week5.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "S57TFN2mPqv2",
        "Ph__UvnDP0rC",
        "SgQKhKt_P0rF",
        "X_sYEjixP0rG",
        "akstnhwAP0rG"
      ],
      "gpuType": "T4",
      "toc_visible": true,
      "authorship_tag": "ABX9TyPgFOTZbtaNAlK2242iOEAS",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/dayoungcho/CAU_DL4AI/blob/main/week5.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ssEAOfLMPqv0"
      },
      "source": [
        "## Predicting house prices: A regression example"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6MtN4c6DPqv0"
      },
      "source": [
        "### The Boston Housing Price dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jtLJRXLUPqv0"
      },
      "source": [
        "**Loading the Boston housing dataset**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hGfdhOdlPqv0"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras.datasets import boston_housing\n",
        "(train_data, train_targets), (test_data, test_targets) = boston_housing.load_data()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Ws2a-WNfPqv0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0686a31e-819a-4d87-c521-c8c80b54565c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(404, 13)"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "train_data.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "RzmJq55IPqv0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9b5238fd-76b7-4e68-ee64-b009699a4d16"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(102, 13)"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ],
      "source": [
        "test_data.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "F2gZmHIYPqv0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4addfc58-a6c4-4df8-e1ea-48c36ce6a407"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([15.2, 42.3, 50. , 21.1, 17.7, 18.5, 11.3, 15.6, 15.6, 14.4, 12.1,\n",
              "       17.9, 23.1, 19.9, 15.7,  8.8, 50. , 22.5, 24.1, 27.5, 10.9, 30.8,\n",
              "       32.9, 24. , 18.5, 13.3, 22.9, 34.7, 16.6, 17.5, 22.3, 16.1, 14.9,\n",
              "       23.1, 34.9, 25. , 13.9, 13.1, 20.4, 20. , 15.2, 24.7, 22.2, 16.7,\n",
              "       12.7, 15.6, 18.4, 21. , 30.1, 15.1, 18.7,  9.6, 31.5, 24.8, 19.1,\n",
              "       22. , 14.5, 11. , 32. , 29.4, 20.3, 24.4, 14.6, 19.5, 14.1, 14.3,\n",
              "       15.6, 10.5,  6.3, 19.3, 19.3, 13.4, 36.4, 17.8, 13.5, 16.5,  8.3,\n",
              "       14.3, 16. , 13.4, 28.6, 43.5, 20.2, 22. , 23. , 20.7, 12.5, 48.5,\n",
              "       14.6, 13.4, 23.7, 50. , 21.7, 39.8, 38.7, 22.2, 34.9, 22.5, 31.1,\n",
              "       28.7, 46. , 41.7, 21. , 26.6, 15. , 24.4, 13.3, 21.2, 11.7, 21.7,\n",
              "       19.4, 50. , 22.8, 19.7, 24.7, 36.2, 14.2, 18.9, 18.3, 20.6, 24.6,\n",
              "       18.2,  8.7, 44. , 10.4, 13.2, 21.2, 37. , 30.7, 22.9, 20. , 19.3,\n",
              "       31.7, 32. , 23.1, 18.8, 10.9, 50. , 19.6,  5. , 14.4, 19.8, 13.8,\n",
              "       19.6, 23.9, 24.5, 25. , 19.9, 17.2, 24.6, 13.5, 26.6, 21.4, 11.9,\n",
              "       22.6, 19.6,  8.5, 23.7, 23.1, 22.4, 20.5, 23.6, 18.4, 35.2, 23.1,\n",
              "       27.9, 20.6, 23.7, 28. , 13.6, 27.1, 23.6, 20.6, 18.2, 21.7, 17.1,\n",
              "        8.4, 25.3, 13.8, 22.2, 18.4, 20.7, 31.6, 30.5, 20.3,  8.8, 19.2,\n",
              "       19.4, 23.1, 23. , 14.8, 48.8, 22.6, 33.4, 21.1, 13.6, 32.2, 13.1,\n",
              "       23.4, 18.9, 23.9, 11.8, 23.3, 22.8, 19.6, 16.7, 13.4, 22.2, 20.4,\n",
              "       21.8, 26.4, 14.9, 24.1, 23.8, 12.3, 29.1, 21. , 19.5, 23.3, 23.8,\n",
              "       17.8, 11.5, 21.7, 19.9, 25. , 33.4, 28.5, 21.4, 24.3, 27.5, 33.1,\n",
              "       16.2, 23.3, 48.3, 22.9, 22.8, 13.1, 12.7, 22.6, 15. , 15.3, 10.5,\n",
              "       24. , 18.5, 21.7, 19.5, 33.2, 23.2,  5. , 19.1, 12.7, 22.3, 10.2,\n",
              "       13.9, 16.3, 17. , 20.1, 29.9, 17.2, 37.3, 45.4, 17.8, 23.2, 29. ,\n",
              "       22. , 18. , 17.4, 34.6, 20.1, 25. , 15.6, 24.8, 28.2, 21.2, 21.4,\n",
              "       23.8, 31. , 26.2, 17.4, 37.9, 17.5, 20. ,  8.3, 23.9,  8.4, 13.8,\n",
              "        7.2, 11.7, 17.1, 21.6, 50. , 16.1, 20.4, 20.6, 21.4, 20.6, 36.5,\n",
              "        8.5, 24.8, 10.8, 21.9, 17.3, 18.9, 36.2, 14.9, 18.2, 33.3, 21.8,\n",
              "       19.7, 31.6, 24.8, 19.4, 22.8,  7.5, 44.8, 16.8, 18.7, 50. , 50. ,\n",
              "       19.5, 20.1, 50. , 17.2, 20.8, 19.3, 41.3, 20.4, 20.5, 13.8, 16.5,\n",
              "       23.9, 20.6, 31.5, 23.3, 16.8, 14. , 33.8, 36.1, 12.8, 18.3, 18.7,\n",
              "       19.1, 29. , 30.1, 50. , 50. , 22. , 11.9, 37.6, 50. , 22.7, 20.8,\n",
              "       23.5, 27.9, 50. , 19.3, 23.9, 22.6, 15.2, 21.7, 19.2, 43.8, 20.3,\n",
              "       33.2, 19.9, 22.5, 32.7, 22. , 17.1, 19. , 15. , 16.1, 25.1, 23.7,\n",
              "       28.7, 37.2, 22.6, 16.4, 25. , 29.8, 22.1, 17.4, 18.1, 30.3, 17.5,\n",
              "       24.7, 12.6, 26.5, 28.7, 13.3, 10.4, 24.4, 23. , 20. , 17.8,  7. ,\n",
              "       11.8, 24.4, 13.8, 19.4, 25.2, 19.4, 19.4, 29.1])"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "train_targets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bCOf6rQyPqv0"
      },
      "source": [
        "### Preparing the data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qXh9bAGuPqv0"
      },
      "source": [
        "**Normalizing the data**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "mHV8dXCdPqv0"
      },
      "outputs": [],
      "source": [
        "mean = train_data.mean(axis=0)\n",
        "train_data -= mean\n",
        "std = train_data.std(axis=0)\n",
        "train_data /= std\n",
        "test_data -= mean\n",
        "test_data /= std"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fd8uTN6oPqv0"
      },
      "source": [
        "### Building your model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TGwDZEAoPqv0"
      },
      "source": [
        "**Model definition**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Ujnp78w8Pqv0"
      },
      "outputs": [],
      "source": [
        "def build_model():\n",
        "    model = keras.Sequential([\n",
        "        layers.Dense(64, activation=\"relu\"),\n",
        "        layers.Dense(64, activation=\"relu\"),\n",
        "        layers.Dense(1)      # classification과 달리 activation function(softmax) 적용 X, 출력값 1개\n",
        "    ])\n",
        "    model.compile(optimizer=\"rmsprop\", loss=\"mse\", metrics=[\"mae\"])\n",
        "    return model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wHlpxKJXPqv1"
      },
      "source": [
        "### Validating your approach using K-fold validation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jkuzd-HRPqv1"
      },
      "source": [
        "**K-fold validation**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eR0eJd0TPqv1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e7e6a402-10c6-46ba-a7d8-e09037261dd6"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Processing fold #0\n",
            "Processing fold #1\n",
            "Processing fold #2\n",
            "Processing fold #3\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "\n",
        "k = 4\n",
        "num_val_samples = len(train_data) // k\n",
        "num_epochs = 100\n",
        "all_scores = []\n",
        "for i in range(k):\n",
        "    print(f\"Processing fold #{i}\")\n",
        "\n",
        "    val_data = train_data[i * num_val_samples: (i + 1) * num_val_samples]\n",
        "    val_targets = train_targets[i * num_val_samples: (i + 1) * num_val_samples]\n",
        "\n",
        "    partial_train_data = np.concatenate(\n",
        "        [train_data[:i * num_val_samples],\n",
        "         train_data[(i + 1) * num_val_samples:]],\n",
        "        axis=0)\n",
        "\n",
        "    partial_train_targets = np.concatenate(\n",
        "        [train_targets[:i * num_val_samples],\n",
        "         train_targets[(i + 1) * num_val_samples:]],\n",
        "        axis=0)\n",
        "\n",
        "    model = build_model()\n",
        "    model.fit(partial_train_data, partial_train_targets,\n",
        "              epochs=num_epochs, batch_size=16, verbose=0)\n",
        "    val_mse, val_mae = model.evaluate(val_data, val_targets, verbose=0)\n",
        "    all_scores.append(val_mae)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ACmDSFdqPqv1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6bfb50a2-ec45-4f02-e6d8-4bf9dde708cd"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.883969783782959, 2.535027027130127, 2.6299431324005127, 2.351130962371826]"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ],
      "source": [
        "all_scores"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kAbQbvQ3Pqv1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b5994884-bada-4bfa-fe52-099082bd86b6"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2.306292951107025"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "np.mean(all_scores)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2AqDK3ahPqv1"
      },
      "source": [
        "**Saving the validation logs at each fold**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "osHiZWhEPqv1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "35cff1e2-19fe-4973-d07a-452a85113222"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Processing fold #0\n",
            "Processing fold #1\n",
            "Processing fold #2\n",
            "Processing fold #3\n"
          ]
        }
      ],
      "source": [
        "num_epochs = 500\n",
        "all_mae_histories = []\n",
        "for i in range(k):\n",
        "    print(f\"Processing fold #{i}\")\n",
        "\n",
        "    val_data = train_data[i * num_val_samples: (i + 1) * num_val_samples]\n",
        "    val_targets = train_targets[i * num_val_samples: (i + 1) * num_val_samples]\n",
        "\n",
        "    partial_train_data = np.concatenate(\n",
        "        [train_data[:i * num_val_samples],\n",
        "         train_data[(i + 1) * num_val_samples:]],\n",
        "        axis=0)\n",
        "\n",
        "    partial_train_targets = np.concatenate(\n",
        "        [train_targets[:i * num_val_samples],\n",
        "         train_targets[(i + 1) * num_val_samples:]],\n",
        "        axis=0)\n",
        "\n",
        "    model = build_model()\n",
        "\n",
        "    history = model.fit(partial_train_data, partial_train_targets,\n",
        "                        validation_data=(val_data, val_targets),\n",
        "                        epochs=num_epochs, batch_size=16, verbose=0)\n",
        "\n",
        "    mae_history = history.history[\"val_mae\"]\n",
        "    all_mae_histories.append(mae_history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WFpJgEXqPqv1"
      },
      "source": [
        "**Building the history of successive mean K-fold validation scores**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7z8uUpVCPqv1"
      },
      "outputs": [],
      "source": [
        "average_mae_history = [\n",
        "    np.mean([x[i] for x in all_mae_histories]) for i in range(num_epochs)]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7n1QIA59Pqv1"
      },
      "source": [
        "**Plotting validation scores**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "anX3Ukv2Pqv2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "outputId": "b089f6bf-7b62-45b1-84a9-8a46d6c46807"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGwCAYAAABB4NqyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABMf0lEQVR4nO3deXhTVf4/8PdN0qT7vkNLWylboYAstQooQ4dSFAQZxMoMBRn5qehXBWdGXFjUGRydwZXBcUboOC6IjizKoligiOxLZS8FCgXadIMmXdM2Ob8/SgKBtrbQ3Bva9+t58jzkbvnktjTvnHPuPZIQQoCIiIioA1EpXQARERGR3BiAiIiIqMNhACIiIqIOhwGIiIiIOhwGICIiIupwGICIiIiow2EAIiIiog5Ho3QBzshisSA/Px9eXl6QJEnpcoiIiKgFhBAoLy9HeHg4VKrm23gYgBqRn5+PiIgIpcsgIiKiG3Du3Dl07ty52W0YgBrh5eUFoOEEent7K1wNERERtYTRaERERITtc7w5DECNsHZ7eXt7MwARERHdYloyfIWDoImIiKjDYQAiIiKiDocBiIiIiDocBiAiIiLqcBiAiIiIqMNhACIiIqIOhwGIiIiIOhxFA9DChQsxaNAgeHl5ITg4GOPGjUN2drbdNjU1NZg5cyYCAgLg6emJCRMmoLCwsNnjCiEwd+5chIWFwc3NDUlJScjJyXHkWyEiIqJbiKIBKDMzEzNnzsTOnTuxceNG1NXVYeTIkaisrLRt8+yzz+Kbb77Bl19+iczMTOTn5+OBBx5o9rhvvPEG3n33XXzwwQfYtWsXPDw8kJycjJqaGke/JSIiIroFSEIIoXQRVsXFxQgODkZmZiaGDRsGg8GAoKAgfPbZZ/jNb34DADh+/Dh69uyJHTt24I477rjuGEIIhIeHY/bs2XjuuecAAAaDASEhIUhPT8dDDz30i3UYjUb4+PjAYDDwTtBERES3iNZ8fjvVGCCDwQAA8Pf3BwDs27cPdXV1SEpKsm3To0cPREZGYseOHY0eIzc3F3q93m4fHx8fJCQkNLmPyWSC0Wi0exAREVH75TQByGKx4JlnnsFdd92F3r17AwD0ej20Wi18fX3ttg0JCYFer2/0ONblISEhLd5n4cKF8PHxsT04EzwREVH75jQBaObMmTh8+DCWL18u+2vPmTMHBoPB9jh37pxDXqe8pg7nL1XhYmWtQ45PRERELeMUAejJJ5/Et99+i82bN6Nz58625aGhoaitrUVZWZnd9oWFhQgNDW30WNbl114p1tw+Op3ONvO7I2eA/3jHWQz562a8seG4Q45PRERELaNoABJC4Mknn8TKlSuxadMmREdH260fMGAAXFxckJGRYVuWnZ2NvLw8JCYmNnrM6OhohIaG2u1jNBqxa9euJveRi0qSAAD1FqcZd05ERNQhKRqAZs6ciU8++QSfffYZvLy8oNfrodfrUV1dDaBh8PL06dMxa9YsbN68Gfv27cO0adOQmJhodwVYjx49sHLlSgCAJEl45pln8Nprr2HNmjU4dOgQpkyZgvDwcIwbN06Jt2mjUTUEIDMDEBERkaI0Sr74kiVLAAD33HOP3fJly5Zh6tSpAIC33noLKpUKEyZMgMlkQnJyMv7xj3/YbZ+dnW27ggwA/vjHP6KyshIzZsxAWVkZhgwZgg0bNsDV1dWh7+eXqBmAiIiInIJT3QfIWTjqPkAf7ziDuauP4N4+YVg8+fY2Oy4RERHdwvcBau+sLUD1FovClRAREXVsDEAyUkvsAiMiInIGDEAyutICxABERESkJAYgGWnUbAEiIiJyBgxAMlKrGk43AxAREZGyGIBkpGEXGBERkVNgAJKRioOgiYiInAIDkIzYAkREROQcGIBkpL48CNrCAERERKQoBiAZsQWIiIjIOTAAyejKjRB5J2giIiIlMQDJiDdCJCIicg4MQDLijRCJiIicAwOQjHgjRCIiIufAACQj6yBoBiAiIiJlMQDJyHojRI4BIiIiUhYDkIw4BoiIiMg5MADJyHYVmJmXwRMRESmJAUhG1jFAbAAiIiJSFgOQjK6MAWILEBERkZIYgGTEMUBERETOgQFIRrwTNBERkXNgAJKR5vKNEIXgjPBERERKYgCSkbUFCADMggGIiIhIKQxAMrILQGwBIiIiUgwDkIw0VwUgjgMiIiJSDgOQjOxagMwMQEREREphAJKRWrq6BYj3AiIiIlIKA5CMVCoJ1kYgDoImIiJSDgOQzKzdYBwETUREpBwGIJldmRCVAYiIiEgpDEAys94MkS1AREREymEAkhmnwyAiIlIeA5DMrAHIwkHQREREimEAkhnHABERESmPAUhmGl4FRkREpDgGIJldGQPEGyESEREpRdEAtHXrVowZMwbh4eGQJAmrVq2yWy9JUqOPN998s8ljzp8//7rte/To4eB30nJsASIiIlKeogGosrISffv2xeLFixtdX1BQYPdYunQpJEnChAkTmj1uXFyc3X7btm1zRPk3RMUAREREpDiNki+ekpKClJSUJteHhobaPV+9ejWGDx+OmJiYZo+r0Wiu29dZsAWIiIhIebfMGKDCwkKsXbsW06dP/8Vtc3JyEB4ejpiYGEyePBl5eXnNbm8ymWA0Gu0ejqK+fCNE3geIiIhIObdMAPrPf/4DLy8vPPDAA81ul5CQgPT0dGzYsAFLlixBbm4uhg4divLy8ib3WbhwIXx8fGyPiIiIti7fhi1AREREyrtlAtDSpUsxefJkuLq6NrtdSkoKJk6ciPj4eCQnJ2PdunUoKyvDihUrmtxnzpw5MBgMtse5c+faunwbFe8ETUREpDhFxwC11I8//ojs7Gx88cUXrd7X19cX3bp1w8mTJ5vcRqfTQafT3UyJLcYWICIiIuXdEi1AH330EQYMGIC+ffu2et+KigqcOnUKYWFhDqis9dQMQERERIpTNABVVFQgKysLWVlZAIDc3FxkZWXZDVo2Go348ssv8fvf/77RY4wYMQLvv/++7flzzz2HzMxMnDlzBtu3b8f48eOhVquRmprq0PfSUhreCJGIiEhxinaB7d27F8OHD7c9nzVrFgAgLS0N6enpAIDly5dDCNFkgDl16hRKSkpsz8+fP4/U1FSUlpYiKCgIQ4YMwc6dOxEUFOS4N9IKbAEiIiJSniQEpyW/ltFohI+PDwwGA7y9vdv02FOX7caW7GK88Zt4PDjQcVebERERdTSt+fy+JcYAtSfWLjALW4CIiIgUwwAkMzUvgyciIlIcA5DMNJfvBM0xQERERMphAJIZW4CIiIiUxwAksytXgfEyeCIiIqUwAMnsSgBSuBAiIqIOjAFIZhq2ABERESmOAUhmHANERESkPAYgmdkCkJkBiIiISCkMQDKzjQHiDbiJiIgUwwAkM7XEO0ETEREpjQFIZpwMlYiISHkMQDJTsQuMiIhIcQxAMmMXGBERkfIYgGTGFiAiIiLlMQDJzNoCxDtBExERKYcBSGbqy2ecXWBERETKYQCSGbvAiIiIlMcAJDMOgiYiIlIeA5DMeCdoIiIi5TEAyUwl8UaIRERESmMAkpm1BcjCFiAiIiLFMADJTMWpMIiIiBTHACQz3geIiIhIeQxAMrPdB4hdYERERIphAJIZB0ETEREpjwFIZho1B0ETEREpjQFIZtYWoHozAxAREZFSGIBkxhshEhERKY8BSGacCoOIiEh5DEAy42SoREREymMAkhlbgIiIiJTHACQzjgEiIiJSHgOQzK5MhaFwIURERB0YA5DM2AVGRESkPAYgmakun3F2gRERESmHAUhmbAEiIiJSnqIBaOvWrRgzZgzCw8MhSRJWrVplt37q1KmQJMnuMWrUqF887uLFixEVFQVXV1ckJCRg9+7dDnoHrcdB0ERERMpTNABVVlaib9++WLx4cZPbjBo1CgUFBbbH559/3uwxv/jiC8yaNQvz5s3D/v370bdvXyQnJ6OoqKity78hVwZBMwAREREpRaPki6ekpCAlJaXZbXQ6HUJDQ1t8zEWLFuHRRx/FtGnTAAAffPAB1q5di6VLl+L5559vdB+TyQSTyWR7bjQaW/x6rcUuMCIiIuU5/RigLVu2IDg4GN27d8fjjz+O0tLSJretra3Fvn37kJSUZFumUqmQlJSEHTt2NLnfwoUL4ePjY3tERES06Xu4GrvAiIiIlOfUAWjUqFH4+OOPkZGRgb/+9a/IzMxESkoKzGZzo9uXlJTAbDYjJCTEbnlISAj0en2TrzNnzhwYDAbb49y5c236Pq5mnQ2e9wEiIiJSjqJdYL/koYcesv27T58+iI+Px2233YYtW7ZgxIgRbfY6Op0OOp2uzY7XHGsLkIUtQERERIpx6haga8XExCAwMBAnT55sdH1gYCDUajUKCwvtlhcWFrZqHJEjqTkImoiISHG3VAA6f/48SktLERYW1uh6rVaLAQMGICMjw7bMYrEgIyMDiYmJcpXZLFsLEAMQERGRYhQNQBUVFcjKykJWVhYAIDc3F1lZWcjLy0NFRQX+8Ic/YOfOnThz5gwyMjJw//33o2vXrkhOTrYdY8SIEXj//fdtz2fNmoV//etf+M9//oNjx47h8ccfR2Vlpe2qMKVZrwLjIGgiIiLlKDoGaO/evRg+fLjt+axZswAAaWlpWLJkCQ4ePIj//Oc/KCsrQ3h4OEaOHIlXX33VbrzOqVOnUFJSYns+adIkFBcXY+7cudDr9ejXrx82bNhw3cBopVinwqhnCxAREZFiJCHYFHEto9EIHx8fGAwGeHt7t+mxCwzVSFy4CRqVhJN/Gd2mxyYiIurIWvP5fUuNAWoP2AVGRESkPAYgmVmnwhACYOMbERGRMhiAZGZtAQJ4KTwREZFSGIBkZm0BAtgNRkREpBQGIJmprwpAFk6HQUREpAgGIJnZdYGxBYiIiEgRDEAyU111xjkGiIiISBkMQDK7ugWI02EQEREpgwFIZmoOgiYiIlIcA5DMJEmCtRGILUBERETKYABSAO8GTUREpCwGIAVYu8E4CJqIiEgZDEAKsAYg3geIiIhIGQxACmAXGBERkbIYgBSgYhcYERGRohiAFGDrAmMLEBERkSIYgBSgktgCREREpCQGIAWoL591BiAiIiJlMAApQM0WICIiIkW1OAC98cYbqK6utj3/6aefYDKZbM/Ly8vxxBNPtG117ZRtEDTHABERESmixQFozpw5KC8vtz1PSUnBhQsXbM+rqqrwz3/+s22ra6eu3AeIAYiIiEgJLQ5A4prWimufU8uxC4yIiEhZHAOkAHaBERERKYsBSAHWFiBOhUFERKQMTWs2/ve//w1PT08AQH19PdLT0xEYGAgAduODqHlsASIiIlJWiwNQZGQk/vWvf9meh4aG4r///e9129Avs94HiIOgiYiIlNHiAHTmzBkHltGxcBA0ERGRstpsDFBZWRnef//9tjpcu8YuMCIiImXddADKyMjAww8/jLCwMMybN68tamr3NLwPEBERkaJuKACdO3cOr7zyCqKjozFy5EhIkoSVK1dCr9e3dX3tkm0yVLYAERERKaLFAaiurg5ffvklkpOT0b17d2RlZeHNN9+ESqXCiy++iFGjRsHFxcWRtbYb1jtBcwwQERGRMlo8CLpTp07o0aMHfvvb32L58uXw8/MDAKSmpjqsuPbKNhUGW4CIiIgU0eIWoPr6ekiSBEmSoFarHVlTu2frAuONEImIiBTR4gCUn5+PGTNm4PPPP0doaCgmTJiAlStXQrr8YU4tx8lQiYiIlNXiAOTq6orJkydj06ZNOHToEHr27In/+7//Q319Pf785z9j48aNMJvNjqy13eAgaCIiImXd0FVgt912G1577TWcPXsWa9euhclkwn333YeQkJC2rq9dst4JmoOgiYiIlHFT9wFSqVRISUnBV199hfPnz+OFF15o1f5bt27FmDFjEB4eDkmSsGrVKtu6uro6/OlPf0KfPn3g4eGB8PBwTJkyBfn5+c0ec/78+baxStZHjx49buTtOQyvAiMiIlJWm90JOigoCLNmzWrVPpWVlejbty8WL1583bqqqirs378fL7/8Mvbv34+vv/4a2dnZGDt27C8eNy4uDgUFBbbHtm3bWlWXo6k4FQYREZGiWnwZfExMTIu2O336dItfPCUlBSkpKY2u8/HxwcaNG+2Wvf/++xg8eDDy8vKanXhVo9EgNDS0xXXIjZfBExERKatVk6F26dIFDz/8MIKDgx1ZU5MMBgMkSYKvr2+z2+Xk5CA8PByurq5ITEzEwoULmw1MJpMJJpPJ9txoNLZVyY3iZKhERETKanEA+uKLL7B06VIsWrQIKSkpeOSRRzB69GioVG3Wi9asmpoa/OlPf0Jqaiq8vb2b3C4hIQHp6eno3r07CgoKsGDBAgwdOhSHDx+Gl5dXo/ssXLgQCxYscFTp1+FkqERERMpqcXqZOHEi1q9fj5MnT2LAgAF49tlnERERgeeffx45OTmOrBF1dXV48MEHIYTAkiVLmt02JSUFEydORHx8PJKTk7Fu3TqUlZVhxYoVTe4zZ84cGAwG2+PcuXNt/RbsWFuAeB8gIiIiZbS6+aZTp0548cUXkZOTg88++wy7du1Cjx49cOnSJUfUZws/Z8+excaNG5tt/WmMr68vunXrhpMnTza5jU6ng7e3t93DkWwtQLwTNBERkSJuqP+qpqYGn3zyCRYsWIBdu3Zh4sSJcHd3b+vabOEnJycHP/zwAwICAlp9jIqKCpw6dQphYWFtXt+Nst0HiF1gREREimhVANq1axdmzJiB0NBQLFq0CA888AAuXLiA5cuXQ6fTtfrFKyoqkJWVhaysLABAbm4usrKykJeXh7q6OvzmN7/B3r178emnn8JsNkOv10Ov16O2ttZ2jBEjRuD999+3PX/uueeQmZmJM2fOYPv27Rg/fjzUarVTTdrKLjAiIiJltXgQdFxcHIqKivDwww8jMzMTffv2vekX37t3L4YPH257br2PUFpaGubPn481a9YAAPr162e33+bNm3HPPfcAAE6dOoWSkhLbuvPnzyM1NRWlpaUICgrCkCFDsHPnTgQFBd10vW1FfXngOFuAiIiIlCEJ0bJPYZVKBQ8PD2g0mmYnQL148WKbFacUo9EIHx8fGAwGh4wH+vPao/jXj7n4f8NiMGd0zzY/PhERUUfUms/vFrcALVu27KYLowYqToVBRESkqBYHoLS0NEfW0aGoORs8ERGRouS5iyHZsU2FwRYgIiIiRTAAKUDFFiAiIiJFMQApQM0bIRIRESmKAUgB7AIjIiJSFgOQAtgFRkREpKwWXwVmZTabkZ6ejoyMDBQVFcFise/H2bRpU5sV115Zp8JgCxAREZEyWh2Ann76aaSnp+Pee+9F7969m70pIjXO2gJUzwBERESkiFYHoOXLl2PFihUYPXq0I+rpEGyDoNkFRkREpIhWjwHSarXo2rWrI2rpMDgImoiISFmtDkCzZ8/GO++8gxZOIUaNsA2CZgAiIiJSRKu7wLZt24bNmzdj/fr1iIuLg4uLi936r7/+us2Ka69sLUAMkURERIpodQDy9fXF+PHjHVFLh6FmCxAREZGiWh2AOCv8zbPNBs/8Q0REpIhWByCr4uJiZGdnAwC6d++OoKCgNiuqvdNwEDQREZGiWj0IurKyEo888gjCwsIwbNgwDBs2DOHh4Zg+fTqqqqocUWO7Y2sBYgAiIiJSRKsD0KxZs5CZmYlvvvkGZWVlKCsrw+rVq5GZmYnZs2c7osZ2R82pMIiIiBTV6i6w//3vf/jqq69wzz332JaNHj0abm5uePDBB7FkyZK2rK9d4lQYREREymp1C1BVVRVCQkKuWx4cHMwusBbiZKhERETKanUASkxMxLx581BTU2NbVl1djQULFiAxMbFNi2uveCdoIiIiZbW6C+ydd95BcnIyOnfujL59+wIAfv75Z7i6uuK7775r8wLbIxXnAiMiIlJUqwNQ7969kZOTg08//RTHjx8HAKSmpmLy5Mlwc3Nr8wLboys3QlS4ECIiog7qhu4D5O7ujkcffbSta+kw2AVGRESkrBYFoDVr1iAlJQUuLi5Ys2ZNs9uOHTu2TQprzzgImoiISFktCkDjxo2DXq9HcHAwxo0b1+R2kiTBbDa3VW3tFluAiIiIlNWiAGSxWBr9N90Y632A2AJERESkjFZfBv/xxx/DZDJdt7y2thYff/xxmxTV3lm7wOo5GyoREZEiWh2Apk2bBoPBcN3y8vJyTJs2rU2Kau9sXWBsASIiIlJEqwOQEALS5RaMq50/fx4+Pj5tUlR7ZxsEzTFAREREimjxZfD9+/eHJEmQJAkjRoyARnNlV7PZjNzcXIwaNcohRbY3bAEiIiJSVosDkPXqr6ysLCQnJ8PT09O2TqvVIioqChMmTGjzAtsjawBiCxAREZEyWhyA5s2bBwCIiorCpEmT4Orq6rCi2jsGICIiImW1+k7QaWlpjqijQ7FOhcH8Q0REpIxWByCz2Yy33noLK1asQF5eHmpra+3WX7x4sc2Ka6/YAkRERKSsVl8FtmDBAixatAiTJk2CwWDArFmz8MADD0ClUmH+/PkOKLH94WzwREREymp1APr000/xr3/9C7Nnz4ZGo0Fqair+/e9/Y+7cudi5c6cjamx3bF1gbAEiIiJSRKsDkF6vR58+fQAAnp6etpsi3nfffVi7dm2rjrV161aMGTMG4eHhkCQJq1atslsvhMDcuXMRFhYGNzc3JCUlIScn5xePu3jxYkRFRcHV1RUJCQnYvXt3q+pyNBWnwiAiIlJUqwNQ586dUVBQAAC47bbb8P333wMA9uzZA51O16pjVVZWom/fvli8eHGj69944w28++67+OCDD7Br1y54eHggOTkZNTU1TR7ziy++wKxZszBv3jzs378fffv2RXJyMoqKilpVmyNZW4CEaAh5REREJK9WB6Dx48cjIyMDAPDUU0/h5ZdfRmxsLKZMmYJHHnmkVcdKSUnBa6+9hvHjx1+3TgiBt99+Gy+99BLuv/9+xMfH4+OPP0Z+fv51LUVXW7RoER599FFMmzYNvXr1wgcffAB3d3csXbq0VbU5knUQNMCB0EREREpo9VVgr7/+uu3fkyZNQmRkJHbs2IHY2FiMGTOmzQrLzc2FXq9HUlKSbZmPjw8SEhKwY8cOPPTQQ9ftU1tbi3379mHOnDm2ZSqVCklJSdixY0eTr2UymewmeDUajW30LhqnujoACdH6HwIRERHdlJv+7E1MTERiYmJb1GJHr9cDAEJCQuyWh4SE2NZdq6SkBGazudF9jh8/3uRrLVy4EAsWLLjJiltOfdVcahaLbC9LREREl7UoAK1Zs6bFBxw7duwNF6OUOXPmYNasWbbnRqMRERERDns99TUtQERERCSvFgUg6zxgVpIkXTd41zpDvNlsbpPCQkNDAQCFhYUICwuzLS8sLES/fv0a3ScwMBBqtRqFhYV2ywsLC23Ha4xOp2v1AO6boZI4BoiIiEhJLRoEbbFYbI/vv/8e/fr1w/r161FWVoaysjKsX78et99+OzZs2NBmhUVHRyM0NNQ24BpoaJnZtWtXk11uWq0WAwYMsNvHYrEgIyPDId10N+rqFiDeC4iIiEh+rR4D9Mwzz+CDDz7AkCFDbMuSk5Ph7u6OGTNm4NixYy0+VkVFBU6ePGl7npubi6ysLPj7+yMyMhLPPPMMXnvtNcTGxiI6Ohovv/wywsPD7VqkRowYgfHjx+PJJ58EAMyaNQtpaWkYOHAgBg8ejLfffhuVlZWYNm1aa9+qw1yVf9gFRkREpIBWB6BTp07B19f3uuU+Pj44c+ZMq461d+9eDB8+3PbcOg4nLS0N6enp+OMf/4jKykrMmDEDZWVlGDJkCDZs2GA3E/2pU6dQUlJiez5p0iQUFxdj7ty50Ov16NevHzZs2HDdwGglSZIEldQwGSq7wIiIiOQniVbeiW/YsGFwdXXFf//7X1uoKCwsxJQpU1BTU4PMzEyHFCono9EIHx8fGAwGeHt7O+Q1Yl9chzqzwPbnf4VwXzeHvAYREVFH0prP71bfCHHp0qUoKChAZGQkunbtiq5duyIyMhIXLlzARx99dMNFdzScEZ6IiEg5re4C69q1Kw4ePIiNGzfa7q3Ts2dPJCUl2a4Eo19mmxCVY4CIiIhkd0M3QpQkCSNHjsTIkSPbup4OQ8UWICIiIsW0KAC9++67mDFjBlxdXfHuu+82u+3//d//tUlh7Z21C4wtQERERPJrUQB66623MHnyZLi6uuKtt95qcjtJkhiAWsjaBWbmVBhERESya1EAys3NbfTfdOPYBUZERKScVl8FRm2Dg6CJiIiU06IWoKsnCv0lixYtuuFiOhJeBk9ERKScFgWgAwcOtOhgvAy+5VSX2944FQYREZH8WhSANm/e7Og6OhxbFxhbgIiIiGTHMUAK4SBoIiIi5dzQjRD37t2LFStWIC8vD7W1tXbrvv766zYprL2zXQbPLjAiIiLZtboFaPny5bjzzjtx7NgxrFy5EnV1dThy5Ag2bdoEHx8fR9TYLtluhMj7ABEREcmu1QHoL3/5C9566y1888030Gq1eOedd3D8+HE8+OCDiIyMdESN7ZKKLUBERESKaXUAOnXqFO69914AgFarRWVlJSRJwrPPPosPP/ywzQtsr660ADEAERERya3VAcjPzw/l5eUAgE6dOuHw4cMAgLKyMlRVVbVtde0YB0ETEREpp9WDoIcNG4aNGzeiT58+mDhxIp5++mls2rQJGzduxIgRIxxRY7ukvnzLpHoGICIiItm1OAAdPnwYvXv3xvvvv4+amhoAwIsvvggXFxds374dEyZMwEsvveSwQtsbzeU7IXIqDCIiIvm1OADFx8dj0KBB+P3vf4+HHnoIAKBSqfD88887rLj2THO5CaiO08ETERHJrsVjgDIzMxEXF4fZs2cjLCwMaWlp+PHHHx1ZW7vmom449fVmtgARERHJrcUBaOjQoVi6dCkKCgrw3nvv4cyZM7j77rvRrVs3/PWvf4Ver3dkne2OC1uAiIiIFNPqq8A8PDwwbdo0ZGZm4sSJE5g4cSIWL16MyMhIjB071hE1tkvWMUB1HARNREQku5uaC6xr16544YUX8NJLL8HLywtr165tq7raPesYoHq2ABEREcnuhuYCA4CtW7di6dKl+N///geVSoUHH3wQ06dPb8va2jXt5TFA7AIjIiKSX6sCUH5+PtLT05Geno6TJ0/izjvvxLvvvosHH3wQHh4ejqqxXbpyFRi7wIiIiOTW4gCUkpKCH374AYGBgZgyZQoeeeQRdO/e3ZG1tWsaXgVGRESkmBYHIBcXF3z11Ve47777oFarHVlTh2DtAqvndPBERESya3EAWrNmjSPr6HA0l+cCq+UYICIiItnd1FVgdOPYBUZERKQcBiCFuPAyeCIiIsUwACnEOhVGLVuAiIiIZMcApBDeCJGIiEg5DEAKcVFZrwJjCxAREZHcGIAUYh0DxKvAiIiI5McApJArV4ExABEREcmNAUghV64CYxcYERGR3BiAFKK5PAaojmOAiIiIZOf0ASgqKgqSJF33mDlzZqPbp6enX7etq6urzFX/MhfN5QBUzy4wIiIiubVqNngl7NmzB2az2fb88OHD+PWvf42JEyc2uY+3tzeys7NtzyVJcmiNN8Ll8lQYnAuMiIhIfk4fgIKCguyev/7667jttttw9913N7mPJEkIDQ11dGk3xToIuo5jgIiIiGTn9F1gV6utrcUnn3yCRx55pNlWnYqKCnTp0gURERG4//77ceTIkWaPazKZYDQa7R6OZh0EXcerwIiIiGR3SwWgVatWoaysDFOnTm1ym+7du2Pp0qVYvXo1PvnkE1gsFtx55504f/58k/ssXLgQPj4+tkdERIQDqrfnwslQiYiIFCMJIW6ZT+Dk5GRotVp88803Ld6nrq4OPXv2RGpqKl599dVGtzGZTDCZTLbnRqMRERERMBgM8Pb2vum6G7PrdCkmfbgTMUEe2DT7Hoe8BhERUUdiNBrh4+PTos9vpx8DZHX27Fn88MMP+Prrr1u1n4uLC/r374+TJ082uY1Op4NOp7vZEltFwxYgIiIixdwyXWDLli1DcHAw7r333lbtZzabcejQIYSFhTmoshujtQ2C5hggIiIiud0SAchisWDZsmVIS0uDRmPfaDVlyhTMmTPH9vyVV17B999/j9OnT2P//v347W9/i7Nnz+L3v/+93GU3S2MbBM0WICIiIrndEl1gP/zwA/Ly8vDII49cty4vLw8q1ZUcd+nSJTz66KPQ6/Xw8/PDgAEDsH37dvTq1UvOkn+RbSoM3geIiIhIdrfUIGi5tGYQ1Y06W1qJu9/cAg+tGkdeGeWQ1yAiIupIWvP5fUt0gbVHthshci4wIiIi2TEAKcQ2FQYHQRMREcmOAUgh1hYgiwDMbAUiIiKSFQOQQqyDoAFeCk9ERCQ3BiCFWKfCAIB6tgARERHJigFIIRrVlRYgjgMiIiKSFwOQQtQqCdYJ7WsZgIiIiGTFAKQQSZLgouJ8YEREREpgAFKQdToMBiAiIiJ5MQApyDoOqI7TYRAREcmKAUhBWg1nhCciIlICA5CCNBwDREREpAgGIAVZxwCxBYiIiEheDEAKsnaBmeoZgIiIiOTEAKQgNxc1AKCmzqxwJURERB0LA5CC3LUMQEREREpgAFKQ6+UWoKpaBiAiIiI5MQApyNoFVs0WICIiIlkxACnI7XIXWDVbgIiIiGTFAKQgDoImIiJSBgOQglzZBUZERKQIBiAFWa8C4yBoIiIieTEAKYhdYERERMpgAFIQB0ETEREpgwFIQRwDREREpAwGIAVduQ8Q5wIjIiKSEwOQgqxdYDXsAiMiIpIVA5CCrAGoqq5e4UqIiIg6FgYgBdm6wNgCREREJCsGIAVduQyeY4CIiIjkxACkINtl8LwKjIiISFYMQApiFxgREZEyGIAUdPV9gCwWoXA1REREHQcDkIKsc4EBgKme44CIiIjkwgCkIGsLEMBxQERERHJiAFKQWiVBq2n4EVTV8l5AREREcmEAUpinTgMAqDSxBYiIiEguTh2A5s+fD0mS7B49evRodp8vv/wSPXr0gKurK/r06YN169bJVO2N8XJtCEDlNXUKV0JERNRxOHUAAoC4uDgUFBTYHtu2bWty2+3btyM1NRXTp0/HgQMHMG7cOIwbNw6HDx+WseLW8XZ1AQAYGYCIiIhko1G6gF+i0WgQGhraom3feecdjBo1Cn/4wx8AAK+++io2btyI999/Hx988EGT+5lMJphMJttzo9F4c0W3wpUWII4BIiIikovTtwDl5OQgPDwcMTExmDx5MvLy8prcdseOHUhKSrJblpycjB07djT7GgsXLoSPj4/tERER0Sa1t4StBaiaLUBERERyceoAlJCQgPT0dGzYsAFLlixBbm4uhg4divLy8ka31+v1CAkJsVsWEhICvV7f7OvMmTMHBoPB9jh37lybvYdfYm0BMrIFiIiISDZO3QWWkpJi+3d8fDwSEhLQpUsXrFixAtOnT2+z19HpdNDpdG12vNbwduMYICIiIrk5dQvQtXx9fdGtWzecPHmy0fWhoaEoLCy0W1ZYWNjiMURK4BggIiIi+d1SAaiiogKnTp1CWFhYo+sTExORkZFht2zjxo1ITEyUo7wb4nV5DBADEBERkXycOgA999xzyMzMxJkzZ7B9+3aMHz8earUaqampAIApU6Zgzpw5tu2ffvppbNiwAX//+99x/PhxzJ8/H3v37sWTTz6p1Fv4Rd7WMUAcBE1ERCQbpx4DdP78eaSmpqK0tBRBQUEYMmQIdu7ciaCgIABAXl4eVKorGe7OO+/EZ599hpdeegkvvPACYmNjsWrVKvTu3Vupt/CLrrQAMQARERHJxakD0PLly5tdv2XLluuWTZw4ERMnTnRQRW3P241XgREREcnNqbvAOgJvtgARERHJjgFIYVduhMgWICIiIrkwACnMehl8dZ0ZdWaLwtUQERF1DAxACvN2c4FKavj3papaZYshIiLqIBiAFKZWSfD30AIASsoZgIiIiOTAAOQEAjwapuEorTT9wpZERETUFhiAnECgV0MLUGkFW4CIiIjkwADkBKwtQCUVbAEiIiKSAwOQEwjwvDwGiC1AREREsmAAcgKBnmwBIiIikhMDkBMI9LSOAWIAIiIikgMDkBO40gLELjAiIiI5MAA5gYDLAYgtQERERPJgAHICId4NAaiw3MTpMIiIiGTAAOQEQrxc4a5Vw2wRyLtYpXQ5RERE7R4DkBNQqSTEBHkAAE4VVShcDRERUfvHAOQkYgI9AQCnSyoVroSIiKj9YwByErcFNQQgtgARERE5HgOQk7gt+HIXWDEDEBERkaMxADmJbiFeAIBsfTksFqFwNURERO0bA5CTiAn0gKuLCpW1Zo4DIiIicjAGICehUavQK8wbAHD4gkHhaoiIiNo3BiAn0qeTDwAGICIiIkdjAHIivS8HoD1nLylcCRERUfvGAORE7u4WBEkCfj5XhnO8IzQREZHDMAA5kWBvVyRE+wMAvj1YoHA1RERE7RcDkJMZ27cTAGD5njxeDk9EROQgDEBO5v5+4fB21eBsaRUyjhcpXQ4REVG7xADkZDx0Gjyc0AUA8PYPJ9gKRERE5AAMQE5oxrAYeOo0OJJvxLeHOBaIiIiorTEAOSF/Dy1mDIsBAPz9+2zU1lsUroiIiKh9YQByUtOHRCPQU4ezpVX4IPOU0uUQERG1KwxATspDp8GL9/YAALyTkYOl23IVroiIiKj9YAByYuP6dULq4AiYLQKvfHsUS7acQp2Z3WFEREQ3iwHIiUmShL+M74OZw28DAPx1w3E8+M8dHBNERER0kxiAnJwkSZj96+5IS2y4NP5AXhm6vbQe72bkKFwZERHRrcupA9DChQsxaNAgeHl5ITg4GOPGjUN2dnaz+6Snp0OSJLuHq6urTBU7hkolYcH9vfHK/XG2ZYs2nkCBoRoAcKGsGhuPFkII3jOIiIioJTRKF9CczMxMzJw5E4MGDUJ9fT1eeOEFjBw5EkePHoWHh0eT+3l7e9sFJUmS5CjX4VIHR6Kk3IR3N50EAKQt3Y3+EX74Yu85AMAfkrvjdHElKk31mDm8K/p09lGyXCIiIqcliVuo2aC4uBjBwcHIzMzEsGHDGt0mPT0dzzzzDMrKym74dYxGI3x8fGAwGODt7X3Dx3GUXadL8duPdqHO3PSPztVFhRfv7YVeYV6ICfSEn4dWxgqJiNqeEAJ1ZgGtxqk7L25abkklOvu5wUXd+Ps01ZuhVasa/XK/6sAF+HtoMaxbUJPHX511Aa4uaiTHhdqWCSHw83kDTujL0T/SF7EhXrBYBL45mI9If3dsP1WKgV38MDja3/a6ZVW1qK23YN2hAnQJ9IChqg79I33RJaDpBgqrnMJyxAR5Qq1q2waK1nx+O3UL0LUMBgMAwN/fv9ntKioq0KVLF1gsFtx+++34y1/+gri4uCa3N5lMMJlMtudGo7FtCnaQhJgAbHz2bqz5OR+5JZXILam0/fJa1dRZ8PKqwwAAtUrCqLhQxHXyRpHRBA+dGhNu74wATx3+/n02hvcIxvDuwaipM6O8ph7+HlrUmS0oMNRAq1Ghk6+b7bhmi0BZVS3e23QSFytr8fqEPnDX3lK/Ru2Gqd4MtSRB08QfSeq4iow12HBEjwcHRsDVRa10OW3iTEklHliyHZeqavH8qB74f3ff5rDXsk5BVGCsQaCnFjrNL5/DkgoT/rfvPCYNikBRuQnHCoxI6R2G5XvyEOylw6Aof/h7aJvtkag01WPlgQt46fLfbi9XDWb/uhs8dBr0jfBFtxAv7Dt7Cakf7sSwboG4IyYA98aHIcyn4W/0ztOleOaLLADAB7+9HUNjg+Ch06C23oKconIEeepQYKjB08sbtukZ5o1nkmKRHBeKheuP48OtpwEAKgn4d9pA7Dx90bbMqkuAO+aN6QVJkvDUZwdQYaq3W69VqzD1rii4uqhx7mIVft0rBLtzL8JNq8aE2zthdVY+jhWUI+N4Id6YEI+JAyN++QfiILdMC5DFYsHYsWNRVlaGbdu2Nbndjh07kJOTg/j4eBgMBvztb3/D1q1bceTIEXTu3LnRfebPn48FCxZct9xZW4AaI4TAukN61NSZcW98GN7blIMDeWU4U1KJfEPNddt7aNWI6+SD3bkXIUnAwC5++Pm8odErzDr5umFKYhdcKKvGf3eexdW/MfPG9EJK7zB8vjsPl6pqcaKwHAGeOjybFIvbgjzxjy2ncPiCAeP7d8LIy982austOJJvgFolYe3BAkwfGo1grxsfp1VntkBvqEGEv/t16ypM9dCqVcg8UYwhXQPhpm0fHwalFSaMfGsreoV7Y+nUQVh/WI/h3YPg5eri8Neurbdg64liJMT4t/r1Kk31KDTWINLf3RbchBBO3U1tsQio2uhb6o22YJjqzVi++xzu6R5k+3Z97Xk7W1qJBd8cxbj+nbDqwAVsOl6EyQmR+PP4Po0es7ymDocuGDA4yr/JEF1UXoOPtuXCUFWHGcNioJIkRAVe/+2+uZ/h2dJKVJjqEertiswTxUjqFQLvJn5v8kqrEOKjg0alsrUMCCGw7WQJlm7LxebsYgBAoKcWO+aMsGsh2XGqFPvzLmFwtD8GRV35kmyxCJRUmnCmpAq3BXnAIoCfTpagdydvdAnwgItaBb2hBieLKpBfVo1yUz0++vG03d/NPp188Puh0RgU5Y/vjujh567FmL7h0BtrMHtFFk4WVaKkouFLdI9QL1woq0Z5jX0wAIDEmAA8n9IDaw8V4ExJJS6UVePubkHYn3cJP58zoLrO3Oh5AQBJAibc3hlf7Ttvt9xDq0aEvzsM1XUouOZvva+7C0b0CMGGwwWorG04trerBsZravvtHZH4ZGceAMDP3QWXquqarKMtTb0zCvPHNt04cSNa0wJ0ywSgxx9/HOvXr8e2bduaDDKNqaurQ8+ePZGamopXX3210W0aawGKiIi4pQJQc97LyMHSn3KReFsAIvzdsfZgAc5fqr6pYwZ6alFSUQsA0GlUMF0TnPw9tIgL98aPOSW2ZR/+bgBKK2vx/qaTuFB25fWjAz2w6om74OWqwdECIxauP4bymnpMTohEQnQAvvk5H4Mu/1Gz/lEsMtbgVHElaurNmLf6CPIuVmHB2Dik3RmF+sutV5uzi/Da2mN2oS7AQ4uPpg5CdIAH6i0W6FzUyNaXI8LfzRbCrv5jLoTA2kMF+PbnAgzrFoQHB3ZGflkNvtx3Dml3RsHfXQuLEM22wgghUG6qR129BQGeOtvy3JJKRPi52e1rqjdjTVY+vFw1iA70xNzVhzHtriiM6h1md8xPd53FiysbviVa/3jd2ycMiyffDlO9GVUmc6PdnlW19dh5uhRDugb94odwobEG2fpyDI0NhBDAsu1n0MnXDcv35GFLdjHcXNQI9tahps6MpVMHwd9DCx83F1TVmvHKN0cxMi4E98WH2453qbIW9723DRfKqtHJ1w1/HNUdhcYa/O27E7g3Pgy+7i54aFAkzBaBnmFedj+DDYf12JV7Eb8fGo3Ofg1B99B5A6pq63F7Fz9sOKxHzzAvZJ0zYFhsIEz1FnT2c8PZ0iq4a9UI9r4+YAshUGu2YO6qIyipMOHd1P7w0F1pzdQbavCn/x3E/rxLSJ82GAO6+KG23oLN2UWICfRAbIhXkz9vY009ThaVw1hTj2+y8hHq44oTheXIL6vBuYtVeDopFil9wiCEwKoDF3B3t2CY6s3YmlOC8f07oZOvm93P55VvjmLpT7noFeaN/zwyGK9+exQ/nSzB2H7h6Bfhi76dfTFl6W7kXayCVq1C7VX3C8ua+2scumCAq4sat0f6YUt2EXbnXsT/9p9HSUUtEmMC0C/SF3X1FgyM8kNVrRnbckowJDYQmSeKsTor3+79TbsrCnNSesJUb8am40X4y7pjuFRZh9fG98Z98WE4dN6AfEM1hAAOnjcgffsZu/1vC/LAQ4MiMSDKD3/7LhuG6jrEBHniQN4l298lF7WEv03si6MFRnx/pBC5JZWNnuuJAzpDrZJQUlGLH44V2pZr1Sok9QrGuYvVOHTB0Oi+zu4Pyd2x7KdclFTUwkOrtgWYtuLlqrkupKUOjsS8Mb0wddlu7Dx9Ee5aNboGe+LgeQMGdvHDR1MHYd7qw/jmYAECPLQI83XDofNl8NBqMKx7EEbFhcJFLWHF3vMw1ZthqrOguMIEX3ctLlaacO5iNaIC3DEyLhSj+4ShX4Rvm74noB0GoCeffBKrV6/G1q1bER0d3er9J06cCI1Gg88//7xF2zv7GKCbdfiCAeMW/wQAePbX3WC2CIR6u2JAlB+0ahV+Pl+G+WuOoKSiFsFeOhSVN4RDX3cXvJDSE7Ehnuga7Im739yCi5UNIahXmDeigzwQE+iB9Yf1OFlU0eq61CoJZov9r6OXToPyy02sWo0K3q4auLqocaGsGtf+5rpr1Rgc7Y8fc0quO861rg1tKgl4/J7b4KHT4J0fcvBwQiS8dBocLSi3+8N6tW4hnnDTapBXWolh3YLg565FSu9QVNWZkVNYjiFdg/D2DyewObsIdWYBL1cNnh4Ri15h3th0vAj/3paLX/UIRp9OPjhRWI7TxZU4e7ESNXX2YVIlAd8+NRR6YzV+OFaErLwyHC1ovJv2g98OwHubcnAk34joQA/0j/TFvDFx0KpV+OlkCV5cdQiFxoafZydfN/xj8u3o08kHy/ecQ3WdGeU1dYgK8ECEvxtmrfgZZ0urMKJHMGrNFrsw2xStRgV/dy30xoZvovfFh6Go3IS59/XCki2nsLaFk/v6ubsgKtADST1DcLq4Ev/b3/Ctd0AXP/QM88KmY0WNtmw2ZWAXP0gScP5SdUMroABON/KhGhPkgXqzsHUBX23igM744VghLlXVQaOSMLxHMBJjAnCmtBLbT5ViaGwgtp8sRXZheYvrak6kvzviwr1hqrdg0/GiGz6Ou1aNqlozJAno7OeGcxdv7stPc1QS8Av/9W7a5IRIfLor77rlkoTr/ia0RqS/O6ICPZBXWokzpVW25b+9IxJnS6uw/VQpzBaBEG+d7f8Q0PBlb+KAztiSXYzSShNKKmrRp5MPJg2KsHVlWbm5qG2tPC5qCYOj/XGmpAoXyqrhopbgqlHDVG/BzOFd8XRSLGrqzDh/qQpRAR5487tsnCyqwLlLVUjpHYY7YgLQM8wLB88bUGiswXdH9LhYWYtnf90NiTEBsAhg64lifH9Ujx2nS7FgbBz6dPJFxrFC+HlokRwXinMXq/Cn/x1EdZ0ZyXGheOSuaGg1KgghkG+ogY+bCzy0avx0shS9wr3hf/lLldkibF9Gy6pqoVGr4Kn75aEQhuo6eLtqHNri224CkBACTz31FFauXIktW7YgNja21ccwm82Ii4vD6NGjsWjRohbt094DEACcKCyHm4u60W4jADDW1MFiEfB118JsEdh0vAjxnX0QctU36dPFFdh5+iIi/d1xV9cA2y/1pcpavPrtUezKvYhnkmJxb3wY/t9/9+HHnBIEeuoagoZWjX9uPY0Jt3fC374/YffaST2DEe7rho93nAXQELyEaPjPczXrH/bBUf4w1tThuP7KB49WrYLORdVoM/S1XNRSswPK5fijfjWtWgUB0WxNbUWnUSE2xBOHL9zYuDdfdxeUtaK5XCUBbz/UH6+vO9aqAKOU3p28caqostmuiaa4uqiuC7PNkSTAVaNu0Wt1DfZEuK8b6uotKDBU40xpFQI8tBjdJwz/3dnw/2Z0n1BkZhejstZs9zvu6+6CX/cMwT3dg+Hj5oL07WeQcbywyfBw++UQnXmiGCoJdv9fdRoVHk6IxJ4zF22/Q6HerogJ8oDZIuDqosaEAZ2REO2Pb37Ox+Bof/x0shS7ckux5XJ31t3dghDq7YotJ4rwwO2d8UD/Tnj80/04WVSBLgHumJIYhdhgTyzZcgqJtwXg0aEx+ObnfFiEQIGhBnVmC8xCYEx8OEJ9XDFvzRGoJAkBHlp08nVD/0hflFXVYXiPYAANXeaFxhq8tOowOvu5I9hLhzF9w9E12NP2viwWgeV7ziE2xNPWnVZgqMZxfTmGdg3EqeJKFJeboHNRIcLPHaE+rrb9ispNCPHWQZIkfL47D2pJwth+4XBRq2CqN+PwBSP6R/oCAFzUDWHDImALFFeHC2q9dhOAnnjiCXz22WdYvXo1unfvblvu4+MDN7eGQV9TpkxBp06dsHDhQgDAK6+8gjvuuANdu3ZFWVkZ3nzzTaxatQr79u1Dr169WvS6HSEAKcFQVQd3nfq6Kxv+u+MMDl0wYNKgCHT2c7eFrHWHCvDZrjzMGtkN8Z18UGCoQXlNPY4VGBHh746BXfxwXF+O2BBP6A01uO+9bTBU1+H5lB6YMTQGKpWEU8UVmPVFFgoMNSgqN0GrUSHU2xVP/aortmQXQ6dR4c/j++B/+89f920tLtwb/SN98dCgSNSaLXjonzttXQv9I32RV1qF0sstYJIEJPUMwZ4zF+Hr5oLqOjMKjSYEeuoghLBtZ3VtC9Tv7ugCnUYFrUaFx+65DUIAaw8W4KdTJVh7sAAalYROfm4Y0jUQ1bVmfH3ggt3x3LVqSADqLeK67kgrSQI8tRr8qmcw6swW7M69aOvGBBoGN/YO98H5smqcKWkYt9EwPkRCaUWtXavTD7OGYf0hPR4aHIlTxRXYd/YSfpfYBfll1diSXYyoAHfsPH0RAR5aLN9zztblOWlgBP76m3gADV9wcooqoLr8QTG8ezDyLlZhVO9Q5Jc1dF2k/3QG+WXVmHJnF7hq1Pj7xhO4u1sQugZ7wsfNBamDI3G6uAJmi8DC9ceRHBcCD50GUQEeOF9Wjd7h3pj04U7U1lvQP9IX98WHIzbYE2aLwInCchzJN2Lm8K44dMGA0goT4jv7QquRoFGpYBYCfTr54PsjhVjz8wXkFFXAzUWND6cMxL6zl5BbXIndZ0pRZxawWATctGrEBHrgQlkNgry0mDcmrqF7VKVC1rkyXCirwtqDeix8oA8EBKYu3YMQbx0WPhCPr/adw5DYIPSL8EWFqR41dWZsOKxHkbEGGrUK3UI8MbxHMNYf0iO+sw9igq58WAshkF1YjgAPHfw9tNh4tBD+HloM7OKHnKIKHDxfhpFxofh8dx40KgkPDY5s9Nv6v7aexsWqWjzQvxPcdRqs2HMOuSWVmDEsBr07+dhe6x9bTsFsEXg4IRLeri7QalQwVNch41ghugR4oH+Eb4vGTO05cxE1dWYMjb3+aiVjTR2O5hsxsIsfB/lTq7SbANRUM9myZcswdepUAMA999yDqKgopKenAwCeffZZfP3119Dr9fDz88OAAQPw2muvoX///i1+XQagW9NxvRFH840Y16/TdX+ATfVm7Dt7CXdEBzT5x/nQeQO+O6JHakIkzGaByAD71rFKUz1cXdQ4kHfJ9kFVYKiBi1qFmjqz7UMCaBg/8u3BfNzfrxMCPbWoqbNgxn/3QiVJmD4kGtGBHqi3CHyw5RQeHRaNrsGNjycxWwS2nihGvwhf25geIQQ+3ZUHbzcXqCRAp1Hjnu5BcFGrUG+2oNJkhodODbMQcFGpUFZdh0tVtYjwc7cbVyKEwNECIzJPFCMqwAOj+1wZZ2SxCFTW1tsNcr5UWYv/98k+3NM9CE/c07WFP5WGlsI3v8tGcbkJiyffbteK2Fo1deZWX9V0sqgCLmqpRZfmysnZB38T3YraTQBSCgMQERHRrac1n99sWyQiIqIOhwGIiIiIOhwGICIiIupwGICIiIiow2EAIiIiog6HAYiIiIg6HAYgIiIi6nAYgIiIiKjDYQAiIiKiDocBiIiIiDocBiAiIiLqcBiAiIiIqMNhACIiIqIOhwGIiIiIOhyN0gU4IyEEAMBoNCpcCREREbWU9XPb+jneHAagRpSXlwMAIiIiFK6EiIiIWqu8vBw+Pj7NbiOJlsSkDsZisSA/Px9eXl6QJKnNjms0GhEREYFz587B29u7zY5L9nie5cHzLB+ea3nwPMvDkedZCIHy8nKEh4dDpWp+lA9bgBqhUqnQuXNnhx3f29ub/7lkwPMsD55n+fBcy4PnWR6OOs+/1PJjxUHQRERE1OEwABEREVGHwwAkI51Oh3nz5kGn0yldSrvG8ywPnmf58FzLg+dZHs5ynjkImoiIiDoctgARERFRh8MARERERB0OAxARERF1OAxARERE1OEwAMlk8eLFiIqKgqurKxISErB7926lS7qlbN26FWPGjEF4eDgkScKqVavs1gshMHfuXISFhcHNzQ1JSUnIycmx2+bixYuYPHkyvL294evri+nTp6OiokLGd+H8Fi5ciEGDBsHLywvBwcEYN24csrOz7bapqanBzJkzERAQAE9PT0yYMAGFhYV22+Tl5eHee++Fu7s7goOD8Yc//AH19fVyvhWnt2TJEsTHx9tuBpeYmIj169fb1vM8O8brr78OSZLwzDPP2JbxXN+8+fPnQ5Iku0ePHj1s653yHAtyuOXLlwutViuWLl0qjhw5Ih599FHh6+srCgsLlS7tlrFu3Trx4osviq+//loAECtXrrRb//rrrwsfHx+xatUq8fPPP4uxY8eK6OhoUV1dbdtm1KhRom/fvmLnzp3ixx9/FF27dhWpqakyvxPnlpycLJYtWyYOHz4ssrKyxOjRo0VkZKSoqKiwbfPYY4+JiIgIkZGRIfbu3SvuuOMOceedd9rW19fXi969e4ukpCRx4MABsW7dOhEYGCjmzJmjxFtyWmvWrBFr164VJ06cENnZ2eKFF14QLi4u4vDhw0IInmdH2L17t4iKihLx8fHi6aefti3nub558+bNE3FxcaKgoMD2KC4utq13xnPMACSDwYMHi5kzZ9qem81mER4eLhYuXKhgVbeuawOQxWIRoaGh4s0337QtKysrEzqdTnz++edCCCGOHj0qAIg9e/bYtlm/fr2QJElcuHBBttpvNUVFRQKAyMzMFEI0nFcXFxfx5Zdf2rY5duyYACB27NghhGgIqyqVSuj1ets2S5YsEd7e3sJkMsn7Bm4xfn5+4t///jfPswOUl5eL2NhYsXHjRnH33XfbAhDPdduYN2+e6Nu3b6PrnPUcswvMwWpra7Fv3z4kJSXZlqlUKiQlJWHHjh0KVtZ+5ObmQq/X251jHx8fJCQk2M7xjh074Ovri4EDB9q2SUpKgkqlwq5du2Sv+VZhMBgAAP7+/gCAffv2oa6uzu5c9+jRA5GRkXbnuk+fPggJCbFtk5ycDKPRiCNHjshY/a3DbDZj+fLlqKysRGJiIs+zA8ycORP33nuv3TkF+DvdlnJychAeHo6YmBhMnjwZeXl5AJz3HHMyVAcrKSmB2Wy2+6ECQEhICI4fP65QVe2LXq8HgEbPsXWdXq9HcHCw3XqNRgN/f3/bNmTPYrHgmWeewV133YXevXsDaDiPWq0Wvr6+dttee64b+1lY19EVhw4dQmJiImpqauDp6YmVK1eiV69eyMrK4nluQ8uXL8f+/fuxZ8+e69bxd7ptJCQkID09Hd27d0dBQQEWLFiAoUOH4vDhw057jhmAiKhRM2fOxOHDh7Ft2zalS2m3unfvjqysLBgMBnz11VdIS0tDZmam0mW1K+fOncPTTz+NjRs3wtXVVely2q2UlBTbv+Pj45GQkIAuXbpgxYoVcHNzU7CyprELzMECAwOhVquvG+1eWFiI0NBQhapqX6znsblzHBoaiqKiIrv19fX1uHjxIn8OjXjyySfx7bffYvPmzejcubNteWhoKGpra1FWVma3/bXnurGfhXUdXaHVatG1a1cMGDAACxcuRN++ffHOO+/wPLehffv2oaioCLfffjs0Gg00Gg0yMzPx7rvvQqPRICQkhOfaAXx9fdGtWzecPHnSaX+fGYAcTKvVYsCAAcjIyLAts1gsyMjIQGJiooKVtR/R0dEIDQ21O8dGoxG7du2ynePExESUlZVh3759tm02bdoEi8WChIQE2Wt2VkIIPPnkk1i5ciU2bdqE6Ohou/UDBgyAi4uL3bnOzs5GXl6e3bk+dOiQXeDcuHEjvL290atXL3neyC3KYrHAZDLxPLehESNG4NChQ8jKyrI9Bg4ciMmTJ9v+zXPd9ioqKnDq1CmEhYU57++zQ4ZWk53ly5cLnU4n0tPTxdGjR8WMGTOEr6+v3Wh3al55ebk4cOCAOHDggAAgFi1aJA4cOCDOnj0rhGi4DN7X11esXr1aHDx4UNx///2NXgbfv39/sWvXLrFt2zYRGxvLy+Cv8fjjjwsfHx+xZcsWu8tZq6qqbNs89thjIjIyUmzatEns3btXJCYmisTERNt66+WsI0eOFFlZWWLDhg0iKCiIlwxf4/nnnxeZmZkiNzdXHDx4UDz//PNCkiTx/fffCyF4nh3p6qvAhOC5bguzZ88WW7ZsEbm5ueKnn34SSUlJIjAwUBQVFQkhnPMcMwDJ5L333hORkZFCq9WKwYMHi507dypd0i1l8+bNAsB1j7S0NCFEw6XwL7/8sggJCRE6nU6MGDFCZGdn2x2jtLRUpKamCk9PT+Ht7S2mTZsmysvLFXg3zquxcwxALFu2zLZNdXW1eOKJJ4Sfn59wd3cX48ePFwUFBXbHOXPmjEhJSRFubm4iMDBQzJ49W9TV1cn8bpzbI488Irp06SK0Wq0ICgoSI0aMsIUfIXieHenaAMRzffMmTZokwsLChFarFZ06dRKTJk0SJ0+etK13xnMsCSGEY9qWiIiIiJwTxwARERFRh8MARERERB0OAxARERF1OAxARERE1OEwABEREVGHwwBEREREHQ4DEBEREXU4DEBERETU4TAAERE1QZIkrFq1SukyiMgBGICIyClNnToVkiRd9xg1apTSpRFRO6BRugAioqaMGjUKy5Yts1um0+kUqoaI2hO2ABGR09LpdAgNDbV7+Pn5AWjonlqyZAlSUlLg5uaGmJgYfPXVV3b7Hzp0CL/61a/g5uaGgIAAzJgxAxUVFXbbLF26FHFxcdDpdAgLC8OTTz5pt76kpATjx4+Hu7s7YmNjsWbNGtu6S5cuYfLkyQgKCoKbmxtiY2OvC2xE5JwYgIjolvXyyy9jwoQJ+PnnnzF58mQ89NBDOHbsGACgsrISycnJ8PPzw549e/Dll1/ihx9+sAs4S5YswcyZMzFjxgwcOnQIa9asQdeuXe1eY8GCBXjwwQdx8OBBjB49GpMnT8bFixdtr3/06FGsX78ex44dw5IlSxAYGCjfCSCiG+eweeaJiG5CWlqaUKvVwsPDw+7x5z//WQghBADx2GOP2e2TkJAgHn/8cSGEEB9++KHw8/MTFRUVtvVr164VKpVK6PV6IYQQ4eHh4sUXX2yyBgDipZdesj2vqKgQAMT69euFEEKMGTNGTJs2rW3eMBHJimOAiMhpDR8+HEuWLLFb5u/vb/t3YmKi3brExERkZWUBAI4dO4a+ffvCw8PDtv6uu+6CxWJBdnY2JElCfn4+RowY0WwN8fHxtn97eHjA29sbRUVFAIDHH38cEyZMwP79+zFy5EiMGzcOd9555w29VyKSFwMQETktDw+P67qk2oqbm1uLtnNxcbF7LkkSLBYLACAlJQVnz57FunXrsHHjRowYMQIzZ87E3/72tzavl4jaFscAEdEta+fOndc979mzJwCgZ8+e+Pnnn1FZWWlb/9NPP0GlUqF79+7w8vJCVFQUMjIybqqGoKAgpKWl4ZNPPsHbb7+NDz/88KaOR0TyYAsQETktk8kEvV5vt0yj0dgGGn/55ZcYOHAghgwZgk8//RS7d+/GRx99BACYPHky5s2bh7S0NMyfPx/FxcV46qmn8Lvf/Q4hISEAgPnz5+Oxxx5DcHAwUlJSUF5ejp9++glPPfVUi+qbO3cuBgwYgLi4OJhMJnz77be2AEZEzo0BiIic1oYNGxAWFma3rHv37jh+/DiAhiu0li9fjieeeAJhYWH4/PPP0atXLwCAu7s7vvvuOzz99NMYNGgQ3N3dMWHCBCxatMh2rLS0NNTU1OCtt97Cc889h8DAQPzmN79pcX1arRZz5szBmTNn4ObmhqFDh2L58uVt8M6JyNEkIYRQuggiotaSJAkrV67EuHHjlC6FiG5BHANEREREHQ4DEBEREXU4HANERLck9t4T0c1gCxARERF1OAxARERE1OEwABEREVGHwwBEREREHQ4DEBEREXU4DEBERETU4TAAERERUYfDAEREREQdzv8H8mM8NQEnHzMAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.plot(range(1, len(average_mae_history) + 1), average_mae_history)\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Validation MAE\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3XOIUQgSPqv2"
      },
      "source": [
        "**Plotting validation scores, excluding the first 10 data points**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FkQ5ziacPqv2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "outputId": "e5d83977-4e36-48cf-c0c3-967d5ecbcdb8"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "truncated_mae_history = average_mae_history[10:]\n",
        "plt.plot(range(1, len(truncated_mae_history) + 1), truncated_mae_history)\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Validation MAE\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1UX2bMwlPqv2"
      },
      "source": [
        "**Training the final model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dEopCmnaPqv2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "991ec7e5-e39c-48af-bbcd-3e160a809a92"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 4ms/step - loss: 13.5250 - mae: 2.3609\n"
          ]
        }
      ],
      "source": [
        "model = build_model()\n",
        "model.fit(train_data, train_targets,\n",
        "          epochs=130, batch_size=16, verbose=0)\n",
        "test_mse_score, test_mae_score = model.evaluate(test_data, test_targets)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "z6afCYwDPqv2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "771a5c4f-f431-4b51-8027-20b5a00d2714"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2.3609158992767334"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ],
      "source": [
        "test_mae_score"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S57TFN2mPqv2"
      },
      "source": [
        "### Generating predictions on new data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "rRTzYEuQPqv2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "327a1b13-8bca-4c68-f1a8-d0e36ebeec82"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4/4 [==============================] - 0s 2ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([8.507302], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "predictions = model.predict(test_data)\n",
        "predictions[0]"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "uaBz8ZvTWmR6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s-cpR_JlP0rC"
      },
      "source": [
        "# Fundamentals of machine learning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A1OoXVRSP0rC"
      },
      "source": [
        "## Generalization: The goal of machine learning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ph__UvnDP0rC"
      },
      "source": [
        "### Underfitting and overfitting"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XnNsMOxjP0rC"
      },
      "source": [
        "#### Noisy training data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "--6KClhcP0rD"
      },
      "source": [
        "#### Ambiguous features"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9cAd_QPeP0rD"
      },
      "source": [
        "#### Rare features and spurious correlations"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HdqS86_SP0rD"
      },
      "source": [
        "**Adding white-noise channels or all-zeros channels to MNIST**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TXy195b6P0rD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b56d4613-b8a9-4558-f48a-2fcb8218a3da"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11490434/11490434 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras.datasets import mnist\n",
        "import numpy as np\n",
        "\n",
        "(train_images, train_labels), _ = mnist.load_data()\n",
        "train_images = train_images.reshape((60000, 28 * 28))\n",
        "train_images = train_images.astype(\"float32\") / 255\n",
        "\n",
        "train_images_with_noise_channels = np.concatenate(\n",
        "    [train_images, np.random.random((len(train_images), 784))], axis=1)\n",
        "\n",
        "train_images_with_zeros_channels = np.concatenate(\n",
        "    [train_images, np.zeros((len(train_images), 784))], axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_images.shape)\n",
        "print(train_images_with_noise_channels.shape)\n",
        "print(train_images_with_zeros_channels.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R0A8wyOhXRrg",
        "outputId": "5ac889c6-20bd-41d0-9572-1a0f15130004"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 784)\n",
            "(60000, 1568)\n",
            "(60000, 1568)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9I_6W908P0rE"
      },
      "source": [
        "**Training the same model on MNIST data with noise channels or all-zero channels**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "S_2_Xz_lP0rE",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "01a2a3b3-40ed-4d1e-c133-ba17358a8a5d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.6142 - accuracy: 0.8125 - val_loss: 0.3400 - val_accuracy: 0.8953\n",
            "Epoch 2/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2565 - accuracy: 0.9211 - val_loss: 0.1782 - val_accuracy: 0.9488\n",
            "Epoch 3/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.1684 - accuracy: 0.9475 - val_loss: 0.1679 - val_accuracy: 0.9509\n",
            "Epoch 4/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.1191 - accuracy: 0.9626 - val_loss: 0.1682 - val_accuracy: 0.9486\n",
            "Epoch 5/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.0880 - accuracy: 0.9721 - val_loss: 0.1238 - val_accuracy: 0.9642\n",
            "Epoch 6/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0639 - accuracy: 0.9797 - val_loss: 0.1157 - val_accuracy: 0.9671\n",
            "Epoch 7/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0464 - accuracy: 0.9853 - val_loss: 0.1149 - val_accuracy: 0.9679\n",
            "Epoch 8/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0350 - accuracy: 0.9887 - val_loss: 0.1495 - val_accuracy: 0.9576\n",
            "Epoch 9/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0266 - accuracy: 0.9915 - val_loss: 0.1372 - val_accuracy: 0.9654\n",
            "Epoch 10/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0198 - accuracy: 0.9942 - val_loss: 0.1258 - val_accuracy: 0.9688\n",
            "Epoch 1/10\n",
            "375/375 [==============================] - 3s 5ms/step - loss: 0.2939 - accuracy: 0.9155 - val_loss: 0.1550 - val_accuracy: 0.9545\n",
            "Epoch 2/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.1213 - accuracy: 0.9650 - val_loss: 0.1061 - val_accuracy: 0.9688\n",
            "Epoch 3/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0795 - accuracy: 0.9763 - val_loss: 0.0836 - val_accuracy: 0.9743\n",
            "Epoch 4/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0580 - accuracy: 0.9833 - val_loss: 0.0840 - val_accuracy: 0.9743\n",
            "Epoch 5/10\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0426 - accuracy: 0.9878 - val_loss: 0.0742 - val_accuracy: 0.9781\n",
            "Epoch 6/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0322 - accuracy: 0.9910 - val_loss: 0.0820 - val_accuracy: 0.9780\n",
            "Epoch 7/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0243 - accuracy: 0.9933 - val_loss: 0.0743 - val_accuracy: 0.9787\n",
            "Epoch 8/10\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0182 - accuracy: 0.9948 - val_loss: 0.0792 - val_accuracy: 0.9787\n",
            "Epoch 9/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0140 - accuracy: 0.9962 - val_loss: 0.0730 - val_accuracy: 0.9798\n",
            "Epoch 10/10\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0104 - accuracy: 0.9974 - val_loss: 0.0774 - val_accuracy: 0.9803\n",
            "Epoch 1/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.2945 - accuracy: 0.9154 - val_loss: 0.1471 - val_accuracy: 0.9576\n",
            "Epoch 2/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.1222 - accuracy: 0.9637 - val_loss: 0.1070 - val_accuracy: 0.9687\n",
            "Epoch 3/10\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0811 - accuracy: 0.9764 - val_loss: 0.0889 - val_accuracy: 0.9738\n",
            "Epoch 4/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.0586 - accuracy: 0.9826 - val_loss: 0.0783 - val_accuracy: 0.9768\n",
            "Epoch 5/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0430 - accuracy: 0.9874 - val_loss: 0.0827 - val_accuracy: 0.9747\n",
            "Epoch 6/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0316 - accuracy: 0.9914 - val_loss: 0.0811 - val_accuracy: 0.9751\n",
            "Epoch 7/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0239 - accuracy: 0.9934 - val_loss: 0.0754 - val_accuracy: 0.9796\n",
            "Epoch 8/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0177 - accuracy: 0.9949 - val_loss: 0.0781 - val_accuracy: 0.9777\n",
            "Epoch 9/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0132 - accuracy: 0.9966 - val_loss: 0.0732 - val_accuracy: 0.9794\n",
            "Epoch 10/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0100 - accuracy: 0.9975 - val_loss: 0.0797 - val_accuracy: 0.9780\n"
          ]
        }
      ],
      "source": [
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "def get_model():\n",
        "    model = keras.Sequential([\n",
        "        layers.Dense(512, activation=\"relu\"),\n",
        "        layers.Dense(10, activation=\"softmax\")\n",
        "    ])\n",
        "    model.compile(optimizer=\"rmsprop\",\n",
        "                  loss=\"sparse_categorical_crossentropy\",\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model\n",
        "\n",
        "\n",
        "\n",
        "model = get_model()\n",
        "history_noise = model.fit(\n",
        "    train_images_with_noise_channels, train_labels,\n",
        "    epochs=10,\n",
        "    batch_size=128,\n",
        "    validation_split=0.2)\n",
        "\n",
        "model = get_model()\n",
        "history_zeros = model.fit(\n",
        "    train_images_with_zeros_channels, train_labels,\n",
        "    epochs=10,\n",
        "    batch_size=128,\n",
        "    validation_split=0.2)\n",
        "\n",
        "model = get_model()\n",
        "history = model.fit(\n",
        "    train_images, train_labels,\n",
        "    epochs=10,\n",
        "    batch_size=128,\n",
        "    validation_split=0.2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aRPQ11HQP0rE"
      },
      "source": [
        "**Plotting a validation accuracy comparison**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "61Ixm2_XP0rE",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "outputId": "4b29c5df-1e9e-45f3-f1c1-d6e3d5970042"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7b528c0ab160>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAHHCAYAAABXx+fLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACdQ0lEQVR4nOzdd1hT1xsH8G/YIFNAlgiIKKgIKuAqgqs4K6gVRxW3dY9q1ap11drh3qOKFvfWah2IG/cAB6iIOIogigKispLz++P8khASRhQI4/08Tx6Tm5N733uJ3JczBYwxBkIIIYSQSkRN1QEQQgghhJQ2SoAIIYQQUulQAkQIIYSQSocSIEIIIYRUOpQAEUIIIaTSoQSIEEIIIZUOJUCEEEIIqXQoASKEEEJIpUMJECGEEEIqHUqASL7S09MxZMgQWFpaQiAQYPz48QCAV69eoUePHjA1NYVAIMDSpUtVGqcy8jun0nD27FkIBAKcPXu21I5ZFJs3b4ZAIMCNGzdUHUqxEJ/P06dPVR2Kytjb22PAgAGS18p893x9feHr61us8cyePRsCgaBY90nIl9JQdQCkdG3evBkDBw7M9/3Lly+jadOmAIBff/0VmzdvxsyZM+Ho6AgXFxcAwIQJE3DixAnMmjULlpaW8PDwKPY4f/31V9StWxf+/v7Fvl9F50QI+TIfP37EH3/8USIJFCElgRKgSmru3LlwcHCQ216rVi3J89OnT6Np06aYNWuWTJnTp0+ja9eumDRpUonF9+uvv6JHjx7FngDld06loWXLlvj06RO0tLRK/dikciuN797Hjx8xZ84cAJBLgGbMmIGpU6eW2LEJ+RyUAFVSHTp0KLTmJikpCXXr1lW43djYuIQiK1n5nVNpUFNTg46OjkqOTSo3VX/3NDQ0oKFBt5vCZGRkQEtLC2pq1DulNNBVJnLE/QXi4uJw9OhRCAQCCAQCSd8KxhhWrVol2S6WkpKC8ePHw9bWFtra2qhVqxZ+//13iEQimf2LRCIsW7YMrq6u0NHRgbm5Odq3by/pgyIQCPDhwwds2bJFcozc/RkUSUpKwuDBg2FhYQEdHR24ublhy5YthZ5TQf1EBAIBRo8ejYMHD6J+/frQ1tZGvXr1cPz4cbmyt2/fRocOHWBoaAh9fX20adMGV65cUXhdc/fDiImJQffu3WFpaQkdHR1Ur14dvXr1Qmpqqsxnt27disaNG0NXVxdVq1ZFr1698OLFiwKviVh8fDwGDx4Ma2traGtrw8HBASNGjEBWVpZMuczMTEycOBHm5uaoUqUKAgIC8Pr1a5kyhw4dQqdOnST7cnR0xLx58yAUCmXK+fr6on79+oiKikKrVq2gp6cHGxsb/PHHHwqvye7duzF//nxUr14dOjo6aNOmDR4/fix3LlevXkX79u1hZGQEPT09+Pj4IDw8vNBrcOPGDfj5+cHMzAy6urpwcHDAoEGDinT9Vq9ejXr16kFbWxvW1tYYNWoUUlJSPut8Falfvz5atWolt10kEsHGxgY9evSQbFu4cCGaN28OU1NT6OrqonHjxti7d2+hx8ivD9D69evh6OgIXV1deHl54cKFC3KfzcrKws8//4zGjRvDyMgIVapUgbe3N86cOSMp8/TpU5ibmwMA5syZI/n/NXv2bACK+wDl5ORg3rx5cHR0hLa2Nuzt7fHTTz8hMzNTppy9vT06d+6MixcvwsvLCzo6OqhZsyb+/vvvQs8bUO6abd26FV5eXtDT04OJiQlatmyJkydPypQ5duwYfHx8YGBgAENDQ3h6emL79u0y8Sr6fZW3aVD8M9m5cydmzJgBGxsb6OnpIS0tDW/fvsWkSZPg6uoKfX19GBoaokOHDoiMjJTbb0ZGBmbPno3atWtDR0cHVlZW6NatG2JjY8EYg729Pbp27arwc0ZGRhg+fHiRrmOFxEilEhwczACwU6dOsdevX8s83rx5wxhjLDExkYWEhDAzMzPm7u7OQkJCWEhICLt37x4LCQlhAFi7du0k2xlj7MOHD6xBgwbM1NSU/fTTT2zt2rWsf//+TCAQsHHjxsnEMGDAAAaAdejQgS1dupQtXLiQde3ala1YsYIxxlhISAjT1tZm3t7ekmNcunQp33P6+PEjc3FxYZqammzChAls+fLlzNvbmwFgS5cuLfCc0tPT890vAObm5sasrKzYvHnz2NKlS1nNmjWZnp6e5Foxxti9e/dYlSpVJOV+++035uDgwLS1tdmVK1ck5c6cOcMAsDNnzjDGGMvMzGQODg7M2tqa/fLLL+yvv/5ic+bMYZ6enuzp06eSz/3yyy9MIBCwwMBAtnr1ajZnzhxmZmbG7O3t2bt37wr8ecfHxzNra2ump6fHxo8fz9auXctmzpzJXFxcJJ8VfycaNmzIWrduzVasWMF++OEHpq6uznr27CmzP39/f9azZ0/2559/sjVr1rBvv/2WAWCTJk2SKefj48Osra2Zra0tGzduHFu9ejVr3bo1A8D+/fdfuWvSsGFD1rhxY7ZkyRI2e/Zspqenx7y8vGT2GRYWxrS0tFizZs3YokWL2JIlS1iDBg2YlpYWu3r1qqSc+Hzi4uIYY4y9evWKmZiYsNq1a7M///yTbdiwgU2fPp25uLgUeO0YY2zWrFkMAGvbti1bsWIFGz16NFNXV2eenp4sKytL6fNVZO7cuUxNTY0lJCTIbD937hwDwPbs2SPZVr16dTZy5Ei2cuVKtnjxYubl5cUAsCNHjsh81s7OjgUFBcldZ/F3jzHG/vrrLwaANW/enC1fvpyNHz+eGRsbs5o1azIfHx9JudevXzMrKys2ceJEtmbNGvbHH3+wOnXqME1NTXb79m3GGGPp6elszZo1DAALCAiQ/P+KjIyUuY65BQUFMQCsR48ebNWqVax///4MAPP395c7lzp16jALCwv2008/sZUrV7JGjRoxgUDA7t27V+C1VeaazZ49W3I9/vzzT7Zs2TLWp08fNmXKFEmZ4OBgJhAIWP369dn8+fPZqlWr2JAhQ1i/fv3yvfZiPj4+MtdV/DOpW7cuc3d3Z4sXL2YLFixgHz58YNevX2eOjo5s6tSpbN26dWzu3LnMxsaGGRkZsfj4eMk+cnJyWJs2bRgA1qtXL7Zy5Uq2YMEC1rp1a3bw4EHGGGPTp09nmpqaLDk5WSae3bt3MwDs/PnzhV7DiooSoEpGfHNQ9NDW1pYpa2dnxzp16iS3DwBs1KhRMtvmzZvHqlSpwh49eiSzferUqUxdXZ09f/6cMcbY6dOnGQA2duxYuf2KRCLJ8ypVqij8JaLI0qVLGQC2detWybasrCzWrFkzpq+vz9LS0go9J0UAMC0tLfb48WPJtsjISAZAkqwxxpMCLS0tFhsbK9n28uVLZmBgwFq2bCnZlvcmdPv2bbkbXF5Pnz5l6urqbP78+TLb7969yzQ0NOS259W/f3+mpqbGrl+/Lvee+HqLvxNt27aV+RlMmDCBqaurs5SUFMm2jx8/yu1n+PDhTE9Pj2VkZEi2+fj4MADs77//lmzLzMxklpaWrHv37pJt4mvi4uLCMjMzJduXLVvGALC7d+9KYnVycmJ+fn4yMX78+JE5ODiwdu3aSbblTYAOHDjAACi8BgVJSkpiWlpa7Ouvv2ZCoVCyfeXKlQwA27Rpk9Lnq8jDhw/lvlOMMTZy5Eimr68vc83zXv+srCxWv3591rp1a5nthSVAWVlZrFq1aszd3V3muq9fv54BkLlR5+TkyJRhjLF3794xCwsLNmjQIMm2169fMwBs1qxZcueYNwGKiIhgANiQIUNkyk2aNIkBYKdPn5Y5l7w36qSkJKatrc1++OEHuWPlVZRrFhMTw9TU1FhAQIDMz5ox6f+TlJQUZmBgwJo0acI+ffqksIw4XmUSoJo1a8rFmJGRIRdHXFwc09bWZnPnzpVs27RpEwPAFi9eLHc8cUzi79eaNWtk3v/mm2+Yvb29TOyVDTWBVVKrVq1CaGiozOPYsWOfvb89e/bA29sbJiYmePPmjeTRtm1bCIVCnD9/HgCwb98+CAQChZ2QP3eY7L///gtLS0v07t1bsk1TUxNjx45Feno6zp0793knBaBt27ZwdHSUvG7QoAEMDQ3x5MkTAIBQKMTJkyfh7++PmjVrSspZWVmhT58+uHjxItLS0hTu28jICABw4sQJfPz4UWGZ/fv3QyQSoWfPnjLX1dLSEk5OTjLNEHmJRCIcPHgQXbp0UdjfK+/1HjZsmMw2b29vCIVCPHv2TLJNV1dX8vz9+/d48+YNvL298fHjRzx48EBmf/r6+vjuu+8kr7W0tODl5SW5drkNHDhQpoOut7c3AEjKRkREICYmBn369EFycrLkOnz48AFt2rTB+fPn5ZpaxcT91Y4cOYLs7GyFZRQ5deoUsrKyMH78eJk+GUOHDoWhoSGOHj362eebW+3ateHu7o5du3ZJtgmFQuzduxddunSRuea5n7979w6pqanw9vbGrVu3inxeAG8STEpKwvfffy9z3QcMGCD5Xoqpq6tLyohEIrx9+xY5OTnw8PBQ+rhi//77LwBg4sSJMtt/+OEHAJC7tnXr1pV8JwDA3NwcderUKfTaAkW7ZgcPHoRIJMLPP/8s1/9G/H8iNDQU79+/x9SpU+X6U33JEP+goCCZGAFAW1tbEodQKERycjL09fVRp04dmbj37dsHMzMzjBkzRm6/4phq166NJk2aYNu2bZL33r59i2PHjqFv376VenoC6pVWSXl5eRXr8PWYmBjcuXNH0g8gr6SkJABAbGwsrK2tUbVq1WI79rNnz+Dk5CT3i0s8xD33DVxZNWrUkNtmYmKCd+/eAQBev36Njx8/ok6dOnLlXFxcIBKJ8OLFC9SrV0/ufQcHB0ycOBGLFy/Gtm3b4O3tjW+++Qbfffed5CYUExMDxhicnJwUxqepqZlv7K9fv0ZaWhrq16//WedqYmICAJJzBYD79+9jxowZOH36tFxil7ffUvXq1eV+uZqYmODOnTtKHzsmJgYAv1nkJzU1VfK53Hx8fNC9e3fMmTMHS5Ysga+vL/z9/dGnTx9oa2vnuz/x9ybvz1ZLSws1a9aU+14pc755BQYG4qeffkJ8fDxsbGxw9uxZJCUlITAwUKbckSNH8MsvvyAiIkKmr4yyNzFx7Hm/V5qamjKJvNiWLVuwaNEiPHjwQCaJVDSStKjHV1NTkxl1CgCWlpYwNjaWu7aF/T8sSFGuWWxsLNTU1AocIBEbGwsARf7/VFSKrqG4n+Tq1asRFxcn08fO1NRUJqY6deoU2sG8f//+GD16NJ49ewY7Ozvs2bMH2dnZ6NevX/GdSDlECRApFiKRCO3atcOPP/6o8P3atWuXckTFQ11dXeF2xlix7H/RokUYMGAADh06hJMnT2Ls2LFYsGABrly5gurVq0MkEkEgEODYsWMKY9HX1y+WOIDCzzUlJQU+Pj4wNDTE3Llz4ejoCB0dHdy6dQtTpkyRq4FR5toVVla87z///BPu7u4Ky+Z3LQQCAfbu3YsrV67gn3/+wYkTJzBo0CAsWrQIV65cKbZr+CXflcDAQEybNg179uzB+PHjsXv3bhgZGaF9+/aSMhcuXMA333yDli1bYvXq1bCysoKmpiaCg4NlOuEWt61bt2LAgAHw9/fH5MmTUa1aNairq2PBggWSpOBzFTVx+9xrq4prlt85CYVCheeRt/YH4NOAzJw5E4MGDcK8efNQtWpVqKmpYfz48fnWdBakV69emDBhArZt24affvoJW7duhYeHh8I/3CoTSoBIsXB0dER6ejratm1baLkTJ07g7du3BdYCKfMXrZ2dHe7cuQORSCRTCyRukrGzsyvyvpRlbm4OPT09PHz4UO69Bw8eQE1NDba2tgXuw9XVFa6urpgxYwYuXbqEFi1aYO3atfjll1/g6OgIxhgcHByUTiLNzc1haGiIe/fuKfW5/Jw9exbJycnYv38/WrZsKdkeFxdXLPsviLgZ0tDQsNDvWH6aNm2Kpk2bYv78+di+fTv69u2LnTt3YsiQIQrLi783Dx8+lKkVycrKQlxc3GfHoYiDgwO8vLywa9cujB49Gvv374e/v79MDdW+ffugo6ODEydOyGwPDg5W+njic4uJiUHr1q0l27OzsxEXFwc3NzfJtr1796JmzZrYv3+/zP/LvM3Yyv6fFYlEiImJkZmM9NWrV0hJSSm2/7NFvWaOjo4QiUSIiorKN8EWfwfv3bsnV3OVm4mJidwoQYDXeimqXVNk7969aNWqFTZu3CizPSUlBWZmZjIxXb16FdnZ2QXWBletWhWdOnXCtm3b0LdvX4SHh5erGfxLCvUBIsWiZ8+euHz5Mk6cOCH3XkpKCnJycgAA3bt3B2NMMmFabrn/mqtSpYrCXyKKdOzYEYmJiTJ9KHJycrBixQro6+vDx8dHybMpOnV1dXz99dc4dOiQzJD6V69eYfv27fjqq69gaGio8LNpaWmS6yLm6uoKNTU1SVV9t27doK6ujjlz5sj9tcsYQ3Jycr6xqampwd/fH//884/CZS6UrcUS//Wa+3NZWVlYvXq1Uvv5HI0bN4ajoyMWLlyI9PR0uffzDtfP7d27d3LnKr7J5R1ynVvbtm2hpaWF5cuXy3x+48aNSE1NRadOnZQ8i4IFBgbiypUr2LRpE968eSPX/KWurg6BQCDTHPL06VMcPHhQ6WN5eHjA3Nwca9eulZkOYfPmzXL/7xT93K9evYrLly/LlNPT0wOAIv2/7dixIwDI3YQXL14MAMV2bYt6zfz9/aGmpoa5c+fK1bCIz/vrr7+GgYEBFixYgIyMDIVlAJ6UXLlyRea6HjlypMjTVojjzvud3bNnD+Lj42W2de/eHW/evMHKlSvl9pH38/369UNUVBQmT54MdXV19OrVq8jxVFRUA1RJHTt2TK7TKgA0b968yH+l5DZ58mQcPnwYnTt3xoABA9C4cWN8+PABd+/exd69e/H06VOYmZmhVatW6NevH5YvX46YmBi0b98eIpEIFy5cQKtWrTB69GgA/IZ36tQpLF68GNbW1nBwcECTJk0UHnvYsGFYt24dBgwYgJs3b8Le3h579+6V/JVjYGCg9Pko45dffkFoaCi++uorjBw5EhoaGli3bh0yMzMLnAfm9OnTGD16NL799lvUrl0bOTk5CAkJgbq6Orp37w6A/zL95ZdfMG3aNDx9+hT+/v4wMDBAXFwcDhw4gGHDhhU4I/evv/6KkydPwsfHB8OGDYOLiwsSEhKwZ88eXLx4UakJLZs3bw4TExMEBQVh7NixEAgECAkJKbbmwIKoqanhr7/+QocOHVCvXj0MHDgQNjY2iI+Px5kzZ2BoaIh//vlH4We3bNmC1atXIyAgAI6Ojnj//j02bNgAQ0NDyY1YEXNzc0ybNg1z5sxB+/bt8c033+Dhw4dYvXo1PD09ZTo8F4eePXti0qRJmDRpEqpWrSpXw9SpUycsXrwY7du3R58+fZCUlIRVq1ahVq1aRepnlJumpiZ++eUXDB8+HK1bt0ZgYCDi4uIQHBws9/+/c+fO2L9/PwICAtCpUyfExcVh7dq1qFu3rkwyqquri7p162LXrl2oXbs2qlativr16yvsM+Pm5oagoCCsX79e0rR67do1bNmyBf7+/grnRfocRb1mtWrVwvTp0zFv3jx4e3ujW7du0NbWxvXr12FtbY0FCxbA0NAQS5YswZAhQ+Dp6Yk+ffrAxMQEkZGR+Pjxo2TesSFDhmDv3r1o3749evbsidjYWGzdulVmMEVhOnfujLlz52LgwIFo3rw57t69i23btsn9bPr374+///4bEydOxLVr1+Dt7Y0PHz7g1KlTGDlypMz8P506dYKpqSn27NmDDh06oFq1al94dSuA0hxyRlSvoGHwAFhwcLCkrDLD4Blj7P3792zatGmsVq1aTEtLi5mZmbHmzZuzhQsXysyZkpOTw/7880/m7OzMtLS0mLm5OevQoQO7efOmpMyDBw9Yy5Ytma6uLgNQ6JD4V69esYEDBzIzMzOmpaXFXF1dZc6lsHNSJL/zVDTM9datW8zPz4/p6+szPT091qpVK7m5i/IORX7y5AkbNGgQc3R0ZDo6Oqxq1aqsVatW7NSpU3LH3LdvH/vqq69YlSpVWJUqVZizszMbNWoUe/jwYaHn8ezZM9a/f39mbm7OtLW1Wc2aNdmoUaMkQ5vF34m8w8QVzR0THh7OmjZtynR1dZm1tTX78ccf2YkTJ+TK+fj4sHr16snFEhQUxOzs7OSOkXcqgLi4OLnvI2N86oBu3boxU1NTpq2tzezs7FjPnj1ZWFiYpEzeYfC3bt1ivXv3ZjVq1GDa2tqsWrVqrHPnzuzGjRuFXjvG+LB3Z2dnpqmpySwsLNiIESPk5l8q6vkWpkWLFgqHh4tt3LiROTk5MW1tbebs7MyCg4MVzrFTlHmAGGNs9erVkjmrPDw82Pnz5+WGa4tEIvbrr78yOzs7pq2tzRo2bMiOHDmi8NwuXbrEGjduzLS0tGSGxCuKMTs7m82ZM4c5ODgwTU1NZmtry6ZNmyYznYL4XBT9n80bZ36Kes0Y48PKGzZsyLS1tZmJiQnz8fFhoaGhMmUOHz7MmjdvznR1dZmhoSHz8vJiO3bskCmzaNEiZmNjw7S1tVmLFi3YjRs38h0Gr2gajIyMDPbDDz8wKysrpqury1q0aMEuX76s8Jw/fvzIpk+fLrmOlpaWrEePHjLTcoiNHDmSAWDbt28v9LpVBgLGSuHPN0IIIYSo1IQJE7Bx40YkJiZKmiwrM+oDRAghhFRwGRkZ2Lp1K7p3707Jz/9RHyBCCCGkgkpKSsKpU6ewd+9eJCcnY9y4caoOqcygBIgQQgipoKKiotC3b19Uq1YNy5cvz3eYf2VEfYAIIYQQUulQHyBCCCGEVDqUABFCCCGk0qE+QAqIRCK8fPkSBgYGlXqlXEIIIaQ8YYzh/fv3sLa2llsgOy9KgBR4+fJloes3EUIIIaRsevHiBapXr15gGUqAFBAvnfDixYt813EihBBCSNmSlpYGW1vbIi2BRAmQAuJmL0NDQ0qACCGEkHKmKN1XqBM0IYQQQiodSoAIIYQQUulQAkQIIYSQSocSIEIIIYRUOpQAEUIIIaTSoQSIEEIIIZUOJUCEEEIIqXQoASKEEEJIpUMJECGEEEIqHUqACCGEEFLpUAJECCGEkEpH5QnQqlWrYG9vDx0dHTRp0gTXrl3Lt2x2djbmzp0LR0dH6OjowM3NDcePH5cpIxQKMXPmTDg4OEBXVxeOjo6YN28eGGMlfSqEEEIIKSdUmgDt2rULEydOxKxZs3Dr1i24ubnBz88PSUlJCsvPmDED69atw4oVKxAVFYXvv/8eAQEBuH37tqTM77//jjVr1mDlypWIjo7G77//jj/++AMrVqwordMihBBCSC6MAVlZQFoaf5QFAqbCqpEmTZrA09MTK1euBACIRCLY2tpizJgxmDp1qlx5a2trTJ8+HaNGjZJs6969O3R1dbF161YAQOfOnWFhYYGNGzfmW6YwaWlpMDIyQmpqKq0GTwghpELKyQGePwc+fQIyMvhD/PzTJ8DBAfDw4GXT04FFi+TLiD/n6wtMmCAt27ChfDmRiL8fGAjs3Fky56TM/VujZEIoXFZWFm7evIlp06ZJtqmpqaFt27a4fPmyws9kZmZCR0dHZpuuri4uXrwoed28eXOsX78ejx49Qu3atREZGYmLFy9i8eLF+caSmZmJzMxMyeu0spKeEkIIqfRSUoDz54GkJODNG+DjR9nEokMHoFs3XvbpU6BvX/nkQ/zvmDHAn3/ysvHxgKNj/scdPlyaAGVmArNn51/WwED6XFMTePw4/7KfPhXhpEuByhKgN2/eQCgUwsLCQma7hYUFHjx4oPAzfn5+WLx4MVq2bAlHR0eEhYVh//79EAqFkjJTp05FWloanJ2doa6uDqFQiPnz56Nv3775xrJgwQLMmTOneE6MEEIqiAcPgGPHAHt7oEsXQENld4yKgTFAIODP37zh1zYpSf7x+jUwfjx/AMCTJ0DXrvnv19RUmgAJhcClS/mXzZ186OoCVarwf3V0+EP8XFcXqFVLWlZfH/j+e/ky4s/Vri0tq6UFXLwoX078r7a2Mlet5JSrr/OyZcswdOhQODs7QyAQwNHREQMHDsSmTZskZXbv3o1t27Zh+/btqFevHiIiIjB+/HhYW1sjKChI4X6nTZuGiRMnSl6npaXB1ta2xM+HEELKEsaA7Gx+AwN4U8bEifzGlZ4uLbdhAxAVBfTsCTRrpppYywKRiDcjia/Xq1fAnj35JzXTpgE//sjLvngB9O+f/75fvJA+t7ICrK15ctOiBX+tq8v71Dx/Dlhaypbdvz//5MPISFq2WjXZn2tBtLWBNWuKVlYg4HGWdSpLgMzMzKCuro5Xr17JbH/16hUsc/80czE3N8fBgweRkZGB5ORkWFtbY+rUqahZs6akzOTJkzF16lT06tULAODq6opnz55hwYIF+SZA2tra0C4rKSkhhKjA7t3ArFnAt98Cc+fybY0bAwMG8ButWq4hM/v3A8ePA3XrShOgp095LYSbGxAcLC0rFALq6qV1Fl+GMeD9e34DFzfpxMcDmzcrTmrevAHmzwfEXVYTEngTU35yj++xtgZat+ZJib09YGvLE5IXL4D164HISGlZKyv+XkQEMHAg0Lkz337wIBAQALx8Ka0t0tMDDh3iydjcuYCnJ9+elsbLmZp+8WWqMFSWAGlpaaFx48YICwuDv78/AN4JOiwsDKNHjy7wszo6OrCxsUF2djb27duHnj17St77+PEj1NRkB7epq6tDJO59RQghBK9f8+YPPT3+WijkTV779/Mb54cP/IYuvrHmviG3asX7jjRtyl+npABHjwK3b/Mbbe6yY8cCjx7x2gN/f/5+VBRPHqyt5ZMjGxvAzIw/T08HYmPzPwdxYgDwfjExMfmXtbYGzM3582fPgFWrFCc1mZnAr7/y2hqAxzljRv77zZ3U2NgAPXrwmHI/NDV5bMbG0rIWFkByMr9WR48CHTvy7efP8742ebq7YsIE/jNxdZVuMzTk/X9cXGTLnjvHE9LccYeF8QTVywu4elW6feFCfs59+/JEDJBtqqvQmArt3LmTaWtrs82bN7OoqCg2bNgwZmxszBITExljjPXr149NnTpVUv7KlSts3759LDY2lp0/f561bt2aOTg4sHfv3knKBAUFMRsbG3bkyBEWFxfH9u/fz8zMzNiPP/5Y5LhSU1MZAJaamlps50oIIaUhK4ux168Ze/yYsZs3Gcv9a+zmTcbmzmXMzY0xgYAxT0/G2rVjzMuLsVq1GJs2TVp+/XrG+K1Q8WPnTul+9+wpuCzA2KlTvOy//xZcbsIExt6/52XPnSu47G+/SWO4fr3gsvPnS8tGRBRcdtIkadk3bxgbPJhfmyVLGNu2jbHQUMYiIxl7+ZJf79xWrmSsVy/GLl2Sbjtzhu+3Vi3Zst26MWZiwtiOHdJtaWmMnT/P9/25Tp1i7K+/GHv7Vrrt778ZMzBgrHt32bL29jy28HDptv37GTM1ZaxfP9myx44xdvq07HeqrFHm/q3SPkCBgYF4/fo1fv75ZyQmJsLd3R3Hjx+XdIx+/vy5TG1ORkYGZsyYgSdPnkBfXx8dO3ZESEgIjHOl1StWrMDMmTMxcuRIJCUlwdraGsOHD8fPP/9c2qdHCCGf5e1b3rcjNZU/UlKkz1NTgVGjpH+tb93Km2HE7338KLuvY8d45+U2bYAbN4DcvwqvX5ct26gRr1UAeG2Fvj5/KKoNyF1DoaPDa2MUEYl4vF5e/LW2Nt93Sgp/XrWqtOzr18CSJYCfH39oafHaoMxMXouSt6eCvr70uaZm/jEAvBlPzNaW923KW1NTrRqvJRLXigG8yWjtWv7zMDKSNiHdvQu0a8ePm2sqOoSG8iaor76SNg/Wrg14e8vX1GzfLn9OBga87Jdo00Z+W79+/JH7OgDA4MG8k3Xu0WAvXvDaqbyjtb7/nteehYcDzZvzbWfOAEuXAi1bAj/8IC0rbm4ry71LVDoPUFlF8wARohqbNwP37vH+JgIB/1f8EAiAn36S3niPHuXNB7nfz/18yBDpDfLiReD+/fzLdu0qvfHfvQs8fJh/DF99JS0bF8cfecuoqfHOsQ0bSmM4dw7491/ZRCZ3YvPPP7w8APzxBzBlSv7X6dQp6U1u/Xo+XDkvPT0eJ2O8P8jVq7yZa/Nm3uykpcVH+RgZSR/165deHxGhkDdxiTvlMgZ88w1w5w6/VuIEb906fuPt2JH/zMVGjuTnMHEiUKNG8cSUng6cPs1v/gMHSrf7+/OkZvVqYMQIvu3ZMx6jpiZPFMRNefv28eandu2ABg2KJ67Slp7Ov9caGtKkjTH+M3jyhDenVa/Oty9fDowbx5v+9uyR7sPenl+jq1elyW9EBHDyJP+et2tXMrErdf8u8fqocoiawAgpeUIhb0bI7ZtvCm6aSEuTlh00qOCyCQnSsqNHF1z28WNp2SlTCi6bO+a5cwsue+OGtOzvvxetiYgxxjZtYszSkrE6dXjzVLt2jPXowdiQIYz98ANjUVHSsv/9x5tYbt1i7ORJ3gSTu1mmTx/GzM0Z27v3s35MKnfyJGN9+zK2dKl0m1DImI4Ov24xMdLt27Yx1rAhY7/+KruPjx/l97t7N/8OHT4s3fbkCd+ntjY/htiECfx4uZvccnIYO36cH18k+rJzLM/u32dszRrGjh6VbhOJGDM05Nfy+XPp9qVL+bYePUounnLTBEYIqVw+feJ/YR8+DBw5wqvJY2MB8UDOYcN4rYSaGm86EYl4eiB+Lh5uDPCaGED2/dzPczfRNGjAR8uI38tbNneTh709r85XVC5vWTMzoF49+bKMSc9BzMuLd2TNXeNibCx9nnselYEDZWsgCmJjwx9Pn/ImLHV1PpJL3Dl42TJ+nPI6h0+7dvK1BTk5vObh0SNpTRHAawRv35Ydgs0Yvxa6ukBionQ0W3g4sGkTr/Hq0oVvs7XlPyd7e97hWDwSbP583lk49/gadXXeTFfZ1a3LH7kJBLx2MzlZtomzdm3gu+/KztQJ1ASmADWBEVJ83rzhzQf//MP7R+Tuo6KvD+zYIR3WS4omLY03tWRlyTZ/NW/ObziLF8smVJXFixc8AapenSeDAB/JZm3Nn//3H08WAd6MeOkSX8KhZUuVhEtKgDL3b0qAFKAEiJDPJ170UNz58cAB6Sy1AP8r+5tv+F/dvr5lu5NkWXXkCL9+lpb8pi+u3cnJKb81PSUpJYXX/jg68j47pOIqF2uBEUKKh1DIO/hmZfE5QlSRUGRl8flLDh/mj6AgQLy6TLt2vGaifXt+03ZzqyRzjBSTmzd5R+AWLfh1BXjTy1df8WualSVNeij5UczYWHYOHkIAqgFSiGqASFn2/j1vAhFX5aek8GYPxngfCPHIk4gI4NYt3qehfv3ij+PtWz6q6Z9/+KzAudcQbtIEuHKl+I9ZWeSeiG7xYj68uEULPpqNEJI/Ze7fagW+SwhRudx/oqxbx/+SzT3fhrEx71RoaSnb7+PgQT7Hx9KlsvubPp2v5VTUNYAUEYkAZ2c+r8ju3Tz5sbDgxzt4kA+TJcpbvZrX4uW+fr16AYMGSZenIIQUD6owJaSMGj2aNyft3i1dcsDJiScfuRdKBHjNQN5mJXt7oG1b6WcBPirj11/58969pdt37AAuX+bz4eSeRC0nh3cU/ecf6Rwe4nlu2rXjc+aI+/N4esqOkiGFe/9eOtII4HPg3LvHfx5t2/Jt1tbAxo2qiY+QiowSIEJULCGB1+y8fcuH9oq9eMEf4eHSJKZFC75NPAmZmKI+NQMG8EduQiGfOC4pSXYW3aNHgW3b+Ey6np7AiRN8lNG+fTwJErt3T7oWUXCw7LB0UnQ5OXwl9aNH+VBuOzu+/fvv+fXP3WmcEFIyqA+QAtQHiJSU+HjeWdjJCfDw4NuePgUcHPjolJQU6Twzly4BGRm8D0/uZKUkHDnC4xIKgRUrgOxs6XsCAV8osUsXPhPszz/zJRR+/FE6f4r4twh1blZMKOSJTu6lEFq35ssIrF2reCZnQojyaBQYIWVATg5v0mjYUJoY/Pknn5hu1ChpAmRnx6fXd3WVnThPvNZOSRCJeBLzzz98mv/Onfnj5EnpHDJt2/L+RdWrS6f/B3hT2ZUrfJVvsTt3AB8fXkOVe7mC1FS+HENlTozi4vjP8tMnPhRbPEHj77/zZLdePdXGR0hlRQkQIcVEKJSuByQU8k7Jycn8L38nJ77dx4c3aYlnPgZ4crB6dcnH9/Ejn/xNPAvzq1d8e0aGNBnz9QUePADq1Ml/Pxs28L4/4pmYAT4MX7ymVW7ffMNHou3cCXTqxLe9f88nR7SzK3t9htLS+AzAVatKpxNISuLna2AgXdMI4E2GL1/ypixxE1ZEBE9wa9SQTgNgZ8eHp2to8Gvr7s63e3qW1lkRQhQquRU5yi9aC4wo48wZxtzcGOvcWXZ7ixaMGRkxduKECoLK5e1bHpt47STxw8CAsW+/lV0L6XNlZDB29y5j16/Lbrey4se6eVO6be9evu2rr2TLhocz9uiR7BpMycmM3bnD12jKbd8+xjZvZiwlRbrtyhXGJk9mLDhYtmy/foy1aSO7ZtTu3XytrcBA2bKOjjy2ixel23bt4ttatpQt6+7Otx8/Lt327798W6NGsmWjoxnLzGSEkBJGa4ERUkK2buU1KGPGAN7efJuBAZ9/59kz3rQkrtXYv5+vFVWatRyM8eaoZ8947QvA15m6eZPX9NjZSUdt+fgUXydmbW3Fcw09ewY8fsxn4BV7/ZofN/c2gNekxMfzz4hX996yhXfa7tuXX3uxYcN47dq9e9LVxO/c4U2MXbvKdv4+f57vMzmZrzMG8ObJxEReE5Wbri6vkcvKkm4zM+Pn5uAgW7ZTJz7nkqWldJuzM2/aEi+9kHs7IaRsoQSIVFopKbKzw/74I2+uUlPjiURyMm8ScXHhSc769Xx01J49fMRWjRrS5KZtWz6CaswYfnNfskS6GOWmTUB0NC+rpiYdRi5+PmuWdD8HDxZcdsQIadJy7hzw8KH0vdu3eXL24gVgbs5Hl6mr8/f/+osvQVG/fun2x9HUlO34C/CRTkOGyM5DlJnJr1dyMu8rI2ZoyLfn7QTeujVvSsu94KmbG58fSTxKTWzhQp7Q5G529PPj18vERLbsrVu8qSr3NWrdmjeB5fXLL/LbHBz494gQUvbRKDAFaBRYxSUU8o6/S5bw/hsPH/Kb3dOnvDPwnTuKP2dqymsLjh3jN879+3mtiiI6OrI38c6dZTsGK4pJnAAFBvJ5f/Lz/r00GRg0iA9Fz0tXl8/Rs3Ejr70oT4RC/q+4LxUhhCiDRoERksf79zxZWL4ciI3l2zQ0eDJz8iTw00+8A/C6dbwZizE+HFxfnw/9trLin+nQgT/s7XmzCmPS8uJ/8968u3XjtSDiMrnLi0SytQ2tWgFVquRfNvdaTw0b8pooxnjiUL06T7batOFJUHlEiQ8hpLRQDZACVANUccTH86UgNmyQjlDS1QUGDgSmTeNJw4kTvE+Mv3/BtS+EEELKNqoBIuT/YmN5HxCAz22jpsaHIlerJp1NuVUrnhyV11oTQgghyitjs3AQ8vlycoBdu4BVq/jrDx/4CKRhw/i8N9HRfBbj+vWlo4EA3qmYkh9CCKlcqAlMAWoCK1/eveNNXCtX8hFQBgb83xYt+AR927YBffrwsuI+N5V5ZmJCCKmoqAmMVAoxMXzW3c2beW0PwId/jxzJn3/7LZ/7JnfH4bI28zAhhBDVoBogBagGqOzbtInPJSP+9mpq8gU8L17kNT8AT360tCjpIYSQykKZ+zfdGki5kJnJJ/YDeF8fc3M+ZLpTJ76+Vc+egI0NHxYupqNDyQ8hhBDFqAmMlGlJScDatXyxUC8vPrlfo0Z8Ic/bt6XLL7i78yURNOgbTQghpAjodkHKpHv3+Pw9ISHSdZkiI/loLSsr3ryVex0nU1OVhEkIIaScogSIlCnnzwPz5vFmLTFNTV7z06sXf75zJ2/u0tZWXZyEEELKN+ohQcoMxoDr13nyo6bGl5AwNua1Pg0a8OQH4ItaUvJDCCHkS1ANEFGZ+Hjet6dePT5L86BBfA6fqVOB4cP5elvXr/N1tPKuBk4IIYR8CUqASKm7eZOvxr5zJ1/Es25dviBpVBQftr53r3TxUU9P1cZKCCGkYqIEiJQKoRA4fJgnPhcuSLd7ewMTJvCE5+BBPoePiYnKwiSEEFJJUAJESsWQIXzGZoDP3yMU8matEyek63B17qyy8AghhFQy1AmalIinT/kcPqdPA23a8E7NVasCP/3E31uzhi9USouQEkIIUQWqASLFhjEgPJzP33PgAPDjj0CdOjwJcnQEnj8HqlThZb//XqWhEkIIqeQoASJfLDsb2LMHmDMHePRIuv3pU2DmTP7voEHS5IcQQghRNUqAyBdZsQL4/Xc+pF3M2pqP6qpXj7+ePVsloRFCCCmjcnKAT5/41CeqQgkQ+SwXL/LkR02NJz8WFkCfPsDHj3xUV506qo6QEEJIWcEYEBMDhIbyx5kzwKhRwK+/qi4mSoCIUkQinrWPHs3X5vruOz66q1cvmp2ZEEKI1Js3QFiYNOl5/lz2/atXVROXGCVARCmTJgEREcCOHXwW5yFDADc3VUdFCCFE1TIz+UCY0FDeDeL2bV7zI6alxed6a9eOPxo2VF2sACVApIjS0oBp03jSA/DOzitWqDYmQkjxEgqBlBTg3Tvpw8AAaNoUEAhUHR0paxgD7t6V1vCcP89bCHKrX1+a8LRsWbYGw1ACRAr15g3QuLG0+vKPP4CuXVUbEyFEsaws+SRG/Chse1qa4n327g389Regp1eKJ0LKpJcvpQnPqVPAq1ey71taShOetm2lyxqVRZQAkUJFRgIvXvDngYG8GYwQUnI+fSo4YSnovY8fv/z4+vqAsTFfliY6mjd5P3zIl6uxtf3y/ZPyIz0dOHdOmvRERcm+r6cH+PhIk5569cpPbSElQKRA9+4B3bvzqs4OHYBt28rPl5sQVWEM+PChaAmLovcyM788BiMjnsCYmEiTmbwPRduNjQFNTel+zp3jvwNu3eKLE+/fDzRv/uXxkbJJKOQLVosTnkuX+FxvYgIBbxFo1w74+mugWbPyOwCGEiCi0JYtQEICT3hSU4GvvuK/+NTVVR0ZIWVPTg6vKQ0Plz5yz431OdTU5BOU/BKZvO8ZGRXf/1UfH+DGDd7sfecO4OvLl7IZPLh49k9U78kTacJz+jRPwnOzt5fW8LRuDZiaqiTMYkcJEJFz4wYwYAB/HhzMq7937AB0dFQaFiFlRmoqcPky/+s4PJwP5/3wQb6chkbREhZF2w0MeBJUFtjb8/McMADYt4+P/oyMBBYtkq0tIuXDu3c80REnPU+eyL5vZMQTHXHS4+hYMWv+BYzlHqRGACAtLQ1GRkZITU2FoaGhqsNRicmT+S/v+fPLzi9hQlSBMb6cS+7anXv3ZIf3Avym0bw5H+bbogUf4mtoWLFuHCIR8MsvwKxZ/HXr1sDu3RWnRqCiysoCrlzhQ9NDQ/kfuSKR9H0NDd6UJU54PDz4tvJImft3OT1FUtzev+dzNGhqAiNG8FEfvr6qjoqQ0pedzee6yp3wJCTIl6tZU5rstGgB1K1b8f9YUFMDfv4ZcHUF+vXjtQheXsChQ3y4MykbGOOd18U1PGfPytdQOjtLEx5fX9UuSaEqVAOkQGWrAYqPBzp25L/ULCyAxYv5KJCnT+kvO1LxpaTw5ixxsnPtmvxIKg0NoFEjabLTvHnZHt5bGu7e5f2C4uL474uQEMDfX9VRVV6vXvFh6eKk5+VL2ffNzfmwdPHw9Io6mo9qgIhSoqOB+/f5L7L37/m2NWso+SEVD2P8e567duf+ffnmLGNj2eYsT0+aAycvV1fg+nXg22/5uk4BAcCcOcCMGRW/Jqws+PgRuHBBmvDcuSP7vo4O4O0treVp0IB+LnlRDZACla0GCAAmTgSWLOHPf/sNmDJFtfEQUhyys/l0/LkTnsRE+XKOjrLNWS4udLMoquxs/vtj5Ur+unt3Poq0LM34W1EkJwMbN/K+PBcvyk+X4O4uHZ7eogWgq6uSMFWKaoBIofbsAVq1AszM+Dwfq1bx7aNHAz/+qNrYCPlc797JN2flnZpfU1O2OatFC970Sz6PpiZfFsfNDRg5ko8Si4nh/YLs7VUdXcXAGPD333wS2jdvpNurV5fW8LRpA1SrproYyyNKgCqhlSuBMWN4r/9163g7flYW0K0bsHRpxRq1Qiouxvjw3bzNWXlVrSrbnOXhUTn/Mi5pQ4bwmrNu3XhzjKcnsHcvn0eIfL7oaD4w5dw5/rpePWD4cJ701KlDv6+/BCVAlVCbNnyekZYtgVq1+AzPL14AW7fSRIek7MrKkm/OyrsOEQA4OcnW7tSpQ81ZpaVFC94vKCCAzxzdti2wfDm/gRPlfPrEpyH54w/ezKirC8yeDUyYQHMvFRfqA6RAZegDlJAgHcUiEvEhkpVxGCQpu96+lW/OysiQLaOpyWt0co/OomYA1fv4kc8UvXMnfz18OE+EtLRUG1d5cfw4MGqUdILCTp14zT01KRaO+gARGfHxwHff8f9A9erxjnMHD/JfSmpq/EHJDykLLl0CNm/mCU/eRRcBPjIxb3MWzVBe9ujpAdu3835BP/3Em9qjoniTGCWo+Xv5ktfw7N7NX9vY8P5V/v7U1FUSKAGqBCZO5BNhDRrEbzBBQcCuXbyKesMGVUdHCHf5Mu+Yn5Ul3Va7tnxzFt0IygeBAJg6lU+Q2KcPH7Lt6ck7R7u7qzq6skUo5FOPTJ8OpKXxP0rHjePTCtAfpyWHEqBKYNUq3p68bBkf3r5rF5/YLTBQ1ZERwsXH886zWVm8c+eoUbymx9xc1ZGRL9W5M18r7ZtvgMePeSK7eTOfP4jwldeHD+f/Anxm7bVr+VIqpGRR18AKKjZW+tzMDDh8mP/ltWgR3xYczDsoEqJqGRm802xiIq8t2L+fj0yk5KficHHhfbi+/pr3D+rZE5g5U3Y9qsomLY3X8nh58eTHyAhYvZrX0lPyUzooAapgGAPmzePrvBw/Lt2+Zw9vCgP4RIfffaea+AjJjTFg2DA+cqhqVZ6k6+urOipSEkxMgKNHpb+HfvmF1/qJZ5+vLBjjv4+dnXnHcJGIr7344AEfLUcjcUsPJUAVUEwMkJPDV/8FgPPnecLDGE10SMqWpUv5GlLq6rzjZ82aqo6IlCQNDV4LvWULoK3NE95mzWRrrCuyJ0/4iK6ePflI3Fq1+KzO27cDlpaqjq7yoQSoghEIgL/+4s0Is2fzbcnJfHtAAE10SMqO0FA+sy3AF+Bt00a18ZDS078/n9jPyopPXunpyRfyrKiysoAFC/go3GPH+HQAP//MF5Rt107V0VVeNA+QAuVtHqD4eD68dNy4/MvcuMH/89EMuKQsePyY93149w4YOJCvb0SJeeXz8iX/w+zaNV4LuHgxn6W+In0Xzp8Hvv+ez+gMAK1b874+deqoNq6KSpn7N9UAlXPv3/Mq5PHj+VwbYqmpwH//SV/T9P+krHj/nndyfvcOaNqUD/+tSDc8UnTW1rwmqF8/PhR83Di+pEbeRT7Lozdv+NQjPj48+TE35829p05R8lNWUAJUzhkY8I5zzs58hAXAf3kEBPCby927qo2PkNxEIn6zi4riN7/9+3lfEFJ56ejwPkGLFvH5bzZt4vNBJSaqOrLPIxLxc6hTh4+2BXhH/4cPeV9MSvbLDkqAyimhUPp86lQ+isbBgf/nGzgQOHOGD7PMXY4QVZszh3d81dYGDhyQLsdCKjeBgI8O+/dfPhz88mVea33jhqojU879+4CvL18G5O1bwNWVz2q+bh0fBUfKFkqAyhnG+PBRf38+0gvgvzzEQ4enTAF27OCjLfbvpxlXSdmxbx8wdy5/vn497wNESG5+frw/kLMz79vo7c1HSJV1Hz8C06bx37cXLvClQP78k8/v07y5qqMj+aEEqJyJjeUrBB85Avzzj+x7y5YBCxfy5zTRISlL7tzhS7AAfK2j/v1VGw8pu2rX5lN4dOrEJ8ns25f/YVdWa7P//ZcPMPntN/5H6Tff8D4/kybRqu1lHSVA5UytWvwvolWreD8fsT17+I0F4MMtaaJDUla8ecM7PX/4wJPyP/5QdUSkrDMy4k2lU6fy13/8wROL1FTVxpVbfDxfzqNTJ+DpU8DWli8yfegQUKOGqqMjRaHyBGjVqlWwt7eHjo4OmjRpgmvXruVbNjs7G3PnzoWjoyN0dHTg5uaG47mnO/6/+Ph4fPfddzA1NYWuri5cXV1xo7w1Jufy8qXsiK6AAGDkSOlrkYjX/DDG11CaMqX0YyREkexsPunb06d8kkPxOnSEFEZdnf8xt3077yj9779AkybAo0eqjSsnh9e2Ozvz6UfU1YEffuAd+7t2VW1sRElMhXbu3Mm0tLTYpk2b2P3799nQoUOZsbExe/XqlcLyP/74I7O2tmZHjx5lsbGxbPXq1UxHR4fdunVLUubt27fMzs6ODRgwgF29epU9efKEnThxgj1+/LjIcaWmpjIALDU19YvP8Uvdv8+YrS1jrq6MpaTkXy4lhbF58xjLySm92AgpzNixjAGM6eszdu+eqqMh5dWNG4xVr86/S0ZGjB07ppo4rl1jrFEjHgfAWNOmjEVEqCYWopgy92+VJkBeXl5s1KhRktdCoZBZW1uzBQsWKCxvZWXFVq5cKbOtW7durG/fvpLXU6ZMYV999dUXxVWWEqBnzxizsmKsTh3+PLfMTNXEREhRbNwovVEcOKDqaEh5l5DAWPPm/PukpsbYn38yJhKVzrFTUhgbNYoxgYAf39iYsbVrGRMKS+f4pOiUuX+rrAksKysLN2/eRNtcPXXV1NTQtm1bXL58WeFnMjMzoaOjI7NNV1cXFy9elLw+fPgwPDw88O2336JatWpo2LAhNmzYUGAsmZmZSEtLk3mUFTVq8CUDLl2SbVdOTeWjaBYt4rcYQsqSy5f5/FQAH/ru76/ScEgFYGkJnD7Nh5iLRMDkybwz/adPJXdMxnizrbMz73fJGO+U/eABMHw4n7eIlF8q+/G9efMGQqEQFhYWMtstLCyQmM8MWH5+fli8eDFiYmIgEokQGhqK/fv3IyEhQVLmyZMnWLNmDZycnHDixAmMGDECY8eOxZYtW/KNZcGCBTAyMpI8bG1ti+ckPwNjfDTB9evSbfXq8ZWyxcQTHUZG8r4/b9+WfpyE5Cc+nq/ynZXF/50xQ9URkYpCWxvYsIGvoq6uDmzdymdajo8v/mPFxgIdOgC9evFJGZ2c+CzOW7cCeW5bpLwqhRopheLj4xkAdunSJZntkydPZl5eXgo/k5SUxLp27crU1NSYuro6q127Nhs5ciTT0dGRlNHU1GTNmjWT+dyYMWNY06ZN840lIyODpaamSh4vXrxQWRPYihW8irVaNcaSk+XfFwoZ691b2q8iV/cnQlTu0yfGPD3597N+fcbev1d1RKSiOnWKsapV+XfN0pKxy5eLZ78ZGbw/pY4O37e2NmNz5vDvNin7ykUTmJmZGdTV1fHq1SuZ7a9evYKlpaXCz5ibm+PgwYP48OEDnj17hgcPHkBfXx81a9aUlLGyskLdunVlPufi4oLnz5/nG4u2tjYMDQ1lHqoSFMRnQJ01S7bWRyzvRIcNG5Z+jIQowhif8v/6df7dPXRIOkEnIcWtTRv+XatXj9fQ+PjwJTW+xNmzfDLDmTP5HERt2vDlhH7+mY9EIxWLyhIgLS0tNG7cGGFhYZJtIpEIYWFhaNasWYGf1dHRgY2NDXJycrBv3z50zTX2sEWLFnj48KFM+UePHsHOzq54T6AY5W7DNjDg/SdyD3MXyz3R4aZNQLt2pRMfIUWxdClf7FFdHdi9mw97J6Qk1azJf1/6+/Mm1wED+Hxo4lnyi+r1a/7HZ6tWvH9PtWrAtm28/6WTU0lETsqEUqiRytfOnTuZtrY227x5M4uKimLDhg1jxsbGLDExkTHGWL9+/djUqVMl5a9cucL27dvHYmNj2fnz51nr1q2Zg4MDe/funaTMtWvXmIaGBps/fz6LiYlh27ZtY3p6emzr1q1Fjqs0R4Hdu8eYnR1jO3cWXO7OHekIhF9/LfGwCFHKyZN8ZA7A2LJlqo6GVDZCIWM//ywdddiuneIuBIo+t2EDYyYm/HMCAWPff8/Y27clHzMpGeVmGDxjjK1YsYLVqFGDaWlpMS8vL3blyhXJez4+PiwoKEjy+uzZs8zFxYVpa2szU1NT1q9fPxYfHy+3z3/++YfVr1+faWtrM2dnZ7Z+/XqlYirNBGjqVP4fz82t8Dl8li9nbPTo0hv6SUhRxMRIbyADB9L3k6jOnj2M6enx72KtWnwetfzcvctYixbSpMnNrfj6ERHVUeb+LWCMBlHnlZaWBiMjI6SmppZ4fyChEJg3DxgzBjA1LdFDEVLs3r8Hmjbls+A2bcr7UGhrqzoqUplFRvIZmZ89410Ktm0DunSRvv/hA1+Ud/Fi3lRWpYr0dzDNUl7+KXP/plkMShljfBFTcdqprg7Mnq04+YmPB/r0Ad69K9UQCSkSkQjo148nP9bWvFM+JT9E1dzceOdoHx+eoHftCvz6K/+de+QI7zT9xx88+QkI4AuXTphAyU9lRAlQKRs1ii/qN29eweVSU4GOHfmIr8GDSyc2QpQxZw4f6aWtDRw4AFhZqToiQjhzc96BecQInvhMn84nM+zShdcM1agBHD7Mk3YVTvtGVIwSoFLm5sZnDy2ouUs8gdydO3z208WLSy8+Qopi3z7ejAAA69fzWckJKUs0NYHVq4G1a3ntzqNHvMZ98mRea5m7WYxUTtQHSIGS7gMUHQ24uCh+T9yssH07n0Pl/Hma64eULXfuAM2b874UEyZQgk7KvkuXeG36sGGAq6uqoyElSZn7N7V6qkB+yQ8ATJ3Kkx8NDf5XNiU/pCx584b3qfjwAWjblvelIKSsa96cPwjJjZrAypA1a4A//+TP//oL+Ppr1cZDSG7Z2UDPnsDTp3wCul27qOMoIaT8ol9fZYivL2Bnx6tpg4JUHQ0hsiZNAs6c4U2zhw8rXqqFEELKC0qAyhAXFyAiAjAyUnUkhMjatImvwA3w5S7q1VNtPIQQ8qWoCUzFoqKA06elr42NAYFAZeEQIufyZT6cGOBD3/39VRoOIYQUC0qAVCg+Hmjfnj+OHVN1NITIi4/nUzJkZQHduwMzZqg6IkIIKR6UAKmIeKLDFy8ABweaR4WUPRkZfKbcxEQ+dHjzZj6HFSGEVAT060wFck90aGEBHD9O64CRsoUx3hn/+nXe2fnQId75mRBCKgpKgEqZSAQMHMj7/ejrA//+y2uACClLli7lnZ3V1YE9e+g7SgipeCgBKmXTpslOdNiokaojIkRWaCgf8g4AS5YArVurNh5CCCkJlACVIpEISEjgz2miQ1IWPX4MBAby7+qgQcDo0aqOiBBCSgYlQKVITQ3YsoU3f9FEh6Ssef+eL3Px7h3QtClfSJKmZCCEVFSUAJUygQBo1UrVURAiS7wIb1QUYG0N7N8PaGurOipCCCk5lAARQjBnDh/ppa0NHDgAWFmpOiJCCClZlAARUsnt2wfMncufr19Pc1IRQioHSoAIqcTu3JH2R5s4EejfX7XxEEJIaaEEiJBK6s0b3un5wwegXTvg999VHREhhJQeSoAIqYSys4GePYGnTwFHR2DnTj43FSGEVBaUABFSCU2aBJw5w2cjP3SIL3dBCCGVCSVAhFQymzYBy5fz51u3AvXqqTYeQghRBUqACKlELl8GRozgz+fO5X2ACCGkMqIEiJBK4r//gIAAICsL6N4dmD5d1RERQojqUAJESCXw6RNPfl69Alxdgc2b+dIshBBSWdGvQFKp5eTwZSAqMsaAYcOAGzd4Z+dDh3jnZ0IIqcxo4CupFDIygEeP+FpX9+/zf6OigJgYQEeH14q4uUkfDRpUnCRhyRLe2VldHdizB3BwUHVEhBCiepQAkQrl40fgwQNpgiN+xMbmX9Pz4QNw5Qp/5FarlmxS5OYG1KhRvlZIP3kSmDyZP1+yBGjdWrXxEEJIWSFgjDFVB1HWpKWlwcjICKmpqTA0NFR1OESB9HQgOlo2ybl/n0/sl9832tiYD/muW1f6cHHh+4qMlH28fJn/PvImRfXq8VqksubxY8DTE0hJAQYNAv76q3wlb4QQoixl7t+UAClACVDZkZoqTXRyN109f57/Z0xN5ROdunUBS8uiJwCvX8snRVFRvM9QXurqgLOzfGJkafl551wc0tKApk35tWvaFDh7lq/0TgghFRklQF+IEqDS9/atfLNVVBQQH5//Zyws5JOcevUAc/OSiTEzkycUeROj5GTF5atVA9zdZZOiOnUATc2SiU9MJOIjvg4fBqyteednK6uSPSYhhJQFlAB9IUqASs7r14oTncTE/D9jbS2b4IibrkxNSy/u/DDGk7S8SdGjR4qb4rS0gPr15WuLTEyKL6affwbmzeM1PufPA15exbdvQggpyygB+kKUAH0Zxvh8M3n750RF8RXI81OjhnyNjosL73dT3nz4ANy7J5sU3bkDvH+vuLytrXxtkaOj8nP17N0LfPstf75lC9C//xedBiGElCuUAH0hSoCKhjHeWThv/5yoKODdu/w/5+CgONExMCi92FVBJOKdtCMiZBOjp08Vl69ShQ/Hz50UubrmPzz/zh2gWTM+Em7iRGDRohI6EUIIKaMoAfpClAAV7MMHoGdP4OJF3tlWEYGA12Dk7Z9Tpw6/sROp1FSevOROjO7d43MX5SW+rnlri/T0+Iivp0+Bdu2Af/8FNGiSC0JIJUMJ0BeiBKhgu3YBvXrx5+rqfL6cvKOuatcGdHVVG2d5lpPDJ2nMW1uUkKC4vIYG/4yjI3DtGp/xmRBCKhtl7t/0NyJR2sWL/N/Bg4FVq2h4dUnQ0ODNgi4uQO/e0u3i4fm5E6PoaJ78GBjwZS4o+SGEkMJRAkSUduEC/9fPj5Kf0mZuDrRtyx9imZl89mszM8DGRnWxEUJIeUIJEFGKuL8KAHz1lWpjIZy2Nu8HRAghpOhoNXiilEuX+OgvR0eaXI8QQkj5RQkQUYq4/4+3t2rjIIQQQr4EJUBEKeL+P9T8RQghpDyjBIgUWWYmH2INUA0QIYSQ8o0SIFJkN27wJKhaNcDJSdXREEIIIZ+PEiBSZLmbvwQC1cZCCCGEfAlKgEiRUQdoQgghFQUlQKRIRCIgPJw/pw7QhBBCyjtKgEiR3LsHpKTwlcjd3VUdDSGEEPJlKAEiRSJu/mrWjFYZJ4QQUv5RAkSKhOb/IYQQUpEonQDZ29tj7ty5eP78eUnEQ8ogxqQJEHWAJoQQUhEonQCNHz8e+/fvR82aNdGuXTvs3LkTmZmZJREbKSOePQPi43nTV5Mmqo6GEEII+XKflQBFRETg2rVrcHFxwZgxY2BlZYXRo0fj1q1bJREjUTFx7U/jxoCenmpjIYQQQorDZ/cBatSoEZYvX46XL19i1qxZ+Ouvv+Dp6Ql3d3ds2rQJjLHijJOokLgDNPX/IYQQUlF89nie7OxsHDhwAMHBwQgNDUXTpk0xePBg/Pfff/jpp59w6tQpbN++vThjJSpC/X8IIYRUNEonQLdu3UJwcDB27NgBNTU19O/fH0uWLIGzs7OkTEBAADw9PYs1UKIab94A0dH8eYsWqo2FEEIIKS5KJ0Cenp5o164d1qxZA39/f2hqasqVcXBwQK9evYolQKJa4tmfXVwAMzPVxkIIIYQUF6UToCdPnsDOzq7AMlWqVEFwcPBnB0XKDlr/ixBCSEWkdCfopKQkXL16VW771atXcePGjWIJipQdNAEiIYSQikjpBGjUqFF48eKF3Pb4+HiMGjWqWIIiZcOHD8DNm/w51QARQgipSJROgKKiotCoUSO57Q0bNkRUVFSxBEXKhmvXgJwcoHp1oJBWT0IIIaRcUToB0tbWxqtXr+S2JyQkQINWyaxQcjd/CQSqjYUQQggpTkonQF9//TWmTZuG1NRUybaUlBT89NNPaNeuXbEGR1SLOkATQgipqJSuslm4cCFatmwJOzs7NGzYEAAQEREBCwsLhISEFHuARDVycoDLl/lz6gBNCCGkolE6AbKxscGdO3ewbds2REZGQldXFwMHDkTv3r0VzglEyqeICCA9HTA2BurXV3U0hBBCSPH6rE47VapUwbBhw4o7FlKGiJu/WrQA1D57xThCCCGkbPrsW1tUVBSOHz+Ow4cPyzw+x6pVq2Bvbw8dHR00adIE165dy7dsdnY25s6dC0dHR+jo6MDNzQ3Hjx/Pt/xvv/0GgUCA8ePHf1ZslRXN/0MIIaQi+6yZoAMCAnD37l0IBALJqu+C/w8TEgqFSu1v165dmDhxItauXYsmTZpg6dKl8PPzw8OHD1GtWjW58jNmzMDWrVuxYcMGODs748SJEwgICMClS5ckfZLErl+/jnXr1qFBgwbKnmalxhh1gCaEEFKxKV0DNG7cODg4OCApKQl6enq4f/8+zp8/Dw8PD5w9e1bpABYvXoyhQ4di4MCBqFu3LtauXQs9PT1s2rRJYfmQkBD89NNP6NixI2rWrIkRI0agY8eOWLRokUy59PR09O3bFxs2bICJiYnScVVmMTFAUhKgrQ14eKg6GkIIIaT4KZ0AXb58GXPnzoWZmRnU1NSgpqaGr776CgsWLMDYsWOV2ldWVhZu3ryJtm3bSgNSU0Pbtm1xWTwEKY/MzEzo6OjIbNPV1cVFcZXF/40aNQqdOnWS2Xd+MjMzkZaWJvOozMTNX15ePAkihBBCKhqlEyChUAgDAwMAgJmZGV6+fAkAsLOzw8OHD5Xa15s3byAUCmFhYSGz3cLCAomJiQo/4+fnh8WLFyMmJgYikQihoaHYv38/EhISJGV27tyJW7duYcGCBUWKY8GCBTAyMpI8bG1tlTqPikacS1L/H0IIIRWV0glQ/fr1ERkZCQBo0qQJ/vjjD4SHh2Pu3LmoWbNmsQeY17Jly+Dk5ARnZ2doaWlh9OjRGDhwINT+P1TpxYsXGDduHLZt2yZXU5Qf8cSO4oeitc4qE3ENEPX/IYQQUlEp3Ql6xowZ+PDhAwBg7ty56Ny5M7y9vWFqaopdu3YptS8zMzOoq6vLLa3x6tUrWFpaKvyMubk5Dh48iIyMDCQnJ8Pa2hpTp06VJF83b95EUlKSzHplQqEQ58+fx8qVK5GZmQl1dXWZfWpra0Ob2noAAAkJQGwsX/qieXNVR0MIIYSUDKUTID8/P8nzWrVq4cGDB3j79i1MTEwkI8GKSktLC40bN0ZYWBj8/f0BACKRCGFhYRg9enSBn9XR0YGNjQ2ys7Oxb98+9OzZEwDQpk0b3L17V6bswIED4ezsjClTpsglP0SWuPmrQQPAyEi1sRBCCCElRakEKDs7G7q6uoiIiED9XNMDV61a9bMDmDhxIoKCguDh4QEvLy8sXboUHz58wMCBAwEA/fv3h42NjaQ/z9WrVxEfHw93d3fEx8dj9uzZEIlE+PHHHwEABgYGMrEBfOJGU1NTue1EHg1/J4QQUhkolQBpamqiRo0aSs/1U5DAwEC8fv0aP//8MxITE+Hu7o7jx49LOkY/f/5c0r8HADIyMjBjxgw8efIE+vr66NixI0JCQmBsbFxsMVVmNAEiIYSQykDAxDMZFtHGjRuxf/9+hISEfFHNT1mWlpYGIyMjpKamwtDQUNXhlJq0NMDEBBCJgPh4wNpa1RERQgghRafM/VvpPkArV67E48ePYW1tDTs7O1SpUkXm/Vu3bim7S1JGXL7Mk5+aNSn5IYQQUrEpnQCJOyuTioeavwghhFQWSidAs2bNKok4SBlAHaAJIYRUFp+9GjypWDIzgatX+XOqASKEEFLRKV0DpKamVuB8P8U5QoyUnps3gYwMwMwMqFNH1dEQQgghJUvpBOjAgQMyr7Ozs3H79m1s2bIFc+bMKbbASOnKvf6XkvNZEkIIIeWO0glQ165d5bb16NED9erVw65duzB48OBiCYyULlr/ixBCSGVSbH2AmjZtirCwsOLaHSlFIhEQHs6fUwJECCGkMiiWBOjTp09Yvnw5bGxsimN3pJRFRQHv3gF6eoC7u6qjIYQQQkqe0k1geRc9ZYzh/fv30NPTw9atW4s1OFI6xP1/mjUDNDVVGwshhBBSGpROgJYsWSKTAKmpqcHc3BxNmjSBiYlJsQZHSgdNgEgIIaSyUToBGjBgQAmEQVSJOkATQgipbJTuAxQcHIw9e/bIbd+zZw+2bNlSLEGR0vP8OfDiBaCuDjRtqupoCCGEkNKhdAK0YMECmJmZyW2vVq0afv3112IJipQece1Po0ZAnnVtCSGEkApL6QTo+fPncHBwkNtuZ2eH58+fF0tQpPTQ+l+EEEIqI6UToGrVquHOnTty2yMjI2FqalosQZHSQx2gCSGEVEZKJ0C9e/fG2LFjcebMGQiFQgiFQpw+fRrjxo1Dr169SiJGUkKSk4H79/lzSoAIIYRUJkqPAps3bx6ePn2KNm3aQEODf1wkEqF///7UB6icuXSJ/+vsDJibqzYWQgghpDQpnQBpaWlh165d+OWXXxAREQFdXV24urrCzs6uJOIjJYiavwghhFRWSidAYk5OTnBycirOWEgpow7QhBBCKiul+wB1794dv//+u9z2P/74A99++22xBEVK3qdPwI0b/DnVABFCCKlslE6Azp8/j44dO8pt79ChA86fP18sQZGSd/UqkJ0NWFsDCmY1IIQQQio0pROg9PR0aGlpyW3X1NREWlpasQRFSp64+eurr4BcS7sRQgghlYLSCZCrqyt27dolt33nzp2oW7dusQRFSh6t/0UIIaQyU7oT9MyZM9GtWzfExsaidevWAICwsDBs374de/fuLfYASfHLyZEOgacEiBBCSGWkdALUpUsXHDx4EL/++iv27t0LXV1duLm54fTp06hatWpJxEiK2Z07QHo6YGgI1K+v6mgIIYSQ0vdZw+A7deqETp06AQDS0tKwY8cOTJo0CTdv3oRQKCzWAEnxE/f/adGCrwJPCCGEVDZK9wESO3/+PIKCgmBtbY1FixahdevWuHLlSnHGRkoITYBICCGkslOqBigxMRGbN2/Gxo0bkZaWhp49eyIzMxMHDx6kDtDlBGPUAZoQQggpcg1Qly5dUKdOHdy5cwdLly7Fy5cvsWLFipKMjZSA2Fjg1StASwvw9FR1NIQQQohqFLkG6NixYxg7dixGjBhBS2CUY+LaH09PQEdHtbEQQgghqlLkGqCLFy/i/fv3aNy4MZo0aYKVK1fizZs3JRkbKQG0/hchhBCiRALUtGlTbNiwAQkJCRg+fDh27twJa2triEQihIaG4v379yUZJykm1AGaEEIIAQSMMfa5H3748CE2btyIkJAQpKSkoF27djh8+HBxxqcSaWlpMDIyQmpqKgwNDVUdTrFJTASsrPjSF8nJgImJqiMihBBCio8y9+/PHgYPAHXq1MEff/yB//77Dzt27PiSXZFSEB7O/61fn5IfQgghldsXJUBi6urq8Pf3rxC1PxUZDX8nhBBCuGJJgEj5QB2gCSGEEI4SoEri/Xvg9m3+nDpAE0IIqewoAaokLl8GRCLA3h6oXl3V0RBCCCGqRQlQJSFu/qLaH0IIIYQSoEqDOkATQgghUpQAVQJZWcDVq/w5JUCEEEIIJUCVwq1bwKdPgKkp4Oys6mgIIYQQ1aMEqBLI3f9HIFBtLIQQQkhZQAlQJUDrfxFCCCGyKAGq4EQimgCREEIIyYsSoAruwQPg7VtAVxdo1EjV0RBCCCFlAyVAFZy4+atpU0BTU7WxEEIIIWUFJUAVHDV/EUIIIfIoAargqAM0IYQQIo8SoArsxQvg2TNAXZ03gRFCCCGEowSoAhM3f7m7AwYGKg2FEEIIKVMoAarAaP0vQgghRDFKgCow6gBNCCGEKEYJUAX17h1w7x5/3qKFamMhhBBCyhpKgCqoS5cAxoDatQELC1VHQwghhJQtlABVUDT8nRBCCMkfJUAVFHWAJoQQQvJHCVAF9OkTcP06f04JECGEECKPEqAK6Pp1IDsbsLQEatZUdTSEEEJI2UMJUAWUe/i7QKDaWAghhJCyiBKgCog6QBNCCCEFowSoghEK+RB4gPr/EEIIIfmhBKiCuXsXSEvja381aKDqaAghhJCyiRKgCkbc/NW8OV8FnhBCCCHyKAGqYGj9L0IIIaRwlABVIIxRB2hCCCGkKCgBqkCePAESEgBNTcDLS9XREEIIIWVXmUiAVq1aBXt7e+jo6KBJkya4du1avmWzs7Mxd+5cODo6QkdHB25ubjh+/LhMmQULFsDT0xMGBgaoVq0a/P398fDhw5I+DZUTN395eAC6uqqNhRBCCCnLVJ4A7dq1CxMnTsSsWbNw69YtuLm5wc/PD0lJSQrLz5gxA+vWrcOKFSsQFRWF77//HgEBAbh9+7akzLlz5zBq1ChcuXIFoaGhyM7Oxtdff40PHz6U1mmpBK3/RQghhBSNgDHGVBlAkyZN4OnpiZUrVwIARCIRbG1tMWbMGEydOlWuvLW1NaZPn45Ro0ZJtnXv3h26urrYunWrwmO8fv0a1apVw7lz59CyZctCY0pLS4ORkRFSU1NhaGj4mWdW+pydgYcPgX/+ATp3VnU0hBBCSOlS5v6t0hqgrKws3Lx5E23btpVsU1NTQ9u2bXH58mWFn8nMzISOjo7MNl1dXVwUt/8okJqaCgCoWrVqvvtMS0uTeZQ3SUk8+QH4EHhCCCGE5E+lCdCbN28gFAphYWEhs93CwgKJiYkKP+Pn54fFixcjJiYGIpEIoaGh2L9/PxISEhSWF4lEGD9+PFq0aIH69esrLLNgwQIYGRlJHra2tl92YioQHs7/rV8fyCfPI4QQQsj/qbwPkLKWLVsGJycnODs7Q0tLC6NHj8bAgQOhpqb4VEaNGoV79+5h586d+e5z2rRpSE1NlTxevHhRUuGXGBr+TgghhBSdShMgMzMzqKur49WrVzLbX716BUtLS4WfMTc3x8GDB/Hhwwc8e/YMDx48gL6+PmrWrClXdvTo0Thy5AjOnDmD6tWr5xuHtrY2DA0NZR7lDXWAJoQQQopOpQmQlpYWGjdujLCwMMk2kUiEsLAwNGvWrMDP6ujowMbGBjk5Odi3bx+6du0qeY8xhtGjR+PAgQM4ffo0HBwcSuwcyoL0dEA8CI4SIEIIIaRwGqoOYOLEiQgKCoKHhwe8vLywdOlSfPjwAQMHDgQA9O/fHzY2NliwYAEA4OrVq4iPj4e7uzvi4+Mxe/ZsiEQi/Pjjj5J9jho1Ctu3b8ehQ4dgYGAg6U9kZGQE3Qo4Qc6VK3wV+Bo1gHLYfYkQQggpdSpPgAIDA/H69Wv8/PPPSExMhLu7O44fPy7pGP38+XOZ/j0ZGRmYMWMGnjx5An19fXTs2BEhISEwNjaWlFmzZg0AwNfXV+ZYwcHBGDBgQEmfUqmj9b8IIYQQ5ah8HqCyqLzNA9SmDXD6NLBmDfD996qOhhBCCFGNcjMPEPly2dm8CQygGiBCCCGkqCgBKudu3wY+fgRMTAAXF1VHQwghhJQPlACVc7nn/8lnKiRCCCGE5EG3zHKOOkATQgghyqMEqBxjTJoA0QzQhBBCSNFRAlSOPXwIvHkD6OgAjRurOhpCCCGk/KAEqBwT9/9p0gTQ0lJtLIQQQkh5QglQOUbrfxFCCCGfhxKgcow6QBNCCCGfhxKgcio+HoiL40PfmzZVdTSEEEJI+UIJUDklrv1xdwfKwWodhBBCSJlCCVA5lXsCREIIIYQohxKgcoo6QBNCCCGfjxKgciglBbh7lz+nGiBCCCFEeZQAlUOXLvFZoGvVAiwtVR0NIYQQUv5QAlQO0fB3Qggh5MtQAlQOUQdoQggh5MtQAlTOZGQA167x51QDRAghhHweSoDKmRs3gKwsoFo13geIEEIIIcqjBKicyT38XSBQbSyEEEJIeUUJUDlDHaAJIYSQL0cJUDkiFALh4fw5dYAmhBBCPh8lQOXI/ftAaiqgrw+4uak6GkIIIaT8ogSoHBH3/2nWDNDQUG0shBBCSHlGCVA5Qut/EUIIIcWDEqBygjFKgAghhJDiQglQOfH0KfDyJaCpCXh5qToaQgghpHyjBKicEA9/b9wY0NNTbSyEEEJIeUcJUDlB638RQgghxYcSoHKC+v8QQgghxYcSoHLg9WvgwQP+vEUL1cZCCCGEVASUAJUD4tmf69YFTE1VGwshhBBSEVACVA7Q+l+EEEJI8aL5hMsB6gBduQmFQmRnZ6s6DEIIUTlNTU2oq6sXy74oASrjPnwAbt3iz6kGqHJhjCExMREpKSmqDoUQQsoMY2NjWFpaQiAQfNF+KAEq465eBXJygOrVgRo1VB0NKU3i5KdatWrQ09P74v/shBBSnjHG8PHjRyQlJQEArKysvmh/lACVcbmHv9P9r/IQCoWS5MeUer4TQggAQFdXFwCQlJSEatWqfVFzGHWCLuOoA3TlJO7zo0fTfhNCiAzx78Uv7RtJCVAZlpMDXL7Mn1MH6MqJmr0IIURWcf1epASoDIuI4J2gjY2BevVUHQ0hhBBScVACVIaJ+/+0aAGo0U+KVCK+vr4YP3685LW9vT2WLl1a4GcEAgEOHjz4xccurv2QklGU78Ls2bPh7u5eKvHk9vTpUwgEAkRERJT6sQFgwIAB8Pf3V8mxlXX27FkIBAKVjnKl22oZRut/kfKmS5cuaN++vcL3Lly4AIFAgDt37ii93+vXr2PYsGFfGp6M/G6SCQkJ6NChQ7EeixSfvN+FspSw2traIiEhAfXr11d1KKQIKAEqoxijDtCk/Bk8eDBCQ0Px33//yb0XHBwMDw8PNGjQQOn9mpubl1qHcEtLS2hra5fKscqSrKwsVYdQJKX5XVCWuro6LC0toaFBA6zLA0qAyqhHj/giqNraQOPGqo6GlAWM8T5hqngwVrQYO3fuDHNzc2zevFlme3p6Ovbs2YPBgwcjOTkZvXv3ho2NDfT09ODq6oodO3YUuN+8zR4xMTFo2bIldHR0ULduXYSGhsp9ZsqUKahduzb09PRQs2ZNzJw5UzJqZPPmzZgzZw4iIyMhEAggEAgkMeetUbh79y5at24NXV1dmJqaYtiwYUhPT5e8L252WLhwIaysrGBqaopRo0YVOEIlNjYWXbt2hYWFBfT19eHp6YlTp07JlMnMzMSUKVNga2sLbW1t1KpVCxs3bpS8f//+fXTu3BmGhoYwMDCAt7c3YmNjAcg3IQKAv78/BgwYIHNN582bh/79+8PQ0FBSq1LQdRP7559/4OnpCR0dHZiZmSEgIAAAMHfuXIW1H+7u7pg5c6bCa+Hh4YGFCxfKxKmpqSm5xv/99x8EAgEeP34siVv8XbC3twcABAQEQCAQSF6LhYSEwN7eHkZGRujVqxfev3+vMAaAfyeMjY1x4sQJuLi4QF9fH+3bt0dCQoKkjEgkwty5c1G9enVoa2vD3d0dx48fl7yftwns3bt36Nu3L8zNzaGrqwsnJycEBwdLyr948QI9e/aEsbExqlatiq5du+Lp06f5xggU/HMXK+i7GBISAg8PDxgYGMDS0hJ9+vSRzKsDSJumwsLC4OHhAT09PTRv3hwPHz6UlBHXnhZ0fUUiERYsWAAHBwfo6urCzc0Ne/fuzfe8nj17hi5dusDExARVqlRBvXr18O+//xZ4Lb4UJUBllLj2p0kTngQR8vEjoK+vmsfHj0WLUUNDA/3798fmzZvBcmVNe/bsgVAoRO/evZGRkYHGjRvj6NGjuHfvHoYNG4Z+/frh2rVrRTqGSCRCt27doKWlhatXr2Lt2rWYMmWKXDkDAwNs3rwZUVFRWLZsGTZs2IAlS5YAAAIDA/HDDz+gXr16SEhIQEJCAgIDA+X28eHDB/j5+cHExATXr1/Hnj17cOrUKYwePVqm3JkzZxAbG4szZ85gy5Yt2Lx5s1wSmFt6ejo6duyIsLAw3L59G+3bt0eXLl3w/PlzSZn+/ftjx44dWL58OaKjo7Fu3Tro6+sDAOLj49GyZUtoa2vj9OnTuHnzJgYNGoScnJwiXUOxhQsXws3NDbdv35YkKAVdNwA4evQoAgIC0LFjR9y+fRthYWHw8vICAAwaNAjR0dG4fv26pPzt27dx584dDBw4UGEMPj4+OHv2LAA+0d2FCxdgbGyMi///JXju3DnY2NigVq1acp8VHyc4OBgJCQkyx42NjcXBgwdx5MgRHDlyBOfOncNvv/1W4PX4+PEjFi5ciJCQEJw/fx7Pnz/HpEmTJO8vW7YMixYtwsKFC3Hnzh34+fnhm2++QUxMjML9zZw5E1FRUTh27Biio6OxZs0amJmZAeBDuP38/GBgYIALFy4gPDxcknTlVxtXlJ97Yd/F7OxszJs3D5GRkTh48CCePn0qkxiLTZ8+HYsWLcKNGzegoaGBQYMGybxf2PVdsGAB/v77b6xduxb379/HhAkT8N133+HcuXMKz23UqFHIzMzE+fPncffuXfz++++S73uJYUROamoqA8BSU1NVFkNQEGMAYz/9pLIQiAp9+vSJRUVFsU+fPkm2pafz74QqHunpRY89OjqaAWBnzpyRbPP29mbfffddvp/p1KkT++GHHySvfXx82Lhx4ySv7ezs2JIlSxhjjJ04cYJpaGiw+Ph4yfvHjh1jANiBAwfyPcaff/7JGjduLHk9a9Ys5ubmJlcu937Wr1/PTExMWHquC3D06FGmpqbGEhMTGWOMBQUFMTs7O5aTkyMp8+2337LAwMB8Y1GkXr16bMWKFYwxxh4+fMgAsNDQUIVlp02bxhwcHFhWVpbC9/NeP8YY69q1KwsKCpK8trOzY/7+/oXGlfe6NWvWjPXt2zff8h06dGAjRoyQvB4zZgzz9fXNt/zhw4eZkZERy8nJYREREczS0pKNGzeOTZkyhTHG2JAhQ1ifPn1k4hZ/FxhjCn/us2bNYnp6eiwtLU2ybfLkyaxJkyb5xhEcHMwAsMePH0u2rVq1illYWEheW1tbs/nz58t8ztPTk40cOZIxxlhcXBwDwG7fvs0YY6xLly5s4MCBCo8XEhLC6tSpw0QikWRbZmYm09XVZSdOnFD4mcJ+7p/zXbx+/ToDwN6/f88YY+zMmTMMADt16pSkzNGjRxkAye+jwq5vRkYG09PTY5cuXZI51uDBg1nv3r1ljvPu3TvGGGOurq5s9uzZ+caZm6Lfj2LK3L+pobKMog7QJC89PSBXy0upH7uonJ2d0bx5c2zatAm+vr54/PgxLly4gLlz5wLgs1z/+uuv2L17N+Lj45GVlYXMzMwi9+uIjo6Gra0trK2tJduaNWsmV27Xrl1Yvnw5YmNjkZ6ejpycHBgaGhb9RP5/LDc3N1SpUkWyrUWLFhCJRHj48CEsLCwAAPXq1ZOZkdbKygp3797Nd7/p6emYPXs2jh49ioSEBOTk5ODTp0+SGqCIiAioq6vDx8dH4ecjIiLg7e0NTU1Npc4nLw8PD7lthV23iIgIDB06NN99Dh06FIMGDcLixYuhpqaG7du3y9Qg5eXt7Y3379/j9u3buHTpEnx8fODr6yupTTh37hwmT56s9LnZ29vDwMBA8trKykqmqUcRPT09ODo6KvxMWloaXr58iRYtWsh8pkWLFoiMjFS4vxEjRqB79+64desWvv76a/j7+6N58+YAgMjISDx+/FgmRgDIyMiQa9ISK8rPvbDv4s2bNzF79mxERkbi3bt3EIlEAIDnz5+jbt26knK5++qJl5xISkpCjf+vyVTQ9X38+DE+fvyIdu3aycSWlZWFhg0bKox77NixGDFiBE6ePIm2bduie/fun9VfUBmUAJVBL18CT57wpS8U/F4nlZRAAOS6D5dpgwcPxpgxY7Bq1SoEBwfD0dFRcjP/888/sWzZMixduhSurq6oUqUKxo8fX6ydcC9fvoy+fftizpw58PPzg5GREXbu3IlFixYV2zFyy3tDEggEkhuLIpMmTUJoaCgWLlyIWrVqQVdXFz169JBcA/F0//kp7H01NTWZJkhA8ay5VfJ8oYpy3Qo7dpcuXaCtrY0DBw5AS0sL2dnZ6NGjR77ljY2N4ebmhrNnz+Ly5cto164dWrZsicDAQDx69AgxMTH5JoIFUfZnkt9n8l5HZXTo0AHPnj3Dv//+i9DQULRp0wajRo3CwoULkZ6ejsaNG2Pbtm1ynzM3N1e4v8KuPVDweYubdP38/LBt2zaYm5vj+fPn8PPzk/v/l3s/4okHc1+/go4j7r919OhR2NjYyJTLb4DBkCFD4Ofnh6NHj+LkyZNYsGABFi1ahDFjxhR6zp+L+gCVQeL+P25ugJGRamMh5HP07NlT8tf/33//jUGDBkl+iYaHh6Nr16747rvv4Obmhpo1a+LRo0dF3reLiwtevHgh0zn1ypUrMmUuXboEOzs7TJ8+HR4eHnBycsKzZ89kymhpaUEoFBZ6rMjISHz48EGyLTw8HGpqaqhTp06RY84rPDwcAwYMQEBAAFxdXWFpaSnT+dXV1RUikSjf/hINGjTAhQsX8u1obW5uLnN9hEIh7t27V2hcRbluDRo0QFhYWL770NDQQFBQEIKDgxEcHIxevXoVeuP28fHBmTNncP78efj6+qJq1apwcXHB/PnzYWVlhdq1a+f7WU1NzUJ/jsXB0NAQ1tbWCA8Pl9keHh4uU3OSl7m5OYKCgrB161YsXboU69evBwA0atQIMTExqFatGmrVqiXzMMrnF39hP/fCPHjwAMnJyfjtt9/g7e0NZ2fnQmvFPkfdunWhra2N58+fy52bra1tvp+ztbXF999/j/379+OHH37Ahg0bij223CgBKoNo+Dsp7/T19REYGIhp06YhISFBppOlk5MTQkNDcenSJURHR2P48OF49epVkffdtm1b1K5dG0FBQYiMjMSFCxcwffp0mTJOTk54/vw5du7cidjYWCxfvhwHDhyQKWNvb4+4uDhERETgzZs3yMzMlDtW3759oaOjg6CgINy7dw9nzpzBmDFj0K9fP0nz1+dwcnLC/v37ERERgcjISPTp00fmr2t7e3sEBQVh0KBBOHjwIOLi4nD27Fns3r0bADB69GikpaWhV69euHHjBmJiYhASEiIZqdO6dWscPXoUR48exYMHDzBixIgiTThXlOs2a9Ys7NixA7NmzUJ0dLSkw2puQ4YMwenTp3H8+HG5zrOK+Pr64sSJE9DQ0ICzs7Nk27Zt2wqt/bG3t0dYWBgSExPx7t27Qo/1JSZPnozff/8du3btwsOHDzF16lRERERg3LhxCsv//PPPOHToEB4/foz79+/jyJEjcHFxAcC/W2ZmZujatSsuXLgg+RmPHTtW4TQSQOE/98LUqFEDWlpaWLFiBZ48eYLDhw9j3rx5n3cxCmBgYIBJkyZhwoQJ2LJlC2JjY3Hr1i2sWLECW7ZsUfiZ8ePH48SJE4iLi8OtW7dw5swZybUqKZQAlUHi/j+0/hcpzwYPHox3797Bz89Ppr/OjBkz0KhRI/j5+cHX1xeWlpZKzV6rpqaGAwcO4NOnT/Dy8sKQIUMwf/58mTLffPMNJkyYgNGjR8Pd3R2XLl2SG4bdvXt3tG/fHq1atYK5ubnCofh6eno4ceIE3r59C09PT/To0QNt2rTBypUrlbsYeSxevBgmJiZo3rw5unTpAj8/PzRq1EimzJo1a9CjRw+MHDkSzs7OGDp0qKQmytTUFKdPn0Z6ejp8fHzQuHFjbNiwQdIsMWjQIAQFBaF///7w8fFBzZo10apVq0LjKsp18/X1xZ49e3D48GG4u7ujdevWciP4nJyc0Lx5czg7O6NJkyaFHtfb2xsikUgm2fH19YVQKISvr2+Bn120aBFCQ0Nha2ubb/+S4jJ27FhMnDgRP/zwA1xdXXH8+HEcPnwYTk5OCstraWlh2rRpaNCgAVq2bAl1dXXs3LkTAP9unT9/HjVq1EC3bt3g4uKCwYMHIyMjI9++aoX93AsjnqJiz549qFu3Ln777TeZKQiK07x58zBz5kwsWLAALi4uaN++PY4ePQoHBweF5YVCIUaNGiUpW7t2baxevbpEYhMTsC9p4Kyg0tLSYGRkhNTUVKU7TX6p1FSgalVAJALi44Fc9w1SiWRkZCAuLg4ODg7Q0dFRdTiEKIUxBicnJ4wcORITJ05UdTikgino96My92/qBF3GXL7Mk5+aNSn5IYSUP69fv8bOnTuRmJiY79w/hJQFlACVMTT8nRBSnlWrVg1mZmZYv349TExMVB0OIfmiBKiMoQ7QhJDyjHpVkPKCOkGXIZmZwNWr/Dl1gCaEEEJKDiVAZcjNmzwJMjcHCpj2ghBCCCFfiBKgMiT38Pf/zxlHCCGEkBJACVAZQh2gCSGEkNJBCVAZIRIB4hnWqf8PIYQQUrIoASoj7t8HUlL4YpclPJkpIWWer68vxo8fL3ltb2+PpUuXFvgZgUCAgwcPfvGxi2s/pGQU5bswe/ZsuLu7l0o8ZcnmzZthbGys6jCKTNX/1ygBKiPEw9+bNQM0aHICUk516dIF7du3V/jehQsXIBAIcOfOHaX3e/36dQwbNuxLw5OR300yISEBHTp0KNZjkeKT97ug6psoKb8oASojaP0vUhEMHjwYoaGhChdzDA4OhoeHBxo0aKD0fs3NzaGnp1ccIRbK0tIS2trapXKssiQrK0vVIRRJaX4XvkR5uZ6VGSVAZQBj1AGaVAydO3eWLLiYW3p6Ovbs2YPBgwcjOTkZvXv3ho2NDfT09ODq6qpwIdLc8jZ7xMTEoGXLltDR0UHdunURGhoq95kpU6agdu3a0NPTQ82aNTFz5kxkZ2cD4E0Fc+bMQWRkJAQCAQQCgSTmvDUKd+/eRevWraGrqwtTU1MMGzYM6enpkvcHDBgAf39/LFy4EFZWVjA1NcWoUaMkx1IkNjYWXbt2hYWFBfT19eHp6YlTp07JlMnMzMSUKVNga2sLbW1t1KpVCxs3bpS8f//+fXTu3BmGhoYwMDCAt7c3YmNjAcg3IQKAv78/BgwYIHNN582bh/79+8PQ0FBSq1LQdRP7559/4OnpCR0dHZiZmSEgIAAAMHfuXNSvX1/ufN3d3eUWVRXz8PCQWZDT398fmpqakmv833//QSAQ4PHjx5K4xd8Fe3t7AEBAQAAEAoHktVhISAjs7e1hZGSEXr164f379wpjEF8z8Xch9+Pp06cAgJSUFAwZMgTm5uYwNDRE69atERkZKfm8uEbxr7/+klmj6vnz5+jatSv09fVhaGiInj174tWrV5LPRUZGolWrVjAwMIChoSEaN26MGzdu5BtnSkoKhg8fDgsLC+jo6KB+/fo4cuSITJkTJ07AxcUF+vr6aN++PRISEiTvXb9+He3atYOZmRmMjIzg4+ODW7duyXxeIBDgr7/+QkBAAPT09ODk5ITDhw9L3j979iwEAgHCwsLg4eEBPT09NG/eXG5V+kOHDqFRo0bQ0dFBzZo1MWfOHOTk5Cg8r6ysLIwePRpWVlbQ0dGBnZ0dFixYkO91KA6UAJUBz58D//3Hm76KsHAyqeQ+fMj/kZFR9LKfPhWtrDI0NDTQv39/bN68WWZG4D179kAoFKJ3797IyMhA48aNcfToUdy7dw/Dhg1Dv3795FYUz49IJEK3bt2gpaWFq1evYu3atZgyZYpcOQMDA2zevBlRUVFYtmwZNmzYgCVLlgAAAgMD8cMPP6BevXpISEhAQkICAgMD5fbx4cMH+Pn5wcTEBNevX8eePXtw6tQpjB49WqbcmTNnEBsbizNnzmDLli3YvHmzXBKYW3p6Ojp27IiwsDDcvn0b7du3R5cuXfD8+XNJmf79+2PHjh1Yvnw5oqOjsW7dOujr6wMA4uPj0bJlS2hra+P06dO4efMmBg0alO/NJT8LFy6Em5sbbt++LUlQCrpuAHD06FEEBASgY8eOuH37NsLCwuDl5QWAr0IfHR2N69evS8rfvn0bd+7cyXddMB8fH5w9exYAn0X6woULMDY2xsX/9ws4d+4cbGxsUKtWLbnPio8THByMhIQEmePGxsbi4MGDOHLkCI4cOYJz587ht99+y/da7N+/X/JdSEhIQLdu3VCnTh1YWFgAAL799lskJSXh2LFjuHnzJho1aoQ2bdrg7du3kn08fvwY+/btw/79+xEREQGRSISuXbvi7du3OHfuHEJDQ/HkyROZ71rfvn1RvXp1XL9+HTdv3sTUqVPzXd1dJBKhQ4cOCA8Px9atWxEVFYXffvsN6urqkjIfP37EwoULERISgvPnz+P58+eYNGmS5P33798jKCgIFy9exJUrV+Dk5ISOHTvKJYdz5sxBz549cefOHXTs2BF9+/aVOVcAmD59OhYtWoQbN25AQ0MDgwYNkrx34cIF9O/fH+PGjUNUVBTWrVuHzZs3Y/78+QrPbfny5Th8+DB2796Nhw8fYtu2bXIJbbFjRE5qaioDwFJTU0vleCEhjAGMeXmVyuFIOfDp0ycWFRXFPn36JPcerzNU/OjYUbasnl7+ZX18ZMuamSkup6zo6GgGgJ05c0ayzdvbm3333Xf5fqZTp07shx9+kLz28fFh48aNk7y2s7NjS5YsYYwxduLECaahocHi4+Ml7x87dowBYAcOHMj3GH/++Sdr3Lix5PWsWbOYm5ubXLnc+1m/fj0zMTFh6enpkvePHj3K1NTUWGJiImOMsaCgIGZnZ8dycnIkZb799lsWGBiYbyyK1KtXj61YsYIxxtjDhw8ZABYaGqqw7LRp05iDgwPLyspS+H7e68cYY127dmVBQUGS13Z2dszf37/QuPJet2bNmrG+ffvmW75Dhw5sxIgRktdjxoxhvr6++ZY/fPgwMzIyYjk5OSwiIoJZWlqycePGsSlTpjDGGBsyZAjr06ePTNzi7wJjTOHPfdasWUxPT4+lpaVJtk2ePJk1adKk0PNljLHFixczY2Nj9vDhQ8YYYxcuXGCGhoYsIyNDppyjoyNbt26d5JiamposKSlJ8v7JkyeZuro6e/78uWTb/fv3GQB27do1xhhjBgYGbPPmzUWK68SJE0xNTU0SV17BwcEMAHv8+LFk26pVq5iFhUW++xQKhczAwID9888/km0A2IwZMySv09PTGQB27NgxxhhjZ86cYQDYqVOnJGWOHj3KAEh+Z7Vp04b9+uuvMscKCQlhVlZWMscR/+zGjBnDWrduzUQiUWGXocDfj8rcv6kGqAyg9b9IReLs7IzmzZtj06ZNAPhfxRcuXMDgwYMBAEKhEPPmzYOrqyuqVq0KfX19nDhxQqb2oyDR0dGwtbWFtbW1ZFuzZs3kyu3atQstWrSApaUl9PX1MWPGjCIfI/ex3NzcUKVKFcm2Fi1aQCQSyVT316tXT+avcCsrKyQlJeW73/T0dEyaNAkuLi4wNjaGvr4+oqOjJfFFRERAXV0dPj4+Cj8fEREBb2/vfGsKisrDw0NuW2HXLSIiAm3atMl3n0OHDsWOHTuQkZGBrKwsbN++XaZmIC9vb2+8f/8et2/fxrlz5+Dj4wNfX19JrdC5c+fg6+ur9LnZ29vDwMBA8rqwn4nYsWPHMHXqVOzatQu1/z8lf2RkJNLT02Fqagp9fX3JIy4uTtLsCAB2dnYwNzeXvBZ/V21tbSXb6tatC2NjY0RHRwMAJk6ciCFDhqBt27b47bffZPaXV0REBKpXry6JSxE9PT04Ojrme96vXr3C0KFD4eTkBCMjIxgaGiI9PV3u/0buvnpVqlSBoaGh3PXLXcbKygoAJGUiIyMxd+5cmes1dOhQJCQk4OPHj3JxDxgwABEREahTpw7Gjh2LkydP5nuOxYXGG5UB1AGaKCNX9xM5ue7BAICCft+r5fnz5/9dHYrF4MGDMWbMGKxatQrBwcFwdHSU3Mz//PNPLFu2DEuXLoWrqyuqVKmC8ePHF2un0cuXL6Nv376YM2cO/Pz8YGRkhJ07d2LRokXFdozc8iYiAoEAIpEo3/KTJk1CaGgoFi5ciFq1akFXVxc9evSQXANdXd0Cj1fY+2pqanKLkirqk5Q7sQOKdt0KO3aXLl2gra2NAwcOQEtLC9nZ2ejRo0e+5Y2NjeHm5oazZ8/i8uXLaNeuHVq2bInAwEA8evQIMTEx+SaCBVH2ZwIAUVFR6NWrF3777Td8/fXXku3p6emwsrKSJGV54xfLez2LYvbs2ejTpw+OHj2KY8eOYdasWdi5c6ekX1VuhV17QPF55/4uBAUFITk5GcuWLYOdnR20tbXRrFkzuf9/Rbl+ucsI/r98gbhMeno65syZg27dusnFKO4flVujRo0QFxeHY8eO4dSpU+jZsyfatm2LvXv3FnrOn6tM1ACtWrUK9vb20NHRQZMmTQrsC5CdnY25c+fC0dEROjo6cHNzw/Hjx79on6qUnAxERfHnLVqoNhZSPlSpkv8j7++Vgsrm/V2aX7nP0bNnT6ipqWH79u34+++/MWjQIMkvyPDwcHTt2hXfffcd3NzcULNmTTx69KjI+3ZxccGLFy9kOnZeuXJFpsylS5dgZ2eH6dOnw8PDA05OTnj27JlMGS0tLQiFwkKPFRkZiQ+5OkOFh4dDTU0NderUKXLMeYWHh2PAgAEICAiAq6srLC0tJZ1tAcDV1RUikQjnzp1T+PkGDRrgwoUL+Xa0Njc3l7k+QqEQ9+7dKzSuoly3Bg0aICwsLN99aGhoICgoCMHBwQgODkavXr0KvXH7+PjgzJkzOH/+PHx9fVG1alW4uLhg/vz5sLKyKrDGQ1NTs9CfY1G8efMGXbp0Qffu3TFhwgSZ9xo1aoTExERoaGigVq1aMg8zM7N89yn+rr548UKyLSoqCikpKahbt65kW+3atTFhwgScPHkS3bp1Q3BwsML9NWjQAP/9959S/1/yCg8Px9ixY9GxY0fUq1cP2traePPmzWfvLz+NGjXCw4cP5a5XrVq1oJb3r6//MzQ0RGBgIDZs2IBdu3Zh3759cv2OipPKE6Bdu3Zh4sSJmDVrFm7dugU3Nzf4+fnlW1U5Y8YMrFu3DitWrEBUVBS+//57BAQE4Pbt25+9T1USz/7s7MwXQSWkItDX10dgYCCmTZuGhIQEmdFHTk5OCA0NxaVLlxAdHY3hw4fLjIopTNu2bVG7dm0EBQUhMjISFy5cwPTp02XKODk54fnz59i5cydiY2OxfPlyHDhwQKaMvb094uLiEBERgTdv3iAzM1PuWH379oWOjg6CgoJw7949nDlzBmPGjEG/fv0knWM/h5OTk6SjbGRkJPr06SPz17W9vT2CgoIwaNAgHDx4EHFxcTh79ix2794NABg9ejTS0tLQq1cv3LhxAzExMQgJCZE0y7Vu3RpHjx7F0aNH8eDBA4wYMQIpKSlFiquw6zZr1izs2LEDs2bNQnR0NO7evYvff/9dpsyQIUNw+vRpHD9+vMDmLzFfX1+cOHECGhoacHZ2lmzbtm1bobU/9vb2CAsLQ2JiIt69e1fosfLTvXt36OnpYfbs2UhMTJQ8hEIh2rZti2bNmsHf3x8nT57E06dPcenSJUyfPr3AEVtt27aFq6sr+vbti1u3buHatWvo378/fHx84OHhgU+fPmH06NE4e/Ysnj17hvDwcFy/fh0uLi4K9+fj44OWLVuie/fuCA0NldSYKKoEyI+TkxNCQkIQHR2Nq1evom/fvkWqWVLWzz//jL///htz5szB/fv3ER0djZ07d2LGjBkKyy9evBg7duzAgwcP8OjRI+zZsweWlpYlO7Fjob2ESpiXlxcbNWqU5LVQKGTW1tZswYIFCstbWVmxlStXymzr1q2bTKc8ZfeZV2l2gp40iXc0HTq0xA9FypGCOvmVF5cuXWIAWMc8PbOTk5NZ165dmb6+PqtWrRqbMWMG69+/P+vataukTEGdoBnjnYS/+uorpqWlxWrXrs2OHz8u1xl28uTJzNTUlOnr67PAwEC2ZMkSZmRkJHk/IyODde/enRkbGzMALDg4mDEm36n2zp07rFWrVkxHR4dVrVqVDR06lL1//17yflBQkEzsjDE2btw45pO3l3kucXFxrFWrVkxXV5fZ2tqylStXyp3zp0+f2IQJE5iVlRXT0tJitWrVYps2bZK8HxkZyb7++mump6fHDAwMmLe3N4uNjWWMMZaVlcVGjBjBqlatyqpVq8YWLFigsBN07mta1OvGGGP79u1j7u7uTEtLi5mZmbFu3brJ7cfb25vVq1cv32uQW3JyMhMIBDIdxw8cOMAAsLVr18qUzRv34cOHWa1atZiGhgazs7NjjCnu4L5kyRLJ+4oAUPiIi4tjjDGWlpbGxowZw6ytrZmmpiaztbVlffv2lXRwzq9T/bNnz9g333zDqlSpwgwMDNi3334r6UCfmZnJevXqxWxtbZmWlhaztrZmo0ePLvD/fXJyMhs4cCAzNTVlOjo6rH79+uzIkSOMMd4JOu/PSnwdxW7dusU8PDyYjo4Oc3JyYnv27ClSx3IjIyPJ/xFxJ+h3795J3r99+7bM9WKMsePHj7PmzZszXV1dZmhoyLy8vNj69esVHmf9+vXM3d2dValShRkaGrI2bdqwW7duKbwGxdUJWvD/IFQiKysLenp62Lt3L/z9/SXbg4KCkJKSgkOHDsl9xtTUFH/88YekQyUAfPfdd7h48SKePn36WfvMzMyU+esvLS0Ntra2SE1NhaGhYfGcbD6aNQOuXAH+/hvo169ED0XKkYyMDMTFxcnMJ0JIecEYg5OTE0aOHImJEyeqOhxSwRT0+zEtLQ1GRkZFun+rtAnszZs3EAqFclXJFhYWSExMVPgZPz8/LF68GDExMRCJRAgNDZXM3/C5+1ywYAGMjIwkj9w99kvSx4+AuPaUOkATQiqC169fY+XKlUhMTMx37h9CygKV9wFS1rJly+Dk5ARnZ2doaWlh9OjRGDhwYL6dqopi2rRpSE1NlTxyd1grSdeuATk5gI0NUNLzPRFCSGmoVq0a5s6di/Xr18PExETV4RCSL5UOgzczM4O6urpcB8hXr17B0tJS4WfMzc1x8OBBZGRkIDk5GdbW1pg6dSpq1qz52fvU1tZWydo/uYe//3+ADCGElGsq7FVBiFJUWgOkpaWFxo0bywypFIlECAsLUzixWW46OjqwsbFBTk4O9u3bh65du37xPksbrf9FCCGEqIbKJ0KcOHEigoKC4OHhAS8vLyxduhQfPnyQtB33798fNjY2kkXRrl69ivj4eLi7uyM+Ph6zZ8+GSCTCjz/+WOR9lgU5OcDly/w59f8hhBBCSpfKE6DAwEC8fv0aP//8MxITE+Hu7o7jx49LOjE/f/5cpn9PRkYGZsyYgSdPnkBfXx8dO3ZESEiIzFwBhe2zLIiM5DP6GhkBChZPJoQQQkgJUukw+LJKmWF0n2vZMmD8eKBjR+Do0RI5BCnHaBg8IYQoViGGwVdmtP4XIYQQojqUAKkAY9QBmhBCCFElSoBU4PFjvkq3lhbg4aHqaAgpW54+fQqBQICIiIgif2bz5s3FvmbQ58TxOezt7bF06VLJa4FAgIMHD5Z4XKV1foSUVZQAqYC49sfLS371bkIqghcvXmDQoEGwtraGlpYW7OzsMG7cOCQnJxf6WVtbWyQkJKC+EqMDAgMDv2iF7M/h6uqK77//XuF7ISEhn73KdkJCAjp06PCl4ckYMGCAzNJAwOddZ0IqEkqAVODiRf4vNX+RiujJkyfw8PBATEwMduzYgcePH2Pt2rWSubjevn2b72ezsrKgrq4OS0tLaGgUfZCqrq4uqlWrVhzhF9ngwYOxc+dOfPr0Se694OBgfPPNNzAzM1N6v5aWlqUyMevnXOeKIjs7W9UhkDKAEiAVoA7QpCIbNWoUtLS0cPLkSfj4+KBGjRro0KEDTp06hfj4eEyfPl1S1t7eHvPmzUP//v1haGiIYcOGKWyaOXz4MJycnKCjo4NWrVphy5YtEAgESElJASDfBDZ79my4u7sjJCQE9vb2MDIyQq9evfD+/XtJmePHj+Orr76CsbExTE1N0blzZ8TGxhb5PL/77jt8+vQJ+/btk9keFxeHs2fPYvDgwYiNjUXXrl1hYWEBfX19eHp64tSpUwXuN28T2LVr19CwYUPo6OjAw8MDt2/flikvFAoxePBgODg4QFdXF3Xq1MGyZctkrsWWLVtw6NAhCAQCCAQCnD17VuF1PnfuHLy8vKCtrQ0rKytMnToVOTk5kvd9fX0xduxY/Pjjj6hatSosLS0xe/bsAs/n+vXraNeuHczMzGBkZAQfHx/cunVLpkxKSgqGDx8OCwsL6OjooH79+jhy5Ijk/fDwcPj6+kJPTw8mJibw8/PDu3fvAMg3IQKAu7u7TFwCgQBr1qzBN998gypVqmD+/PmFXjexTZs2oV69epJrMnr0aADAoEGD0LlzZ5my2dnZqFatGjZu3FjgNSFlAyVApSwxkfcBEgiA5s1VHQ0pjz584I/cE1hkZfFtmZmKy4pE0m3Z2XxbRkbRyirj7du3OHHiBEaOHAldXV2Z9ywtLdG3b1/s2rVLZrmEhQsXws3NDbdv38bMmTPl9hkXF4cePXrA398fkZGRGD58uEwSlZ/Y2FgcPHgQR44cwZEjR3Du3Dn89ttvuc73AyZOnIgbN24gLCwMampqCAgIgCj3BSiAmZkZunbtik2bNsls37x5M6pXr46vv/4a6enp6NixI8LCwnD79m20b98eXbp0wfPnz4t0jPT0dHTu3Bl169bFzZs3MXv2bEyaNEmmjEgkQvXq1bFnzx5ERUXh559/xk8//YTdu3cDACZNmoSePXuiffv2SEhIQEJCApor+OUTHx+Pjh07wtPTE5GRkVizZg02btyIX375Rabcli1bUKVKFVy9ehV//PEH5s6di9DQ0HzP4f379wgKCsLFixdx5coVODk5oWPHjpJkVCQSoUOHDggPD8fWrVsRFRWF3377Derq6gCAiIgItGnTBnXr1sXly5dx8eJFdOnSBUKhsEjXUGz27NkICAjA3bt3MWjQoEKvGwCsWbMGo0aNwrBhw3D37l0cPnwYtWrVAgAMGTIEx48flyzEDQBHjhzBx48fERgYqFRsREUYkZOamsoAsNTU1GLf9549jAGMNWhQ7LsmFcinT59YVFQU+/Tpk9x7PPVhLClJuu2XX/i2IUNky+rp8e1xcdJtS5bwbX36yJY1M+Pb792Tblu/Xrm4r1y5wgCwAwcOKHx/8eLFDAB79eoVY4wxOzs75u/vL1MmLi6OAWC3b99mjDE2ZcoUVr9+fZky06dPZwDYu3fvGGOMBQcHMyMjI8n7s2bNYnp6eiwtLU2ybfLkyaxJkyb5xv769WsGgN29e1dhHIocP36cCQQC9uTJE8YYYyKRiNnZ2bEZM2bk+5l69eqxFStWSF7b2dmxJUuWSF7nvn7r1q1jpqamMt+DNWvWFBrXqFGjWPfu3SWvg4KCWNeuXWXK5D2/n376idWpU4eJRCJJmVWrVjF9fX0mFAoZY4z5+Piwr776SmY/np6ebMqUKfnGkpdQKGQGBgbsn3/+YYwxduLECaampsYePnyosHzv3r1ZixYt8t1f3uvHGGNubm5s1qxZktcA2Pjx4wuNLe91s7a2ZtOnT8+3fN26ddnvv/8ued2lSxc2YMCAQo9DvkxBvx+VuX9TDVApo+HvpDJgSsyv6lHIUMiHDx/C09NTZpuXl1eh+7W3t4eBgYHktZWVFZKSkiSvY2Ji0Lt3b9SsWROGhoawt7cHgCLXzgBAu3btUL16dQQHBwMAwsLC8Pz5c8myO+np6Zg0aRJcXFxgbGwMfX19REdHF/kY0dHRaNCggcxkb4rWNFy1ahUaN24Mc3Nz6OvrY/369Uqdh/hYzZo1gyDXyswtWrRAeno6/vvvP8m2Bg0ayHwu73XN69WrVxg6dCicnJxgZGQEQ0NDpKenS+KLiIhA9erVUbt2bYWfF9cAfSlF37OCrltSUhJevnxZ4LGHDBki+dm/evUKx44dw6BBg744VlI6KAEqZdQBmnyp9HT+yN2/dvJkvm3lStmySUl8e40a0m2jRvFtebspPH3Kt7u4SLcNGKBcbLVq1YJAIEB0dLTC96Ojo2FiYgJzc3PJtipVqih3kCLS1NSUeS0QCGSat7p06YK3b99iw4YNuHr1Kq5evQqAd8QuKjU1NQwYMABbtmyBSCRCcHAwWrVqhZo1awLgzU8HDhzAr7/+igsXLiAiIgKurq5KHaMwO3fuxKRJkzB48GCcPHkSERERGDhwYLEeI7fCrmteQUFBiIiIwLJly3Dp0iVERETA1NRUEl/eptK8CntfTU1NLuFW1Mk57/essOtW2HEBvlblkydPcPnyZWzduhUODg7wpl/u5QYlQKUoLQ0Q9zekDtDkc1Wpwh+5/lCHlhbflnfwkLhsruX0oKnJt+WdgiG/ssowNTVFu3btsHr1arnRUYmJidi2bRsCAwNlahkKU6dOHdy4cUNm2/Xr15ULLI/k5GQ8fPgQM2bMQJs2beDi4iLpVKusgQMH4sWLF9i/fz8OHDiAwYMHS94LDw/HgAEDEBAQAFdXV1haWuLp06dF3reLiwvu3LmDjFwdtq5cuSJTJjw8HM2bN8fIkSPRsGFD1KpVS64zt5aWVqF9ZlxcXHD58mWZZCI8PBwGBgaoXr16kWPOKzw8HGPHjkXHjh0lnYlzTw/QoEED/Pfff/lOY9CgQQOEhYXlu39zc3OZfjhpaWmIi4srUlwFXTcDAwPY29sXeGxTU1P4+/sjODgYmzdvLlMLbpPCUQJUiq5c4R1MHRwAGxtVR0NIyVi5ciUyMzPh5+eH8+fP48WLFzh+/DjatWsHGxsbzJ8/X6n9DR8+HA8ePMCUKVPw6NEj7N69G5s3bwYApRKp3ExMTGBqaor169fj8ePHOH36NCZOnPhZ+3JwcEDr1q0xbNgwaGtro1u3bpL3nJycsH//fkRERCAyMhJ9+vQpcidrAOjTpw8EAgGGDh2KqKgo/Pvvv1i4cKFMGScnJ9y4cQMnTpzAo0ePMHPmTLkE0d7eHnfu3MHDhw/x5s0bhTUkI0eOxIsXLzBmzBg8ePAAhw4dwqxZszBx4kSZBamV5eTkhJCQEERHR+Pq1avo27evTO2Kj48PWrZsie7duyM0NBRxcXE4duwYjh8/DgCYNm0arl+/jpEjR+LOnTt48OAB1qxZI0miWrdujZCQEFy4cAF3795FUFCQpAN1YXEVdt1mz56NRYsWYfny5YiJicGtW7ewYsUKmTJDhgzBli1bEB0djaCgoM++TqT0UQJUihIT+ervVPtDKjLxjaVmzZro2bMnHB0dMWzYMLRq1QqXL19G1apVldqfg4MD9u7di/3796NBgwZYs2aNZBTY586Xo6amhp07d+LmzZuoX78+JkyYgD///POz9gXwOYHevXuHPn36yPTXWbx4MUxMTNC8eXN06dIFfn5+aNSoUZH3q6+vj3/++Qd3795Fw4YNMX36dPz+++8yZYYPH45u3bohMDAQTZo0QXJyMkaOHClTZujQoahTpw48PDxgbm6O8PBwuWPZ2Njg33//xbVr1+Dm5obvv/8egwcPxowZM5S8GrI2btyId+/eoVGjRujXrx/Gjh0rN2fTvn374Onpid69e6Nu3br48ccfJTVWtWvXxsmTJxEZGQkvLy80a9YMhw4dksxfNG3aNPj4+KBz587o1KkT/P394ejoWGhcRbluQUFBWLp0KVavXo169er9r737j4myfuAA/j4OPB9u/NY77kLyLKaIP9JgDujHCiZS0miUY7vcVX847TAwbcMMtQWaluKsPLpm+IcayzaMLGpEjpKlUnaEk5CWJcshOp3HwWSN+3z/8Ott95X8AgLPY5/3a7uNe54D3sdn3r19ns89HyxduhSdnZ1Bj8nOzobFYkFOTg6sVuud/KlognE1+CGM52rwg4NAby8wxlftp38ZrgZ/exUVFaiqqkJXV5faUUhyPp8P99xzD6qrq4OO/tH4GavV4OW7BKjK9HqWH6KR2rNnD9LS0hAXF4fm5ma8/fbbgQvSEanB7/fj8uXL2LFjB6Kjo/HUU0+pHYlGiAWIiDSvs7MT5eXluHLlChITE7F27VqsX79e7VgksfPnz8NmsyEhIQH79u2TckmRux1HjIg0r7KyEpWVlWrHIAqYPn36iK53RdrDSdBEREQkHRYgIiIikg4LEJGG8RA7EVGwsXpdZAEi0qCbyw309/ernISISFtuvi7+77IsI8VJ0EQapNfrER0dHVhkMjw8fNRXPSYi+jcQQqC/vx89PT2Ijo4e1hW/b4cFiEij4uPjAeC2K20TEckmOjo68Pp4J1iAiDRKp9PBYrHAZDINuXYTEZFswsLC7vjIz00sQEQap9frx+wfPBER3cBJ0ERERCQdFiAiIiKSDgsQERERSYdzgIZw8yJLXq9X5SREREQ0XDfft4dzsUQWoCH09vYCAKZNm6ZyEiIiIhqp3t5eREVF3fYxOsFr7d/C7/fjwoULiIiI4MXn/oHX68W0adPQ1dWFyMhIteNIj+OhLRwPbeF4aM94jYkQAr29vbBarQgJuf0sHx4BGkJISAgSEhLUjnFXiIyM5AuKhnA8tIXjoS0cD+0ZjzH5f0d+buIkaCIiIpIOCxARERFJhwWIRsVgMGDTpk0wGAxqRyFwPLSG46EtHA/t0cKYcBI0ERERSYdHgIiIiEg6LEBEREQkHRYgIiIikg4LEBEREUmHBYiGbevWrUhLS0NERARMJhPy8/PR0dGhdiz6r7feegs6nQ4lJSVqR5HaX3/9heeeew5xcXFQFAVz587Fjz/+qHYsKQ0ODqKsrAw2mw2KouC+++7Dm2++Oax1oujOfffdd8jLy4PVaoVOp8Phw4eD9gshsHHjRlgsFiiKguzsbHR2dk5YPhYgGrampiY4nU4cP34cDQ0N+Pvvv7F48WL09fWpHU16LS0t+OCDDzBv3jy1o0jt6tWryMzMRFhYGOrr63HmzBns2LEDMTExakeT0rZt2+ByufDee++hvb0d27Ztw/bt2/Huu++qHU0KfX19mD9/Pt5///0h92/fvh27d+9GVVUVTpw4AaPRiJycHFy/fn1C8vFj8DRqly5dgslkQlNTEx555BG140jL5/Nh4cKF2LNnD8rLy/HAAw9g165daseSUmlpKZqbm/H999+rHYUALF26FGazGXv37g1sKygogKIo2L9/v4rJ5KPT6VBbW4v8/HwAN47+WK1WrF27FuvWrQMAXLt2DWazGfv27UNhYeG4Z+IRIBq1a9euAQBiY2NVTiI3p9OJJ598EtnZ2WpHkV5dXR1SU1Px7LPPwmQyYcGCBfjwww/VjiWtjIwMNDY24uzZswCA1tZWHDt2DLm5uSono3PnzqG7uzvodSsqKgqLFi3CDz/8MCEZuBgqjYrf70dJSQkyMzMxZ84cteNIq6amBqdOnUJLS4vaUQjA77//DpfLhVdeeQWvvfYaWlpa8PLLL2PSpElwOBxqx5NOaWkpvF4vZs2aBb1ej8HBQVRUVMBut6sdTXrd3d0AALPZHLTdbDYH9o03FiAaFafTidOnT+PYsWNqR5FWV1cXiouL0dDQgMmTJ6sdh3DjPwapqanYsmULAGDBggU4ffo0qqqqWIBU8Mknn+DAgQM4ePAgUlJS4PF4UFJSAqvVyvEgngKjkSsqKsKRI0dw9OhRJCQkqB1HWj/99BN6enqwcOFChIaGIjQ0FE1NTdi9ezdCQ0MxODiodkTpWCwWzJ49O2hbcnIyzp8/r1Iiub366qsoLS1FYWEh5s6di+XLl2PNmjXYunWr2tGkFx8fDwC4ePFi0PaLFy8G9o03FiAaNiEEioqKUFtbi2+//RY2m03tSFLLyspCW1sbPB5P4Jaamgq73Q6PxwO9Xq92ROlkZmbecmmIs2fP4t5771Upkdz6+/sREhL8NqfX6+H3+1VKRDfZbDbEx8ejsbExsM3r9eLEiRNIT0+fkAw8BUbD5nQ6cfDgQXz22WeIiIgInKeNioqCoigqp5NPRETELfOvjEYj4uLiOC9LJWvWrEFGRga2bNmCZcuW4eTJk3C73XC73WpHk1JeXh4qKiqQmJiIlJQU/Pzzz9i5cydefPFFtaNJwefz4bfffgvcP3fuHDweD2JjY5GYmIiSkhKUl5cjKSkJNpsNZWVlsFqtgU+KjTtBNEwAhrxVV1erHY3+69FHHxXFxcVqx5Da559/LubMmSMMBoOYNWuWcLvdakeSltfrFcXFxSIxMVFMnjxZzJgxQ2zYsEEMDAyoHU0KR48eHfI9w+FwCCGE8Pv9oqysTJjNZmEwGERWVpbo6OiYsHy8DhARERFJh3OAiIiISDosQERERCQdFiAiIiKSDgsQERERSYcFiIiIiKTDAkRERETSYQEiIiIi6bAAERH9A51Oh8OHD6sdg4jGAQsQEWnS888/D51Od8ttyZIlakcjon8BrgVGRJq1ZMkSVFdXB20zGAwqpSGifxMeASIizTIYDIiPjw+6xcTEALhxesrlciE3NxeKomDGjBn49NNPg76/ra0Njz/+OBRFQVxcHFasWAGfzxf0mI8++ggpKSkwGAywWCwoKioK2n/58mU8/fTTCA8PR1JSEurq6gL7rl69CrvdjqlTp0JRFCQlJd1S2IhIm1iAiOiuVVZWhoKCArS2tsJut6OwsBDt7e0AgL6+PuTk5CAmJgYtLS04dOgQvvnmm6CC43K54HQ6sWLFCrS1taGurg73339/0O944403sGzZMvzyyy944oknYLfbceXKlcDvP3PmDOrr69He3g6Xy4UpU6ZM3B+AiEZvwpZdJSIaAYfDIfR6vTAajUG3iooKIYQQAMTKlSuDvmfRokVi1apVQggh3G63iImJET6fL7D/iy++ECEhIaK7u1sIIYTVahUbNmz4xwwAxOuvvx647/P5BABRX18vhBAiLy9PvPDCC2PzhIloQnEOEBFp1mOPPQaXyxW0LTY2NvB1enp60L709HR4PB4AQHt7O+bPnw+j0RjYn5mZCb/fj46ODuh0Oly4cAFZWVm3zTBv3rzA10ajEZGRkejp6QEArFq1CgUFBTh16hQWL16M/Px8ZGRkjOq5EtHEYgEiIs0yGo23nJIaK4qiDOtxYWFhQfd1Oh38fj8AIDc3F3/++Se+/PJLNDQ0ICsrC06nE++8886Y5yWiscU5QER01zp+/Pgt95OTkwEAycnJaG1tRV9fX2B/c3MzQkJCMHPmTERERGD69OlobGy8owxTp06Fw+HA/v37sWvXLrjd7jv6eUQ0MXgEiIg0a2BgAN3d3UHbQkNDAxONDx06hNTUVDz00EM4cOAATp48ib179wIA7HY7Nm3aBIfDgc2bN+PSpUtYvXo1li9fDrPZDADYvHkzVq5cCZPJhNzcXPT29qK5uRmrV68eVr6NGzfiwQcfREpKCgYGBnDkyJFAASMibWMBIiLN+uqrr2CxWIK2zZw5E7/++iuAG5/QqqmpwUsvvQSLxYKPP/4Ys2fPBgCEh4fj66+/RnFxMdLS0hAeHo6CggLs3Lkz8LMcDgeuX7+OyspKrFu3DlOmTMEzzzwz7HyTJk3C+vXr8ccff0BRFDz88MOoqakZg2dORONNJ4QQaocgIhopnU6H2tpa5Ofnqx2FiO5CnANERERE0mEBIiIiIulwDhAR3ZV49p6I7gSPABEREZF0WICIiIhIOixAREREJB0WICIiIpIOCxARERFJhwWIiIiIpMMCRERERNJhASIiIiLpsAARERGRdP4DBCkl25mIK7oAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "val_acc_noise = history_noise.history[\"val_accuracy\"]\n",
        "val_acc_zeros = history_zeros.history[\"val_accuracy\"]\n",
        "vla_acc = history.history['val_accuracy']\n",
        "\n",
        "epochs = range(1, 11)\n",
        "\n",
        "plt.plot(epochs, val_acc_noise, \"b-\",\n",
        "         label=\"Validation accuracy with noise channels\")\n",
        "plt.plot(epochs, val_acc_zeros, \"b--\",\n",
        "         label=\"Validation accuracy with zeros channels\")\n",
        "plt.plot(epochs, vla_acc, 'b:',\n",
        "         label = 'Original Validation accuracy')\n",
        "\n",
        "plt.title(\"Effect of noise channels on validation accuracy\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VtPK5DKHP0rE"
      },
      "source": [
        "### The nature of generalization in deep learning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FVUr7aS7P0rF"
      },
      "source": [
        "**Fitting a MNIST model with randomly shuffled labels**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iJzgcGPnP0rF",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e08f8ca1-1ee1-40e4-927a-caeeecd6e1c4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 2.3160 - accuracy: 0.1035 - val_loss: 2.3048 - val_accuracy: 0.1034\n",
            "Epoch 2/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.2992 - accuracy: 0.1154 - val_loss: 2.3102 - val_accuracy: 0.1032\n",
            "Epoch 3/30\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 2.2907 - accuracy: 0.1273 - val_loss: 2.3187 - val_accuracy: 0.1018\n",
            "Epoch 4/30\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 2.2790 - accuracy: 0.1380 - val_loss: 2.3219 - val_accuracy: 0.1058\n",
            "Epoch 5/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.2621 - accuracy: 0.1533 - val_loss: 2.3386 - val_accuracy: 0.0970\n",
            "Epoch 6/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.2426 - accuracy: 0.1650 - val_loss: 2.3552 - val_accuracy: 0.0964\n",
            "Epoch 7/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.2185 - accuracy: 0.1831 - val_loss: 2.3637 - val_accuracy: 0.0990\n",
            "Epoch 8/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.1931 - accuracy: 0.1985 - val_loss: 2.3823 - val_accuracy: 0.0987\n",
            "Epoch 9/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.1639 - accuracy: 0.2119 - val_loss: 2.4009 - val_accuracy: 0.0989\n",
            "Epoch 10/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.1317 - accuracy: 0.2299 - val_loss: 2.4245 - val_accuracy: 0.0978\n",
            "Epoch 11/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.0968 - accuracy: 0.2462 - val_loss: 2.4519 - val_accuracy: 0.0988\n",
            "Epoch 12/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.0588 - accuracy: 0.2618 - val_loss: 2.4816 - val_accuracy: 0.0975\n",
            "Epoch 13/30\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 2.0225 - accuracy: 0.2808 - val_loss: 2.4998 - val_accuracy: 0.0991\n",
            "Epoch 14/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.9860 - accuracy: 0.2953 - val_loss: 2.5481 - val_accuracy: 0.0954\n",
            "Epoch 15/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.9472 - accuracy: 0.3113 - val_loss: 2.5653 - val_accuracy: 0.0967\n",
            "Epoch 16/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.9067 - accuracy: 0.3313 - val_loss: 2.6073 - val_accuracy: 0.0997\n",
            "Epoch 17/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.8679 - accuracy: 0.3447 - val_loss: 2.6500 - val_accuracy: 0.0999\n",
            "Epoch 18/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.8279 - accuracy: 0.3599 - val_loss: 2.6635 - val_accuracy: 0.0986\n",
            "Epoch 19/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.7880 - accuracy: 0.3747 - val_loss: 2.7052 - val_accuracy: 0.0970\n",
            "Epoch 20/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.7497 - accuracy: 0.3935 - val_loss: 2.7494 - val_accuracy: 0.0973\n",
            "Epoch 21/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.7111 - accuracy: 0.4075 - val_loss: 2.7974 - val_accuracy: 0.0990\n",
            "Epoch 22/30\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 1.6752 - accuracy: 0.4218 - val_loss: 2.8474 - val_accuracy: 0.0974\n",
            "Epoch 23/30\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 1.6361 - accuracy: 0.4348 - val_loss: 2.8897 - val_accuracy: 0.0983\n",
            "Epoch 24/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.6015 - accuracy: 0.4499 - val_loss: 2.9338 - val_accuracy: 0.1007\n",
            "Epoch 25/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.5649 - accuracy: 0.4660 - val_loss: 2.9728 - val_accuracy: 0.0980\n",
            "Epoch 26/30\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 1.5314 - accuracy: 0.4746 - val_loss: 3.0177 - val_accuracy: 0.1008\n",
            "Epoch 27/30\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 1.4984 - accuracy: 0.4868 - val_loss: 3.0503 - val_accuracy: 0.0980\n",
            "Epoch 28/30\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 1.4631 - accuracy: 0.5005 - val_loss: 3.1210 - val_accuracy: 0.0986\n",
            "Epoch 29/30\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 1.4312 - accuracy: 0.5127 - val_loss: 3.1796 - val_accuracy: 0.0944\n",
            "Epoch 30/30\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 1.4006 - accuracy: 0.5225 - val_loss: 3.2414 - val_accuracy: 0.0996\n"
          ]
        }
      ],
      "source": [
        "(train_images, train_labels), _ = mnist.load_data()\n",
        "train_images = train_images.reshape((60000, 28 * 28))\n",
        "train_images = train_images.astype(\"float32\") / 255\n",
        "\n",
        "random_train_labels = train_labels[:]\n",
        "np.random.shuffle(random_train_labels)\n",
        "\n",
        "model = keras.Sequential([\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(10, activation=\"softmax\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "hisory_shuffled = model.fit(train_images, random_train_labels,\n",
        "          epochs=30,\n",
        "          batch_size=128,\n",
        "          validation_split=0.2)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "val_acc_shuffled = hisory_shuffled.history[\"val_accuracy\"]\n",
        "acc_shuffled = hisory_shuffled.history[\"accuracy\"]\n",
        "\n",
        "epochs = range(1, 31)\n",
        "\n",
        "plt.plot(epochs, val_acc_shuffled, \"r-\",\n",
        "         label=\"Validation accuracy with shuffled data\")\n",
        "plt.plot(epochs, acc_shuffled, \"b-\",\n",
        "         label=\"Train accuracy with shuffled data\")\n",
        "\n",
        "plt.title(\"Effect of shuffling on validation accuracy\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Accuracy\")\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "id": "KGCs_JsZZ5pR",
        "outputId": "16507071-a87d-46b5-ca96-f064b5960781"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7b5228c632e0>"
            ]
          },
          "metadata": {},
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uOiTG1RAP0rF"
      },
      "source": [
        "#### The manifold hypothesis"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3gxz63JPP0rF"
      },
      "source": [
        "#### Interpolation as a source of generalization"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SnlMusryP0rF"
      },
      "source": [
        "#### Why deep learning works"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HWxDeIzPP0rF"
      },
      "source": [
        "#### Training data is paramount"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B1RAKc50P0rF"
      },
      "source": [
        "## Evaluating machine-learning models"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SgQKhKt_P0rF"
      },
      "source": [
        "### Training, validation, and test sets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6lbtwW3ZP0rF"
      },
      "source": [
        "#### Simple hold-out validation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2i0P__HZP0rF"
      },
      "source": [
        "#### K-fold validation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gOw71f0pP0rF"
      },
      "source": [
        "#### Iterated K-fold validation with shuffling"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KwKdQ25jP0rF"
      },
      "source": [
        "### Beating a common-sense baseline"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NqD0AVw-P0rG"
      },
      "source": [
        "### Things to keep in mind about model evaluation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mYDqJGqZP0rG"
      },
      "source": [
        "## Improving model fit"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X_sYEjixP0rG"
      },
      "source": [
        "### Tuning key gradient descent parameters"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kz4t1dlvP0rG"
      },
      "source": [
        "**Training a MNIST model with an incorrectly high learning rate**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zK-vRPCbP0rG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "59ed6a06-9111-4888-afcd-09c79b0b1900"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 562.1663 - accuracy: 0.2844 - val_loss: 2.2045 - val_accuracy: 0.1990\n",
            "Epoch 2/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.9698 - accuracy: 0.2156 - val_loss: 2.2131 - val_accuracy: 0.1975\n",
            "Epoch 3/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.4900 - accuracy: 0.1853 - val_loss: 2.4843 - val_accuracy: 0.1887\n",
            "Epoch 4/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.3822 - accuracy: 0.1834 - val_loss: 2.6636 - val_accuracy: 0.2368\n",
            "Epoch 5/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.6268 - accuracy: 0.2145 - val_loss: 2.2319 - val_accuracy: 0.1872\n",
            "Epoch 6/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.4395 - accuracy: 0.2074 - val_loss: 2.4861 - val_accuracy: 0.1747\n",
            "Epoch 7/10\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 2.2511 - accuracy: 0.2148 - val_loss: 5.5053 - val_accuracy: 0.1773\n",
            "Epoch 8/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.5048 - accuracy: 0.1985 - val_loss: 2.0286 - val_accuracy: 0.2286\n",
            "Epoch 9/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 2.3396 - accuracy: 0.2194 - val_loss: 2.1369 - val_accuracy: 0.2035\n",
            "Epoch 10/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 2.3292 - accuracy: 0.2116 - val_loss: 4.9792 - val_accuracy: 0.2587\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7b5264db8b20>"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ],
      "source": [
        "(train_images, train_labels), _ = mnist.load_data()\n",
        "train_images = train_images.reshape((60000, 28 * 28))\n",
        "train_images = train_images.astype(\"float32\") / 255\n",
        "\n",
        "model = keras.Sequential([\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(10, activation=\"softmax\")\n",
        "])\n",
        "model.compile(optimizer=keras.optimizers.RMSprop(1.),\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "model.fit(train_images, train_labels,\n",
        "          epochs=10,\n",
        "          batch_size=128,\n",
        "          validation_split=0.2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "08JBPCnbP0rG"
      },
      "source": [
        "**The same model with a more appropriate learning rate**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "N079QQhLP0rG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d5fdbc36-203a-4c31-ea2c-5b558d822ad8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.3436 - accuracy: 0.9109 - val_loss: 0.1453 - val_accuracy: 0.9591\n",
            "Epoch 2/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.1253 - accuracy: 0.9641 - val_loss: 0.1271 - val_accuracy: 0.9680\n",
            "Epoch 3/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0981 - accuracy: 0.9738 - val_loss: 0.1281 - val_accuracy: 0.9712\n",
            "Epoch 4/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0801 - accuracy: 0.9793 - val_loss: 0.2236 - val_accuracy: 0.9578\n",
            "Epoch 5/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0664 - accuracy: 0.9827 - val_loss: 0.1810 - val_accuracy: 0.9693\n",
            "Epoch 6/10\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.0608 - accuracy: 0.9851 - val_loss: 0.1739 - val_accuracy: 0.9723\n",
            "Epoch 7/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0520 - accuracy: 0.9867 - val_loss: 0.1864 - val_accuracy: 0.9723\n",
            "Epoch 8/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0478 - accuracy: 0.9890 - val_loss: 0.2116 - val_accuracy: 0.9708\n",
            "Epoch 9/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0416 - accuracy: 0.9901 - val_loss: 0.1922 - val_accuracy: 0.9753\n",
            "Epoch 10/10\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0375 - accuracy: 0.9914 - val_loss: 0.2253 - val_accuracy: 0.9728\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7b523eb13760>"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ],
      "source": [
        "model = keras.Sequential([\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(10, activation=\"softmax\")\n",
        "])\n",
        "model.compile(optimizer=keras.optimizers.RMSprop(1e-2),\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "model.fit(train_images, train_labels,\n",
        "          epochs=10,\n",
        "          batch_size=128,\n",
        "          validation_split=0.2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "08DF_5I6P0rG"
      },
      "source": [
        "### Leveraging better architecture priors"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "akstnhwAP0rG"
      },
      "source": [
        "### Increasing model capacity"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6i9419KzP0rG"
      },
      "source": [
        "**A simple logistic regression on MNIST**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OzOQrzH5P0rG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "26144d06-665d-44fd-9fbc-d8bda6699f48"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.6620 - accuracy: 0.8408 - val_loss: 0.3583 - val_accuracy: 0.9043\n",
            "Epoch 2/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.3524 - accuracy: 0.9037 - val_loss: 0.3106 - val_accuracy: 0.9147\n",
            "Epoch 3/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.3180 - accuracy: 0.9120 - val_loss: 0.2921 - val_accuracy: 0.9183\n",
            "Epoch 4/20\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.3019 - accuracy: 0.9160 - val_loss: 0.2849 - val_accuracy: 0.9213\n",
            "Epoch 5/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2922 - accuracy: 0.9184 - val_loss: 0.2798 - val_accuracy: 0.9240\n",
            "Epoch 6/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2854 - accuracy: 0.9207 - val_loss: 0.2756 - val_accuracy: 0.9236\n",
            "Epoch 7/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2810 - accuracy: 0.9215 - val_loss: 0.2717 - val_accuracy: 0.9277\n",
            "Epoch 8/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2769 - accuracy: 0.9231 - val_loss: 0.2709 - val_accuracy: 0.9258\n",
            "Epoch 9/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2733 - accuracy: 0.9249 - val_loss: 0.2688 - val_accuracy: 0.9260\n",
            "Epoch 10/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2708 - accuracy: 0.9249 - val_loss: 0.2694 - val_accuracy: 0.9258\n",
            "Epoch 11/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2687 - accuracy: 0.9259 - val_loss: 0.2666 - val_accuracy: 0.9282\n",
            "Epoch 12/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2665 - accuracy: 0.9264 - val_loss: 0.2674 - val_accuracy: 0.9272\n",
            "Epoch 13/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2647 - accuracy: 0.9270 - val_loss: 0.2652 - val_accuracy: 0.9299\n",
            "Epoch 14/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.2630 - accuracy: 0.9279 - val_loss: 0.2659 - val_accuracy: 0.9273\n",
            "Epoch 15/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2619 - accuracy: 0.9281 - val_loss: 0.2647 - val_accuracy: 0.9287\n",
            "Epoch 16/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2606 - accuracy: 0.9285 - val_loss: 0.2647 - val_accuracy: 0.9285\n",
            "Epoch 17/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2594 - accuracy: 0.9287 - val_loss: 0.2642 - val_accuracy: 0.9279\n",
            "Epoch 18/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2583 - accuracy: 0.9289 - val_loss: 0.2648 - val_accuracy: 0.9300\n",
            "Epoch 19/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2570 - accuracy: 0.9292 - val_loss: 0.2647 - val_accuracy: 0.9280\n",
            "Epoch 20/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.2560 - accuracy: 0.9294 - val_loss: 0.2635 - val_accuracy: 0.9308\n"
          ]
        }
      ],
      "source": [
        "model = keras.Sequential([layers.Dense(10, activation=\"softmax\")])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_small_model = model.fit(\n",
        "    train_images, train_labels,\n",
        "    epochs=20,\n",
        "    batch_size=128,\n",
        "    validation_split=0.2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dtstl7NfP0rG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "outputId": "31b85c8b-170e-4219-8efd-defe0127975a"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7b5264847190>"
            ]
          },
          "metadata": {},
          "execution_count": 32
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "val_loss = history_small_model.history[\"val_loss\"]\n",
        "epochs = range(1, 21)\n",
        "plt.plot(epochs, val_loss, \"b--\",\n",
        "         label=\"Validation loss\")\n",
        "plt.title(\"Effect of insufficient model capacity on validation loss\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.legend()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Qp_rGrvtP0rG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6f647975-ccf4-4628-b57e-f1801df72a7d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "375/375 [==============================] - 3s 5ms/step - loss: 0.3757 - accuracy: 0.8926 - val_loss: 0.1984 - val_accuracy: 0.9436\n",
            "Epoch 2/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.1675 - accuracy: 0.9503 - val_loss: 0.1444 - val_accuracy: 0.9586\n",
            "Epoch 3/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.1160 - accuracy: 0.9652 - val_loss: 0.1085 - val_accuracy: 0.9687\n",
            "Epoch 4/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0890 - accuracy: 0.9735 - val_loss: 0.0982 - val_accuracy: 0.9713\n",
            "Epoch 5/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0712 - accuracy: 0.9789 - val_loss: 0.0891 - val_accuracy: 0.9725\n",
            "Epoch 6/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0594 - accuracy: 0.9812 - val_loss: 0.0915 - val_accuracy: 0.9719\n",
            "Epoch 7/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0498 - accuracy: 0.9849 - val_loss: 0.0926 - val_accuracy: 0.9736\n",
            "Epoch 8/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0414 - accuracy: 0.9873 - val_loss: 0.0979 - val_accuracy: 0.9716\n",
            "Epoch 9/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0357 - accuracy: 0.9891 - val_loss: 0.1004 - val_accuracy: 0.9723\n",
            "Epoch 10/20\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 0.0284 - accuracy: 0.9913 - val_loss: 0.1113 - val_accuracy: 0.9712\n",
            "Epoch 11/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0252 - accuracy: 0.9921 - val_loss: 0.1045 - val_accuracy: 0.9736\n",
            "Epoch 12/20\n",
            "375/375 [==============================] - 2s 6ms/step - loss: 0.0208 - accuracy: 0.9936 - val_loss: 0.1046 - val_accuracy: 0.9743\n",
            "Epoch 13/20\n",
            "375/375 [==============================] - 2s 6ms/step - loss: 0.0169 - accuracy: 0.9947 - val_loss: 0.1034 - val_accuracy: 0.9738\n",
            "Epoch 14/20\n",
            "375/375 [==============================] - 1s 3ms/step - loss: 0.0150 - accuracy: 0.9954 - val_loss: 0.1022 - val_accuracy: 0.9762\n",
            "Epoch 15/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0131 - accuracy: 0.9958 - val_loss: 0.1104 - val_accuracy: 0.9736\n",
            "Epoch 16/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0107 - accuracy: 0.9966 - val_loss: 0.1206 - val_accuracy: 0.9737\n",
            "Epoch 17/20\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.0088 - accuracy: 0.9972 - val_loss: 0.1227 - val_accuracy: 0.9756\n",
            "Epoch 18/20\n",
            "375/375 [==============================] - 2s 4ms/step - loss: 0.0084 - accuracy: 0.9975 - val_loss: 0.1180 - val_accuracy: 0.9756\n",
            "Epoch 19/20\n",
            "375/375 [==============================] - 1s 4ms/step - loss: 0.0068 - accuracy: 0.9980 - val_loss: 0.1290 - val_accuracy: 0.9751\n",
            "Epoch 20/20\n",
            "375/375 [==============================] - 2s 5ms/step - loss: 0.0057 - accuracy: 0.9982 - val_loss: 0.1320 - val_accuracy: 0.9751\n"
          ]
        }
      ],
      "source": [
        "model = keras.Sequential([\n",
        "    layers.Dense(96, activation=\"relu\"),\n",
        "    layers.Dense(96, activation=\"relu\"),\n",
        "    layers.Dense(10, activation=\"softmax\"),\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_large_model = model.fit(\n",
        "    train_images, train_labels,\n",
        "    epochs=20,\n",
        "    batch_size=128,\n",
        "    validation_split=0.2)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "val_loss = history_large_model.history[\"val_loss\"]\n",
        "epochs = range(1, 21)\n",
        "plt.plot(epochs, val_loss, \"b--\",\n",
        "         label=\"Validation loss\")\n",
        "plt.title(\"Effect of large model capacity on validation loss\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Loss\")\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "id": "0_jWaQwvblel",
        "outputId": "c175343e-571b-4f59-ee63-69f8578b870c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7b526519c6a0>"
            ]
          },
          "metadata": {},
          "execution_count": 34
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ueWEC4vRP0rH"
      },
      "source": [
        "## Improving generalization"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DgnV6pGNP0rH"
      },
      "source": [
        "### Dataset curation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wmcipE_SP0rH"
      },
      "source": [
        "### Feature engineering"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8K79vwobP0rH"
      },
      "source": [
        "### Using early stopping"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F89OXQanP0rH"
      },
      "source": [
        "### Regularizing your model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MP-K62yEP0rI"
      },
      "source": [
        "#### Reducing the network's size"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dt_c0GkGP0rI"
      },
      "source": [
        "**Original model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7lpuG4owP0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b97606ae-a947-43c0-bf94-7d3528909ee2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/imdb.npz\n",
            "17464789/17464789 [==============================] - 2s 0us/step\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras.datasets import imdb\n",
        "(train_data, train_labels), _ = imdb.load_data(num_words=10000)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_data.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y8rOCb5scgIp",
        "outputId": "bde0a3f6-52d6-433b-b5f4-345f180c9b11"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000,)"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "def vectorize_sequences(sequences, dimension=10000):\n",
        "    results = np.zeros((len(sequences), dimension))\n",
        "    for i, sequence in enumerate(sequences):\n",
        "        results[i, sequence] = 1.\n",
        "    return results\n",
        "\n",
        "train_data = vectorize_sequences(train_data)"
      ],
      "metadata": {
        "id": "sEi7PhuLco8P"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_data.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "05z18iiDcUAJ",
        "outputId": "623bc7bb-7b5a-47ab-9e5d-703291761ea5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 10000)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f0VpQP9AcY2S",
        "outputId": "96046742-91c0-4883-8266-60dded9ae02e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 0, 0, ..., 0, 1, 0])"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "model = keras.Sequential([\n",
        "    layers.Dense(16, activation=\"relu\"),\n",
        "    layers.Dense(16, activation=\"relu\"),\n",
        "    layers.Dense(1, activation=\"sigmoid\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"binary_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_original = model.fit(train_data, train_labels,\n",
        "                             epochs=20, batch_size=512, validation_split=0.4)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U7c7XF0ecOAO",
        "outputId": "a504bf6f-9a49-4732-f017-176f1dced3db"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "30/30 [==============================] - 5s 70ms/step - loss: 0.5232 - accuracy: 0.7670 - val_loss: 0.3834 - val_accuracy: 0.8664\n",
            "Epoch 2/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.3180 - accuracy: 0.8959 - val_loss: 0.3486 - val_accuracy: 0.8574\n",
            "Epoch 3/20\n",
            "30/30 [==============================] - 1s 21ms/step - loss: 0.2379 - accuracy: 0.9237 - val_loss: 0.2863 - val_accuracy: 0.8866\n",
            "Epoch 4/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.1886 - accuracy: 0.9399 - val_loss: 0.3092 - val_accuracy: 0.8737\n",
            "Epoch 5/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.1611 - accuracy: 0.9485 - val_loss: 0.2779 - val_accuracy: 0.8880\n",
            "Epoch 6/20\n",
            "30/30 [==============================] - 1s 21ms/step - loss: 0.1365 - accuracy: 0.9567 - val_loss: 0.3047 - val_accuracy: 0.8814\n",
            "Epoch 7/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.1157 - accuracy: 0.9655 - val_loss: 0.3074 - val_accuracy: 0.8821\n",
            "Epoch 8/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.0999 - accuracy: 0.9702 - val_loss: 0.3124 - val_accuracy: 0.8828\n",
            "Epoch 9/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.0862 - accuracy: 0.9768 - val_loss: 0.3786 - val_accuracy: 0.8681\n",
            "Epoch 10/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.0770 - accuracy: 0.9805 - val_loss: 0.3483 - val_accuracy: 0.8779\n",
            "Epoch 11/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.0658 - accuracy: 0.9833 - val_loss: 0.3767 - val_accuracy: 0.8751\n",
            "Epoch 12/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.0578 - accuracy: 0.9854 - val_loss: 0.3794 - val_accuracy: 0.8767\n",
            "Epoch 13/20\n",
            "30/30 [==============================] - 1s 41ms/step - loss: 0.0485 - accuracy: 0.9903 - val_loss: 0.4040 - val_accuracy: 0.8742\n",
            "Epoch 14/20\n",
            "30/30 [==============================] - 1s 30ms/step - loss: 0.0447 - accuracy: 0.9897 - val_loss: 0.4354 - val_accuracy: 0.8728\n",
            "Epoch 15/20\n",
            "30/30 [==============================] - 1s 26ms/step - loss: 0.0369 - accuracy: 0.9916 - val_loss: 0.4478 - val_accuracy: 0.8724\n",
            "Epoch 16/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.0294 - accuracy: 0.9953 - val_loss: 0.4675 - val_accuracy: 0.8721\n",
            "Epoch 17/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.0261 - accuracy: 0.9959 - val_loss: 0.4896 - val_accuracy: 0.8720\n",
            "Epoch 18/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.0241 - accuracy: 0.9949 - val_loss: 0.5105 - val_accuracy: 0.8710\n",
            "Epoch 19/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.0211 - accuracy: 0.9958 - val_loss: 0.5379 - val_accuracy: 0.8716\n",
            "Epoch 20/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.0139 - accuracy: 0.9987 - val_loss: 0.5835 - val_accuracy: 0.8654\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e6DEEwcGP0rI"
      },
      "source": [
        "**Version of the model with lower capacity**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GpRzHuDjP0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d8e8705a-f07f-47d8-881c-16456b274676"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "30/30 [==============================] - 3s 67ms/step - loss: 0.5895 - accuracy: 0.7610 - val_loss: 0.5019 - val_accuracy: 0.8505\n",
            "Epoch 2/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.4395 - accuracy: 0.8761 - val_loss: 0.4117 - val_accuracy: 0.8692\n",
            "Epoch 3/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.3550 - accuracy: 0.8985 - val_loss: 0.3594 - val_accuracy: 0.8786\n",
            "Epoch 4/20\n",
            "30/30 [==============================] - 2s 54ms/step - loss: 0.2973 - accuracy: 0.9127 - val_loss: 0.3265 - val_accuracy: 0.8806\n",
            "Epoch 5/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.2564 - accuracy: 0.9221 - val_loss: 0.3047 - val_accuracy: 0.8871\n",
            "Epoch 6/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.2251 - accuracy: 0.9325 - val_loss: 0.2872 - val_accuracy: 0.8917\n",
            "Epoch 7/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.2004 - accuracy: 0.9396 - val_loss: 0.2802 - val_accuracy: 0.8902\n",
            "Epoch 8/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.1795 - accuracy: 0.9471 - val_loss: 0.2761 - val_accuracy: 0.8895\n",
            "Epoch 9/20\n",
            "30/30 [==============================] - 1s 32ms/step - loss: 0.1622 - accuracy: 0.9527 - val_loss: 0.2739 - val_accuracy: 0.8901\n",
            "Epoch 10/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.1465 - accuracy: 0.9582 - val_loss: 0.2765 - val_accuracy: 0.8892\n",
            "Epoch 11/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.1336 - accuracy: 0.9617 - val_loss: 0.2772 - val_accuracy: 0.8902\n",
            "Epoch 12/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.1211 - accuracy: 0.9665 - val_loss: 0.2924 - val_accuracy: 0.8856\n",
            "Epoch 13/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.1113 - accuracy: 0.9703 - val_loss: 0.2886 - val_accuracy: 0.8866\n",
            "Epoch 14/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.1020 - accuracy: 0.9729 - val_loss: 0.2979 - val_accuracy: 0.8863\n",
            "Epoch 15/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.0932 - accuracy: 0.9761 - val_loss: 0.3005 - val_accuracy: 0.8863\n",
            "Epoch 16/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.0853 - accuracy: 0.9785 - val_loss: 0.3097 - val_accuracy: 0.8856\n",
            "Epoch 17/20\n",
            "30/30 [==============================] - 1s 21ms/step - loss: 0.0780 - accuracy: 0.9817 - val_loss: 0.3328 - val_accuracy: 0.8803\n",
            "Epoch 18/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.0715 - accuracy: 0.9842 - val_loss: 0.3360 - val_accuracy: 0.8808\n",
            "Epoch 19/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.0651 - accuracy: 0.9858 - val_loss: 0.3424 - val_accuracy: 0.8809\n",
            "Epoch 20/20\n",
            "30/30 [==============================] - 1s 39ms/step - loss: 0.0593 - accuracy: 0.9876 - val_loss: 0.3532 - val_accuracy: 0.8806\n"
          ]
        }
      ],
      "source": [
        "model = keras.Sequential([\n",
        "    layers.Dense(4, activation=\"relu\"),\n",
        "    layers.Dense(4, activation=\"relu\"),\n",
        "    layers.Dense(1, activation=\"sigmoid\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"binary_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_smaller_model = model.fit(\n",
        "    train_data, train_labels,\n",
        "    epochs=20, batch_size=512, validation_split=0.4)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GVyeLcgQP0rI"
      },
      "source": [
        "**Version of the model with higher capacity**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nWZR8NeQP0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "12948be1-d53e-4ac3-bd5e-89a6b3b8ccf3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "30/30 [==============================] - 3s 70ms/step - loss: 0.5791 - accuracy: 0.7038 - val_loss: 0.3581 - val_accuracy: 0.8673\n",
            "Epoch 2/20\n",
            "30/30 [==============================] - 1s 28ms/step - loss: 0.3205 - accuracy: 0.8697 - val_loss: 0.2784 - val_accuracy: 0.8869\n",
            "Epoch 3/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.2452 - accuracy: 0.9033 - val_loss: 0.2716 - val_accuracy: 0.8856\n",
            "Epoch 4/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.1809 - accuracy: 0.9321 - val_loss: 0.2763 - val_accuracy: 0.8855\n",
            "Epoch 5/20\n",
            "30/30 [==============================] - 1s 26ms/step - loss: 0.1474 - accuracy: 0.9444 - val_loss: 0.2811 - val_accuracy: 0.8864\n",
            "Epoch 6/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.1057 - accuracy: 0.9617 - val_loss: 0.2977 - val_accuracy: 0.8868\n",
            "Epoch 7/20\n",
            "30/30 [==============================] - 1s 37ms/step - loss: 0.0820 - accuracy: 0.9735 - val_loss: 0.3304 - val_accuracy: 0.8830\n",
            "Epoch 8/20\n",
            "30/30 [==============================] - 1s 40ms/step - loss: 0.0717 - accuracy: 0.9793 - val_loss: 0.3414 - val_accuracy: 0.8827\n",
            "Epoch 9/20\n",
            "30/30 [==============================] - 1s 32ms/step - loss: 0.0146 - accuracy: 0.9982 - val_loss: 0.4657 - val_accuracy: 0.8818\n",
            "Epoch 10/20\n",
            "30/30 [==============================] - 1s 44ms/step - loss: 0.1029 - accuracy: 0.9780 - val_loss: 0.3628 - val_accuracy: 0.8816\n",
            "Epoch 11/20\n",
            "30/30 [==============================] - 1s 31ms/step - loss: 0.0071 - accuracy: 0.9997 - val_loss: 0.4720 - val_accuracy: 0.8827\n",
            "Epoch 12/20\n",
            "30/30 [==============================] - 1s 39ms/step - loss: 0.0027 - accuracy: 0.9999 - val_loss: 0.5419 - val_accuracy: 0.8834\n",
            "Epoch 13/20\n",
            "30/30 [==============================] - 1s 47ms/step - loss: 0.0893 - accuracy: 0.9814 - val_loss: 0.4435 - val_accuracy: 0.8824\n",
            "Epoch 14/20\n",
            "30/30 [==============================] - 1s 45ms/step - loss: 0.0026 - accuracy: 1.0000 - val_loss: 0.5173 - val_accuracy: 0.8842\n",
            "Epoch 15/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.0012 - accuracy: 1.0000 - val_loss: 0.5937 - val_accuracy: 0.8833\n",
            "Epoch 16/20\n",
            "30/30 [==============================] - 1s 39ms/step - loss: 7.3900e-04 - accuracy: 1.0000 - val_loss: 0.6262 - val_accuracy: 0.8822\n",
            "Epoch 17/20\n",
            "30/30 [==============================] - 1s 33ms/step - loss: 4.6447e-04 - accuracy: 1.0000 - val_loss: 0.6666 - val_accuracy: 0.8831\n",
            "Epoch 18/20\n",
            "30/30 [==============================] - 1s 34ms/step - loss: 3.2623e-04 - accuracy: 1.0000 - val_loss: 0.6934 - val_accuracy: 0.8834\n",
            "Epoch 19/20\n",
            "30/30 [==============================] - 1s 45ms/step - loss: 2.4786e-04 - accuracy: 1.0000 - val_loss: 0.7097 - val_accuracy: 0.8833\n",
            "Epoch 20/20\n",
            "30/30 [==============================] - 1s 50ms/step - loss: 2.0167e-04 - accuracy: 1.0000 - val_loss: 0.7316 - val_accuracy: 0.8825\n"
          ]
        }
      ],
      "source": [
        "model = keras.Sequential([\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(1, activation=\"sigmoid\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"binary_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_larger_model = model.fit(\n",
        "    train_data, train_labels,\n",
        "    epochs=20, batch_size=512, validation_split=0.4)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gHcSVrh3P0rI"
      },
      "source": [
        "#### Adding weight regularization"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kRk98ZF4P0rI"
      },
      "source": [
        "**Adding L2 weight regularization to the model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "AAxxOgz8P0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "320658fe-cf48-4f44-9787-1acded160592"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "30/30 [==============================] - 3s 84ms/step - loss: 0.5997 - accuracy: 0.7759 - val_loss: 0.4647 - val_accuracy: 0.8653\n",
            "Epoch 2/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.4017 - accuracy: 0.8920 - val_loss: 0.4024 - val_accuracy: 0.8737\n",
            "Epoch 3/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.3296 - accuracy: 0.9139 - val_loss: 0.4372 - val_accuracy: 0.8473\n",
            "Epoch 4/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.2907 - accuracy: 0.9268 - val_loss: 0.3571 - val_accuracy: 0.8883\n",
            "Epoch 5/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.2723 - accuracy: 0.9349 - val_loss: 0.3724 - val_accuracy: 0.8783\n",
            "Epoch 6/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.2544 - accuracy: 0.9412 - val_loss: 0.3575 - val_accuracy: 0.8863\n",
            "Epoch 7/20\n",
            "30/30 [==============================] - 1s 22ms/step - loss: 0.2429 - accuracy: 0.9447 - val_loss: 0.3754 - val_accuracy: 0.8806\n",
            "Epoch 8/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.2342 - accuracy: 0.9490 - val_loss: 0.4071 - val_accuracy: 0.8688\n",
            "Epoch 9/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.2336 - accuracy: 0.9493 - val_loss: 0.3738 - val_accuracy: 0.8824\n",
            "Epoch 10/20\n",
            "30/30 [==============================] - 1s 44ms/step - loss: 0.2247 - accuracy: 0.9505 - val_loss: 0.3839 - val_accuracy: 0.8798\n",
            "Epoch 11/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.2189 - accuracy: 0.9538 - val_loss: 0.3836 - val_accuracy: 0.8789\n",
            "Epoch 12/20\n",
            "30/30 [==============================] - 1s 31ms/step - loss: 0.2137 - accuracy: 0.9561 - val_loss: 0.3992 - val_accuracy: 0.8775\n",
            "Epoch 13/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.2070 - accuracy: 0.9602 - val_loss: 0.3987 - val_accuracy: 0.8771\n",
            "Epoch 14/20\n",
            "30/30 [==============================] - 1s 47ms/step - loss: 0.2112 - accuracy: 0.9564 - val_loss: 0.3967 - val_accuracy: 0.8782\n",
            "Epoch 15/20\n",
            "30/30 [==============================] - 1s 45ms/step - loss: 0.2018 - accuracy: 0.9608 - val_loss: 0.4291 - val_accuracy: 0.8714\n",
            "Epoch 16/20\n",
            "30/30 [==============================] - 1s 48ms/step - loss: 0.1984 - accuracy: 0.9647 - val_loss: 0.4056 - val_accuracy: 0.8778\n",
            "Epoch 17/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.1988 - accuracy: 0.9617 - val_loss: 0.4262 - val_accuracy: 0.8731\n",
            "Epoch 18/20\n",
            "30/30 [==============================] - 1s 28ms/step - loss: 0.1953 - accuracy: 0.9631 - val_loss: 0.5152 - val_accuracy: 0.8460\n",
            "Epoch 19/20\n",
            "30/30 [==============================] - 1s 39ms/step - loss: 0.1899 - accuracy: 0.9671 - val_loss: 0.4361 - val_accuracy: 0.8732\n",
            "Epoch 20/20\n",
            "30/30 [==============================] - 1s 24ms/step - loss: 0.1945 - accuracy: 0.9624 - val_loss: 0.4214 - val_accuracy: 0.8757\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras import regularizers\n",
        "model = keras.Sequential([\n",
        "    layers.Dense(16,\n",
        "                 kernel_regularizer=regularizers.l2(0.002),\n",
        "                 activation=\"relu\"),\n",
        "    layers.Dense(16,\n",
        "                 kernel_regularizer=regularizers.l2(0.002),\n",
        "                 activation=\"relu\"),\n",
        "    layers.Dense(1, activation=\"sigmoid\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"binary_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_l2_reg = model.fit(\n",
        "    train_data, train_labels,\n",
        "    epochs=20, batch_size=512, validation_split=0.4)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o5clDm9yP0rI"
      },
      "source": [
        "**Different weight regularizers available in Keras**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gjNxWW1OP0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0477144a-1559-481e-ac6f-ca8efab08320"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.regularizers.L1L2 at 0x7edc8ae5f4c0>"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "from tensorflow.keras import regularizers\n",
        "regularizers.l1(0.001)\n",
        "regularizers.l1_l2(l1=0.001, l2=0.001)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2JNvON9rP0rI"
      },
      "source": [
        "#### Adding dropout"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sK9V2ThiP0rI"
      },
      "source": [
        "**Adding dropout to the IMDB model**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4mRAQzwtP0rI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "af2e2c3c-1442-4dae-f355-cbdf619fe63d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "30/30 [==============================] - 7s 75ms/step - loss: 0.6389 - accuracy: 0.6223 - val_loss: 0.5412 - val_accuracy: 0.8145\n",
            "Epoch 2/20\n",
            "30/30 [==============================] - 1s 39ms/step - loss: 0.5416 - accuracy: 0.7469 - val_loss: 0.4560 - val_accuracy: 0.8546\n",
            "Epoch 3/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.4651 - accuracy: 0.8077 - val_loss: 0.4069 - val_accuracy: 0.8651\n",
            "Epoch 4/20\n",
            "30/30 [==============================] - 1s 27ms/step - loss: 0.4055 - accuracy: 0.8465 - val_loss: 0.3386 - val_accuracy: 0.8850\n",
            "Epoch 5/20\n",
            "30/30 [==============================] - 1s 26ms/step - loss: 0.3534 - accuracy: 0.8737 - val_loss: 0.3133 - val_accuracy: 0.8800\n",
            "Epoch 6/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.3174 - accuracy: 0.8885 - val_loss: 0.2899 - val_accuracy: 0.8908\n",
            "Epoch 7/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.2866 - accuracy: 0.9042 - val_loss: 0.2832 - val_accuracy: 0.8911\n",
            "Epoch 8/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.2587 - accuracy: 0.9143 - val_loss: 0.2813 - val_accuracy: 0.8919\n",
            "Epoch 9/20\n",
            "30/30 [==============================] - 1s 31ms/step - loss: 0.2337 - accuracy: 0.9233 - val_loss: 0.2804 - val_accuracy: 0.8919\n",
            "Epoch 10/20\n",
            "30/30 [==============================] - 1s 32ms/step - loss: 0.2169 - accuracy: 0.9317 - val_loss: 0.2877 - val_accuracy: 0.8920\n",
            "Epoch 11/20\n",
            "30/30 [==============================] - 1s 40ms/step - loss: 0.1961 - accuracy: 0.9368 - val_loss: 0.3120 - val_accuracy: 0.8903\n",
            "Epoch 12/20\n",
            "30/30 [==============================] - 1s 48ms/step - loss: 0.1839 - accuracy: 0.9400 - val_loss: 0.3427 - val_accuracy: 0.8864\n",
            "Epoch 13/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.1722 - accuracy: 0.9445 - val_loss: 0.3355 - val_accuracy: 0.8862\n",
            "Epoch 14/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.1488 - accuracy: 0.9543 - val_loss: 0.3415 - val_accuracy: 0.8882\n",
            "Epoch 15/20\n",
            "30/30 [==============================] - 1s 29ms/step - loss: 0.1469 - accuracy: 0.9529 - val_loss: 0.3759 - val_accuracy: 0.8878\n",
            "Epoch 16/20\n",
            "30/30 [==============================] - 1s 47ms/step - loss: 0.1347 - accuracy: 0.9567 - val_loss: 0.3720 - val_accuracy: 0.8883\n",
            "Epoch 17/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.1279 - accuracy: 0.9607 - val_loss: 0.4279 - val_accuracy: 0.8879\n",
            "Epoch 18/20\n",
            "30/30 [==============================] - 1s 25ms/step - loss: 0.1192 - accuracy: 0.9603 - val_loss: 0.4255 - val_accuracy: 0.8865\n",
            "Epoch 19/20\n",
            "30/30 [==============================] - 1s 23ms/step - loss: 0.1155 - accuracy: 0.9636 - val_loss: 0.4599 - val_accuracy: 0.8863\n",
            "Epoch 20/20\n",
            "30/30 [==============================] - 1s 21ms/step - loss: 0.1122 - accuracy: 0.9646 - val_loss: 0.4625 - val_accuracy: 0.8863\n"
          ]
        }
      ],
      "source": [
        "model = keras.Sequential([\n",
        "    layers.Dense(16, activation=\"relu\"),\n",
        "    layers.Dropout(0.5),\n",
        "    layers.Dense(16, activation=\"relu\"),\n",
        "    layers.Dropout(0.5),\n",
        "    layers.Dense(1, activation=\"sigmoid\")\n",
        "])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"binary_crossentropy\",\n",
        "              metrics=[\"accuracy\"])\n",
        "history_dropout = model.fit(\n",
        "    train_data, train_labels,\n",
        "    epochs=20, batch_size=512, validation_split=0.4)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# train accuracy 시각화\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "acc = history_original.history[\"accuracy\"]\n",
        "smaller_acc = history_smaller_model.history[\"accuracy\"]\n",
        "larger_acc = history_larger_model.history[\"accuracy\"]\n",
        "l2_acc = history_l2_reg.history[\"accuracy\"]\n",
        "dropout_acc = history_dropout.history[\"accuracy\"]\n",
        "\n",
        "\n",
        "epochs = range(1, 21)\n",
        "\n",
        "plt.plot(epochs, acc, \"k\",\n",
        "         label=\"original model\")\n",
        "plt.plot(epochs, smaller_acc, \"b\",\n",
        "         label=\"lower capacity\")\n",
        "plt.plot(epochs,larger_acc, \"r\",\n",
        "         label = \"higher capacity\")\n",
        "plt.plot(epochs, l2_acc, \"g\",\n",
        "         label = \"with L2 regularization\")\n",
        "plt.plot(epochs, dropout_acc, \"y\",\n",
        "         label = \"with dropout\")\n",
        "\n",
        "plt.title(\"Comparing train accuracy\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\" Accuracy\")\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "id": "Q7Ysyrmcml62",
        "outputId": "d34bd45e-f620-424b-c31d-9079674876c2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7edc8cd39a80>"
            ]
          },
          "metadata": {},
          "execution_count": 11
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# validation accuracy 시각화\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "val_acc = history_original.history[\"val_accuracy\"]\n",
        "smaller_val_acc = history_smaller_model.history[\"val_accuracy\"]\n",
        "larger_val_acc = history_larger_model.history[\"val_accuracy\"]\n",
        "l2_val_acc = history_l2_reg.history[\"val_accuracy\"]\n",
        "dropout_val_acc = history_dropout.history[\"val_accuracy\"]\n",
        "\n",
        "\n",
        "epochs = range(1, 21)\n",
        "\n",
        "plt.plot(epochs, val_acc, \"k\",\n",
        "         label=\"original model\")\n",
        "plt.plot(epochs, smaller_val_acc, \"b\",\n",
        "         label=\"lower capacity\")\n",
        "plt.plot(epochs,larger_val_acc, \"r\",\n",
        "         label = \"higher capacity\")\n",
        "plt.plot(epochs, l2_val_acc, \"g\",\n",
        "         label = \"with L2 regularization\")\n",
        "plt.plot(epochs, dropout_val_acc, \"y\",\n",
        "         label = \"with dropout\")\n",
        "\n",
        "plt.title(\"Comparing validation accuracy\")\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel(\"Validation Accuracy\")\n",
        "plt.legend()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "id": "Ig81AYjOeZYK",
        "outputId": "4cfff869-be55-4ba8-bed0-2a5fb50878f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7edc861ee4d0>"
            ]
          },
          "metadata": {},
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "762m5lEvP0rJ"
      },
      "source": [
        "## Summary"
      ]
    }
  ]
}