emotiv_custom_auto_rf.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import mne\n",
    "import os\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3nine_EPOCX_243632_2024.11.20T22.15.58.05.00.md.csv\n",
      "Processing file:  3nine_EPOCX_243632_2024.11.20T22.15.58.05.00.md.csv\n",
      "4nine_EPOCX_243632_2024.11.20T22.16.54.05.00.md.csv\n",
      "Processing file:  4nine_EPOCX_243632_2024.11.20T22.16.54.05.00.md.csv\n",
      "1nine_EPOCX_243632_2024.11.20T22.14.14.05.00.md.csv\n",
      "Processing file:  1nine_EPOCX_243632_2024.11.20T22.14.14.05.00.md.csv\n",
      "nine3_EPOCX_243632_2024.11.20T01.20.37.05.00.md.csv\n",
      "Processing file:  nine3_EPOCX_243632_2024.11.20T01.20.37.05.00.md.csv\n",
      "2nine_EPOCX_243632_2024.11.20T22.15.09.05.00.md.csv\n",
      "Processing file:  2nine_EPOCX_243632_2024.11.20T22.15.09.05.00.md.csv\n",
      "nine2_EPOCX_243632_2024.11.20T00.53.45.05.00.md.csv\n",
      "Processing file:  nine2_EPOCX_243632_2024.11.20T00.53.45.05.00.md.csv\n",
      "nine_EPOCX_243632_2024.11.20T00.53.06.05.00.md.csv\n",
      "Processing file:  nine_EPOCX_243632_2024.11.20T00.53.06.05.00.md.csv\n",
      "nine4_EPOCX_243632_2024.11.20T01.21.26.05.00.md.csv\n",
      "Processing file:  nine4_EPOCX_243632_2024.11.20T01.21.26.05.00.md.csv\n",
      "2zero_EPOCX_243632_2024.11.20T22.18.48.05.00.md.csv\n",
      "Processing file:  2zero_EPOCX_243632_2024.11.20T22.18.48.05.00.md.csv\n",
      "zero_EPOCX_243632_2024.11.20T00.54.20.05.00.md.csv\n",
      "Processing file:  zero_EPOCX_243632_2024.11.20T00.54.20.05.00.md.csv\n",
      "zero4_EPOCX_243632_2024.11.20T01.23.06.05.00.md.csv\n",
      "Processing file:  zero4_EPOCX_243632_2024.11.20T01.23.06.05.00.md.csv\n",
      "3zero_EPOCX_243632_2024.11.20T22.19.44.05.00.md.csv\n",
      "Processing file:  3zero_EPOCX_243632_2024.11.20T22.19.44.05.00.md.csv\n",
      "zero2_EPOCX_243632_2024.11.20T00.55.04.05.00.md.csv\n",
      "Processing file:  zero2_EPOCX_243632_2024.11.20T00.55.04.05.00.md.csv\n",
      "4zero_EPOCX_243632_2024.11.20T22.20.32.05.00.md.csv\n",
      "Processing file:  4zero_EPOCX_243632_2024.11.20T22.20.32.05.00.md.csv\n",
      "1zero_EPOCX_243632_2024.11.20T22.17.56.05.00.md.csv\n",
      "Processing file:  1zero_EPOCX_243632_2024.11.20T22.17.56.05.00.md.csv\n",
      "zero3_EPOCX_243632_2024.11.20T01.22.12.05.00.md.csv\n",
      "Processing file:  zero3_EPOCX_243632_2024.11.20T01.22.12.05.00.md.csv\n",
      "seven_EPOCX_243632_2024.11.20T00.50.31.05.00.md.csv\n",
      "Processing file:  seven_EPOCX_243632_2024.11.20T00.50.31.05.00.md.csv\n",
      "1seven_EPOCX_243632_2024.11.20T22.06.08.05.00.md.csv\n",
      "Processing file:  1seven_EPOCX_243632_2024.11.20T22.06.08.05.00.md.csv\n",
      "seven4_EPOCX_243632_2024.11.20T01.19.01.05.00.md.csv\n",
      "Processing file:  seven4_EPOCX_243632_2024.11.20T01.19.01.05.00.md.csv\n",
      "4seven_EPOCX_243632_2024.11.20T22.09.14.05.00.md.csv\n",
      "Processing file:  4seven_EPOCX_243632_2024.11.20T22.09.14.05.00.md.csv\n",
      "3seven_EPOCX_243632_2024.11.20T22.08.06.05.00.md.csv\n",
      "Processing file:  3seven_EPOCX_243632_2024.11.20T22.08.06.05.00.md.csv\n",
      "2seven_EPOCX_243632_2024.11.20T22.06.57.05.00.md.csv\n",
      "Processing file:  2seven_EPOCX_243632_2024.11.20T22.06.57.05.00.md.csv\n",
      "seven3_EPOCX_243632_2024.11.20T01.18.12.05.00.md.csv\n",
      "Processing file:  seven3_EPOCX_243632_2024.11.20T01.18.12.05.00.md.csv\n",
      "seven2_EPOCX_243632_2024.11.20T00.51.08.05.00.md.csv\n",
      "Processing file:  seven2_EPOCX_243632_2024.11.20T00.51.08.05.00.md.csv\n",
      "six4_EPOCX_243632_2024.11.20T01.17.19.05.00.md.csv\n",
      "Processing file:  six4_EPOCX_243632_2024.11.20T01.17.19.05.00.md.csv\n",
      "six2_EPOCX_243632_2024.11.20T00.49.52.05.00.md.csv\n",
      "Processing file:  six2_EPOCX_243632_2024.11.20T00.49.52.05.00.md.csv\n",
      "2six_EPOCX_243632_2024.11.20T22.02.35.05.00.md.csv\n",
      "Processing file:  2six_EPOCX_243632_2024.11.20T22.02.35.05.00.md.csv\n",
      "3six_EPOCX_243632_2024.11.20T22.03.23.05.00.md.csv\n",
      "Processing file:  3six_EPOCX_243632_2024.11.20T22.03.23.05.00.md.csv\n",
      "3six_EPOCX_243632_2024.11.20T22.04.22.05.00.md.csv\n",
      "Processing file:  3six_EPOCX_243632_2024.11.20T22.04.22.05.00.md.csv\n",
      "six_EPOCX_243632_2024.11.20T00.49.13.05.00.md.csv\n",
      "Processing file:  six_EPOCX_243632_2024.11.20T00.49.13.05.00.md.csv\n",
      "4six_EPOCX_243632_2024.11.20T22.05.07.05.00.md.csv\n",
      "Processing file:  4six_EPOCX_243632_2024.11.20T22.05.07.05.00.md.csv\n",
      "1six_EPOCX_243632_2024.11.20T22.01.36.05.00.md.csv\n",
      "Processing file:  1six_EPOCX_243632_2024.11.20T22.01.36.05.00.md.csv\n",
      "six3_EPOCX_243632_2024.11.20T01.16.23.05.00.md.csv\n",
      "Processing file:  six3_EPOCX_243632_2024.11.20T01.16.23.05.00.md.csv\n",
      "3one_EPOCX_243632_2024.11.20T21.41.40.05.00.md.csv\n",
      "Processing file:  3one_EPOCX_243632_2024.11.20T21.41.40.05.00.md.csv\n",
      "one2_EPOCX_243632_2024.11.20T00.42.26.05.00.md.csv\n",
      "Processing file:  one2_EPOCX_243632_2024.11.20T00.42.26.05.00.md.csv\n",
      "1one_EPOCX_243632_2024.11.20T21.39.16.05.00.md.csv\n",
      "Processing file:  1one_EPOCX_243632_2024.11.20T21.39.16.05.00.md.csv\n",
      "one_EPOCX_247201_2024.11.20T00.39.57.05.00.md.csv\n",
      "Processing file:  one_EPOCX_247201_2024.11.20T00.39.57.05.00.md.csv\n",
      "4one_EPOCX_243632_2024.11.20T21.42.49.05.00.md.csv\n",
      "Processing file:  4one_EPOCX_243632_2024.11.20T21.42.49.05.00.md.csv\n",
      "one4_EPOCX_243632_2024.11.20T01.07.32.05.00.csv\n",
      "Processing file:  one4_EPOCX_243632_2024.11.20T01.07.32.05.00.csv\n",
      "one3_EPOCX_243632_2024.11.20T01.06.29.05.00.csv\n",
      "Processing file:  one3_EPOCX_243632_2024.11.20T01.06.29.05.00.csv\n",
      "2one_EPOCX_243632_2024.11.20T21.40.36.05.00.md.csv\n",
      "Processing file:  2one_EPOCX_243632_2024.11.20T21.40.36.05.00.md.csv\n",
      "1eight_EPOCX_243632_2024.11.20T22.10.10.05.00.md.csv\n",
      "Processing file:  1eight_EPOCX_243632_2024.11.20T22.10.10.05.00.md.csv\n",
      "eight2_EPOCX_243632_2024.11.20T00.52.26.05.00.md.csv\n",
      "Processing file:  eight2_EPOCX_243632_2024.11.20T00.52.26.05.00.md.csv\n",
      "3eight_EPOCX_243632_2024.11.20T22.12.21.05.00.md.csv\n",
      "Processing file:  3eight_EPOCX_243632_2024.11.20T22.12.21.05.00.md.csv\n",
      "eight3_EPOCX_243632_2024.11.20T01.19.57.05.00.md.csv\n",
      "Processing file:  eight3_EPOCX_243632_2024.11.20T01.19.57.05.00.md.csv\n",
      "4eight_EPOCX_243632_2024.11.20T22.13.25.05.00.md.csv\n",
      "Processing file:  4eight_EPOCX_243632_2024.11.20T22.13.25.05.00.md.csv\n",
      "2eight_EPOCX_243632_2024.11.20T22.11.15.05.00.md.csv\n",
      "Processing file:  2eight_EPOCX_243632_2024.11.20T22.11.15.05.00.md.csv\n",
      "eight_EPOCX_243632_2024.11.20T00.51.45.05.00.md.csv\n",
      "Processing file:  eight_EPOCX_243632_2024.11.20T00.51.45.05.00.md.csv\n",
      "four2_EPOCX_243632_2024.11.20T00.46.41.05.00.md.csv\n",
      "Processing file:  four2_EPOCX_243632_2024.11.20T00.46.41.05.00.md.csv\n",
      "2four_EPOCX_243632_2024.11.20T21.54.06.05.00.md.csv\n",
      "Processing file:  2four_EPOCX_243632_2024.11.20T21.54.06.05.00.md.csv\n",
      "4four_EPOCX_243632_2024.11.20T21.56.12.05.00.md.csv\n",
      "Processing file:  4four_EPOCX_243632_2024.11.20T21.56.12.05.00.md.csv\n",
      "3four_EPOCX_243632_2024.11.20T21.55.07.05.00.md.csv\n",
      "Processing file:  3four_EPOCX_243632_2024.11.20T21.55.07.05.00.md.csv\n",
      "four3_EPOCX_243632_2024.11.20T01.13.46.05.00.md.csv\n",
      "Processing file:  four3_EPOCX_243632_2024.11.20T01.13.46.05.00.md.csv\n",
      "1four_EPOCX_243632_2024.11.20T21.52.55.05.00.md.csv\n",
      "Processing file:  1four_EPOCX_243632_2024.11.20T21.52.55.05.00.md.csv\n",
      "four_EPOCX_243632_2024.11.20T00.46.01.05.00.md.csv\n",
      "Processing file:  four_EPOCX_243632_2024.11.20T00.46.01.05.00.md.csv\n",
      "three3_EPOCX_243632_2024.11.20T01.10.53.05.00.md.csv\n",
      "Processing file:  three3_EPOCX_243632_2024.11.20T01.10.53.05.00.md.csv\n",
      "2three_EPOCX_243632_2024.11.20T21.49.56.05.00.md.csv\n",
      "Processing file:  2three_EPOCX_243632_2024.11.20T21.49.56.05.00.md.csv\n",
      "4three_EPOCX_243632_2024.11.20T21.52.02.05.00.md.csv\n",
      "Processing file:  4three_EPOCX_243632_2024.11.20T21.52.02.05.00.md.csv\n",
      "three4_EPOCX_243632_2024.11.20T01.12.52.05.00.md.csv\n",
      "Processing file:  three4_EPOCX_243632_2024.11.20T01.12.52.05.00.md.csv\n",
      "three_EPOCX_243632_2024.11.20T00.44.38.05.00.md.csv\n",
      "Processing file:  three_EPOCX_243632_2024.11.20T00.44.38.05.00.md.csv\n",
      "1three_EPOCX_243632_2024.11.20T21.48.56.05.00.md.csv\n",
      "Processing file:  1three_EPOCX_243632_2024.11.20T21.48.56.05.00.md.csv\n",
      "three2_EPOCX_243632_2024.11.20T00.45.15.05.00.md.csv\n",
      "Processing file:  three2_EPOCX_243632_2024.11.20T00.45.15.05.00.md.csv\n",
      "three4_EPOCX_243632_2024.11.20T01.11.53.05.00.md.csv\n",
      "Processing file:  three4_EPOCX_243632_2024.11.20T01.11.53.05.00.md.csv\n",
      "3three_EPOCX_243632_2024.11.20T21.51.02.05.00.md.csv\n",
      "Processing file:  3three_EPOCX_243632_2024.11.20T21.51.02.05.00.md.csv\n",
      "two4_EPOCX_243632_2024.11.20T01.10.02.05.00.md.csv\n",
      "Processing file:  two4_EPOCX_243632_2024.11.20T01.10.02.05.00.md.csv\n",
      "4two_EPOCX_243632_2024.11.20T21.48.04.05.00.md.csv\n",
      "Processing file:  4two_EPOCX_243632_2024.11.20T21.48.04.05.00.md.csv\n",
      "1two_EPOCX_243632_2024.11.20T21.44.03.05.00.md.csv\n",
      "Processing file:  1two_EPOCX_243632_2024.11.20T21.44.03.05.00.md.csv\n",
      "3two_EPOCX_243632_2024.11.20T21.46.16.05.00.md.csv\n",
      "Processing file:  3two_EPOCX_243632_2024.11.20T21.46.16.05.00.md.csv\n",
      "two2_EPOCX_247204_2024.11.20T00.43.59.05.00.md.csv\n",
      "Processing file:  two2_EPOCX_247204_2024.11.20T00.43.59.05.00.md.csv\n",
      "2two_EPOCX_243632_2024.11.20T21.45.16.05.00.md.csv\n",
      "Processing file:  2two_EPOCX_243632_2024.11.20T21.45.16.05.00.md.csv\n",
      "two_EPOCX_247203_2024.11.20T00.43.19.05.00.md.csv\n",
      "Processing file:  two_EPOCX_247203_2024.11.20T00.43.19.05.00.md.csv\n",
      "3two_EPOCX_243632_2024.11.20T21.47.17.05.00.md.csv\n",
      "Processing file:  3two_EPOCX_243632_2024.11.20T21.47.17.05.00.md.csv\n",
      "two3_EPOCX_243632_2024.11.20T01.09.03.05.00.md.csv\n",
      "Processing file:  two3_EPOCX_243632_2024.11.20T01.09.03.05.00.md.csv\n",
      "five3_EPOCX_243632_2024.11.20T01.14.40.05.00.md.csv\n",
      "Processing file:  five3_EPOCX_243632_2024.11.20T01.14.40.05.00.md.csv\n",
      "4five_EPOCX_243632_2024.11.20T22.00.49.05.00.md.csv\n",
      "Processing file:  4five_EPOCX_243632_2024.11.20T22.00.49.05.00.md.csv\n",
      "3five_EPOCX_243632_2024.11.20T21.59.24.05.00.md.csv\n",
      "Processing file:  3five_EPOCX_243632_2024.11.20T21.59.24.05.00.md.csv\n",
      "2five_EPOCX_243632_2024.11.20T21.58.28.05.00.md.csv\n",
      "Processing file:  2five_EPOCX_243632_2024.11.20T21.58.28.05.00.md.csv\n",
      "1five_EPOCX_243632_2024.11.20T21.57.18.05.00.md.csv\n",
      "Processing file:  1five_EPOCX_243632_2024.11.20T21.57.18.05.00.md.csv\n",
      "five_EPOCX_243632_2024.11.20T00.47.25.05.00.md.csv\n",
      "Processing file:  five_EPOCX_243632_2024.11.20T00.47.25.05.00.md.csv\n",
      "five2_EPOCX_243632_2024.11.20T00.48.08.05.00.md.csv\n",
      "Processing file:  five2_EPOCX_243632_2024.11.20T00.48.08.05.00.md.csv\n",
      "five2_EPOCX_243632_2024.11.20T01.15.33.05.00.md.csv\n",
      "Processing file:  five2_EPOCX_243632_2024.11.20T01.15.33.05.00.md.csv\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "folder_name = \"digit\"\n",
    "root_folder = f\"./data/emotiv/custom/{folder_name}\"\n",
    "\n",
    "dfs = {\n",
    "    \n",
    "}\n",
    "\n",
    "pad = 600\n",
    "\n",
    "for folder in os.listdir(root_folder):\n",
    "    # check if folder is a directory\n",
    "    isFolder = os.path.isdir(os.path.join(root_folder, folder))\n",
    "    folder_name = folder\n",
    "    dfs[folder_name] = []\n",
    "    if isFolder:\n",
    "        for file in os.listdir(os.path.join(root_folder, folder)):\n",
    "            print(file)\n",
    "            if file.endswith(\".csv\"):\n",
    "                print(\"Processing file: \", file)\n",
    "                df = pd.read_csv(os.path.join(root_folder, folder, file), header=1)\n",
    "                # merge with previous dataframes\n",
    "                # remove the fist 256 rows since there might be some noise in the beginning\n",
    "                df = df[pad:-pad]\n",
    "                dfs[folder_name].append(df) \n",
    "                "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "eeg_columns = [\"EEG.AF3\", \"EEG.F7\", \"EEG.F3\", \"EEG.FC5\", \"EEG.T7\", \"EEG.P7\", \"EEG.O1\", \"EEG.O2\", \"EEG.P8\", \"EEG.T8\", \"EEG.FC6\", \"EEG.F4\", \"EEG.F8\", \"EEG.AF4\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy.signal import butter, filtfilt, lfilter\n",
    "\n",
    "def butter_bandpass(lowcut, highcut, fs, order=4):\n",
    "    nyquist = 0.5 * fs\n",
    "    low = lowcut / nyquist\n",
    "    high = highcut / nyquist\n",
    "    b, a = butter(order, [low, high], btype='band')\n",
    "    return b, a\n",
    "\n",
    "def bandpass_filter(data, lowcut, highcut, fs, order=4):\n",
    "    b, a = butter_bandpass(lowcut, highcut, fs, order=order)\n",
    "    y = lfilter(b, a, data)\n",
    "    # y = filtfilt(b, a, data)\n",
    "    return y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.signal import stft\n",
    "\n",
    "def compute_stft_for_segment(segment, fs, n_fft=256, hop_length=128):\n",
    "    \"\"\"\n",
    "    Compute STFT for a multi-channel EEG segment.\n",
    "    Args:\n",
    "        segment: 2D array of shape (256, 14), where rows are time points and columns are channels.\n",
    "        fs: Sampling frequency in Hz.\n",
    "        n_fft: Number of FFT points.\n",
    "        hop_length: Number of overlapping samples between windows.\n",
    "    Returns:\n",
    "        stft_results: 3D array of shape (n_channels, n_freq_bins, n_time_frames)\n",
    "                      containing the magnitude spectrograms for each channel.\n",
    "    \"\"\"\n",
    "    n_channels = segment.shape[1]\n",
    "    stft_results = []\n",
    "    \n",
    "    for ch in range(n_channels):\n",
    "        # Compute STFT for each channel\n",
    "        _, _, Zxx = stft(segment[:, ch], fs, nperseg=n_fft, noverlap=hop_length)\n",
    "        stft_results.append(np.abs(Zxx))  # Use magnitude spectrogram\n",
    "    \n",
    "    return np.array(stft_results)  # Shape: (n_channels, n_freq_bins, n_time_frames)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "0\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "7\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "6\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "1\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "8\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "4\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "3\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "2\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "5\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n",
      "(256, 14)\n"
     ]
    }
   ],
   "source": [
    "train = {}\n",
    "\n",
    "lowcut = 0.5\n",
    "highcut = 30\n",
    "fs = 128\n",
    "\n",
    "for key in dfs.keys():\n",
    "    train[key] = []\n",
    "    print(key)\n",
    "    min = np.inf\n",
    "    max = -np.inf\n",
    "    for df in dfs[key]:\n",
    "        # split the data by 256 samples\n",
    "        for i in range(0, len(df), 256):\n",
    "            if i + 256 < len(df):\n",
    "                data = df[i:i+256][eeg_columns].values\n",
    "                print(data.shape)\n",
    "                # apply bandpass filter\n",
    "                data = bandpass_filter(data, lowcut, highcut, fs)\n",
    "                extracted = compute_stft_for_segment(data, fs)\n",
    "                if extracted.min() < min:\n",
    "                    min = extracted.min()\n",
    "                if extracted.max() > max:\n",
    "                    max = extracted.max()\n",
    "                train[key].append(extracted)\n",
    "    # normalize the data for consistency\n",
    "    train[key] = (train[key] - min) / (max - min)\n",
    "    # change to cnn input\n",
    "    train[key] = np.array(train[key])\n",
    "    num_samples, num_channels, height, width = train[key].shape\n",
    "    # train[key] = train[key].reshape(num_samples, num_channels, height, width)\n",
    "    # merge the 14 channels in to one picture\n",
    "\n",
    "# save min and max values\n",
    "np.save(\"min_stft.npy\", min)\n",
    "np.save(\"max_stft.npy\", max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(523, 14, 129, 3)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[\"0\"].shape # we treat them as image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5469, 14, 129, 3)\n",
      "[9 9 9 ... 5 5 5]\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# Combine data from the dictionary\n",
    "X = []\n",
    "y = []\n",
    "labels = []\n",
    "\n",
    "for label, data in train.items():\n",
    "    X.append(data)  # Add all the spectrograms\n",
    "    y.extend([label] * data.shape[0])  # Extend labels for each sample\n",
    "    # create a mapping of the labels\n",
    "    labels.append(label)\n",
    "    \n",
    "# use categorical encoding\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "le = LabelEncoder()\n",
    "y = le.fit_transform(y)\n",
    "\n",
    "# Convert to numpy arrays\n",
    "X = np.concatenate(X, axis=0)\n",
    "print(X.shape)\n",
    "y = np.array(y)\n",
    "# change y to categorical\n",
    "y = le.transform(y)\n",
    "\n",
    "print(y)\n",
    "# labels count\n",
    "print(len(le.classes_))\n",
    "\n",
    "outputs = len(le.classes_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 9, 9, ..., 5, 5, 5])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5469, 14, 129, 3)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X shape: (5469, 14, 129, 3)\n",
      "y shape: (5469,)\n",
      "Training set shape: (3828, 14, 129, 3)\n",
      "Validation set shape: (820, 14, 129, 3)\n",
      "Test set shape: (821, 14, 129, 3)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "print(\"X shape:\", X.shape)\n",
    "print(\"y shape:\", y.shape)\n",
    "\n",
    "# Split data into train, validation, and test sets\n",
    "X_train, X_temp, y_train, y_temp = train_test_split(X, y, test_size=0.3, random_state=42)\n",
    "X_val, X_test, y_val, y_test = train_test_split(X_temp, y_temp, test_size=0.5, random_state=42)\n",
    "\n",
    "print(\"Training set shape:\", X_train.shape)\n",
    "print(\"Validation set shape:\", X_val.shape)\n",
    "print(\"Test set shape:\", X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3828, 5418)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.reshape(X_train.shape[0], -1).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train accuracy: 1.0000\n",
      "Validation accuracy: 0.9146\n",
      "Test accuracy: 0.9220\n",
      "(821, 5418)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwwAAAK9CAYAAACJnusfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB4K0lEQVR4nO3dd3gUZff/8c8mkEILIXSR0Am9CEJAwEIRFalSRA1NUaMCEURApChEEbDRVKSIIGKBRwUpglI0IL2LNAHpCSSUFCDZ3x/+2O+OZDTB3Z2w+34911zXk3snM+fsJJKz575nbHa73S4AAAAAyISf1QEAAAAAyLkoGAAAAACYomAAAAAAYIqCAQAAAIApCgYAAAAApigYAAAAAJiiYAAAAABgioIBAAAAgCkKBgAAAACmKBgAIBP79+9Xy5YtFRISIpvNpkWLFrn0+H/88YdsNptmzZrl0uPeyu6++27dfffdVocBAPgbCgYAOdbBgwfVt29flStXTkFBQSpQoIAaN26sd999VykpKW49d1RUlHbu3KkxY8Zozpw5qlevnlvP50k9evSQzWZTgQIFMn0f9+/fL5vNJpvNpvHjx2f7+CdOnNDIkSO1bds2F0QLALBaLqsDAIDMLF68WI888ogCAwP1xBNPqHr16rpy5YrWrVunQYMGaffu3frwww/dcu6UlBTFxcVp2LBheu6559xyjvDwcKWkpCh37txuOf6/yZUrl5KTk/Xtt9+qc+fOhtfmzp2roKAgpaam3tSxT5w4oVGjRqlMmTKqXbt2lr9v+fLlN3U+AIB7UTAAyHEOHz6srl27Kjw8XKtWrVKJEiUcr0VHR+vAgQNavHix285/9uxZSVLBggXddg6bzaagoCC3Hf/fBAYGqnHjxvrss89uKBjmzZunBx98UF999ZVHYklOTlaePHkUEBDgkfMBALKHKUkAcpxx48bp0qVL+vjjjw3FwnUVKlRQv379HF9fu3ZNr732msqXL6/AwECVKVNGQ4cOVVpamuH7ypQpo4ceekjr1q3TnXfeqaCgIJUrV06ffPKJY5+RI0cqPDxckjRo0CDZbDaVKVNG0l9Tea7/f2cjR46UzWYzjK1YsUJ33XWXChYsqHz58qly5coaOnSo43WzNQyrVq1SkyZNlDdvXhUsWFBt27bV3r17Mz3fgQMH1KNHDxUsWFAhISHq2bOnkpOTzd/Yv3n00Uf1/fffKzEx0TG2ceNG7d+/X48++ugN+587d04DBw5UjRo1lC9fPhUoUECtW7fW9u3bHfv89NNPql+/viSpZ8+ejqlN1/O8++67Vb16dW3evFlNmzZVnjx5HO/L39cwREVFKSgo6Ib8W7VqpdDQUJ04cSLLuQIAbh4FA4Ac59tvv1W5cuXUqFGjLO3fp08fvfrqq6pbt67efvttNWvWTLGxseratesN+x44cECdOnVSixYtNGHCBIWGhqpHjx7avXu3JKlDhw56++23JUndunXTnDlz9M4772Qr/t27d+uhhx5SWlqaRo8erQkTJujhhx/Wzz///I/f98MPP6hVq1Y6c+aMRo4cqZiYGP3yyy9q3Lix/vjjjxv279y5sy5evKjY2Fh17txZs2bN0qhRo7IcZ4cOHWSz2fT11187xubNm6eIiAjVrVv3hv0PHTqkRYsW6aGHHtLEiRM1aNAg7dy5U82aNXP88V6lShWNHj1akvTUU09pzpw5mjNnjpo2beo4TkJCglq3bq3atWvrnXfe0T333JNpfO+++66KFCmiqKgopaenS5I++OADLV++XO+//75KliyZ5VwBAP+BHQBykKSkJLske9u2bbO0/7Zt2+yS7H369DGMDxw40C7JvmrVKsdYeHi4XZJ9zZo1jrEzZ87YAwMD7S+++KJj7PDhw3ZJ9rfeestwzKioKHt4ePgNMYwYMcLu/J/Tt99+2y7JfvbsWdO4r59j5syZjrHatWvbixYtak9ISHCMbd++3e7n52d/4oknbjhfr169DMds3769PSwszPScznnkzZvXbrfb7Z06dbLfd999drvdbk9PT7cXL17cPmrUqEzfg9TUVHt6evoNeQQGBtpHjx7tGNu4ceMNuV3XrFkzuyT7tGnTMn2tWbNmhrFly5bZJdlff/11+6FDh+z58uWzt2vX7l9zBAC4Dh0GADnKhQsXJEn58+fP0v5LliyRJMXExBjGX3zxRUm6Ya1D1apV1aRJE8fXRYoUUeXKlXXo0KGbjvnvrq99+N///qeMjIwsfc/Jkye1bds29ejRQ4UKFXKM16xZUy1atHDk6ezpp582fN2kSRMlJCQ43sOsePTRR/XTTz/p1KlTWrVqlU6dOpXpdCTpr3UPfn5//bORnp6uhIQEx3SrLVu2ZPmcgYGB6tmzZ5b2bdmypfr27avRo0erQ4cOCgoK0gcffJDlcwEA/jsKBgA5SoECBSRJFy9ezNL+R44ckZ+fnypUqGAYL168uAoWLKgjR44YxkuXLn3DMUJDQ3X+/PmbjPhGXbp0UePGjdWnTx8VK1ZMXbt21YIFC/6xeLgeZ+XKlW94rUqVKoqPj9fly5cN43/PJTQ0VJKylcsDDzyg/Pnz6/PPP9fcuXNVv379G97L6zIyMvT222+rYsWKCgwMVOHChVWkSBHt2LFDSUlJWT7nbbfdlq0FzuPHj1ehQoW0bds2vffeeypatGiWvxcA8N9RMADIUQoUKKCSJUtq165d2fq+vy86NuPv75/puN1uv+lzXJ9ff11wcLDWrFmjH374QY8//rh27NihLl26qEWLFjfs+1/8l1yuCwwMVIcOHTR79mwtXLjQtLsgSWPHjlVMTIyaNm2qTz/9VMuWLdOKFStUrVq1LHdSpL/en+zYunWrzpw5I0nauXNntr4XAPDfUTAAyHEeeughHTx4UHFxcf+6b3h4uDIyMrR//37D+OnTp5WYmOi445ErhIaGGu4odN3fuxiS5Ofnp/vuu08TJ07Unj17NGbMGK1atUo//vhjpse+Hue+fftueO23335T4cKFlTdv3v+WgIlHH31UW7du1cWLFzNdKH7dl19+qXvuuUcff/yxunbtqpYtW6p58+Y3vCdZLd6y4vLly+rZs6eqVq2qp556SuPGjdPGjRtddnwAwL+jYACQ47z00kvKmzev+vTpo9OnT9/w+sGDB/Xuu+9K+mtKjaQb7mQ0ceJESdKDDz7osrjKly+vpKQk7dixwzF28uRJLVy40LDfuXPnbvje6w8w+/utXq8rUaKEateurdmzZxv+AN+1a5eWL1/uyNMd7rnnHr322muaNGmSihcvbrqfv7//Dd2LL774QsePHzeMXS9sMiuusmvw4ME6evSoZs+erYkTJ6pMmTKKiooyfR8BAK7Hg9sA5Djly5fXvHnz1KVLF1WpUsXwpOdffvlFX3zxhXr06CFJqlWrlqKiovThhx8qMTFRzZo106+//qrZs2erXbt2prfsvBldu3bV4MGD1b59e73wwgtKTk7W1KlTValSJcOi39GjR2vNmjV68MEHFR4erjNnzmjKlCkqVaqU7rrrLtPjv/XWW2rdurUiIyPVu3dvpaSk6P3331dISIhGjhzpsjz+zs/PT6+88sq/7vfQQw9p9OjR6tmzpxo1aqSdO3dq7ty5KleunGG/8uXLq2DBgpo2bZry58+vvHnzqkGDBipbtmy24lq1apWmTJmiESNGOG7zOnPmTN19990aPny4xo0bl63jAQBuDh0GADnSww8/rB07dqhTp0763//+p+joaL388sv6448/NGHCBL333nuOfadPn65Ro0Zp48aN6t+/v1atWqUhQ4Zo/vz5Lo0pLCxMCxcuVJ48efTSSy9p9uzZio2NVZs2bW6IvXTp0poxY4aio6M1efJkNW3aVKtWrVJISIjp8Zs3b66lS5cqLCxMr776qsaPH6+GDRvq559/zvYf2+4wdOhQvfjii1q2bJn69eunLVu2aPHixbr99tsN++XOnVuzZ8+Wv7+/nn76aXXr1k2rV6/O1rkuXryoXr16qU6dOho2bJhjvEmTJurXr58mTJig9evXuyQvAMA/s9mzszoOAAAAgE+hwwAAAADAFAUDAAAAAFMUDAAAAABMUTAAAAAAMEXBAAAAAMAUBQMAAAAAUxQMAAAAAEx55ZOeiz/5pdUhWOKPqZ2sDsESGT76KBE/m83qEAAAuGlBOfiv0OA6z1l27pStkyw7txk6DAAAAABM5eDaDgAAALCAjc/UnfFuAAAAADBFwQAAAADAFFOSAAAAAGfcWMSADgMAAAAAU3QYAAAAAGcsejbg3QAAAABgig4DAAAA4Iw1DAZ0GAAAAACYomAAAAAAYIopSQAAAIAzFj0b8G4AAAAAMEWHAQAAAHDGomcDOgwAAAAATFEwAAAAADDFlCQAAADAGYueDXg3AAAAAJiiwwAAAAA4Y9GzAR0GAAAAAKboMAAAAADOWMNgwLsBAAAAwBQFAwAAAABTTEkCAAAAnLHo2YAOw03YGNtapz7qdMMW+2htSdLXA5vd8Nqbj9WxNmg3mj9vrlq3uFf169RQ966PaOeOHVaH5FabN21Uv+in1eKeJqpTPUI/rvzB6pA8yteu93XkTd6+gLzJG8gMBcNNuH/MStV48VvH9sjENZKkbzcdd+wzZ80hwz6vfbnTqnDdaun3SzR+XKz6Phut+V8sVOXKEXqmb28lJCRYHZrbpKSkqFLlCA0Z9qrVoXicL15vibzJm7y9GXn7Vt5ZZvOzbsuBcmZUOVzCpSs6eyHNsbWoWUKHz1zSL7+fdeyTciXdsM+l1GsWRuw+c2bPVIdOndWufUeVr1BBr4wYpaCgIC36+iurQ3Obu5o0VfQL/XVv8xZWh+Jxvni9JfImb/L2ZuTtW3nj5lAw/Ee5/W3q2KC0Pvv5D8N4xwaltXtiG/00soWGtq+u4AB/awJ0o6tXrmjvnt1qGNnIMebn56eGDRtpx/atFkYGd/DV603e5E3e5O1tfDVv3DxLFz2fPHlSU6dO1bp163Ty5En5+fmpXLlyateunXr06CF//5z/R3brOrcpJE9ufe5UMHy94aj+PJesU4kpqloqRK90rKHyxfOr99Q46wJ1g/OJ55Wenq6wsDDDeFhYmA4fPmRRVHAXX73e5E3eEnl7K/L2rbyzhUXPBpYVDJs2bVLz5s1VoUIFBQcHa//+/Xr00Ud15coVDRw4UDNmzNDSpUuVP3/+fzxOWlqa0tLSDGP29Kuy+ed2Z/gO3e4qo1W7Tul0Uqpj7NO1hx3//7fjF3Q6KVVfvdhM4UXy6sjZyx6JCwAAAHAFy6Yk9e/fXwMGDNCmTZu0du1azZo1S7///rvmz5+vQ4cOKTk5Wa+88sq/Hic2NlYhISGG7fK2hR7IQCpVKI+aVimmuU4FQma2HjonSSpbNJ8nwvKY0IKh8vf3v2GBVEJCggoXLmxRVHAXX73e5E3eEnl7K/L2rbyzhUXPBpZFtWXLFj3++OOOrx999FFt2bJFp0+fVmhoqMaNG6cvv/zyX48zZMgQJSUlGba8tdu7M3SHro3LKP5Cqn7Yeeof96t2e0FJ0unE1H/c71aTOyBAVapW04b1/zfVKiMjQxs2xKlmLe+9jayv8tXrTd7kTd7k7W18NW/cPMumJBUtWlQnT55UuXLlJEmnT5/WtWvXVKBAAUlSxYoVde7cuX89TmBgoAIDAw1jnpiOZLNJXRuHa0HcEaVn2B3j4UXyqsOdpbVy50mdv3xFVUqFaHTnWorbd1Z7jye5PS5Pezyqp4YPHaxq1aqreo2a+nTObKWkpKhd+w5Wh+Y2ycmXdezoUcfXx4//qX2/7VWBkBCVKFHSwsjczxevt0Te5E3e3oy8fSvvLMuhn/RbxbKCoV27dnr66af11ltvKTAwUK+99pqaNWum4OBgSdK+fft02223WRXev2papZhKheW94e5IV69lqGmVonqyeQXlCcylE+eStXjLcb29eK81gbrZ/a0f0Plz5zRl0nuKjz+ryhFVNOWD6Qrz4pbmnl279GSvKMfXE8a9IUlq07adRo95w6qwPMIXr7dE3uRN3t6MvH0rb9wcm91ut//7bq536dIl9e7dW19//bXS09MVGRmpTz/9VGXLlpUkLV++XElJSXrkkUeyfeziT/77VCZv9MfUTlaHYIkMa36ELefHHRwAALewIEvv1fnPgpuNtuzcKatz3oNhLbtU+fLl0+eff67U1FRdu3ZN+fIZFwS3bNnSosgAAADg0/z4UM6Z5bVdUFCQ1SEAAAAAMGF5wQAAAADkKCx6NuDdAAAAAGCKggEAAACAKaYkAQAAAM64E6EBHQYAAAAApugwAAAAAM5Y9GzAuwEAAADAFB0GAAAAwBlrGAzoMAAAAAAwRcEAAAAAwBRTkgAAAABnLHo24N0AAAAAYIoOAwAAAOCMRc8GdBgAAAAAmKJgAAAAAGCKKUkAAACAMxY9G/BuAAAAADBFhwEAAABwxqJnAzoMAAAAAEzRYQAAAACcsYbBgHcDAAAAgCkKBgAAAACmmJIEAAAAOGPRs4FXFgx/TO1kdQiWCOs20+oQLJHwWU+rQ4AHXbmWYXUIlgjIRUMYAGANrywYAAAAgJvGomcD3g0AAAAApigYAAAAAJhiShIAAADgjClJBrwbAAAAAEzRYQAAAACccVtVAzoMAAAAAExRMAAAAAAwxZQkAAAAwBmLng14NwAAAACYosMAAAAAOGPRswEdBgAAAACm6DAAAAAAzljDYMC7AQAAAMAUBQMAAAAAU0xJAgAAAJyx6NmADgMAAAAAU3QYAAAAACc2OgwGdBgAAAAAmKJgAAAAAGCKKUkAAACAE6YkGdFhAAAAAGCKDgMAAADgjAaDAR0GAAAAAKboMAAAAABOWMNgRIfBhebPm6vWLe5V/To11L3rI9q5Y4fVIbmMn59Nw7vU0e7JnRQ/93HtfL+jBnesZdjn4TvD9c0rLXV0Rjdd/qKnapYpZFG0nuHN1/uf+FreXy74TN06tdXdjerp7kb11Ovxrvp53Rqrw/IYX7ve15E3efsCX80b2UfB4CJLv1+i8eNi1ffZaM3/YqEqV47QM317KyEhwerQXCKmbQ31aRmhmI/Xq27/hRo+d5MGtK2hZ1pXceyTNyiXfvnttIZ/usnCSD3D26+3GV/Mu2jR4nquX4w++exLzZ73herd2VAD+z2ngwf2Wx2a2/ni9ZbIm7zJG7eO9PR0DR8+XGXLllVwcLDKly+v1157TXa73bGP3W7Xq6++qhIlSig4OFjNmzfX/v3Z+zeMgsFF5syeqQ6dOqtd+44qX6GCXhkxSkFBQVr09VdWh+YSDSsX1eJNR7Vsy586evaSFq0/opXbj6tehSKOfT5bc1BvfLldP+48aWGknuHt19uML+bd9O571LhJM5UOL6PwMmX17PP9lSdPHu3asd3q0NzOF6+3RN7kTd74a0qSVVt2vPnmm5o6daomTZqkvXv36s0339S4ceP0/vvvO/YZN26c3nvvPU2bNk0bNmxQ3rx51apVK6Wmpmb5PJYWDPHx8Ro3bpzat2+vyMhIRUZGqn379nrrrbd09uxZK0PLlqtXrmjvnt1qGNnIMebn56eGDRtpx/atFkbmOuv3ndHd1UuoQokCkqQa4aFqFFFMy7f+aXFknucL1zszvpq3s/T0dC3/frFSUpJVo1Ztq8NxK1+93uRN3uTtvXnfKtLS0nThwgXDlpaWlum+v/zyi9q2basHH3xQZcqUUadOndSyZUv9+uuvkv7qLrzzzjt65ZVX1LZtW9WsWVOffPKJTpw4oUWLFmU5JssKho0bN6pSpUp67733FBISoqZNm6pp06YKCQnRe++9p4iICG3a9O9TW7LzprrL+cTzSk9PV1hYmGE8LCxM8fHxHo3FXSYs2qEvfzmsre90UOJnUfplXFtNXrxbn687ZHVoHucL1zszvpq3JB3Y/7uaNrxDjevXUuyYUXrr7fdVrnwFq8NyK1+93uRN3hJ5w9oOQ2xsrEJCQgxbbGxspnE2atRIK1eu1O+//y5J2r59u9atW6fWrVtLkg4fPqxTp06pefPmju8JCQlRgwYNFBcXl+X3w7K7JD3//PN65JFHNG3atBvaL3a7XU8//bSef/75f00mNjZWo0aNMowNGz5Cr7w60tUh+7SOkWXV5a7y6vnuau39M1E1yxTSmz3u1MnzKZq7+oDV4QFuFV6mjOYu+FqXLl3SyhXLNHL4EH3w8SdeXzQAADxvyJAhiomJMYwFBgZmuu/LL7+sCxcuKCIiQv7+/kpPT9eYMWPUvXt3SdKpU6ckScWKFTN8X7FixRyvZYVlBcP27ds1a9asTOdq2Ww2DRgwQHXq1PnX42T2ptr9M39T3SW0YKj8/f1vWCiUkJCgwoULezQWdxnzeH1Hl0GSdh89r9sL59OL7Wv4XMHgC9c7M76atyTlzh2g20uHS5KqVK2mPbt3av7cORr66qh/+c5bl69eb/Imb4m8Ya3AwEDTAuHvFixYoLlz52revHmqVq2atm3bpv79+6tkyZKKiopyWUyWTUkqXry4Y35VZn799dcbqqHMBAYGqkCBAoYtq2+yq+QOCFCVqtW0Yf3/dUMyMjK0YUOcatb696LnVhAc6K8MpxX30l85+vngfYp94Xpnxlfzzow9w64rV69YHYZb+er1Jm/yJm/vzTs7bpVFz4MGDdLLL7+srl27qkaNGnr88cc1YMAAxxSm4sWLS5JOnz5t+L7Tp087XssKyzoMAwcO1FNPPaXNmzfrvvvucxQHp0+f1sqVK/XRRx9p/PjxVoWXbY9H9dTwoYNVrVp1Va9RU5/Oma2UlBS1a9/B6tBc4vvNx/RSh1o6Fn9Ze48lqlbZQnquTXXNWfV/t+UKzReg2wvnU4nQPJKkiiVDJEmnE1N0OjHFkrjdxduvtxlfzHvSuxPV6K4mKl68pJKTL2vpku+0edOven/qR1aH5na+eL0l8iZv8satIzk5WX5+xs///f39lZGRIUkqW7asihcvrpUrV6p27dqSpAsXLmjDhg165plnsnweywqG6OhoFS5cWG+//bamTJmi9PR0SX8leccdd2jWrFnq3LmzVeFl2/2tH9D5c+c0ZdJ7io8/q8oRVTTlg+kK85LW3osfr9erXevqnT6RKhISpJPnkjVjxT7FfrnNsc+D9Urrg+gmjq8/GXC3JGnMgq0a+8U2eRNvv95mfDHv8+cSNPKVlxV/9qzy5cuvCpUq6f2pH6lBZGOrQ3M7X7zeEnmTN3lD0i0ygaJNmzYaM2aMSpcurWrVqmnr1q2aOHGievXqJemvTkn//v31+uuvq2LFiipbtqyGDx+ukiVLql27dlk+j81u/9s8EwtcvXrVsSq/cOHCyp079386Xuo1V0R16wnrNtPqECyR8FlPq0OAB125lmF1CJYIyMVjcwB4lyDLPrb+dyGPzrHs3EnzHs/yvhcvXtTw4cO1cOFCnTlzRiVLllS3bt306quvKiAgQNJfNxMaMWKEPvzwQyUmJuquu+7SlClTVKlSpSyfJ0cUDK5GweBbKBh8CwUDAHiHnFwwFOz+qWXnTpz7mGXnNsO/QAAAAABMUTAAAAAAMJWDm0EAAACA52X39qbejg4DAAAAAFN0GAAAAAAndBiM6DAAAAAAMEXBAAAAAMAUU5IAAAAAJ0xJMqLDAAAAAMAUHQYAAADAGQ0GAzoMAAAAAEzRYQAAAACcsIbBiA4DAAAAAFMUDAAAAABMMSUJAAAAcMKUJCM6DAAAAABM0WEAAAAAnNBhMKLDAAAAAMAUBQMAAAAAU0xJAgAAAJwxI8mADgMAAAAAU3QYAAAAACcsejaiwwAAAADAFB0GAAAAwAkdBiOvLBgy7HarQ7BEwmc9rQ7BEqH1n7M6BEuc3zjJ6hAsEZCLxigAAJ7Ev7wAAAAATHllhwEAAAC4WUxJMqLDAAAAAMAUHQYAAADACR0GIzoMAAAAAExRMAAAAAAwxZQkAAAAwBkzkgzoMAAAAAAwRYcBAAAAcMKiZyM6DAAAAABM0WEAAAAAnNBhMKLDAAAAAMAUBQMAAAAAU0xJAgAAAJwwJcmIDgMAAAAAU3QYAAAAAGc0GAzoMAAAAAAwRcEAAAAAwBRTkgAAAAAnLHo2osMAAAAAwBQdBgAAAMAJHQYjOgwAAAAATFEwAAAAADDFlCQAAADACVOSjOgwuMDmTRvVL/pptbiniepUj9CPK3+wOiSPmj9vrlq3uFf169RQ966PaOeOHVaH5FL58gTqrYEdtW/JaJ2Lm6gfZ8XojqqlM933vWFdlbJ1kp579G7PBulB3n69zZA3efsC8iZvIDMUDC6QkpKiSpUjNGTYq1aH4nFLv1+i8eNi1ffZaM3/YqEqV47QM317KyEhwerQXGbqq4/q3oYR6vXKbNXrPFY/xP2mxdOeV8kiIYb9Hr6npu6sUUYnziRaE6gH+ML1zgx5kzd5ey/y9q28s8pms1m25UQUDC5wV5Omin6hv+5t3sLqUDxuzuyZ6tCps9q176jyFSrolRGjFBQUpEVff2V1aC4RFJhb7e6rrWHvLNLPWw7q0LF4jflgiQ4eO6snH2ni2K9kkRBNHPyIeg6dpavX0i2M2L28/XqbIW/yJm/vRd6+lTduDgUDbtrVK1e0d89uNYxs5Bjz8/NTw4aNtGP7Vgsjc51c/n7KlctfqVeuGsZT066qUZ3ykv76FOLj15/Q27NXau+hU1aE6RG+cL0zQ97kTd7k7W18Ne9ssVm45UAUDLhp5xPPKz09XWFhYYbxsLAwxcfHWxSVa11KTtP67Yc05MnWKlEkRH5+NnV9oL4a1Cyr4oULSJJe7NlC19IzNPmzn6wN1s184XpnhrzJWyJvb0XevpU3bp6lBcPzzz+vtWvX/qdjpKWl6cKFC4YtLS3NRRECUq9XPpHNJh1aPkZJG95RdLdmWrB0kzIy7KpT5XZFd7tbT4341OowAQAA3MLSgmHy5Mm6++67ValSJb355ps6dSr70zliY2MVEhJi2Ma/GeuGaPF3oQVD5e/vf8MCqYSEBBUuXNiiqFzv8J/xatnnXYVFxqhi6+Fq8vh45c7lr8PH49W4TnkVLZRPvy8ZrYsb39XFje8qvGSY3ojpoN8Wj7I6dJfylev9d+RN3hJ5eyvy9q28s4NFz0aWT0lavny5HnjgAY0fP16lS5dW27Zt9d133ykjIyNL3z9kyBAlJSUZtoGDh7g5akhS7oAAValaTRvWxznGMjIytGFDnGrWqmNhZO6RnHpFp+IvqGD+YDVvVEXf/bRT8xZvVP3OsWrQ9Q3HduJMot7+5Ae1eXay1SG7lK9d7+vIm7zJm7y9ja/mjZtn+YPbatSoofvuu09vvfWWFi5cqBkzZqhdu3YqVqyYevTooZ49e6pChQqm3x8YGKjAwEDDWPJVu7vDNp4v+bKOHT3q+Pr48T+177e9KhASohIlSno0Fk97PKqnhg8drGrVqqt6jZr6dM5spaSkqF37DlaH5jLNI6vIZpN+/+OMyt9eRGMHtNPvh0/rk2/idO1ahs4lXTbsf/Vauk7HX9D+I2csith9fOF6Z4a8yZu8vRd5+1beWZVTP+m3iuUFw3W5c+dW586d1blzZx09elQzZszQrFmz9MYbbyg9PWffpnLPrl16sleU4+sJ496QJLVp206jx7xhVVgecX/rB3T+3DlNmfSe4uPPqnJEFU35YLrCvKilGZIvSKOff1i3FSuoc0nJ+t/KbRox+Vtdu5a1Lpg38YXrnRnyJm/y9l7k7Vt54+bY7Ha7Zz+Od+Ln56dTp06paNGimb5ut9v1ww8/qEWL7D3fwNMdhpzCz0er4dD6z1kdgiXOb5xkdQgAANy0oBzzsfWNyr/4vWXnPjihtWXnNmPppQoPD5e/v7/p6zabLdvFAgAAAPBf+OhnsKYsLRgOHz5s5ekBAAAA/Isc3AwCAAAAPI9Fz0aW31YVAAAAQM5FhwEAAABwQoPBiA4DAAAAAFMUDAAAAABMMSUJAAAAcMKiZyM6DAAAAABM0WEAAAAAnNBgMKLDAAAAAMAUBQMAAAAAU0xJAgAAAJz4+TEnyRkdBgAAAACm6DAAAAAATlj0bESHAQAAAIApOgwAAACAEx7cZkSHAQAAAIApCgYAAAAAppiSBAAAADhhRpIRHQYAAAAApugwAAAAAE5Y9GxEhwEAAACAKQoGAAAAAKaYkgQAAAA4YUqSER0GAAAAAKa8ssPg56NVYYbdbnUIlji/cZLVIVgitOUYq0OwRMKyoVaHYIk/z6VYHYIlSoflsToEAD7IR/+UNEWHAQAAAIApr+wwAAAAADeLNQxGdBgAAAAAmKJgAAAAAGCKKUkAAACAE2YkGdFhAAAAAGCKDgMAAADghEXPRnQYAAAAAJiiYAAAAABgiilJAAAAgBNmJBnRYQAAAABgig4DAAAA4IRFz0Z0GAAAAACYosMAAAAAOKHBYESHAQAAAIApCgYAAAAAppiSBAAAADhh0bMRHQYAAAAApugwAAAAAE5oMBjRYQAAAABgioIBAAAAgCmmJAEAAABOWPRsRIcBAAAAgCk6DAAAAIATGgxGdBhcaP68uWrd4l7Vr1ND3bs+op07dlgdkltt3rRR/aKfVot7mqhO9Qj9uPIHq0PyKG++3n5+Nr3as5n2zo3Wue9f0u5Pn9XLj91l2CdvUG69/UIrHfj8eZ37/iVtmfGU+rSpa1HE7uMrP+e7tm3W6Jf7Kap9C7VpWkdxa3803Xfy+NfVpmkd/W/BXA9G6Fne/Pv9T8ibvIHMUDC4yNLvl2j8uFj1fTZa879YqMqVI/RM395KSEiwOjS3SUlJUaXKERoy7FWrQ/E4b7/eL3aN1JMP19WA95apdo8P9MqHqxTTtaGebV/Psc+bz7ZQi/rl1HPs/1S7xwea9NVGvf1CKz3YqKKFkbuer/ycp6amqGz5Snp6wJB/3C9uzSrt27NThQoX8VBknuftv99myJu8fSHvrLLZbJZtOREFg4vMmT1THTp1Vrv2HVW+QgW9MmKUgoKCtOjrr6wOzW3uatJU0S/0173NW1gdisd5+/VuWK2Uvvv5dy3dcEBHTydp4ZrftHLTYdWLKOm0z236dNlOrd1+VEdPJ2nG4q3acfC0YR9v4Cs/5/Ua3qXHn4xWZNN7TfdJOHtGH7z7pl4cPla5cnnvjFZv//02Q97k7Qt54+ZQMLjA1StXtHfPbjWMbOQY8/PzU8OGjbRj+1YLI4M7+ML1Xr/7T91Tt4wqlCokSapRrqgiq5fS8l8POu1zXA81qqiShfNLkprWDlfFUoX0w6ZDlsQM98rIyNDE119Rh65RCi9b3upw3MYXfr8zQ97k7Qt54+ZZ+hHRli1bFBoaqrJly0qS5syZo2nTpuno0aMKDw/Xc889p65du/7jMdLS0pSWlmYYs/sHKjAw0G1x/935xPNKT09XWFiYYTwsLEyHD/PHk7fxhes9/rNfVCBvoLbPelrpGRny9/PTiI9/0vyVux37xLy/TJNjHtDBBS/o6rV0ZWTY9eyEJfp5xzELI4e7fDVvpvz8/dWmUzerQ3ErX/j9zgx5k7fk/XlnRw6dGWQZSzsMPXv21MGDf31iOX36dPXt21f16tXTsGHDVL9+fT355JOaMWPGPx4jNjZWISEhhu2tN2M9ET7gtTrdXVVd76uuHmMWKbLvx+rz5jfq37mBures4djn2fb1dGfV29Rx2AI1enqGXp62Uu/0a6V76paxLnC4xYF9e/TNl5+p/9BROXZ+LQDAfSztMOzfv18VK/61QHLKlCl699139eSTTzper1+/vsaMGaNevXqZHmPIkCGKiYkxjNn9PdddkKTQgqHy9/e/YaFQQkKCChcu7NFY4H6+cL3H9r1P4z/7RV/8uEeStPvwWZUuFqJBjzbS3OU7FRSQS6N636Mur36ppRsOSJJ2HTqjmuWLqX/nhvpxyx8WRg9X2719q5LOn1OvRx5wjGWkp2vGlIn65su5+njBEgujcy1f+P3ODHmTt+T9eWcHH44YWdphyJMnj+Lj4yVJx48f15133ml4vUGDBjp8+PA/HiMwMFAFChQwbJ6cjiRJuQMCVKVqNW1YH+cYy8jI0IYNcapZq45HY4H7+cL1Dg7MpQy73TCWnm6X3///D2juXH4KyO1/4z4ZGfLz4z+y3uaeVg/q/ZkL9N7H8x1bocJF1L7rExo1forV4bmUL/x+Z4a8ydsX8sbNs7TD0Lp1a02dOlXTp09Xs2bN9OWXX6pWrVqO1xcsWKAKFSpYGGHWPR7VU8OHDla1atVVvUZNfTpntlJSUtSufQerQ3Ob5OTLOnb0qOPr48f/1L7f9qpASIhKlPCuO+X8nbdf7yVx+zW4e2MdO31Be/44q9oVi+uFR+7UJ99vlyRdTL6iNduOaGzfe5WSdlVHTyepSa1wdW9ZQ4OnetdzCnzl5zwlOVknj//f+pPTJ4/r0P59yleggIoWK6ECIQUN++fKlUuhhQqrVOkyng3UA7z999sMeZO3L+SNm2NpwfDmm2+qcePGatasmerVq6cJEybop59+UpUqVbRv3z6tX79eCxcutDLELLu/9QM6f+6cpkx6T/HxZ1U5ooqmfDBdYV7c2tuza5ee7BXl+HrCuDckSW3attPoMW9YFZZHePv1jnl/uUb0aqZ3+9+vIgXz6GTCJX383VaN/WStY58nXluo0U/eo1nD2ik0f5COnk7SyI9/0kffbLEwctfzlZ/zA/v2aGi//5sS+vGkCZKke+9vowFDR1sVliW8/ffbDHmTty/knVW30pSk48ePa/Dgwfr++++VnJysChUqaObMmapX769nJ9ntdo0YMUIfffSREhMT1bhxY02dOtWxLCArbHb73+YUeFhiYqLeeOMNffvttzp06JAyMjJUokQJNW7cWAMGDHAkmx2p19wQ6C3g79NDfIXfLfRL7UqhLcdYHYIlEpYNtToES/x5LsXqECxROiyP1SEAcJOgHPw4l6YTf7bs3GtiGmd53/Pnz6tOnTq655579Mwzz6hIkSLav3+/ypcvr/Ll/7oF9ptvvqnY2FjNnj1bZcuW1fDhw7Vz507t2bNHQUFBWTqP5QWDO1Aw+BYKBt9CweBbKBgA75WTC4Zmb1tXMKwekPWC4eWXX9bPP/+stWvXZvq63W5XyZIl9eKLL2rgwIGSpKSkJBUrVkyzZs3618cXXMeD2wAAAIAcIi0tTRcuXDBsf3/m2HXffPON6tWrp0ceeURFixZVnTp19NFHHzleP3z4sE6dOqXmzZs7xkJCQtSgQQPFxcVldshMUTAAAAAAOURmzxiLjc38GWOHDh1yrEdYtmyZnnnmGb3wwguaPXu2JOnUqVOSpGLFihm+r1ixYo7XsiIHN4MAAAAAz7Ny0XNmzxgze2RARkaG6tWrp7Fjx0qS6tSpo127dmnatGmKiorK9HtuBh0GAAAAIIfIzjPGSpQooapVqxrGqlSpoqP//3bgxYsXlySdPn3asM/p06cdr2UFBQMAAADgxGazbsuOxo0ba9++fYax33//XeHh4ZKksmXLqnjx4lq5cqXj9QsXLmjDhg2KjIzM8nmYkgQAAADcggYMGKBGjRpp7Nix6ty5s3799Vd9+OGH+vDDDyX9NbWqf//+ev3111WxYkXHbVVLliypdu3aZfk8FAwAAACAk1vlwW3169fXwoULNWTIEI0ePVply5bVO++8o+7duzv2eemll3T58mU99dRTSkxM1F133aWlS5dm+RkMEs9h8Co8h8G38BwG38JzGAB4m5z8HIZ738v6LUddbdULWZ8q5CmsYQAAAABgKgfXdgAAAIDn+ejkBVN0GAAAAACYosMAAAAAOPHV9ZFm6DAAAAAAMEXBAAAAAMAUU5IAAAAAJ8xIMqLDAAAAAMAUHQYAAADAya3ypGdPocMAAAAAwBQdBgAAAMCJHw0GAzoMAAAAAExRMAAAAAAwxZQkAAAAwAmLno3oMAAAAAAwRYcBAAAAcEKDwYiCwYv48dPtU84vH2Z1CJZo8NpKq0OwxIbh91kdAjzoWrrd6hAskcuff8eAnIgpSQAAAABM0WEAAAAAnNhEt8sZHQYAAAAApugwAAAAAE540rMRHQYAAAAApugwAAAAAE54cJsRHQYAAAAApigYAAAAAJhiShIAAADghBlJRnQYAAAAAJiiwwAAAAA48aPFYECHAQAAAIApCgYAAAAAppiSBAAAADhhRpIRHQYAAAAApugwAAAAAE540rMRHQYAAAAApugwAAAAAE5oMBjRYQAAAABgioIBAAAAgCmmJAEAAABOeNKzER0GAAAAAKboMAAAAABO6C8Y0WEAAAAAYCrbBcPs2bO1ePFix9cvvfSSChYsqEaNGunIkSMuDQ4AAACAtbJdMIwdO1bBwcGSpLi4OE2ePFnjxo1T4cKFNWDAAJcHeCuZP2+uWre4V/Xr1FD3ro9o544dVofkEeRN3t6kaP5Aje1QVasHN9WGV+7Wl882UNWS+Q37lC2cR+92q6l1Q5pp/bC7Nfep+ioeEmhRxO7l7dfbjK/lPWP6B3q8Wyc1aVhXzZs1Uky/aP1x+JDVYXmMr13v63w176yw2WyWbTlRtguGY8eOqUKFCpKkRYsWqWPHjnrqqacUGxurtWvXujzAW8XS75do/LhY9X02WvO/WKjKlSP0TN/eSkhIsDo0tyJv8vamvPMH5dKs3nfoWoZd0Z9uU4dJ6zVh2X5dSLnm2KdUaLBm9a6nw/HJ6jNzszpN2aAPVx/WlWsZFkbuHt5+vc34Yt5bNm3UI10f1axPP9eUD2fo2rVrin66j1KSk60Oze188XpLvps3bk62C4Z8+fI5fpiWL1+uFi1aSJKCgoKUkpLi2uhuIXNmz1SHTp3Vrn1Hla9QQa+MGKWgoCAt+vorq0NzK/Imb2/Ku9dd4Tp9IU2vLtqrXccv6HhiquIOntOf5//vv23P31de6/bH650VB/TbqUv683yKVu+L17nLVy2M3D28/Xqb8cW8J02brofbdlD5ChVVqXKERr0Wq1MnT2jvnt1Wh+Z2vni9Jd/NO6v8bNZtOVG2C4YWLVqoT58+6tOnj37//Xc98MADkqTdu3erTJkyro7vlnD1yhXt3bNbDSMbOcb8/PzUsGEj7di+1cLI3Iu8ydvb8m5WuYh2n7igtzpX14+Dmujzp+9UhztKOl632aQmlcJ0JCFZUx+vrR8HNdGnT9bTPRGFLYzaPXzhemfGV/P+u0uXLkqSCoSEWByJe/nq9fbVvHHzsl0wTJ48WZGRkTp79qy++uorhYWFSZI2b96sbt26uTS4Y8eOqVevXv+4T1pami5cuGDY0tLSXBrHvzmfeF7p6emO9+K6sLAwxcfHezQWTyJv8pa8K+9SoUHqXO82HU1I0TNztmrBxj81uHUltalVXJJUKG+A8gbmUq+7yujnAwl6es5Wrdp7VhO71NQd4QWtDd7FfOF6Z8ZX83aWkZGh8ePGqladuqpQsZLV4biVr15vX807O1jDYJTt5zAULFhQkyZNumF81KhRLgnI2blz5zR79mzNmDHDdJ/Y2Ngbzj1s+Ai98upIl8cDwLv52WzafeKC3l95UJL026lLqlA0nx6pX0rfbj/laBX/+NtZfRp3TJK079Ql1bo9RI/Uv02bjyRaFDngOm+MGa2DB/br41nzrA4FQA6RpYJhRzZWzdesWTPL+37zzTf/+PqhQ/9+h4YhQ4YoJibGMGb39+zdSkILhsrf3/+GhUIJCQkqXNj7pipcR97kLXlX3mcvpenQ2cuGsUPxl9W8ahFJ0vnkq7qannHDPofjL6t26YKeCtMjfOF6Z8ZX877uzbGjtW7NT/po5qcqVry41eG4na9eb1/NGzcvSwVD7dq1ZbPZZLfbM339+ms2m03p6elZPnm7du3+8bjXj/1PAgMDFRhoLBBSr5ns7Ca5AwJUpWo1bVgfp3vvay7pr5buhg1x6trtMc8G40HkTd7elve2o0kqUzivYSw8LI9OJKZKkq6l27X7+AWVKZznhn1O/v99vIUvXO/M+Gredrtd42Jf04+rftCHH3+i20qVsjokj/DV6+2reWdHDp0ZZJksFQyHDx92y8lLlCihKVOmqG3btpm+vm3bNt1xxx1uOberPR7VU8OHDla1atVVvUZNfTpntlJSUtSufQerQ3Mr8iZvb8r707ijmt2nnno3Cdfy3WdU/bYC6nTHbRr9zV7HPrN/Pqpxj1TX5iOJ2nj4vBpXCFPTSoXVZ9YWCyN3D2+/3mZ8Me83xozW0u+/08R3JytP3ryKjz8rScqXL7+CgoIsjs69fPF6S76bN25OlgqG8PBwt5z8jjvu0ObNm00Lhn/rPuQk97d+QOfPndOUSe8pPv6sKkdU0ZQPpivMy1t75E3e3pT37hMXFTN/h15oXkF9m5XV8cRUjVv6u5bsPO3YZ9VvZ/X6d7+pV5MyGty6kv6IT9aLn+/U1qNJFkbuHt5+vc34Yt5fLvhMkvRUrycM4yNeG6uH23r3H5C+eL0l3807q3Lq4mOr2Ow38Rf5nDlzNG3aNB0+fFhxcXEKDw/XO++8o7Jly5r+8Z+ZtWvX6vLly7r//vszff3y5cvatGmTmjVrlq34PD0lCYDnNHhtpdUhWGLD8PusDgEedC391viwzNVy+fNHmi8JyvatdzzniXnWPfX6k0ezvh7YU7J9W9WpU6cqJiZGDzzwgBITEx1rFgoWLKh33nknW8dq0qSJabEgSXnz5s12sQAAAADAdbJdMLz//vv66KOPNGzYMPn7+zvG69Wrp507d7o0OAAAAMDTeNKzUbYLhsOHD6tOnTo3jAcGBury5cuZfAcAAACAW1W2C4ayZctq27ZtN4wvXbpUVapUcUVMAAAAgGV40rNRtpebxMTEKDo6WqmpqbLb7fr111/12WefKTY2VtOnT3dHjAAAAAAsku2CoU+fPgoODtYrr7yi5ORkPfrooypZsqTeffddde3a1R0xAgAAAB6TMz/nt85N3dCqe/fu6t69u5KTk3Xp0iUVLVrU1XEBAAAAyAFu+g64Z86c0b59+yT9Nc+rSJEiLgsKAAAAQM6Q7YLh4sWLevbZZ/XZZ58pIyNDkuTv768uXbpo8uTJCgkJcXmQAAAAgKf45dDFx1bJ9l2S+vTpow0bNmjx4sVKTExUYmKivvvuO23atEl9+/Z1R4wAAAAALJLtDsN3332nZcuW6a677nKMtWrVSh999NE/PrUZAAAAuBXQYDDKdochLCws02lHISEhCg0NdUlQAAAAAHKGbBcMr7zyimJiYnTq1CnH2KlTpzRo0CANHz7cpcEBAAAAsFaWpiTVqVPH8OS5/fv3q3Tp0ipdurQk6ejRowoMDNTZs2dZxwAAAIBbWk594rJVslQwtGvXzs1hAAAAAMiJslQwjBgxwt1xAAAAADkCDQajbK9hAAAAAOA7sn1b1fT0dL399ttasGCBjh49qitXrhheP3funMuCAwAAAGCtbHcYRo0apYkTJ6pLly5KSkpSTEyMOnToID8/P40cOdINIQIAAACe42ezWbblRNkuGObOnauPPvpIL774onLlyqVu3bpp+vTpevXVV7V+/Xp3xAgAAADAItkuGE6dOqUaNWpIkvLly6ekpCRJ0kMPPaTFixe7NjoAAADAw2w267acKNsFQ6lSpXTy5ElJUvny5bV8+XJJ0saNGxUYGOja6AAAAABYKtsFQ/v27bVy5UpJ0vPPP6/hw4erYsWKeuKJJ9SrVy+XBwgAAAB4ks1ms2zLibJ9l6Q33njD8f+7dOmi8PBw/fLLL6pYsaLatGnj0uAAAAAAWOs/P4ehYcOGiomJUYMGDTR27FhXxAQAAAAgh7DZ7Xa7Kw60fft21a1bV+np6a443H+Ses3qCAD323Yk0eoQLFE7vKDVIVjimS93Wh2CJaZ2qmF1CADcJCjb81w85/mFey079/vtq1h2bjM86RkAAACAqRxc2wEAAACel1MXH1uFDgMAAAAAU1nuMMTExPzj62fPnv3PwQAAAADIWbJcMGzduvVf92natOl/CgYAAACwmh8zkgyyXDD8+OOP7owDAAAAQA7EomcAAADACR0GIxY9AwAAADBFhwEAAABwwm1VjegwAAAAADBFwQAAAADA1E0VDGvXrtVjjz2myMhIHT9+XJI0Z84crVu3zqXBAQAAAJ7mZ7Nuy4myXTB89dVXatWqlYKDg7V161alpaVJkpKSkjR27FiXBwgAAADAOtkuGF5//XVNmzZNH330kXLnzu0Yb9y4sbZs2eLS4AAAAABPs9ms23KibBcM+/bty/SJziEhIUpMTHRFTAAAAAByiGwXDMWLF9eBAwduGF+3bp3KlSvnkqAAAAAA5AzZfg7Dk08+qX79+mnGjBmy2Ww6ceKE4uLiNHDgQA0fPtwdMQIAAAAe45dT5wZZJNsFw8svv6yMjAzdd999Sk5OVtOmTRUYGKiBAwfq+eefd0eMAAAAACyS7YLBZrNp2LBhGjRokA4cOKBLly6patWqypcvnzviAwAAADyKB5UZZbtguC4gIEBVq1Z1ZSwAAAAAcphsFwz33HOPbP8wr2vVqlX/KSAAAADASixhMMp2wVC7dm3D11evXtW2bdu0a9cuRUVFuSouAAAAADlAtguGt99+O9PxkSNH6tKlS/85IAAAAAA5h8vWdDz22GOaMWOGqw4HAAAAWMLPZrNsy4luetHz38XFxSkoKMhVh7slzZ83V7Nnfqz4+LOqVDlCLw8drho1a1odltuRt3fm/e3ns7T5l5908s8jyh0QqIpVaqhzr+dUolS4Y5+Z78dq99aNSjwXr6CgYFWoWkOdez6nkreXsS5wN/H26922elG1q17MMHbyQqqGLtkvScrlZ1PXOiXUoHSIcvnZtOvUJc3ZdEIX0q5ZEa7befv1NkPe5O0LeSP7st1h6NChg2Fr3769GjZsqJ49e6pv377uiPGWsPT7JRo/LlZ9n43W/C8WqnLlCD3Tt7cSEhKsDs2tyNt78963a6vue6iThk/8WC+NeU/p6df01rAXlJaa4tinTIUI9RkwXLEfzNfA19+V3S699coLykhPtzBy1/OF6y1Jfyamqt+ivY5t7A+HHK91q1NCtUvm15Sfj+qNVYdUMDi3nrurtIXRuo+vXO+/I2/y9oW8s8pms27LibJdMISEhBi2QoUK6e6779aSJUs0YsQId8R4S5gze6Y6dOqsdu07qnyFCnplxCgFBQVp0ddfWR2aW5G39+Y98LV31aTFQyoVXk6ly1VSn5hXlXD2lA7v/82xzz2t2yuiRh0VKVZSZSpEqOMTfXXu7GmdPXPSwshdzxeutyRl2O26kHrNsV268lfhF5zbT03LhWr+1pPae+ayjpxP1ccb/lTFInlVLizY4qhdz1eu99+RN3n7Qt64OdmakpSenq6ePXuqRo0aCg0NdVdMt5yrV65o757d6v3k/3VY/Pz81LBhI+3YvtXCyNyLvH0r75TLf93UIF/+Apm+npaaorUrvlOR4iUVVrhYpvvcinzpehfLH6iJbSN0Nd2ug/HJ+nLHKZ1LvqoyocHK5e+n3af/78YWpy6mKf7yFVUIy6NDCSn/cNRbiy9db2fkTd6+kDduXrY6DP7+/mrZsqUSExPdFM6t6XzieaWnpyssLMwwHhYWpvj4eIuicj/y9p28MzIyNPeDt1Wxak2VKlPe8NrK777UUx3u1lMd7tbOTXEaNOZ95cqd26JIXc9XrvehhGRN33BME3/6Q3M2HVeRfLk15L5yCsrlp5DgXLqanqGUqxmG77mQek0hwd5zrSXfud5/R97kLXl/3tnhZ7Nuy4myPSWpevXqOnTo0L/vmEWTJk3SE088ofnz50uS5syZo6pVqyoiIkJDhw7VtWv/vKAuLS1NFy5cMGxpaWkuiw+A9MmUt3T8yCE9+/LrN7wWec/9Gv3+Jxry5jQVu620JscO1ZUr/A7eanaevKRNxy7oz6RU7Tp1SRNX/6E8uf1Vv3SI1aEBACyW7YLh9ddf18CBA/Xdd9/p5MmTN/yxnt1jDR06VMnJyRowYIDefPNNDRgwQN27d1dUVJSmT5+u11577R+PERsbe8O6irfejM1uWv9JaMFQ+fv737BQKCEhQYULF/ZoLJ5E3r6R9ydT3tL2X9fp5TemqFAmU43y5M2n4reVVkSNOnp+aKxOHjuizb/85PlA3cTXrvd1KVczdPpimorlC1BSyjXl9vdTcG7jPxkFgnIpKeWqRRG6h69eb/Imb8n7884ObqtqlOWCYfTo0bp8+bIeeOABbd++XQ8//LBKlSql0NBQhYaGqmDBgtle1zBr1izNmjVLX375pZYuXaphw4bp3Xff1bBhwzRkyBB98MEHmjdv3j8eY8iQIUpKSjJsgwYPyVYc/1XugABVqVpNG9bHOcYyMjK0YUOcataq49FYPIm8vTtvu92uT6a8pc1xqzU4drKKFC/5798juyS7rl31nj8ifeV6/11gLj8VyRegxJRr+uN8iq6lZ6hqsXyO14vnD1DhvAE6kJBsYZSu56vXm7zJ2xfyxs3L8qLnUaNG6emnn9aPP/7ospOfOHFC9erVkyTVqlVLfn5+ql27tuP1unXr6sSJE/94jMDAQAUGBhrGUi24LfjjUT01fOhgVatWXdVr1NSnc2YrJSVF7dp38HwwHkTe3pv3J1Pe0vqflqnfq28pKDivEs/99UlUnrx5FRAYpDMnj2vDmhWqXreBCoSE6lz8GX33xSfKHRCoWvUbWRy9a/nC9e5Su7i2Hb+o+OQrCg3KrXY1ispulzYcTVTK1QytOXReXeuU0OUr6Uq5mq7H7iipA/GXvWrB83W+cL0zQ97k7Qt5Z1UO/aDfMlkuGOx2uySpWbNmLjt58eLFtWfPHpUuXVr79+9Xenq69uzZo2rVqkmSdu/eraJFi7rsfO50f+sHdP7cOU2Z9J7i48+qckQVTflgusK8vLVH3t6b96rFf91aL3bwM4bxPgOGq0mLh5Q7IEC/796m5f+br8uXLiqkYCFVrl5HwydMV4GChawI2W184XqHBudW30a3K1+Avy6mpWv/2ct67YeDupj2161VP9t6UnZJ0Y1LK7e/n3advKhPNv/zBzq3Kl+43pkhb/L2hbxxc2z265XAv/Dz89Pp06dVpEgRl518+PDh+uCDD9S2bVutXLlSXbp00bx58zRkyBDZbDaNGTNGnTp10sSJE7N1XCs6DICnbTuSaHUIlqgdXtDqECzxzJc7rQ7BElM71bA6BABuEpStm/t71ms/HLDs3MObV7Ds3GaydakqVaok27/0aM6dO5fl440aNUrBwcGKi4vTk08+qZdfflm1atXSSy+9pOTkZLVp0+ZfFz0DAAAArpRTb29qlWwVDKNGjVJIiOtusefn56ehQ4caxrp27aquXbu67BwAAAAAbl62CoauXbveMmsKAAAAgJthEy0GZ1m+req/TUUCAAAA4H2yXDBkcW00AAAAAC+S5SlJGRkZ7owDAAAAyBFY9GyU5Q4DAAAAgJzpjTfekM1mU//+/R1jqampio6OVlhYmPLly6eOHTvq9OnT2T42BQMAAADgxM9m3XYzNm7cqA8++EA1a9Y0jA8YMEDffvutvvjiC61evVonTpxQhw7Zf5o3BQMAAABwi7p06ZK6d++ujz76SKGhoY7xpKQkffzxx5o4caLuvfde3XHHHZo5c6Z++eUXrV+/PlvnoGAAAAAAnNhsNsu2tLQ0XbhwwbClpaWZxhodHa0HH3xQzZs3N4xv3rxZV69eNYxHRESodOnSiouLy9b7QcEAAAAA5BCxsbEKCQkxbLGxsZnuO3/+fG3ZsiXT10+dOqWAgAAVLFjQMF6sWDGdOnUqWzFl68FtAAAAANxnyJAhiomJMYwFBgbesN+xY8fUr18/rVixQkFBQW6NiYIBAAAAcGLlbVUDAwMzLRD+bvPmzTpz5ozq1q3rGEtPT9eaNWs0adIkLVu2TFeuXFFiYqKhy3D69GkVL148WzFRMAAAAAC3mPvuu087d+40jPXs2VMREREaPHiwbr/9duXOnVsrV65Ux44dJUn79u3T0aNHFRkZma1zUTAAAAAATmy3wIPb8ufPr+rVqxvG8ubNq7CwMMd47969FRMTo0KFCqlAgQJ6/vnnFRkZqYYNG2brXBQMAAAAgBd6++235efnp44dOyotLU2tWrXSlClTsn0cCgYAAADAC/z000+Gr4OCgjR58mRNnjz5Px2XggEAAABw4ncrzEnyIJ7DAAAAAMAUHQYAAADAiZW3Vc2J6DAAAAAAMEWHAQAAAHDCEgYjOgwAAAAATFEwAAAAADDFlCQAAADAiZ+Yk+SMggG4RdUOL2h1CPCgqZ1qWB2CJUr2nGd1CJbYN7Wz1SFYIn8Qf5YAORG/mQAAAIATFj0bsYYBAAAAgCkKBgAAAACmmJIEAAAAOOFJz0Z0GAAAAACYosMAAAAAOPFj1bMBHQYAAAAApigYAAAAAJhiShIAAADghBlJRnQYAAAAAJiiwwAAAAA4YdGzER0GAAAAAKboMAAAAABOaDAY0WEAAAAAYIqCAQAAAIAppiQBAAAATvhE3Yj3AwAAAIApOgwAAACAExurng3oMAAAAAAwRcEAAAAAwBRTkgAAAAAnTEgyosMAAAAAwBQdBgAAAMCJH4ueDegwAAAAADBFhwEAAABwQn/BiA6DC82fN1etW9yr+nVqqHvXR7Rzxw6rQ/II8iZvX0De3pn3tokP69ycR2/YxkXVkyQVDQnS1L6R2vt+ex2b3lk/vna/2tS73eKo3ePsmdMa/cpgPXBvI93bqK6e6NxOv+3ZZXVYHuHtP+dmfDVvZB8Fg4ss/X6Jxo+LVd9nozX/i4WqXDlCz/TtrYSEBKtDcyvyJm/y9l6+kPd9I5Yp4rmvHVv7N1ZKkv634agkaWrfSFUoUUDd316ju4Ys1nebjmnG841VIzzUyrBd7sKFJD3T6zHlypVL49+bpk+/+EbPDRik/PkLWB2a2/nCz3lmfDVv3JwcVzDY7XarQ7gpc2bPVIdOndWufUeVr1BBr4wYpaCgIC36+iurQ3Mr8iZv8vZevpB3wsU0nUlKdWytat+mQ6cv6uffzkiS6lcsrI9W7NOWQwk6cvayJvxvt5IuX1XtMoUsjty15s76WEWLFdfQkWNUtXpNlbytlO6MbKzbbi9tdWhu5ws/55nx1byzymazbsuJclzBEBgYqL1791odRrZcvXJFe/fsVsPIRo4xPz8/NWzYSDu2b7UwMvcib/Imb/L2Jrn9/fRI4zKau/qgY2zj/ni1bxCugnkDZLNJHRqGKzDAX+v2nrYwUtf7ec2PiqhaTa+8NEAPNW+ino921Ddff2F1WG7niz/nku/mjZtn2aLnmJiYTMfT09P1xhtvKCwsTJI0ceLEfzxOWlqa0tLSDGN2/0AFBga6JtAsOJ94Xunp6Y6YrwsLC9Phw4c8FoenkTd5S+TtrXwx7wfvKKWQPAH6bO1hx1jPSes0I/ouHZrWSVevZSjlyjU98c4aHT5zycJIXe/E8T+16MvP1aV7lJ7o9ZT27tmpd8bHKnfu3Grdpp3V4bmNL/6cS76bd3bYcupH/RaxrGB45513VKtWLRUsWNAwbrfbtXfvXuXNmzdLFys2NlajRo0yjA0bPkKvvDrShdECALzdY83K64cdJ3UqMcUxNrRjTYXkza12sSuVcClND95RSjOeu0sPvL5Ce/9MsjBa18rIyFBE1erq+1x/SVKliCo6fOCAFn21wKsLBgBZY1nBMHbsWH344YeaMGGC7r33Xsd47ty5NWvWLFWtWjVLxxkyZMgN3Qq7v+e6C5IUWjBU/v7+NywUSkhIUOHChT0aiyeRN3lL5O2tfC3vUmF51Kx6MT3x7lrHWJmi+fRUy8pq9PJi/Xb8r+Jg99FENaxUVH2aV9KLszZaFa7LhRUuojJlyxvGwsuW00+rVlgUkWf42s/5db6aN26eZWsYXn75ZX3++ed65plnNHDgQF29evWmjhMYGKgCBQoYNk9OR5Kk3AEBqlK1mjasj3OMZWRkaMOGONWsVcejsXgSeZM3eZO3t+jetLzOXkjT8m0nHGPBAf6SpIy/3YwjI8MuPz/vmq5Qo1YdHT1y2DB27OgfKl6ipEUReYav/Zxf56t5Z4efhVtOZGlc9evX1+bNm3X27FnVq1dPu3btumXnjD0e1VNff7lA3yxaqEMHD+r10SOVkpKidu07WB2aW5E3eZO39/KVvG026dGm5TR/7SGlZ/xfcbD/5AUdPHVRE3veqbrlwlSmaD5Ft47Q3dWLa/HmPy2M2PW6dH9Cu3fu0CczPtSfx45o+fff6Zuvv1SHR7pZHZrb+crP+d/5at64OZY/6TlfvnyaPXu25s+fr+bNmys9Pd3qkG7K/a0f0Plz5zRl0nuKjz+ryhFVNOWD6Qrz8tYeeZM3eXsvX8n77mrFdXvhvJq7xrjY81q6XV3G/6QRXWppXkxT5Q3KrcOnL+rZD+P0w/YTJke7NVWpVkNjx7+rDya9o1kfTVWJkqX0wouD1fKBh6wOze185ef873w176y6VT/AdhebPQc9+ODPP//U5s2b1bx5c+XNm/emj5N6zYVBAQAsU7LnPKtDsMS+qZ2tDsES+YMs/xwTHpSTL/eCbdZ9KNC5ds6bCpijLlWpUqVUqlQpq8MAAACAD6O/YJRT11YAAAAAyAEoGAAAAACYylFTkgAAAACrsejZiA4DAAAAAFN0GAAAAAAnfKJuxPsBAAAAwBQFAwAAAABTTEkCAAAAnLDo2YgOAwAAAABTdBgAAAAAJ/QXjOgwAAAAADBFhwEAAABwwhIGIzoMAAAAAExRMAAAAAAwxZQkAAAAwIkfy54N6DAAAAAAMEWHAQAAAHDComcjOgwAAAAATFEwAAAAADDFlCQAAADAiY1FzwZ0GAAAAACYosMAAAAAOGHRsxEdBgAAAACm6DAAAAAATnhwmxEFA255GXa71SFYws9H+6VJyVetDsESIXlyWx2CJQ592MXqECxR4ZkvrA7BEn9O72p1CAAywZQkAAAAAKboMAAAAABOfLSJb4oOAwAAAABTdBgAAAAAJ3QYjOgwAAAAADBFwQAAAADAFFOSAAAAACc2nsNgQIcBAAAAgCk6DAAAAIATPxoMBnQYAAAAAJiiwwAAAAA4YQ2DER0GAAAAAKYoGAAAAACYYkoSAAAA4IQnPRvRYQAAAABgig4DAAAA4IRFz0Z0GAAAAACYomAAAAAAYIopSQAAAIATnvRsRIcBAAAAgCk6DAAAAIATFj0b0WEAAAAAYIqCAQAAAIAppiQBAAAATnjSsxEFgwvNnzdXs2d+rPj4s6pUOUIvDx2uGjVrWh2W2/la3ps3bdQnMz/Wnj27FX/2rCa+O0n33Nfc6rA8xteud+eHW+rUyRM3jLfr1FUxg1+xICLP8rXr/XezZ3ykKe+9rS6PPq6Yl4ZYHY7LbBnfRqUL571h/OOV+zXp+9+0dXybTL+v1+Sf9c3GY+4Oz+N89efcV/NG9jElyUWWfr9E48fFqu+z0Zr/xUJVrhyhZ/r2VkJCgtWhuZUv5p2SkqJKlSM0ZNirVoficb54vT+cPV8Lv//JsU2c9JEk6Z7mLS2OzP188Xo727NrpxZ+uUAVKlW2OhSXazFquar2W+TYOoz7UZL0zcZjOp6QbHitar9FeuPrnbqUclUrd5y0OHLX89Wfc1/NO6tsFm45EQWDi8yZPVMdOnVWu/YdVb5CBb0yYpSCgoK06OuvrA7NrXwx77uaNFX0C/11b/MWVoficb54vQuGFlJY4cKO7Zd1q3VbqdtVu259q0NzO1+83tclJ1/Wq0Nf0tBXR6lA/gJWh+NyCRfTdCYp1bG1rF1Sh05f1M+/nVGG3W547UxSqh64o5QWbTymy2nXrA7d5Xz159xX88bNoWBwgatXrmjvnt1qGNnIMebn56eGDRtpx/atFkbmXr6at6/iektXr17Viu+/0wMPt5fNyye4+vr1fmvs62rcpJnubNjo33e+xeX299MjkWU0b+3hTF+vFR6qmuGhmrvmoIcjcz9f/Tn31byzw89ms2zLiSgYXOB84nmlp6crLCzMMB4WFqb4+HiLonI/X83bV3G9pbU/rdSlSxfV+qF2Vofidr58vZcvXaJ9v+3Rsy8MsDoUj3ig7m0KyZNb89cdyvT17k3Lad/xJG084H1TVXz159xX88bNs3zR8969e7V+/XpFRkYqIiJCv/32m959912lpaXpscce07333vuP35+Wlqa0tDTDmN0/UIGBge4MG4APWvzN12oQeZcKFylqdShwk9OnTmriuFi9P226z/w70r1pOa3ceVKnElNveC0ot786RoZrwje7LYgMQE5haYdh6dKlql27tgYOHKg6depo6dKlatq0qQ4cOKAjR46oZcuWWrVq1T8eIzY2ViEhIYbtrTdjPZTBX0ILhsrf3/+GhUIJCQkqXLiwR2PxJF/N21f5+vU+dfKENv+6Xg+262h1KB7hq9f7tz27df5cgqK6dVKjO2qo0R01tGXzRi347FM1uqOG0tPTrQ7RpUqF5VGzasX06erMuwtt6t+u4AB/ff7zH54NzEN89efcV/PODhY9G1laMIwePVqDBg1SQkKCZs6cqUcffVRPPvmkVqxYoZUrV2rQoEF64403/vEYQ4YMUVJSkmEbNNizt77LHRCgKlWracP6OMdYRkaGNmyIU81adTwaiyf5at6+ytev95JvF6pgaCFFNm5qdSge4avXu16DSM378n+a8/nXjq1K1epq9cBDmvP51/L397c6RJd6tEk5xV9I0/LtN946WJIea1pOS7eeUMLFtExfv9X56s+5r+aNm2fplKTdu3frk08+kSR17txZjz/+uDp16uR4vXv37po5c+Y/HiMw8MbpR6kW3MTh8aieGj50sKpVq67qNWrq0zmzlZKSonbtO3g+GA/yxbyTky/r2NGjjq+PH/9T+37bqwIhISpRoqSFkbmfL15v6a9/SL//dpHuf7CtcuWyfCanx/ji9c6bN6/KV6hoGAsODlZISMEbxm91NpvU7a6ymv/zYaVn2G94vWzRfIqsVERd315tQXSe44s/55Lv5p1lOfWjfotY/i/f9TuN+Pn5KSgoSCEhIY7X8ufPr6SkJKtCy5b7Wz+g8+fOacqk9xQff1aVI6poygfTFeblrT1fzHvPrl16sleU4+sJ4/7qgrVp206jx/xzR+xW54vXW5I2/Rqn06dO6sGH21sdikf56vX2Fc2qFtfthfNq3prM7470aJNyOnE+WT/uOuXhyDzLV3/OfTVv3Byb3W6/8WMFD6lVq5befPNN3X///ZKkXbt2KSIiwvEJ3tq1axUVFaVDhzKfW2nGig4DrJNh3Y+wpXLqrdfcLSn5qtUhWCIkT26rQ7BE6lXvWjOQVRWe+cLqECzx5/SuVocADwqy/GNrc+sPJlp27oblC1p2bjOWXqpnnnnGsICsevXqhte///77f71LEgAAAOBKNuYkGVhaMDz99NP/+PrYsWM9FAkAAACAzOTgZhAAAADgeT4669cUT3oGAAAAYIoOAwAAAOCEBoMRHQYAAAAApigYAAAAAJhiShIAAADgjDlJBnQYAAAAAJiiwwAAAAA44cFtRnQYAAAAAJiiYAAAAABgiilJAAAAgBOe9GxEhwEAAACAKQoGAAAAwInNwi07YmNjVb9+feXPn19FixZVu3bttG/fPsM+qampio6OVlhYmPLly6eOHTvq9OnT2ToPBQMAAABwC1q9erWio6O1fv16rVixQlevXlXLli11+fJlxz4DBgzQt99+qy+++EKrV6/WiRMn1KFDh2ydx2a32+2uDt5qqdesjgCelOF9P8JZ4uejEyyTkq9aHYIlQvLktjoES6ReTbc6BEtUeOYLq0OwxJ/Tu1odAjwoKAevpN1y5IJl564bXuCmv/fs2bMqWrSoVq9eraZNmyopKUlFihTRvHnz1KlTJ0nSb7/9pipVqiguLk4NGzbM0nHpMAAAAAA5RFpami5cuGDY0tLSsvS9SUlJkqRChQpJkjZv3qyrV6+qefPmjn0iIiJUunRpxcXFZTkmCgYAAAAgh4iNjVVISIhhi42N/dfvy8jIUP/+/dW4cWNVr15dknTq1CkFBASoYMGChn2LFSumU6dOZTmmHNwMAgAAADzPyic9DxkyRDExMYaxwMDAf/2+6Oho7dq1S+vWrXN5TBQMAAAAQA4RGBiYpQLB2XPPPafvvvtOa9asUalSpRzjxYsX15UrV5SYmGjoMpw+fVrFixfP8vGZkgQAAAA4sdms27LDbrfrueee08KFC7Vq1SqVLVvW8Podd9yh3Llza+XKlY6xffv26ejRo4qMjMzyeegwAAAAALeg6OhozZs3T//73/+UP39+x7qEkJAQBQcHKyQkRL1791ZMTIwKFSqkAgUK6Pnnn1dkZGSW75AkUTAAAAAAt6SpU6dKku6++27D+MyZM9WjRw9J0ttvvy0/Pz917NhRaWlpatWqlaZMmZKt81AwAAAAAE5ulScdZeVxakFBQZo8ebImT5580+dhDQMAAAAAUzzpGQAA5AjN31lrdQiW+KF/E6tDsEROftLz9mMXLTt3rdvzW3ZuM3QYAAAAAJjKwbUdAAAA4HlWPrgtJ6LDAAAAAMAUBQMAAAAAU0xJAgAAAJxk94nL3o4OAwAAAABTdBgAAAAAJzQYjOgwAAAAADBFwQAAAADAFFOSAAAAAGfMSTKgwwAAAADAFB0GAAAAwAlPejaiwwAAAADAFB0GAAAAwAkPbjOiwwAAAADAFAUDAAAAAFNMSQIAAACcMCPJiA4DAAAAAFN0GAAAAABntBgM6DAAAAAAMEXBAAAAAMAUU5IAAAAAJzzp2YgOAwAAAABTdBgAAAAAJzzp2YgOgwvNnzdXrVvcq/p1aqh710e0c8cOq0PyCPImb19A3uTtC7w978L5AjT8gcpaHN1QK/s10uyouqpcLJ/j9aYVwzSxU3Utjm6odQObqEKRvBZG637efr3hOhQMLrL0+yUaPy5WfZ+N1vwvFqpy5Qg907e3EhISrA7NrcibvMnbe5E3eXtT3vkDc2lqt1q6lpGhgV/t0mOzNmvST4d0MfWaY5/g3P7acfyCpq45bGGknuHt1/u/slm45UQUDC4yZ/ZMdejUWe3ad1T5ChX0yohRCgoK0qKvv7I6NLcib/Imb+9F3uTtTXl3v7OUzlxMU+zS/dp76pJOJqVp45FEnUhKdeyzbM8ZzYo7qk1HEq0L1EO8/XrDtSgYXODqlSvau2e3GkY2coz5+fmpYcNG2rF9q4WRuRd5kzd5k7e3IW/vzbtxhTD9duqSXmsToW+fbaAZj9dRmxrFrQ7LEr5wveFalhcMKSkpWrdunfbs2XPDa6mpqfrkk0/+8fvT0tJ04cIFw5aWluaucDN1PvG80tPTFRYWZhgPCwtTfHy8R2PxJPImb4m8vRV5k7fkXXmXDAlSu9oldOx8imK+3KVF20+q/73ldH+1olaH5nG+cL3/M+YkGVhaMPz++++qUqWKmjZtqho1aqhZs2Y6efKk4/WkpCT17NnzH48RGxurkJAQw/bWm7HuDh0AANxC/GzS76cv6cN1R7T/zGV9s+OUvtl5Su1qlbA6NCDHs7RgGDx4sKpXr64zZ85o3759yp8/vxo3bqyjR49m+RhDhgxRUlKSYRs0eIgbo75RaMFQ+fv737BQKCEhQYULF/ZoLJ5E3uQtkbe3Im/ylrwr74TLV/RHQrJh7EhCiorlD7QoIuv4wvX+r2wW/i8nsrRg+OWXXxQbG6vChQurQoUK+vbbb9WqVSs1adJEhw4dytIxAgMDVaBAAcMWGOjZX/7cAQGqUrWaNqyPc4xlZGRow4Y41axVx6OxeBJ5kzd5k7e3IW/vzXvn8QsqXSjYMHZ7aLBOXfDsNOacwBeuN1zL0ge3paSkKFeu/wvBZrNp6tSpeu6559SsWTPNmzfPwuiy5/Gonho+dLCqVauu6jVq6tM5s5WSkqJ27TtYHZpbkTd5k7f3Im/y9qa8P998XNO61dLjDW7Xqn1nVbV4fj1cq7jGLd/v2Cd/UC4Vyx+owvkCJMlRYJy7fEXnkq9aEre7ePv1hmtZWjBERERo06ZNqlKlimF80qRJkqSHH37YirBuyv2tH9D5c+c0ZdJ7io8/q8oRVTTlg+kK8/LWHnmTN3l7L/Imb2/K+7dTlzT0f3vVt0kZ9YgsrZNJqXpv1SGt2HvWsc9d5QtpWOvKjq9Ht/nr75MZvxzRjF+yPl36VuDt1/u/4knPRja73W636uSxsbFau3atlixZkunrzz77rKZNm6aMjIxsHdfpGSwAAOAW0fydtVaHYIkf+jexOgRLBFn6sfU/O3AmxbJzVyga/O87eZilBYO7UDAAAHDroWDwLTm5YDhoYcFQPgcWDJY/hwEAAABAzkXBAAAAAMBUDm4GAQAAABZg0bMBHQYAAAAApugwAAAAAE5y6hOXrUKHAQAAAIApOgwAAACAEx7cZkSHAQAAAIApCgYAAAAAppiSBAAAADhhRpIRHQYAAAAApugwAAAAAM5oMRjQYQAAAABgioIBAAAAgCmmJAEAAABOeNKzER0GAAAAAKboMAAAAABOeNKzER0GAAAAAKboMAAAAABOaDAY0WEAAAAAYIqCAQAAAIAppiQBAAAATlj0bESHAQAAAIApOgwAAACAAS0GZza73W63OghXS71mdQQAAABZE9r6TatDsETKisFWh2Dqz/NXLDt3qdAAy85thilJAAAAAEwxJQkAAABwwqJnIzoMAAAAAEzRYQAAAACc0GAwosMAAAAAwBQdBgAAAMAJaxiM6DAAAAAAMEXBAAAAAMAUU5IAAAAAJzaWPRvQYQAAAABgig4DAAAA4IwGgwEdBgAAAACmKBgAAAAAmGJKEgAAAOCEGUlGdBgAAAAAmKLDAAAAADjhSc9GdBgAAAAAmKLDAAAAADjhwW1GdBgAAAAAmKJgAAAAAGCKKUkAAACAM2YkGdBhAAAAAGCKDgMAAADghAaDER0GAAAAAKYoGAAAAACYomBwofnz5qp1i3tVv04Nde/6iHbu2GF1SB5B3uTtC8ibvH0BeXtf3n5+Nr0a1UR7P+mrc9/FaPfsp/Ry90aGfVJWDM50G/DInRZFbT2bzbotJ6JgcJGl3y/R+HGx6vtstOZ/sVCVK0fomb69lZCQYHVobkXe5E3e3ou8yZu8b30vdmmgJ9vU1oBJK1S793S9Mn21YjrfqWfb3eHYp0znSYbtqfFLlJFh18K1+yyMHDkJBYOLzJk9Ux06dVa79h1VvkIFvTJilIKCgrTo66+sDs2tyJu8ydt7kTd5k/etr2HV2/TdLwe09NdDOnr6ghau3aeVm/9QvcolHPucPn/ZsLWJrKDV24/oj1NJFkZuLZuF/8uJKBhc4OqVK9q7Z7caRv5fi8/Pz08NGzbSju1bLYzMvcibvMmbvL0NeZO3t+W9fs9x3VMnXBVuC5Uk1ShXRJHVS2n5xkOZ7l+0YB7d36C8Zn/vPdOy8N/lqNuqXr58WQsWLNCBAwdUokQJdevWTWFhYf/4PWlpaUpLSzOM2f0DFRgY6M5QDc4nnld6evoNsYaFhenw4cx/Ib0BeZO3RN7eirzJWyJvbzB+/noVyBOo7TOeVHpGhvz9/DRi5hrNX7Un0/0fa1ldF5OvaNG63z0cac6SU9cSWMXSDkPVqlV17tw5SdKxY8dUvXp1DRgwQCtWrNCIESNUtWpVHT58+B+PERsbq5CQEMP21puxnggfAAAgR+vUrIq63ltVPWK/VeQzs9TnrcXq/8id6t6ieqb7P9Gqpj5ftUdpV9M9HClyMks7DL/99puuXbsmSRoyZIhKliypbdu2KSQkRJcuXVL79u01bNgwzZs3z/QYQ4YMUUxMjGHM7u+57oIkhRYMlb+//w0LpBISElS4cGGPxuJJ5E3eEnl7K/Imb4m8vcHYJ+/W+M/X64uf9kqSdv8Rr9JFC2hQ14aau2KXYd/G1UupcukwPT7mf1aEihwsx6xhiIuL08iRIxUSEiJJypcvn0aNGqV169b94/cFBgaqQIEChs2T05EkKXdAgKpUraYN6+McYxkZGdqwIU41a9XxaCyeRN7kTd7k7W3Im7y9Le/goNzKyLAbxtIz7PLzu3HOTVTrmtr8+0ntPHTWU+HhFmH5Ggbb/58klpqaqhIlShheu+2223T27K3xQ/t4VE8NHzpY1apVV/UaNfXpnNlKSUlRu/YdrA7NrcibvMnbe5E3eZP3rW/J+gMa/GgjHTtzQXuOxKt2hWJ6oWN9fbLMuKg5f54AdWhSWS9/+KNFkSIns7xguO+++5QrVy5duHBB+/btU/Xq/zen7siRI/+66DmnuL/1Azp/7pymTHpP8fFnVTmiiqZ8MF1hXtLSNEPe5E3e3ou8yZu8b30xk37QiB5N9O4LLVWkYB6dTLikjxdv09hPfzbs98jdVWSz2bTAZDG0r2HRs5HNbrfb/3039xg1apTh64YNG6pVq1aOrwcNGqQ///xTn332WbaOm3rNJeEBAAC4XWjrN60OwRIpKwZbHYKpxBTrFn0XDPa37NxmLC0Y3IWCAQAA3CooGHIeCgYjy6ckAQAAADlJTn3islVyzF2SAAAAAOQ8dBgAAAAAJyx6NqLDAAAAAMAUHQYAAADACQ0GIzoMAAAAAExRMAAAAAAwxZQkAAAAwBlzkgzoMAAAAAAwRYcBAAAAcMKD24zoMAAAAAAwRcEAAAAAwBRTkgAAAAAnPOnZiA4DAAAAAFN0GAAAAAAnNBiM6DAAAAAAMEXBAAAAAMAUU5IAAAAAZ8xJMqDDAAAAAMAUHQYAAADACU96NqLDAAAAANyiJk+erDJlyigoKEgNGjTQr7/+6vJzUDAAAAAATmw267bs+PzzzxUTE6MRI0Zoy5YtqlWrllq1aqUzZ8649P2gYAAAAABuQRMnTtSTTz6pnj17qmrVqpo2bZry5MmjGTNmuPQ8FAwAAABADpGWlqYLFy4YtrS0tBv2u3LlijZv3qzmzZs7xvz8/NS8eXPFxcW5Nig7XCY1NdU+YsQIe2pqqtWheBR5k7cvIG/y9gXkTd6w3ogRI+ySDNuIESNu2O/48eN2SfZffvnFMD5o0CD7nXfe6dKYbHa73e7aEsR3XbhwQSEhIUpKSlKBAgWsDsdjyJu8fQF5k7cvIG/yhvXS0tJu6CgEBgYqMDDQMHbixAnddttt+uWXXxQZGekYf+mll7R69Wpt2LDBZTFxW1UAAAAgh8isOMhM4cKF5e/vr9OnTxvGT58+reLFi7s0JtYwAAAAALeYgIAA3XHHHVq5cqVjLCMjQytXrjR0HFyBDgMAAABwC4qJiVFUVJTq1aunO++8U++8844uX76snj17uvQ8FAwuFBgYqBEjRmSpjeRNyJu8fQF5k7cvIG/yxq2lS5cuOnv2rF599VWdOnVKtWvX1tKlS1WsWDGXnodFzwAAAABMsYYBAAAAgCkKBgAAAACmKBgAAAAAmKJgAAAAAGCKgsFFLl68qP79+ys8PFzBwcFq1KiRNm7caHVYHjF58mSVKVNGQUFBatCggX799VerQ3KrMmXKyGaz3bBFR0dbHZrbHT9+XI899pjCwsIUHBysGjVqaNOmTVaH5VZr1qxRmzZtVLJkSdlsNi1atMjqkDwiPT1dw4cPV9myZRUcHKzy5cvrtddek6/dJ+ONN96QzWZT//79rQ7FraZOnaqaNWuqQIECKlCggCIjI/X9999bHZbbxcbGqn79+sqfP7+KFi2qdu3aad++fVaH5XYjR4684d+wiIgIq8NCDkbB4CJ9+vTRihUrNGfOHO3cuVMtW7ZU8+bNdfz4catDc6vPP/9cMTExGjFihLZs2aJatWqpVatWOnPmjNWhuc3GjRt18uRJx7ZixQpJ0iOPPGJxZO51/vx5NW7cWLlz59b333+vPXv2aMKECQoNDbU6NLe6fPmyatWqpcmTJ1sdike9+eabmjp1qiZNmqS9e/fqzTff1Lhx4/T+++9bHZrHbNy4UR988IFq1qxpdShuV6pUKb3xxhvavHmzNm3apHvvvVdt27bV7t27rQ7NrVavXq3o6GitX79eK1as0NWrV9WyZUtdvnzZ6tDcrlq1aoZ/y9atW2d1SMjJ7PjPkpOT7f7+/vbvvvvOMF63bl37sGHDLIrKM+688057dHS04+v09HR7yZIl7bGxsRZG5Vn9+vWzly9f3p6RkWF1KG41ePBg+1133WV1GJaSZF+4cKHVYXjEgw8+aO/Vq5dhrEOHDvbu3btbFJFnXbx40V6xYkX7ihUr7M2aNbP369fP6pA8LjQ01D59+nSrw/CoM2fO2CXZV69ebXUobjVixAh7rVq1rA4DtxA6DC5w7do1paenKygoyDAeHBzs1RX7lStXtHnzZjVv3twx5ufnp+bNmysuLs7CyDznypUr+vTTT9WrVy/ZbDarw3Grb775RvXq1dMjjzyiokWLqk6dOvroo4+sDgtu0qhRI61cuVK///67JGn79u1at26dWrdubXFknhEdHa0HH3zQ8N83X5Genq758+fr8uXLioyMtDocj0pKSpIkFSpUyOJI3G///v0qWbKkypUrp+7du+vo0aNWh4QcjCc9u0D+/PkVGRmp1157TVWqVFGxYsX02WefKS4uThUqVLA6PLeJj49Xenr6DU8TLFasmH777TeLovKsRYsWKTExUT169LA6FLc7dOiQpk6dqpiYGA0dOlQbN27UCy+8oICAAEVFRVkdHlzs5Zdf1oULFxQRESF/f3+lp6drzJgx6t69u9Whud38+fO1ZcsWn1mHdt3OnTsVGRmp1NRU5cuXTwsXLlTVqlWtDstjMjIy1L9/fzVu3FjVq1e3Ohy3atCggWbNmqXKlSvr5MmTGjVqlJo0aaJdu3Ypf/78VoeHHIiCwUXmzJmjXr166bbbbpO/v7/q1q2rbt26afPmzVaHBjf6+OOP1bp1a5UsWdLqUNwuIyND9erV09ixYyVJderU0a5duzRt2jQKBi+0YMECzZ07V/PmzVO1atW0bds29e/fXyVLlvTq633s2DH169dPK1asuKFr7O0qV66sbdu2KSkpSV9++aWioqK0evVqnykaoqOjtWvXLq+eGXCdc6ewZs2aatCggcLDw7VgwQL17t3bwsiQU1EwuEj58uW1evVqXb58WRcuXFCJEiXUpUsXlStXzurQ3KZw4cLy9/fX6dOnDeOnT59W8eLFLYrKc44cOaIffvhBX3/9tdWheESJEiVu+MOhSpUq+uqrryyKCO40aNAgvfzyy+rataskqUaNGjpy5IhiY2O9umDYvHmzzpw5o7p16zrG0tPTtWbNGk2aNElpaWny9/e3MEL3CQgIcHTF77jjDm3cuFHvvvuuPvjgA4sjc7/nnntO3333ndasWaNSpUpZHY7HFSxYUJUqVdKBAwesDgU5FGsYXCxv3rwqUaKEzp8/r2XLlqlt27ZWh+Q2AQEBuuOOO7Ry5UrHWEZGhlauXOkT815nzpypokWL6sEHH7Q6FI9o3LjxDbcb/P333xUeHm5RRHCn5ORk+fkZ/4nw9/dXRkaGRRF5xn333aedO3dq27Ztjq1evXrq3r27tm3b5rXFQmYyMjKUlpZmdRhuZbfb9dxzz2nhwoVatWqVypYta3VIlrh06ZIOHjyoEiVKWB0Kcig6DC6ybNky2e12Va5cWQcOHNCgQYMUERGhnj17Wh2aW8XExCgqKkr16tXTnXfeqXfeeUeXL1/2+rwzMjI0c+ZMRUVFKVcu3/g1GjBggBo1aqSxY8eqc+fO+vXXX/Xhhx/qww8/tDo0t7p06ZLhU7fDhw9r27ZtKlSokEqXLm1hZO7Vpk0bjRkzRqVLl1a1atW0detWTZw4Ub169bI6NLfKnz//DfPX8+bNq7CwMK+e1z5kyBC1bt1apUuX1sWLFzVv3jz99NNPWrZsmdWhuVV0dLTmzZun//3vf8qfP79OnTolSQoJCVFwcLDF0bnPwIED1aZNG4WHh+vEiRMaMWKE/P391a1bN6tDQ05l9W2avMXnn39uL1eunD0gIMBevHhxe3R0tD0xMdHqsDzi/ffft5cuXdoeEBBgv/POO+3r16+3OiS3W7ZsmV2Sfd++fVaH4lHffvutvXr16vbAwEB7RESE/cMPP7Q6JLf78ccf7ZJu2KKioqwOza0uXLhg79evn7106dL2oKAge7ly5ezDhg2zp6WlWR2ax/nCbVV79eplDw8PtwcEBNiLFCliv+++++zLly+3Oiy3y+x3W5J95syZVofmVl26dLGXKFHCHhAQYL/tttvsXbp0sR84cMDqsJCD2ex2H3tsJwAAAIAsYw0DAAAAAFMUDAAAAABMUTAAAAAAMEXBAAAAAMAUBQMAAAAAUxQMAAAAAExRMAAAAAAwRcEAAAAAwBQFAwBkU48ePdSuXTvH13fffbf69+/v8Th++ukn2Ww2JSYmuu0cf8/1ZngiTgCA+1AwAPAKPXr0kM1mk81mU0BAgCpUqKDRo0fr2rVrbj/3119/rddeey1L+3r6j+cyZcronXfe8ci5AADeKZfVAQCAq9x///2aOXOm0tLStGTJEkVHRyt37twaMmTIDfteuXJFAQEBLjlvoUKFXHIcAAByIjoMALxGYGCgihcvrvDwcD3zzDNq3ry5vvnmG0n/N7VmzJgxKlmypCpXrixJOnbsmDp37qyCBQuqUKFCatu2rf744w/HMdPT0xUTE6OCBQsqLCxML730kux2u+G8f5+SlJaWpsGDB+v2229XYGCgKlSooI8//lh//PGH7rnnHklSaGiobDabevToIUnKyMhQbGysypYtq+DgYNWqVUtffvml4TxLlixRpUqVFBwcrHvuuccQ581IT09X7969HeesXLmy3n333Uz3HTVqlIoUKaICBQro6aef1pUrVxyvZSV2Z0eOHFGbNm0UGhqqvHnzqlq1alqyZMl/ygUA4D50GAB4reDgYCUkJDi+XrlypQoUKKAVK1ZIkq5evapWrVopMjJSa9euVa5cufT666/r/vvv144dOxQQEKAJEyZo1qxZmjFjhqpUqaIJEyZo4cKFuvfee03P+8QTTyguLk7vvfeeatWqpcOHDys+Pl633367vvrqK3Xs2FH79u1TgQIFFBwcLEmKjY3Vp59+qmnTpqlixYpas2aNHnvsMRUpUkTNmjXTsWPH1KFDB0VHR+upp57Spk2b9OKLL/6n9ycjI0OlSpXSF198obCwMP3yyy966qmnVKJECXXu3NnwvgUFBemnn37SH3/8oZ49eyosLExjxozJUux/Fx0drStXrmjNmjXKmzev9uzZo3z58v2nXAAAbmQHAC8QFRVlb9u2rd1ut9szMjLsK1assAcGBtoHDhzoeL1YsWL2tLQ0x/fMmTPHXrlyZXtGRoZjLC0tzR4cHGxftmyZ3W6320uUKGEfN26c4/WrV6/aS5Uq5TiX3W63N2vWzN6vXz+73W6379u3zy7JvmLFikzj/PHHH+2S7OfPn3eMpaam2vPkyWP/5ZdfDPv27t3b3q1bN7vdbrcPGTLEXrVqVcPrgwcPvuFYfxceHm5/++23TV//u+joaHvHjh0dX0dFRdkLFSpkv3z5smNs6tSp9nz58tnT09OzFPvfc65Ro4Z95MiRWY4JAGAtOgwAvMZ3332nfPny6erVq8rIyNCjjz6qkSNHOl6vUaOGYd3C9u3bdeDAAeXPn99wnNTUVB08eFBJSUk6efKkGjRo4HgtV65cqlev3g3Tkq7btm2b/P39M/1k3cyBAweUnJysFi1aGMavXLmiOnXqSJL27t1riEOSIiMjs3wOM5MnT9aMGTN09OhRpaSk6MqVK6pdu7Zhn1q1ailPnjyG8166dEnHjh3TpUuX/jX2v3vhhRf0zDPPaPny5WrevLk6duyomjVr/udcAADuQcEAwGvcc889mjp1qgICAlSyZEnlymX8T1zevHkNX1+6dEl33HGH5s6de8OxihQpclMxXJ9ilB2XLl2SJC1evFi33Xab4bXAwMCbiiMr5s+fr4EDB2rChAmKjIxU/vz59dZbb2nDhg1ZPsbNxN6nTx+1atVKixcv1vLlyxUbG6sJEybo+eefv/lkAABuQ8EAwGvkzZtXFSpUyPL+devW1eeff66iRYuqQIECme5TokQJbdiwQU2bNpUkXbt2TZs3b1bdunUz3b9GjRrKyMjQ6tWr1bx58xtev97hSE9Pd4xVrVpVgYGBOnr0qGlnokqVKo4F3NetX7/+35P8Bz///LMaNWqkZ5991jF28ODBG/bbvn27UlJSHMXQ+vXrlS9fPt1+++0qVKjQv8aemdtvv11PP/20nn76aQ0ZMkQfffQRBQMA5FDcJQmAz+revbsKFy6stm3bau3atTp8+LB++uknvfDCC/rzzz8lSf369dMbb7yhRYsW6bffftOzzz77j89QKFOmjKKiotSrVy8tWrTIccwFCxZIksLDw2Wz2fTdd9/p7NmzunTpkvLnz6+BAwdqwIABmj17tg4ePKgtW7bo/fff1+zZsyVJTz/9tPbv369BgwZp3759mjdvnmbNmpWlPI8fP65t27YZtvPnz6tixYratGmTli1bpt9//13Dhw/Xxo0bb/j+K1euqHfv3tqzZ4+WLFmiESNG6LnnnpOfn1+WYv+7/v37a9myZTp8+LC2bNmiH3/8UVWqVMlSLgAAz6NgAOCz8uTJozVr1qh06dLq0KGDqlSpot69eys1NdXRcXjxxRf1+OOPKyoqyjFtp3379v943KlTp6pTp0569tlnFRERoSeffFKXL1+WJN12220aNWqUXn75ZRUrVkzPPfecJOm1117T8OHDFRsbqypVquj+++/X4sWLVbZsWUlS6dKl9dVXX2nRokWqVauWpk2bprFjx2Ypz/Hjx6tOnTqGbfHixerbt686dOigLl26qEGDBkpISDB0G6677777VLFiRTVt2lRdunTRww8/bFgb8m+x/116erqio6Md+1aqVElTpkzJUi4AAM+z2c1W7gEAAADweXQYAAAAAJiiYAAAAABgioIBAAAAgCkKBgAAAACmKBgAAAAAmKJgAAAAAGCKggEAAACAKQoGAAAAAKYoGAAAAACYomAAAAAAYIqCAQAAAICp/wfZTQjP6alpEQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create a random forest model\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "# Create a random forest classifier\n",
    "rf = RandomForestClassifier(n_estimators=100, random_state=42)\n",
    "\n",
    "# Train the classifier\n",
    "rf.fit(X_train.reshape(X_train.shape[0], -1), y_train)\n",
    "\n",
    "# Evaluate the classifier\n",
    "train_score = rf.score(X_train.reshape(X_train.shape[0], -1), y_train)\n",
    "val_score = rf.score(X_val.reshape(X_val.shape[0], -1), y_val)\n",
    "test_score = rf.score(X_test.reshape(X_test.shape[0], -1), y_test)\n",
    "\n",
    "print(f\"Train accuracy: {train_score:.4f}\")\n",
    "print(f\"Validation accuracy: {val_score:.4f}\")\n",
    "print(f\"Test accuracy: {test_score:.4f}\")\n",
    "\n",
    "# plot confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Predict labels for the test set\n",
    "print(X_test.reshape(X_test.shape[0], -1).shape)\n",
    "y_pred = rf.predict(X_test.reshape(X_test.shape[0], -1))\n",
    "\n",
    "# Compute the confusion matrix\n",
    "conf_matrix = confusion_matrix(y_test, y_pred)\n",
    "\n",
    "# Plot the confusion matrix\n",
    "plt.figure(figsize=(10, 8))\n",
    "sns.heatmap(conf_matrix, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=labels, yticklabels=labels)\n",
    "plt.title(\"Confusion Matrix\")\n",
    "plt.xlabel(\"Predicted Labels\")\n",
    "plt.ylabel(\"True Labels\")\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['rf_model_custom.joblib']"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# save the rf model\n",
    "import joblib\n",
    "\n",
    "joblib.dump(rf, \"rf_model_custom.joblib\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "\n",
    "class EEGDataset(Dataset):\n",
    "    def __init__(self, data, labels):\n",
    "        \"\"\"\n",
    "        Args:\n",
    "            data: A NumPy array of shape (num_samples, channels, height, width).\n",
    "            labels: A NumPy array of shape (num_samples,) with classification labels.\n",
    "        \"\"\"\n",
    "        self.data = torch.tensor(data, dtype=torch.float32)  # Convert data to tensor\n",
    "        self.labels = torch.tensor(labels, dtype=torch.long)  # Convert labels to tensor\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.data)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        x = self.data[idx]  # Shape: (channels, height, width)\n",
    "        y = self.labels[idx]  # Shape: ()\n",
    "        return x, y\n",
    "\n",
    "# Create PyTorch Datasets\n",
    "train_dataset = EEGDataset(X_train, y_train)\n",
    "val_dataset = EEGDataset(X_val, y_val)\n",
    "test_dataset = EEGDataset(X_test, y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "EEGCNN(\n",
      "  (conv1): Conv2d(14, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (pool1): MaxPool2d(kernel_size=(2, 1), stride=(2, 1), padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv2): Conv2d(16, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (pool2): MaxPool2d(kernel_size=(2, 1), stride=(2, 1), padding=0, dilation=1, ceil_mode=False)\n",
      "  (conv3): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (pool3): MaxPool2d(kernel_size=(2, 1), stride=(2, 1), padding=0, dilation=1, ceil_mode=False)\n",
      "  (fc1): Linear(in_features=3072, out_features=128, bias=True)\n",
      "  (fc2): Linear(in_features=128, out_features=10, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "class EEGCNN(nn.Module):\n",
    "    def __init__(self, input_channels, num_classes):\n",
    "        super(EEGCNN, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(input_channels, 16, kernel_size=3, stride=1, padding=(1, 1))  # Conv1\n",
    "        self.pool1 = nn.MaxPool2d(kernel_size=(2, 1))  # Pool only height\n",
    "\n",
    "        self.conv2 = nn.Conv2d(16, 32, kernel_size=3, stride=1, padding=(1, 1))  # Conv2\n",
    "        self.pool2 = nn.MaxPool2d(kernel_size=(2, 1))  # Pool only height\n",
    "\n",
    "        self.conv3 = nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=(1, 1))  # Conv3\n",
    "        self.pool3 = nn.MaxPool2d(kernel_size=(2, 1))  # Pool only height\n",
    "\n",
    "        # Calculate flattened size\n",
    "        self.fc1 = nn.Linear(64 * (129 // 8) * 3, 128)  # Adjust for reduced height and fixed width\n",
    "        self.fc2 = nn.Linear(128, num_classes)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.conv1(x))\n",
    "        x = self.pool1(x)\n",
    "        x = F.relu(self.conv2(x))\n",
    "        x = self.pool2(x)\n",
    "        x = F.relu(self.conv3(x))\n",
    "        x = self.pool3(x)\n",
    "        x = x.view(x.size(0), -1)  # Flatten\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = self.fc2(x)\n",
    "        return x\n",
    "\n",
    "# Define model\n",
    "input_channels = X_train.shape[1]  # Number of channels in the input data\n",
    "num_classes = len(le.classes_)  # Number of classes\n",
    "model = EEGCNN(input_channels, num_classes)\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.optim as optim\n",
    "\n",
    "# Define loss and optimizer\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.001)\n",
    "\n",
    "# Training function\n",
    "def train_model(model, train_loader, val_loader, num_epochs):\n",
    "    model.train()\n",
    "    for epoch in range(num_epochs):\n",
    "        running_loss = 0.0\n",
    "        for inputs, labels in train_loader:\n",
    "            optimizer.zero_grad()\n",
    "            outputs = model(inputs)\n",
    "            loss = criterion(outputs, labels)\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            running_loss += loss.item()\n",
    "        \n",
    "        val_loss, val_acc = validate_model(model, val_loader)\n",
    "        print(f\"Epoch {epoch + 1}/{num_epochs}, Loss: {running_loss:.4f}, Val Loss: {val_loss:.4f}, Val Acc: {val_acc:.4f}\")\n",
    "\n",
    "# Validation function\n",
    "def validate_model(model, val_loader):\n",
    "    model.eval()\n",
    "    val_loss = 0.0\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for inputs, labels in val_loader:\n",
    "            outputs = model(inputs)\n",
    "            loss = criterion(outputs, labels)\n",
    "            val_loss += loss.item()\n",
    "            _, predicted = torch.max(outputs, 1)\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "    return val_loss / len(val_loader), correct / total\n",
    "\n",
    "# Train the model\n",
    "def test_model(model, test_loader):\n",
    "    model.eval()\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    with torch.no_grad():\n",
    "        for inputs, labels in test_loader:\n",
    "            outputs = model(inputs)\n",
    "            _, predicted = torch.max(outputs, 1)\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "    print(f\"Test Accuracy: {correct / total:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10, Loss: 11.4874, Val Loss: 2.3276, Val Acc: 0.1000\n",
      "Epoch 2/10, Loss: 11.4958, Val Loss: 2.3644, Val Acc: 0.0667\n",
      "Epoch 3/10, Loss: 11.4278, Val Loss: 2.3736, Val Acc: 0.0667\n",
      "Epoch 4/10, Loss: 11.4276, Val Loss: 2.3821, Val Acc: 0.0667\n",
      "Epoch 5/10, Loss: 11.4344, Val Loss: 2.3720, Val Acc: 0.0667\n",
      "Epoch 6/10, Loss: 11.3780, Val Loss: 2.3673, Val Acc: 0.0667\n",
      "Epoch 7/10, Loss: 11.4234, Val Loss: 2.3676, Val Acc: 0.0667\n",
      "Epoch 8/10, Loss: 11.3601, Val Loss: 2.3749, Val Acc: 0.0667\n",
      "Epoch 9/10, Loss: 11.4704, Val Loss: 2.3949, Val Acc: 0.0667\n",
      "Epoch 10/10, Loss: 11.4208, Val Loss: 2.3805, Val Acc: 0.0667\n",
      "Test Accuracy: 0.0667\n"
     ]
    }
   ],
   "source": [
    "train_loader = DataLoader(train_dataset, batch_size=32, shuffle=True)\n",
    "val_loader = DataLoader(val_dataset, batch_size=32)\n",
    "test_loader = DataLoader(test_dataset, batch_size=32)\n",
    "\n",
    "train_model(model, train_loader, val_loader, num_epochs=10)\n",
    "test_model(model, test_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x700 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# draw confusion matrix\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "def plot_confusion_matrix(y_true, y_pred, labels):\n",
    "    cm = confusion_matrix(y_true, y_pred)\n",
    "    cm = cm / cm.sum(axis=1)[:, np.newaxis]  # Normalize\n",
    "    df_cm = pd.DataFrame(cm, index=labels, columns=labels)\n",
    "    plt.figure(figsize=(10, 7))\n",
    "    sns.heatmap(df_cm, annot=True, cmap=\"Blues\", fmt=\".2f\")\n",
    "    plt.title(\"Confusion Matrix\")\n",
    "    plt.xlabel(\"Predicted Label\")\n",
    "    plt.ylabel(\"True Label\")\n",
    "    plt.show()\n",
    "\n",
    "# Get predictions\n",
    "y_true = []\n",
    "y_pred = []\n",
    "with torch.no_grad():\n",
    "    for inputs, labels in test_loader:\n",
    "        outputs = model(inputs)\n",
    "        _, predicted = torch.max(outputs, 1)\n",
    "        y_true.extend(labels.tolist())\n",
    "        y_pred.extend(predicted.tolist())\n",
    "\n",
    "# Plot confusion matrix\n",
    "plot_confusion_matrix(y_true, y_pred, le.classes_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}