRbtQcgkbNIWjUHHoCdYegUXMIGjXXO+3QNpP6+notXrw4/vmyZcv04Ycfqri4WDvttJMuvfRS/fKXv9SoUaM0YsQIXXPNNSorK9OMGTMkSbvvvruOPPJInX/++brjjjvU2tqqiy++WKeccorKysq2bzAbJKVL/3KFpudZOZ9HHzdt5w+KXqNX1ZxkpvunyTMH7D0r5fDVMvchRij1uprLl9mzV7piRVbOOrGfC7FeV3MDZFrqRrpig6ycMZJaJS0WQqpX1V2bpIikKlfsQCtnSZVU14UfFL1Gb6q5ZplDa7UrZt+msVLSxu3/MdGL9Kaaa41uOa5bbVRbOdPEbm6HLs36yiuvOIouF+PezjzzTMdxzBI311xzjTNo0CAnKyvLOeyww5yFCxd6vkZlZaVz6qmnOn379nUKCgqcs88+26mrq9vmMdTU1JjlavaW4+wn5xIlNudAa4su0VpzaGovcRNmvaHmHMdVd+VynGFynF0T2/Pybk6unJocai6sel3NTZbjTPEuB+ycZm3s50Kt19XcsXKcE6LLs8a2O63tBJNHzYVXb6i7eM2Nk+NMtJYDPtbaFsup+YCaC7PeVHNPRpdePde13Wxtt0nO/6T4Mplh15tqbqXk1MgsbR7b7pB3WyE581O85iKO4yT134Rra2tVWFiomr2lgnTpR64/i//BvoqfG33NJqnwZammpkYFBQWBjRXJI1535VJBmjw3KXvBupHP4blSrSMVbqTm0HXxmpssFdgzM0ZbyevYz6H74jV3rFSQIe+si5lW8r+k2lap8ElqDl0Xr7lx0f2c+3hqv6f7nVRbJxXuTc2h62I196SkPEkPup4ba+VmyMzM+ImoOXRdrOZWytxfu8A10/HP3tt1xGdmjFHq1lzSrWayWf0k9ZE+csf2s3Iejz6ysgT80irTZuKaA7bUSmlrNDNmAV+0yOzDhrtirVZOq2gzgf8qXB/bO7USmdoE/LBJ5m+RTa5YtZVzj1idDr5plukcznTFiqycSlFy8M8mRd++ud7DrbZydsoyfxBN5eMri/MBAAAAAIBQ4WIGAAAAAAAIldRpM9lVUpY08XlXzO4jvyv6mNR3EUGgmiVFJMd1x3X7Jv/paVJ67BZDQHfVysyFdU/5t3osaTOBr76ReTfhvlH7Zz00FqSGWBtTcSK0/HVvyvAa0cMJ3zTJ/AXY/R7OLq888Vdi+KcturlbSOw7JKhEprXY7j9JIfyfAwAAAAAAocLFDAAAAAAAECqp02YySFK2NNs9DbbOyondgpjp/vBLRFKadzGJQ+2cIpma2xDQmJDcmmTaTNJdsXwrp1pMv4Z/GmXqbbkr9oyVc4w6rqoDdFWGTM31T4T+8JU35ffLxfs5+CYtuo1yxRqsnHx5D71AdzTJ7Ooa6hOxY/a2krJl2oZpMwEAAAAAAAgHLmYAAAAAAIBQSZ02kyWSMiX9zRX7vZUTW92kTdL8IAaFpBe9tXXG+kQo82ArZ6XMnYhpM4EfYtOvs12xEitnuUzNAX4YIPNuYkUi9ES9N2XGItHaBP+kqcN8/hVWypc1HbuJga4aLKmvpEJX7AMrJ1e0mcA/tTJv1dwL0o2bbiUtUMq3cDIzAwAAAAAAhAoXMwAAAAAAQKikTpvJf2Qu3dyVGQ9tmNLiSel3QfSDFtFmAn+0qePd1O0pYnOiefYcWaAr+sns2TO2kNMqpvzDPzvJtHG6aq5xoZXTJGoO/mmWuYN/UyI01UpZqo6rTQBd1U9mtZKRAxKxxes75tFmAr+0RLcFrtg4+zjappQ/tjIzAwAAAAAAhAoXMwAAAAAAQKikTpvJBkkRSW8lWkvsTpKDghwPUsNGSRHvjYYzC62cEqX8nYjho0yZPbtr+rUWWTlM+Yef+krKknRoIjTs/6ycvWT2c58FNioksyaZ+fyuP8ldOMyb8skKU5aAHzZp62/VqiU17vihIEUUyrQ2VbqD5VbSazLFmcKYmQEAAAAAAEKFixkAAAAAACBUUqfNpI/MpZsXE6EKOyc2FTvFp+vAf567W//denKQaDOBf9aqw/RrfWTl5EtqD2xESBWjEh8eYM/vnyLTdvdIgONB8mqW2ccNccVu8qaMnSnVtktaGdywkLxao9t/XCuY5Fo5K2V2c4AfBksqkFTkDh5iJf1RKd82zMwMAAAAAAAQKlzMAAAAAAAAoZI6bSZtkhxJP02EhsyxcmJTEZl+Db9EJKVJ6cWJ0HNveFOOnCJam+CftTKXqasSoQ1rvSn9SpTy0xLho2aZ46t7zvXvrJxSSQ2BjQjJrk5mP1fiih21izfn2S+kFkl/CW5YSF4t0e0hV+yHVs4CcWiFf3JypJyINNy9RE6pldSilD9vZWYGAAAAAAAIFS5mAAAAAACAUEmdNpNhMnf4z3QSsf0i3pxPoo/MEYNfYqtKlCdCD1R5U478XCk/RQw+ypapudWJULOdUyxam+C/ItfHFxV4n7upVmoKcjBIao7McdPVwqn6L7w568RKYfDNRplD65uumN1mkq9UOrHCDtdXUppU4m4z+ZeVky7T0p7CmJkBAAAAAABChYsZAAAAAAAgVFJnNtQYSZlWbIX1+bjoY4uk93f4iJAKNslM/+qfCKVbKau+MjdmB3yxk8yefV0iVJrdU4NBSsiRlCXvzm1DrTfnadHaBP9EVwrztAVf401Z9QjHVvinSqZTrsYVs7uY8pTyM/7hpwyZNhN37Med5KV4qzozMwAAAAAAQKhwMQMAAAAAAIRK6rSZbJSZ4rrMNQFsuZUzIfrY4db/QBfFpsK6ZlxPsVLWSKoPbkRIdjvJTE1c44qNtXKWi1Wb4J/M6PaJK/ZPb8o3rzPlHz6qlzm+LnXF3vamvCzz1g/wwzcyHXWjXTF7gaY08Vdi+KhJUpp30abfrfam/Ei8neP/HAAAAAAACBUuZgAAAAAAgFBJnTaTNyWlSbUViVDBNCtnQfTRvj0x0FWDZO7wvzYROv1Ab0rtG54uFKB72qPbrq7YrlbO52JeIvzTINPG+bgr1uJN+SiaBvhhXYuZgT1wfiLWYrUIV4iag38aZQ6t7rdw6zaTC/giXVKalJmVCL1i7efOFC2czMwAAAAAAAChwsUMAAAAAAAQKinTZrJxtbnB/8Ou2PllVlJsiqwTzJiQAkbKFN67rti/T/WkFEx4yEz5fz/AcSF5rZbZs3/LFWu0cvJEmwn8s0FmNZOXXbGbvCmj/4+psPDPepk2k+yqRMxuHT7gTanWEX2c8EWuzGombvahNVemMwDwRaFMQbmWzSm32kyKlUIn85vBzAwAAAAAABAqXMwAAAAAAAChkjIzU+olRSQNcQet2xB/E52uyFRY+KZAZvp1kTv4oDdnt4fMnf9pM4Ef6mWmJbpWbtI8KydfZvUJwA+N6rgK2IneT4f8TardJOmtgMaEpFYv0xHsfr9WMMVKWifTTvdBUKNCMmuLbtmuWIaVky7aTOCjWJtJZSJ0iJUSGSFF2iWtCG5YvQ0zMwAAAAAAQKhwMQMAAAAAAIRKyrSZ9JGZDua5gf8Cb07sptj1gYwIKWGjzHR+z1zEvJ4ZC1JDrswOz71ak7WvU7Y6tgUA3TXK9bE95XUvmXY62kzggzR1MqX/IyupQazaBN+sk5Ql6QhXbK2VU62OK5wAXZahDmfqw+2ckTLnGbSZAAAAAAAAhAMXMwAAAAAAQKikTJtJvxypICLNc83/OqbKmxOblV0b2KiQ9L6S+V/mqrtVEe8kxCEniCn/8E9adFvtijVYOfYt2IHuaI9uxa7Y7VZOS3QDfDBQZlGmfFds1d+8OUOGytQl4IMmmRV0ttRNVyfTXQz4olmmhcS1bFOhnRN7z5fCUvzHBwAAAAAAYcPFDAAAAAAAECop02aifpLSpLXuGf4V3pSC2BTZdklLgxkWklyNzO3WXVPETrJS3mwUbSbwT6vMXNg5rpjdZlIe3HCQAr6W2c8VuGLzrJzRYso/fFOeaVqH3fuyhxd7c84JdERIdhnRzb0eHYvlYIeqk5QuNTQnQrsM6iRnU4Bj6oWYmQEAAAAAAEKFixkAAAAAACBUUqfNZL2kiHUXWHuqzrroI1Nh4ZdNMvXkuov//nZOq1J+ihh8lG62b95IhAbat78uEjUH/yyT+dPIJFfMPr4OkrkzO+CHQTI15zqgjrXaTJQh3s/BNwMl5Uha44rlWTmNAnxULynNrKQTk7erlbNAKb+fY2YGAAAAAAAIFS5mAAAAAACAUEmZNpOPW6S+kvLdwbe9ObXro48BjQkpINrepOxE6H/sS4ht4pbY8E9fSRnSaldoYJOVUyVqDr5pqTKddJnuFcJYoQk7Ulp0y02ExtnTr4tl2ulWBDYqJLEBMuW2wBXLsHKaRDcdfNSiDtMOnNe9n7+mjgvWpRpmZgAAAAAAgFDhYgYAAAAAAAiVlGkzWSUzPazGFfv3em9ObIZsfTBDQgpoqDWLS3jueD3FSmoUK0vAP9mSMqRR7lhZD40FKWGDzC6s1L2j+5eVVCJaT+CfVpk/x811xexVmxZKcgIbEZJcRnTzrCxh5eTKdBYDvnAktXtvkfCslfKaaG1iZgYAAAAAAAgVLmYAAAAAAIBQSZk2kyEyN/lvccUWWDmjo4/MhIVfqmTqKa/RFZxvJRWKlSXguzx3a8lu1pNNorUJvokvyOQ+qL5nJe0s9nPwT7qkNGn5R4nQcGtpif9p9bYEAN1RqI5tJfZqJmPFyhLwUZukiJQ5LBGaYK3OtEDs55iZAQAAAAAAQoWLGQAAAAAAIFRSps1kN0kFkka6YrlWTv++5jHDEfPE4Iv06Oaei7hmtTentEnccR3+aZHkSN+46mzgRCsnW/TTwTeDZY6vck3573CAbZLUHtSIkPQyJKVJ81yhXGufdp/obIJ/cqObu8xKrJxiSXWBjQhJL11meRzX8nT9j/OmnHqzqblrghtVr7NDZ2a8/vrrOuaYY1RWVqZIJKInnngi/lxra6uuvPJKjR07Vnl5eSorK9MZZ5yh1au9Z3rDhw9XJBLxbDfeeOOOHDZCjJpD0Kg5BI2aQ0+g7hA0ag5Bo+bCZ4dezGhoaNBee+2l2267rcNzjY2Nmjdvnq655hrNmzdPjz32mBYuXKhjjz22Q+4NN9ygr7/+Or7NmjVrRw4bIUbNIWjUHIJGzaEnUHcIGjWHoFFz4bND20ymTZumadOmdfpcYWGhXnjhBU/s1ltv1fjx4/Xll19qp512isfz8/NVWlrarbFkZkqZEWnX5kRsgp0Um8bTJunjbn079JDeVHOSNEDR6dfFiVipPb2fNpNQ6201p3pJfaR/ukLn2yvojBF3TAqx3lZzkV2kSLynLmp/K2mpmPMfcr2q7rIkpXs7gu2VJsbKtAR81r3vhB7Um2ouTWYXV+SK5Vs5TWJlibDrTTWnBpk2kyxX7PvelCEvSrVtSukdXa96O1tTU6NIJKKioiJP/MYbb1T//v21995766abbtKmTZtfU7C5uVm1tbWeDdgcP2pOou6w7ag5BI2aQ0/gPR2CRs0haNRcz+s1NwBtamrSlVdeqVNPPVUFBQXx+CWXXKJ99tlHxcXFevPNN3X11Vfr66+/1u9+97tOv87s2bN1/fXXBzVshJhfNSdRd9g21ByCRs2hJ/CeDkGj5hA0aq53iDiOE8gE90gkoscff1wzZszo8Fxra6tOOOEEffXVV3r11Vc9BWH761//qgsvvFD19fXKysrq8Hxzc7OamxO9JLW1tSovL9camen+b7tyJ59svfiT6GvapMIvzNW2LY0FvVtQNSdtvu5qCqSCiKThrmR7qvUCqdaRCh1qLux6Rc3tLhWkS/9wtZbYM/6HTJNqW6XCF6m5sOsVNTdNKsiQ99b+9lzrBdFj68fUXDLo6fd0sf3cM6793FTrtW/KzNKeJmouGfR0zf1bUl9Ju7pyl1qvrZapuami5pJBT9fcaplz17zxruT7rRefK9VukgrfSt2a6/GZGa2trTr55JO1YsUKvfzyy1v9R5gwYYI2bdqk5cuXa9ddd+3wfFZW1mbfiAGS/zUnUXfYMmoOQaPm0BN4T4egUXMIGjXXu/ToxYxYMSxatEivvPKK+vfvv9XXfPjhh0pLS9PAgQMDGCGSDTWHoFFzCBo1h55A3SFo1ByCRs31Pjv0YkZ9fb0WL14c/3zZsmX68MMPVVxcrMGDB+vEE0/UvHnz9PTTT6utrU1r1qyRJBUXFyszM1Nz587V22+/rUMOOUT5+fmaO3euLrvsMn3/+99Xv379tmss30jaKGmtOzjKSno3+ti+nT8oeo3eVHOSzK2vI5JyXbGF3pSN7aY2EU69ruY2SEqTThqQCH263psypFrSlu/1iF6s19XcIEmZkoa5Yv9r5ewjai7kelXd1UtK83Y2ped6Uw4qlmrbJa3u0o+LXqA31VxudMtx/QF9UbM3J09SY9d+VPQSvanm2qKbalzBeVbSOrFSmLMDvfLKK47MopOe7cwzz3SWLVvW6XOSnFdeecVxHMd5//33nQkTJjiFhYVOdna2s/vuuzu//vWvnaampm0eQ01NjSPJ+VhylknO/3Ntzn9b2wiz1Qwz46ipqdlBvxnsKL2h5hwnUXc1/eQ4xXKcia6t2Ls1Ss4aUXNh1etqrlSOUybHGZDY5su7ORPl1Iyj5sKq19XcWXKcC+Q4v3Jt1n7OmSKnZjI1F2a9oe7iNVcuxxkm5x0lNifX2obKqSmj5sKsN9Xc+5KzUHKcrMT2uLzb89FHai68elPNrZScGslxdnVtD1nbrnJqdk7tmtuhMzMmT54sZwv3F93Sc5K0zz776K233vJ7WEhi1ByCRs0haNQcegJ1h6BRcwgaNRc+PX4D0KDEpuqMdgerraQiVzLgh4ikNEnZidDGKm9KtqSWAIeEJFcnU3d/SoSGn2HlNIr9HPyTEd1WJEJfWPu5XXIltQY4JiS39ZIipsMprsTK2V+m5uYENSgks/zopvREbJ2VM0QpdGKFHa6gT3RFxGJX8BEraaXM3JAUltbTAwAAAAAAANgeXMwAAAAAAAChkjKzoSoi0ak6R7uC9h1h0wX4K1vmkmFGIpRTZuWkSRHuuA6/xFbQOX2feChvlr2zA3yUFd3eTIRK7ZwmsZoJfONsNDOrd3IdWzu0zuWKHk74Ji26uduGa6ylS3LFgojw0UCZoitPhL6w2kx2SRNtJj09AAAAAAAAgO3BxQwAAAAAABAqKdNmom/JTL+ucMWetXKGRh+ZIwa/9JWpO3cLU4aVky1WloB/ChW9TP3rRKzkSG8O0xLhp4Ey+zHXCiYFuVZOf7GaCXzTHt3S3UvUrbKSqkXNwTet0c29ak6atWpTY3QDfBE7H3WdNzxupVyZL/N+rjaYIfVGzMwAAAAAAAChwsUMAAAAAAAQKqnTZtJX5qd9wBUrsXIyo49M+YdfsmVaTFxTXT9f4U3ZbahobYJ/ShXds09NxIZbOZWBjQapYJCkHHlb6IqsnAMlbZQ0J6AxIam1RLecXV1Be5p1rBcF8MEGmbdyQ/ITMXvVpkWizQQ+qpFZna4uEcqzc7KV8vs5ZmYAAAAAAIBQ4WIGAAAAAAAIldRpM1kuKU16f30itO+BVk7s0s4mSYsDGRWSXbvMFLGGROhzK2W3dWJlCfinUNHp/g8nYmVWToPMfg7wQ5bMVFd3m8kRVs5oefaDQHfUyhw2c9wrhWVbSfbKYUA3rJfplNPKRMxetGmFpKbARoSk1yRzDuFaNWe0ndOglD+HYGYGAAAAAAAIFS5mAAAAAACAUEmZNpN1K81sHc8NXxdZSXtFH5l+Db+kRTfXnYh3s1I2NkenLgJ++uaUxMf2Zet8sZ+DfxrVcZrr8dbn2WKlMPimSmY1k9LlrqA9v79a7OfgmzqZXdiStYlYppWTrZSf8Q8/5cq0mVQnQofubeXY57IpiJkZAAAAAAAgVLiYAQAAAAAAQiVl2kyWS8qTVOSKfb7Wm7PbiugHTIWF31x38a+wnlopTxcK0D3NMvuwX7li9lxY7vIPP22Q6ZVz78jsFXSKxY4OvmlS9A3sAlewyEpaKqu3GOi69uj2nis21MphcTr4qlBm2oH7vPRRK+eE6POfBTWo3oeZGQAAAAAAIFS4mAEAAAAAAEIlZdpM0iSlSypxxV6zcnZdaB6ZIgbfNMkUXnMilJnrTalqlOqDHBNSw4Ouj7/XY6NAKmiV2c+525c+sXK+K3NndsAHhTKLMm2oScT6FXtz2hbTNQz/tEU39+IRdjddLAfwRbYSJ7Ax1n5ORUr5VZuYmQEAAAAAAEKFixkAAAAAACBUUqbNpEJSgbzTEPOrvDkro4/ccB2+2SRz++sWV2ySN2XQs8y+hv++WZ/4eGCT9WSzUn5aInyUFt2Gu2IPWTnZkhqDGhCSXUWmVBCR1rhaOFm5BEFwH07XWM81yizsBPiir0yLSbUrdpWVs0opv+9jZgYAAAAAAAgVLmYAAAAAAIBQSZk2k36lUkGapKGJ2F7veHNiN4tN8dk68FOuTGG5p1ef6E3Z6TWp1hFzE+GPJkl9Em1zkjSw0sqpFLdch38aZQ6cw1yx570pzqmsFAYfjZWULpW438dVe1PSh0np7fLuDIEuao9u7g716p4ZClJFscyZ+oJE6NO7vSl7FCrlD67MzAAAAAAAAKHCxQwAAAAAABAqKdNmoiEy0/3TE6H+Vkp6oXnMcCTVBjMsJLkcmf9l7uVKdrVyimXmLtJmAj+0S2qT3nCF9rV759aJfjr4p1JSprwHVWs/d9Zq76JOQLcMkZQhpbvm/Lct9qak7yOzahNtJvBBjcw+bJErVmHl5IuFwuCj6EphDa5Vm/5ppewxXKZteH5go+p1mJkBAAAAAABChYsZAAAAAAAgVFKnzaRQ5qddmwg1WCkFRdEP2kWbCfwRa23Kd8UWWjmxW2QDPlru/iTberJOKX/3a/ioRlKGpBJXrMibMkeUHHxUKNPaNCoR6tBmUiqpNcAxIalVyRxK17li9qJgbZ3EgC5rlNTHe75aYudMlOl/os0EAAAAAAAgHLiYAQAAAAAAQiV12kxqZKb7750IVX/kTSkIcjxIDc0yt7bOcMVetnKaxPxr+KddUkTa3x3LsHLyRTsd/NMkM7fafZv/D7wpFdGUzwIbFFKCa6WwzFzruVViaQn4JtYRXOeKlVs5vJ2Dr76RlGZanGL2t3NGyRReCmNmBgAAAAAACBUuZgAAAAAAgFBJnTaTOpk2k4MSoaZ7rZzYHf+5FTH8Eqs79zR/ezUTVpaAn1oltUsnDbJibsNlpl+vDmpQSGq1Mu8m1iRCjlVbN8nckf3E4EaFZFYvc1x1rwRW5E1peL3jqnVAVw2WlCPvKcIwK6dO1Bx89KWkiHcFnTorZY8VMi3tKYyZGQAAAAAAIFS4mAEAAAAAAEIlddpMmmUu3bju+Jpp58RaAbjEA780ytSTeyqs1cbU1kpnE3zUJNPaNM0VW2fllKhj6wnQVU0y7yZcq0lEdvWmHJkv1bapwyonQJfUybxny3fFrNVMlsp0owB+GCRTYu5Dp70KYqPM4RfwQ210NSZ317C9IOL0h+U9x0hBnLYDAAAAAIBQ4WIGAAAAAAAIldRpM6mVFJG0PBHKsHNityBO8ek68FGTpDSpZX0iZLeU5BRL6Y6kDQGOC8kr1trkngu7wsqpEL1N8E+GzLuJYlcsz8rZVVKLaDOBPzbIzOcvcsWKvClji6Vajq3wSZ3MYbPaFbMPo+tkDsGAn3YpS3x8sbVS2FtrWUGHmRkAAAAAACBUuJgBAAAAAABCJXXaTGK2dAf/5ugjbSbwS6akNO8UMLu8cspl5ioyFRZ+2CjTUjc/EWqZ703JLBermcA/heq4soQ1FVZFShxjge6KLRvhXsGkyMr5sUyr5y8CGhOS2ipJ2dEtpsnKaZA5BAN+2KToW7X9ErH+1v7s33vR2sTMDAAAAAAAECpczAAAAAAAAKGSOm0mJTJTEl1LmAzpa+XEnqPNBD7rt6XLhsUyc8kAP2TJXKauTIQ+sVL2bRI1B//0k2mpc69mMs/KKVLHOdlAVzXL7OeGu2LrrJxTZZagoM0EPvhSZje3vytWbeVkS3KCGhCSXmt08/Q27XmFJ2c33aT6AMfUGzEzAwAAAAAAhAoXMwAAAAAAQKikTpvJTjJtJOmu2EFWTmyKYpuklUEMCkmvWWZliVJXLMvKyRcrS8A/A2T2c0sSoTw7p0FmPwf4KdP1cYb13K7iluvwT5PMn+Pcx9N0K2etvEuJAd2wQWa3drQrxi4NO1JfRRcJc+/H1t/kydktV6p1lNLL6DAzAwAAAAAAhAoXMwAAAAAAQKikTpvJAJkpsNWu2DArJzaNh7v8wy8RmUuG7suGQ6wcezo20B0FkvpIG1y3t96tzMqpE20m8M9GmeOmuz3TnvJ/tKTawEaEZBdbdW6uK1Zp5fyXeD8H3/SVOY0Y44qttnKaxKJN8E9ePykvIu9qJntZSbky+0PaTAAAAAAAAMKBixkAAAAAACBUUqfNpF5mOv+WpvvHbkvM9Gv4JVOm5upcsXwrp0msZgL/ZErqYxbSiTvWynlWiWnaQHdVyxxfq10xu42zz/lSnxZJ9wU0KCS1XElpUsvjiVDmIG/Kv9eymAn801dm8ZzsLeS0irdz8FGOOkw7eNjqbfpusSQnqAH1TszMAAAAAAAAocLFDAAAAAAAECqp02ZSJfPT7uOK2bccLoo+cvdr+KVNZvpXnRVzq+wkBnRVm6SIVVITrJxXRM3BPzUyx9dqV+w7dtJXYgI2fFMoKV1a4AqNXutNeVesLAH/bJDp4nTvxeyWkyZZLZ5Ad2yQWRVxRSJUYufky7QNbwhqUL3PDp2Z8frrr+uYY45RWVmZIpGInnjiCc/zZ511liKRiGc78sgjPTlVVVU67bTTVFBQoKKiIp177rmqr68X0BlqDkGj5hA0ag49gbpD0Kg5BI2aC58dejGjoaFBe+21l2677bbN5hx55JH6+uuv49tDDz3kef60007Tp59+qhdeeEFPP/20Xn/9dV1wwQU7ctgIMWoOQaPmEDRqDj2BukPQqDkEjZoLnx3aZjJt2jRNmzZtizlZWVkqLS3t9LnPPvtMzz33nN59913tt99+kqRbbrlF06dP1//8z/+orKxs2wczQOZu6zu7Yo9YOYXRR2bChlavqjnJzEFMU6KFSdbHkvSeUv5OxGHW62quVZIjuW94PSTTyikUbSYh1utqrl2mnvJcsZ2tnCeeTawYhlDqVXXXX1KGWdQkJjPDm5LdyqE17HpTzVXJnEZsqc0kW9Rc2PWmmmvYKKVLyvskESu2k7KV8u/nevwGoK+++qoGDhyoXXfdVT/84Q9VWVkZf27u3LkqKiqKF4MkTZkyRWlpaXr77bc7/XrNzc2qra31bICb3zUnUXfYMmoOQaPm0BN4T4egUXMIGjXXu/ToxYwjjzxSf/vb3/TSSy/pN7/5jV577TVNmzZNbW3mEtOaNWs0cOBAz2v69Omj4uJirVmzptOvOXv2bBUWFsa38vLyHf5zIDx2RM1J1B02j5pD0Kg59ATe0yFo1ByCRs31Pj26mskpp5wS/3js2LHac889NXLkSL366qs67LDDuvQ1r776al1++eXxz2tra01RDJaUJe+qEqusF6dHH1nNJGntiJqTtlB30TuuK9+VvJf14qfFvMQkFnjNpUlKs2YdLrJenCf2c0ks8JrLlXk34Z7/arc2fU/s55JcoO/p8iVlSEPcyUO9r61YJjV06bsiLIKsuWjJaZArd6H12gpRc8kuyJrLlTm8LnctkVNlv7hSptUzhfV4m4lbRUWFBgwYoMWLF0uSSktL9c0333hyNm3apKqqqs32KmVlZamgoMCzAZvjR81J1B22HTWHoFFz6Am8p0PQqDkEjZrreb3qYsZXX32lyspKDR48WJI0ceJEVVdX6/3334/nvPzyy2pvb9eECRN6aphIItQcgkbNIWjUHHoCdYegUXMIGjXX83Zom0l9fX38SpUkLVu2TB9++KGKi4tVXFys66+/XieccIJKS0u1ZMkS/fSnP9XOO++sqVOnSpJ23313HXnkkTr//PN1xx13qLW1VRdffLFOOeWU7b/beszjro+LrOfejD4yFTa0el3NtSjRvhRjf76PzJT/t7b/y6Pn9bqay5bURxrljtlzYZtEm0mI9bqay5GZf+3+o5PVGvx+vVS//V8ZvUivqrvokhJ57uVMirwpFaLmwq431dwBMru6XVztdK9Yc/73keloR3j1ppqLFEuRNKl6fSJmLdqkyvXeOyikJGcHeuWVVxyZSwOe7cwzz3QaGxudI444wikpKXEyMjKcYcOGOeeff76zZs0az9eorKx0Tj31VKdv375OQUGBc/bZZzt1dXXbPIaamhpHklNzoRznEjnONNd2hLX1NVtNnhlnTU2N378S7GC9oeYcx1V3u8hxdre2G63tQDk136bmwqrX1dyhZn+2XonNOdXaJsqpGUfNhVWvq7kj5DhHyXFmurY7vdt7kvOqqLkw6w11F6+5o+U4x8lxcl3b3t5toeS8T82FWm+quT9Izp2S4xQntjvk3d6RnJepuVDrTTVXUyzHGSDnAyW2161tveQsS/GaiziOk9TzEGpra1VYWKiaC6WCLHlvhGevyxudmVHrSIUNUk1NDX1L6JJ43e0iFdgzMc60Pn9aqt0kFb5FzaHr4jV3qFTQR6p8PvFc/1Ot5OXRmnuXmkPXxWvuCKkgQ9Jw15PWjY7fv8D8lXyyqDl0Xbzmjo7W3L9cT+7qzf3iA1Nz+4qaQ9fFau4PMjMzznfNzPhzJzMz6iUdKmoOXRffzxVLBWnSh66ZGfYsjNHR2Ailbs316GomgaqT1CwteTYRGnmBN2XN84lUwBfrJEXkvTuNvcz0gZKaRZsJ/JErKUPqX+iK2Ss3NarjxVygu5pcH1sH0n2zzB8K1BLkgJC0vlbHlcJavSklYso//HOEouXmumhWOdeb06IOZQh0nSOp3dtakm+lLBftdL3qBqAAAAAAAABbw8UMAAAAAAAQKqnTZrJGUh/pn67Q5Q3elHXRx1SfrgP/OBvMLLGI67Jh2+PenPTfyjs9G+iObJk5ie57tay0cvJEmwn8U6uO7yYOtD4fLlNziwV03zcyf45zz7ku9qb0211Kb5P0RXDDQvIqK5AKIvIcO7N7bDRICS2SIt6Fmuy3c6WSMgMbUO/EzAwAAAAAABAqXMwAAAAAAAChkjptJt9ISpc8qyg97U3JsB6B7opEzOaei/jPRm/O8R+I21/DP5lma3Pt7NLtNqaKIAeEpLdK5k8j7jqbb+VUS2oPakBIel/LrBS2tys21sqZIGmjpAuDGhRSworEh/2tp9rFbg4+ypeUJhW77n+wzkoZzkphzMwAAAAAAADhwsUMAAAAAAAQKqnTZpIjqY93FuzyGm/KbtFLO7WOzBIUQHf1k7lkWJoILbenXz8s6g2+cy/WVMBla+xI62Wm/A9yxe6zcobLrAJgz5EFumDDJlNO/YpcwWor6fRi84buwg2BjQvJa2NttA3dde5gLwpWLcnqJAa6bqCkdCmnMhEqarZySmR6m1YHN6zehre4AAAAAAAgVLiYAQAAAAAAQiV12kxGScqUKuYmQtl2zn7RxzZJ7wcyKiS7PJlLhsNcIavN5D/t3pYAoFvqJWVIBa6aU6WV06aO82OBrsqQaTMpT4Ra3vGmZF4oqVnSB8ENCynAvRLY49ZzTtX/b+/ug+O6yjuO/67WWsmSV2spka2RkePYhDgZGOO81ECSUk9DGdNkGshMZ0LTptTpBMo7f7RlWhiSdkhmXBhImtJ0BkPoMDQhUxNISkPTpo2LHRcCDi8OzovjYOKXOLa1K+v95faPc1f33CPb2NLR1d6938/MGVnPPmuO4OHe1fV5zqGFE96MR2PAirldc62izQQetcj8pt4eh5rdNhOO0GFlBgAAAAAAyBYeZgAAAAAAgEzhYQYAAAAAAMiU/OyZUTDjz6zHN6HbY9QdfR0X4EezZjwy7HZSotIE/HhV5sp+XRyavDeZQr3BqxUyRVWKQ8XNTs7VYnMgeDOh6KPa83Hsv53NCn7rYzL7tAAedBSljkA6btVUyckpifsrPKrIFFQ5DrUdd3ImxZ4ZCz0BAAAAAACAc8HDDAAAAAAAkCn5aTPZL/PT/lEcCp5wcmpLFCdSmRHyYFLmaLhfxaGlTkqvkkd9AXNyXGZZ4roz5LSLo1nhT1nm/lqxYv/6m8mcf38yxQmh0dXaMwete+s+J+fSu7m3wqOCpEAasUKXOSnNMqejA168JrPsoCcOzWhjKsm0mbjnBOcIKzMAAAAAAECm8DADAAAAAABkSn7aTF6ReXTze1Zsj5NT2yGW5dfwpbbD8IE41OykdEtqSWk6yBHrOlZY4rxWEu108GdIZu1rmx38RDLnR08m12cDc9ApqUNSvxVzD6IbESUHf6rD5qtdZ+7pdBKnmcCjUUmBEp/nDjspHX0yn+fcPrscYWUGAAAAAADIFB5mAAAAAACATMlPm8kvZZbq7LdifU5O7XSTMI0JIRei00wmj8cht+wWd0njoaQTKc4LjatdZp3rXiu2xsmZFO108KfWZtJrxZ6+IZnzM83sAwBmKeiQgkAqWCfoLHdyVi6Xqjnf5R/+vCpzqStZsWNOTlHScGozQsMryyw7sK5zQ25Or3J/b2VlBgAAAAAAyBQeZgAAAAAAgEzJTZvJvgmzNGzN41ZwQzKnGrUCVNOaFBpfdJqJvaP6qt91cg7KLPmnzQQ+NEXDbjN5q5OzV7TTwZ9WmTaTi+LQi1ckU9a8X9JYinNCY+uU1CQdtZZfu910epvM8utHUpsVGtghmS5O6zI3o4OpJGkwtRmh4ZVl7q3W6ZtvvtDJaVfu762szAAAAAAAAJnCwwwAAAAAAJApuWkzOSGz2nBNqxU8nMypLRcbSGdKyIMxSU1Smx271cm5V9JEajNCo6vILEssWLHrnJwjyv3u1/CoU+bThHV/vdtJ+eJaJfvtgLlYJmmRdPSlOPQW95/nrpCpOdpM4EFF5rZ52InZWsWtFR5Fp869OBWH1nzayfmacv87BCszAAAAAABApvAwAwAAAAAAZEpu2kymFK3WudoKPpbMqa2AHU1lRsiFUUmBFHRZsR1OzjFNLyUD5qwi85i614o1Ozkl5X73a3jUKVNj++LQjH8p2S9urvCnJGmRZN9a9VdOzsPi3gpvxqNhXebU4uT0SxpKa0JofCclNUnHrdAa5yRObVbuT6djZQYAAAAAAMgUHmYAAAAAAIBMyU2bSY/MqkRdagW/kczpjr62CvDMOs7k+1uSL13VptwvEYNHBZnH1PaW6u4RTSNiy3X40yrTZmKdGvF2N+dxseQf/oxLCqU3lK3YHe9LpPzF33yFziZ4U2szecaKude5EZm2dsCLZklNkn2Z08Fkys+nTDdKnrEyAwAAAAAAZAoPMwAAAAAAQKbkps2kL5A6AklHrGBXMqc7es3dnRiYtYKkQIl1hw86KVcOcbAEPOqUqbtXrNhxJ+eopInUZoScePHl+M83rHdePCDWX8OfSZl7a2Jn/62JlL/XV+jghDet0fiBFbvYyVmlmYeHAbNWllSIt0GQJH07mVJrf8ozVmYAAAAAAIBM4WEGAAAAAADIlNy0mahZZknidivW5+Q8m950kBNDMnVnrTu8yEkpLpeKUzJL/4G5Wi5zZd9jxQ47OYPiZAn4M2XGiB27ycn5pyjPbXkCZqMoc517kx28PZHSI1Ny+9OaExpap6R2mS7OGvdjW5+4tcKjE5Kakh2ag3cnU7pEaxMrMwAAAAAAQKbwMAMAAAAAAGRK7tpMBr8Xh9o3LthskBNjE+akkmJ7HFv/mpNUW5dImwl86JVZgm2fZnJggeaCfBiTFDotdAednG6ZE3T2pTUpNLSizOe6Xiv2t59JpFwrU5pfTWtOaGi100zWWbGCk3NY0nBqM0KjGz5gLnP9VuwhJ+dDos2ElRkAAAAAACBTeJgBAAAAAAAyJT9tJmVJTdJDJ+PQLc6u6sHy6CsnS8CTYmCG2uLYVUucpDaZ5deAD72SWpQ8NWLIyVme3nSQA9GS/+LVVuxrTs4V4joH/6yjI178VPKl1XJO2AHmYEk0LrZibpvJeDQAH07KHIhoo2t4JlZmAAAAAACATOFhBgAAAAAAyJT8tJmskLRI+umvrJi7Vqe2KzYnS8CXTplHhvZa1y1OzoOSwtRmhEbXIqlVGjsSh9yl1h29oubgTzEaN8Wh73wwmXL9uGgzgT9VmU+w1qk5e5yUbnGyBPyZikaPFXM7OKWZbQHAbHUvkjoCqWD1Ll3p5AzItKPkGSszAAAAAABApvAwAwAAAAAAZEp+2kyaJS1KLg8bc04zKXZFf5hKaU5ofO0yjwztWnu/8wzx61Msv4Y/Q5ImpeKFcejwS8mUjoJoM4E/rTJtJr8fh/qdNhPRZgKfXpW5t+6NQ71OyrtkulE+lNqk0MhekTl8zjpAR1Un5zyZXzcAL86X1CQttdrpbnEKbGw8T7/MnxorMwAAAAAAQKbwMAMAAAAAAGRKflamHJDUJL3RCvU7KctqJ5iw/Bq+jEtqkiYrcahwwuljKkV5gA8HZNa5/nkcav2Ak9MstlyHP4VonL9yOvSHLb9M5rSKNhP485rMNcw6qmmVk1L4klQYlvSJ1GaFBvaypMVKtpEMODk9os0EHlUkBVLhfCv218mU4n1ScVLScynOq86wMgMAAAAAAGQKDzMAAAAAAECm5KfN5JCkQHqTFep3UrqjVgC6TODNpKRQGrRCHd91ci6SNCrp8dRmhUb2osyS/0/FoW43p020NsGfEZlTwJ6zWktudnKOi5qDP0OSgmQLZ7+Tct4tMsdN0GYCD05IGnZiFzjf1zruAB8Gh009tV9hBT/alkz6jyFzb6XNBAAAAAAAIBt4mAEAAAAAADJlXh9mPPnkk7r++uvV29urIAj0rW99K/F6EASnHFu2bJnOWbVq1YzX77rrrnOfTLMZK9bHo1XJEbRFY/Hsf2YsrLqqOcmcVFKKlx4WJOm/nDEps0QbmVR3NXdI0kFJj8QjeF1yqBgNZFLd1dwRSYcl3WmNfmc0iX8+ybi6qzuZFs7a2OYMHZapTWRWPdXcqExHnT1WOKNVUsvsflTUiXqqufgvtIaWOAPz+tFicHBQ69at07333nvK1w8dOpQYW7duVRAEuvHGGxN5d9xxRyLvwx/+8HxOGxlGzSFt1BzSRs1hIVB3SBs1h7RRc9kzrxuAbtq0SZs2bTrt6z09PYnvH374YW3cuFGrV69OxEul0oxc4FSoOaSNmkPaqDksBOoOaaPmkDZqLnvqZtHnkSNH9Oijj2rz5s0zXrvrrrt03nnnaf369dqyZYsmJiZO+/eMjo6qWq0mhiTpjZLWSXp7PFY2JYcui8Y67z8e6pCvmpPOUHdlSUuT7Ux6xhmPSfpPbz8W6lgqNVdbC/ugNbqdMRYNNLxUam5Y5nSJb1tjrzOGooFcmPfPdJNm2C2cDzlDd0v6R78/F+rXfNfceDSWWqPsjKijHTkx3zU3/XvDgDX+4dXkGNLMY3Zypm6OZr3//vtVKpX0nve8JxH/yEc+ossuu0xdXV3asWOHPvnJT+rQoUP6/Oc/f8q/584779Ttt9+expSRcb5qTqLucHaoOaSNmsNC4DMd0kbNIW3UXH0IwjAMU/kPCgJt27ZNN9xwwylfX7t2rd7xjnfonnvuOePfs3XrVt122206efKkWlpmbrMzOjqq0dHR6e+r1ar6+vpUuVLqWCRpg5V8t/Pmt0XvmZDKT0mVSkUdHR2//odDXUqr5qQz1N16qaMgTf4wzi1c4by5X6pOSeV91FzW1UXNdUodgSR7xeOk8+ZeqToulb9HzWVdXdTcNdH99Rkrudd5c19Uc49Tc41gwT/TBeY6N2htnv3bznuf+phUHZXKX6LmGsFC19xHZTb37LZyNzrvHZLZjHaTqLlGsNA1d1xSh6TCu63ka503PxT93ro9vzVXFysztm/frr179+qBBx74tbkbNmzQxMSE9u/fr4svvnjG6y0tLaf+IHaBzNqvESvW5+TU/roxSU+d3dyRTT5rTjpD3bVIWuT0c+1Ppvzba6y+zoNUa65J0nIr5p6WMyDpzB0FaACp1VxtfbXtJuf7R0TN5UQqn+laJQVSu/Wgtnk0mVL9glQ9t6kjo9KouRFJoZK3VvdQsIKk4FwmjsxKo+YKi6RCIOmVODb2wWRO8VqZwsyxutgz48tf/rIuv/xyrVv36zer2L17t5qamrRs2bIUZoZGRc0hbdQc0kbNYSFQd0gbNYe0UXP1Y15XZpw8eVIvvPDC9PcvvfSSdu/era6uLq1cuVKSWUrzzW9+U5/73OdmvH/nzp3atWuXNm7cqFKppJ07d+rjH/+4br75ZnV2ds7n1JFR1BzSRs0hbdQcFgJ1h7RRc0gbNZdB4Tx64oknQpnFL4lxyy23TOfcd9994eLFi8P+/v4Z73/66afDDRs2hOVyOWxtbQ0vueSS8LOf/Ww4MjJy1nOoVCqhpLByjcJwo8LwfGtc6Iz3m1H5EzPPSqXi478GpKgeai4MrbpbrzC8QmHYEo8jSo73SeEfiJrLqrqrudcrDC9WGH7DGn/sjEsUVt5AzWVV3dXce6O6WmeN8NLk+A2FlcupuSyrh7qbrrkLos9tl8TjC0qOZ6XwB9xbM62eam6zFH5ACr9tjf9zxgtS+GNqLtPqqeYqHQrDcvR7RDR2KjnCTQor1+a75lLbAHShVKtVlcvleIOyn1ovlpzkd0bvGZPKW/O7kQrmbrruog1A7bp71enr/UuZbVq+LmoOszddc6+Pau4z1ouPOcm7pOqkVH6OmsPsTdfce6WOoqQfWy/uvjSZvGGPqbmnqTnM3nTNXSB1NCk6t9D44rPJ3HdKOinpSlFzmL1azW2W2SNjk/Vaj5PbJbMl1XpRc5i96etcR7Sh+0Xxa0/9MJn7lk1srl0Xe2YAAAAAAACcrbo4zSQVz0tqkra/FoeuaXdyxp2vwFwNyGxvvTQOHT2STGlW7jcihk/LZK7sb7di25ycEc084QSYralodNlBZ8f2qT3UHPxpk7m3Wisz3BML177OHHuug+lNC41ricxhYfaibvfU8yFJw6nNCA1vicyyg7Y4tMHNOaaZhZgzrMwAAAAAAACZwsMMAAAAAACQKflpMzkuKXBiXc73BecrMFejMo8Mm+NQq5NynczSxK+mNSc0tpUyu5RZNaddTk5JuV+WCI9GZVpI7Ivbz53epn7RZgJ/Dst8prPW/K92c3pkrnO0mcCDFZIWK3kZc3+NeEHm8xzgRUHmdwjr3hqUnZxfKff3VlZmAAAAAACATOFhBgAAAAAAyJT8tJmMSwqc/dXd00xqy7I5WgK+TMrU00gcWuM8QlzTLFVDSWMpzguNqykaL8Sh4ZeTKYvfKmkixTmhsbXItDYNWrEfOTnjyv1SWPhTPWG+th4/Q9I+8XkO3nTKHCpxwIo5ZzZpRImPe8DcVGTa6ayi+n4lmXKVlPvrHCszAAAAAABApvAwAwAAAAAAZEpu2kzC0Ixly63gUSeptoyH5f7wpbasetyKXX2KvAlJO+Z/OsiBfTJX9sNxaMYBTQXlflkiPGqTaTPZY8VKTk6XTNvdAQFzNirzka3firkdJ29uFq1N8KY5Gvut2OBpcgAfwqr5qBZYbcOPOzlXLZW5zlVTm1bdYWUGAAAAAADIFB5mAAAAAACATMlNm8lUNArrrOD/OknPR1/Z5R++1E6W6LZiq5ycY0q2oQBz8YzM7tf9cajoPrYek1nyD/hQiMYzVuyHTk6fzHXuJ2lNCo1siUwn00ErdthNukjm85zbUgzMwhKZQxBbrZhzUJgmxa0V/gSLpSCQjlkXuhmdcxfIXOdy3MLJygwAAAAAAJApPMwAAAAAAACZkps2k0K7VAicYJfz/d7oK7tfw5c2meXXdq25a2GnRGsTvAmHo92vd8WxYeeatnhcrIWFP9Ha6l9a7XIr/87JuVWcFAZvitEYsWJ9blK/uM7Bm/NlWk1WWbF9Tk63+BUCHvVJKkjjz8ahGQcirpa5t+b4RERWZgAAAAAAgEzhYQYAAAAAAMiU3LSZ6EKZ5f721tdvcnJqS3TCVGaEPKi1mZxp++vV6U0HjS+4QAqaJD0Sx1pPmw14MCJpyiz7r/n+aDLlqtq1EPDIvratdV4b+xmdTfCndmjTUiu2x8lZLv6VGB71Slok9RyJQz3u7wztkppTnFMd4v9zAAAAAAAgU3iYAQAAAAAAMiU/bSadMj/tUSs27uQMRF9pM4EvQzLrEnut2FEnhzYT+LRC5lq3Mw4Fv+PkDIgTdODPSUnNUk9bHNo95OTU1mgDHjRFwz4orNCSzHlx1JQm4JN9go77L8JT4jQTeDQpKZD0Viv2XifnfuX+8xwrMwAAAAAAQKbwMAMAAAAAAGRKftpMXpNZ4mov6T+eTBmM1oYNpjQl5MAhmSViG+LQL5y6W9uv3C8Rwzywr3XXOa99VzPb7IDZqtXSO+PQeducnCFJzgknwGwF7VIQSEvtPhJnR//iaO43+cc8KFl/Lp42C/Cg9rvrtVaslEyZ/J7pRskzVmYAAAAAAIBM4WEGAAAAAADIlPy0mfxCZrn/rVZsRzKlthE7K/7hy/CgWea6uDWO/Y+TszbNCaHxjcucyNRtxUqnyQV8el/8xyt3Oa8dE61N8GelpILU+jMr1p1M6T0pVdOcExraeDSusY7Q2eG0DddO2QG8qMgUlPU7hB5KpnxHposzz/j/HAAAAAAAyBQeZgAAAAAAgEzJTZvJkVAaCqWe9VZwezIniJaOBaGkE2nNDI2sEA17q+EZK/6bZVqgAB+mZOptjxU76uSMi346+DMqU3Mbrdi7nZy9oubgzyqZe6fdZnJlMqW4VCpOSvpJWpNCIytJWiJJN8expXcnc1rFyRLwKDplUwfj0LGvJVNeljSS1nzqFCszAAAAAABApjT8yowwDCVJtaPI2watF93NyEyqqmHyvcC5qtXOQPR9cSx+zd2opzphhv0+4FzVaqda+2chu5Tcx/bUHDyYrrnaigt7t8UxJ5magwfTNRd9frPLrOh+ppuMr4fUHGbL/T2iOhq/NuzkDir+jEfNYbamr3O1lRnWhW7AyR1R/BEvrzXX8A8zBgbM/+yX1QLvOrf3lstl73NC46vV3epa4J9Pn/un1qk61Bxmq1ZzfbtP8eKnz/w+ag6zMV1zT0aBFWf/PmoOszFdc4+f4sVtZ34fNYfZqNXcdBfdfWf/PmoOszF9nXs1CvzL2b8vjzUXhA3+GGdqakoHDx5UqVRSEJzdxgRhGGpgYEC9vb1qaqITB+fuXOuOmsNcUXNIGzWHtFFzSBs1h7RRc+em4R9mAAAAAACAxpK/xzcAAAAAACDTeJgBAAAAAAAyhYcZAAAAAAAgU3iYAQAAAAAAMoWHGQAAAAAAIFN4mAEAAAAAADKFhxkAAAAAACBT/h9G2WqRVSC3qgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1100x600 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "_f, axarr = plt.subplots(1, 7, constrained_layout=True, figsize=(11, 6))\n",
    "for i, word in enumerate(emg_words[:7]):\n",
    "    axarr[i].imshow(emg_data[i], cmap='hot', interpolation='none')\n",
    "    axarr[i].title.set_text(word + ' %d'% (1 + i/5))\n",
    "    axarr[i].set(xticks=[])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fbceea5",
   "metadata": {},
   "source": [
    "### Training the Neural Network\n",
    "\n",
    "We need to factorize the words array so we can get an unique list of the words used, and also retreive a list the model can read as an expected output:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4b4465a6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This is the expected output, containing only 5 possibilities (Open hand, Closed hand, Lateral pinch, Signaling sign, Rock sign):\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open hand</th>\n",
       "      <th>Closed hand</th>\n",
       "      <th>Lateral pinch</th>\n",
       "      <th>Signaling sign</th>\n",
       "      <th>Rock sign</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Word 1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Word 2</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Word 3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Word 4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Word 5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Word 6</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Open hand  Closed hand  Lateral pinch  Signaling sign  Rock sign\n",
       "Word 1          1            0              0               0          0\n",
       "Word 2          0            1              0               0          0\n",
       "Word 3          0            0              1               0          0\n",
       "Word 4          0            0              0               1          0\n",
       "Word 5          0            0              0               0          1\n",
       "Word 6          1            0              0               0          0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words, words_map = pd.factorize(emg_words)\n",
    "words = to_categorical(words)\n",
    "possibilities = len(words_map)\n",
    "index = list(map(lambda x: \"Word %s\" % (x), range(1, 7)))\n",
    "\n",
    "print(\"This is the expected output, containing only %s possibilities (%s):\" % (possibilities, ', '.join(words_map)))\n",
    "\n",
    "pd.DataFrame(words[:6].astype('int'), columns=words_map, index=index)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c57c7fcc",
   "metadata": {},
   "source": [
    "Finally, the colormap only has one color, so all we have to do now is to split the words that will be used to train (90%) and the words used to validate (10%) and submit all these \"words\" into a single dimensional Convolutional Neural Network (CNN) to create our model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b70ce024",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "14/14 [==============================] - 1s 21ms/step - loss: 0.6872 - acc: 0.2058 - val_loss: 0.6793 - val_acc: 0.1711\n",
      "Epoch 2/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.6686 - acc: 0.2067 - val_loss: 0.6571 - val_acc: 0.1711\n",
      "Epoch 3/1000\n",
      "14/14 [==============================] - 0s 12ms/step - loss: 0.6383 - acc: 0.2079 - val_loss: 0.6217 - val_acc: 0.1711\n",
      "Epoch 4/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.5939 - acc: 0.2129 - val_loss: 0.5731 - val_acc: 0.1711\n",
      "Epoch 5/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.5443 - acc: 0.2085 - val_loss: 0.5282 - val_acc: 0.1711\n",
      "Epoch 6/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.5135 - acc: 0.2181 - val_loss: 0.5056 - val_acc: 0.1711\n",
      "Epoch 7/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.5046 - acc: 0.2345 - val_loss: 0.4974 - val_acc: 0.3842\n",
      "Epoch 8/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.4988 - acc: 0.2564 - val_loss: 0.4941 - val_acc: 0.2342\n",
      "Epoch 9/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.4964 - acc: 0.2667 - val_loss: 0.4908 - val_acc: 0.3105\n",
      "Epoch 10/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.4917 - acc: 0.2909 - val_loss: 0.4858 - val_acc: 0.4053\n",
      "Epoch 11/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.4852 - acc: 0.3319 - val_loss: 0.4798 - val_acc: 0.4053\n",
      "Epoch 12/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.4788 - acc: 0.3751 - val_loss: 0.4724 - val_acc: 0.4316\n",
      "Epoch 13/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.4712 - acc: 0.4175 - val_loss: 0.4633 - val_acc: 0.4500\n",
      "Epoch 14/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.4601 - acc: 0.4447 - val_loss: 0.4523 - val_acc: 0.4395\n",
      "Epoch 15/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.4498 - acc: 0.4784 - val_loss: 0.4399 - val_acc: 0.4947\n",
      "Epoch 16/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.4353 - acc: 0.5094 - val_loss: 0.4274 - val_acc: 0.5237\n",
      "Epoch 17/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.4222 - acc: 0.5199 - val_loss: 0.4135 - val_acc: 0.5158\n",
      "Epoch 18/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.4089 - acc: 0.5342 - val_loss: 0.4009 - val_acc: 0.5632\n",
      "Epoch 19/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.3957 - acc: 0.5406 - val_loss: 0.3873 - val_acc: 0.6000\n",
      "Epoch 20/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.3833 - acc: 0.5591 - val_loss: 0.3756 - val_acc: 0.6342\n",
      "Epoch 21/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.3714 - acc: 0.5871 - val_loss: 0.3647 - val_acc: 0.6711\n",
      "Epoch 22/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.3609 - acc: 0.6009 - val_loss: 0.3530 - val_acc: 0.6947\n",
      "Epoch 23/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.3503 - acc: 0.6213 - val_loss: 0.3441 - val_acc: 0.7105\n",
      "Epoch 24/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.3403 - acc: 0.6488 - val_loss: 0.3346 - val_acc: 0.7132\n",
      "Epoch 25/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.3330 - acc: 0.6526 - val_loss: 0.3252 - val_acc: 0.7289\n",
      "Epoch 26/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.3248 - acc: 0.6617 - val_loss: 0.3164 - val_acc: 0.7474\n",
      "Epoch 27/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.3166 - acc: 0.6930 - val_loss: 0.3087 - val_acc: 0.7526\n",
      "Epoch 28/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.3072 - acc: 0.6901 - val_loss: 0.2996 - val_acc: 0.7553\n",
      "Epoch 29/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.3005 - acc: 0.7076 - val_loss: 0.2917 - val_acc: 0.7658\n",
      "Epoch 30/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2917 - acc: 0.7161 - val_loss: 0.2837 - val_acc: 0.7737\n",
      "Epoch 31/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.2847 - acc: 0.7155 - val_loss: 0.2760 - val_acc: 0.7737\n",
      "Epoch 32/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2755 - acc: 0.7313 - val_loss: 0.2671 - val_acc: 0.8026\n",
      "Epoch 33/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2681 - acc: 0.7398 - val_loss: 0.2591 - val_acc: 0.8000\n",
      "Epoch 34/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2607 - acc: 0.7477 - val_loss: 0.2524 - val_acc: 0.7947\n",
      "Epoch 35/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2517 - acc: 0.7529 - val_loss: 0.2443 - val_acc: 0.8079\n",
      "Epoch 36/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.2459 - acc: 0.7617 - val_loss: 0.2362 - val_acc: 0.8211\n",
      "Epoch 37/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2389 - acc: 0.7673 - val_loss: 0.2294 - val_acc: 0.8211\n",
      "Epoch 38/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2320 - acc: 0.7734 - val_loss: 0.2253 - val_acc: 0.8211\n",
      "Epoch 39/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2273 - acc: 0.7807 - val_loss: 0.2175 - val_acc: 0.8316\n",
      "Epoch 40/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.2191 - acc: 0.7804 - val_loss: 0.2136 - val_acc: 0.8211\n",
      "Epoch 41/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2162 - acc: 0.7865 - val_loss: 0.2077 - val_acc: 0.8316\n",
      "Epoch 42/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.2087 - acc: 0.7944 - val_loss: 0.2036 - val_acc: 0.8342\n",
      "Epoch 43/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2064 - acc: 0.7930 - val_loss: 0.1988 - val_acc: 0.8316\n",
      "Epoch 44/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.2024 - acc: 0.7918 - val_loss: 0.1956 - val_acc: 0.8289\n",
      "Epoch 45/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1973 - acc: 0.7982 - val_loss: 0.1921 - val_acc: 0.8263\n",
      "Epoch 46/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1926 - acc: 0.8067 - val_loss: 0.1886 - val_acc: 0.8342\n",
      "Epoch 47/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.1915 - acc: 0.8123 - val_loss: 0.1858 - val_acc: 0.8342\n",
      "Epoch 48/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.1886 - acc: 0.8082 - val_loss: 0.1814 - val_acc: 0.8421\n",
      "Epoch 49/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1849 - acc: 0.8167 - val_loss: 0.1803 - val_acc: 0.8342\n",
      "Epoch 50/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1829 - acc: 0.8164 - val_loss: 0.1765 - val_acc: 0.8474\n",
      "Epoch 51/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1803 - acc: 0.8108 - val_loss: 0.1733 - val_acc: 0.8579\n",
      "Epoch 52/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1797 - acc: 0.8070 - val_loss: 0.1733 - val_acc: 0.8447\n",
      "Epoch 53/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1747 - acc: 0.8278 - val_loss: 0.1708 - val_acc: 0.8395\n",
      "Epoch 54/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.1748 - acc: 0.8173 - val_loss: 0.1696 - val_acc: 0.8474\n",
      "Epoch 55/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.1716 - acc: 0.8301 - val_loss: 0.1659 - val_acc: 0.8553\n",
      "Epoch 56/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1690 - acc: 0.8234 - val_loss: 0.1638 - val_acc: 0.8526\n",
      "Epoch 57/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1652 - acc: 0.8316 - val_loss: 0.1631 - val_acc: 0.8526\n",
      "Epoch 58/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1661 - acc: 0.8272 - val_loss: 0.1607 - val_acc: 0.8658\n",
      "Epoch 59/1000\n",
      "14/14 [==============================] - 0s 13ms/step - loss: 0.1636 - acc: 0.8243 - val_loss: 0.1595 - val_acc: 0.8447\n",
      "Epoch 60/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.1603 - acc: 0.8348 - val_loss: 0.1580 - val_acc: 0.8658\n",
      "Epoch 61/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1594 - acc: 0.8380 - val_loss: 0.1562 - val_acc: 0.8579\n",
      "Epoch 62/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1594 - acc: 0.8368 - val_loss: 0.1547 - val_acc: 0.8526\n",
      "Epoch 63/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1574 - acc: 0.8272 - val_loss: 0.1529 - val_acc: 0.8684\n",
      "Epoch 64/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1564 - acc: 0.8383 - val_loss: 0.1519 - val_acc: 0.8711\n",
      "Epoch 65/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1537 - acc: 0.8389 - val_loss: 0.1503 - val_acc: 0.8658\n",
      "Epoch 66/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1546 - acc: 0.8310 - val_loss: 0.1499 - val_acc: 0.8658\n",
      "Epoch 67/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1513 - acc: 0.8465 - val_loss: 0.1484 - val_acc: 0.8684\n",
      "Epoch 68/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1497 - acc: 0.8480 - val_loss: 0.1460 - val_acc: 0.8684\n",
      "Epoch 69/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1499 - acc: 0.8421 - val_loss: 0.1463 - val_acc: 0.8684\n",
      "Epoch 70/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1490 - acc: 0.8494 - val_loss: 0.1463 - val_acc: 0.8684\n",
      "Epoch 71/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1472 - acc: 0.8450 - val_loss: 0.1417 - val_acc: 0.8711\n",
      "Epoch 72/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1457 - acc: 0.8459 - val_loss: 0.1430 - val_acc: 0.8711\n",
      "Epoch 73/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1434 - acc: 0.8602 - val_loss: 0.1423 - val_acc: 0.8684\n",
      "Epoch 74/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1422 - acc: 0.8523 - val_loss: 0.1399 - val_acc: 0.8763\n",
      "Epoch 75/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1413 - acc: 0.8474 - val_loss: 0.1397 - val_acc: 0.8711\n",
      "Epoch 76/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1408 - acc: 0.8532 - val_loss: 0.1379 - val_acc: 0.8737\n",
      "Epoch 77/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1397 - acc: 0.8573 - val_loss: 0.1387 - val_acc: 0.8684\n",
      "Epoch 78/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.1390 - acc: 0.8608 - val_loss: 0.1358 - val_acc: 0.8763\n",
      "Epoch 79/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1378 - acc: 0.8547 - val_loss: 0.1375 - val_acc: 0.8737\n",
      "Epoch 80/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1372 - acc: 0.8547 - val_loss: 0.1341 - val_acc: 0.8737\n",
      "Epoch 81/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1345 - acc: 0.8664 - val_loss: 0.1328 - val_acc: 0.8737\n",
      "Epoch 82/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1356 - acc: 0.8596 - val_loss: 0.1344 - val_acc: 0.8737\n",
      "Epoch 83/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1345 - acc: 0.8632 - val_loss: 0.1318 - val_acc: 0.8816\n",
      "Epoch 84/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1335 - acc: 0.8637 - val_loss: 0.1337 - val_acc: 0.8763\n",
      "Epoch 85/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1333 - acc: 0.8643 - val_loss: 0.1310 - val_acc: 0.8737\n",
      "Epoch 86/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1315 - acc: 0.8699 - val_loss: 0.1304 - val_acc: 0.8842\n",
      "Epoch 87/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1295 - acc: 0.8649 - val_loss: 0.1288 - val_acc: 0.8842\n",
      "Epoch 88/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1307 - acc: 0.8649 - val_loss: 0.1282 - val_acc: 0.8868\n",
      "Epoch 89/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1276 - acc: 0.8675 - val_loss: 0.1293 - val_acc: 0.8737\n",
      "Epoch 90/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1281 - acc: 0.8699 - val_loss: 0.1262 - val_acc: 0.8868\n",
      "Epoch 91/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1267 - acc: 0.8673 - val_loss: 0.1275 - val_acc: 0.8763\n",
      "Epoch 92/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1266 - acc: 0.8722 - val_loss: 0.1249 - val_acc: 0.8868\n",
      "Epoch 93/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1243 - acc: 0.8734 - val_loss: 0.1259 - val_acc: 0.8763\n",
      "Epoch 94/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1253 - acc: 0.8678 - val_loss: 0.1242 - val_acc: 0.8895\n",
      "Epoch 95/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1247 - acc: 0.8743 - val_loss: 0.1264 - val_acc: 0.8737\n",
      "Epoch 96/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1240 - acc: 0.8699 - val_loss: 0.1205 - val_acc: 0.8816\n",
      "Epoch 97/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1238 - acc: 0.8795 - val_loss: 0.1246 - val_acc: 0.8789\n",
      "Epoch 98/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1195 - acc: 0.8775 - val_loss: 0.1216 - val_acc: 0.8816\n",
      "Epoch 99/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1211 - acc: 0.8746 - val_loss: 0.1203 - val_acc: 0.8895\n",
      "Epoch 100/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1203 - acc: 0.8769 - val_loss: 0.1201 - val_acc: 0.8842\n",
      "Epoch 101/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1191 - acc: 0.8787 - val_loss: 0.1204 - val_acc: 0.8842\n",
      "Epoch 102/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1193 - acc: 0.8766 - val_loss: 0.1191 - val_acc: 0.8816\n",
      "Epoch 103/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1167 - acc: 0.8798 - val_loss: 0.1196 - val_acc: 0.8842\n",
      "Epoch 104/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1175 - acc: 0.8763 - val_loss: 0.1174 - val_acc: 0.8868\n",
      "Epoch 105/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1163 - acc: 0.8836 - val_loss: 0.1186 - val_acc: 0.8816\n",
      "Epoch 106/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.1155 - acc: 0.8871 - val_loss: 0.1180 - val_acc: 0.8868\n",
      "Epoch 107/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1158 - acc: 0.8804 - val_loss: 0.1168 - val_acc: 0.8816\n",
      "Epoch 108/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1138 - acc: 0.8851 - val_loss: 0.1152 - val_acc: 0.8868\n",
      "Epoch 109/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1151 - acc: 0.8787 - val_loss: 0.1153 - val_acc: 0.8895\n",
      "Epoch 110/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1168 - acc: 0.8775 - val_loss: 0.1161 - val_acc: 0.8816\n",
      "Epoch 111/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.1122 - acc: 0.8825 - val_loss: 0.1140 - val_acc: 0.8868\n",
      "Epoch 112/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1129 - acc: 0.8854 - val_loss: 0.1134 - val_acc: 0.8842\n",
      "Epoch 113/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1118 - acc: 0.8839 - val_loss: 0.1144 - val_acc: 0.8868\n",
      "Epoch 114/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1113 - acc: 0.8898 - val_loss: 0.1123 - val_acc: 0.8868\n",
      "Epoch 115/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1100 - acc: 0.8889 - val_loss: 0.1132 - val_acc: 0.8868\n",
      "Epoch 116/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1114 - acc: 0.8854 - val_loss: 0.1122 - val_acc: 0.8895\n",
      "Epoch 117/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1100 - acc: 0.8901 - val_loss: 0.1129 - val_acc: 0.8868\n",
      "Epoch 118/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1085 - acc: 0.8860 - val_loss: 0.1117 - val_acc: 0.8895\n",
      "Epoch 119/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1095 - acc: 0.8874 - val_loss: 0.1128 - val_acc: 0.8921\n",
      "Epoch 120/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1075 - acc: 0.8942 - val_loss: 0.1108 - val_acc: 0.8895\n",
      "Epoch 121/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1096 - acc: 0.8877 - val_loss: 0.1104 - val_acc: 0.8895\n",
      "Epoch 122/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1073 - acc: 0.8906 - val_loss: 0.1103 - val_acc: 0.8921\n",
      "Epoch 123/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.1073 - acc: 0.8906 - val_loss: 0.1090 - val_acc: 0.8868\n",
      "Epoch 124/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1051 - acc: 0.8974 - val_loss: 0.1082 - val_acc: 0.8895\n",
      "Epoch 125/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1040 - acc: 0.8956 - val_loss: 0.1074 - val_acc: 0.8895\n",
      "Epoch 126/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1048 - acc: 0.8915 - val_loss: 0.1100 - val_acc: 0.8842\n",
      "Epoch 127/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1050 - acc: 0.8939 - val_loss: 0.1058 - val_acc: 0.8921\n",
      "Epoch 128/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1048 - acc: 0.8942 - val_loss: 0.1096 - val_acc: 0.8842\n",
      "Epoch 129/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1025 - acc: 0.8968 - val_loss: 0.1035 - val_acc: 0.8868\n",
      "Epoch 130/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1032 - acc: 0.8942 - val_loss: 0.1083 - val_acc: 0.8895\n",
      "Epoch 131/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1012 - acc: 0.8959 - val_loss: 0.1041 - val_acc: 0.8921\n",
      "Epoch 132/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1012 - acc: 0.8977 - val_loss: 0.1055 - val_acc: 0.8921\n",
      "Epoch 133/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.1029 - acc: 0.8906 - val_loss: 0.1041 - val_acc: 0.8921\n",
      "Epoch 134/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1011 - acc: 0.8930 - val_loss: 0.1045 - val_acc: 0.8947\n",
      "Epoch 135/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.1015 - acc: 0.8980 - val_loss: 0.1031 - val_acc: 0.8947\n",
      "Epoch 136/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.1004 - acc: 0.8950 - val_loss: 0.1073 - val_acc: 0.8895\n",
      "Epoch 137/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0992 - acc: 0.8980 - val_loss: 0.1036 - val_acc: 0.8895\n",
      "Epoch 138/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0997 - acc: 0.8942 - val_loss: 0.1035 - val_acc: 0.8895\n",
      "Epoch 139/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0971 - acc: 0.8985 - val_loss: 0.1049 - val_acc: 0.8895\n",
      "Epoch 140/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0994 - acc: 0.8988 - val_loss: 0.1025 - val_acc: 0.8974\n",
      "Epoch 141/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0979 - acc: 0.9023 - val_loss: 0.1052 - val_acc: 0.8868\n",
      "Epoch 142/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0979 - acc: 0.9015 - val_loss: 0.1017 - val_acc: 0.8921\n",
      "Epoch 143/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0977 - acc: 0.8997 - val_loss: 0.1027 - val_acc: 0.8895\n",
      "Epoch 144/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0973 - acc: 0.9026 - val_loss: 0.1006 - val_acc: 0.8974\n",
      "Epoch 145/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0991 - acc: 0.9012 - val_loss: 0.1019 - val_acc: 0.8895\n",
      "Epoch 146/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0979 - acc: 0.9012 - val_loss: 0.1008 - val_acc: 0.8974\n",
      "Epoch 147/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0958 - acc: 0.9047 - val_loss: 0.1027 - val_acc: 0.8895\n",
      "Epoch 148/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0963 - acc: 0.9061 - val_loss: 0.1002 - val_acc: 0.8947\n",
      "Epoch 149/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0961 - acc: 0.9026 - val_loss: 0.0994 - val_acc: 0.9000\n",
      "Epoch 150/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0958 - acc: 0.8982 - val_loss: 0.1025 - val_acc: 0.8842\n",
      "Epoch 151/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0962 - acc: 0.9015 - val_loss: 0.0975 - val_acc: 0.8974\n",
      "Epoch 152/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0965 - acc: 0.9015 - val_loss: 0.1006 - val_acc: 0.8895\n",
      "Epoch 153/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0934 - acc: 0.9041 - val_loss: 0.1004 - val_acc: 0.8895\n",
      "Epoch 154/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0951 - acc: 0.9023 - val_loss: 0.0979 - val_acc: 0.8842\n",
      "Epoch 155/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0954 - acc: 0.8994 - val_loss: 0.0985 - val_acc: 0.8921\n",
      "Epoch 156/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0942 - acc: 0.8991 - val_loss: 0.1006 - val_acc: 0.8947\n",
      "Epoch 157/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0951 - acc: 0.9032 - val_loss: 0.0970 - val_acc: 0.9000\n",
      "Epoch 158/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0931 - acc: 0.9088 - val_loss: 0.1010 - val_acc: 0.8947\n",
      "Epoch 159/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0929 - acc: 0.9041 - val_loss: 0.0967 - val_acc: 0.9026\n",
      "Epoch 160/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0899 - acc: 0.9105 - val_loss: 0.0981 - val_acc: 0.8895\n",
      "Epoch 161/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0919 - acc: 0.9018 - val_loss: 0.0959 - val_acc: 0.8947\n",
      "Epoch 162/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0920 - acc: 0.9064 - val_loss: 0.0995 - val_acc: 0.8921\n",
      "Epoch 163/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0893 - acc: 0.9120 - val_loss: 0.0960 - val_acc: 0.8895\n",
      "Epoch 164/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0897 - acc: 0.9061 - val_loss: 0.0960 - val_acc: 0.8947\n",
      "Epoch 165/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0922 - acc: 0.9061 - val_loss: 0.0959 - val_acc: 0.8921\n",
      "Epoch 166/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0896 - acc: 0.9094 - val_loss: 0.0950 - val_acc: 0.8868\n",
      "Epoch 167/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0909 - acc: 0.9061 - val_loss: 0.0991 - val_acc: 0.8895\n",
      "Epoch 168/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0903 - acc: 0.9050 - val_loss: 0.0939 - val_acc: 0.8895\n",
      "Epoch 169/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0894 - acc: 0.9085 - val_loss: 0.0948 - val_acc: 0.9000\n",
      "Epoch 170/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0885 - acc: 0.9102 - val_loss: 0.0956 - val_acc: 0.8921\n",
      "Epoch 171/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0875 - acc: 0.9096 - val_loss: 0.0936 - val_acc: 0.8947\n",
      "Epoch 172/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0893 - acc: 0.9050 - val_loss: 0.0948 - val_acc: 0.8921\n",
      "Epoch 173/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0892 - acc: 0.9070 - val_loss: 0.0943 - val_acc: 0.8921\n",
      "Epoch 174/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0898 - acc: 0.9032 - val_loss: 0.0934 - val_acc: 0.8921\n",
      "Epoch 175/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0884 - acc: 0.9061 - val_loss: 0.0952 - val_acc: 0.8947\n",
      "Epoch 176/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0880 - acc: 0.9091 - val_loss: 0.0934 - val_acc: 0.8921\n",
      "Epoch 177/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0883 - acc: 0.9164 - val_loss: 0.0952 - val_acc: 0.8921\n",
      "Epoch 178/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0862 - acc: 0.9073 - val_loss: 0.0920 - val_acc: 0.8947\n",
      "Epoch 179/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0872 - acc: 0.9120 - val_loss: 0.0963 - val_acc: 0.8895\n",
      "Epoch 180/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0863 - acc: 0.9129 - val_loss: 0.0908 - val_acc: 0.8974\n",
      "Epoch 181/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0865 - acc: 0.9120 - val_loss: 0.0941 - val_acc: 0.8947\n",
      "Epoch 182/1000\n",
      "14/14 [==============================] - 0s 30ms/step - loss: 0.0878 - acc: 0.9094 - val_loss: 0.0931 - val_acc: 0.8974\n",
      "Epoch 183/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0856 - acc: 0.9102 - val_loss: 0.0921 - val_acc: 0.8921\n",
      "Epoch 184/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0861 - acc: 0.9120 - val_loss: 0.0918 - val_acc: 0.8974\n",
      "Epoch 185/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0865 - acc: 0.9105 - val_loss: 0.0913 - val_acc: 0.9026\n",
      "Epoch 186/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0851 - acc: 0.9094 - val_loss: 0.0914 - val_acc: 0.8947\n",
      "Epoch 187/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0859 - acc: 0.9117 - val_loss: 0.0920 - val_acc: 0.8947\n",
      "Epoch 188/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0868 - acc: 0.9117 - val_loss: 0.0904 - val_acc: 0.8974\n",
      "Epoch 189/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0825 - acc: 0.9178 - val_loss: 0.0911 - val_acc: 0.9000\n",
      "Epoch 190/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0854 - acc: 0.9094 - val_loss: 0.0903 - val_acc: 0.8974\n",
      "Epoch 191/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0850 - acc: 0.9079 - val_loss: 0.0903 - val_acc: 0.8947\n",
      "Epoch 192/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0826 - acc: 0.9161 - val_loss: 0.0913 - val_acc: 0.9000\n",
      "Epoch 193/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0839 - acc: 0.9146 - val_loss: 0.0924 - val_acc: 0.8947\n",
      "Epoch 194/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0813 - acc: 0.9120 - val_loss: 0.0891 - val_acc: 0.9000\n",
      "Epoch 195/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0809 - acc: 0.9199 - val_loss: 0.0895 - val_acc: 0.8947\n",
      "Epoch 196/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0837 - acc: 0.9123 - val_loss: 0.0900 - val_acc: 0.8947\n",
      "Epoch 197/1000\n",
      "14/14 [==============================] - 0s 30ms/step - loss: 0.0827 - acc: 0.9120 - val_loss: 0.0872 - val_acc: 0.9026\n",
      "Epoch 198/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0839 - acc: 0.9070 - val_loss: 0.0890 - val_acc: 0.9000\n",
      "Epoch 199/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0818 - acc: 0.9164 - val_loss: 0.0892 - val_acc: 0.8947\n",
      "Epoch 200/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0826 - acc: 0.9167 - val_loss: 0.0888 - val_acc: 0.9000\n",
      "Epoch 201/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0831 - acc: 0.9146 - val_loss: 0.0881 - val_acc: 0.9053\n",
      "Epoch 202/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0833 - acc: 0.9135 - val_loss: 0.0882 - val_acc: 0.8974\n",
      "Epoch 203/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0806 - acc: 0.9164 - val_loss: 0.0881 - val_acc: 0.9053\n",
      "Epoch 204/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0794 - acc: 0.9196 - val_loss: 0.0889 - val_acc: 0.9026\n",
      "Epoch 205/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0815 - acc: 0.9184 - val_loss: 0.0881 - val_acc: 0.9000\n",
      "Epoch 206/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0822 - acc: 0.9135 - val_loss: 0.0860 - val_acc: 0.9026\n",
      "Epoch 207/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0796 - acc: 0.9181 - val_loss: 0.0881 - val_acc: 0.8974\n",
      "Epoch 208/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0818 - acc: 0.9175 - val_loss: 0.0872 - val_acc: 0.9000\n",
      "Epoch 209/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0814 - acc: 0.9181 - val_loss: 0.0868 - val_acc: 0.8974\n",
      "Epoch 210/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0785 - acc: 0.9193 - val_loss: 0.0882 - val_acc: 0.8974\n",
      "Epoch 211/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0808 - acc: 0.9129 - val_loss: 0.0867 - val_acc: 0.8947\n",
      "Epoch 212/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0788 - acc: 0.9173 - val_loss: 0.0898 - val_acc: 0.8947\n",
      "Epoch 213/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0807 - acc: 0.9155 - val_loss: 0.0854 - val_acc: 0.9000\n",
      "Epoch 214/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0806 - acc: 0.9190 - val_loss: 0.0888 - val_acc: 0.8974\n",
      "Epoch 215/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0806 - acc: 0.9164 - val_loss: 0.0867 - val_acc: 0.9053\n",
      "Epoch 216/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0807 - acc: 0.9187 - val_loss: 0.0867 - val_acc: 0.9026\n",
      "Epoch 217/1000\n",
      "14/14 [==============================] - 0s 25ms/step - loss: 0.0771 - acc: 0.9228 - val_loss: 0.0855 - val_acc: 0.8974\n",
      "Epoch 218/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0772 - acc: 0.9216 - val_loss: 0.0853 - val_acc: 0.8974\n",
      "Epoch 219/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0762 - acc: 0.9251 - val_loss: 0.0846 - val_acc: 0.9105\n",
      "Epoch 220/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.0801 - acc: 0.9193 - val_loss: 0.0846 - val_acc: 0.8947\n",
      "Epoch 221/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0782 - acc: 0.9211 - val_loss: 0.0862 - val_acc: 0.9000\n",
      "Epoch 222/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0784 - acc: 0.9216 - val_loss: 0.0859 - val_acc: 0.8974\n",
      "Epoch 223/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.0760 - acc: 0.9243 - val_loss: 0.0857 - val_acc: 0.9026\n",
      "Epoch 224/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0783 - acc: 0.9231 - val_loss: 0.0855 - val_acc: 0.8974\n",
      "Epoch 225/1000\n",
      "14/14 [==============================] - 0s 27ms/step - loss: 0.0790 - acc: 0.9216 - val_loss: 0.0822 - val_acc: 0.9053\n",
      "Epoch 226/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0766 - acc: 0.9249 - val_loss: 0.0850 - val_acc: 0.9000\n",
      "Epoch 227/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0770 - acc: 0.9225 - val_loss: 0.0861 - val_acc: 0.8974\n",
      "Epoch 228/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0781 - acc: 0.9205 - val_loss: 0.0834 - val_acc: 0.9053\n",
      "Epoch 229/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0766 - acc: 0.9225 - val_loss: 0.0831 - val_acc: 0.9105\n",
      "Epoch 230/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0771 - acc: 0.9193 - val_loss: 0.0831 - val_acc: 0.9026\n",
      "Epoch 231/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0757 - acc: 0.9211 - val_loss: 0.0843 - val_acc: 0.8974\n",
      "Epoch 232/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0744 - acc: 0.9254 - val_loss: 0.0827 - val_acc: 0.9105\n",
      "Epoch 233/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0761 - acc: 0.9228 - val_loss: 0.0860 - val_acc: 0.9000\n",
      "Epoch 234/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0772 - acc: 0.9251 - val_loss: 0.0818 - val_acc: 0.9079\n",
      "Epoch 235/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0762 - acc: 0.9275 - val_loss: 0.0846 - val_acc: 0.8974\n",
      "Epoch 236/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0766 - acc: 0.9193 - val_loss: 0.0817 - val_acc: 0.9105\n",
      "Epoch 237/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0734 - acc: 0.9246 - val_loss: 0.0815 - val_acc: 0.9079\n",
      "Epoch 238/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0787 - acc: 0.9193 - val_loss: 0.0809 - val_acc: 0.9105\n",
      "Epoch 239/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0746 - acc: 0.9249 - val_loss: 0.0822 - val_acc: 0.9079\n",
      "Epoch 240/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0744 - acc: 0.9243 - val_loss: 0.0831 - val_acc: 0.9026\n",
      "Epoch 241/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0777 - acc: 0.9240 - val_loss: 0.0801 - val_acc: 0.9132\n",
      "Epoch 242/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0755 - acc: 0.9225 - val_loss: 0.0823 - val_acc: 0.9105\n",
      "Epoch 243/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0762 - acc: 0.9222 - val_loss: 0.0847 - val_acc: 0.9000\n",
      "Epoch 244/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0750 - acc: 0.9249 - val_loss: 0.0799 - val_acc: 0.9105\n",
      "Epoch 245/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0760 - acc: 0.9175 - val_loss: 0.0859 - val_acc: 0.9026\n",
      "Epoch 246/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0739 - acc: 0.9213 - val_loss: 0.0803 - val_acc: 0.9053\n",
      "Epoch 247/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0751 - acc: 0.9213 - val_loss: 0.0810 - val_acc: 0.9105\n",
      "Epoch 248/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0751 - acc: 0.9219 - val_loss: 0.0808 - val_acc: 0.9026\n",
      "Epoch 249/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0737 - acc: 0.9249 - val_loss: 0.0790 - val_acc: 0.9132\n",
      "Epoch 250/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0732 - acc: 0.9284 - val_loss: 0.0837 - val_acc: 0.8947\n",
      "Epoch 251/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0771 - acc: 0.9225 - val_loss: 0.0777 - val_acc: 0.9158\n",
      "Epoch 252/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0748 - acc: 0.9231 - val_loss: 0.0823 - val_acc: 0.9053\n",
      "Epoch 253/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0737 - acc: 0.9266 - val_loss: 0.0779 - val_acc: 0.9053\n",
      "Epoch 254/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0744 - acc: 0.9249 - val_loss: 0.0831 - val_acc: 0.9026\n",
      "Epoch 255/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0736 - acc: 0.9266 - val_loss: 0.0794 - val_acc: 0.9132\n",
      "Epoch 256/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0752 - acc: 0.9167 - val_loss: 0.0823 - val_acc: 0.9053\n",
      "Epoch 257/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0751 - acc: 0.9263 - val_loss: 0.0780 - val_acc: 0.9105\n",
      "Epoch 258/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0726 - acc: 0.9266 - val_loss: 0.0801 - val_acc: 0.9105\n",
      "Epoch 259/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0712 - acc: 0.9266 - val_loss: 0.0806 - val_acc: 0.9079\n",
      "Epoch 260/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0737 - acc: 0.9287 - val_loss: 0.0784 - val_acc: 0.9132\n",
      "Epoch 261/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0729 - acc: 0.9287 - val_loss: 0.0779 - val_acc: 0.9132\n",
      "Epoch 262/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0718 - acc: 0.9257 - val_loss: 0.0798 - val_acc: 0.9105\n",
      "Epoch 263/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0710 - acc: 0.9263 - val_loss: 0.0791 - val_acc: 0.9132\n",
      "Epoch 264/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0723 - acc: 0.9266 - val_loss: 0.0787 - val_acc: 0.9105\n",
      "Epoch 265/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0728 - acc: 0.9254 - val_loss: 0.0786 - val_acc: 0.9158\n",
      "Epoch 266/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0711 - acc: 0.9289 - val_loss: 0.0811 - val_acc: 0.9000\n",
      "Epoch 267/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0711 - acc: 0.9301 - val_loss: 0.0765 - val_acc: 0.9132\n",
      "Epoch 268/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0717 - acc: 0.9263 - val_loss: 0.0806 - val_acc: 0.9026\n",
      "Epoch 269/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0699 - acc: 0.9325 - val_loss: 0.0766 - val_acc: 0.9158\n",
      "Epoch 270/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0721 - acc: 0.9260 - val_loss: 0.0782 - val_acc: 0.9105\n",
      "Epoch 271/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0729 - acc: 0.9254 - val_loss: 0.0802 - val_acc: 0.9053\n",
      "Epoch 272/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0724 - acc: 0.9269 - val_loss: 0.0749 - val_acc: 0.9132\n",
      "Epoch 273/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0723 - acc: 0.9263 - val_loss: 0.0779 - val_acc: 0.9105\n",
      "Epoch 274/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0714 - acc: 0.9251 - val_loss: 0.0753 - val_acc: 0.9184\n",
      "Epoch 275/1000\n",
      "14/14 [==============================] - 0s 30ms/step - loss: 0.0714 - acc: 0.9246 - val_loss: 0.0793 - val_acc: 0.9105\n",
      "Epoch 276/1000\n",
      "14/14 [==============================] - 0s 28ms/step - loss: 0.0700 - acc: 0.9298 - val_loss: 0.0766 - val_acc: 0.9132\n",
      "Epoch 277/1000\n",
      "14/14 [==============================] - 0s 28ms/step - loss: 0.0710 - acc: 0.9269 - val_loss: 0.0777 - val_acc: 0.9132\n",
      "Epoch 278/1000\n",
      "14/14 [==============================] - 0s 30ms/step - loss: 0.0694 - acc: 0.9298 - val_loss: 0.0781 - val_acc: 0.9105\n",
      "Epoch 279/1000\n",
      "14/14 [==============================] - 0s 32ms/step - loss: 0.0709 - acc: 0.9284 - val_loss: 0.0768 - val_acc: 0.9158\n",
      "Epoch 280/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0695 - acc: 0.9289 - val_loss: 0.0754 - val_acc: 0.9184\n",
      "Epoch 281/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0688 - acc: 0.9313 - val_loss: 0.0766 - val_acc: 0.9132\n",
      "Epoch 282/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0692 - acc: 0.9313 - val_loss: 0.0745 - val_acc: 0.9211\n",
      "Epoch 283/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0688 - acc: 0.9289 - val_loss: 0.0738 - val_acc: 0.9237\n",
      "Epoch 284/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0700 - acc: 0.9260 - val_loss: 0.0757 - val_acc: 0.9158\n",
      "Epoch 285/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0701 - acc: 0.9310 - val_loss: 0.0755 - val_acc: 0.9158\n",
      "Epoch 286/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0703 - acc: 0.9295 - val_loss: 0.0784 - val_acc: 0.9132\n",
      "Epoch 287/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0712 - acc: 0.9269 - val_loss: 0.0717 - val_acc: 0.9211\n",
      "Epoch 288/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0677 - acc: 0.9298 - val_loss: 0.0753 - val_acc: 0.9184\n",
      "Epoch 289/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0699 - acc: 0.9289 - val_loss: 0.0736 - val_acc: 0.9263\n",
      "Epoch 290/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0691 - acc: 0.9295 - val_loss: 0.0730 - val_acc: 0.9289\n",
      "Epoch 291/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0700 - acc: 0.9289 - val_loss: 0.0759 - val_acc: 0.9132\n",
      "Epoch 292/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0683 - acc: 0.9351 - val_loss: 0.0755 - val_acc: 0.9158\n",
      "Epoch 293/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0683 - acc: 0.9275 - val_loss: 0.0746 - val_acc: 0.9184\n",
      "Epoch 294/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0685 - acc: 0.9304 - val_loss: 0.0742 - val_acc: 0.9211\n",
      "Epoch 295/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0704 - acc: 0.9284 - val_loss: 0.0768 - val_acc: 0.9184\n",
      "Epoch 296/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0682 - acc: 0.9292 - val_loss: 0.0735 - val_acc: 0.9184\n",
      "Epoch 297/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0666 - acc: 0.9377 - val_loss: 0.0740 - val_acc: 0.9211\n",
      "Epoch 298/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0666 - acc: 0.9336 - val_loss: 0.0719 - val_acc: 0.9237\n",
      "Epoch 299/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0678 - acc: 0.9316 - val_loss: 0.0725 - val_acc: 0.9237\n",
      "Epoch 300/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0692 - acc: 0.9322 - val_loss: 0.0734 - val_acc: 0.9237\n",
      "Epoch 301/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0675 - acc: 0.9313 - val_loss: 0.0708 - val_acc: 0.9237\n",
      "Epoch 302/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0679 - acc: 0.9330 - val_loss: 0.0744 - val_acc: 0.9211\n",
      "Epoch 303/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0690 - acc: 0.9269 - val_loss: 0.0703 - val_acc: 0.9289\n",
      "Epoch 304/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0679 - acc: 0.9325 - val_loss: 0.0742 - val_acc: 0.9211\n",
      "Epoch 305/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0698 - acc: 0.9298 - val_loss: 0.0712 - val_acc: 0.9237\n",
      "Epoch 306/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0668 - acc: 0.9325 - val_loss: 0.0745 - val_acc: 0.9211\n",
      "Epoch 307/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0660 - acc: 0.9351 - val_loss: 0.0693 - val_acc: 0.9237\n",
      "Epoch 308/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0668 - acc: 0.9377 - val_loss: 0.0729 - val_acc: 0.9237\n",
      "Epoch 309/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0653 - acc: 0.9322 - val_loss: 0.0704 - val_acc: 0.9211\n",
      "Epoch 310/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0676 - acc: 0.9360 - val_loss: 0.0739 - val_acc: 0.9211\n",
      "Epoch 311/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0672 - acc: 0.9319 - val_loss: 0.0724 - val_acc: 0.9263\n",
      "Epoch 312/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0668 - acc: 0.9342 - val_loss: 0.0697 - val_acc: 0.9289\n",
      "Epoch 313/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0663 - acc: 0.9339 - val_loss: 0.0734 - val_acc: 0.9237\n",
      "Epoch 314/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0663 - acc: 0.9327 - val_loss: 0.0686 - val_acc: 0.9263\n",
      "Epoch 315/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0662 - acc: 0.9325 - val_loss: 0.0718 - val_acc: 0.9289\n",
      "Epoch 316/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0661 - acc: 0.9348 - val_loss: 0.0733 - val_acc: 0.9263\n",
      "Epoch 317/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0663 - acc: 0.9325 - val_loss: 0.0692 - val_acc: 0.9263\n",
      "Epoch 318/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0647 - acc: 0.9339 - val_loss: 0.0718 - val_acc: 0.9184\n",
      "Epoch 319/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0635 - acc: 0.9371 - val_loss: 0.0675 - val_acc: 0.9263\n",
      "Epoch 320/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0664 - acc: 0.9389 - val_loss: 0.0736 - val_acc: 0.9263\n",
      "Epoch 321/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0645 - acc: 0.9354 - val_loss: 0.0676 - val_acc: 0.9263\n",
      "Epoch 322/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0668 - acc: 0.9319 - val_loss: 0.0725 - val_acc: 0.9289\n",
      "Epoch 323/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0651 - acc: 0.9336 - val_loss: 0.0686 - val_acc: 0.9263\n",
      "Epoch 324/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0658 - acc: 0.9345 - val_loss: 0.0723 - val_acc: 0.9263\n",
      "Epoch 325/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0620 - acc: 0.9398 - val_loss: 0.0681 - val_acc: 0.9263\n",
      "Epoch 326/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0632 - acc: 0.9386 - val_loss: 0.0706 - val_acc: 0.9237\n",
      "Epoch 327/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0632 - acc: 0.9374 - val_loss: 0.0681 - val_acc: 0.9289\n",
      "Epoch 328/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0642 - acc: 0.9398 - val_loss: 0.0679 - val_acc: 0.9316\n",
      "Epoch 329/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0657 - acc: 0.9345 - val_loss: 0.0695 - val_acc: 0.9316\n",
      "Epoch 330/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0641 - acc: 0.9368 - val_loss: 0.0704 - val_acc: 0.9237\n",
      "Epoch 331/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0645 - acc: 0.9363 - val_loss: 0.0670 - val_acc: 0.9289\n",
      "Epoch 332/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0655 - acc: 0.9345 - val_loss: 0.0726 - val_acc: 0.9237\n",
      "Epoch 333/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0656 - acc: 0.9360 - val_loss: 0.0644 - val_acc: 0.9342\n",
      "Epoch 334/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0624 - acc: 0.9365 - val_loss: 0.0718 - val_acc: 0.9263\n",
      "Epoch 335/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0659 - acc: 0.9360 - val_loss: 0.0655 - val_acc: 0.9368\n",
      "Epoch 336/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0642 - acc: 0.9345 - val_loss: 0.0693 - val_acc: 0.9289\n",
      "Epoch 337/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0624 - acc: 0.9427 - val_loss: 0.0676 - val_acc: 0.9289\n",
      "Epoch 338/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0642 - acc: 0.9377 - val_loss: 0.0679 - val_acc: 0.9263\n",
      "Epoch 339/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0621 - acc: 0.9395 - val_loss: 0.0687 - val_acc: 0.9316\n",
      "Epoch 340/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0614 - acc: 0.9377 - val_loss: 0.0672 - val_acc: 0.9263\n",
      "Epoch 341/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0638 - acc: 0.9360 - val_loss: 0.0682 - val_acc: 0.9289\n",
      "Epoch 342/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0614 - acc: 0.9386 - val_loss: 0.0684 - val_acc: 0.9263\n",
      "Epoch 343/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0634 - acc: 0.9380 - val_loss: 0.0659 - val_acc: 0.9289\n",
      "Epoch 344/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0634 - acc: 0.9395 - val_loss: 0.0668 - val_acc: 0.9289\n",
      "Epoch 345/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0630 - acc: 0.9365 - val_loss: 0.0693 - val_acc: 0.9289\n",
      "Epoch 346/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0639 - acc: 0.9368 - val_loss: 0.0669 - val_acc: 0.9342\n",
      "Epoch 347/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0613 - acc: 0.9398 - val_loss: 0.0660 - val_acc: 0.9342\n",
      "Epoch 348/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0617 - acc: 0.9377 - val_loss: 0.0672 - val_acc: 0.9316\n",
      "Epoch 349/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0610 - acc: 0.9380 - val_loss: 0.0656 - val_acc: 0.9316\n",
      "Epoch 350/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0642 - acc: 0.9368 - val_loss: 0.0677 - val_acc: 0.9316\n",
      "Epoch 351/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0628 - acc: 0.9342 - val_loss: 0.0642 - val_acc: 0.9342\n",
      "Epoch 352/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0617 - acc: 0.9418 - val_loss: 0.0669 - val_acc: 0.9289\n",
      "Epoch 353/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0623 - acc: 0.9392 - val_loss: 0.0658 - val_acc: 0.9316\n",
      "Epoch 354/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0638 - acc: 0.9386 - val_loss: 0.0647 - val_acc: 0.9342\n",
      "Epoch 355/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0611 - acc: 0.9395 - val_loss: 0.0665 - val_acc: 0.9368\n",
      "Epoch 356/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0601 - acc: 0.9409 - val_loss: 0.0637 - val_acc: 0.9368\n",
      "Epoch 357/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0608 - acc: 0.9406 - val_loss: 0.0650 - val_acc: 0.9368\n",
      "Epoch 358/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0621 - acc: 0.9383 - val_loss: 0.0678 - val_acc: 0.9289\n",
      "Epoch 359/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0626 - acc: 0.9380 - val_loss: 0.0659 - val_acc: 0.9342\n",
      "Epoch 360/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0615 - acc: 0.9392 - val_loss: 0.0663 - val_acc: 0.9289\n",
      "Epoch 361/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0612 - acc: 0.9418 - val_loss: 0.0641 - val_acc: 0.9342\n",
      "Epoch 362/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0606 - acc: 0.9424 - val_loss: 0.0636 - val_acc: 0.9316\n",
      "Epoch 363/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0622 - acc: 0.9386 - val_loss: 0.0605 - val_acc: 0.9421\n",
      "Epoch 364/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0613 - acc: 0.9412 - val_loss: 0.0668 - val_acc: 0.9316\n",
      "Epoch 365/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0611 - acc: 0.9406 - val_loss: 0.0629 - val_acc: 0.9368\n",
      "Epoch 366/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0617 - acc: 0.9415 - val_loss: 0.0676 - val_acc: 0.9289\n",
      "Epoch 367/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0617 - acc: 0.9404 - val_loss: 0.0613 - val_acc: 0.9447\n",
      "Epoch 368/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0631 - acc: 0.9363 - val_loss: 0.0644 - val_acc: 0.9342\n",
      "Epoch 369/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0596 - acc: 0.9398 - val_loss: 0.0661 - val_acc: 0.9289\n",
      "Epoch 370/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0615 - acc: 0.9383 - val_loss: 0.0652 - val_acc: 0.9342\n",
      "Epoch 371/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0595 - acc: 0.9433 - val_loss: 0.0660 - val_acc: 0.9289\n",
      "Epoch 372/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0604 - acc: 0.9418 - val_loss: 0.0606 - val_acc: 0.9395\n",
      "Epoch 373/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0606 - acc: 0.9398 - val_loss: 0.0643 - val_acc: 0.9342\n",
      "Epoch 374/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0580 - acc: 0.9406 - val_loss: 0.0626 - val_acc: 0.9395\n",
      "Epoch 375/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0584 - acc: 0.9401 - val_loss: 0.0631 - val_acc: 0.9368\n",
      "Epoch 376/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0601 - acc: 0.9436 - val_loss: 0.0628 - val_acc: 0.9368\n",
      "Epoch 377/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0594 - acc: 0.9368 - val_loss: 0.0630 - val_acc: 0.9316\n",
      "Epoch 378/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0604 - acc: 0.9406 - val_loss: 0.0634 - val_acc: 0.9342\n",
      "Epoch 379/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0597 - acc: 0.9427 - val_loss: 0.0621 - val_acc: 0.9368\n",
      "Epoch 380/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0606 - acc: 0.9395 - val_loss: 0.0598 - val_acc: 0.9421\n",
      "Epoch 381/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0602 - acc: 0.9383 - val_loss: 0.0626 - val_acc: 0.9368\n",
      "Epoch 382/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0583 - acc: 0.9442 - val_loss: 0.0621 - val_acc: 0.9342\n",
      "Epoch 383/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0583 - acc: 0.9433 - val_loss: 0.0618 - val_acc: 0.9368\n",
      "Epoch 384/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0613 - acc: 0.9401 - val_loss: 0.0629 - val_acc: 0.9342\n",
      "Epoch 385/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0594 - acc: 0.9430 - val_loss: 0.0623 - val_acc: 0.9368\n",
      "Epoch 386/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0596 - acc: 0.9418 - val_loss: 0.0618 - val_acc: 0.9342\n",
      "Epoch 387/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0580 - acc: 0.9436 - val_loss: 0.0593 - val_acc: 0.9447\n",
      "Epoch 388/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0586 - acc: 0.9418 - val_loss: 0.0622 - val_acc: 0.9342\n",
      "Epoch 389/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0584 - acc: 0.9404 - val_loss: 0.0650 - val_acc: 0.9289\n",
      "Epoch 390/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0597 - acc: 0.9421 - val_loss: 0.0594 - val_acc: 0.9395\n",
      "Epoch 391/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0591 - acc: 0.9442 - val_loss: 0.0622 - val_acc: 0.9368\n",
      "Epoch 392/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0584 - acc: 0.9406 - val_loss: 0.0614 - val_acc: 0.9368\n",
      "Epoch 393/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0616 - acc: 0.9377 - val_loss: 0.0604 - val_acc: 0.9395\n",
      "Epoch 394/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0601 - acc: 0.9404 - val_loss: 0.0623 - val_acc: 0.9368\n",
      "Epoch 395/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0589 - acc: 0.9430 - val_loss: 0.0619 - val_acc: 0.9342\n",
      "Epoch 396/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0577 - acc: 0.9427 - val_loss: 0.0615 - val_acc: 0.9395\n",
      "Epoch 397/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0580 - acc: 0.9436 - val_loss: 0.0612 - val_acc: 0.9316\n",
      "Epoch 398/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0587 - acc: 0.9433 - val_loss: 0.0618 - val_acc: 0.9368\n",
      "Epoch 399/1000\n",
      "14/14 [==============================] - 0s 25ms/step - loss: 0.0584 - acc: 0.9471 - val_loss: 0.0610 - val_acc: 0.9368\n",
      "Epoch 400/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0588 - acc: 0.9389 - val_loss: 0.0586 - val_acc: 0.9421\n",
      "Epoch 401/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0578 - acc: 0.9427 - val_loss: 0.0601 - val_acc: 0.9395\n",
      "Epoch 402/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0588 - acc: 0.9395 - val_loss: 0.0603 - val_acc: 0.9342\n",
      "Epoch 403/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0577 - acc: 0.9406 - val_loss: 0.0597 - val_acc: 0.9342\n",
      "Epoch 404/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0571 - acc: 0.9453 - val_loss: 0.0605 - val_acc: 0.9342\n",
      "Epoch 405/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0572 - acc: 0.9421 - val_loss: 0.0589 - val_acc: 0.9395\n",
      "Epoch 406/1000\n",
      "14/14 [==============================] - 0s 34ms/step - loss: 0.0566 - acc: 0.9421 - val_loss: 0.0632 - val_acc: 0.9316\n",
      "Epoch 407/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0553 - acc: 0.9444 - val_loss: 0.0581 - val_acc: 0.9421\n",
      "Epoch 408/1000\n",
      "14/14 [==============================] - 0s 29ms/step - loss: 0.0562 - acc: 0.9456 - val_loss: 0.0598 - val_acc: 0.9368\n",
      "Epoch 409/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0563 - acc: 0.9450 - val_loss: 0.0576 - val_acc: 0.9395\n",
      "Epoch 410/1000\n",
      "14/14 [==============================] - 0s 31ms/step - loss: 0.0558 - acc: 0.9462 - val_loss: 0.0628 - val_acc: 0.9395\n",
      "Epoch 411/1000\n",
      "14/14 [==============================] - 0s 28ms/step - loss: 0.0558 - acc: 0.9459 - val_loss: 0.0584 - val_acc: 0.9395\n",
      "Epoch 412/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0586 - acc: 0.9401 - val_loss: 0.0581 - val_acc: 0.9421\n",
      "Epoch 413/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0569 - acc: 0.9453 - val_loss: 0.0592 - val_acc: 0.9395\n",
      "Epoch 414/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0560 - acc: 0.9439 - val_loss: 0.0591 - val_acc: 0.9395\n",
      "Epoch 415/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0567 - acc: 0.9459 - val_loss: 0.0564 - val_acc: 0.9474\n",
      "Epoch 416/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0565 - acc: 0.9439 - val_loss: 0.0622 - val_acc: 0.9316\n",
      "Epoch 417/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0563 - acc: 0.9453 - val_loss: 0.0576 - val_acc: 0.9421\n",
      "Epoch 418/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0589 - acc: 0.9415 - val_loss: 0.0593 - val_acc: 0.9395\n",
      "Epoch 419/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0548 - acc: 0.9468 - val_loss: 0.0569 - val_acc: 0.9447\n",
      "Epoch 420/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0572 - acc: 0.9447 - val_loss: 0.0603 - val_acc: 0.9316\n",
      "Epoch 421/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0561 - acc: 0.9421 - val_loss: 0.0562 - val_acc: 0.9447\n",
      "Epoch 422/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0545 - acc: 0.9450 - val_loss: 0.0596 - val_acc: 0.9368\n",
      "Epoch 423/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0553 - acc: 0.9462 - val_loss: 0.0553 - val_acc: 0.9474\n",
      "Epoch 424/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0594 - acc: 0.9433 - val_loss: 0.0587 - val_acc: 0.9421\n",
      "Epoch 425/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0556 - acc: 0.9468 - val_loss: 0.0574 - val_acc: 0.9421\n",
      "Epoch 426/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0540 - acc: 0.9482 - val_loss: 0.0569 - val_acc: 0.9395\n",
      "Epoch 427/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0563 - acc: 0.9444 - val_loss: 0.0562 - val_acc: 0.9395\n",
      "Epoch 428/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0545 - acc: 0.9462 - val_loss: 0.0573 - val_acc: 0.9395\n",
      "Epoch 429/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0562 - acc: 0.9436 - val_loss: 0.0570 - val_acc: 0.9395\n",
      "Epoch 430/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0553 - acc: 0.9468 - val_loss: 0.0587 - val_acc: 0.9368\n",
      "Epoch 431/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0553 - acc: 0.9444 - val_loss: 0.0570 - val_acc: 0.9395\n",
      "Epoch 432/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0535 - acc: 0.9477 - val_loss: 0.0598 - val_acc: 0.9342\n",
      "Epoch 433/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0565 - acc: 0.9424 - val_loss: 0.0565 - val_acc: 0.9395\n",
      "Epoch 434/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0570 - acc: 0.9453 - val_loss: 0.0576 - val_acc: 0.9395\n",
      "Epoch 435/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0559 - acc: 0.9477 - val_loss: 0.0555 - val_acc: 0.9421\n",
      "Epoch 436/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0548 - acc: 0.9491 - val_loss: 0.0576 - val_acc: 0.9368\n",
      "Epoch 437/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0544 - acc: 0.9488 - val_loss: 0.0551 - val_acc: 0.9447\n",
      "Epoch 438/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0556 - acc: 0.9465 - val_loss: 0.0566 - val_acc: 0.9395\n",
      "Epoch 439/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0559 - acc: 0.9465 - val_loss: 0.0562 - val_acc: 0.9421\n",
      "Epoch 440/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0534 - acc: 0.9459 - val_loss: 0.0568 - val_acc: 0.9421\n",
      "Epoch 441/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0554 - acc: 0.9456 - val_loss: 0.0550 - val_acc: 0.9421\n",
      "Epoch 442/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0555 - acc: 0.9433 - val_loss: 0.0563 - val_acc: 0.9421\n",
      "Epoch 443/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0546 - acc: 0.9459 - val_loss: 0.0570 - val_acc: 0.9421\n",
      "Epoch 444/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0549 - acc: 0.9482 - val_loss: 0.0536 - val_acc: 0.9447\n",
      "Epoch 445/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0524 - acc: 0.9518 - val_loss: 0.0561 - val_acc: 0.9447\n",
      "Epoch 446/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0548 - acc: 0.9471 - val_loss: 0.0553 - val_acc: 0.9447\n",
      "Epoch 447/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0532 - acc: 0.9515 - val_loss: 0.0555 - val_acc: 0.9395\n",
      "Epoch 448/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0527 - acc: 0.9453 - val_loss: 0.0547 - val_acc: 0.9421\n",
      "Epoch 449/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0539 - acc: 0.9474 - val_loss: 0.0532 - val_acc: 0.9447\n",
      "Epoch 450/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0552 - acc: 0.9468 - val_loss: 0.0575 - val_acc: 0.9395\n",
      "Epoch 451/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0548 - acc: 0.9468 - val_loss: 0.0529 - val_acc: 0.9500\n",
      "Epoch 452/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0543 - acc: 0.9491 - val_loss: 0.0563 - val_acc: 0.9395\n",
      "Epoch 453/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0541 - acc: 0.9494 - val_loss: 0.0537 - val_acc: 0.9447\n",
      "Epoch 454/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0534 - acc: 0.9447 - val_loss: 0.0538 - val_acc: 0.9421\n",
      "Epoch 455/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0561 - acc: 0.9439 - val_loss: 0.0556 - val_acc: 0.9421\n",
      "Epoch 456/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0543 - acc: 0.9474 - val_loss: 0.0564 - val_acc: 0.9421\n",
      "Epoch 457/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0552 - acc: 0.9477 - val_loss: 0.0517 - val_acc: 0.9526\n",
      "Epoch 458/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0543 - acc: 0.9497 - val_loss: 0.0560 - val_acc: 0.9421\n",
      "Epoch 459/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0542 - acc: 0.9477 - val_loss: 0.0529 - val_acc: 0.9447\n",
      "Epoch 460/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0519 - acc: 0.9497 - val_loss: 0.0563 - val_acc: 0.9421\n",
      "Epoch 461/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0541 - acc: 0.9471 - val_loss: 0.0524 - val_acc: 0.9447\n",
      "Epoch 462/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0532 - acc: 0.9456 - val_loss: 0.0573 - val_acc: 0.9447\n",
      "Epoch 463/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0514 - acc: 0.9500 - val_loss: 0.0535 - val_acc: 0.9447\n",
      "Epoch 464/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0514 - acc: 0.9477 - val_loss: 0.0542 - val_acc: 0.9474\n",
      "Epoch 465/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0532 - acc: 0.9456 - val_loss: 0.0533 - val_acc: 0.9474\n",
      "Epoch 466/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0528 - acc: 0.9523 - val_loss: 0.0518 - val_acc: 0.9447\n",
      "Epoch 467/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0520 - acc: 0.9506 - val_loss: 0.0532 - val_acc: 0.9447\n",
      "Epoch 468/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0534 - acc: 0.9488 - val_loss: 0.0541 - val_acc: 0.9447\n",
      "Epoch 469/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0534 - acc: 0.9468 - val_loss: 0.0504 - val_acc: 0.9474\n",
      "Epoch 470/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0520 - acc: 0.9491 - val_loss: 0.0547 - val_acc: 0.9474\n",
      "Epoch 471/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0505 - acc: 0.9518 - val_loss: 0.0507 - val_acc: 0.9500\n",
      "Epoch 472/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0530 - acc: 0.9477 - val_loss: 0.0544 - val_acc: 0.9447\n",
      "Epoch 473/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0519 - acc: 0.9468 - val_loss: 0.0553 - val_acc: 0.9447\n",
      "Epoch 474/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0541 - acc: 0.9459 - val_loss: 0.0497 - val_acc: 0.9500\n",
      "Epoch 475/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0504 - acc: 0.9523 - val_loss: 0.0538 - val_acc: 0.9447\n",
      "Epoch 476/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0532 - acc: 0.9485 - val_loss: 0.0540 - val_acc: 0.9447\n",
      "Epoch 477/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0532 - acc: 0.9500 - val_loss: 0.0522 - val_acc: 0.9447\n",
      "Epoch 478/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0517 - acc: 0.9529 - val_loss: 0.0518 - val_acc: 0.9447\n",
      "Epoch 479/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0519 - acc: 0.9518 - val_loss: 0.0517 - val_acc: 0.9447\n",
      "Epoch 480/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0509 - acc: 0.9497 - val_loss: 0.0525 - val_acc: 0.9421\n",
      "Epoch 481/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0525 - acc: 0.9500 - val_loss: 0.0540 - val_acc: 0.9421\n",
      "Epoch 482/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0512 - acc: 0.9494 - val_loss: 0.0508 - val_acc: 0.9421\n",
      "Epoch 483/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0508 - acc: 0.9491 - val_loss: 0.0529 - val_acc: 0.9474\n",
      "Epoch 484/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0522 - acc: 0.9497 - val_loss: 0.0532 - val_acc: 0.9474\n",
      "Epoch 485/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0513 - acc: 0.9520 - val_loss: 0.0499 - val_acc: 0.9447\n",
      "Epoch 486/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0509 - acc: 0.9506 - val_loss: 0.0526 - val_acc: 0.9474\n",
      "Epoch 487/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0521 - acc: 0.9494 - val_loss: 0.0514 - val_acc: 0.9447\n",
      "Epoch 488/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0515 - acc: 0.9503 - val_loss: 0.0505 - val_acc: 0.9447\n",
      "Epoch 489/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0544 - acc: 0.9471 - val_loss: 0.0554 - val_acc: 0.9421\n",
      "Epoch 490/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0510 - acc: 0.9506 - val_loss: 0.0504 - val_acc: 0.9447\n",
      "Epoch 491/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0522 - acc: 0.9518 - val_loss: 0.0537 - val_acc: 0.9447\n",
      "Epoch 492/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0505 - acc: 0.9512 - val_loss: 0.0501 - val_acc: 0.9395\n",
      "Epoch 493/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0513 - acc: 0.9491 - val_loss: 0.0537 - val_acc: 0.9474\n",
      "Epoch 494/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0512 - acc: 0.9506 - val_loss: 0.0506 - val_acc: 0.9447\n",
      "Epoch 495/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0501 - acc: 0.9520 - val_loss: 0.0530 - val_acc: 0.9447\n",
      "Epoch 496/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0520 - acc: 0.9503 - val_loss: 0.0489 - val_acc: 0.9500\n",
      "Epoch 497/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0497 - acc: 0.9541 - val_loss: 0.0532 - val_acc: 0.9474\n",
      "Epoch 498/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.0518 - acc: 0.9480 - val_loss: 0.0505 - val_acc: 0.9447\n",
      "Epoch 499/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0515 - acc: 0.9512 - val_loss: 0.0503 - val_acc: 0.9526\n",
      "Epoch 500/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0506 - acc: 0.9547 - val_loss: 0.0516 - val_acc: 0.9395\n",
      "Epoch 501/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0506 - acc: 0.9491 - val_loss: 0.0517 - val_acc: 0.9474\n",
      "Epoch 502/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0513 - acc: 0.9488 - val_loss: 0.0512 - val_acc: 0.9421\n",
      "Epoch 503/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0533 - acc: 0.9468 - val_loss: 0.0502 - val_acc: 0.9447\n",
      "Epoch 504/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0513 - acc: 0.9512 - val_loss: 0.0505 - val_acc: 0.9447\n",
      "Epoch 505/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0511 - acc: 0.9512 - val_loss: 0.0499 - val_acc: 0.9500\n",
      "Epoch 506/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0506 - acc: 0.9488 - val_loss: 0.0531 - val_acc: 0.9474\n",
      "Epoch 507/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0501 - acc: 0.9526 - val_loss: 0.0489 - val_acc: 0.9474\n",
      "Epoch 508/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0492 - acc: 0.9538 - val_loss: 0.0521 - val_acc: 0.9474\n",
      "Epoch 509/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0511 - acc: 0.9509 - val_loss: 0.0520 - val_acc: 0.9447\n",
      "Epoch 510/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0512 - acc: 0.9512 - val_loss: 0.0498 - val_acc: 0.9474\n",
      "Epoch 511/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0507 - acc: 0.9497 - val_loss: 0.0514 - val_acc: 0.9474\n",
      "Epoch 512/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0521 - acc: 0.9523 - val_loss: 0.0485 - val_acc: 0.9500\n",
      "Epoch 513/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0525 - acc: 0.9480 - val_loss: 0.0513 - val_acc: 0.9474\n",
      "Epoch 514/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0489 - acc: 0.9541 - val_loss: 0.0500 - val_acc: 0.9447\n",
      "Epoch 515/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0499 - acc: 0.9485 - val_loss: 0.0489 - val_acc: 0.9500\n",
      "Epoch 516/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0486 - acc: 0.9523 - val_loss: 0.0492 - val_acc: 0.9474\n",
      "Epoch 517/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0490 - acc: 0.9503 - val_loss: 0.0495 - val_acc: 0.9421\n",
      "Epoch 518/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0480 - acc: 0.9520 - val_loss: 0.0514 - val_acc: 0.9474\n",
      "Epoch 519/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0500 - acc: 0.9509 - val_loss: 0.0483 - val_acc: 0.9474\n",
      "Epoch 520/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0496 - acc: 0.9535 - val_loss: 0.0526 - val_acc: 0.9474\n",
      "Epoch 521/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0489 - acc: 0.9538 - val_loss: 0.0495 - val_acc: 0.9447\n",
      "Epoch 522/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0509 - acc: 0.9480 - val_loss: 0.0521 - val_acc: 0.9474\n",
      "Epoch 523/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0525 - acc: 0.9485 - val_loss: 0.0472 - val_acc: 0.9500\n",
      "Epoch 524/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0487 - acc: 0.9570 - val_loss: 0.0509 - val_acc: 0.9421\n",
      "Epoch 525/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0510 - acc: 0.9529 - val_loss: 0.0494 - val_acc: 0.9500\n",
      "Epoch 526/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0494 - acc: 0.9523 - val_loss: 0.0505 - val_acc: 0.9474\n",
      "Epoch 527/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0482 - acc: 0.9529 - val_loss: 0.0469 - val_acc: 0.9526\n",
      "Epoch 528/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0490 - acc: 0.9512 - val_loss: 0.0496 - val_acc: 0.9500\n",
      "Epoch 529/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0501 - acc: 0.9500 - val_loss: 0.0474 - val_acc: 0.9500\n",
      "Epoch 530/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0493 - acc: 0.9515 - val_loss: 0.0499 - val_acc: 0.9474\n",
      "Epoch 531/1000\n",
      "14/14 [==============================] - 0s 25ms/step - loss: 0.0494 - acc: 0.9485 - val_loss: 0.0482 - val_acc: 0.9526\n",
      "Epoch 532/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0505 - acc: 0.9491 - val_loss: 0.0489 - val_acc: 0.9474\n",
      "Epoch 533/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0515 - acc: 0.9515 - val_loss: 0.0476 - val_acc: 0.9500\n",
      "Epoch 534/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0508 - acc: 0.9523 - val_loss: 0.0492 - val_acc: 0.9500\n",
      "Epoch 535/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0475 - acc: 0.9556 - val_loss: 0.0486 - val_acc: 0.9474\n",
      "Epoch 536/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0483 - acc: 0.9538 - val_loss: 0.0494 - val_acc: 0.9500\n",
      "Epoch 537/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0488 - acc: 0.9515 - val_loss: 0.0498 - val_acc: 0.9474\n",
      "Epoch 538/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0489 - acc: 0.9529 - val_loss: 0.0460 - val_acc: 0.9500\n",
      "Epoch 539/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0492 - acc: 0.9550 - val_loss: 0.0512 - val_acc: 0.9474\n",
      "Epoch 540/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0503 - acc: 0.9515 - val_loss: 0.0476 - val_acc: 0.9500\n",
      "Epoch 541/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0501 - acc: 0.9523 - val_loss: 0.0488 - val_acc: 0.9474\n",
      "Epoch 542/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0490 - acc: 0.9526 - val_loss: 0.0498 - val_acc: 0.9474\n",
      "Epoch 543/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0463 - acc: 0.9558 - val_loss: 0.0469 - val_acc: 0.9500\n",
      "Epoch 544/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0475 - acc: 0.9544 - val_loss: 0.0485 - val_acc: 0.9474\n",
      "Epoch 545/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0495 - acc: 0.9526 - val_loss: 0.0503 - val_acc: 0.9500\n",
      "Epoch 546/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0456 - acc: 0.9570 - val_loss: 0.0498 - val_acc: 0.9447\n",
      "Epoch 547/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0467 - acc: 0.9547 - val_loss: 0.0486 - val_acc: 0.9526\n",
      "Epoch 548/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0487 - acc: 0.9547 - val_loss: 0.0473 - val_acc: 0.9500\n",
      "Epoch 549/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0485 - acc: 0.9570 - val_loss: 0.0462 - val_acc: 0.9500\n",
      "Epoch 550/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0479 - acc: 0.9535 - val_loss: 0.0496 - val_acc: 0.9500\n",
      "Epoch 551/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0480 - acc: 0.9532 - val_loss: 0.0440 - val_acc: 0.9579\n",
      "Epoch 552/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0461 - acc: 0.9576 - val_loss: 0.0483 - val_acc: 0.9500\n",
      "Epoch 553/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0475 - acc: 0.9541 - val_loss: 0.0459 - val_acc: 0.9500\n",
      "Epoch 554/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0477 - acc: 0.9523 - val_loss: 0.0509 - val_acc: 0.9474\n",
      "Epoch 555/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0505 - acc: 0.9503 - val_loss: 0.0466 - val_acc: 0.9474\n",
      "Epoch 556/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0476 - acc: 0.9532 - val_loss: 0.0484 - val_acc: 0.9500\n",
      "Epoch 557/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0472 - acc: 0.9526 - val_loss: 0.0471 - val_acc: 0.9500\n",
      "Epoch 558/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0466 - acc: 0.9582 - val_loss: 0.0478 - val_acc: 0.9526\n",
      "Epoch 559/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0483 - acc: 0.9556 - val_loss: 0.0465 - val_acc: 0.9474\n",
      "Epoch 560/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0480 - acc: 0.9550 - val_loss: 0.0477 - val_acc: 0.9447\n",
      "Epoch 561/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0477 - acc: 0.9541 - val_loss: 0.0467 - val_acc: 0.9500\n",
      "Epoch 562/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0469 - acc: 0.9538 - val_loss: 0.0483 - val_acc: 0.9500\n",
      "Epoch 563/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0484 - acc: 0.9579 - val_loss: 0.0458 - val_acc: 0.9500\n",
      "Epoch 564/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0461 - acc: 0.9532 - val_loss: 0.0481 - val_acc: 0.9500\n",
      "Epoch 565/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0461 - acc: 0.9582 - val_loss: 0.0452 - val_acc: 0.9526\n",
      "Epoch 566/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0476 - acc: 0.9553 - val_loss: 0.0484 - val_acc: 0.9500\n",
      "Epoch 567/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0480 - acc: 0.9535 - val_loss: 0.0469 - val_acc: 0.9500\n",
      "Epoch 568/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0476 - acc: 0.9518 - val_loss: 0.0464 - val_acc: 0.9500\n",
      "Epoch 569/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0461 - acc: 0.9541 - val_loss: 0.0476 - val_acc: 0.9526\n",
      "Epoch 570/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0498 - acc: 0.9538 - val_loss: 0.0446 - val_acc: 0.9500\n",
      "Epoch 571/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0467 - acc: 0.9547 - val_loss: 0.0475 - val_acc: 0.9500\n",
      "Epoch 572/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0441 - acc: 0.9556 - val_loss: 0.0474 - val_acc: 0.9526\n",
      "Epoch 573/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0456 - acc: 0.9541 - val_loss: 0.0480 - val_acc: 0.9474\n",
      "Epoch 574/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0479 - acc: 0.9497 - val_loss: 0.0464 - val_acc: 0.9474\n",
      "Epoch 575/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0486 - acc: 0.9497 - val_loss: 0.0476 - val_acc: 0.9500\n",
      "Epoch 576/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0462 - acc: 0.9535 - val_loss: 0.0462 - val_acc: 0.9474\n",
      "Epoch 577/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0472 - acc: 0.9576 - val_loss: 0.0443 - val_acc: 0.9526\n",
      "Epoch 578/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0483 - acc: 0.9497 - val_loss: 0.0491 - val_acc: 0.9421\n",
      "Epoch 579/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0456 - acc: 0.9588 - val_loss: 0.0448 - val_acc: 0.9500\n",
      "Epoch 580/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0488 - acc: 0.9547 - val_loss: 0.0472 - val_acc: 0.9500\n",
      "Epoch 581/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0463 - acc: 0.9564 - val_loss: 0.0480 - val_acc: 0.9474\n",
      "Epoch 582/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0466 - acc: 0.9585 - val_loss: 0.0454 - val_acc: 0.9526\n",
      "Epoch 583/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0479 - acc: 0.9561 - val_loss: 0.0487 - val_acc: 0.9474\n",
      "Epoch 584/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0480 - acc: 0.9535 - val_loss: 0.0445 - val_acc: 0.9500\n",
      "Epoch 585/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0468 - acc: 0.9526 - val_loss: 0.0471 - val_acc: 0.9500\n",
      "Epoch 586/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0471 - acc: 0.9553 - val_loss: 0.0454 - val_acc: 0.9500\n",
      "Epoch 587/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0467 - acc: 0.9564 - val_loss: 0.0452 - val_acc: 0.9526\n",
      "Epoch 588/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0451 - acc: 0.9550 - val_loss: 0.0467 - val_acc: 0.9500\n",
      "Epoch 589/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0480 - acc: 0.9544 - val_loss: 0.0463 - val_acc: 0.9474\n",
      "Epoch 590/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0479 - acc: 0.9567 - val_loss: 0.0452 - val_acc: 0.9526\n",
      "Epoch 591/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0464 - acc: 0.9573 - val_loss: 0.0477 - val_acc: 0.9474\n",
      "Epoch 592/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0477 - acc: 0.9558 - val_loss: 0.0463 - val_acc: 0.9500\n",
      "Epoch 593/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0451 - acc: 0.9585 - val_loss: 0.0454 - val_acc: 0.9474\n",
      "Epoch 594/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0479 - acc: 0.9523 - val_loss: 0.0468 - val_acc: 0.9474\n",
      "Epoch 595/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0449 - acc: 0.9579 - val_loss: 0.0482 - val_acc: 0.9474\n",
      "Epoch 596/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0459 - acc: 0.9547 - val_loss: 0.0449 - val_acc: 0.9500\n",
      "Epoch 597/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0456 - acc: 0.9558 - val_loss: 0.0454 - val_acc: 0.9500\n",
      "Epoch 598/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0448 - acc: 0.9570 - val_loss: 0.0445 - val_acc: 0.9500\n",
      "Epoch 599/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0467 - acc: 0.9550 - val_loss: 0.0460 - val_acc: 0.9500\n",
      "Epoch 600/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0453 - acc: 0.9558 - val_loss: 0.0452 - val_acc: 0.9500\n",
      "Epoch 601/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0476 - acc: 0.9532 - val_loss: 0.0467 - val_acc: 0.9474\n",
      "Epoch 602/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0458 - acc: 0.9582 - val_loss: 0.0441 - val_acc: 0.9553\n",
      "Epoch 603/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0458 - acc: 0.9547 - val_loss: 0.0483 - val_acc: 0.9447\n",
      "Epoch 604/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0454 - acc: 0.9585 - val_loss: 0.0421 - val_acc: 0.9579\n",
      "Epoch 605/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0463 - acc: 0.9529 - val_loss: 0.0496 - val_acc: 0.9474\n",
      "Epoch 606/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0490 - acc: 0.9547 - val_loss: 0.0439 - val_acc: 0.9474\n",
      "Epoch 607/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0441 - acc: 0.9623 - val_loss: 0.0458 - val_acc: 0.9500\n",
      "Epoch 608/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0455 - acc: 0.9558 - val_loss: 0.0443 - val_acc: 0.9526\n",
      "Epoch 609/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0455 - acc: 0.9558 - val_loss: 0.0447 - val_acc: 0.9526\n",
      "Epoch 610/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0448 - acc: 0.9567 - val_loss: 0.0450 - val_acc: 0.9447\n",
      "Epoch 611/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0437 - acc: 0.9579 - val_loss: 0.0428 - val_acc: 0.9579\n",
      "Epoch 612/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0450 - acc: 0.9576 - val_loss: 0.0442 - val_acc: 0.9500\n",
      "Epoch 613/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0447 - acc: 0.9564 - val_loss: 0.0450 - val_acc: 0.9500\n",
      "Epoch 614/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0446 - acc: 0.9570 - val_loss: 0.0443 - val_acc: 0.9500\n",
      "Epoch 615/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0445 - acc: 0.9564 - val_loss: 0.0427 - val_acc: 0.9553\n",
      "Epoch 616/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0453 - acc: 0.9570 - val_loss: 0.0430 - val_acc: 0.9526\n",
      "Epoch 617/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0451 - acc: 0.9567 - val_loss: 0.0464 - val_acc: 0.9500\n",
      "Epoch 618/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0459 - acc: 0.9576 - val_loss: 0.0428 - val_acc: 0.9553\n",
      "Epoch 619/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0460 - acc: 0.9520 - val_loss: 0.0441 - val_acc: 0.9526\n",
      "Epoch 620/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9582 - val_loss: 0.0457 - val_acc: 0.9474\n",
      "Epoch 621/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0440 - acc: 0.9579 - val_loss: 0.0441 - val_acc: 0.9526\n",
      "Epoch 622/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0438 - acc: 0.9579 - val_loss: 0.0440 - val_acc: 0.9500\n",
      "Epoch 623/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0452 - acc: 0.9561 - val_loss: 0.0438 - val_acc: 0.9526\n",
      "Epoch 624/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0432 - acc: 0.9585 - val_loss: 0.0437 - val_acc: 0.9500\n",
      "Epoch 625/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0425 - acc: 0.9611 - val_loss: 0.0461 - val_acc: 0.9474\n",
      "Epoch 626/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0444 - acc: 0.9553 - val_loss: 0.0421 - val_acc: 0.9579\n",
      "Epoch 627/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0457 - acc: 0.9558 - val_loss: 0.0455 - val_acc: 0.9500\n",
      "Epoch 628/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0463 - acc: 0.9564 - val_loss: 0.0434 - val_acc: 0.9553\n",
      "Epoch 629/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9579 - val_loss: 0.0453 - val_acc: 0.9526\n",
      "Epoch 630/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0450 - acc: 0.9585 - val_loss: 0.0419 - val_acc: 0.9632\n",
      "Epoch 631/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0453 - acc: 0.9573 - val_loss: 0.0450 - val_acc: 0.9526\n",
      "Epoch 632/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0459 - acc: 0.9535 - val_loss: 0.0445 - val_acc: 0.9500\n",
      "Epoch 633/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0454 - acc: 0.9576 - val_loss: 0.0441 - val_acc: 0.9474\n",
      "Epoch 634/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0426 - acc: 0.9599 - val_loss: 0.0431 - val_acc: 0.9526\n",
      "Epoch 635/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0423 - acc: 0.9605 - val_loss: 0.0431 - val_acc: 0.9526\n",
      "Epoch 636/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0445 - acc: 0.9564 - val_loss: 0.0435 - val_acc: 0.9526\n",
      "Epoch 637/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0436 - acc: 0.9576 - val_loss: 0.0456 - val_acc: 0.9500\n",
      "Epoch 638/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0451 - acc: 0.9602 - val_loss: 0.0430 - val_acc: 0.9500\n",
      "Epoch 639/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0432 - acc: 0.9608 - val_loss: 0.0453 - val_acc: 0.9474\n",
      "Epoch 640/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0438 - acc: 0.9602 - val_loss: 0.0420 - val_acc: 0.9553\n",
      "Epoch 641/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0443 - acc: 0.9602 - val_loss: 0.0472 - val_acc: 0.9474\n",
      "Epoch 642/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0444 - acc: 0.9550 - val_loss: 0.0413 - val_acc: 0.9605\n",
      "Epoch 643/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0454 - acc: 0.9544 - val_loss: 0.0439 - val_acc: 0.9500\n",
      "Epoch 644/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0460 - acc: 0.9573 - val_loss: 0.0432 - val_acc: 0.9579\n",
      "Epoch 645/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0456 - acc: 0.9556 - val_loss: 0.0501 - val_acc: 0.9421\n",
      "Epoch 646/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0456 - acc: 0.9558 - val_loss: 0.0405 - val_acc: 0.9605\n",
      "Epoch 647/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0463 - acc: 0.9576 - val_loss: 0.0460 - val_acc: 0.9500\n",
      "Epoch 648/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0427 - acc: 0.9602 - val_loss: 0.0407 - val_acc: 0.9605\n",
      "Epoch 649/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0451 - acc: 0.9561 - val_loss: 0.0462 - val_acc: 0.9447\n",
      "Epoch 650/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0435 - acc: 0.9564 - val_loss: 0.0410 - val_acc: 0.9579\n",
      "Epoch 651/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0461 - acc: 0.9500 - val_loss: 0.0419 - val_acc: 0.9553\n",
      "Epoch 652/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0463 - acc: 0.9582 - val_loss: 0.0416 - val_acc: 0.9579\n",
      "Epoch 653/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0426 - acc: 0.9596 - val_loss: 0.0435 - val_acc: 0.9500\n",
      "Epoch 654/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0431 - acc: 0.9588 - val_loss: 0.0425 - val_acc: 0.9553\n",
      "Epoch 655/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0434 - acc: 0.9594 - val_loss: 0.0416 - val_acc: 0.9526\n",
      "Epoch 656/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0435 - acc: 0.9617 - val_loss: 0.0421 - val_acc: 0.9579\n",
      "Epoch 657/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0446 - acc: 0.9518 - val_loss: 0.0430 - val_acc: 0.9500\n",
      "Epoch 658/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0429 - acc: 0.9588 - val_loss: 0.0419 - val_acc: 0.9579\n",
      "Epoch 659/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9588 - val_loss: 0.0424 - val_acc: 0.9553\n",
      "Epoch 660/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0412 - acc: 0.9614 - val_loss: 0.0423 - val_acc: 0.9500\n",
      "Epoch 661/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0417 - acc: 0.9596 - val_loss: 0.0436 - val_acc: 0.9526\n",
      "Epoch 662/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0450 - acc: 0.9567 - val_loss: 0.0424 - val_acc: 0.9526\n",
      "Epoch 663/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0429 - acc: 0.9567 - val_loss: 0.0406 - val_acc: 0.9526\n",
      "Epoch 664/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0443 - acc: 0.9570 - val_loss: 0.0421 - val_acc: 0.9553\n",
      "Epoch 665/1000\n",
      "14/14 [==============================] - 0s 23ms/step - loss: 0.0429 - acc: 0.9599 - val_loss: 0.0405 - val_acc: 0.9579\n",
      "Epoch 666/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0436 - acc: 0.9596 - val_loss: 0.0424 - val_acc: 0.9579\n",
      "Epoch 667/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0429 - acc: 0.9608 - val_loss: 0.0412 - val_acc: 0.9579\n",
      "Epoch 668/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0436 - acc: 0.9564 - val_loss: 0.0454 - val_acc: 0.9500\n",
      "Epoch 669/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0424 - acc: 0.9608 - val_loss: 0.0399 - val_acc: 0.9605\n",
      "Epoch 670/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0430 - acc: 0.9596 - val_loss: 0.0439 - val_acc: 0.9526\n",
      "Epoch 671/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0429 - acc: 0.9594 - val_loss: 0.0405 - val_acc: 0.9526\n",
      "Epoch 672/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9550 - val_loss: 0.0425 - val_acc: 0.9579\n",
      "Epoch 673/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0426 - acc: 0.9614 - val_loss: 0.0421 - val_acc: 0.9553\n",
      "Epoch 674/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0424 - acc: 0.9620 - val_loss: 0.0409 - val_acc: 0.9579\n",
      "Epoch 675/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0405 - acc: 0.9643 - val_loss: 0.0425 - val_acc: 0.9553\n",
      "Epoch 676/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9599 - val_loss: 0.0391 - val_acc: 0.9632\n",
      "Epoch 677/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0437 - acc: 0.9591 - val_loss: 0.0435 - val_acc: 0.9553\n",
      "Epoch 678/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0424 - acc: 0.9582 - val_loss: 0.0404 - val_acc: 0.9579\n",
      "Epoch 679/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0415 - acc: 0.9605 - val_loss: 0.0429 - val_acc: 0.9526\n",
      "Epoch 680/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0426 - acc: 0.9614 - val_loss: 0.0430 - val_acc: 0.9526\n",
      "Epoch 681/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0443 - acc: 0.9561 - val_loss: 0.0404 - val_acc: 0.9553\n",
      "Epoch 682/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0431 - acc: 0.9579 - val_loss: 0.0419 - val_acc: 0.9526\n",
      "Epoch 683/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0437 - acc: 0.9564 - val_loss: 0.0425 - val_acc: 0.9526\n",
      "Epoch 684/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0423 - acc: 0.9596 - val_loss: 0.0412 - val_acc: 0.9579\n",
      "Epoch 685/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0425 - acc: 0.9564 - val_loss: 0.0423 - val_acc: 0.9553\n",
      "Epoch 686/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0420 - acc: 0.9605 - val_loss: 0.0419 - val_acc: 0.9526\n",
      "Epoch 687/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0412 - acc: 0.9605 - val_loss: 0.0430 - val_acc: 0.9500\n",
      "Epoch 688/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0428 - acc: 0.9605 - val_loss: 0.0405 - val_acc: 0.9553\n",
      "Epoch 689/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0428 - acc: 0.9605 - val_loss: 0.0426 - val_acc: 0.9553\n",
      "Epoch 690/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0414 - acc: 0.9620 - val_loss: 0.0418 - val_acc: 0.9553\n",
      "Epoch 691/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0445 - acc: 0.9582 - val_loss: 0.0397 - val_acc: 0.9553\n",
      "Epoch 692/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0414 - acc: 0.9626 - val_loss: 0.0403 - val_acc: 0.9579\n",
      "Epoch 693/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0430 - acc: 0.9573 - val_loss: 0.0387 - val_acc: 0.9632\n",
      "Epoch 694/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0411 - acc: 0.9596 - val_loss: 0.0398 - val_acc: 0.9579\n",
      "Epoch 695/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0411 - acc: 0.9617 - val_loss: 0.0430 - val_acc: 0.9526\n",
      "Epoch 696/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0434 - acc: 0.9576 - val_loss: 0.0392 - val_acc: 0.9605\n",
      "Epoch 697/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0417 - acc: 0.9591 - val_loss: 0.0399 - val_acc: 0.9579\n",
      "Epoch 698/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0397 - acc: 0.9632 - val_loss: 0.0405 - val_acc: 0.9553\n",
      "Epoch 699/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0406 - acc: 0.9599 - val_loss: 0.0405 - val_acc: 0.9553\n",
      "Epoch 700/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0450 - acc: 0.9588 - val_loss: 0.0416 - val_acc: 0.9553\n",
      "Epoch 701/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0411 - acc: 0.9594 - val_loss: 0.0403 - val_acc: 0.9553\n",
      "Epoch 702/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0430 - acc: 0.9605 - val_loss: 0.0395 - val_acc: 0.9579\n",
      "Epoch 703/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0404 - acc: 0.9626 - val_loss: 0.0413 - val_acc: 0.9579\n",
      "Epoch 704/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0425 - acc: 0.9591 - val_loss: 0.0406 - val_acc: 0.9553\n",
      "Epoch 705/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0444 - acc: 0.9588 - val_loss: 0.0408 - val_acc: 0.9526\n",
      "Epoch 706/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0401 - acc: 0.9617 - val_loss: 0.0388 - val_acc: 0.9579\n",
      "Epoch 707/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0413 - acc: 0.9632 - val_loss: 0.0390 - val_acc: 0.9553\n",
      "Epoch 708/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0422 - acc: 0.9596 - val_loss: 0.0433 - val_acc: 0.9553\n",
      "Epoch 709/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0404 - acc: 0.9629 - val_loss: 0.0388 - val_acc: 0.9605\n",
      "Epoch 710/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0411 - acc: 0.9602 - val_loss: 0.0412 - val_acc: 0.9553\n",
      "Epoch 711/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0428 - acc: 0.9588 - val_loss: 0.0412 - val_acc: 0.9579\n",
      "Epoch 712/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0410 - acc: 0.9620 - val_loss: 0.0412 - val_acc: 0.9579\n",
      "Epoch 713/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0421 - acc: 0.9596 - val_loss: 0.0394 - val_acc: 0.9553\n",
      "Epoch 714/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0389 - acc: 0.9635 - val_loss: 0.0414 - val_acc: 0.9526\n",
      "Epoch 715/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0421 - acc: 0.9585 - val_loss: 0.0379 - val_acc: 0.9658\n",
      "Epoch 716/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0411 - acc: 0.9626 - val_loss: 0.0431 - val_acc: 0.9526\n",
      "Epoch 717/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0399 - acc: 0.9608 - val_loss: 0.0388 - val_acc: 0.9579\n",
      "Epoch 718/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0405 - acc: 0.9626 - val_loss: 0.0427 - val_acc: 0.9526\n",
      "Epoch 719/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0398 - acc: 0.9635 - val_loss: 0.0395 - val_acc: 0.9605\n",
      "Epoch 720/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0413 - acc: 0.9611 - val_loss: 0.0393 - val_acc: 0.9605\n",
      "Epoch 721/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0421 - acc: 0.9558 - val_loss: 0.0411 - val_acc: 0.9553\n",
      "Epoch 722/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0402 - acc: 0.9643 - val_loss: 0.0391 - val_acc: 0.9579\n",
      "Epoch 723/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0417 - acc: 0.9617 - val_loss: 0.0411 - val_acc: 0.9553\n",
      "Epoch 724/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0402 - acc: 0.9629 - val_loss: 0.0402 - val_acc: 0.9579\n",
      "Epoch 725/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0425 - acc: 0.9620 - val_loss: 0.0380 - val_acc: 0.9605\n",
      "Epoch 726/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0434 - acc: 0.9582 - val_loss: 0.0417 - val_acc: 0.9526\n",
      "Epoch 727/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0430 - acc: 0.9576 - val_loss: 0.0391 - val_acc: 0.9632\n",
      "Epoch 728/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0425 - acc: 0.9582 - val_loss: 0.0397 - val_acc: 0.9553\n",
      "Epoch 729/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0412 - acc: 0.9617 - val_loss: 0.0421 - val_acc: 0.9500\n",
      "Epoch 730/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0403 - acc: 0.9635 - val_loss: 0.0393 - val_acc: 0.9553\n",
      "Epoch 731/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0415 - acc: 0.9617 - val_loss: 0.0395 - val_acc: 0.9605\n",
      "Epoch 732/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0392 - acc: 0.9623 - val_loss: 0.0401 - val_acc: 0.9553\n",
      "Epoch 733/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0405 - acc: 0.9605 - val_loss: 0.0389 - val_acc: 0.9632\n",
      "Epoch 734/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0406 - acc: 0.9617 - val_loss: 0.0390 - val_acc: 0.9579\n",
      "Epoch 735/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0415 - acc: 0.9620 - val_loss: 0.0392 - val_acc: 0.9553\n",
      "Epoch 736/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0405 - acc: 0.9626 - val_loss: 0.0396 - val_acc: 0.9553\n",
      "Epoch 737/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0407 - acc: 0.9594 - val_loss: 0.0380 - val_acc: 0.9605\n",
      "Epoch 738/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0399 - acc: 0.9599 - val_loss: 0.0393 - val_acc: 0.9553\n",
      "Epoch 739/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0409 - acc: 0.9655 - val_loss: 0.0385 - val_acc: 0.9605\n",
      "Epoch 740/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0390 - acc: 0.9629 - val_loss: 0.0407 - val_acc: 0.9553\n",
      "Epoch 741/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0391 - acc: 0.9611 - val_loss: 0.0367 - val_acc: 0.9658\n",
      "Epoch 742/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0407 - acc: 0.9617 - val_loss: 0.0407 - val_acc: 0.9579\n",
      "Epoch 743/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0427 - acc: 0.9596 - val_loss: 0.0401 - val_acc: 0.9553\n",
      "Epoch 744/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0414 - acc: 0.9596 - val_loss: 0.0384 - val_acc: 0.9553\n",
      "Epoch 745/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0412 - acc: 0.9591 - val_loss: 0.0395 - val_acc: 0.9553\n",
      "Epoch 746/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0409 - acc: 0.9629 - val_loss: 0.0384 - val_acc: 0.9605\n",
      "Epoch 747/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0408 - acc: 0.9614 - val_loss: 0.0385 - val_acc: 0.9579\n",
      "Epoch 748/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0392 - acc: 0.9626 - val_loss: 0.0396 - val_acc: 0.9579\n",
      "Epoch 749/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0400 - acc: 0.9632 - val_loss: 0.0391 - val_acc: 0.9579\n",
      "Epoch 750/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0406 - acc: 0.9579 - val_loss: 0.0393 - val_acc: 0.9579\n",
      "Epoch 751/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0416 - acc: 0.9599 - val_loss: 0.0403 - val_acc: 0.9579\n",
      "Epoch 752/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0404 - acc: 0.9637 - val_loss: 0.0377 - val_acc: 0.9605\n",
      "Epoch 753/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0401 - acc: 0.9617 - val_loss: 0.0387 - val_acc: 0.9605\n",
      "Epoch 754/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0421 - acc: 0.9605 - val_loss: 0.0384 - val_acc: 0.9579\n",
      "Epoch 755/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0412 - acc: 0.9614 - val_loss: 0.0386 - val_acc: 0.9579\n",
      "Epoch 756/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0387 - acc: 0.9649 - val_loss: 0.0380 - val_acc: 0.9605\n",
      "Epoch 757/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0392 - acc: 0.9629 - val_loss: 0.0378 - val_acc: 0.9632\n",
      "Epoch 758/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0379 - acc: 0.9664 - val_loss: 0.0388 - val_acc: 0.9553\n",
      "Epoch 759/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0411 - acc: 0.9585 - val_loss: 0.0416 - val_acc: 0.9500\n",
      "Epoch 760/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0411 - acc: 0.9614 - val_loss: 0.0371 - val_acc: 0.9632\n",
      "Epoch 761/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0411 - acc: 0.9640 - val_loss: 0.0408 - val_acc: 0.9553\n",
      "Epoch 762/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0401 - acc: 0.9611 - val_loss: 0.0382 - val_acc: 0.9579\n",
      "Epoch 763/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0384 - acc: 0.9629 - val_loss: 0.0390 - val_acc: 0.9553\n",
      "Epoch 764/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0388 - acc: 0.9608 - val_loss: 0.0385 - val_acc: 0.9605\n",
      "Epoch 765/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0404 - acc: 0.9599 - val_loss: 0.0369 - val_acc: 0.9658\n",
      "Epoch 766/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0372 - acc: 0.9652 - val_loss: 0.0411 - val_acc: 0.9553\n",
      "Epoch 767/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0426 - acc: 0.9579 - val_loss: 0.0380 - val_acc: 0.9553\n",
      "Epoch 768/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0390 - acc: 0.9640 - val_loss: 0.0383 - val_acc: 0.9553\n",
      "Epoch 769/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0413 - acc: 0.9591 - val_loss: 0.0387 - val_acc: 0.9605\n",
      "Epoch 770/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0392 - acc: 0.9623 - val_loss: 0.0384 - val_acc: 0.9553\n",
      "Epoch 771/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0382 - acc: 0.9602 - val_loss: 0.0372 - val_acc: 0.9632\n",
      "Epoch 772/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0407 - acc: 0.9605 - val_loss: 0.0385 - val_acc: 0.9553\n",
      "Epoch 773/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0415 - acc: 0.9605 - val_loss: 0.0392 - val_acc: 0.9526\n",
      "Epoch 774/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0402 - acc: 0.9623 - val_loss: 0.0370 - val_acc: 0.9658\n",
      "Epoch 775/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0388 - acc: 0.9632 - val_loss: 0.0379 - val_acc: 0.9632\n",
      "Epoch 776/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0376 - acc: 0.9664 - val_loss: 0.0380 - val_acc: 0.9579\n",
      "Epoch 777/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0381 - acc: 0.9664 - val_loss: 0.0379 - val_acc: 0.9658\n",
      "Epoch 778/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0399 - acc: 0.9620 - val_loss: 0.0392 - val_acc: 0.9526\n",
      "Epoch 779/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0392 - acc: 0.9635 - val_loss: 0.0364 - val_acc: 0.9684\n",
      "Epoch 780/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0393 - acc: 0.9608 - val_loss: 0.0370 - val_acc: 0.9684\n",
      "Epoch 781/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0396 - acc: 0.9637 - val_loss: 0.0390 - val_acc: 0.9579\n",
      "Epoch 782/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0390 - acc: 0.9620 - val_loss: 0.0365 - val_acc: 0.9684\n",
      "Epoch 783/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0396 - acc: 0.9640 - val_loss: 0.0393 - val_acc: 0.9526\n",
      "Epoch 784/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0388 - acc: 0.9658 - val_loss: 0.0371 - val_acc: 0.9658\n",
      "Epoch 785/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0385 - acc: 0.9617 - val_loss: 0.0375 - val_acc: 0.9605\n",
      "Epoch 786/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0385 - acc: 0.9646 - val_loss: 0.0383 - val_acc: 0.9605\n",
      "Epoch 787/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0412 - acc: 0.9608 - val_loss: 0.0375 - val_acc: 0.9605\n",
      "Epoch 788/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0406 - acc: 0.9608 - val_loss: 0.0372 - val_acc: 0.9658\n",
      "Epoch 789/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0404 - acc: 0.9594 - val_loss: 0.0402 - val_acc: 0.9526\n",
      "Epoch 790/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0392 - acc: 0.9620 - val_loss: 0.0375 - val_acc: 0.9632\n",
      "Epoch 791/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0374 - acc: 0.9646 - val_loss: 0.0373 - val_acc: 0.9632\n",
      "Epoch 792/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0403 - acc: 0.9640 - val_loss: 0.0367 - val_acc: 0.9658\n",
      "Epoch 793/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0376 - acc: 0.9667 - val_loss: 0.0372 - val_acc: 0.9605\n",
      "Epoch 794/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0401 - acc: 0.9599 - val_loss: 0.0383 - val_acc: 0.9632\n",
      "Epoch 795/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0404 - acc: 0.9629 - val_loss: 0.0376 - val_acc: 0.9526\n",
      "Epoch 796/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0386 - acc: 0.9632 - val_loss: 0.0364 - val_acc: 0.9632\n",
      "Epoch 797/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0375 - acc: 0.9652 - val_loss: 0.0407 - val_acc: 0.9553\n",
      "Epoch 798/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0377 - acc: 0.9635 - val_loss: 0.0384 - val_acc: 0.9632\n",
      "Epoch 799/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0402 - acc: 0.9602 - val_loss: 0.0363 - val_acc: 0.9684\n",
      "Epoch 800/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0383 - acc: 0.9690 - val_loss: 0.0408 - val_acc: 0.9500\n",
      "Epoch 801/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0407 - acc: 0.9623 - val_loss: 0.0380 - val_acc: 0.9553\n",
      "Epoch 802/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0404 - acc: 0.9617 - val_loss: 0.0376 - val_acc: 0.9553\n",
      "Epoch 803/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0399 - acc: 0.9596 - val_loss: 0.0371 - val_acc: 0.9658\n",
      "Epoch 804/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0402 - acc: 0.9599 - val_loss: 0.0402 - val_acc: 0.9553\n",
      "Epoch 805/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0372 - acc: 0.9649 - val_loss: 0.0372 - val_acc: 0.9658\n",
      "Epoch 806/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0375 - acc: 0.9643 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 807/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0364 - acc: 0.9667 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 808/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0397 - acc: 0.9632 - val_loss: 0.0416 - val_acc: 0.9500\n",
      "Epoch 809/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0380 - acc: 0.9664 - val_loss: 0.0362 - val_acc: 0.9711\n",
      "Epoch 810/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0387 - acc: 0.9614 - val_loss: 0.0394 - val_acc: 0.9526\n",
      "Epoch 811/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0385 - acc: 0.9652 - val_loss: 0.0381 - val_acc: 0.9579\n",
      "Epoch 812/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0390 - acc: 0.9643 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 813/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0367 - acc: 0.9652 - val_loss: 0.0384 - val_acc: 0.9500\n",
      "Epoch 814/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0378 - acc: 0.9629 - val_loss: 0.0353 - val_acc: 0.9658\n",
      "Epoch 815/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0367 - acc: 0.9675 - val_loss: 0.0387 - val_acc: 0.9632\n",
      "Epoch 816/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0383 - acc: 0.9629 - val_loss: 0.0371 - val_acc: 0.9579\n",
      "Epoch 817/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0404 - acc: 0.9623 - val_loss: 0.0383 - val_acc: 0.9579\n",
      "Epoch 818/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0385 - acc: 0.9667 - val_loss: 0.0369 - val_acc: 0.9605\n",
      "Epoch 819/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0359 - acc: 0.9670 - val_loss: 0.0416 - val_acc: 0.9526\n",
      "Epoch 820/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0349 - acc: 0.9673 - val_loss: 0.0379 - val_acc: 0.9605\n",
      "Epoch 821/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0369 - acc: 0.9664 - val_loss: 0.0379 - val_acc: 0.9553\n",
      "Epoch 822/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0377 - acc: 0.9640 - val_loss: 0.0375 - val_acc: 0.9526\n",
      "Epoch 823/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0397 - acc: 0.9640 - val_loss: 0.0373 - val_acc: 0.9605\n",
      "Epoch 824/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0369 - acc: 0.9643 - val_loss: 0.0386 - val_acc: 0.9553\n",
      "Epoch 825/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0395 - acc: 0.9626 - val_loss: 0.0369 - val_acc: 0.9658\n",
      "Epoch 826/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0378 - acc: 0.9652 - val_loss: 0.0374 - val_acc: 0.9632\n",
      "Epoch 827/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0391 - acc: 0.9643 - val_loss: 0.0382 - val_acc: 0.9553\n",
      "Epoch 828/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0387 - acc: 0.9640 - val_loss: 0.0385 - val_acc: 0.9605\n",
      "Epoch 829/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0367 - acc: 0.9658 - val_loss: 0.0374 - val_acc: 0.9605\n",
      "Epoch 830/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0409 - acc: 0.9623 - val_loss: 0.0382 - val_acc: 0.9579\n",
      "Epoch 831/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0397 - acc: 0.9608 - val_loss: 0.0386 - val_acc: 0.9605\n",
      "Epoch 832/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0365 - acc: 0.9658 - val_loss: 0.0354 - val_acc: 0.9658\n",
      "Epoch 833/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0388 - acc: 0.9626 - val_loss: 0.0381 - val_acc: 0.9605\n",
      "Epoch 834/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0379 - acc: 0.9655 - val_loss: 0.0366 - val_acc: 0.9632\n",
      "Epoch 835/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0364 - acc: 0.9667 - val_loss: 0.0426 - val_acc: 0.9500\n",
      "Epoch 836/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0367 - acc: 0.9658 - val_loss: 0.0368 - val_acc: 0.9632\n",
      "Epoch 837/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0357 - acc: 0.9643 - val_loss: 0.0385 - val_acc: 0.9632\n",
      "Epoch 838/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0380 - acc: 0.9661 - val_loss: 0.0362 - val_acc: 0.9658\n",
      "Epoch 839/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0380 - acc: 0.9643 - val_loss: 0.0377 - val_acc: 0.9579\n",
      "Epoch 840/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0357 - acc: 0.9667 - val_loss: 0.0364 - val_acc: 0.9658\n",
      "Epoch 841/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0379 - acc: 0.9667 - val_loss: 0.0363 - val_acc: 0.9632\n",
      "Epoch 842/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0360 - acc: 0.9626 - val_loss: 0.0386 - val_acc: 0.9605\n",
      "Epoch 843/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0367 - acc: 0.9649 - val_loss: 0.0358 - val_acc: 0.9684\n",
      "Epoch 844/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0370 - acc: 0.9646 - val_loss: 0.0375 - val_acc: 0.9605\n",
      "Epoch 845/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0366 - acc: 0.9626 - val_loss: 0.0374 - val_acc: 0.9579\n",
      "Epoch 846/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0390 - acc: 0.9643 - val_loss: 0.0379 - val_acc: 0.9632\n",
      "Epoch 847/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0394 - acc: 0.9611 - val_loss: 0.0392 - val_acc: 0.9553\n",
      "Epoch 848/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0386 - acc: 0.9687 - val_loss: 0.0358 - val_acc: 0.9632\n",
      "Epoch 849/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0369 - acc: 0.9637 - val_loss: 0.0377 - val_acc: 0.9632\n",
      "Epoch 850/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0352 - acc: 0.9649 - val_loss: 0.0359 - val_acc: 0.9632\n",
      "Epoch 851/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0361 - acc: 0.9629 - val_loss: 0.0381 - val_acc: 0.9553\n",
      "Epoch 852/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0375 - acc: 0.9664 - val_loss: 0.0372 - val_acc: 0.9632\n",
      "Epoch 853/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0383 - acc: 0.9678 - val_loss: 0.0361 - val_acc: 0.9632\n",
      "Epoch 854/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0376 - acc: 0.9667 - val_loss: 0.0375 - val_acc: 0.9605\n",
      "Epoch 855/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0346 - acc: 0.9664 - val_loss: 0.0365 - val_acc: 0.9632\n",
      "Epoch 856/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0365 - acc: 0.9667 - val_loss: 0.0383 - val_acc: 0.9553\n",
      "Epoch 857/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0360 - acc: 0.9713 - val_loss: 0.0376 - val_acc: 0.9579\n",
      "Epoch 858/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0368 - acc: 0.9646 - val_loss: 0.0354 - val_acc: 0.9684\n",
      "Epoch 859/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0372 - acc: 0.9661 - val_loss: 0.0392 - val_acc: 0.9500\n",
      "Epoch 860/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0387 - acc: 0.9643 - val_loss: 0.0351 - val_acc: 0.9684\n",
      "Epoch 861/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0397 - acc: 0.9626 - val_loss: 0.0401 - val_acc: 0.9526\n",
      "Epoch 862/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0363 - acc: 0.9623 - val_loss: 0.0365 - val_acc: 0.9658\n",
      "Epoch 863/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0376 - acc: 0.9637 - val_loss: 0.0397 - val_acc: 0.9605\n",
      "Epoch 864/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0366 - acc: 0.9652 - val_loss: 0.0363 - val_acc: 0.9684\n",
      "Epoch 865/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0373 - acc: 0.9664 - val_loss: 0.0365 - val_acc: 0.9632\n",
      "Epoch 866/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0379 - acc: 0.9655 - val_loss: 0.0359 - val_acc: 0.9684\n",
      "Epoch 867/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0367 - acc: 0.9635 - val_loss: 0.0386 - val_acc: 0.9500\n",
      "Epoch 868/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0358 - acc: 0.9670 - val_loss: 0.0368 - val_acc: 0.9632\n",
      "Epoch 869/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0371 - acc: 0.9649 - val_loss: 0.0359 - val_acc: 0.9632\n",
      "Epoch 870/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0363 - acc: 0.9675 - val_loss: 0.0356 - val_acc: 0.9658\n",
      "Epoch 871/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0361 - acc: 0.9655 - val_loss: 0.0369 - val_acc: 0.9632\n",
      "Epoch 872/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0346 - acc: 0.9667 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 873/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0362 - acc: 0.9687 - val_loss: 0.0389 - val_acc: 0.9553\n",
      "Epoch 874/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0355 - acc: 0.9678 - val_loss: 0.0351 - val_acc: 0.9658\n",
      "Epoch 875/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0357 - acc: 0.9652 - val_loss: 0.0370 - val_acc: 0.9632\n",
      "Epoch 876/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0364 - acc: 0.9658 - val_loss: 0.0343 - val_acc: 0.9711\n",
      "Epoch 877/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0361 - acc: 0.9646 - val_loss: 0.0385 - val_acc: 0.9605\n",
      "Epoch 878/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0356 - acc: 0.9649 - val_loss: 0.0365 - val_acc: 0.9632\n",
      "Epoch 879/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0366 - acc: 0.9652 - val_loss: 0.0364 - val_acc: 0.9658\n",
      "Epoch 880/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0348 - acc: 0.9687 - val_loss: 0.0360 - val_acc: 0.9632\n",
      "Epoch 881/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0347 - acc: 0.9667 - val_loss: 0.0361 - val_acc: 0.9658\n",
      "Epoch 882/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0370 - acc: 0.9658 - val_loss: 0.0389 - val_acc: 0.9605\n",
      "Epoch 883/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0355 - acc: 0.9667 - val_loss: 0.0355 - val_acc: 0.9684\n",
      "Epoch 884/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0355 - acc: 0.9655 - val_loss: 0.0354 - val_acc: 0.9632\n",
      "Epoch 885/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0352 - acc: 0.9690 - val_loss: 0.0355 - val_acc: 0.9684\n",
      "Epoch 886/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0373 - acc: 0.9658 - val_loss: 0.0362 - val_acc: 0.9632\n",
      "Epoch 887/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0353 - acc: 0.9670 - val_loss: 0.0364 - val_acc: 0.9632\n",
      "Epoch 888/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0366 - acc: 0.9646 - val_loss: 0.0346 - val_acc: 0.9658\n",
      "Epoch 889/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0338 - acc: 0.9678 - val_loss: 0.0371 - val_acc: 0.9605\n",
      "Epoch 890/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0371 - acc: 0.9658 - val_loss: 0.0348 - val_acc: 0.9632\n",
      "Epoch 891/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0364 - acc: 0.9643 - val_loss: 0.0377 - val_acc: 0.9632\n",
      "Epoch 892/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0331 - acc: 0.9705 - val_loss: 0.0352 - val_acc: 0.9684\n",
      "Epoch 893/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0343 - acc: 0.9675 - val_loss: 0.0353 - val_acc: 0.9658\n",
      "Epoch 894/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0361 - acc: 0.9640 - val_loss: 0.0349 - val_acc: 0.9658\n",
      "Epoch 895/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0367 - acc: 0.9629 - val_loss: 0.0373 - val_acc: 0.9605\n",
      "Epoch 896/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0345 - acc: 0.9687 - val_loss: 0.0347 - val_acc: 0.9684\n",
      "Epoch 897/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0353 - acc: 0.9658 - val_loss: 0.0378 - val_acc: 0.9553\n",
      "Epoch 898/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0330 - acc: 0.9667 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 899/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0359 - acc: 0.9661 - val_loss: 0.0368 - val_acc: 0.9605\n",
      "Epoch 900/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0362 - acc: 0.9690 - val_loss: 0.0361 - val_acc: 0.9579\n",
      "Epoch 901/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0360 - acc: 0.9640 - val_loss: 0.0354 - val_acc: 0.9711\n",
      "Epoch 902/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0350 - acc: 0.9702 - val_loss: 0.0355 - val_acc: 0.9605\n",
      "Epoch 903/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0342 - acc: 0.9699 - val_loss: 0.0385 - val_acc: 0.9605\n",
      "Epoch 904/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0348 - acc: 0.9713 - val_loss: 0.0333 - val_acc: 0.9711\n",
      "Epoch 905/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0357 - acc: 0.9637 - val_loss: 0.0385 - val_acc: 0.9526\n",
      "Epoch 906/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0370 - acc: 0.9646 - val_loss: 0.0361 - val_acc: 0.9658\n",
      "Epoch 907/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0352 - acc: 0.9687 - val_loss: 0.0362 - val_acc: 0.9632\n",
      "Epoch 908/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0353 - acc: 0.9675 - val_loss: 0.0382 - val_acc: 0.9605\n",
      "Epoch 909/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0329 - acc: 0.9708 - val_loss: 0.0342 - val_acc: 0.9684\n",
      "Epoch 910/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0350 - acc: 0.9658 - val_loss: 0.0358 - val_acc: 0.9632\n",
      "Epoch 911/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0345 - acc: 0.9693 - val_loss: 0.0346 - val_acc: 0.9684\n",
      "Epoch 912/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0372 - acc: 0.9629 - val_loss: 0.0353 - val_acc: 0.9632\n",
      "Epoch 913/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0365 - acc: 0.9673 - val_loss: 0.0360 - val_acc: 0.9658\n",
      "Epoch 914/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0397 - acc: 0.9588 - val_loss: 0.0343 - val_acc: 0.9711\n",
      "Epoch 915/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0353 - acc: 0.9675 - val_loss: 0.0348 - val_acc: 0.9658\n",
      "Epoch 916/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0347 - acc: 0.9705 - val_loss: 0.0347 - val_acc: 0.9711\n",
      "Epoch 917/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0365 - acc: 0.9675 - val_loss: 0.0350 - val_acc: 0.9632\n",
      "Epoch 918/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0347 - acc: 0.9664 - val_loss: 0.0353 - val_acc: 0.9684\n",
      "Epoch 919/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0369 - acc: 0.9640 - val_loss: 0.0342 - val_acc: 0.9684\n",
      "Epoch 920/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0345 - acc: 0.9699 - val_loss: 0.0368 - val_acc: 0.9632\n",
      "Epoch 921/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0358 - acc: 0.9646 - val_loss: 0.0350 - val_acc: 0.9632\n",
      "Epoch 922/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0328 - acc: 0.9725 - val_loss: 0.0347 - val_acc: 0.9632\n",
      "Epoch 923/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0367 - acc: 0.9678 - val_loss: 0.0362 - val_acc: 0.9684\n",
      "Epoch 924/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0358 - acc: 0.9658 - val_loss: 0.0343 - val_acc: 0.9632\n",
      "Epoch 925/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0345 - acc: 0.9664 - val_loss: 0.0375 - val_acc: 0.9658\n",
      "Epoch 926/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0360 - acc: 0.9670 - val_loss: 0.0339 - val_acc: 0.9684\n",
      "Epoch 927/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0336 - acc: 0.9675 - val_loss: 0.0383 - val_acc: 0.9605\n",
      "Epoch 928/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0345 - acc: 0.9670 - val_loss: 0.0336 - val_acc: 0.9684\n",
      "Epoch 929/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0350 - acc: 0.9684 - val_loss: 0.0358 - val_acc: 0.9605\n",
      "Epoch 930/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0350 - acc: 0.9652 - val_loss: 0.0343 - val_acc: 0.9684\n",
      "Epoch 931/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0359 - acc: 0.9646 - val_loss: 0.0373 - val_acc: 0.9658\n",
      "Epoch 932/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0351 - acc: 0.9678 - val_loss: 0.0341 - val_acc: 0.9684\n",
      "Epoch 933/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0336 - acc: 0.9696 - val_loss: 0.0335 - val_acc: 0.9711\n",
      "Epoch 934/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0358 - acc: 0.9675 - val_loss: 0.0368 - val_acc: 0.9605\n",
      "Epoch 935/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0359 - acc: 0.9652 - val_loss: 0.0347 - val_acc: 0.9632\n",
      "Epoch 936/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0351 - acc: 0.9655 - val_loss: 0.0360 - val_acc: 0.9711\n",
      "Epoch 937/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0356 - acc: 0.9670 - val_loss: 0.0350 - val_acc: 0.9658\n",
      "Epoch 938/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0349 - acc: 0.9670 - val_loss: 0.0363 - val_acc: 0.9658\n",
      "Epoch 939/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0347 - acc: 0.9693 - val_loss: 0.0343 - val_acc: 0.9632\n",
      "Epoch 940/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0346 - acc: 0.9687 - val_loss: 0.0349 - val_acc: 0.9632\n",
      "Epoch 941/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0342 - acc: 0.9690 - val_loss: 0.0346 - val_acc: 0.9658\n",
      "Epoch 942/1000\n",
      "14/14 [==============================] - 0s 21ms/step - loss: 0.0346 - acc: 0.9693 - val_loss: 0.0358 - val_acc: 0.9658\n",
      "Epoch 943/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.0348 - acc: 0.9681 - val_loss: 0.0352 - val_acc: 0.9632\n",
      "Epoch 944/1000\n",
      "14/14 [==============================] - 0s 30ms/step - loss: 0.0326 - acc: 0.9687 - val_loss: 0.0334 - val_acc: 0.9658\n",
      "Epoch 945/1000\n",
      "14/14 [==============================] - 0s 25ms/step - loss: 0.0334 - acc: 0.9690 - val_loss: 0.0352 - val_acc: 0.9658\n",
      "Epoch 946/1000\n",
      "14/14 [==============================] - 0s 27ms/step - loss: 0.0349 - acc: 0.9673 - val_loss: 0.0352 - val_acc: 0.9658\n",
      "Epoch 947/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0352 - acc: 0.9678 - val_loss: 0.0346 - val_acc: 0.9684\n",
      "Epoch 948/1000\n",
      "14/14 [==============================] - 0s 25ms/step - loss: 0.0350 - acc: 0.9643 - val_loss: 0.0339 - val_acc: 0.9684\n",
      "Epoch 949/1000\n",
      "14/14 [==============================] - 0s 26ms/step - loss: 0.0335 - acc: 0.9681 - val_loss: 0.0354 - val_acc: 0.9658\n",
      "Epoch 950/1000\n",
      "14/14 [==============================] - 0s 24ms/step - loss: 0.0361 - acc: 0.9652 - val_loss: 0.0358 - val_acc: 0.9632\n",
      "Epoch 951/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0343 - acc: 0.9678 - val_loss: 0.0348 - val_acc: 0.9684\n",
      "Epoch 952/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0345 - acc: 0.9667 - val_loss: 0.0358 - val_acc: 0.9658\n",
      "Epoch 953/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0339 - acc: 0.9725 - val_loss: 0.0348 - val_acc: 0.9684\n",
      "Epoch 954/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0361 - acc: 0.9675 - val_loss: 0.0340 - val_acc: 0.9632\n",
      "Epoch 955/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0343 - acc: 0.9658 - val_loss: 0.0342 - val_acc: 0.9632\n",
      "Epoch 956/1000\n",
      "14/14 [==============================] - 0s 20ms/step - loss: 0.0370 - acc: 0.9661 - val_loss: 0.0340 - val_acc: 0.9684\n",
      "Epoch 957/1000\n",
      "14/14 [==============================] - 0s 22ms/step - loss: 0.0353 - acc: 0.9673 - val_loss: 0.0347 - val_acc: 0.9684\n",
      "Epoch 958/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0350 - acc: 0.9684 - val_loss: 0.0335 - val_acc: 0.9658\n",
      "Epoch 959/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0352 - acc: 0.9655 - val_loss: 0.0337 - val_acc: 0.9632\n",
      "Epoch 960/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0349 - acc: 0.9687 - val_loss: 0.0344 - val_acc: 0.9684\n",
      "Epoch 961/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0353 - acc: 0.9690 - val_loss: 0.0354 - val_acc: 0.9658\n",
      "Epoch 962/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0345 - acc: 0.9661 - val_loss: 0.0344 - val_acc: 0.9684\n",
      "Epoch 963/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0329 - acc: 0.9681 - val_loss: 0.0334 - val_acc: 0.9684\n",
      "Epoch 964/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0352 - acc: 0.9673 - val_loss: 0.0356 - val_acc: 0.9632\n",
      "Epoch 965/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0349 - acc: 0.9664 - val_loss: 0.0340 - val_acc: 0.9658\n",
      "Epoch 966/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0354 - acc: 0.9673 - val_loss: 0.0342 - val_acc: 0.9711\n",
      "Epoch 967/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0328 - acc: 0.9687 - val_loss: 0.0352 - val_acc: 0.9632\n",
      "Epoch 968/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0337 - acc: 0.9690 - val_loss: 0.0326 - val_acc: 0.9684\n",
      "Epoch 969/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0345 - acc: 0.9670 - val_loss: 0.0344 - val_acc: 0.9658\n",
      "Epoch 970/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0333 - acc: 0.9681 - val_loss: 0.0340 - val_acc: 0.9658\n",
      "Epoch 971/1000\n",
      "14/14 [==============================] - 0s 14ms/step - loss: 0.0335 - acc: 0.9661 - val_loss: 0.0337 - val_acc: 0.9658\n",
      "Epoch 972/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0328 - acc: 0.9716 - val_loss: 0.0336 - val_acc: 0.9658\n",
      "Epoch 973/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0333 - acc: 0.9675 - val_loss: 0.0339 - val_acc: 0.9658\n",
      "Epoch 974/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0338 - acc: 0.9661 - val_loss: 0.0346 - val_acc: 0.9711\n",
      "Epoch 975/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0326 - acc: 0.9678 - val_loss: 0.0319 - val_acc: 0.9711\n",
      "Epoch 976/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0338 - acc: 0.9693 - val_loss: 0.0368 - val_acc: 0.9632\n",
      "Epoch 977/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0374 - acc: 0.9667 - val_loss: 0.0344 - val_acc: 0.9658\n",
      "Epoch 978/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0350 - acc: 0.9699 - val_loss: 0.0332 - val_acc: 0.9711\n",
      "Epoch 979/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0334 - acc: 0.9678 - val_loss: 0.0361 - val_acc: 0.9605\n",
      "Epoch 980/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0338 - acc: 0.9684 - val_loss: 0.0333 - val_acc: 0.9711\n",
      "Epoch 981/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0335 - acc: 0.9702 - val_loss: 0.0336 - val_acc: 0.9737\n",
      "Epoch 982/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0335 - acc: 0.9690 - val_loss: 0.0338 - val_acc: 0.9658\n",
      "Epoch 983/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0346 - acc: 0.9655 - val_loss: 0.0331 - val_acc: 0.9711\n",
      "Epoch 984/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0336 - acc: 0.9702 - val_loss: 0.0361 - val_acc: 0.9658\n",
      "Epoch 985/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0340 - acc: 0.9722 - val_loss: 0.0337 - val_acc: 0.9658\n",
      "Epoch 986/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0322 - acc: 0.9711 - val_loss: 0.0329 - val_acc: 0.9684\n",
      "Epoch 987/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0337 - acc: 0.9696 - val_loss: 0.0361 - val_acc: 0.9658\n",
      "Epoch 988/1000\n",
      "14/14 [==============================] - 0s 17ms/step - loss: 0.0364 - acc: 0.9673 - val_loss: 0.0330 - val_acc: 0.9658\n",
      "Epoch 989/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0336 - acc: 0.9643 - val_loss: 0.0331 - val_acc: 0.9711\n",
      "Epoch 990/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0342 - acc: 0.9693 - val_loss: 0.0338 - val_acc: 0.9684\n",
      "Epoch 991/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0339 - acc: 0.9678 - val_loss: 0.0343 - val_acc: 0.9658\n",
      "Epoch 992/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0331 - acc: 0.9711 - val_loss: 0.0347 - val_acc: 0.9684\n",
      "Epoch 993/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0332 - acc: 0.9673 - val_loss: 0.0336 - val_acc: 0.9658\n",
      "Epoch 994/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0325 - acc: 0.9690 - val_loss: 0.0338 - val_acc: 0.9711\n",
      "Epoch 995/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0367 - acc: 0.9649 - val_loss: 0.0341 - val_acc: 0.9711\n",
      "Epoch 996/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0326 - acc: 0.9684 - val_loss: 0.0319 - val_acc: 0.9737\n",
      "Epoch 997/1000\n",
      "14/14 [==============================] - 0s 15ms/step - loss: 0.0350 - acc: 0.9667 - val_loss: 0.0366 - val_acc: 0.9658\n",
      "Epoch 998/1000\n",
      "14/14 [==============================] - 0s 19ms/step - loss: 0.0339 - acc: 0.9699 - val_loss: 0.0348 - val_acc: 0.9632\n",
      "Epoch 999/1000\n",
      "14/14 [==============================] - 0s 16ms/step - loss: 0.0341 - acc: 0.9670 - val_loss: 0.0320 - val_acc: 0.9658\n",
      "Epoch 1000/1000\n",
      "14/14 [==============================] - 0s 18ms/step - loss: 0.0339 - acc: 0.9699 - val_loss: 0.0349 - val_acc: 0.9658\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(emg_data, words, test_size=.10)\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Conv1D(40, 10, strides=2, padding='same', activation='relu', input_shape=(samples, 4)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(MaxPooling1D(3))\n",
    "model.add(GlobalAveragePooling1D())\n",
    "model.add(Dense(50, activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(possibilities, activation='softmax'))\n",
    "\n",
    "model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['acc'])\n",
    "\n",
    "history = model.fit(X_train, y_train, validation_data=(X_test, y_test), batch_size=256, epochs=1000, verbose=1, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "108dcd66-95b4-4aaf-8758-dee3ddc8bd0c",
   "metadata": {},
   "source": [
    "Here's the loss chart using this model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1209bad0-d7c4-43b1-9e44-d8ee6f1a2de8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f6807fa0c50>]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15dc640b",
   "metadata": {},
   "source": [
    "Now that our model is trained, we can predict single words with a stream of samples gathered from EMG recording:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f8d651ff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 74ms/step\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Open hand</th>\n",
       "      <th>Closed hand</th>\n",
       "      <th>Lateral pinch</th>\n",
       "      <th>Signaling sign</th>\n",
       "      <th>Rock sign</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>7.571407e-24</td>\n",
       "      <td>8.682827e-21</td>\n",
       "      <td>2.022934e-20</td>\n",
       "      <td>1.319103e-21</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Open hand   Closed hand  Lateral pinch  Signaling sign     Rock sign\n",
       "0        1.0  7.571407e-24   8.682827e-21    2.022934e-20  1.319103e-21"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data = emg_data[0].astype('float32')\n",
    "test_data = np.expand_dims(test_data, axis=0)\n",
    "prediction_table = model.predict(test_data)\n",
    "\n",
    "pd.DataFrame(prediction_table, columns=words_map)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "fc3496e4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predicted word: Open hand\n"
     ]
    }
   ],
   "source": [
    "predicted_word_index = np.argmax(prediction_table, axis=1)\n",
    "predicted_word = pd.DataFrame(words_map).iloc[predicted_word_index].values[0][0]\n",
    "\n",
    "print(\"Predicted word: %s\" % (predicted_word))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f87cbe17",
   "metadata": {},
   "source": [
    "When predicting in real time with a classifier stream, to guarantee the predicted word is right, we can also create a history list with the last ~10 predictions or so, and assume the word being subvocalized is the most seen in the history."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "language": "python",
   "name": "venv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}