5-FDI_NN_Basics_FCAutoencoder_Bearing_Signal_Dimensionality_Reduction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "source": [
        "\"\"\"\n",
        "An implementation of Fully Connected Autoencoder architecture for dimensionality\n",
        "reduction for bearing fault pattern dimensionality reduction,\n",
        "for Fault Detection and Identification Course, KNTU, Spring 2023.\n",
        "\n",
        "Programmed by Ardavan Modarres\n",
        "ardavan.modarres@email.kntu.ac.ir\n",
        "* 2023-05-20: Initial coding\n",
        "\"\"\""
      ],
      "metadata": {
        "id": "2-t60Kx9pShd"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8BnLOPI83zD5"
      },
      "outputs": [],
      "source": [
        "# Imports\n",
        "import torch\n",
        "import torch.nn.functional as F                          # Parameterless functions, like activation functions\n",
        "from torch import optim                                  # For optimizers like SGD, Adam, etc.\n",
        "from torch import nn                                     # neural network modules\n",
        "from torch.utils.data import Dataset, DataLoader         # dataset managment by creating mini batches etc.\n",
        "from tqdm import tqdm                                    # For nice progress bar!\n",
        "import matplotlib.pyplot as plt                          # for plotting!\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.manifold import TSNE\n",
        "import seaborn as sn"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Reading data in .csv format using pandas"
      ],
      "metadata": {
        "id": "TVRxilK-j9nL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x = np.array(pd.read_csv('/content/drive/MyDrive/bearingFaultData.csv'))\n",
        "print(x.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7OGbxABE4QFx",
        "outputId": "3bd88fe3-b13a-4b52-bc0e-73ab9290c37b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(959, 12)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## t-sne for visualization"
      ],
      "metadata": {
        "id": "evQMa74JkB7H"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sn.set_theme(style=\"darkgrid\")\n",
        "x_embedded = TSNE(n_components=2, learning_rate='auto', init='random', perplexity=3).fit_transform(x)\n",
        "plt.scatter(x_embedded[:, 0], x_embedded[:, 1])\n",
        "plt.title('2D visualization of raw data')\n",
        "plt.xlabel(\"Dimension 1\")\n",
        "plt.ylabel(\"Dimension 2\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 480
        },
        "id": "U6hUuo764QKx",
        "outputId": "00e9fae3-4b3a-431a-c67b-fb2efad3dcef"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Investigating data before normalization"
      ],
      "metadata": {
        "id": "h2U-Zkk1kNIK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(x.min())\n",
        "print(x.max())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dNFVdwTr4QNH",
        "outputId": "11361238-b573-4be4-8b31-520906fa19f2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "-3806.1\n",
            "3713700.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Normalizing data"
      ],
      "metadata": {
        "id": "liWr-x3qkRLN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x_min = np.amin(x, axis=0)\n",
        "x_max = np.amax(x, axis=0)\n",
        "\n",
        "normalized_x = (x - x_min)/(x_max - x_min)"
      ],
      "metadata": {
        "id": "V4d_hMbT_32c"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(normalized_x.min())\n",
        "print(normalized_x.max())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Azkmg7iSAdm9",
        "outputId": "45b2d75f-f22b-42a7-e6c2-f609122630b1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.0\n",
            "1.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Defining custome dataset class"
      ],
      "metadata": {
        "id": "7pfSyre1kUuu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class AEDataset(Dataset):\n",
        "    def __init__(self, x):\n",
        "        self.x = torch.from_numpy(x).to(torch.float32)\n",
        "\n",
        "    def __getitem__(self, index):\n",
        "        return self.x[index]\n",
        "\n",
        "    def __len__(self):\n",
        "        return self.x.shape[0]\n",
        "\n",
        "# Just for test\n",
        "ds = AEDataset(normalized_x)\n",
        "print(len(ds))\n",
        "print(ds[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1r4fhxBOA12R",
        "outputId": "eafb76e6-22e9-4270-f926-e0095dd3b465"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "959\n",
            "tensor([3.4343e-01, 2.4537e-02, 4.6557e-01, 2.4537e-02, 9.3741e-04, 3.0720e-01,\n",
            "        4.7025e-05, 2.7687e-02, 5.8099e-01, 9.6940e-01, 2.5042e-02, 0.0000e+00])\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "batch_size = 64\n",
        "dataset = AEDataset(normalized_x)\n",
        "train_set, validation_set, test_set = torch.utils.data.random_split(dataset, [0.8, 0.1, 0.1])\n",
        "train_loader = DataLoader(dataset=train_set, batch_size=batch_size, shuffle=True)\n",
        "validation_loader = DataLoader(dataset=validation_set, batch_size=batch_size, shuffle=False)\n",
        "test_loader = DataLoader(dataset=test_set, batch_size=batch_size, shuffle=False)"
      ],
      "metadata": {
        "id": "9CxmamVlBJwX"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# dataiter = iter(data_loader)\n",
        "# features = next(dataiter)\n",
        "# print(torch.min(features), torch.max(features))\n",
        "# print(features.shape)"
      ],
      "metadata": {
        "id": "SHN5E0kuCJMW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Defining autoencoder model"
      ],
      "metadata": {
        "id": "2v2FBFdNlJVv"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![The-schema-of-the-used-Hinton-and-Salakhutdinov-autoencoder.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "FEY9bQFopj-9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "class Autoencoder(nn.Module):\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "        self.encoder = nn.Sequential(\n",
        "            nn.Linear(12, 12),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(12, 8),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(8, 4),\n",
        "        )\n",
        "\n",
        "        self.decoder = nn.Sequential(\n",
        "            nn.Linear(4, 8),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(8, 12),\n",
        "            nn.ReLU(),\n",
        "            nn.Linear(12, 12),\n",
        "        )\n",
        "\n",
        "    def forward(self, x):\n",
        "        encoded = self.encoder(x)\n",
        "        decoded = self.decoder(encoded)\n",
        "        return encoded, decoded"
      ],
      "metadata": {
        "id": "8NUysOUaCJ8g"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = Autoencoder()\n",
        "\n",
        "criterion = nn.MSELoss()\n",
        "optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)"
      ],
      "metadata": {
        "id": "UDMMpnq9DqCx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Error histogram before training"
      ],
      "metadata": {
        "id": "swie8n65EcxN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_samples = torch.zeros(len(train_set), 12)\n",
        "validation_samples = torch.zeros(len(validation_set), 12)\n",
        "test_samples = torch.zeros(len(test_set), 12)\n",
        "\n",
        "train_recon = torch.zeros(len(train_set), 12)\n",
        "validation_recon = torch.zeros(len(validation_set), 12)\n",
        "test_recon = torch.zeros(len(test_set), 12)\n",
        "\n",
        "model.eval()\n",
        "with torch.no_grad():\n",
        "  # Getting reconstructed signals for train set\n",
        "  pointer = 0\n",
        "  for batch in train_loader:\n",
        "      train_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      train_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting reconstructed signals for validation set\n",
        "  pointer = 0\n",
        "  for batch in validation_loader:\n",
        "      validation_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      validation_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting reconstructed signals for test set\n",
        "  pointer = 0\n",
        "  for batch in test_loader:\n",
        "      test_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      test_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "model.train()\n",
        "\n",
        "train_error = torch.sum((train_samples - train_recon)**2, 1)\n",
        "validation_error = torch.sum((validation_samples - validation_recon)**2, 1)\n",
        "test_error = torch.sum((test_samples - test_recon)**2, 1)\n",
        "\n",
        "plt.figure(figsize=(8, 12))\n",
        "#plot 1:\n",
        "plt.subplot(3, 1, 1)\n",
        "plt.hist(train_error, 50)\n",
        "plt.title('Train Error Histogram')\n",
        "#plot 2:\n",
        "plt.subplot(3, 1, 2)\n",
        "plt.hist(validation_error, 50)\n",
        "plt.title('Validation Error Histogram')\n",
        "#plot 3:\n",
        "plt.subplot(3, 1, 3)\n",
        "plt.hist(test_error, 50)\n",
        "plt.title('Test Error Histogram')\n",
        "plt.suptitle(\"Performance\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "3nYYZ9eYEnFw",
        "outputId": "7e7fe65d-8eb0-436d-a1e7-609406fb75d4"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x1200 with 3 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAqoAAARECAYAAAC9GzOiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACRj0lEQVR4nOzde1yUdd7/8fcMgoI4gq7aJmqiCyqgoCZS5oEs81hbtyWptB087HrKcm/JvdMsy7rXbUvxXOl6rHbTu1Qys4O1Zu5mpWmmJpWCt2y3yUFBBWZ+f/RjtgnUARmu78Dr+Xj0qPle37muz8xnoDfXaWwul8slAAAAwDB2qwsAAAAAKkJQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAEmrVq3SoEGD1LlzZ0VHR2vlypVWlwQAdV49qwsAgOjoaI/HdrtdDodD0dHRGj58uIYOHerT7W/ZskVPPvmkOnXqpHvuuUdBQUGKj4/36TYBAJdHUAVgjIkTJ0qSSkpKlJmZqXfeeUe7d+/W/v379cgjj/hsu++9954kacmSJWrRooXPtgMAqByby+VyWV0EgLqtbI/qoUOHPMZ37dqle++9V5K0fft2RURE+GT7qamp2r17d7ntAwCsxTmqAIyVlJSkyMhIuVwuffHFF+7xkydP6vHHH9eNN96o2NhYJSYmavz48dq3b1+5dSxYsEDR0dHavXu3Nm3apOHDhyshIUHJyckey6QfA3PZPz+1a9cu3X///erRo4diY2M1YMAAzZs3TwUFBeW2N3r0aEVHR+vChQtKT0/XgAEDFBsbq7S0NI/lxcXFSk9PV//+/RUXF6cBAwbo1Vdfda9n/fr1Gjp0qDp37qzevXtr/vz5cjqd5ba3YcMGTZo0STfeeKM6d+6srl27asSIEXr99dcrfE/Ltl9SUqIlS5bo5ptvVmxsrPr06aM//vGPunDhQoXPO3r0qB555BElJycrNjZWSUlJuvvuu7Vu3boK56alpalPnz6KjY3Vddddp4cffliZmZkVrhsALoZD/wCMVnbQx2azSZIOHDig++67T3l5eerVq5duvvlmnT59Wtu3b9fdd9+thQsXqk+fPuXWs2LFCu3cuVP9+vVTYmKiCgoK1KNHD02cOFEbN25Udna2+9SDn3r55Zf12GOPKTg4WLfccouaNm2qf/zjH1q+fLnee+89rV+/Xg6Ho9zzJk+erC+++EK9e/dW//791bRpU4/lDz30kPbu3as+ffqoXr16euutt/Too4+qXr16OnTokP7nf/5Hffv2Vc+ePfXuu+9q4cKFatCggcaOHeuxnscee0zt27fXtddeq2bNmik3N1c7duzQf/7nf+qbb77Rgw8+WOH7+vDDD2vPnj264YYb1KdPH33wwQd64YUX9MMPP2ju3Lkec99//31NmTJFFy5c0A033KDBgwcrPz9fhw4d0gsvvKC7777bPfeDDz7QpEmTVFJSon79+ql169bKycnRtm3b9P7772vVqlWKiYmpsCYAKMcFABaLiopyRUVFlRvfuXOnKzo62hUdHe3KyspyFRcXu/r37++KjY117d6922PuyZMnXb169XJdf/31rvPnz7vH58+f74qKinJ16dLFdeDAgQq3P2rUqAq3n5WV5YqJiXElJCS4vv76a49ls2bNckVFRbn+67/+q8J1DRkyxHXq1KmLbuv222935eXlucePHTvmiomJcXXv3t3Vr18/18mTJ93L8vLyXD169HAlJia6iouLPdb33XffldvG+fPnXampqa5OnTp5rOen2//1r3/tOn36tHv87Nmzrv79+7s6dOjg+te//uUeP3XqlKtr166umJiYcu+5y+Vy/e///q/7v3Nzc13du3d39ejRw3XkyBGPeYcOHXLFx8e7brvttnLrAICL4dA/AGMsWLBACxYs0J///GdNnjxZDzzwgFwul+655x61bNlS77//vo4dO6ZRo0apR48eHs9t0aKFHnjgAX3//ffatWtXuXXfeeed6tSpU6XqeeONN1RcXKxRo0apXbt2HsumTp2qhg0b6vXXX6/wcPmUKVPUpEmTi6572rRpHntiW7Vqpa5duyo/P1+/+93vPC7qcjgcSk5O1unTp5WTk+OxntatW5dbd1BQkEaOHKmSkpIK34uy7YeFhbkfh4SEaOjQoXI6ndq/f797/H/+53905swZjRgxotx7LklXXXWVx9z8/HxNnjxZ7du395gXFRWl4cOH68svv9TXX399kXcFADxx6B+AMdLT0yX9eJjf4XCoW7du+o//+A/deuutkqTPP/9cknTixAktWLCg3PO//fZbST+eI/nzw/+dO3eudD1ffvmlJKlnz57lljVu3FidOnXSP//5T2VmZqpDhw6V2l5sbGy5sebNm1922cmTJ9WyZUv3+IkTJ7R8+XLt2rVL//u//6tz5855PO/nwfZS2//lL38pScrLy3OPlb3nvXv3vtTL8Zj71VdfXbY/Pw+yAFARgioAY1zuqvvc3FxJ0tatWy85r7CwsNzYL37xi0rXU3axVLNmzSpcXjaen59/0WUX06hRo3Jj9er9+Cs5NDT0ostKSkrcY8ePH9d//Md/KD8/X927d1evXr0UGhqqgIAAZWdna+PGjRe9OKqi82oDAgIkyeOirbL3wJvbdpX156cXhVWkov4AQEUIqgD8Rlm4W7RokW688cZKPbfsYqyqbO///u//9Ktf/arc8u+//95j3pVur7JWrFih3NxczZ07V7fffrvHss2bN2vjxo1XvI2y15aTk1PubggXm/v666+X28MMAFXBOaoA/EaXLl0kSZ988kmNbK9jx46S5L591U/l5+fr4MGDql+/frnzV2vKd999J0m6+eabyy37xz/+US3bKPuGrg8++OCyc8v6s2fPnmrZNgAQVAH4jRtvvFGtW7fWunXrtGPHjgrnfPbZZyoqKqqW7Q0bNkyBgYFas2aNOxSWef7553XmzBkNGzZMQUFB1bK9yio7V/XnofTDDz/U3/72t2rZxm233abQ0FC9/PLL+uc//1lu+cmTJ93/ffvtt8vhcCg9Pb3Ce9o6nc4KQz8AXAyH/gH4jcDAQC1YsEAPPPCAxo4dq4SEBHXs2FENGjTQyZMn9cUXX+j48eP6+9//ruDg4CveXkREhB555BE9/vjj+vWvf62BAweqSZMm+uc//6nPPvtMkZGRmjZtWjW8sqq5++67tWHDBk2ZMkUDBgxQ8+bNdeTIEX344YcaOHCgMjIyrngbTZo00Z/+9CdNnjxZqamp6t27t6Kjo3XmzBkdOnRI//u//6t3331XkhQeHq758+drwoQJuvPOO5WUlKT27dvLZrPp5MmT+uyzz5Sbm+vx5Q0AcCkEVQB+pUOHDnr99de1YsUKvf/++9qwYYPsdruaNWumTp06adKkSQoPD6+27Y0cOVJt2rTRSy+9pG3btqmoqEi//OUvdf/992v8+PEVXpRUUzp06KBVq1bpueee044dO1RSUqIOHTooPT1djRo1qpagKkl9+/bVa6+95r67wM6dO+VwOBQZGalx48Z5zE1KStIbb7yhl156SX//+9/1ySefKDAwUM2bN1fPnj01YMCAaqkJQN1gc7n+/9e+AAAAAAbhHFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQB+I20tDQlJydbXYZfio6O1oIFC6wuAwAqpZ7VBQDwf9HR0V7NW7VqlRITE31cjfd2796t1NTUiy5/9tlnNXjw4BqsyDsLFixQenq6du3apSZNmpRbnpycrF/96ldaunTpFW1n06ZNOnXqlH7zm99c0XoAoKoIqgCu2H//9397PH799de1c+fOcuPt2rW7ou088cQTcrlcV7SOiowePVpxcXHlxuPj46t9W1bZt2+fAgICKvWczZs368iRIwRVAJYhqAK4YrfeeqvH471792rnzp3lxn+uqKhIwcHBXm8nMDCwSvVdTvfu3XXLLbdU6jlOp1PFxcWqX79+uWWFhYUKCQm5opoq+95cTkV1mq463kcA/o1zVAHUiNGjR2vIkCHav3+/Ro4cqS5duujZZ5+VJG3fvl1jx45Vr169FBsbq/79+2vhwoUqLS31WMfPz1HNyspSdHS0XnzxRb3yyivq37+/YmNjdccdd2jfvn3VWn90dLQef/xxvfHGGxo8eLDi4uL04YcfasOGDYqOjtY//vEPPfbYY0pKSlKfPn3cz1u7dq0GDx6s2NhY9erVS7Nnz1Z+fr7X70111v/Tc1TPnDmjJ598UsnJyYqNjVVSUpLuvfdeHThwwF3T+++/r+zsbEVHRys6OtrjvT916pRmzJih6667TnFxcRo2bJg2btxYbrunT5/W73//e3Xt2lXdu3fX9OnT9dVXXyk6OlobNmxwz0tLS1NCQoKOHTumMWPGKCEhQdOmTZMkffLJJ5o8ebL69u2r2NhY9enTR0899ZTOnTvnsa2ydZw4cULjxo1TQkKCbrjhBq1du1aSdOjQIaWmpio+Pl79+vXTpk2bqu8NBuAT7FEFUGNyc3M1ZswYDR48WMOGDVPTpk0lSRs3blRISIjuvfdehYSE6OOPP9b8+fN15swZTZ8+/bLr3bx5s86ePau77rpLNptNL7zwgiZNmqTt27d7tRf27Nmz+uGHH8qNh4eHy2azuR9//PHHevPNNzVy5EiFh4erZcuW7tA5e/ZsNWnSRBMmTFBhYaGkf59Let111yklJUXffPON1q9fry+++ELr16/3qO1i782l5OXlVTjudDov+9xZs2bprbfe0qhRo9SuXTvl5uZqz549Onr0qGJiYjR+/HgVFBTo5MmTeuSRRyRJDRs2lCSdO3dOo0eP1rFjxzRy5EhFRERo69atSktLU35+vu655x53Hb/97W+1b98+paSkKDIyUu+8885Fe1pSUqL7779f3bp10/Tp09WgQQNJ0tatW3Xu3DmlpKQoLCxM+/bt05o1a3Ty5EnNnz/fYx2lpaUaM2aMunfvrmnTpmnTpk16/PHHFRwcrD//+c8aOnSobr75Zr388suaPn264uPj1apVq8u+XwCsQVAFUGO+//57zZ49WyNGjPAY/9Of/uQOJZKUkpKimTNnav369Zo6daqCgoIuud4TJ05o27Ztaty4sSSpbdu2+t3vfqe///3v6tev32XrmjFjRoXjf//739WsWTP342+++UabNm1S+/bt3WMHDx6UJDVu3FgrV650nwf6ww8/aOnSperVq5eWL18uu/3HA1iRkZHuPbN33HHHZd+bS7nU6QqXu8Btx44duvPOO5WWluYeGzNmjPu/r7/+eq1atUr5+fnlTuF45ZVXdPToUf3xj3/UsGHDJEkjRozQ6NGj9dxzz+mOO+5QaGiotm/frs8++0wzZsxwh9eUlBTde++9FdZ04cIF3XLLLXr44Yc9xqdNm+bx+bjrrrvUpk0bPfvsszpx4oSuvvpq97Lz589r2LBhGjdunCRp6NChuuGGGzRjxgw9++yzGjRokCTpuuuu08CBA/U///M/mjRp0iXfKwDWIagCqDFBQUG6/fbby43/NIScOXNGFy5cUPfu3fXKK68oMzNTHTp0uOR6Bw0a5A6p0o/nnErS8ePHvaprwoQJ7uf81E/XKUnXXnutR0j9qTvvvNPjYqWPPvpIxcXFSk1NdYdUSRo+fLieffZZ7dixwyOoXuy9uZQFCxYoNDS03Pjvf//7yz7X4XBo7969ysnJUYsWLSq13Q8++EDNmjXTkCFD3GOBgYEaPXq0HnroIf3zn/9Uv3799OGHHyowMFB33nmne57dbtfIkSP18ccfV7julJSUcmM//XwUFhbq3LlzSkhIkMvl0pdffukRVKUf3+Ofvs62bdvqu+++08CBA93jkZGRcjgcXn9GAFiDoAqgxrRo0aLCvaNHjhzRc889p48//lhnzpzxWFZQUHDZ9f7yl7/0eFwWMH9+LujFREVF6brrrrvsvIiICK+XnThxQtKPgeingoKC1KpVK2VnZ3uMX+y9uZTu3btXeHsqby6cmjZtmtLS0tS3b1/FxMSoT58+uu2227w6DJ6dna02bdp4BHDp33d1KHvtJ06cULNmzcpdFNa6desK11uvXj1dddVV5cZPnDih+fPn69133y13usPPPy/169cv9540atRIV111lcdpHGXj3n5GAFiDoAqgxvx0z1iZ/Px8jRo1SqGhoZo8ebJat26t+vXr68CBA5o3b55X51te7LZL1X0rq4rqL3OlV9Vfat2+MGjQIHXv3l1vv/22du7cqRdffFHLly/XggULPC4Gq0lBQUHlwm9paanuvfde5eXl6YEHHlBkZKRCQkKUk5OjtLS0cp+Pi30WauozAqB6EVQBWOof//iHcnNzlZ6ermuvvdY9npWVZWFVV67scHRmZqbHXsoLFy4oKyvLqz24vta8eXONHDlSI0eO1KlTp/TrX/9aS5YscQfVn++BLNOyZUsdOnRITqfTI1hmZmZK+vdrv/rqq7V79+5yt9o6duyY1zUePnxY3377rZ555hnddttt7vGdO3d6vQ4A/ovbUwGwVFnQ+emerQsXLmjdunVWlVQtrrvuOgUGBmr16tUer+1vf/ubCgoKLNtrKf24l/Lnp1Q0bdpUzZs314ULF9xjwcHBFZ560bt3b33//ffKyMhwj5WUlGj16tUKCQlx/8HRq1cvFRcX69VXX3XPczqd7ttFeaOiz4fL5dKqVau8XgcA/8UeVQCWSkhIUOPGjZWWlqbRo0fLZrPp9ddfr9FDsp988onOnz9fbjw6OvqyF3JdTJMmTTRu3Dilp6frgQceUHJysr755hutW7fOfd9Rq5w9e1Z9+vTRgAED1KFDB4WEhOijjz7SF1984XEXgJiYGGVkZGju3LmKi4tTSEiIkpOTddddd+mVV15RWlqaDhw4oJYtW+qtt97Sp59+qhkzZrgv8Orfv786d+6sZ555RseOHVNkZKTHeaYX22P7U5GRkWrdurWeeeYZ5eTkKDQ0VG+99RbnlgJ1BEEVgKXCw8O1ZMkSPfPMM3ruuefkcDg0bNgwJSUl6f7776+RGlavXl3h+MSJE6scVCVp0qRJatKkidasWaO5c+eqcePGuvPOO/XQQw/57Fu2vNGgQQOlpKRo586d2rZtm1wul1q3bq1Zs2bp7rvvds+7++67dfDgQW3YsEErV65Uy5YtlZycrAYNGmj16tWaN2+eNm7cqDNnzqht27aaO3eux50LAgICtHTpUj355JPauHGj7Ha7brrpJk2YMEEpKSlendcbGBioJUuWaM6cOVq6dKnq16+vm266SSNHjrzsN58B8H82F2eSAwBq0Pbt2zVhwgStW7dO3bp1s7ocAAbjHFUAgM/8/GtOS0tLtXr1aoWGhiomJsaiqgD4Cw79AwB85oknnnDfoP/ChQvatm2bPvvsMz300EM1fksuAP6HQ/8AAJ/ZtGmTVqxYoe+++07nz59XmzZtlJKSolGjRlldGgA/QFAFAACAkThHFQAAAEYiqAIAAMBIBFUAAAAYqdZd9e9yueR0ep52a7fbyo3Bv9BD/0cP/R899H/00P/Vhh7a7TavvplOqoVB1el06Ycfzrof16tnV3h4Q+XnF6qkxGlhZagqeuj/6KH/o4f+jx76v9rSwyZNGiogwLugyqF/AAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIlQ6q3333nWbOnKlbb71VnTp10pAhQyqc99e//lUDBgxQXFychg0bpvfee6/cnIKCAs2YMUM9evRQQkKCJk+erH/961+VfxUAAACodSodVI8cOaIdO3aoTZs2ateuXYVztmzZokcffVQDBw7U8uXLFR8fr4kTJ+rzzz/3mPfggw9q586deuyxxzRv3jx98803GjNmjEpKSqr0YgAAAFB7VPorVJOTk9W/f39JUlpamvbv319uzvz58zV48GA9+OCDkqSePXvq8OHDWrhwoZYvXy5J+uyzz/T3v/9dL774onr16iVJatu2rQYNGqRt27Zp0KBBVX1NsJDdbpPd7t3XojmdLr//vmIAAOA7ld6jardf+inHjx/Xt99+q4EDB3qMDxo0SLt27dKFCxckSR988IEcDoeuv/5695zIyEh17NhRH3zwQWXLggHsdpvCwkIUHt7Qq3/CwkK8DrUAAKDuqfQe1cvJzMyU9OPe0Z9q166diouLdfz4cbVr106ZmZlq27atbDbPoBIZGeleR1XVq/fvMB0QYPf4N3wnIMCugAC75q3do6ycgkvOjWjRSNNGdlNgYIBKS52XXe9P/w3/Qw/9Hz30f/TQ/9XFHlZ7UM3Ly5MkORwOj/Gyx2XL8/Pz1ahRo3LPb9y4cYWnE3jLbrcpPLxhuXGHI7jK60TlZOUU6Gh2nldzK9MXeuj/6KH/o4f+jx76v7rUw2oPqlZzOl3Kzy90Pw4IsMvhCFZ+ftFl99zhypS915XhTV/oof+jh/6PHvo/euj/aksPHY5gr/cKV3tQbdy4saQfbz3VrFkz93h+fr7HcofDoZMnT5Z7fl5enntOVZWUlG9eaamzwnFYqzJ9oYf+jx76P3ro/+ih/6tLPaz2kxwiIyMlqdx5ppmZmQoMDFSrVq3c87755hu5XJ5XfX/zzTfudQAAAKDuqvag2qpVK11zzTXaunWrx3hGRoaSkpIUFBQkSerdu7fy8vK0a9cu95xvvvlGX375pXr37l3dZQEAAMDPVPrQf1FRkXbs2CFJys7O1pkzZ9yhtEePHmrSpIkmTZqkadOmqXXr1kpMTFRGRob27dunNWvWuNeTkJCgXr16acaMGZo+fbrq16+vP//5z4qOjtbNN99cTS8PAAAA/qrSQfXUqVOaMmWKx1jZ41WrVikxMVFDhgxRUVGRli9frmXLlqlt27ZKT09XQkKCx/Oee+45zZ07VzNnzlRJSYl69eql//qv/1K9erXuGi8AAABUUqUTYUREhA4dOnTZecOHD9fw4cMvOadRo0Z66qmn9NRTT1W2DAAAANRydeeOsQAAAPArBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAI9WzugDUbQEBl/9byZs5AACg9iGowhJhjerL6XTJ4Qj2ar7T6ZLNZvNxVQAAwCQEVVgiNDhQdrtN89buUVZOwSXnRrRopGkju8luJ6gCAFCXEFRhqaycAh3NzrO6DAAAYCBO/gMAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjcXsqXJbdbvPqHqZ8gxQAAKhOPguq77zzjpYsWaKvv/5aDRs2VLdu3TRt2jS1atXKY95f//pXvfDCCzpx4oTatm2rqVOnql+/fr4qC5Vkt9sUFhZCCAUAADXOJ0F19+7dmjhxom677TZNnTpVubm5ev7553Xfffdp06ZNatCggSRpy5YtevTRRzV+/Hj17NlTGRkZmjhxotauXav4+HhflIZKstttCgiwe/UNUl07NFfqoE41VBkAAKjtfBJUt2zZoquvvlpPPfWU+/vZmzRponvuuUf79+9X9+7dJUnz58/X4MGD9eCDD0qSevbsqcOHD2vhwoVavny5L0pDFXnzDVIRzUNrqBoAAFAX+OR4bklJiRo2bOgOqZLUqFEjSZLL5ZIkHT9+XN9++60GDhzo8dxBgwZp165dunDhgi9KAwAAgJ/wSVC9/fbbdfToUa1du1YFBQU6fvy4nn32WXXq1Eldu3aVJGVmZkqS2rZt6/Hcdu3aqbi4WMePH/dFaQAAAPATPjn03717d6Wnp+vhhx/W448/Lknq2LGjXnjhBQUEBEiS8vJ+PIzscDg8nlv2uGx5VdSr9+/8XXYREBcDVY1J75vdbvPoLfwHP4f+jx76P3ro/+piD30SVD/99FP953/+p+6880717dtXubm5WrRokcaOHat169a5L6byBbvdpvDwhuXGHY5gn20TNSM01HefG9QMfg79Hz30f/TQ/9WlHvokqM6ZM0c9e/ZUWlqaeyw+Pl59+/bV66+/rrvuukuNGzeWJBUUFKhZs2buefn5+ZLkXl5ZTqdL+fmF7scBAXY5HMHKzy9SaamzSuusy8rePxOcOXNOxcWlVpeBKuDn0P/RQ/9HD/1fbemhwxHs9V5hnwTVo0eP6sYbb/QYu+qqqxQeHq5jx45JkiIjIyX9eK5q2X+XPQ4MDCx3v9XKKCkp37zSUmeF4/AfTqeLHvo5fg79Hz30f/TQ/9WlHvrkJIerr75aX375pcdYdna2Tp8+rZYtW0qSWrVqpWuuuUZbt271mJeRkaGkpCQFBQX5ojQAAAD4CZ/sUR0xYoSeeuopzZkzR8nJycrNzdXixYvVtGlTj9tRTZo0SdOmTVPr1q2VmJiojIwM7du3T2vWrPFFWQAAAPAjPgmqqampCgoK0vr16/Xaa6+pYcOGio+P13PPPafw8HD3vCFDhqioqEjLly/XsmXL1LZtW6WnpyshIcEXZQEAAMCP+CSo2mw2paSkKCUl5bJzhw8fruHDh/uiDAAAAPixunMjLgAAAPgVgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABipntUFoObZ7TbZ7Tav5gYE8LcMAACwBkG1jrHbbQoLCyGAAgAA4xFU6xi73aaAALvmrd2jrJyCy87v2qG5Ugd1qoHKAAAAPBFU66isnAIdzc677LyI5qE1UA0AAEB5HP8FAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIzEDf9rCbvdJrvddtl5fHUqAADwFwTVWsButyksLIQQCgAAahWCai1gt9sUEGDXvLV7lJVTcMm5XTs0V+qgTjVUGQAAQNURVGuRrJwCHc3Ou+SciOahNVQNAADAleFYMQAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAj+TSobty4Ubfddpvi4uKUmJioBx54QOfOnXMvf/fddzVs2DDFxcVpwIABeu2113xZDgAAAPyIz274v3jxYi1fvlzjx49XfHy8Tp8+rV27dqm0tFSS9Mknn2jixIn6j//4D82YMUMff/yx/vCHP6hhw4a65ZZbfFUWAAAA/IRPgmpmZqbS09O1aNEi9enTxz0+YMAA938vXrxYnTt31uOPPy5J6tmzp44fP6758+cTVAEAAOCbQ/8bNmxQRESER0j9qQsXLmj37t3lAumgQYN09OhRZWVl+aIsAAAA+BGfBNW9e/cqKipKixYtUlJSkmJjYzVixAjt3btXknTs2DEVFxcrMjLS43nt2rWT9OMeWQAAANRtPjn0//3332v//v06fPiwZs2apeDgYC1ZskT33Xeftm3bpry8PEmSw+HweF7Z47LlVVWv3r/zd0CA3ePftVFtfm0/ZbfbPHoL/1EXfg5rO3ro/+ih/6uLPfRJUHW5XCosLNTzzz+vDh06SJK6dOmi5ORkrVmzRr169fLFZiX9GGbCwxuWG3c4gn22TdSM0NAGVpeAK8TPof+jh/6PHvq/utRDnwRVh8OhsLAwd0iVpLCwMHXq1Elff/21Bg8eLEkqKCjweF5+fr4kqXHjxlXettPpUn5+oftxQIBdDkew8vOLVFrqrPJ6TVb2Gmu7wsLzKinxrodOp0sul8vHFcFbdeHnsLajh/6PHvq/2tJDhyPY673CPgmq7du317Fjxypcdv78ebVu3VqBgYHKzMzUDTfc4F5Wdm7qz89drayKwkxpqdPrkAOzhDWqL6fTpZCQ+l4/p7TUqdzcQjmdhFWT8HPo/+ih/6OH/q8u9dAnQbVfv37asGGDDh48qI4dO0qSTp8+rQMHDug3v/mNgoKClJiYqLfeekv33HOP+3kZGRlq166dIiIifFEW/FRocKDsdpvmrd2jrJyCy86PaNFI00Z2k91uI6gCAODHfBJU+/fvr7i4OE2ePFlTp05V/fr1tWzZMgUFBenuu++WJP32t79VamqqHnvsMQ0cOFC7d+/W5s2b9ec//9kXJaEWyMop0NHsK7vQDgAA+A+fXDZmt9u1bNkyxcfHa+bMmXrooYcUGhqqtWvXqlmzZpKk7t27a8GCBdqzZ4/uv/9+bd68WXPmzNHAgQN9URIAAAD8jM++QrVJkyb64x//eMk5N954o2688UZflQAAAAA/VnduxAUAAAC/QlAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABjJ50H17Nmz6t27t6Kjo/XFF194LPvrX/+qAQMGKC4uTsOGDdN7773n63IAAADgJ3weVBctWqTS0tJy41u2bNGjjz6qgQMHavny5YqPj9fEiRP1+eef+7okAAAA+AGfBtWjR49q3bp1mjRpUrll8+fP1+DBg/Xggw+qZ8+eevzxxxUXF6eFCxf6siQAAAD4CZ8G1Tlz5mjEiBFq27atx/jx48f17bffauDAgR7jgwYN0q5du3ThwgVfluU37Hab6tWzX/afgABONQYAALVPPV+teOvWrTp8+LAWLFigAwcOeCzLzMyUpHIBtl27diouLtbx48fVrl07X5XmF+x2m8LCQgihAACgzvJJUC0qKtLTTz+tqVOnKjQ0tNzyvLw8SZLD4fAYL3tctryq6tX7d7grC3r+FvgCAn7cUzpv7R5l5RRccm7XDs2VOqhTDVXmP/yt57WZv/4c4t/oof+jh/6vLvbQJ0F18eLFatq0qe644w5frP6S7HabwsMblht3OIJrvJbqkJVToKPZlw7uEc3L/zEA/+15bUZP/B899H/00P/VpR5We1DNzs7WSy+9pIULF6qg4Mc9gYWFhe5/nz17Vo0bN5YkFRQUqFmzZu7n5ufnS5J7eVU4nS7l5xe6HwcE2OVwBCs/v0ilpc4qr7emldWNqvO3ntdm/vpziH+jh/6PHvq/2tJDhyPY673C1R5Us7KyVFxcrLFjx5Zblpqaqi5duuhPf/qTpB/PVY2MjHQvz8zMVGBgoFq1anVFNZSUlG9eaamzwnHUXvTcPPTE/9FD/0cP/V9d6mG1B9WOHTtq1apVHmMHDx7U3LlzNXv2bMXFxalVq1a65pprtHXrVvXv3989LyMjQ0lJSQoKCqrusgAAAOBnqj2oOhwOJSYmVrgsJiZGMTExkqRJkyZp2rRpat26tRITE5WRkaF9+/ZpzZo11V0SAAAA/JDPbk91OUOGDFFRUZGWL1+uZcuWqW3btkpPT1dCQoJVJQEAAMAgNRJUExMTdejQoXLjw4cP1/Dhw2uiBAAAAPiZunMjLgAAAPgVgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMFI9qwsAUHl2u012u82ruU6nS06ny8cVAQBQ/QiqgJ+x220KCwtRQIB3B0RKS53KzS0krAIA/A5BFfAzdrtNAQF2zVu7R1k5BZecG9GikaaN7Ca73UZQBQD4HYIq4Keycgp0NDvP6jIAAPAZLqYCAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAI3F7KtRa3t4Qn29uAgDATD4Jqm+++abeeOMNHThwQPn5+WrTpo1Gjx6tO+64Qzbbv7/28a9//ateeOEFnThxQm3bttXUqVPVr18/X5SEOiSsUX05nS45HMFezeebmwAAMJNPgurKlSvVsmVLpaWlKTw8XB999JEeffRRnTx5UhMnTpQkbdmyRY8++qjGjx+vnj17KiMjQxMnTtTatWsVHx/vi7JQR4QGB8put/HNTQAA+DmfBNXFixerSZMm7sdJSUnKzc3VihUr9Lvf/U52u13z58/X4MGD9eCDD0qSevbsqcOHD2vhwoVavny5L8pCHcM3N/2bt6dBSJwKAQAwh08upvppSC3TsWNHnTlzRoWFhTp+/Li+/fZbDRw40GPOoEGDtGvXLl24cMEXZQF1zk9PgwgPb+jVP2FhIbLbbZdfOQAAPlZjF1Pt2bNHLVq0UGhoqPbs2SNJatu2rcecdu3aqbi4WMePH1e7du2qvK169f6dv8v2JFVmj5IJ/K1ef+dP73dlaq3MaRDSv0+FCAwMUGmp80rK9OCvP4f4N3ro/+ih/6uLPayRoPrJJ58oIyND06dPlyTl5f14ONbhcHjMK3tctrwq7HabwsMblhv39sIa1E21/fNR2dMgfPV+1Pb3uS6gh/6PHvq/utRDnwfVkydPaurUqUpMTFRqaqqvNyen06X8/EL344AAuxyOYOXnF1XrHiJfK6sbNcOfPh818dmo7vfDX38O8W/00P/RQ/9XW3rocAR7vVfYp0E1Pz9fY8aMUVhYmBYsWCC7/ceiGjduLEkqKChQs2bNPOb/dHlVlZSUb15pqbPCcUDi8/Fzvno/eJ/9Hz30f/TQ/9WlHvrsJIdz585p3LhxKigo0AsvvKBGjRq5l0VGRkqSMjMzPZ6TmZmpwMBAtWrVyldlAQAAwE/4JKiWlJTowQcfVGZmpl544QW1aNHCY3mrVq10zTXXaOvWrR7jGRkZSkpKUlBQkC/KAgAAgB/xyaH/2bNn67333lNaWprOnDmjzz//3L2sU6dOCgoK0qRJkzRt2jS1bt1aiYmJysjI0L59+7RmzRpflAQAAAA/45OgunPnTknS008/XW7ZO++8o4iICA0ZMkRFRUVavny5li1bprZt2yo9PV0JCQm+KAkAAAB+xidB9d133/Vq3vDhwzV8+HBflAAAAAA/V3fuGAsAAAC/QlAFAACAkWrsK1TxI7vd5tX3qNelr0cDAACoCEG1BtntNoWFhRBCAQAAvEBQrUF2u00BAXbNW7tHWTkFl5zbtUNzpQ7qVEOVAQAAmIegaoGsnAIdzc675JyI5qE1VA0AAICZOAYNAAAAI7FHFUA53p5H7XS65HS6fFwNAKCuIqgCcAtrVF9Op0sOR7BX80tLncrNLSSsAgB8gqAKGMKEW5eFBgfKbrd5dcFfRItGmjaym+x2m+VB1dv3TmIvMAD4E4IqYADTbl3mzQV/pqjse8deYADwHwRVwADcuqzqKvPembQXGABweQRVwCDcuqzq/GkvMADAO2YcZwQAAAB+hj2qAK6IN+eGls3htlcAgMogqAKokkrfyorbXgEAKomgCqBKKnMrq7ILwLjgCQBQGQRVAFekMheAccETAKAyuJgKAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJG4jyoAI3n7davezoO57Hab7Hab1/P5il2g7iCoAjBKZb+aFf7NbrcpLCykUn9w8BW7QN1BUAVglMp8Nav0769n9TeV2YtYmT2I/rZ30m63KSDA7nW/+YpdT5XpN0cf4I8IqgCM5O3XrZZ9Pas/qexeRG/3IPrz3km+XrfyqtJvp9Mlm837P2QAqxFUAaCGVWYvYmX2ILJ3sm65kn4D/oKgeoU47AL4H29/Fn19WLwyexG9qblsDnsn/81Xp1iYhH6jNiOoXoGqHHYBYJ3KXqhlwmFxLi6rOl+dYgGg5hBUr0BlD7v460UfQG1RmQu1yg6TBgYGqLTUWa11VOaP28rUXNXfMZXZW+stb/dk+vIP/aqcYlGZfvvrHlj4r7Lzi005KlQTCKrVoDZf9AHURt78zFZlT2ap06UAH53/503Nlf0dU9nX6O2FOKYdbfJZv9kDixpkt9vUqFEDSfKro0JXiqAKABWo6m2yfLnns7pVZQ+zN3tJK7Mn0x/fC4kL0VDz6urFkgRVQL47jGLC4U9cmcoeMfHFnk9fq8zFOHa7TfXqXfrzWpmLuvz5vfBX3vSwsnx5iLm2XxBX2f9P1IXP6E8RVFGn+fLiGtMOfwJXouxnJTS0gdWloIoq28PKnMriq0PMtf2COP4/cXkEVdRpVT306W1Q9bfDn8DF1MRFXZXhiwvAfFWHKduuSg99cSFaZb9pzRf3HK4KX+zZ5f8Tl2dpUD169KjmzJmjzz77TA0bNtStt96qBx98UEFBQVaWhTrIl4dS/PHwJ3AxVn+eTbldV6WPxlRm76QPL8qTKtdDX1yIVpW9nlYf7vb1nl2rf65MZllQzcvL0z333KNrrrlGCxYsUE5Ojp5++mmdO3dOM2fOtKosAIDBTNmz66u9k3Xpojx/OTwvmbVnt66xLKi+/PLLOnv2rNLT0xUWFiZJKi0t1ezZszVu3Di1aNHCqtKAS/L2L2rOOQJ8x5Q9UNW9d7KuXJTny9+jvvjdW5ULmfzt9BRTL0SzLKh+8MEHSkpKcodUSRo4cKBmzZqlnTt36vbbb7eqNKBCphxyBAB/5cvfo748HcOXdfiKP34TX0UsC6qZmZm64447PMYcDoeaNWumzMxMi6oCLq6q99UEAPzIl79HfXU6Rk3U4Qu15ZQMm8vlsqSimJgYTZkyRWPHjvUYHzJkiBISEvTEE09Uab0ul+eua5tNstvtcjqdqu5XWrbu3ILzKvHiSsf6QQFqFBLk1XwT5ppShz/WbEod1Fy36qDmulUHNdetOnxZc70A+//fA1v9WakidrvNq2+5k2phUAUAAEDtYNnVHg6HQwUF5XdF5+XlqXHjxhZUBAAAAJNYFlQjIyPLnYtaUFCg77//XpGRkRZVBQAAAFNYFlR79+6tjz76SPn5+e6xrVu3ym636/rrr7eqLAAAABjCsnNU8/LyNHjwYLVt21bjxo1z3/B/6NCh3PAfAAAA1gVV6cevUH3iiSc8vkJ16tSpfIUqAAAArA2qAAAAwMXwHY8AAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgpHpWF+ALb775pt544w0dOHBA+fn5atOmjUaPHq077rhDNpvN6vLghR07dmj58uX6+uuvdebMGbVo0UL9+/fXxIkT1ahRI6vLQxWcPXtWAwcOVE5Ojv72t78pLi7O6pJwGRs2bNAjjzxSbnzMmDGaNm2aBRWhqjZu3Ki//OUvOnr0qEJCQhQXF6f09HQ1aNDA6tJwGaNHj9Y//vGPCpc9++yzGjx4cA1XVLNqZVBduXKlWrZsqbS0NIWHh+ujjz7So48+qpMnT2rixIlWlwcv5ObmqnPnzho9erTCwsJ05MgRLViwQEeOHNFLL71kdXmogkWLFqm0tNTqMlAFL7zwgscfiC1atLCwGlTW4sWLtXz5co0fP17x8fE6ffq0du3axc+jn5g1a5bOnDnjMfaXv/xF27ZtU1JSkkVV1ZxaGVQXL16sJk2auB8nJSUpNzdXK1as0O9+9zvZ7ZzxYLpbb73V43FiYqKCgoL06KOPKicnh/9R+pmjR49q3bp1mj59umbNmmV1OaikmJgYj9+p8B+ZmZlKT0/XokWL1KdPH/f4gAEDLKwKldG+fftyYw8//LCuv/76OvFzWSsTW0WN69ixo86cOaPCwkILKkJ1CAsLkyQVFxdbWwgqbc6cORoxYoTatm1rdSlAnbJhwwZFRER4hFT4t08//VRZWVkaOnSo1aXUiFoZVCuyZ88etWjRQqGhoVaXgkooLS3V+fPndeDAAS1cuFDJycmKiIiwuixUwtatW3X48GFNmDDB6lJQRUOGDFHHjh114403aunSpRwy9iN79+5VVFSUFi1apKSkJMXGxmrEiBHau3ev1aWhijZv3qyQkBDdeOONVpdSI2rlof+f++STT5SRkaHp06dbXQoqqV+/fsrJyZEk3XDDDfrTn/5kcUWojKKiIj399NOaOnUqfyT6oWbNmmnSpEnq0qWLbDab3n33XT333HPKycnRzJkzrS4PXvj++++1f/9+HT58WLNmzVJwcLCWLFmi++67T9u2bVPTpk2tLhGVUFJSojfffFPJyckKCQmxupwaUeuD6smTJzV16lQlJiYqNTXV6nJQScuWLVNRUZG+/vprLV68WOPHj9eKFSsUEBBgdWnwwuLFi9W0aVPdcccdVpeCKrjhhht0ww03uB/36tVL9evX11/+8heNHz9ezZs3t7A6eMPlcqmwsFDPP/+8OnToIEnq0qWLkpOTtWbNGk2ZMsXiClEZO3fu1A8//KAhQ4ZYXUqNqdWH/vPz8zVmzBiFhYVpwYIFXETlhzp06KCEhAQNHz5cixYt0u7du/X2229bXRa8kJ2drZdeekmTJ09WQUGB8vPz3eeIFxYW6uzZsxZXiKoYOHCgSktLdfDgQatLgRccDofCwsLcIVX68Xz/Tp066euvv7awMlTF5s2bFRYWpl69elldSo2ptXtUz507p3HjxqmgoECvvPIK996sBaKjoxUYGKhjx45ZXQq8kJWVpeLiYo0dO7bcstTUVHXp0kWvvvqqBZUBdUf79u0v+jvz/PnzNVwNrsS5c+e0fft2DRs2TIGBgVaXU2NqZVAtKSnRgw8+qMzMTK1du5ZbGdUSe/fuVXFxMRdT+YmOHTtq1apVHmMHDx7U3LlzNXv2bG7476cyMjIUEBCgTp06WV0KvNCvXz9t2LBBBw8eVMeOHSVJp0+f1oEDB/Sb3/zG2uJQKe+++64KCwvrzNX+ZWplUJ09e7bee+89paWl6cyZM/r888/dyzp16qSgoCDrioNXJk6cqNjYWEVHR6tBgwb66quv9OKLLyo6Olr9+/e3ujx4weFwKDExscJlMTExiomJqeGKUFn333+/EhMTFR0dLUl655139Oqrryo1NVXNmjWzuDp4o3///oqLi9PkyZM1depU1a9fX8uWLVNQUJDuvvtuq8tDJWzatElXX321unXrZnUpNapWBtWdO3dKkp5++ulyy9555x32yPmBzp07KyMjQ8uWLZPL5VLLli01fPhw3X///fyhAdSQtm3b6rXXXtPJkyfldDp1zTXXaMaMGRo9erTVpcFLdrtdy5Yt09y5czVz5kwVFxere/fuWrt2LX9s+JG8vDx9+OGHuueee+rcV8HbXC6Xy+oiAAAAgJ/jMngAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVwBXLyspSdHS0NmzY4B5bsGCB+0bxlxMdHa0FCxZUa02jR4/mfp8+5ou+AcBPEVSBOmb8+PHq0qWLzpw5c9E5Dz/8sGJjY3X69OkarKzyvv76ay1YsEBZWVlWl+K2e/duRUdHX/SfLVu2WF1ihcr+sPjhhx8qXJ6cnKxx48Zd8XY2bdqklStXXvF6ANQNtfKbqQBc3LBhw/Tee+9p+/btuu2228otLyoq0rvvvqtevXopPDy8ytv57W9/q7Fjx15BpZf39ddfKz09XT169Cj3jXMvvviiT7d9OaNHj1ZcXFy58fj4+Jovxkf27dungICASj1n8+bNOnLkCN8zD8ArBFWgjklOTlbDhg21adOmCoPqO++8o8LCQg0bNuyKtlOvXj3Vq2fdrxirv2q3e/fuuuWWWyr1HKfTqeLiYtWvX7/cssLCQoWEhFxRTUVFRQoODr6idfxURXWarjreRwA1h0P/QB3ToEED3Xzzzfr444916tSpcss3b96shg0bKjk5Wbm5uXrmmWc0dOhQJSQkqGvXrnrggQf01VdfXXY7FZ2jeuHCBT311FPq2bOnEhISNH78eJ08ebLcc7Ozs/XYY49pwIAB6ty5sxITEzV58mSPQ/wbNmzQlClTJEmpqanuQ+u7d++WVPE5qqdOndKMGTN03XXXKS4uTsOGDdPGjRs95pSdb/viiy/qlVdeUf/+/RUbG6s77rhD+/btu+zrrozo6Gg9/vjjeuONNzR48GDFxcXpww8/1IYNGxQdHa1//OMfeuyxx5SUlKQ+ffq4n7d27VoNHjxYsbGx6tWrl2bPnq38/HyPdY8ePVpDhgzR/v37NXLkSHXp0kXPPvtstdf/03NUz5w5oyeffFLJycmKjY1VUlKS7r33Xh04cMBd0/vvv6/s7Gx3v5KTk93P96Y/knT69Gn9/ve/V9euXdW9e3dNnz5dX331VbnzpNPS0pSQkKBjx45pzJgxSkhI0LRp0yRJn3zyiSZPnqy+ffsqNjZWffr00VNPPaVz5855bKtsHSdOnNC4ceOUkJCgG264QWvXrpUkHTp0SKmpqYqPj1e/fv20adOm6nuDAbBHFaiLhg4dqo0bN+rNN9/UqFGj3OO5ubn6+9//rsGDB6tBgwY6cuSItm/frltuuUURERH6v//7P73yyisaNWqUtmzZohYtWlRqu3/4wx/0xhtvaMiQIeratas+/vjjCk8P+OKLL/TZZ59p8ODBuuqqq5Sdna3169crNTVVW7ZsUXBwsK699lqNHj1aq1ev1vjx4xUZGSlJateuXYXbPnfunEaPHq1jx45p5MiRioiI0NatW5WWlqb8/Hzdc889HvM3b96ss2fP6q677pLNZtMLL7ygSZMmafv27QoMDLzsaz179myF53uGh4fLZrO5H3/88cd68803NXLkSIWHh6tly5bu0Dl79mw1adJEEyZMUGFhoaQf/wBIT0/Xddddp5SUFH3zzTdav369vvjiC61fv96jttzcXI0ZM0aDBw/WsGHD1LRp08vWnZeXV+G40+m87HNnzZqlt956S6NGjVK7du2Um5urPXv26OjRo4qJidH48eNVUFCgkydP6pFHHpEkNWzYUJL3/XE6nfrtb3+rffv2KSUlRZGRkXrnnXc0ffr0CmsqKSnR/fffr27dumn69Olq0KCBJGnr1q06d+6cUlJSFBYWpn379mnNmjU6efKk5s+f77GO0tJSjRkzRt27d9e0adO0adMmPf744woODtaf//xnDR06VDfffLNefvllTZ8+XfHx8WrVqtVl3y8AXnABqHNKSkpc119/veuuu+7yGF+/fr0rKirK9eGHH7pcLpfr/PnzrtLSUo85x48fd8XGxrrS09M9xqKiolyvvfaae2z+/PmuqKgo9+ODBw+6oqKiXI899pjH+h566CFXVFSUa/78+e6xoqKicjV/9tlnrqioKNfGjRvdY2+++aYrKirK9fHHH5ebP2rUKNeoUaPcj1euXOmKiopyvf766+6xCxcuuO666y5XfHy8q6CgwOO19OjRw5Wbm+ueu337dldUVJTr3XffLbetn/r4449dUVFRF/3nX//6l3tuVFSUq0OHDq4jR454rOO1115zRUVFuVJSUlwlJSXu8VOnTrliYmJc9913n0df1qxZ44qKinL97W9/83j9UVFRrvXr11+y3jJl/brUP2PHjvV4zs/71q1bN9fs2bMvuZ2xY8e6+vXrV27c2/689dZbrqioKNfKlSvd80pLS12pqanlPoPTp093RUVFuebNm1duexV9xpYuXeqKjo52ZWdnl1vHkiVL3GN5eXmuzp07u6Kjo11btmxxjx89erTcewLgynDoH6iDAgICNHjwYH322Wceh9M3b96sX/ziF0pKSpL043medvuPvyZKS0t1+vRphYSEqG3btvryyy8rtc0dO3ZIUrnD8T/fkynJvddLkoqLi3X69Gm1bt1aDoej0tst88EHH6hZs2YaMmSIeywwMFCjR49WYWGh/vnPf3rMHzRokBo3bux+3L17d0nS8ePHvdrehAkTtGLFinL//HSdknTttdeqffv2Fa7jzjvv9LhY6aOPPlJxcbFSU1PdfZGk4cOHKzQ01P0elwkKCtLtt9/uVb1lFixYUGHdv/jFLy77XIfDob179yonJ6dS25S878+HH36owMBA3Xnnne55drtdI0eOvOi6U1JSyo399DNWWFioH374QQkJCXK5XBV+xoYPH+7+b4fDobZt2yo4OFgDBw50j0dGRsrhcHj9GQFweRz6B+qooUOHauXKldq8ebP7XNFPPvlEo0ePdocjp9OpVatWad26dcrKylJpaan7+WFhYZXaXnZ2tux2u1q3bu0xXnbI/qfOnTunpUuXasOGDcrJyZHL5XIvKygoqNR2f7r9Nm3aeAQ86d+nCpw4ccJj/Je//KXH47KA+fNzQS8mKipK11133WXn/fxuBZdaVlbjz9+zoKAgtWrVStnZ2R7jLVq0qPRFZd27d1eTJk3KjXtz4dS0adOUlpamvn37KiYmRn369NFtt93m1WFwb/tz4sQJNWvWrNxFYT//XJWpV6+errrqqnLjJ06c0Pz58/Xuu++WO93h57duq1+/frn3pFGjRrrqqqs8TuMoG/f2MwLg8giqQB0VGxuryMhIbdmyRePHj9fmzZvlcrk0dOhQ95wlS5bo+eef1x133KEpU6aocePGstvteuqppzzCY3V74okntGHDBt1zzz2Kj49Xo0aNZLPZNHXqVJ9u96cudtul6t7+T/fs/dyVXlV/qXX7wqBBg9S9e3e9/fbb2rlzp1588UUtX75cCxYs8LgYrCb99KhAmdLSUt17773Ky8vTAw88oMjISIWEhCgnJ0dpaWnlzse92Gehpj4jQF1GUAXqsKFDh+r555/XV199pc2bN+uaa65R586d3cvfeustJSYm6qmnnvJ4Xn5+fqXvsdqyZUs5nU4dO3bMY49gZmZmublvvfWWbrvtNqWlpbnHzp8/X25v6s/3Zl1u+4cOHZLT6fQILmXbv/rqq71el1XKaszMzPTYS3nhwgVlZWV5tQfX15o3b66RI0dq5MiROnXqlH79619ryZIl7qB6sZ5525+rr75au3fvLnerrWPHjnld4+HDh/Xtt9/qmWee8bhF286dO71eB4CawTmqQB1Wtvd0/vz5OnjwoMfeVOnHPUY/3zv05ptvVukcxN69e0uSVq9e7TH+l7/8pdzcivZUrV692uPUA0nuoOLN6QC9e/fW999/r4yMDPdYSUmJVq9erZCQEF177bWXfxEWu+666xQYGKjVq1d79OVvf/ubCgoKLNtrKf24l/LnfWjatKmaN2+uCxcuuMeCg4Mr7Je3/enVq5eKi4v16quvuuc5nU737aK8URaEf/oeulwurVq1yut1AKgZ7FEF6rBWrVopISFB77zzjiSVC6p9+/bVwoUL9cgjjyghIUGHDx/Wpk2bqnTrnY4dO2rIkCFat26dCgoKlJCQoI8//ljfffddubl9+/bV66+/rtDQULVv316ff/65Pvroo3LnxXbs2FEBAQFavny5CgoKFBQUpJ49e1Z4G6a77rpLr7zyitLS0nTgwAG1bNlSb731lj799FPNmDFDoaGhlX5Nl/LJJ5/o/Pnz5cajo6PVoUOHKq2zSZMmGjdunNLT0/XAAw8oOTlZ33zzjdatW+e+76hVzp49qz59+mjAgAHq0KGDQkJC9NFHH+mLL77w2DMeExOjjIwMzZ07V3FxcQoJCVFycrLX/enfv786d+6sZ555xr13/qfnmXqzlz0yMlKtW7fWM888o5ycHIWGhuqtt97i3FLAQARVoI4bOnSoPvvsM3Xu3Flt2rTxWDZ+/HgVFRVp06ZNysjIUKdOnbR06VL96U9/qtK2nnrqKYWHh2vTpk165513lJiYqGXLlpXbE/iHP/xBdrtdmzZt0vnz59W1a1etWLFCDzzwgMe8Zs2aafbs2Vq6dKn+8Ic/qLS0VKtWraowqDZo0ECrV6/WvHnztHHjRp05c0Zt27bV3LlzK31lvDd+vue4zMSJE6scVCVp0qRJatKkidasWaO5c+eqcePGuvPOO/XQQw95dX9XX2nQoIFSUlK0c+dObdu2TS6XS61bt9asWbN09913u+fdfffdOnjwoDZs2KCVK1eqZcuWSk5O9ro/AQEBWrp0qZ588klt3LhRdrtdN910kyZMmKCUlBSvzusNDAzUkiVLNGfOHC1dulT169fXTTfdpJEjR+rWW2/1yfsDoGpsLs76BgD4ue3bt2vChAlat26dunXrZnU5AKoJ56gCAPzKz7/mtLS0VKtXr1ZoaKhiYmIsqgqAL3DoHwDgV5544gmdO3dOCQkJunDhgrZt26bPPvtMDz30UI3fkguAb3HoHwDgVzZt2qQVK1bou+++0/nz59WmTRulpKRo1KhRVpcGoJoRVAEAAGAkzlEFAACAkQiqAAAAMBJBFQAAAEaqdVf9u1wuOZ2cdusv7HYb/apj6HndQ8/rJvpe93jbc7vd5tW3yEm1MKg6nS798MNZq8uAF+rVsys8vKHy8wtVUuK0uhzUAHpe99Dzuom+1z2V6XmTJg0VEOBdUOXQPwAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJFq3Veo1iZ2u012u3dfMeZ0unz2ncqm1AEAAOoWgqqh7HabwsJCFBDg3U7v0lKncnMLqz0kmlIHAACoewiqhrLbbQoIsGve2j3Kyim45NyIFo00bWQ32e02nwRVE+oAAAB1D0HVcFk5BTqanWd1GcbUAQAA6g4upgIAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAknwTV7777TjNnztStt96qTp06aciQIRXO++tf/6oBAwYoLi5Ow4YN03vvveeLcgAAAOCHfBJUjxw5oh07dqhNmzZq165dhXO2bNmiRx99VAMHDtTy5csVHx+viRMn6vPPP/dFSQAAAPAz9Xyx0uTkZPXv31+SlJaWpv3795ebM3/+fA0ePFgPPvigJKlnz546fPiwFi5cqOXLl/uiLAAAAPgRn+xRtdsvvdrjx4/r22+/1cCBAz3GBw0apF27dunChQu+KAsAAAB+xCd7VC8nMzNTktS2bVuP8Xbt2qm4uFjHjx+/6CkD3qhXz/+vEQsIqPxrqMpzfLHOwMAAr55ns9mqvA34p7Je0/O6g57XTfS97vFVzy0Jqnl5eZIkh8PhMV72uGx5VdjtNoWHN6x6cX7M4Qi2dPthjerL6XQpNLSBV/OdTpck6+tGzaPndQ89r5voe91T3T23JKj6ktPpUn5+odVlXLGAAHulm52fX6TSUqdldYQGB8put2ne2j3Kyim45NyIFo00bWQ3Sb6pG2Yq+zzR87qDntdN9L3uqUzPHY5gr/e8WhJUGzduLEkqKChQs2bN3OP5+fkey6uqpKRu/lCUljqNeO1ZOQU6mu39XnFT6kbNoed1Dz2vm+h73VPdPbfk5JHIyEhJ/z5XtUxmZqYCAwPVqlUrK8oCAACAQSwJqq1atdI111yjrVu3eoxnZGQoKSlJQUFBVpQFAAAAg/jk0H9RUZF27NghScrOztaZM2fcobRHjx5q0qSJJk2apGnTpql169ZKTExURkaG9u3bpzVr1viiJAAAAPgZnwTVU6dOacqUKR5jZY9XrVqlxMREDRkyREVFRVq+fLmWLVumtm3bKj09XQkJCb4oCQAAAH7GJ0E1IiJChw4duuy84cOHa/jw4b4oAQAAAH6OO/ECAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwUj2rC0DNs9ttstttXs0NCOBvGQAAYA2Cah1jt9sUFhZCAAUAAMYjqNYxdrtNAQF2zVu7R1k5BZed37VDc6UO6lQDlQEAAHgiqNZRWTkFOpqdd9l5Ec1Da6AaAACA8jj+CwAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEiWBtV33nlHw4cPV0JCgnr16qUpU6bo+PHjVpYEAAAAQ1gWVHfv3q2JEyeqffv2WrhwoWbMmKGvvvpK9913n86dO2dVWQAAADBEPas2vGXLFl199dV66qmnZLPZJElNmjTRPffco/3796t79+5WlQYAAAADWLZHtaSkRA0bNnSHVElq1KiRJMnlcllVFgAAAAxh2R7V22+/Xa+//rrWrl2rYcOGKTc3V88++6w6deqkrl27XtG669Xz/2vEAgIq/xoCAwMu+zy73XbJ5VaoymuFfyrrNT2vO+h53UTf6x5f9dyyoNq9e3elp6fr4Ycf1uOPPy5J6tixo1544QUFBARUeb12u03h4Q2rq0y/ENaovpxOl0JDG1hdSpU4HMFWl4AaRs/rHnpeN9H3uqe6e25ZUP3000/1n//5n7rzzjvVt29f5ebmatGiRRo7dqzWrVunBg2qFrqcTpfy8wurudqaFxBg97rZocGBstttmrd2j7JyCi45t2uH5kod1Kk6Sqw2+flFKi11Wl0GakDZ55qe1x30vG6i73VPZXrucAR7vefVsqA6Z84c9ezZU2lpae6x+Ph49e3bV6+//rruuuuuKq+7pKRu/lBk5RToaHbeJedENA+toWq8V1rqrLM9q6voed1Dz+sm+l73VHfPLTt55OjRo+rQoYPH2FVXXaXw8HAdO3bMoqoAAABgCsuC6tVXX60vv/zSYyw7O1unT59Wy5YtLaoKAAAAprAsqI4YMULbt2/XnDlz9NFHHykjI0Pjx49X06ZNNXDgQKvKAgAAgCEsO0c1NTVVQUFBWr9+vV577TU1bNhQ8fHxeu655xQeHm5VWQAAADCEZUHVZrMpJSVFKSkpVpUAAAAAg3EnXgAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEaqZ3UBdY3dbpPdbrvsvICAuvM3hDev1el0yel01UA1AADAFATVGmS32xQWFlKnQujFhDWqL6fTJbvdJocj+LLzS0udys0tJKwCAFCHEFRrkN1uU0CAXfPW7lFWTsEl53bt0FypgzrVUGU1LzQ4UHa7zav3IqJFI00b2U12u42gCgBAHUJQtUBWToGOZuddck5E89AaqsZa3rwXAACgbuIYNAAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACNZHlQ3btyo2267TXFxcUpMTNQDDzygc+fOWV0WAAAALFbPyo0vXrxYy5cv1/jx4xUfH6/Tp09r165dKi0ttbIsAAAAGMCyoJqZman09HQtWrRIffr0cY8PGDDAqpIAAABgEMsO/W/YsEEREREeIRUAAAAoY1lQ3bt3r6KiorRo0SIlJSUpNjZWI0aM0N69e60qCQAAAAax7ND/999/r/379+vw4cOaNWuWgoODtWTJEt13333atm2bmjZtWuV116tn+TViFQoIMLMuf8H75//Kekgv6w56XjfR97rHVz23LKi6XC4VFhbq+eefV4cOHSRJXbp0UXJystasWaMpU6ZUab12u03h4Q2rs1QYwuEItroEVBN6WffQ87qJvtc91d1zy4Kqw+FQWFiYO6RKUlhYmDp16qSvv/66yut1Ol3Kzy+sjhKrXUCAnR/aK5CfX6TSUqfVZeAKlP0M0Mu6g57XTfS97qlMzx2OYK/3vFoWVNu3b69jx45VuOz8+fNXtO6SEn4oaqPSUie9rSXoZd1Dz+sm+l73VHfPLTt5pF+/fsrNzdXBgwfdY6dPn9aBAwcUExNjVVkAAAAwhGV7VPv376+4uDhNnjxZU6dOVf369bVs2TIFBQXp7rvvtqosAAAAGMKyPap2u13Lli1TfHy8Zs6cqYceekihoaFau3atmjVrZlVZAAAAMISlX6HapEkT/fGPf7SyBAAAABiKG5wBAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwUj2rCwDgv+x2m+x2m1dznU6Xj6sBANQ2BFUAVWK32xQWFqKAAO8OzJSWOlVQcM7HVQEAahOCKoAqsdttCgiwa97aPcrKKbjk3IgWjTRtZDev974CACARVAFcoaycAh3NzrO6DABALcTFVAAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxkTFA9e/asevfurejoaH3xxRdWlwMAAACLGRNUFy1apNLSUqvLAAAAgCGMCKpHjx7VunXrNGnSJKtLAQAAgCGMCKpz5szRiBEj1LZtW6tLAQAAgCEsD6pbt27V4cOHNWHCBKtLAQAAgEHqWbnxoqIiPf3005o6dapCQ0Orbb316lmevysUEGBmXf6C988sVemH3W6r8nPhn8p6Tc/rFvpe9/iq55YG1cWLF6tp06a64447qm2ddrtN4eENq219MIfDEWx1CbhCoaENJNHLuoie1030ve6p7p5bFlSzs7P10ksvaeHChSooKJAkFRYWuv999uxZNWxY+cDpdLqUn19YrbVWl4AAOz+0VyA/v0ilpU6ry8D/V5XP85kz5xQa2oBe1iFlnxN6XrfQ97qnMj13OIK93vNqWVDNyspScXGxxo4dW25ZamqqunTpoldffbVK6y4p4YeiNiotddJbP+d0uiTRy7qIntdN9L3uqe6eWxZUO3bsqFWrVnmMHTx4UHPnztXs2bMVFxdnUWUAAAAwgWVB1eFwKDExscJlMTExiomJqeGKAAAAYBIuxwMAAICRLL3q/+cSExN16NAhq8sAAACAAdijCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEj1rC7A39ntNtntNq/mBgTwdwGqR2U+d06nS06ny8cVwRSV+WxIfD4AmI2gegXsdpvCwkIIoKhRlf3clZY6lZtbSBipA6ryO4nPBwCTEVSvgN1uU0CAXfPW7lFWTsFl53ft0FypgzrVQGWozSrzuYto0UjTRnaT3W4jiNQBlf2dxOcDgOkIqtUgK6dAR7PzLjsvonloDVSDusLbzx3qHj4bAGoLjlkDAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwUj2rNvzmm2/qjTfe0IEDB5Sfn682bdpo9OjRuuOOO2Sz2awqCwAAAIawLKiuXLlSLVu2VFpamsLDw/XRRx/p0Ucf1cmTJzVx4kSrygIAAIAhLAuqixcvVpMmTdyPk5KSlJubqxUrVuh3v/ud7HbOSgAAAKjLLEuDPw2pZTp27KgzZ86osLDQgooAAABgEqN2W+7Zs0ctWrRQaGio1aUAAADAYpYd+v+5Tz75RBkZGZo+ffoVr6tevZrJ3wEBRuX8Wi8wMMDr99zpdMnlcvm4ImtU5XPni89qVdZZ9rN5uV76sn82m012u3cXbFa2Dl+u2xtV7bMvf5eVrbuu/760+rNR0+h73eOrnhsRVE+ePKmpU6cqMTFRqampV7Quu92m8PCG1VQZTBDWqL6cTpdCQxt4/Ryn0+X1/xTqAocj2NLtl/UwJKS+JF22l77sX2XWXdk6fLluX6qJz4fVn0Gr+etn40rV9b7XRdXdc8uDan5+vsaMGaOwsDAtWLDgii+icjpdys+vmXNcAwLs/BDWgNDgQNntNs1bu0dZOQWXnR/RopGmjeym/PwilZY6a6DCmlWVz50v3ovK1FGZHvqyf2U1+6IOX67bW1X9neTLn5Wymmrrz6M3TPhs1DT6XvdUpucOR7DXe14tDarnzp3TuHHjVFBQoFdeeUWNGjWqlvWWlPBDURtl5RToaHae1/NLS518Fv4/U96LyvTQlzX7sg5TXmNl1EQdprxWK/njZ+NK1ZbXAe9Vd88tC6olJSV68MEHlZmZqbVr16pFixZWlQIAAAADWRZUZ8+erffee09paWk6c+aMPv/8c/eyTp06KSgoyKrSAAAAYADLgurOnTslSU8//XS5Ze+8844iIiJquiQAAAAYxLKg+u6771q1aQAAAPgBbnAGAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASPWsLgDwlYAA7/4OczpdcjpdXs21222y221e1+CLdXv7uqr6nMrUjCtTlV7W5PoAE1Tm964pv79MqdmUOq4EQRW1Tlij+nI6XXI4gr2aX1rqVG5u4WV/QO12m8LCQioVBny5bm9U9r2QvK8ZVVfpz6jTpYBK/IEE1BaV/d1owu8vU2o2pY4rRVBFrRMaHCi73aZ5a/coK6fgknMjWjTStJHdZLfbvAqTAQF2r9bry3V37dBcqYM6XXb7UuXei8rWjKqrTF/K+l3dnw3AH1Tmd6Mpv79MqdmUOq4UQRW1VlZOgY5m5/nNer1dd0TzUJ+sFzWvMv321WcD8Af++DvMlJpNqaOqOKEJAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkSwNqkePHtW9996r+Ph4XX/99frv//5vXbhwwcqSAAAAYIh6Vm04Ly9P99xzj6655hotWLBAOTk5evrpp3Xu3DnNnDnTqrIAAABgCMuC6ssvv6yzZ88qPT1dYWFhkqTS0lLNnj1b48aNU4sWLawqDQAAAAaw7ND/Bx98oKSkJHdIlaSBAwfK6XRq586dVpUFAAAAQ1gWVDMzMxUZGekx5nA41KxZM2VmZlpUFQAAAExhc7lcLis2HBMToylTpmjs2LEe40OGDFFCQoKeeOKJKq3X5XLJ6ayZl2SzSXa7XbkF51VS6rzs/PpBAWoUEuTVfBPmmlKHL2uuF2BXWKP6cjovv16pcv321bp9+d6ZUHNla6gsb2uuSh1W99DX/a4qu93u822YzpefO1NVV9/98b0zpeaq1FHVVFiWibxZh91uk81m82q9lp2j6is2m00BAd69+OoS1qi+z+abMNeUOnxZs93u/cGFytbhq3X78r0zoebK1FBZvqzDhB76st9VVRPbMJ0pn/+aVF2vwx/fO1Nqruk6qvu1WNZNh8OhgoKCcuN5eXlq3LixBRUBAADAJJYF1cjIyHLnohYUFOj7778vd+4qAAAA6h7Lgmrv3r310UcfKT8/3z22detW2e12XX/99VaVBQAAAENYdjFVXl6eBg8erLZt22rcuHHuG/4PHTqUG/4DAADAuqAq/fgVqk888YQ+++wzNWzYULfeequmTp2qoKAgq0oCAACAISwNqgAAAMDFmHEPBwAAAOBnCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjEVRRo3bs2KFRo0apZ8+eio2N1Y033qi5c+eqoKDA6tJQQ86ePavevXsrOjpaX3zxhdXlwEc2bNig6Ojocv/MmzfP6tLgYxs3btRtt92muLg4JSYm6oEHHtC5c+esLgs+MHr06Ap/zqOjo7Vly5Zq2Ua9alkL4KXc3Fx17txZo0ePVlhYmI4cOaIFCxboyJEjeumll6wuDzVg0aJFKi0ttboM1JAXXnhBjRo1cj9u0aKFhdXA1xYvXqzly5dr/Pjxio+P1+nTp7Vr1y5+5mupWbNm6cyZMx5jf/nLX7Rt2zYlJSVVyzYIqqhRt956q8fjxMREBQUF6dFHH1VOTg7/E6vljh49qnXr1mn69OmaNWuW1eWgBsTExKhJkyZWl4EakJmZqfT0dC1atEh9+vRxjw8YMMDCquBL7du3Lzf28MMP6/rrr6+2n3sO/cNyYWFhkqTi4mJrC4HPzZkzRyNGjFDbtm2tLgVANduwYYMiIiI8Qirqlk8//VRZWVkaOnRota2ToApLlJaW6vz58zpw4IAWLlyo5ORkRUREWF0WfGjr1q06fPiwJkyYYHUpqEFDhgxRx44ddeONN2rp0qUcAq7F9u7dq6ioKC1atEhJSUmKjY3ViBEjtHfvXqtLQw3ZvHmzQkJCdOONN1bbOjn0D0v069dPOTk5kqQbbrhBf/rTnyyuCL5UVFSkp59+WlOnTlVoaKjV5aAGNGvWTJMmTVKXLl1ks9n07rvv6rnnnlNOTo5mzpxpdXnwge+//1779+/X4cOHNWvWLAUHB2vJkiW67777tG3bNjVt2tTqEuFDJSUlevPNN5WcnKyQkJBqWy9BFZZYtmyZioqK9PXXX2vx4sUaP368VqxYoYCAAKtLgw8sXrxYTZs21R133GF1KaghN9xwg2644Qb34169eql+/fr6y1/+ovHjx6t58+YWVgdfcLlcKiws1PPPP68OHTpIkrp06aLk5GStWbNGU6ZMsbhC+NLOnTv1ww8/aMiQIdW6Xg79wxIdOnRQQkKChg8frkWLFmn37t16++23rS4LPpCdna2XXnpJkydPVkFBgfLz81VYWChJKiws1NmzZy2uEDVl4MCBKi0t1cGDB60uBT7gcDgUFhbmDqnSj9cgdOrUSV9//bWFlaEmbN68WWFhYerVq1e1rpc9qrBcdHS0AgMDdezYMatLgQ9kZWWpuLhYY8eOLbcsNTVVXbp00auvvmpBZQCqU/v27S/6e/z8+fM1XA1q0rlz57R9+3YNGzZMgYGB1bpugiost3fvXhUXF3MxVS3VsWNHrVq1ymPs4MGDmjt3rmbPnq24uDiLKkNNy8jIUEBAgDp16mR1KfCBfv36acOGDTp48KA6duwoSTp9+rQOHDig3/zmN9YWB5969913VVhYWK1X+5chqKJGTZw4UbGxsYqOjlaDBg301Vdf6cUXX1R0dLT69+9vdXnwAYfDocTExAqXxcTEKCYmpoYrQk24//77lZiYqOjoaEnSO++8o1dffVWpqalq1qyZxdXBF/r376+4uDhNnjxZU6dOVf369bVs2TIFBQXp7rvvtro8+NCmTZt09dVXq1u3btW+boIqalTnzp2VkZGhZcuWyeVyqWXLlho+fLjuv/9+BQUFWV0egGrStm1bvfbaazp58qScTqeuueYazZgxQ6NHj7a6NPiI3W7XsmXLNHfuXM2cOVPFxcXq3r271q5dyx8ntVheXp4+/PBD3XPPPbLZbNW+fpvL5XJV+1oBAACAK8RV/wAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioA1ELR0dFasGCB1WUAwBXhhv8AqlXZNxFdzqpVqy76jVXeKioq0gsvvKAePXp4ta7du3crNTX1osufffZZDR48+Ipq8oUFCxYoPT1du3btUpMmTcotT05O1q9+9SstXbr0irazadMmnTp1iq+7BGAMgiqAavXf//3fHo9ff/117dy5s9x4u3btrnhbRUVFSk9P18SJEysVekePHq24uLhy4/Hx8Vdckyn27dungICASj1n8+bNOnLkCEEVgDEIqgCq1a233urxeO/evdq5c2e5cSt1795dt9xyS6We43Q6VVxcrPr165dbVlhYqJCQkCuqqaioSMHBwVe0jp+qqE7TVcf7CKB24RxVADXO6XRq5cqVGjx4sOLi4nTddddp5syZysvL85j3xRdf6P7771diYqI6d+6s5ORkPfLII5KkrKwsJSUlSZLS09MVHR1dredlRkdH6/HHH9cbb7zhrvPDDz/Uhg0bFB0drX/84x967LHHlJSUpD59+rift3btWg0ePFixsbHq1auXZs+erfz8fI91jx49WkOGDNH+/fs1cuRIdenSRc8++2y11P3T+n/6Xpw5c0ZPPvmkkpOTFRsbq6SkJN177706cOCAu6b3339f2dnZ7vcyOTnZ/fxTp05pxowZuu666xQXF6dhw4Zp48aN5bZ7+vRp/f73v1fXrl3VvXt3TZ8+XV999ZWio6O1YcMG97y0tDQlJCTo2LFjGjNmjBISEjRt2jRJ0ieffKLJkyerb9++io2NVZ8+ffTUU0/p3LlzHtsqW8eJEyc0btw4JSQk6IYbbtDatWslSYcOHVJqaqri4+PVr18/bdq0qfreYAA1gj2qAGrczJkztXHjRt1+++0aPXq0srKytHbtWn355Zdav369AgMDderUKd1///0KDw/X2LFj5XA4lJWVpbfffluS1KRJEz322GN67LHHdNNNN+mmm26S5N05smfPntUPP/xQbjw8PFw2m839+OOPP9abb76pkSNHKjw8XC1btnSHztmzZ6tJkyaaMGGCCgsLJf37XNLrrrtOKSkp+uabb7R+/Xp98cUX7tdVJjc3V2PGjNHgwYM1bNgwNW3a9LJ1/zzIl3E6nZd97qxZs/TWW29p1KhRateunXJzc7Vnzx4dPXpUMTExGj9+vAoKCnTy5En3HwMNGzaUJJ07d06jR4/WsWPHNHLkSEVERGjr1q1KS0tTfn6+7rnnHncdv/3tb7Vv3z6lpKQoMjJS77zzjqZPn15hTSUlJbr//vvVrVs3TZ8+XQ0aNJAkbd26VefOnVNKSorCwsK0b98+rVmzRidPntT8+fM91lFaWqoxY8aoe/fumjZtmjZt2qTHH39cwcHB+vOf/6yhQ4fq5ptv1ssvv6zp06crPj5erVq1uuz7BcAQLgDwodmzZ7uioqLcj//5z3+6oqKiXG+88YbHvA8++MBj/O2333ZFRUW59u3bd9F1nzp1yhUVFeWaP3++V7V8/PHHrqioqIv+869//cs9NyoqytWhQwfXkSNHPNbx2muvuaKiolwpKSmukpISj1piYmJc9913n6u0tNQ9vmbNGldUVJTrb3/7m3ts1KhRrqioKNf69eu9qnv+/PmXrDsqKso1duxYj+f8/H3p1q2ba/bs2ZfcztixY139+vUrN75y5UpXVFSU6/XXX3ePXbhwwXXXXXe54uPjXQUFBS6Xy+V66623XFFRUa6VK1e655WWlrpSU1NdUVFRrtdee809Pn36dFdUVJRr3rx55bZXVFRUbmzp0qWu6OhoV3Z2drl1LFmyxD2Wl5fn6ty5sys6Otq1ZcsW9/jRo0cr9VkBYAb2qAKoUVu3blWjRo10/fXXe+zVjImJUUhIiHbv3q2hQ4eqUaNGkqT3339fHTp08NgbeaUmTJig7t27lxtv3Lixx+Nrr71W7du3r3Add955p8fFSh999JGKi4uVmpoqu/3fZ1UNHz5czz77rHbs2KE77rjDPR4UFKTbb7+9UnUvWLBAoaGh5cZ///vfX/a5DodDe/fuVU5Ojlq0aFGp7X7wwQdq1qyZhgwZ4h4LDAzU6NGj9dBDD+mf//yn+vXrpw8//FCBgYG688473fPsdrtGjhypjz/+uMJ1p6SklBsr27Mq/Xje6rlz55SQkCCXy6Uvv/xSV199tcf84cOHe7zOtm3b6rvvvtPAgQPd45GRkXI4HDp+/HilXjsAaxFUAdSo7777TgUFBe7zS3/u1KlTkqQePXpowIABSk9P18qVK9WjRw/1799fQ4cOVVBQ0BXVEBUVpeuuu+6y8yIiIrxeduLECUk/BqKfCgoKUqtWrZSdne0x3qJFi0q/ju7du1d4eypvLpyaNm2a0tLS1LdvX8XExKhPnz667bbbvDoMnp2drTZt2ngEcOnfd24oe+0nTpxQs2bNyl0U1rp16wrXW69ePV111VXlxk+cOKH58+fr3XffLXe6w5kzZzwe169fv9x70qhRI1111VUep3GUjf/8fGEAZiOoAqhRTqdTTZs21bx58ypcXhY6bDab5s+fr88//1zvvfeePvzwQ82YMUMrVqzQK6+84j5/0pd+umfv5670qvpLrdsXBg0apO7du+vtt9/Wzp079eKLL2r58uVasGCBx8VgNSkoKKhc+C0tLdW9996rvLw8PfDAA4qMjFRISIhycnKUlpZW7nzci92C62LjLpereooHUCMIqgBqVOvWrbVr1y517drVq7AWHx+v+Ph4TZ06VZs2bdK0adOUkZGh4cOHl9tjZqWyw9GZmZkeeykvXLigrKwsr/bg+lrz5s01cuRIjRw5UqdOndKvf/1rLVmyxB1UL/Z+tmzZUocOHZLT6fQIlpmZmZL+/dqvvvpq7d69u9ytto4dO+Z1jYcPH9a3336rZ555Rrfddpt7fOfOnV6vA0Dtwe2pANSogQMHqrS0VIsWLSq3rKSkxH1oNi8vr9zer44dO0r6MfxJcochEw7nXnfddQoMDNTq1as96v7b3/6mgoICy/ZaSj/upSwoKPAYa9q0qZo3b+5+L6Uf38+fz5Ok3r176/vvv1dGRoZ7rKSkRKtXr1ZISIiuvfZaSVKvXr1UXFysV1991T3P6XS6bxfljbIg/NP30OVyadWqVV6vA0DtwR5VADWqR48euuuuu7R06VIdPHhQ119/vQIDA/Xtt99q69at+sMf/qBbbrlFGzdu1Pr169W/f3+1bt1aZ8+e1auvvqrQ0FD17t1b0o+Hz9u3b68333xT11xzjcLCwvSrX/1KUVFRl6zhk08+0fnz58uNR0dHq0OHDlV6XU2aNNG4ceOUnp6uBx54QMnJyfrmm2+0bt06931HrXL27Fn16dNHAwYMUIcOHRQSEqKPPvpIX3zxhdLS0tzzYmJilJGRoblz5youLk4hISFKTk7WXXfdpVdeeUVpaWk6cOCAWrZsqbfeekuffvqpZsyY4b7Aq3///urcubOeeeYZHTt2TJGRkR7nmXqzBzwyMlKtW7fWM888o5ycHIWGhuqtt94y4o8RADWPoAqgxj3++OOKjY3Vyy+/rD//+c8KCAhQy5YtNWzYMHXt2lXSj4H2iy++UEZGhv7v//5PjRo1UufOnTVv3jyPQ+tz5szRE088oblz56q4uFgTJ068bFBdvXp1heMTJ06sclCVpEmTJqlJkyZas2aN5s6dq8aNG+vOO+/UQw89VK13LaisBg0aKCUlRTt37tS2bdvkcrnUunVrzZo1S3fffbd73t13362DBw9qw4YNWrlypVq2bKnk5GQ1aNBAq1ev1rx587Rx40adOXNGbdu21dy5cz3uXBAQEKClS5fqySef1MaNG2W323XTTTdpwoQJSklJ8eq83sDAQC1ZskRz5szR0qVLVb9+fd10000aOXKkUd9uBqBm2FycWQ4A8KHt27drwoQJWrdunbp162Z1OQD8COeoAgCqzc+/5rS0tFSrV69WaGioYmJiLKoKgL/i0D8AoNo88cQT7hv0X7hwQdu2bdNnn32mhx56qMZvyQXA/3HoHwBQbTZt2qQVK1bou+++0/nz59WmTRulpKRo1KhRVpcGwA8RVAEAAGAkzlEFAACAkQiqAAAAMBJBFQAAAEaqdVf9u1wuOZ2cdlsb2O02elnL0ePajf7WbvS3dvNlf+12m1ffVCfVwqDqdLr0ww9nrS4DV6hePbvCwxsqP79QJSVOq8uBD9Dj2o3+1m70t3bzdX+bNGmogADvgiqH/gEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxU675CFday222y2737WjSn08X3RAMAgIsiqKLa2O02hYWFKCDAux31paVO5eYWElYBAECFCKqoNna7TQEBds1bu0dZOQWXnBvRopGmjewmu91GUAUAABUiqKLaZeUU6Gh2ntVlAAAAP8fFVAAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkXwSVL/77jvNnDlTt956qzp16qQhQ4Zccv727dsVHR192XkAAACoO3zyzVRHjhzRjh071KVLFzmdTrlcF/+KzHPnzumpp57SL37xC1+UAgAAAD/lkz2qycnJ2rFjh+bPn6+YmJhLzl26dKmuvvpq3XDDDb4oBQAAAH7KJ0HVbvdutceOHdOKFSv0X//1X74oAwAAAH7M0oupnnzySd16663q0KGDlWUAAADAQD45R9Ub7777rj777DNt3bq12tddrx43M7BCQEDl3/eLPadsvCrrhH+gx7Ub/a3d6G/tZlJ/LQmq58+f11NPPaVJkyapSZMm1bpuu92m8PCG1bpO+I7DEXxFy+H/6HHtRn9rN/pbu5nQX0uC6l/+8hfZ7XYNHjxY+fn5kqTi4mI5nU7l5+erQYMGCgoKqtK6nU6X8vMLq7NceCkgwF7pD3V+fpFKS50XXdfFlsP/0ePajf7WbvS3dvN1fx2OYK/31loSVDMzM/Xdd98pKSmp3LJrr71Wjz32mFJSUqq8/pISfmj8RWmp85L9utxy+D96XLvR39qN/tZuJvTXkqA6ZswY/frXv/YYW7Zsmb755hvNnTtX11xzjRVlAQAAwCA+CapFRUXasWOHJCk7O1tnzpxxXzTVo0cPtWvXTu3atfN4zsaNG5WTk6PExERflAQAAAA/45OgeurUKU2ZMsVjrOzxqlWrCKMAAAC4LJ8E1YiICB06dKhSz3n66ad9UQoAAAD8lPU3yAIAAAAqQFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIzkk2+mQu1it9tkt9suOy8ggL97AABA9SGo4pLsdpvCwkIIoQAAoMYRVHFJdrtNAQF2zVu7R1k5BZec27VDc6UO6lRDlQEAgNqOoAqvZOUU6Gh23iXnRDQPraFqAABAXcDxXAAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGMkn30z13Xff6cUXX9TevXt15MgRRUZGavPmze7lZ86c0YoVK7Rjxw59++23CgoKUufOnTV16lRFR0f7oiQAAAD4GZ/sUT1y5Ih27NihNm3aqF27duWWnzhxQq+88oquv/56Pffcc3riiSdUUFCgu+66S0ePHvVFSQAAAPAzPtmjmpycrP79+0uS0tLStH//fo/lERERevvttxUcHOwe69mzp5KTk7Vu3To9+uijvigLAAAAfsQnQdVuv/SO2pCQkHJjDRs2VOvWrfWvf/3LFyUBAADAzxhzMVV+fr77fFYAAADAJ3tUq+KPf/yjbDabUlJSrnhd9eoZk7/9XkCAb9/LwMCACrdht9s8ljudLrlcLp/WgppV1ndff8ZgDfpbu9Hf2s2k/hoRVF977TW9+uqrevrpp3XVVVdd0brsdpvCwxtWU2XwlbBG9eV0uhQa2uCS88qWO50ud3hF7eJwBF9+EvwW/a3d6G/tZkJ/LQ+qO3bs0MyZM/W73/1Ov/71r694fU6nS/n5hdVQGaQf/5ryxQc1NDhQdrtN89buUVZOwSXnRrRopGkjuyk/v0ilpc5qrwXWKPts0dfaif7WbvS3dvN1fx2OYK/31loaVD///HNNmTJFt912m6ZMmVJt6y0p4YfGX2TlFOhodp5Xc0tLnfS2FqKvtRv9rd3ob+1mQn8tO/ng66+/1rhx49SzZ0/Nnj3bqjIAAABgKJ/sUS0qKtKOHTskSdnZ2Tpz5oy2bt0qSerRo4dcLpfuv/9+1a9fX/fcc4/HfVZDQ0PVvn17X5QFAAAAP+KToHrq1Klyh/LLHq9atUqSdPLkSUnSb37zG495PXr00OrVq31RFgAAAPyIT4JqRESEDh06dMk5l1sOAACAus36G2QBAAAAFSCoAgAAwEgEVQAAABiJoAoAAAAjEVQBAABgJIIqAAAAjERQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSfBNXvvvtOM2fO1K233qpOnTppyJAhFc7761//qgEDBiguLk7Dhg3Te++954tyAAAA4Id8ElSPHDmiHTt2qE2bNmrXrl2Fc7Zs2aJHH31UAwcO1PLlyxUfH6+JEyfq888/90VJAAAA8DP1fLHS5ORk9e/fX5KUlpam/fv3l5szf/58DR48WA8++KAkqWfPnjp8+LAWLlyo5cuX+6IsAAAA+BGf7FG12y+92uPHj+vbb7/VwIEDPcYHDRqkXbt26cKFC74oCwAAAH7EkoupMjMzJUlt27b1GG/Xrp2Ki4t1/PhxK8oCAACAQXxy6P9y8vLyJEkOh8NjvOxx2fKqqlePmxlcis1mk91u82qut/NqQkAAfa1NyvpJX2sn+lu70d/azaT+WhJUfclutyk8vKHVZRjN6XQZFUC95XAEW10CfIC+1m70t3ajv7WbCf21JKg2btxYklRQUKBmzZq5x/Pz8z2WV4XT6VJ+fuGVFViLBQTY5XAEa97aPcrKKbjs/K4dmit1UKcaqOzy8vOLVFrqtLoMVJOyzyJ9rZ3ob+1Gf2s3X/fX4Qj2em+tJUE1MjJS0o/nqpb9d9njwMBAtWrV6orWX1LCD83lZOUU6Gj25U+xiGgeWgPVeKe01ElvayH6WrvR39qN/tZuJvTXkpMPWrVqpWuuuUZbt271GM/IyFBSUpKCgoKsKAsAAAAG8cke1aKiIu3YsUOSlJ2drTNnzrhDaY8ePdSkSRNNmjRJ06ZNU+vWrZWYmKiMjAzt27dPa9as8UVJAAAA8DM+CaqnTp3SlClTPMbKHq9atUqJiYkaMmSIioqKtHz5ci1btkxt27ZVenq6EhISfFESAAAA/IxPgmpERIQOHTp02XnDhw/X8OHDfVECAAAA/Jz1N8gCAAAAKkBQBQAAgJEIqgAAADASQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABjJ0qD6zjvvaPjw4UpISFCvXr00ZcoUHT9+3MqSAAAAYAjLguru3bs1ceJEtW/fXgsXLtSMGTP01Vdf6b777tO5c+esKgsAAACGqGfVhrds2aKrr75aTz31lGw2mySpSZMmuueee7R//351797dqtIAAABgAMv2qJaUlKhhw4bukCpJjRo1kiS5XC6rygIAAIAhLAuqt99+u44ePaq1a9eqoKBAx48f17PPPqtOnTqpa9euVpUFAAAAQ1h26L979+5KT0/Xww8/rMcff1yS1LFjR73wwgsKCAi4onXXq8fNDC4mIMB/3xt/rh3llfWTvtZO9Ld2o7+1m0n9tSyofvrpp/rP//xP3Xnnnerbt69yc3O1aNEijR07VuvWrVODBg2qtF673abw8IbVXC1M4HAEW10CfIC+1m70t3ajv7WbCf21LKjOmTNHPXv2VFpamnssPj5effv21euvv6677rqrSut1Ol3Kzy+srjJrnYAAuxEfvKrIzy9SaanT6jJQTco+i/S1dqK/tRv9rd183V+HI9jrvbWWBdWjR4/qxhtv9Bi76qqrFB4ermPHjl3RuktK+KGpjUpLnfS2FqKvtRv9rd3ob+1mQn8tO/ng6quv1pdffukxlp2drdOnT6tly5YWVQUAAABTWBZUR4wYoe3bt2vOnDn66KOPlJGRofHjx6tp06YaOHCgVWUBAADAEJYd+k9NTVVQUJDWr1+v1157TQ0bNlR8fLyee+45hYeHW1UWAAAADGFZULXZbEpJSVFKSopVJQAAAMBg1t8gCwAAAKgAQRUAAABGIqgCAADASARVAAAAGImgCgAAACMRVAEAAGAkgioAAACMRFAFAACAkQiqAAAAMBJBFQAAAEYiqAIAAMBIBFUAAAAYiaAKAAAAIxFUAQAAYCSCKgAAAIxEUAUAAICRCKoAAAAwEkEVAAAARiKoAgAAwEgEVQAAABiJoAoAAAAjWR5UN27cqNtuu01xcXFKTEzUAw88oHPnzlldFgAAACxWz8qNL168WMuXL9f48eMVHx+v06dPa9euXSotLbWyLAAAABjAsqCamZmp9PR0LVq0SH369HGPDxgwwKqSAAAAYBDLDv1v2LBBERERHiEVAAAAKGNZUN27d6+ioqK0aNEiJSUlKTY2ViNGjNDevXutKgkAAAAGsezQ//fff6/9+/fr8OHDmjVrloKDg7VkyRLdd9992rZtm5o2bVrldderZ/k1YsYKCPDf98afa0d5Zf2kr7UT/a3d6G/tZlJ/LQuqLpdLhYWFev7559WhQwdJUpcuXZScnKw1a9ZoypQpVVqv3W5TeHjD6iwVhnA4gq0uAT5AX2s3+lu70d/azYT+WhZUHQ6HwsLC3CFVksLCwtTp/7V3/6FVlv8fx1/nPrqlbqejMgc5yIl03OacZrCGP8IfEJZmIIJYm5Wlkr9aBUrgZChslEGhbrXKPkqTEjIoU7E0KjSCrMTGSt3+kAkOIbcd98O2c873j5jfTNP7bOfede34fEDEzu7uvc71vu6zV+fsbLm5On/+fJ/PG43G1NbWkYiIScnvd6zYeH3R1tapSCRqOgYSpHcvMtfkxHyTG/NNbl7PNxAY5vrZWmNFdcKECbpw4cItP3ft2rV+nbunh4smGUUiUWabhJhrcmO+yY35Jjcb5mvshw9mz56tlpYW1dfXX7/typUrqqurU15enqlYAAAAsISxZ1TnzZun/Px8rV+/XqWlpUpNTVVNTY1SUlK0bNkyU7EAAABgCWPPqDqOo5qaGk2ZMkVlZWV6+eWXlZaWptraWmVkZJiKBQAAAEsY/ROqo0aN0htvvGEyAgAAACxl/hdkAQAAALdAUQUAAICVKKoAAACwEkUVAAAAVqKoAgAAwEoUVQAAAFiJogoAAAArUVQBAABgJYoqAAAArERRBQAAgJUoqgAAALASRRUAAABWoqgCAADAShRVAAAAWImiCgAAACtRVAEAAGAliioAAACsRFEFAACAlSiqAAAAsBJFFQAAAFaiqAIAAMBKFFUAAABYyZqi2t7erlmzZikUCunMmTOm4wAAAMAwa4pqVVWVIpGI6RgAAACwhBVFtaGhQfv27dO6detMRwEAAIAlrCiq27Zt09KlS5WdnW06CgAAACxhvKgeOXJEZ8+e1Zo1a0xHAQAAgEWGmPzinZ2dqqysVGlpqdLS0hJ23iFDjPdva/n9g3dtBnN2N3w+nxzH5+rYWEzyuTtU0WhMsVisH8m80TvPZJ/r3Yr5Jjfmm9xsmq/RolpdXa3Ro0dr8eLFCTun4/g0cuSIhJ0P9ggEhpmO4KloNOa6qEaiMfldHhvPeU1I9rne7ZhvcmO+yc2G+RorqhcvXtTu3bu1a9cuhcNhSVJHR8f1f7e3t2vEiPgLZzQaU1tbR0KzJhO/37Fi4/VFW1unIpGo6Rie6J3L9tpTamoO3/bYByeOUcljua6OzcpM16tPTbNy7Xrvs43Z0H/MN7kx3+Tm9XwDgWGun601VlSbmprU3d2tlStX3vS5kpISFRQUaP/+/X06d08PF00yikSiST/bpuawGi623vaYrDFpro/tZfPa2ZwN/cd8kxvzTW42zNdYUc3JydHevXtvuK2+vl4VFRUqLy9Xfn6+oWQAAACwgbGiGggEVFhYeMvP5eXlKS8vb4ATAQAAwCbm384FAAAA3ILRd/3/W2Fhof744w/TMQAAAGABnlEFAACAlSiqAAAAsBJFFQAAAFaiqAIAAMBKFFUAAABYiaIKAAAAK1FUAQAAYCWKKgAAAKxEUQUAAICVrPrLVIOR4/jkOD7Xx0ejMUWjMQ8TAQMnnv0fz97nugIASBTVfnEcn4LB4fL73T8xHYlE1dLSwTdVDHrx7n+3e5/rCgDQi6LaD47jk9/vaHvtKTU1h+94fFZmul59apocx8c3VAx68ez/ePY+1xUAoBdFNQGamsNquNhqOgZghFf7n+sKAMCbqQAAAGAliioAAACsRFEFAACAlSiqAAAAsBJFFQAAAFaiqAIAAMBKFFUAAABYiaIKAAAAKxn7hf+HDx/W559/rrq6OrW1ten+++9XcXGxFi9eLJ/P/d/4BgAAQHIyVlT/97//aezYsdq0aZNGjhypkydPavPmzbp06ZLWrl1rKhYAAAAsYayoVldXa9SoUdc/LioqUktLiz788EO9+OKLchx+KgEAAOBuZqwN/rOk9srJydHVq1fV0dFhIBEAAABsYtXTlqdOnVJmZqbS0tJMRwEAAIBhxl76/7effvpJhw4d0saNG/t9riFDBqZ/+/19+zpDh/pd/bexmOT2fWXRaEyxWOyOx/U1sw3crpvkfj1s4fVc4jm/z+eT49x547k55t/+OcPe//7fc+3LeaXBt7fdrrMU336O57zxnjsevfMYbHOBO8kwX1uuFRvZNF8riuqlS5dUWlqqwsJClZSU9OtcjuPTyJEjEpQssYLpqYpGY0pLu8fV8ZFoTP44vpH19Ru87eJdNym516MvAoFhro/1Yu1uN8N45no78dxHG8Szzl4d25fj4zXY5oL4DOb52nat2MiG+Rovqm1tbXrhhRcUDAa1Y8eOfr+JKhqNqa1tYH7G1e934hpi2rChchyftteeUlNz+LbHPjhxjEoey3V1bFZmul59apra2joViUQTmtkG8aybFN962MLrubhdi94c8exRN/qy9+M1GOft1fVtw7XSm2UwzQXuDfb52nSt2Mjr+QYCw1w/W2u0qHZ1dWnVqlUKh8P65JNPlJ6enpDz9vTYvYmamsNquNh622OyxqS5PrZXJBK1/r73RzxrISX/esQj3rWIZ4/Gw6vzSoNz3l5d3zZdK4NxLnBvsM/XpmvFRjbcX2NFtaenRy+99JIaGxtVW1urzMxMU1EAAABgIWNFtby8XN988402bdqkq1ev6tdff73+udzcXKWkpJiKBgAAAAsYK6onTpyQJFVWVt70uWPHjikrK2ugIwEAAMAixorq8ePHTX1pAAAADALmf0EWAAAAcAsUVQAAAFiJogoAAAArUVQBAABgJYoqAAAArERRBQAAgJUoqgAAALASRRUAAABWoqgCAADAShRVAAAAWMnYn1BF4vn9d/7/DjfHJAu39zUajSkajXmSwXF8chzfHY/zei5uz3837Q8veDlvG65vt/dPknw+d8fdLeJZu3gek+I5b7zn9oqXmW15zLWFV/tuIFFUk0AwPVXRaEyBwDDTUawQ73pEIlG1tHQk/AJ1HJ+CweFGHxDZGwPHq3nbMsN4718kEvU40eDRl7Vz85jUlz3n1eOdW15mtuEx1yZe7buBRlFNAmnDhspxfNpee0pNzeHbHvvgxDEqeSx3gJKZEc96ZGWm69WnpslxfJ4UVb/fMTqXeNbCyxx3A6/mbcv1Hc/9672u8Le+rJ2bx6R4zhvvub3iZWYbHnNt4tW+G2gU1STS1BxWw8XW2x6TNSZtgNKY52Y9bMnh9VzcrsXdtD+84tW8bdhHbnPg1rxau8E4Ey8z23Kt2GIw7o9/4vlxAAAAWImiCgAAACtRVAEAAGAliioAAACsRFEFAACAlSiqAAAAsBJFFQAAAFaiqAIAAMBKRotqQ0ODnn32WU2ZMkXTp0/X66+/rr/++stkJAAAAFjC2F+mam1t1fLlyzVu3Djt2LFDzc3NqqysVFdXl8rKykzFAgAAgCWMFdWPP/5Y7e3t2rlzp4LBoCQpEomovLxcq1atUmZmpqloAAAAsICxl/6/++47FRUVXS+pkjR//nxFo1GdOHHCVCwAAABYwlhRbWxs1Pjx42+4LRAIKCMjQ42NjYZSAQAAwBa+WCwWM/GF8/LytGHDBq1cufKG2xcsWKCpU6dq69atfTpvLBZTNDowd8nnkxzHUUv4mnoi0Tsen5riV/rwFFfH23CsLTm8zDzE7yiYnqpo9M7n7Qu3+2Mwrp0Nx0rezzAeg23e8a6d2/vXe15JVszFBvGuXaJn0pdzu/nafTmXl5m9uAb7ksMWfdl3sdj/95vejxOfyyefz+fqWGM/o+oVn88nv9/dnU+U3gdkL4634VhbcniZ2XG8e3HBhvW4G+bt5QzjYcN6eLl2g3UuNvDqMcnkTPp6Llv26N2wn/uz72y4v8YSBAIBhcPhm25vbW3VvffeayARAAAAbGKsqI4fP/6mn0UNh8O6fPnyTT+7CgAAgLuPsaI6a9YsnTx5Um1tbddvO3LkiBzH0fTp003FAgAAgCWMvZmqtbVVjz/+uLKzs7Vq1arrv/B/4cKF/MJ/AAAAmCuq0t9/QnXr1q365ZdfNGLECC1atEilpaVKSUkxFQkAAACWMFpUAQAAgP9i/vcOAAAAALdAUQUAAICVKKoAAACwEkUVAAAAVqKoAgAAwEoUVQAAAFiJogqrfPvtt3r66af18MMPa9KkSZo7d64qKioUDodNR4MH2tvbNWvWLIVCIZ05c8Z0HPTTgQMHFAqFbvpn+/btpqMhgT777DM9+eSTys/PV2FhoZ5//nl1dXWZjoUEKC4uvuU1HAqF9OWXXxrJNMTIVwX+Q0tLiyZPnqzi4mIFg0GdO3dOO3bs0Llz57R7927T8ZBgVVVVikQipmMgwd5//32lp6df/zgzM9NgGiRSdXW13nvvPa1evVpTpkzRlStX9MMPP3AdJ4ktW7bo6tWrN9y2Z88eHT16VEVFRUYyUVRhlUWLFt3wcWFhoVJSUrR582Y1NzfzDS+JNDQ0aN++fdq4caO2bNliOg4SKC8vT6NGjTIdAwnW2NionTt3qqqqSo888sj12x999FGDqZBIEyZMuOm2V155RdOnTzd2TfPSP6wXDAYlSd3d3WaDIKG2bdumpUuXKjs723QUAC4cOHBAWVlZN5RUJLeff/5ZTU1NWrhwobEMFFVYKRKJ6Nq1a6qrq9OuXbs0Z84cZWVlmY6FBDly5IjOnj2rNWvWmI4CDyxYsEA5OTmaO3eu3n33XV4WThKnT5/WAw88oKqqKhUVFWnSpElaunSpTp8+bToaPHLw4EENHz5cc+fONZaBl/5hpdmzZ6u5uVmSNHPmTL355puGEyFROjs7VVlZqdLSUqWlpZmOgwTKyMjQunXrVFBQIJ/Pp+PHj+utt95Sc3OzysrKTMdDP12+fFm//fabzp49qy1btmjYsGF655139Nxzz+no0aMaPXq06YhIoJ6eHh0+fFhz5szR8OHDjeWgqMJKNTU16uzs1Pnz51VdXa3Vq1frww8/lN/vNx0N/VRdXa3Ro0dr8eLFpqMgwWbOnKmZM2de/3jGjBlKTU3Vnj17tHr1ao0ZM8ZgOvRXLBZTR0eH3n77bU2cOFGSVFBQoDlz5uijjz7Shg0bDCdEIp04cUJ//vmnFixYYDQHL/3DShMnTtTUqVO1ZMkSVVVV6ccff9RXX31lOhb66eLFi9q9e7fWr1+vcDistrY2dXR0SJI6OjrU3t5uOCESbf78+YpEIqqvrzcdBf0UCAQUDAavl1Tp7/cQ5Obm6vz58waTwQsHDx5UMBjUjBkzjObgGVVYLxQKaejQobpw4YLpKOinpqYmdXd3a+XKlTd9rqSkRAUFBdq/f7+BZADuZMKECf/5OHzt2rUBTgMvdXV16euvv9YTTzyhoUOHGs1CUYX1Tp8+re7ubt5MlQRycnK0d+/eG26rr69XRUWFysvLlZ+fbygZvHLo0CH5/X7l5uaajoJ+mj17tg4cOKD6+nrl5ORIkq5cuaK6ujo988wzZsMhoY4fP66Ojg6j7/bvRVGFVdauXatJkyYpFArpnnvu0e+//64PPvhAoVBI8+bNMx0P/RQIBFRYWHjLz+Xl5SkvL2+AEyGRVqxYocLCQoVCIUnSsWPHtH//fpWUlCgjI8NwOvTXvHnzlJ+fr/Xr16u0tFSpqamqqalRSkqKli1bZjoeEuiLL77Qfffdp2nTppmOQlGFXSZPnqxDhw6ppqZGsVhMY8eO1ZIlS7RixQqlpKSYjgfgNrKzs/Xpp5/q0qVLikajGjdunF577TUVFxebjoYEcBxHNTU1qqioUFlZmbq7u/XQQw+ptraW/xFJIq2trfr++++1fPly+Xw+03Hki8ViMdMhAAAAgH/jXf8AAACwEkUVAAAAVqKoAgAAwEoUVQAAAFiJogoAAAArUVQBAABgJYoqAAAArERRBQAAgJUoqgAAALASRRUAAABWoqgCAADAShRVAAAAWOn/ALzqJjfKxLr7AAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Trainer Function and other useful function during training procedure"
      ],
      "metadata": {
        "id": "4dEfQggr7kPM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_info(model, loader):\n",
        "  batch_loss = []\n",
        "  epoch_loss = []\n",
        "  model.eval()\n",
        "  with torch.no_grad():\n",
        "    for data in loader:\n",
        "      # Get data to device\n",
        "      data = data.to(device=device)\n",
        "      # Forward\n",
        "      _, recon = model(data)\n",
        "      loss = criterion(recon, data)\n",
        "      # Taking record of losses\n",
        "      batch_loss.append(loss.item())\n",
        "  model.train()\n",
        "  return sum(batch_loss)/len(batch_loss)"
      ],
      "metadata": {
        "id": "Lg4UgLmw7iI2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Train Network\n",
        "def Train(model, device, criterion, optimizer, num_epochs):\n",
        "  train_loss = []\n",
        "  validation_loss = []\n",
        "  for epoch in range(num_epochs):\n",
        "      for batch_idx, (data) in enumerate(tqdm(train_loader)):\n",
        "          # Get data to device\n",
        "          data = data.to(device=device)\n",
        "          # Forward\n",
        "          _, recon = model(data)\n",
        "          loss = criterion(recon, data)\n",
        "          # Backward\n",
        "          optimizer.zero_grad()\n",
        "          loss.backward()\n",
        "\n",
        "          # Gradient descent or adam step\n",
        "          optimizer.step()\n",
        "\n",
        "      # Taking record of losses\n",
        "      tr_loss = get_info(model, train_loader)\n",
        "      val_loss = get_info(model, validation_loader)\n",
        "      train_loss.append(tr_loss)\n",
        "      validation_loss.append(val_loss)\n",
        "\n",
        "  plt.plot(train_loss)\n",
        "  plt.plot(validation_loss)\n",
        "  plt.title(\"Performance/Loss\")\n",
        "  plt.legend(['Train Loss', 'Validation Loss'])\n",
        "  plt.show()"
      ],
      "metadata": {
        "id": "76I-SNyR8BLc"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Set device cuda for GPU if it's available otherwise run on the CPU\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "num_epochs = 500\n",
        "Train(model, device, criterion, optimizer, num_epochs)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "odxiAWdv9NB_",
        "outputId": "4de89753-01bf-4249-edd4-a51e2591cc29"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 12/12 [00:00<00:00, 701.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 732.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 668.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 751.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 756.23it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 820.82it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 746.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 685.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 581.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 708.55it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 744.98it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 755.14it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 726.30it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 722.65it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 706.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 674.55it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 674.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 720.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 657.02it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 642.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 727.91it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 613.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 699.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 748.30it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 719.61it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 717.83it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 729.08it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 729.82it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 718.62it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 527.55it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 602.41it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 673.30it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 443.06it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 731.63it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 702.95it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 706.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 706.35it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 763.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 748.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 709.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 727.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 732.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 724.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 725.66it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 723.95it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 706.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 605.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 672.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 774.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 569.53it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 701.69it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 733.16it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 701.36it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 740.42it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 553.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 206.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 418.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 410.58it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 447.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 470.37it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 445.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 404.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 461.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 452.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 296.83it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 432.91it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 224.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 544.39it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 393.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 405.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 401.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 333.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 295.81it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 335.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 359.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 374.08it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 348.01it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 350.34it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 230.78it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 243.34it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 403.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 362.17it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 484.55it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 417.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 357.34it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 473.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 358.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 444.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 360.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 366.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 324.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 443.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 360.00it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 476.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 288.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 351.01it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 421.60it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 324.37it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 449.51it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 427.53it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 383.66it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 353.96it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 399.49it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 424.13it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 415.80it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 245.89it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 510.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 394.60it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 433.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 412.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 431.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 307.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 382.47it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 217.97it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 458.77it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 269.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 231.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 162.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 143.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 159.14it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 225.77it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 283.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 234.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 171.87it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 122.83it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 194.10it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 197.55it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 159.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 288.98it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 243.41it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 270.96it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 172.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 273.08it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 233.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 189.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 308.28it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 436.47it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 429.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 426.76it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 354.01it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 473.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 355.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 191.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 122.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 102.18it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 144.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 241.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 219.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 149.57it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 144.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 180.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 177.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 167.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 282.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 481.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 375.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 395.89it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 478.65it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 391.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 240.49it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 451.03it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 392.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 459.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 492.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 500.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 405.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 274.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 175.42it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 423.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 257.51it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 441.79it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 330.02it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 417.51it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 484.30it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 455.02it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 445.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 360.95it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 441.01it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 479.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 292.28it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 472.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 382.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 431.81it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 479.23it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 325.79it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 474.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 414.92it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 114.00it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 470.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 318.18it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 287.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 457.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.92it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 413.91it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 500.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 465.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 415.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 459.91it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 263.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 425.26it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 346.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 465.43it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 472.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 366.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 441.69it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 481.43it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 473.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 466.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 507.13it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 462.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 435.58it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 479.42it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 487.65it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 499.60it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 439.09it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 419.25it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 417.72it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 455.33it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 435.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 417.31it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 430.17it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 296.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 354.57it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 394.96it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 322.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 413.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 449.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 377.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 311.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 148.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 371.26it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 326.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 304.79it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 206.62it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 318.89it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 174.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 198.26it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 129.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 398.18it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 391.42it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 449.17it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 442.61it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 422.02it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 441.96it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 396.95it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 407.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 478.33it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 353.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 390.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 447.62it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 286.51it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 148.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 298.33it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 264.47it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 336.39it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 451.00it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 225.92it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 370.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 420.03it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 410.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 437.31it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 334.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 378.10it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 416.43it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 249.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 466.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 345.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 427.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 221.95it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 336.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 173.58it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 422.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 292.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 100.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 287.32it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 360.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 473.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 316.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 337.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 305.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 315.34it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 246.42it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 405.58it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 247.26it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 446.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 293.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 470.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 358.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 407.97it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 292.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 260.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 197.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 439.25it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 430.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 407.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 396.97it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 454.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 411.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 324.97it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 448.37it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 386.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 356.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 174.23it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 465.06it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 288.72it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 193.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 287.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 179.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 370.33it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 180.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 248.36it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 177.36it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 377.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 197.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 396.77it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 313.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 449.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 397.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 306.30it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 128.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 116.59it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 163.78it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 143.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 144.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 182.49it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 180.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 72.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 63.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 79.87it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 131.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 478.50it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 394.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 481.32it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 516.44it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 470.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 527.31it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 525.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 519.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 498.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 423.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 511.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 460.18it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 556.79it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 507.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 539.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 501.25it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 488.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 561.86it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 488.43it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 503.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 479.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 470.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 515.46it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 538.28it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 536.26it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 471.37it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 531.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 517.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 460.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 516.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 510.03it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 457.66it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 441.92it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 486.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 494.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 519.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 451.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 472.38it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 492.76it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 450.48it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 406.62it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 425.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 467.90it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 464.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 416.03it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 403.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 488.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 368.82it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 521.01it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 356.47it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 425.35it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 659.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 597.52it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 554.50it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 682.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 636.10it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 588.87it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 607.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 655.77it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 556.82it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 701.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 581.16it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 641.64it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 423.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 694.80it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 592.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 626.35it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 644.82it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 614.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 673.34it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 629.14it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 541.52it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 631.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 648.05it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 668.51it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 639.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 521.89it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 663.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 543.45it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 664.17it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 693.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 679.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 701.11it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 537.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 615.24it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 700.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 698.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 640.54it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 665.74it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 680.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 660.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 622.19it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 628.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 704.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 641.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 411.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 599.97it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 683.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 631.90it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 671.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 611.43it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 630.31it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 586.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 657.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 513.85it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 707.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 731.92it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 666.49it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 655.31it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 690.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 666.65it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 616.60it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 638.49it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 655.04it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 681.47it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 576.68it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 687.84it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 616.18it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 653.62it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 668.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 632.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 397.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 568.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 585.27it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 695.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 657.69it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 628.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 493.56it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 633.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 574.20it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 651.12it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 594.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 593.70it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 694.69it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 663.08it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 683.63it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 717.50it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 726.88it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 630.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 609.25it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 640.65it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 697.73it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 666.99it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 663.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 586.75it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 688.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 670.22it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 706.06it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 673.21it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 686.17it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 635.06it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 607.14it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 670.15it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 700.29it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 693.61it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 632.06it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 708.23it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 630.71it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 686.93it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 533.07it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 675.33it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 623.67it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 613.79it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 626.94it/s]\n",
            "100%|██████████| 12/12 [00:00<00:00, 675.00it/s]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Post-train error histogram"
      ],
      "metadata": {
        "id": "xL-c566fHRgO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_samples = torch.zeros(len(train_set), 12)\n",
        "validation_samples = torch.zeros(len(validation_set), 12)\n",
        "test_samples = torch.zeros(len(test_set), 12)\n",
        "\n",
        "train_recon = torch.zeros(len(train_set), 12)\n",
        "validation_recon = torch.zeros(len(validation_set), 12)\n",
        "test_recon = torch.zeros(len(test_set), 12)\n",
        "\n",
        "model.eval()\n",
        "with torch.no_grad():\n",
        "  # Getting reconstructed signals for train set\n",
        "  pointer = 0\n",
        "  for batch in train_loader:\n",
        "      train_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      train_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting reconstructed signals for validation set\n",
        "  pointer = 0\n",
        "  for batch in validation_loader:\n",
        "      validation_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      validation_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting reconstructed signals for test set\n",
        "  pointer = 0\n",
        "  for batch in test_loader:\n",
        "      test_samples[pointer:pointer+batch.shape[0]] = batch\n",
        "      test_recon[pointer:pointer+batch.shape[0]] = model(batch)[1]\n",
        "      pointer+=batch.shape[0]\n",
        "model.train()\n",
        "\n",
        "train_error = torch.sum((train_samples - train_recon)**2, 1)\n",
        "validation_error = torch.sum((validation_samples - validation_recon)**2, 1)\n",
        "test_error = torch.sum((test_samples - test_recon)**2, 1)\n",
        "\n",
        "plt.figure(figsize=(8, 12))\n",
        "#plot 1:\n",
        "plt.subplot(3, 1, 1)\n",
        "plt.hist(train_error, 50)\n",
        "plt.title('Train Error Histogram')\n",
        "#plot 2:\n",
        "plt.subplot(3, 1, 2)\n",
        "plt.hist(validation_error, 50)\n",
        "plt.title('Validation Error Histogram')\n",
        "#plot 3:\n",
        "plt.subplot(3, 1, 3)\n",
        "plt.hist(test_error, 50)\n",
        "plt.title('Test Error Histogram')\n",
        "plt.suptitle(\"Performance\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "6B43uDS4ZT9_",
        "outputId": "644b9182-288c-4f68-b9f4-dd6b2292d3b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x1200 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Getting the latent representation of the data for training a classifier"
      ],
      "metadata": {
        "id": "U3k7LkH9lcNP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_code = torch.zeros(len(train_set), 4)\n",
        "validation_code = torch.zeros(len(validation_set), 4)\n",
        "test_code = torch.zeros(len(test_set), 4)\n",
        "\n",
        "model.eval()\n",
        "with torch.no_grad():\n",
        "  # Getting train set latent representation\n",
        "  pointer = 0\n",
        "  for batch in train_loader:\n",
        "      train_code[pointer:pointer+batch.shape[0]] = model(batch)[0]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting validation set latent representation\n",
        "  pointer = 0\n",
        "  for batch in validation_loader:\n",
        "      validation_code[pointer:pointer+batch.shape[0]] = model(batch)[0]\n",
        "      pointer+=batch.shape[0]\n",
        "  # Getting test set latent representation\n",
        "  pointer = 0\n",
        "  for batch in test_loader:\n",
        "      test_code[pointer:pointer+batch.shape[0]] = model(batch)[0]\n",
        "      pointer+=batch.shape[0]\n",
        "model.train()\n",
        "\n",
        "print(train_code.shape)\n",
        "print(validation_code.shape)\n",
        "print(test_code.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Smg5kE-OlfhV",
        "outputId": "c7aa92f8-39f7-4cec-8578-b03ff4c26b82"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "torch.Size([768, 4])\n",
            "torch.Size([96, 4])\n",
            "torch.Size([95, 4])\n"
          ]
        }
      ]
    }
  ]
}